{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.6, "eval_steps": 500, "global_step": 60000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.693756103515625, "learning_rate": 1.8e-07, "loss": 1.6892, "step": 10 }, { "grad_norm": 0.6003186702728271, "learning_rate": 3.8e-07, "loss": 1.6916, "step": 20 }, { "grad_norm": 0.6004471778869629, "learning_rate": 5.8e-07, "loss": 1.6847, "step": 30 }, { "grad_norm": 0.5414710640907288, "learning_rate": 7.8e-07, "loss": 1.689, "step": 40 }, { "grad_norm": 0.6172371506690979, "learning_rate": 9.8e-07, "loss": 1.6692, "step": 50 }, { "grad_norm": 0.729831337928772, "learning_rate": 1.18e-06, "loss": 1.6154, "step": 60 }, { "grad_norm": 1.3932980298995972, "learning_rate": 1.3800000000000001e-06, "loss": 1.5266, "step": 70 }, { "grad_norm": 1.3984190225601196, "learning_rate": 1.5800000000000003e-06, "loss": 1.3497, "step": 80 }, { "grad_norm": 0.8466391563415527, "learning_rate": 1.7800000000000001e-06, "loss": 1.2074, "step": 90 }, { "grad_norm": 0.4408414363861084, "learning_rate": 1.98e-06, "loss": 1.1551, "step": 100 }, { "grad_norm": 0.4921647012233734, "learning_rate": 2.1800000000000003e-06, "loss": 1.134, "step": 110 }, { "grad_norm": 0.5542495250701904, "learning_rate": 2.38e-06, "loss": 1.1293, "step": 120 }, { "grad_norm": 0.4530753493309021, "learning_rate": 2.5800000000000003e-06, "loss": 1.1149, "step": 130 }, { "grad_norm": 0.5203211307525635, "learning_rate": 2.78e-06, "loss": 1.1042, "step": 140 }, { "grad_norm": 0.7891977429389954, "learning_rate": 2.9800000000000003e-06, "loss": 1.0894, "step": 150 }, { "grad_norm": 0.6363332867622375, "learning_rate": 3.1800000000000005e-06, "loss": 1.0839, "step": 160 }, { "grad_norm": 0.5244449377059937, "learning_rate": 3.38e-06, "loss": 1.068, "step": 170 }, { "grad_norm": 0.49814116954803467, "learning_rate": 3.58e-06, "loss": 1.0607, "step": 180 }, { "grad_norm": 0.4419620931148529, "learning_rate": 3.7800000000000002e-06, "loss": 1.0612, "step": 190 }, { "grad_norm": 0.41190576553344727, "learning_rate": 3.98e-06, "loss": 1.0582, "step": 200 }, { "grad_norm": 0.45068758726119995, "learning_rate": 4.18e-06, "loss": 1.0527, "step": 210 }, { "grad_norm": 0.5831984281539917, "learning_rate": 4.38e-06, "loss": 1.0498, "step": 220 }, { "grad_norm": 0.5598687529563904, "learning_rate": 4.58e-06, "loss": 1.0378, "step": 230 }, { "grad_norm": 0.4275733530521393, "learning_rate": 4.780000000000001e-06, "loss": 1.0387, "step": 240 }, { "grad_norm": 0.5040018558502197, "learning_rate": 4.98e-06, "loss": 1.0355, "step": 250 }, { "grad_norm": 0.6816054582595825, "learning_rate": 5.18e-06, "loss": 1.0295, "step": 260 }, { "grad_norm": 0.8895726799964905, "learning_rate": 5.38e-06, "loss": 1.0163, "step": 270 }, { "grad_norm": 0.9921588897705078, "learning_rate": 5.580000000000001e-06, "loss": 0.974, "step": 280 }, { "grad_norm": 1.0930176973342896, "learning_rate": 5.78e-06, "loss": 0.9334, "step": 290 }, { "grad_norm": 1.1918584108352661, "learning_rate": 5.98e-06, "loss": 0.898, "step": 300 }, { "grad_norm": 1.069115161895752, "learning_rate": 6.18e-06, "loss": 0.8546, "step": 310 }, { "grad_norm": 0.9410675168037415, "learning_rate": 6.38e-06, "loss": 0.8149, "step": 320 }, { "grad_norm": 1.5173364877700806, "learning_rate": 6.58e-06, "loss": 0.767, "step": 330 }, { "grad_norm": 1.1821058988571167, "learning_rate": 6.78e-06, "loss": 0.7077, "step": 340 }, { "grad_norm": 1.2589993476867676, "learning_rate": 6.98e-06, "loss": 0.6553, "step": 350 }, { "grad_norm": 1.2846978902816772, "learning_rate": 7.180000000000001e-06, "loss": 0.6008, "step": 360 }, { "grad_norm": 1.5558958053588867, "learning_rate": 7.3800000000000005e-06, "loss": 0.5642, "step": 370 }, { "grad_norm": 1.5295792818069458, "learning_rate": 7.580000000000001e-06, "loss": 0.53, "step": 380 }, { "grad_norm": 1.4677554368972778, "learning_rate": 7.78e-06, "loss": 0.4928, "step": 390 }, { "grad_norm": 1.7895665168762207, "learning_rate": 7.98e-06, "loss": 0.4453, "step": 400 }, { "grad_norm": 1.7666951417922974, "learning_rate": 8.18e-06, "loss": 0.3966, "step": 410 }, { "grad_norm": 1.8350659608840942, "learning_rate": 8.380000000000001e-06, "loss": 0.364, "step": 420 }, { "grad_norm": 1.4781367778778076, "learning_rate": 8.580000000000001e-06, "loss": 0.322, "step": 430 }, { "grad_norm": 1.908348798751831, "learning_rate": 8.78e-06, "loss": 0.2884, "step": 440 }, { "grad_norm": 1.8030550479888916, "learning_rate": 8.98e-06, "loss": 0.2636, "step": 450 }, { "grad_norm": 1.69098961353302, "learning_rate": 9.180000000000002e-06, "loss": 0.2299, "step": 460 }, { "grad_norm": 1.2861294746398926, "learning_rate": 9.38e-06, "loss": 0.2144, "step": 470 }, { "grad_norm": 1.6657389402389526, "learning_rate": 9.58e-06, "loss": 0.1975, "step": 480 }, { "grad_norm": 1.8318380117416382, "learning_rate": 9.78e-06, "loss": 0.1842, "step": 490 }, { "grad_norm": 1.814424753189087, "learning_rate": 9.980000000000001e-06, "loss": 0.1813, "step": 500 }, { "grad_norm": 1.571799397468567, "learning_rate": 1.018e-05, "loss": 0.1765, "step": 510 }, { "grad_norm": 1.9887003898620605, "learning_rate": 1.038e-05, "loss": 0.1573, "step": 520 }, { "grad_norm": 1.8860033750534058, "learning_rate": 1.058e-05, "loss": 0.1503, "step": 530 }, { "grad_norm": 1.725490927696228, "learning_rate": 1.0780000000000002e-05, "loss": 0.157, "step": 540 }, { "grad_norm": 1.2211230993270874, "learning_rate": 1.098e-05, "loss": 0.1401, "step": 550 }, { "grad_norm": 1.692814826965332, "learning_rate": 1.118e-05, "loss": 0.1389, "step": 560 }, { "grad_norm": 1.8478869199752808, "learning_rate": 1.1380000000000001e-05, "loss": 0.1407, "step": 570 }, { "grad_norm": 1.5018110275268555, "learning_rate": 1.1580000000000001e-05, "loss": 0.1306, "step": 580 }, { "grad_norm": 1.7675715684890747, "learning_rate": 1.178e-05, "loss": 0.1313, "step": 590 }, { "grad_norm": 1.6987860202789307, "learning_rate": 1.198e-05, "loss": 0.1252, "step": 600 }, { "grad_norm": 1.8356386423110962, "learning_rate": 1.2180000000000002e-05, "loss": 0.121, "step": 610 }, { "grad_norm": 1.8774702548980713, "learning_rate": 1.238e-05, "loss": 0.121, "step": 620 }, { "grad_norm": 1.2651643753051758, "learning_rate": 1.258e-05, "loss": 0.1163, "step": 630 }, { "grad_norm": 1.6558728218078613, "learning_rate": 1.278e-05, "loss": 0.1206, "step": 640 }, { "grad_norm": 1.8063287734985352, "learning_rate": 1.2980000000000001e-05, "loss": 0.1146, "step": 650 }, { "grad_norm": 1.576841950416565, "learning_rate": 1.3180000000000001e-05, "loss": 0.1069, "step": 660 }, { "grad_norm": 1.6164145469665527, "learning_rate": 1.338e-05, "loss": 0.1148, "step": 670 }, { "grad_norm": 1.3057990074157715, "learning_rate": 1.358e-05, "loss": 0.1075, "step": 680 }, { "grad_norm": 1.9110182523727417, "learning_rate": 1.3780000000000002e-05, "loss": 0.1057, "step": 690 }, { "grad_norm": 1.6646548509597778, "learning_rate": 1.3980000000000002e-05, "loss": 0.104, "step": 700 }, { "grad_norm": 1.7710074186325073, "learning_rate": 1.4180000000000001e-05, "loss": 0.1008, "step": 710 }, { "grad_norm": 1.826549768447876, "learning_rate": 1.4380000000000001e-05, "loss": 0.1015, "step": 720 }, { "grad_norm": 1.9241018295288086, "learning_rate": 1.4580000000000003e-05, "loss": 0.1052, "step": 730 }, { "grad_norm": 1.8638265132904053, "learning_rate": 1.4779999999999999e-05, "loss": 0.0999, "step": 740 }, { "grad_norm": 1.4784295558929443, "learning_rate": 1.4979999999999999e-05, "loss": 0.097, "step": 750 }, { "grad_norm": 1.446042537689209, "learning_rate": 1.518e-05, "loss": 0.0971, "step": 760 }, { "grad_norm": 1.9359536170959473, "learning_rate": 1.538e-05, "loss": 0.1013, "step": 770 }, { "grad_norm": 1.9544962644577026, "learning_rate": 1.558e-05, "loss": 0.0946, "step": 780 }, { "grad_norm": 1.8630268573760986, "learning_rate": 1.578e-05, "loss": 0.0907, "step": 790 }, { "grad_norm": 2.144911527633667, "learning_rate": 1.598e-05, "loss": 0.0957, "step": 800 }, { "grad_norm": 1.913624882698059, "learning_rate": 1.618e-05, "loss": 0.0913, "step": 810 }, { "grad_norm": 1.580592155456543, "learning_rate": 1.6380000000000002e-05, "loss": 0.0876, "step": 820 }, { "grad_norm": 1.991773009300232, "learning_rate": 1.658e-05, "loss": 0.0811, "step": 830 }, { "grad_norm": 2.168091058731079, "learning_rate": 1.6780000000000002e-05, "loss": 0.0879, "step": 840 }, { "grad_norm": 1.40365731716156, "learning_rate": 1.698e-05, "loss": 0.0946, "step": 850 }, { "grad_norm": 1.99235200881958, "learning_rate": 1.718e-05, "loss": 0.0839, "step": 860 }, { "grad_norm": 1.6337414979934692, "learning_rate": 1.7380000000000003e-05, "loss": 0.0842, "step": 870 }, { "grad_norm": 1.4455708265304565, "learning_rate": 1.758e-05, "loss": 0.0857, "step": 880 }, { "grad_norm": 1.5116981267929077, "learning_rate": 1.7780000000000003e-05, "loss": 0.0852, "step": 890 }, { "grad_norm": 1.4903907775878906, "learning_rate": 1.798e-05, "loss": 0.0852, "step": 900 }, { "grad_norm": 1.9278842210769653, "learning_rate": 1.818e-05, "loss": 0.0809, "step": 910 }, { "grad_norm": 1.855885624885559, "learning_rate": 1.838e-05, "loss": 0.0842, "step": 920 }, { "grad_norm": 1.8897618055343628, "learning_rate": 1.858e-05, "loss": 0.0847, "step": 930 }, { "grad_norm": 1.6090404987335205, "learning_rate": 1.878e-05, "loss": 0.0808, "step": 940 }, { "grad_norm": 2.008901834487915, "learning_rate": 1.898e-05, "loss": 0.0789, "step": 950 }, { "grad_norm": 1.4659730195999146, "learning_rate": 1.918e-05, "loss": 0.0773, "step": 960 }, { "grad_norm": 1.4381332397460938, "learning_rate": 1.938e-05, "loss": 0.0779, "step": 970 }, { "grad_norm": 1.7711352109909058, "learning_rate": 1.9580000000000002e-05, "loss": 0.0763, "step": 980 }, { "grad_norm": 1.4798812866210938, "learning_rate": 1.978e-05, "loss": 0.0769, "step": 990 }, { "grad_norm": 1.9653278589248657, "learning_rate": 1.9980000000000002e-05, "loss": 0.0752, "step": 1000 }, { "grad_norm": 1.6752407550811768, "learning_rate": 2.0180000000000003e-05, "loss": 0.0774, "step": 1010 }, { "grad_norm": 1.3617324829101562, "learning_rate": 2.038e-05, "loss": 0.0744, "step": 1020 }, { "grad_norm": 1.2933712005615234, "learning_rate": 2.0580000000000003e-05, "loss": 0.0747, "step": 1030 }, { "grad_norm": 1.71421480178833, "learning_rate": 2.078e-05, "loss": 0.074, "step": 1040 }, { "grad_norm": 1.8509901762008667, "learning_rate": 2.098e-05, "loss": 0.0762, "step": 1050 }, { "grad_norm": 2.285193681716919, "learning_rate": 2.118e-05, "loss": 0.0731, "step": 1060 }, { "grad_norm": 1.8731358051300049, "learning_rate": 2.138e-05, "loss": 0.0763, "step": 1070 }, { "grad_norm": 1.7692893743515015, "learning_rate": 2.158e-05, "loss": 0.0695, "step": 1080 }, { "grad_norm": 1.8922609090805054, "learning_rate": 2.178e-05, "loss": 0.0679, "step": 1090 }, { "grad_norm": 1.8896149396896362, "learning_rate": 2.198e-05, "loss": 0.0728, "step": 1100 }, { "grad_norm": 1.4519813060760498, "learning_rate": 2.218e-05, "loss": 0.075, "step": 1110 }, { "grad_norm": 1.4566588401794434, "learning_rate": 2.2380000000000003e-05, "loss": 0.0674, "step": 1120 }, { "grad_norm": 2.1822526454925537, "learning_rate": 2.258e-05, "loss": 0.0727, "step": 1130 }, { "grad_norm": 1.610540509223938, "learning_rate": 2.2780000000000002e-05, "loss": 0.0648, "step": 1140 }, { "grad_norm": 1.361895203590393, "learning_rate": 2.298e-05, "loss": 0.0665, "step": 1150 }, { "grad_norm": 1.015256404876709, "learning_rate": 2.318e-05, "loss": 0.0702, "step": 1160 }, { "grad_norm": 1.8129799365997314, "learning_rate": 2.3380000000000003e-05, "loss": 0.0708, "step": 1170 }, { "grad_norm": 1.131244421005249, "learning_rate": 2.358e-05, "loss": 0.0696, "step": 1180 }, { "grad_norm": 1.350728988647461, "learning_rate": 2.3780000000000003e-05, "loss": 0.0649, "step": 1190 }, { "grad_norm": 1.6572991609573364, "learning_rate": 2.398e-05, "loss": 0.0663, "step": 1200 }, { "grad_norm": 1.6831896305084229, "learning_rate": 2.418e-05, "loss": 0.0691, "step": 1210 }, { "grad_norm": 1.935192584991455, "learning_rate": 2.438e-05, "loss": 0.0668, "step": 1220 }, { "grad_norm": 1.4159165620803833, "learning_rate": 2.4580000000000002e-05, "loss": 0.0661, "step": 1230 }, { "grad_norm": 1.5825340747833252, "learning_rate": 2.478e-05, "loss": 0.0658, "step": 1240 }, { "grad_norm": 1.8843193054199219, "learning_rate": 2.498e-05, "loss": 0.0692, "step": 1250 }, { "grad_norm": 1.329866647720337, "learning_rate": 2.5180000000000003e-05, "loss": 0.0648, "step": 1260 }, { "grad_norm": 1.6016234159469604, "learning_rate": 2.5380000000000004e-05, "loss": 0.0623, "step": 1270 }, { "grad_norm": 1.58713698387146, "learning_rate": 2.5580000000000002e-05, "loss": 0.07, "step": 1280 }, { "grad_norm": 1.4170138835906982, "learning_rate": 2.5779999999999997e-05, "loss": 0.0626, "step": 1290 }, { "grad_norm": 1.2571437358856201, "learning_rate": 2.598e-05, "loss": 0.0637, "step": 1300 }, { "grad_norm": 1.4791704416275024, "learning_rate": 2.618e-05, "loss": 0.0647, "step": 1310 }, { "grad_norm": 1.5405535697937012, "learning_rate": 2.6379999999999998e-05, "loss": 0.0638, "step": 1320 }, { "grad_norm": 1.5608922243118286, "learning_rate": 2.658e-05, "loss": 0.0645, "step": 1330 }, { "grad_norm": 1.6547499895095825, "learning_rate": 2.678e-05, "loss": 0.0645, "step": 1340 }, { "grad_norm": 1.3953131437301636, "learning_rate": 2.698e-05, "loss": 0.0638, "step": 1350 }, { "grad_norm": 1.5092405080795288, "learning_rate": 2.718e-05, "loss": 0.0586, "step": 1360 }, { "grad_norm": 1.5479133129119873, "learning_rate": 2.738e-05, "loss": 0.0613, "step": 1370 }, { "grad_norm": 1.4852193593978882, "learning_rate": 2.758e-05, "loss": 0.0576, "step": 1380 }, { "grad_norm": 1.558904767036438, "learning_rate": 2.778e-05, "loss": 0.0605, "step": 1390 }, { "grad_norm": 1.59416925907135, "learning_rate": 2.798e-05, "loss": 0.0597, "step": 1400 }, { "grad_norm": 1.2178386449813843, "learning_rate": 2.818e-05, "loss": 0.0607, "step": 1410 }, { "grad_norm": 1.540867805480957, "learning_rate": 2.8380000000000003e-05, "loss": 0.0572, "step": 1420 }, { "grad_norm": 1.736190676689148, "learning_rate": 2.858e-05, "loss": 0.0654, "step": 1430 }, { "grad_norm": 1.3736093044281006, "learning_rate": 2.8780000000000002e-05, "loss": 0.0619, "step": 1440 }, { "grad_norm": 1.390933871269226, "learning_rate": 2.898e-05, "loss": 0.0607, "step": 1450 }, { "grad_norm": 1.4792113304138184, "learning_rate": 2.9180000000000002e-05, "loss": 0.0601, "step": 1460 }, { "grad_norm": 1.763543963432312, "learning_rate": 2.9380000000000003e-05, "loss": 0.0661, "step": 1470 }, { "grad_norm": 1.5734057426452637, "learning_rate": 2.958e-05, "loss": 0.0589, "step": 1480 }, { "grad_norm": 1.6969457864761353, "learning_rate": 2.9780000000000003e-05, "loss": 0.0553, "step": 1490 }, { "grad_norm": 1.5006678104400635, "learning_rate": 2.998e-05, "loss": 0.0554, "step": 1500 }, { "grad_norm": 1.380702257156372, "learning_rate": 3.0180000000000002e-05, "loss": 0.0577, "step": 1510 }, { "grad_norm": 1.7511974573135376, "learning_rate": 3.0380000000000004e-05, "loss": 0.0551, "step": 1520 }, { "grad_norm": 1.5342329740524292, "learning_rate": 3.058e-05, "loss": 0.057, "step": 1530 }, { "grad_norm": 1.6250231266021729, "learning_rate": 3.078e-05, "loss": 0.0551, "step": 1540 }, { "grad_norm": 1.5715347528457642, "learning_rate": 3.0980000000000005e-05, "loss": 0.0564, "step": 1550 }, { "grad_norm": 1.52754545211792, "learning_rate": 3.118e-05, "loss": 0.0557, "step": 1560 }, { "grad_norm": 1.2226603031158447, "learning_rate": 3.138e-05, "loss": 0.0549, "step": 1570 }, { "grad_norm": 1.2831780910491943, "learning_rate": 3.1580000000000006e-05, "loss": 0.0542, "step": 1580 }, { "grad_norm": 1.372782826423645, "learning_rate": 3.1780000000000004e-05, "loss": 0.0534, "step": 1590 }, { "grad_norm": 1.2618694305419922, "learning_rate": 3.198e-05, "loss": 0.0554, "step": 1600 }, { "grad_norm": 1.5687618255615234, "learning_rate": 3.218e-05, "loss": 0.0549, "step": 1610 }, { "grad_norm": 1.39605712890625, "learning_rate": 3.238e-05, "loss": 0.0516, "step": 1620 }, { "grad_norm": 1.534334659576416, "learning_rate": 3.2579999999999996e-05, "loss": 0.056, "step": 1630 }, { "grad_norm": 1.6379858255386353, "learning_rate": 3.278e-05, "loss": 0.0569, "step": 1640 }, { "grad_norm": 1.3880635499954224, "learning_rate": 3.298e-05, "loss": 0.061, "step": 1650 }, { "grad_norm": 1.49595046043396, "learning_rate": 3.318e-05, "loss": 0.0585, "step": 1660 }, { "grad_norm": 1.4962018728256226, "learning_rate": 3.338e-05, "loss": 0.0545, "step": 1670 }, { "grad_norm": 1.3349930047988892, "learning_rate": 3.358e-05, "loss": 0.0523, "step": 1680 }, { "grad_norm": 1.3363659381866455, "learning_rate": 3.378e-05, "loss": 0.0542, "step": 1690 }, { "grad_norm": 1.409330129623413, "learning_rate": 3.398e-05, "loss": 0.053, "step": 1700 }, { "grad_norm": 1.1951149702072144, "learning_rate": 3.418e-05, "loss": 0.0488, "step": 1710 }, { "grad_norm": 1.7074482440948486, "learning_rate": 3.438e-05, "loss": 0.0554, "step": 1720 }, { "grad_norm": 1.5195300579071045, "learning_rate": 3.4580000000000004e-05, "loss": 0.0567, "step": 1730 }, { "grad_norm": 0.8830610513687134, "learning_rate": 3.478e-05, "loss": 0.0519, "step": 1740 }, { "grad_norm": 1.792742371559143, "learning_rate": 3.498e-05, "loss": 0.0555, "step": 1750 }, { "grad_norm": 1.2809474468231201, "learning_rate": 3.518e-05, "loss": 0.0541, "step": 1760 }, { "grad_norm": 1.0412020683288574, "learning_rate": 3.5380000000000003e-05, "loss": 0.0551, "step": 1770 }, { "grad_norm": 1.5938231945037842, "learning_rate": 3.558e-05, "loss": 0.0546, "step": 1780 }, { "grad_norm": 0.9111155271530151, "learning_rate": 3.578e-05, "loss": 0.0571, "step": 1790 }, { "grad_norm": 1.274709701538086, "learning_rate": 3.5980000000000004e-05, "loss": 0.0538, "step": 1800 }, { "grad_norm": 1.3212766647338867, "learning_rate": 3.618e-05, "loss": 0.0521, "step": 1810 }, { "grad_norm": 1.0994735956192017, "learning_rate": 3.638e-05, "loss": 0.0516, "step": 1820 }, { "grad_norm": 1.1335134506225586, "learning_rate": 3.6580000000000006e-05, "loss": 0.052, "step": 1830 }, { "grad_norm": 1.2332730293273926, "learning_rate": 3.6780000000000004e-05, "loss": 0.0532, "step": 1840 }, { "grad_norm": 1.759627103805542, "learning_rate": 3.698e-05, "loss": 0.0525, "step": 1850 }, { "grad_norm": 1.1046327352523804, "learning_rate": 3.7180000000000007e-05, "loss": 0.0518, "step": 1860 }, { "grad_norm": 1.1078822612762451, "learning_rate": 3.7380000000000005e-05, "loss": 0.0506, "step": 1870 }, { "grad_norm": 1.5564062595367432, "learning_rate": 3.758e-05, "loss": 0.0536, "step": 1880 }, { "grad_norm": 1.0875248908996582, "learning_rate": 3.778000000000001e-05, "loss": 0.0505, "step": 1890 }, { "grad_norm": 1.3859227895736694, "learning_rate": 3.7980000000000006e-05, "loss": 0.054, "step": 1900 }, { "grad_norm": 1.3205198049545288, "learning_rate": 3.818e-05, "loss": 0.0511, "step": 1910 }, { "grad_norm": 1.0833168029785156, "learning_rate": 3.838e-05, "loss": 0.0547, "step": 1920 }, { "grad_norm": 1.4245330095291138, "learning_rate": 3.858e-05, "loss": 0.0509, "step": 1930 }, { "grad_norm": 1.5774645805358887, "learning_rate": 3.878e-05, "loss": 0.0517, "step": 1940 }, { "grad_norm": 1.0717997550964355, "learning_rate": 3.898e-05, "loss": 0.0497, "step": 1950 }, { "grad_norm": 1.1655628681182861, "learning_rate": 3.918e-05, "loss": 0.0555, "step": 1960 }, { "grad_norm": 1.2361398935317993, "learning_rate": 3.938e-05, "loss": 0.0496, "step": 1970 }, { "grad_norm": 1.1725077629089355, "learning_rate": 3.958e-05, "loss": 0.0498, "step": 1980 }, { "grad_norm": 1.6416962146759033, "learning_rate": 3.978e-05, "loss": 0.0537, "step": 1990 }, { "grad_norm": 1.4302135705947876, "learning_rate": 3.998e-05, "loss": 0.0549, "step": 2000 }, { "grad_norm": 1.1555523872375488, "learning_rate": 4.018e-05, "loss": 0.0508, "step": 2010 }, { "grad_norm": 1.2920221090316772, "learning_rate": 4.038e-05, "loss": 0.0479, "step": 2020 }, { "grad_norm": 1.4682244062423706, "learning_rate": 4.058e-05, "loss": 0.0484, "step": 2030 }, { "grad_norm": 1.0421990156173706, "learning_rate": 4.078e-05, "loss": 0.0491, "step": 2040 }, { "grad_norm": 1.3361893892288208, "learning_rate": 4.0980000000000004e-05, "loss": 0.0499, "step": 2050 }, { "grad_norm": 1.4599676132202148, "learning_rate": 4.118e-05, "loss": 0.0505, "step": 2060 }, { "grad_norm": 1.043084979057312, "learning_rate": 4.138e-05, "loss": 0.0496, "step": 2070 }, { "grad_norm": 1.2248846292495728, "learning_rate": 4.1580000000000005e-05, "loss": 0.0474, "step": 2080 }, { "grad_norm": 1.3510767221450806, "learning_rate": 4.178e-05, "loss": 0.0464, "step": 2090 }, { "grad_norm": 1.060725450515747, "learning_rate": 4.198e-05, "loss": 0.0471, "step": 2100 }, { "grad_norm": 1.114605188369751, "learning_rate": 4.2180000000000006e-05, "loss": 0.0479, "step": 2110 }, { "grad_norm": 1.1400954723358154, "learning_rate": 4.2380000000000004e-05, "loss": 0.0476, "step": 2120 }, { "grad_norm": 1.2667913436889648, "learning_rate": 4.258e-05, "loss": 0.0494, "step": 2130 }, { "grad_norm": 0.746956467628479, "learning_rate": 4.278e-05, "loss": 0.0475, "step": 2140 }, { "grad_norm": 0.8724976778030396, "learning_rate": 4.2980000000000005e-05, "loss": 0.0478, "step": 2150 }, { "grad_norm": 0.9687085151672363, "learning_rate": 4.318e-05, "loss": 0.0471, "step": 2160 }, { "grad_norm": 1.1828197240829468, "learning_rate": 4.338e-05, "loss": 0.0491, "step": 2170 }, { "grad_norm": 1.2612067461013794, "learning_rate": 4.3580000000000006e-05, "loss": 0.0445, "step": 2180 }, { "grad_norm": 0.98041170835495, "learning_rate": 4.3780000000000004e-05, "loss": 0.0478, "step": 2190 }, { "grad_norm": 1.1112871170043945, "learning_rate": 4.398e-05, "loss": 0.0467, "step": 2200 }, { "grad_norm": 1.0798406600952148, "learning_rate": 4.418000000000001e-05, "loss": 0.0471, "step": 2210 }, { "grad_norm": 1.6303399801254272, "learning_rate": 4.438e-05, "loss": 0.0465, "step": 2220 }, { "grad_norm": 0.6615476012229919, "learning_rate": 4.458e-05, "loss": 0.0469, "step": 2230 }, { "grad_norm": 0.719767153263092, "learning_rate": 4.478e-05, "loss": 0.0457, "step": 2240 }, { "grad_norm": 0.8425790071487427, "learning_rate": 4.498e-05, "loss": 0.0444, "step": 2250 }, { "grad_norm": 0.8923128247261047, "learning_rate": 4.518e-05, "loss": 0.0413, "step": 2260 }, { "grad_norm": 1.1897622346878052, "learning_rate": 4.538e-05, "loss": 0.0447, "step": 2270 }, { "grad_norm": 0.9154702425003052, "learning_rate": 4.558e-05, "loss": 0.0472, "step": 2280 }, { "grad_norm": 1.0249792337417603, "learning_rate": 4.578e-05, "loss": 0.0454, "step": 2290 }, { "grad_norm": 0.8096370697021484, "learning_rate": 4.5980000000000004e-05, "loss": 0.0474, "step": 2300 }, { "grad_norm": 0.9660692811012268, "learning_rate": 4.618e-05, "loss": 0.0472, "step": 2310 }, { "grad_norm": 1.0357638597488403, "learning_rate": 4.638e-05, "loss": 0.0455, "step": 2320 }, { "grad_norm": 0.9514322280883789, "learning_rate": 4.6580000000000005e-05, "loss": 0.0439, "step": 2330 }, { "grad_norm": 1.105112075805664, "learning_rate": 4.678e-05, "loss": 0.0449, "step": 2340 }, { "grad_norm": 1.021741509437561, "learning_rate": 4.698e-05, "loss": 0.0468, "step": 2350 }, { "grad_norm": 0.6840607523918152, "learning_rate": 4.718e-05, "loss": 0.0445, "step": 2360 }, { "grad_norm": 1.2649844884872437, "learning_rate": 4.7380000000000004e-05, "loss": 0.0435, "step": 2370 }, { "grad_norm": 0.9851300716400146, "learning_rate": 4.758e-05, "loss": 0.0471, "step": 2380 }, { "grad_norm": 0.8768832087516785, "learning_rate": 4.778e-05, "loss": 0.0466, "step": 2390 }, { "grad_norm": 0.9125208258628845, "learning_rate": 4.7980000000000005e-05, "loss": 0.0448, "step": 2400 }, { "grad_norm": 0.8409683108329773, "learning_rate": 4.818e-05, "loss": 0.0472, "step": 2410 }, { "grad_norm": 0.7289535403251648, "learning_rate": 4.838e-05, "loss": 0.0449, "step": 2420 }, { "grad_norm": 0.8793744444847107, "learning_rate": 4.8580000000000006e-05, "loss": 0.0437, "step": 2430 }, { "grad_norm": 0.9384137988090515, "learning_rate": 4.8780000000000004e-05, "loss": 0.0443, "step": 2440 }, { "grad_norm": 0.971282958984375, "learning_rate": 4.898e-05, "loss": 0.0438, "step": 2450 }, { "grad_norm": 1.3028326034545898, "learning_rate": 4.918000000000001e-05, "loss": 0.0455, "step": 2460 }, { "grad_norm": 0.8423455357551575, "learning_rate": 4.9380000000000005e-05, "loss": 0.0442, "step": 2470 }, { "grad_norm": 0.8896216154098511, "learning_rate": 4.958e-05, "loss": 0.0455, "step": 2480 }, { "grad_norm": 0.9048638343811035, "learning_rate": 4.978e-05, "loss": 0.0423, "step": 2490 }, { "grad_norm": 0.8092187643051147, "learning_rate": 4.9980000000000006e-05, "loss": 0.0419, "step": 2500 }, { "grad_norm": 0.9305392503738403, "learning_rate": 5.0180000000000004e-05, "loss": 0.0432, "step": 2510 }, { "grad_norm": 0.8799822926521301, "learning_rate": 5.038e-05, "loss": 0.0422, "step": 2520 }, { "grad_norm": 1.0047564506530762, "learning_rate": 5.058000000000001e-05, "loss": 0.0424, "step": 2530 }, { "grad_norm": 0.5929383039474487, "learning_rate": 5.0780000000000005e-05, "loss": 0.04, "step": 2540 }, { "grad_norm": 0.891604483127594, "learning_rate": 5.098e-05, "loss": 0.0407, "step": 2550 }, { "grad_norm": 1.1013257503509521, "learning_rate": 5.118000000000001e-05, "loss": 0.0405, "step": 2560 }, { "grad_norm": 1.1506165266036987, "learning_rate": 5.1380000000000006e-05, "loss": 0.0424, "step": 2570 }, { "grad_norm": 0.7693191766738892, "learning_rate": 5.1580000000000004e-05, "loss": 0.0421, "step": 2580 }, { "grad_norm": 0.9716805219650269, "learning_rate": 5.178000000000001e-05, "loss": 0.0393, "step": 2590 }, { "grad_norm": 0.8578951954841614, "learning_rate": 5.198000000000001e-05, "loss": 0.0413, "step": 2600 }, { "grad_norm": 0.8432899713516235, "learning_rate": 5.2180000000000005e-05, "loss": 0.0426, "step": 2610 }, { "grad_norm": 0.7629929184913635, "learning_rate": 5.238000000000001e-05, "loss": 0.042, "step": 2620 }, { "grad_norm": 0.9284157156944275, "learning_rate": 5.258000000000001e-05, "loss": 0.0446, "step": 2630 }, { "grad_norm": 0.8470019698143005, "learning_rate": 5.2780000000000006e-05, "loss": 0.042, "step": 2640 }, { "grad_norm": 0.8543501496315002, "learning_rate": 5.2980000000000004e-05, "loss": 0.0395, "step": 2650 }, { "grad_norm": 1.070601463317871, "learning_rate": 5.318000000000001e-05, "loss": 0.0393, "step": 2660 }, { "grad_norm": 0.8198586702346802, "learning_rate": 5.338000000000001e-05, "loss": 0.0398, "step": 2670 }, { "grad_norm": 0.8744573593139648, "learning_rate": 5.3580000000000005e-05, "loss": 0.0386, "step": 2680 }, { "grad_norm": 0.8825362324714661, "learning_rate": 5.378e-05, "loss": 0.039, "step": 2690 }, { "grad_norm": 0.9426615834236145, "learning_rate": 5.3979999999999995e-05, "loss": 0.0404, "step": 2700 }, { "grad_norm": 0.8418384790420532, "learning_rate": 5.418e-05, "loss": 0.0425, "step": 2710 }, { "grad_norm": 0.9775055050849915, "learning_rate": 5.438e-05, "loss": 0.0384, "step": 2720 }, { "grad_norm": 0.6742515563964844, "learning_rate": 5.4579999999999996e-05, "loss": 0.0363, "step": 2730 }, { "grad_norm": 1.088884949684143, "learning_rate": 5.478e-05, "loss": 0.0377, "step": 2740 }, { "grad_norm": 0.8638517260551453, "learning_rate": 5.498e-05, "loss": 0.0399, "step": 2750 }, { "grad_norm": 0.6636028289794922, "learning_rate": 5.518e-05, "loss": 0.0383, "step": 2760 }, { "grad_norm": 1.0866152048110962, "learning_rate": 5.538e-05, "loss": 0.0394, "step": 2770 }, { "grad_norm": 0.7441537380218506, "learning_rate": 5.558e-05, "loss": 0.0415, "step": 2780 }, { "grad_norm": 0.6915196180343628, "learning_rate": 5.578e-05, "loss": 0.0421, "step": 2790 }, { "grad_norm": 0.8099533915519714, "learning_rate": 5.5979999999999996e-05, "loss": 0.0415, "step": 2800 }, { "grad_norm": 0.7210959196090698, "learning_rate": 5.618e-05, "loss": 0.0429, "step": 2810 }, { "grad_norm": 0.7099542617797852, "learning_rate": 5.638e-05, "loss": 0.0387, "step": 2820 }, { "grad_norm": 0.7680811882019043, "learning_rate": 5.658e-05, "loss": 0.0405, "step": 2830 }, { "grad_norm": 0.6909506320953369, "learning_rate": 5.678e-05, "loss": 0.0391, "step": 2840 }, { "grad_norm": 0.9611960053443909, "learning_rate": 5.698e-05, "loss": 0.0406, "step": 2850 }, { "grad_norm": 0.790065586566925, "learning_rate": 5.718e-05, "loss": 0.0405, "step": 2860 }, { "grad_norm": 0.747916579246521, "learning_rate": 5.738e-05, "loss": 0.0436, "step": 2870 }, { "grad_norm": 0.8887858390808105, "learning_rate": 5.758e-05, "loss": 0.0399, "step": 2880 }, { "grad_norm": 0.7003387212753296, "learning_rate": 5.778e-05, "loss": 0.0384, "step": 2890 }, { "grad_norm": 0.8714925050735474, "learning_rate": 5.7980000000000004e-05, "loss": 0.0371, "step": 2900 }, { "grad_norm": 0.6318824291229248, "learning_rate": 5.818e-05, "loss": 0.0387, "step": 2910 }, { "grad_norm": 0.9169565439224243, "learning_rate": 5.838e-05, "loss": 0.0396, "step": 2920 }, { "grad_norm": 0.8051956295967102, "learning_rate": 5.858e-05, "loss": 0.0408, "step": 2930 }, { "grad_norm": 0.7737332582473755, "learning_rate": 5.878e-05, "loss": 0.0373, "step": 2940 }, { "grad_norm": 0.902993381023407, "learning_rate": 5.898e-05, "loss": 0.038, "step": 2950 }, { "grad_norm": 0.7519251704216003, "learning_rate": 5.918e-05, "loss": 0.0385, "step": 2960 }, { "grad_norm": 0.8388000130653381, "learning_rate": 5.9380000000000004e-05, "loss": 0.0389, "step": 2970 }, { "grad_norm": 0.7788883447647095, "learning_rate": 5.958e-05, "loss": 0.0371, "step": 2980 }, { "grad_norm": 0.5345686674118042, "learning_rate": 5.978e-05, "loss": 0.0364, "step": 2990 }, { "grad_norm": 0.607315719127655, "learning_rate": 5.9980000000000005e-05, "loss": 0.0343, "step": 3000 }, { "grad_norm": 0.8990206718444824, "learning_rate": 6.018e-05, "loss": 0.0395, "step": 3010 }, { "grad_norm": 0.5942394137382507, "learning_rate": 6.038e-05, "loss": 0.039, "step": 3020 }, { "grad_norm": 0.7041111588478088, "learning_rate": 6.0580000000000006e-05, "loss": 0.0386, "step": 3030 }, { "grad_norm": 0.6832917332649231, "learning_rate": 6.0780000000000004e-05, "loss": 0.0395, "step": 3040 }, { "grad_norm": 0.8025203347206116, "learning_rate": 6.098e-05, "loss": 0.0376, "step": 3050 }, { "grad_norm": 1.0283249616622925, "learning_rate": 6.118000000000001e-05, "loss": 0.0396, "step": 3060 }, { "grad_norm": 0.6958217620849609, "learning_rate": 6.138e-05, "loss": 0.0396, "step": 3070 }, { "grad_norm": 0.725710928440094, "learning_rate": 6.158e-05, "loss": 0.0363, "step": 3080 }, { "grad_norm": 0.8787626028060913, "learning_rate": 6.178000000000001e-05, "loss": 0.0353, "step": 3090 }, { "grad_norm": 0.7741187810897827, "learning_rate": 6.198e-05, "loss": 0.0384, "step": 3100 }, { "grad_norm": 0.7863332033157349, "learning_rate": 6.218e-05, "loss": 0.0384, "step": 3110 }, { "grad_norm": 0.6019672155380249, "learning_rate": 6.238000000000001e-05, "loss": 0.0389, "step": 3120 }, { "grad_norm": 0.6868641376495361, "learning_rate": 6.258e-05, "loss": 0.0372, "step": 3130 }, { "grad_norm": 0.6221737861633301, "learning_rate": 6.278e-05, "loss": 0.0356, "step": 3140 }, { "grad_norm": 0.8705888986587524, "learning_rate": 6.298000000000001e-05, "loss": 0.0366, "step": 3150 }, { "grad_norm": 0.6587945818901062, "learning_rate": 6.318e-05, "loss": 0.0347, "step": 3160 }, { "grad_norm": 0.8290166258811951, "learning_rate": 6.338e-05, "loss": 0.0354, "step": 3170 }, { "grad_norm": 0.6827037334442139, "learning_rate": 6.358000000000001e-05, "loss": 0.0392, "step": 3180 }, { "grad_norm": 0.7603594064712524, "learning_rate": 6.378e-05, "loss": 0.0372, "step": 3190 }, { "grad_norm": 0.6990677714347839, "learning_rate": 6.398000000000001e-05, "loss": 0.0401, "step": 3200 }, { "grad_norm": 0.5744336843490601, "learning_rate": 6.418000000000001e-05, "loss": 0.0395, "step": 3210 }, { "grad_norm": 0.6342930197715759, "learning_rate": 6.438e-05, "loss": 0.037, "step": 3220 }, { "grad_norm": 0.6847115159034729, "learning_rate": 6.458000000000001e-05, "loss": 0.0369, "step": 3230 }, { "grad_norm": 0.6115511059761047, "learning_rate": 6.478000000000001e-05, "loss": 0.0351, "step": 3240 }, { "grad_norm": 0.7588299512863159, "learning_rate": 6.498e-05, "loss": 0.0389, "step": 3250 }, { "grad_norm": 0.7638617157936096, "learning_rate": 6.518000000000001e-05, "loss": 0.0385, "step": 3260 }, { "grad_norm": 0.7376441955566406, "learning_rate": 6.538000000000001e-05, "loss": 0.0359, "step": 3270 }, { "grad_norm": 0.7770403027534485, "learning_rate": 6.558e-05, "loss": 0.0375, "step": 3280 }, { "grad_norm": 0.4765743911266327, "learning_rate": 6.578000000000001e-05, "loss": 0.0364, "step": 3290 }, { "grad_norm": 0.747207760810852, "learning_rate": 6.598e-05, "loss": 0.0355, "step": 3300 }, { "grad_norm": 0.8424144983291626, "learning_rate": 6.618e-05, "loss": 0.0391, "step": 3310 }, { "grad_norm": 0.7289816737174988, "learning_rate": 6.638e-05, "loss": 0.0389, "step": 3320 }, { "grad_norm": 0.4892016351222992, "learning_rate": 6.658e-05, "loss": 0.0359, "step": 3330 }, { "grad_norm": 0.6606920957565308, "learning_rate": 6.678e-05, "loss": 0.0342, "step": 3340 }, { "grad_norm": 0.5098496675491333, "learning_rate": 6.698e-05, "loss": 0.0344, "step": 3350 }, { "grad_norm": 0.9544999599456787, "learning_rate": 6.718e-05, "loss": 0.0394, "step": 3360 }, { "grad_norm": 0.733701229095459, "learning_rate": 6.738e-05, "loss": 0.0364, "step": 3370 }, { "grad_norm": 0.640524685382843, "learning_rate": 6.758e-05, "loss": 0.0395, "step": 3380 }, { "grad_norm": 0.553862452507019, "learning_rate": 6.778e-05, "loss": 0.0372, "step": 3390 }, { "grad_norm": 0.6884357333183289, "learning_rate": 6.798e-05, "loss": 0.0342, "step": 3400 }, { "grad_norm": 0.7827233076095581, "learning_rate": 6.818e-05, "loss": 0.0353, "step": 3410 }, { "grad_norm": 0.6895797252655029, "learning_rate": 6.838e-05, "loss": 0.0335, "step": 3420 }, { "grad_norm": 0.604799747467041, "learning_rate": 6.858e-05, "loss": 0.0335, "step": 3430 }, { "grad_norm": 0.7528464198112488, "learning_rate": 6.878e-05, "loss": 0.0349, "step": 3440 }, { "grad_norm": 0.5410658121109009, "learning_rate": 6.898e-05, "loss": 0.0341, "step": 3450 }, { "grad_norm": 0.7903725504875183, "learning_rate": 6.918e-05, "loss": 0.038, "step": 3460 }, { "grad_norm": 0.5671918392181396, "learning_rate": 6.938e-05, "loss": 0.0351, "step": 3470 }, { "grad_norm": 0.6357104778289795, "learning_rate": 6.958e-05, "loss": 0.0355, "step": 3480 }, { "grad_norm": 0.7853618264198303, "learning_rate": 6.978e-05, "loss": 0.0349, "step": 3490 }, { "grad_norm": 0.5347936749458313, "learning_rate": 6.998e-05, "loss": 0.0342, "step": 3500 }, { "grad_norm": 0.6255249977111816, "learning_rate": 7.018e-05, "loss": 0.0341, "step": 3510 }, { "grad_norm": 0.7741848230361938, "learning_rate": 7.038e-05, "loss": 0.0329, "step": 3520 }, { "grad_norm": 0.6770904660224915, "learning_rate": 7.058e-05, "loss": 0.0322, "step": 3530 }, { "grad_norm": 0.6510072350502014, "learning_rate": 7.078e-05, "loss": 0.0355, "step": 3540 }, { "grad_norm": 0.5878069400787354, "learning_rate": 7.098e-05, "loss": 0.0351, "step": 3550 }, { "grad_norm": 0.7459149360656738, "learning_rate": 7.118e-05, "loss": 0.0333, "step": 3560 }, { "grad_norm": 0.5417954921722412, "learning_rate": 7.138e-05, "loss": 0.0329, "step": 3570 }, { "grad_norm": 0.5595616102218628, "learning_rate": 7.158e-05, "loss": 0.0342, "step": 3580 }, { "grad_norm": 0.8437157273292542, "learning_rate": 7.178000000000001e-05, "loss": 0.0357, "step": 3590 }, { "grad_norm": 0.8173953294754028, "learning_rate": 7.198e-05, "loss": 0.0344, "step": 3600 }, { "grad_norm": 0.56256502866745, "learning_rate": 7.218e-05, "loss": 0.0317, "step": 3610 }, { "grad_norm": 0.5586937665939331, "learning_rate": 7.238000000000001e-05, "loss": 0.0359, "step": 3620 }, { "grad_norm": 0.46543535590171814, "learning_rate": 7.258e-05, "loss": 0.0308, "step": 3630 }, { "grad_norm": 0.5688590407371521, "learning_rate": 7.278e-05, "loss": 0.034, "step": 3640 }, { "grad_norm": 0.6466200947761536, "learning_rate": 7.298000000000001e-05, "loss": 0.0333, "step": 3650 }, { "grad_norm": 0.5055368542671204, "learning_rate": 7.318e-05, "loss": 0.0362, "step": 3660 }, { "grad_norm": 0.5036585330963135, "learning_rate": 7.338e-05, "loss": 0.0315, "step": 3670 }, { "grad_norm": 0.6525483131408691, "learning_rate": 7.358000000000001e-05, "loss": 0.036, "step": 3680 }, { "grad_norm": 0.6674254536628723, "learning_rate": 7.378e-05, "loss": 0.0349, "step": 3690 }, { "grad_norm": 0.6171066761016846, "learning_rate": 7.398e-05, "loss": 0.0343, "step": 3700 }, { "grad_norm": 0.6047539710998535, "learning_rate": 7.418000000000001e-05, "loss": 0.0325, "step": 3710 }, { "grad_norm": 0.6130427718162537, "learning_rate": 7.438e-05, "loss": 0.0321, "step": 3720 }, { "grad_norm": 0.5674771070480347, "learning_rate": 7.458000000000001e-05, "loss": 0.0352, "step": 3730 }, { "grad_norm": 0.8181855082511902, "learning_rate": 7.478e-05, "loss": 0.0334, "step": 3740 }, { "grad_norm": 0.6460356712341309, "learning_rate": 7.498e-05, "loss": 0.0352, "step": 3750 }, { "grad_norm": 0.4272245466709137, "learning_rate": 7.518000000000001e-05, "loss": 0.0352, "step": 3760 }, { "grad_norm": 0.5770817995071411, "learning_rate": 7.538e-05, "loss": 0.0348, "step": 3770 }, { "grad_norm": 0.6084820032119751, "learning_rate": 7.558e-05, "loss": 0.0377, "step": 3780 }, { "grad_norm": 0.5465871691703796, "learning_rate": 7.578000000000001e-05, "loss": 0.0335, "step": 3790 }, { "grad_norm": 0.7102587223052979, "learning_rate": 7.598e-05, "loss": 0.0324, "step": 3800 }, { "grad_norm": 0.6593692898750305, "learning_rate": 7.618e-05, "loss": 0.0321, "step": 3810 }, { "grad_norm": 0.65223228931427, "learning_rate": 7.638000000000001e-05, "loss": 0.0336, "step": 3820 }, { "grad_norm": 0.711927592754364, "learning_rate": 7.658e-05, "loss": 0.0344, "step": 3830 }, { "grad_norm": 0.5448877811431885, "learning_rate": 7.678000000000001e-05, "loss": 0.0319, "step": 3840 }, { "grad_norm": 0.769862174987793, "learning_rate": 7.698000000000001e-05, "loss": 0.0335, "step": 3850 }, { "grad_norm": 0.563822329044342, "learning_rate": 7.718e-05, "loss": 0.0346, "step": 3860 }, { "grad_norm": 0.5054752230644226, "learning_rate": 7.738000000000001e-05, "loss": 0.0343, "step": 3870 }, { "grad_norm": 0.645063042640686, "learning_rate": 7.758000000000001e-05, "loss": 0.0331, "step": 3880 }, { "grad_norm": 0.46488696336746216, "learning_rate": 7.778e-05, "loss": 0.035, "step": 3890 }, { "grad_norm": 0.565007209777832, "learning_rate": 7.798000000000001e-05, "loss": 0.0313, "step": 3900 }, { "grad_norm": 0.436413049697876, "learning_rate": 7.818000000000001e-05, "loss": 0.029, "step": 3910 }, { "grad_norm": 0.7115194797515869, "learning_rate": 7.838e-05, "loss": 0.0314, "step": 3920 }, { "grad_norm": 0.4010336399078369, "learning_rate": 7.858000000000001e-05, "loss": 0.0321, "step": 3930 }, { "grad_norm": 0.6771782040596008, "learning_rate": 7.878e-05, "loss": 0.0317, "step": 3940 }, { "grad_norm": 0.7222819924354553, "learning_rate": 7.897999999999999e-05, "loss": 0.0335, "step": 3950 }, { "grad_norm": 0.5548766255378723, "learning_rate": 7.918e-05, "loss": 0.0312, "step": 3960 }, { "grad_norm": 0.5138027667999268, "learning_rate": 7.938e-05, "loss": 0.0316, "step": 3970 }, { "grad_norm": 0.639278769493103, "learning_rate": 7.958e-05, "loss": 0.0336, "step": 3980 }, { "grad_norm": 0.5436372756958008, "learning_rate": 7.978e-05, "loss": 0.036, "step": 3990 }, { "grad_norm": 0.6359093189239502, "learning_rate": 7.998e-05, "loss": 0.0304, "step": 4000 }, { "grad_norm": 0.4705268442630768, "learning_rate": 8.018e-05, "loss": 0.0294, "step": 4010 }, { "grad_norm": 0.4110027551651001, "learning_rate": 8.038e-05, "loss": 0.0326, "step": 4020 }, { "grad_norm": 0.5733879804611206, "learning_rate": 8.058e-05, "loss": 0.0305, "step": 4030 }, { "grad_norm": 0.5632413029670715, "learning_rate": 8.078e-05, "loss": 0.0316, "step": 4040 }, { "grad_norm": 0.5736110210418701, "learning_rate": 8.098e-05, "loss": 0.0327, "step": 4050 }, { "grad_norm": 0.6374561786651611, "learning_rate": 8.118e-05, "loss": 0.0306, "step": 4060 }, { "grad_norm": 0.6369447112083435, "learning_rate": 8.138e-05, "loss": 0.0321, "step": 4070 }, { "grad_norm": 0.584868311882019, "learning_rate": 8.158e-05, "loss": 0.0333, "step": 4080 }, { "grad_norm": 0.5931523442268372, "learning_rate": 8.178e-05, "loss": 0.0323, "step": 4090 }, { "grad_norm": 0.55528724193573, "learning_rate": 8.198e-05, "loss": 0.0308, "step": 4100 }, { "grad_norm": 0.5637499094009399, "learning_rate": 8.218e-05, "loss": 0.0296, "step": 4110 }, { "grad_norm": 0.519040584564209, "learning_rate": 8.238000000000001e-05, "loss": 0.0301, "step": 4120 }, { "grad_norm": 0.4287347197532654, "learning_rate": 8.258e-05, "loss": 0.0307, "step": 4130 }, { "grad_norm": 0.43425047397613525, "learning_rate": 8.278e-05, "loss": 0.0293, "step": 4140 }, { "grad_norm": 0.4664963185787201, "learning_rate": 8.298000000000001e-05, "loss": 0.0294, "step": 4150 }, { "grad_norm": 0.5327455997467041, "learning_rate": 8.318e-05, "loss": 0.0299, "step": 4160 }, { "grad_norm": 0.6587020754814148, "learning_rate": 8.338e-05, "loss": 0.0303, "step": 4170 }, { "grad_norm": 0.5751709938049316, "learning_rate": 8.358e-05, "loss": 0.031, "step": 4180 }, { "grad_norm": 0.39325276017189026, "learning_rate": 8.378e-05, "loss": 0.0281, "step": 4190 }, { "grad_norm": 0.4431546628475189, "learning_rate": 8.398e-05, "loss": 0.0319, "step": 4200 }, { "grad_norm": 0.5153656601905823, "learning_rate": 8.418e-05, "loss": 0.0315, "step": 4210 }, { "grad_norm": 0.41890907287597656, "learning_rate": 8.438e-05, "loss": 0.0296, "step": 4220 }, { "grad_norm": 0.5988590121269226, "learning_rate": 8.458e-05, "loss": 0.0325, "step": 4230 }, { "grad_norm": 0.5646542906761169, "learning_rate": 8.478e-05, "loss": 0.0294, "step": 4240 }, { "grad_norm": 0.5910056233406067, "learning_rate": 8.498e-05, "loss": 0.0312, "step": 4250 }, { "grad_norm": 0.49524375796318054, "learning_rate": 8.518000000000001e-05, "loss": 0.029, "step": 4260 }, { "grad_norm": 0.48719412088394165, "learning_rate": 8.538e-05, "loss": 0.0326, "step": 4270 }, { "grad_norm": 0.6445849537849426, "learning_rate": 8.558e-05, "loss": 0.0319, "step": 4280 }, { "grad_norm": 0.5413854718208313, "learning_rate": 8.578000000000001e-05, "loss": 0.0354, "step": 4290 }, { "grad_norm": 0.6239173412322998, "learning_rate": 8.598e-05, "loss": 0.0334, "step": 4300 }, { "grad_norm": 0.4903818666934967, "learning_rate": 8.618e-05, "loss": 0.0343, "step": 4310 }, { "grad_norm": 0.6508244872093201, "learning_rate": 8.638000000000001e-05, "loss": 0.0325, "step": 4320 }, { "grad_norm": 0.4555196166038513, "learning_rate": 8.658e-05, "loss": 0.0322, "step": 4330 }, { "grad_norm": 0.507361114025116, "learning_rate": 8.678e-05, "loss": 0.031, "step": 4340 }, { "grad_norm": 0.4571037292480469, "learning_rate": 8.698000000000001e-05, "loss": 0.0283, "step": 4350 }, { "grad_norm": 0.6517669558525085, "learning_rate": 8.718e-05, "loss": 0.0304, "step": 4360 }, { "grad_norm": 0.5762364864349365, "learning_rate": 8.738000000000001e-05, "loss": 0.0293, "step": 4370 }, { "grad_norm": 0.6050146818161011, "learning_rate": 8.758000000000001e-05, "loss": 0.0279, "step": 4380 }, { "grad_norm": 0.537484884262085, "learning_rate": 8.778e-05, "loss": 0.032, "step": 4390 }, { "grad_norm": 0.6340864896774292, "learning_rate": 8.798000000000001e-05, "loss": 0.0298, "step": 4400 }, { "grad_norm": 0.5194920301437378, "learning_rate": 8.818000000000001e-05, "loss": 0.0316, "step": 4410 }, { "grad_norm": 0.5937759280204773, "learning_rate": 8.838e-05, "loss": 0.0334, "step": 4420 }, { "grad_norm": 0.6219013929367065, "learning_rate": 8.858000000000001e-05, "loss": 0.0348, "step": 4430 }, { "grad_norm": 0.5902090072631836, "learning_rate": 8.878000000000001e-05, "loss": 0.0343, "step": 4440 }, { "grad_norm": 0.47670236229896545, "learning_rate": 8.898e-05, "loss": 0.0336, "step": 4450 }, { "grad_norm": 0.6736308932304382, "learning_rate": 8.918000000000001e-05, "loss": 0.0326, "step": 4460 }, { "grad_norm": 0.5302193760871887, "learning_rate": 8.938e-05, "loss": 0.0344, "step": 4470 }, { "grad_norm": 0.4352051019668579, "learning_rate": 8.958e-05, "loss": 0.0313, "step": 4480 }, { "grad_norm": 0.3776344954967499, "learning_rate": 8.978000000000001e-05, "loss": 0.03, "step": 4490 }, { "grad_norm": 0.5243827700614929, "learning_rate": 8.998e-05, "loss": 0.0324, "step": 4500 }, { "grad_norm": 0.5967336297035217, "learning_rate": 9.018000000000001e-05, "loss": 0.0335, "step": 4510 }, { "grad_norm": 0.3916470408439636, "learning_rate": 9.038000000000001e-05, "loss": 0.0315, "step": 4520 }, { "grad_norm": 0.5931420922279358, "learning_rate": 9.058e-05, "loss": 0.0306, "step": 4530 }, { "grad_norm": 0.5418990850448608, "learning_rate": 9.078000000000001e-05, "loss": 0.0306, "step": 4540 }, { "grad_norm": 0.6257407069206238, "learning_rate": 9.098000000000001e-05, "loss": 0.0308, "step": 4550 }, { "grad_norm": 0.5218918323516846, "learning_rate": 9.118e-05, "loss": 0.0312, "step": 4560 }, { "grad_norm": 0.3779914975166321, "learning_rate": 9.138e-05, "loss": 0.0321, "step": 4570 }, { "grad_norm": 0.597557544708252, "learning_rate": 9.158e-05, "loss": 0.0324, "step": 4580 }, { "grad_norm": 0.3704134225845337, "learning_rate": 9.178e-05, "loss": 0.0292, "step": 4590 }, { "grad_norm": 0.628476083278656, "learning_rate": 9.198e-05, "loss": 0.0304, "step": 4600 }, { "grad_norm": 0.4951136112213135, "learning_rate": 9.218e-05, "loss": 0.0294, "step": 4610 }, { "grad_norm": 0.47080492973327637, "learning_rate": 9.238e-05, "loss": 0.0294, "step": 4620 }, { "grad_norm": 0.5058643221855164, "learning_rate": 9.258e-05, "loss": 0.0314, "step": 4630 }, { "grad_norm": 0.46928760409355164, "learning_rate": 9.278e-05, "loss": 0.0321, "step": 4640 }, { "grad_norm": 0.41207289695739746, "learning_rate": 9.298e-05, "loss": 0.0293, "step": 4650 }, { "grad_norm": 0.36348530650138855, "learning_rate": 9.318e-05, "loss": 0.0266, "step": 4660 }, { "grad_norm": 0.5266436338424683, "learning_rate": 9.338e-05, "loss": 0.0304, "step": 4670 }, { "grad_norm": 0.5364354848861694, "learning_rate": 9.358e-05, "loss": 0.0328, "step": 4680 }, { "grad_norm": 0.5660241842269897, "learning_rate": 9.378e-05, "loss": 0.0336, "step": 4690 }, { "grad_norm": 0.32322123646736145, "learning_rate": 9.398e-05, "loss": 0.0351, "step": 4700 }, { "grad_norm": 0.5041892528533936, "learning_rate": 9.418e-05, "loss": 0.0321, "step": 4710 }, { "grad_norm": 0.6384232044219971, "learning_rate": 9.438e-05, "loss": 0.0305, "step": 4720 }, { "grad_norm": 0.5535484552383423, "learning_rate": 9.458e-05, "loss": 0.0309, "step": 4730 }, { "grad_norm": 0.5543848872184753, "learning_rate": 9.478e-05, "loss": 0.0364, "step": 4740 }, { "grad_norm": 0.4363686442375183, "learning_rate": 9.498e-05, "loss": 0.0292, "step": 4750 }, { "grad_norm": 0.4807997941970825, "learning_rate": 9.518000000000001e-05, "loss": 0.0321, "step": 4760 }, { "grad_norm": 0.5859388709068298, "learning_rate": 9.538e-05, "loss": 0.0282, "step": 4770 }, { "grad_norm": 0.4279617369174957, "learning_rate": 9.558e-05, "loss": 0.0315, "step": 4780 }, { "grad_norm": 0.48466935753822327, "learning_rate": 9.578000000000001e-05, "loss": 0.0321, "step": 4790 }, { "grad_norm": 0.4677855670452118, "learning_rate": 9.598e-05, "loss": 0.0302, "step": 4800 }, { "grad_norm": 0.5192621350288391, "learning_rate": 9.618e-05, "loss": 0.0279, "step": 4810 }, { "grad_norm": 0.5519497990608215, "learning_rate": 9.638000000000001e-05, "loss": 0.0289, "step": 4820 }, { "grad_norm": 0.5007696151733398, "learning_rate": 9.658e-05, "loss": 0.0283, "step": 4830 }, { "grad_norm": 0.4573926627635956, "learning_rate": 9.678e-05, "loss": 0.03, "step": 4840 }, { "grad_norm": 0.5700302124023438, "learning_rate": 9.698000000000001e-05, "loss": 0.0293, "step": 4850 }, { "grad_norm": 0.6004430651664734, "learning_rate": 9.718e-05, "loss": 0.0296, "step": 4860 }, { "grad_norm": 0.5468766093254089, "learning_rate": 9.738e-05, "loss": 0.0297, "step": 4870 }, { "grad_norm": 0.43055883049964905, "learning_rate": 9.758000000000001e-05, "loss": 0.0307, "step": 4880 }, { "grad_norm": 0.39746373891830444, "learning_rate": 9.778e-05, "loss": 0.0301, "step": 4890 }, { "grad_norm": 0.41065651178359985, "learning_rate": 9.798000000000001e-05, "loss": 0.0288, "step": 4900 }, { "grad_norm": 0.42012137174606323, "learning_rate": 9.818000000000001e-05, "loss": 0.0269, "step": 4910 }, { "grad_norm": 0.37514424324035645, "learning_rate": 9.838e-05, "loss": 0.0288, "step": 4920 }, { "grad_norm": 0.4821118414402008, "learning_rate": 9.858000000000001e-05, "loss": 0.0291, "step": 4930 }, { "grad_norm": 0.4632747769355774, "learning_rate": 9.878e-05, "loss": 0.0307, "step": 4940 }, { "grad_norm": 0.46894511580467224, "learning_rate": 9.898e-05, "loss": 0.0292, "step": 4950 }, { "grad_norm": 0.4519664943218231, "learning_rate": 9.918000000000001e-05, "loss": 0.0299, "step": 4960 }, { "grad_norm": 0.5138817429542542, "learning_rate": 9.938e-05, "loss": 0.029, "step": 4970 }, { "grad_norm": 0.65997713804245, "learning_rate": 9.958e-05, "loss": 0.0272, "step": 4980 }, { "grad_norm": 0.5350005626678467, "learning_rate": 9.978000000000001e-05, "loss": 0.0269, "step": 4990 }, { "grad_norm": 0.4568028151988983, "learning_rate": 9.998e-05, "loss": 0.0277, "step": 5000 }, { "grad_norm": 0.5644394755363464, "learning_rate": 9.999999778549045e-05, "loss": 0.0273, "step": 5010 }, { "grad_norm": 0.3524456322193146, "learning_rate": 9.999999013039593e-05, "loss": 0.0265, "step": 5020 }, { "grad_norm": 0.4231429994106293, "learning_rate": 9.999997700737766e-05, "loss": 0.0271, "step": 5030 }, { "grad_norm": 0.6364794373512268, "learning_rate": 9.999995841643709e-05, "loss": 0.0293, "step": 5040 }, { "grad_norm": 0.5041588544845581, "learning_rate": 9.999993435757623e-05, "loss": 0.0301, "step": 5050 }, { "grad_norm": 0.4282458424568176, "learning_rate": 9.999990483079773e-05, "loss": 0.0276, "step": 5060 }, { "grad_norm": 0.38724684715270996, "learning_rate": 9.999986983610481e-05, "loss": 0.0256, "step": 5070 }, { "grad_norm": 0.421709269285202, "learning_rate": 9.99998293735013e-05, "loss": 0.0263, "step": 5080 }, { "grad_norm": 0.5024977326393127, "learning_rate": 9.999978344299161e-05, "loss": 0.0296, "step": 5090 }, { "grad_norm": 0.3776308596134186, "learning_rate": 9.99997320445808e-05, "loss": 0.026, "step": 5100 }, { "grad_norm": 0.5071274042129517, "learning_rate": 9.999967517827444e-05, "loss": 0.0275, "step": 5110 }, { "grad_norm": 0.4356124699115753, "learning_rate": 9.999961284407879e-05, "loss": 0.026, "step": 5120 }, { "grad_norm": 0.3497796654701233, "learning_rate": 9.999954504200067e-05, "loss": 0.0255, "step": 5130 }, { "grad_norm": 0.5008280277252197, "learning_rate": 9.999947177204744e-05, "loss": 0.0254, "step": 5140 }, { "grad_norm": 0.5429554581642151, "learning_rate": 9.999939303422718e-05, "loss": 0.0261, "step": 5150 }, { "grad_norm": 0.4061000347137451, "learning_rate": 9.999930882854847e-05, "loss": 0.0257, "step": 5160 }, { "grad_norm": 0.48910266160964966, "learning_rate": 9.999921915502051e-05, "loss": 0.0265, "step": 5170 }, { "grad_norm": 0.5816081762313843, "learning_rate": 9.99991240136531e-05, "loss": 0.0254, "step": 5180 }, { "grad_norm": 0.38539668917655945, "learning_rate": 9.999902340445668e-05, "loss": 0.0248, "step": 5190 }, { "grad_norm": 0.5046334862709045, "learning_rate": 9.999891732744224e-05, "loss": 0.029, "step": 5200 }, { "grad_norm": 0.5540731549263, "learning_rate": 9.999880578262135e-05, "loss": 0.0263, "step": 5210 }, { "grad_norm": 0.43131741881370544, "learning_rate": 9.999868877000624e-05, "loss": 0.0252, "step": 5220 }, { "grad_norm": 0.4904332160949707, "learning_rate": 9.99985662896097e-05, "loss": 0.026, "step": 5230 }, { "grad_norm": 0.4706449806690216, "learning_rate": 9.999843834144513e-05, "loss": 0.0261, "step": 5240 }, { "grad_norm": 0.5267513990402222, "learning_rate": 9.99983049255265e-05, "loss": 0.0338, "step": 5250 }, { "grad_norm": 0.343492716550827, "learning_rate": 9.999816604186843e-05, "loss": 0.0268, "step": 5260 }, { "grad_norm": 0.32301095128059387, "learning_rate": 9.999802169048609e-05, "loss": 0.0285, "step": 5270 }, { "grad_norm": 0.3012272119522095, "learning_rate": 9.999787187139527e-05, "loss": 0.0253, "step": 5280 }, { "grad_norm": 0.3644024431705475, "learning_rate": 9.999771658461234e-05, "loss": 0.025, "step": 5290 }, { "grad_norm": 0.37684720754623413, "learning_rate": 9.999755583015431e-05, "loss": 0.0264, "step": 5300 }, { "grad_norm": 0.4297749698162079, "learning_rate": 9.999738960803874e-05, "loss": 0.026, "step": 5310 }, { "grad_norm": 0.4212551414966583, "learning_rate": 9.99972179182838e-05, "loss": 0.0282, "step": 5320 }, { "grad_norm": 0.32434460520744324, "learning_rate": 9.99970407609083e-05, "loss": 0.0244, "step": 5330 }, { "grad_norm": 0.49371930956840515, "learning_rate": 9.999685813593159e-05, "loss": 0.0275, "step": 5340 }, { "grad_norm": 0.41701656579971313, "learning_rate": 9.999667004337362e-05, "loss": 0.0272, "step": 5350 }, { "grad_norm": 0.43799638748168945, "learning_rate": 9.9996476483255e-05, "loss": 0.0286, "step": 5360 }, { "grad_norm": 0.36568671464920044, "learning_rate": 9.999627745559688e-05, "loss": 0.0303, "step": 5370 }, { "grad_norm": 0.4717998504638672, "learning_rate": 9.999607296042101e-05, "loss": 0.0277, "step": 5380 }, { "grad_norm": 0.38087159395217896, "learning_rate": 9.99958629977498e-05, "loss": 0.0238, "step": 5390 }, { "grad_norm": 0.46726328134536743, "learning_rate": 9.999564756760615e-05, "loss": 0.0256, "step": 5400 }, { "grad_norm": 0.4451892673969269, "learning_rate": 9.999542667001366e-05, "loss": 0.0251, "step": 5410 }, { "grad_norm": 0.4331384003162384, "learning_rate": 9.999520030499647e-05, "loss": 0.0243, "step": 5420 }, { "grad_norm": 0.45206892490386963, "learning_rate": 9.999496847257936e-05, "loss": 0.0273, "step": 5430 }, { "grad_norm": 0.505297064781189, "learning_rate": 9.999473117278764e-05, "loss": 0.0272, "step": 5440 }, { "grad_norm": 0.36388128995895386, "learning_rate": 9.999448840564731e-05, "loss": 0.0262, "step": 5450 }, { "grad_norm": 0.41026735305786133, "learning_rate": 9.999424017118488e-05, "loss": 0.0258, "step": 5460 }, { "grad_norm": 0.4279789924621582, "learning_rate": 9.999398646942751e-05, "loss": 0.0262, "step": 5470 }, { "grad_norm": 0.4262969493865967, "learning_rate": 9.999372730040296e-05, "loss": 0.0244, "step": 5480 }, { "grad_norm": 0.4336332678794861, "learning_rate": 9.999346266413953e-05, "loss": 0.0257, "step": 5490 }, { "grad_norm": 0.5243339538574219, "learning_rate": 9.99931925606662e-05, "loss": 0.0253, "step": 5500 }, { "grad_norm": 0.43605703115463257, "learning_rate": 9.99929169900125e-05, "loss": 0.0265, "step": 5510 }, { "grad_norm": 0.43022871017456055, "learning_rate": 9.999263595220855e-05, "loss": 0.026, "step": 5520 }, { "grad_norm": 0.3363441824913025, "learning_rate": 9.99923494472851e-05, "loss": 0.0255, "step": 5530 }, { "grad_norm": 0.585462749004364, "learning_rate": 9.999205747527348e-05, "loss": 0.026, "step": 5540 }, { "grad_norm": 0.4439209997653961, "learning_rate": 9.999176003620561e-05, "loss": 0.0317, "step": 5550 }, { "grad_norm": 0.4321342408657074, "learning_rate": 9.999145713011405e-05, "loss": 0.0289, "step": 5560 }, { "grad_norm": 0.38081446290016174, "learning_rate": 9.999114875703186e-05, "loss": 0.0281, "step": 5570 }, { "grad_norm": 0.5579205751419067, "learning_rate": 9.999083491699281e-05, "loss": 0.0279, "step": 5580 }, { "grad_norm": 0.3291715979576111, "learning_rate": 9.999051561003123e-05, "loss": 0.0273, "step": 5590 }, { "grad_norm": 0.46259555220603943, "learning_rate": 9.999019083618202e-05, "loss": 0.0274, "step": 5600 }, { "grad_norm": 0.36873337626457214, "learning_rate": 9.99898605954807e-05, "loss": 0.0251, "step": 5610 }, { "grad_norm": 0.3598043918609619, "learning_rate": 9.998952488796338e-05, "loss": 0.0291, "step": 5620 }, { "grad_norm": 0.4554140865802765, "learning_rate": 9.998918371366676e-05, "loss": 0.0257, "step": 5630 }, { "grad_norm": 0.36672282218933105, "learning_rate": 9.99888370726282e-05, "loss": 0.0262, "step": 5640 }, { "grad_norm": 0.4443296194076538, "learning_rate": 9.998848496488556e-05, "loss": 0.0278, "step": 5650 }, { "grad_norm": 0.3850274980068207, "learning_rate": 9.998812739047736e-05, "loss": 0.0262, "step": 5660 }, { "grad_norm": 0.4227592647075653, "learning_rate": 9.99877643494427e-05, "loss": 0.0255, "step": 5670 }, { "grad_norm": 0.37681159377098083, "learning_rate": 9.998739584182128e-05, "loss": 0.0285, "step": 5680 }, { "grad_norm": 0.5118816494941711, "learning_rate": 9.998702186765342e-05, "loss": 0.0277, "step": 5690 }, { "grad_norm": 0.41797110438346863, "learning_rate": 9.998664242698e-05, "loss": 0.0269, "step": 5700 }, { "grad_norm": 0.31711915135383606, "learning_rate": 9.998625751984251e-05, "loss": 0.026, "step": 5710 }, { "grad_norm": 0.4247218370437622, "learning_rate": 9.998586714628307e-05, "loss": 0.0246, "step": 5720 }, { "grad_norm": 0.29255521297454834, "learning_rate": 9.998547130634432e-05, "loss": 0.0255, "step": 5730 }, { "grad_norm": 0.36039289832115173, "learning_rate": 9.99850700000696e-05, "loss": 0.0235, "step": 5740 }, { "grad_norm": 0.44994354248046875, "learning_rate": 9.998466322750278e-05, "loss": 0.0235, "step": 5750 }, { "grad_norm": 0.25855234265327454, "learning_rate": 9.998425098868834e-05, "loss": 0.024, "step": 5760 }, { "grad_norm": 0.2773844599723816, "learning_rate": 9.998383328367136e-05, "loss": 0.0255, "step": 5770 }, { "grad_norm": 0.4123058021068573, "learning_rate": 9.99834101124975e-05, "loss": 0.0243, "step": 5780 }, { "grad_norm": 0.47598984837532043, "learning_rate": 9.998298147521309e-05, "loss": 0.0255, "step": 5790 }, { "grad_norm": 0.41940170526504517, "learning_rate": 9.998254737186496e-05, "loss": 0.0244, "step": 5800 }, { "grad_norm": 0.3085149824619293, "learning_rate": 9.99821078025006e-05, "loss": 0.0298, "step": 5810 }, { "grad_norm": 0.29555901885032654, "learning_rate": 9.998166276716807e-05, "loss": 0.0273, "step": 5820 }, { "grad_norm": 0.3924579620361328, "learning_rate": 9.998121226591606e-05, "loss": 0.0245, "step": 5830 }, { "grad_norm": 0.3428288698196411, "learning_rate": 9.998075629879382e-05, "loss": 0.0234, "step": 5840 }, { "grad_norm": 0.32596588134765625, "learning_rate": 9.99802948658512e-05, "loss": 0.024, "step": 5850 }, { "grad_norm": 0.2924344837665558, "learning_rate": 9.99798279671387e-05, "loss": 0.024, "step": 5860 }, { "grad_norm": 0.5519469976425171, "learning_rate": 9.997935560270734e-05, "loss": 0.0261, "step": 5870 }, { "grad_norm": 0.4803103506565094, "learning_rate": 9.997887777260879e-05, "loss": 0.025, "step": 5880 }, { "grad_norm": 0.42842745780944824, "learning_rate": 9.997839447689532e-05, "loss": 0.0228, "step": 5890 }, { "grad_norm": 0.3348039984703064, "learning_rate": 9.997790571561978e-05, "loss": 0.0249, "step": 5900 }, { "grad_norm": 0.4422380030155182, "learning_rate": 9.99774114888356e-05, "loss": 0.023, "step": 5910 }, { "grad_norm": 0.40746331214904785, "learning_rate": 9.997691179659684e-05, "loss": 0.0232, "step": 5920 }, { "grad_norm": 0.33477163314819336, "learning_rate": 9.997640663895815e-05, "loss": 0.0217, "step": 5930 }, { "grad_norm": 0.33074548840522766, "learning_rate": 9.997589601597477e-05, "loss": 0.0232, "step": 5940 }, { "grad_norm": 0.3518698513507843, "learning_rate": 9.997537992770252e-05, "loss": 0.0218, "step": 5950 }, { "grad_norm": 0.35154789686203003, "learning_rate": 9.997485837419788e-05, "loss": 0.0224, "step": 5960 }, { "grad_norm": 0.3214685022830963, "learning_rate": 9.997433135551786e-05, "loss": 0.0275, "step": 5970 }, { "grad_norm": 0.362356960773468, "learning_rate": 9.997379887172009e-05, "loss": 0.0258, "step": 5980 }, { "grad_norm": 0.35591787099838257, "learning_rate": 9.997326092286281e-05, "loss": 0.0241, "step": 5990 }, { "grad_norm": 0.3122444748878479, "learning_rate": 9.997271750900486e-05, "loss": 0.0239, "step": 6000 }, { "grad_norm": 0.2878078520298004, "learning_rate": 9.997216863020565e-05, "loss": 0.0257, "step": 6010 }, { "grad_norm": 0.30182477831840515, "learning_rate": 9.99716142865252e-05, "loss": 0.0241, "step": 6020 }, { "grad_norm": 0.3354875147342682, "learning_rate": 9.997105447802415e-05, "loss": 0.0244, "step": 6030 }, { "grad_norm": 0.3615817129611969, "learning_rate": 9.997048920476373e-05, "loss": 0.0219, "step": 6040 }, { "grad_norm": 0.3197365701198578, "learning_rate": 9.996991846680572e-05, "loss": 0.0225, "step": 6050 }, { "grad_norm": 0.40437784790992737, "learning_rate": 9.996934226421257e-05, "loss": 0.0223, "step": 6060 }, { "grad_norm": 0.2911413013935089, "learning_rate": 9.996876059704726e-05, "loss": 0.0224, "step": 6070 }, { "grad_norm": 0.3493189811706543, "learning_rate": 9.996817346537343e-05, "loss": 0.0226, "step": 6080 }, { "grad_norm": 0.43082720041275024, "learning_rate": 9.996758086925526e-05, "loss": 0.0223, "step": 6090 }, { "grad_norm": 0.3501473665237427, "learning_rate": 9.996698280875759e-05, "loss": 0.0234, "step": 6100 }, { "grad_norm": 0.35927942395210266, "learning_rate": 9.99663792839458e-05, "loss": 0.023, "step": 6110 }, { "grad_norm": 0.5549718737602234, "learning_rate": 9.99657702948859e-05, "loss": 0.0238, "step": 6120 }, { "grad_norm": 0.40139240026474, "learning_rate": 9.996515584164448e-05, "loss": 0.0221, "step": 6130 }, { "grad_norm": 0.2797819972038269, "learning_rate": 9.996453592428873e-05, "loss": 0.0254, "step": 6140 }, { "grad_norm": 0.32555073499679565, "learning_rate": 9.996391054288646e-05, "loss": 0.0235, "step": 6150 }, { "grad_norm": 0.5943684577941895, "learning_rate": 9.996327969750605e-05, "loss": 0.0241, "step": 6160 }, { "grad_norm": 0.3752487003803253, "learning_rate": 9.996264338821649e-05, "loss": 0.025, "step": 6170 }, { "grad_norm": 0.29227644205093384, "learning_rate": 9.996200161508735e-05, "loss": 0.0228, "step": 6180 }, { "grad_norm": 0.26861056685447693, "learning_rate": 9.996135437818885e-05, "loss": 0.0208, "step": 6190 }, { "grad_norm": 0.4290105104446411, "learning_rate": 9.996070167759175e-05, "loss": 0.0209, "step": 6200 }, { "grad_norm": 0.3552893400192261, "learning_rate": 9.996004351336743e-05, "loss": 0.0204, "step": 6210 }, { "grad_norm": 0.3955668807029724, "learning_rate": 9.995937988558785e-05, "loss": 0.0208, "step": 6220 }, { "grad_norm": 0.30653905868530273, "learning_rate": 9.995871079432561e-05, "loss": 0.0191, "step": 6230 }, { "grad_norm": 0.3273443579673767, "learning_rate": 9.995803623965389e-05, "loss": 0.0218, "step": 6240 }, { "grad_norm": 0.4455917775630951, "learning_rate": 9.995735622164641e-05, "loss": 0.0229, "step": 6250 }, { "grad_norm": 0.3069714605808258, "learning_rate": 9.995667074037758e-05, "loss": 0.0223, "step": 6260 }, { "grad_norm": 0.3425945043563843, "learning_rate": 9.995597979592232e-05, "loss": 0.0204, "step": 6270 }, { "grad_norm": 0.3615414798259735, "learning_rate": 9.995528338835625e-05, "loss": 0.0226, "step": 6280 }, { "grad_norm": 0.3400874435901642, "learning_rate": 9.995458151775547e-05, "loss": 0.0228, "step": 6290 }, { "grad_norm": 0.41797780990600586, "learning_rate": 9.995387418419677e-05, "loss": 0.0217, "step": 6300 }, { "grad_norm": 0.33953192830085754, "learning_rate": 9.99531613877575e-05, "loss": 0.0202, "step": 6310 }, { "grad_norm": 0.34940478205680847, "learning_rate": 9.995244312851559e-05, "loss": 0.0221, "step": 6320 }, { "grad_norm": 0.28944751620292664, "learning_rate": 9.995171940654961e-05, "loss": 0.0213, "step": 6330 }, { "grad_norm": 0.43870216608047485, "learning_rate": 9.995099022193871e-05, "loss": 0.0215, "step": 6340 }, { "grad_norm": 0.30585211515426636, "learning_rate": 9.995025557476261e-05, "loss": 0.0213, "step": 6350 }, { "grad_norm": 0.2818921208381653, "learning_rate": 9.994951546510165e-05, "loss": 0.0227, "step": 6360 }, { "grad_norm": 0.4078730344772339, "learning_rate": 9.994876989303679e-05, "loss": 0.0217, "step": 6370 }, { "grad_norm": 0.40389347076416016, "learning_rate": 9.994801885864955e-05, "loss": 0.0236, "step": 6380 }, { "grad_norm": 0.311278760433197, "learning_rate": 9.994726236202205e-05, "loss": 0.025, "step": 6390 }, { "grad_norm": 0.41170766949653625, "learning_rate": 9.994650040323704e-05, "loss": 0.0249, "step": 6400 }, { "grad_norm": 0.34045401215553284, "learning_rate": 9.994573298237784e-05, "loss": 0.0264, "step": 6410 }, { "grad_norm": 0.48640140891075134, "learning_rate": 9.994496009952837e-05, "loss": 0.0267, "step": 6420 }, { "grad_norm": 0.36458516120910645, "learning_rate": 9.994418175477316e-05, "loss": 0.0266, "step": 6430 }, { "grad_norm": 0.3395686447620392, "learning_rate": 9.994339794819733e-05, "loss": 0.0259, "step": 6440 }, { "grad_norm": 0.3906483054161072, "learning_rate": 9.994260867988658e-05, "loss": 0.0249, "step": 6450 }, { "grad_norm": 0.46869608759880066, "learning_rate": 9.994181394992723e-05, "loss": 0.0248, "step": 6460 }, { "grad_norm": 0.3485550880432129, "learning_rate": 9.994101375840618e-05, "loss": 0.0249, "step": 6470 }, { "grad_norm": 0.29368147253990173, "learning_rate": 9.994020810541098e-05, "loss": 0.0235, "step": 6480 }, { "grad_norm": 0.34801164269447327, "learning_rate": 9.99393969910297e-05, "loss": 0.0233, "step": 6490 }, { "grad_norm": 0.3225986361503601, "learning_rate": 9.993858041535104e-05, "loss": 0.0231, "step": 6500 }, { "grad_norm": 0.28575190901756287, "learning_rate": 9.99377583784643e-05, "loss": 0.0223, "step": 6510 }, { "grad_norm": 0.333723783493042, "learning_rate": 9.993693088045939e-05, "loss": 0.0228, "step": 6520 }, { "grad_norm": 0.39766430854797363, "learning_rate": 9.99360979214268e-05, "loss": 0.0227, "step": 6530 }, { "grad_norm": 0.262091726064682, "learning_rate": 9.99352595014576e-05, "loss": 0.0204, "step": 6540 }, { "grad_norm": 0.36905229091644287, "learning_rate": 9.993441562064354e-05, "loss": 0.0226, "step": 6550 }, { "grad_norm": 0.2716531455516815, "learning_rate": 9.993356627907685e-05, "loss": 0.0218, "step": 6560 }, { "grad_norm": 0.3039698898792267, "learning_rate": 9.99327114768504e-05, "loss": 0.0237, "step": 6570 }, { "grad_norm": 0.33795052766799927, "learning_rate": 9.99318512140577e-05, "loss": 0.0238, "step": 6580 }, { "grad_norm": 0.41775211691856384, "learning_rate": 9.993098549079284e-05, "loss": 0.0238, "step": 6590 }, { "grad_norm": 0.36130961775779724, "learning_rate": 9.993011430715047e-05, "loss": 0.0229, "step": 6600 }, { "grad_norm": 0.4719183146953583, "learning_rate": 9.992923766322586e-05, "loss": 0.0227, "step": 6610 }, { "grad_norm": 0.3385705351829529, "learning_rate": 9.99283555591149e-05, "loss": 0.0224, "step": 6620 }, { "grad_norm": 0.416936457157135, "learning_rate": 9.992746799491404e-05, "loss": 0.0234, "step": 6630 }, { "grad_norm": 0.3626827597618103, "learning_rate": 9.992657497072033e-05, "loss": 0.0226, "step": 6640 }, { "grad_norm": 0.3189864754676819, "learning_rate": 9.992567648663147e-05, "loss": 0.0248, "step": 6650 }, { "grad_norm": 0.3454511761665344, "learning_rate": 9.992477254274568e-05, "loss": 0.0246, "step": 6660 }, { "grad_norm": 0.43259403109550476, "learning_rate": 9.992386313916183e-05, "loss": 0.0258, "step": 6670 }, { "grad_norm": 0.3271529972553253, "learning_rate": 9.992294827597934e-05, "loss": 0.0247, "step": 6680 }, { "grad_norm": 0.287916362285614, "learning_rate": 9.992202795329831e-05, "loss": 0.0216, "step": 6690 }, { "grad_norm": 0.31315359473228455, "learning_rate": 9.992110217121936e-05, "loss": 0.0237, "step": 6700 }, { "grad_norm": 0.2915042042732239, "learning_rate": 9.992017092984372e-05, "loss": 0.0228, "step": 6710 }, { "grad_norm": 0.38848090171813965, "learning_rate": 9.991923422927326e-05, "loss": 0.0244, "step": 6720 }, { "grad_norm": 0.42219841480255127, "learning_rate": 9.991829206961037e-05, "loss": 0.0214, "step": 6730 }, { "grad_norm": 0.4490721821784973, "learning_rate": 9.991734445095813e-05, "loss": 0.0234, "step": 6740 }, { "grad_norm": 0.3552895188331604, "learning_rate": 9.991639137342015e-05, "loss": 0.0216, "step": 6750 }, { "grad_norm": 0.3143540322780609, "learning_rate": 9.991543283710064e-05, "loss": 0.0199, "step": 6760 }, { "grad_norm": 0.30248597264289856, "learning_rate": 9.991446884210445e-05, "loss": 0.0202, "step": 6770 }, { "grad_norm": 0.3862917423248291, "learning_rate": 9.9913499388537e-05, "loss": 0.0216, "step": 6780 }, { "grad_norm": 0.29088759422302246, "learning_rate": 9.99125244765043e-05, "loss": 0.0214, "step": 6790 }, { "grad_norm": 0.34585270285606384, "learning_rate": 9.991154410611296e-05, "loss": 0.0227, "step": 6800 }, { "grad_norm": 0.349284827709198, "learning_rate": 9.99105582774702e-05, "loss": 0.0192, "step": 6810 }, { "grad_norm": 0.2346998006105423, "learning_rate": 9.990956699068384e-05, "loss": 0.0201, "step": 6820 }, { "grad_norm": 0.3574236333370209, "learning_rate": 9.990857024586224e-05, "loss": 0.0203, "step": 6830 }, { "grad_norm": 0.3526395857334137, "learning_rate": 9.990756804311446e-05, "loss": 0.0217, "step": 6840 }, { "grad_norm": 0.33956801891326904, "learning_rate": 9.990656038255006e-05, "loss": 0.0221, "step": 6850 }, { "grad_norm": 0.32819968461990356, "learning_rate": 9.990554726427926e-05, "loss": 0.0215, "step": 6860 }, { "grad_norm": 0.3939273953437805, "learning_rate": 9.990452868841284e-05, "loss": 0.0219, "step": 6870 }, { "grad_norm": 0.36858758330345154, "learning_rate": 9.99035046550622e-05, "loss": 0.0219, "step": 6880 }, { "grad_norm": 0.46127235889434814, "learning_rate": 9.99024751643393e-05, "loss": 0.0204, "step": 6890 }, { "grad_norm": 0.28423747420310974, "learning_rate": 9.990144021635677e-05, "loss": 0.0204, "step": 6900 }, { "grad_norm": 0.3164880573749542, "learning_rate": 9.990039981122775e-05, "loss": 0.0205, "step": 6910 }, { "grad_norm": 0.37718653678894043, "learning_rate": 9.989935394906602e-05, "loss": 0.0219, "step": 6920 }, { "grad_norm": 0.28659796714782715, "learning_rate": 9.989830262998598e-05, "loss": 0.0184, "step": 6930 }, { "grad_norm": 0.3209061324596405, "learning_rate": 9.989724585410259e-05, "loss": 0.0213, "step": 6940 }, { "grad_norm": 0.2850815951824188, "learning_rate": 9.989618362153139e-05, "loss": 0.0199, "step": 6950 }, { "grad_norm": 0.2577317953109741, "learning_rate": 9.989511593238859e-05, "loss": 0.0211, "step": 6960 }, { "grad_norm": 0.3463124632835388, "learning_rate": 9.98940427867909e-05, "loss": 0.0188, "step": 6970 }, { "grad_norm": 0.27325567603111267, "learning_rate": 9.989296418485573e-05, "loss": 0.0199, "step": 6980 }, { "grad_norm": 0.34366700053215027, "learning_rate": 9.989188012670101e-05, "loss": 0.023, "step": 6990 }, { "grad_norm": 0.24359220266342163, "learning_rate": 9.989079061244528e-05, "loss": 0.0203, "step": 7000 }, { "grad_norm": 0.287850946187973, "learning_rate": 9.988969564220769e-05, "loss": 0.0221, "step": 7010 }, { "grad_norm": 0.2846655547618866, "learning_rate": 9.988859521610801e-05, "loss": 0.019, "step": 7020 }, { "grad_norm": 0.28150951862335205, "learning_rate": 9.988748933426656e-05, "loss": 0.0212, "step": 7030 }, { "grad_norm": 0.3349640369415283, "learning_rate": 9.988637799680428e-05, "loss": 0.018, "step": 7040 }, { "grad_norm": 0.2609870433807373, "learning_rate": 9.98852612038427e-05, "loss": 0.0187, "step": 7050 }, { "grad_norm": 0.3491242527961731, "learning_rate": 9.988413895550397e-05, "loss": 0.0202, "step": 7060 }, { "grad_norm": 0.2620899975299835, "learning_rate": 9.98830112519108e-05, "loss": 0.0197, "step": 7070 }, { "grad_norm": 0.35392022132873535, "learning_rate": 9.98818780931865e-05, "loss": 0.0191, "step": 7080 }, { "grad_norm": 0.2575293183326721, "learning_rate": 9.988073947945502e-05, "loss": 0.0195, "step": 7090 }, { "grad_norm": 0.29257574677467346, "learning_rate": 9.987959541084087e-05, "loss": 0.0185, "step": 7100 }, { "grad_norm": 0.3553456962108612, "learning_rate": 9.987844588746915e-05, "loss": 0.0206, "step": 7110 }, { "grad_norm": 0.2824777066707611, "learning_rate": 9.987729090946558e-05, "loss": 0.0203, "step": 7120 }, { "grad_norm": 0.33149823546409607, "learning_rate": 9.987613047695647e-05, "loss": 0.0194, "step": 7130 }, { "grad_norm": 0.436980664730072, "learning_rate": 9.987496459006871e-05, "loss": 0.0205, "step": 7140 }, { "grad_norm": 0.3107277452945709, "learning_rate": 9.987379324892982e-05, "loss": 0.02, "step": 7150 }, { "grad_norm": 0.35453248023986816, "learning_rate": 9.987261645366788e-05, "loss": 0.0224, "step": 7160 }, { "grad_norm": 0.2748722732067108, "learning_rate": 9.987143420441158e-05, "loss": 0.0188, "step": 7170 }, { "grad_norm": 0.3042769432067871, "learning_rate": 9.987024650129022e-05, "loss": 0.0179, "step": 7180 }, { "grad_norm": 0.3619014024734497, "learning_rate": 9.986905334443368e-05, "loss": 0.0192, "step": 7190 }, { "grad_norm": 0.306188702583313, "learning_rate": 9.986785473397245e-05, "loss": 0.0216, "step": 7200 }, { "grad_norm": 0.23124317824840546, "learning_rate": 9.98666506700376e-05, "loss": 0.0199, "step": 7210 }, { "grad_norm": 0.370084673166275, "learning_rate": 9.986544115276081e-05, "loss": 0.0222, "step": 7220 }, { "grad_norm": 0.2566327452659607, "learning_rate": 9.986422618227433e-05, "loss": 0.0217, "step": 7230 }, { "grad_norm": 0.32059773802757263, "learning_rate": 9.986300575871106e-05, "loss": 0.0204, "step": 7240 }, { "grad_norm": 0.3489284813404083, "learning_rate": 9.986177988220444e-05, "loss": 0.0217, "step": 7250 }, { "grad_norm": 0.3318527340888977, "learning_rate": 9.986054855288856e-05, "loss": 0.0216, "step": 7260 }, { "grad_norm": 0.3417718708515167, "learning_rate": 9.985931177089802e-05, "loss": 0.0187, "step": 7270 }, { "grad_norm": 0.2371060699224472, "learning_rate": 9.985806953636814e-05, "loss": 0.0221, "step": 7280 }, { "grad_norm": 0.2601771652698517, "learning_rate": 9.985682184943471e-05, "loss": 0.0199, "step": 7290 }, { "grad_norm": 0.28185680508613586, "learning_rate": 9.98555687102342e-05, "loss": 0.0193, "step": 7300 }, { "grad_norm": 0.24549731612205505, "learning_rate": 9.985431011890367e-05, "loss": 0.0194, "step": 7310 }, { "grad_norm": 0.3810310959815979, "learning_rate": 9.985304607558075e-05, "loss": 0.0192, "step": 7320 }, { "grad_norm": 0.3539576232433319, "learning_rate": 9.985177658040364e-05, "loss": 0.0179, "step": 7330 }, { "grad_norm": 0.33971911668777466, "learning_rate": 9.985050163351119e-05, "loss": 0.0195, "step": 7340 }, { "grad_norm": 0.29823407530784607, "learning_rate": 9.984922123504286e-05, "loss": 0.0185, "step": 7350 }, { "grad_norm": 0.26644206047058105, "learning_rate": 9.984793538513862e-05, "loss": 0.0177, "step": 7360 }, { "grad_norm": 0.32062622904777527, "learning_rate": 9.984664408393912e-05, "loss": 0.018, "step": 7370 }, { "grad_norm": 0.3298834562301636, "learning_rate": 9.984534733158556e-05, "loss": 0.0204, "step": 7380 }, { "grad_norm": 0.3816513419151306, "learning_rate": 9.984404512821977e-05, "loss": 0.0184, "step": 7390 }, { "grad_norm": 0.2893332242965698, "learning_rate": 9.984273747398411e-05, "loss": 0.0196, "step": 7400 }, { "grad_norm": 0.2794543504714966, "learning_rate": 9.984142436902165e-05, "loss": 0.0168, "step": 7410 }, { "grad_norm": 0.4255591928958893, "learning_rate": 9.984010581347596e-05, "loss": 0.0184, "step": 7420 }, { "grad_norm": 0.3247465193271637, "learning_rate": 9.983878180749121e-05, "loss": 0.0173, "step": 7430 }, { "grad_norm": 0.25612232089042664, "learning_rate": 9.983745235121222e-05, "loss": 0.0191, "step": 7440 }, { "grad_norm": 0.38008326292037964, "learning_rate": 9.983611744478438e-05, "loss": 0.0195, "step": 7450 }, { "grad_norm": 0.30437397956848145, "learning_rate": 9.983477708835365e-05, "loss": 0.0198, "step": 7460 }, { "grad_norm": 0.31053197383880615, "learning_rate": 9.983343128206664e-05, "loss": 0.0191, "step": 7470 }, { "grad_norm": 0.3771955966949463, "learning_rate": 9.983208002607049e-05, "loss": 0.0174, "step": 7480 }, { "grad_norm": 0.40361288189888, "learning_rate": 9.9830723320513e-05, "loss": 0.0184, "step": 7490 }, { "grad_norm": 0.4022625684738159, "learning_rate": 9.982936116554254e-05, "loss": 0.0186, "step": 7500 }, { "grad_norm": 0.4314488470554352, "learning_rate": 9.982799356130803e-05, "loss": 0.0221, "step": 7510 }, { "grad_norm": 0.25763651728630066, "learning_rate": 9.982662050795908e-05, "loss": 0.0187, "step": 7520 }, { "grad_norm": 0.23892146348953247, "learning_rate": 9.982524200564583e-05, "loss": 0.02, "step": 7530 }, { "grad_norm": 0.3567933440208435, "learning_rate": 9.982385805451901e-05, "loss": 0.0218, "step": 7540 }, { "grad_norm": 0.3633735179901123, "learning_rate": 9.982246865472998e-05, "loss": 0.0218, "step": 7550 }, { "grad_norm": 0.2628510892391205, "learning_rate": 9.982107380643069e-05, "loss": 0.0222, "step": 7560 }, { "grad_norm": 0.3215087354183197, "learning_rate": 9.981967350977368e-05, "loss": 0.0195, "step": 7570 }, { "grad_norm": 0.23922666907310486, "learning_rate": 9.981826776491208e-05, "loss": 0.019, "step": 7580 }, { "grad_norm": 0.30434489250183105, "learning_rate": 9.98168565719996e-05, "loss": 0.0183, "step": 7590 }, { "grad_norm": 0.26028624176979065, "learning_rate": 9.98154399311906e-05, "loss": 0.0203, "step": 7600 }, { "grad_norm": 0.257759153842926, "learning_rate": 9.981401784263997e-05, "loss": 0.019, "step": 7610 }, { "grad_norm": 0.24626532196998596, "learning_rate": 9.981259030650326e-05, "loss": 0.0192, "step": 7620 }, { "grad_norm": 0.25954270362854004, "learning_rate": 9.981115732293655e-05, "loss": 0.0177, "step": 7630 }, { "grad_norm": 0.2578761875629425, "learning_rate": 9.980971889209659e-05, "loss": 0.0194, "step": 7640 }, { "grad_norm": 0.23871487379074097, "learning_rate": 9.980827501414064e-05, "loss": 0.0194, "step": 7650 }, { "grad_norm": 0.27502280473709106, "learning_rate": 9.980682568922663e-05, "loss": 0.0166, "step": 7660 }, { "grad_norm": 0.3007548451423645, "learning_rate": 9.980537091751304e-05, "loss": 0.0186, "step": 7670 }, { "grad_norm": 0.3150603175163269, "learning_rate": 9.980391069915897e-05, "loss": 0.0203, "step": 7680 }, { "grad_norm": 0.2945225238800049, "learning_rate": 9.98024450343241e-05, "loss": 0.0208, "step": 7690 }, { "grad_norm": 0.27505093812942505, "learning_rate": 9.980097392316872e-05, "loss": 0.0211, "step": 7700 }, { "grad_norm": 0.34897759556770325, "learning_rate": 9.97994973658537e-05, "loss": 0.0193, "step": 7710 }, { "grad_norm": 0.27814221382141113, "learning_rate": 9.979801536254054e-05, "loss": 0.0187, "step": 7720 }, { "grad_norm": 0.333171010017395, "learning_rate": 9.979652791339127e-05, "loss": 0.0173, "step": 7730 }, { "grad_norm": 0.3978307545185089, "learning_rate": 9.97950350185686e-05, "loss": 0.0179, "step": 7740 }, { "grad_norm": 0.26187318563461304, "learning_rate": 9.979353667823574e-05, "loss": 0.0214, "step": 7750 }, { "grad_norm": 0.3258225619792938, "learning_rate": 9.979203289255658e-05, "loss": 0.0199, "step": 7760 }, { "grad_norm": 0.329476535320282, "learning_rate": 9.979052366169557e-05, "loss": 0.0209, "step": 7770 }, { "grad_norm": 0.20415279269218445, "learning_rate": 9.978900898581775e-05, "loss": 0.0209, "step": 7780 }, { "grad_norm": 0.3139052987098694, "learning_rate": 9.978748886508875e-05, "loss": 0.0218, "step": 7790 }, { "grad_norm": 0.24102438986301422, "learning_rate": 9.978596329967484e-05, "loss": 0.0193, "step": 7800 }, { "grad_norm": 0.42182081937789917, "learning_rate": 9.978443228974284e-05, "loss": 0.0196, "step": 7810 }, { "grad_norm": 0.26666098833084106, "learning_rate": 9.978289583546015e-05, "loss": 0.0181, "step": 7820 }, { "grad_norm": 0.37027299404144287, "learning_rate": 9.978135393699484e-05, "loss": 0.0205, "step": 7830 }, { "grad_norm": 0.389800488948822, "learning_rate": 9.977980659451548e-05, "loss": 0.0228, "step": 7840 }, { "grad_norm": 0.46813610196113586, "learning_rate": 9.977825380819135e-05, "loss": 0.0229, "step": 7850 }, { "grad_norm": 0.32199352979660034, "learning_rate": 9.97766955781922e-05, "loss": 0.0214, "step": 7860 }, { "grad_norm": 0.31542009115219116, "learning_rate": 9.977513190468848e-05, "loss": 0.0203, "step": 7870 }, { "grad_norm": 0.30758538842201233, "learning_rate": 9.977356278785116e-05, "loss": 0.0209, "step": 7880 }, { "grad_norm": 0.27027931809425354, "learning_rate": 9.977198822785184e-05, "loss": 0.0208, "step": 7890 }, { "grad_norm": 0.3520703613758087, "learning_rate": 9.977040822486273e-05, "loss": 0.0235, "step": 7900 }, { "grad_norm": 0.2837248742580414, "learning_rate": 9.97688227790566e-05, "loss": 0.0209, "step": 7910 }, { "grad_norm": 0.2720511853694916, "learning_rate": 9.976723189060684e-05, "loss": 0.0206, "step": 7920 }, { "grad_norm": 0.274930864572525, "learning_rate": 9.976563555968742e-05, "loss": 0.0208, "step": 7930 }, { "grad_norm": 0.30582451820373535, "learning_rate": 9.976403378647292e-05, "loss": 0.0198, "step": 7940 }, { "grad_norm": 0.37607356905937195, "learning_rate": 9.97624265711385e-05, "loss": 0.0205, "step": 7950 }, { "grad_norm": 0.301155686378479, "learning_rate": 9.976081391385993e-05, "loss": 0.0197, "step": 7960 }, { "grad_norm": 0.3497355580329895, "learning_rate": 9.975919581481356e-05, "loss": 0.0195, "step": 7970 }, { "grad_norm": 0.34948280453681946, "learning_rate": 9.975757227417634e-05, "loss": 0.0195, "step": 7980 }, { "grad_norm": 0.37196457386016846, "learning_rate": 9.975594329212586e-05, "loss": 0.0203, "step": 7990 }, { "grad_norm": 0.3156779408454895, "learning_rate": 9.97543088688402e-05, "loss": 0.0192, "step": 8000 }, { "grad_norm": 0.3664815425872803, "learning_rate": 9.975266900449814e-05, "loss": 0.0185, "step": 8010 }, { "grad_norm": 0.21214348077774048, "learning_rate": 9.975102369927898e-05, "loss": 0.0177, "step": 8020 }, { "grad_norm": 0.2834559381008148, "learning_rate": 9.974937295336269e-05, "loss": 0.0196, "step": 8030 }, { "grad_norm": 0.2223299741744995, "learning_rate": 9.974771676692975e-05, "loss": 0.019, "step": 8040 }, { "grad_norm": 0.31630441546440125, "learning_rate": 9.974605514016131e-05, "loss": 0.0198, "step": 8050 }, { "grad_norm": 0.2838837206363678, "learning_rate": 9.974438807323907e-05, "loss": 0.0188, "step": 8060 }, { "grad_norm": 0.3149281442165375, "learning_rate": 9.974271556634535e-05, "loss": 0.0184, "step": 8070 }, { "grad_norm": 0.2038646638393402, "learning_rate": 9.974103761966302e-05, "loss": 0.018, "step": 8080 }, { "grad_norm": 0.2556861937046051, "learning_rate": 9.973935423337563e-05, "loss": 0.0185, "step": 8090 }, { "grad_norm": 0.32131287455558777, "learning_rate": 9.973766540766722e-05, "loss": 0.0205, "step": 8100 }, { "grad_norm": 0.3030673861503601, "learning_rate": 9.97359711427225e-05, "loss": 0.0202, "step": 8110 }, { "grad_norm": 0.4263915717601776, "learning_rate": 9.973427143872677e-05, "loss": 0.0181, "step": 8120 }, { "grad_norm": 0.3235431909561157, "learning_rate": 9.973256629586589e-05, "loss": 0.0208, "step": 8130 }, { "grad_norm": 0.24135863780975342, "learning_rate": 9.973085571432632e-05, "loss": 0.0199, "step": 8140 }, { "grad_norm": 0.28634727001190186, "learning_rate": 9.972913969429513e-05, "loss": 0.0219, "step": 8150 }, { "grad_norm": 0.21846061944961548, "learning_rate": 9.972741823596e-05, "loss": 0.0178, "step": 8160 }, { "grad_norm": 0.3792208433151245, "learning_rate": 9.972569133950917e-05, "loss": 0.0187, "step": 8170 }, { "grad_norm": 0.315467894077301, "learning_rate": 9.972395900513151e-05, "loss": 0.0172, "step": 8180 }, { "grad_norm": 0.2661696970462799, "learning_rate": 9.972222123301645e-05, "loss": 0.0177, "step": 8190 }, { "grad_norm": 0.2705479562282562, "learning_rate": 9.972047802335403e-05, "loss": 0.0201, "step": 8200 }, { "grad_norm": 0.2746221423149109, "learning_rate": 9.971872937633488e-05, "loss": 0.018, "step": 8210 }, { "grad_norm": 0.21674226224422455, "learning_rate": 9.971697529215024e-05, "loss": 0.018, "step": 8220 }, { "grad_norm": 0.245545893907547, "learning_rate": 9.971521577099192e-05, "loss": 0.0186, "step": 8230 }, { "grad_norm": 0.36647629737854004, "learning_rate": 9.971345081305236e-05, "loss": 0.0194, "step": 8240 }, { "grad_norm": 0.3345620036125183, "learning_rate": 9.971168041852456e-05, "loss": 0.0186, "step": 8250 }, { "grad_norm": 0.2692772448062897, "learning_rate": 9.970990458760215e-05, "loss": 0.0181, "step": 8260 }, { "grad_norm": 0.16554459929466248, "learning_rate": 9.970812332047929e-05, "loss": 0.0174, "step": 8270 }, { "grad_norm": 0.3065255284309387, "learning_rate": 9.97063366173508e-05, "loss": 0.0174, "step": 8280 }, { "grad_norm": 0.24563468992710114, "learning_rate": 9.970454447841207e-05, "loss": 0.0185, "step": 8290 }, { "grad_norm": 0.2021968811750412, "learning_rate": 9.970274690385909e-05, "loss": 0.0192, "step": 8300 }, { "grad_norm": 0.26922428607940674, "learning_rate": 9.970094389388844e-05, "loss": 0.0163, "step": 8310 }, { "grad_norm": 0.20902825891971588, "learning_rate": 9.969913544869728e-05, "loss": 0.0178, "step": 8320 }, { "grad_norm": 0.250581830739975, "learning_rate": 9.96973215684834e-05, "loss": 0.0175, "step": 8330 }, { "grad_norm": 0.2839232385158539, "learning_rate": 9.969550225344513e-05, "loss": 0.0195, "step": 8340 }, { "grad_norm": 0.33806636929512024, "learning_rate": 9.969367750378147e-05, "loss": 0.0202, "step": 8350 }, { "grad_norm": 0.25317683815956116, "learning_rate": 9.969184731969194e-05, "loss": 0.0181, "step": 8360 }, { "grad_norm": 0.3361537754535675, "learning_rate": 9.96900117013767e-05, "loss": 0.0177, "step": 8370 }, { "grad_norm": 0.30429309606552124, "learning_rate": 9.96881706490365e-05, "loss": 0.0188, "step": 8380 }, { "grad_norm": 0.28665247559547424, "learning_rate": 9.968632416287265e-05, "loss": 0.0177, "step": 8390 }, { "grad_norm": 0.26523521542549133, "learning_rate": 9.96844722430871e-05, "loss": 0.0189, "step": 8400 }, { "grad_norm": 0.2917472720146179, "learning_rate": 9.968261488988235e-05, "loss": 0.0202, "step": 8410 }, { "grad_norm": 0.24319495260715485, "learning_rate": 9.968075210346155e-05, "loss": 0.018, "step": 8420 }, { "grad_norm": 0.34090954065322876, "learning_rate": 9.967888388402839e-05, "loss": 0.0188, "step": 8430 }, { "grad_norm": 0.3555627465248108, "learning_rate": 9.967701023178717e-05, "loss": 0.02, "step": 8440 }, { "grad_norm": 0.28944170475006104, "learning_rate": 9.967513114694282e-05, "loss": 0.0201, "step": 8450 }, { "grad_norm": 0.27466654777526855, "learning_rate": 9.967324662970079e-05, "loss": 0.0176, "step": 8460 }, { "grad_norm": 0.27529585361480713, "learning_rate": 9.96713566802672e-05, "loss": 0.0174, "step": 8470 }, { "grad_norm": 0.33888354897499084, "learning_rate": 9.966946129884873e-05, "loss": 0.0177, "step": 8480 }, { "grad_norm": 0.274924635887146, "learning_rate": 9.966756048565265e-05, "loss": 0.0169, "step": 8490 }, { "grad_norm": 0.22880931198596954, "learning_rate": 9.966565424088681e-05, "loss": 0.0193, "step": 8500 }, { "grad_norm": 0.2532312870025635, "learning_rate": 9.96637425647597e-05, "loss": 0.0179, "step": 8510 }, { "grad_norm": 0.36537966132164, "learning_rate": 9.966182545748038e-05, "loss": 0.0175, "step": 8520 }, { "grad_norm": 0.2979459762573242, "learning_rate": 9.96599029192585e-05, "loss": 0.0179, "step": 8530 }, { "grad_norm": 0.322125107049942, "learning_rate": 9.965797495030428e-05, "loss": 0.0167, "step": 8540 }, { "grad_norm": 0.3158606290817261, "learning_rate": 9.96560415508286e-05, "loss": 0.017, "step": 8550 }, { "grad_norm": 0.31564781069755554, "learning_rate": 9.965410272104286e-05, "loss": 0.0168, "step": 8560 }, { "grad_norm": 0.29807499051094055, "learning_rate": 9.96521584611591e-05, "loss": 0.0177, "step": 8570 }, { "grad_norm": 0.2545613646507263, "learning_rate": 9.965020877138994e-05, "loss": 0.0172, "step": 8580 }, { "grad_norm": 0.35109516978263855, "learning_rate": 9.964825365194861e-05, "loss": 0.0192, "step": 8590 }, { "grad_norm": 0.2957920730113983, "learning_rate": 9.96462931030489e-05, "loss": 0.0196, "step": 8600 }, { "grad_norm": 0.23707249760627747, "learning_rate": 9.96443271249052e-05, "loss": 0.0212, "step": 8610 }, { "grad_norm": 0.35603106021881104, "learning_rate": 9.964235571773255e-05, "loss": 0.019, "step": 8620 }, { "grad_norm": 0.3715573251247406, "learning_rate": 9.96403788817465e-05, "loss": 0.0188, "step": 8630 }, { "grad_norm": 0.2976357042789459, "learning_rate": 9.963839661716325e-05, "loss": 0.0183, "step": 8640 }, { "grad_norm": 0.3034140467643738, "learning_rate": 9.963640892419958e-05, "loss": 0.0186, "step": 8650 }, { "grad_norm": 0.3584252595901489, "learning_rate": 9.963441580307286e-05, "loss": 0.0192, "step": 8660 }, { "grad_norm": 0.32320553064346313, "learning_rate": 9.963241725400104e-05, "loss": 0.0178, "step": 8670 }, { "grad_norm": 0.5107633471488953, "learning_rate": 9.963041327720271e-05, "loss": 0.0181, "step": 8680 }, { "grad_norm": 0.3026471436023712, "learning_rate": 9.962840387289697e-05, "loss": 0.0187, "step": 8690 }, { "grad_norm": 0.20108556747436523, "learning_rate": 9.962638904130363e-05, "loss": 0.0207, "step": 8700 }, { "grad_norm": 0.3223501145839691, "learning_rate": 9.962436878264298e-05, "loss": 0.0172, "step": 8710 }, { "grad_norm": 0.3039383888244629, "learning_rate": 9.962234309713598e-05, "loss": 0.0165, "step": 8720 }, { "grad_norm": 0.29420343041419983, "learning_rate": 9.962031198500414e-05, "loss": 0.0176, "step": 8730 }, { "grad_norm": 0.22005166113376617, "learning_rate": 9.961827544646958e-05, "loss": 0.0163, "step": 8740 }, { "grad_norm": 0.18435801565647125, "learning_rate": 9.961623348175501e-05, "loss": 0.0173, "step": 8750 }, { "grad_norm": 0.24678201973438263, "learning_rate": 9.961418609108377e-05, "loss": 0.0174, "step": 8760 }, { "grad_norm": 0.3067871630191803, "learning_rate": 9.961213327467971e-05, "loss": 0.0182, "step": 8770 }, { "grad_norm": 0.26831233501434326, "learning_rate": 9.961007503276736e-05, "loss": 0.0172, "step": 8780 }, { "grad_norm": 0.24441158771514893, "learning_rate": 9.960801136557179e-05, "loss": 0.0177, "step": 8790 }, { "grad_norm": 0.35039690136909485, "learning_rate": 9.960594227331866e-05, "loss": 0.0172, "step": 8800 }, { "grad_norm": 0.40154021978378296, "learning_rate": 9.960386775623429e-05, "loss": 0.0151, "step": 8810 }, { "grad_norm": 0.3658040463924408, "learning_rate": 9.96017878145455e-05, "loss": 0.0205, "step": 8820 }, { "grad_norm": 0.30907127261161804, "learning_rate": 9.959970244847977e-05, "loss": 0.021, "step": 8830 }, { "grad_norm": 0.3922678530216217, "learning_rate": 9.959761165826518e-05, "loss": 0.019, "step": 8840 }, { "grad_norm": 0.29884037375450134, "learning_rate": 9.959551544413033e-05, "loss": 0.0211, "step": 8850 }, { "grad_norm": 0.2681359648704529, "learning_rate": 9.959341380630448e-05, "loss": 0.0199, "step": 8860 }, { "grad_norm": 0.24521541595458984, "learning_rate": 9.959130674501746e-05, "loss": 0.0192, "step": 8870 }, { "grad_norm": 0.25174644589424133, "learning_rate": 9.958919426049968e-05, "loss": 0.0181, "step": 8880 }, { "grad_norm": 0.44031113386154175, "learning_rate": 9.958707635298219e-05, "loss": 0.0175, "step": 8890 }, { "grad_norm": 0.2148972600698471, "learning_rate": 9.958495302269657e-05, "loss": 0.0177, "step": 8900 }, { "grad_norm": 0.29566437005996704, "learning_rate": 9.958282426987503e-05, "loss": 0.0194, "step": 8910 }, { "grad_norm": 0.3301193118095398, "learning_rate": 9.95806900947504e-05, "loss": 0.018, "step": 8920 }, { "grad_norm": 0.23485635221004486, "learning_rate": 9.957855049755604e-05, "loss": 0.0183, "step": 8930 }, { "grad_norm": 0.23761619627475739, "learning_rate": 9.957640547852593e-05, "loss": 0.0176, "step": 8940 }, { "grad_norm": 0.24499033391475677, "learning_rate": 9.957425503789466e-05, "loss": 0.0159, "step": 8950 }, { "grad_norm": 0.30238571763038635, "learning_rate": 9.957209917589738e-05, "loss": 0.0173, "step": 8960 }, { "grad_norm": 0.3057126998901367, "learning_rate": 9.956993789276987e-05, "loss": 0.0169, "step": 8970 }, { "grad_norm": 0.2512117326259613, "learning_rate": 9.956777118874847e-05, "loss": 0.0155, "step": 8980 }, { "grad_norm": 0.2788311243057251, "learning_rate": 9.956559906407016e-05, "loss": 0.0159, "step": 8990 }, { "grad_norm": 0.2576052248477936, "learning_rate": 9.956342151897245e-05, "loss": 0.0164, "step": 9000 }, { "grad_norm": 0.23956488072872162, "learning_rate": 9.956123855369346e-05, "loss": 0.0189, "step": 9010 }, { "grad_norm": 0.35117825865745544, "learning_rate": 9.955905016847196e-05, "loss": 0.019, "step": 9020 }, { "grad_norm": 0.3078348934650421, "learning_rate": 9.955685636354723e-05, "loss": 0.0193, "step": 9030 }, { "grad_norm": 0.32739898562431335, "learning_rate": 9.95546571391592e-05, "loss": 0.0185, "step": 9040 }, { "grad_norm": 0.35153278708457947, "learning_rate": 9.955245249554837e-05, "loss": 0.0179, "step": 9050 }, { "grad_norm": 0.3841295838356018, "learning_rate": 9.955024243295582e-05, "loss": 0.0187, "step": 9060 }, { "grad_norm": 0.22844354808330536, "learning_rate": 9.954802695162328e-05, "loss": 0.019, "step": 9070 }, { "grad_norm": 0.3079792261123657, "learning_rate": 9.954580605179302e-05, "loss": 0.0164, "step": 9080 }, { "grad_norm": 0.31906673312187195, "learning_rate": 9.954357973370788e-05, "loss": 0.0214, "step": 9090 }, { "grad_norm": 0.26310449838638306, "learning_rate": 9.954134799761135e-05, "loss": 0.0169, "step": 9100 }, { "grad_norm": 0.376852810382843, "learning_rate": 9.953911084374748e-05, "loss": 0.0184, "step": 9110 }, { "grad_norm": 0.2945413589477539, "learning_rate": 9.953686827236093e-05, "loss": 0.0177, "step": 9120 }, { "grad_norm": 0.3071431815624237, "learning_rate": 9.953462028369695e-05, "loss": 0.0169, "step": 9130 }, { "grad_norm": 0.24073046445846558, "learning_rate": 9.953236687800136e-05, "loss": 0.0165, "step": 9140 }, { "grad_norm": 0.21852049231529236, "learning_rate": 9.95301080555206e-05, "loss": 0.0162, "step": 9150 }, { "grad_norm": 0.2315206527709961, "learning_rate": 9.952784381650171e-05, "loss": 0.0162, "step": 9160 }, { "grad_norm": 0.23340865969657898, "learning_rate": 9.952557416119226e-05, "loss": 0.0167, "step": 9170 }, { "grad_norm": 0.36923736333847046, "learning_rate": 9.95232990898405e-05, "loss": 0.0167, "step": 9180 }, { "grad_norm": 0.2886420488357544, "learning_rate": 9.95210186026952e-05, "loss": 0.0189, "step": 9190 }, { "grad_norm": 0.25378021597862244, "learning_rate": 9.951873270000576e-05, "loss": 0.0172, "step": 9200 }, { "grad_norm": 0.2012024223804474, "learning_rate": 9.951644138202216e-05, "loss": 0.0194, "step": 9210 }, { "grad_norm": 0.20530185103416443, "learning_rate": 9.951414464899498e-05, "loss": 0.0168, "step": 9220 }, { "grad_norm": 0.2471402883529663, "learning_rate": 9.951184250117538e-05, "loss": 0.0166, "step": 9230 }, { "grad_norm": 0.20736733078956604, "learning_rate": 9.950953493881513e-05, "loss": 0.0156, "step": 9240 }, { "grad_norm": 0.24648983776569366, "learning_rate": 9.950722196216658e-05, "loss": 0.0158, "step": 9250 }, { "grad_norm": 0.24400390684604645, "learning_rate": 9.950490357148265e-05, "loss": 0.0182, "step": 9260 }, { "grad_norm": 0.3627861440181732, "learning_rate": 9.950257976701692e-05, "loss": 0.0182, "step": 9270 }, { "grad_norm": 0.3737737536430359, "learning_rate": 9.950025054902348e-05, "loss": 0.0166, "step": 9280 }, { "grad_norm": 0.2540437579154968, "learning_rate": 9.949791591775706e-05, "loss": 0.0176, "step": 9290 }, { "grad_norm": 0.20788393914699554, "learning_rate": 9.949557587347298e-05, "loss": 0.0173, "step": 9300 }, { "grad_norm": 0.2577945590019226, "learning_rate": 9.949323041642713e-05, "loss": 0.0168, "step": 9310 }, { "grad_norm": 0.21470674872398376, "learning_rate": 9.949087954687602e-05, "loss": 0.0161, "step": 9320 }, { "grad_norm": 0.3139008581638336, "learning_rate": 9.948852326507672e-05, "loss": 0.015, "step": 9330 }, { "grad_norm": 0.3291078209877014, "learning_rate": 9.948616157128694e-05, "loss": 0.0152, "step": 9340 }, { "grad_norm": 0.3354476988315582, "learning_rate": 9.948379446576493e-05, "loss": 0.0178, "step": 9350 }, { "grad_norm": 0.27564436197280884, "learning_rate": 9.948142194876952e-05, "loss": 0.0158, "step": 9360 }, { "grad_norm": 0.2868430018424988, "learning_rate": 9.947904402056024e-05, "loss": 0.0169, "step": 9370 }, { "grad_norm": 0.20462003350257874, "learning_rate": 9.947666068139708e-05, "loss": 0.0155, "step": 9380 }, { "grad_norm": 0.24453580379486084, "learning_rate": 9.947427193154071e-05, "loss": 0.0158, "step": 9390 }, { "grad_norm": 0.2421475201845169, "learning_rate": 9.947187777125233e-05, "loss": 0.016, "step": 9400 }, { "grad_norm": 0.22447450459003448, "learning_rate": 9.946947820079377e-05, "loss": 0.016, "step": 9410 }, { "grad_norm": 0.30506131052970886, "learning_rate": 9.946707322042747e-05, "loss": 0.0162, "step": 9420 }, { "grad_norm": 0.3491334915161133, "learning_rate": 9.94646628304164e-05, "loss": 0.0167, "step": 9430 }, { "grad_norm": 0.23410949110984802, "learning_rate": 9.946224703102418e-05, "loss": 0.0151, "step": 9440 }, { "grad_norm": 0.20782046020030975, "learning_rate": 9.945982582251498e-05, "loss": 0.0154, "step": 9450 }, { "grad_norm": 0.35906970500946045, "learning_rate": 9.94573992051536e-05, "loss": 0.0157, "step": 9460 }, { "grad_norm": 0.38662251830101013, "learning_rate": 9.94549671792054e-05, "loss": 0.0158, "step": 9470 }, { "grad_norm": 0.2211717963218689, "learning_rate": 9.945252974493635e-05, "loss": 0.0158, "step": 9480 }, { "grad_norm": 0.2667981684207916, "learning_rate": 9.9450086902613e-05, "loss": 0.0167, "step": 9490 }, { "grad_norm": 0.22725015878677368, "learning_rate": 9.944763865250248e-05, "loss": 0.0162, "step": 9500 }, { "grad_norm": 0.22456516325473785, "learning_rate": 9.944518499487254e-05, "loss": 0.0157, "step": 9510 }, { "grad_norm": 0.2490440458059311, "learning_rate": 9.944272592999151e-05, "loss": 0.0176, "step": 9520 }, { "grad_norm": 0.2996194660663605, "learning_rate": 9.94402614581283e-05, "loss": 0.0173, "step": 9530 }, { "grad_norm": 0.24834595620632172, "learning_rate": 9.943779157955244e-05, "loss": 0.0167, "step": 9540 }, { "grad_norm": 0.3393700122833252, "learning_rate": 9.943531629453403e-05, "loss": 0.017, "step": 9550 }, { "grad_norm": 0.206892728805542, "learning_rate": 9.943283560334375e-05, "loss": 0.018, "step": 9560 }, { "grad_norm": 0.2678886353969574, "learning_rate": 9.943034950625288e-05, "loss": 0.0172, "step": 9570 }, { "grad_norm": 0.27539652585983276, "learning_rate": 9.942785800353332e-05, "loss": 0.0159, "step": 9580 }, { "grad_norm": 0.30443432927131653, "learning_rate": 9.942536109545751e-05, "loss": 0.0143, "step": 9590 }, { "grad_norm": 0.26075059175491333, "learning_rate": 9.942285878229853e-05, "loss": 0.0179, "step": 9600 }, { "grad_norm": 0.530485212802887, "learning_rate": 9.942035106433001e-05, "loss": 0.0173, "step": 9610 }, { "grad_norm": 0.2801876664161682, "learning_rate": 9.94178379418262e-05, "loss": 0.0174, "step": 9620 }, { "grad_norm": 0.3105754554271698, "learning_rate": 9.941531941506194e-05, "loss": 0.0189, "step": 9630 }, { "grad_norm": 0.2784695625305176, "learning_rate": 9.941279548431263e-05, "loss": 0.0176, "step": 9640 }, { "grad_norm": 0.24263812601566315, "learning_rate": 9.941026614985431e-05, "loss": 0.0165, "step": 9650 }, { "grad_norm": 0.2616410553455353, "learning_rate": 9.940773141196357e-05, "loss": 0.0172, "step": 9660 }, { "grad_norm": 0.24367229640483856, "learning_rate": 9.94051912709176e-05, "loss": 0.0164, "step": 9670 }, { "grad_norm": 0.2599971890449524, "learning_rate": 9.940264572699421e-05, "loss": 0.0176, "step": 9680 }, { "grad_norm": 0.24200700223445892, "learning_rate": 9.940009478047174e-05, "loss": 0.0197, "step": 9690 }, { "grad_norm": 0.24534928798675537, "learning_rate": 9.939753843162918e-05, "loss": 0.0184, "step": 9700 }, { "grad_norm": 0.2852005362510681, "learning_rate": 9.939497668074609e-05, "loss": 0.0177, "step": 9710 }, { "grad_norm": 0.3317963480949402, "learning_rate": 9.93924095281026e-05, "loss": 0.0168, "step": 9720 }, { "grad_norm": 0.3046674430370331, "learning_rate": 9.938983697397948e-05, "loss": 0.0152, "step": 9730 }, { "grad_norm": 0.24024662375450134, "learning_rate": 9.938725901865805e-05, "loss": 0.0154, "step": 9740 }, { "grad_norm": 0.2012353539466858, "learning_rate": 9.93846756624202e-05, "loss": 0.0144, "step": 9750 }, { "grad_norm": 0.257767915725708, "learning_rate": 9.938208690554849e-05, "loss": 0.017, "step": 9760 }, { "grad_norm": 0.21000288426876068, "learning_rate": 9.9379492748326e-05, "loss": 0.0148, "step": 9770 }, { "grad_norm": 0.3015901744365692, "learning_rate": 9.937689319103641e-05, "loss": 0.0164, "step": 9780 }, { "grad_norm": 0.3207058310508728, "learning_rate": 9.937428823396404e-05, "loss": 0.0164, "step": 9790 }, { "grad_norm": 0.3000504970550537, "learning_rate": 9.937167787739372e-05, "loss": 0.0144, "step": 9800 }, { "grad_norm": 0.2727981209754944, "learning_rate": 9.936906212161095e-05, "loss": 0.0147, "step": 9810 }, { "grad_norm": 0.2632080018520355, "learning_rate": 9.936644096690176e-05, "loss": 0.0151, "step": 9820 }, { "grad_norm": 0.23362122476100922, "learning_rate": 9.936381441355282e-05, "loss": 0.0181, "step": 9830 }, { "grad_norm": 0.24809476733207703, "learning_rate": 9.936118246185136e-05, "loss": 0.0159, "step": 9840 }, { "grad_norm": 0.23193863034248352, "learning_rate": 9.935854511208518e-05, "loss": 0.0145, "step": 9850 }, { "grad_norm": 0.21321390569210052, "learning_rate": 9.935590236454272e-05, "loss": 0.0145, "step": 9860 }, { "grad_norm": 0.2556403577327728, "learning_rate": 9.935325421951298e-05, "loss": 0.0156, "step": 9870 }, { "grad_norm": 0.2816809415817261, "learning_rate": 9.935060067728557e-05, "loss": 0.0172, "step": 9880 }, { "grad_norm": 0.2333536148071289, "learning_rate": 9.934794173815067e-05, "loss": 0.0144, "step": 9890 }, { "grad_norm": 0.2337859869003296, "learning_rate": 9.934527740239906e-05, "loss": 0.017, "step": 9900 }, { "grad_norm": 0.27501100301742554, "learning_rate": 9.934260767032209e-05, "loss": 0.0157, "step": 9910 }, { "grad_norm": 0.3522106409072876, "learning_rate": 9.933993254221172e-05, "loss": 0.0162, "step": 9920 }, { "grad_norm": 0.24080047011375427, "learning_rate": 9.933725201836053e-05, "loss": 0.0155, "step": 9930 }, { "grad_norm": 0.2780868411064148, "learning_rate": 9.933456609906162e-05, "loss": 0.0159, "step": 9940 }, { "grad_norm": 0.2815173864364624, "learning_rate": 9.933187478460875e-05, "loss": 0.0151, "step": 9950 }, { "grad_norm": 0.2979986369609833, "learning_rate": 9.93291780752962e-05, "loss": 0.017, "step": 9960 }, { "grad_norm": 0.2597820460796356, "learning_rate": 9.932647597141893e-05, "loss": 0.0162, "step": 9970 }, { "grad_norm": 0.29755333065986633, "learning_rate": 9.932376847327239e-05, "loss": 0.0177, "step": 9980 }, { "grad_norm": 0.3163885772228241, "learning_rate": 9.932105558115268e-05, "loss": 0.015, "step": 9990 }, { "grad_norm": 0.258001446723938, "learning_rate": 9.931833729535651e-05, "loss": 0.0146, "step": 10000 }, { "grad_norm": 0.32228603959083557, "learning_rate": 9.931561361618111e-05, "loss": 0.0148, "step": 10010 }, { "grad_norm": 0.36085060238838196, "learning_rate": 9.931288454392435e-05, "loss": 0.0176, "step": 10020 }, { "grad_norm": 0.20614176988601685, "learning_rate": 9.931015007888467e-05, "loss": 0.0146, "step": 10030 }, { "grad_norm": 0.22143316268920898, "learning_rate": 9.930741022136112e-05, "loss": 0.0155, "step": 10040 }, { "grad_norm": 0.19099435210227966, "learning_rate": 9.930466497165333e-05, "loss": 0.0148, "step": 10050 }, { "grad_norm": 0.32930469512939453, "learning_rate": 9.93019143300615e-05, "loss": 0.0152, "step": 10060 }, { "grad_norm": 0.2640717029571533, "learning_rate": 9.929915829688644e-05, "loss": 0.0154, "step": 10070 }, { "grad_norm": 0.28400880098342896, "learning_rate": 9.929639687242955e-05, "loss": 0.0169, "step": 10080 }, { "grad_norm": 0.2741788625717163, "learning_rate": 9.929363005699281e-05, "loss": 0.0169, "step": 10090 }, { "grad_norm": 0.2454707771539688, "learning_rate": 9.92908578508788e-05, "loss": 0.0165, "step": 10100 }, { "grad_norm": 0.28543856739997864, "learning_rate": 9.928808025439069e-05, "loss": 0.0182, "step": 10110 }, { "grad_norm": 0.2909379303455353, "learning_rate": 9.928529726783223e-05, "loss": 0.0159, "step": 10120 }, { "grad_norm": 0.22887840867042542, "learning_rate": 9.928250889150774e-05, "loss": 0.0161, "step": 10130 }, { "grad_norm": 0.21903342008590698, "learning_rate": 9.92797151257222e-05, "loss": 0.0142, "step": 10140 }, { "grad_norm": 0.525456428527832, "learning_rate": 9.927691597078108e-05, "loss": 0.0154, "step": 10150 }, { "grad_norm": 0.20753927528858185, "learning_rate": 9.927411142699053e-05, "loss": 0.015, "step": 10160 }, { "grad_norm": 0.1965390294790268, "learning_rate": 9.927130149465725e-05, "loss": 0.016, "step": 10170 }, { "grad_norm": 0.28857019543647766, "learning_rate": 9.92684861740885e-05, "loss": 0.0165, "step": 10180 }, { "grad_norm": 0.2858971357345581, "learning_rate": 9.926566546559217e-05, "loss": 0.0186, "step": 10190 }, { "grad_norm": 0.2336043119430542, "learning_rate": 9.926283936947673e-05, "loss": 0.0155, "step": 10200 }, { "grad_norm": 0.3077205419540405, "learning_rate": 9.926000788605126e-05, "loss": 0.0157, "step": 10210 }, { "grad_norm": 0.27110835909843445, "learning_rate": 9.92571710156254e-05, "loss": 0.0158, "step": 10220 }, { "grad_norm": 0.25070109963417053, "learning_rate": 9.925432875850936e-05, "loss": 0.0155, "step": 10230 }, { "grad_norm": 0.268465131521225, "learning_rate": 9.925148111501396e-05, "loss": 0.0168, "step": 10240 }, { "grad_norm": 0.2542147636413574, "learning_rate": 9.924862808545066e-05, "loss": 0.0141, "step": 10250 }, { "grad_norm": 0.2879466414451599, "learning_rate": 9.924576967013141e-05, "loss": 0.0143, "step": 10260 }, { "grad_norm": 0.21206286549568176, "learning_rate": 9.924290586936887e-05, "loss": 0.0163, "step": 10270 }, { "grad_norm": 0.19928139448165894, "learning_rate": 9.924003668347614e-05, "loss": 0.014, "step": 10280 }, { "grad_norm": 0.25687408447265625, "learning_rate": 9.923716211276704e-05, "loss": 0.0155, "step": 10290 }, { "grad_norm": 0.2527884840965271, "learning_rate": 9.923428215755594e-05, "loss": 0.0161, "step": 10300 }, { "grad_norm": 0.2520071864128113, "learning_rate": 9.923139681815775e-05, "loss": 0.0157, "step": 10310 }, { "grad_norm": 0.285568505525589, "learning_rate": 9.922850609488801e-05, "loss": 0.0166, "step": 10320 }, { "grad_norm": 0.2695990204811096, "learning_rate": 9.922560998806287e-05, "loss": 0.0143, "step": 10330 }, { "grad_norm": 0.25177469849586487, "learning_rate": 9.922270849799905e-05, "loss": 0.0151, "step": 10340 }, { "grad_norm": 0.20304124057292938, "learning_rate": 9.92198016250138e-05, "loss": 0.0157, "step": 10350 }, { "grad_norm": 0.29364073276519775, "learning_rate": 9.921688936942506e-05, "loss": 0.0171, "step": 10360 }, { "grad_norm": 0.28188857436180115, "learning_rate": 9.921397173155129e-05, "loss": 0.0152, "step": 10370 }, { "grad_norm": 0.25441840291023254, "learning_rate": 9.921104871171157e-05, "loss": 0.0139, "step": 10380 }, { "grad_norm": 0.1947430670261383, "learning_rate": 9.920812031022554e-05, "loss": 0.0153, "step": 10390 }, { "grad_norm": 0.209956094622612, "learning_rate": 9.920518652741348e-05, "loss": 0.0139, "step": 10400 }, { "grad_norm": 0.2237226963043213, "learning_rate": 9.920224736359618e-05, "loss": 0.0152, "step": 10410 }, { "grad_norm": 0.24993698298931122, "learning_rate": 9.91993028190951e-05, "loss": 0.0149, "step": 10420 }, { "grad_norm": 0.27009063959121704, "learning_rate": 9.919635289423222e-05, "loss": 0.0147, "step": 10430 }, { "grad_norm": 0.2329999953508377, "learning_rate": 9.919339758933015e-05, "loss": 0.0139, "step": 10440 }, { "grad_norm": 0.24801871180534363, "learning_rate": 9.919043690471209e-05, "loss": 0.0146, "step": 10450 }, { "grad_norm": 0.2771431803703308, "learning_rate": 9.91874708407018e-05, "loss": 0.0157, "step": 10460 }, { "grad_norm": 0.31572237610816956, "learning_rate": 9.918449939762367e-05, "loss": 0.0148, "step": 10470 }, { "grad_norm": 0.20726406574249268, "learning_rate": 9.91815225758026e-05, "loss": 0.0151, "step": 10480 }, { "grad_norm": 0.2919332683086395, "learning_rate": 9.917854037556419e-05, "loss": 0.0151, "step": 10490 }, { "grad_norm": 0.23278391361236572, "learning_rate": 9.917555279723454e-05, "loss": 0.0148, "step": 10500 }, { "grad_norm": 0.19689282774925232, "learning_rate": 9.917255984114036e-05, "loss": 0.0156, "step": 10510 }, { "grad_norm": 0.2739807069301605, "learning_rate": 9.916956150760896e-05, "loss": 0.015, "step": 10520 }, { "grad_norm": 0.2788967192173004, "learning_rate": 9.916655779696826e-05, "loss": 0.0138, "step": 10530 }, { "grad_norm": 0.3157488703727722, "learning_rate": 9.916354870954671e-05, "loss": 0.0145, "step": 10540 }, { "grad_norm": 0.3259623646736145, "learning_rate": 9.91605342456734e-05, "loss": 0.016, "step": 10550 }, { "grad_norm": 0.2635427713394165, "learning_rate": 9.915751440567795e-05, "loss": 0.0157, "step": 10560 }, { "grad_norm": 0.31188496947288513, "learning_rate": 9.915448918989066e-05, "loss": 0.0177, "step": 10570 }, { "grad_norm": 0.3222777247428894, "learning_rate": 9.915145859864232e-05, "loss": 0.0168, "step": 10580 }, { "grad_norm": 0.22599849104881287, "learning_rate": 9.914842263226437e-05, "loss": 0.0154, "step": 10590 }, { "grad_norm": 0.26232391595840454, "learning_rate": 9.914538129108882e-05, "loss": 0.0146, "step": 10600 }, { "grad_norm": 0.21785789728164673, "learning_rate": 9.914233457544825e-05, "loss": 0.0151, "step": 10610 }, { "grad_norm": 0.2521976828575134, "learning_rate": 9.913928248567586e-05, "loss": 0.0158, "step": 10620 }, { "grad_norm": 0.3067465126514435, "learning_rate": 9.913622502210542e-05, "loss": 0.0142, "step": 10630 }, { "grad_norm": 0.17162366211414337, "learning_rate": 9.913316218507128e-05, "loss": 0.0142, "step": 10640 }, { "grad_norm": 0.25397101044654846, "learning_rate": 9.91300939749084e-05, "loss": 0.0136, "step": 10650 }, { "grad_norm": 0.3129657804965973, "learning_rate": 9.91270203919523e-05, "loss": 0.0157, "step": 10660 }, { "grad_norm": 0.29491421580314636, "learning_rate": 9.912394143653912e-05, "loss": 0.0152, "step": 10670 }, { "grad_norm": 0.20970402657985687, "learning_rate": 9.912085710900555e-05, "loss": 0.0157, "step": 10680 }, { "grad_norm": 0.27624744176864624, "learning_rate": 9.911776740968892e-05, "loss": 0.0142, "step": 10690 }, { "grad_norm": 0.3094635009765625, "learning_rate": 9.911467233892709e-05, "loss": 0.0144, "step": 10700 }, { "grad_norm": 0.210626021027565, "learning_rate": 9.911157189705853e-05, "loss": 0.0147, "step": 10710 }, { "grad_norm": 0.2794250249862671, "learning_rate": 9.910846608442229e-05, "loss": 0.0138, "step": 10720 }, { "grad_norm": 0.2972165048122406, "learning_rate": 9.910535490135805e-05, "loss": 0.0148, "step": 10730 }, { "grad_norm": 0.21979990601539612, "learning_rate": 9.910223834820603e-05, "loss": 0.0172, "step": 10740 }, { "grad_norm": 0.2358979880809784, "learning_rate": 9.909911642530703e-05, "loss": 0.016, "step": 10750 }, { "grad_norm": 0.2711038291454315, "learning_rate": 9.909598913300249e-05, "loss": 0.0171, "step": 10760 }, { "grad_norm": 0.19755075871944427, "learning_rate": 9.909285647163438e-05, "loss": 0.0146, "step": 10770 }, { "grad_norm": 0.23286263644695282, "learning_rate": 9.908971844154531e-05, "loss": 0.015, "step": 10780 }, { "grad_norm": 0.22706519067287445, "learning_rate": 9.908657504307843e-05, "loss": 0.0142, "step": 10790 }, { "grad_norm": 0.2793072462081909, "learning_rate": 9.908342627657751e-05, "loss": 0.0143, "step": 10800 }, { "grad_norm": 0.24047715961933136, "learning_rate": 9.908027214238689e-05, "loss": 0.0152, "step": 10810 }, { "grad_norm": 0.2542651295661926, "learning_rate": 9.90771126408515e-05, "loss": 0.0138, "step": 10820 }, { "grad_norm": 0.18645359575748444, "learning_rate": 9.907394777231685e-05, "loss": 0.0129, "step": 10830 }, { "grad_norm": 0.2633582055568695, "learning_rate": 9.907077753712905e-05, "loss": 0.0142, "step": 10840 }, { "grad_norm": 0.34340110421180725, "learning_rate": 9.906760193563482e-05, "loss": 0.0147, "step": 10850 }, { "grad_norm": 0.20651701092720032, "learning_rate": 9.906442096818139e-05, "loss": 0.0146, "step": 10860 }, { "grad_norm": 0.2666710913181305, "learning_rate": 9.906123463511665e-05, "loss": 0.0138, "step": 10870 }, { "grad_norm": 0.2815028131008148, "learning_rate": 9.905804293678907e-05, "loss": 0.0171, "step": 10880 }, { "grad_norm": 0.31709644198417664, "learning_rate": 9.905484587354766e-05, "loss": 0.0145, "step": 10890 }, { "grad_norm": 0.22829753160476685, "learning_rate": 9.905164344574205e-05, "loss": 0.013, "step": 10900 }, { "grad_norm": 0.25730475783348083, "learning_rate": 9.904843565372248e-05, "loss": 0.0137, "step": 10910 }, { "grad_norm": 0.22910776734352112, "learning_rate": 9.904522249783972e-05, "loss": 0.0133, "step": 10920 }, { "grad_norm": 0.20808281004428864, "learning_rate": 9.904200397844517e-05, "loss": 0.0148, "step": 10930 }, { "grad_norm": 0.20149046182632446, "learning_rate": 9.903878009589078e-05, "loss": 0.0144, "step": 10940 }, { "grad_norm": 0.21762453019618988, "learning_rate": 9.903555085052915e-05, "loss": 0.0138, "step": 10950 }, { "grad_norm": 0.27042174339294434, "learning_rate": 9.903231624271338e-05, "loss": 0.0144, "step": 10960 }, { "grad_norm": 0.19069916009902954, "learning_rate": 9.902907627279724e-05, "loss": 0.0152, "step": 10970 }, { "grad_norm": 0.21448993682861328, "learning_rate": 9.902583094113504e-05, "loss": 0.0127, "step": 10980 }, { "grad_norm": 0.1439875215291977, "learning_rate": 9.902258024808168e-05, "loss": 0.0142, "step": 10990 }, { "grad_norm": 0.3219401240348816, "learning_rate": 9.901932419399264e-05, "loss": 0.0147, "step": 11000 }, { "grad_norm": 0.16731825470924377, "learning_rate": 9.9016062779224e-05, "loss": 0.0146, "step": 11010 }, { "grad_norm": 0.17249910533428192, "learning_rate": 9.901279600413242e-05, "loss": 0.014, "step": 11020 }, { "grad_norm": 0.5008352398872375, "learning_rate": 9.900952386907518e-05, "loss": 0.0128, "step": 11030 }, { "grad_norm": 0.22247684001922607, "learning_rate": 9.90062463744101e-05, "loss": 0.0163, "step": 11040 }, { "grad_norm": 0.2287120819091797, "learning_rate": 9.900296352049558e-05, "loss": 0.0149, "step": 11050 }, { "grad_norm": 0.20348404347896576, "learning_rate": 9.899967530769065e-05, "loss": 0.0159, "step": 11060 }, { "grad_norm": 0.22785106301307678, "learning_rate": 9.899638173635489e-05, "loss": 0.0162, "step": 11070 }, { "grad_norm": 0.2223448008298874, "learning_rate": 9.899308280684849e-05, "loss": 0.0169, "step": 11080 }, { "grad_norm": 0.25927576422691345, "learning_rate": 9.898977851953222e-05, "loss": 0.0164, "step": 11090 }, { "grad_norm": 0.32149845361709595, "learning_rate": 9.898646887476741e-05, "loss": 0.0157, "step": 11100 }, { "grad_norm": 0.2234102487564087, "learning_rate": 9.898315387291603e-05, "loss": 0.0164, "step": 11110 }, { "grad_norm": 0.20914439857006073, "learning_rate": 9.89798335143406e-05, "loss": 0.015, "step": 11120 }, { "grad_norm": 0.23667700588703156, "learning_rate": 9.897650779940419e-05, "loss": 0.0143, "step": 11130 }, { "grad_norm": 0.2349025309085846, "learning_rate": 9.897317672847054e-05, "loss": 0.0158, "step": 11140 }, { "grad_norm": 0.28725162148475647, "learning_rate": 9.89698403019039e-05, "loss": 0.016, "step": 11150 }, { "grad_norm": 0.2271767556667328, "learning_rate": 9.896649852006917e-05, "loss": 0.0152, "step": 11160 }, { "grad_norm": 0.23056001961231232, "learning_rate": 9.896315138333177e-05, "loss": 0.0161, "step": 11170 }, { "grad_norm": 0.26108092069625854, "learning_rate": 9.895979889205774e-05, "loss": 0.0157, "step": 11180 }, { "grad_norm": 0.21951349079608917, "learning_rate": 9.895644104661372e-05, "loss": 0.0186, "step": 11190 }, { "grad_norm": 0.4313497543334961, "learning_rate": 9.895307784736691e-05, "loss": 0.0205, "step": 11200 }, { "grad_norm": 0.2945649325847626, "learning_rate": 9.894970929468512e-05, "loss": 0.0196, "step": 11210 }, { "grad_norm": 0.2854219377040863, "learning_rate": 9.89463353889367e-05, "loss": 0.0169, "step": 11220 }, { "grad_norm": 0.2871972322463989, "learning_rate": 9.894295613049065e-05, "loss": 0.0179, "step": 11230 }, { "grad_norm": 0.2533280551433563, "learning_rate": 9.893957151971649e-05, "loss": 0.0146, "step": 11240 }, { "grad_norm": 0.22026951611042023, "learning_rate": 9.893618155698436e-05, "loss": 0.0142, "step": 11250 }, { "grad_norm": 0.283952921628952, "learning_rate": 9.8932786242665e-05, "loss": 0.0162, "step": 11260 }, { "grad_norm": 0.2551984190940857, "learning_rate": 9.89293855771297e-05, "loss": 0.0152, "step": 11270 }, { "grad_norm": 0.24354222416877747, "learning_rate": 9.892597956075036e-05, "loss": 0.0156, "step": 11280 }, { "grad_norm": 0.23060327768325806, "learning_rate": 9.892256819389947e-05, "loss": 0.0164, "step": 11290 }, { "grad_norm": 0.17072127759456635, "learning_rate": 9.891915147695006e-05, "loss": 0.0141, "step": 11300 }, { "grad_norm": 0.23244796693325043, "learning_rate": 9.891572941027577e-05, "loss": 0.0153, "step": 11310 }, { "grad_norm": 0.25288403034210205, "learning_rate": 9.89123019942509e-05, "loss": 0.0141, "step": 11320 }, { "grad_norm": 0.16815036535263062, "learning_rate": 9.89088692292502e-05, "loss": 0.0152, "step": 11330 }, { "grad_norm": 0.23084062337875366, "learning_rate": 9.89054311156491e-05, "loss": 0.0145, "step": 11340 }, { "grad_norm": 0.22587306797504425, "learning_rate": 9.890198765382357e-05, "loss": 0.015, "step": 11350 }, { "grad_norm": 0.3014427423477173, "learning_rate": 9.889853884415021e-05, "loss": 0.0145, "step": 11360 }, { "grad_norm": 0.25574415922164917, "learning_rate": 9.889508468700614e-05, "loss": 0.0135, "step": 11370 }, { "grad_norm": 0.2641591727733612, "learning_rate": 9.889162518276915e-05, "loss": 0.0142, "step": 11380 }, { "grad_norm": 0.1573314368724823, "learning_rate": 9.888816033181752e-05, "loss": 0.0145, "step": 11390 }, { "grad_norm": 0.30820000171661377, "learning_rate": 9.888469013453018e-05, "loss": 0.0131, "step": 11400 }, { "grad_norm": 0.2356191724538803, "learning_rate": 9.888121459128663e-05, "loss": 0.0136, "step": 11410 }, { "grad_norm": 0.2512969374656677, "learning_rate": 9.887773370246693e-05, "loss": 0.0121, "step": 11420 }, { "grad_norm": 0.18511033058166504, "learning_rate": 9.887424746845177e-05, "loss": 0.0126, "step": 11430 }, { "grad_norm": 0.34741291403770447, "learning_rate": 9.887075588962239e-05, "loss": 0.0143, "step": 11440 }, { "grad_norm": 0.28018856048583984, "learning_rate": 9.88672589663606e-05, "loss": 0.0137, "step": 11450 }, { "grad_norm": 0.18155984580516815, "learning_rate": 9.886375669904886e-05, "loss": 0.0138, "step": 11460 }, { "grad_norm": 0.2120228260755539, "learning_rate": 9.886024908807014e-05, "loss": 0.0152, "step": 11470 }, { "grad_norm": 0.19268271327018738, "learning_rate": 9.885673613380806e-05, "loss": 0.0124, "step": 11480 }, { "grad_norm": 0.21648232638835907, "learning_rate": 9.885321783664676e-05, "loss": 0.0131, "step": 11490 }, { "grad_norm": 0.25704047083854675, "learning_rate": 9.884969419697101e-05, "loss": 0.0137, "step": 11500 }, { "grad_norm": 0.2597517669200897, "learning_rate": 9.884616521516614e-05, "loss": 0.0159, "step": 11510 }, { "grad_norm": 0.26661059260368347, "learning_rate": 9.88426308916181e-05, "loss": 0.0161, "step": 11520 }, { "grad_norm": 0.23200424015522003, "learning_rate": 9.883909122671335e-05, "loss": 0.0154, "step": 11530 }, { "grad_norm": 0.29601478576660156, "learning_rate": 9.883554622083904e-05, "loss": 0.0144, "step": 11540 }, { "grad_norm": 0.17355036735534668, "learning_rate": 9.88319958743828e-05, "loss": 0.0142, "step": 11550 }, { "grad_norm": 0.23594626784324646, "learning_rate": 9.882844018773291e-05, "loss": 0.0148, "step": 11560 }, { "grad_norm": 0.1879127323627472, "learning_rate": 9.882487916127823e-05, "loss": 0.0133, "step": 11570 }, { "grad_norm": 0.24873629212379456, "learning_rate": 9.882131279540815e-05, "loss": 0.0157, "step": 11580 }, { "grad_norm": 0.2911539375782013, "learning_rate": 9.881774109051271e-05, "loss": 0.014, "step": 11590 }, { "grad_norm": 0.2647499442100525, "learning_rate": 9.881416404698252e-05, "loss": 0.0146, "step": 11600 }, { "grad_norm": 0.24471800029277802, "learning_rate": 9.881058166520873e-05, "loss": 0.0136, "step": 11610 }, { "grad_norm": 0.2521738111972809, "learning_rate": 9.880699394558311e-05, "loss": 0.0144, "step": 11620 }, { "grad_norm": 0.2612912952899933, "learning_rate": 9.880340088849801e-05, "loss": 0.0143, "step": 11630 }, { "grad_norm": 0.22425077855587006, "learning_rate": 9.879980249434637e-05, "loss": 0.0139, "step": 11640 }, { "grad_norm": 0.35825684666633606, "learning_rate": 9.879619876352168e-05, "loss": 0.014, "step": 11650 }, { "grad_norm": 0.23443958163261414, "learning_rate": 9.879258969641809e-05, "loss": 0.0139, "step": 11660 }, { "grad_norm": 0.3240991234779358, "learning_rate": 9.878897529343023e-05, "loss": 0.0155, "step": 11670 }, { "grad_norm": 0.2454058825969696, "learning_rate": 9.878535555495338e-05, "loss": 0.0153, "step": 11680 }, { "grad_norm": 0.230025053024292, "learning_rate": 9.87817304813834e-05, "loss": 0.0154, "step": 11690 }, { "grad_norm": 0.26871457695961, "learning_rate": 9.877810007311671e-05, "loss": 0.0165, "step": 11700 }, { "grad_norm": 0.333758682012558, "learning_rate": 9.877446433055035e-05, "loss": 0.0173, "step": 11710 }, { "grad_norm": 0.21920257806777954, "learning_rate": 9.877082325408191e-05, "loss": 0.0168, "step": 11720 }, { "grad_norm": 0.30103984475135803, "learning_rate": 9.876717684410954e-05, "loss": 0.0149, "step": 11730 }, { "grad_norm": 0.29045918583869934, "learning_rate": 9.876352510103204e-05, "loss": 0.0161, "step": 11740 }, { "grad_norm": 0.20774438977241516, "learning_rate": 9.875986802524875e-05, "loss": 0.0146, "step": 11750 }, { "grad_norm": 0.26061001420021057, "learning_rate": 9.87562056171596e-05, "loss": 0.0159, "step": 11760 }, { "grad_norm": 0.23562626540660858, "learning_rate": 9.875253787716511e-05, "loss": 0.0146, "step": 11770 }, { "grad_norm": 0.25376975536346436, "learning_rate": 9.874886480566637e-05, "loss": 0.0143, "step": 11780 }, { "grad_norm": 0.28476616740226746, "learning_rate": 9.874518640306507e-05, "loss": 0.0147, "step": 11790 }, { "grad_norm": 0.19576746225357056, "learning_rate": 9.874150266976347e-05, "loss": 0.014, "step": 11800 }, { "grad_norm": 0.26166126132011414, "learning_rate": 9.873781360616443e-05, "loss": 0.0141, "step": 11810 }, { "grad_norm": 0.23106759786605835, "learning_rate": 9.873411921267137e-05, "loss": 0.0136, "step": 11820 }, { "grad_norm": 0.21811750531196594, "learning_rate": 9.873041948968829e-05, "loss": 0.0157, "step": 11830 }, { "grad_norm": 0.25676190853118896, "learning_rate": 9.872671443761981e-05, "loss": 0.0136, "step": 11840 }, { "grad_norm": 0.2844957411289215, "learning_rate": 9.872300405687109e-05, "loss": 0.0135, "step": 11850 }, { "grad_norm": 0.21109387278556824, "learning_rate": 9.871928834784792e-05, "loss": 0.0139, "step": 11860 }, { "grad_norm": 0.18493986129760742, "learning_rate": 9.871556731095661e-05, "loss": 0.0135, "step": 11870 }, { "grad_norm": 0.26599571108818054, "learning_rate": 9.871184094660411e-05, "loss": 0.016, "step": 11880 }, { "grad_norm": 0.2864406704902649, "learning_rate": 9.870810925519791e-05, "loss": 0.0137, "step": 11890 }, { "grad_norm": 0.2703743577003479, "learning_rate": 9.870437223714612e-05, "loss": 0.0145, "step": 11900 }, { "grad_norm": 0.22905109822750092, "learning_rate": 9.87006298928574e-05, "loss": 0.0155, "step": 11910 }, { "grad_norm": 0.2804114520549774, "learning_rate": 9.869688222274103e-05, "loss": 0.0147, "step": 11920 }, { "grad_norm": 0.16699552536010742, "learning_rate": 9.869312922720681e-05, "loss": 0.0141, "step": 11930 }, { "grad_norm": 0.2631762623786926, "learning_rate": 9.868937090666521e-05, "loss": 0.0129, "step": 11940 }, { "grad_norm": 0.29210761189460754, "learning_rate": 9.86856072615272e-05, "loss": 0.0137, "step": 11950 }, { "grad_norm": 0.32307296991348267, "learning_rate": 9.868183829220438e-05, "loss": 0.0129, "step": 11960 }, { "grad_norm": 0.3061976134777069, "learning_rate": 9.867806399910893e-05, "loss": 0.0132, "step": 11970 }, { "grad_norm": 0.31700649857521057, "learning_rate": 9.867428438265356e-05, "loss": 0.0139, "step": 11980 }, { "grad_norm": 0.22278541326522827, "learning_rate": 9.867049944325165e-05, "loss": 0.0147, "step": 11990 }, { "grad_norm": 0.168077290058136, "learning_rate": 9.86667091813171e-05, "loss": 0.0138, "step": 12000 }, { "grad_norm": 0.2116209715604782, "learning_rate": 9.866291359726438e-05, "loss": 0.013, "step": 12010 }, { "grad_norm": 0.21500882506370544, "learning_rate": 9.865911269150861e-05, "loss": 0.0116, "step": 12020 }, { "grad_norm": 0.18610474467277527, "learning_rate": 9.865530646446544e-05, "loss": 0.0124, "step": 12030 }, { "grad_norm": 0.17981843650341034, "learning_rate": 9.86514949165511e-05, "loss": 0.0122, "step": 12040 }, { "grad_norm": 0.25416257977485657, "learning_rate": 9.864767804818243e-05, "loss": 0.013, "step": 12050 }, { "grad_norm": 0.2737232446670532, "learning_rate": 9.86438558597768e-05, "loss": 0.0129, "step": 12060 }, { "grad_norm": 0.15721261501312256, "learning_rate": 9.864002835175225e-05, "loss": 0.0122, "step": 12070 }, { "grad_norm": 0.19181647896766663, "learning_rate": 9.863619552452734e-05, "loss": 0.0128, "step": 12080 }, { "grad_norm": 0.23911133408546448, "learning_rate": 9.863235737852119e-05, "loss": 0.0133, "step": 12090 }, { "grad_norm": 0.24073180556297302, "learning_rate": 9.862851391415356e-05, "loss": 0.0131, "step": 12100 }, { "grad_norm": 0.18554577231407166, "learning_rate": 9.862466513184477e-05, "loss": 0.0133, "step": 12110 }, { "grad_norm": 0.1795523464679718, "learning_rate": 9.86208110320157e-05, "loss": 0.0112, "step": 12120 }, { "grad_norm": 0.16649512946605682, "learning_rate": 9.861695161508784e-05, "loss": 0.0112, "step": 12130 }, { "grad_norm": 0.22208404541015625, "learning_rate": 9.861308688148324e-05, "loss": 0.012, "step": 12140 }, { "grad_norm": 0.20636676251888275, "learning_rate": 9.860921683162455e-05, "loss": 0.0132, "step": 12150 }, { "grad_norm": 0.22210949659347534, "learning_rate": 9.860534146593499e-05, "loss": 0.0141, "step": 12160 }, { "grad_norm": 0.2443423569202423, "learning_rate": 9.860146078483836e-05, "loss": 0.0125, "step": 12170 }, { "grad_norm": 0.21483193337917328, "learning_rate": 9.859757478875905e-05, "loss": 0.0143, "step": 12180 }, { "grad_norm": 0.21856382489204407, "learning_rate": 9.859368347812204e-05, "loss": 0.0131, "step": 12190 }, { "grad_norm": 0.26489317417144775, "learning_rate": 9.858978685335285e-05, "loss": 0.0135, "step": 12200 }, { "grad_norm": 0.23099908232688904, "learning_rate": 9.858588491487763e-05, "loss": 0.0135, "step": 12210 }, { "grad_norm": 0.19843743741512299, "learning_rate": 9.858197766312308e-05, "loss": 0.0141, "step": 12220 }, { "grad_norm": 0.2211918830871582, "learning_rate": 9.857806509851649e-05, "loss": 0.0132, "step": 12230 }, { "grad_norm": 0.18531571328639984, "learning_rate": 9.857414722148574e-05, "loss": 0.0126, "step": 12240 }, { "grad_norm": 0.3205604553222656, "learning_rate": 9.857022403245928e-05, "loss": 0.0135, "step": 12250 }, { "grad_norm": 0.19540753960609436, "learning_rate": 9.856629553186615e-05, "loss": 0.0134, "step": 12260 }, { "grad_norm": 0.23207958042621613, "learning_rate": 9.856236172013595e-05, "loss": 0.0128, "step": 12270 }, { "grad_norm": 0.17090840637683868, "learning_rate": 9.85584225976989e-05, "loss": 0.0122, "step": 12280 }, { "grad_norm": 0.26570260524749756, "learning_rate": 9.855447816498575e-05, "loss": 0.0127, "step": 12290 }, { "grad_norm": 0.2504376471042633, "learning_rate": 9.855052842242787e-05, "loss": 0.0122, "step": 12300 }, { "grad_norm": 0.2759188413619995, "learning_rate": 9.85465733704572e-05, "loss": 0.0126, "step": 12310 }, { "grad_norm": 0.20871835947036743, "learning_rate": 9.854261300950624e-05, "loss": 0.0123, "step": 12320 }, { "grad_norm": 0.21691074967384338, "learning_rate": 9.853864734000813e-05, "loss": 0.0141, "step": 12330 }, { "grad_norm": 0.20647557079792023, "learning_rate": 9.85346763623965e-05, "loss": 0.0128, "step": 12340 }, { "grad_norm": 0.23820720613002777, "learning_rate": 9.853070007710564e-05, "loss": 0.0134, "step": 12350 }, { "grad_norm": 0.17293718457221985, "learning_rate": 9.85267184845704e-05, "loss": 0.0131, "step": 12360 }, { "grad_norm": 0.18172688782215118, "learning_rate": 9.852273158522616e-05, "loss": 0.0132, "step": 12370 }, { "grad_norm": 0.2167576551437378, "learning_rate": 9.851873937950896e-05, "loss": 0.0143, "step": 12380 }, { "grad_norm": 0.2033420354127884, "learning_rate": 9.851474186785537e-05, "loss": 0.0161, "step": 12390 }, { "grad_norm": 0.24741144478321075, "learning_rate": 9.851073905070254e-05, "loss": 0.0149, "step": 12400 }, { "grad_norm": 0.272916704416275, "learning_rate": 9.850673092848824e-05, "loss": 0.0156, "step": 12410 }, { "grad_norm": 0.26633894443511963, "learning_rate": 9.850271750165077e-05, "loss": 0.014, "step": 12420 }, { "grad_norm": 0.27061256766319275, "learning_rate": 9.849869877062902e-05, "loss": 0.015, "step": 12430 }, { "grad_norm": 0.2063382863998413, "learning_rate": 9.849467473586252e-05, "loss": 0.0138, "step": 12440 }, { "grad_norm": 0.23243993520736694, "learning_rate": 9.849064539779127e-05, "loss": 0.0125, "step": 12450 }, { "grad_norm": 0.384926974773407, "learning_rate": 9.848661075685594e-05, "loss": 0.0152, "step": 12460 }, { "grad_norm": 0.3200766146183014, "learning_rate": 9.848257081349778e-05, "loss": 0.0151, "step": 12470 }, { "grad_norm": 0.19691230356693268, "learning_rate": 9.847852556815856e-05, "loss": 0.0144, "step": 12480 }, { "grad_norm": 0.2384326159954071, "learning_rate": 9.847447502128067e-05, "loss": 0.013, "step": 12490 }, { "grad_norm": 0.265292227268219, "learning_rate": 9.847041917330708e-05, "loss": 0.0154, "step": 12500 }, { "grad_norm": 0.26887932419776917, "learning_rate": 9.846635802468132e-05, "loss": 0.0139, "step": 12510 }, { "grad_norm": 0.19238696992397308, "learning_rate": 9.84622915758475e-05, "loss": 0.015, "step": 12520 }, { "grad_norm": 0.20145373046398163, "learning_rate": 9.845821982725034e-05, "loss": 0.0135, "step": 12530 }, { "grad_norm": 0.27242207527160645, "learning_rate": 9.845414277933514e-05, "loss": 0.0141, "step": 12540 }, { "grad_norm": 0.1757490634918213, "learning_rate": 9.845006043254771e-05, "loss": 0.0131, "step": 12550 }, { "grad_norm": 0.16725552082061768, "learning_rate": 9.844597278733451e-05, "loss": 0.0133, "step": 12560 }, { "grad_norm": 0.27945786714553833, "learning_rate": 9.844187984414259e-05, "loss": 0.0143, "step": 12570 }, { "grad_norm": 0.264500230550766, "learning_rate": 9.84377816034195e-05, "loss": 0.0131, "step": 12580 }, { "grad_norm": 0.19910408556461334, "learning_rate": 9.843367806561345e-05, "loss": 0.0139, "step": 12590 }, { "grad_norm": 0.26948630809783936, "learning_rate": 9.842956923117317e-05, "loss": 0.013, "step": 12600 }, { "grad_norm": 0.22799746692180634, "learning_rate": 9.842545510054802e-05, "loss": 0.0137, "step": 12610 }, { "grad_norm": 0.24071767926216125, "learning_rate": 9.842133567418792e-05, "loss": 0.012, "step": 12620 }, { "grad_norm": 0.2805996835231781, "learning_rate": 9.841721095254333e-05, "loss": 0.0133, "step": 12630 }, { "grad_norm": 0.23690693080425262, "learning_rate": 9.841308093606537e-05, "loss": 0.0135, "step": 12640 }, { "grad_norm": 0.24245215952396393, "learning_rate": 9.840894562520565e-05, "loss": 0.0127, "step": 12650 }, { "grad_norm": 0.20243440568447113, "learning_rate": 9.840480502041642e-05, "loss": 0.0128, "step": 12660 }, { "grad_norm": 0.24563559889793396, "learning_rate": 9.840065912215049e-05, "loss": 0.0129, "step": 12670 }, { "grad_norm": 0.289917916059494, "learning_rate": 9.839650793086124e-05, "loss": 0.0127, "step": 12680 }, { "grad_norm": 0.2573428750038147, "learning_rate": 9.839235144700265e-05, "loss": 0.0131, "step": 12690 }, { "grad_norm": 0.21911217272281647, "learning_rate": 9.838818967102926e-05, "loss": 0.013, "step": 12700 }, { "grad_norm": 0.28590813279151917, "learning_rate": 9.83840226033962e-05, "loss": 0.0158, "step": 12710 }, { "grad_norm": 0.2550654411315918, "learning_rate": 9.837985024455918e-05, "loss": 0.0157, "step": 12720 }, { "grad_norm": 0.18412666022777557, "learning_rate": 9.837567259497447e-05, "loss": 0.0146, "step": 12730 }, { "grad_norm": 0.21522681415081024, "learning_rate": 9.837148965509894e-05, "loss": 0.0151, "step": 12740 }, { "grad_norm": 0.2135860025882721, "learning_rate": 9.836730142539001e-05, "loss": 0.0146, "step": 12750 }, { "grad_norm": 0.15280574560165405, "learning_rate": 9.836310790630574e-05, "loss": 0.0127, "step": 12760 }, { "grad_norm": 0.23979805409908295, "learning_rate": 9.83589090983047e-05, "loss": 0.0146, "step": 12770 }, { "grad_norm": 0.2978084981441498, "learning_rate": 9.835470500184605e-05, "loss": 0.0142, "step": 12780 }, { "grad_norm": 0.2862471044063568, "learning_rate": 9.835049561738957e-05, "loss": 0.0135, "step": 12790 }, { "grad_norm": 0.2603139579296112, "learning_rate": 9.834628094539558e-05, "loss": 0.0141, "step": 12800 }, { "grad_norm": 0.2597595155239105, "learning_rate": 9.834206098632499e-05, "loss": 0.0123, "step": 12810 }, { "grad_norm": 0.19988499581813812, "learning_rate": 9.833783574063931e-05, "loss": 0.0121, "step": 12820 }, { "grad_norm": 0.3246777653694153, "learning_rate": 9.833360520880058e-05, "loss": 0.0128, "step": 12830 }, { "grad_norm": 0.2184208780527115, "learning_rate": 9.832936939127144e-05, "loss": 0.0133, "step": 12840 }, { "grad_norm": 0.2401234358549118, "learning_rate": 9.832512828851515e-05, "loss": 0.0142, "step": 12850 }, { "grad_norm": 0.2528693974018097, "learning_rate": 9.832088190099546e-05, "loss": 0.013, "step": 12860 }, { "grad_norm": 0.18755407631397247, "learning_rate": 9.831663022917679e-05, "loss": 0.0127, "step": 12870 }, { "grad_norm": 0.23070955276489258, "learning_rate": 9.831237327352407e-05, "loss": 0.013, "step": 12880 }, { "grad_norm": 0.199567973613739, "learning_rate": 9.830811103450286e-05, "loss": 0.0126, "step": 12890 }, { "grad_norm": 0.25560224056243896, "learning_rate": 9.830384351257924e-05, "loss": 0.0145, "step": 12900 }, { "grad_norm": 0.25703680515289307, "learning_rate": 9.829957070821993e-05, "loss": 0.0125, "step": 12910 }, { "grad_norm": 0.29031941294670105, "learning_rate": 9.829529262189218e-05, "loss": 0.0128, "step": 12920 }, { "grad_norm": 0.2962847352027893, "learning_rate": 9.829100925406385e-05, "loss": 0.0136, "step": 12930 }, { "grad_norm": 0.28992924094200134, "learning_rate": 9.828672060520333e-05, "loss": 0.0135, "step": 12940 }, { "grad_norm": 0.21572037041187286, "learning_rate": 9.828242667577966e-05, "loss": 0.015, "step": 12950 }, { "grad_norm": 0.3277944326400757, "learning_rate": 9.82781274662624e-05, "loss": 0.0138, "step": 12960 }, { "grad_norm": 0.15410295128822327, "learning_rate": 9.82738229771217e-05, "loss": 0.014, "step": 12970 }, { "grad_norm": 0.21713721752166748, "learning_rate": 9.826951320882829e-05, "loss": 0.0131, "step": 12980 }, { "grad_norm": 0.26298972964286804, "learning_rate": 9.826519816185351e-05, "loss": 0.0142, "step": 12990 }, { "grad_norm": 0.24391184747219086, "learning_rate": 9.826087783666921e-05, "loss": 0.014, "step": 13000 }, { "grad_norm": 0.19885005056858063, "learning_rate": 9.825655223374787e-05, "loss": 0.0134, "step": 13010 }, { "grad_norm": 0.22776226699352264, "learning_rate": 9.825222135356253e-05, "loss": 0.0131, "step": 13020 }, { "grad_norm": 0.22271022200584412, "learning_rate": 9.82478851965868e-05, "loss": 0.0135, "step": 13030 }, { "grad_norm": 0.2298460304737091, "learning_rate": 9.82435437632949e-05, "loss": 0.0118, "step": 13040 }, { "grad_norm": 0.2070944905281067, "learning_rate": 9.823919705416158e-05, "loss": 0.0107, "step": 13050 }, { "grad_norm": 0.2463218867778778, "learning_rate": 9.82348450696622e-05, "loss": 0.0132, "step": 13060 }, { "grad_norm": 0.16038070619106293, "learning_rate": 9.823048781027268e-05, "loss": 0.0114, "step": 13070 }, { "grad_norm": 0.18800543248653412, "learning_rate": 9.822612527646953e-05, "loss": 0.0121, "step": 13080 }, { "grad_norm": 0.20703552663326263, "learning_rate": 9.822175746872984e-05, "loss": 0.0112, "step": 13090 }, { "grad_norm": 0.2668860852718353, "learning_rate": 9.821738438753123e-05, "loss": 0.0125, "step": 13100 }, { "grad_norm": 0.2946722209453583, "learning_rate": 9.821300603335196e-05, "loss": 0.012, "step": 13110 }, { "grad_norm": 0.14577847719192505, "learning_rate": 9.820862240667085e-05, "loss": 0.0129, "step": 13120 }, { "grad_norm": 0.17903171479701996, "learning_rate": 9.820423350796726e-05, "loss": 0.0141, "step": 13130 }, { "grad_norm": 0.24009037017822266, "learning_rate": 9.819983933772118e-05, "loss": 0.0132, "step": 13140 }, { "grad_norm": 0.19827337563037872, "learning_rate": 9.819543989641314e-05, "loss": 0.0118, "step": 13150 }, { "grad_norm": 0.22732138633728027, "learning_rate": 9.819103518452423e-05, "loss": 0.014, "step": 13160 }, { "grad_norm": 0.2407030463218689, "learning_rate": 9.818662520253618e-05, "loss": 0.0149, "step": 13170 }, { "grad_norm": 0.21390321850776672, "learning_rate": 9.818220995093126e-05, "loss": 0.0116, "step": 13180 }, { "grad_norm": 0.1788872480392456, "learning_rate": 9.817778943019228e-05, "loss": 0.0124, "step": 13190 }, { "grad_norm": 0.23030714690685272, "learning_rate": 9.81733636408027e-05, "loss": 0.0143, "step": 13200 }, { "grad_norm": 0.21357592940330505, "learning_rate": 9.816893258324649e-05, "loss": 0.0125, "step": 13210 }, { "grad_norm": 0.23491565883159637, "learning_rate": 9.816449625800823e-05, "loss": 0.0143, "step": 13220 }, { "grad_norm": 0.2857644259929657, "learning_rate": 9.816005466557308e-05, "loss": 0.0135, "step": 13230 }, { "grad_norm": 0.27715498208999634, "learning_rate": 9.815560780642674e-05, "loss": 0.0153, "step": 13240 }, { "grad_norm": 0.25934192538261414, "learning_rate": 9.815115568105555e-05, "loss": 0.0158, "step": 13250 }, { "grad_norm": 0.2660696804523468, "learning_rate": 9.814669828994638e-05, "loss": 0.0148, "step": 13260 }, { "grad_norm": 0.21391214430332184, "learning_rate": 9.814223563358665e-05, "loss": 0.0144, "step": 13270 }, { "grad_norm": 0.19007708132266998, "learning_rate": 9.813776771246443e-05, "loss": 0.0133, "step": 13280 }, { "grad_norm": 0.22393938899040222, "learning_rate": 9.813329452706829e-05, "loss": 0.0158, "step": 13290 }, { "grad_norm": 0.2683499753475189, "learning_rate": 9.812881607788744e-05, "loss": 0.014, "step": 13300 }, { "grad_norm": 0.20959272980690002, "learning_rate": 9.812433236541163e-05, "loss": 0.0133, "step": 13310 }, { "grad_norm": 0.2870536148548126, "learning_rate": 9.811984339013116e-05, "loss": 0.0146, "step": 13320 }, { "grad_norm": 0.21350917220115662, "learning_rate": 9.811534915253698e-05, "loss": 0.0139, "step": 13330 }, { "grad_norm": 0.18041466176509857, "learning_rate": 9.811084965312056e-05, "loss": 0.013, "step": 13340 }, { "grad_norm": 0.21774601936340332, "learning_rate": 9.810634489237396e-05, "loss": 0.0132, "step": 13350 }, { "grad_norm": 0.24614877998828888, "learning_rate": 9.81018348707898e-05, "loss": 0.0138, "step": 13360 }, { "grad_norm": 0.28024691343307495, "learning_rate": 9.809731958886131e-05, "loss": 0.0152, "step": 13370 }, { "grad_norm": 0.23713688552379608, "learning_rate": 9.809279904708224e-05, "loss": 0.0141, "step": 13380 }, { "grad_norm": 0.21428658068180084, "learning_rate": 9.808827324594699e-05, "loss": 0.0135, "step": 13390 }, { "grad_norm": 0.2368648797273636, "learning_rate": 9.808374218595046e-05, "loss": 0.015, "step": 13400 }, { "grad_norm": 0.21166899800300598, "learning_rate": 9.80792058675882e-05, "loss": 0.0117, "step": 13410 }, { "grad_norm": 0.20480701327323914, "learning_rate": 9.807466429135627e-05, "loss": 0.0124, "step": 13420 }, { "grad_norm": 0.24093954265117645, "learning_rate": 9.807011745775132e-05, "loss": 0.0141, "step": 13430 }, { "grad_norm": 0.28310757875442505, "learning_rate": 9.806556536727061e-05, "loss": 0.0123, "step": 13440 }, { "grad_norm": 0.21303394436836243, "learning_rate": 9.806100802041193e-05, "loss": 0.0118, "step": 13450 }, { "grad_norm": 0.21888339519500732, "learning_rate": 9.805644541767368e-05, "loss": 0.0123, "step": 13460 }, { "grad_norm": 0.21048320829868317, "learning_rate": 9.805187755955478e-05, "loss": 0.0123, "step": 13470 }, { "grad_norm": 0.22224034368991852, "learning_rate": 9.804730444655483e-05, "loss": 0.0116, "step": 13480 }, { "grad_norm": 0.2146947979927063, "learning_rate": 9.804272607917388e-05, "loss": 0.0118, "step": 13490 }, { "grad_norm": 0.1977873593568802, "learning_rate": 9.803814245791265e-05, "loss": 0.0144, "step": 13500 }, { "grad_norm": 0.25108447670936584, "learning_rate": 9.803355358327239e-05, "loss": 0.0139, "step": 13510 }, { "grad_norm": 0.23172199726104736, "learning_rate": 9.802895945575492e-05, "loss": 0.0146, "step": 13520 }, { "grad_norm": 0.26351556181907654, "learning_rate": 9.802436007586266e-05, "loss": 0.0128, "step": 13530 }, { "grad_norm": 0.2291017472743988, "learning_rate": 9.801975544409858e-05, "loss": 0.0146, "step": 13540 }, { "grad_norm": 0.22872671484947205, "learning_rate": 9.801514556096625e-05, "loss": 0.0138, "step": 13550 }, { "grad_norm": 0.2094402015209198, "learning_rate": 9.801053042696977e-05, "loss": 0.0143, "step": 13560 }, { "grad_norm": 0.2691579759120941, "learning_rate": 9.800591004261388e-05, "loss": 0.0134, "step": 13570 }, { "grad_norm": 0.29525279998779297, "learning_rate": 9.800128440840385e-05, "loss": 0.0128, "step": 13580 }, { "grad_norm": 0.22668291628360748, "learning_rate": 9.799665352484552e-05, "loss": 0.0157, "step": 13590 }, { "grad_norm": 0.2276436984539032, "learning_rate": 9.799201739244532e-05, "loss": 0.014, "step": 13600 }, { "grad_norm": 0.25906285643577576, "learning_rate": 9.798737601171025e-05, "loss": 0.0135, "step": 13610 }, { "grad_norm": 0.21044225990772247, "learning_rate": 9.79827293831479e-05, "loss": 0.0126, "step": 13620 }, { "grad_norm": 0.22688473761081696, "learning_rate": 9.797807750726638e-05, "loss": 0.0134, "step": 13630 }, { "grad_norm": 0.27555733919143677, "learning_rate": 9.797342038457446e-05, "loss": 0.0137, "step": 13640 }, { "grad_norm": 0.22256653010845184, "learning_rate": 9.796875801558141e-05, "loss": 0.0137, "step": 13650 }, { "grad_norm": 0.30228328704833984, "learning_rate": 9.79640904007971e-05, "loss": 0.0135, "step": 13660 }, { "grad_norm": 0.262550413608551, "learning_rate": 9.795941754073199e-05, "loss": 0.0134, "step": 13670 }, { "grad_norm": 0.21632637083530426, "learning_rate": 9.795473943589705e-05, "loss": 0.0118, "step": 13680 }, { "grad_norm": 0.21213434636592865, "learning_rate": 9.795005608680394e-05, "loss": 0.0118, "step": 13690 }, { "grad_norm": 0.17519758641719818, "learning_rate": 9.794536749396477e-05, "loss": 0.0115, "step": 13700 }, { "grad_norm": 0.2408963441848755, "learning_rate": 9.79406736578923e-05, "loss": 0.0116, "step": 13710 }, { "grad_norm": 0.30178171396255493, "learning_rate": 9.793597457909984e-05, "loss": 0.0143, "step": 13720 }, { "grad_norm": 0.2596440315246582, "learning_rate": 9.793127025810127e-05, "loss": 0.0153, "step": 13730 }, { "grad_norm": 0.18327201902866364, "learning_rate": 9.792656069541104e-05, "loss": 0.0128, "step": 13740 }, { "grad_norm": 0.24333733320236206, "learning_rate": 9.79218458915442e-05, "loss": 0.0117, "step": 13750 }, { "grad_norm": 0.18439877033233643, "learning_rate": 9.791712584701634e-05, "loss": 0.0127, "step": 13760 }, { "grad_norm": 0.23056749999523163, "learning_rate": 9.791240056234364e-05, "loss": 0.0126, "step": 13770 }, { "grad_norm": 0.20939090847969055, "learning_rate": 9.790767003804283e-05, "loss": 0.0138, "step": 13780 }, { "grad_norm": 0.17865107953548431, "learning_rate": 9.790293427463126e-05, "loss": 0.0126, "step": 13790 }, { "grad_norm": 0.18988095223903656, "learning_rate": 9.789819327262684e-05, "loss": 0.013, "step": 13800 }, { "grad_norm": 0.20205850899219513, "learning_rate": 9.7893447032548e-05, "loss": 0.0147, "step": 13810 }, { "grad_norm": 0.22701039910316467, "learning_rate": 9.78886955549138e-05, "loss": 0.0146, "step": 13820 }, { "grad_norm": 0.3321824371814728, "learning_rate": 9.788393884024387e-05, "loss": 0.013, "step": 13830 }, { "grad_norm": 0.22702951729297638, "learning_rate": 9.787917688905836e-05, "loss": 0.012, "step": 13840 }, { "grad_norm": 0.1560211479663849, "learning_rate": 9.787440970187807e-05, "loss": 0.0121, "step": 13850 }, { "grad_norm": 0.284342497587204, "learning_rate": 9.786963727922429e-05, "loss": 0.0155, "step": 13860 }, { "grad_norm": 0.34029144048690796, "learning_rate": 9.786485962161897e-05, "loss": 0.0141, "step": 13870 }, { "grad_norm": 0.24445633590221405, "learning_rate": 9.786007672958455e-05, "loss": 0.0153, "step": 13880 }, { "grad_norm": 0.2506732642650604, "learning_rate": 9.78552886036441e-05, "loss": 0.0135, "step": 13890 }, { "grad_norm": 0.3579540550708771, "learning_rate": 9.785049524432124e-05, "loss": 0.0139, "step": 13900 }, { "grad_norm": 0.23187214136123657, "learning_rate": 9.784569665214016e-05, "loss": 0.0131, "step": 13910 }, { "grad_norm": 0.2514662742614746, "learning_rate": 9.784089282762563e-05, "loss": 0.0129, "step": 13920 }, { "grad_norm": 0.27887025475502014, "learning_rate": 9.7836083771303e-05, "loss": 0.0129, "step": 13930 }, { "grad_norm": 0.17780601978302002, "learning_rate": 9.783126948369817e-05, "loss": 0.0141, "step": 13940 }, { "grad_norm": 0.22976762056350708, "learning_rate": 9.78264499653376e-05, "loss": 0.0133, "step": 13950 }, { "grad_norm": 0.20139671862125397, "learning_rate": 9.782162521674838e-05, "loss": 0.0143, "step": 13960 }, { "grad_norm": 0.25634583830833435, "learning_rate": 9.781679523845812e-05, "loss": 0.0121, "step": 13970 }, { "grad_norm": 0.19379256665706635, "learning_rate": 9.781196003099502e-05, "loss": 0.0138, "step": 13980 }, { "grad_norm": 0.16900914907455444, "learning_rate": 9.780711959488786e-05, "loss": 0.0118, "step": 13990 }, { "grad_norm": 0.2084251344203949, "learning_rate": 9.780227393066599e-05, "loss": 0.0146, "step": 14000 }, { "grad_norm": 0.20739564299583435, "learning_rate": 9.77974230388593e-05, "loss": 0.013, "step": 14010 }, { "grad_norm": 0.2627558708190918, "learning_rate": 9.779256691999829e-05, "loss": 0.0125, "step": 14020 }, { "grad_norm": 0.241111159324646, "learning_rate": 9.778770557461403e-05, "loss": 0.0132, "step": 14030 }, { "grad_norm": 0.23169122636318207, "learning_rate": 9.778283900323812e-05, "loss": 0.0126, "step": 14040 }, { "grad_norm": 0.16425998508930206, "learning_rate": 9.777796720640277e-05, "loss": 0.0125, "step": 14050 }, { "grad_norm": 0.23831380903720856, "learning_rate": 9.777309018464078e-05, "loss": 0.0135, "step": 14060 }, { "grad_norm": 0.23817141354084015, "learning_rate": 9.776820793848547e-05, "loss": 0.0131, "step": 14070 }, { "grad_norm": 0.2082299292087555, "learning_rate": 9.776332046847075e-05, "loss": 0.0119, "step": 14080 }, { "grad_norm": 0.1662086546421051, "learning_rate": 9.775842777513111e-05, "loss": 0.0126, "step": 14090 }, { "grad_norm": 0.2360081523656845, "learning_rate": 9.775352985900163e-05, "loss": 0.011, "step": 14100 }, { "grad_norm": 0.1850288361310959, "learning_rate": 9.774862672061791e-05, "loss": 0.0112, "step": 14110 }, { "grad_norm": 0.16144044697284698, "learning_rate": 9.774371836051616e-05, "loss": 0.0104, "step": 14120 }, { "grad_norm": 0.20173223316669464, "learning_rate": 9.773880477923315e-05, "loss": 0.0134, "step": 14130 }, { "grad_norm": 0.18036983907222748, "learning_rate": 9.773388597730623e-05, "loss": 0.0116, "step": 14140 }, { "grad_norm": 0.19358958303928375, "learning_rate": 9.77289619552733e-05, "loss": 0.0111, "step": 14150 }, { "grad_norm": 0.19823656976222992, "learning_rate": 9.772403271367285e-05, "loss": 0.0119, "step": 14160 }, { "grad_norm": 0.15721914172172546, "learning_rate": 9.771909825304396e-05, "loss": 0.0118, "step": 14170 }, { "grad_norm": 0.2141619622707367, "learning_rate": 9.771415857392619e-05, "loss": 0.0112, "step": 14180 }, { "grad_norm": 0.25704342126846313, "learning_rate": 9.770921367685978e-05, "loss": 0.0112, "step": 14190 }, { "grad_norm": 0.24680233001708984, "learning_rate": 9.770426356238551e-05, "loss": 0.0124, "step": 14200 }, { "grad_norm": 0.1920509934425354, "learning_rate": 9.769930823104469e-05, "loss": 0.0113, "step": 14210 }, { "grad_norm": 0.17581354081630707, "learning_rate": 9.769434768337926e-05, "loss": 0.0112, "step": 14220 }, { "grad_norm": 0.18436987698078156, "learning_rate": 9.768938191993164e-05, "loss": 0.0111, "step": 14230 }, { "grad_norm": 0.2296726256608963, "learning_rate": 9.768441094124494e-05, "loss": 0.0118, "step": 14240 }, { "grad_norm": 0.21140798926353455, "learning_rate": 9.767943474786275e-05, "loss": 0.0125, "step": 14250 }, { "grad_norm": 0.24951909482479095, "learning_rate": 9.767445334032923e-05, "loss": 0.0115, "step": 14260 }, { "grad_norm": 0.24279069900512695, "learning_rate": 9.766946671918919e-05, "loss": 0.012, "step": 14270 }, { "grad_norm": 0.2403528094291687, "learning_rate": 9.766447488498796e-05, "loss": 0.0123, "step": 14280 }, { "grad_norm": 0.2404792159795761, "learning_rate": 9.765947783827139e-05, "loss": 0.0148, "step": 14290 }, { "grad_norm": 0.25733086466789246, "learning_rate": 9.765447557958599e-05, "loss": 0.0129, "step": 14300 }, { "grad_norm": 0.254571795463562, "learning_rate": 9.764946810947879e-05, "loss": 0.0145, "step": 14310 }, { "grad_norm": 0.23505890369415283, "learning_rate": 9.764445542849738e-05, "loss": 0.0145, "step": 14320 }, { "grad_norm": 0.2632765769958496, "learning_rate": 9.763943753718998e-05, "loss": 0.0128, "step": 14330 }, { "grad_norm": 0.2570973038673401, "learning_rate": 9.76344144361053e-05, "loss": 0.0132, "step": 14340 }, { "grad_norm": 0.245000422000885, "learning_rate": 9.762938612579269e-05, "loss": 0.0127, "step": 14350 }, { "grad_norm": 0.2523738741874695, "learning_rate": 9.762435260680202e-05, "loss": 0.0132, "step": 14360 }, { "grad_norm": 0.22236236929893494, "learning_rate": 9.761931387968373e-05, "loss": 0.0127, "step": 14370 }, { "grad_norm": 0.21433699131011963, "learning_rate": 9.76142699449889e-05, "loss": 0.0142, "step": 14380 }, { "grad_norm": 0.22866030037403107, "learning_rate": 9.760922080326908e-05, "loss": 0.0123, "step": 14390 }, { "grad_norm": 0.22727347910404205, "learning_rate": 9.760416645507644e-05, "loss": 0.0113, "step": 14400 }, { "grad_norm": 0.19756031036376953, "learning_rate": 9.759910690096375e-05, "loss": 0.0136, "step": 14410 }, { "grad_norm": 0.16626620292663574, "learning_rate": 9.759404214148429e-05, "loss": 0.0126, "step": 14420 }, { "grad_norm": 0.1716020554304123, "learning_rate": 9.758897217719191e-05, "loss": 0.013, "step": 14430 }, { "grad_norm": 0.24008162319660187, "learning_rate": 9.758389700864113e-05, "loss": 0.0147, "step": 14440 }, { "grad_norm": 0.24852292239665985, "learning_rate": 9.757881663638688e-05, "loss": 0.0134, "step": 14450 }, { "grad_norm": 0.29128679633140564, "learning_rate": 9.757373106098478e-05, "loss": 0.0141, "step": 14460 }, { "grad_norm": 0.21704760193824768, "learning_rate": 9.756864028299097e-05, "loss": 0.0132, "step": 14470 }, { "grad_norm": 0.22721920907497406, "learning_rate": 9.75635443029622e-05, "loss": 0.0151, "step": 14480 }, { "grad_norm": 0.19514808058738708, "learning_rate": 9.755844312145572e-05, "loss": 0.0132, "step": 14490 }, { "grad_norm": 0.22157619893550873, "learning_rate": 9.755333673902941e-05, "loss": 0.0127, "step": 14500 }, { "grad_norm": 0.22774189710617065, "learning_rate": 9.75482251562417e-05, "loss": 0.0136, "step": 14510 }, { "grad_norm": 0.18632622063159943, "learning_rate": 9.754310837365155e-05, "loss": 0.0124, "step": 14520 }, { "grad_norm": 0.2172672301530838, "learning_rate": 9.753798639181856e-05, "loss": 0.0134, "step": 14530 }, { "grad_norm": 0.18262220919132233, "learning_rate": 9.753285921130286e-05, "loss": 0.012, "step": 14540 }, { "grad_norm": 0.18493621051311493, "learning_rate": 9.752772683266512e-05, "loss": 0.0109, "step": 14550 }, { "grad_norm": 0.20669351518154144, "learning_rate": 9.752258925646665e-05, "loss": 0.0116, "step": 14560 }, { "grad_norm": 0.23536472022533417, "learning_rate": 9.751744648326926e-05, "loss": 0.0134, "step": 14570 }, { "grad_norm": 0.22055462002754211, "learning_rate": 9.751229851363536e-05, "loss": 0.0135, "step": 14580 }, { "grad_norm": 0.2225675731897354, "learning_rate": 9.750714534812793e-05, "loss": 0.0128, "step": 14590 }, { "grad_norm": 0.28034117817878723, "learning_rate": 9.750198698731053e-05, "loss": 0.0115, "step": 14600 }, { "grad_norm": 0.21085470914840698, "learning_rate": 9.749682343174722e-05, "loss": 0.0146, "step": 14610 }, { "grad_norm": 0.23164887726306915, "learning_rate": 9.749165468200272e-05, "loss": 0.0136, "step": 14620 }, { "grad_norm": 0.28503042459487915, "learning_rate": 9.748648073864229e-05, "loss": 0.0132, "step": 14630 }, { "grad_norm": 0.24298714101314545, "learning_rate": 9.748130160223168e-05, "loss": 0.0133, "step": 14640 }, { "grad_norm": 0.21789710223674774, "learning_rate": 9.747611727333734e-05, "loss": 0.0136, "step": 14650 }, { "grad_norm": 0.24831292033195496, "learning_rate": 9.74709277525262e-05, "loss": 0.0137, "step": 14660 }, { "grad_norm": 0.24151383340358734, "learning_rate": 9.746573304036576e-05, "loss": 0.0124, "step": 14670 }, { "grad_norm": 0.22693979740142822, "learning_rate": 9.746053313742412e-05, "loss": 0.0136, "step": 14680 }, { "grad_norm": 0.21503475308418274, "learning_rate": 9.745532804426994e-05, "loss": 0.0117, "step": 14690 }, { "grad_norm": 0.16514329612255096, "learning_rate": 9.745011776147242e-05, "loss": 0.0112, "step": 14700 }, { "grad_norm": 0.1643703281879425, "learning_rate": 9.744490228960138e-05, "loss": 0.0124, "step": 14710 }, { "grad_norm": 0.1936594694852829, "learning_rate": 9.743968162922713e-05, "loss": 0.0127, "step": 14720 }, { "grad_norm": 0.23914769291877747, "learning_rate": 9.743445578092064e-05, "loss": 0.0131, "step": 14730 }, { "grad_norm": 0.16315220296382904, "learning_rate": 9.742922474525338e-05, "loss": 0.0127, "step": 14740 }, { "grad_norm": 0.194951593875885, "learning_rate": 9.742398852279741e-05, "loss": 0.0124, "step": 14750 }, { "grad_norm": 0.19749221205711365, "learning_rate": 9.741874711412535e-05, "loss": 0.0124, "step": 14760 }, { "grad_norm": 0.2164951115846634, "learning_rate": 9.741350051981042e-05, "loss": 0.012, "step": 14770 }, { "grad_norm": 0.20128485560417175, "learning_rate": 9.740824874042633e-05, "loss": 0.0114, "step": 14780 }, { "grad_norm": 0.21634626388549805, "learning_rate": 9.740299177654746e-05, "loss": 0.0125, "step": 14790 }, { "grad_norm": 0.21022899448871613, "learning_rate": 9.739772962874867e-05, "loss": 0.0111, "step": 14800 }, { "grad_norm": 0.24238379299640656, "learning_rate": 9.739246229760541e-05, "loss": 0.011, "step": 14810 }, { "grad_norm": 0.235488623380661, "learning_rate": 9.738718978369376e-05, "loss": 0.0131, "step": 14820 }, { "grad_norm": 0.23971058428287506, "learning_rate": 9.738191208759025e-05, "loss": 0.0112, "step": 14830 }, { "grad_norm": 0.19238930940628052, "learning_rate": 9.73766292098721e-05, "loss": 0.0106, "step": 14840 }, { "grad_norm": 0.2576615810394287, "learning_rate": 9.737134115111699e-05, "loss": 0.0116, "step": 14850 }, { "grad_norm": 0.1926189661026001, "learning_rate": 9.736604791190323e-05, "loss": 0.0107, "step": 14860 }, { "grad_norm": 0.25894609093666077, "learning_rate": 9.73607494928097e-05, "loss": 0.0112, "step": 14870 }, { "grad_norm": 0.1742730587720871, "learning_rate": 9.735544589441581e-05, "loss": 0.0119, "step": 14880 }, { "grad_norm": 0.18665170669555664, "learning_rate": 9.735013711730154e-05, "loss": 0.0117, "step": 14890 }, { "grad_norm": 0.2493816614151001, "learning_rate": 9.734482316204747e-05, "loss": 0.0111, "step": 14900 }, { "grad_norm": 0.1946944147348404, "learning_rate": 9.733950402923473e-05, "loss": 0.0113, "step": 14910 }, { "grad_norm": 0.21196520328521729, "learning_rate": 9.7334179719445e-05, "loss": 0.0112, "step": 14920 }, { "grad_norm": 0.16605153679847717, "learning_rate": 9.732885023326053e-05, "loss": 0.0122, "step": 14930 }, { "grad_norm": 0.21592025458812714, "learning_rate": 9.732351557126418e-05, "loss": 0.0115, "step": 14940 }, { "grad_norm": 0.2558695077896118, "learning_rate": 9.731817573403929e-05, "loss": 0.012, "step": 14950 }, { "grad_norm": 0.1938157081604004, "learning_rate": 9.731283072216985e-05, "loss": 0.0118, "step": 14960 }, { "grad_norm": 0.1247243732213974, "learning_rate": 9.730748053624039e-05, "loss": 0.0122, "step": 14970 }, { "grad_norm": 0.23953388631343842, "learning_rate": 9.730212517683598e-05, "loss": 0.0113, "step": 14980 }, { "grad_norm": 0.1467108428478241, "learning_rate": 9.729676464454228e-05, "loss": 0.0119, "step": 14990 }, { "grad_norm": 0.19572418928146362, "learning_rate": 9.72913989399455e-05, "loss": 0.0096, "step": 15000 }, { "grad_norm": 0.23253875970840454, "learning_rate": 9.728602806363242e-05, "loss": 0.0122, "step": 15010 }, { "grad_norm": 0.2057059109210968, "learning_rate": 9.728065201619043e-05, "loss": 0.0114, "step": 15020 }, { "grad_norm": 0.207879438996315, "learning_rate": 9.727527079820742e-05, "loss": 0.0118, "step": 15030 }, { "grad_norm": 0.14693601429462433, "learning_rate": 9.726988441027186e-05, "loss": 0.0136, "step": 15040 }, { "grad_norm": 0.21833258867263794, "learning_rate": 9.726449285297281e-05, "loss": 0.0129, "step": 15050 }, { "grad_norm": 0.16854070127010345, "learning_rate": 9.72590961268999e-05, "loss": 0.0106, "step": 15060 }, { "grad_norm": 0.28356319665908813, "learning_rate": 9.725369423264328e-05, "loss": 0.0128, "step": 15070 }, { "grad_norm": 0.17275458574295044, "learning_rate": 9.72482871707937e-05, "loss": 0.0129, "step": 15080 }, { "grad_norm": 0.1827777922153473, "learning_rate": 9.724287494194247e-05, "loss": 0.012, "step": 15090 }, { "grad_norm": 0.1900530755519867, "learning_rate": 9.723745754668147e-05, "loss": 0.0102, "step": 15100 }, { "grad_norm": 0.14755183458328247, "learning_rate": 9.723203498560313e-05, "loss": 0.0104, "step": 15110 }, { "grad_norm": 0.1864483803510666, "learning_rate": 9.722660725930046e-05, "loss": 0.0115, "step": 15120 }, { "grad_norm": 0.16236206889152527, "learning_rate": 9.722117436836702e-05, "loss": 0.0112, "step": 15130 }, { "grad_norm": 0.22836904227733612, "learning_rate": 9.721573631339696e-05, "loss": 0.0125, "step": 15140 }, { "grad_norm": 0.2360716462135315, "learning_rate": 9.721029309498494e-05, "loss": 0.0121, "step": 15150 }, { "grad_norm": 0.16937436163425446, "learning_rate": 9.720484471372627e-05, "loss": 0.0106, "step": 15160 }, { "grad_norm": 0.2602883577346802, "learning_rate": 9.719939117021673e-05, "loss": 0.0122, "step": 15170 }, { "grad_norm": 0.21171659231185913, "learning_rate": 9.719393246505275e-05, "loss": 0.012, "step": 15180 }, { "grad_norm": 0.22818003594875336, "learning_rate": 9.718846859883128e-05, "loss": 0.0111, "step": 15190 }, { "grad_norm": 0.24655230343341827, "learning_rate": 9.718299957214982e-05, "loss": 0.0145, "step": 15200 }, { "grad_norm": 0.20170089602470398, "learning_rate": 9.717752538560646e-05, "loss": 0.0139, "step": 15210 }, { "grad_norm": 0.24755193293094635, "learning_rate": 9.717204603979986e-05, "loss": 0.0106, "step": 15220 }, { "grad_norm": 0.20934604108333588, "learning_rate": 9.716656153532922e-05, "loss": 0.0116, "step": 15230 }, { "grad_norm": 0.16902285814285278, "learning_rate": 9.716107187279434e-05, "loss": 0.0116, "step": 15240 }, { "grad_norm": 0.1202966570854187, "learning_rate": 9.715557705279555e-05, "loss": 0.0102, "step": 15250 }, { "grad_norm": 0.16270792484283447, "learning_rate": 9.715007707593372e-05, "loss": 0.0117, "step": 15260 }, { "grad_norm": 0.21259623765945435, "learning_rate": 9.714457194281036e-05, "loss": 0.0123, "step": 15270 }, { "grad_norm": 0.3299506604671478, "learning_rate": 9.713906165402751e-05, "loss": 0.0143, "step": 15280 }, { "grad_norm": 0.29110580682754517, "learning_rate": 9.713354621018774e-05, "loss": 0.0156, "step": 15290 }, { "grad_norm": 0.2421134114265442, "learning_rate": 9.712802561189422e-05, "loss": 0.0153, "step": 15300 }, { "grad_norm": 0.22315023839473724, "learning_rate": 9.712249985975069e-05, "loss": 0.0143, "step": 15310 }, { "grad_norm": 0.2141207456588745, "learning_rate": 9.71169689543614e-05, "loss": 0.0134, "step": 15320 }, { "grad_norm": 0.27937594056129456, "learning_rate": 9.711143289633123e-05, "loss": 0.0124, "step": 15330 }, { "grad_norm": 0.2762961685657501, "learning_rate": 9.710589168626561e-05, "loss": 0.0137, "step": 15340 }, { "grad_norm": 0.14826983213424683, "learning_rate": 9.710034532477048e-05, "loss": 0.0123, "step": 15350 }, { "grad_norm": 0.2019694298505783, "learning_rate": 9.709479381245239e-05, "loss": 0.0129, "step": 15360 }, { "grad_norm": 0.25537723302841187, "learning_rate": 9.708923714991847e-05, "loss": 0.0127, "step": 15370 }, { "grad_norm": 0.2756803631782532, "learning_rate": 9.708367533777638e-05, "loss": 0.0133, "step": 15380 }, { "grad_norm": 0.25228866934776306, "learning_rate": 9.707810837663431e-05, "loss": 0.012, "step": 15390 }, { "grad_norm": 0.16712751984596252, "learning_rate": 9.707253626710113e-05, "loss": 0.0115, "step": 15400 }, { "grad_norm": 0.18791979551315308, "learning_rate": 9.706695900978613e-05, "loss": 0.0115, "step": 15410 }, { "grad_norm": 0.2649623155593872, "learning_rate": 9.706137660529926e-05, "loss": 0.0109, "step": 15420 }, { "grad_norm": 0.18762795627117157, "learning_rate": 9.705578905425101e-05, "loss": 0.0112, "step": 15430 }, { "grad_norm": 0.2606821656227112, "learning_rate": 9.705019635725241e-05, "loss": 0.0114, "step": 15440 }, { "grad_norm": 0.18972517549991608, "learning_rate": 9.704459851491508e-05, "loss": 0.0139, "step": 15450 }, { "grad_norm": 0.26453378796577454, "learning_rate": 9.703899552785118e-05, "loss": 0.0115, "step": 15460 }, { "grad_norm": 0.16790460050106049, "learning_rate": 9.703338739667346e-05, "loss": 0.0125, "step": 15470 }, { "grad_norm": 0.24278780817985535, "learning_rate": 9.70277741219952e-05, "loss": 0.0129, "step": 15480 }, { "grad_norm": 0.18431472778320312, "learning_rate": 9.702215570443027e-05, "loss": 0.011, "step": 15490 }, { "grad_norm": 0.21146179735660553, "learning_rate": 9.701653214459309e-05, "loss": 0.0129, "step": 15500 }, { "grad_norm": 0.2041109949350357, "learning_rate": 9.701090344309865e-05, "loss": 0.013, "step": 15510 }, { "grad_norm": 0.2250318080186844, "learning_rate": 9.700526960056247e-05, "loss": 0.014, "step": 15520 }, { "grad_norm": 0.215574711561203, "learning_rate": 9.699963061760068e-05, "loss": 0.0135, "step": 15530 }, { "grad_norm": 0.2023819237947464, "learning_rate": 9.699398649482997e-05, "loss": 0.0125, "step": 15540 }, { "grad_norm": 0.23792865872383118, "learning_rate": 9.698833723286753e-05, "loss": 0.0122, "step": 15550 }, { "grad_norm": 0.20080120861530304, "learning_rate": 9.698268283233118e-05, "loss": 0.0123, "step": 15560 }, { "grad_norm": 0.14788153767585754, "learning_rate": 9.697702329383929e-05, "loss": 0.011, "step": 15570 }, { "grad_norm": 0.17040768265724182, "learning_rate": 9.697135861801074e-05, "loss": 0.0111, "step": 15580 }, { "grad_norm": 0.2531072199344635, "learning_rate": 9.696568880546505e-05, "loss": 0.0122, "step": 15590 }, { "grad_norm": 0.24870076775550842, "learning_rate": 9.696001385682223e-05, "loss": 0.0139, "step": 15600 }, { "grad_norm": 0.18638542294502258, "learning_rate": 9.695433377270291e-05, "loss": 0.0126, "step": 15610 }, { "grad_norm": 0.13910992443561554, "learning_rate": 9.694864855372824e-05, "loss": 0.0103, "step": 15620 }, { "grad_norm": 0.19185547530651093, "learning_rate": 9.694295820051995e-05, "loss": 0.0116, "step": 15630 }, { "grad_norm": 0.1797342151403427, "learning_rate": 9.693726271370032e-05, "loss": 0.0111, "step": 15640 }, { "grad_norm": 0.1737290471792221, "learning_rate": 9.693156209389221e-05, "loss": 0.0109, "step": 15650 }, { "grad_norm": 0.23409968614578247, "learning_rate": 9.692585634171905e-05, "loss": 0.011, "step": 15660 }, { "grad_norm": 0.18658974766731262, "learning_rate": 9.692014545780476e-05, "loss": 0.0127, "step": 15670 }, { "grad_norm": 0.1984906643629074, "learning_rate": 9.691442944277393e-05, "loss": 0.0103, "step": 15680 }, { "grad_norm": 0.15723039209842682, "learning_rate": 9.690870829725162e-05, "loss": 0.0108, "step": 15690 }, { "grad_norm": 0.18143655359745026, "learning_rate": 9.69029820218635e-05, "loss": 0.0111, "step": 15700 }, { "grad_norm": 0.17899133265018463, "learning_rate": 9.689725061723579e-05, "loss": 0.0115, "step": 15710 }, { "grad_norm": 0.2211613804101944, "learning_rate": 9.689151408399527e-05, "loss": 0.0114, "step": 15720 }, { "grad_norm": 0.14833475649356842, "learning_rate": 9.688577242276924e-05, "loss": 0.0106, "step": 15730 }, { "grad_norm": 0.24205030500888824, "learning_rate": 9.688002563418566e-05, "loss": 0.0105, "step": 15740 }, { "grad_norm": 0.1493901014328003, "learning_rate": 9.687427371887293e-05, "loss": 0.0111, "step": 15750 }, { "grad_norm": 0.16179107129573822, "learning_rate": 9.686851667746012e-05, "loss": 0.0105, "step": 15760 }, { "grad_norm": 0.20462487637996674, "learning_rate": 9.686275451057677e-05, "loss": 0.0104, "step": 15770 }, { "grad_norm": 0.22517411410808563, "learning_rate": 9.685698721885308e-05, "loss": 0.0112, "step": 15780 }, { "grad_norm": 0.23273544013500214, "learning_rate": 9.68512148029197e-05, "loss": 0.011, "step": 15790 }, { "grad_norm": 0.19252069294452667, "learning_rate": 9.684543726340791e-05, "loss": 0.0103, "step": 15800 }, { "grad_norm": 0.1994853913784027, "learning_rate": 9.683965460094952e-05, "loss": 0.0104, "step": 15810 }, { "grad_norm": 0.15791279077529907, "learning_rate": 9.683386681617694e-05, "loss": 0.0112, "step": 15820 }, { "grad_norm": 0.1344049572944641, "learning_rate": 9.68280739097231e-05, "loss": 0.0112, "step": 15830 }, { "grad_norm": 0.2765013873577118, "learning_rate": 9.682227588222148e-05, "loss": 0.0109, "step": 15840 }, { "grad_norm": 0.17752443253993988, "learning_rate": 9.681647273430618e-05, "loss": 0.0108, "step": 15850 }, { "grad_norm": 0.17245036363601685, "learning_rate": 9.681066446661182e-05, "loss": 0.0112, "step": 15860 }, { "grad_norm": 0.221096470952034, "learning_rate": 9.680485107977357e-05, "loss": 0.0114, "step": 15870 }, { "grad_norm": 0.23016834259033203, "learning_rate": 9.679903257442716e-05, "loss": 0.0104, "step": 15880 }, { "grad_norm": 0.17777329683303833, "learning_rate": 9.679320895120891e-05, "loss": 0.0119, "step": 15890 }, { "grad_norm": 0.21727551519870758, "learning_rate": 9.67873802107557e-05, "loss": 0.0122, "step": 15900 }, { "grad_norm": 0.2253246158361435, "learning_rate": 9.67815463537049e-05, "loss": 0.0114, "step": 15910 }, { "grad_norm": 0.18594872951507568, "learning_rate": 9.677570738069457e-05, "loss": 0.0113, "step": 15920 }, { "grad_norm": 0.19622908532619476, "learning_rate": 9.676986329236318e-05, "loss": 0.0114, "step": 15930 }, { "grad_norm": 0.20216889679431915, "learning_rate": 9.676401408934987e-05, "loss": 0.0112, "step": 15940 }, { "grad_norm": 0.15227536857128143, "learning_rate": 9.675815977229428e-05, "loss": 0.0118, "step": 15950 }, { "grad_norm": 0.18668270111083984, "learning_rate": 9.675230034183664e-05, "loss": 0.0099, "step": 15960 }, { "grad_norm": 0.22950753569602966, "learning_rate": 9.674643579861773e-05, "loss": 0.0108, "step": 15970 }, { "grad_norm": 0.2016977071762085, "learning_rate": 9.674056614327886e-05, "loss": 0.012, "step": 15980 }, { "grad_norm": 0.19632188975811005, "learning_rate": 9.673469137646198e-05, "loss": 0.012, "step": 15990 }, { "grad_norm": 0.15610116720199585, "learning_rate": 9.67288114988095e-05, "loss": 0.0101, "step": 16000 }, { "grad_norm": 0.16910767555236816, "learning_rate": 9.672292651096447e-05, "loss": 0.0114, "step": 16010 }, { "grad_norm": 0.18051639199256897, "learning_rate": 9.671703641357042e-05, "loss": 0.0116, "step": 16020 }, { "grad_norm": 0.21376143395900726, "learning_rate": 9.67111412072715e-05, "loss": 0.0106, "step": 16030 }, { "grad_norm": 0.1774631142616272, "learning_rate": 9.670524089271242e-05, "loss": 0.0097, "step": 16040 }, { "grad_norm": 0.1559000462293625, "learning_rate": 9.669933547053842e-05, "loss": 0.0101, "step": 16050 }, { "grad_norm": 0.1721155196428299, "learning_rate": 9.669342494139531e-05, "loss": 0.011, "step": 16060 }, { "grad_norm": 0.23424826562404633, "learning_rate": 9.668750930592943e-05, "loss": 0.0114, "step": 16070 }, { "grad_norm": 0.16892950236797333, "learning_rate": 9.668158856478775e-05, "loss": 0.0108, "step": 16080 }, { "grad_norm": 0.19219817221164703, "learning_rate": 9.66756627186177e-05, "loss": 0.0099, "step": 16090 }, { "grad_norm": 0.1784491091966629, "learning_rate": 9.666973176806737e-05, "loss": 0.0099, "step": 16100 }, { "grad_norm": 0.27330583333969116, "learning_rate": 9.666379571378534e-05, "loss": 0.0105, "step": 16110 }, { "grad_norm": 0.2357868105173111, "learning_rate": 9.665785455642076e-05, "loss": 0.0105, "step": 16120 }, { "grad_norm": 0.23660613596439362, "learning_rate": 9.665190829662337e-05, "loss": 0.0104, "step": 16130 }, { "grad_norm": 0.1820640116930008, "learning_rate": 9.664595693504342e-05, "loss": 0.0107, "step": 16140 }, { "grad_norm": 0.15566982328891754, "learning_rate": 9.664000047233175e-05, "loss": 0.011, "step": 16150 }, { "grad_norm": 0.19300054013729095, "learning_rate": 9.663403890913976e-05, "loss": 0.0111, "step": 16160 }, { "grad_norm": 0.1715439409017563, "learning_rate": 9.662807224611938e-05, "loss": 0.0104, "step": 16170 }, { "grad_norm": 0.17512451112270355, "learning_rate": 9.662210048392311e-05, "loss": 0.0097, "step": 16180 }, { "grad_norm": 0.2570105791091919, "learning_rate": 9.661612362320405e-05, "loss": 0.0102, "step": 16190 }, { "grad_norm": 0.30118367075920105, "learning_rate": 9.661014166461579e-05, "loss": 0.0105, "step": 16200 }, { "grad_norm": 0.1838231235742569, "learning_rate": 9.66041546088125e-05, "loss": 0.0124, "step": 16210 }, { "grad_norm": 0.21977104246616364, "learning_rate": 9.659816245644895e-05, "loss": 0.011, "step": 16220 }, { "grad_norm": 0.21320068836212158, "learning_rate": 9.65921652081804e-05, "loss": 0.0098, "step": 16230 }, { "grad_norm": 0.1649157851934433, "learning_rate": 9.658616286466271e-05, "loss": 0.0096, "step": 16240 }, { "grad_norm": 0.18682563304901123, "learning_rate": 9.65801554265523e-05, "loss": 0.0099, "step": 16250 }, { "grad_norm": 0.1854579895734787, "learning_rate": 9.657414289450612e-05, "loss": 0.0106, "step": 16260 }, { "grad_norm": 0.15347005426883698, "learning_rate": 9.656812526918171e-05, "loss": 0.011, "step": 16270 }, { "grad_norm": 0.16445504128932953, "learning_rate": 9.656210255123712e-05, "loss": 0.0112, "step": 16280 }, { "grad_norm": 0.18991799652576447, "learning_rate": 9.6556074741331e-05, "loss": 0.0118, "step": 16290 }, { "grad_norm": 0.17990277707576752, "learning_rate": 9.655004184012256e-05, "loss": 0.0094, "step": 16300 }, { "grad_norm": 0.16808989644050598, "learning_rate": 9.654400384827152e-05, "loss": 0.011, "step": 16310 }, { "grad_norm": 0.17789548635482788, "learning_rate": 9.653796076643818e-05, "loss": 0.0097, "step": 16320 }, { "grad_norm": 0.22630704939365387, "learning_rate": 9.653191259528344e-05, "loss": 0.0104, "step": 16330 }, { "grad_norm": 0.1932152658700943, "learning_rate": 9.65258593354687e-05, "loss": 0.011, "step": 16340 }, { "grad_norm": 0.1871771365404129, "learning_rate": 9.651980098765591e-05, "loss": 0.0106, "step": 16350 }, { "grad_norm": 0.2512326240539551, "learning_rate": 9.651373755250765e-05, "loss": 0.0104, "step": 16360 }, { "grad_norm": 0.2139042764902115, "learning_rate": 9.650766903068697e-05, "loss": 0.0112, "step": 16370 }, { "grad_norm": 0.14687085151672363, "learning_rate": 9.650159542285753e-05, "loss": 0.0112, "step": 16380 }, { "grad_norm": 0.1641414612531662, "learning_rate": 9.649551672968353e-05, "loss": 0.0124, "step": 16390 }, { "grad_norm": 0.236938938498497, "learning_rate": 9.648943295182973e-05, "loss": 0.0128, "step": 16400 }, { "grad_norm": 0.25046074390411377, "learning_rate": 9.648334408996144e-05, "loss": 0.012, "step": 16410 }, { "grad_norm": 0.26107388734817505, "learning_rate": 9.647725014474452e-05, "loss": 0.0118, "step": 16420 }, { "grad_norm": 0.23789626359939575, "learning_rate": 9.64711511168454e-05, "loss": 0.0128, "step": 16430 }, { "grad_norm": 0.19504399597644806, "learning_rate": 9.646504700693108e-05, "loss": 0.0119, "step": 16440 }, { "grad_norm": 0.18813838064670563, "learning_rate": 9.645893781566907e-05, "loss": 0.012, "step": 16450 }, { "grad_norm": 0.18608303368091583, "learning_rate": 9.645282354372744e-05, "loss": 0.0112, "step": 16460 }, { "grad_norm": 0.190056711435318, "learning_rate": 9.644670419177491e-05, "loss": 0.0104, "step": 16470 }, { "grad_norm": 0.1830611526966095, "learning_rate": 9.644057976048062e-05, "loss": 0.0114, "step": 16480 }, { "grad_norm": 0.1762164831161499, "learning_rate": 9.643445025051435e-05, "loss": 0.0096, "step": 16490 }, { "grad_norm": 0.25879591703414917, "learning_rate": 9.642831566254641e-05, "loss": 0.0117, "step": 16500 }, { "grad_norm": 0.18168506026268005, "learning_rate": 9.642217599724769e-05, "loss": 0.0118, "step": 16510 }, { "grad_norm": 0.15754057466983795, "learning_rate": 9.64160312552896e-05, "loss": 0.0109, "step": 16520 }, { "grad_norm": 0.25008824467658997, "learning_rate": 9.64098814373441e-05, "loss": 0.0121, "step": 16530 }, { "grad_norm": 0.176801398396492, "learning_rate": 9.640372654408374e-05, "loss": 0.0114, "step": 16540 }, { "grad_norm": 0.2552355229854584, "learning_rate": 9.639756657618162e-05, "loss": 0.0119, "step": 16550 }, { "grad_norm": 0.575444221496582, "learning_rate": 9.639140153431138e-05, "loss": 0.0144, "step": 16560 }, { "grad_norm": 0.1522151529788971, "learning_rate": 9.638523141914721e-05, "loss": 0.0146, "step": 16570 }, { "grad_norm": 0.2928134500980377, "learning_rate": 9.637905623136388e-05, "loss": 0.0139, "step": 16580 }, { "grad_norm": 0.2818327248096466, "learning_rate": 9.637287597163669e-05, "loss": 0.0144, "step": 16590 }, { "grad_norm": 0.22643108665943146, "learning_rate": 9.63666906406415e-05, "loss": 0.013, "step": 16600 }, { "grad_norm": 0.28937214612960815, "learning_rate": 9.636050023905473e-05, "loss": 0.0144, "step": 16610 }, { "grad_norm": 0.2980504035949707, "learning_rate": 9.635430476755336e-05, "loss": 0.0149, "step": 16620 }, { "grad_norm": 0.25468671321868896, "learning_rate": 9.63481042268149e-05, "loss": 0.0141, "step": 16630 }, { "grad_norm": 0.27711036801338196, "learning_rate": 9.634189861751745e-05, "loss": 0.0123, "step": 16640 }, { "grad_norm": 0.2558392286300659, "learning_rate": 9.633568794033967e-05, "loss": 0.0127, "step": 16650 }, { "grad_norm": 0.22336597740650177, "learning_rate": 9.63294721959607e-05, "loss": 0.0122, "step": 16660 }, { "grad_norm": 0.22562703490257263, "learning_rate": 9.63232513850603e-05, "loss": 0.0119, "step": 16670 }, { "grad_norm": 0.24069470167160034, "learning_rate": 9.631702550831878e-05, "loss": 0.0131, "step": 16680 }, { "grad_norm": 0.16450072824954987, "learning_rate": 9.631079456641698e-05, "loss": 0.0105, "step": 16690 }, { "grad_norm": 0.18716001510620117, "learning_rate": 9.630455856003632e-05, "loss": 0.0101, "step": 16700 }, { "grad_norm": 0.15917038917541504, "learning_rate": 9.629831748985876e-05, "loss": 0.0115, "step": 16710 }, { "grad_norm": 0.23869073390960693, "learning_rate": 9.629207135656679e-05, "loss": 0.0127, "step": 16720 }, { "grad_norm": 0.18490181863307953, "learning_rate": 9.628582016084353e-05, "loss": 0.0122, "step": 16730 }, { "grad_norm": 0.16459880769252777, "learning_rate": 9.627956390337254e-05, "loss": 0.012, "step": 16740 }, { "grad_norm": 0.208418071269989, "learning_rate": 9.627330258483802e-05, "loss": 0.0116, "step": 16750 }, { "grad_norm": 0.3049144148826599, "learning_rate": 9.62670362059247e-05, "loss": 0.0125, "step": 16760 }, { "grad_norm": 0.20229986310005188, "learning_rate": 9.626076476731786e-05, "loss": 0.0134, "step": 16770 }, { "grad_norm": 0.2079063355922699, "learning_rate": 9.625448826970336e-05, "loss": 0.0124, "step": 16780 }, { "grad_norm": 0.22997985780239105, "learning_rate": 9.624820671376755e-05, "loss": 0.0117, "step": 16790 }, { "grad_norm": 0.17269666492938995, "learning_rate": 9.62419201001974e-05, "loss": 0.0103, "step": 16800 }, { "grad_norm": 0.140580952167511, "learning_rate": 9.623562842968037e-05, "loss": 0.0119, "step": 16810 }, { "grad_norm": 0.17830225825309753, "learning_rate": 9.622933170290454e-05, "loss": 0.0124, "step": 16820 }, { "grad_norm": 0.1969769448041916, "learning_rate": 9.622302992055849e-05, "loss": 0.0116, "step": 16830 }, { "grad_norm": 0.21995198726654053, "learning_rate": 9.62167230833314e-05, "loss": 0.0131, "step": 16840 }, { "grad_norm": 0.2807694971561432, "learning_rate": 9.621041119191295e-05, "loss": 0.014, "step": 16850 }, { "grad_norm": 0.2302873134613037, "learning_rate": 9.620409424699342e-05, "loss": 0.0117, "step": 16860 }, { "grad_norm": 0.21992981433868408, "learning_rate": 9.619777224926359e-05, "loss": 0.0119, "step": 16870 }, { "grad_norm": 0.2031913548707962, "learning_rate": 9.619144519941485e-05, "loss": 0.0121, "step": 16880 }, { "grad_norm": 0.16475898027420044, "learning_rate": 9.618511309813912e-05, "loss": 0.0114, "step": 16890 }, { "grad_norm": 0.2763487994670868, "learning_rate": 9.617877594612886e-05, "loss": 0.0119, "step": 16900 }, { "grad_norm": 0.18874309957027435, "learning_rate": 9.617243374407707e-05, "loss": 0.0107, "step": 16910 }, { "grad_norm": 0.14915448427200317, "learning_rate": 9.616608649267736e-05, "loss": 0.011, "step": 16920 }, { "grad_norm": 0.18729594349861145, "learning_rate": 9.615973419262385e-05, "loss": 0.0113, "step": 16930 }, { "grad_norm": 0.2600887417793274, "learning_rate": 9.615337684461119e-05, "loss": 0.0139, "step": 16940 }, { "grad_norm": 0.19840534031391144, "learning_rate": 9.614701444933465e-05, "loss": 0.0142, "step": 16950 }, { "grad_norm": 0.23401248455047607, "learning_rate": 9.614064700748997e-05, "loss": 0.011, "step": 16960 }, { "grad_norm": 0.2681317627429962, "learning_rate": 9.613427451977352e-05, "loss": 0.0129, "step": 16970 }, { "grad_norm": 0.17918500304222107, "learning_rate": 9.612789698688216e-05, "loss": 0.0127, "step": 16980 }, { "grad_norm": 0.1539125293493271, "learning_rate": 9.612151440951334e-05, "loss": 0.0117, "step": 16990 }, { "grad_norm": 0.1751260906457901, "learning_rate": 9.611512678836506e-05, "loss": 0.0118, "step": 17000 }, { "grad_norm": 0.1835380345582962, "learning_rate": 9.610873412413584e-05, "loss": 0.0113, "step": 17010 }, { "grad_norm": 0.21966581046581268, "learning_rate": 9.610233641752476e-05, "loss": 0.0125, "step": 17020 }, { "grad_norm": 0.14782889187335968, "learning_rate": 9.609593366923151e-05, "loss": 0.0129, "step": 17030 }, { "grad_norm": 0.15814237296581268, "learning_rate": 9.608952587995625e-05, "loss": 0.0109, "step": 17040 }, { "grad_norm": 0.19711320102214813, "learning_rate": 9.608311305039972e-05, "loss": 0.0119, "step": 17050 }, { "grad_norm": 0.137323260307312, "learning_rate": 9.607669518126326e-05, "loss": 0.0103, "step": 17060 }, { "grad_norm": 0.16633591055870056, "learning_rate": 9.607027227324866e-05, "loss": 0.0113, "step": 17070 }, { "grad_norm": 0.18767675757408142, "learning_rate": 9.606384432705837e-05, "loss": 0.0103, "step": 17080 }, { "grad_norm": 0.19804571568965912, "learning_rate": 9.60574113433953e-05, "loss": 0.0112, "step": 17090 }, { "grad_norm": 0.17675839364528656, "learning_rate": 9.6050973322963e-05, "loss": 0.0116, "step": 17100 }, { "grad_norm": 0.15908025205135345, "learning_rate": 9.604453026646547e-05, "loss": 0.0103, "step": 17110 }, { "grad_norm": 0.22675122320652008, "learning_rate": 9.603808217460735e-05, "loss": 0.0106, "step": 17120 }, { "grad_norm": 0.1884813904762268, "learning_rate": 9.603162904809377e-05, "loss": 0.0105, "step": 17130 }, { "grad_norm": 0.2524895668029785, "learning_rate": 9.602517088763045e-05, "loss": 0.0116, "step": 17140 }, { "grad_norm": 0.195269376039505, "learning_rate": 9.601870769392365e-05, "loss": 0.0114, "step": 17150 }, { "grad_norm": 0.1861712485551834, "learning_rate": 9.601223946768017e-05, "loss": 0.0128, "step": 17160 }, { "grad_norm": 0.1772746741771698, "learning_rate": 9.600576620960734e-05, "loss": 0.0145, "step": 17170 }, { "grad_norm": 0.1923585683107376, "learning_rate": 9.599928792041308e-05, "loss": 0.0128, "step": 17180 }, { "grad_norm": 0.23340526223182678, "learning_rate": 9.599280460080587e-05, "loss": 0.0117, "step": 17190 }, { "grad_norm": 0.18748220801353455, "learning_rate": 9.59863162514947e-05, "loss": 0.0115, "step": 17200 }, { "grad_norm": 0.21903656423091888, "learning_rate": 9.597982287318911e-05, "loss": 0.0128, "step": 17210 }, { "grad_norm": 0.18793684244155884, "learning_rate": 9.597332446659923e-05, "loss": 0.0116, "step": 17220 }, { "grad_norm": 0.1892538070678711, "learning_rate": 9.59668210324357e-05, "loss": 0.0109, "step": 17230 }, { "grad_norm": 0.16644920408725739, "learning_rate": 9.596031257140974e-05, "loss": 0.0108, "step": 17240 }, { "grad_norm": 0.18829156458377838, "learning_rate": 9.59537990842331e-05, "loss": 0.0121, "step": 17250 }, { "grad_norm": 0.14341312646865845, "learning_rate": 9.594728057161806e-05, "loss": 0.0105, "step": 17260 }, { "grad_norm": 0.23102553188800812, "learning_rate": 9.594075703427752e-05, "loss": 0.0126, "step": 17270 }, { "grad_norm": 0.1745271533727646, "learning_rate": 9.593422847292486e-05, "loss": 0.0118, "step": 17280 }, { "grad_norm": 0.22146660089492798, "learning_rate": 9.592769488827402e-05, "loss": 0.0109, "step": 17290 }, { "grad_norm": 0.20354753732681274, "learning_rate": 9.592115628103952e-05, "loss": 0.013, "step": 17300 }, { "grad_norm": 0.15396736562252045, "learning_rate": 9.591461265193643e-05, "loss": 0.0097, "step": 17310 }, { "grad_norm": 0.17184826731681824, "learning_rate": 9.590806400168032e-05, "loss": 0.0095, "step": 17320 }, { "grad_norm": 0.1493832767009735, "learning_rate": 9.590151033098735e-05, "loss": 0.0098, "step": 17330 }, { "grad_norm": 0.17730262875556946, "learning_rate": 9.589495164057423e-05, "loss": 0.0102, "step": 17340 }, { "grad_norm": 0.15623579919338226, "learning_rate": 9.58883879311582e-05, "loss": 0.0102, "step": 17350 }, { "grad_norm": 0.21163241565227509, "learning_rate": 9.588181920345705e-05, "loss": 0.0099, "step": 17360 }, { "grad_norm": 0.2862675189971924, "learning_rate": 9.587524545818913e-05, "loss": 0.0107, "step": 17370 }, { "grad_norm": 0.17467767000198364, "learning_rate": 9.586866669607335e-05, "loss": 0.0127, "step": 17380 }, { "grad_norm": 0.17380842566490173, "learning_rate": 9.586208291782915e-05, "loss": 0.0119, "step": 17390 }, { "grad_norm": 0.1538197100162506, "learning_rate": 9.58554941241765e-05, "loss": 0.0118, "step": 17400 }, { "grad_norm": 0.14204463362693787, "learning_rate": 9.584890031583596e-05, "loss": 0.0123, "step": 17410 }, { "grad_norm": 0.21430082619190216, "learning_rate": 9.584230149352861e-05, "loss": 0.0106, "step": 17420 }, { "grad_norm": 0.2128249704837799, "learning_rate": 9.58356976579761e-05, "loss": 0.013, "step": 17430 }, { "grad_norm": 0.20480459928512573, "learning_rate": 9.58290888099006e-05, "loss": 0.0112, "step": 17440 }, { "grad_norm": 0.14445088803768158, "learning_rate": 9.582247495002486e-05, "loss": 0.0109, "step": 17450 }, { "grad_norm": 0.16061285138130188, "learning_rate": 9.581585607907214e-05, "loss": 0.0108, "step": 17460 }, { "grad_norm": 0.20844657719135284, "learning_rate": 9.580923219776628e-05, "loss": 0.0131, "step": 17470 }, { "grad_norm": 0.19231542944908142, "learning_rate": 9.580260330683167e-05, "loss": 0.0129, "step": 17480 }, { "grad_norm": 0.20752599835395813, "learning_rate": 9.579596940699322e-05, "loss": 0.0137, "step": 17490 }, { "grad_norm": 0.16795960068702698, "learning_rate": 9.578933049897643e-05, "loss": 0.0135, "step": 17500 }, { "grad_norm": 0.174673393368721, "learning_rate": 9.578268658350728e-05, "loss": 0.0115, "step": 17510 }, { "grad_norm": 0.21259771287441254, "learning_rate": 9.577603766131235e-05, "loss": 0.0114, "step": 17520 }, { "grad_norm": 0.12957175076007843, "learning_rate": 9.576938373311878e-05, "loss": 0.0113, "step": 17530 }, { "grad_norm": 0.16701385378837585, "learning_rate": 9.576272479965421e-05, "loss": 0.0101, "step": 17540 }, { "grad_norm": 0.21828460693359375, "learning_rate": 9.575606086164687e-05, "loss": 0.0102, "step": 17550 }, { "grad_norm": 0.1395043134689331, "learning_rate": 9.57493919198255e-05, "loss": 0.0102, "step": 17560 }, { "grad_norm": 0.16974198818206787, "learning_rate": 9.57427179749194e-05, "loss": 0.0082, "step": 17570 }, { "grad_norm": 0.18131670355796814, "learning_rate": 9.573603902765846e-05, "loss": 0.0109, "step": 17580 }, { "grad_norm": 0.22254151105880737, "learning_rate": 9.572935507877304e-05, "loss": 0.0094, "step": 17590 }, { "grad_norm": 0.17811879515647888, "learning_rate": 9.57226661289941e-05, "loss": 0.0098, "step": 17600 }, { "grad_norm": 0.1860174834728241, "learning_rate": 9.571597217905315e-05, "loss": 0.0114, "step": 17610 }, { "grad_norm": 0.17837651073932648, "learning_rate": 9.57092732296822e-05, "loss": 0.0114, "step": 17620 }, { "grad_norm": 0.14318256080150604, "learning_rate": 9.570256928161385e-05, "loss": 0.0094, "step": 17630 }, { "grad_norm": 0.16155192255973816, "learning_rate": 9.569586033558126e-05, "loss": 0.0107, "step": 17640 }, { "grad_norm": 0.1698993742465973, "learning_rate": 9.568914639231807e-05, "loss": 0.0102, "step": 17650 }, { "grad_norm": 0.22866611182689667, "learning_rate": 9.568242745255852e-05, "loss": 0.0095, "step": 17660 }, { "grad_norm": 0.1355152726173401, "learning_rate": 9.567570351703739e-05, "loss": 0.0104, "step": 17670 }, { "grad_norm": 0.17662350833415985, "learning_rate": 9.566897458649001e-05, "loss": 0.01, "step": 17680 }, { "grad_norm": 0.19381467998027802, "learning_rate": 9.566224066165221e-05, "loss": 0.011, "step": 17690 }, { "grad_norm": 0.1754765510559082, "learning_rate": 9.565550174326043e-05, "loss": 0.0098, "step": 17700 }, { "grad_norm": 0.15657131373882294, "learning_rate": 9.564875783205162e-05, "loss": 0.0102, "step": 17710 }, { "grad_norm": 0.20131660997867584, "learning_rate": 9.564200892876328e-05, "loss": 0.0103, "step": 17720 }, { "grad_norm": 0.1544792205095291, "learning_rate": 9.563525503413348e-05, "loss": 0.01, "step": 17730 }, { "grad_norm": 0.1629495918750763, "learning_rate": 9.562849614890079e-05, "loss": 0.0099, "step": 17740 }, { "grad_norm": 0.17180415987968445, "learning_rate": 9.562173227380436e-05, "loss": 0.0101, "step": 17750 }, { "grad_norm": 0.22659704089164734, "learning_rate": 9.561496340958389e-05, "loss": 0.0107, "step": 17760 }, { "grad_norm": 0.1552792638540268, "learning_rate": 9.560818955697959e-05, "loss": 0.0104, "step": 17770 }, { "grad_norm": 0.19508251547813416, "learning_rate": 9.560141071673228e-05, "loss": 0.0106, "step": 17780 }, { "grad_norm": 0.1960969865322113, "learning_rate": 9.559462688958323e-05, "loss": 0.01, "step": 17790 }, { "grad_norm": 0.13830940425395966, "learning_rate": 9.558783807627434e-05, "loss": 0.0098, "step": 17800 }, { "grad_norm": 0.24553833901882172, "learning_rate": 9.558104427754801e-05, "loss": 0.0097, "step": 17810 }, { "grad_norm": 0.1213269829750061, "learning_rate": 9.557424549414722e-05, "loss": 0.0089, "step": 17820 }, { "grad_norm": 0.1791766881942749, "learning_rate": 9.556744172681546e-05, "loss": 0.0098, "step": 17830 }, { "grad_norm": 0.13641193509101868, "learning_rate": 9.556063297629677e-05, "loss": 0.0094, "step": 17840 }, { "grad_norm": 0.22744251787662506, "learning_rate": 9.555381924333578e-05, "loss": 0.0099, "step": 17850 }, { "grad_norm": 0.17527607083320618, "learning_rate": 9.554700052867758e-05, "loss": 0.0107, "step": 17860 }, { "grad_norm": 0.1670922487974167, "learning_rate": 9.554017683306789e-05, "loss": 0.0092, "step": 17870 }, { "grad_norm": 0.1404394805431366, "learning_rate": 9.553334815725294e-05, "loss": 0.0094, "step": 17880 }, { "grad_norm": 0.1387302279472351, "learning_rate": 9.552651450197949e-05, "loss": 0.0091, "step": 17890 }, { "grad_norm": 0.14957861602306366, "learning_rate": 9.551967586799486e-05, "loss": 0.0088, "step": 17900 }, { "grad_norm": 0.1867150068283081, "learning_rate": 9.551283225604692e-05, "loss": 0.0102, "step": 17910 }, { "grad_norm": 0.13435617089271545, "learning_rate": 9.550598366688406e-05, "loss": 0.0091, "step": 17920 }, { "grad_norm": 0.12972667813301086, "learning_rate": 9.549913010125526e-05, "loss": 0.0089, "step": 17930 }, { "grad_norm": 0.19569513201713562, "learning_rate": 9.549227155990999e-05, "loss": 0.0088, "step": 17940 }, { "grad_norm": 0.17173153162002563, "learning_rate": 9.548540804359828e-05, "loss": 0.0093, "step": 17950 }, { "grad_norm": 0.15054760873317719, "learning_rate": 9.547853955307077e-05, "loss": 0.0104, "step": 17960 }, { "grad_norm": 0.2109326720237732, "learning_rate": 9.547166608907853e-05, "loss": 0.0099, "step": 17970 }, { "grad_norm": 0.1606817990541458, "learning_rate": 9.546478765237326e-05, "loss": 0.0094, "step": 17980 }, { "grad_norm": 0.2118864357471466, "learning_rate": 9.545790424370715e-05, "loss": 0.0117, "step": 17990 }, { "grad_norm": 0.19261671602725983, "learning_rate": 9.5451015863833e-05, "loss": 0.0114, "step": 18000 }, { "grad_norm": 0.22829847037792206, "learning_rate": 9.544412251350408e-05, "loss": 0.0093, "step": 18010 }, { "grad_norm": 0.1440560221672058, "learning_rate": 9.543722419347422e-05, "loss": 0.0094, "step": 18020 }, { "grad_norm": 0.1524229645729065, "learning_rate": 9.543032090449788e-05, "loss": 0.0102, "step": 18030 }, { "grad_norm": 0.1614827811717987, "learning_rate": 9.542341264732992e-05, "loss": 0.0097, "step": 18040 }, { "grad_norm": 0.1570865660905838, "learning_rate": 9.541649942272585e-05, "loss": 0.0096, "step": 18050 }, { "grad_norm": 0.16349926590919495, "learning_rate": 9.54095812314417e-05, "loss": 0.0096, "step": 18060 }, { "grad_norm": 0.19771401584148407, "learning_rate": 9.540265807423401e-05, "loss": 0.0093, "step": 18070 }, { "grad_norm": 0.20345675945281982, "learning_rate": 9.53957299518599e-05, "loss": 0.0092, "step": 18080 }, { "grad_norm": 0.1503521203994751, "learning_rate": 9.5388796865077e-05, "loss": 0.01, "step": 18090 }, { "grad_norm": 0.20039938390254974, "learning_rate": 9.538185881464353e-05, "loss": 0.01, "step": 18100 }, { "grad_norm": 0.20401495695114136, "learning_rate": 9.537491580131821e-05, "loss": 0.0097, "step": 18110 }, { "grad_norm": 0.1688227504491806, "learning_rate": 9.53679678258603e-05, "loss": 0.0111, "step": 18120 }, { "grad_norm": 0.21751724183559418, "learning_rate": 9.536101488902966e-05, "loss": 0.0097, "step": 18130 }, { "grad_norm": 0.18390890955924988, "learning_rate": 9.535405699158663e-05, "loss": 0.0112, "step": 18140 }, { "grad_norm": 0.1684410125017166, "learning_rate": 9.53470941342921e-05, "loss": 0.0105, "step": 18150 }, { "grad_norm": 0.16323792934417725, "learning_rate": 9.534012631790756e-05, "loss": 0.0097, "step": 18160 }, { "grad_norm": 0.24965029954910278, "learning_rate": 9.533315354319494e-05, "loss": 0.0104, "step": 18170 }, { "grad_norm": 0.19208869338035583, "learning_rate": 9.532617581091682e-05, "loss": 0.0104, "step": 18180 }, { "grad_norm": 0.2299148440361023, "learning_rate": 9.531919312183629e-05, "loss": 0.0119, "step": 18190 }, { "grad_norm": 0.14616616070270538, "learning_rate": 9.531220547671688e-05, "loss": 0.0122, "step": 18200 }, { "grad_norm": 0.16412314772605896, "learning_rate": 9.530521287632285e-05, "loss": 0.0096, "step": 18210 }, { "grad_norm": 0.18200397491455078, "learning_rate": 9.529821532141884e-05, "loss": 0.0093, "step": 18220 }, { "grad_norm": 0.25551965832710266, "learning_rate": 9.52912128127701e-05, "loss": 0.0104, "step": 18230 }, { "grad_norm": 0.1320711076259613, "learning_rate": 9.528420535114244e-05, "loss": 0.0092, "step": 18240 }, { "grad_norm": 0.19634662568569183, "learning_rate": 9.527719293730215e-05, "loss": 0.0105, "step": 18250 }, { "grad_norm": 0.1791931539773941, "learning_rate": 9.527017557201611e-05, "loss": 0.0088, "step": 18260 }, { "grad_norm": 0.15491622686386108, "learning_rate": 9.526315325605176e-05, "loss": 0.0097, "step": 18270 }, { "grad_norm": 0.19133855402469635, "learning_rate": 9.525612599017699e-05, "loss": 0.0097, "step": 18280 }, { "grad_norm": 0.19605523347854614, "learning_rate": 9.524909377516033e-05, "loss": 0.0097, "step": 18290 }, { "grad_norm": 0.18383873999118805, "learning_rate": 9.524205661177081e-05, "loss": 0.0097, "step": 18300 }, { "grad_norm": 0.16632068157196045, "learning_rate": 9.523501450077801e-05, "loss": 0.0098, "step": 18310 }, { "grad_norm": 0.2650834321975708, "learning_rate": 9.522796744295202e-05, "loss": 0.0129, "step": 18320 }, { "grad_norm": 0.32055070996284485, "learning_rate": 9.522091543906352e-05, "loss": 0.0103, "step": 18330 }, { "grad_norm": 0.17877161502838135, "learning_rate": 9.521385848988369e-05, "loss": 0.0097, "step": 18340 }, { "grad_norm": 0.16583648324012756, "learning_rate": 9.520679659618428e-05, "loss": 0.0097, "step": 18350 }, { "grad_norm": 0.1570948362350464, "learning_rate": 9.519972975873754e-05, "loss": 0.01, "step": 18360 }, { "grad_norm": 0.1891234964132309, "learning_rate": 9.519265797831633e-05, "loss": 0.0105, "step": 18370 }, { "grad_norm": 0.21736666560173035, "learning_rate": 9.518558125569399e-05, "loss": 0.0102, "step": 18380 }, { "grad_norm": 0.14207126200199127, "learning_rate": 9.517849959164442e-05, "loss": 0.011, "step": 18390 }, { "grad_norm": 0.1910778284072876, "learning_rate": 9.517141298694205e-05, "loss": 0.0114, "step": 18400 }, { "grad_norm": 0.19590921700000763, "learning_rate": 9.516432144236188e-05, "loss": 0.0146, "step": 18410 }, { "grad_norm": 0.2700670063495636, "learning_rate": 9.515722495867941e-05, "loss": 0.0133, "step": 18420 }, { "grad_norm": 0.22560973465442657, "learning_rate": 9.515012353667072e-05, "loss": 0.0126, "step": 18430 }, { "grad_norm": 0.2950184643268585, "learning_rate": 9.51430171771124e-05, "loss": 0.0119, "step": 18440 }, { "grad_norm": 0.16549274325370789, "learning_rate": 9.513590588078159e-05, "loss": 0.0133, "step": 18450 }, { "grad_norm": 0.1888149529695511, "learning_rate": 9.512878964845597e-05, "loss": 0.0123, "step": 18460 }, { "grad_norm": 0.18088245391845703, "learning_rate": 9.512166848091377e-05, "loss": 0.0113, "step": 18470 }, { "grad_norm": 0.19750791788101196, "learning_rate": 9.511454237893376e-05, "loss": 0.0111, "step": 18480 }, { "grad_norm": 0.1741161197423935, "learning_rate": 9.51074113432952e-05, "loss": 0.0105, "step": 18490 }, { "grad_norm": 0.13691462576389313, "learning_rate": 9.510027537477797e-05, "loss": 0.0095, "step": 18500 }, { "grad_norm": 0.14130686223506927, "learning_rate": 9.509313447416242e-05, "loss": 0.0092, "step": 18510 }, { "grad_norm": 0.14991557598114014, "learning_rate": 9.508598864222949e-05, "loss": 0.011, "step": 18520 }, { "grad_norm": 0.1949160248041153, "learning_rate": 9.507883787976062e-05, "loss": 0.011, "step": 18530 }, { "grad_norm": 0.20636270940303802, "learning_rate": 9.507168218753781e-05, "loss": 0.0115, "step": 18540 }, { "grad_norm": 0.23547756671905518, "learning_rate": 9.506452156634362e-05, "loss": 0.0099, "step": 18550 }, { "grad_norm": 0.20602673292160034, "learning_rate": 9.505735601696109e-05, "loss": 0.0108, "step": 18560 }, { "grad_norm": 0.26373860239982605, "learning_rate": 9.505018554017385e-05, "loss": 0.0104, "step": 18570 }, { "grad_norm": 0.15175524353981018, "learning_rate": 9.504301013676604e-05, "loss": 0.0098, "step": 18580 }, { "grad_norm": 0.1535380780696869, "learning_rate": 9.503582980752238e-05, "loss": 0.0099, "step": 18590 }, { "grad_norm": 0.15074928104877472, "learning_rate": 9.502864455322809e-05, "loss": 0.0116, "step": 18600 }, { "grad_norm": 0.16166898608207703, "learning_rate": 9.502145437466891e-05, "loss": 0.0105, "step": 18610 }, { "grad_norm": 0.1436203271150589, "learning_rate": 9.501425927263116e-05, "loss": 0.0107, "step": 18620 }, { "grad_norm": 0.14462606608867645, "learning_rate": 9.500705924790172e-05, "loss": 0.011, "step": 18630 }, { "grad_norm": 0.17848186194896698, "learning_rate": 9.499985430126794e-05, "loss": 0.0109, "step": 18640 }, { "grad_norm": 0.16061215102672577, "learning_rate": 9.499264443351775e-05, "loss": 0.0111, "step": 18650 }, { "grad_norm": 0.17781542241573334, "learning_rate": 9.498542964543961e-05, "loss": 0.0097, "step": 18660 }, { "grad_norm": 0.13510878384113312, "learning_rate": 9.497820993782252e-05, "loss": 0.0099, "step": 18670 }, { "grad_norm": 0.1749648153781891, "learning_rate": 9.497098531145601e-05, "loss": 0.0094, "step": 18680 }, { "grad_norm": 0.16859044134616852, "learning_rate": 9.496375576713017e-05, "loss": 0.0101, "step": 18690 }, { "grad_norm": 0.2143586128950119, "learning_rate": 9.49565213056356e-05, "loss": 0.0102, "step": 18700 }, { "grad_norm": 0.2238079160451889, "learning_rate": 9.494928192776342e-05, "loss": 0.0104, "step": 18710 }, { "grad_norm": 0.17613276839256287, "learning_rate": 9.494203763430538e-05, "loss": 0.0093, "step": 18720 }, { "grad_norm": 0.1438503861427307, "learning_rate": 9.493478842605366e-05, "loss": 0.0112, "step": 18730 }, { "grad_norm": 0.17679812014102936, "learning_rate": 9.492753430380105e-05, "loss": 0.0087, "step": 18740 }, { "grad_norm": 0.1670365035533905, "learning_rate": 9.492027526834083e-05, "loss": 0.0096, "step": 18750 }, { "grad_norm": 0.26633238792419434, "learning_rate": 9.491301132046684e-05, "loss": 0.012, "step": 18760 }, { "grad_norm": 0.16485227644443512, "learning_rate": 9.490574246097345e-05, "loss": 0.0107, "step": 18770 }, { "grad_norm": 0.19716720283031464, "learning_rate": 9.48984686906556e-05, "loss": 0.009, "step": 18780 }, { "grad_norm": 0.119265116751194, "learning_rate": 9.489119001030871e-05, "loss": 0.0094, "step": 18790 }, { "grad_norm": 0.18573243916034698, "learning_rate": 9.488390642072878e-05, "loss": 0.0114, "step": 18800 }, { "grad_norm": 0.18593639135360718, "learning_rate": 9.48766179227123e-05, "loss": 0.0095, "step": 18810 }, { "grad_norm": 0.16165070235729218, "learning_rate": 9.486932451705636e-05, "loss": 0.0104, "step": 18820 }, { "grad_norm": 0.1620328724384308, "learning_rate": 9.486202620455857e-05, "loss": 0.0097, "step": 18830 }, { "grad_norm": 0.14284709095954895, "learning_rate": 9.485472298601704e-05, "loss": 0.008, "step": 18840 }, { "grad_norm": 0.15886665880680084, "learning_rate": 9.484741486223043e-05, "loss": 0.0093, "step": 18850 }, { "grad_norm": 0.15933750569820404, "learning_rate": 9.484010183399797e-05, "loss": 0.0092, "step": 18860 }, { "grad_norm": 0.2678402066230774, "learning_rate": 9.483278390211938e-05, "loss": 0.0106, "step": 18870 }, { "grad_norm": 0.18660908937454224, "learning_rate": 9.482546106739496e-05, "loss": 0.0089, "step": 18880 }, { "grad_norm": 0.20102353394031525, "learning_rate": 9.48181333306255e-05, "loss": 0.0093, "step": 18890 }, { "grad_norm": 0.21325315535068512, "learning_rate": 9.481080069261237e-05, "loss": 0.0092, "step": 18900 }, { "grad_norm": 0.19465576112270355, "learning_rate": 9.480346315415745e-05, "loss": 0.0094, "step": 18910 }, { "grad_norm": 0.19744229316711426, "learning_rate": 9.479612071606314e-05, "loss": 0.0103, "step": 18920 }, { "grad_norm": 0.21225108206272125, "learning_rate": 9.478877337913244e-05, "loss": 0.0098, "step": 18930 }, { "grad_norm": 0.1969701200723648, "learning_rate": 9.478142114416881e-05, "loss": 0.0113, "step": 18940 }, { "grad_norm": 0.13553233444690704, "learning_rate": 9.47740640119763e-05, "loss": 0.0088, "step": 18950 }, { "grad_norm": 0.15718425810337067, "learning_rate": 9.476670198335947e-05, "loss": 0.0103, "step": 18960 }, { "grad_norm": 0.17393910884857178, "learning_rate": 9.47593350591234e-05, "loss": 0.0101, "step": 18970 }, { "grad_norm": 0.17955558001995087, "learning_rate": 9.475196324007376e-05, "loss": 0.0096, "step": 18980 }, { "grad_norm": 0.17178526520729065, "learning_rate": 9.474458652701669e-05, "loss": 0.01, "step": 18990 }, { "grad_norm": 0.19269959628582, "learning_rate": 9.473720492075892e-05, "loss": 0.009, "step": 19000 }, { "grad_norm": 0.16335366666316986, "learning_rate": 9.472981842210768e-05, "loss": 0.0079, "step": 19010 }, { "grad_norm": 0.11290016025304794, "learning_rate": 9.472242703187074e-05, "loss": 0.0092, "step": 19020 }, { "grad_norm": 0.1895931214094162, "learning_rate": 9.471503075085643e-05, "loss": 0.0097, "step": 19030 }, { "grad_norm": 0.15840895473957062, "learning_rate": 9.470762957987359e-05, "loss": 0.0102, "step": 19040 }, { "grad_norm": 0.14669564366340637, "learning_rate": 9.470022351973158e-05, "loss": 0.0086, "step": 19050 }, { "grad_norm": 0.12486725300550461, "learning_rate": 9.469281257124034e-05, "loss": 0.0088, "step": 19060 }, { "grad_norm": 0.14166709780693054, "learning_rate": 9.46853967352103e-05, "loss": 0.0098, "step": 19070 }, { "grad_norm": 0.1858864575624466, "learning_rate": 9.467797601245246e-05, "loss": 0.0107, "step": 19080 }, { "grad_norm": 0.14585264027118683, "learning_rate": 9.467055040377834e-05, "loss": 0.0088, "step": 19090 }, { "grad_norm": 0.13434240221977234, "learning_rate": 9.466311990999999e-05, "loss": 0.0105, "step": 19100 }, { "grad_norm": 0.19618697464466095, "learning_rate": 9.465568453193e-05, "loss": 0.01, "step": 19110 }, { "grad_norm": 0.18251225352287292, "learning_rate": 9.464824427038148e-05, "loss": 0.0091, "step": 19120 }, { "grad_norm": 0.16379736363887787, "learning_rate": 9.46407991261681e-05, "loss": 0.0091, "step": 19130 }, { "grad_norm": 0.18653489649295807, "learning_rate": 9.463334910010404e-05, "loss": 0.0099, "step": 19140 }, { "grad_norm": 0.22089675068855286, "learning_rate": 9.462589419300403e-05, "loss": 0.0109, "step": 19150 }, { "grad_norm": 0.20906281471252441, "learning_rate": 9.461843440568333e-05, "loss": 0.0109, "step": 19160 }, { "grad_norm": 0.22351329028606415, "learning_rate": 9.461096973895773e-05, "loss": 0.0113, "step": 19170 }, { "grad_norm": 0.17052225768566132, "learning_rate": 9.460350019364355e-05, "loss": 0.0107, "step": 19180 }, { "grad_norm": 0.18446506559848785, "learning_rate": 9.459602577055764e-05, "loss": 0.0088, "step": 19190 }, { "grad_norm": 0.1566491425037384, "learning_rate": 9.45885464705174e-05, "loss": 0.0112, "step": 19200 }, { "grad_norm": 0.17418654263019562, "learning_rate": 9.458106229434076e-05, "loss": 0.0094, "step": 19210 }, { "grad_norm": 0.22951753437519073, "learning_rate": 9.457357324284617e-05, "loss": 0.0102, "step": 19220 }, { "grad_norm": 0.15011373162269592, "learning_rate": 9.456607931685262e-05, "loss": 0.011, "step": 19230 }, { "grad_norm": 0.15097928047180176, "learning_rate": 9.455858051717965e-05, "loss": 0.0093, "step": 19240 }, { "grad_norm": 0.14453156292438507, "learning_rate": 9.45510768446473e-05, "loss": 0.0098, "step": 19250 }, { "grad_norm": 0.18378205597400665, "learning_rate": 9.454356830007618e-05, "loss": 0.0111, "step": 19260 }, { "grad_norm": 0.24509593844413757, "learning_rate": 9.45360548842874e-05, "loss": 0.0111, "step": 19270 }, { "grad_norm": 0.17619983851909637, "learning_rate": 9.452853659810261e-05, "loss": 0.0107, "step": 19280 }, { "grad_norm": 0.13976174592971802, "learning_rate": 9.452101344234401e-05, "loss": 0.0097, "step": 19290 }, { "grad_norm": 0.1642160564661026, "learning_rate": 9.451348541783431e-05, "loss": 0.0109, "step": 19300 }, { "grad_norm": 0.1877174973487854, "learning_rate": 9.450595252539678e-05, "loss": 0.0103, "step": 19310 }, { "grad_norm": 0.19971206784248352, "learning_rate": 9.449841476585518e-05, "loss": 0.0103, "step": 19320 }, { "grad_norm": 0.17967082560062408, "learning_rate": 9.449087214003384e-05, "loss": 0.0121, "step": 19330 }, { "grad_norm": 0.14415095746517181, "learning_rate": 9.448332464875765e-05, "loss": 0.0123, "step": 19340 }, { "grad_norm": 0.14014208316802979, "learning_rate": 9.447577229285192e-05, "loss": 0.0115, "step": 19350 }, { "grad_norm": 0.24443596601486206, "learning_rate": 9.446821507314261e-05, "loss": 0.0097, "step": 19360 }, { "grad_norm": 0.22829031944274902, "learning_rate": 9.446065299045617e-05, "loss": 0.0119, "step": 19370 }, { "grad_norm": 0.21552030742168427, "learning_rate": 9.445308604561955e-05, "loss": 0.0109, "step": 19380 }, { "grad_norm": 0.18232518434524536, "learning_rate": 9.444551423946028e-05, "loss": 0.0099, "step": 19390 }, { "grad_norm": 0.1479436159133911, "learning_rate": 9.443793757280638e-05, "loss": 0.0091, "step": 19400 }, { "grad_norm": 0.15661188960075378, "learning_rate": 9.443035604648646e-05, "loss": 0.0097, "step": 19410 }, { "grad_norm": 0.16862265765666962, "learning_rate": 9.44227696613296e-05, "loss": 0.0107, "step": 19420 }, { "grad_norm": 0.17634165287017822, "learning_rate": 9.441517841816542e-05, "loss": 0.0094, "step": 19430 }, { "grad_norm": 0.1435561329126358, "learning_rate": 9.440758231782413e-05, "loss": 0.0115, "step": 19440 }, { "grad_norm": 0.15840230882167816, "learning_rate": 9.439998136113639e-05, "loss": 0.0093, "step": 19450 }, { "grad_norm": 0.1913565993309021, "learning_rate": 9.439237554893344e-05, "loss": 0.009, "step": 19460 }, { "grad_norm": 0.16982191801071167, "learning_rate": 9.438476488204705e-05, "loss": 0.0089, "step": 19470 }, { "grad_norm": 0.16392023861408234, "learning_rate": 9.43771493613095e-05, "loss": 0.0087, "step": 19480 }, { "grad_norm": 0.1784169226884842, "learning_rate": 9.436952898755362e-05, "loss": 0.009, "step": 19490 }, { "grad_norm": 0.15978828072547913, "learning_rate": 9.436190376161276e-05, "loss": 0.0112, "step": 19500 }, { "grad_norm": 0.16075000166893005, "learning_rate": 9.43542736843208e-05, "loss": 0.0094, "step": 19510 }, { "grad_norm": 0.1785104125738144, "learning_rate": 9.434663875651216e-05, "loss": 0.0094, "step": 19520 }, { "grad_norm": 0.15932713449001312, "learning_rate": 9.433899897902177e-05, "loss": 0.0095, "step": 19530 }, { "grad_norm": 0.15300703048706055, "learning_rate": 9.433135435268511e-05, "loss": 0.0086, "step": 19540 }, { "grad_norm": 0.19616608321666718, "learning_rate": 9.432370487833819e-05, "loss": 0.009, "step": 19550 }, { "grad_norm": 0.16888833045959473, "learning_rate": 9.431605055681756e-05, "loss": 0.0092, "step": 19560 }, { "grad_norm": 0.1753423810005188, "learning_rate": 9.430839138896026e-05, "loss": 0.0092, "step": 19570 }, { "grad_norm": 0.21287831664085388, "learning_rate": 9.43007273756039e-05, "loss": 0.0092, "step": 19580 }, { "grad_norm": 0.14594799280166626, "learning_rate": 9.429305851758658e-05, "loss": 0.0104, "step": 19590 }, { "grad_norm": 0.16358442604541779, "learning_rate": 9.428538481574699e-05, "loss": 0.0091, "step": 19600 }, { "grad_norm": 0.12878918647766113, "learning_rate": 9.42777062709243e-05, "loss": 0.0106, "step": 19610 }, { "grad_norm": 0.18954183161258698, "learning_rate": 9.427002288395821e-05, "loss": 0.009, "step": 19620 }, { "grad_norm": 0.21322575211524963, "learning_rate": 9.426233465568898e-05, "loss": 0.0101, "step": 19630 }, { "grad_norm": 0.14909735321998596, "learning_rate": 9.42546415869574e-05, "loss": 0.0092, "step": 19640 }, { "grad_norm": 0.13134635984897614, "learning_rate": 9.424694367860473e-05, "loss": 0.0078, "step": 19650 }, { "grad_norm": 0.12421927601099014, "learning_rate": 9.423924093147284e-05, "loss": 0.0114, "step": 19660 }, { "grad_norm": 0.1706709861755371, "learning_rate": 9.423153334640407e-05, "loss": 0.0099, "step": 19670 }, { "grad_norm": 0.15384641289710999, "learning_rate": 9.42238209242413e-05, "loss": 0.0104, "step": 19680 }, { "grad_norm": 0.149454727768898, "learning_rate": 9.421610366582798e-05, "loss": 0.0108, "step": 19690 }, { "grad_norm": 0.11142045259475708, "learning_rate": 9.420838157200803e-05, "loss": 0.0097, "step": 19700 }, { "grad_norm": 0.1510130912065506, "learning_rate": 9.420065464362594e-05, "loss": 0.0104, "step": 19710 }, { "grad_norm": 0.12997452914714813, "learning_rate": 9.419292288152673e-05, "loss": 0.008, "step": 19720 }, { "grad_norm": 0.14194519817829132, "learning_rate": 9.418518628655588e-05, "loss": 0.0097, "step": 19730 }, { "grad_norm": 0.18236598372459412, "learning_rate": 9.417744485955951e-05, "loss": 0.0086, "step": 19740 }, { "grad_norm": 0.15847846865653992, "learning_rate": 9.41696986013842e-05, "loss": 0.0077, "step": 19750 }, { "grad_norm": 0.1708521544933319, "learning_rate": 9.416194751287705e-05, "loss": 0.0089, "step": 19760 }, { "grad_norm": 0.1592836081981659, "learning_rate": 9.415419159488572e-05, "loss": 0.0095, "step": 19770 }, { "grad_norm": 0.1811426430940628, "learning_rate": 9.414643084825837e-05, "loss": 0.0085, "step": 19780 }, { "grad_norm": 0.16485364735126495, "learning_rate": 9.413866527384372e-05, "loss": 0.0102, "step": 19790 }, { "grad_norm": 0.2318023443222046, "learning_rate": 9.4130894872491e-05, "loss": 0.009, "step": 19800 }, { "grad_norm": 0.16261111199855804, "learning_rate": 9.412311964504998e-05, "loss": 0.0102, "step": 19810 }, { "grad_norm": 0.13489551842212677, "learning_rate": 9.411533959237091e-05, "loss": 0.009, "step": 19820 }, { "grad_norm": 0.19034025073051453, "learning_rate": 9.410755471530464e-05, "loss": 0.0106, "step": 19830 }, { "grad_norm": 0.18840956687927246, "learning_rate": 9.40997650147025e-05, "loss": 0.0094, "step": 19840 }, { "grad_norm": 0.19710715115070343, "learning_rate": 9.409197049141637e-05, "loss": 0.0126, "step": 19850 }, { "grad_norm": 0.15210489928722382, "learning_rate": 9.408417114629863e-05, "loss": 0.0105, "step": 19860 }, { "grad_norm": 0.18500924110412598, "learning_rate": 9.40763669802022e-05, "loss": 0.0103, "step": 19870 }, { "grad_norm": 0.19751158356666565, "learning_rate": 9.406855799398056e-05, "loss": 0.0095, "step": 19880 }, { "grad_norm": 0.1815962940454483, "learning_rate": 9.406074418848767e-05, "loss": 0.0106, "step": 19890 }, { "grad_norm": 0.17778117954730988, "learning_rate": 9.405292556457805e-05, "loss": 0.0103, "step": 19900 }, { "grad_norm": 0.12664110958576202, "learning_rate": 9.404510212310671e-05, "loss": 0.0107, "step": 19910 }, { "grad_norm": 0.1691967397928238, "learning_rate": 9.403727386492924e-05, "loss": 0.0093, "step": 19920 }, { "grad_norm": 0.17190593481063843, "learning_rate": 9.40294407909017e-05, "loss": 0.0091, "step": 19930 }, { "grad_norm": 0.1564832329750061, "learning_rate": 9.40216029018807e-05, "loss": 0.0102, "step": 19940 }, { "grad_norm": 0.23100435733795166, "learning_rate": 9.401376019872338e-05, "loss": 0.0097, "step": 19950 }, { "grad_norm": 0.196359783411026, "learning_rate": 9.400591268228746e-05, "loss": 0.0095, "step": 19960 }, { "grad_norm": 0.16099031269550323, "learning_rate": 9.399806035343106e-05, "loss": 0.0086, "step": 19970 }, { "grad_norm": 0.13398101925849915, "learning_rate": 9.399020321301294e-05, "loss": 0.0085, "step": 19980 }, { "grad_norm": 0.20347674190998077, "learning_rate": 9.398234126189234e-05, "loss": 0.0098, "step": 19990 }, { "grad_norm": 0.2523788809776306, "learning_rate": 9.397447450092902e-05, "loss": 0.0109, "step": 20000 }, { "grad_norm": 0.22429604828357697, "learning_rate": 9.39666029309833e-05, "loss": 0.0107, "step": 20010 }, { "grad_norm": 0.21326661109924316, "learning_rate": 9.395872655291596e-05, "loss": 0.0117, "step": 20020 }, { "grad_norm": 0.20775826275348663, "learning_rate": 9.395084536758838e-05, "loss": 0.0106, "step": 20030 }, { "grad_norm": 0.25186416506767273, "learning_rate": 9.394295937586243e-05, "loss": 0.014, "step": 20040 }, { "grad_norm": 0.2591722011566162, "learning_rate": 9.393506857860052e-05, "loss": 0.0121, "step": 20050 }, { "grad_norm": 0.16492199897766113, "learning_rate": 9.392717297666555e-05, "loss": 0.0108, "step": 20060 }, { "grad_norm": 0.24875392019748688, "learning_rate": 9.391927257092101e-05, "loss": 0.0114, "step": 20070 }, { "grad_norm": 0.15318168699741364, "learning_rate": 9.391136736223085e-05, "loss": 0.0113, "step": 20080 }, { "grad_norm": 0.19688403606414795, "learning_rate": 9.390345735145956e-05, "loss": 0.0107, "step": 20090 }, { "grad_norm": 0.2512843906879425, "learning_rate": 9.389554253947219e-05, "loss": 0.0106, "step": 20100 }, { "grad_norm": 0.20893898606300354, "learning_rate": 9.388762292713428e-05, "loss": 0.0118, "step": 20110 }, { "grad_norm": 0.18091663718223572, "learning_rate": 9.38796985153119e-05, "loss": 0.0116, "step": 20120 }, { "grad_norm": 0.19070163369178772, "learning_rate": 9.387176930487169e-05, "loss": 0.012, "step": 20130 }, { "grad_norm": 0.25302889943122864, "learning_rate": 9.386383529668072e-05, "loss": 0.0102, "step": 20140 }, { "grad_norm": 0.20228581130504608, "learning_rate": 9.385589649160669e-05, "loss": 0.0123, "step": 20150 }, { "grad_norm": 0.19780932366847992, "learning_rate": 9.384795289051775e-05, "loss": 0.0105, "step": 20160 }, { "grad_norm": 0.2206864356994629, "learning_rate": 9.384000449428261e-05, "loss": 0.0099, "step": 20170 }, { "grad_norm": 0.15768930315971375, "learning_rate": 9.383205130377048e-05, "loss": 0.0104, "step": 20180 }, { "grad_norm": 0.1593555063009262, "learning_rate": 9.382409331985114e-05, "loss": 0.0106, "step": 20190 }, { "grad_norm": 0.18868689239025116, "learning_rate": 9.381613054339482e-05, "loss": 0.0108, "step": 20200 }, { "grad_norm": 0.1489286571741104, "learning_rate": 9.380816297527235e-05, "loss": 0.0106, "step": 20210 }, { "grad_norm": 0.21533000469207764, "learning_rate": 9.380019061635506e-05, "loss": 0.0112, "step": 20220 }, { "grad_norm": 0.14731520414352417, "learning_rate": 9.379221346751474e-05, "loss": 0.0087, "step": 20230 }, { "grad_norm": 0.22245115041732788, "learning_rate": 9.378423152962382e-05, "loss": 0.0089, "step": 20240 }, { "grad_norm": 0.2350260317325592, "learning_rate": 9.377624480355517e-05, "loss": 0.0097, "step": 20250 }, { "grad_norm": 0.19475798308849335, "learning_rate": 9.376825329018219e-05, "loss": 0.0092, "step": 20260 }, { "grad_norm": 0.200778990983963, "learning_rate": 9.376025699037884e-05, "loss": 0.0096, "step": 20270 }, { "grad_norm": 0.13431207835674286, "learning_rate": 9.37522559050196e-05, "loss": 0.0096, "step": 20280 }, { "grad_norm": 0.18370072543621063, "learning_rate": 9.37442500349794e-05, "loss": 0.01, "step": 20290 }, { "grad_norm": 0.16952632367610931, "learning_rate": 9.373623938113381e-05, "loss": 0.0097, "step": 20300 }, { "grad_norm": 0.162045419216156, "learning_rate": 9.372822394435883e-05, "loss": 0.0118, "step": 20310 }, { "grad_norm": 0.206332266330719, "learning_rate": 9.372020372553102e-05, "loss": 0.0105, "step": 20320 }, { "grad_norm": 0.2361873835325241, "learning_rate": 9.371217872552746e-05, "loss": 0.0103, "step": 20330 }, { "grad_norm": 0.20331986248493195, "learning_rate": 9.370414894522576e-05, "loss": 0.0103, "step": 20340 }, { "grad_norm": 0.14655916392803192, "learning_rate": 9.369611438550406e-05, "loss": 0.0109, "step": 20350 }, { "grad_norm": 0.21974323689937592, "learning_rate": 9.368807504724095e-05, "loss": 0.0131, "step": 20360 }, { "grad_norm": 0.19910161197185516, "learning_rate": 9.368003093131565e-05, "loss": 0.0121, "step": 20370 }, { "grad_norm": 0.1845090389251709, "learning_rate": 9.367198203860785e-05, "loss": 0.0108, "step": 20380 }, { "grad_norm": 0.17811629176139832, "learning_rate": 9.366392836999774e-05, "loss": 0.0108, "step": 20390 }, { "grad_norm": 0.1750430017709732, "learning_rate": 9.365586992636607e-05, "loss": 0.0101, "step": 20400 }, { "grad_norm": 0.1815274953842163, "learning_rate": 9.364780670859412e-05, "loss": 0.0103, "step": 20410 }, { "grad_norm": 0.2179901897907257, "learning_rate": 9.363973871756364e-05, "loss": 0.0111, "step": 20420 }, { "grad_norm": 0.14467275142669678, "learning_rate": 9.363166595415696e-05, "loss": 0.0104, "step": 20430 }, { "grad_norm": 0.15888115763664246, "learning_rate": 9.362358841925686e-05, "loss": 0.0117, "step": 20440 }, { "grad_norm": 0.1375856250524521, "learning_rate": 9.361550611374674e-05, "loss": 0.0102, "step": 20450 }, { "grad_norm": 0.18407678604125977, "learning_rate": 9.360741903851043e-05, "loss": 0.0094, "step": 20460 }, { "grad_norm": 0.18444198369979858, "learning_rate": 9.359932719443236e-05, "loss": 0.0099, "step": 20470 }, { "grad_norm": 0.16331854462623596, "learning_rate": 9.35912305823974e-05, "loss": 0.0106, "step": 20480 }, { "grad_norm": 0.16082921624183655, "learning_rate": 9.358312920329101e-05, "loss": 0.0107, "step": 20490 }, { "grad_norm": 0.17170919477939606, "learning_rate": 9.357502305799914e-05, "loss": 0.0099, "step": 20500 }, { "grad_norm": 0.15338964760303497, "learning_rate": 9.356691214740824e-05, "loss": 0.0092, "step": 20510 }, { "grad_norm": 0.1528092473745346, "learning_rate": 9.355879647240535e-05, "loss": 0.0103, "step": 20520 }, { "grad_norm": 0.2043410837650299, "learning_rate": 9.355067603387798e-05, "loss": 0.01, "step": 20530 }, { "grad_norm": 0.16032710671424866, "learning_rate": 9.354255083271412e-05, "loss": 0.0091, "step": 20540 }, { "grad_norm": 0.20409049093723297, "learning_rate": 9.353442086980239e-05, "loss": 0.0099, "step": 20550 }, { "grad_norm": 0.12796539068222046, "learning_rate": 9.352628614603185e-05, "loss": 0.0122, "step": 20560 }, { "grad_norm": 0.11143138259649277, "learning_rate": 9.351814666229209e-05, "loss": 0.0093, "step": 20570 }, { "grad_norm": 0.15694740414619446, "learning_rate": 9.351000241947324e-05, "loss": 0.0088, "step": 20580 }, { "grad_norm": 0.15202389657497406, "learning_rate": 9.350185341846594e-05, "loss": 0.0101, "step": 20590 }, { "grad_norm": 0.18859243392944336, "learning_rate": 9.349369966016134e-05, "loss": 0.0097, "step": 20600 }, { "grad_norm": 0.29688119888305664, "learning_rate": 9.348554114545117e-05, "loss": 0.0098, "step": 20610 }, { "grad_norm": 0.15279871225357056, "learning_rate": 9.347737787522758e-05, "loss": 0.0101, "step": 20620 }, { "grad_norm": 0.16371332108974457, "learning_rate": 9.346920985038332e-05, "loss": 0.01, "step": 20630 }, { "grad_norm": 0.21854664385318756, "learning_rate": 9.346103707181162e-05, "loss": 0.0108, "step": 20640 }, { "grad_norm": 0.1649589091539383, "learning_rate": 9.345285954040626e-05, "loss": 0.0115, "step": 20650 }, { "grad_norm": 0.19366753101348877, "learning_rate": 9.34446772570615e-05, "loss": 0.0097, "step": 20660 }, { "grad_norm": 0.1717393398284912, "learning_rate": 9.343649022267214e-05, "loss": 0.0107, "step": 20670 }, { "grad_norm": 0.13275302946567535, "learning_rate": 9.342829843813353e-05, "loss": 0.0103, "step": 20680 }, { "grad_norm": 0.1401652991771698, "learning_rate": 9.342010190434149e-05, "loss": 0.0104, "step": 20690 }, { "grad_norm": 0.20857037603855133, "learning_rate": 9.34119006221924e-05, "loss": 0.0115, "step": 20700 }, { "grad_norm": 0.1962824910879135, "learning_rate": 9.340369459258313e-05, "loss": 0.0083, "step": 20710 }, { "grad_norm": 0.1609586626291275, "learning_rate": 9.339548381641106e-05, "loss": 0.0094, "step": 20720 }, { "grad_norm": 0.16859553754329681, "learning_rate": 9.338726829457413e-05, "loss": 0.0089, "step": 20730 }, { "grad_norm": 0.22321061789989471, "learning_rate": 9.337904802797078e-05, "loss": 0.0102, "step": 20740 }, { "grad_norm": 0.13047470152378082, "learning_rate": 9.337082301749993e-05, "loss": 0.0092, "step": 20750 }, { "grad_norm": 0.1986130177974701, "learning_rate": 9.336259326406109e-05, "loss": 0.0094, "step": 20760 }, { "grad_norm": 0.13697689771652222, "learning_rate": 9.335435876855427e-05, "loss": 0.0086, "step": 20770 }, { "grad_norm": 0.20492780208587646, "learning_rate": 9.334611953187994e-05, "loss": 0.0106, "step": 20780 }, { "grad_norm": 0.1311679184436798, "learning_rate": 9.333787555493914e-05, "loss": 0.0099, "step": 20790 }, { "grad_norm": 0.11869297176599503, "learning_rate": 9.332962683863345e-05, "loss": 0.0084, "step": 20800 }, { "grad_norm": 0.17186711728572845, "learning_rate": 9.332137338386489e-05, "loss": 0.0082, "step": 20810 }, { "grad_norm": 0.13585296273231506, "learning_rate": 9.33131151915361e-05, "loss": 0.0089, "step": 20820 }, { "grad_norm": 0.11158602684736252, "learning_rate": 9.330485226255012e-05, "loss": 0.0098, "step": 20830 }, { "grad_norm": 0.210873082280159, "learning_rate": 9.329658459781061e-05, "loss": 0.0105, "step": 20840 }, { "grad_norm": 0.19031836092472076, "learning_rate": 9.328831219822172e-05, "loss": 0.0111, "step": 20850 }, { "grad_norm": 0.19481559097766876, "learning_rate": 9.328003506468808e-05, "loss": 0.0099, "step": 20860 }, { "grad_norm": 0.16924981772899628, "learning_rate": 9.327175319811488e-05, "loss": 0.0115, "step": 20870 }, { "grad_norm": 0.1528737097978592, "learning_rate": 9.326346659940781e-05, "loss": 0.0102, "step": 20880 }, { "grad_norm": 0.17284691333770752, "learning_rate": 9.325517526947308e-05, "loss": 0.0115, "step": 20890 }, { "grad_norm": 0.1707906275987625, "learning_rate": 9.32468792092174e-05, "loss": 0.0095, "step": 20900 }, { "grad_norm": 0.1706528514623642, "learning_rate": 9.323857841954803e-05, "loss": 0.0095, "step": 20910 }, { "grad_norm": 0.1423109769821167, "learning_rate": 9.323027290137276e-05, "loss": 0.0103, "step": 20920 }, { "grad_norm": 0.1655084788799286, "learning_rate": 9.322196265559981e-05, "loss": 0.0096, "step": 20930 }, { "grad_norm": 0.16853223741054535, "learning_rate": 9.321364768313803e-05, "loss": 0.01, "step": 20940 }, { "grad_norm": 0.16523230075836182, "learning_rate": 9.32053279848967e-05, "loss": 0.0096, "step": 20950 }, { "grad_norm": 0.16207270324230194, "learning_rate": 9.319700356178567e-05, "loss": 0.008, "step": 20960 }, { "grad_norm": 0.16721084713935852, "learning_rate": 9.318867441471527e-05, "loss": 0.0092, "step": 20970 }, { "grad_norm": 0.22384536266326904, "learning_rate": 9.318034054459637e-05, "loss": 0.0091, "step": 20980 }, { "grad_norm": 0.18484896421432495, "learning_rate": 9.317200195234034e-05, "loss": 0.01, "step": 20990 }, { "grad_norm": 0.18504755198955536, "learning_rate": 9.316365863885909e-05, "loss": 0.0093, "step": 21000 }, { "grad_norm": 0.20106838643550873, "learning_rate": 9.315531060506502e-05, "loss": 0.0109, "step": 21010 }, { "grad_norm": 0.17755690217018127, "learning_rate": 9.314695785187108e-05, "loss": 0.0084, "step": 21020 }, { "grad_norm": 0.17890842258930206, "learning_rate": 9.313860038019069e-05, "loss": 0.0106, "step": 21030 }, { "grad_norm": 0.16777192056179047, "learning_rate": 9.313023819093782e-05, "loss": 0.0131, "step": 21040 }, { "grad_norm": 0.12670856714248657, "learning_rate": 9.312187128502695e-05, "loss": 0.0081, "step": 21050 }, { "grad_norm": 0.12277147173881531, "learning_rate": 9.311349966337307e-05, "loss": 0.0095, "step": 21060 }, { "grad_norm": 0.1535506546497345, "learning_rate": 9.310512332689169e-05, "loss": 0.0096, "step": 21070 }, { "grad_norm": 0.1429797261953354, "learning_rate": 9.309674227649883e-05, "loss": 0.0097, "step": 21080 }, { "grad_norm": 0.17093244194984436, "learning_rate": 9.308835651311103e-05, "loss": 0.0097, "step": 21090 }, { "grad_norm": 0.17232996225357056, "learning_rate": 9.307996603764533e-05, "loss": 0.0096, "step": 21100 }, { "grad_norm": 0.14849649369716644, "learning_rate": 9.307157085101932e-05, "loss": 0.0092, "step": 21110 }, { "grad_norm": 0.1577005237340927, "learning_rate": 9.306317095415109e-05, "loss": 0.0088, "step": 21120 }, { "grad_norm": 0.1506040245294571, "learning_rate": 9.305476634795922e-05, "loss": 0.0093, "step": 21130 }, { "grad_norm": 0.19628740847110748, "learning_rate": 9.304635703336284e-05, "loss": 0.0095, "step": 21140 }, { "grad_norm": 0.21910879015922546, "learning_rate": 9.303794301128157e-05, "loss": 0.009, "step": 21150 }, { "grad_norm": 0.18472698330879211, "learning_rate": 9.302952428263555e-05, "loss": 0.0097, "step": 21160 }, { "grad_norm": 0.13688117265701294, "learning_rate": 9.302110084834545e-05, "loss": 0.0093, "step": 21170 }, { "grad_norm": 0.16983084380626678, "learning_rate": 9.301267270933245e-05, "loss": 0.0091, "step": 21180 }, { "grad_norm": 0.13853226602077484, "learning_rate": 9.300423986651823e-05, "loss": 0.0114, "step": 21190 }, { "grad_norm": 0.15152952075004578, "learning_rate": 9.299580232082501e-05, "loss": 0.0094, "step": 21200 }, { "grad_norm": 0.18275150656700134, "learning_rate": 9.298736007317547e-05, "loss": 0.01, "step": 21210 }, { "grad_norm": 0.1454230099916458, "learning_rate": 9.297891312449288e-05, "loss": 0.0109, "step": 21220 }, { "grad_norm": 0.24118955433368683, "learning_rate": 9.297046147570094e-05, "loss": 0.0088, "step": 21230 }, { "grad_norm": 0.14540399610996246, "learning_rate": 9.296200512772396e-05, "loss": 0.0091, "step": 21240 }, { "grad_norm": 0.20258063077926636, "learning_rate": 9.295354408148668e-05, "loss": 0.0091, "step": 21250 }, { "grad_norm": 0.1572381854057312, "learning_rate": 9.294507833791441e-05, "loss": 0.0089, "step": 21260 }, { "grad_norm": 0.17996597290039062, "learning_rate": 9.293660789793295e-05, "loss": 0.0096, "step": 21270 }, { "grad_norm": 0.1788841336965561, "learning_rate": 9.292813276246858e-05, "loss": 0.0096, "step": 21280 }, { "grad_norm": 0.12297702580690384, "learning_rate": 9.291965293244816e-05, "loss": 0.0086, "step": 21290 }, { "grad_norm": 0.1550878882408142, "learning_rate": 9.291116840879904e-05, "loss": 0.0088, "step": 21300 }, { "grad_norm": 0.16961181163787842, "learning_rate": 9.290267919244904e-05, "loss": 0.0086, "step": 21310 }, { "grad_norm": 0.16646499931812286, "learning_rate": 9.289418528432655e-05, "loss": 0.0098, "step": 21320 }, { "grad_norm": 0.14008162915706635, "learning_rate": 9.288568668536045e-05, "loss": 0.0088, "step": 21330 }, { "grad_norm": 0.15183411538600922, "learning_rate": 9.287718339648013e-05, "loss": 0.0091, "step": 21340 }, { "grad_norm": 0.13991408050060272, "learning_rate": 9.28686754186155e-05, "loss": 0.0095, "step": 21350 }, { "grad_norm": 0.16051584482192993, "learning_rate": 9.286016275269698e-05, "loss": 0.0087, "step": 21360 }, { "grad_norm": 0.14373275637626648, "learning_rate": 9.285164539965551e-05, "loss": 0.0083, "step": 21370 }, { "grad_norm": 0.12775205075740814, "learning_rate": 9.284312336042251e-05, "loss": 0.009, "step": 21380 }, { "grad_norm": 0.14377985894680023, "learning_rate": 9.283459663592996e-05, "loss": 0.0097, "step": 21390 }, { "grad_norm": 0.18501754105091095, "learning_rate": 9.282606522711033e-05, "loss": 0.0088, "step": 21400 }, { "grad_norm": 0.1841363161802292, "learning_rate": 9.281752913489657e-05, "loss": 0.0089, "step": 21410 }, { "grad_norm": 0.18358008563518524, "learning_rate": 9.280898836022222e-05, "loss": 0.0083, "step": 21420 }, { "grad_norm": 0.14864686131477356, "learning_rate": 9.280044290402126e-05, "loss": 0.0091, "step": 21430 }, { "grad_norm": 0.17151790857315063, "learning_rate": 9.279189276722821e-05, "loss": 0.0096, "step": 21440 }, { "grad_norm": 0.18916946649551392, "learning_rate": 9.278333795077812e-05, "loss": 0.0101, "step": 21450 }, { "grad_norm": 0.12748438119888306, "learning_rate": 9.27747784556065e-05, "loss": 0.0096, "step": 21460 }, { "grad_norm": 0.17594024538993835, "learning_rate": 9.276621428264942e-05, "loss": 0.0083, "step": 21470 }, { "grad_norm": 0.1758812963962555, "learning_rate": 9.275764543284345e-05, "loss": 0.0085, "step": 21480 }, { "grad_norm": 0.1918814331293106, "learning_rate": 9.274907190712566e-05, "loss": 0.0087, "step": 21490 }, { "grad_norm": 0.15680693089962006, "learning_rate": 9.274049370643363e-05, "loss": 0.0088, "step": 21500 }, { "grad_norm": 0.23330819606781006, "learning_rate": 9.273191083170547e-05, "loss": 0.0093, "step": 21510 }, { "grad_norm": 0.2050314098596573, "learning_rate": 9.27233232838798e-05, "loss": 0.0099, "step": 21520 }, { "grad_norm": 0.20624102652072906, "learning_rate": 9.27147310638957e-05, "loss": 0.0094, "step": 21530 }, { "grad_norm": 0.1472322642803192, "learning_rate": 9.270613417269286e-05, "loss": 0.0095, "step": 21540 }, { "grad_norm": 0.18603834509849548, "learning_rate": 9.269753261121138e-05, "loss": 0.0102, "step": 21550 }, { "grad_norm": 0.17027023434638977, "learning_rate": 9.268892638039194e-05, "loss": 0.0092, "step": 21560 }, { "grad_norm": 0.1398957073688507, "learning_rate": 9.268031548117569e-05, "loss": 0.0093, "step": 21570 }, { "grad_norm": 0.18045954406261444, "learning_rate": 9.26716999145043e-05, "loss": 0.009, "step": 21580 }, { "grad_norm": 0.2263951152563095, "learning_rate": 9.266307968131998e-05, "loss": 0.0099, "step": 21590 }, { "grad_norm": 0.15258587896823883, "learning_rate": 9.26544547825654e-05, "loss": 0.0102, "step": 21600 }, { "grad_norm": 0.1840156614780426, "learning_rate": 9.264582521918376e-05, "loss": 0.0099, "step": 21610 }, { "grad_norm": 0.1663631945848465, "learning_rate": 9.263719099211881e-05, "loss": 0.01, "step": 21620 }, { "grad_norm": 0.1801300048828125, "learning_rate": 9.262855210231476e-05, "loss": 0.0083, "step": 21630 }, { "grad_norm": 0.14262908697128296, "learning_rate": 9.261990855071633e-05, "loss": 0.0108, "step": 21640 }, { "grad_norm": 0.19491545855998993, "learning_rate": 9.261126033826878e-05, "loss": 0.0083, "step": 21650 }, { "grad_norm": 0.1562386006116867, "learning_rate": 9.260260746591786e-05, "loss": 0.009, "step": 21660 }, { "grad_norm": 0.14422211050987244, "learning_rate": 9.259394993460985e-05, "loss": 0.0088, "step": 21670 }, { "grad_norm": 0.1700417697429657, "learning_rate": 9.258528774529151e-05, "loss": 0.0091, "step": 21680 }, { "grad_norm": 0.18267838656902313, "learning_rate": 9.257662089891013e-05, "loss": 0.0101, "step": 21690 }, { "grad_norm": 0.09926625341176987, "learning_rate": 9.25679493964135e-05, "loss": 0.0076, "step": 21700 }, { "grad_norm": 0.12677571177482605, "learning_rate": 9.255927323874994e-05, "loss": 0.0082, "step": 21710 }, { "grad_norm": 0.13829880952835083, "learning_rate": 9.255059242686822e-05, "loss": 0.0074, "step": 21720 }, { "grad_norm": 0.1966012716293335, "learning_rate": 9.254190696171769e-05, "loss": 0.0084, "step": 21730 }, { "grad_norm": 0.18231911957263947, "learning_rate": 9.25332168442482e-05, "loss": 0.0085, "step": 21740 }, { "grad_norm": 0.11940547823905945, "learning_rate": 9.252452207541004e-05, "loss": 0.0087, "step": 21750 }, { "grad_norm": 0.20057843625545502, "learning_rate": 9.251582265615409e-05, "loss": 0.0086, "step": 21760 }, { "grad_norm": 0.17527148127555847, "learning_rate": 9.250711858743169e-05, "loss": 0.0087, "step": 21770 }, { "grad_norm": 0.2022380381822586, "learning_rate": 9.24984098701947e-05, "loss": 0.0105, "step": 21780 }, { "grad_norm": 0.14482811093330383, "learning_rate": 9.248969650539552e-05, "loss": 0.0081, "step": 21790 }, { "grad_norm": 0.20539534091949463, "learning_rate": 9.2480978493987e-05, "loss": 0.0102, "step": 21800 }, { "grad_norm": 0.1345069706439972, "learning_rate": 9.247225583692256e-05, "loss": 0.0087, "step": 21810 }, { "grad_norm": 0.12574522197246552, "learning_rate": 9.246352853515607e-05, "loss": 0.0087, "step": 21820 }, { "grad_norm": 0.16518893837928772, "learning_rate": 9.245479658964194e-05, "loss": 0.0088, "step": 21830 }, { "grad_norm": 0.17075280845165253, "learning_rate": 9.244606000133507e-05, "loss": 0.0087, "step": 21840 }, { "grad_norm": 0.1482645571231842, "learning_rate": 9.24373187711909e-05, "loss": 0.0101, "step": 21850 }, { "grad_norm": 0.16837802529335022, "learning_rate": 9.242857290016537e-05, "loss": 0.0089, "step": 21860 }, { "grad_norm": 0.1501701921224594, "learning_rate": 9.241982238921488e-05, "loss": 0.0087, "step": 21870 }, { "grad_norm": 0.16795340180397034, "learning_rate": 9.24110672392964e-05, "loss": 0.0098, "step": 21880 }, { "grad_norm": 0.19904601573944092, "learning_rate": 9.240230745136737e-05, "loss": 0.0095, "step": 21890 }, { "grad_norm": 0.14920617640018463, "learning_rate": 9.239354302638575e-05, "loss": 0.0099, "step": 21900 }, { "grad_norm": 0.19926013052463531, "learning_rate": 9.238477396531e-05, "loss": 0.0088, "step": 21910 }, { "grad_norm": 0.22300592064857483, "learning_rate": 9.23760002690991e-05, "loss": 0.0096, "step": 21920 }, { "grad_norm": 0.17374642193317413, "learning_rate": 9.236722193871252e-05, "loss": 0.0099, "step": 21930 }, { "grad_norm": 0.14547494053840637, "learning_rate": 9.235843897511023e-05, "loss": 0.0091, "step": 21940 }, { "grad_norm": 0.22848014533519745, "learning_rate": 9.234965137925276e-05, "loss": 0.0095, "step": 21950 }, { "grad_norm": 0.19437569379806519, "learning_rate": 9.234085915210108e-05, "loss": 0.0106, "step": 21960 }, { "grad_norm": 0.18296454846858978, "learning_rate": 9.23320622946167e-05, "loss": 0.0092, "step": 21970 }, { "grad_norm": 0.14153355360031128, "learning_rate": 9.232326080776163e-05, "loss": 0.009, "step": 21980 }, { "grad_norm": 0.15713992714881897, "learning_rate": 9.23144546924984e-05, "loss": 0.01, "step": 21990 }, { "grad_norm": 0.19422633945941925, "learning_rate": 9.230564394979e-05, "loss": 0.0093, "step": 22000 }, { "grad_norm": 0.1392928510904312, "learning_rate": 9.22968285806e-05, "loss": 0.0088, "step": 22010 }, { "grad_norm": 0.15768958628177643, "learning_rate": 9.228800858589242e-05, "loss": 0.0093, "step": 22020 }, { "grad_norm": 0.17602980136871338, "learning_rate": 9.227918396663179e-05, "loss": 0.0098, "step": 22030 }, { "grad_norm": 0.14835621416568756, "learning_rate": 9.227035472378319e-05, "loss": 0.0087, "step": 22040 }, { "grad_norm": 0.19053179025650024, "learning_rate": 9.226152085831213e-05, "loss": 0.0098, "step": 22050 }, { "grad_norm": 0.13902178406715393, "learning_rate": 9.22526823711847e-05, "loss": 0.0105, "step": 22060 }, { "grad_norm": 0.16866059601306915, "learning_rate": 9.224383926336745e-05, "loss": 0.0091, "step": 22070 }, { "grad_norm": 0.15777687728405, "learning_rate": 9.223499153582744e-05, "loss": 0.0109, "step": 22080 }, { "grad_norm": 0.18874762952327728, "learning_rate": 9.222613918953226e-05, "loss": 0.0091, "step": 22090 }, { "grad_norm": 0.13824252784252167, "learning_rate": 9.221728222544999e-05, "loss": 0.0086, "step": 22100 }, { "grad_norm": 0.2059050351381302, "learning_rate": 9.22084206445492e-05, "loss": 0.0106, "step": 22110 }, { "grad_norm": 0.17148636281490326, "learning_rate": 9.2199554447799e-05, "loss": 0.0117, "step": 22120 }, { "grad_norm": 0.22490613162517548, "learning_rate": 9.219068363616897e-05, "loss": 0.0104, "step": 22130 }, { "grad_norm": 0.1628367304801941, "learning_rate": 9.218180821062919e-05, "loss": 0.0108, "step": 22140 }, { "grad_norm": 0.19530777633190155, "learning_rate": 9.21729281721503e-05, "loss": 0.0087, "step": 22150 }, { "grad_norm": 0.16038210690021515, "learning_rate": 9.216404352170339e-05, "loss": 0.0085, "step": 22160 }, { "grad_norm": 0.17047326266765594, "learning_rate": 9.215515426026007e-05, "loss": 0.0088, "step": 22170 }, { "grad_norm": 0.15595506131649017, "learning_rate": 9.214626038879246e-05, "loss": 0.0089, "step": 22180 }, { "grad_norm": 0.1351913958787918, "learning_rate": 9.21373619082732e-05, "loss": 0.0083, "step": 22190 }, { "grad_norm": 0.17323438823223114, "learning_rate": 9.212845881967535e-05, "loss": 0.008, "step": 22200 }, { "grad_norm": 0.1826910823583603, "learning_rate": 9.211955112397262e-05, "loss": 0.0087, "step": 22210 }, { "grad_norm": 0.17421509325504303, "learning_rate": 9.211063882213909e-05, "loss": 0.008, "step": 22220 }, { "grad_norm": 0.16434291005134583, "learning_rate": 9.210172191514942e-05, "loss": 0.0089, "step": 22230 }, { "grad_norm": 0.1450960636138916, "learning_rate": 9.209280040397874e-05, "loss": 0.0086, "step": 22240 }, { "grad_norm": 0.22899457812309265, "learning_rate": 9.208387428960268e-05, "loss": 0.008, "step": 22250 }, { "grad_norm": 0.18697397410869598, "learning_rate": 9.20749435729974e-05, "loss": 0.0078, "step": 22260 }, { "grad_norm": 0.16412536799907684, "learning_rate": 9.206600825513957e-05, "loss": 0.0099, "step": 22270 }, { "grad_norm": 0.21262352168560028, "learning_rate": 9.20570683370063e-05, "loss": 0.0094, "step": 22280 }, { "grad_norm": 0.14025256037712097, "learning_rate": 9.204812381957528e-05, "loss": 0.0099, "step": 22290 }, { "grad_norm": 0.14159905910491943, "learning_rate": 9.203917470382465e-05, "loss": 0.0074, "step": 22300 }, { "grad_norm": 0.12692593038082123, "learning_rate": 9.203022099073309e-05, "loss": 0.0083, "step": 22310 }, { "grad_norm": 0.12604834139347076, "learning_rate": 9.202126268127976e-05, "loss": 0.0074, "step": 22320 }, { "grad_norm": 0.14346778392791748, "learning_rate": 9.20122997764443e-05, "loss": 0.0085, "step": 22330 }, { "grad_norm": 0.1564849317073822, "learning_rate": 9.200333227720692e-05, "loss": 0.0098, "step": 22340 }, { "grad_norm": 0.16071976721286774, "learning_rate": 9.199436018454826e-05, "loss": 0.0094, "step": 22350 }, { "grad_norm": 0.16288761794567108, "learning_rate": 9.198538349944952e-05, "loss": 0.0093, "step": 22360 }, { "grad_norm": 0.18473760783672333, "learning_rate": 9.197640222289234e-05, "loss": 0.0094, "step": 22370 }, { "grad_norm": 0.23101818561553955, "learning_rate": 9.196741635585895e-05, "loss": 0.0086, "step": 22380 }, { "grad_norm": 0.18981976807117462, "learning_rate": 9.195842589933199e-05, "loss": 0.0091, "step": 22390 }, { "grad_norm": 0.15837155282497406, "learning_rate": 9.194943085429466e-05, "loss": 0.0089, "step": 22400 }, { "grad_norm": 0.13335229456424713, "learning_rate": 9.194043122173065e-05, "loss": 0.0098, "step": 22410 }, { "grad_norm": 0.14874950051307678, "learning_rate": 9.193142700262413e-05, "loss": 0.0093, "step": 22420 }, { "grad_norm": 0.15908941626548767, "learning_rate": 9.192241819795979e-05, "loss": 0.0088, "step": 22430 }, { "grad_norm": 0.19252198934555054, "learning_rate": 9.191340480872284e-05, "loss": 0.0078, "step": 22440 }, { "grad_norm": 0.14270135760307312, "learning_rate": 9.190438683589895e-05, "loss": 0.0095, "step": 22450 }, { "grad_norm": 0.21423569321632385, "learning_rate": 9.189536428047432e-05, "loss": 0.0105, "step": 22460 }, { "grad_norm": 0.13468681275844574, "learning_rate": 9.188633714343564e-05, "loss": 0.0098, "step": 22470 }, { "grad_norm": 0.16464863717556, "learning_rate": 9.18773054257701e-05, "loss": 0.0085, "step": 22480 }, { "grad_norm": 0.11985877901315689, "learning_rate": 9.18682691284654e-05, "loss": 0.0073, "step": 22490 }, { "grad_norm": 0.14697109162807465, "learning_rate": 9.185922825250974e-05, "loss": 0.0093, "step": 22500 }, { "grad_norm": 0.17661546170711517, "learning_rate": 9.185018279889181e-05, "loss": 0.0102, "step": 22510 }, { "grad_norm": 0.20878343284130096, "learning_rate": 9.184113276860082e-05, "loss": 0.0095, "step": 22520 }, { "grad_norm": 0.11553771793842316, "learning_rate": 9.183207816262645e-05, "loss": 0.0084, "step": 22530 }, { "grad_norm": 0.12403970956802368, "learning_rate": 9.182301898195891e-05, "loss": 0.0094, "step": 22540 }, { "grad_norm": 0.13209772109985352, "learning_rate": 9.181395522758889e-05, "loss": 0.0091, "step": 22550 }, { "grad_norm": 0.16710568964481354, "learning_rate": 9.180488690050759e-05, "loss": 0.0092, "step": 22560 }, { "grad_norm": 0.1577199250459671, "learning_rate": 9.179581400170671e-05, "loss": 0.0094, "step": 22570 }, { "grad_norm": 0.1266585737466812, "learning_rate": 9.178673653217845e-05, "loss": 0.0084, "step": 22580 }, { "grad_norm": 0.1526888906955719, "learning_rate": 9.177765449291551e-05, "loss": 0.0071, "step": 22590 }, { "grad_norm": 0.119814433157444, "learning_rate": 9.176856788491109e-05, "loss": 0.0082, "step": 22600 }, { "grad_norm": 0.14025399088859558, "learning_rate": 9.175947670915887e-05, "loss": 0.0074, "step": 22610 }, { "grad_norm": 0.1567646861076355, "learning_rate": 9.175038096665309e-05, "loss": 0.0072, "step": 22620 }, { "grad_norm": 0.1310335397720337, "learning_rate": 9.17412806583884e-05, "loss": 0.0089, "step": 22630 }, { "grad_norm": 0.15423446893692017, "learning_rate": 9.173217578536002e-05, "loss": 0.0088, "step": 22640 }, { "grad_norm": 0.16332681477069855, "learning_rate": 9.172306634856362e-05, "loss": 0.0094, "step": 22650 }, { "grad_norm": 0.23044775426387787, "learning_rate": 9.171395234899545e-05, "loss": 0.0091, "step": 22660 }, { "grad_norm": 0.1436641961336136, "learning_rate": 9.170483378765214e-05, "loss": 0.0082, "step": 22670 }, { "grad_norm": 0.18167544901371002, "learning_rate": 9.169571066553091e-05, "loss": 0.0088, "step": 22680 }, { "grad_norm": 0.218098446726799, "learning_rate": 9.168658298362946e-05, "loss": 0.0087, "step": 22690 }, { "grad_norm": 0.1287192404270172, "learning_rate": 9.167745074294598e-05, "loss": 0.0081, "step": 22700 }, { "grad_norm": 0.15724912285804749, "learning_rate": 9.166831394447913e-05, "loss": 0.0085, "step": 22710 }, { "grad_norm": 0.16478367149829865, "learning_rate": 9.165917258922812e-05, "loss": 0.0079, "step": 22720 }, { "grad_norm": 0.14786866307258606, "learning_rate": 9.165002667819262e-05, "loss": 0.0097, "step": 22730 }, { "grad_norm": 0.17982381582260132, "learning_rate": 9.164087621237282e-05, "loss": 0.009, "step": 22740 }, { "grad_norm": 0.1274188756942749, "learning_rate": 9.163172119276942e-05, "loss": 0.0087, "step": 22750 }, { "grad_norm": 0.2665190100669861, "learning_rate": 9.162256162038358e-05, "loss": 0.0102, "step": 22760 }, { "grad_norm": 0.16847071051597595, "learning_rate": 9.161339749621698e-05, "loss": 0.0085, "step": 22770 }, { "grad_norm": 0.1634378731250763, "learning_rate": 9.160422882127177e-05, "loss": 0.0096, "step": 22780 }, { "grad_norm": 0.1395900994539261, "learning_rate": 9.159505559655069e-05, "loss": 0.0082, "step": 22790 }, { "grad_norm": 0.12643614411354065, "learning_rate": 9.158587782305684e-05, "loss": 0.0081, "step": 22800 }, { "grad_norm": 0.1492277979850769, "learning_rate": 9.157669550179391e-05, "loss": 0.0085, "step": 22810 }, { "grad_norm": 0.15140338242053986, "learning_rate": 9.156750863376609e-05, "loss": 0.0094, "step": 22820 }, { "grad_norm": 0.19123752415180206, "learning_rate": 9.155831721997801e-05, "loss": 0.0088, "step": 22830 }, { "grad_norm": 0.19994108378887177, "learning_rate": 9.154912126143484e-05, "loss": 0.0088, "step": 22840 }, { "grad_norm": 0.16809916496276855, "learning_rate": 9.153992075914224e-05, "loss": 0.0083, "step": 22850 }, { "grad_norm": 0.12025898694992065, "learning_rate": 9.153071571410635e-05, "loss": 0.0085, "step": 22860 }, { "grad_norm": 0.17882534861564636, "learning_rate": 9.152150612733384e-05, "loss": 0.0081, "step": 22870 }, { "grad_norm": 0.19265460968017578, "learning_rate": 9.151229199983184e-05, "loss": 0.0092, "step": 22880 }, { "grad_norm": 0.16676577925682068, "learning_rate": 9.150307333260802e-05, "loss": 0.0087, "step": 22890 }, { "grad_norm": 0.19259877502918243, "learning_rate": 9.149385012667048e-05, "loss": 0.0076, "step": 22900 }, { "grad_norm": 0.14600910246372223, "learning_rate": 9.148462238302788e-05, "loss": 0.0082, "step": 22910 }, { "grad_norm": 0.2023015171289444, "learning_rate": 9.147539010268936e-05, "loss": 0.0082, "step": 22920 }, { "grad_norm": 0.14134448766708374, "learning_rate": 9.14661532866645e-05, "loss": 0.0079, "step": 22930 }, { "grad_norm": 0.12305139750242233, "learning_rate": 9.145691193596348e-05, "loss": 0.0081, "step": 22940 }, { "grad_norm": 0.1517171859741211, "learning_rate": 9.144766605159691e-05, "loss": 0.0085, "step": 22950 }, { "grad_norm": 0.1353994607925415, "learning_rate": 9.14384156345759e-05, "loss": 0.0091, "step": 22960 }, { "grad_norm": 0.20241208374500275, "learning_rate": 9.142916068591204e-05, "loss": 0.0083, "step": 22970 }, { "grad_norm": 0.12334839999675751, "learning_rate": 9.141990120661746e-05, "loss": 0.0081, "step": 22980 }, { "grad_norm": 0.15685944259166718, "learning_rate": 9.141063719770475e-05, "loss": 0.0101, "step": 22990 }, { "grad_norm": 0.17755351960659027, "learning_rate": 9.140136866018704e-05, "loss": 0.009, "step": 23000 }, { "grad_norm": 0.16483500599861145, "learning_rate": 9.139209559507788e-05, "loss": 0.0089, "step": 23010 }, { "grad_norm": 0.14808738231658936, "learning_rate": 9.13828180033914e-05, "loss": 0.0106, "step": 23020 }, { "grad_norm": 0.1959356814622879, "learning_rate": 9.137353588614212e-05, "loss": 0.0099, "step": 23030 }, { "grad_norm": 0.21873877942562103, "learning_rate": 9.136424924434519e-05, "loss": 0.0096, "step": 23040 }, { "grad_norm": 0.15184485912322998, "learning_rate": 9.135495807901615e-05, "loss": 0.0083, "step": 23050 }, { "grad_norm": 0.14126572012901306, "learning_rate": 9.134566239117108e-05, "loss": 0.0091, "step": 23060 }, { "grad_norm": 0.12237631529569626, "learning_rate": 9.13363621818265e-05, "loss": 0.0073, "step": 23070 }, { "grad_norm": 0.17606523633003235, "learning_rate": 9.132705745199953e-05, "loss": 0.0093, "step": 23080 }, { "grad_norm": 0.17328594624996185, "learning_rate": 9.131774820270768e-05, "loss": 0.0093, "step": 23090 }, { "grad_norm": 0.16404691338539124, "learning_rate": 9.130843443496901e-05, "loss": 0.0078, "step": 23100 }, { "grad_norm": 0.12548796832561493, "learning_rate": 9.129911614980206e-05, "loss": 0.0085, "step": 23110 }, { "grad_norm": 0.16071180999279022, "learning_rate": 9.128979334822584e-05, "loss": 0.0095, "step": 23120 }, { "grad_norm": 0.2083810567855835, "learning_rate": 9.128046603125992e-05, "loss": 0.008, "step": 23130 }, { "grad_norm": 0.19423028826713562, "learning_rate": 9.12711341999243e-05, "loss": 0.0084, "step": 23140 }, { "grad_norm": 0.14758457243442535, "learning_rate": 9.12617978552395e-05, "loss": 0.0095, "step": 23150 }, { "grad_norm": 0.15780118107795715, "learning_rate": 9.12524569982265e-05, "loss": 0.0086, "step": 23160 }, { "grad_norm": 0.14760665595531464, "learning_rate": 9.124311162990684e-05, "loss": 0.0083, "step": 23170 }, { "grad_norm": 0.14734487235546112, "learning_rate": 9.12337617513025e-05, "loss": 0.0082, "step": 23180 }, { "grad_norm": 0.12068001925945282, "learning_rate": 9.122440736343596e-05, "loss": 0.0083, "step": 23190 }, { "grad_norm": 0.15443062782287598, "learning_rate": 9.12150484673302e-05, "loss": 0.0088, "step": 23200 }, { "grad_norm": 0.14673691987991333, "learning_rate": 9.120568506400873e-05, "loss": 0.0088, "step": 23210 }, { "grad_norm": 0.1610925793647766, "learning_rate": 9.119631715449548e-05, "loss": 0.0093, "step": 23220 }, { "grad_norm": 0.21469277143478394, "learning_rate": 9.118694473981493e-05, "loss": 0.0091, "step": 23230 }, { "grad_norm": 0.22963270545005798, "learning_rate": 9.117756782099203e-05, "loss": 0.0125, "step": 23240 }, { "grad_norm": 0.19749261438846588, "learning_rate": 9.11681863990522e-05, "loss": 0.0084, "step": 23250 }, { "grad_norm": 0.12493160367012024, "learning_rate": 9.115880047502142e-05, "loss": 0.0079, "step": 23260 }, { "grad_norm": 0.2042454183101654, "learning_rate": 9.114941004992609e-05, "loss": 0.0094, "step": 23270 }, { "grad_norm": 0.1835818886756897, "learning_rate": 9.114001512479317e-05, "loss": 0.0093, "step": 23280 }, { "grad_norm": 0.15949799120426178, "learning_rate": 9.113061570065003e-05, "loss": 0.0089, "step": 23290 }, { "grad_norm": 0.14619660377502441, "learning_rate": 9.112121177852459e-05, "loss": 0.0087, "step": 23300 }, { "grad_norm": 0.1348583698272705, "learning_rate": 9.111180335944527e-05, "loss": 0.0086, "step": 23310 }, { "grad_norm": 0.1394542008638382, "learning_rate": 9.110239044444093e-05, "loss": 0.008, "step": 23320 }, { "grad_norm": 0.10509855300188065, "learning_rate": 9.109297303454099e-05, "loss": 0.0087, "step": 23330 }, { "grad_norm": 0.20017892122268677, "learning_rate": 9.108355113077526e-05, "loss": 0.0081, "step": 23340 }, { "grad_norm": 0.10406965017318726, "learning_rate": 9.107412473417419e-05, "loss": 0.0074, "step": 23350 }, { "grad_norm": 0.1732950657606125, "learning_rate": 9.106469384576858e-05, "loss": 0.0077, "step": 23360 }, { "grad_norm": 0.20216189324855804, "learning_rate": 9.105525846658978e-05, "loss": 0.0081, "step": 23370 }, { "grad_norm": 0.16337306797504425, "learning_rate": 9.104581859766965e-05, "loss": 0.009, "step": 23380 }, { "grad_norm": 0.17604592442512512, "learning_rate": 9.10363742400405e-05, "loss": 0.0093, "step": 23390 }, { "grad_norm": 0.17455902695655823, "learning_rate": 9.102692539473518e-05, "loss": 0.0082, "step": 23400 }, { "grad_norm": 0.15875540673732758, "learning_rate": 9.101747206278697e-05, "loss": 0.0094, "step": 23410 }, { "grad_norm": 0.13638915121555328, "learning_rate": 9.100801424522968e-05, "loss": 0.0075, "step": 23420 }, { "grad_norm": 0.15235696732997894, "learning_rate": 9.099855194309762e-05, "loss": 0.0083, "step": 23430 }, { "grad_norm": 0.1766950488090515, "learning_rate": 9.098908515742554e-05, "loss": 0.0084, "step": 23440 }, { "grad_norm": 0.17550094425678253, "learning_rate": 9.097961388924873e-05, "loss": 0.0085, "step": 23450 }, { "grad_norm": 0.18487758934497833, "learning_rate": 9.097013813960298e-05, "loss": 0.0098, "step": 23460 }, { "grad_norm": 0.1020863950252533, "learning_rate": 9.09606579095245e-05, "loss": 0.0086, "step": 23470 }, { "grad_norm": 0.1041572168469429, "learning_rate": 9.095117320005008e-05, "loss": 0.0079, "step": 23480 }, { "grad_norm": 0.1244489923119545, "learning_rate": 9.094168401221691e-05, "loss": 0.0086, "step": 23490 }, { "grad_norm": 0.18120847642421722, "learning_rate": 9.093219034706273e-05, "loss": 0.01, "step": 23500 }, { "grad_norm": 0.13532039523124695, "learning_rate": 9.092269220562577e-05, "loss": 0.0085, "step": 23510 }, { "grad_norm": 0.14048053324222565, "learning_rate": 9.09131895889447e-05, "loss": 0.0093, "step": 23520 }, { "grad_norm": 0.13409264385700226, "learning_rate": 9.090368249805873e-05, "loss": 0.0094, "step": 23530 }, { "grad_norm": 0.2105754166841507, "learning_rate": 9.089417093400754e-05, "loss": 0.0099, "step": 23540 }, { "grad_norm": 0.19830316305160522, "learning_rate": 9.088465489783131e-05, "loss": 0.0091, "step": 23550 }, { "grad_norm": 0.20426787436008453, "learning_rate": 9.087513439057068e-05, "loss": 0.0105, "step": 23560 }, { "grad_norm": 0.15517622232437134, "learning_rate": 9.08656094132668e-05, "loss": 0.0105, "step": 23570 }, { "grad_norm": 0.12144474685192108, "learning_rate": 9.085607996696134e-05, "loss": 0.0094, "step": 23580 }, { "grad_norm": 0.167647585272789, "learning_rate": 9.084654605269639e-05, "loss": 0.0107, "step": 23590 }, { "grad_norm": 0.16796265542507172, "learning_rate": 9.083700767151457e-05, "loss": 0.0086, "step": 23600 }, { "grad_norm": 0.10144126415252686, "learning_rate": 9.082746482445898e-05, "loss": 0.0088, "step": 23610 }, { "grad_norm": 0.13897864520549774, "learning_rate": 9.081791751257325e-05, "loss": 0.009, "step": 23620 }, { "grad_norm": 0.1269090473651886, "learning_rate": 9.080836573690142e-05, "loss": 0.0081, "step": 23630 }, { "grad_norm": 0.14354906976222992, "learning_rate": 9.079880949848805e-05, "loss": 0.009, "step": 23640 }, { "grad_norm": 0.14526760578155518, "learning_rate": 9.078924879837822e-05, "loss": 0.0083, "step": 23650 }, { "grad_norm": 0.1517181247472763, "learning_rate": 9.077968363761747e-05, "loss": 0.0082, "step": 23660 }, { "grad_norm": 0.14884768426418304, "learning_rate": 9.077011401725182e-05, "loss": 0.0091, "step": 23670 }, { "grad_norm": 0.1598699986934662, "learning_rate": 9.07605399383278e-05, "loss": 0.0085, "step": 23680 }, { "grad_norm": 0.21434487402439117, "learning_rate": 9.075096140189243e-05, "loss": 0.0089, "step": 23690 }, { "grad_norm": 0.16633571684360504, "learning_rate": 9.074137840899318e-05, "loss": 0.0101, "step": 23700 }, { "grad_norm": 0.1666557788848877, "learning_rate": 9.073179096067804e-05, "loss": 0.0084, "step": 23710 }, { "grad_norm": 0.1287163645029068, "learning_rate": 9.072219905799549e-05, "loss": 0.0111, "step": 23720 }, { "grad_norm": 0.15168415009975433, "learning_rate": 9.071260270199447e-05, "loss": 0.013, "step": 23730 }, { "grad_norm": 0.17583772540092468, "learning_rate": 9.070300189372441e-05, "loss": 0.0122, "step": 23740 }, { "grad_norm": 0.1411771923303604, "learning_rate": 9.069339663423528e-05, "loss": 0.011, "step": 23750 }, { "grad_norm": 0.11866027116775513, "learning_rate": 9.068378692457747e-05, "loss": 0.0115, "step": 23760 }, { "grad_norm": 0.15022221207618713, "learning_rate": 9.067417276580189e-05, "loss": 0.0095, "step": 23770 }, { "grad_norm": 0.16497844457626343, "learning_rate": 9.066455415895993e-05, "loss": 0.0106, "step": 23780 }, { "grad_norm": 0.1723879724740982, "learning_rate": 9.065493110510346e-05, "loss": 0.0106, "step": 23790 }, { "grad_norm": 0.17972181737422943, "learning_rate": 9.064530360528484e-05, "loss": 0.0102, "step": 23800 }, { "grad_norm": 0.153659850358963, "learning_rate": 9.063567166055695e-05, "loss": 0.0101, "step": 23810 }, { "grad_norm": 0.17102384567260742, "learning_rate": 9.062603527197308e-05, "loss": 0.0095, "step": 23820 }, { "grad_norm": 0.15165136754512787, "learning_rate": 9.06163944405871e-05, "loss": 0.0085, "step": 23830 }, { "grad_norm": 0.1929195076227188, "learning_rate": 9.060674916745327e-05, "loss": 0.0085, "step": 23840 }, { "grad_norm": 0.1727643609046936, "learning_rate": 9.05970994536264e-05, "loss": 0.0086, "step": 23850 }, { "grad_norm": 0.14371885359287262, "learning_rate": 9.05874453001618e-05, "loss": 0.0082, "step": 23860 }, { "grad_norm": 0.1260405033826828, "learning_rate": 9.057778670811517e-05, "loss": 0.0095, "step": 23870 }, { "grad_norm": 0.20240002870559692, "learning_rate": 9.056812367854281e-05, "loss": 0.0082, "step": 23880 }, { "grad_norm": 0.14286670088768005, "learning_rate": 9.055845621250143e-05, "loss": 0.0089, "step": 23890 }, { "grad_norm": 0.14618799090385437, "learning_rate": 9.054878431104825e-05, "loss": 0.0076, "step": 23900 }, { "grad_norm": 0.15490609407424927, "learning_rate": 9.0539107975241e-05, "loss": 0.0084, "step": 23910 }, { "grad_norm": 0.15604621171951294, "learning_rate": 9.052942720613784e-05, "loss": 0.0079, "step": 23920 }, { "grad_norm": 0.14420585334300995, "learning_rate": 9.051974200479745e-05, "loss": 0.0078, "step": 23930 }, { "grad_norm": 0.14455249905586243, "learning_rate": 9.051005237227901e-05, "loss": 0.0081, "step": 23940 }, { "grad_norm": 0.1473190039396286, "learning_rate": 9.050035830964215e-05, "loss": 0.0092, "step": 23950 }, { "grad_norm": 0.12258944660425186, "learning_rate": 9.049065981794698e-05, "loss": 0.0085, "step": 23960 }, { "grad_norm": 0.14998552203178406, "learning_rate": 9.048095689825414e-05, "loss": 0.0087, "step": 23970 }, { "grad_norm": 0.15901386737823486, "learning_rate": 9.047124955162472e-05, "loss": 0.0083, "step": 23980 }, { "grad_norm": 0.15681679546833038, "learning_rate": 9.046153777912028e-05, "loss": 0.0085, "step": 23990 }, { "grad_norm": 0.13661688566207886, "learning_rate": 9.045182158180292e-05, "loss": 0.0078, "step": 24000 }, { "grad_norm": 0.1753833293914795, "learning_rate": 9.044210096073516e-05, "loss": 0.0077, "step": 24010 }, { "grad_norm": 0.12909455597400665, "learning_rate": 9.043237591698004e-05, "loss": 0.0074, "step": 24020 }, { "grad_norm": 0.1409531831741333, "learning_rate": 9.04226464516011e-05, "loss": 0.0079, "step": 24030 }, { "grad_norm": 0.12366678565740585, "learning_rate": 9.041291256566229e-05, "loss": 0.0075, "step": 24040 }, { "grad_norm": 0.13060279190540314, "learning_rate": 9.040317426022814e-05, "loss": 0.008, "step": 24050 }, { "grad_norm": 0.14428642392158508, "learning_rate": 9.03934315363636e-05, "loss": 0.0086, "step": 24060 }, { "grad_norm": 0.1530367136001587, "learning_rate": 9.038368439513409e-05, "loss": 0.0082, "step": 24070 }, { "grad_norm": 0.13706941902637482, "learning_rate": 9.03739328376056e-05, "loss": 0.0092, "step": 24080 }, { "grad_norm": 0.11682993918657303, "learning_rate": 9.036417686484451e-05, "loss": 0.0096, "step": 24090 }, { "grad_norm": 0.16345453262329102, "learning_rate": 9.035441647791773e-05, "loss": 0.0086, "step": 24100 }, { "grad_norm": 0.18851859867572784, "learning_rate": 9.034465167789263e-05, "loss": 0.009, "step": 24110 }, { "grad_norm": 0.1274908483028412, "learning_rate": 9.033488246583706e-05, "loss": 0.0089, "step": 24120 }, { "grad_norm": 0.10438261926174164, "learning_rate": 9.032510884281941e-05, "loss": 0.0079, "step": 24130 }, { "grad_norm": 0.12147900462150574, "learning_rate": 9.031533080990848e-05, "loss": 0.008, "step": 24140 }, { "grad_norm": 0.12839244306087494, "learning_rate": 9.030554836817358e-05, "loss": 0.0095, "step": 24150 }, { "grad_norm": 0.17381422221660614, "learning_rate": 9.029576151868451e-05, "loss": 0.0091, "step": 24160 }, { "grad_norm": 0.19154809415340424, "learning_rate": 9.028597026251155e-05, "loss": 0.0072, "step": 24170 }, { "grad_norm": 0.23587465286254883, "learning_rate": 9.027617460072547e-05, "loss": 0.0081, "step": 24180 }, { "grad_norm": 0.2001708298921585, "learning_rate": 9.026637453439745e-05, "loss": 0.0084, "step": 24190 }, { "grad_norm": 0.13423030078411102, "learning_rate": 9.025657006459927e-05, "loss": 0.0079, "step": 24200 }, { "grad_norm": 0.1128801479935646, "learning_rate": 9.024676119240311e-05, "loss": 0.008, "step": 24210 }, { "grad_norm": 0.14877845346927643, "learning_rate": 9.023694791888166e-05, "loss": 0.0083, "step": 24220 }, { "grad_norm": 0.13631127774715424, "learning_rate": 9.022713024510808e-05, "loss": 0.0085, "step": 24230 }, { "grad_norm": 0.13263216614723206, "learning_rate": 9.021730817215601e-05, "loss": 0.0078, "step": 24240 }, { "grad_norm": 0.15907993912696838, "learning_rate": 9.02074817010996e-05, "loss": 0.008, "step": 24250 }, { "grad_norm": 0.13974817097187042, "learning_rate": 9.019765083301342e-05, "loss": 0.0084, "step": 24260 }, { "grad_norm": 0.18171891570091248, "learning_rate": 9.01878155689726e-05, "loss": 0.0075, "step": 24270 }, { "grad_norm": 0.17008894681930542, "learning_rate": 9.017797591005268e-05, "loss": 0.0108, "step": 24280 }, { "grad_norm": 0.159986212849617, "learning_rate": 9.016813185732972e-05, "loss": 0.0102, "step": 24290 }, { "grad_norm": 0.2395729273557663, "learning_rate": 9.015828341188027e-05, "loss": 0.0096, "step": 24300 }, { "grad_norm": 0.14036789536476135, "learning_rate": 9.01484305747813e-05, "loss": 0.0096, "step": 24310 }, { "grad_norm": 0.15938794612884521, "learning_rate": 9.013857334711033e-05, "loss": 0.0089, "step": 24320 }, { "grad_norm": 0.1512121856212616, "learning_rate": 9.012871172994534e-05, "loss": 0.0087, "step": 24330 }, { "grad_norm": 0.1256473809480667, "learning_rate": 9.011884572436476e-05, "loss": 0.0072, "step": 24340 }, { "grad_norm": 0.13776539266109467, "learning_rate": 9.010897533144754e-05, "loss": 0.0082, "step": 24350 }, { "grad_norm": 0.13613012433052063, "learning_rate": 9.009910055227306e-05, "loss": 0.0078, "step": 24360 }, { "grad_norm": 0.17274096608161926, "learning_rate": 9.008922138792124e-05, "loss": 0.0088, "step": 24370 }, { "grad_norm": 0.14228080213069916, "learning_rate": 9.007933783947244e-05, "loss": 0.0096, "step": 24380 }, { "grad_norm": 0.233407124876976, "learning_rate": 9.006944990800752e-05, "loss": 0.0097, "step": 24390 }, { "grad_norm": 0.14563706517219543, "learning_rate": 9.005955759460779e-05, "loss": 0.0103, "step": 24400 }, { "grad_norm": 0.15131103992462158, "learning_rate": 9.004966090035508e-05, "loss": 0.0086, "step": 24410 }, { "grad_norm": 0.16293971240520477, "learning_rate": 9.003975982633166e-05, "loss": 0.0088, "step": 24420 }, { "grad_norm": 0.14471088349819183, "learning_rate": 9.00298543736203e-05, "loss": 0.0082, "step": 24430 }, { "grad_norm": 0.15577177703380585, "learning_rate": 9.001994454330427e-05, "loss": 0.0079, "step": 24440 }, { "grad_norm": 0.17007428407669067, "learning_rate": 9.001003033646727e-05, "loss": 0.0075, "step": 24450 }, { "grad_norm": 0.1861376315355301, "learning_rate": 9.00001117541935e-05, "loss": 0.0069, "step": 24460 }, { "grad_norm": 0.20416826009750366, "learning_rate": 8.999018879756764e-05, "loss": 0.0084, "step": 24470 }, { "grad_norm": 0.1401083767414093, "learning_rate": 8.998026146767487e-05, "loss": 0.0074, "step": 24480 }, { "grad_norm": 0.15444660186767578, "learning_rate": 8.99703297656008e-05, "loss": 0.008, "step": 24490 }, { "grad_norm": 0.22016826272010803, "learning_rate": 8.996039369243156e-05, "loss": 0.0079, "step": 24500 }, { "grad_norm": 0.1883348673582077, "learning_rate": 8.995045324925378e-05, "loss": 0.0107, "step": 24510 }, { "grad_norm": 0.1653984934091568, "learning_rate": 8.994050843715448e-05, "loss": 0.0082, "step": 24520 }, { "grad_norm": 0.1356526017189026, "learning_rate": 8.993055925722121e-05, "loss": 0.0095, "step": 24530 }, { "grad_norm": 0.12228257954120636, "learning_rate": 8.992060571054202e-05, "loss": 0.008, "step": 24540 }, { "grad_norm": 0.11155253648757935, "learning_rate": 8.991064779820542e-05, "loss": 0.0078, "step": 24550 }, { "grad_norm": 0.2174905687570572, "learning_rate": 8.990068552130036e-05, "loss": 0.0078, "step": 24560 }, { "grad_norm": 0.17636722326278687, "learning_rate": 8.989071888091634e-05, "loss": 0.0072, "step": 24570 }, { "grad_norm": 0.1439332813024521, "learning_rate": 8.988074787814329e-05, "loss": 0.0085, "step": 24580 }, { "grad_norm": 0.1391952782869339, "learning_rate": 8.987077251407158e-05, "loss": 0.008, "step": 24590 }, { "grad_norm": 0.15085750818252563, "learning_rate": 8.986079278979216e-05, "loss": 0.0082, "step": 24600 }, { "grad_norm": 0.1666959673166275, "learning_rate": 8.985080870639635e-05, "loss": 0.0081, "step": 24610 }, { "grad_norm": 0.1332613080739975, "learning_rate": 8.984082026497603e-05, "loss": 0.0086, "step": 24620 }, { "grad_norm": 0.13018910586833954, "learning_rate": 8.98308274666235e-05, "loss": 0.0081, "step": 24630 }, { "grad_norm": 0.1682562530040741, "learning_rate": 8.982083031243155e-05, "loss": 0.0089, "step": 24640 }, { "grad_norm": 0.15348809957504272, "learning_rate": 8.98108288034935e-05, "loss": 0.0095, "step": 24650 }, { "grad_norm": 0.1435629278421402, "learning_rate": 8.980082294090305e-05, "loss": 0.0072, "step": 24660 }, { "grad_norm": 0.15667319297790527, "learning_rate": 8.979081272575443e-05, "loss": 0.0098, "step": 24670 }, { "grad_norm": 0.11826327443122864, "learning_rate": 8.978079815914236e-05, "loss": 0.0096, "step": 24680 }, { "grad_norm": 0.15555565059185028, "learning_rate": 8.977077924216202e-05, "loss": 0.0085, "step": 24690 }, { "grad_norm": 0.20965053141117096, "learning_rate": 8.976075597590905e-05, "loss": 0.0081, "step": 24700 }, { "grad_norm": 0.1793166548013687, "learning_rate": 8.975072836147958e-05, "loss": 0.0092, "step": 24710 }, { "grad_norm": 0.11650002747774124, "learning_rate": 8.974069639997025e-05, "loss": 0.0087, "step": 24720 }, { "grad_norm": 0.11777836829423904, "learning_rate": 8.973066009247808e-05, "loss": 0.0079, "step": 24730 }, { "grad_norm": 0.17465081810951233, "learning_rate": 8.972061944010066e-05, "loss": 0.008, "step": 24740 }, { "grad_norm": 0.1850181519985199, "learning_rate": 8.971057444393603e-05, "loss": 0.0087, "step": 24750 }, { "grad_norm": 0.15345896780490875, "learning_rate": 8.970052510508268e-05, "loss": 0.0074, "step": 24760 }, { "grad_norm": 0.14842146635055542, "learning_rate": 8.969047142463959e-05, "loss": 0.0079, "step": 24770 }, { "grad_norm": 0.1495223492383957, "learning_rate": 8.968041340370621e-05, "loss": 0.0093, "step": 24780 }, { "grad_norm": 0.10455884039402008, "learning_rate": 8.96703510433825e-05, "loss": 0.0085, "step": 24790 }, { "grad_norm": 0.14209671318531036, "learning_rate": 8.966028434476883e-05, "loss": 0.0086, "step": 24800 }, { "grad_norm": 0.1220848485827446, "learning_rate": 8.96502133089661e-05, "loss": 0.0075, "step": 24810 }, { "grad_norm": 0.1713375449180603, "learning_rate": 8.964013793707564e-05, "loss": 0.0088, "step": 24820 }, { "grad_norm": 0.12051782757043839, "learning_rate": 8.963005823019932e-05, "loss": 0.0083, "step": 24830 }, { "grad_norm": 0.1622859239578247, "learning_rate": 8.961997418943939e-05, "loss": 0.0081, "step": 24840 }, { "grad_norm": 0.11426564306020737, "learning_rate": 8.960988581589865e-05, "loss": 0.0089, "step": 24850 }, { "grad_norm": 0.1611398607492447, "learning_rate": 8.959979311068037e-05, "loss": 0.0082, "step": 24860 }, { "grad_norm": 0.16943195462226868, "learning_rate": 8.958969607488823e-05, "loss": 0.0076, "step": 24870 }, { "grad_norm": 0.14177308976650238, "learning_rate": 8.957959470962647e-05, "loss": 0.0089, "step": 24880 }, { "grad_norm": 0.11231102794408798, "learning_rate": 8.956948901599971e-05, "loss": 0.0082, "step": 24890 }, { "grad_norm": 0.1434362381696701, "learning_rate": 8.955937899511315e-05, "loss": 0.0082, "step": 24900 }, { "grad_norm": 0.17850051820278168, "learning_rate": 8.954926464807238e-05, "loss": 0.0084, "step": 24910 }, { "grad_norm": 0.15144143998622894, "learning_rate": 8.953914597598347e-05, "loss": 0.0093, "step": 24920 }, { "grad_norm": 0.29501914978027344, "learning_rate": 8.952902297995303e-05, "loss": 0.0097, "step": 24930 }, { "grad_norm": 0.16945432126522064, "learning_rate": 8.951889566108804e-05, "loss": 0.0091, "step": 24940 }, { "grad_norm": 0.15754476189613342, "learning_rate": 8.950876402049606e-05, "loss": 0.0077, "step": 24950 }, { "grad_norm": 0.2218533605337143, "learning_rate": 8.949862805928504e-05, "loss": 0.0082, "step": 24960 }, { "grad_norm": 0.20224086940288544, "learning_rate": 8.948848777856343e-05, "loss": 0.0088, "step": 24970 }, { "grad_norm": 0.14087682962417603, "learning_rate": 8.947834317944017e-05, "loss": 0.0083, "step": 24980 }, { "grad_norm": 0.1683323085308075, "learning_rate": 8.946819426302466e-05, "loss": 0.0083, "step": 24990 }, { "grad_norm": 0.19490621984004974, "learning_rate": 8.945804103042676e-05, "loss": 0.0083, "step": 25000 }, { "grad_norm": 0.16727855801582336, "learning_rate": 8.944788348275681e-05, "loss": 0.0087, "step": 25010 }, { "grad_norm": 0.22090843319892883, "learning_rate": 8.943772162112565e-05, "loss": 0.0081, "step": 25020 }, { "grad_norm": 0.1353777050971985, "learning_rate": 8.942755544664454e-05, "loss": 0.0084, "step": 25030 }, { "grad_norm": 0.15036822855472565, "learning_rate": 8.941738496042525e-05, "loss": 0.0094, "step": 25040 }, { "grad_norm": 0.13364489376544952, "learning_rate": 8.940721016357999e-05, "loss": 0.0085, "step": 25050 }, { "grad_norm": 0.13454101979732513, "learning_rate": 8.939703105722148e-05, "loss": 0.0087, "step": 25060 }, { "grad_norm": 0.18703971803188324, "learning_rate": 8.93868476424629e-05, "loss": 0.0091, "step": 25070 }, { "grad_norm": 0.1643713116645813, "learning_rate": 8.937665992041786e-05, "loss": 0.0082, "step": 25080 }, { "grad_norm": 0.1525220423936844, "learning_rate": 8.93664678922005e-05, "loss": 0.009, "step": 25090 }, { "grad_norm": 0.11937591433525085, "learning_rate": 8.93562715589254e-05, "loss": 0.0093, "step": 25100 }, { "grad_norm": 0.13617990911006927, "learning_rate": 8.934607092170762e-05, "loss": 0.0079, "step": 25110 }, { "grad_norm": 0.17603912949562073, "learning_rate": 8.933586598166266e-05, "loss": 0.0076, "step": 25120 }, { "grad_norm": 0.1675972044467926, "learning_rate": 8.932565673990655e-05, "loss": 0.0076, "step": 25130 }, { "grad_norm": 0.16377148032188416, "learning_rate": 8.931544319755574e-05, "loss": 0.0087, "step": 25140 }, { "grad_norm": 0.15126337110996246, "learning_rate": 8.930522535572718e-05, "loss": 0.0077, "step": 25150 }, { "grad_norm": 0.14974018931388855, "learning_rate": 8.929500321553826e-05, "loss": 0.0079, "step": 25160 }, { "grad_norm": 0.13813196122646332, "learning_rate": 8.928477677810686e-05, "loss": 0.0086, "step": 25170 }, { "grad_norm": 0.13397498428821564, "learning_rate": 8.927454604455137e-05, "loss": 0.0077, "step": 25180 }, { "grad_norm": 0.17894554138183594, "learning_rate": 8.926431101599053e-05, "loss": 0.0088, "step": 25190 }, { "grad_norm": 0.13466942310333252, "learning_rate": 8.925407169354369e-05, "loss": 0.0081, "step": 25200 }, { "grad_norm": 0.142214834690094, "learning_rate": 8.92438280783306e-05, "loss": 0.009, "step": 25210 }, { "grad_norm": 0.15861237049102783, "learning_rate": 8.923358017147146e-05, "loss": 0.0074, "step": 25220 }, { "grad_norm": 0.12812714278697968, "learning_rate": 8.922332797408697e-05, "loss": 0.0088, "step": 25230 }, { "grad_norm": 0.1228962242603302, "learning_rate": 8.921307148729831e-05, "loss": 0.0095, "step": 25240 }, { "grad_norm": 0.1960867941379547, "learning_rate": 8.920281071222712e-05, "loss": 0.0072, "step": 25250 }, { "grad_norm": 0.12465748190879822, "learning_rate": 8.919254564999548e-05, "loss": 0.008, "step": 25260 }, { "grad_norm": 0.12997716665267944, "learning_rate": 8.918227630172598e-05, "loss": 0.0074, "step": 25270 }, { "grad_norm": 0.14880435168743134, "learning_rate": 8.917200266854165e-05, "loss": 0.0086, "step": 25280 }, { "grad_norm": 0.17510482668876648, "learning_rate": 8.9161724751566e-05, "loss": 0.0081, "step": 25290 }, { "grad_norm": 0.1638181507587433, "learning_rate": 8.915144255192302e-05, "loss": 0.0077, "step": 25300 }, { "grad_norm": 0.13819944858551025, "learning_rate": 8.914115607073714e-05, "loss": 0.0092, "step": 25310 }, { "grad_norm": 0.16815386712551117, "learning_rate": 8.913086530913327e-05, "loss": 0.0079, "step": 25320 }, { "grad_norm": 0.15536823868751526, "learning_rate": 8.912057026823681e-05, "loss": 0.0079, "step": 25330 }, { "grad_norm": 0.1650146245956421, "learning_rate": 8.91102709491736e-05, "loss": 0.0081, "step": 25340 }, { "grad_norm": 0.17495454847812653, "learning_rate": 8.909996735306996e-05, "loss": 0.0083, "step": 25350 }, { "grad_norm": 0.13479213416576385, "learning_rate": 8.908965948105268e-05, "loss": 0.0098, "step": 25360 }, { "grad_norm": 0.14546601474285126, "learning_rate": 8.907934733424901e-05, "loss": 0.0076, "step": 25370 }, { "grad_norm": 0.17054681479930878, "learning_rate": 8.906903091378666e-05, "loss": 0.0089, "step": 25380 }, { "grad_norm": 0.1639559268951416, "learning_rate": 8.905871022079384e-05, "loss": 0.0089, "step": 25390 }, { "grad_norm": 0.13274939358234406, "learning_rate": 8.90483852563992e-05, "loss": 0.0091, "step": 25400 }, { "grad_norm": 0.1315070539712906, "learning_rate": 8.903805602173185e-05, "loss": 0.0079, "step": 25410 }, { "grad_norm": 0.14126163721084595, "learning_rate": 8.902772251792137e-05, "loss": 0.0079, "step": 25420 }, { "grad_norm": 0.14394767582416534, "learning_rate": 8.901738474609786e-05, "loss": 0.008, "step": 25430 }, { "grad_norm": 0.19075530767440796, "learning_rate": 8.900704270739179e-05, "loss": 0.0081, "step": 25440 }, { "grad_norm": 0.12643659114837646, "learning_rate": 8.89966964029342e-05, "loss": 0.0077, "step": 25450 }, { "grad_norm": 0.16842436790466309, "learning_rate": 8.898634583385652e-05, "loss": 0.0083, "step": 25460 }, { "grad_norm": 0.1677578240633011, "learning_rate": 8.897599100129065e-05, "loss": 0.0086, "step": 25470 }, { "grad_norm": 0.12547354400157928, "learning_rate": 8.896563190636903e-05, "loss": 0.0079, "step": 25480 }, { "grad_norm": 0.17743037641048431, "learning_rate": 8.895526855022448e-05, "loss": 0.0075, "step": 25490 }, { "grad_norm": 0.11454357951879501, "learning_rate": 8.894490093399033e-05, "loss": 0.0078, "step": 25500 }, { "grad_norm": 0.13091816008090973, "learning_rate": 8.893452905880035e-05, "loss": 0.0066, "step": 25510 }, { "grad_norm": 0.13114504516124725, "learning_rate": 8.892415292578883e-05, "loss": 0.0071, "step": 25520 }, { "grad_norm": 0.1464730203151703, "learning_rate": 8.891377253609046e-05, "loss": 0.0081, "step": 25530 }, { "grad_norm": 0.14844109117984772, "learning_rate": 8.890338789084043e-05, "loss": 0.0077, "step": 25540 }, { "grad_norm": 0.1496584564447403, "learning_rate": 8.88929989911744e-05, "loss": 0.008, "step": 25550 }, { "grad_norm": 0.19189856946468353, "learning_rate": 8.888260583822847e-05, "loss": 0.0083, "step": 25560 }, { "grad_norm": 0.13372492790222168, "learning_rate": 8.887220843313921e-05, "loss": 0.0079, "step": 25570 }, { "grad_norm": 0.11388476938009262, "learning_rate": 8.88618067770437e-05, "loss": 0.0084, "step": 25580 }, { "grad_norm": 0.1805977076292038, "learning_rate": 8.885140087107942e-05, "loss": 0.0089, "step": 25590 }, { "grad_norm": 0.13377003371715546, "learning_rate": 8.884099071638436e-05, "loss": 0.0068, "step": 25600 }, { "grad_norm": 0.15977860987186432, "learning_rate": 8.883057631409695e-05, "loss": 0.0081, "step": 25610 }, { "grad_norm": 0.12520799040794373, "learning_rate": 8.882015766535608e-05, "loss": 0.008, "step": 25620 }, { "grad_norm": 0.13959363102912903, "learning_rate": 8.880973477130115e-05, "loss": 0.0074, "step": 25630 }, { "grad_norm": 0.1753527671098709, "learning_rate": 8.879930763307197e-05, "loss": 0.0069, "step": 25640 }, { "grad_norm": 0.16140234470367432, "learning_rate": 8.878887625180884e-05, "loss": 0.0074, "step": 25650 }, { "grad_norm": 0.1131998673081398, "learning_rate": 8.877844062865253e-05, "loss": 0.0074, "step": 25660 }, { "grad_norm": 0.15715248882770538, "learning_rate": 8.876800076474424e-05, "loss": 0.0076, "step": 25670 }, { "grad_norm": 0.10850641876459122, "learning_rate": 8.875755666122568e-05, "loss": 0.0071, "step": 25680 }, { "grad_norm": 0.1387833058834076, "learning_rate": 8.8747108319239e-05, "loss": 0.0086, "step": 25690 }, { "grad_norm": 0.09985536336898804, "learning_rate": 8.87366557399268e-05, "loss": 0.007, "step": 25700 }, { "grad_norm": 0.09391814470291138, "learning_rate": 8.872619892443217e-05, "loss": 0.0073, "step": 25710 }, { "grad_norm": 0.1598110944032669, "learning_rate": 8.871573787389865e-05, "loss": 0.0073, "step": 25720 }, { "grad_norm": 0.1502944827079773, "learning_rate": 8.870527258947024e-05, "loss": 0.0076, "step": 25730 }, { "grad_norm": 0.11343889683485031, "learning_rate": 8.869480307229143e-05, "loss": 0.0077, "step": 25740 }, { "grad_norm": 0.0997815728187561, "learning_rate": 8.868432932350712e-05, "loss": 0.0074, "step": 25750 }, { "grad_norm": 0.1473969668149948, "learning_rate": 8.867385134426272e-05, "loss": 0.0077, "step": 25760 }, { "grad_norm": 0.16320417821407318, "learning_rate": 8.866336913570407e-05, "loss": 0.0088, "step": 25770 }, { "grad_norm": 0.12540525197982788, "learning_rate": 8.865288269897751e-05, "loss": 0.0079, "step": 25780 }, { "grad_norm": 0.11892130225896835, "learning_rate": 8.864239203522981e-05, "loss": 0.0075, "step": 25790 }, { "grad_norm": 0.12485922127962112, "learning_rate": 8.863189714560822e-05, "loss": 0.0075, "step": 25800 }, { "grad_norm": 0.14458946883678436, "learning_rate": 8.862139803126043e-05, "loss": 0.0087, "step": 25810 }, { "grad_norm": 0.18350283801555634, "learning_rate": 8.861089469333463e-05, "loss": 0.0091, "step": 25820 }, { "grad_norm": 0.13274246454238892, "learning_rate": 8.860038713297944e-05, "loss": 0.0083, "step": 25830 }, { "grad_norm": 0.14494988322257996, "learning_rate": 8.858987535134394e-05, "loss": 0.0081, "step": 25840 }, { "grad_norm": 0.10963384062051773, "learning_rate": 8.857935934957769e-05, "loss": 0.0076, "step": 25850 }, { "grad_norm": 0.16727659106254578, "learning_rate": 8.856883912883071e-05, "loss": 0.0083, "step": 25860 }, { "grad_norm": 0.1388489007949829, "learning_rate": 8.855831469025346e-05, "loss": 0.0081, "step": 25870 }, { "grad_norm": 0.15055860579013824, "learning_rate": 8.854778603499689e-05, "loss": 0.0076, "step": 25880 }, { "grad_norm": 0.17883305251598358, "learning_rate": 8.85372531642124e-05, "loss": 0.0073, "step": 25890 }, { "grad_norm": 0.11683842539787292, "learning_rate": 8.852671607905185e-05, "loss": 0.008, "step": 25900 }, { "grad_norm": 0.1309782713651657, "learning_rate": 8.851617478066754e-05, "loss": 0.0073, "step": 25910 }, { "grad_norm": 0.15100020170211792, "learning_rate": 8.850562927021227e-05, "loss": 0.0083, "step": 25920 }, { "grad_norm": 0.16336481273174286, "learning_rate": 8.849507954883928e-05, "loss": 0.0085, "step": 25930 }, { "grad_norm": 0.12507537007331848, "learning_rate": 8.848452561770226e-05, "loss": 0.0092, "step": 25940 }, { "grad_norm": 0.18579131364822388, "learning_rate": 8.847396747795538e-05, "loss": 0.0079, "step": 25950 }, { "grad_norm": 0.17117254436016083, "learning_rate": 8.846340513075327e-05, "loss": 0.0074, "step": 25960 }, { "grad_norm": 0.13435855507850647, "learning_rate": 8.845283857725099e-05, "loss": 0.0083, "step": 25970 }, { "grad_norm": 0.13331136107444763, "learning_rate": 8.844226781860409e-05, "loss": 0.0081, "step": 25980 }, { "grad_norm": 0.0965200737118721, "learning_rate": 8.84316928559686e-05, "loss": 0.0082, "step": 25990 }, { "grad_norm": 0.12848319113254547, "learning_rate": 8.842111369050094e-05, "loss": 0.0088, "step": 26000 }, { "grad_norm": 0.12373735755681992, "learning_rate": 8.841053032335808e-05, "loss": 0.0072, "step": 26010 }, { "grad_norm": 0.14372475445270538, "learning_rate": 8.839994275569735e-05, "loss": 0.0085, "step": 26020 }, { "grad_norm": 0.16122998297214508, "learning_rate": 8.838935098867662e-05, "loss": 0.0078, "step": 26030 }, { "grad_norm": 0.13809846341609955, "learning_rate": 8.837875502345418e-05, "loss": 0.0068, "step": 26040 }, { "grad_norm": 0.10697141289710999, "learning_rate": 8.83681548611888e-05, "loss": 0.0079, "step": 26050 }, { "grad_norm": 0.1156516894698143, "learning_rate": 8.835755050303969e-05, "loss": 0.0075, "step": 26060 }, { "grad_norm": 0.14100581407546997, "learning_rate": 8.834694195016653e-05, "loss": 0.0074, "step": 26070 }, { "grad_norm": 0.16894115507602692, "learning_rate": 8.833632920372942e-05, "loss": 0.0074, "step": 26080 }, { "grad_norm": 0.16566412150859833, "learning_rate": 8.832571226488903e-05, "loss": 0.0081, "step": 26090 }, { "grad_norm": 0.13482648134231567, "learning_rate": 8.831509113480634e-05, "loss": 0.0082, "step": 26100 }, { "grad_norm": 0.13588018715381622, "learning_rate": 8.83044658146429e-05, "loss": 0.0083, "step": 26110 }, { "grad_norm": 0.11667335033416748, "learning_rate": 8.829383630556067e-05, "loss": 0.0086, "step": 26120 }, { "grad_norm": 0.16757740080356598, "learning_rate": 8.828320260872207e-05, "loss": 0.0079, "step": 26130 }, { "grad_norm": 0.12729577720165253, "learning_rate": 8.827256472529e-05, "loss": 0.0079, "step": 26140 }, { "grad_norm": 0.16336241364479065, "learning_rate": 8.826192265642778e-05, "loss": 0.0081, "step": 26150 }, { "grad_norm": 0.19166091084480286, "learning_rate": 8.825127640329923e-05, "loss": 0.0079, "step": 26160 }, { "grad_norm": 0.16278253495693207, "learning_rate": 8.824062596706861e-05, "loss": 0.0093, "step": 26170 }, { "grad_norm": 0.11973316967487335, "learning_rate": 8.822997134890062e-05, "loss": 0.0082, "step": 26180 }, { "grad_norm": 0.17476452887058258, "learning_rate": 8.821931254996044e-05, "loss": 0.0084, "step": 26190 }, { "grad_norm": 0.10918039083480835, "learning_rate": 8.82086495714137e-05, "loss": 0.0096, "step": 26200 }, { "grad_norm": 0.16743890941143036, "learning_rate": 8.81979824144265e-05, "loss": 0.0087, "step": 26210 }, { "grad_norm": 0.1564018875360489, "learning_rate": 8.818731108016536e-05, "loss": 0.0078, "step": 26220 }, { "grad_norm": 0.13045324385166168, "learning_rate": 8.81766355697973e-05, "loss": 0.0084, "step": 26230 }, { "grad_norm": 0.1048620194196701, "learning_rate": 8.816595588448977e-05, "loss": 0.0073, "step": 26240 }, { "grad_norm": 0.19820736348628998, "learning_rate": 8.81552720254107e-05, "loss": 0.0085, "step": 26250 }, { "grad_norm": 0.13257871568202972, "learning_rate": 8.814458399372842e-05, "loss": 0.0078, "step": 26260 }, { "grad_norm": 0.15083110332489014, "learning_rate": 8.813389179061181e-05, "loss": 0.0074, "step": 26270 }, { "grad_norm": 0.12439427524805069, "learning_rate": 8.812319541723012e-05, "loss": 0.0073, "step": 26280 }, { "grad_norm": 0.13075503706932068, "learning_rate": 8.811249487475309e-05, "loss": 0.0082, "step": 26290 }, { "grad_norm": 0.16863104701042175, "learning_rate": 8.810179016435092e-05, "loss": 0.0084, "step": 26300 }, { "grad_norm": 0.17373210191726685, "learning_rate": 8.809108128719428e-05, "loss": 0.0074, "step": 26310 }, { "grad_norm": 0.16726697981357574, "learning_rate": 8.808036824445424e-05, "loss": 0.0082, "step": 26320 }, { "grad_norm": 0.13306967914104462, "learning_rate": 8.806965103730238e-05, "loss": 0.0084, "step": 26330 }, { "grad_norm": 0.18426786363124847, "learning_rate": 8.805892966691074e-05, "loss": 0.0097, "step": 26340 }, { "grad_norm": 0.12152470648288727, "learning_rate": 8.804820413445175e-05, "loss": 0.0075, "step": 26350 }, { "grad_norm": 0.13440097868442535, "learning_rate": 8.803747444109837e-05, "loss": 0.0079, "step": 26360 }, { "grad_norm": 0.1052052304148674, "learning_rate": 8.802674058802399e-05, "loss": 0.0072, "step": 26370 }, { "grad_norm": 0.11313182860612869, "learning_rate": 8.801600257640241e-05, "loss": 0.0076, "step": 26380 }, { "grad_norm": 0.11854977905750275, "learning_rate": 8.800526040740795e-05, "loss": 0.0069, "step": 26390 }, { "grad_norm": 0.1682843267917633, "learning_rate": 8.799451408221535e-05, "loss": 0.0078, "step": 26400 }, { "grad_norm": 0.13672131299972534, "learning_rate": 8.798376360199982e-05, "loss": 0.0077, "step": 26410 }, { "grad_norm": 0.11173857748508453, "learning_rate": 8.797300896793701e-05, "loss": 0.0078, "step": 26420 }, { "grad_norm": 0.14398108422756195, "learning_rate": 8.796225018120302e-05, "loss": 0.0075, "step": 26430 }, { "grad_norm": 0.15845546126365662, "learning_rate": 8.795148724297444e-05, "loss": 0.0073, "step": 26440 }, { "grad_norm": 0.1991671621799469, "learning_rate": 8.794072015442825e-05, "loss": 0.0085, "step": 26450 }, { "grad_norm": 0.1634817272424698, "learning_rate": 8.792994891674198e-05, "loss": 0.0068, "step": 26460 }, { "grad_norm": 0.14310713112354279, "learning_rate": 8.79191735310935e-05, "loss": 0.0082, "step": 26470 }, { "grad_norm": 0.1430201232433319, "learning_rate": 8.790839399866122e-05, "loss": 0.008, "step": 26480 }, { "grad_norm": 0.14020046591758728, "learning_rate": 8.789761032062397e-05, "loss": 0.0071, "step": 26490 }, { "grad_norm": 0.13456101715564728, "learning_rate": 8.788682249816103e-05, "loss": 0.0071, "step": 26500 }, { "grad_norm": 0.18193018436431885, "learning_rate": 8.787603053245215e-05, "loss": 0.0075, "step": 26510 }, { "grad_norm": 0.1244933009147644, "learning_rate": 8.78652344246775e-05, "loss": 0.0095, "step": 26520 }, { "grad_norm": 0.1649419367313385, "learning_rate": 8.785443417601776e-05, "loss": 0.0078, "step": 26530 }, { "grad_norm": 0.10872617363929749, "learning_rate": 8.784362978765401e-05, "loss": 0.0075, "step": 26540 }, { "grad_norm": 0.13876934349536896, "learning_rate": 8.783282126076779e-05, "loss": 0.0069, "step": 26550 }, { "grad_norm": 0.13305112719535828, "learning_rate": 8.782200859654112e-05, "loss": 0.0083, "step": 26560 }, { "grad_norm": 0.13253909349441528, "learning_rate": 8.781119179615646e-05, "loss": 0.0074, "step": 26570 }, { "grad_norm": 0.1336115300655365, "learning_rate": 8.780037086079674e-05, "loss": 0.0061, "step": 26580 }, { "grad_norm": 0.16143354773521423, "learning_rate": 8.778954579164527e-05, "loss": 0.0075, "step": 26590 }, { "grad_norm": 0.1771334409713745, "learning_rate": 8.777871658988588e-05, "loss": 0.0088, "step": 26600 }, { "grad_norm": 0.13656537234783173, "learning_rate": 8.776788325670285e-05, "loss": 0.0078, "step": 26610 }, { "grad_norm": 0.17401467263698578, "learning_rate": 8.775704579328089e-05, "loss": 0.0089, "step": 26620 }, { "grad_norm": 0.1404183804988861, "learning_rate": 8.774620420080517e-05, "loss": 0.0076, "step": 26630 }, { "grad_norm": 0.14471124112606049, "learning_rate": 8.773535848046131e-05, "loss": 0.0076, "step": 26640 }, { "grad_norm": 0.10323827713727951, "learning_rate": 8.772450863343538e-05, "loss": 0.0075, "step": 26650 }, { "grad_norm": 0.12989918887615204, "learning_rate": 8.77136546609139e-05, "loss": 0.0094, "step": 26660 }, { "grad_norm": 0.12913921475410461, "learning_rate": 8.770279656408385e-05, "loss": 0.0074, "step": 26670 }, { "grad_norm": 0.12421948462724686, "learning_rate": 8.769193434413265e-05, "loss": 0.0074, "step": 26680 }, { "grad_norm": 0.15325655043125153, "learning_rate": 8.76810680022482e-05, "loss": 0.0071, "step": 26690 }, { "grad_norm": 0.14071017503738403, "learning_rate": 8.767019753961878e-05, "loss": 0.0079, "step": 26700 }, { "grad_norm": 0.2150823026895523, "learning_rate": 8.765932295743321e-05, "loss": 0.0078, "step": 26710 }, { "grad_norm": 0.16427209973335266, "learning_rate": 8.764844425688068e-05, "loss": 0.0079, "step": 26720 }, { "grad_norm": 0.17675507068634033, "learning_rate": 8.763756143915092e-05, "loss": 0.0089, "step": 26730 }, { "grad_norm": 0.13329023122787476, "learning_rate": 8.7626674505434e-05, "loss": 0.0084, "step": 26740 }, { "grad_norm": 0.15171171724796295, "learning_rate": 8.761578345692053e-05, "loss": 0.0085, "step": 26750 }, { "grad_norm": 0.17949149012565613, "learning_rate": 8.760488829480156e-05, "loss": 0.0103, "step": 26760 }, { "grad_norm": 0.13038291037082672, "learning_rate": 8.759398902026854e-05, "loss": 0.0076, "step": 26770 }, { "grad_norm": 0.11246337741613388, "learning_rate": 8.758308563451339e-05, "loss": 0.008, "step": 26780 }, { "grad_norm": 0.1378759741783142, "learning_rate": 8.75721781387285e-05, "loss": 0.008, "step": 26790 }, { "grad_norm": 0.13235783576965332, "learning_rate": 8.75612665341067e-05, "loss": 0.0076, "step": 26800 }, { "grad_norm": 0.15121938288211823, "learning_rate": 8.755035082184126e-05, "loss": 0.0086, "step": 26810 }, { "grad_norm": 0.13522326946258545, "learning_rate": 8.753943100312592e-05, "loss": 0.0086, "step": 26820 }, { "grad_norm": 0.1752253770828247, "learning_rate": 8.752850707915484e-05, "loss": 0.0082, "step": 26830 }, { "grad_norm": 0.15930940210819244, "learning_rate": 8.751757905112264e-05, "loss": 0.0092, "step": 26840 }, { "grad_norm": 0.17840637266635895, "learning_rate": 8.75066469202244e-05, "loss": 0.0083, "step": 26850 }, { "grad_norm": 0.11059217154979706, "learning_rate": 8.749571068765567e-05, "loss": 0.0084, "step": 26860 }, { "grad_norm": 0.16799266636371613, "learning_rate": 8.748477035461238e-05, "loss": 0.0079, "step": 26870 }, { "grad_norm": 0.10219292342662811, "learning_rate": 8.747382592229095e-05, "loss": 0.0082, "step": 26880 }, { "grad_norm": 0.13539357483386993, "learning_rate": 8.746287739188828e-05, "loss": 0.0082, "step": 26890 }, { "grad_norm": 0.1359178125858307, "learning_rate": 8.745192476460165e-05, "loss": 0.0076, "step": 26900 }, { "grad_norm": 0.09030469506978989, "learning_rate": 8.744096804162882e-05, "loss": 0.0076, "step": 26910 }, { "grad_norm": 0.14469914138317108, "learning_rate": 8.743000722416804e-05, "loss": 0.0074, "step": 26920 }, { "grad_norm": 0.18300847709178925, "learning_rate": 8.741904231341793e-05, "loss": 0.0077, "step": 26930 }, { "grad_norm": 0.15257498621940613, "learning_rate": 8.740807331057762e-05, "loss": 0.0082, "step": 26940 }, { "grad_norm": 0.16667547821998596, "learning_rate": 8.739710021684667e-05, "loss": 0.0084, "step": 26950 }, { "grad_norm": 0.18488825857639313, "learning_rate": 8.738612303342503e-05, "loss": 0.0092, "step": 26960 }, { "grad_norm": 0.17847099900245667, "learning_rate": 8.73751417615132e-05, "loss": 0.0083, "step": 26970 }, { "grad_norm": 0.17925255000591278, "learning_rate": 8.736415640231208e-05, "loss": 0.0073, "step": 26980 }, { "grad_norm": 0.16275613009929657, "learning_rate": 8.735316695702297e-05, "loss": 0.0091, "step": 26990 }, { "grad_norm": 0.1704971045255661, "learning_rate": 8.734217342684769e-05, "loss": 0.0077, "step": 27000 }, { "grad_norm": 0.15271736681461334, "learning_rate": 8.733117581298847e-05, "loss": 0.0079, "step": 27010 }, { "grad_norm": 0.12274172157049179, "learning_rate": 8.732017411664796e-05, "loss": 0.0081, "step": 27020 }, { "grad_norm": 0.15182676911354065, "learning_rate": 8.730916833902936e-05, "loss": 0.0072, "step": 27030 }, { "grad_norm": 0.10561959445476532, "learning_rate": 8.729815848133618e-05, "loss": 0.0068, "step": 27040 }, { "grad_norm": 0.13994552195072174, "learning_rate": 8.728714454477247e-05, "loss": 0.0078, "step": 27050 }, { "grad_norm": 0.13087248802185059, "learning_rate": 8.727612653054269e-05, "loss": 0.0081, "step": 27060 }, { "grad_norm": 0.10348384082317352, "learning_rate": 8.726510443985176e-05, "loss": 0.008, "step": 27070 }, { "grad_norm": 0.1250406801700592, "learning_rate": 8.725407827390503e-05, "loss": 0.0073, "step": 27080 }, { "grad_norm": 0.12451039999723434, "learning_rate": 8.724304803390833e-05, "loss": 0.007, "step": 27090 }, { "grad_norm": 0.14508526027202606, "learning_rate": 8.723201372106788e-05, "loss": 0.0071, "step": 27100 }, { "grad_norm": 0.09704602509737015, "learning_rate": 8.722097533659038e-05, "loss": 0.0067, "step": 27110 }, { "grad_norm": 0.14302432537078857, "learning_rate": 8.720993288168299e-05, "loss": 0.0077, "step": 27120 }, { "grad_norm": 0.12204158306121826, "learning_rate": 8.719888635755327e-05, "loss": 0.0081, "step": 27130 }, { "grad_norm": 0.1362055540084839, "learning_rate": 8.718783576540928e-05, "loss": 0.0075, "step": 27140 }, { "grad_norm": 0.12135270982980728, "learning_rate": 8.717678110645948e-05, "loss": 0.0082, "step": 27150 }, { "grad_norm": 0.1774645894765854, "learning_rate": 8.716572238191279e-05, "loss": 0.0075, "step": 27160 }, { "grad_norm": 0.18263955414295197, "learning_rate": 8.715465959297857e-05, "loss": 0.0078, "step": 27170 }, { "grad_norm": 0.1374088078737259, "learning_rate": 8.714359274086665e-05, "loss": 0.0086, "step": 27180 }, { "grad_norm": 0.1256033331155777, "learning_rate": 8.713252182678726e-05, "loss": 0.0069, "step": 27190 }, { "grad_norm": 0.14040157198905945, "learning_rate": 8.712144685195112e-05, "loss": 0.0074, "step": 27200 }, { "grad_norm": 0.13720540702342987, "learning_rate": 8.711036781756936e-05, "loss": 0.0082, "step": 27210 }, { "grad_norm": 0.16261553764343262, "learning_rate": 8.709928472485357e-05, "loss": 0.0079, "step": 27220 }, { "grad_norm": 0.10236656665802002, "learning_rate": 8.708819757501579e-05, "loss": 0.008, "step": 27230 }, { "grad_norm": 0.11398598551750183, "learning_rate": 8.707710636926846e-05, "loss": 0.0079, "step": 27240 }, { "grad_norm": 0.1087375208735466, "learning_rate": 8.706601110882455e-05, "loss": 0.0071, "step": 27250 }, { "grad_norm": 0.1289469450712204, "learning_rate": 8.705491179489738e-05, "loss": 0.0072, "step": 27260 }, { "grad_norm": 0.10007114708423615, "learning_rate": 8.704380842870077e-05, "loss": 0.0064, "step": 27270 }, { "grad_norm": 0.12485840916633606, "learning_rate": 8.703270101144895e-05, "loss": 0.0072, "step": 27280 }, { "grad_norm": 0.1487235575914383, "learning_rate": 8.702158954435664e-05, "loss": 0.0079, "step": 27290 }, { "grad_norm": 0.12201929837465286, "learning_rate": 8.701047402863896e-05, "loss": 0.0082, "step": 27300 }, { "grad_norm": 0.16330960392951965, "learning_rate": 8.699935446551148e-05, "loss": 0.0084, "step": 27310 }, { "grad_norm": 0.13228438794612885, "learning_rate": 8.698823085619022e-05, "loss": 0.0085, "step": 27320 }, { "grad_norm": 0.19815684854984283, "learning_rate": 8.697710320189166e-05, "loss": 0.0093, "step": 27330 }, { "grad_norm": 0.13061118125915527, "learning_rate": 8.696597150383268e-05, "loss": 0.0086, "step": 27340 }, { "grad_norm": 0.09311086684465408, "learning_rate": 8.695483576323063e-05, "loss": 0.0079, "step": 27350 }, { "grad_norm": 0.156179279088974, "learning_rate": 8.69436959813033e-05, "loss": 0.008, "step": 27360 }, { "grad_norm": 0.14810021221637726, "learning_rate": 8.693255215926892e-05, "loss": 0.0086, "step": 27370 }, { "grad_norm": 0.1533357948064804, "learning_rate": 8.692140429834617e-05, "loss": 0.0087, "step": 27380 }, { "grad_norm": 0.15866902470588684, "learning_rate": 8.691025239975415e-05, "loss": 0.008, "step": 27390 }, { "grad_norm": 0.1315397024154663, "learning_rate": 8.689909646471243e-05, "loss": 0.0071, "step": 27400 }, { "grad_norm": 0.16005051136016846, "learning_rate": 8.688793649444099e-05, "loss": 0.0076, "step": 27410 }, { "grad_norm": 0.12340503185987473, "learning_rate": 8.687677249016029e-05, "loss": 0.0088, "step": 27420 }, { "grad_norm": 0.14079703390598297, "learning_rate": 8.686560445309118e-05, "loss": 0.0073, "step": 27430 }, { "grad_norm": 0.1427820920944214, "learning_rate": 8.685443238445499e-05, "loss": 0.0077, "step": 27440 }, { "grad_norm": 0.19003134965896606, "learning_rate": 8.68432562854735e-05, "loss": 0.0087, "step": 27450 }, { "grad_norm": 0.14652332663536072, "learning_rate": 8.683207615736887e-05, "loss": 0.009, "step": 27460 }, { "grad_norm": 0.13217362761497498, "learning_rate": 8.682089200136379e-05, "loss": 0.0072, "step": 27470 }, { "grad_norm": 0.15331730246543884, "learning_rate": 8.680970381868132e-05, "loss": 0.0086, "step": 27480 }, { "grad_norm": 0.14591416716575623, "learning_rate": 8.679851161054498e-05, "loss": 0.0075, "step": 27490 }, { "grad_norm": 0.1536940038204193, "learning_rate": 8.678731537817873e-05, "loss": 0.0079, "step": 27500 }, { "grad_norm": 0.1475682407617569, "learning_rate": 8.677611512280697e-05, "loss": 0.0083, "step": 27510 }, { "grad_norm": 0.1499038189649582, "learning_rate": 8.676491084565457e-05, "loss": 0.0083, "step": 27520 }, { "grad_norm": 0.1114799901843071, "learning_rate": 8.675370254794678e-05, "loss": 0.0071, "step": 27530 }, { "grad_norm": 0.1594182550907135, "learning_rate": 8.674249023090935e-05, "loss": 0.0094, "step": 27540 }, { "grad_norm": 0.13529030978679657, "learning_rate": 8.673127389576843e-05, "loss": 0.0083, "step": 27550 }, { "grad_norm": 0.16672037541866302, "learning_rate": 8.67200535437506e-05, "loss": 0.0091, "step": 27560 }, { "grad_norm": 0.1715506613254547, "learning_rate": 8.670882917608296e-05, "loss": 0.0073, "step": 27570 }, { "grad_norm": 0.13231390714645386, "learning_rate": 8.669760079399292e-05, "loss": 0.008, "step": 27580 }, { "grad_norm": 0.16979984939098358, "learning_rate": 8.668636839870845e-05, "loss": 0.0092, "step": 27590 }, { "grad_norm": 0.15412236750125885, "learning_rate": 8.667513199145789e-05, "loss": 0.0101, "step": 27600 }, { "grad_norm": 0.17395809292793274, "learning_rate": 8.666389157347002e-05, "loss": 0.0087, "step": 27610 }, { "grad_norm": 0.15584135055541992, "learning_rate": 8.66526471459741e-05, "loss": 0.0079, "step": 27620 }, { "grad_norm": 0.1519521027803421, "learning_rate": 8.66413987101998e-05, "loss": 0.0082, "step": 27630 }, { "grad_norm": 0.1161729097366333, "learning_rate": 8.663014626737723e-05, "loss": 0.0086, "step": 27640 }, { "grad_norm": 0.14099548757076263, "learning_rate": 8.661888981873691e-05, "loss": 0.0084, "step": 27650 }, { "grad_norm": 0.14624129235744476, "learning_rate": 8.660762936550988e-05, "loss": 0.009, "step": 27660 }, { "grad_norm": 0.1552497297525406, "learning_rate": 8.659636490892753e-05, "loss": 0.0082, "step": 27670 }, { "grad_norm": 0.1443503350019455, "learning_rate": 8.658509645022174e-05, "loss": 0.0103, "step": 27680 }, { "grad_norm": 0.1993781328201294, "learning_rate": 8.657382399062481e-05, "loss": 0.0088, "step": 27690 }, { "grad_norm": 0.17977629601955414, "learning_rate": 8.656254753136946e-05, "loss": 0.0097, "step": 27700 }, { "grad_norm": 0.1529873162508011, "learning_rate": 8.655126707368891e-05, "loss": 0.0092, "step": 27710 }, { "grad_norm": 0.12224581837654114, "learning_rate": 8.653998261881672e-05, "loss": 0.0075, "step": 27720 }, { "grad_norm": 0.16351446509361267, "learning_rate": 8.652869416798699e-05, "loss": 0.0079, "step": 27730 }, { "grad_norm": 0.14394736289978027, "learning_rate": 8.651740172243417e-05, "loss": 0.0084, "step": 27740 }, { "grad_norm": 0.12375658750534058, "learning_rate": 8.65061052833932e-05, "loss": 0.0084, "step": 27750 }, { "grad_norm": 0.16905707120895386, "learning_rate": 8.649480485209945e-05, "loss": 0.0075, "step": 27760 }, { "grad_norm": 0.13579285144805908, "learning_rate": 8.64835004297887e-05, "loss": 0.0075, "step": 27770 }, { "grad_norm": 0.12203359603881836, "learning_rate": 8.64721920176972e-05, "loss": 0.0074, "step": 27780 }, { "grad_norm": 0.12588268518447876, "learning_rate": 8.646087961706164e-05, "loss": 0.0086, "step": 27790 }, { "grad_norm": 0.13499648869037628, "learning_rate": 8.644956322911908e-05, "loss": 0.0079, "step": 27800 }, { "grad_norm": 0.17560259997844696, "learning_rate": 8.643824285510709e-05, "loss": 0.0096, "step": 27810 }, { "grad_norm": 0.12819072604179382, "learning_rate": 8.642691849626364e-05, "loss": 0.0076, "step": 27820 }, { "grad_norm": 0.12453342974185944, "learning_rate": 8.641559015382717e-05, "loss": 0.0075, "step": 27830 }, { "grad_norm": 0.13140557706356049, "learning_rate": 8.640425782903649e-05, "loss": 0.0084, "step": 27840 }, { "grad_norm": 0.14756855368614197, "learning_rate": 8.639292152313091e-05, "loss": 0.0077, "step": 27850 }, { "grad_norm": 0.12422213703393936, "learning_rate": 8.638158123735015e-05, "loss": 0.0078, "step": 27860 }, { "grad_norm": 0.11814700812101364, "learning_rate": 8.637023697293436e-05, "loss": 0.0096, "step": 27870 }, { "grad_norm": 0.10796556621789932, "learning_rate": 8.635888873112414e-05, "loss": 0.0076, "step": 27880 }, { "grad_norm": 0.1155673936009407, "learning_rate": 8.634753651316052e-05, "loss": 0.0071, "step": 27890 }, { "grad_norm": 0.1062009260058403, "learning_rate": 8.633618032028496e-05, "loss": 0.0075, "step": 27900 }, { "grad_norm": 0.17179283499717712, "learning_rate": 8.632482015373934e-05, "loss": 0.0067, "step": 27910 }, { "grad_norm": 0.15015096962451935, "learning_rate": 8.6313456014766e-05, "loss": 0.0078, "step": 27920 }, { "grad_norm": 0.139568492770195, "learning_rate": 8.630208790460771e-05, "loss": 0.0069, "step": 27930 }, { "grad_norm": 0.15317389369010925, "learning_rate": 8.629071582450768e-05, "loss": 0.0076, "step": 27940 }, { "grad_norm": 0.12648728489875793, "learning_rate": 8.62793397757095e-05, "loss": 0.0071, "step": 27950 }, { "grad_norm": 0.12229068577289581, "learning_rate": 8.626795975945729e-05, "loss": 0.0074, "step": 27960 }, { "grad_norm": 0.13552813231945038, "learning_rate": 8.625657577699551e-05, "loss": 0.0072, "step": 27970 }, { "grad_norm": 0.15231432020664215, "learning_rate": 8.624518782956914e-05, "loss": 0.0076, "step": 27980 }, { "grad_norm": 0.10863625258207321, "learning_rate": 8.62337959184235e-05, "loss": 0.0067, "step": 27990 }, { "grad_norm": 0.09993485361337662, "learning_rate": 8.622240004480441e-05, "loss": 0.007, "step": 28000 }, { "grad_norm": 0.15676292777061462, "learning_rate": 8.621100020995814e-05, "loss": 0.0073, "step": 28010 }, { "grad_norm": 0.17116820812225342, "learning_rate": 8.619959641513132e-05, "loss": 0.0082, "step": 28020 }, { "grad_norm": 0.16126638650894165, "learning_rate": 8.618818866157105e-05, "loss": 0.0083, "step": 28030 }, { "grad_norm": 0.10411401838064194, "learning_rate": 8.617677695052487e-05, "loss": 0.0079, "step": 28040 }, { "grad_norm": 0.12381000816822052, "learning_rate": 8.616536128324078e-05, "loss": 0.0088, "step": 28050 }, { "grad_norm": 0.16428142786026, "learning_rate": 8.615394166096712e-05, "loss": 0.008, "step": 28060 }, { "grad_norm": 0.1641220897436142, "learning_rate": 8.614251808495279e-05, "loss": 0.0085, "step": 28070 }, { "grad_norm": 0.1585456132888794, "learning_rate": 8.6131090556447e-05, "loss": 0.0079, "step": 28080 }, { "grad_norm": 0.13337475061416626, "learning_rate": 8.611965907669947e-05, "loss": 0.0073, "step": 28090 }, { "grad_norm": 0.148769348859787, "learning_rate": 8.610822364696034e-05, "loss": 0.0073, "step": 28100 }, { "grad_norm": 0.18941718339920044, "learning_rate": 8.609678426848015e-05, "loss": 0.0092, "step": 28110 }, { "grad_norm": 0.13679680228233337, "learning_rate": 8.60853409425099e-05, "loss": 0.0079, "step": 28120 }, { "grad_norm": 0.1444157510995865, "learning_rate": 8.607389367030104e-05, "loss": 0.008, "step": 28130 }, { "grad_norm": 0.13691595196723938, "learning_rate": 8.606244245310538e-05, "loss": 0.0084, "step": 28140 }, { "grad_norm": 0.16340813040733337, "learning_rate": 8.605098729217525e-05, "loss": 0.0079, "step": 28150 }, { "grad_norm": 0.17737796902656555, "learning_rate": 8.603952818876335e-05, "loss": 0.01, "step": 28160 }, { "grad_norm": 0.11908183991909027, "learning_rate": 8.602806514412281e-05, "loss": 0.0075, "step": 28170 }, { "grad_norm": 0.1506173461675644, "learning_rate": 8.601659815950726e-05, "loss": 0.0077, "step": 28180 }, { "grad_norm": 0.12745793163776398, "learning_rate": 8.600512723617067e-05, "loss": 0.0072, "step": 28190 }, { "grad_norm": 0.13162916898727417, "learning_rate": 8.59936523753675e-05, "loss": 0.0088, "step": 28200 }, { "grad_norm": 0.13166114687919617, "learning_rate": 8.598217357835264e-05, "loss": 0.0079, "step": 28210 }, { "grad_norm": 0.14714224636554718, "learning_rate": 8.597069084638135e-05, "loss": 0.0086, "step": 28220 }, { "grad_norm": 0.12168154120445251, "learning_rate": 8.595920418070939e-05, "loss": 0.0076, "step": 28230 }, { "grad_norm": 0.11263018846511841, "learning_rate": 8.594771358259295e-05, "loss": 0.0067, "step": 28240 }, { "grad_norm": 0.1452581137418747, "learning_rate": 8.593621905328858e-05, "loss": 0.0068, "step": 28250 }, { "grad_norm": 0.17834822833538055, "learning_rate": 8.592472059405333e-05, "loss": 0.007, "step": 28260 }, { "grad_norm": 0.14355655014514923, "learning_rate": 8.591321820614464e-05, "loss": 0.0086, "step": 28270 }, { "grad_norm": 0.15324264764785767, "learning_rate": 8.590171189082041e-05, "loss": 0.0071, "step": 28280 }, { "grad_norm": 0.16897420585155487, "learning_rate": 8.589020164933894e-05, "loss": 0.0073, "step": 28290 }, { "grad_norm": 0.1442059576511383, "learning_rate": 8.587868748295898e-05, "loss": 0.0071, "step": 28300 }, { "grad_norm": 0.10639975965023041, "learning_rate": 8.586716939293971e-05, "loss": 0.0069, "step": 28310 }, { "grad_norm": 0.1358734518289566, "learning_rate": 8.58556473805407e-05, "loss": 0.007, "step": 28320 }, { "grad_norm": 0.16487981379032135, "learning_rate": 8.584412144702202e-05, "loss": 0.0081, "step": 28330 }, { "grad_norm": 0.13734032213687897, "learning_rate": 8.58325915936441e-05, "loss": 0.0083, "step": 28340 }, { "grad_norm": 0.1993238478899002, "learning_rate": 8.582105782166783e-05, "loss": 0.0076, "step": 28350 }, { "grad_norm": 0.18159916996955872, "learning_rate": 8.580952013235455e-05, "loss": 0.0093, "step": 28360 }, { "grad_norm": 0.16298042237758636, "learning_rate": 8.579797852696596e-05, "loss": 0.0087, "step": 28370 }, { "grad_norm": 0.12838385999202728, "learning_rate": 8.578643300676428e-05, "loss": 0.0079, "step": 28380 }, { "grad_norm": 0.1367899775505066, "learning_rate": 8.577488357301209e-05, "loss": 0.0084, "step": 28390 }, { "grad_norm": 0.1294841170310974, "learning_rate": 8.576333022697242e-05, "loss": 0.0082, "step": 28400 }, { "grad_norm": 0.1871757060289383, "learning_rate": 8.575177296990873e-05, "loss": 0.0078, "step": 28410 }, { "grad_norm": 0.149122953414917, "learning_rate": 8.574021180308489e-05, "loss": 0.0077, "step": 28420 }, { "grad_norm": 0.21082761883735657, "learning_rate": 8.572864672776523e-05, "loss": 0.0081, "step": 28430 }, { "grad_norm": 0.16954901814460754, "learning_rate": 8.571707774521447e-05, "loss": 0.0085, "step": 28440 }, { "grad_norm": 0.10935863107442856, "learning_rate": 8.57055048566978e-05, "loss": 0.0076, "step": 28450 }, { "grad_norm": 0.14645960927009583, "learning_rate": 8.569392806348078e-05, "loss": 0.0072, "step": 28460 }, { "grad_norm": 0.08904128521680832, "learning_rate": 8.568234736682947e-05, "loss": 0.0073, "step": 28470 }, { "grad_norm": 0.10507151484489441, "learning_rate": 8.567076276801029e-05, "loss": 0.0072, "step": 28480 }, { "grad_norm": 0.17946937680244446, "learning_rate": 8.565917426829013e-05, "loss": 0.0076, "step": 28490 }, { "grad_norm": 0.15126672387123108, "learning_rate": 8.564758186893628e-05, "loss": 0.0074, "step": 28500 }, { "grad_norm": 0.14836737513542175, "learning_rate": 8.563598557121649e-05, "loss": 0.0071, "step": 28510 }, { "grad_norm": 0.14126166701316833, "learning_rate": 8.562438537639888e-05, "loss": 0.0072, "step": 28520 }, { "grad_norm": 0.17398639023303986, "learning_rate": 8.561278128575206e-05, "loss": 0.0066, "step": 28530 }, { "grad_norm": 0.13990390300750732, "learning_rate": 8.5601173300545e-05, "loss": 0.0072, "step": 28540 }, { "grad_norm": 0.11567962169647217, "learning_rate": 8.558956142204717e-05, "loss": 0.0071, "step": 28550 }, { "grad_norm": 0.13875697553157806, "learning_rate": 8.55779456515284e-05, "loss": 0.0073, "step": 28560 }, { "grad_norm": 0.1324213594198227, "learning_rate": 8.556632599025898e-05, "loss": 0.0067, "step": 28570 }, { "grad_norm": 0.13843007385730743, "learning_rate": 8.555470243950964e-05, "loss": 0.0075, "step": 28580 }, { "grad_norm": 0.10165523737668991, "learning_rate": 8.554307500055148e-05, "loss": 0.0066, "step": 28590 }, { "grad_norm": 0.12807458639144897, "learning_rate": 8.553144367465609e-05, "loss": 0.0068, "step": 28600 }, { "grad_norm": 0.14452458918094635, "learning_rate": 8.551980846309544e-05, "loss": 0.0074, "step": 28610 }, { "grad_norm": 0.21114321053028107, "learning_rate": 8.550816936714193e-05, "loss": 0.009, "step": 28620 }, { "grad_norm": 0.1272326558828354, "learning_rate": 8.549652638806841e-05, "loss": 0.0081, "step": 28630 }, { "grad_norm": 0.119441457092762, "learning_rate": 8.548487952714812e-05, "loss": 0.0073, "step": 28640 }, { "grad_norm": 0.11183171719312668, "learning_rate": 8.547322878565478e-05, "loss": 0.0063, "step": 28650 }, { "grad_norm": 0.12269280850887299, "learning_rate": 8.546157416486245e-05, "loss": 0.0077, "step": 28660 }, { "grad_norm": 0.13262729346752167, "learning_rate": 8.54499156660457e-05, "loss": 0.007, "step": 28670 }, { "grad_norm": 0.18941143155097961, "learning_rate": 8.543825329047947e-05, "loss": 0.0087, "step": 28680 }, { "grad_norm": 0.15724332630634308, "learning_rate": 8.542658703943913e-05, "loss": 0.0082, "step": 28690 }, { "grad_norm": 0.11051396280527115, "learning_rate": 8.541491691420051e-05, "loss": 0.0076, "step": 28700 }, { "grad_norm": 0.1768072098493576, "learning_rate": 8.54032429160398e-05, "loss": 0.0078, "step": 28710 }, { "grad_norm": 0.17498603463172913, "learning_rate": 8.539156504623369e-05, "loss": 0.0076, "step": 28720 }, { "grad_norm": 0.15346749126911163, "learning_rate": 8.537988330605923e-05, "loss": 0.0071, "step": 28730 }, { "grad_norm": 0.11536207795143127, "learning_rate": 8.536819769679393e-05, "loss": 0.0069, "step": 28740 }, { "grad_norm": 0.11268126219511032, "learning_rate": 8.53565082197157e-05, "loss": 0.0068, "step": 28750 }, { "grad_norm": 0.1364465057849884, "learning_rate": 8.534481487610289e-05, "loss": 0.0065, "step": 28760 }, { "grad_norm": 0.14567922055721283, "learning_rate": 8.533311766723428e-05, "loss": 0.007, "step": 28770 }, { "grad_norm": 0.13000105321407318, "learning_rate": 8.532141659438901e-05, "loss": 0.0074, "step": 28780 }, { "grad_norm": 0.16071349382400513, "learning_rate": 8.530971165884675e-05, "loss": 0.0062, "step": 28790 }, { "grad_norm": 0.12081096321344376, "learning_rate": 8.529800286188752e-05, "loss": 0.007, "step": 28800 }, { "grad_norm": 0.15392830967903137, "learning_rate": 8.528629020479175e-05, "loss": 0.0085, "step": 28810 }, { "grad_norm": 0.11589876562356949, "learning_rate": 8.527457368884033e-05, "loss": 0.0083, "step": 28820 }, { "grad_norm": 0.0857432410120964, "learning_rate": 8.526285331531458e-05, "loss": 0.0072, "step": 28830 }, { "grad_norm": 0.17145630717277527, "learning_rate": 8.525112908549621e-05, "loss": 0.0076, "step": 28840 }, { "grad_norm": 0.15725861489772797, "learning_rate": 8.523940100066735e-05, "loss": 0.0077, "step": 28850 }, { "grad_norm": 0.14344920217990875, "learning_rate": 8.52276690621106e-05, "loss": 0.0066, "step": 28860 }, { "grad_norm": 0.12946847081184387, "learning_rate": 8.521593327110889e-05, "loss": 0.007, "step": 28870 }, { "grad_norm": 0.1334114968776703, "learning_rate": 8.520419362894569e-05, "loss": 0.0071, "step": 28880 }, { "grad_norm": 0.12055898457765579, "learning_rate": 8.51924501369048e-05, "loss": 0.0061, "step": 28890 }, { "grad_norm": 0.1656540334224701, "learning_rate": 8.518070279627047e-05, "loss": 0.0075, "step": 28900 }, { "grad_norm": 0.12991204857826233, "learning_rate": 8.516895160832737e-05, "loss": 0.0061, "step": 28910 }, { "grad_norm": 0.1008758470416069, "learning_rate": 8.515719657436061e-05, "loss": 0.0066, "step": 28920 }, { "grad_norm": 0.12977322936058044, "learning_rate": 8.514543769565568e-05, "loss": 0.0072, "step": 28930 }, { "grad_norm": 0.16574089229106903, "learning_rate": 8.513367497349853e-05, "loss": 0.0061, "step": 28940 }, { "grad_norm": 0.14490680396556854, "learning_rate": 8.51219084091755e-05, "loss": 0.0077, "step": 28950 }, { "grad_norm": 0.1135653406381607, "learning_rate": 8.511013800397338e-05, "loss": 0.0067, "step": 28960 }, { "grad_norm": 0.08651597797870636, "learning_rate": 8.509836375917937e-05, "loss": 0.0062, "step": 28970 }, { "grad_norm": 0.10844499617815018, "learning_rate": 8.508658567608104e-05, "loss": 0.0063, "step": 28980 }, { "grad_norm": 0.12821407616138458, "learning_rate": 8.507480375596647e-05, "loss": 0.0067, "step": 28990 }, { "grad_norm": 0.06379446387290955, "learning_rate": 8.506301800012408e-05, "loss": 0.0065, "step": 29000 }, { "grad_norm": 0.12304133176803589, "learning_rate": 8.505122840984278e-05, "loss": 0.0061, "step": 29010 }, { "grad_norm": 0.11828873306512833, "learning_rate": 8.503943498641182e-05, "loss": 0.006, "step": 29020 }, { "grad_norm": 0.15516303479671478, "learning_rate": 8.502763773112095e-05, "loss": 0.0077, "step": 29030 }, { "grad_norm": 0.13738249242305756, "learning_rate": 8.501583664526026e-05, "loss": 0.0072, "step": 29040 }, { "grad_norm": 0.10618148744106293, "learning_rate": 8.500403173012032e-05, "loss": 0.0077, "step": 29050 }, { "grad_norm": 0.16776059567928314, "learning_rate": 8.499222298699211e-05, "loss": 0.0064, "step": 29060 }, { "grad_norm": 0.14552220702171326, "learning_rate": 8.498041041716701e-05, "loss": 0.007, "step": 29070 }, { "grad_norm": 0.12928874790668488, "learning_rate": 8.496859402193681e-05, "loss": 0.0071, "step": 29080 }, { "grad_norm": 0.10197131335735321, "learning_rate": 8.495677380259374e-05, "loss": 0.0072, "step": 29090 }, { "grad_norm": 0.16821448504924774, "learning_rate": 8.494494976043045e-05, "loss": 0.0094, "step": 29100 }, { "grad_norm": 0.13312409818172455, "learning_rate": 8.493312189673998e-05, "loss": 0.0087, "step": 29110 }, { "grad_norm": 0.13442139327526093, "learning_rate": 8.492129021281584e-05, "loss": 0.0074, "step": 29120 }, { "grad_norm": 0.14908969402313232, "learning_rate": 8.490945470995188e-05, "loss": 0.0079, "step": 29130 }, { "grad_norm": 0.12934087216854095, "learning_rate": 8.489761538944247e-05, "loss": 0.0065, "step": 29140 }, { "grad_norm": 0.15128228068351746, "learning_rate": 8.48857722525823e-05, "loss": 0.009, "step": 29150 }, { "grad_norm": 0.1558387130498886, "learning_rate": 8.487392530066652e-05, "loss": 0.0083, "step": 29160 }, { "grad_norm": 0.1325654238462448, "learning_rate": 8.486207453499069e-05, "loss": 0.0073, "step": 29170 }, { "grad_norm": 0.1438884288072586, "learning_rate": 8.485021995685082e-05, "loss": 0.0072, "step": 29180 }, { "grad_norm": 0.16736945509910583, "learning_rate": 8.483836156754328e-05, "loss": 0.0077, "step": 29190 }, { "grad_norm": 0.15048682689666748, "learning_rate": 8.482649936836491e-05, "loss": 0.0073, "step": 29200 }, { "grad_norm": 0.10931472480297089, "learning_rate": 8.481463336061293e-05, "loss": 0.0073, "step": 29210 }, { "grad_norm": 0.12344759702682495, "learning_rate": 8.480276354558496e-05, "loss": 0.008, "step": 29220 }, { "grad_norm": 0.2846308946609497, "learning_rate": 8.479088992457913e-05, "loss": 0.0085, "step": 29230 }, { "grad_norm": 0.1412172019481659, "learning_rate": 8.477901249889387e-05, "loss": 0.0097, "step": 29240 }, { "grad_norm": 0.15631601214408875, "learning_rate": 8.47671312698281e-05, "loss": 0.0081, "step": 29250 }, { "grad_norm": 0.13835377991199493, "learning_rate": 8.475524623868112e-05, "loss": 0.0075, "step": 29260 }, { "grad_norm": 0.15253086388111115, "learning_rate": 8.474335740675266e-05, "loss": 0.0092, "step": 29270 }, { "grad_norm": 0.19671227037906647, "learning_rate": 8.473146477534289e-05, "loss": 0.0083, "step": 29280 }, { "grad_norm": 0.14138169586658478, "learning_rate": 8.471956834575232e-05, "loss": 0.0093, "step": 29290 }, { "grad_norm": 0.19989930093288422, "learning_rate": 8.470766811928197e-05, "loss": 0.0077, "step": 29300 }, { "grad_norm": 0.15125343203544617, "learning_rate": 8.469576409723323e-05, "loss": 0.0077, "step": 29310 }, { "grad_norm": 0.11010654270648956, "learning_rate": 8.468385628090788e-05, "loss": 0.0071, "step": 29320 }, { "grad_norm": 0.13006815314292908, "learning_rate": 8.467194467160815e-05, "loss": 0.0066, "step": 29330 }, { "grad_norm": 0.14966726303100586, "learning_rate": 8.466002927063667e-05, "loss": 0.0082, "step": 29340 }, { "grad_norm": 0.1300755888223648, "learning_rate": 8.464811007929651e-05, "loss": 0.007, "step": 29350 }, { "grad_norm": 0.09455937147140503, "learning_rate": 8.463618709889114e-05, "loss": 0.0072, "step": 29360 }, { "grad_norm": 0.10834109783172607, "learning_rate": 8.462426033072442e-05, "loss": 0.0068, "step": 29370 }, { "grad_norm": 0.1296132653951645, "learning_rate": 8.461232977610061e-05, "loss": 0.0074, "step": 29380 }, { "grad_norm": 0.150991290807724, "learning_rate": 8.46003954363245e-05, "loss": 0.0076, "step": 29390 }, { "grad_norm": 0.1391124427318573, "learning_rate": 8.458845731270115e-05, "loss": 0.0093, "step": 29400 }, { "grad_norm": 0.12769164144992828, "learning_rate": 8.45765154065361e-05, "loss": 0.0083, "step": 29410 }, { "grad_norm": 0.12336011976003647, "learning_rate": 8.456456971913532e-05, "loss": 0.0073, "step": 29420 }, { "grad_norm": 0.15719464421272278, "learning_rate": 8.455262025180517e-05, "loss": 0.0075, "step": 29430 }, { "grad_norm": 0.1301058828830719, "learning_rate": 8.454066700585242e-05, "loss": 0.0076, "step": 29440 }, { "grad_norm": 0.12295348942279816, "learning_rate": 8.452870998258423e-05, "loss": 0.0079, "step": 29450 }, { "grad_norm": 0.12385734915733337, "learning_rate": 8.451674918330825e-05, "loss": 0.0076, "step": 29460 }, { "grad_norm": 0.2337786704301834, "learning_rate": 8.450478460933246e-05, "loss": 0.0085, "step": 29470 }, { "grad_norm": 0.12770624458789825, "learning_rate": 8.449281626196532e-05, "loss": 0.0077, "step": 29480 }, { "grad_norm": 0.12381754070520401, "learning_rate": 8.448084414251564e-05, "loss": 0.0065, "step": 29490 }, { "grad_norm": 0.1795097291469574, "learning_rate": 8.446886825229271e-05, "loss": 0.0068, "step": 29500 }, { "grad_norm": 0.14774173498153687, "learning_rate": 8.445688859260615e-05, "loss": 0.0079, "step": 29510 }, { "grad_norm": 0.16271081566810608, "learning_rate": 8.444490516476606e-05, "loss": 0.0081, "step": 29520 }, { "grad_norm": 0.1373639553785324, "learning_rate": 8.443291797008293e-05, "loss": 0.0072, "step": 29530 }, { "grad_norm": 0.1521013230085373, "learning_rate": 8.442092700986765e-05, "loss": 0.0076, "step": 29540 }, { "grad_norm": 0.12646295130252838, "learning_rate": 8.440893228543156e-05, "loss": 0.0064, "step": 29550 }, { "grad_norm": 0.12290258705615997, "learning_rate": 8.439693379808638e-05, "loss": 0.0066, "step": 29560 }, { "grad_norm": 0.12623745203018188, "learning_rate": 8.43849315491442e-05, "loss": 0.0072, "step": 29570 }, { "grad_norm": 0.1329209953546524, "learning_rate": 8.437292553991763e-05, "loss": 0.0062, "step": 29580 }, { "grad_norm": 0.16310182213783264, "learning_rate": 8.436091577171959e-05, "loss": 0.0071, "step": 29590 }, { "grad_norm": 0.176676407456398, "learning_rate": 8.434890224586347e-05, "loss": 0.0095, "step": 29600 }, { "grad_norm": 0.11622826009988785, "learning_rate": 8.433688496366303e-05, "loss": 0.0073, "step": 29610 }, { "grad_norm": 0.13570396602153778, "learning_rate": 8.432486392643248e-05, "loss": 0.006, "step": 29620 }, { "grad_norm": 0.11845974624156952, "learning_rate": 8.431283913548643e-05, "loss": 0.0065, "step": 29630 }, { "grad_norm": 0.1342480480670929, "learning_rate": 8.430081059213985e-05, "loss": 0.0077, "step": 29640 }, { "grad_norm": 0.09965535253286362, "learning_rate": 8.428877829770823e-05, "loss": 0.0072, "step": 29650 }, { "grad_norm": 0.10912018269300461, "learning_rate": 8.427674225350735e-05, "loss": 0.0071, "step": 29660 }, { "grad_norm": 0.15249763429164886, "learning_rate": 8.426470246085347e-05, "loss": 0.0077, "step": 29670 }, { "grad_norm": 0.1384955495595932, "learning_rate": 8.425265892106324e-05, "loss": 0.0082, "step": 29680 }, { "grad_norm": 0.12903928756713867, "learning_rate": 8.424061163545374e-05, "loss": 0.0066, "step": 29690 }, { "grad_norm": 0.11429282277822495, "learning_rate": 8.422856060534243e-05, "loss": 0.0066, "step": 29700 }, { "grad_norm": 0.12619264423847198, "learning_rate": 8.421650583204718e-05, "loss": 0.0082, "step": 29710 }, { "grad_norm": 0.09101539850234985, "learning_rate": 8.420444731688633e-05, "loss": 0.0065, "step": 29720 }, { "grad_norm": 0.1521563082933426, "learning_rate": 8.419238506117852e-05, "loss": 0.0065, "step": 29730 }, { "grad_norm": 0.13761566579341888, "learning_rate": 8.418031906624289e-05, "loss": 0.0081, "step": 29740 }, { "grad_norm": 0.13803337514400482, "learning_rate": 8.416824933339898e-05, "loss": 0.0075, "step": 29750 }, { "grad_norm": 0.14771395921707153, "learning_rate": 8.415617586396667e-05, "loss": 0.0069, "step": 29760 }, { "grad_norm": 0.1342165619134903, "learning_rate": 8.414409865926632e-05, "loss": 0.0077, "step": 29770 }, { "grad_norm": 0.13623952865600586, "learning_rate": 8.413201772061867e-05, "loss": 0.0079, "step": 29780 }, { "grad_norm": 0.14887875318527222, "learning_rate": 8.411993304934488e-05, "loss": 0.0079, "step": 29790 }, { "grad_norm": 0.17352059483528137, "learning_rate": 8.410784464676654e-05, "loss": 0.0071, "step": 29800 }, { "grad_norm": 0.1534147709608078, "learning_rate": 8.409575251420556e-05, "loss": 0.0076, "step": 29810 }, { "grad_norm": 0.10850135236978531, "learning_rate": 8.408365665298435e-05, "loss": 0.0071, "step": 29820 }, { "grad_norm": 0.1515093296766281, "learning_rate": 8.40715570644257e-05, "loss": 0.0075, "step": 29830 }, { "grad_norm": 0.1371680647134781, "learning_rate": 8.40594537498528e-05, "loss": 0.0065, "step": 29840 }, { "grad_norm": 0.14638707041740417, "learning_rate": 8.404734671058924e-05, "loss": 0.0066, "step": 29850 }, { "grad_norm": 0.16776421666145325, "learning_rate": 8.403523594795902e-05, "loss": 0.0067, "step": 29860 }, { "grad_norm": 0.15623517334461212, "learning_rate": 8.402312146328659e-05, "loss": 0.0068, "step": 29870 }, { "grad_norm": 0.15834347903728485, "learning_rate": 8.401100325789675e-05, "loss": 0.0062, "step": 29880 }, { "grad_norm": 0.14886532723903656, "learning_rate": 8.399888133311472e-05, "loss": 0.0073, "step": 29890 }, { "grad_norm": 0.14393746852874756, "learning_rate": 8.398675569026613e-05, "loss": 0.0064, "step": 29900 }, { "grad_norm": 0.12963348627090454, "learning_rate": 8.397462633067705e-05, "loss": 0.0069, "step": 29910 }, { "grad_norm": 0.12475918978452682, "learning_rate": 8.396249325567392e-05, "loss": 0.0072, "step": 29920 }, { "grad_norm": 0.08920488506555557, "learning_rate": 8.395035646658357e-05, "loss": 0.0062, "step": 29930 }, { "grad_norm": 0.13461513817310333, "learning_rate": 8.39382159647333e-05, "loss": 0.0064, "step": 29940 }, { "grad_norm": 0.12167056649923325, "learning_rate": 8.392607175145075e-05, "loss": 0.0069, "step": 29950 }, { "grad_norm": 0.11343743652105331, "learning_rate": 8.3913923828064e-05, "loss": 0.0074, "step": 29960 }, { "grad_norm": 0.1298273801803589, "learning_rate": 8.390177219590152e-05, "loss": 0.0063, "step": 29970 }, { "grad_norm": 0.08908066898584366, "learning_rate": 8.388961685629222e-05, "loss": 0.0065, "step": 29980 }, { "grad_norm": 0.11231346428394318, "learning_rate": 8.387745781056536e-05, "loss": 0.0072, "step": 29990 }, { "grad_norm": 0.12323284149169922, "learning_rate": 8.386529506005065e-05, "loss": 0.0066, "step": 30000 }, { "grad_norm": 0.11942236125469208, "learning_rate": 8.38531286060782e-05, "loss": 0.0075, "step": 30010 }, { "grad_norm": 0.12996123731136322, "learning_rate": 8.384095844997849e-05, "loss": 0.007, "step": 30020 }, { "grad_norm": 0.12418239563703537, "learning_rate": 8.382878459308245e-05, "loss": 0.006, "step": 30030 }, { "grad_norm": 0.11697758734226227, "learning_rate": 8.381660703672138e-05, "loss": 0.0076, "step": 30040 }, { "grad_norm": 0.1283719539642334, "learning_rate": 8.380442578222702e-05, "loss": 0.006, "step": 30050 }, { "grad_norm": 0.1128285676240921, "learning_rate": 8.379224083093146e-05, "loss": 0.0073, "step": 30060 }, { "grad_norm": 0.1260465830564499, "learning_rate": 8.378005218416727e-05, "loss": 0.0069, "step": 30070 }, { "grad_norm": 0.1457066386938095, "learning_rate": 8.376785984326735e-05, "loss": 0.006, "step": 30080 }, { "grad_norm": 0.09756246954202652, "learning_rate": 8.375566380956506e-05, "loss": 0.0063, "step": 30090 }, { "grad_norm": 0.09727320075035095, "learning_rate": 8.374346408439411e-05, "loss": 0.0063, "step": 30100 }, { "grad_norm": 0.13968881964683533, "learning_rate": 8.373126066908868e-05, "loss": 0.0072, "step": 30110 }, { "grad_norm": 0.17277395725250244, "learning_rate": 8.371905356498326e-05, "loss": 0.0069, "step": 30120 }, { "grad_norm": 0.2021387815475464, "learning_rate": 8.370684277341288e-05, "loss": 0.0079, "step": 30130 }, { "grad_norm": 0.17488236725330353, "learning_rate": 8.369462829571282e-05, "loss": 0.0075, "step": 30140 }, { "grad_norm": 0.11667768657207489, "learning_rate": 8.36824101332189e-05, "loss": 0.0067, "step": 30150 }, { "grad_norm": 0.1439897119998932, "learning_rate": 8.367018828726721e-05, "loss": 0.0071, "step": 30160 }, { "grad_norm": 0.11682911962270737, "learning_rate": 8.365796275919438e-05, "loss": 0.0066, "step": 30170 }, { "grad_norm": 0.12388362735509872, "learning_rate": 8.364573355033734e-05, "loss": 0.0075, "step": 30180 }, { "grad_norm": 0.11124937236309052, "learning_rate": 8.363350066203346e-05, "loss": 0.0082, "step": 30190 }, { "grad_norm": 0.12340359389781952, "learning_rate": 8.362126409562053e-05, "loss": 0.0067, "step": 30200 }, { "grad_norm": 0.07514071464538574, "learning_rate": 8.360902385243667e-05, "loss": 0.0064, "step": 30210 }, { "grad_norm": 0.13383550941944122, "learning_rate": 8.359677993382052e-05, "loss": 0.0091, "step": 30220 }, { "grad_norm": 0.151775062084198, "learning_rate": 8.358453234111103e-05, "loss": 0.0068, "step": 30230 }, { "grad_norm": 0.1410267949104309, "learning_rate": 8.357228107564756e-05, "loss": 0.0078, "step": 30240 }, { "grad_norm": 0.1349225491285324, "learning_rate": 8.356002613876993e-05, "loss": 0.0064, "step": 30250 }, { "grad_norm": 0.12963569164276123, "learning_rate": 8.35477675318183e-05, "loss": 0.0077, "step": 30260 }, { "grad_norm": 0.12032514065504074, "learning_rate": 8.353550525613323e-05, "loss": 0.0068, "step": 30270 }, { "grad_norm": 0.12856827676296234, "learning_rate": 8.352323931305572e-05, "loss": 0.0059, "step": 30280 }, { "grad_norm": 0.13282321393489838, "learning_rate": 8.351096970392717e-05, "loss": 0.0078, "step": 30290 }, { "grad_norm": 0.13029298186302185, "learning_rate": 8.349869643008937e-05, "loss": 0.0061, "step": 30300 }, { "grad_norm": 0.1301354616880417, "learning_rate": 8.348641949288449e-05, "loss": 0.0078, "step": 30310 }, { "grad_norm": 0.1489444524049759, "learning_rate": 8.34741388936551e-05, "loss": 0.0079, "step": 30320 }, { "grad_norm": 0.09608953446149826, "learning_rate": 8.346185463374423e-05, "loss": 0.0066, "step": 30330 }, { "grad_norm": 0.14212679862976074, "learning_rate": 8.344956671449524e-05, "loss": 0.0073, "step": 30340 }, { "grad_norm": 0.14395371079444885, "learning_rate": 8.343727513725192e-05, "loss": 0.006, "step": 30350 }, { "grad_norm": 0.11996032297611237, "learning_rate": 8.342497990335847e-05, "loss": 0.007, "step": 30360 }, { "grad_norm": 0.11185632646083832, "learning_rate": 8.341268101415946e-05, "loss": 0.0067, "step": 30370 }, { "grad_norm": 0.11116403341293335, "learning_rate": 8.34003784709999e-05, "loss": 0.0074, "step": 30380 }, { "grad_norm": 0.11299311369657516, "learning_rate": 8.338807227522516e-05, "loss": 0.0067, "step": 30390 }, { "grad_norm": 0.16758935153484344, "learning_rate": 8.337576242818103e-05, "loss": 0.0072, "step": 30400 }, { "grad_norm": 0.14515629410743713, "learning_rate": 8.336344893121372e-05, "loss": 0.0068, "step": 30410 }, { "grad_norm": 0.08649928867816925, "learning_rate": 8.335113178566977e-05, "loss": 0.0068, "step": 30420 }, { "grad_norm": 0.10490816831588745, "learning_rate": 8.33388109928962e-05, "loss": 0.007, "step": 30430 }, { "grad_norm": 0.1420324593782425, "learning_rate": 8.33264865542404e-05, "loss": 0.0063, "step": 30440 }, { "grad_norm": 0.13810671865940094, "learning_rate": 8.331415847105013e-05, "loss": 0.0073, "step": 30450 }, { "grad_norm": 0.18822793662548065, "learning_rate": 8.330182674467357e-05, "loss": 0.0067, "step": 30460 }, { "grad_norm": 0.13157331943511963, "learning_rate": 8.32894913764593e-05, "loss": 0.0063, "step": 30470 }, { "grad_norm": 0.14945253729820251, "learning_rate": 8.327715236775633e-05, "loss": 0.0079, "step": 30480 }, { "grad_norm": 0.11199159920215607, "learning_rate": 8.326480971991398e-05, "loss": 0.0072, "step": 30490 }, { "grad_norm": 0.1483849436044693, "learning_rate": 8.325246343428206e-05, "loss": 0.0065, "step": 30500 }, { "grad_norm": 0.1368429809808731, "learning_rate": 8.324011351221072e-05, "loss": 0.0075, "step": 30510 }, { "grad_norm": 0.10424398630857468, "learning_rate": 8.322775995505057e-05, "loss": 0.008, "step": 30520 }, { "grad_norm": 0.12237724661827087, "learning_rate": 8.321540276415254e-05, "loss": 0.0074, "step": 30530 }, { "grad_norm": 0.2398490160703659, "learning_rate": 8.320304194086798e-05, "loss": 0.0074, "step": 30540 }, { "grad_norm": 0.16334767639636993, "learning_rate": 8.31906774865487e-05, "loss": 0.007, "step": 30550 }, { "grad_norm": 0.13422945141792297, "learning_rate": 8.317830940254682e-05, "loss": 0.0073, "step": 30560 }, { "grad_norm": 0.0891033262014389, "learning_rate": 8.316593769021491e-05, "loss": 0.0072, "step": 30570 }, { "grad_norm": 0.18598884344100952, "learning_rate": 8.315356235090592e-05, "loss": 0.0076, "step": 30580 }, { "grad_norm": 0.14568237960338593, "learning_rate": 8.314118338597319e-05, "loss": 0.0094, "step": 30590 }, { "grad_norm": 0.10956599563360214, "learning_rate": 8.312880079677048e-05, "loss": 0.0063, "step": 30600 }, { "grad_norm": 0.16499947011470795, "learning_rate": 8.311641458465191e-05, "loss": 0.0086, "step": 30610 }, { "grad_norm": 0.12104317545890808, "learning_rate": 8.310402475097205e-05, "loss": 0.0073, "step": 30620 }, { "grad_norm": 0.16347579658031464, "learning_rate": 8.309163129708581e-05, "loss": 0.0078, "step": 30630 }, { "grad_norm": 0.12892505526542664, "learning_rate": 8.307923422434852e-05, "loss": 0.0073, "step": 30640 }, { "grad_norm": 0.13842590153217316, "learning_rate": 8.30668335341159e-05, "loss": 0.008, "step": 30650 }, { "grad_norm": 0.14205628633499146, "learning_rate": 8.305442922774408e-05, "loss": 0.0056, "step": 30660 }, { "grad_norm": 0.19207651913166046, "learning_rate": 8.304202130658959e-05, "loss": 0.0069, "step": 30670 }, { "grad_norm": 0.1450698971748352, "learning_rate": 8.302960977200931e-05, "loss": 0.0074, "step": 30680 }, { "grad_norm": 0.12486214190721512, "learning_rate": 8.301719462536058e-05, "loss": 0.0063, "step": 30690 }, { "grad_norm": 0.09028539061546326, "learning_rate": 8.300477586800108e-05, "loss": 0.0067, "step": 30700 }, { "grad_norm": 0.15526728332042694, "learning_rate": 8.299235350128892e-05, "loss": 0.0064, "step": 30710 }, { "grad_norm": 0.11180320382118225, "learning_rate": 8.297992752658258e-05, "loss": 0.0067, "step": 30720 }, { "grad_norm": 0.12400521337985992, "learning_rate": 8.296749794524094e-05, "loss": 0.0076, "step": 30730 }, { "grad_norm": 0.1394391655921936, "learning_rate": 8.295506475862332e-05, "loss": 0.0074, "step": 30740 }, { "grad_norm": 0.1308317929506302, "learning_rate": 8.294262796808933e-05, "loss": 0.0069, "step": 30750 }, { "grad_norm": 0.1562187373638153, "learning_rate": 8.293018757499909e-05, "loss": 0.0065, "step": 30760 }, { "grad_norm": 0.13360807299613953, "learning_rate": 8.291774358071305e-05, "loss": 0.0074, "step": 30770 }, { "grad_norm": 0.09947826713323593, "learning_rate": 8.290529598659205e-05, "loss": 0.0063, "step": 30780 }, { "grad_norm": 0.09074632823467255, "learning_rate": 8.289284479399738e-05, "loss": 0.0067, "step": 30790 }, { "grad_norm": 0.11944101750850677, "learning_rate": 8.288039000429064e-05, "loss": 0.006, "step": 30800 }, { "grad_norm": 0.1284421980381012, "learning_rate": 8.286793161883388e-05, "loss": 0.006, "step": 30810 }, { "grad_norm": 0.09638094156980515, "learning_rate": 8.285546963898954e-05, "loss": 0.0057, "step": 30820 }, { "grad_norm": 0.07468719780445099, "learning_rate": 8.284300406612044e-05, "loss": 0.0063, "step": 30830 }, { "grad_norm": 0.11443408578634262, "learning_rate": 8.283053490158978e-05, "loss": 0.0076, "step": 30840 }, { "grad_norm": 0.11100952327251434, "learning_rate": 8.28180621467612e-05, "loss": 0.0057, "step": 30850 }, { "grad_norm": 0.10236162692308426, "learning_rate": 8.280558580299867e-05, "loss": 0.0064, "step": 30860 }, { "grad_norm": 0.11894407868385315, "learning_rate": 8.279310587166661e-05, "loss": 0.0073, "step": 30870 }, { "grad_norm": 0.1386343538761139, "learning_rate": 8.278062235412978e-05, "loss": 0.0074, "step": 30880 }, { "grad_norm": 0.11623381823301315, "learning_rate": 8.276813525175339e-05, "loss": 0.0067, "step": 30890 }, { "grad_norm": 0.09949006140232086, "learning_rate": 8.2755644565903e-05, "loss": 0.0073, "step": 30900 }, { "grad_norm": 0.13044373691082, "learning_rate": 8.274315029794454e-05, "loss": 0.0085, "step": 30910 }, { "grad_norm": 0.12487705796957016, "learning_rate": 8.273065244924443e-05, "loss": 0.0079, "step": 30920 }, { "grad_norm": 0.11295270919799805, "learning_rate": 8.271815102116936e-05, "loss": 0.0066, "step": 30930 }, { "grad_norm": 0.11586359888315201, "learning_rate": 8.270564601508648e-05, "loss": 0.0073, "step": 30940 }, { "grad_norm": 0.1567217856645584, "learning_rate": 8.269313743236333e-05, "loss": 0.0068, "step": 30950 }, { "grad_norm": 0.16219525039196014, "learning_rate": 8.268062527436783e-05, "loss": 0.0065, "step": 30960 }, { "grad_norm": 0.13512387871742249, "learning_rate": 8.266810954246828e-05, "loss": 0.0075, "step": 30970 }, { "grad_norm": 0.13448630273342133, "learning_rate": 8.265559023803338e-05, "loss": 0.0073, "step": 30980 }, { "grad_norm": 0.11518614739179611, "learning_rate": 8.264306736243223e-05, "loss": 0.0061, "step": 30990 }, { "grad_norm": 0.12725655734539032, "learning_rate": 8.263054091703432e-05, "loss": 0.0064, "step": 31000 }, { "grad_norm": 0.1958981305360794, "learning_rate": 8.26180109032095e-05, "loss": 0.0067, "step": 31010 }, { "grad_norm": 0.14870670437812805, "learning_rate": 8.260547732232807e-05, "loss": 0.0063, "step": 31020 }, { "grad_norm": 0.09787233173847198, "learning_rate": 8.259294017576063e-05, "loss": 0.0068, "step": 31030 }, { "grad_norm": 0.1438794732093811, "learning_rate": 8.258039946487828e-05, "loss": 0.0075, "step": 31040 }, { "grad_norm": 0.12246469408273697, "learning_rate": 8.256785519105241e-05, "loss": 0.0063, "step": 31050 }, { "grad_norm": 0.12073912471532822, "learning_rate": 8.255530735565488e-05, "loss": 0.0065, "step": 31060 }, { "grad_norm": 0.1306726485490799, "learning_rate": 8.254275596005787e-05, "loss": 0.006, "step": 31070 }, { "grad_norm": 0.12070835381746292, "learning_rate": 8.2530201005634e-05, "loss": 0.0068, "step": 31080 }, { "grad_norm": 0.13756163418293, "learning_rate": 8.251764249375626e-05, "loss": 0.006, "step": 31090 }, { "grad_norm": 0.11350195854902267, "learning_rate": 8.250508042579803e-05, "loss": 0.0075, "step": 31100 }, { "grad_norm": 0.12970386445522308, "learning_rate": 8.249251480313307e-05, "loss": 0.0064, "step": 31110 }, { "grad_norm": 0.11800684779882431, "learning_rate": 8.247994562713555e-05, "loss": 0.0068, "step": 31120 }, { "grad_norm": 0.1628960371017456, "learning_rate": 8.246737289917998e-05, "loss": 0.0067, "step": 31130 }, { "grad_norm": 0.12571236491203308, "learning_rate": 8.245479662064135e-05, "loss": 0.0086, "step": 31140 }, { "grad_norm": 0.15049311518669128, "learning_rate": 8.244221679289496e-05, "loss": 0.0076, "step": 31150 }, { "grad_norm": 0.11848881840705872, "learning_rate": 8.242963341731652e-05, "loss": 0.0065, "step": 31160 }, { "grad_norm": 0.11658526211977005, "learning_rate": 8.24170464952821e-05, "loss": 0.0082, "step": 31170 }, { "grad_norm": 0.1275898665189743, "learning_rate": 8.240445602816824e-05, "loss": 0.0069, "step": 31180 }, { "grad_norm": 0.09102784097194672, "learning_rate": 8.239186201735179e-05, "loss": 0.0066, "step": 31190 }, { "grad_norm": 0.1491519957780838, "learning_rate": 8.237926446420998e-05, "loss": 0.0075, "step": 31200 }, { "grad_norm": 0.13003870844841003, "learning_rate": 8.236666337012052e-05, "loss": 0.0064, "step": 31210 }, { "grad_norm": 0.13437822461128235, "learning_rate": 8.23540587364614e-05, "loss": 0.0077, "step": 31220 }, { "grad_norm": 0.09930375218391418, "learning_rate": 8.234145056461107e-05, "loss": 0.0087, "step": 31230 }, { "grad_norm": 0.12784281373023987, "learning_rate": 8.232883885594832e-05, "loss": 0.0062, "step": 31240 }, { "grad_norm": 0.15705618262290955, "learning_rate": 8.231622361185236e-05, "loss": 0.0077, "step": 31250 }, { "grad_norm": 0.11634062975645065, "learning_rate": 8.230360483370278e-05, "loss": 0.0073, "step": 31260 }, { "grad_norm": 0.149638369679451, "learning_rate": 8.229098252287953e-05, "loss": 0.007, "step": 31270 }, { "grad_norm": 0.15719419717788696, "learning_rate": 8.2278356680763e-05, "loss": 0.0077, "step": 31280 }, { "grad_norm": 0.14850200712680817, "learning_rate": 8.22657273087339e-05, "loss": 0.0067, "step": 31290 }, { "grad_norm": 0.12550845742225647, "learning_rate": 8.225309440817336e-05, "loss": 0.007, "step": 31300 }, { "grad_norm": 0.13013266026973724, "learning_rate": 8.224045798046293e-05, "loss": 0.0075, "step": 31310 }, { "grad_norm": 0.11139975488185883, "learning_rate": 8.22278180269845e-05, "loss": 0.0076, "step": 31320 }, { "grad_norm": 0.12080329656600952, "learning_rate": 8.221517454912031e-05, "loss": 0.0064, "step": 31330 }, { "grad_norm": 0.19104637205600739, "learning_rate": 8.220252754825308e-05, "loss": 0.0086, "step": 31340 }, { "grad_norm": 0.1386386603116989, "learning_rate": 8.218987702576586e-05, "loss": 0.0073, "step": 31350 }, { "grad_norm": 0.10174199193716049, "learning_rate": 8.217722298304207e-05, "loss": 0.0088, "step": 31360 }, { "grad_norm": 0.1337379515171051, "learning_rate": 8.216456542146557e-05, "loss": 0.0066, "step": 31370 }, { "grad_norm": 0.15537770092487335, "learning_rate": 8.215190434242055e-05, "loss": 0.0091, "step": 31380 }, { "grad_norm": 0.1220746710896492, "learning_rate": 8.213923974729161e-05, "loss": 0.0074, "step": 31390 }, { "grad_norm": 0.12173815071582794, "learning_rate": 8.212657163746373e-05, "loss": 0.0076, "step": 31400 }, { "grad_norm": 0.13403762876987457, "learning_rate": 8.211390001432227e-05, "loss": 0.0064, "step": 31410 }, { "grad_norm": 0.09254489094018936, "learning_rate": 8.210122487925297e-05, "loss": 0.0065, "step": 31420 }, { "grad_norm": 0.0962701067328453, "learning_rate": 8.208854623364202e-05, "loss": 0.006, "step": 31430 }, { "grad_norm": 0.0977090373635292, "learning_rate": 8.207586407887589e-05, "loss": 0.0076, "step": 31440 }, { "grad_norm": 0.11224830895662308, "learning_rate": 8.206317841634148e-05, "loss": 0.0066, "step": 31450 }, { "grad_norm": 0.14225444197654724, "learning_rate": 8.205048924742609e-05, "loss": 0.008, "step": 31460 }, { "grad_norm": 0.15445426106452942, "learning_rate": 8.203779657351738e-05, "loss": 0.0064, "step": 31470 }, { "grad_norm": 0.12442678213119507, "learning_rate": 8.20251003960034e-05, "loss": 0.0069, "step": 31480 }, { "grad_norm": 0.12767526507377625, "learning_rate": 8.201240071627258e-05, "loss": 0.0066, "step": 31490 }, { "grad_norm": 0.09826187789440155, "learning_rate": 8.199969753571377e-05, "loss": 0.0079, "step": 31500 }, { "grad_norm": 0.11495046317577362, "learning_rate": 8.198699085571615e-05, "loss": 0.0075, "step": 31510 }, { "grad_norm": 0.11492829024791718, "learning_rate": 8.197428067766928e-05, "loss": 0.0077, "step": 31520 }, { "grad_norm": 0.10763975977897644, "learning_rate": 8.196156700296316e-05, "loss": 0.0062, "step": 31530 }, { "grad_norm": 0.09763116389513016, "learning_rate": 8.194884983298814e-05, "loss": 0.0062, "step": 31540 }, { "grad_norm": 0.13887472450733185, "learning_rate": 8.193612916913491e-05, "loss": 0.0066, "step": 31550 }, { "grad_norm": 0.14644241333007812, "learning_rate": 8.192340501279463e-05, "loss": 0.006, "step": 31560 }, { "grad_norm": 0.1110975369811058, "learning_rate": 8.191067736535876e-05, "loss": 0.0081, "step": 31570 }, { "grad_norm": 0.11460452526807785, "learning_rate": 8.18979462282192e-05, "loss": 0.0074, "step": 31580 }, { "grad_norm": 0.1165570542216301, "learning_rate": 8.188521160276819e-05, "loss": 0.0076, "step": 31590 }, { "grad_norm": 0.1148766502737999, "learning_rate": 8.187247349039837e-05, "loss": 0.0065, "step": 31600 }, { "grad_norm": 0.14815595746040344, "learning_rate": 8.185973189250278e-05, "loss": 0.0066, "step": 31610 }, { "grad_norm": 0.12720662355422974, "learning_rate": 8.184698681047482e-05, "loss": 0.0073, "step": 31620 }, { "grad_norm": 0.13019543886184692, "learning_rate": 8.183423824570827e-05, "loss": 0.0078, "step": 31630 }, { "grad_norm": 0.11978695541620255, "learning_rate": 8.182148619959725e-05, "loss": 0.0066, "step": 31640 }, { "grad_norm": 0.14430351555347443, "learning_rate": 8.180873067353636e-05, "loss": 0.0067, "step": 31650 }, { "grad_norm": 0.11301806569099426, "learning_rate": 8.179597166892052e-05, "loss": 0.0063, "step": 31660 }, { "grad_norm": 0.11093707382678986, "learning_rate": 8.178320918714501e-05, "loss": 0.0061, "step": 31670 }, { "grad_norm": 0.1331539899110794, "learning_rate": 8.177044322960554e-05, "loss": 0.0076, "step": 31680 }, { "grad_norm": 0.14935049414634705, "learning_rate": 8.175767379769816e-05, "loss": 0.0071, "step": 31690 }, { "grad_norm": 0.14524702727794647, "learning_rate": 8.174490089281932e-05, "loss": 0.0077, "step": 31700 }, { "grad_norm": 0.14424173533916473, "learning_rate": 8.173212451636584e-05, "loss": 0.0069, "step": 31710 }, { "grad_norm": 0.1056654304265976, "learning_rate": 8.171934466973493e-05, "loss": 0.0064, "step": 31720 }, { "grad_norm": 0.10184909403324127, "learning_rate": 8.170656135432418e-05, "loss": 0.0064, "step": 31730 }, { "grad_norm": 0.16472692787647247, "learning_rate": 8.169377457153155e-05, "loss": 0.0063, "step": 31740 }, { "grad_norm": 0.1071547344326973, "learning_rate": 8.168098432275539e-05, "loss": 0.0055, "step": 31750 }, { "grad_norm": 0.1307533085346222, "learning_rate": 8.166819060939442e-05, "loss": 0.006, "step": 31760 }, { "grad_norm": 0.15478108823299408, "learning_rate": 8.165539343284772e-05, "loss": 0.0064, "step": 31770 }, { "grad_norm": 0.11441052705049515, "learning_rate": 8.16425927945148e-05, "loss": 0.0088, "step": 31780 }, { "grad_norm": 0.11615051329135895, "learning_rate": 8.162978869579551e-05, "loss": 0.0082, "step": 31790 }, { "grad_norm": 0.15067508816719055, "learning_rate": 8.161698113809007e-05, "loss": 0.0073, "step": 31800 }, { "grad_norm": 0.15109051764011383, "learning_rate": 8.160417012279911e-05, "loss": 0.007, "step": 31810 }, { "grad_norm": 0.1162022277712822, "learning_rate": 8.159135565132363e-05, "loss": 0.0073, "step": 31820 }, { "grad_norm": 0.12635065615177155, "learning_rate": 8.157853772506498e-05, "loss": 0.0081, "step": 31830 }, { "grad_norm": 0.12727048993110657, "learning_rate": 8.156571634542494e-05, "loss": 0.0068, "step": 31840 }, { "grad_norm": 0.14204905927181244, "learning_rate": 8.15528915138056e-05, "loss": 0.0079, "step": 31850 }, { "grad_norm": 0.1420241892337799, "learning_rate": 8.154006323160949e-05, "loss": 0.0075, "step": 31860 }, { "grad_norm": 0.12919826805591583, "learning_rate": 8.152723150023949e-05, "loss": 0.0074, "step": 31870 }, { "grad_norm": 0.17980872094631195, "learning_rate": 8.151439632109886e-05, "loss": 0.0075, "step": 31880 }, { "grad_norm": 0.16208136081695557, "learning_rate": 8.150155769559122e-05, "loss": 0.0074, "step": 31890 }, { "grad_norm": 0.13449178636074066, "learning_rate": 8.148871562512058e-05, "loss": 0.0069, "step": 31900 }, { "grad_norm": 0.13300488889217377, "learning_rate": 8.147587011109136e-05, "loss": 0.0077, "step": 31910 }, { "grad_norm": 0.15464560687541962, "learning_rate": 8.14630211549083e-05, "loss": 0.0069, "step": 31920 }, { "grad_norm": 0.1277027279138565, "learning_rate": 8.145016875797655e-05, "loss": 0.0077, "step": 31930 }, { "grad_norm": 0.119478240609169, "learning_rate": 8.143731292170164e-05, "loss": 0.0066, "step": 31940 }, { "grad_norm": 0.12288009375333786, "learning_rate": 8.142445364748944e-05, "loss": 0.0064, "step": 31950 }, { "grad_norm": 0.12097617983818054, "learning_rate": 8.141159093674624e-05, "loss": 0.008, "step": 31960 }, { "grad_norm": 0.14301814138889313, "learning_rate": 8.139872479087869e-05, "loss": 0.0062, "step": 31970 }, { "grad_norm": 0.12577451765537262, "learning_rate": 8.13858552112938e-05, "loss": 0.0059, "step": 31980 }, { "grad_norm": 0.10989046096801758, "learning_rate": 8.137298219939895e-05, "loss": 0.0073, "step": 31990 }, { "grad_norm": 0.11713874340057373, "learning_rate": 8.136010575660196e-05, "loss": 0.0069, "step": 32000 }, { "grad_norm": 0.13857042789459229, "learning_rate": 8.134722588431095e-05, "loss": 0.0073, "step": 32010 }, { "grad_norm": 0.16801074147224426, "learning_rate": 8.133434258393444e-05, "loss": 0.0069, "step": 32020 }, { "grad_norm": 0.12481174618005753, "learning_rate": 8.132145585688134e-05, "loss": 0.0075, "step": 32030 }, { "grad_norm": 0.10495954006910324, "learning_rate": 8.130856570456093e-05, "loss": 0.0075, "step": 32040 }, { "grad_norm": 0.08916697651147842, "learning_rate": 8.129567212838283e-05, "loss": 0.0069, "step": 32050 }, { "grad_norm": 0.14539183676242828, "learning_rate": 8.128277512975708e-05, "loss": 0.0075, "step": 32060 }, { "grad_norm": 0.13017134368419647, "learning_rate": 8.126987471009408e-05, "loss": 0.0078, "step": 32070 }, { "grad_norm": 0.11197181791067123, "learning_rate": 8.125697087080459e-05, "loss": 0.0092, "step": 32080 }, { "grad_norm": 0.15958702564239502, "learning_rate": 8.124406361329976e-05, "loss": 0.0079, "step": 32090 }, { "grad_norm": 0.16742649674415588, "learning_rate": 8.123115293899111e-05, "loss": 0.0082, "step": 32100 }, { "grad_norm": 0.15533463656902313, "learning_rate": 8.121823884929055e-05, "loss": 0.0077, "step": 32110 }, { "grad_norm": 0.11625643074512482, "learning_rate": 8.12053213456103e-05, "loss": 0.0077, "step": 32120 }, { "grad_norm": 0.11818306148052216, "learning_rate": 8.119240042936303e-05, "loss": 0.0066, "step": 32130 }, { "grad_norm": 0.12041525542736053, "learning_rate": 8.117947610196175e-05, "loss": 0.0068, "step": 32140 }, { "grad_norm": 0.15039707720279694, "learning_rate": 8.116654836481982e-05, "loss": 0.0077, "step": 32150 }, { "grad_norm": 0.1409413367509842, "learning_rate": 8.115361721935106e-05, "loss": 0.0081, "step": 32160 }, { "grad_norm": 0.138546884059906, "learning_rate": 8.114068266696953e-05, "loss": 0.007, "step": 32170 }, { "grad_norm": 0.11864488571882248, "learning_rate": 8.112774470908978e-05, "loss": 0.0078, "step": 32180 }, { "grad_norm": 0.15305237472057343, "learning_rate": 8.111480334712665e-05, "loss": 0.0084, "step": 32190 }, { "grad_norm": 0.14636535942554474, "learning_rate": 8.110185858249542e-05, "loss": 0.0076, "step": 32200 }, { "grad_norm": 0.13145054876804352, "learning_rate": 8.108891041661168e-05, "loss": 0.0066, "step": 32210 }, { "grad_norm": 0.10279174149036407, "learning_rate": 8.107595885089146e-05, "loss": 0.0065, "step": 32220 }, { "grad_norm": 0.10997382551431656, "learning_rate": 8.106300388675108e-05, "loss": 0.0067, "step": 32230 }, { "grad_norm": 0.13961416482925415, "learning_rate": 8.105004552560731e-05, "loss": 0.0067, "step": 32240 }, { "grad_norm": 0.15259325504302979, "learning_rate": 8.103708376887724e-05, "loss": 0.0066, "step": 32250 }, { "grad_norm": 0.13716909289360046, "learning_rate": 8.102411861797836e-05, "loss": 0.0063, "step": 32260 }, { "grad_norm": 0.0929180458188057, "learning_rate": 8.101115007432851e-05, "loss": 0.0067, "step": 32270 }, { "grad_norm": 0.13320930302143097, "learning_rate": 8.09981781393459e-05, "loss": 0.0064, "step": 32280 }, { "grad_norm": 0.122689388692379, "learning_rate": 8.098520281444915e-05, "loss": 0.0058, "step": 32290 }, { "grad_norm": 0.11652141064405441, "learning_rate": 8.09722241010572e-05, "loss": 0.008, "step": 32300 }, { "grad_norm": 0.14782989025115967, "learning_rate": 8.095924200058939e-05, "loss": 0.0069, "step": 32310 }, { "grad_norm": 0.1565481424331665, "learning_rate": 8.094625651446541e-05, "loss": 0.0068, "step": 32320 }, { "grad_norm": 0.17951302230358124, "learning_rate": 8.093326764410536e-05, "loss": 0.0068, "step": 32330 }, { "grad_norm": 0.12093459069728851, "learning_rate": 8.092027539092966e-05, "loss": 0.0074, "step": 32340 }, { "grad_norm": 0.12228143215179443, "learning_rate": 8.090727975635913e-05, "loss": 0.0066, "step": 32350 }, { "grad_norm": 0.11127287894487381, "learning_rate": 8.089428074181497e-05, "loss": 0.0066, "step": 32360 }, { "grad_norm": 0.15090179443359375, "learning_rate": 8.088127834871871e-05, "loss": 0.0064, "step": 32370 }, { "grad_norm": 0.12641382217407227, "learning_rate": 8.086827257849226e-05, "loss": 0.0069, "step": 32380 }, { "grad_norm": 0.12986333668231964, "learning_rate": 8.085526343255795e-05, "loss": 0.0071, "step": 32390 }, { "grad_norm": 0.10123847424983978, "learning_rate": 8.084225091233842e-05, "loss": 0.008, "step": 32400 }, { "grad_norm": 0.1308261752128601, "learning_rate": 8.082923501925668e-05, "loss": 0.0074, "step": 32410 }, { "grad_norm": 0.16567769646644592, "learning_rate": 8.081621575473617e-05, "loss": 0.0081, "step": 32420 }, { "grad_norm": 0.15705186128616333, "learning_rate": 8.080319312020064e-05, "loss": 0.0078, "step": 32430 }, { "grad_norm": 0.12693491578102112, "learning_rate": 8.079016711707421e-05, "loss": 0.0068, "step": 32440 }, { "grad_norm": 0.10887563973665237, "learning_rate": 8.077713774678139e-05, "loss": 0.0062, "step": 32450 }, { "grad_norm": 0.13414539396762848, "learning_rate": 8.076410501074707e-05, "loss": 0.0056, "step": 32460 }, { "grad_norm": 0.1142486035823822, "learning_rate": 8.075106891039647e-05, "loss": 0.0063, "step": 32470 }, { "grad_norm": 0.15933136641979218, "learning_rate": 8.073802944715523e-05, "loss": 0.0057, "step": 32480 }, { "grad_norm": 0.109934501349926, "learning_rate": 8.072498662244929e-05, "loss": 0.0075, "step": 32490 }, { "grad_norm": 0.11574388295412064, "learning_rate": 8.0711940437705e-05, "loss": 0.0077, "step": 32500 }, { "grad_norm": 0.09264020621776581, "learning_rate": 8.06988908943491e-05, "loss": 0.0071, "step": 32510 }, { "grad_norm": 0.14012658596038818, "learning_rate": 8.068583799380863e-05, "loss": 0.0064, "step": 32520 }, { "grad_norm": 0.11410577595233917, "learning_rate": 8.067278173751104e-05, "loss": 0.0062, "step": 32530 }, { "grad_norm": 0.10951803624629974, "learning_rate": 8.065972212688417e-05, "loss": 0.0061, "step": 32540 }, { "grad_norm": 0.08514055609703064, "learning_rate": 8.064665916335618e-05, "loss": 0.0061, "step": 32550 }, { "grad_norm": 0.1333349347114563, "learning_rate": 8.063359284835564e-05, "loss": 0.0056, "step": 32560 }, { "grad_norm": 0.11516246944665909, "learning_rate": 8.062052318331142e-05, "loss": 0.0067, "step": 32570 }, { "grad_norm": 0.1281438171863556, "learning_rate": 8.060745016965283e-05, "loss": 0.0064, "step": 32580 }, { "grad_norm": 0.11589547246694565, "learning_rate": 8.059437380880952e-05, "loss": 0.0062, "step": 32590 }, { "grad_norm": 0.1326325237751007, "learning_rate": 8.058129410221146e-05, "loss": 0.007, "step": 32600 }, { "grad_norm": 0.11895734071731567, "learning_rate": 8.056821105128908e-05, "loss": 0.0065, "step": 32610 }, { "grad_norm": 0.13953326642513275, "learning_rate": 8.05551246574731e-05, "loss": 0.0066, "step": 32620 }, { "grad_norm": 0.1412997543811798, "learning_rate": 8.05420349221946e-05, "loss": 0.0071, "step": 32630 }, { "grad_norm": 0.08424060791730881, "learning_rate": 8.05289418468851e-05, "loss": 0.0059, "step": 32640 }, { "grad_norm": 0.13537907600402832, "learning_rate": 8.051584543297642e-05, "loss": 0.0077, "step": 32650 }, { "grad_norm": 0.18904221057891846, "learning_rate": 8.050274568190074e-05, "loss": 0.0075, "step": 32660 }, { "grad_norm": 0.1447908580303192, "learning_rate": 8.048964259509067e-05, "loss": 0.0071, "step": 32670 }, { "grad_norm": 0.11493213474750519, "learning_rate": 8.047653617397914e-05, "loss": 0.0073, "step": 32680 }, { "grad_norm": 0.1081603467464447, "learning_rate": 8.046342641999941e-05, "loss": 0.0072, "step": 32690 }, { "grad_norm": 0.11879462003707886, "learning_rate": 8.045031333458517e-05, "loss": 0.0067, "step": 32700 }, { "grad_norm": 0.10415942966938019, "learning_rate": 8.043719691917047e-05, "loss": 0.008, "step": 32710 }, { "grad_norm": 0.11386924982070923, "learning_rate": 8.042407717518966e-05, "loss": 0.0063, "step": 32720 }, { "grad_norm": 0.11402254551649094, "learning_rate": 8.041095410407751e-05, "loss": 0.0071, "step": 32730 }, { "grad_norm": 0.1285354048013687, "learning_rate": 8.039782770726913e-05, "loss": 0.0074, "step": 32740 }, { "grad_norm": 0.13599970936775208, "learning_rate": 8.038469798620004e-05, "loss": 0.0067, "step": 32750 }, { "grad_norm": 0.08652687817811966, "learning_rate": 8.037156494230604e-05, "loss": 0.006, "step": 32760 }, { "grad_norm": 0.13031582534313202, "learning_rate": 8.035842857702338e-05, "loss": 0.0063, "step": 32770 }, { "grad_norm": 0.11109879612922668, "learning_rate": 8.03452888917886e-05, "loss": 0.0055, "step": 32780 }, { "grad_norm": 0.13020342588424683, "learning_rate": 8.033214588803866e-05, "loss": 0.0064, "step": 32790 }, { "grad_norm": 0.11346656084060669, "learning_rate": 8.031899956721083e-05, "loss": 0.0066, "step": 32800 }, { "grad_norm": 0.1571444422006607, "learning_rate": 8.030584993074282e-05, "loss": 0.0075, "step": 32810 }, { "grad_norm": 0.14322051405906677, "learning_rate": 8.02926969800726e-05, "loss": 0.0091, "step": 32820 }, { "grad_norm": 0.16732682287693024, "learning_rate": 8.027954071663859e-05, "loss": 0.0073, "step": 32830 }, { "grad_norm": 0.16690075397491455, "learning_rate": 8.026638114187954e-05, "loss": 0.0075, "step": 32840 }, { "grad_norm": 0.12871545553207397, "learning_rate": 8.025321825723456e-05, "loss": 0.0082, "step": 32850 }, { "grad_norm": 0.11363900452852249, "learning_rate": 8.02400520641431e-05, "loss": 0.0082, "step": 32860 }, { "grad_norm": 0.12127317488193512, "learning_rate": 8.022688256404501e-05, "loss": 0.0066, "step": 32870 }, { "grad_norm": 0.15987958014011383, "learning_rate": 8.02137097583805e-05, "loss": 0.0075, "step": 32880 }, { "grad_norm": 0.13187865912914276, "learning_rate": 8.02005336485901e-05, "loss": 0.008, "step": 32890 }, { "grad_norm": 0.14134861528873444, "learning_rate": 8.018735423611476e-05, "loss": 0.0069, "step": 32900 }, { "grad_norm": 0.15199163556098938, "learning_rate": 8.017417152239574e-05, "loss": 0.0063, "step": 32910 }, { "grad_norm": 0.12528982758522034, "learning_rate": 8.01609855088747e-05, "loss": 0.0067, "step": 32920 }, { "grad_norm": 0.1317427009344101, "learning_rate": 8.014779619699362e-05, "loss": 0.0059, "step": 32930 }, { "grad_norm": 0.11823000758886337, "learning_rate": 8.013460358819489e-05, "loss": 0.0071, "step": 32940 }, { "grad_norm": 0.11734212189912796, "learning_rate": 8.01214076839212e-05, "loss": 0.0069, "step": 32950 }, { "grad_norm": 0.1175585389137268, "learning_rate": 8.010820848561565e-05, "loss": 0.0067, "step": 32960 }, { "grad_norm": 0.1194191724061966, "learning_rate": 8.009500599472171e-05, "loss": 0.006, "step": 32970 }, { "grad_norm": 0.12749655544757843, "learning_rate": 8.008180021268314e-05, "loss": 0.0058, "step": 32980 }, { "grad_norm": 0.10546735674142838, "learning_rate": 8.006859114094414e-05, "loss": 0.0069, "step": 32990 }, { "grad_norm": 0.09174380451440811, "learning_rate": 8.005537878094921e-05, "loss": 0.0071, "step": 33000 }, { "grad_norm": 0.09942766278982162, "learning_rate": 8.004216313414323e-05, "loss": 0.006, "step": 33010 }, { "grad_norm": 0.13726584613323212, "learning_rate": 8.002894420197149e-05, "loss": 0.007, "step": 33020 }, { "grad_norm": 0.1916765719652176, "learning_rate": 8.001572198587954e-05, "loss": 0.0071, "step": 33030 }, { "grad_norm": 0.0900643840432167, "learning_rate": 8.000249648731338e-05, "loss": 0.0069, "step": 33040 }, { "grad_norm": 0.10018372535705566, "learning_rate": 7.998926770771928e-05, "loss": 0.0063, "step": 33050 }, { "grad_norm": 0.14239977300167084, "learning_rate": 7.997603564854397e-05, "loss": 0.0066, "step": 33060 }, { "grad_norm": 0.14168304204940796, "learning_rate": 7.996280031123448e-05, "loss": 0.0063, "step": 33070 }, { "grad_norm": 0.1254909485578537, "learning_rate": 7.994956169723818e-05, "loss": 0.006, "step": 33080 }, { "grad_norm": 0.1173924133181572, "learning_rate": 7.993631980800285e-05, "loss": 0.0063, "step": 33090 }, { "grad_norm": 0.12037171423435211, "learning_rate": 7.992307464497659e-05, "loss": 0.006, "step": 33100 }, { "grad_norm": 0.1451815813779831, "learning_rate": 7.990982620960787e-05, "loss": 0.0064, "step": 33110 }, { "grad_norm": 0.11690037697553635, "learning_rate": 7.989657450334554e-05, "loss": 0.0059, "step": 33120 }, { "grad_norm": 0.10830927640199661, "learning_rate": 7.988331952763877e-05, "loss": 0.0072, "step": 33130 }, { "grad_norm": 0.1231347993016243, "learning_rate": 7.987006128393709e-05, "loss": 0.0065, "step": 33140 }, { "grad_norm": 0.1603824347257614, "learning_rate": 7.985679977369043e-05, "loss": 0.0061, "step": 33150 }, { "grad_norm": 0.10284818708896637, "learning_rate": 7.984353499834902e-05, "loss": 0.0068, "step": 33160 }, { "grad_norm": 0.21156799793243408, "learning_rate": 7.983026695936351e-05, "loss": 0.007, "step": 33170 }, { "grad_norm": 0.17751508951187134, "learning_rate": 7.981699565818486e-05, "loss": 0.0067, "step": 33180 }, { "grad_norm": 0.1156994104385376, "learning_rate": 7.980372109626437e-05, "loss": 0.0062, "step": 33190 }, { "grad_norm": 0.12988238036632538, "learning_rate": 7.979044327505375e-05, "loss": 0.0076, "step": 33200 }, { "grad_norm": 0.1467040330171585, "learning_rate": 7.977716219600506e-05, "loss": 0.0062, "step": 33210 }, { "grad_norm": 0.14481347799301147, "learning_rate": 7.97638778605707e-05, "loss": 0.0064, "step": 33220 }, { "grad_norm": 0.10314248502254486, "learning_rate": 7.975059027020338e-05, "loss": 0.0066, "step": 33230 }, { "grad_norm": 0.10775463283061981, "learning_rate": 7.973729942635623e-05, "loss": 0.0065, "step": 33240 }, { "grad_norm": 0.13260510563850403, "learning_rate": 7.972400533048273e-05, "loss": 0.0057, "step": 33250 }, { "grad_norm": 0.1426638513803482, "learning_rate": 7.97107079840367e-05, "loss": 0.0069, "step": 33260 }, { "grad_norm": 0.09983289986848831, "learning_rate": 7.969740738847231e-05, "loss": 0.0063, "step": 33270 }, { "grad_norm": 0.1347036361694336, "learning_rate": 7.968410354524411e-05, "loss": 0.0071, "step": 33280 }, { "grad_norm": 0.09972184896469116, "learning_rate": 7.967079645580697e-05, "loss": 0.0064, "step": 33290 }, { "grad_norm": 0.1361885666847229, "learning_rate": 7.965748612161612e-05, "loss": 0.0066, "step": 33300 }, { "grad_norm": 0.11876700818538666, "learning_rate": 7.96441725441272e-05, "loss": 0.0066, "step": 33310 }, { "grad_norm": 0.1368328481912613, "learning_rate": 7.963085572479614e-05, "loss": 0.0077, "step": 33320 }, { "grad_norm": 0.2232883721590042, "learning_rate": 7.961753566507924e-05, "loss": 0.0075, "step": 33330 }, { "grad_norm": 0.10775700956583023, "learning_rate": 7.960421236643316e-05, "loss": 0.0066, "step": 33340 }, { "grad_norm": 0.10093480348587036, "learning_rate": 7.959088583031496e-05, "loss": 0.0076, "step": 33350 }, { "grad_norm": 0.11896884441375732, "learning_rate": 7.957755605818194e-05, "loss": 0.0066, "step": 33360 }, { "grad_norm": 0.10184609889984131, "learning_rate": 7.956422305149185e-05, "loss": 0.0066, "step": 33370 }, { "grad_norm": 0.11045308411121368, "learning_rate": 7.95508868117028e-05, "loss": 0.0063, "step": 33380 }, { "grad_norm": 0.13680621981620789, "learning_rate": 7.953754734027318e-05, "loss": 0.0064, "step": 33390 }, { "grad_norm": 0.10587238520383835, "learning_rate": 7.952420463866182e-05, "loss": 0.0062, "step": 33400 }, { "grad_norm": 0.14949437975883484, "learning_rate": 7.951085870832782e-05, "loss": 0.0072, "step": 33410 }, { "grad_norm": 0.09581635892391205, "learning_rate": 7.949750955073067e-05, "loss": 0.0076, "step": 33420 }, { "grad_norm": 0.15239235758781433, "learning_rate": 7.948415716733022e-05, "loss": 0.0073, "step": 33430 }, { "grad_norm": 0.08328022062778473, "learning_rate": 7.947080155958669e-05, "loss": 0.0069, "step": 33440 }, { "grad_norm": 0.12175051867961884, "learning_rate": 7.94574427289606e-05, "loss": 0.0076, "step": 33450 }, { "grad_norm": 0.16573859751224518, "learning_rate": 7.944408067691284e-05, "loss": 0.0075, "step": 33460 }, { "grad_norm": 0.18428535759449005, "learning_rate": 7.943071540490473e-05, "loss": 0.0076, "step": 33470 }, { "grad_norm": 0.12602035701274872, "learning_rate": 7.94173469143978e-05, "loss": 0.0058, "step": 33480 }, { "grad_norm": 0.1313425451517105, "learning_rate": 7.940397520685406e-05, "loss": 0.0065, "step": 33490 }, { "grad_norm": 0.1256580650806427, "learning_rate": 7.939060028373577e-05, "loss": 0.0068, "step": 33500 }, { "grad_norm": 0.11961676925420761, "learning_rate": 7.937722214650565e-05, "loss": 0.0054, "step": 33510 }, { "grad_norm": 0.13777881860733032, "learning_rate": 7.936384079662666e-05, "loss": 0.0052, "step": 33520 }, { "grad_norm": 0.08901271224021912, "learning_rate": 7.93504562355622e-05, "loss": 0.0055, "step": 33530 }, { "grad_norm": 0.09018532186746597, "learning_rate": 7.933706846477599e-05, "loss": 0.006, "step": 33540 }, { "grad_norm": 0.1407826840877533, "learning_rate": 7.932367748573206e-05, "loss": 0.0059, "step": 33550 }, { "grad_norm": 0.12793388962745667, "learning_rate": 7.931028329989485e-05, "loss": 0.0072, "step": 33560 }, { "grad_norm": 0.11837304383516312, "learning_rate": 7.929688590872913e-05, "loss": 0.0071, "step": 33570 }, { "grad_norm": 0.18248304724693298, "learning_rate": 7.928348531370003e-05, "loss": 0.0072, "step": 33580 }, { "grad_norm": 0.14196133613586426, "learning_rate": 7.927008151627297e-05, "loss": 0.0085, "step": 33590 }, { "grad_norm": 0.11733519285917282, "learning_rate": 7.925667451791383e-05, "loss": 0.0071, "step": 33600 }, { "grad_norm": 0.19862116873264313, "learning_rate": 7.924326432008874e-05, "loss": 0.0085, "step": 33610 }, { "grad_norm": 0.1667778491973877, "learning_rate": 7.922985092426422e-05, "loss": 0.0073, "step": 33620 }, { "grad_norm": 0.10621780157089233, "learning_rate": 7.921643433190717e-05, "loss": 0.0075, "step": 33630 }, { "grad_norm": 0.1314547061920166, "learning_rate": 7.920301454448478e-05, "loss": 0.0077, "step": 33640 }, { "grad_norm": 0.1848745495080948, "learning_rate": 7.918959156346461e-05, "loss": 0.0072, "step": 33650 }, { "grad_norm": 0.11527696996927261, "learning_rate": 7.91761653903146e-05, "loss": 0.0089, "step": 33660 }, { "grad_norm": 0.15181447565555573, "learning_rate": 7.916273602650302e-05, "loss": 0.0076, "step": 33670 }, { "grad_norm": 0.11320127546787262, "learning_rate": 7.914930347349847e-05, "loss": 0.0075, "step": 33680 }, { "grad_norm": 0.1078757792711258, "learning_rate": 7.913586773276992e-05, "loss": 0.006, "step": 33690 }, { "grad_norm": 0.11547825485467911, "learning_rate": 7.912242880578667e-05, "loss": 0.0062, "step": 33700 }, { "grad_norm": 0.15767574310302734, "learning_rate": 7.910898669401839e-05, "loss": 0.0067, "step": 33710 }, { "grad_norm": 0.15845586359500885, "learning_rate": 7.909554139893511e-05, "loss": 0.0071, "step": 33720 }, { "grad_norm": 0.18723519146442413, "learning_rate": 7.908209292200715e-05, "loss": 0.0071, "step": 33730 }, { "grad_norm": 0.11065255105495453, "learning_rate": 7.906864126470523e-05, "loss": 0.0064, "step": 33740 }, { "grad_norm": 0.10075601190328598, "learning_rate": 7.905518642850041e-05, "loss": 0.0065, "step": 33750 }, { "grad_norm": 0.09718336164951324, "learning_rate": 7.904172841486409e-05, "loss": 0.0068, "step": 33760 }, { "grad_norm": 0.1103060320019722, "learning_rate": 7.902826722526801e-05, "loss": 0.0062, "step": 33770 }, { "grad_norm": 0.12307845056056976, "learning_rate": 7.901480286118427e-05, "loss": 0.006, "step": 33780 }, { "grad_norm": 0.1175466850399971, "learning_rate": 7.900133532408531e-05, "loss": 0.0062, "step": 33790 }, { "grad_norm": 0.08910369127988815, "learning_rate": 7.898786461544395e-05, "loss": 0.0055, "step": 33800 }, { "grad_norm": 0.10903053730726242, "learning_rate": 7.897439073673325e-05, "loss": 0.0071, "step": 33810 }, { "grad_norm": 0.13405585289001465, "learning_rate": 7.896091368942677e-05, "loss": 0.0055, "step": 33820 }, { "grad_norm": 0.12409459054470062, "learning_rate": 7.894743347499832e-05, "loss": 0.006, "step": 33830 }, { "grad_norm": 0.12278617918491364, "learning_rate": 7.893395009492203e-05, "loss": 0.0066, "step": 33840 }, { "grad_norm": 0.1060931459069252, "learning_rate": 7.892046355067248e-05, "loss": 0.006, "step": 33850 }, { "grad_norm": 0.11168690025806427, "learning_rate": 7.890697384372451e-05, "loss": 0.0063, "step": 33860 }, { "grad_norm": 0.11063462495803833, "learning_rate": 7.889348097555336e-05, "loss": 0.0067, "step": 33870 }, { "grad_norm": 0.14417998492717743, "learning_rate": 7.887998494763455e-05, "loss": 0.0063, "step": 33880 }, { "grad_norm": 0.12876257300376892, "learning_rate": 7.886648576144404e-05, "loss": 0.0059, "step": 33890 }, { "grad_norm": 0.1225515604019165, "learning_rate": 7.885298341845802e-05, "loss": 0.0065, "step": 33900 }, { "grad_norm": 0.16965551674365997, "learning_rate": 7.883947792015311e-05, "loss": 0.0065, "step": 33910 }, { "grad_norm": 0.18963435292243958, "learning_rate": 7.882596926800628e-05, "loss": 0.0065, "step": 33920 }, { "grad_norm": 0.10310326516628265, "learning_rate": 7.881245746349477e-05, "loss": 0.0061, "step": 33930 }, { "grad_norm": 0.09196997433900833, "learning_rate": 7.879894250809623e-05, "loss": 0.0051, "step": 33940 }, { "grad_norm": 0.1371212899684906, "learning_rate": 7.878542440328865e-05, "loss": 0.0072, "step": 33950 }, { "grad_norm": 0.1347423940896988, "learning_rate": 7.877190315055031e-05, "loss": 0.0065, "step": 33960 }, { "grad_norm": 0.13878977298736572, "learning_rate": 7.875837875135991e-05, "loss": 0.0083, "step": 33970 }, { "grad_norm": 0.1379505842924118, "learning_rate": 7.874485120719646e-05, "loss": 0.0069, "step": 33980 }, { "grad_norm": 0.12230727821588516, "learning_rate": 7.873132051953928e-05, "loss": 0.006, "step": 33990 }, { "grad_norm": 0.12935218214988708, "learning_rate": 7.87177866898681e-05, "loss": 0.0064, "step": 34000 }, { "grad_norm": 0.13631802797317505, "learning_rate": 7.870424971966294e-05, "loss": 0.0064, "step": 34010 }, { "grad_norm": 0.09757626056671143, "learning_rate": 7.869070961040419e-05, "loss": 0.0065, "step": 34020 }, { "grad_norm": 0.12246251106262207, "learning_rate": 7.867716636357257e-05, "loss": 0.0079, "step": 34030 }, { "grad_norm": 0.13995176553726196, "learning_rate": 7.866361998064915e-05, "loss": 0.0067, "step": 34040 }, { "grad_norm": 0.15549933910369873, "learning_rate": 7.865007046311534e-05, "loss": 0.0075, "step": 34050 }, { "grad_norm": 0.12559598684310913, "learning_rate": 7.86365178124529e-05, "loss": 0.007, "step": 34060 }, { "grad_norm": 0.07490981370210648, "learning_rate": 7.862296203014394e-05, "loss": 0.0062, "step": 34070 }, { "grad_norm": 0.17449365556240082, "learning_rate": 7.860940311767088e-05, "loss": 0.0067, "step": 34080 }, { "grad_norm": 0.131960928440094, "learning_rate": 7.85958410765165e-05, "loss": 0.0086, "step": 34090 }, { "grad_norm": 0.11517252027988434, "learning_rate": 7.858227590816394e-05, "loss": 0.0077, "step": 34100 }, { "grad_norm": 0.0818404033780098, "learning_rate": 7.856870761409664e-05, "loss": 0.007, "step": 34110 }, { "grad_norm": 0.14117716252803802, "learning_rate": 7.855513619579846e-05, "loss": 0.0068, "step": 34120 }, { "grad_norm": 0.10889007896184921, "learning_rate": 7.85415616547535e-05, "loss": 0.006, "step": 34130 }, { "grad_norm": 0.13380059599876404, "learning_rate": 7.852798399244627e-05, "loss": 0.0063, "step": 34140 }, { "grad_norm": 0.15119265019893646, "learning_rate": 7.851440321036161e-05, "loss": 0.0066, "step": 34150 }, { "grad_norm": 0.15078745782375336, "learning_rate": 7.850081930998468e-05, "loss": 0.0067, "step": 34160 }, { "grad_norm": 0.15043607354164124, "learning_rate": 7.8487232292801e-05, "loss": 0.0066, "step": 34170 }, { "grad_norm": 0.09404760599136353, "learning_rate": 7.847364216029642e-05, "loss": 0.0068, "step": 34180 }, { "grad_norm": 0.11696093529462814, "learning_rate": 7.846004891395716e-05, "loss": 0.0067, "step": 34190 }, { "grad_norm": 0.10026758164167404, "learning_rate": 7.844645255526972e-05, "loss": 0.0056, "step": 34200 }, { "grad_norm": 0.1325603574514389, "learning_rate": 7.843285308572101e-05, "loss": 0.007, "step": 34210 }, { "grad_norm": 0.13349772989749908, "learning_rate": 7.841925050679823e-05, "loss": 0.0071, "step": 34220 }, { "grad_norm": 0.13044774532318115, "learning_rate": 7.840564481998895e-05, "loss": 0.0063, "step": 34230 }, { "grad_norm": 0.10378646850585938, "learning_rate": 7.839203602678105e-05, "loss": 0.0056, "step": 34240 }, { "grad_norm": 0.11932441592216492, "learning_rate": 7.837842412866279e-05, "loss": 0.0056, "step": 34250 }, { "grad_norm": 0.0919903814792633, "learning_rate": 7.836480912712272e-05, "loss": 0.0063, "step": 34260 }, { "grad_norm": 0.10854244232177734, "learning_rate": 7.835119102364976e-05, "loss": 0.0063, "step": 34270 }, { "grad_norm": 0.14053495228290558, "learning_rate": 7.83375698197332e-05, "loss": 0.0059, "step": 34280 }, { "grad_norm": 0.08833645284175873, "learning_rate": 7.83239455168626e-05, "loss": 0.0055, "step": 34290 }, { "grad_norm": 0.11330215632915497, "learning_rate": 7.83103181165279e-05, "loss": 0.0064, "step": 34300 }, { "grad_norm": 0.1314748078584671, "learning_rate": 7.829668762021937e-05, "loss": 0.007, "step": 34310 }, { "grad_norm": 0.1065087839961052, "learning_rate": 7.828305402942764e-05, "loss": 0.0052, "step": 34320 }, { "grad_norm": 0.1352338194847107, "learning_rate": 7.826941734564363e-05, "loss": 0.0064, "step": 34330 }, { "grad_norm": 0.1205352395772934, "learning_rate": 7.825577757035865e-05, "loss": 0.0057, "step": 34340 }, { "grad_norm": 0.1400795727968216, "learning_rate": 7.824213470506431e-05, "loss": 0.0054, "step": 34350 }, { "grad_norm": 0.12524889409542084, "learning_rate": 7.822848875125257e-05, "loss": 0.0068, "step": 34360 }, { "grad_norm": 0.14097005128860474, "learning_rate": 7.821483971041576e-05, "loss": 0.0068, "step": 34370 }, { "grad_norm": 0.12515510618686676, "learning_rate": 7.820118758404649e-05, "loss": 0.008, "step": 34380 }, { "grad_norm": 0.12267692387104034, "learning_rate": 7.818753237363776e-05, "loss": 0.0079, "step": 34390 }, { "grad_norm": 0.12241930514574051, "learning_rate": 7.817387408068286e-05, "loss": 0.0074, "step": 34400 }, { "grad_norm": 0.10721238702535629, "learning_rate": 7.816021270667544e-05, "loss": 0.0064, "step": 34410 }, { "grad_norm": 0.118561290204525, "learning_rate": 7.81465482531095e-05, "loss": 0.0064, "step": 34420 }, { "grad_norm": 0.1529906988143921, "learning_rate": 7.813288072147938e-05, "loss": 0.0077, "step": 34430 }, { "grad_norm": 0.13121643662452698, "learning_rate": 7.811921011327972e-05, "loss": 0.0071, "step": 34440 }, { "grad_norm": 0.10871895402669907, "learning_rate": 7.810553643000549e-05, "loss": 0.0061, "step": 34450 }, { "grad_norm": 0.11882684379816055, "learning_rate": 7.809185967315206e-05, "loss": 0.0061, "step": 34460 }, { "grad_norm": 0.11050104349851608, "learning_rate": 7.80781798442151e-05, "loss": 0.0067, "step": 34470 }, { "grad_norm": 0.13163872063159943, "learning_rate": 7.806449694469061e-05, "loss": 0.0067, "step": 34480 }, { "grad_norm": 0.12924450635910034, "learning_rate": 7.805081097607492e-05, "loss": 0.0063, "step": 34490 }, { "grad_norm": 0.10969337821006775, "learning_rate": 7.803712193986474e-05, "loss": 0.0079, "step": 34500 }, { "grad_norm": 0.10354539006948471, "learning_rate": 7.802342983755702e-05, "loss": 0.0076, "step": 34510 }, { "grad_norm": 0.10672355443239212, "learning_rate": 7.800973467064918e-05, "loss": 0.0072, "step": 34520 }, { "grad_norm": 0.12130634486675262, "learning_rate": 7.799603644063884e-05, "loss": 0.0078, "step": 34530 }, { "grad_norm": 0.10888217389583588, "learning_rate": 7.798233514902405e-05, "loss": 0.0061, "step": 34540 }, { "grad_norm": 0.1529541313648224, "learning_rate": 7.796863079730318e-05, "loss": 0.0063, "step": 34550 }, { "grad_norm": 0.10879548639059067, "learning_rate": 7.795492338697488e-05, "loss": 0.0062, "step": 34560 }, { "grad_norm": 0.13302510976791382, "learning_rate": 7.794121291953819e-05, "loss": 0.0075, "step": 34570 }, { "grad_norm": 0.15601933002471924, "learning_rate": 7.792749939649246e-05, "loss": 0.0071, "step": 34580 }, { "grad_norm": 0.1098509132862091, "learning_rate": 7.79137828193374e-05, "loss": 0.0075, "step": 34590 }, { "grad_norm": 0.0815686285495758, "learning_rate": 7.790006318957301e-05, "loss": 0.0064, "step": 34600 }, { "grad_norm": 0.09483987092971802, "learning_rate": 7.788634050869965e-05, "loss": 0.0066, "step": 34610 }, { "grad_norm": 0.06691727787256241, "learning_rate": 7.787261477821803e-05, "loss": 0.0063, "step": 34620 }, { "grad_norm": 0.08516352623701096, "learning_rate": 7.785888599962916e-05, "loss": 0.0057, "step": 34630 }, { "grad_norm": 0.10396038740873337, "learning_rate": 7.784515417443439e-05, "loss": 0.0059, "step": 34640 }, { "grad_norm": 0.10256504267454147, "learning_rate": 7.783141930413545e-05, "loss": 0.0065, "step": 34650 }, { "grad_norm": 0.11051181703805923, "learning_rate": 7.78176813902343e-05, "loss": 0.0059, "step": 34660 }, { "grad_norm": 0.116960808634758, "learning_rate": 7.780394043423336e-05, "loss": 0.0074, "step": 34670 }, { "grad_norm": 0.10280166566371918, "learning_rate": 7.77901964376353e-05, "loss": 0.0066, "step": 34680 }, { "grad_norm": 0.089612677693367, "learning_rate": 7.777644940194316e-05, "loss": 0.0069, "step": 34690 }, { "grad_norm": 0.0918358787894249, "learning_rate": 7.776269932866023e-05, "loss": 0.0064, "step": 34700 }, { "grad_norm": 0.16089710593223572, "learning_rate": 7.774894621929026e-05, "loss": 0.0062, "step": 34710 }, { "grad_norm": 0.1142783984541893, "learning_rate": 7.773519007533725e-05, "loss": 0.0062, "step": 34720 }, { "grad_norm": 0.1346464604139328, "learning_rate": 7.772143089830556e-05, "loss": 0.0063, "step": 34730 }, { "grad_norm": 0.09125851094722748, "learning_rate": 7.770766868969985e-05, "loss": 0.0062, "step": 34740 }, { "grad_norm": 0.09479278326034546, "learning_rate": 7.769390345102518e-05, "loss": 0.0057, "step": 34750 }, { "grad_norm": 0.09486401081085205, "learning_rate": 7.768013518378683e-05, "loss": 0.0071, "step": 34760 }, { "grad_norm": 0.1051890179514885, "learning_rate": 7.766636388949053e-05, "loss": 0.0065, "step": 34770 }, { "grad_norm": 0.09385871887207031, "learning_rate": 7.765258956964229e-05, "loss": 0.007, "step": 34780 }, { "grad_norm": 0.16137346625328064, "learning_rate": 7.76388122257484e-05, "loss": 0.0065, "step": 34790 }, { "grad_norm": 0.11920218169689178, "learning_rate": 7.762503185931558e-05, "loss": 0.0058, "step": 34800 }, { "grad_norm": 0.10390755534172058, "learning_rate": 7.76112484718508e-05, "loss": 0.0068, "step": 34810 }, { "grad_norm": 0.1228112205862999, "learning_rate": 7.75974620648614e-05, "loss": 0.0057, "step": 34820 }, { "grad_norm": 0.11804575473070145, "learning_rate": 7.758367263985503e-05, "loss": 0.0059, "step": 34830 }, { "grad_norm": 0.0986645370721817, "learning_rate": 7.75698801983397e-05, "loss": 0.0048, "step": 34840 }, { "grad_norm": 0.13183291256427765, "learning_rate": 7.755608474182372e-05, "loss": 0.0062, "step": 34850 }, { "grad_norm": 0.12342321127653122, "learning_rate": 7.754228627181574e-05, "loss": 0.0068, "step": 34860 }, { "grad_norm": 0.13576790690422058, "learning_rate": 7.752848478982476e-05, "loss": 0.0063, "step": 34870 }, { "grad_norm": 0.1381283402442932, "learning_rate": 7.751468029736006e-05, "loss": 0.0065, "step": 34880 }, { "grad_norm": 0.09377340972423553, "learning_rate": 7.750087279593129e-05, "loss": 0.005, "step": 34890 }, { "grad_norm": 0.11712776124477386, "learning_rate": 7.748706228704843e-05, "loss": 0.0057, "step": 34900 }, { "grad_norm": 0.1298222541809082, "learning_rate": 7.747324877222176e-05, "loss": 0.0059, "step": 34910 }, { "grad_norm": 0.21800321340560913, "learning_rate": 7.745943225296188e-05, "loss": 0.0071, "step": 34920 }, { "grad_norm": 0.1749444454908371, "learning_rate": 7.744561273077981e-05, "loss": 0.0072, "step": 34930 }, { "grad_norm": 0.12685760855674744, "learning_rate": 7.743179020718678e-05, "loss": 0.0072, "step": 34940 }, { "grad_norm": 0.1292799562215805, "learning_rate": 7.741796468369443e-05, "loss": 0.0068, "step": 34950 }, { "grad_norm": 0.10344397276639938, "learning_rate": 7.740413616181466e-05, "loss": 0.0069, "step": 34960 }, { "grad_norm": 0.13140039145946503, "learning_rate": 7.739030464305978e-05, "loss": 0.006, "step": 34970 }, { "grad_norm": 0.11632467061281204, "learning_rate": 7.737647012894235e-05, "loss": 0.0062, "step": 34980 }, { "grad_norm": 0.11338090896606445, "learning_rate": 7.736263262097532e-05, "loss": 0.0053, "step": 34990 }, { "grad_norm": 0.10875946283340454, "learning_rate": 7.734879212067192e-05, "loss": 0.0063, "step": 35000 }, { "grad_norm": 0.09879937022924423, "learning_rate": 7.733494862954573e-05, "loss": 0.0054, "step": 35010 }, { "grad_norm": 0.09489672631025314, "learning_rate": 7.732110214911066e-05, "loss": 0.0058, "step": 35020 }, { "grad_norm": 0.09360397607088089, "learning_rate": 7.730725268088094e-05, "loss": 0.0057, "step": 35030 }, { "grad_norm": 0.1332530528306961, "learning_rate": 7.729340022637112e-05, "loss": 0.0062, "step": 35040 }, { "grad_norm": 0.11166030913591385, "learning_rate": 7.727954478709607e-05, "loss": 0.0058, "step": 35050 }, { "grad_norm": 0.11150902509689331, "learning_rate": 7.726568636457103e-05, "loss": 0.0052, "step": 35060 }, { "grad_norm": 0.09673286974430084, "learning_rate": 7.725182496031153e-05, "loss": 0.0052, "step": 35070 }, { "grad_norm": 0.12654714286327362, "learning_rate": 7.723796057583341e-05, "loss": 0.0068, "step": 35080 }, { "grad_norm": 0.11322834342718124, "learning_rate": 7.722409321265287e-05, "loss": 0.0058, "step": 35090 }, { "grad_norm": 0.12761905789375305, "learning_rate": 7.721022287228645e-05, "loss": 0.0051, "step": 35100 }, { "grad_norm": 0.13264024257659912, "learning_rate": 7.719634955625094e-05, "loss": 0.0068, "step": 35110 }, { "grad_norm": 0.14356178045272827, "learning_rate": 7.718247326606355e-05, "loss": 0.0083, "step": 35120 }, { "grad_norm": 0.12752538919448853, "learning_rate": 7.716859400324175e-05, "loss": 0.0059, "step": 35130 }, { "grad_norm": 0.14207114279270172, "learning_rate": 7.715471176930335e-05, "loss": 0.0065, "step": 35140 }, { "grad_norm": 0.1018320620059967, "learning_rate": 7.714082656576651e-05, "loss": 0.0058, "step": 35150 }, { "grad_norm": 0.11583293229341507, "learning_rate": 7.712693839414968e-05, "loss": 0.0065, "step": 35160 }, { "grad_norm": 0.11953124403953552, "learning_rate": 7.711304725597164e-05, "loss": 0.0065, "step": 35170 }, { "grad_norm": 0.16725598275661469, "learning_rate": 7.709915315275151e-05, "loss": 0.0059, "step": 35180 }, { "grad_norm": 0.16685223579406738, "learning_rate": 7.708525608600876e-05, "loss": 0.0069, "step": 35190 }, { "grad_norm": 0.13809694349765778, "learning_rate": 7.707135605726311e-05, "loss": 0.0072, "step": 35200 }, { "grad_norm": 0.1079254001379013, "learning_rate": 7.705745306803466e-05, "loss": 0.0059, "step": 35210 }, { "grad_norm": 0.11351119726896286, "learning_rate": 7.704354711984383e-05, "loss": 0.0074, "step": 35220 }, { "grad_norm": 0.1120767816901207, "learning_rate": 7.702963821421133e-05, "loss": 0.0061, "step": 35230 }, { "grad_norm": 0.1342155486345291, "learning_rate": 7.701572635265826e-05, "loss": 0.0076, "step": 35240 }, { "grad_norm": 0.10798729211091995, "learning_rate": 7.700181153670596e-05, "loss": 0.0052, "step": 35250 }, { "grad_norm": 0.09962303191423416, "learning_rate": 7.698789376787616e-05, "loss": 0.0053, "step": 35260 }, { "grad_norm": 0.14744950830936432, "learning_rate": 7.697397304769085e-05, "loss": 0.0067, "step": 35270 }, { "grad_norm": 0.1596262902021408, "learning_rate": 7.696004937767241e-05, "loss": 0.0064, "step": 35280 }, { "grad_norm": 0.13623487949371338, "learning_rate": 7.694612275934352e-05, "loss": 0.0064, "step": 35290 }, { "grad_norm": 0.10433323681354523, "learning_rate": 7.693219319422714e-05, "loss": 0.0073, "step": 35300 }, { "grad_norm": 0.13455712795257568, "learning_rate": 7.69182606838466e-05, "loss": 0.0071, "step": 35310 }, { "grad_norm": 0.11760633438825607, "learning_rate": 7.690432522972558e-05, "loss": 0.0063, "step": 35320 }, { "grad_norm": 0.1446632593870163, "learning_rate": 7.689038683338796e-05, "loss": 0.0065, "step": 35330 }, { "grad_norm": 0.14730150997638702, "learning_rate": 7.687644549635808e-05, "loss": 0.0067, "step": 35340 }, { "grad_norm": 0.1358775496482849, "learning_rate": 7.686250122016053e-05, "loss": 0.006, "step": 35350 }, { "grad_norm": 0.1385977864265442, "learning_rate": 7.684855400632022e-05, "loss": 0.0074, "step": 35360 }, { "grad_norm": 0.12291950732469559, "learning_rate": 7.683460385636243e-05, "loss": 0.0063, "step": 35370 }, { "grad_norm": 0.13759197294712067, "learning_rate": 7.68206507718127e-05, "loss": 0.0057, "step": 35380 }, { "grad_norm": 0.10849986225366592, "learning_rate": 7.680669475419693e-05, "loss": 0.0063, "step": 35390 }, { "grad_norm": 0.11799272149801254, "learning_rate": 7.679273580504132e-05, "loss": 0.0065, "step": 35400 }, { "grad_norm": 0.13348709046840668, "learning_rate": 7.67787739258724e-05, "loss": 0.0054, "step": 35410 }, { "grad_norm": 0.14521455764770508, "learning_rate": 7.676480911821705e-05, "loss": 0.006, "step": 35420 }, { "grad_norm": 0.10093061625957489, "learning_rate": 7.675084138360238e-05, "loss": 0.0057, "step": 35430 }, { "grad_norm": 0.09810534864664078, "learning_rate": 7.673687072355592e-05, "loss": 0.0064, "step": 35440 }, { "grad_norm": 0.10253936052322388, "learning_rate": 7.67228971396055e-05, "loss": 0.0051, "step": 35450 }, { "grad_norm": 0.11237771809101105, "learning_rate": 7.670892063327922e-05, "loss": 0.0063, "step": 35460 }, { "grad_norm": 0.15497805178165436, "learning_rate": 7.669494120610552e-05, "loss": 0.006, "step": 35470 }, { "grad_norm": 0.13329753279685974, "learning_rate": 7.66809588596132e-05, "loss": 0.0062, "step": 35480 }, { "grad_norm": 0.1253460794687271, "learning_rate": 7.666697359533132e-05, "loss": 0.0059, "step": 35490 }, { "grad_norm": 0.15232780575752258, "learning_rate": 7.665298541478932e-05, "loss": 0.0052, "step": 35500 }, { "grad_norm": 0.12435264140367508, "learning_rate": 7.66389943195169e-05, "loss": 0.0051, "step": 35510 }, { "grad_norm": 0.0850544348359108, "learning_rate": 7.66250003110441e-05, "loss": 0.0052, "step": 35520 }, { "grad_norm": 0.15992675721645355, "learning_rate": 7.66110033909013e-05, "loss": 0.0062, "step": 35530 }, { "grad_norm": 0.125422403216362, "learning_rate": 7.659700356061918e-05, "loss": 0.0057, "step": 35540 }, { "grad_norm": 0.13770076632499695, "learning_rate": 7.658300082172875e-05, "loss": 0.007, "step": 35550 }, { "grad_norm": 0.12196115404367447, "learning_rate": 7.65689951757613e-05, "loss": 0.0061, "step": 35560 }, { "grad_norm": 0.13783328235149384, "learning_rate": 7.655498662424849e-05, "loss": 0.0064, "step": 35570 }, { "grad_norm": 0.14415955543518066, "learning_rate": 7.654097516872227e-05, "loss": 0.0059, "step": 35580 }, { "grad_norm": 0.18884100019931793, "learning_rate": 7.65269608107149e-05, "loss": 0.007, "step": 35590 }, { "grad_norm": 0.09514040499925613, "learning_rate": 7.6512943551759e-05, "loss": 0.0068, "step": 35600 }, { "grad_norm": 0.08533284813165665, "learning_rate": 7.649892339338743e-05, "loss": 0.006, "step": 35610 }, { "grad_norm": 0.11185498535633087, "learning_rate": 7.648490033713344e-05, "loss": 0.0065, "step": 35620 }, { "grad_norm": 0.130506694316864, "learning_rate": 7.647087438453058e-05, "loss": 0.0071, "step": 35630 }, { "grad_norm": 0.12283299118280411, "learning_rate": 7.64568455371127e-05, "loss": 0.0058, "step": 35640 }, { "grad_norm": 0.1057375892996788, "learning_rate": 7.644281379641396e-05, "loss": 0.006, "step": 35650 }, { "grad_norm": 0.09179438650608063, "learning_rate": 7.642877916396887e-05, "loss": 0.0058, "step": 35660 }, { "grad_norm": 0.10082694888114929, "learning_rate": 7.641474164131221e-05, "loss": 0.0055, "step": 35670 }, { "grad_norm": 0.09338641911745071, "learning_rate": 7.640070122997913e-05, "loss": 0.0059, "step": 35680 }, { "grad_norm": 0.09636767208576202, "learning_rate": 7.638665793150505e-05, "loss": 0.007, "step": 35690 }, { "grad_norm": 0.13261191546916962, "learning_rate": 7.637261174742574e-05, "loss": 0.006, "step": 35700 }, { "grad_norm": 0.09275521337985992, "learning_rate": 7.635856267927725e-05, "loss": 0.006, "step": 35710 }, { "grad_norm": 0.0999288409948349, "learning_rate": 7.634451072859597e-05, "loss": 0.0055, "step": 35720 }, { "grad_norm": 0.10175952315330505, "learning_rate": 7.633045589691863e-05, "loss": 0.0063, "step": 35730 }, { "grad_norm": 0.09868396073579788, "learning_rate": 7.63163981857822e-05, "loss": 0.0056, "step": 35740 }, { "grad_norm": 0.12349274009466171, "learning_rate": 7.630233759672403e-05, "loss": 0.0063, "step": 35750 }, { "grad_norm": 0.10961664468050003, "learning_rate": 7.628827413128177e-05, "loss": 0.0066, "step": 35760 }, { "grad_norm": 0.11899574100971222, "learning_rate": 7.627420779099338e-05, "loss": 0.0055, "step": 35770 }, { "grad_norm": 0.11969181895256042, "learning_rate": 7.626013857739711e-05, "loss": 0.0055, "step": 35780 }, { "grad_norm": 0.1359441578388214, "learning_rate": 7.624606649203158e-05, "loss": 0.0058, "step": 35790 }, { "grad_norm": 0.10808342695236206, "learning_rate": 7.623199153643569e-05, "loss": 0.0052, "step": 35800 }, { "grad_norm": 0.09740789979696274, "learning_rate": 7.621791371214863e-05, "loss": 0.0051, "step": 35810 }, { "grad_norm": 0.11298861354589462, "learning_rate": 7.620383302070995e-05, "loss": 0.0059, "step": 35820 }, { "grad_norm": 0.13124465942382812, "learning_rate": 7.61897494636595e-05, "loss": 0.0059, "step": 35830 }, { "grad_norm": 0.10491568595170975, "learning_rate": 7.617566304253739e-05, "loss": 0.0053, "step": 35840 }, { "grad_norm": 0.10002461820840836, "learning_rate": 7.616157375888416e-05, "loss": 0.0052, "step": 35850 }, { "grad_norm": 0.10843578726053238, "learning_rate": 7.614748161424053e-05, "loss": 0.006, "step": 35860 }, { "grad_norm": 0.1310075968503952, "learning_rate": 7.613338661014763e-05, "loss": 0.0066, "step": 35870 }, { "grad_norm": 0.16593864560127258, "learning_rate": 7.611928874814686e-05, "loss": 0.0054, "step": 35880 }, { "grad_norm": 0.13723163306713104, "learning_rate": 7.610518802977996e-05, "loss": 0.0066, "step": 35890 }, { "grad_norm": 0.1291668564081192, "learning_rate": 7.609108445658893e-05, "loss": 0.0053, "step": 35900 }, { "grad_norm": 0.07890098541975021, "learning_rate": 7.607697803011612e-05, "loss": 0.0053, "step": 35910 }, { "grad_norm": 0.10803528875112534, "learning_rate": 7.606286875190421e-05, "loss": 0.0065, "step": 35920 }, { "grad_norm": 0.09269905835390091, "learning_rate": 7.604875662349617e-05, "loss": 0.0061, "step": 35930 }, { "grad_norm": 0.09776517748832703, "learning_rate": 7.603464164643525e-05, "loss": 0.0051, "step": 35940 }, { "grad_norm": 0.11961233615875244, "learning_rate": 7.602052382226507e-05, "loss": 0.0054, "step": 35950 }, { "grad_norm": 0.09818261116743088, "learning_rate": 7.600640315252954e-05, "loss": 0.0048, "step": 35960 }, { "grad_norm": 0.13387471437454224, "learning_rate": 7.599227963877284e-05, "loss": 0.0055, "step": 35970 }, { "grad_norm": 0.12491993606090546, "learning_rate": 7.597815328253954e-05, "loss": 0.0061, "step": 35980 }, { "grad_norm": 0.12412962317466736, "learning_rate": 7.596402408537443e-05, "loss": 0.0069, "step": 35990 }, { "grad_norm": 0.14429309964179993, "learning_rate": 7.59498920488227e-05, "loss": 0.0066, "step": 36000 }, { "grad_norm": 0.1076829805970192, "learning_rate": 7.593575717442979e-05, "loss": 0.0071, "step": 36010 }, { "grad_norm": 0.10404311120510101, "learning_rate": 7.592161946374147e-05, "loss": 0.0056, "step": 36020 }, { "grad_norm": 0.10055079311132431, "learning_rate": 7.590747891830381e-05, "loss": 0.0061, "step": 36030 }, { "grad_norm": 0.08753074705600739, "learning_rate": 7.58933355396632e-05, "loss": 0.0061, "step": 36040 }, { "grad_norm": 0.10868678241968155, "learning_rate": 7.587918932936636e-05, "loss": 0.006, "step": 36050 }, { "grad_norm": 0.1115369200706482, "learning_rate": 7.586504028896028e-05, "loss": 0.0067, "step": 36060 }, { "grad_norm": 0.1273895651102066, "learning_rate": 7.585088841999228e-05, "loss": 0.0054, "step": 36070 }, { "grad_norm": 0.14231516420841217, "learning_rate": 7.583673372400999e-05, "loss": 0.0056, "step": 36080 }, { "grad_norm": 0.14648154377937317, "learning_rate": 7.582257620256134e-05, "loss": 0.0059, "step": 36090 }, { "grad_norm": 0.13540002703666687, "learning_rate": 7.580841585719458e-05, "loss": 0.0056, "step": 36100 }, { "grad_norm": 0.12006038427352905, "learning_rate": 7.579425268945825e-05, "loss": 0.0061, "step": 36110 }, { "grad_norm": 0.13703030347824097, "learning_rate": 7.578008670090127e-05, "loss": 0.0062, "step": 36120 }, { "grad_norm": 0.12211182713508606, "learning_rate": 7.576591789307272e-05, "loss": 0.0069, "step": 36130 }, { "grad_norm": 0.13761399686336517, "learning_rate": 7.575174626752216e-05, "loss": 0.0067, "step": 36140 }, { "grad_norm": 0.12006507813930511, "learning_rate": 7.573757182579934e-05, "loss": 0.0054, "step": 36150 }, { "grad_norm": 0.13572953641414642, "learning_rate": 7.572339456945435e-05, "loss": 0.0066, "step": 36160 }, { "grad_norm": 0.11273593455553055, "learning_rate": 7.570921450003762e-05, "loss": 0.0058, "step": 36170 }, { "grad_norm": 0.14989975094795227, "learning_rate": 7.569503161909983e-05, "loss": 0.0051, "step": 36180 }, { "grad_norm": 0.11752274632453918, "learning_rate": 7.568084592819202e-05, "loss": 0.0058, "step": 36190 }, { "grad_norm": 0.11396855115890503, "learning_rate": 7.566665742886551e-05, "loss": 0.0065, "step": 36200 }, { "grad_norm": 0.1438816338777542, "learning_rate": 7.565246612267194e-05, "loss": 0.0063, "step": 36210 }, { "grad_norm": 0.1458183228969574, "learning_rate": 7.563827201116325e-05, "loss": 0.0063, "step": 36220 }, { "grad_norm": 0.1103445291519165, "learning_rate": 7.562407509589164e-05, "loss": 0.0062, "step": 36230 }, { "grad_norm": 0.09794566035270691, "learning_rate": 7.560987537840973e-05, "loss": 0.0052, "step": 36240 }, { "grad_norm": 0.10150843113660812, "learning_rate": 7.559567286027036e-05, "loss": 0.0053, "step": 36250 }, { "grad_norm": 0.11198678612709045, "learning_rate": 7.558146754302668e-05, "loss": 0.0059, "step": 36260 }, { "grad_norm": 0.1387559473514557, "learning_rate": 7.556725942823217e-05, "loss": 0.0062, "step": 36270 }, { "grad_norm": 0.11383487284183502, "learning_rate": 7.555304851744061e-05, "loss": 0.0068, "step": 36280 }, { "grad_norm": 0.11653272807598114, "learning_rate": 7.553883481220608e-05, "loss": 0.0066, "step": 36290 }, { "grad_norm": 0.12985864281654358, "learning_rate": 7.552461831408298e-05, "loss": 0.0077, "step": 36300 }, { "grad_norm": 0.0926237478852272, "learning_rate": 7.551039902462599e-05, "loss": 0.0061, "step": 36310 }, { "grad_norm": 0.1418367475271225, "learning_rate": 7.549617694539011e-05, "loss": 0.0074, "step": 36320 }, { "grad_norm": 0.14373426139354706, "learning_rate": 7.548195207793066e-05, "loss": 0.0076, "step": 36330 }, { "grad_norm": 0.11734005808830261, "learning_rate": 7.546772442380323e-05, "loss": 0.0065, "step": 36340 }, { "grad_norm": 0.13950848579406738, "learning_rate": 7.545349398456376e-05, "loss": 0.0066, "step": 36350 }, { "grad_norm": 0.11165682226419449, "learning_rate": 7.543926076176845e-05, "loss": 0.006, "step": 36360 }, { "grad_norm": 0.11496810615062714, "learning_rate": 7.542502475697385e-05, "loss": 0.006, "step": 36370 }, { "grad_norm": 0.13959918916225433, "learning_rate": 7.541078597173675e-05, "loss": 0.0066, "step": 36380 }, { "grad_norm": 0.10158625990152359, "learning_rate": 7.539654440761431e-05, "loss": 0.007, "step": 36390 }, { "grad_norm": 0.16620126366615295, "learning_rate": 7.538230006616395e-05, "loss": 0.0069, "step": 36400 }, { "grad_norm": 0.10984403640031815, "learning_rate": 7.536805294894342e-05, "loss": 0.0057, "step": 36410 }, { "grad_norm": 0.08917874097824097, "learning_rate": 7.535380305751076e-05, "loss": 0.0054, "step": 36420 }, { "grad_norm": 0.12820148468017578, "learning_rate": 7.533955039342431e-05, "loss": 0.0059, "step": 36430 }, { "grad_norm": 0.11788023263216019, "learning_rate": 7.532529495824274e-05, "loss": 0.0062, "step": 36440 }, { "grad_norm": 0.09852693229913712, "learning_rate": 7.531103675352497e-05, "loss": 0.006, "step": 36450 }, { "grad_norm": 0.13794606924057007, "learning_rate": 7.529677578083029e-05, "loss": 0.0066, "step": 36460 }, { "grad_norm": 0.16700516641139984, "learning_rate": 7.528251204171823e-05, "loss": 0.0059, "step": 36470 }, { "grad_norm": 0.12388519197702408, "learning_rate": 7.526824553774867e-05, "loss": 0.0066, "step": 36480 }, { "grad_norm": 0.11774983257055283, "learning_rate": 7.525397627048178e-05, "loss": 0.0062, "step": 36490 }, { "grad_norm": 0.11851491779088974, "learning_rate": 7.523970424147802e-05, "loss": 0.0066, "step": 36500 }, { "grad_norm": 0.09873486310243607, "learning_rate": 7.522542945229813e-05, "loss": 0.0062, "step": 36510 }, { "grad_norm": 0.11046583205461502, "learning_rate": 7.521115190450321e-05, "loss": 0.007, "step": 36520 }, { "grad_norm": 0.10894443839788437, "learning_rate": 7.519687159965462e-05, "loss": 0.0064, "step": 36530 }, { "grad_norm": 0.11272689700126648, "learning_rate": 7.518258853931403e-05, "loss": 0.0068, "step": 36540 }, { "grad_norm": 0.16956019401550293, "learning_rate": 7.516830272504342e-05, "loss": 0.0062, "step": 36550 }, { "grad_norm": 0.11412351578474045, "learning_rate": 7.515401415840509e-05, "loss": 0.0086, "step": 36560 }, { "grad_norm": 0.1222461387515068, "learning_rate": 7.513972284096155e-05, "loss": 0.0061, "step": 36570 }, { "grad_norm": 0.09770426154136658, "learning_rate": 7.512542877427574e-05, "loss": 0.0056, "step": 36580 }, { "grad_norm": 0.11352312564849854, "learning_rate": 7.51111319599108e-05, "loss": 0.0062, "step": 36590 }, { "grad_norm": 0.09659270197153091, "learning_rate": 7.509683239943024e-05, "loss": 0.0065, "step": 36600 }, { "grad_norm": 0.11506203562021255, "learning_rate": 7.50825300943978e-05, "loss": 0.0071, "step": 36610 }, { "grad_norm": 0.11150076240301132, "learning_rate": 7.50682250463776e-05, "loss": 0.0062, "step": 36620 }, { "grad_norm": 0.09939960390329361, "learning_rate": 7.5053917256934e-05, "loss": 0.0064, "step": 36630 }, { "grad_norm": 0.141446053981781, "learning_rate": 7.503960672763166e-05, "loss": 0.0074, "step": 36640 }, { "grad_norm": 0.14001169800758362, "learning_rate": 7.502529346003559e-05, "loss": 0.0058, "step": 36650 }, { "grad_norm": 0.10778339952230453, "learning_rate": 7.501097745571106e-05, "loss": 0.0056, "step": 36660 }, { "grad_norm": 0.11133506149053574, "learning_rate": 7.499665871622361e-05, "loss": 0.0063, "step": 36670 }, { "grad_norm": 0.09503626823425293, "learning_rate": 7.498233724313917e-05, "loss": 0.0054, "step": 36680 }, { "grad_norm": 0.1127639189362526, "learning_rate": 7.496801303802387e-05, "loss": 0.0063, "step": 36690 }, { "grad_norm": 0.08539437502622604, "learning_rate": 7.495368610244423e-05, "loss": 0.0075, "step": 36700 }, { "grad_norm": 0.10891201347112656, "learning_rate": 7.493935643796697e-05, "loss": 0.0055, "step": 36710 }, { "grad_norm": 0.1492001712322235, "learning_rate": 7.49250240461592e-05, "loss": 0.0068, "step": 36720 }, { "grad_norm": 0.13757193088531494, "learning_rate": 7.491068892858826e-05, "loss": 0.0066, "step": 36730 }, { "grad_norm": 0.13605473935604095, "learning_rate": 7.489635108682184e-05, "loss": 0.0063, "step": 36740 }, { "grad_norm": 0.10558265447616577, "learning_rate": 7.48820105224279e-05, "loss": 0.0065, "step": 36750 }, { "grad_norm": 0.1673320084810257, "learning_rate": 7.48676672369747e-05, "loss": 0.007, "step": 36760 }, { "grad_norm": 0.11093104630708694, "learning_rate": 7.485332123203079e-05, "loss": 0.006, "step": 36770 }, { "grad_norm": 0.09739326685667038, "learning_rate": 7.483897250916504e-05, "loss": 0.0052, "step": 36780 }, { "grad_norm": 0.11447561532258987, "learning_rate": 7.48246210699466e-05, "loss": 0.0058, "step": 36790 }, { "grad_norm": 0.14336061477661133, "learning_rate": 7.481026691594492e-05, "loss": 0.0066, "step": 36800 }, { "grad_norm": 0.1299479454755783, "learning_rate": 7.479591004872975e-05, "loss": 0.0063, "step": 36810 }, { "grad_norm": 0.155448779463768, "learning_rate": 7.478155046987114e-05, "loss": 0.0074, "step": 36820 }, { "grad_norm": 0.14339999854564667, "learning_rate": 7.476718818093944e-05, "loss": 0.0062, "step": 36830 }, { "grad_norm": 0.11522053927183151, "learning_rate": 7.475282318350527e-05, "loss": 0.0065, "step": 36840 }, { "grad_norm": 0.10876724869012833, "learning_rate": 7.473845547913957e-05, "loss": 0.0059, "step": 36850 }, { "grad_norm": 0.137939453125, "learning_rate": 7.472408506941357e-05, "loss": 0.0064, "step": 36860 }, { "grad_norm": 0.11618257313966751, "learning_rate": 7.470971195589881e-05, "loss": 0.0078, "step": 36870 }, { "grad_norm": 0.12415950745344162, "learning_rate": 7.46953361401671e-05, "loss": 0.0071, "step": 36880 }, { "grad_norm": 0.13990280032157898, "learning_rate": 7.468095762379055e-05, "loss": 0.0068, "step": 36890 }, { "grad_norm": 0.1286458522081375, "learning_rate": 7.466657640834158e-05, "loss": 0.007, "step": 36900 }, { "grad_norm": 0.0997919887304306, "learning_rate": 7.46521924953929e-05, "loss": 0.0055, "step": 36910 }, { "grad_norm": 0.1217709332704544, "learning_rate": 7.463780588651752e-05, "loss": 0.0066, "step": 36920 }, { "grad_norm": 0.10099524259567261, "learning_rate": 7.462341658328873e-05, "loss": 0.0059, "step": 36930 }, { "grad_norm": 0.09283298254013062, "learning_rate": 7.460902458728012e-05, "loss": 0.0062, "step": 36940 }, { "grad_norm": 0.10029538720846176, "learning_rate": 7.459462990006558e-05, "loss": 0.0065, "step": 36950 }, { "grad_norm": 0.08681619167327881, "learning_rate": 7.458023252321929e-05, "loss": 0.0057, "step": 36960 }, { "grad_norm": 0.1199355274438858, "learning_rate": 7.456583245831573e-05, "loss": 0.0055, "step": 36970 }, { "grad_norm": 0.10797956585884094, "learning_rate": 7.455142970692968e-05, "loss": 0.0064, "step": 36980 }, { "grad_norm": 0.14753253757953644, "learning_rate": 7.453702427063615e-05, "loss": 0.0076, "step": 36990 }, { "grad_norm": 0.12937891483306885, "learning_rate": 7.452261615101057e-05, "loss": 0.0073, "step": 37000 }, { "grad_norm": 0.12350115925073624, "learning_rate": 7.450820534962856e-05, "loss": 0.007, "step": 37010 }, { "grad_norm": 0.08832090348005295, "learning_rate": 7.449379186806604e-05, "loss": 0.0056, "step": 37020 }, { "grad_norm": 0.11206740885972977, "learning_rate": 7.44793757078993e-05, "loss": 0.007, "step": 37030 }, { "grad_norm": 0.11938875913619995, "learning_rate": 7.44649568707048e-05, "loss": 0.0064, "step": 37040 }, { "grad_norm": 0.10423044115304947, "learning_rate": 7.445053535805942e-05, "loss": 0.0058, "step": 37050 }, { "grad_norm": 0.10693187266588211, "learning_rate": 7.443611117154024e-05, "loss": 0.0063, "step": 37060 }, { "grad_norm": 0.1108604297041893, "learning_rate": 7.442168431272469e-05, "loss": 0.0051, "step": 37070 }, { "grad_norm": 0.10983329266309738, "learning_rate": 7.440725478319044e-05, "loss": 0.0072, "step": 37080 }, { "grad_norm": 0.12790772318840027, "learning_rate": 7.439282258451553e-05, "loss": 0.0059, "step": 37090 }, { "grad_norm": 0.13204064965248108, "learning_rate": 7.43783877182782e-05, "loss": 0.0066, "step": 37100 }, { "grad_norm": 0.09724520146846771, "learning_rate": 7.436395018605707e-05, "loss": 0.0061, "step": 37110 }, { "grad_norm": 0.09602734446525574, "learning_rate": 7.434950998943094e-05, "loss": 0.0072, "step": 37120 }, { "grad_norm": 0.12353581190109253, "learning_rate": 7.433506712997902e-05, "loss": 0.0053, "step": 37130 }, { "grad_norm": 0.11347132921218872, "learning_rate": 7.432062160928077e-05, "loss": 0.0066, "step": 37140 }, { "grad_norm": 0.09529774636030197, "learning_rate": 7.430617342891588e-05, "loss": 0.006, "step": 37150 }, { "grad_norm": 0.09369580447673798, "learning_rate": 7.429172259046442e-05, "loss": 0.0054, "step": 37160 }, { "grad_norm": 0.10749886929988861, "learning_rate": 7.42772690955067e-05, "loss": 0.006, "step": 37170 }, { "grad_norm": 0.12162792682647705, "learning_rate": 7.426281294562332e-05, "loss": 0.0055, "step": 37180 }, { "grad_norm": 0.08003608137369156, "learning_rate": 7.424835414239519e-05, "loss": 0.0056, "step": 37190 }, { "grad_norm": 0.0801018550992012, "learning_rate": 7.423389268740352e-05, "loss": 0.006, "step": 37200 }, { "grad_norm": 0.11545222997665405, "learning_rate": 7.42194285822298e-05, "loss": 0.0065, "step": 37210 }, { "grad_norm": 0.09457919001579285, "learning_rate": 7.420496182845576e-05, "loss": 0.0059, "step": 37220 }, { "grad_norm": 0.15143252909183502, "learning_rate": 7.419049242766348e-05, "loss": 0.006, "step": 37230 }, { "grad_norm": 0.08930440992116928, "learning_rate": 7.417602038143533e-05, "loss": 0.0053, "step": 37240 }, { "grad_norm": 0.08484935015439987, "learning_rate": 7.416154569135393e-05, "loss": 0.0058, "step": 37250 }, { "grad_norm": 0.14504918456077576, "learning_rate": 7.414706835900222e-05, "loss": 0.0068, "step": 37260 }, { "grad_norm": 0.12093955278396606, "learning_rate": 7.413258838596345e-05, "loss": 0.0063, "step": 37270 }, { "grad_norm": 0.1504545658826828, "learning_rate": 7.411810577382106e-05, "loss": 0.0063, "step": 37280 }, { "grad_norm": 0.15230019390583038, "learning_rate": 7.410362052415889e-05, "loss": 0.0059, "step": 37290 }, { "grad_norm": 0.1410602480173111, "learning_rate": 7.408913263856102e-05, "loss": 0.0071, "step": 37300 }, { "grad_norm": 0.15622903406620026, "learning_rate": 7.407464211861183e-05, "loss": 0.0068, "step": 37310 }, { "grad_norm": 0.10796761512756348, "learning_rate": 7.406014896589597e-05, "loss": 0.0054, "step": 37320 }, { "grad_norm": 0.1145251914858818, "learning_rate": 7.404565318199838e-05, "loss": 0.0069, "step": 37330 }, { "grad_norm": 0.12686535716056824, "learning_rate": 7.403115476850432e-05, "loss": 0.0068, "step": 37340 }, { "grad_norm": 0.11582756042480469, "learning_rate": 7.401665372699932e-05, "loss": 0.007, "step": 37350 }, { "grad_norm": 0.10448924452066422, "learning_rate": 7.400215005906916e-05, "loss": 0.0066, "step": 37360 }, { "grad_norm": 0.09856623411178589, "learning_rate": 7.398764376629996e-05, "loss": 0.0071, "step": 37370 }, { "grad_norm": 0.11323129385709763, "learning_rate": 7.39731348502781e-05, "loss": 0.0059, "step": 37380 }, { "grad_norm": 0.09396188706159592, "learning_rate": 7.395862331259028e-05, "loss": 0.0062, "step": 37390 }, { "grad_norm": 0.09897931665182114, "learning_rate": 7.394410915482341e-05, "loss": 0.0057, "step": 37400 }, { "grad_norm": 0.0651230588555336, "learning_rate": 7.392959237856476e-05, "loss": 0.0049, "step": 37410 }, { "grad_norm": 0.11391793191432953, "learning_rate": 7.391507298540187e-05, "loss": 0.0063, "step": 37420 }, { "grad_norm": 0.12124557048082352, "learning_rate": 7.390055097692257e-05, "loss": 0.0053, "step": 37430 }, { "grad_norm": 0.12877273559570312, "learning_rate": 7.388602635471494e-05, "loss": 0.0048, "step": 37440 }, { "grad_norm": 0.112977035343647, "learning_rate": 7.387149912036738e-05, "loss": 0.0063, "step": 37450 }, { "grad_norm": 0.1137915626168251, "learning_rate": 7.385696927546854e-05, "loss": 0.0068, "step": 37460 }, { "grad_norm": 0.12082614749670029, "learning_rate": 7.384243682160743e-05, "loss": 0.0069, "step": 37470 }, { "grad_norm": 0.14285549521446228, "learning_rate": 7.382790176037329e-05, "loss": 0.0058, "step": 37480 }, { "grad_norm": 0.11557119339704514, "learning_rate": 7.381336409335562e-05, "loss": 0.0066, "step": 37490 }, { "grad_norm": 0.10839226096868515, "learning_rate": 7.379882382214426e-05, "loss": 0.006, "step": 37500 }, { "grad_norm": 0.11437410861253738, "learning_rate": 7.378428094832931e-05, "loss": 0.0058, "step": 37510 }, { "grad_norm": 0.11403106898069382, "learning_rate": 7.376973547350114e-05, "loss": 0.0062, "step": 37520 }, { "grad_norm": 0.11495572328567505, "learning_rate": 7.375518739925046e-05, "loss": 0.0055, "step": 37530 }, { "grad_norm": 0.102845199406147, "learning_rate": 7.374063672716818e-05, "loss": 0.0057, "step": 37540 }, { "grad_norm": 0.13492153584957123, "learning_rate": 7.372608345884558e-05, "loss": 0.0068, "step": 37550 }, { "grad_norm": 0.10424568504095078, "learning_rate": 7.371152759587415e-05, "loss": 0.0064, "step": 37560 }, { "grad_norm": 0.11182733625173569, "learning_rate": 7.369696913984572e-05, "loss": 0.0054, "step": 37570 }, { "grad_norm": 0.1316032111644745, "learning_rate": 7.368240809235237e-05, "loss": 0.0063, "step": 37580 }, { "grad_norm": 0.12060572952032089, "learning_rate": 7.366784445498649e-05, "loss": 0.0059, "step": 37590 }, { "grad_norm": 0.10468947142362595, "learning_rate": 7.36532782293407e-05, "loss": 0.0051, "step": 37600 }, { "grad_norm": 0.08506745100021362, "learning_rate": 7.363870941700797e-05, "loss": 0.0052, "step": 37610 }, { "grad_norm": 0.09239199012517929, "learning_rate": 7.362413801958152e-05, "loss": 0.0055, "step": 37620 }, { "grad_norm": 0.12391476333141327, "learning_rate": 7.360956403865486e-05, "loss": 0.0053, "step": 37630 }, { "grad_norm": 0.07882674783468246, "learning_rate": 7.359498747582177e-05, "loss": 0.0051, "step": 37640 }, { "grad_norm": 0.12753765285015106, "learning_rate": 7.358040833267634e-05, "loss": 0.005, "step": 37650 }, { "grad_norm": 0.10221296548843384, "learning_rate": 7.356582661081288e-05, "loss": 0.0053, "step": 37660 }, { "grad_norm": 0.09682764112949371, "learning_rate": 7.355124231182607e-05, "loss": 0.0046, "step": 37670 }, { "grad_norm": 0.0930761992931366, "learning_rate": 7.353665543731083e-05, "loss": 0.0047, "step": 37680 }, { "grad_norm": 0.08242903649806976, "learning_rate": 7.352206598886229e-05, "loss": 0.005, "step": 37690 }, { "grad_norm": 0.08367427438497543, "learning_rate": 7.350747396807601e-05, "loss": 0.0053, "step": 37700 }, { "grad_norm": 0.15704859793186188, "learning_rate": 7.349287937654772e-05, "loss": 0.0059, "step": 37710 }, { "grad_norm": 0.13873982429504395, "learning_rate": 7.347828221587345e-05, "loss": 0.0061, "step": 37720 }, { "grad_norm": 0.1388111710548401, "learning_rate": 7.346368248764955e-05, "loss": 0.0053, "step": 37730 }, { "grad_norm": 0.0999591276049614, "learning_rate": 7.34490801934726e-05, "loss": 0.0052, "step": 37740 }, { "grad_norm": 0.15327167510986328, "learning_rate": 7.343447533493947e-05, "loss": 0.0068, "step": 37750 }, { "grad_norm": 0.12024202197790146, "learning_rate": 7.341986791364739e-05, "loss": 0.0062, "step": 37760 }, { "grad_norm": 0.10615366697311401, "learning_rate": 7.340525793119374e-05, "loss": 0.006, "step": 37770 }, { "grad_norm": 0.07091643661260605, "learning_rate": 7.33906453891763e-05, "loss": 0.0046, "step": 37780 }, { "grad_norm": 0.1340220421552658, "learning_rate": 7.337603028919303e-05, "loss": 0.0058, "step": 37790 }, { "grad_norm": 0.09290491789579391, "learning_rate": 7.336141263284225e-05, "loss": 0.0054, "step": 37800 }, { "grad_norm": 0.09778096526861191, "learning_rate": 7.334679242172249e-05, "loss": 0.0062, "step": 37810 }, { "grad_norm": 0.09765278548002243, "learning_rate": 7.33321696574326e-05, "loss": 0.0057, "step": 37820 }, { "grad_norm": 0.13651853799819946, "learning_rate": 7.331754434157174e-05, "loss": 0.007, "step": 37830 }, { "grad_norm": 0.11153820902109146, "learning_rate": 7.330291647573928e-05, "loss": 0.0064, "step": 37840 }, { "grad_norm": 0.14490196108818054, "learning_rate": 7.32882860615349e-05, "loss": 0.0057, "step": 37850 }, { "grad_norm": 0.12809105217456818, "learning_rate": 7.327365310055858e-05, "loss": 0.006, "step": 37860 }, { "grad_norm": 0.11469395458698273, "learning_rate": 7.325901759441055e-05, "loss": 0.0059, "step": 37870 }, { "grad_norm": 0.11935126036405563, "learning_rate": 7.324437954469132e-05, "loss": 0.0062, "step": 37880 }, { "grad_norm": 0.12650208175182343, "learning_rate": 7.32297389530017e-05, "loss": 0.0062, "step": 37890 }, { "grad_norm": 0.14109931886196136, "learning_rate": 7.321509582094274e-05, "loss": 0.006, "step": 37900 }, { "grad_norm": 0.09318337589502335, "learning_rate": 7.320045015011582e-05, "loss": 0.0066, "step": 37910 }, { "grad_norm": 0.10580393671989441, "learning_rate": 7.318580194212253e-05, "loss": 0.0062, "step": 37920 }, { "grad_norm": 0.15257132053375244, "learning_rate": 7.317115119856483e-05, "loss": 0.0057, "step": 37930 }, { "grad_norm": 0.10268186777830124, "learning_rate": 7.315649792104485e-05, "loss": 0.0053, "step": 37940 }, { "grad_norm": 0.1800701916217804, "learning_rate": 7.314184211116507e-05, "loss": 0.0064, "step": 37950 }, { "grad_norm": 0.10882632434368134, "learning_rate": 7.312718377052823e-05, "loss": 0.0062, "step": 37960 }, { "grad_norm": 0.10665184259414673, "learning_rate": 7.311252290073737e-05, "loss": 0.0062, "step": 37970 }, { "grad_norm": 0.13040611147880554, "learning_rate": 7.309785950339572e-05, "loss": 0.0063, "step": 37980 }, { "grad_norm": 0.13393619656562805, "learning_rate": 7.308319358010692e-05, "loss": 0.0076, "step": 37990 }, { "grad_norm": 0.12956292927265167, "learning_rate": 7.306852513247474e-05, "loss": 0.0071, "step": 38000 }, { "grad_norm": 0.1569199115037918, "learning_rate": 7.305385416210336e-05, "loss": 0.0066, "step": 38010 }, { "grad_norm": 0.13865959644317627, "learning_rate": 7.303918067059711e-05, "loss": 0.0069, "step": 38020 }, { "grad_norm": 0.12164576351642609, "learning_rate": 7.302450465956075e-05, "loss": 0.0068, "step": 38030 }, { "grad_norm": 0.10733721405267715, "learning_rate": 7.300982613059914e-05, "loss": 0.0054, "step": 38040 }, { "grad_norm": 0.14004214107990265, "learning_rate": 7.299514508531757e-05, "loss": 0.0074, "step": 38050 }, { "grad_norm": 0.10100464522838593, "learning_rate": 7.298046152532148e-05, "loss": 0.0066, "step": 38060 }, { "grad_norm": 0.11846383661031723, "learning_rate": 7.296577545221668e-05, "loss": 0.007, "step": 38070 }, { "grad_norm": 0.1254892349243164, "learning_rate": 7.29510868676092e-05, "loss": 0.0063, "step": 38080 }, { "grad_norm": 0.10888401418924332, "learning_rate": 7.293639577310538e-05, "loss": 0.0054, "step": 38090 }, { "grad_norm": 0.10044363141059875, "learning_rate": 7.292170217031179e-05, "loss": 0.0066, "step": 38100 }, { "grad_norm": 0.12404169142246246, "learning_rate": 7.290700606083532e-05, "loss": 0.0053, "step": 38110 }, { "grad_norm": 0.11056475341320038, "learning_rate": 7.28923074462831e-05, "loss": 0.0061, "step": 38120 }, { "grad_norm": 0.09566143155097961, "learning_rate": 7.287760632826255e-05, "loss": 0.0059, "step": 38130 }, { "grad_norm": 0.12832757830619812, "learning_rate": 7.286290270838139e-05, "loss": 0.0061, "step": 38140 }, { "grad_norm": 0.08698752522468567, "learning_rate": 7.284819658824756e-05, "loss": 0.0053, "step": 38150 }, { "grad_norm": 0.1156487911939621, "learning_rate": 7.283348796946929e-05, "loss": 0.0051, "step": 38160 }, { "grad_norm": 0.08805545419454575, "learning_rate": 7.28187768536551e-05, "loss": 0.0057, "step": 38170 }, { "grad_norm": 0.08831384032964706, "learning_rate": 7.280406324241379e-05, "loss": 0.0058, "step": 38180 }, { "grad_norm": 0.10750152915716171, "learning_rate": 7.278934713735441e-05, "loss": 0.0055, "step": 38190 }, { "grad_norm": 0.10237433761358261, "learning_rate": 7.277462854008629e-05, "loss": 0.0062, "step": 38200 }, { "grad_norm": 0.10504653304815292, "learning_rate": 7.275990745221904e-05, "loss": 0.0049, "step": 38210 }, { "grad_norm": 0.11946065723896027, "learning_rate": 7.274518387536252e-05, "loss": 0.0057, "step": 38220 }, { "grad_norm": 0.09439391642808914, "learning_rate": 7.273045781112689e-05, "loss": 0.0048, "step": 38230 }, { "grad_norm": 0.0860804095864296, "learning_rate": 7.271572926112257e-05, "loss": 0.0054, "step": 38240 }, { "grad_norm": 0.17195716500282288, "learning_rate": 7.270099822696024e-05, "loss": 0.0071, "step": 38250 }, { "grad_norm": 0.11454486846923828, "learning_rate": 7.26862647102509e-05, "loss": 0.0066, "step": 38260 }, { "grad_norm": 0.08808185160160065, "learning_rate": 7.267152871260573e-05, "loss": 0.006, "step": 38270 }, { "grad_norm": 0.11455556750297546, "learning_rate": 7.265679023563629e-05, "loss": 0.0058, "step": 38280 }, { "grad_norm": 0.12753058969974518, "learning_rate": 7.264204928095433e-05, "loss": 0.0058, "step": 38290 }, { "grad_norm": 0.09710785001516342, "learning_rate": 7.262730585017188e-05, "loss": 0.0062, "step": 38300 }, { "grad_norm": 0.13962583243846893, "learning_rate": 7.261255994490131e-05, "loss": 0.0048, "step": 38310 }, { "grad_norm": 0.12649782001972198, "learning_rate": 7.259781156675517e-05, "loss": 0.0054, "step": 38320 }, { "grad_norm": 0.10611587017774582, "learning_rate": 7.258306071734633e-05, "loss": 0.0049, "step": 38330 }, { "grad_norm": 0.09747183322906494, "learning_rate": 7.256830739828792e-05, "loss": 0.0056, "step": 38340 }, { "grad_norm": 0.11565370857715607, "learning_rate": 7.255355161119336e-05, "loss": 0.0051, "step": 38350 }, { "grad_norm": 0.1168985515832901, "learning_rate": 7.253879335767628e-05, "loss": 0.0057, "step": 38360 }, { "grad_norm": 0.10204081237316132, "learning_rate": 7.252403263935067e-05, "loss": 0.006, "step": 38370 }, { "grad_norm": 0.09338969737291336, "learning_rate": 7.25092694578307e-05, "loss": 0.0051, "step": 38380 }, { "grad_norm": 0.10742547363042831, "learning_rate": 7.249450381473087e-05, "loss": 0.0052, "step": 38390 }, { "grad_norm": 0.10909625887870789, "learning_rate": 7.247973571166593e-05, "loss": 0.0057, "step": 38400 }, { "grad_norm": 0.08641140162944794, "learning_rate": 7.246496515025089e-05, "loss": 0.0052, "step": 38410 }, { "grad_norm": 0.0983629897236824, "learning_rate": 7.245019213210101e-05, "loss": 0.0052, "step": 38420 }, { "grad_norm": 0.10906147211790085, "learning_rate": 7.24354166588319e-05, "loss": 0.0055, "step": 38430 }, { "grad_norm": 0.14115838706493378, "learning_rate": 7.242063873205935e-05, "loss": 0.005, "step": 38440 }, { "grad_norm": 0.08148624002933502, "learning_rate": 7.240585835339946e-05, "loss": 0.0054, "step": 38450 }, { "grad_norm": 0.09700973331928253, "learning_rate": 7.239107552446857e-05, "loss": 0.0066, "step": 38460 }, { "grad_norm": 0.10274405777454376, "learning_rate": 7.237629024688334e-05, "loss": 0.0056, "step": 38470 }, { "grad_norm": 0.09711938351392746, "learning_rate": 7.236150252226064e-05, "loss": 0.0052, "step": 38480 }, { "grad_norm": 0.14088916778564453, "learning_rate": 7.234671235221765e-05, "loss": 0.0062, "step": 38490 }, { "grad_norm": 0.125071182847023, "learning_rate": 7.233191973837179e-05, "loss": 0.0064, "step": 38500 }, { "grad_norm": 0.1227295994758606, "learning_rate": 7.231712468234078e-05, "loss": 0.0056, "step": 38510 }, { "grad_norm": 0.14208553731441498, "learning_rate": 7.230232718574254e-05, "loss": 0.0073, "step": 38520 }, { "grad_norm": 0.14084388315677643, "learning_rate": 7.228752725019535e-05, "loss": 0.0056, "step": 38530 }, { "grad_norm": 0.11270364373922348, "learning_rate": 7.227272487731769e-05, "loss": 0.0066, "step": 38540 }, { "grad_norm": 0.08358514308929443, "learning_rate": 7.225792006872831e-05, "loss": 0.0057, "step": 38550 }, { "grad_norm": 0.08108430355787277, "learning_rate": 7.224311282604628e-05, "loss": 0.0054, "step": 38560 }, { "grad_norm": 0.11535665392875671, "learning_rate": 7.222830315089085e-05, "loss": 0.0059, "step": 38570 }, { "grad_norm": 0.14369796216487885, "learning_rate": 7.22134910448816e-05, "loss": 0.0052, "step": 38580 }, { "grad_norm": 0.0925617441534996, "learning_rate": 7.219867650963839e-05, "loss": 0.005, "step": 38590 }, { "grad_norm": 0.11994349211454391, "learning_rate": 7.21838595467813e-05, "loss": 0.0057, "step": 38600 }, { "grad_norm": 0.10269375145435333, "learning_rate": 7.216904015793067e-05, "loss": 0.0061, "step": 38610 }, { "grad_norm": 0.11202152073383331, "learning_rate": 7.215421834470713e-05, "loss": 0.0065, "step": 38620 }, { "grad_norm": 0.10855945944786072, "learning_rate": 7.213939410873158e-05, "loss": 0.0046, "step": 38630 }, { "grad_norm": 0.11831972748041153, "learning_rate": 7.212456745162519e-05, "loss": 0.0062, "step": 38640 }, { "grad_norm": 0.11361639946699142, "learning_rate": 7.210973837500937e-05, "loss": 0.0054, "step": 38650 }, { "grad_norm": 0.10030423849821091, "learning_rate": 7.209490688050578e-05, "loss": 0.0061, "step": 38660 }, { "grad_norm": 0.10646272450685501, "learning_rate": 7.208007296973641e-05, "loss": 0.0051, "step": 38670 }, { "grad_norm": 0.127053901553154, "learning_rate": 7.206523664432345e-05, "loss": 0.0055, "step": 38680 }, { "grad_norm": 0.10547782480716705, "learning_rate": 7.205039790588939e-05, "loss": 0.0059, "step": 38690 }, { "grad_norm": 0.13685771822929382, "learning_rate": 7.203555675605697e-05, "loss": 0.0056, "step": 38700 }, { "grad_norm": 0.0971313938498497, "learning_rate": 7.202071319644917e-05, "loss": 0.0057, "step": 38710 }, { "grad_norm": 0.10270702093839645, "learning_rate": 7.200586722868932e-05, "loss": 0.0049, "step": 38720 }, { "grad_norm": 0.10728299617767334, "learning_rate": 7.199101885440088e-05, "loss": 0.0053, "step": 38730 }, { "grad_norm": 0.11163533478975296, "learning_rate": 7.197616807520771e-05, "loss": 0.005, "step": 38740 }, { "grad_norm": 0.12102509289979935, "learning_rate": 7.196131489273381e-05, "loss": 0.0052, "step": 38750 }, { "grad_norm": 0.10761256515979767, "learning_rate": 7.194645930860355e-05, "loss": 0.0053, "step": 38760 }, { "grad_norm": 0.1545538306236267, "learning_rate": 7.19316013244415e-05, "loss": 0.0058, "step": 38770 }, { "grad_norm": 0.12975965440273285, "learning_rate": 7.191674094187248e-05, "loss": 0.0066, "step": 38780 }, { "grad_norm": 0.1255815625190735, "learning_rate": 7.190187816252165e-05, "loss": 0.0062, "step": 38790 }, { "grad_norm": 0.12213873863220215, "learning_rate": 7.188701298801435e-05, "loss": 0.0053, "step": 38800 }, { "grad_norm": 0.10509737581014633, "learning_rate": 7.18721454199762e-05, "loss": 0.0046, "step": 38810 }, { "grad_norm": 0.0805273950099945, "learning_rate": 7.185727546003312e-05, "loss": 0.0052, "step": 38820 }, { "grad_norm": 0.11718805134296417, "learning_rate": 7.184240310981126e-05, "loss": 0.0067, "step": 38830 }, { "grad_norm": 0.10003478825092316, "learning_rate": 7.182752837093704e-05, "loss": 0.0056, "step": 38840 }, { "grad_norm": 0.13156971335411072, "learning_rate": 7.181265124503711e-05, "loss": 0.006, "step": 38850 }, { "grad_norm": 0.10001566261053085, "learning_rate": 7.179777173373847e-05, "loss": 0.0056, "step": 38860 }, { "grad_norm": 0.08611496537923813, "learning_rate": 7.178288983866826e-05, "loss": 0.0049, "step": 38870 }, { "grad_norm": 0.25533735752105713, "learning_rate": 7.176800556145397e-05, "loss": 0.0059, "step": 38880 }, { "grad_norm": 0.11466415226459503, "learning_rate": 7.175311890372334e-05, "loss": 0.0063, "step": 38890 }, { "grad_norm": 0.12264101207256317, "learning_rate": 7.17382298671043e-05, "loss": 0.0074, "step": 38900 }, { "grad_norm": 0.09156889468431473, "learning_rate": 7.172333845322515e-05, "loss": 0.0065, "step": 38910 }, { "grad_norm": 0.13397395610809326, "learning_rate": 7.170844466371436e-05, "loss": 0.0059, "step": 38920 }, { "grad_norm": 0.13959771394729614, "learning_rate": 7.16935485002007e-05, "loss": 0.0055, "step": 38930 }, { "grad_norm": 0.11965499818325043, "learning_rate": 7.167864996431319e-05, "loss": 0.0053, "step": 38940 }, { "grad_norm": 0.12131497263908386, "learning_rate": 7.166374905768111e-05, "loss": 0.0052, "step": 38950 }, { "grad_norm": 0.1038970872759819, "learning_rate": 7.164884578193401e-05, "loss": 0.0057, "step": 38960 }, { "grad_norm": 0.11527242511510849, "learning_rate": 7.163394013870168e-05, "loss": 0.0066, "step": 38970 }, { "grad_norm": 0.10591328144073486, "learning_rate": 7.161903212961419e-05, "loss": 0.0061, "step": 38980 }, { "grad_norm": 0.1280706226825714, "learning_rate": 7.160412175630183e-05, "loss": 0.0063, "step": 38990 }, { "grad_norm": 0.15963640809059143, "learning_rate": 7.158920902039521e-05, "loss": 0.0057, "step": 39000 }, { "grad_norm": 0.1347988396883011, "learning_rate": 7.157429392352514e-05, "loss": 0.0078, "step": 39010 }, { "grad_norm": 0.1080172136425972, "learning_rate": 7.155937646732275e-05, "loss": 0.0054, "step": 39020 }, { "grad_norm": 0.17013685405254364, "learning_rate": 7.154445665341933e-05, "loss": 0.0061, "step": 39030 }, { "grad_norm": 0.11221487075090408, "learning_rate": 7.152953448344654e-05, "loss": 0.0057, "step": 39040 }, { "grad_norm": 0.12315509468317032, "learning_rate": 7.151460995903624e-05, "loss": 0.0046, "step": 39050 }, { "grad_norm": 0.08166947215795517, "learning_rate": 7.149968308182052e-05, "loss": 0.006, "step": 39060 }, { "grad_norm": 0.06913135200738907, "learning_rate": 7.14847538534318e-05, "loss": 0.0067, "step": 39070 }, { "grad_norm": 0.12617498636245728, "learning_rate": 7.14698222755027e-05, "loss": 0.0068, "step": 39080 }, { "grad_norm": 0.17445158958435059, "learning_rate": 7.14548883496661e-05, "loss": 0.0058, "step": 39090 }, { "grad_norm": 0.10837750881910324, "learning_rate": 7.143995207755517e-05, "loss": 0.0059, "step": 39100 }, { "grad_norm": 0.1031530499458313, "learning_rate": 7.142501346080333e-05, "loss": 0.006, "step": 39110 }, { "grad_norm": 0.08876395970582962, "learning_rate": 7.141007250104421e-05, "loss": 0.0071, "step": 39120 }, { "grad_norm": 0.11121109873056412, "learning_rate": 7.139512919991176e-05, "loss": 0.0058, "step": 39130 }, { "grad_norm": 0.06889376789331436, "learning_rate": 7.138018355904014e-05, "loss": 0.0054, "step": 39140 }, { "grad_norm": 0.10334022343158722, "learning_rate": 7.13652355800638e-05, "loss": 0.0052, "step": 39150 }, { "grad_norm": 0.08230440318584442, "learning_rate": 7.135028526461743e-05, "loss": 0.0062, "step": 39160 }, { "grad_norm": 0.10027958452701569, "learning_rate": 7.133533261433594e-05, "loss": 0.005, "step": 39170 }, { "grad_norm": 0.15320193767547607, "learning_rate": 7.132037763085457e-05, "loss": 0.0056, "step": 39180 }, { "grad_norm": 0.09017224609851837, "learning_rate": 7.130542031580875e-05, "loss": 0.005, "step": 39190 }, { "grad_norm": 0.12023971974849701, "learning_rate": 7.12904606708342e-05, "loss": 0.006, "step": 39200 }, { "grad_norm": 0.08444627374410629, "learning_rate": 7.127549869756687e-05, "loss": 0.0051, "step": 39210 }, { "grad_norm": 0.09734588861465454, "learning_rate": 7.126053439764299e-05, "loss": 0.005, "step": 39220 }, { "grad_norm": 0.07744438201189041, "learning_rate": 7.124556777269904e-05, "loss": 0.0056, "step": 39230 }, { "grad_norm": 0.1016877144575119, "learning_rate": 7.123059882437174e-05, "loss": 0.0051, "step": 39240 }, { "grad_norm": 0.10585721582174301, "learning_rate": 7.121562755429807e-05, "loss": 0.0056, "step": 39250 }, { "grad_norm": 0.11876864731311798, "learning_rate": 7.120065396411528e-05, "loss": 0.0054, "step": 39260 }, { "grad_norm": 0.08898599445819855, "learning_rate": 7.118567805546084e-05, "loss": 0.0063, "step": 39270 }, { "grad_norm": 0.1225041076540947, "learning_rate": 7.117069982997248e-05, "loss": 0.0055, "step": 39280 }, { "grad_norm": 0.16073578596115112, "learning_rate": 7.115571928928825e-05, "loss": 0.006, "step": 39290 }, { "grad_norm": 0.13231855630874634, "learning_rate": 7.114073643504635e-05, "loss": 0.0076, "step": 39300 }, { "grad_norm": 0.08586809784173965, "learning_rate": 7.11257512688853e-05, "loss": 0.0063, "step": 39310 }, { "grad_norm": 0.0748850554227829, "learning_rate": 7.111076379244384e-05, "loss": 0.005, "step": 39320 }, { "grad_norm": 0.10492711514234543, "learning_rate": 7.109577400736101e-05, "loss": 0.0055, "step": 39330 }, { "grad_norm": 0.08426028490066528, "learning_rate": 7.108078191527605e-05, "loss": 0.0056, "step": 39340 }, { "grad_norm": 0.1126338467001915, "learning_rate": 7.106578751782847e-05, "loss": 0.0061, "step": 39350 }, { "grad_norm": 0.10079307109117508, "learning_rate": 7.105079081665803e-05, "loss": 0.0056, "step": 39360 }, { "grad_norm": 0.11480690538883209, "learning_rate": 7.103579181340476e-05, "loss": 0.0057, "step": 39370 }, { "grad_norm": 0.11685722321271896, "learning_rate": 7.102079050970893e-05, "loss": 0.0062, "step": 39380 }, { "grad_norm": 0.08992310613393784, "learning_rate": 7.100578690721105e-05, "loss": 0.005, "step": 39390 }, { "grad_norm": 0.10650692135095596, "learning_rate": 7.09907810075519e-05, "loss": 0.0051, "step": 39400 }, { "grad_norm": 0.1272568553686142, "learning_rate": 7.097577281237249e-05, "loss": 0.0059, "step": 39410 }, { "grad_norm": 0.08602641522884369, "learning_rate": 7.09607623233141e-05, "loss": 0.0058, "step": 39420 }, { "grad_norm": 0.1014733836054802, "learning_rate": 7.094574954201827e-05, "loss": 0.0053, "step": 39430 }, { "grad_norm": 0.10779013484716415, "learning_rate": 7.093073447012675e-05, "loss": 0.0063, "step": 39440 }, { "grad_norm": 0.11527079343795776, "learning_rate": 7.09157171092816e-05, "loss": 0.0057, "step": 39450 }, { "grad_norm": 0.09465707838535309, "learning_rate": 7.090069746112504e-05, "loss": 0.0052, "step": 39460 }, { "grad_norm": 0.10866574198007584, "learning_rate": 7.088567552729965e-05, "loss": 0.0053, "step": 39470 }, { "grad_norm": 0.14674772322177887, "learning_rate": 7.087065130944818e-05, "loss": 0.0057, "step": 39480 }, { "grad_norm": 0.08613763004541397, "learning_rate": 7.085562480921366e-05, "loss": 0.0056, "step": 39490 }, { "grad_norm": 0.09855638444423676, "learning_rate": 7.084059602823937e-05, "loss": 0.0076, "step": 39500 }, { "grad_norm": 0.10805720835924149, "learning_rate": 7.082556496816882e-05, "loss": 0.0054, "step": 39510 }, { "grad_norm": 0.11214123666286469, "learning_rate": 7.081053163064582e-05, "loss": 0.0046, "step": 39520 }, { "grad_norm": 0.10742761194705963, "learning_rate": 7.079549601731434e-05, "loss": 0.006, "step": 39530 }, { "grad_norm": 0.0978492796421051, "learning_rate": 7.07804581298187e-05, "loss": 0.005, "step": 39540 }, { "grad_norm": 0.14970046281814575, "learning_rate": 7.07654179698034e-05, "loss": 0.0055, "step": 39550 }, { "grad_norm": 0.11560052633285522, "learning_rate": 7.075037553891321e-05, "loss": 0.0052, "step": 39560 }, { "grad_norm": 0.10143466293811798, "learning_rate": 7.073533083879315e-05, "loss": 0.0049, "step": 39570 }, { "grad_norm": 0.10514481365680695, "learning_rate": 7.072028387108849e-05, "loss": 0.0061, "step": 39580 }, { "grad_norm": 0.10524038225412369, "learning_rate": 7.070523463744472e-05, "loss": 0.0054, "step": 39590 }, { "grad_norm": 0.10328555852174759, "learning_rate": 7.069018313950763e-05, "loss": 0.0061, "step": 39600 }, { "grad_norm": 0.11263197660446167, "learning_rate": 7.067512937892322e-05, "loss": 0.0049, "step": 39610 }, { "grad_norm": 0.17636577785015106, "learning_rate": 7.066007335733775e-05, "loss": 0.0051, "step": 39620 }, { "grad_norm": 0.1399187296628952, "learning_rate": 7.064501507639772e-05, "loss": 0.006, "step": 39630 }, { "grad_norm": 0.10645653307437897, "learning_rate": 7.062995453774987e-05, "loss": 0.0057, "step": 39640 }, { "grad_norm": 0.13833992183208466, "learning_rate": 7.061489174304121e-05, "loss": 0.0059, "step": 39650 }, { "grad_norm": 0.13510362803936005, "learning_rate": 7.0599826693919e-05, "loss": 0.0056, "step": 39660 }, { "grad_norm": 0.1177968680858612, "learning_rate": 7.05847593920307e-05, "loss": 0.0059, "step": 39670 }, { "grad_norm": 0.11637958884239197, "learning_rate": 7.056968983902406e-05, "loss": 0.0062, "step": 39680 }, { "grad_norm": 0.08907154202461243, "learning_rate": 7.055461803654706e-05, "loss": 0.0073, "step": 39690 }, { "grad_norm": 0.1430126279592514, "learning_rate": 7.053954398624794e-05, "loss": 0.0056, "step": 39700 }, { "grad_norm": 0.1098453477025032, "learning_rate": 7.052446768977518e-05, "loss": 0.0051, "step": 39710 }, { "grad_norm": 0.09227579832077026, "learning_rate": 7.050938914877748e-05, "loss": 0.0058, "step": 39720 }, { "grad_norm": 0.11459232866764069, "learning_rate": 7.049430836490382e-05, "loss": 0.0057, "step": 39730 }, { "grad_norm": 0.10923866182565689, "learning_rate": 7.047922533980342e-05, "loss": 0.0054, "step": 39740 }, { "grad_norm": 0.10661699622869492, "learning_rate": 7.046414007512571e-05, "loss": 0.0062, "step": 39750 }, { "grad_norm": 0.10760572552680969, "learning_rate": 7.044905257252042e-05, "loss": 0.006, "step": 39760 }, { "grad_norm": 0.12085415422916412, "learning_rate": 7.04339628336375e-05, "loss": 0.0062, "step": 39770 }, { "grad_norm": 0.1062241941690445, "learning_rate": 7.041887086012711e-05, "loss": 0.0062, "step": 39780 }, { "grad_norm": 0.1121019721031189, "learning_rate": 7.040377665363969e-05, "loss": 0.0061, "step": 39790 }, { "grad_norm": 0.11565796285867691, "learning_rate": 7.038868021582594e-05, "loss": 0.0065, "step": 39800 }, { "grad_norm": 0.11009250581264496, "learning_rate": 7.037358154833679e-05, "loss": 0.0057, "step": 39810 }, { "grad_norm": 0.10160514712333679, "learning_rate": 7.035848065282339e-05, "loss": 0.0058, "step": 39820 }, { "grad_norm": 0.12079977989196777, "learning_rate": 7.034337753093714e-05, "loss": 0.0057, "step": 39830 }, { "grad_norm": 0.11943700909614563, "learning_rate": 7.032827218432972e-05, "loss": 0.0058, "step": 39840 }, { "grad_norm": 0.11762431263923645, "learning_rate": 7.031316461465302e-05, "loss": 0.0056, "step": 39850 }, { "grad_norm": 0.11618249863386154, "learning_rate": 7.029805482355915e-05, "loss": 0.0054, "step": 39860 }, { "grad_norm": 0.10056586563587189, "learning_rate": 7.028294281270055e-05, "loss": 0.0065, "step": 39870 }, { "grad_norm": 0.09100984781980515, "learning_rate": 7.026782858372982e-05, "loss": 0.0056, "step": 39880 }, { "grad_norm": 0.1277647763490677, "learning_rate": 7.025271213829982e-05, "loss": 0.0053, "step": 39890 }, { "grad_norm": 0.10772635042667389, "learning_rate": 7.023759347806366e-05, "loss": 0.005, "step": 39900 }, { "grad_norm": 0.10800643265247345, "learning_rate": 7.022247260467472e-05, "loss": 0.0058, "step": 39910 }, { "grad_norm": 0.10702487826347351, "learning_rate": 7.020734951978658e-05, "loss": 0.0046, "step": 39920 }, { "grad_norm": 0.12860959768295288, "learning_rate": 7.019222422505307e-05, "loss": 0.0077, "step": 39930 }, { "grad_norm": 0.10868074744939804, "learning_rate": 7.017709672212827e-05, "loss": 0.0059, "step": 39940 }, { "grad_norm": 0.10483425855636597, "learning_rate": 7.016196701266652e-05, "loss": 0.0066, "step": 39950 }, { "grad_norm": 0.12176865339279175, "learning_rate": 7.014683509832238e-05, "loss": 0.0055, "step": 39960 }, { "grad_norm": 0.10662732273340225, "learning_rate": 7.013170098075063e-05, "loss": 0.0065, "step": 39970 }, { "grad_norm": 0.10351800173521042, "learning_rate": 7.011656466160632e-05, "loss": 0.0048, "step": 39980 }, { "grad_norm": 0.10871163010597229, "learning_rate": 7.010142614254475e-05, "loss": 0.0058, "step": 39990 }, { "grad_norm": 0.0867559015750885, "learning_rate": 7.008628542522147e-05, "loss": 0.0054, "step": 40000 }, { "grad_norm": 0.09196704626083374, "learning_rate": 7.007114251129218e-05, "loss": 0.0059, "step": 40010 }, { "grad_norm": 0.1101231649518013, "learning_rate": 7.005599740241293e-05, "loss": 0.0064, "step": 40020 }, { "grad_norm": 0.13772179186344147, "learning_rate": 7.004085010023996e-05, "loss": 0.0056, "step": 40030 }, { "grad_norm": 0.07705149054527283, "learning_rate": 7.002570060642976e-05, "loss": 0.0049, "step": 40040 }, { "grad_norm": 0.11573393642902374, "learning_rate": 7.001054892263903e-05, "loss": 0.0054, "step": 40050 }, { "grad_norm": 0.09649534523487091, "learning_rate": 6.999539505052477e-05, "loss": 0.0051, "step": 40060 }, { "grad_norm": 0.12977083027362823, "learning_rate": 6.998023899174418e-05, "loss": 0.0049, "step": 40070 }, { "grad_norm": 0.1129840835928917, "learning_rate": 6.996508074795467e-05, "loss": 0.0058, "step": 40080 }, { "grad_norm": 0.10716467350721359, "learning_rate": 6.994992032081396e-05, "loss": 0.005, "step": 40090 }, { "grad_norm": 0.09632185101509094, "learning_rate": 6.993475771197995e-05, "loss": 0.0062, "step": 40100 }, { "grad_norm": 0.11804657429456711, "learning_rate": 6.991959292311082e-05, "loss": 0.0057, "step": 40110 }, { "grad_norm": 0.09750214964151382, "learning_rate": 6.990442595586495e-05, "loss": 0.0047, "step": 40120 }, { "grad_norm": 0.08631893992424011, "learning_rate": 6.988925681190098e-05, "loss": 0.0054, "step": 40130 }, { "grad_norm": 0.07344982028007507, "learning_rate": 6.987408549287778e-05, "loss": 0.0044, "step": 40140 }, { "grad_norm": 0.10510838031768799, "learning_rate": 6.985891200045449e-05, "loss": 0.0052, "step": 40150 }, { "grad_norm": 0.12809526920318604, "learning_rate": 6.984373633629045e-05, "loss": 0.0045, "step": 40160 }, { "grad_norm": 0.09600845724344254, "learning_rate": 6.98285585020452e-05, "loss": 0.0056, "step": 40170 }, { "grad_norm": 0.12113814055919647, "learning_rate": 6.981337849937864e-05, "loss": 0.0048, "step": 40180 }, { "grad_norm": 0.09097032994031906, "learning_rate": 6.979819632995078e-05, "loss": 0.006, "step": 40190 }, { "grad_norm": 0.1294672042131424, "learning_rate": 6.978301199542193e-05, "loss": 0.0048, "step": 40200 }, { "grad_norm": 0.13128550350666046, "learning_rate": 6.976782549745263e-05, "loss": 0.0056, "step": 40210 }, { "grad_norm": 0.16153103113174438, "learning_rate": 6.975263683770367e-05, "loss": 0.0066, "step": 40220 }, { "grad_norm": 0.11484497040510178, "learning_rate": 6.973744601783604e-05, "loss": 0.0059, "step": 40230 }, { "grad_norm": 0.11539538204669952, "learning_rate": 6.972225303951097e-05, "loss": 0.0059, "step": 40240 }, { "grad_norm": 0.11627104133367538, "learning_rate": 6.970705790438998e-05, "loss": 0.0052, "step": 40250 }, { "grad_norm": 0.1013978123664856, "learning_rate": 6.969186061413477e-05, "loss": 0.0049, "step": 40260 }, { "grad_norm": 0.11454508453607559, "learning_rate": 6.967666117040727e-05, "loss": 0.0046, "step": 40270 }, { "grad_norm": 0.11754677444696426, "learning_rate": 6.966145957486972e-05, "loss": 0.005, "step": 40280 }, { "grad_norm": 0.10173064470291138, "learning_rate": 6.964625582918449e-05, "loss": 0.0053, "step": 40290 }, { "grad_norm": 0.1387171745300293, "learning_rate": 6.963104993501425e-05, "loss": 0.006, "step": 40300 }, { "grad_norm": 0.10206766426563263, "learning_rate": 6.961584189402192e-05, "loss": 0.0054, "step": 40310 }, { "grad_norm": 0.10483420640230179, "learning_rate": 6.96006317078706e-05, "loss": 0.0059, "step": 40320 }, { "grad_norm": 0.10082763433456421, "learning_rate": 6.958541937822367e-05, "loss": 0.0059, "step": 40330 }, { "grad_norm": 0.15334056317806244, "learning_rate": 6.957020490674473e-05, "loss": 0.0051, "step": 40340 }, { "grad_norm": 0.11722998321056366, "learning_rate": 6.95549882950976e-05, "loss": 0.0062, "step": 40350 }, { "grad_norm": 0.12077126652002335, "learning_rate": 6.953976954494635e-05, "loss": 0.0056, "step": 40360 }, { "grad_norm": 0.11687763035297394, "learning_rate": 6.952454865795528e-05, "loss": 0.0059, "step": 40370 }, { "grad_norm": 0.1103697344660759, "learning_rate": 6.950932563578892e-05, "loss": 0.0047, "step": 40380 }, { "grad_norm": 0.09938892722129822, "learning_rate": 6.949410048011206e-05, "loss": 0.0049, "step": 40390 }, { "grad_norm": 0.11649433523416519, "learning_rate": 6.947887319258966e-05, "loss": 0.0049, "step": 40400 }, { "grad_norm": 0.10782214254140854, "learning_rate": 6.946364377488696e-05, "loss": 0.006, "step": 40410 }, { "grad_norm": 0.14795908331871033, "learning_rate": 6.944841222866947e-05, "loss": 0.0058, "step": 40420 }, { "grad_norm": 0.1087593212723732, "learning_rate": 6.943317855560284e-05, "loss": 0.0062, "step": 40430 }, { "grad_norm": 0.11592411994934082, "learning_rate": 6.941794275735302e-05, "loss": 0.0055, "step": 40440 }, { "grad_norm": 0.13258948922157288, "learning_rate": 6.94027048355862e-05, "loss": 0.0056, "step": 40450 }, { "grad_norm": 0.09192229062318802, "learning_rate": 6.938746479196871e-05, "loss": 0.0063, "step": 40460 }, { "grad_norm": 0.07689564675092697, "learning_rate": 6.937222262816724e-05, "loss": 0.0055, "step": 40470 }, { "grad_norm": 0.0740559920668602, "learning_rate": 6.935697834584865e-05, "loss": 0.0045, "step": 40480 }, { "grad_norm": 0.0813165009021759, "learning_rate": 6.934173194667998e-05, "loss": 0.0045, "step": 40490 }, { "grad_norm": 0.09503662586212158, "learning_rate": 6.93264834323286e-05, "loss": 0.0049, "step": 40500 }, { "grad_norm": 0.09809502959251404, "learning_rate": 6.931123280446206e-05, "loss": 0.0049, "step": 40510 }, { "grad_norm": 0.09500142186880112, "learning_rate": 6.929598006474811e-05, "loss": 0.0045, "step": 40520 }, { "grad_norm": 0.09986644983291626, "learning_rate": 6.92807252148548e-05, "loss": 0.0061, "step": 40530 }, { "grad_norm": 0.11296145617961884, "learning_rate": 6.926546825645037e-05, "loss": 0.0054, "step": 40540 }, { "grad_norm": 0.12996657192707062, "learning_rate": 6.92502091912033e-05, "loss": 0.0051, "step": 40550 }, { "grad_norm": 0.10815262794494629, "learning_rate": 6.923494802078229e-05, "loss": 0.0049, "step": 40560 }, { "grad_norm": 0.11138399690389633, "learning_rate": 6.921968474685628e-05, "loss": 0.0053, "step": 40570 }, { "grad_norm": 0.09028409421443939, "learning_rate": 6.920441937109445e-05, "loss": 0.0056, "step": 40580 }, { "grad_norm": 0.1289171278476715, "learning_rate": 6.91891518951662e-05, "loss": 0.0051, "step": 40590 }, { "grad_norm": 0.12233063578605652, "learning_rate": 6.917388232074114e-05, "loss": 0.0052, "step": 40600 }, { "grad_norm": 0.08974894881248474, "learning_rate": 6.915861064948914e-05, "loss": 0.0048, "step": 40610 }, { "grad_norm": 0.11511185765266418, "learning_rate": 6.914333688308025e-05, "loss": 0.0062, "step": 40620 }, { "grad_norm": 0.09918301552534103, "learning_rate": 6.912806102318487e-05, "loss": 0.0058, "step": 40630 }, { "grad_norm": 0.12303448468446732, "learning_rate": 6.911278307147347e-05, "loss": 0.0046, "step": 40640 }, { "grad_norm": 0.10957936942577362, "learning_rate": 6.909750302961684e-05, "loss": 0.0056, "step": 40650 }, { "grad_norm": 0.13204191625118256, "learning_rate": 6.9082220899286e-05, "loss": 0.005, "step": 40660 }, { "grad_norm": 0.10995671153068542, "learning_rate": 6.906693668215219e-05, "loss": 0.0057, "step": 40670 }, { "grad_norm": 0.07995610684156418, "learning_rate": 6.905165037988683e-05, "loss": 0.0045, "step": 40680 }, { "grad_norm": 0.11147044599056244, "learning_rate": 6.903636199416164e-05, "loss": 0.0051, "step": 40690 }, { "grad_norm": 0.11410573124885559, "learning_rate": 6.902107152664851e-05, "loss": 0.0059, "step": 40700 }, { "grad_norm": 0.0992329940199852, "learning_rate": 6.90057789790196e-05, "loss": 0.0053, "step": 40710 }, { "grad_norm": 0.10844872891902924, "learning_rate": 6.899048435294728e-05, "loss": 0.0053, "step": 40720 }, { "grad_norm": 0.09480102360248566, "learning_rate": 6.897518765010415e-05, "loss": 0.005, "step": 40730 }, { "grad_norm": 0.06841043382883072, "learning_rate": 6.895988887216302e-05, "loss": 0.0054, "step": 40740 }, { "grad_norm": 0.10505059361457825, "learning_rate": 6.894458802079694e-05, "loss": 0.0057, "step": 40750 }, { "grad_norm": 0.07535727322101593, "learning_rate": 6.892928509767922e-05, "loss": 0.0047, "step": 40760 }, { "grad_norm": 0.10870581120252609, "learning_rate": 6.891398010448333e-05, "loss": 0.005, "step": 40770 }, { "grad_norm": 0.12603630125522614, "learning_rate": 6.8898673042883e-05, "loss": 0.0047, "step": 40780 }, { "grad_norm": 0.11573583632707596, "learning_rate": 6.888336391455222e-05, "loss": 0.0053, "step": 40790 }, { "grad_norm": 0.10310239344835281, "learning_rate": 6.886805272116513e-05, "loss": 0.0055, "step": 40800 }, { "grad_norm": 0.0985519215464592, "learning_rate": 6.885273946439617e-05, "loss": 0.0048, "step": 40810 }, { "grad_norm": 0.10068129003047943, "learning_rate": 6.883742414591998e-05, "loss": 0.0046, "step": 40820 }, { "grad_norm": 0.08023469150066376, "learning_rate": 6.88221067674114e-05, "loss": 0.0049, "step": 40830 }, { "grad_norm": 0.10142893344163895, "learning_rate": 6.88067873305455e-05, "loss": 0.0047, "step": 40840 }, { "grad_norm": 0.10130134969949722, "learning_rate": 6.879146583699765e-05, "loss": 0.0055, "step": 40850 }, { "grad_norm": 0.1039264127612114, "learning_rate": 6.877614228844334e-05, "loss": 0.0065, "step": 40860 }, { "grad_norm": 0.11989353597164154, "learning_rate": 6.876081668655832e-05, "loss": 0.0052, "step": 40870 }, { "grad_norm": 0.12799063324928284, "learning_rate": 6.874548903301861e-05, "loss": 0.0058, "step": 40880 }, { "grad_norm": 0.11306850612163544, "learning_rate": 6.873015932950039e-05, "loss": 0.0063, "step": 40890 }, { "grad_norm": 0.1124669760465622, "learning_rate": 6.871482757768012e-05, "loss": 0.0061, "step": 40900 }, { "grad_norm": 0.09178759902715683, "learning_rate": 6.869949377923442e-05, "loss": 0.0053, "step": 40910 }, { "grad_norm": 0.09334874898195267, "learning_rate": 6.868415793584022e-05, "loss": 0.0053, "step": 40920 }, { "grad_norm": 0.12158958613872528, "learning_rate": 6.86688200491746e-05, "loss": 0.0059, "step": 40930 }, { "grad_norm": 0.10077165812253952, "learning_rate": 6.865348012091488e-05, "loss": 0.0055, "step": 40940 }, { "grad_norm": 0.0939776673913002, "learning_rate": 6.86381381527386e-05, "loss": 0.0055, "step": 40950 }, { "grad_norm": 0.11776447296142578, "learning_rate": 6.862279414632359e-05, "loss": 0.0049, "step": 40960 }, { "grad_norm": 0.1253868192434311, "learning_rate": 6.860744810334778e-05, "loss": 0.0058, "step": 40970 }, { "grad_norm": 0.12610246241092682, "learning_rate": 6.859210002548943e-05, "loss": 0.0051, "step": 40980 }, { "grad_norm": 0.11296239495277405, "learning_rate": 6.8576749914427e-05, "loss": 0.0059, "step": 40990 }, { "grad_norm": 0.1192389726638794, "learning_rate": 6.85613977718391e-05, "loss": 0.0073, "step": 41000 }, { "grad_norm": 0.09726574271917343, "learning_rate": 6.854604359940464e-05, "loss": 0.006, "step": 41010 }, { "grad_norm": 0.099971242249012, "learning_rate": 6.853068739880276e-05, "loss": 0.0055, "step": 41020 }, { "grad_norm": 0.08640437573194504, "learning_rate": 6.851532917171276e-05, "loss": 0.0069, "step": 41030 }, { "grad_norm": 0.11688221991062164, "learning_rate": 6.849996891981418e-05, "loss": 0.0049, "step": 41040 }, { "grad_norm": 0.13613061606884003, "learning_rate": 6.848460664478684e-05, "loss": 0.0056, "step": 41050 }, { "grad_norm": 0.1329372227191925, "learning_rate": 6.846924234831069e-05, "loss": 0.0068, "step": 41060 }, { "grad_norm": 0.13099923729896545, "learning_rate": 6.845387603206597e-05, "loss": 0.0058, "step": 41070 }, { "grad_norm": 0.09969031810760498, "learning_rate": 6.843850769773312e-05, "loss": 0.0049, "step": 41080 }, { "grad_norm": 0.06747399270534515, "learning_rate": 6.842313734699278e-05, "loss": 0.0055, "step": 41090 }, { "grad_norm": 0.10454236716032028, "learning_rate": 6.840776498152584e-05, "loss": 0.0056, "step": 41100 }, { "grad_norm": 0.08075451105833054, "learning_rate": 6.83923906030134e-05, "loss": 0.0049, "step": 41110 }, { "grad_norm": 0.09543301165103912, "learning_rate": 6.837701421313677e-05, "loss": 0.0057, "step": 41120 }, { "grad_norm": 0.105460524559021, "learning_rate": 6.83616358135775e-05, "loss": 0.0064, "step": 41130 }, { "grad_norm": 0.12198846787214279, "learning_rate": 6.834625540601734e-05, "loss": 0.0053, "step": 41140 }, { "grad_norm": 0.11922716349363327, "learning_rate": 6.833087299213829e-05, "loss": 0.0066, "step": 41150 }, { "grad_norm": 0.08811408281326294, "learning_rate": 6.83154885736225e-05, "loss": 0.006, "step": 41160 }, { "grad_norm": 0.10445096343755722, "learning_rate": 6.830010215215244e-05, "loss": 0.0066, "step": 41170 }, { "grad_norm": 0.10903418064117432, "learning_rate": 6.828471372941072e-05, "loss": 0.0054, "step": 41180 }, { "grad_norm": 0.09388116002082825, "learning_rate": 6.826932330708018e-05, "loss": 0.006, "step": 41190 }, { "grad_norm": 0.08917802572250366, "learning_rate": 6.825393088684393e-05, "loss": 0.005, "step": 41200 }, { "grad_norm": 0.08848781138658524, "learning_rate": 6.823853647038524e-05, "loss": 0.0066, "step": 41210 }, { "grad_norm": 0.09580101817846298, "learning_rate": 6.822314005938762e-05, "loss": 0.0052, "step": 41220 }, { "grad_norm": 0.09197316318750381, "learning_rate": 6.820774165553479e-05, "loss": 0.0051, "step": 41230 }, { "grad_norm": 0.0672156810760498, "learning_rate": 6.819234126051072e-05, "loss": 0.0048, "step": 41240 }, { "grad_norm": 0.0842500627040863, "learning_rate": 6.817693887599956e-05, "loss": 0.0053, "step": 41250 }, { "grad_norm": 0.08965382725000381, "learning_rate": 6.816153450368568e-05, "loss": 0.0049, "step": 41260 }, { "grad_norm": 0.09946135431528091, "learning_rate": 6.814612814525372e-05, "loss": 0.0054, "step": 41270 }, { "grad_norm": 0.09024674445390701, "learning_rate": 6.813071980238846e-05, "loss": 0.0056, "step": 41280 }, { "grad_norm": 0.08593938499689102, "learning_rate": 6.811530947677492e-05, "loss": 0.005, "step": 41290 }, { "grad_norm": 0.10254093259572983, "learning_rate": 6.809989717009839e-05, "loss": 0.0058, "step": 41300 }, { "grad_norm": 0.08345038443803787, "learning_rate": 6.808448288404431e-05, "loss": 0.0065, "step": 41310 }, { "grad_norm": 0.11643955856561661, "learning_rate": 6.806906662029838e-05, "loss": 0.0064, "step": 41320 }, { "grad_norm": 0.11025074124336243, "learning_rate": 6.805364838054647e-05, "loss": 0.0056, "step": 41330 }, { "grad_norm": 0.10317238420248032, "learning_rate": 6.803822816647474e-05, "loss": 0.005, "step": 41340 }, { "grad_norm": 0.1036328449845314, "learning_rate": 6.802280597976949e-05, "loss": 0.0077, "step": 41350 }, { "grad_norm": 0.0712483823299408, "learning_rate": 6.800738182211727e-05, "loss": 0.0047, "step": 41360 }, { "grad_norm": 0.0960918441414833, "learning_rate": 6.799195569520485e-05, "loss": 0.0051, "step": 41370 }, { "grad_norm": 0.13079319894313812, "learning_rate": 6.79765276007192e-05, "loss": 0.0061, "step": 41380 }, { "grad_norm": 0.08273229748010635, "learning_rate": 6.796109754034753e-05, "loss": 0.0055, "step": 41390 }, { "grad_norm": 0.12626299262046814, "learning_rate": 6.794566551577724e-05, "loss": 0.0059, "step": 41400 }, { "grad_norm": 0.11933738738298416, "learning_rate": 6.793023152869594e-05, "loss": 0.0069, "step": 41410 }, { "grad_norm": 0.10064753144979477, "learning_rate": 6.791479558079147e-05, "loss": 0.0064, "step": 41420 }, { "grad_norm": 0.11361371725797653, "learning_rate": 6.789935767375191e-05, "loss": 0.005, "step": 41430 }, { "grad_norm": 0.1073908880352974, "learning_rate": 6.78839178092655e-05, "loss": 0.0057, "step": 41440 }, { "grad_norm": 0.12360487133264542, "learning_rate": 6.786847598902072e-05, "loss": 0.0062, "step": 41450 }, { "grad_norm": 0.12279359251260757, "learning_rate": 6.785303221470627e-05, "loss": 0.0058, "step": 41460 }, { "grad_norm": 0.09303024411201477, "learning_rate": 6.783758648801108e-05, "loss": 0.0063, "step": 41470 }, { "grad_norm": 0.10622885078191757, "learning_rate": 6.782213881062422e-05, "loss": 0.0052, "step": 41480 }, { "grad_norm": 0.08482249081134796, "learning_rate": 6.780668918423508e-05, "loss": 0.0048, "step": 41490 }, { "grad_norm": 0.12056839466094971, "learning_rate": 6.779123761053317e-05, "loss": 0.0061, "step": 41500 }, { "grad_norm": 0.09202564507722855, "learning_rate": 6.777578409120826e-05, "loss": 0.0054, "step": 41510 }, { "grad_norm": 0.08447383344173431, "learning_rate": 6.776032862795033e-05, "loss": 0.0048, "step": 41520 }, { "grad_norm": 0.08752156049013138, "learning_rate": 6.774487122244956e-05, "loss": 0.0049, "step": 41530 }, { "grad_norm": 0.11010599136352539, "learning_rate": 6.772941187639637e-05, "loss": 0.0048, "step": 41540 }, { "grad_norm": 0.10934262722730637, "learning_rate": 6.771395059148134e-05, "loss": 0.0056, "step": 41550 }, { "grad_norm": 0.13422133028507233, "learning_rate": 6.76984873693953e-05, "loss": 0.0063, "step": 41560 }, { "grad_norm": 0.1340823471546173, "learning_rate": 6.768302221182931e-05, "loss": 0.005, "step": 41570 }, { "grad_norm": 0.08905881643295288, "learning_rate": 6.766755512047457e-05, "loss": 0.0055, "step": 41580 }, { "grad_norm": 0.13015854358673096, "learning_rate": 6.765208609702259e-05, "loss": 0.0055, "step": 41590 }, { "grad_norm": 0.11544706672430038, "learning_rate": 6.763661514316499e-05, "loss": 0.005, "step": 41600 }, { "grad_norm": 0.09588227421045303, "learning_rate": 6.76211422605937e-05, "loss": 0.0051, "step": 41610 }, { "grad_norm": 0.09847301989793777, "learning_rate": 6.760566745100076e-05, "loss": 0.0048, "step": 41620 }, { "grad_norm": 0.0955430194735527, "learning_rate": 6.75901907160785e-05, "loss": 0.0046, "step": 41630 }, { "grad_norm": 0.09602897614240646, "learning_rate": 6.757471205751943e-05, "loss": 0.0046, "step": 41640 }, { "grad_norm": 0.11535916477441788, "learning_rate": 6.75592314770163e-05, "loss": 0.0061, "step": 41650 }, { "grad_norm": 0.09368907660245895, "learning_rate": 6.7543748976262e-05, "loss": 0.0073, "step": 41660 }, { "grad_norm": 0.13715550303459167, "learning_rate": 6.752826455694968e-05, "loss": 0.0052, "step": 41670 }, { "grad_norm": 0.13368305563926697, "learning_rate": 6.751277822077271e-05, "loss": 0.0052, "step": 41680 }, { "grad_norm": 0.13797102868556976, "learning_rate": 6.749728996942465e-05, "loss": 0.0056, "step": 41690 }, { "grad_norm": 0.14689983427524567, "learning_rate": 6.748179980459924e-05, "loss": 0.0056, "step": 41700 }, { "grad_norm": 0.10567792505025864, "learning_rate": 6.746630772799052e-05, "loss": 0.0049, "step": 41710 }, { "grad_norm": 0.09423746913671494, "learning_rate": 6.745081374129262e-05, "loss": 0.0046, "step": 41720 }, { "grad_norm": 0.08479741215705872, "learning_rate": 6.743531784619998e-05, "loss": 0.0047, "step": 41730 }, { "grad_norm": 0.1058184951543808, "learning_rate": 6.741982004440719e-05, "loss": 0.0046, "step": 41740 }, { "grad_norm": 0.09507264941930771, "learning_rate": 6.740432033760907e-05, "loss": 0.0052, "step": 41750 }, { "grad_norm": 0.0809001699090004, "learning_rate": 6.738881872750066e-05, "loss": 0.0056, "step": 41760 }, { "grad_norm": 0.12253642827272415, "learning_rate": 6.737331521577715e-05, "loss": 0.0053, "step": 41770 }, { "grad_norm": 0.11053606122732162, "learning_rate": 6.735780980413403e-05, "loss": 0.0054, "step": 41780 }, { "grad_norm": 0.10880466550588608, "learning_rate": 6.734230249426692e-05, "loss": 0.0046, "step": 41790 }, { "grad_norm": 0.08600013703107834, "learning_rate": 6.732679328787168e-05, "loss": 0.0055, "step": 41800 }, { "grad_norm": 0.088872991502285, "learning_rate": 6.731128218664438e-05, "loss": 0.0045, "step": 41810 }, { "grad_norm": 0.11176474392414093, "learning_rate": 6.729576919228129e-05, "loss": 0.0048, "step": 41820 }, { "grad_norm": 0.10741955786943436, "learning_rate": 6.728025430647888e-05, "loss": 0.0048, "step": 41830 }, { "grad_norm": 0.12355684489011765, "learning_rate": 6.726473753093383e-05, "loss": 0.0054, "step": 41840 }, { "grad_norm": 0.11473313719034195, "learning_rate": 6.724921886734305e-05, "loss": 0.0052, "step": 41850 }, { "grad_norm": 0.13875678181648254, "learning_rate": 6.723369831740363e-05, "loss": 0.0058, "step": 41860 }, { "grad_norm": 0.10658717900514603, "learning_rate": 6.721817588281288e-05, "loss": 0.0047, "step": 41870 }, { "grad_norm": 0.10040878504514694, "learning_rate": 6.720265156526828e-05, "loss": 0.0049, "step": 41880 }, { "grad_norm": 0.1108081191778183, "learning_rate": 6.718712536646758e-05, "loss": 0.0054, "step": 41890 }, { "grad_norm": 0.10339483618736267, "learning_rate": 6.71715972881087e-05, "loss": 0.0054, "step": 41900 }, { "grad_norm": 0.12259113788604736, "learning_rate": 6.715606733188974e-05, "loss": 0.0049, "step": 41910 }, { "grad_norm": 0.11494473367929459, "learning_rate": 6.714053549950909e-05, "loss": 0.0047, "step": 41920 }, { "grad_norm": 0.08585824817419052, "learning_rate": 6.71250017926652e-05, "loss": 0.0053, "step": 41930 }, { "grad_norm": 0.09604554623365402, "learning_rate": 6.71094662130569e-05, "loss": 0.0057, "step": 41940 }, { "grad_norm": 0.10514195263385773, "learning_rate": 6.709392876238307e-05, "loss": 0.0056, "step": 41950 }, { "grad_norm": 0.07597783207893372, "learning_rate": 6.707838944234289e-05, "loss": 0.0046, "step": 41960 }, { "grad_norm": 0.07471833378076553, "learning_rate": 6.706284825463574e-05, "loss": 0.0049, "step": 41970 }, { "grad_norm": 0.10352704674005508, "learning_rate": 6.704730520096114e-05, "loss": 0.0053, "step": 41980 }, { "grad_norm": 0.09880291670560837, "learning_rate": 6.703176028301888e-05, "loss": 0.0053, "step": 41990 }, { "grad_norm": 0.10625690966844559, "learning_rate": 6.701621350250892e-05, "loss": 0.0053, "step": 42000 }, { "grad_norm": 0.1090213730931282, "learning_rate": 6.700066486113144e-05, "loss": 0.0051, "step": 42010 }, { "grad_norm": 0.11885274946689606, "learning_rate": 6.69851143605868e-05, "loss": 0.0056, "step": 42020 }, { "grad_norm": 0.10735109448432922, "learning_rate": 6.696956200257561e-05, "loss": 0.0065, "step": 42030 }, { "grad_norm": 0.11377254128456116, "learning_rate": 6.695400778879863e-05, "loss": 0.004, "step": 42040 }, { "grad_norm": 0.09947672486305237, "learning_rate": 6.693845172095683e-05, "loss": 0.0046, "step": 42050 }, { "grad_norm": 0.09550686925649643, "learning_rate": 6.692289380075142e-05, "loss": 0.0045, "step": 42060 }, { "grad_norm": 0.13052166998386383, "learning_rate": 6.690733402988379e-05, "loss": 0.0051, "step": 42070 }, { "grad_norm": 0.11403823643922806, "learning_rate": 6.689177241005553e-05, "loss": 0.0056, "step": 42080 }, { "grad_norm": 0.10541490465402603, "learning_rate": 6.687620894296844e-05, "loss": 0.0056, "step": 42090 }, { "grad_norm": 0.11482509225606918, "learning_rate": 6.686064363032451e-05, "loss": 0.0056, "step": 42100 }, { "grad_norm": 0.10803543031215668, "learning_rate": 6.684507647382596e-05, "loss": 0.0049, "step": 42110 }, { "grad_norm": 0.09364014118909836, "learning_rate": 6.682950747517513e-05, "loss": 0.0055, "step": 42120 }, { "grad_norm": 0.1154339537024498, "learning_rate": 6.68139366360747e-05, "loss": 0.0057, "step": 42130 }, { "grad_norm": 0.08068038523197174, "learning_rate": 6.679836395822744e-05, "loss": 0.0043, "step": 42140 }, { "grad_norm": 0.06397140771150589, "learning_rate": 6.678278944333633e-05, "loss": 0.0051, "step": 42150 }, { "grad_norm": 0.12924525141716003, "learning_rate": 6.676721309310462e-05, "loss": 0.0055, "step": 42160 }, { "grad_norm": 0.09302754700183868, "learning_rate": 6.67516349092357e-05, "loss": 0.0046, "step": 42170 }, { "grad_norm": 0.08628649264574051, "learning_rate": 6.673605489343315e-05, "loss": 0.0048, "step": 42180 }, { "grad_norm": 0.07535620033740997, "learning_rate": 6.672047304740081e-05, "loss": 0.0051, "step": 42190 }, { "grad_norm": 0.09296178072690964, "learning_rate": 6.670488937284268e-05, "loss": 0.0042, "step": 42200 }, { "grad_norm": 0.08316181600093842, "learning_rate": 6.668930387146294e-05, "loss": 0.0042, "step": 42210 }, { "grad_norm": 0.10010123252868652, "learning_rate": 6.667371654496605e-05, "loss": 0.0051, "step": 42220 }, { "grad_norm": 0.09273414313793182, "learning_rate": 6.665812739505659e-05, "loss": 0.005, "step": 42230 }, { "grad_norm": 0.1353113055229187, "learning_rate": 6.664253642343935e-05, "loss": 0.0051, "step": 42240 }, { "grad_norm": 0.1206509917974472, "learning_rate": 6.662694363181935e-05, "loss": 0.0055, "step": 42250 }, { "grad_norm": 0.08700916171073914, "learning_rate": 6.66113490219018e-05, "loss": 0.0046, "step": 42260 }, { "grad_norm": 0.1271519958972931, "learning_rate": 6.659575259539207e-05, "loss": 0.0058, "step": 42270 }, { "grad_norm": 0.0977679118514061, "learning_rate": 6.65801543539958e-05, "loss": 0.0046, "step": 42280 }, { "grad_norm": 0.09643791615962982, "learning_rate": 6.65645542994188e-05, "loss": 0.0051, "step": 42290 }, { "grad_norm": 0.09284186363220215, "learning_rate": 6.654895243336702e-05, "loss": 0.0052, "step": 42300 }, { "grad_norm": 0.08436049520969391, "learning_rate": 6.65333487575467e-05, "loss": 0.0049, "step": 42310 }, { "grad_norm": 0.10036439448595047, "learning_rate": 6.651774327366422e-05, "loss": 0.0053, "step": 42320 }, { "grad_norm": 0.15554645657539368, "learning_rate": 6.650213598342613e-05, "loss": 0.0059, "step": 42330 }, { "grad_norm": 0.10774260014295578, "learning_rate": 6.648652688853931e-05, "loss": 0.0052, "step": 42340 }, { "grad_norm": 0.09579364955425262, "learning_rate": 6.647091599071066e-05, "loss": 0.0057, "step": 42350 }, { "grad_norm": 0.10733473300933838, "learning_rate": 6.645530329164743e-05, "loss": 0.0056, "step": 42360 }, { "grad_norm": 0.08382973819971085, "learning_rate": 6.643968879305694e-05, "loss": 0.0054, "step": 42370 }, { "grad_norm": 0.12489687651395798, "learning_rate": 6.642407249664683e-05, "loss": 0.005, "step": 42380 }, { "grad_norm": 0.09208178520202637, "learning_rate": 6.640845440412483e-05, "loss": 0.0049, "step": 42390 }, { "grad_norm": 0.1028316393494606, "learning_rate": 6.639283451719893e-05, "loss": 0.0053, "step": 42400 }, { "grad_norm": 0.12333592027425766, "learning_rate": 6.637721283757729e-05, "loss": 0.005, "step": 42410 }, { "grad_norm": 0.10086662322282791, "learning_rate": 6.636158936696828e-05, "loss": 0.0054, "step": 42420 }, { "grad_norm": 0.10297305881977081, "learning_rate": 6.634596410708047e-05, "loss": 0.0053, "step": 42430 }, { "grad_norm": 0.1197078675031662, "learning_rate": 6.633033705962257e-05, "loss": 0.0044, "step": 42440 }, { "grad_norm": 0.10293760150671005, "learning_rate": 6.631470822630359e-05, "loss": 0.0047, "step": 42450 }, { "grad_norm": 0.10128215700387955, "learning_rate": 6.629907760883263e-05, "loss": 0.0046, "step": 42460 }, { "grad_norm": 0.10444486886262894, "learning_rate": 6.628344520891905e-05, "loss": 0.0044, "step": 42470 }, { "grad_norm": 0.09938621520996094, "learning_rate": 6.62678110282724e-05, "loss": 0.0049, "step": 42480 }, { "grad_norm": 0.09954918920993805, "learning_rate": 6.625217506860237e-05, "loss": 0.0049, "step": 42490 }, { "grad_norm": 0.11206275969743729, "learning_rate": 6.623653733161892e-05, "loss": 0.0049, "step": 42500 }, { "grad_norm": 0.0729346051812172, "learning_rate": 6.622089781903217e-05, "loss": 0.0055, "step": 42510 }, { "grad_norm": 0.0916706770658493, "learning_rate": 6.620525653255241e-05, "loss": 0.0066, "step": 42520 }, { "grad_norm": 0.09087477624416351, "learning_rate": 6.618961347389015e-05, "loss": 0.0053, "step": 42530 }, { "grad_norm": 0.08652769029140472, "learning_rate": 6.617396864475613e-05, "loss": 0.0057, "step": 42540 }, { "grad_norm": 0.09706927090883255, "learning_rate": 6.61583220468612e-05, "loss": 0.005, "step": 42550 }, { "grad_norm": 0.08301892131567001, "learning_rate": 6.614267368191645e-05, "loss": 0.0041, "step": 42560 }, { "grad_norm": 0.07453721016645432, "learning_rate": 6.61270235516332e-05, "loss": 0.0056, "step": 42570 }, { "grad_norm": 0.10702495276927948, "learning_rate": 6.611137165772287e-05, "loss": 0.0047, "step": 42580 }, { "grad_norm": 0.12999506294727325, "learning_rate": 6.60957180018972e-05, "loss": 0.0046, "step": 42590 }, { "grad_norm": 0.09033241122961044, "learning_rate": 6.608006258586797e-05, "loss": 0.0054, "step": 42600 }, { "grad_norm": 0.07506384700536728, "learning_rate": 6.606440541134731e-05, "loss": 0.0048, "step": 42610 }, { "grad_norm": 0.06849875301122665, "learning_rate": 6.60487464800474e-05, "loss": 0.0048, "step": 42620 }, { "grad_norm": 0.11818956583738327, "learning_rate": 6.603308579368071e-05, "loss": 0.0056, "step": 42630 }, { "grad_norm": 0.11089026927947998, "learning_rate": 6.601742335395987e-05, "loss": 0.0055, "step": 42640 }, { "grad_norm": 0.11710329353809357, "learning_rate": 6.600175916259769e-05, "loss": 0.0054, "step": 42650 }, { "grad_norm": 0.11822997033596039, "learning_rate": 6.598609322130718e-05, "loss": 0.006, "step": 42660 }, { "grad_norm": 0.13058333098888397, "learning_rate": 6.597042553180155e-05, "loss": 0.0057, "step": 42670 }, { "grad_norm": 0.09577686339616776, "learning_rate": 6.59547560957942e-05, "loss": 0.0051, "step": 42680 }, { "grad_norm": 0.07675492018461227, "learning_rate": 6.59390849149987e-05, "loss": 0.0053, "step": 42690 }, { "grad_norm": 0.08857733011245728, "learning_rate": 6.592341199112886e-05, "loss": 0.0048, "step": 42700 }, { "grad_norm": 0.10033177584409714, "learning_rate": 6.590773732589861e-05, "loss": 0.0045, "step": 42710 }, { "grad_norm": 0.1140509694814682, "learning_rate": 6.589206092102213e-05, "loss": 0.0043, "step": 42720 }, { "grad_norm": 0.11803193390369415, "learning_rate": 6.587638277821375e-05, "loss": 0.0059, "step": 42730 }, { "grad_norm": 0.08857996761798859, "learning_rate": 6.586070289918803e-05, "loss": 0.0051, "step": 42740 }, { "grad_norm": 0.09004878252744675, "learning_rate": 6.584502128565968e-05, "loss": 0.0057, "step": 42750 }, { "grad_norm": 0.07341990619897842, "learning_rate": 6.582933793934363e-05, "loss": 0.0045, "step": 42760 }, { "grad_norm": 0.10716751217842102, "learning_rate": 6.581365286195498e-05, "loss": 0.0049, "step": 42770 }, { "grad_norm": 0.11940921097993851, "learning_rate": 6.579796605520903e-05, "loss": 0.005, "step": 42780 }, { "grad_norm": 0.08852969110012054, "learning_rate": 6.578227752082127e-05, "loss": 0.0056, "step": 42790 }, { "grad_norm": 0.06641935557126999, "learning_rate": 6.576658726050735e-05, "loss": 0.0047, "step": 42800 }, { "grad_norm": 0.08032628148794174, "learning_rate": 6.575089527598316e-05, "loss": 0.0057, "step": 42810 }, { "grad_norm": 0.10821939259767532, "learning_rate": 6.573520156896476e-05, "loss": 0.0049, "step": 42820 }, { "grad_norm": 0.13893374800682068, "learning_rate": 6.571950614116835e-05, "loss": 0.0062, "step": 42830 }, { "grad_norm": 0.1196555644273758, "learning_rate": 6.570380899431039e-05, "loss": 0.0049, "step": 42840 }, { "grad_norm": 0.10352818667888641, "learning_rate": 6.568811013010749e-05, "loss": 0.0044, "step": 42850 }, { "grad_norm": 0.10356218367815018, "learning_rate": 6.567240955027645e-05, "loss": 0.0047, "step": 42860 }, { "grad_norm": 0.08291316032409668, "learning_rate": 6.565670725653427e-05, "loss": 0.0049, "step": 42870 }, { "grad_norm": 0.10880455374717712, "learning_rate": 6.564100325059812e-05, "loss": 0.005, "step": 42880 }, { "grad_norm": 0.1156800165772438, "learning_rate": 6.562529753418538e-05, "loss": 0.0044, "step": 42890 }, { "grad_norm": 0.10720827430486679, "learning_rate": 6.56095901090136e-05, "loss": 0.0054, "step": 42900 }, { "grad_norm": 0.08322770893573761, "learning_rate": 6.55938809768005e-05, "loss": 0.0062, "step": 42910 }, { "grad_norm": 0.11646448075771332, "learning_rate": 6.557817013926402e-05, "loss": 0.0054, "step": 42920 }, { "grad_norm": 0.09787791967391968, "learning_rate": 6.556245759812227e-05, "loss": 0.0044, "step": 42930 }, { "grad_norm": 0.0987577885389328, "learning_rate": 6.554674335509357e-05, "loss": 0.0042, "step": 42940 }, { "grad_norm": 0.09724939614534378, "learning_rate": 6.553102741189638e-05, "loss": 0.0046, "step": 42950 }, { "grad_norm": 0.11163584887981415, "learning_rate": 6.55153097702494e-05, "loss": 0.0044, "step": 42960 }, { "grad_norm": 0.08751705288887024, "learning_rate": 6.549959043187146e-05, "loss": 0.0048, "step": 42970 }, { "grad_norm": 0.09681777656078339, "learning_rate": 6.54838693984816e-05, "loss": 0.0057, "step": 42980 }, { "grad_norm": 0.1254313588142395, "learning_rate": 6.546814667179909e-05, "loss": 0.0046, "step": 42990 }, { "grad_norm": 0.1251964122056961, "learning_rate": 6.545242225354328e-05, "loss": 0.0055, "step": 43000 }, { "grad_norm": 0.07932652533054352, "learning_rate": 6.543669614543383e-05, "loss": 0.0046, "step": 43010 }, { "grad_norm": 0.10408834367990494, "learning_rate": 6.542096834919049e-05, "loss": 0.0047, "step": 43020 }, { "grad_norm": 0.07488121092319489, "learning_rate": 6.540523886653324e-05, "loss": 0.0053, "step": 43030 }, { "grad_norm": 0.07903538644313812, "learning_rate": 6.538950769918222e-05, "loss": 0.0054, "step": 43040 }, { "grad_norm": 0.13585712015628815, "learning_rate": 6.537377484885779e-05, "loss": 0.0062, "step": 43050 }, { "grad_norm": 0.1282406747341156, "learning_rate": 6.535804031728045e-05, "loss": 0.0057, "step": 43060 }, { "grad_norm": 0.11252239346504211, "learning_rate": 6.534230410617092e-05, "loss": 0.0056, "step": 43070 }, { "grad_norm": 0.11602048575878143, "learning_rate": 6.532656621725007e-05, "loss": 0.0049, "step": 43080 }, { "grad_norm": 0.0859927088022232, "learning_rate": 6.531082665223898e-05, "loss": 0.005, "step": 43090 }, { "grad_norm": 0.06822189688682556, "learning_rate": 6.529508541285889e-05, "loss": 0.0039, "step": 43100 }, { "grad_norm": 0.08541105687618256, "learning_rate": 6.527934250083128e-05, "loss": 0.0043, "step": 43110 }, { "grad_norm": 0.11555551737546921, "learning_rate": 6.526359791787775e-05, "loss": 0.0048, "step": 43120 }, { "grad_norm": 0.105202816426754, "learning_rate": 6.524785166572009e-05, "loss": 0.0055, "step": 43130 }, { "grad_norm": 0.10907065123319626, "learning_rate": 6.523210374608029e-05, "loss": 0.0055, "step": 43140 }, { "grad_norm": 0.13672871887683868, "learning_rate": 6.521635416068054e-05, "loss": 0.0046, "step": 43150 }, { "grad_norm": 0.1104535311460495, "learning_rate": 6.520060291124317e-05, "loss": 0.0053, "step": 43160 }, { "grad_norm": 0.10130341351032257, "learning_rate": 6.518484999949072e-05, "loss": 0.0059, "step": 43170 }, { "grad_norm": 0.09169113636016846, "learning_rate": 6.516909542714591e-05, "loss": 0.0052, "step": 43180 }, { "grad_norm": 0.0932900458574295, "learning_rate": 6.515333919593161e-05, "loss": 0.0048, "step": 43190 }, { "grad_norm": 0.08892053365707397, "learning_rate": 6.513758130757094e-05, "loss": 0.0045, "step": 43200 }, { "grad_norm": 0.09771200269460678, "learning_rate": 6.512182176378713e-05, "loss": 0.005, "step": 43210 }, { "grad_norm": 0.0958702340722084, "learning_rate": 6.510606056630362e-05, "loss": 0.0051, "step": 43220 }, { "grad_norm": 0.08832713216543198, "learning_rate": 6.509029771684405e-05, "loss": 0.0048, "step": 43230 }, { "grad_norm": 0.13449998199939728, "learning_rate": 6.50745332171322e-05, "loss": 0.0075, "step": 43240 }, { "grad_norm": 0.0901903510093689, "learning_rate": 6.505876706889207e-05, "loss": 0.0051, "step": 43250 }, { "grad_norm": 0.10592836141586304, "learning_rate": 6.504299927384781e-05, "loss": 0.0053, "step": 43260 }, { "grad_norm": 0.1346939504146576, "learning_rate": 6.502722983372378e-05, "loss": 0.0063, "step": 43270 }, { "grad_norm": 0.1136755421757698, "learning_rate": 6.501145875024446e-05, "loss": 0.0055, "step": 43280 }, { "grad_norm": 0.08276303857564926, "learning_rate": 6.499568602513459e-05, "loss": 0.0055, "step": 43290 }, { "grad_norm": 0.07319048047065735, "learning_rate": 6.497991166011903e-05, "loss": 0.0058, "step": 43300 }, { "grad_norm": 0.10396593809127808, "learning_rate": 6.49641356569229e-05, "loss": 0.0046, "step": 43310 }, { "grad_norm": 0.11206033825874329, "learning_rate": 6.494835801727135e-05, "loss": 0.0047, "step": 43320 }, { "grad_norm": 0.09805982559919357, "learning_rate": 6.493257874288986e-05, "loss": 0.0053, "step": 43330 }, { "grad_norm": 0.086294025182724, "learning_rate": 6.491679783550399e-05, "loss": 0.0043, "step": 43340 }, { "grad_norm": 0.11246495693922043, "learning_rate": 6.490101529683957e-05, "loss": 0.0056, "step": 43350 }, { "grad_norm": 0.09714291244745255, "learning_rate": 6.488523112862248e-05, "loss": 0.0048, "step": 43360 }, { "grad_norm": 0.0875181257724762, "learning_rate": 6.486944533257891e-05, "loss": 0.0052, "step": 43370 }, { "grad_norm": 0.13137805461883545, "learning_rate": 6.485365791043517e-05, "loss": 0.005, "step": 43380 }, { "grad_norm": 0.0922633484005928, "learning_rate": 6.483786886391772e-05, "loss": 0.0059, "step": 43390 }, { "grad_norm": 0.10132692754268646, "learning_rate": 6.482207819475323e-05, "loss": 0.0054, "step": 43400 }, { "grad_norm": 0.11343558877706528, "learning_rate": 6.480628590466857e-05, "loss": 0.0066, "step": 43410 }, { "grad_norm": 0.11087311059236526, "learning_rate": 6.479049199539074e-05, "loss": 0.0056, "step": 43420 }, { "grad_norm": 0.1261579692363739, "learning_rate": 6.477469646864693e-05, "loss": 0.005, "step": 43430 }, { "grad_norm": 0.09493012726306915, "learning_rate": 6.475889932616454e-05, "loss": 0.0052, "step": 43440 }, { "grad_norm": 0.1097632497549057, "learning_rate": 6.474310056967111e-05, "loss": 0.0055, "step": 43450 }, { "grad_norm": 0.12795881927013397, "learning_rate": 6.472730020089437e-05, "loss": 0.0053, "step": 43460 }, { "grad_norm": 0.12093948572874069, "learning_rate": 6.471149822156222e-05, "loss": 0.0049, "step": 43470 }, { "grad_norm": 0.12099424749612808, "learning_rate": 6.469569463340275e-05, "loss": 0.0049, "step": 43480 }, { "grad_norm": 0.08274838328361511, "learning_rate": 6.467988943814418e-05, "loss": 0.0048, "step": 43490 }, { "grad_norm": 0.11027451604604721, "learning_rate": 6.4664082637515e-05, "loss": 0.0055, "step": 43500 }, { "grad_norm": 0.09344281256198883, "learning_rate": 6.464827423324377e-05, "loss": 0.0046, "step": 43510 }, { "grad_norm": 0.11151709407567978, "learning_rate": 6.463246422705931e-05, "loss": 0.0055, "step": 43520 }, { "grad_norm": 0.10081377625465393, "learning_rate": 6.461665262069055e-05, "loss": 0.0052, "step": 43530 }, { "grad_norm": 0.11016398668289185, "learning_rate": 6.460083941586665e-05, "loss": 0.0055, "step": 43540 }, { "grad_norm": 0.10246555507183075, "learning_rate": 6.45850246143169e-05, "loss": 0.005, "step": 43550 }, { "grad_norm": 0.10157833248376846, "learning_rate": 6.456920821777077e-05, "loss": 0.0057, "step": 43560 }, { "grad_norm": 0.09377210587263107, "learning_rate": 6.455339022795796e-05, "loss": 0.0041, "step": 43570 }, { "grad_norm": 0.1406259834766388, "learning_rate": 6.453757064660825e-05, "loss": 0.0052, "step": 43580 }, { "grad_norm": 0.11068092286586761, "learning_rate": 6.452174947545169e-05, "loss": 0.0056, "step": 43590 }, { "grad_norm": 0.1184028834104538, "learning_rate": 6.450592671621842e-05, "loss": 0.0061, "step": 43600 }, { "grad_norm": 0.14473803341388702, "learning_rate": 6.449010237063882e-05, "loss": 0.0061, "step": 43610 }, { "grad_norm": 0.12239585071802139, "learning_rate": 6.447427644044343e-05, "loss": 0.0066, "step": 43620 }, { "grad_norm": 0.12158411741256714, "learning_rate": 6.44584489273629e-05, "loss": 0.0075, "step": 43630 }, { "grad_norm": 0.13333052396774292, "learning_rate": 6.444261983312817e-05, "loss": 0.008, "step": 43640 }, { "grad_norm": 0.135192409157753, "learning_rate": 6.442678915947023e-05, "loss": 0.0066, "step": 43650 }, { "grad_norm": 0.15352696180343628, "learning_rate": 6.441095690812032e-05, "loss": 0.0078, "step": 43660 }, { "grad_norm": 0.11492375284433365, "learning_rate": 6.439512308080982e-05, "loss": 0.007, "step": 43670 }, { "grad_norm": 0.10956470668315887, "learning_rate": 6.43792876792703e-05, "loss": 0.0065, "step": 43680 }, { "grad_norm": 0.13168172538280487, "learning_rate": 6.43634507052335e-05, "loss": 0.0065, "step": 43690 }, { "grad_norm": 0.14954903721809387, "learning_rate": 6.434761216043135e-05, "loss": 0.0068, "step": 43700 }, { "grad_norm": 0.08347994089126587, "learning_rate": 6.433177204659589e-05, "loss": 0.0068, "step": 43710 }, { "grad_norm": 0.13161778450012207, "learning_rate": 6.431593036545938e-05, "loss": 0.006, "step": 43720 }, { "grad_norm": 0.10853049159049988, "learning_rate": 6.430008711875426e-05, "loss": 0.0072, "step": 43730 }, { "grad_norm": 0.09395390003919601, "learning_rate": 6.42842423082131e-05, "loss": 0.0057, "step": 43740 }, { "grad_norm": 0.07804155349731445, "learning_rate": 6.426839593556868e-05, "loss": 0.0046, "step": 43750 }, { "grad_norm": 0.11172374337911606, "learning_rate": 6.425254800255392e-05, "loss": 0.0063, "step": 43760 }, { "grad_norm": 0.08944377303123474, "learning_rate": 6.423669851090196e-05, "loss": 0.0051, "step": 43770 }, { "grad_norm": 0.12158022820949554, "learning_rate": 6.422084746234604e-05, "loss": 0.0063, "step": 43780 }, { "grad_norm": 0.08774279803037643, "learning_rate": 6.420499485861965e-05, "loss": 0.0045, "step": 43790 }, { "grad_norm": 0.10999347269535065, "learning_rate": 6.418914070145636e-05, "loss": 0.0061, "step": 43800 }, { "grad_norm": 0.09563827514648438, "learning_rate": 6.417328499258996e-05, "loss": 0.0053, "step": 43810 }, { "grad_norm": 0.1339760571718216, "learning_rate": 6.415742773375446e-05, "loss": 0.0055, "step": 43820 }, { "grad_norm": 0.08725477010011673, "learning_rate": 6.414156892668393e-05, "loss": 0.0063, "step": 43830 }, { "grad_norm": 0.12708669900894165, "learning_rate": 6.412570857311267e-05, "loss": 0.0049, "step": 43840 }, { "grad_norm": 0.10706286132335663, "learning_rate": 6.410984667477518e-05, "loss": 0.005, "step": 43850 }, { "grad_norm": 0.07894868403673172, "learning_rate": 6.409398323340607e-05, "loss": 0.0049, "step": 43860 }, { "grad_norm": 0.11055935174226761, "learning_rate": 6.407811825074012e-05, "loss": 0.0053, "step": 43870 }, { "grad_norm": 0.09063178300857544, "learning_rate": 6.406225172851234e-05, "loss": 0.0044, "step": 43880 }, { "grad_norm": 0.10863370448350906, "learning_rate": 6.404638366845786e-05, "loss": 0.0047, "step": 43890 }, { "grad_norm": 0.12006962299346924, "learning_rate": 6.403051407231196e-05, "loss": 0.0049, "step": 43900 }, { "grad_norm": 0.07447808235883713, "learning_rate": 6.401464294181016e-05, "loss": 0.0057, "step": 43910 }, { "grad_norm": 0.11835481226444244, "learning_rate": 6.399877027868808e-05, "loss": 0.0053, "step": 43920 }, { "grad_norm": 0.0998135581612587, "learning_rate": 6.39828960846815e-05, "loss": 0.0055, "step": 43930 }, { "grad_norm": 0.08538269251585007, "learning_rate": 6.396702036152644e-05, "loss": 0.005, "step": 43940 }, { "grad_norm": 0.08584325015544891, "learning_rate": 6.395114311095905e-05, "loss": 0.0057, "step": 43950 }, { "grad_norm": 0.130081444978714, "learning_rate": 6.393526433471559e-05, "loss": 0.0056, "step": 43960 }, { "grad_norm": 0.11104003340005875, "learning_rate": 6.391938403453262e-05, "loss": 0.0053, "step": 43970 }, { "grad_norm": 0.12439735233783722, "learning_rate": 6.390350221214671e-05, "loss": 0.0056, "step": 43980 }, { "grad_norm": 0.11224766075611115, "learning_rate": 6.38876188692947e-05, "loss": 0.0054, "step": 43990 }, { "grad_norm": 0.13368751108646393, "learning_rate": 6.387173400771358e-05, "loss": 0.0047, "step": 44000 }, { "grad_norm": 0.10962747782468796, "learning_rate": 6.385584762914047e-05, "loss": 0.0053, "step": 44010 }, { "grad_norm": 0.10296355187892914, "learning_rate": 6.38399597353127e-05, "loss": 0.0048, "step": 44020 }, { "grad_norm": 0.09158313274383545, "learning_rate": 6.382407032796775e-05, "loss": 0.0048, "step": 44030 }, { "grad_norm": 0.09796541184186935, "learning_rate": 6.380817940884325e-05, "loss": 0.0058, "step": 44040 }, { "grad_norm": 0.12720130383968353, "learning_rate": 6.379228697967702e-05, "loss": 0.0056, "step": 44050 }, { "grad_norm": 0.10900719463825226, "learning_rate": 6.3776393042207e-05, "loss": 0.0054, "step": 44060 }, { "grad_norm": 0.11945760995149612, "learning_rate": 6.376049759817137e-05, "loss": 0.0047, "step": 44070 }, { "grad_norm": 0.0869809091091156, "learning_rate": 6.37446006493084e-05, "loss": 0.0045, "step": 44080 }, { "grad_norm": 0.09067908674478531, "learning_rate": 6.372870219735656e-05, "loss": 0.0048, "step": 44090 }, { "grad_norm": 0.08967103064060211, "learning_rate": 6.371280224405452e-05, "loss": 0.0043, "step": 44100 }, { "grad_norm": 0.09247991442680359, "learning_rate": 6.369690079114101e-05, "loss": 0.0045, "step": 44110 }, { "grad_norm": 0.11749012768268585, "learning_rate": 6.368099784035504e-05, "loss": 0.0057, "step": 44120 }, { "grad_norm": 0.10338347405195236, "learning_rate": 6.366509339343572e-05, "loss": 0.0047, "step": 44130 }, { "grad_norm": 0.10115084052085876, "learning_rate": 6.364918745212232e-05, "loss": 0.0047, "step": 44140 }, { "grad_norm": 0.10203782469034195, "learning_rate": 6.363328001815431e-05, "loss": 0.0061, "step": 44150 }, { "grad_norm": 0.10872375220060349, "learning_rate": 6.361737109327128e-05, "loss": 0.0047, "step": 44160 }, { "grad_norm": 0.1294129341840744, "learning_rate": 6.360146067921301e-05, "loss": 0.0048, "step": 44170 }, { "grad_norm": 0.11468083411455154, "learning_rate": 6.35855487777195e-05, "loss": 0.0051, "step": 44180 }, { "grad_norm": 0.1268087923526764, "learning_rate": 6.356963539053076e-05, "loss": 0.0051, "step": 44190 }, { "grad_norm": 0.11018216609954834, "learning_rate": 6.355372051938711e-05, "loss": 0.0057, "step": 44200 }, { "grad_norm": 0.10060601681470871, "learning_rate": 6.353780416602894e-05, "loss": 0.0045, "step": 44210 }, { "grad_norm": 0.1129390224814415, "learning_rate": 6.352188633219689e-05, "loss": 0.0049, "step": 44220 }, { "grad_norm": 0.10944466292858124, "learning_rate": 6.350596701963166e-05, "loss": 0.0048, "step": 44230 }, { "grad_norm": 0.11503937840461731, "learning_rate": 6.349004623007419e-05, "loss": 0.006, "step": 44240 }, { "grad_norm": 0.08509059250354767, "learning_rate": 6.347412396526555e-05, "loss": 0.0046, "step": 44250 }, { "grad_norm": 0.10269273072481155, "learning_rate": 6.345820022694696e-05, "loss": 0.0054, "step": 44260 }, { "grad_norm": 0.09430182725191116, "learning_rate": 6.344227501685984e-05, "loss": 0.0056, "step": 44270 }, { "grad_norm": 0.09037316590547562, "learning_rate": 6.342634833674572e-05, "loss": 0.0053, "step": 44280 }, { "grad_norm": 0.10145900398492813, "learning_rate": 6.341042018834635e-05, "loss": 0.0051, "step": 44290 }, { "grad_norm": 0.16637147963047028, "learning_rate": 6.339449057340359e-05, "loss": 0.0048, "step": 44300 }, { "grad_norm": 0.0926050916314125, "learning_rate": 6.337855949365945e-05, "loss": 0.0057, "step": 44310 }, { "grad_norm": 0.09887930750846863, "learning_rate": 6.336262695085619e-05, "loss": 0.0053, "step": 44320 }, { "grad_norm": 0.0970158576965332, "learning_rate": 6.334669294673612e-05, "loss": 0.0039, "step": 44330 }, { "grad_norm": 0.07186640053987503, "learning_rate": 6.333075748304179e-05, "loss": 0.0053, "step": 44340 }, { "grad_norm": 0.09766057133674622, "learning_rate": 6.331482056151585e-05, "loss": 0.0058, "step": 44350 }, { "grad_norm": 0.11937277764081955, "learning_rate": 6.329888218390117e-05, "loss": 0.0053, "step": 44360 }, { "grad_norm": 0.15288563072681427, "learning_rate": 6.328294235194072e-05, "loss": 0.005, "step": 44370 }, { "grad_norm": 0.10859627276659012, "learning_rate": 6.326700106737765e-05, "loss": 0.0054, "step": 44380 }, { "grad_norm": 0.1012161448597908, "learning_rate": 6.325105833195531e-05, "loss": 0.0051, "step": 44390 }, { "grad_norm": 0.1029212474822998, "learning_rate": 6.323511414741715e-05, "loss": 0.0051, "step": 44400 }, { "grad_norm": 0.10338257253170013, "learning_rate": 6.321916851550678e-05, "loss": 0.0049, "step": 44410 }, { "grad_norm": 0.10661432892084122, "learning_rate": 6.320322143796806e-05, "loss": 0.005, "step": 44420 }, { "grad_norm": 0.08384723961353302, "learning_rate": 6.318727291654488e-05, "loss": 0.0054, "step": 44430 }, { "grad_norm": 0.08564162999391556, "learning_rate": 6.317132295298134e-05, "loss": 0.0065, "step": 44440 }, { "grad_norm": 0.07847645878791809, "learning_rate": 6.315537154902173e-05, "loss": 0.0057, "step": 44450 }, { "grad_norm": 0.06381380558013916, "learning_rate": 6.313941870641048e-05, "loss": 0.0049, "step": 44460 }, { "grad_norm": 0.10417874902486801, "learning_rate": 6.312346442689214e-05, "loss": 0.0053, "step": 44470 }, { "grad_norm": 0.10868186503648758, "learning_rate": 6.310750871221147e-05, "loss": 0.006, "step": 44480 }, { "grad_norm": 0.10555291920900345, "learning_rate": 6.309155156411335e-05, "loss": 0.0048, "step": 44490 }, { "grad_norm": 0.10116701573133469, "learning_rate": 6.307559298434284e-05, "loss": 0.0038, "step": 44500 }, { "grad_norm": 0.09171325713396072, "learning_rate": 6.305963297464512e-05, "loss": 0.0045, "step": 44510 }, { "grad_norm": 0.07268057763576508, "learning_rate": 6.304367153676561e-05, "loss": 0.0056, "step": 44520 }, { "grad_norm": 0.11320775747299194, "learning_rate": 6.302770867244978e-05, "loss": 0.0052, "step": 44530 }, { "grad_norm": 0.1162404716014862, "learning_rate": 6.301174438344328e-05, "loss": 0.006, "step": 44540 }, { "grad_norm": 0.1095435693860054, "learning_rate": 6.299577867149202e-05, "loss": 0.0061, "step": 44550 }, { "grad_norm": 0.11942032724618912, "learning_rate": 6.297981153834192e-05, "loss": 0.0052, "step": 44560 }, { "grad_norm": 0.08870716392993927, "learning_rate": 6.296384298573916e-05, "loss": 0.0054, "step": 44570 }, { "grad_norm": 0.07845005393028259, "learning_rate": 6.294787301543001e-05, "loss": 0.0037, "step": 44580 }, { "grad_norm": 0.07728194445371628, "learning_rate": 6.293190162916095e-05, "loss": 0.0054, "step": 44590 }, { "grad_norm": 0.0910470262169838, "learning_rate": 6.291592882867855e-05, "loss": 0.0047, "step": 44600 }, { "grad_norm": 0.10239818692207336, "learning_rate": 6.28999546157296e-05, "loss": 0.0052, "step": 44610 }, { "grad_norm": 0.10283772647380829, "learning_rate": 6.288397899206102e-05, "loss": 0.0051, "step": 44620 }, { "grad_norm": 0.10130111873149872, "learning_rate": 6.286800195941984e-05, "loss": 0.0053, "step": 44630 }, { "grad_norm": 0.07659991085529327, "learning_rate": 6.285202351955334e-05, "loss": 0.0044, "step": 44640 }, { "grad_norm": 0.12311125546693802, "learning_rate": 6.283604367420887e-05, "loss": 0.0049, "step": 44650 }, { "grad_norm": 0.09991604089736938, "learning_rate": 6.282006242513394e-05, "loss": 0.0042, "step": 44660 }, { "grad_norm": 0.097651407122612, "learning_rate": 6.280407977407628e-05, "loss": 0.0039, "step": 44670 }, { "grad_norm": 0.06682667881250381, "learning_rate": 6.27880957227837e-05, "loss": 0.0038, "step": 44680 }, { "grad_norm": 0.09009882062673569, "learning_rate": 6.27721102730042e-05, "loss": 0.0047, "step": 44690 }, { "grad_norm": 0.08259434998035431, "learning_rate": 6.275612342648592e-05, "loss": 0.0046, "step": 44700 }, { "grad_norm": 0.11074136942625046, "learning_rate": 6.274013518497716e-05, "loss": 0.0045, "step": 44710 }, { "grad_norm": 0.1127709299325943, "learning_rate": 6.272414555022636e-05, "loss": 0.0053, "step": 44720 }, { "grad_norm": 0.10213283449411392, "learning_rate": 6.270815452398215e-05, "loss": 0.0043, "step": 44730 }, { "grad_norm": 0.09097893536090851, "learning_rate": 6.269216210799326e-05, "loss": 0.004, "step": 44740 }, { "grad_norm": 0.10880596190690994, "learning_rate": 6.26761683040086e-05, "loss": 0.005, "step": 44750 }, { "grad_norm": 0.1051020398736, "learning_rate": 6.266017311377723e-05, "loss": 0.0036, "step": 44760 }, { "grad_norm": 0.11397437751293182, "learning_rate": 6.264417653904839e-05, "loss": 0.0049, "step": 44770 }, { "grad_norm": 0.11573643982410431, "learning_rate": 6.262817858157139e-05, "loss": 0.0042, "step": 44780 }, { "grad_norm": 0.1014668196439743, "learning_rate": 6.261217924309576e-05, "loss": 0.0057, "step": 44790 }, { "grad_norm": 0.10590384155511856, "learning_rate": 6.259617852537118e-05, "loss": 0.0051, "step": 44800 }, { "grad_norm": 0.09508483111858368, "learning_rate": 6.258017643014747e-05, "loss": 0.0051, "step": 44810 }, { "grad_norm": 0.08053506165742874, "learning_rate": 6.256417295917456e-05, "loss": 0.0046, "step": 44820 }, { "grad_norm": 0.113677479326725, "learning_rate": 6.254816811420258e-05, "loss": 0.0044, "step": 44830 }, { "grad_norm": 0.0949171707034111, "learning_rate": 6.253216189698183e-05, "loss": 0.0043, "step": 44840 }, { "grad_norm": 0.07725010812282562, "learning_rate": 6.251615430926267e-05, "loss": 0.0037, "step": 44850 }, { "grad_norm": 0.11548522859811783, "learning_rate": 6.25001453527957e-05, "loss": 0.0048, "step": 44860 }, { "grad_norm": 0.1161985993385315, "learning_rate": 6.248413502933164e-05, "loss": 0.0044, "step": 44870 }, { "grad_norm": 0.10717602074146271, "learning_rate": 6.246812334062133e-05, "loss": 0.0045, "step": 44880 }, { "grad_norm": 0.08988078683614731, "learning_rate": 6.245211028841579e-05, "loss": 0.0053, "step": 44890 }, { "grad_norm": 0.10446520149707794, "learning_rate": 6.24360958744662e-05, "loss": 0.005, "step": 44900 }, { "grad_norm": 0.10604091733694077, "learning_rate": 6.242008010052387e-05, "loss": 0.0044, "step": 44910 }, { "grad_norm": 0.0803682804107666, "learning_rate": 6.240406296834024e-05, "loss": 0.0043, "step": 44920 }, { "grad_norm": 0.11472208797931671, "learning_rate": 6.238804447966694e-05, "loss": 0.0043, "step": 44930 }, { "grad_norm": 0.10221628099679947, "learning_rate": 6.237202463625573e-05, "loss": 0.0054, "step": 44940 }, { "grad_norm": 0.09728352725505829, "learning_rate": 6.235600343985848e-05, "loss": 0.004, "step": 44950 }, { "grad_norm": 0.09527727961540222, "learning_rate": 6.233998089222729e-05, "loss": 0.0044, "step": 44960 }, { "grad_norm": 0.08608711510896683, "learning_rate": 6.232395699511433e-05, "loss": 0.0046, "step": 44970 }, { "grad_norm": 0.09429434686899185, "learning_rate": 6.230793175027199e-05, "loss": 0.0049, "step": 44980 }, { "grad_norm": 0.08130284398794174, "learning_rate": 6.22919051594527e-05, "loss": 0.0039, "step": 44990 }, { "grad_norm": 0.10272777080535889, "learning_rate": 6.227587722440917e-05, "loss": 0.0046, "step": 45000 }, { "grad_norm": 0.1117338091135025, "learning_rate": 6.225984794689414e-05, "loss": 0.0042, "step": 45010 }, { "grad_norm": 0.1008196696639061, "learning_rate": 6.22438173286606e-05, "loss": 0.0047, "step": 45020 }, { "grad_norm": 0.10744157433509827, "learning_rate": 6.222778537146159e-05, "loss": 0.0051, "step": 45030 }, { "grad_norm": 0.08404714614152908, "learning_rate": 6.221175207705037e-05, "loss": 0.0043, "step": 45040 }, { "grad_norm": 0.11551322788000107, "learning_rate": 6.219571744718027e-05, "loss": 0.0047, "step": 45050 }, { "grad_norm": 0.128890722990036, "learning_rate": 6.217968148360487e-05, "loss": 0.0061, "step": 45060 }, { "grad_norm": 0.110353484749794, "learning_rate": 6.216364418807782e-05, "loss": 0.005, "step": 45070 }, { "grad_norm": 0.07953719794750214, "learning_rate": 6.214760556235292e-05, "loss": 0.0054, "step": 45080 }, { "grad_norm": 0.07147657871246338, "learning_rate": 6.213156560818414e-05, "loss": 0.0041, "step": 45090 }, { "grad_norm": 0.07859409600496292, "learning_rate": 6.211552432732559e-05, "loss": 0.0045, "step": 45100 }, { "grad_norm": 0.1067918986082077, "learning_rate": 6.20994817215315e-05, "loss": 0.0048, "step": 45110 }, { "grad_norm": 0.08642381429672241, "learning_rate": 6.208343779255628e-05, "loss": 0.0067, "step": 45120 }, { "grad_norm": 0.08749724179506302, "learning_rate": 6.206739254215449e-05, "loss": 0.0055, "step": 45130 }, { "grad_norm": 0.12707878649234772, "learning_rate": 6.205134597208077e-05, "loss": 0.0045, "step": 45140 }, { "grad_norm": 0.11287384480237961, "learning_rate": 6.203529808409e-05, "loss": 0.005, "step": 45150 }, { "grad_norm": 0.10160811245441437, "learning_rate": 6.201924887993712e-05, "loss": 0.0049, "step": 45160 }, { "grad_norm": 0.10893452912569046, "learning_rate": 6.200319836137725e-05, "loss": 0.0053, "step": 45170 }, { "grad_norm": 0.08935166895389557, "learning_rate": 6.198714653016565e-05, "loss": 0.0044, "step": 45180 }, { "grad_norm": 0.12639068067073822, "learning_rate": 6.197109338805774e-05, "loss": 0.006, "step": 45190 }, { "grad_norm": 0.12567496299743652, "learning_rate": 6.195503893680903e-05, "loss": 0.0044, "step": 45200 }, { "grad_norm": 0.08192409574985504, "learning_rate": 6.193898317817524e-05, "loss": 0.0061, "step": 45210 }, { "grad_norm": 0.11382318288087845, "learning_rate": 6.192292611391221e-05, "loss": 0.0048, "step": 45220 }, { "grad_norm": 0.10970506072044373, "learning_rate": 6.190686774577591e-05, "loss": 0.005, "step": 45230 }, { "grad_norm": 0.10996975004673004, "learning_rate": 6.189080807552245e-05, "loss": 0.0065, "step": 45240 }, { "grad_norm": 0.1268063187599182, "learning_rate": 6.187474710490809e-05, "loss": 0.0039, "step": 45250 }, { "grad_norm": 0.10066217184066772, "learning_rate": 6.185868483568926e-05, "loss": 0.0047, "step": 45260 }, { "grad_norm": 0.10637200623750687, "learning_rate": 6.184262126962245e-05, "loss": 0.0055, "step": 45270 }, { "grad_norm": 0.10297322273254395, "learning_rate": 6.182655640846442e-05, "loss": 0.0061, "step": 45280 }, { "grad_norm": 0.08786343783140182, "learning_rate": 6.181049025397196e-05, "loss": 0.0041, "step": 45290 }, { "grad_norm": 0.10782910883426666, "learning_rate": 6.179442280790202e-05, "loss": 0.0056, "step": 45300 }, { "grad_norm": 0.11524581164121628, "learning_rate": 6.177835407201174e-05, "loss": 0.0048, "step": 45310 }, { "grad_norm": 0.10315055400133133, "learning_rate": 6.176228404805839e-05, "loss": 0.0051, "step": 45320 }, { "grad_norm": 0.09648629277944565, "learning_rate": 6.174621273779932e-05, "loss": 0.006, "step": 45330 }, { "grad_norm": 0.10176539421081543, "learning_rate": 6.17301401429921e-05, "loss": 0.0051, "step": 45340 }, { "grad_norm": 0.10537955164909363, "learning_rate": 6.171406626539436e-05, "loss": 0.0055, "step": 45350 }, { "grad_norm": 0.09645907580852509, "learning_rate": 6.169799110676398e-05, "loss": 0.0043, "step": 45360 }, { "grad_norm": 0.11878026276826859, "learning_rate": 6.168191466885885e-05, "loss": 0.0059, "step": 45370 }, { "grad_norm": 0.08847343176603317, "learning_rate": 6.166583695343711e-05, "loss": 0.0054, "step": 45380 }, { "grad_norm": 0.08574431389570236, "learning_rate": 6.164975796225698e-05, "loss": 0.004, "step": 45390 }, { "grad_norm": 0.0841890424489975, "learning_rate": 6.163367769707683e-05, "loss": 0.0047, "step": 45400 }, { "grad_norm": 0.10195734351873398, "learning_rate": 6.161759615965519e-05, "loss": 0.0043, "step": 45410 }, { "grad_norm": 0.10607840120792389, "learning_rate": 6.16015133517507e-05, "loss": 0.0055, "step": 45420 }, { "grad_norm": 0.11035256832838058, "learning_rate": 6.158542927512214e-05, "loss": 0.0045, "step": 45430 }, { "grad_norm": 0.0953969806432724, "learning_rate": 6.156934393152846e-05, "loss": 0.0044, "step": 45440 }, { "grad_norm": 0.06158720329403877, "learning_rate": 6.15532573227287e-05, "loss": 0.0043, "step": 45450 }, { "grad_norm": 0.08667564392089844, "learning_rate": 6.153716945048212e-05, "loss": 0.0044, "step": 45460 }, { "grad_norm": 0.10393970459699631, "learning_rate": 6.152108031654802e-05, "loss": 0.0053, "step": 45470 }, { "grad_norm": 0.11296308040618896, "learning_rate": 6.15049899226859e-05, "loss": 0.0069, "step": 45480 }, { "grad_norm": 0.13153021037578583, "learning_rate": 6.148889827065537e-05, "loss": 0.0052, "step": 45490 }, { "grad_norm": 0.10812881588935852, "learning_rate": 6.147280536221622e-05, "loss": 0.0048, "step": 45500 }, { "grad_norm": 0.13708002865314484, "learning_rate": 6.145671119912832e-05, "loss": 0.0048, "step": 45510 }, { "grad_norm": 0.11147349327802658, "learning_rate": 6.144061578315169e-05, "loss": 0.0063, "step": 45520 }, { "grad_norm": 0.09617161005735397, "learning_rate": 6.142451911604654e-05, "loss": 0.0049, "step": 45530 }, { "grad_norm": 0.11023347079753876, "learning_rate": 6.140842119957315e-05, "loss": 0.0067, "step": 45540 }, { "grad_norm": 0.08563409745693207, "learning_rate": 6.139232203549197e-05, "loss": 0.0048, "step": 45550 }, { "grad_norm": 0.09777150303125381, "learning_rate": 6.137622162556357e-05, "loss": 0.0037, "step": 45560 }, { "grad_norm": 0.10352298617362976, "learning_rate": 6.13601199715487e-05, "loss": 0.0047, "step": 45570 }, { "grad_norm": 0.11625572293996811, "learning_rate": 6.134401707520816e-05, "loss": 0.0058, "step": 45580 }, { "grad_norm": 0.11724168807268143, "learning_rate": 6.132791293830298e-05, "loss": 0.0046, "step": 45590 }, { "grad_norm": 0.12261715531349182, "learning_rate": 6.131180756259428e-05, "loss": 0.0041, "step": 45600 }, { "grad_norm": 0.07253459841012955, "learning_rate": 6.129570094984331e-05, "loss": 0.0045, "step": 45610 }, { "grad_norm": 0.1445188969373703, "learning_rate": 6.127959310181145e-05, "loss": 0.0047, "step": 45620 }, { "grad_norm": 0.08678780496120453, "learning_rate": 6.126348402026026e-05, "loss": 0.0041, "step": 45630 }, { "grad_norm": 0.08489502221345901, "learning_rate": 6.12473737069514e-05, "loss": 0.0047, "step": 45640 }, { "grad_norm": 0.1201128214597702, "learning_rate": 6.123126216364665e-05, "loss": 0.0057, "step": 45650 }, { "grad_norm": 0.1052638590335846, "learning_rate": 6.121514939210797e-05, "loss": 0.0042, "step": 45660 }, { "grad_norm": 0.08892206102609634, "learning_rate": 6.119903539409741e-05, "loss": 0.0056, "step": 45670 }, { "grad_norm": 0.08034037798643112, "learning_rate": 6.118292017137716e-05, "loss": 0.0051, "step": 45680 }, { "grad_norm": 0.09204373508691788, "learning_rate": 6.116680372570959e-05, "loss": 0.0043, "step": 45690 }, { "grad_norm": 0.09421781450510025, "learning_rate": 6.115068605885713e-05, "loss": 0.0046, "step": 45700 }, { "grad_norm": 0.1110864207148552, "learning_rate": 6.113456717258243e-05, "loss": 0.0057, "step": 45710 }, { "grad_norm": 0.09760359674692154, "learning_rate": 6.11184470686482e-05, "loss": 0.0059, "step": 45720 }, { "grad_norm": 0.06713858246803284, "learning_rate": 6.110232574881731e-05, "loss": 0.004, "step": 45730 }, { "grad_norm": 0.09052541851997375, "learning_rate": 6.108620321485277e-05, "loss": 0.0045, "step": 45740 }, { "grad_norm": 0.07468931376934052, "learning_rate": 6.107007946851773e-05, "loss": 0.0047, "step": 45750 }, { "grad_norm": 0.11762451380491257, "learning_rate": 6.105395451157542e-05, "loss": 0.0054, "step": 45760 }, { "grad_norm": 0.0949929878115654, "learning_rate": 6.103782834578928e-05, "loss": 0.0047, "step": 45770 }, { "grad_norm": 0.0824572890996933, "learning_rate": 6.102170097292281e-05, "loss": 0.0044, "step": 45780 }, { "grad_norm": 0.07820724695920944, "learning_rate": 6.1005572394739716e-05, "loss": 0.0043, "step": 45790 }, { "grad_norm": 0.07733690738677979, "learning_rate": 6.0989442613003765e-05, "loss": 0.0043, "step": 45800 }, { "grad_norm": 0.08868248760700226, "learning_rate": 6.097331162947888e-05, "loss": 0.0041, "step": 45810 }, { "grad_norm": 0.07503602653741837, "learning_rate": 6.095717944592914e-05, "loss": 0.0047, "step": 45820 }, { "grad_norm": 0.10549815744161606, "learning_rate": 6.094104606411873e-05, "loss": 0.0058, "step": 45830 }, { "grad_norm": 0.1105477586388588, "learning_rate": 6.0924911485811966e-05, "loss": 0.0045, "step": 45840 }, { "grad_norm": 0.11223562806844711, "learning_rate": 6.090877571277331e-05, "loss": 0.0044, "step": 45850 }, { "grad_norm": 0.11737807095050812, "learning_rate": 6.089263874676736e-05, "loss": 0.0052, "step": 45860 }, { "grad_norm": 0.09931574016809464, "learning_rate": 6.0876500589558796e-05, "loss": 0.0053, "step": 45870 }, { "grad_norm": 0.11301986873149872, "learning_rate": 6.086036124291248e-05, "loss": 0.0048, "step": 45880 }, { "grad_norm": 0.08384910970926285, "learning_rate": 6.084422070859339e-05, "loss": 0.0046, "step": 45890 }, { "grad_norm": 0.10871277004480362, "learning_rate": 6.082807898836663e-05, "loss": 0.0055, "step": 45900 }, { "grad_norm": 0.08703676611185074, "learning_rate": 6.081193608399742e-05, "loss": 0.0051, "step": 45910 }, { "grad_norm": 0.08454418182373047, "learning_rate": 6.0795791997251164e-05, "loss": 0.0045, "step": 45920 }, { "grad_norm": 0.09563294798135757, "learning_rate": 6.0779646729893294e-05, "loss": 0.0048, "step": 45930 }, { "grad_norm": 0.09794995188713074, "learning_rate": 6.0763500283689476e-05, "loss": 0.0048, "step": 45940 }, { "grad_norm": 0.07375495880842209, "learning_rate": 6.0747352660405455e-05, "loss": 0.005, "step": 45950 }, { "grad_norm": 0.08933544158935547, "learning_rate": 6.073120386180709e-05, "loss": 0.005, "step": 45960 }, { "grad_norm": 0.11379685997962952, "learning_rate": 6.0715053889660425e-05, "loss": 0.0047, "step": 45970 }, { "grad_norm": 0.09559892863035202, "learning_rate": 6.069890274573157e-05, "loss": 0.0045, "step": 45980 }, { "grad_norm": 0.15674126148223877, "learning_rate": 6.068275043178679e-05, "loss": 0.006, "step": 45990 }, { "grad_norm": 0.12556038796901703, "learning_rate": 6.066659694959248e-05, "loss": 0.006, "step": 46000 }, { "grad_norm": 0.09242552518844604, "learning_rate": 6.0650442300915176e-05, "loss": 0.0066, "step": 46010 }, { "grad_norm": 0.10053988546133041, "learning_rate": 6.063428648752152e-05, "loss": 0.0055, "step": 46020 }, { "grad_norm": 0.11393022537231445, "learning_rate": 6.0618129511178266e-05, "loss": 0.0054, "step": 46030 }, { "grad_norm": 0.08715410530567169, "learning_rate": 6.060197137365234e-05, "loss": 0.0052, "step": 46040 }, { "grad_norm": 0.09309086203575134, "learning_rate": 6.058581207671077e-05, "loss": 0.0049, "step": 46050 }, { "grad_norm": 0.16410040855407715, "learning_rate": 6.056965162212072e-05, "loss": 0.0043, "step": 46060 }, { "grad_norm": 0.10894568264484406, "learning_rate": 6.055349001164943e-05, "loss": 0.0049, "step": 46070 }, { "grad_norm": 0.13771766424179077, "learning_rate": 6.0537327247064347e-05, "loss": 0.0058, "step": 46080 }, { "grad_norm": 0.13849230110645294, "learning_rate": 6.0521163330133014e-05, "loss": 0.0055, "step": 46090 }, { "grad_norm": 0.13724832236766815, "learning_rate": 6.050499826262306e-05, "loss": 0.0063, "step": 46100 }, { "grad_norm": 0.08962121605873108, "learning_rate": 6.0488832046302294e-05, "loss": 0.0049, "step": 46110 }, { "grad_norm": 0.09225985407829285, "learning_rate": 6.0472664682938626e-05, "loss": 0.0041, "step": 46120 }, { "grad_norm": 0.07737231254577637, "learning_rate": 6.045649617430009e-05, "loss": 0.0052, "step": 46130 }, { "grad_norm": 0.07447291165590286, "learning_rate": 6.0440326522154866e-05, "loss": 0.0044, "step": 46140 }, { "grad_norm": 0.13529616594314575, "learning_rate": 6.0424155728271224e-05, "loss": 0.0055, "step": 46150 }, { "grad_norm": 0.12942704558372498, "learning_rate": 6.040798379441758e-05, "loss": 0.0039, "step": 46160 }, { "grad_norm": 0.08456612378358841, "learning_rate": 6.0391810722362485e-05, "loss": 0.0044, "step": 46170 }, { "grad_norm": 0.12503841519355774, "learning_rate": 6.037563651387458e-05, "loss": 0.0049, "step": 46180 }, { "grad_norm": 0.10450530797243118, "learning_rate": 6.0359461170722666e-05, "loss": 0.0059, "step": 46190 }, { "grad_norm": 0.10154899954795837, "learning_rate": 6.034328469467566e-05, "loss": 0.0046, "step": 46200 }, { "grad_norm": 0.09695043414831161, "learning_rate": 6.0327107087502596e-05, "loss": 0.0055, "step": 46210 }, { "grad_norm": 0.1397063285112381, "learning_rate": 6.031092835097262e-05, "loss": 0.0058, "step": 46220 }, { "grad_norm": 0.127934068441391, "learning_rate": 6.0294748486855024e-05, "loss": 0.0044, "step": 46230 }, { "grad_norm": 0.09725523740053177, "learning_rate": 6.0278567496919216e-05, "loss": 0.0046, "step": 46240 }, { "grad_norm": 0.09302513301372528, "learning_rate": 6.026238538293472e-05, "loss": 0.0039, "step": 46250 }, { "grad_norm": 0.13232259452342987, "learning_rate": 6.024620214667118e-05, "loss": 0.0041, "step": 46260 }, { "grad_norm": 0.14105218648910522, "learning_rate": 6.0230017789898384e-05, "loss": 0.0051, "step": 46270 }, { "grad_norm": 0.11369029432535172, "learning_rate": 6.021383231438622e-05, "loss": 0.0057, "step": 46280 }, { "grad_norm": 0.1003694161772728, "learning_rate": 6.0197645721904704e-05, "loss": 0.0049, "step": 46290 }, { "grad_norm": 0.10761402547359467, "learning_rate": 6.0181458014224e-05, "loss": 0.0057, "step": 46300 }, { "grad_norm": 0.14610826969146729, "learning_rate": 6.016526919311435e-05, "loss": 0.0051, "step": 46310 }, { "grad_norm": 0.10561851412057877, "learning_rate": 6.014907926034613e-05, "loss": 0.0045, "step": 46320 }, { "grad_norm": 0.10656584799289703, "learning_rate": 6.0132888217689875e-05, "loss": 0.0046, "step": 46330 }, { "grad_norm": 0.0888557955622673, "learning_rate": 6.01166960669162e-05, "loss": 0.0062, "step": 46340 }, { "grad_norm": 0.09059838205575943, "learning_rate": 6.0100502809795845e-05, "loss": 0.005, "step": 46350 }, { "grad_norm": 0.089662566781044, "learning_rate": 6.00843084480997e-05, "loss": 0.0054, "step": 46360 }, { "grad_norm": 0.07834024727344513, "learning_rate": 6.0068112983598736e-05, "loss": 0.0059, "step": 46370 }, { "grad_norm": 0.14149422943592072, "learning_rate": 6.005191641806407e-05, "loss": 0.005, "step": 46380 }, { "grad_norm": 0.09661386907100677, "learning_rate": 6.003571875326694e-05, "loss": 0.0053, "step": 46390 }, { "grad_norm": 0.07071204483509064, "learning_rate": 6.00195199909787e-05, "loss": 0.0061, "step": 46400 }, { "grad_norm": 0.10625626891851425, "learning_rate": 6.000332013297082e-05, "loss": 0.0051, "step": 46410 }, { "grad_norm": 0.09657345712184906, "learning_rate": 5.998711918101487e-05, "loss": 0.0048, "step": 46420 }, { "grad_norm": 0.0826750248670578, "learning_rate": 5.997091713688261e-05, "loss": 0.005, "step": 46430 }, { "grad_norm": 0.08559463918209076, "learning_rate": 5.9954714002345836e-05, "loss": 0.0041, "step": 46440 }, { "grad_norm": 0.11785344779491425, "learning_rate": 5.993850977917649e-05, "loss": 0.0046, "step": 46450 }, { "grad_norm": 0.1108294129371643, "learning_rate": 5.992230446914667e-05, "loss": 0.0058, "step": 46460 }, { "grad_norm": 0.08642691373825073, "learning_rate": 5.990609807402855e-05, "loss": 0.0051, "step": 46470 }, { "grad_norm": 0.10797897726297379, "learning_rate": 5.988989059559443e-05, "loss": 0.0059, "step": 46480 }, { "grad_norm": 0.07747319340705872, "learning_rate": 5.987368203561675e-05, "loss": 0.0043, "step": 46490 }, { "grad_norm": 0.08177728205919266, "learning_rate": 5.9857472395868055e-05, "loss": 0.005, "step": 46500 }, { "grad_norm": 0.09625197947025299, "learning_rate": 5.9841261678120983e-05, "loss": 0.0052, "step": 46510 }, { "grad_norm": 0.11801842600107193, "learning_rate": 5.982504988414834e-05, "loss": 0.0055, "step": 46520 }, { "grad_norm": 0.10524942725896835, "learning_rate": 5.9808837015723015e-05, "loss": 0.0048, "step": 46530 }, { "grad_norm": 0.11124427616596222, "learning_rate": 5.9792623074618016e-05, "loss": 0.0047, "step": 46540 }, { "grad_norm": 0.10871023684740067, "learning_rate": 5.977640806260648e-05, "loss": 0.0044, "step": 46550 }, { "grad_norm": 0.07161051034927368, "learning_rate": 5.976019198146166e-05, "loss": 0.0052, "step": 46560 }, { "grad_norm": 0.09157445281744003, "learning_rate": 5.974397483295692e-05, "loss": 0.0042, "step": 46570 }, { "grad_norm": 0.17888720333576202, "learning_rate": 5.972775661886575e-05, "loss": 0.005, "step": 46580 }, { "grad_norm": 0.08529090136289597, "learning_rate": 5.971153734096173e-05, "loss": 0.0047, "step": 46590 }, { "grad_norm": 0.07699960470199585, "learning_rate": 5.969531700101859e-05, "loss": 0.004, "step": 46600 }, { "grad_norm": 0.08664224296808243, "learning_rate": 5.9679095600810155e-05, "loss": 0.004, "step": 46610 }, { "grad_norm": 0.07860337942838669, "learning_rate": 5.9662873142110384e-05, "loss": 0.0046, "step": 46620 }, { "grad_norm": 0.09537969529628754, "learning_rate": 5.964664962669333e-05, "loss": 0.0046, "step": 46630 }, { "grad_norm": 0.13183732330799103, "learning_rate": 5.9630425056333186e-05, "loss": 0.0056, "step": 46640 }, { "grad_norm": 0.07313346117734909, "learning_rate": 5.961419943280422e-05, "loss": 0.0045, "step": 46650 }, { "grad_norm": 0.09553627669811249, "learning_rate": 5.959797275788087e-05, "loss": 0.0061, "step": 46660 }, { "grad_norm": 0.10046239197254181, "learning_rate": 5.958174503333765e-05, "loss": 0.0055, "step": 46670 }, { "grad_norm": 0.09537644684314728, "learning_rate": 5.9565516260949195e-05, "loss": 0.0056, "step": 46680 }, { "grad_norm": 0.07353519648313522, "learning_rate": 5.9549286442490273e-05, "loss": 0.0038, "step": 46690 }, { "grad_norm": 0.10247659683227539, "learning_rate": 5.953305557973572e-05, "loss": 0.0043, "step": 46700 }, { "grad_norm": 0.10178708285093307, "learning_rate": 5.9516823674460564e-05, "loss": 0.0044, "step": 46710 }, { "grad_norm": 0.11127002537250519, "learning_rate": 5.9500590728439875e-05, "loss": 0.0057, "step": 46720 }, { "grad_norm": 0.10350055247545242, "learning_rate": 5.948435674344886e-05, "loss": 0.0049, "step": 46730 }, { "grad_norm": 0.09551341086626053, "learning_rate": 5.946812172126285e-05, "loss": 0.0055, "step": 46740 }, { "grad_norm": 0.07259585708379745, "learning_rate": 5.945188566365729e-05, "loss": 0.004, "step": 46750 }, { "grad_norm": 0.08992080390453339, "learning_rate": 5.943564857240773e-05, "loss": 0.004, "step": 46760 }, { "grad_norm": 0.05077878385782242, "learning_rate": 5.941941044928981e-05, "loss": 0.0041, "step": 46770 }, { "grad_norm": 0.08549591898918152, "learning_rate": 5.940317129607935e-05, "loss": 0.0049, "step": 46780 }, { "grad_norm": 0.09864175319671631, "learning_rate": 5.9386931114552204e-05, "loss": 0.0049, "step": 46790 }, { "grad_norm": 0.11289366334676743, "learning_rate": 5.937068990648438e-05, "loss": 0.0047, "step": 46800 }, { "grad_norm": 0.10399365425109863, "learning_rate": 5.935444767365199e-05, "loss": 0.0048, "step": 46810 }, { "grad_norm": 0.1150810644030571, "learning_rate": 5.933820441783129e-05, "loss": 0.0056, "step": 46820 }, { "grad_norm": 0.09073639661073685, "learning_rate": 5.932196014079857e-05, "loss": 0.0046, "step": 46830 }, { "grad_norm": 0.133762389421463, "learning_rate": 5.930571484433032e-05, "loss": 0.005, "step": 46840 }, { "grad_norm": 0.1123400554060936, "learning_rate": 5.928946853020309e-05, "loss": 0.0054, "step": 46850 }, { "grad_norm": 0.06814651936292648, "learning_rate": 5.927322120019352e-05, "loss": 0.0046, "step": 46860 }, { "grad_norm": 0.08475922048091888, "learning_rate": 5.925697285607844e-05, "loss": 0.0048, "step": 46870 }, { "grad_norm": 0.08785369247198105, "learning_rate": 5.924072349963472e-05, "loss": 0.0048, "step": 46880 }, { "grad_norm": 0.07188442349433899, "learning_rate": 5.922447313263938e-05, "loss": 0.0039, "step": 46890 }, { "grad_norm": 0.08529481291770935, "learning_rate": 5.920822175686952e-05, "loss": 0.0044, "step": 46900 }, { "grad_norm": 0.11465628445148468, "learning_rate": 5.919196937410239e-05, "loss": 0.004, "step": 46910 }, { "grad_norm": 0.13591796159744263, "learning_rate": 5.917571598611529e-05, "loss": 0.0051, "step": 46920 }, { "grad_norm": 0.1108570322394371, "learning_rate": 5.9159461594685686e-05, "loss": 0.0048, "step": 46930 }, { "grad_norm": 0.08827069401741028, "learning_rate": 5.914320620159114e-05, "loss": 0.0041, "step": 46940 }, { "grad_norm": 0.07920102030038834, "learning_rate": 5.912694980860931e-05, "loss": 0.0046, "step": 46950 }, { "grad_norm": 0.1045524924993515, "learning_rate": 5.9110692417517964e-05, "loss": 0.0059, "step": 46960 }, { "grad_norm": 0.09221614897251129, "learning_rate": 5.9094434030095e-05, "loss": 0.0057, "step": 46970 }, { "grad_norm": 0.10746967047452927, "learning_rate": 5.907817464811841e-05, "loss": 0.0044, "step": 46980 }, { "grad_norm": 0.0962408035993576, "learning_rate": 5.906191427336627e-05, "loss": 0.0055, "step": 46990 }, { "grad_norm": 0.10683713853359222, "learning_rate": 5.9045652907616835e-05, "loss": 0.0048, "step": 47000 }, { "grad_norm": 0.12036668509244919, "learning_rate": 5.902939055264838e-05, "loss": 0.0045, "step": 47010 }, { "grad_norm": 0.08691290020942688, "learning_rate": 5.901312721023935e-05, "loss": 0.0042, "step": 47020 }, { "grad_norm": 0.1206635981798172, "learning_rate": 5.8996862882168294e-05, "loss": 0.0069, "step": 47030 }, { "grad_norm": 0.11532527208328247, "learning_rate": 5.8980597570213826e-05, "loss": 0.0054, "step": 47040 }, { "grad_norm": 0.07978015393018723, "learning_rate": 5.896433127615471e-05, "loss": 0.0049, "step": 47050 }, { "grad_norm": 0.10170197486877441, "learning_rate": 5.894806400176981e-05, "loss": 0.0043, "step": 47060 }, { "grad_norm": 0.08857252448797226, "learning_rate": 5.893179574883808e-05, "loss": 0.0045, "step": 47070 }, { "grad_norm": 0.0977744534611702, "learning_rate": 5.8915526519138585e-05, "loss": 0.0048, "step": 47080 }, { "grad_norm": 0.11444619297981262, "learning_rate": 5.889925631445053e-05, "loss": 0.0056, "step": 47090 }, { "grad_norm": 0.12075652182102203, "learning_rate": 5.888298513655318e-05, "loss": 0.0046, "step": 47100 }, { "grad_norm": 0.10545247048139572, "learning_rate": 5.886671298722592e-05, "loss": 0.0048, "step": 47110 }, { "grad_norm": 0.1106729730963707, "learning_rate": 5.885043986824828e-05, "loss": 0.0042, "step": 47120 }, { "grad_norm": 0.09956379234790802, "learning_rate": 5.883416578139982e-05, "loss": 0.005, "step": 47130 }, { "grad_norm": 0.11979229748249054, "learning_rate": 5.881789072846029e-05, "loss": 0.0049, "step": 47140 }, { "grad_norm": 0.08549641817808151, "learning_rate": 5.8801614711209474e-05, "loss": 0.0045, "step": 47150 }, { "grad_norm": 0.09553147852420807, "learning_rate": 5.878533773142732e-05, "loss": 0.0056, "step": 47160 }, { "grad_norm": 0.0911291092634201, "learning_rate": 5.8769059790893834e-05, "loss": 0.0047, "step": 47170 }, { "grad_norm": 0.11602733284235, "learning_rate": 5.8752780891389145e-05, "loss": 0.0046, "step": 47180 }, { "grad_norm": 0.13730235397815704, "learning_rate": 5.873650103469351e-05, "loss": 0.0054, "step": 47190 }, { "grad_norm": 0.10660237073898315, "learning_rate": 5.872022022258725e-05, "loss": 0.0052, "step": 47200 }, { "grad_norm": 0.10547923296689987, "learning_rate": 5.8703938456850814e-05, "loss": 0.0042, "step": 47210 }, { "grad_norm": 0.0992659479379654, "learning_rate": 5.868765573926477e-05, "loss": 0.0045, "step": 47220 }, { "grad_norm": 0.10620652139186859, "learning_rate": 5.867137207160974e-05, "loss": 0.0055, "step": 47230 }, { "grad_norm": 0.1025053933262825, "learning_rate": 5.86550874556665e-05, "loss": 0.0052, "step": 47240 }, { "grad_norm": 0.0790366381406784, "learning_rate": 5.863880189321592e-05, "loss": 0.0058, "step": 47250 }, { "grad_norm": 0.13566266000270844, "learning_rate": 5.862251538603894e-05, "loss": 0.0046, "step": 47260 }, { "grad_norm": 0.08424380421638489, "learning_rate": 5.8606227935916656e-05, "loss": 0.005, "step": 47270 }, { "grad_norm": 0.08726783096790314, "learning_rate": 5.858993954463021e-05, "loss": 0.0042, "step": 47280 }, { "grad_norm": 0.10044378787279129, "learning_rate": 5.8573650213960896e-05, "loss": 0.004, "step": 47290 }, { "grad_norm": 0.10386170446872711, "learning_rate": 5.8557359945690084e-05, "loss": 0.0043, "step": 47300 }, { "grad_norm": 0.08076567947864532, "learning_rate": 5.8541068741599246e-05, "loss": 0.0045, "step": 47310 }, { "grad_norm": 0.07612980157136917, "learning_rate": 5.8524776603469985e-05, "loss": 0.0045, "step": 47320 }, { "grad_norm": 0.12026484310626984, "learning_rate": 5.850848353308397e-05, "loss": 0.0044, "step": 47330 }, { "grad_norm": 0.10282018035650253, "learning_rate": 5.849218953222297e-05, "loss": 0.0043, "step": 47340 }, { "grad_norm": 0.09389274567365646, "learning_rate": 5.847589460266891e-05, "loss": 0.0055, "step": 47350 }, { "grad_norm": 0.07457968592643738, "learning_rate": 5.845959874620375e-05, "loss": 0.0041, "step": 47360 }, { "grad_norm": 0.06821770966053009, "learning_rate": 5.84433019646096e-05, "loss": 0.0041, "step": 47370 }, { "grad_norm": 0.09034924954175949, "learning_rate": 5.842700425966863e-05, "loss": 0.0044, "step": 47380 }, { "grad_norm": 0.08789434283971786, "learning_rate": 5.841070563316315e-05, "loss": 0.0045, "step": 47390 }, { "grad_norm": 0.11393889039754868, "learning_rate": 5.839440608687554e-05, "loss": 0.0046, "step": 47400 }, { "grad_norm": 0.09666186571121216, "learning_rate": 5.837810562258831e-05, "loss": 0.0045, "step": 47410 }, { "grad_norm": 0.08185364305973053, "learning_rate": 5.836180424208405e-05, "loss": 0.0033, "step": 47420 }, { "grad_norm": 0.0986066460609436, "learning_rate": 5.834550194714543e-05, "loss": 0.0036, "step": 47430 }, { "grad_norm": 0.09471359848976135, "learning_rate": 5.832919873955528e-05, "loss": 0.0043, "step": 47440 }, { "grad_norm": 0.08914124220609665, "learning_rate": 5.831289462109648e-05, "loss": 0.0047, "step": 47450 }, { "grad_norm": 0.17138628661632538, "learning_rate": 5.829658959355202e-05, "loss": 0.005, "step": 47460 }, { "grad_norm": 0.09100646525621414, "learning_rate": 5.8280283658704995e-05, "loss": 0.005, "step": 47470 }, { "grad_norm": 0.10929979383945465, "learning_rate": 5.826397681833859e-05, "loss": 0.0044, "step": 47480 }, { "grad_norm": 0.12849529087543488, "learning_rate": 5.824766907423612e-05, "loss": 0.0045, "step": 47490 }, { "grad_norm": 0.09237167984247208, "learning_rate": 5.8231360428180945e-05, "loss": 0.0043, "step": 47500 }, { "grad_norm": 0.09463310241699219, "learning_rate": 5.821505088195658e-05, "loss": 0.0039, "step": 47510 }, { "grad_norm": 0.09839595854282379, "learning_rate": 5.819874043734661e-05, "loss": 0.0045, "step": 47520 }, { "grad_norm": 0.07233864068984985, "learning_rate": 5.8182429096134695e-05, "loss": 0.0038, "step": 47530 }, { "grad_norm": 0.09591328352689743, "learning_rate": 5.816611686010465e-05, "loss": 0.0046, "step": 47540 }, { "grad_norm": 0.11693355441093445, "learning_rate": 5.814980373104033e-05, "loss": 0.0048, "step": 47550 }, { "grad_norm": 0.09463168680667877, "learning_rate": 5.813348971072572e-05, "loss": 0.0039, "step": 47560 }, { "grad_norm": 0.09530717134475708, "learning_rate": 5.811717480094492e-05, "loss": 0.0045, "step": 47570 }, { "grad_norm": 0.10448838770389557, "learning_rate": 5.810085900348209e-05, "loss": 0.0053, "step": 47580 }, { "grad_norm": 0.08081821352243423, "learning_rate": 5.8084542320121483e-05, "loss": 0.0049, "step": 47590 }, { "grad_norm": 0.10238337516784668, "learning_rate": 5.8068224752647497e-05, "loss": 0.0045, "step": 47600 }, { "grad_norm": 0.10819300264120102, "learning_rate": 5.805190630284457e-05, "loss": 0.0045, "step": 47610 }, { "grad_norm": 0.10519833862781525, "learning_rate": 5.803558697249729e-05, "loss": 0.0049, "step": 47620 }, { "grad_norm": 0.09817441552877426, "learning_rate": 5.8019266763390276e-05, "loss": 0.0047, "step": 47630 }, { "grad_norm": 0.10845950990915298, "learning_rate": 5.800294567730833e-05, "loss": 0.0049, "step": 47640 }, { "grad_norm": 0.12436823546886444, "learning_rate": 5.798662371603628e-05, "loss": 0.0057, "step": 47650 }, { "grad_norm": 0.09107235074043274, "learning_rate": 5.797030088135904e-05, "loss": 0.0044, "step": 47660 }, { "grad_norm": 0.0685734897851944, "learning_rate": 5.7953977175061705e-05, "loss": 0.0041, "step": 47670 }, { "grad_norm": 0.08115196973085403, "learning_rate": 5.793765259892938e-05, "loss": 0.0043, "step": 47680 }, { "grad_norm": 0.10270629823207855, "learning_rate": 5.792132715474729e-05, "loss": 0.0052, "step": 47690 }, { "grad_norm": 0.0979040190577507, "learning_rate": 5.790500084430078e-05, "loss": 0.0047, "step": 47700 }, { "grad_norm": 0.10836992412805557, "learning_rate": 5.7888673669375274e-05, "loss": 0.0044, "step": 47710 }, { "grad_norm": 0.061773721128702164, "learning_rate": 5.787234563175625e-05, "loss": 0.0042, "step": 47720 }, { "grad_norm": 0.11166911572217941, "learning_rate": 5.7856016733229355e-05, "loss": 0.0046, "step": 47730 }, { "grad_norm": 0.07592913508415222, "learning_rate": 5.7839686975580297e-05, "loss": 0.0045, "step": 47740 }, { "grad_norm": 0.06954390555620193, "learning_rate": 5.782335636059484e-05, "loss": 0.0041, "step": 47750 }, { "grad_norm": 0.12492311000823975, "learning_rate": 5.780702489005889e-05, "loss": 0.0044, "step": 47760 }, { "grad_norm": 0.09971880167722702, "learning_rate": 5.779069256575845e-05, "loss": 0.0053, "step": 47770 }, { "grad_norm": 0.10145917534828186, "learning_rate": 5.7774359389479574e-05, "loss": 0.0045, "step": 47780 }, { "grad_norm": 0.09392907470464706, "learning_rate": 5.775802536300845e-05, "loss": 0.0039, "step": 47790 }, { "grad_norm": 0.09165144711732864, "learning_rate": 5.774169048813134e-05, "loss": 0.004, "step": 47800 }, { "grad_norm": 0.11157544702291489, "learning_rate": 5.77253547666346e-05, "loss": 0.0048, "step": 47810 }, { "grad_norm": 0.10965175181627274, "learning_rate": 5.770901820030465e-05, "loss": 0.0046, "step": 47820 }, { "grad_norm": 0.12378010153770447, "learning_rate": 5.769268079092809e-05, "loss": 0.0055, "step": 47830 }, { "grad_norm": 0.0915001854300499, "learning_rate": 5.767634254029151e-05, "loss": 0.0045, "step": 47840 }, { "grad_norm": 0.08894364535808563, "learning_rate": 5.7660003450181655e-05, "loss": 0.0041, "step": 47850 }, { "grad_norm": 0.08410205692052841, "learning_rate": 5.764366352238534e-05, "loss": 0.0041, "step": 47860 }, { "grad_norm": 0.10984165966510773, "learning_rate": 5.7627322758689474e-05, "loss": 0.0046, "step": 47870 }, { "grad_norm": 0.07939690351486206, "learning_rate": 5.761098116088105e-05, "loss": 0.0043, "step": 47880 }, { "grad_norm": 0.11772065609693527, "learning_rate": 5.759463873074717e-05, "loss": 0.0044, "step": 47890 }, { "grad_norm": 0.09344077110290527, "learning_rate": 5.757829547007504e-05, "loss": 0.0052, "step": 47900 }, { "grad_norm": 0.09489050507545471, "learning_rate": 5.756195138065189e-05, "loss": 0.0033, "step": 47910 }, { "grad_norm": 0.08711585402488708, "learning_rate": 5.754560646426511e-05, "loss": 0.0054, "step": 47920 }, { "grad_norm": 0.12669804692268372, "learning_rate": 5.752926072270216e-05, "loss": 0.0055, "step": 47930 }, { "grad_norm": 0.0744752287864685, "learning_rate": 5.7512914157750563e-05, "loss": 0.0038, "step": 47940 }, { "grad_norm": 0.09346362203359604, "learning_rate": 5.749656677119798e-05, "loss": 0.0046, "step": 47950 }, { "grad_norm": 0.0781714916229248, "learning_rate": 5.7480218564832125e-05, "loss": 0.0046, "step": 47960 }, { "grad_norm": 0.11112578213214874, "learning_rate": 5.746386954044082e-05, "loss": 0.0044, "step": 47970 }, { "grad_norm": 0.08262363076210022, "learning_rate": 5.744751969981195e-05, "loss": 0.0059, "step": 47980 }, { "grad_norm": 0.11669697612524033, "learning_rate": 5.7431169044733526e-05, "loss": 0.005, "step": 47990 }, { "grad_norm": 0.07778365910053253, "learning_rate": 5.741481757699364e-05, "loss": 0.0046, "step": 48000 }, { "grad_norm": 0.11618753522634506, "learning_rate": 5.7398465298380434e-05, "loss": 0.0043, "step": 48010 }, { "grad_norm": 0.10322991758584976, "learning_rate": 5.7382112210682193e-05, "loss": 0.004, "step": 48020 }, { "grad_norm": 0.12838755548000336, "learning_rate": 5.7365758315687266e-05, "loss": 0.0042, "step": 48030 }, { "grad_norm": 0.07452085614204407, "learning_rate": 5.734940361518407e-05, "loss": 0.0037, "step": 48040 }, { "grad_norm": 0.07548945397138596, "learning_rate": 5.733304811096116e-05, "loss": 0.0046, "step": 48050 }, { "grad_norm": 0.07351198047399521, "learning_rate": 5.731669180480713e-05, "loss": 0.0031, "step": 48060 }, { "grad_norm": 0.09100805968046188, "learning_rate": 5.730033469851067e-05, "loss": 0.0034, "step": 48070 }, { "grad_norm": 0.08874741196632385, "learning_rate": 5.7283976793860607e-05, "loss": 0.0036, "step": 48080 }, { "grad_norm": 0.08122395724058151, "learning_rate": 5.726761809264577e-05, "loss": 0.0046, "step": 48090 }, { "grad_norm": 0.09027031809091568, "learning_rate": 5.7251258596655155e-05, "loss": 0.0042, "step": 48100 }, { "grad_norm": 0.11930283904075623, "learning_rate": 5.72348983076778e-05, "loss": 0.0052, "step": 48110 }, { "grad_norm": 0.10378551483154297, "learning_rate": 5.7218537227502854e-05, "loss": 0.0047, "step": 48120 }, { "grad_norm": 0.09836720675230026, "learning_rate": 5.720217535791951e-05, "loss": 0.0044, "step": 48130 }, { "grad_norm": 0.090388223528862, "learning_rate": 5.718581270071711e-05, "loss": 0.0045, "step": 48140 }, { "grad_norm": 0.0924922376871109, "learning_rate": 5.716944925768505e-05, "loss": 0.0043, "step": 48150 }, { "grad_norm": 0.10356869548559189, "learning_rate": 5.7153085030612786e-05, "loss": 0.0043, "step": 48160 }, { "grad_norm": 0.09889519959688187, "learning_rate": 5.713672002128989e-05, "loss": 0.0044, "step": 48170 }, { "grad_norm": 0.08525450527667999, "learning_rate": 5.712035423150602e-05, "loss": 0.0036, "step": 48180 }, { "grad_norm": 0.08354917913675308, "learning_rate": 5.710398766305094e-05, "loss": 0.0047, "step": 48190 }, { "grad_norm": 0.09112069755792618, "learning_rate": 5.708762031771442e-05, "loss": 0.0039, "step": 48200 }, { "grad_norm": 0.08736158907413483, "learning_rate": 5.707125219728642e-05, "loss": 0.0053, "step": 48210 }, { "grad_norm": 0.09221621602773666, "learning_rate": 5.7054883303556905e-05, "loss": 0.0049, "step": 48220 }, { "grad_norm": 0.0889502614736557, "learning_rate": 5.703851363831595e-05, "loss": 0.0048, "step": 48230 }, { "grad_norm": 0.09272510558366776, "learning_rate": 5.702214320335374e-05, "loss": 0.0044, "step": 48240 }, { "grad_norm": 0.06744775176048279, "learning_rate": 5.7005772000460514e-05, "loss": 0.0039, "step": 48250 }, { "grad_norm": 0.1033308282494545, "learning_rate": 5.698940003142656e-05, "loss": 0.0044, "step": 48260 }, { "grad_norm": 0.08325274288654327, "learning_rate": 5.697302729804236e-05, "loss": 0.0046, "step": 48270 }, { "grad_norm": 0.09807175397872925, "learning_rate": 5.695665380209837e-05, "loss": 0.0047, "step": 48280 }, { "grad_norm": 0.13161949813365936, "learning_rate": 5.6940279545385165e-05, "loss": 0.0045, "step": 48290 }, { "grad_norm": 0.11375274509191513, "learning_rate": 5.692390452969344e-05, "loss": 0.0044, "step": 48300 }, { "grad_norm": 0.13967648148536682, "learning_rate": 5.690752875681392e-05, "loss": 0.0055, "step": 48310 }, { "grad_norm": 0.12960469722747803, "learning_rate": 5.6891152228537435e-05, "loss": 0.0058, "step": 48320 }, { "grad_norm": 0.07971630990505219, "learning_rate": 5.687477494665492e-05, "loss": 0.0038, "step": 48330 }, { "grad_norm": 0.10350116342306137, "learning_rate": 5.685839691295733e-05, "loss": 0.0041, "step": 48340 }, { "grad_norm": 0.08916611969470978, "learning_rate": 5.6842018129235786e-05, "loss": 0.0042, "step": 48350 }, { "grad_norm": 0.10773084312677383, "learning_rate": 5.6825638597281404e-05, "loss": 0.0047, "step": 48360 }, { "grad_norm": 0.09030091762542725, "learning_rate": 5.680925831888546e-05, "loss": 0.0045, "step": 48370 }, { "grad_norm": 0.11035096645355225, "learning_rate": 5.6792877295839274e-05, "loss": 0.0046, "step": 48380 }, { "grad_norm": 0.12361257523298264, "learning_rate": 5.6776495529934224e-05, "loss": 0.0044, "step": 48390 }, { "grad_norm": 0.08266907185316086, "learning_rate": 5.6760113022961824e-05, "loss": 0.0044, "step": 48400 }, { "grad_norm": 0.10329928994178772, "learning_rate": 5.6743729776713617e-05, "loss": 0.0055, "step": 48410 }, { "grad_norm": 0.12883687019348145, "learning_rate": 5.672734579298126e-05, "loss": 0.0059, "step": 48420 }, { "grad_norm": 0.08991061896085739, "learning_rate": 5.671096107355649e-05, "loss": 0.0044, "step": 48430 }, { "grad_norm": 0.11690416187047958, "learning_rate": 5.66945756202311e-05, "loss": 0.0048, "step": 48440 }, { "grad_norm": 0.08725129067897797, "learning_rate": 5.667818943479699e-05, "loss": 0.0051, "step": 48450 }, { "grad_norm": 0.10947445034980774, "learning_rate": 5.666180251904612e-05, "loss": 0.0044, "step": 48460 }, { "grad_norm": 0.10058600455522537, "learning_rate": 5.6645414874770555e-05, "loss": 0.0041, "step": 48470 }, { "grad_norm": 0.09766378253698349, "learning_rate": 5.66290265037624e-05, "loss": 0.0052, "step": 48480 }, { "grad_norm": 0.08996966481208801, "learning_rate": 5.661263740781386e-05, "loss": 0.0045, "step": 48490 }, { "grad_norm": 0.07778575271368027, "learning_rate": 5.6596247588717254e-05, "loss": 0.0051, "step": 48500 }, { "grad_norm": 0.0915386900305748, "learning_rate": 5.6579857048264926e-05, "loss": 0.0043, "step": 48510 }, { "grad_norm": 0.08101426064968109, "learning_rate": 5.6563465788249314e-05, "loss": 0.0051, "step": 48520 }, { "grad_norm": 0.07980421930551529, "learning_rate": 5.6547073810462956e-05, "loss": 0.0042, "step": 48530 }, { "grad_norm": 0.0709330216050148, "learning_rate": 5.653068111669846e-05, "loss": 0.0043, "step": 48540 }, { "grad_norm": 0.07499140501022339, "learning_rate": 5.651428770874848e-05, "loss": 0.0046, "step": 48550 }, { "grad_norm": 0.07981082051992416, "learning_rate": 5.64978935884058e-05, "loss": 0.0036, "step": 48560 }, { "grad_norm": 0.10341430455446243, "learning_rate": 5.6481498757463244e-05, "loss": 0.0039, "step": 48570 }, { "grad_norm": 0.11402929574251175, "learning_rate": 5.646510321771373e-05, "loss": 0.0045, "step": 48580 }, { "grad_norm": 0.09899064898490906, "learning_rate": 5.644870697095024e-05, "loss": 0.0045, "step": 48590 }, { "grad_norm": 0.08796464651823044, "learning_rate": 5.643231001896586e-05, "loss": 0.0045, "step": 48600 }, { "grad_norm": 0.10460737347602844, "learning_rate": 5.6415912363553726e-05, "loss": 0.0046, "step": 48610 }, { "grad_norm": 0.09238569438457489, "learning_rate": 5.639951400650706e-05, "loss": 0.0046, "step": 48620 }, { "grad_norm": 0.1317245364189148, "learning_rate": 5.6383114949619165e-05, "loss": 0.0056, "step": 48630 }, { "grad_norm": 0.11240151524543762, "learning_rate": 5.636671519468342e-05, "loss": 0.0059, "step": 48640 }, { "grad_norm": 0.10289905220270157, "learning_rate": 5.635031474349327e-05, "loss": 0.0046, "step": 48650 }, { "grad_norm": 0.10253742337226868, "learning_rate": 5.6333913597842246e-05, "loss": 0.0044, "step": 48660 }, { "grad_norm": 0.0854698196053505, "learning_rate": 5.6317511759523955e-05, "loss": 0.0044, "step": 48670 }, { "grad_norm": 0.08214321732521057, "learning_rate": 5.630110923033207e-05, "loss": 0.0052, "step": 48680 }, { "grad_norm": 0.11639022082090378, "learning_rate": 5.628470601206036e-05, "loss": 0.0049, "step": 48690 }, { "grad_norm": 0.09332343935966492, "learning_rate": 5.626830210650266e-05, "loss": 0.0043, "step": 48700 }, { "grad_norm": 0.0983528345823288, "learning_rate": 5.625189751545285e-05, "loss": 0.0042, "step": 48710 }, { "grad_norm": 0.1220572292804718, "learning_rate": 5.6235492240704936e-05, "loss": 0.0044, "step": 48720 }, { "grad_norm": 0.08701884746551514, "learning_rate": 5.621908628405296e-05, "loss": 0.0039, "step": 48730 }, { "grad_norm": 0.09067877382040024, "learning_rate": 5.620267964729106e-05, "loss": 0.0041, "step": 48740 }, { "grad_norm": 0.12030570209026337, "learning_rate": 5.618627233221344e-05, "loss": 0.0043, "step": 48750 }, { "grad_norm": 0.10305389761924744, "learning_rate": 5.616986434061438e-05, "loss": 0.0049, "step": 48760 }, { "grad_norm": 0.0856802687048912, "learning_rate": 5.615345567428822e-05, "loss": 0.0045, "step": 48770 }, { "grad_norm": 0.08937740325927734, "learning_rate": 5.613704633502941e-05, "loss": 0.0048, "step": 48780 }, { "grad_norm": 0.089472696185112, "learning_rate": 5.612063632463245e-05, "loss": 0.0041, "step": 48790 }, { "grad_norm": 0.11005991697311401, "learning_rate": 5.610422564489188e-05, "loss": 0.0047, "step": 48800 }, { "grad_norm": 0.09898482263088226, "learning_rate": 5.608781429760239e-05, "loss": 0.0042, "step": 48810 }, { "grad_norm": 0.09507321566343307, "learning_rate": 5.607140228455866e-05, "loss": 0.0037, "step": 48820 }, { "grad_norm": 0.07658050209283829, "learning_rate": 5.605498960755553e-05, "loss": 0.0045, "step": 48830 }, { "grad_norm": 0.0912744551897049, "learning_rate": 5.603857626838782e-05, "loss": 0.004, "step": 48840 }, { "grad_norm": 0.06519516557455063, "learning_rate": 5.60221622688505e-05, "loss": 0.0036, "step": 48850 }, { "grad_norm": 0.09322293847799301, "learning_rate": 5.6005747610738565e-05, "loss": 0.0055, "step": 48860 }, { "grad_norm": 0.0942782387137413, "learning_rate": 5.59893322958471e-05, "loss": 0.004, "step": 48870 }, { "grad_norm": 0.09065473824739456, "learning_rate": 5.5972916325971256e-05, "loss": 0.005, "step": 48880 }, { "grad_norm": 0.10826472193002701, "learning_rate": 5.595649970290628e-05, "loss": 0.0043, "step": 48890 }, { "grad_norm": 0.08446399867534637, "learning_rate": 5.5940082428447426e-05, "loss": 0.006, "step": 48900 }, { "grad_norm": 0.08231256157159805, "learning_rate": 5.592366450439012e-05, "loss": 0.0045, "step": 48910 }, { "grad_norm": 0.07439698278903961, "learning_rate": 5.590724593252975e-05, "loss": 0.0048, "step": 48920 }, { "grad_norm": 0.07609918713569641, "learning_rate": 5.589082671466184e-05, "loss": 0.0032, "step": 48930 }, { "grad_norm": 0.0936480388045311, "learning_rate": 5.587440685258199e-05, "loss": 0.0042, "step": 48940 }, { "grad_norm": 0.08408848196268082, "learning_rate": 5.585798634808583e-05, "loss": 0.0047, "step": 48950 }, { "grad_norm": 0.09861960262060165, "learning_rate": 5.584156520296909e-05, "loss": 0.0042, "step": 48960 }, { "grad_norm": 0.09050079435110092, "learning_rate": 5.582514341902757e-05, "loss": 0.0042, "step": 48970 }, { "grad_norm": 0.11762312054634094, "learning_rate": 5.580872099805713e-05, "loss": 0.0042, "step": 48980 }, { "grad_norm": 0.09833449870347977, "learning_rate": 5.5792297941853674e-05, "loss": 0.0045, "step": 48990 }, { "grad_norm": 0.09634390473365784, "learning_rate": 5.5775874252213247e-05, "loss": 0.0038, "step": 49000 }, { "grad_norm": 0.09656031429767609, "learning_rate": 5.575944993093189e-05, "loss": 0.0049, "step": 49010 }, { "grad_norm": 0.11711150407791138, "learning_rate": 5.574302497980574e-05, "loss": 0.0045, "step": 49020 }, { "grad_norm": 0.09333209693431854, "learning_rate": 5.5726599400631005e-05, "loss": 0.0044, "step": 49030 }, { "grad_norm": 0.09493009746074677, "learning_rate": 5.571017319520399e-05, "loss": 0.0038, "step": 49040 }, { "grad_norm": 0.09386540949344635, "learning_rate": 5.5693746365320984e-05, "loss": 0.0041, "step": 49050 }, { "grad_norm": 0.11484748125076294, "learning_rate": 5.567731891277848e-05, "loss": 0.0046, "step": 49060 }, { "grad_norm": 0.12457012385129929, "learning_rate": 5.566089083937288e-05, "loss": 0.0039, "step": 49070 }, { "grad_norm": 0.08319355547428131, "learning_rate": 5.564446214690079e-05, "loss": 0.0039, "step": 49080 }, { "grad_norm": 0.09591539949178696, "learning_rate": 5.56280328371588e-05, "loss": 0.0045, "step": 49090 }, { "grad_norm": 0.08989743888378143, "learning_rate": 5.5611602911943596e-05, "loss": 0.0043, "step": 49100 }, { "grad_norm": 0.08379019796848297, "learning_rate": 5.559517237305194e-05, "loss": 0.0044, "step": 49110 }, { "grad_norm": 0.08230851590633392, "learning_rate": 5.557874122228064e-05, "loss": 0.0038, "step": 49120 }, { "grad_norm": 0.10232709348201752, "learning_rate": 5.55623094614266e-05, "loss": 0.0046, "step": 49130 }, { "grad_norm": 0.08077527582645416, "learning_rate": 5.5545877092286743e-05, "loss": 0.0043, "step": 49140 }, { "grad_norm": 0.1012847051024437, "learning_rate": 5.552944411665811e-05, "loss": 0.0047, "step": 49150 }, { "grad_norm": 0.07782634347677231, "learning_rate": 5.5513010536337795e-05, "loss": 0.0039, "step": 49160 }, { "grad_norm": 0.0805051326751709, "learning_rate": 5.5496576353122934e-05, "loss": 0.0044, "step": 49170 }, { "grad_norm": 0.09550353139638901, "learning_rate": 5.548014156881074e-05, "loss": 0.0053, "step": 49180 }, { "grad_norm": 0.07840742915868759, "learning_rate": 5.5463706185198494e-05, "loss": 0.0038, "step": 49190 }, { "grad_norm": 0.11312393099069595, "learning_rate": 5.544727020408358e-05, "loss": 0.0049, "step": 49200 }, { "grad_norm": 0.09045196324586868, "learning_rate": 5.5430833627263366e-05, "loss": 0.0049, "step": 49210 }, { "grad_norm": 0.10728249698877335, "learning_rate": 5.5414396456535354e-05, "loss": 0.0044, "step": 49220 }, { "grad_norm": 0.1136610209941864, "learning_rate": 5.539795869369709e-05, "loss": 0.0036, "step": 49230 }, { "grad_norm": 0.07089407742023468, "learning_rate": 5.538152034054618e-05, "loss": 0.0043, "step": 49240 }, { "grad_norm": 0.07702299952507019, "learning_rate": 5.536508139888028e-05, "loss": 0.005, "step": 49250 }, { "grad_norm": 0.09546732902526855, "learning_rate": 5.534864187049716e-05, "loss": 0.0042, "step": 49260 }, { "grad_norm": 0.0794968530535698, "learning_rate": 5.5332201757194615e-05, "loss": 0.0052, "step": 49270 }, { "grad_norm": 0.07421883195638657, "learning_rate": 5.531576106077049e-05, "loss": 0.0042, "step": 49280 }, { "grad_norm": 0.07879418879747391, "learning_rate": 5.529931978302272e-05, "loss": 0.0042, "step": 49290 }, { "grad_norm": 0.09857694804668427, "learning_rate": 5.528287792574932e-05, "loss": 0.0043, "step": 49300 }, { "grad_norm": 0.088625468313694, "learning_rate": 5.5266435490748324e-05, "loss": 0.0048, "step": 49310 }, { "grad_norm": 0.08099699020385742, "learning_rate": 5.524999247981787e-05, "loss": 0.0035, "step": 49320 }, { "grad_norm": 0.10476887971162796, "learning_rate": 5.523354889475613e-05, "loss": 0.0053, "step": 49330 }, { "grad_norm": 0.10252406448125839, "learning_rate": 5.521710473736134e-05, "loss": 0.0048, "step": 49340 }, { "grad_norm": 0.0890474021434784, "learning_rate": 5.5200660009431826e-05, "loss": 0.0046, "step": 49350 }, { "grad_norm": 0.09629137068986893, "learning_rate": 5.518421471276596e-05, "loss": 0.0039, "step": 49360 }, { "grad_norm": 0.0710076242685318, "learning_rate": 5.5167768849162174e-05, "loss": 0.0045, "step": 49370 }, { "grad_norm": 0.11146054416894913, "learning_rate": 5.515132242041893e-05, "loss": 0.005, "step": 49380 }, { "grad_norm": 0.11189491301774979, "learning_rate": 5.513487542833483e-05, "loss": 0.0035, "step": 49390 }, { "grad_norm": 0.10572689026594162, "learning_rate": 5.5118427874708466e-05, "loss": 0.0043, "step": 49400 }, { "grad_norm": 0.08831118792295456, "learning_rate": 5.510197976133853e-05, "loss": 0.0041, "step": 49410 }, { "grad_norm": 0.07954033464193344, "learning_rate": 5.508553109002376e-05, "loss": 0.0047, "step": 49420 }, { "grad_norm": 0.07446097582578659, "learning_rate": 5.5069081862562957e-05, "loss": 0.0038, "step": 49430 }, { "grad_norm": 0.09940771013498306, "learning_rate": 5.5052632080754965e-05, "loss": 0.0046, "step": 49440 }, { "grad_norm": 0.08022072911262512, "learning_rate": 5.503618174639874e-05, "loss": 0.0042, "step": 49450 }, { "grad_norm": 0.09482819586992264, "learning_rate": 5.501973086129325e-05, "loss": 0.0048, "step": 49460 }, { "grad_norm": 0.07392160594463348, "learning_rate": 5.500327942723753e-05, "loss": 0.0043, "step": 49470 }, { "grad_norm": 0.09805719554424286, "learning_rate": 5.498682744603071e-05, "loss": 0.0041, "step": 49480 }, { "grad_norm": 0.08171055465936661, "learning_rate": 5.497037491947194e-05, "loss": 0.0048, "step": 49490 }, { "grad_norm": 0.07454060763120651, "learning_rate": 5.4953921849360424e-05, "loss": 0.0037, "step": 49500 }, { "grad_norm": 0.09767075628042221, "learning_rate": 5.493746823749547e-05, "loss": 0.0043, "step": 49510 }, { "grad_norm": 0.10609499365091324, "learning_rate": 5.49210140856764e-05, "loss": 0.0044, "step": 49520 }, { "grad_norm": 0.09241557866334915, "learning_rate": 5.4904559395702647e-05, "loss": 0.0041, "step": 49530 }, { "grad_norm": 0.0922064557671547, "learning_rate": 5.488810416937364e-05, "loss": 0.004, "step": 49540 }, { "grad_norm": 0.11352066695690155, "learning_rate": 5.4871648408488926e-05, "loss": 0.0037, "step": 49550 }, { "grad_norm": 0.09512735903263092, "learning_rate": 5.485519211484807e-05, "loss": 0.005, "step": 49560 }, { "grad_norm": 0.10055188089609146, "learning_rate": 5.4838735290250676e-05, "loss": 0.0043, "step": 49570 }, { "grad_norm": 0.10006804764270782, "learning_rate": 5.482227793649649e-05, "loss": 0.0051, "step": 49580 }, { "grad_norm": 0.10614455491304398, "learning_rate": 5.480582005538524e-05, "loss": 0.0039, "step": 49590 }, { "grad_norm": 0.08801618218421936, "learning_rate": 5.478936164871671e-05, "loss": 0.0053, "step": 49600 }, { "grad_norm": 0.08430946618318558, "learning_rate": 5.477290271829081e-05, "loss": 0.0041, "step": 49610 }, { "grad_norm": 0.09361475706100464, "learning_rate": 5.475644326590744e-05, "loss": 0.0065, "step": 49620 }, { "grad_norm": 0.14264467358589172, "learning_rate": 5.473998329336658e-05, "loss": 0.0046, "step": 49630 }, { "grad_norm": 0.08452237397432327, "learning_rate": 5.4723522802468286e-05, "loss": 0.004, "step": 49640 }, { "grad_norm": 0.12126228958368301, "learning_rate": 5.470706179501264e-05, "loss": 0.0049, "step": 49650 }, { "grad_norm": 0.12411905825138092, "learning_rate": 5.4690600272799776e-05, "loss": 0.0044, "step": 49660 }, { "grad_norm": 0.07512776553630829, "learning_rate": 5.467413823762993e-05, "loss": 0.0044, "step": 49670 }, { "grad_norm": 0.09417549520730972, "learning_rate": 5.465767569130335e-05, "loss": 0.0045, "step": 49680 }, { "grad_norm": 0.09971587359905243, "learning_rate": 5.464121263562036e-05, "loss": 0.0046, "step": 49690 }, { "grad_norm": 0.10674320161342621, "learning_rate": 5.4624749072381343e-05, "loss": 0.005, "step": 49700 }, { "grad_norm": 0.12663301825523376, "learning_rate": 5.460828500338672e-05, "loss": 0.0047, "step": 49710 }, { "grad_norm": 0.11532401293516159, "learning_rate": 5.459182043043698e-05, "loss": 0.0044, "step": 49720 }, { "grad_norm": 0.0973854809999466, "learning_rate": 5.457535535533265e-05, "loss": 0.0054, "step": 49730 }, { "grad_norm": 0.11820437759160995, "learning_rate": 5.4558889779874334e-05, "loss": 0.0039, "step": 49740 }, { "grad_norm": 0.12563450634479523, "learning_rate": 5.454242370586269e-05, "loss": 0.0053, "step": 49750 }, { "grad_norm": 0.10659889876842499, "learning_rate": 5.452595713509843e-05, "loss": 0.0044, "step": 49760 }, { "grad_norm": 0.08129243552684784, "learning_rate": 5.4509490069382274e-05, "loss": 0.0057, "step": 49770 }, { "grad_norm": 0.09120311588048935, "learning_rate": 5.449302251051509e-05, "loss": 0.0046, "step": 49780 }, { "grad_norm": 0.09005583077669144, "learning_rate": 5.44765544602977e-05, "loss": 0.0047, "step": 49790 }, { "grad_norm": 0.06002390384674072, "learning_rate": 5.446008592053107e-05, "loss": 0.004, "step": 49800 }, { "grad_norm": 0.09209010004997253, "learning_rate": 5.4443616893016135e-05, "loss": 0.0066, "step": 49810 }, { "grad_norm": 0.08959796279668808, "learning_rate": 5.4427147379553934e-05, "loss": 0.0043, "step": 49820 }, { "grad_norm": 0.10214847326278687, "learning_rate": 5.441067738194556e-05, "loss": 0.0044, "step": 49830 }, { "grad_norm": 0.10164496302604675, "learning_rate": 5.439420690199214e-05, "loss": 0.0045, "step": 49840 }, { "grad_norm": 0.08833476155996323, "learning_rate": 5.437773594149484e-05, "loss": 0.0048, "step": 49850 }, { "grad_norm": 0.09794768691062927, "learning_rate": 5.436126450225495e-05, "loss": 0.0049, "step": 49860 }, { "grad_norm": 0.09254036843776703, "learning_rate": 5.434479258607371e-05, "loss": 0.0046, "step": 49870 }, { "grad_norm": 0.11098392307758331, "learning_rate": 5.43283201947525e-05, "loss": 0.0041, "step": 49880 }, { "grad_norm": 0.11074572801589966, "learning_rate": 5.431184733009268e-05, "loss": 0.0046, "step": 49890 }, { "grad_norm": 0.115729920566082, "learning_rate": 5.4295373993895736e-05, "loss": 0.0049, "step": 49900 }, { "grad_norm": 0.08333197981119156, "learning_rate": 5.4278900187963157e-05, "loss": 0.0052, "step": 49910 }, { "grad_norm": 0.10122589766979218, "learning_rate": 5.426242591409646e-05, "loss": 0.0047, "step": 49920 }, { "grad_norm": 0.09121587872505188, "learning_rate": 5.42459511740973e-05, "loss": 0.0044, "step": 49930 }, { "grad_norm": 0.08261411637067795, "learning_rate": 5.422947596976729e-05, "loss": 0.0039, "step": 49940 }, { "grad_norm": 0.08403968065977097, "learning_rate": 5.4213000302908134e-05, "loss": 0.0041, "step": 49950 }, { "grad_norm": 0.08552147448062897, "learning_rate": 5.419652417532162e-05, "loss": 0.0039, "step": 49960 }, { "grad_norm": 0.10245989263057709, "learning_rate": 5.4180047588809534e-05, "loss": 0.004, "step": 49970 }, { "grad_norm": 0.10728245973587036, "learning_rate": 5.4163570545173704e-05, "loss": 0.0045, "step": 49980 }, { "grad_norm": 0.09799855202436447, "learning_rate": 5.414709304621608e-05, "loss": 0.0042, "step": 49990 }, { "grad_norm": 0.12190021574497223, "learning_rate": 5.413061509373858e-05, "loss": 0.0044, "step": 50000 }, { "grad_norm": 0.07516579329967499, "learning_rate": 5.411413668954324e-05, "loss": 0.0035, "step": 50010 }, { "grad_norm": 0.1250249445438385, "learning_rate": 5.4097657835432083e-05, "loss": 0.0043, "step": 50020 }, { "grad_norm": 0.08242499828338623, "learning_rate": 5.408117853320723e-05, "loss": 0.0041, "step": 50030 }, { "grad_norm": 0.08889293670654297, "learning_rate": 5.406469878467084e-05, "loss": 0.0041, "step": 50040 }, { "grad_norm": 0.0870271846652031, "learning_rate": 5.404821859162509e-05, "loss": 0.004, "step": 50050 }, { "grad_norm": 0.09734456986188889, "learning_rate": 5.403173795587225e-05, "loss": 0.004, "step": 50060 }, { "grad_norm": 0.09439247101545334, "learning_rate": 5.4015256879214606e-05, "loss": 0.0038, "step": 50070 }, { "grad_norm": 0.0882260724902153, "learning_rate": 5.39987753634545e-05, "loss": 0.0045, "step": 50080 }, { "grad_norm": 0.0882822647690773, "learning_rate": 5.3982293410394335e-05, "loss": 0.0043, "step": 50090 }, { "grad_norm": 0.10743638128042221, "learning_rate": 5.396581102183655e-05, "loss": 0.0045, "step": 50100 }, { "grad_norm": 0.11993584781885147, "learning_rate": 5.3949328199583634e-05, "loss": 0.0062, "step": 50110 }, { "grad_norm": 0.09111116826534271, "learning_rate": 5.393284494543813e-05, "loss": 0.0041, "step": 50120 }, { "grad_norm": 0.07807204127311707, "learning_rate": 5.391636126120262e-05, "loss": 0.0041, "step": 50130 }, { "grad_norm": 0.09108944237232208, "learning_rate": 5.389987714867971e-05, "loss": 0.0047, "step": 50140 }, { "grad_norm": 0.1068587526679039, "learning_rate": 5.3883392609672114e-05, "loss": 0.0042, "step": 50150 }, { "grad_norm": 0.07968319952487946, "learning_rate": 5.3866907645982545e-05, "loss": 0.0042, "step": 50160 }, { "grad_norm": 0.06670267879962921, "learning_rate": 5.385042225941375e-05, "loss": 0.0037, "step": 50170 }, { "grad_norm": 0.06907769292593002, "learning_rate": 5.3833936451768585e-05, "loss": 0.0035, "step": 50180 }, { "grad_norm": 0.10038629919290543, "learning_rate": 5.3817450224849884e-05, "loss": 0.0035, "step": 50190 }, { "grad_norm": 0.09696465730667114, "learning_rate": 5.380096358046056e-05, "loss": 0.0039, "step": 50200 }, { "grad_norm": 0.0997084304690361, "learning_rate": 5.378447652040359e-05, "loss": 0.0049, "step": 50210 }, { "grad_norm": 0.0885465145111084, "learning_rate": 5.376798904648195e-05, "loss": 0.0045, "step": 50220 }, { "grad_norm": 0.1319715827703476, "learning_rate": 5.375150116049869e-05, "loss": 0.0052, "step": 50230 }, { "grad_norm": 0.12637147307395935, "learning_rate": 5.373501286425691e-05, "loss": 0.0051, "step": 50240 }, { "grad_norm": 0.11745621263980865, "learning_rate": 5.3718524159559726e-05, "loss": 0.004, "step": 50250 }, { "grad_norm": 0.08742348104715347, "learning_rate": 5.370203504821034e-05, "loss": 0.0046, "step": 50260 }, { "grad_norm": 0.08648231625556946, "learning_rate": 5.368554553201196e-05, "loss": 0.0045, "step": 50270 }, { "grad_norm": 0.12243445962667465, "learning_rate": 5.3669055612767874e-05, "loss": 0.0046, "step": 50280 }, { "grad_norm": 0.09907793998718262, "learning_rate": 5.3652565292281376e-05, "loss": 0.0046, "step": 50290 }, { "grad_norm": 0.09686397761106491, "learning_rate": 5.3636074572355824e-05, "loss": 0.0057, "step": 50300 }, { "grad_norm": 0.07190842926502228, "learning_rate": 5.361958345479463e-05, "loss": 0.0046, "step": 50310 }, { "grad_norm": 0.07819197326898575, "learning_rate": 5.360309194140123e-05, "loss": 0.004, "step": 50320 }, { "grad_norm": 0.09055715054273605, "learning_rate": 5.3586600033979106e-05, "loss": 0.0034, "step": 50330 }, { "grad_norm": 0.09543462842702866, "learning_rate": 5.357010773433181e-05, "loss": 0.0034, "step": 50340 }, { "grad_norm": 0.09314721077680588, "learning_rate": 5.3553615044262895e-05, "loss": 0.0047, "step": 50350 }, { "grad_norm": 0.08882683515548706, "learning_rate": 5.3537121965575964e-05, "loss": 0.0039, "step": 50360 }, { "grad_norm": 0.11077846586704254, "learning_rate": 5.352062850007471e-05, "loss": 0.004, "step": 50370 }, { "grad_norm": 0.09873747080564499, "learning_rate": 5.350413464956282e-05, "loss": 0.0044, "step": 50380 }, { "grad_norm": 0.10959358513355255, "learning_rate": 5.348764041584403e-05, "loss": 0.0051, "step": 50390 }, { "grad_norm": 0.11587606370449066, "learning_rate": 5.347114580072212e-05, "loss": 0.005, "step": 50400 }, { "grad_norm": 0.09225188940763474, "learning_rate": 5.345465080600093e-05, "loss": 0.0041, "step": 50410 }, { "grad_norm": 0.0784643366932869, "learning_rate": 5.3438155433484326e-05, "loss": 0.0038, "step": 50420 }, { "grad_norm": 0.08044765144586563, "learning_rate": 5.3421659684976197e-05, "loss": 0.0038, "step": 50430 }, { "grad_norm": 0.09132184833288193, "learning_rate": 5.340516356228052e-05, "loss": 0.0042, "step": 50440 }, { "grad_norm": 0.10862809419631958, "learning_rate": 5.338866706720128e-05, "loss": 0.0038, "step": 50450 }, { "grad_norm": 0.10340196639299393, "learning_rate": 5.337217020154249e-05, "loss": 0.0044, "step": 50460 }, { "grad_norm": 0.11624950915575027, "learning_rate": 5.335567296710825e-05, "loss": 0.0046, "step": 50470 }, { "grad_norm": 0.10899168998003006, "learning_rate": 5.333917536570265e-05, "loss": 0.0051, "step": 50480 }, { "grad_norm": 0.12634065747261047, "learning_rate": 5.332267739912986e-05, "loss": 0.0042, "step": 50490 }, { "grad_norm": 0.10811767727136612, "learning_rate": 5.330617906919405e-05, "loss": 0.0044, "step": 50500 }, { "grad_norm": 0.09542031586170197, "learning_rate": 5.328968037769949e-05, "loss": 0.0045, "step": 50510 }, { "grad_norm": 0.10378566384315491, "learning_rate": 5.327318132645042e-05, "loss": 0.0038, "step": 50520 }, { "grad_norm": 0.09804895520210266, "learning_rate": 5.325668191725116e-05, "loss": 0.0042, "step": 50530 }, { "grad_norm": 0.10075968503952026, "learning_rate": 5.324018215190606e-05, "loss": 0.0048, "step": 50540 }, { "grad_norm": 0.1077665314078331, "learning_rate": 5.3223682032219515e-05, "loss": 0.0041, "step": 50550 }, { "grad_norm": 0.09496268630027771, "learning_rate": 5.320718155999595e-05, "loss": 0.0039, "step": 50560 }, { "grad_norm": 0.09019001573324203, "learning_rate": 5.3190680737039835e-05, "loss": 0.0037, "step": 50570 }, { "grad_norm": 0.07682912796735764, "learning_rate": 5.317417956515567e-05, "loss": 0.0037, "step": 50580 }, { "grad_norm": 0.0685095489025116, "learning_rate": 5.315767804614798e-05, "loss": 0.0044, "step": 50590 }, { "grad_norm": 0.10092001408338547, "learning_rate": 5.3141176181821395e-05, "loss": 0.0045, "step": 50600 }, { "grad_norm": 0.0881929025053978, "learning_rate": 5.31246739739805e-05, "loss": 0.0048, "step": 50610 }, { "grad_norm": 0.08191286772489548, "learning_rate": 5.310817142442994e-05, "loss": 0.0038, "step": 50620 }, { "grad_norm": 0.10279025137424469, "learning_rate": 5.309166853497445e-05, "loss": 0.0041, "step": 50630 }, { "grad_norm": 0.10232510417699814, "learning_rate": 5.307516530741873e-05, "loss": 0.0045, "step": 50640 }, { "grad_norm": 0.11502587050199509, "learning_rate": 5.305866174356754e-05, "loss": 0.0049, "step": 50650 }, { "grad_norm": 0.11593594402074814, "learning_rate": 5.304215784522571e-05, "loss": 0.005, "step": 50660 }, { "grad_norm": 0.10273311287164688, "learning_rate": 5.302565361419808e-05, "loss": 0.0043, "step": 50670 }, { "grad_norm": 0.07390718907117844, "learning_rate": 5.3009149052289507e-05, "loss": 0.004, "step": 50680 }, { "grad_norm": 0.090695820748806, "learning_rate": 5.299264416130493e-05, "loss": 0.0039, "step": 50690 }, { "grad_norm": 0.09771499037742615, "learning_rate": 5.297613894304928e-05, "loss": 0.0036, "step": 50700 }, { "grad_norm": 0.07897303253412247, "learning_rate": 5.2959633399327534e-05, "loss": 0.0038, "step": 50710 }, { "grad_norm": 0.06735952198505402, "learning_rate": 5.294312753194476e-05, "loss": 0.0036, "step": 50720 }, { "grad_norm": 0.1114363819360733, "learning_rate": 5.292662134270596e-05, "loss": 0.0042, "step": 50730 }, { "grad_norm": 0.09011188894510269, "learning_rate": 5.291011483341626e-05, "loss": 0.004, "step": 50740 }, { "grad_norm": 0.09059348702430725, "learning_rate": 5.2893608005880767e-05, "loss": 0.0036, "step": 50750 }, { "grad_norm": 0.08748103678226471, "learning_rate": 5.287710086190467e-05, "loss": 0.0035, "step": 50760 }, { "grad_norm": 0.0908328965306282, "learning_rate": 5.286059340329314e-05, "loss": 0.0037, "step": 50770 }, { "grad_norm": 0.07388483732938766, "learning_rate": 5.284408563185141e-05, "loss": 0.0039, "step": 50780 }, { "grad_norm": 0.08474207669496536, "learning_rate": 5.282757754938476e-05, "loss": 0.0042, "step": 50790 }, { "grad_norm": 0.0707804262638092, "learning_rate": 5.281106915769849e-05, "loss": 0.0039, "step": 50800 }, { "grad_norm": 0.09401627629995346, "learning_rate": 5.2794560458597897e-05, "loss": 0.0048, "step": 50810 }, { "grad_norm": 0.07874745875597, "learning_rate": 5.27780514538884e-05, "loss": 0.0034, "step": 50820 }, { "grad_norm": 0.09351885318756104, "learning_rate": 5.2761542145375365e-05, "loss": 0.0052, "step": 50830 }, { "grad_norm": 0.08953789621591568, "learning_rate": 5.274503253486421e-05, "loss": 0.0037, "step": 50840 }, { "grad_norm": 0.07319485396146774, "learning_rate": 5.272852262416046e-05, "loss": 0.0046, "step": 50850 }, { "grad_norm": 0.11065840721130371, "learning_rate": 5.2712012415069555e-05, "loss": 0.0057, "step": 50860 }, { "grad_norm": 0.07648120075464249, "learning_rate": 5.269550190939705e-05, "loss": 0.0038, "step": 50870 }, { "grad_norm": 0.08709119260311127, "learning_rate": 5.267899110894852e-05, "loss": 0.0049, "step": 50880 }, { "grad_norm": 0.1080760508775711, "learning_rate": 5.266248001552955e-05, "loss": 0.0045, "step": 50890 }, { "grad_norm": 0.09300179034471512, "learning_rate": 5.264596863094575e-05, "loss": 0.0053, "step": 50900 }, { "grad_norm": 0.09905292838811874, "learning_rate": 5.262945695700282e-05, "loss": 0.004, "step": 50910 }, { "grad_norm": 0.09967433661222458, "learning_rate": 5.261294499550643e-05, "loss": 0.0063, "step": 50920 }, { "grad_norm": 0.08506450802087784, "learning_rate": 5.25964327482623e-05, "loss": 0.0053, "step": 50930 }, { "grad_norm": 0.1041753962635994, "learning_rate": 5.257992021707617e-05, "loss": 0.0053, "step": 50940 }, { "grad_norm": 0.09800084680318832, "learning_rate": 5.256340740375387e-05, "loss": 0.0044, "step": 50950 }, { "grad_norm": 0.09692689776420593, "learning_rate": 5.254689431010117e-05, "loss": 0.0042, "step": 50960 }, { "grad_norm": 0.07375089824199677, "learning_rate": 5.253038093792395e-05, "loss": 0.0048, "step": 50970 }, { "grad_norm": 0.08327540010213852, "learning_rate": 5.251386728902806e-05, "loss": 0.0038, "step": 50980 }, { "grad_norm": 0.10461066663265228, "learning_rate": 5.2497353365219446e-05, "loss": 0.0047, "step": 50990 }, { "grad_norm": 0.11008426547050476, "learning_rate": 5.2480839168304e-05, "loss": 0.0044, "step": 51000 }, { "grad_norm": 0.09280865639448166, "learning_rate": 5.246432470008773e-05, "loss": 0.0044, "step": 51010 }, { "grad_norm": 0.0835452452301979, "learning_rate": 5.2447809962376616e-05, "loss": 0.0063, "step": 51020 }, { "grad_norm": 0.08473426103591919, "learning_rate": 5.243129495697666e-05, "loss": 0.0038, "step": 51030 }, { "grad_norm": 0.09649400413036346, "learning_rate": 5.241477968569397e-05, "loss": 0.0044, "step": 51040 }, { "grad_norm": 0.10769294202327728, "learning_rate": 5.23982641503346e-05, "loss": 0.0059, "step": 51050 }, { "grad_norm": 0.1037197858095169, "learning_rate": 5.238174835270465e-05, "loss": 0.0053, "step": 51060 }, { "grad_norm": 0.13211078941822052, "learning_rate": 5.23652322946103e-05, "loss": 0.0048, "step": 51070 }, { "grad_norm": 0.09075910598039627, "learning_rate": 5.234871597785771e-05, "loss": 0.0047, "step": 51080 }, { "grad_norm": 0.09388072788715363, "learning_rate": 5.233219940425306e-05, "loss": 0.0042, "step": 51090 }, { "grad_norm": 0.11061397939920425, "learning_rate": 5.2315682575602576e-05, "loss": 0.0041, "step": 51100 }, { "grad_norm": 0.08987300097942352, "learning_rate": 5.229916549371254e-05, "loss": 0.0039, "step": 51110 }, { "grad_norm": 0.08380507677793503, "learning_rate": 5.228264816038924e-05, "loss": 0.0035, "step": 51120 }, { "grad_norm": 0.08195719867944717, "learning_rate": 5.2266130577438935e-05, "loss": 0.0038, "step": 51130 }, { "grad_norm": 0.07802686840295792, "learning_rate": 5.224961274666802e-05, "loss": 0.0036, "step": 51140 }, { "grad_norm": 0.06507889181375504, "learning_rate": 5.223309466988283e-05, "loss": 0.0038, "step": 51150 }, { "grad_norm": 0.08098874241113663, "learning_rate": 5.221657634888976e-05, "loss": 0.0039, "step": 51160 }, { "grad_norm": 0.05256766825914383, "learning_rate": 5.2200057785495235e-05, "loss": 0.0044, "step": 51170 }, { "grad_norm": 0.1001376062631607, "learning_rate": 5.2183538981505706e-05, "loss": 0.0036, "step": 51180 }, { "grad_norm": 0.0979323461651802, "learning_rate": 5.216701993872762e-05, "loss": 0.0035, "step": 51190 }, { "grad_norm": 0.10700681805610657, "learning_rate": 5.215050065896752e-05, "loss": 0.0043, "step": 51200 }, { "grad_norm": 0.11061102151870728, "learning_rate": 5.2133981144031875e-05, "loss": 0.0053, "step": 51210 }, { "grad_norm": 0.14792965352535248, "learning_rate": 5.211746139572727e-05, "loss": 0.0043, "step": 51220 }, { "grad_norm": 0.10855967551469803, "learning_rate": 5.210094141586026e-05, "loss": 0.0045, "step": 51230 }, { "grad_norm": 0.0819670632481575, "learning_rate": 5.2084421206237476e-05, "loss": 0.0043, "step": 51240 }, { "grad_norm": 0.08077860623598099, "learning_rate": 5.20679007686655e-05, "loss": 0.0053, "step": 51250 }, { "grad_norm": 0.09169000387191772, "learning_rate": 5.205138010495102e-05, "loss": 0.0051, "step": 51260 }, { "grad_norm": 0.10069452971220016, "learning_rate": 5.20348592169007e-05, "loss": 0.0047, "step": 51270 }, { "grad_norm": 0.10824963450431824, "learning_rate": 5.201833810632123e-05, "loss": 0.0047, "step": 51280 }, { "grad_norm": 0.10214352607727051, "learning_rate": 5.200181677501933e-05, "loss": 0.0043, "step": 51290 }, { "grad_norm": 0.1208028495311737, "learning_rate": 5.1985295224801777e-05, "loss": 0.0053, "step": 51300 }, { "grad_norm": 0.1006908193230629, "learning_rate": 5.196877345747531e-05, "loss": 0.0055, "step": 51310 }, { "grad_norm": 0.10309431701898575, "learning_rate": 5.195225147484674e-05, "loss": 0.0045, "step": 51320 }, { "grad_norm": 0.09739576280117035, "learning_rate": 5.193572927872288e-05, "loss": 0.0049, "step": 51330 }, { "grad_norm": 0.10177081823348999, "learning_rate": 5.191920687091059e-05, "loss": 0.0038, "step": 51340 }, { "grad_norm": 0.07239563763141632, "learning_rate": 5.1902684253216695e-05, "loss": 0.004, "step": 51350 }, { "grad_norm": 0.07492624968290329, "learning_rate": 5.1886161427448134e-05, "loss": 0.0034, "step": 51360 }, { "grad_norm": 0.07504897564649582, "learning_rate": 5.186963839541179e-05, "loss": 0.0046, "step": 51370 }, { "grad_norm": 0.0989358201622963, "learning_rate": 5.185311515891459e-05, "loss": 0.0044, "step": 51380 }, { "grad_norm": 0.07724564522504807, "learning_rate": 5.1836591719763516e-05, "loss": 0.0044, "step": 51390 }, { "grad_norm": 0.11207042634487152, "learning_rate": 5.182006807976554e-05, "loss": 0.0044, "step": 51400 }, { "grad_norm": 0.09797264635562897, "learning_rate": 5.180354424072763e-05, "loss": 0.0039, "step": 51410 }, { "grad_norm": 0.09334935247898102, "learning_rate": 5.178702020445686e-05, "loss": 0.0035, "step": 51420 }, { "grad_norm": 0.07425490766763687, "learning_rate": 5.177049597276024e-05, "loss": 0.004, "step": 51430 }, { "grad_norm": 0.10077635943889618, "learning_rate": 5.175397154744482e-05, "loss": 0.0043, "step": 51440 }, { "grad_norm": 0.06748794764280319, "learning_rate": 5.1737446930317736e-05, "loss": 0.0035, "step": 51450 }, { "grad_norm": 0.09992999583482742, "learning_rate": 5.172092212318604e-05, "loss": 0.0048, "step": 51460 }, { "grad_norm": 0.09028635919094086, "learning_rate": 5.170439712785692e-05, "loss": 0.004, "step": 51470 }, { "grad_norm": 0.09055919200181961, "learning_rate": 5.1687871946137467e-05, "loss": 0.0048, "step": 51480 }, { "grad_norm": 0.08113108575344086, "learning_rate": 5.16713465798349e-05, "loss": 0.0041, "step": 51490 }, { "grad_norm": 0.0770225003361702, "learning_rate": 5.165482103075637e-05, "loss": 0.0038, "step": 51500 }, { "grad_norm": 0.08221600949764252, "learning_rate": 5.163829530070909e-05, "loss": 0.0039, "step": 51510 }, { "grad_norm": 0.07711181044578552, "learning_rate": 5.162176939150032e-05, "loss": 0.0041, "step": 51520 }, { "grad_norm": 0.11195646226406097, "learning_rate": 5.160524330493728e-05, "loss": 0.0055, "step": 51530 }, { "grad_norm": 0.1038472130894661, "learning_rate": 5.158871704282724e-05, "loss": 0.0047, "step": 51540 }, { "grad_norm": 0.09095794707536697, "learning_rate": 5.15721906069775e-05, "loss": 0.005, "step": 51550 }, { "grad_norm": 0.08678235113620758, "learning_rate": 5.155566399919536e-05, "loss": 0.0039, "step": 51560 }, { "grad_norm": 0.06796052306890488, "learning_rate": 5.153913722128813e-05, "loss": 0.0037, "step": 51570 }, { "grad_norm": 0.08475019037723541, "learning_rate": 5.1522610275063176e-05, "loss": 0.0045, "step": 51580 }, { "grad_norm": 0.06781081855297089, "learning_rate": 5.150608316232786e-05, "loss": 0.0039, "step": 51590 }, { "grad_norm": 0.09588771313428879, "learning_rate": 5.148955588488955e-05, "loss": 0.0053, "step": 51600 }, { "grad_norm": 0.07309585064649582, "learning_rate": 5.147302844455564e-05, "loss": 0.0041, "step": 51610 }, { "grad_norm": 0.07339229434728622, "learning_rate": 5.145650084313356e-05, "loss": 0.0037, "step": 51620 }, { "grad_norm": 0.09006980806589127, "learning_rate": 5.1439973082430734e-05, "loss": 0.0044, "step": 51630 }, { "grad_norm": 0.09871643781661987, "learning_rate": 5.142344516425461e-05, "loss": 0.0044, "step": 51640 }, { "grad_norm": 0.08798862993717194, "learning_rate": 5.1406917090412673e-05, "loss": 0.0042, "step": 51650 }, { "grad_norm": 0.08398545533418655, "learning_rate": 5.1390388862712404e-05, "loss": 0.0039, "step": 51660 }, { "grad_norm": 0.08668967336416245, "learning_rate": 5.137386048296129e-05, "loss": 0.0049, "step": 51670 }, { "grad_norm": 0.08966856449842453, "learning_rate": 5.135733195296686e-05, "loss": 0.004, "step": 51680 }, { "grad_norm": 0.07678835093975067, "learning_rate": 5.134080327453664e-05, "loss": 0.0041, "step": 51690 }, { "grad_norm": 0.07342532277107239, "learning_rate": 5.1324274449478204e-05, "loss": 0.0038, "step": 51700 }, { "grad_norm": 0.11111447960138321, "learning_rate": 5.130774547959909e-05, "loss": 0.0041, "step": 51710 }, { "grad_norm": 0.13442447781562805, "learning_rate": 5.129121636670692e-05, "loss": 0.0041, "step": 51720 }, { "grad_norm": 0.07755539566278458, "learning_rate": 5.1274687112609246e-05, "loss": 0.0043, "step": 51730 }, { "grad_norm": 0.09578916430473328, "learning_rate": 5.125815771911372e-05, "loss": 0.0044, "step": 51740 }, { "grad_norm": 0.12448430806398392, "learning_rate": 5.1241628188027956e-05, "loss": 0.0041, "step": 51750 }, { "grad_norm": 0.11733756214380264, "learning_rate": 5.1225098521159585e-05, "loss": 0.0041, "step": 51760 }, { "grad_norm": 0.11670565605163574, "learning_rate": 5.12085687203163e-05, "loss": 0.0044, "step": 51770 }, { "grad_norm": 0.12836019694805145, "learning_rate": 5.119203878730575e-05, "loss": 0.0046, "step": 51780 }, { "grad_norm": 0.10038911551237106, "learning_rate": 5.117550872393564e-05, "loss": 0.004, "step": 51790 }, { "grad_norm": 0.08546270430088043, "learning_rate": 5.1158978532013645e-05, "loss": 0.0036, "step": 51800 }, { "grad_norm": 0.08722522109746933, "learning_rate": 5.114244821334751e-05, "loss": 0.0036, "step": 51810 }, { "grad_norm": 0.09227494895458221, "learning_rate": 5.112591776974497e-05, "loss": 0.0043, "step": 51820 }, { "grad_norm": 0.10903406143188477, "learning_rate": 5.110938720301374e-05, "loss": 0.0041, "step": 51830 }, { "grad_norm": 0.10568851232528687, "learning_rate": 5.1092856514961595e-05, "loss": 0.0052, "step": 51840 }, { "grad_norm": 0.06312890350818634, "learning_rate": 5.107632570739631e-05, "loss": 0.0037, "step": 51850 }, { "grad_norm": 0.0866297110915184, "learning_rate": 5.1059794782125645e-05, "loss": 0.0043, "step": 51860 }, { "grad_norm": 0.0947011336684227, "learning_rate": 5.104326374095744e-05, "loss": 0.004, "step": 51870 }, { "grad_norm": 0.07808498293161392, "learning_rate": 5.102673258569947e-05, "loss": 0.0036, "step": 51880 }, { "grad_norm": 0.08939104527235031, "learning_rate": 5.101020131815956e-05, "loss": 0.0036, "step": 51890 }, { "grad_norm": 0.09843992441892624, "learning_rate": 5.099366994014556e-05, "loss": 0.0046, "step": 51900 }, { "grad_norm": 0.08073978871107101, "learning_rate": 5.097713845346532e-05, "loss": 0.0041, "step": 51910 }, { "grad_norm": 0.10297038406133652, "learning_rate": 5.096060685992668e-05, "loss": 0.0044, "step": 51920 }, { "grad_norm": 0.1408909559249878, "learning_rate": 5.0944075161337514e-05, "loss": 0.0046, "step": 51930 }, { "grad_norm": 0.06322550028562546, "learning_rate": 5.09275433595057e-05, "loss": 0.0034, "step": 51940 }, { "grad_norm": 0.08363864570856094, "learning_rate": 5.0911011456239157e-05, "loss": 0.0042, "step": 51950 }, { "grad_norm": 0.1059315875172615, "learning_rate": 5.089447945334574e-05, "loss": 0.0039, "step": 51960 }, { "grad_norm": 0.07693718373775482, "learning_rate": 5.087794735263343e-05, "loss": 0.0043, "step": 51970 }, { "grad_norm": 0.08706126362085342, "learning_rate": 5.086141515591011e-05, "loss": 0.0045, "step": 51980 }, { "grad_norm": 0.07663895189762115, "learning_rate": 5.084488286498371e-05, "loss": 0.0037, "step": 51990 }, { "grad_norm": 0.09109542518854141, "learning_rate": 5.08283504816622e-05, "loss": 0.0038, "step": 52000 }, { "grad_norm": 0.08394808322191238, "learning_rate": 5.081181800775353e-05, "loss": 0.0034, "step": 52010 }, { "grad_norm": 0.13036267459392548, "learning_rate": 5.079528544506565e-05, "loss": 0.0046, "step": 52020 }, { "grad_norm": 0.07521074265241623, "learning_rate": 5.077875279540657e-05, "loss": 0.0032, "step": 52030 }, { "grad_norm": 0.09464097768068314, "learning_rate": 5.076222006058424e-05, "loss": 0.0037, "step": 52040 }, { "grad_norm": 0.07765763998031616, "learning_rate": 5.074568724240667e-05, "loss": 0.0038, "step": 52050 }, { "grad_norm": 0.0735350102186203, "learning_rate": 5.072915434268187e-05, "loss": 0.0035, "step": 52060 }, { "grad_norm": 0.09284306317567825, "learning_rate": 5.071262136321785e-05, "loss": 0.0041, "step": 52070 }, { "grad_norm": 0.08587570488452911, "learning_rate": 5.069608830582262e-05, "loss": 0.004, "step": 52080 }, { "grad_norm": 0.08862309157848358, "learning_rate": 5.067955517230423e-05, "loss": 0.0047, "step": 52090 }, { "grad_norm": 0.08773086220026016, "learning_rate": 5.06630219644707e-05, "loss": 0.0042, "step": 52100 }, { "grad_norm": 0.09293600916862488, "learning_rate": 5.064648868413008e-05, "loss": 0.0042, "step": 52110 }, { "grad_norm": 0.09368427097797394, "learning_rate": 5.062995533309044e-05, "loss": 0.0037, "step": 52120 }, { "grad_norm": 0.10547982901334763, "learning_rate": 5.061342191315983e-05, "loss": 0.0039, "step": 52130 }, { "grad_norm": 0.07991927117109299, "learning_rate": 5.0596888426146325e-05, "loss": 0.0037, "step": 52140 }, { "grad_norm": 0.09318003803491592, "learning_rate": 5.058035487385798e-05, "loss": 0.0033, "step": 52150 }, { "grad_norm": 0.12160287797451019, "learning_rate": 5.0563821258102915e-05, "loss": 0.0043, "step": 52160 }, { "grad_norm": 0.09033560007810593, "learning_rate": 5.05472875806892e-05, "loss": 0.0044, "step": 52170 }, { "grad_norm": 0.10590124130249023, "learning_rate": 5.053075384342494e-05, "loss": 0.0039, "step": 52180 }, { "grad_norm": 0.08729401230812073, "learning_rate": 5.0514220048118234e-05, "loss": 0.0039, "step": 52190 }, { "grad_norm": 0.09568663686513901, "learning_rate": 5.049768619657721e-05, "loss": 0.0038, "step": 52200 }, { "grad_norm": 0.07918651401996613, "learning_rate": 5.048115229060996e-05, "loss": 0.005, "step": 52210 }, { "grad_norm": 0.07560967653989792, "learning_rate": 5.046461833202462e-05, "loss": 0.0039, "step": 52220 }, { "grad_norm": 0.07003144174814224, "learning_rate": 5.044808432262932e-05, "loss": 0.0037, "step": 52230 }, { "grad_norm": 0.09131006896495819, "learning_rate": 5.043155026423219e-05, "loss": 0.0035, "step": 52240 }, { "grad_norm": 0.09927304834127426, "learning_rate": 5.041501615864137e-05, "loss": 0.0041, "step": 52250 }, { "grad_norm": 0.1039905920624733, "learning_rate": 5.039848200766502e-05, "loss": 0.0039, "step": 52260 }, { "grad_norm": 0.11138800531625748, "learning_rate": 5.038194781311125e-05, "loss": 0.0034, "step": 52270 }, { "grad_norm": 0.08040357381105423, "learning_rate": 5.0365413576788264e-05, "loss": 0.0047, "step": 52280 }, { "grad_norm": 0.0847989022731781, "learning_rate": 5.0348879300504194e-05, "loss": 0.0039, "step": 52290 }, { "grad_norm": 0.0735316351056099, "learning_rate": 5.033234498606721e-05, "loss": 0.0051, "step": 52300 }, { "grad_norm": 0.08020670711994171, "learning_rate": 5.0315810635285476e-05, "loss": 0.0036, "step": 52310 }, { "grad_norm": 0.11801782995462418, "learning_rate": 5.0299276249967164e-05, "loss": 0.0036, "step": 52320 }, { "grad_norm": 0.09997769445180893, "learning_rate": 5.0282741831920454e-05, "loss": 0.005, "step": 52330 }, { "grad_norm": 0.07348169386386871, "learning_rate": 5.026620738295351e-05, "loss": 0.004, "step": 52340 }, { "grad_norm": 0.08118646591901779, "learning_rate": 5.024967290487455e-05, "loss": 0.0044, "step": 52350 }, { "grad_norm": 0.09215475618839264, "learning_rate": 5.023313839949172e-05, "loss": 0.0055, "step": 52360 }, { "grad_norm": 0.08427689224481583, "learning_rate": 5.0216603868613224e-05, "loss": 0.0052, "step": 52370 }, { "grad_norm": 0.09504874795675278, "learning_rate": 5.020006931404726e-05, "loss": 0.0041, "step": 52380 }, { "grad_norm": 0.06809784471988678, "learning_rate": 5.0183534737602014e-05, "loss": 0.0038, "step": 52390 }, { "grad_norm": 0.09441176801919937, "learning_rate": 5.0167000141085686e-05, "loss": 0.0041, "step": 52400 }, { "grad_norm": 0.08736440539360046, "learning_rate": 5.015046552630647e-05, "loss": 0.0043, "step": 52410 }, { "grad_norm": 0.0775066390633583, "learning_rate": 5.013393089507258e-05, "loss": 0.0041, "step": 52420 }, { "grad_norm": 0.09131746739149094, "learning_rate": 5.011739624919219e-05, "loss": 0.0044, "step": 52430 }, { "grad_norm": 0.08732155710458755, "learning_rate": 5.0100861590473545e-05, "loss": 0.0047, "step": 52440 }, { "grad_norm": 0.08722545206546783, "learning_rate": 5.008432692072482e-05, "loss": 0.0039, "step": 52450 }, { "grad_norm": 0.08748410642147064, "learning_rate": 5.006779224175423e-05, "loss": 0.004, "step": 52460 }, { "grad_norm": 0.08462569117546082, "learning_rate": 5.005125755536998e-05, "loss": 0.0046, "step": 52470 }, { "grad_norm": 0.09976218640804291, "learning_rate": 5.003472286338029e-05, "loss": 0.0043, "step": 52480 }, { "grad_norm": 0.07738779485225677, "learning_rate": 5.001818816759335e-05, "loss": 0.0054, "step": 52490 }, { "grad_norm": 0.11113707721233368, "learning_rate": 5.000165346981738e-05, "loss": 0.0048, "step": 52500 }, { "grad_norm": 0.08924438059329987, "learning_rate": 4.9985118771860595e-05, "loss": 0.0038, "step": 52510 }, { "grad_norm": 0.07393690943717957, "learning_rate": 4.996858407553119e-05, "loss": 0.0038, "step": 52520 }, { "grad_norm": 0.09647994488477707, "learning_rate": 4.99520493826374e-05, "loss": 0.0039, "step": 52530 }, { "grad_norm": 0.10077691823244095, "learning_rate": 4.9935514694987384e-05, "loss": 0.0037, "step": 52540 }, { "grad_norm": 0.09280992299318314, "learning_rate": 4.991898001438941e-05, "loss": 0.0043, "step": 52550 }, { "grad_norm": 0.0953446477651596, "learning_rate": 4.990244534265166e-05, "loss": 0.0051, "step": 52560 }, { "grad_norm": 0.11246687918901443, "learning_rate": 4.988591068158234e-05, "loss": 0.0044, "step": 52570 }, { "grad_norm": 0.10439679771661758, "learning_rate": 4.986937603298965e-05, "loss": 0.0042, "step": 52580 }, { "grad_norm": 0.09109127521514893, "learning_rate": 4.985284139868181e-05, "loss": 0.0053, "step": 52590 }, { "grad_norm": 0.11388963460922241, "learning_rate": 4.9836306780467e-05, "loss": 0.0056, "step": 52600 }, { "grad_norm": 0.10621220618486404, "learning_rate": 4.981977218015343e-05, "loss": 0.0044, "step": 52610 }, { "grad_norm": 0.09114708751440048, "learning_rate": 4.980323759954932e-05, "loss": 0.0046, "step": 52620 }, { "grad_norm": 0.09035566449165344, "learning_rate": 4.978670304046284e-05, "loss": 0.0049, "step": 52630 }, { "grad_norm": 0.11180909723043442, "learning_rate": 4.9770168504702195e-05, "loss": 0.0049, "step": 52640 }, { "grad_norm": 0.08459007740020752, "learning_rate": 4.975363399407559e-05, "loss": 0.0039, "step": 52650 }, { "grad_norm": 0.09209837019443512, "learning_rate": 4.9737099510391194e-05, "loss": 0.0045, "step": 52660 }, { "grad_norm": 0.09047085791826248, "learning_rate": 4.972056505545719e-05, "loss": 0.0047, "step": 52670 }, { "grad_norm": 0.07958155870437622, "learning_rate": 4.9704030631081775e-05, "loss": 0.0042, "step": 52680 }, { "grad_norm": 0.10764505714178085, "learning_rate": 4.968749623907313e-05, "loss": 0.0035, "step": 52690 }, { "grad_norm": 0.0976504385471344, "learning_rate": 4.967096188123943e-05, "loss": 0.0049, "step": 52700 }, { "grad_norm": 0.08013731241226196, "learning_rate": 4.965442755938884e-05, "loss": 0.004, "step": 52710 }, { "grad_norm": 0.11531206220388412, "learning_rate": 4.963789327532954e-05, "loss": 0.0042, "step": 52720 }, { "grad_norm": 0.11069618165493011, "learning_rate": 4.962135903086968e-05, "loss": 0.0045, "step": 52730 }, { "grad_norm": 0.09112485498189926, "learning_rate": 4.960482482781741e-05, "loss": 0.0035, "step": 52740 }, { "grad_norm": 0.11477984488010406, "learning_rate": 4.958829066798092e-05, "loss": 0.0046, "step": 52750 }, { "grad_norm": 0.07913622260093689, "learning_rate": 4.957175655316835e-05, "loss": 0.004, "step": 52760 }, { "grad_norm": 0.11328735947608948, "learning_rate": 4.9555222485187834e-05, "loss": 0.0049, "step": 52770 }, { "grad_norm": 0.11862763017416, "learning_rate": 4.953868846584752e-05, "loss": 0.0038, "step": 52780 }, { "grad_norm": 0.0712570995092392, "learning_rate": 4.9522154496955544e-05, "loss": 0.0049, "step": 52790 }, { "grad_norm": 0.08399619162082672, "learning_rate": 4.950562058032001e-05, "loss": 0.0048, "step": 52800 }, { "grad_norm": 0.09710092842578888, "learning_rate": 4.948908671774909e-05, "loss": 0.0044, "step": 52810 }, { "grad_norm": 0.08211096376180649, "learning_rate": 4.947255291105087e-05, "loss": 0.0045, "step": 52820 }, { "grad_norm": 0.0949544832110405, "learning_rate": 4.945601916203348e-05, "loss": 0.0042, "step": 52830 }, { "grad_norm": 0.08360140025615692, "learning_rate": 4.9439485472505016e-05, "loss": 0.0039, "step": 52840 }, { "grad_norm": 0.08766945451498032, "learning_rate": 4.942295184427356e-05, "loss": 0.0045, "step": 52850 }, { "grad_norm": 0.12839500606060028, "learning_rate": 4.9406418279147234e-05, "loss": 0.004, "step": 52860 }, { "grad_norm": 0.1126728281378746, "learning_rate": 4.9389884778934116e-05, "loss": 0.0043, "step": 52870 }, { "grad_norm": 0.0770132765173912, "learning_rate": 4.937335134544228e-05, "loss": 0.0039, "step": 52880 }, { "grad_norm": 0.10260170698165894, "learning_rate": 4.935681798047979e-05, "loss": 0.0043, "step": 52890 }, { "grad_norm": 0.075452521443367, "learning_rate": 4.934028468585473e-05, "loss": 0.0038, "step": 52900 }, { "grad_norm": 0.08542021363973618, "learning_rate": 4.932375146337514e-05, "loss": 0.0043, "step": 52910 }, { "grad_norm": 0.06683231145143509, "learning_rate": 4.930721831484908e-05, "loss": 0.0033, "step": 52920 }, { "grad_norm": 0.06994777917861938, "learning_rate": 4.929068524208456e-05, "loss": 0.0043, "step": 52930 }, { "grad_norm": 0.0806749165058136, "learning_rate": 4.927415224688965e-05, "loss": 0.004, "step": 52940 }, { "grad_norm": 0.09431196004152298, "learning_rate": 4.9257619331072375e-05, "loss": 0.0042, "step": 52950 }, { "grad_norm": 0.10795895755290985, "learning_rate": 4.924108649644072e-05, "loss": 0.0046, "step": 52960 }, { "grad_norm": 0.1016603484749794, "learning_rate": 4.922455374480271e-05, "loss": 0.0036, "step": 52970 }, { "grad_norm": 0.09440841525793076, "learning_rate": 4.920802107796634e-05, "loss": 0.0035, "step": 52980 }, { "grad_norm": 0.07822706550359726, "learning_rate": 4.9191488497739564e-05, "loss": 0.0037, "step": 52990 }, { "grad_norm": 0.09356501698493958, "learning_rate": 4.917495600593042e-05, "loss": 0.0032, "step": 53000 }, { "grad_norm": 0.10902056843042374, "learning_rate": 4.915842360434684e-05, "loss": 0.0047, "step": 53010 }, { "grad_norm": 0.05954146757721901, "learning_rate": 4.914189129479679e-05, "loss": 0.0041, "step": 53020 }, { "grad_norm": 0.09196522831916809, "learning_rate": 4.912535907908822e-05, "loss": 0.0049, "step": 53030 }, { "grad_norm": 0.0988435447216034, "learning_rate": 4.910882695902907e-05, "loss": 0.0043, "step": 53040 }, { "grad_norm": 0.07418229430913925, "learning_rate": 4.9092294936427234e-05, "loss": 0.0053, "step": 53050 }, { "grad_norm": 0.09314417093992233, "learning_rate": 4.9075763013090685e-05, "loss": 0.0047, "step": 53060 }, { "grad_norm": 0.08940799534320831, "learning_rate": 4.9059231190827305e-05, "loss": 0.0038, "step": 53070 }, { "grad_norm": 0.08401785045862198, "learning_rate": 4.904269947144499e-05, "loss": 0.0037, "step": 53080 }, { "grad_norm": 0.10486996918916702, "learning_rate": 4.902616785675161e-05, "loss": 0.004, "step": 53090 }, { "grad_norm": 0.098456472158432, "learning_rate": 4.900963634855505e-05, "loss": 0.0045, "step": 53100 }, { "grad_norm": 0.08467629551887512, "learning_rate": 4.8993104948663173e-05, "loss": 0.0035, "step": 53110 }, { "grad_norm": 0.10124309360980988, "learning_rate": 4.8976573658883823e-05, "loss": 0.003, "step": 53120 }, { "grad_norm": 0.0726514533162117, "learning_rate": 4.896004248102484e-05, "loss": 0.0034, "step": 53130 }, { "grad_norm": 0.069583959877491, "learning_rate": 4.8943511416894045e-05, "loss": 0.0035, "step": 53140 }, { "grad_norm": 0.09039953351020813, "learning_rate": 4.892698046829926e-05, "loss": 0.0043, "step": 53150 }, { "grad_norm": 0.07025423645973206, "learning_rate": 4.8910449637048275e-05, "loss": 0.0035, "step": 53160 }, { "grad_norm": 0.07288399338722229, "learning_rate": 4.8893918924948884e-05, "loss": 0.0044, "step": 53170 }, { "grad_norm": 0.07823940366506577, "learning_rate": 4.8877388333808835e-05, "loss": 0.0035, "step": 53180 }, { "grad_norm": 0.1061486005783081, "learning_rate": 4.8860857865435925e-05, "loss": 0.0056, "step": 53190 }, { "grad_norm": 0.1205587387084961, "learning_rate": 4.884432752163789e-05, "loss": 0.0037, "step": 53200 }, { "grad_norm": 0.06894516944885254, "learning_rate": 4.8827797304222466e-05, "loss": 0.004, "step": 53210 }, { "grad_norm": 0.09429080039262772, "learning_rate": 4.8811267214997366e-05, "loss": 0.0046, "step": 53220 }, { "grad_norm": 0.09364135563373566, "learning_rate": 4.879473725577029e-05, "loss": 0.004, "step": 53230 }, { "grad_norm": 0.0965658649802208, "learning_rate": 4.877820742834891e-05, "loss": 0.0044, "step": 53240 }, { "grad_norm": 0.08134803175926208, "learning_rate": 4.8761677734540956e-05, "loss": 0.0053, "step": 53250 }, { "grad_norm": 0.0757327452301979, "learning_rate": 4.874514817615407e-05, "loss": 0.0059, "step": 53260 }, { "grad_norm": 0.08878692984580994, "learning_rate": 4.872861875499588e-05, "loss": 0.0046, "step": 53270 }, { "grad_norm": 0.10855628550052643, "learning_rate": 4.871208947287404e-05, "loss": 0.004, "step": 53280 }, { "grad_norm": 0.09670670330524445, "learning_rate": 4.869556033159615e-05, "loss": 0.0043, "step": 53290 }, { "grad_norm": 0.07684478163719177, "learning_rate": 4.8679031332969835e-05, "loss": 0.0031, "step": 53300 }, { "grad_norm": 0.0857657864689827, "learning_rate": 4.8662502478802635e-05, "loss": 0.0041, "step": 53310 }, { "grad_norm": 0.10081777721643448, "learning_rate": 4.8645973770902176e-05, "loss": 0.0047, "step": 53320 }, { "grad_norm": 0.097562275826931, "learning_rate": 4.862944521107599e-05, "loss": 0.0043, "step": 53330 }, { "grad_norm": 0.11164116114377975, "learning_rate": 4.861291680113162e-05, "loss": 0.0045, "step": 53340 }, { "grad_norm": 0.10926232486963272, "learning_rate": 4.8596388542876563e-05, "loss": 0.0042, "step": 53350 }, { "grad_norm": 0.10345359146595001, "learning_rate": 4.857986043811836e-05, "loss": 0.0045, "step": 53360 }, { "grad_norm": 0.0792606770992279, "learning_rate": 4.856333248866448e-05, "loss": 0.0041, "step": 53370 }, { "grad_norm": 0.09602776914834976, "learning_rate": 4.85468046963224e-05, "loss": 0.0037, "step": 53380 }, { "grad_norm": 0.07606383413076401, "learning_rate": 4.853027706289957e-05, "loss": 0.0042, "step": 53390 }, { "grad_norm": 0.08040452003479004, "learning_rate": 4.851374959020344e-05, "loss": 0.0034, "step": 53400 }, { "grad_norm": 0.08611761778593063, "learning_rate": 4.849722228004141e-05, "loss": 0.0037, "step": 53410 }, { "grad_norm": 0.0842614397406578, "learning_rate": 4.848069513422091e-05, "loss": 0.0035, "step": 53420 }, { "grad_norm": 0.08643797039985657, "learning_rate": 4.8464168154549294e-05, "loss": 0.0043, "step": 53430 }, { "grad_norm": 0.08330844342708588, "learning_rate": 4.844764134283392e-05, "loss": 0.004, "step": 53440 }, { "grad_norm": 0.13105203211307526, "learning_rate": 4.843111470088218e-05, "loss": 0.0048, "step": 53450 }, { "grad_norm": 0.1292855441570282, "learning_rate": 4.841458823050137e-05, "loss": 0.004, "step": 53460 }, { "grad_norm": 0.0779951959848404, "learning_rate": 4.8398061933498816e-05, "loss": 0.0047, "step": 53470 }, { "grad_norm": 0.09770995378494263, "learning_rate": 4.83815358116818e-05, "loss": 0.0037, "step": 53480 }, { "grad_norm": 0.11235586553812027, "learning_rate": 4.8365009866857584e-05, "loss": 0.0038, "step": 53490 }, { "grad_norm": 0.08469758182764053, "learning_rate": 4.834848410083342e-05, "loss": 0.0042, "step": 53500 }, { "grad_norm": 0.09631899744272232, "learning_rate": 4.833195851541657e-05, "loss": 0.0054, "step": 53510 }, { "grad_norm": 0.08384610712528229, "learning_rate": 4.831543311241423e-05, "loss": 0.0045, "step": 53520 }, { "grad_norm": 0.10026268661022186, "learning_rate": 4.829890789363359e-05, "loss": 0.004, "step": 53530 }, { "grad_norm": 0.08867219090461731, "learning_rate": 4.828238286088182e-05, "loss": 0.0037, "step": 53540 }, { "grad_norm": 0.10102944076061249, "learning_rate": 4.826585801596609e-05, "loss": 0.0045, "step": 53550 }, { "grad_norm": 0.06341370195150375, "learning_rate": 4.8249333360693486e-05, "loss": 0.0033, "step": 53560 }, { "grad_norm": 0.08832486718893051, "learning_rate": 4.8232808896871185e-05, "loss": 0.004, "step": 53570 }, { "grad_norm": 0.07799579948186874, "learning_rate": 4.821628462630624e-05, "loss": 0.0039, "step": 53580 }, { "grad_norm": 0.078346848487854, "learning_rate": 4.8199760550805735e-05, "loss": 0.0035, "step": 53590 }, { "grad_norm": 0.07267038524150848, "learning_rate": 4.818323667217669e-05, "loss": 0.0034, "step": 53600 }, { "grad_norm": 0.07289882749319077, "learning_rate": 4.8166712992226165e-05, "loss": 0.0047, "step": 53610 }, { "grad_norm": 0.08302772790193558, "learning_rate": 4.8150189512761154e-05, "loss": 0.0035, "step": 53620 }, { "grad_norm": 0.08869656920433044, "learning_rate": 4.813366623558862e-05, "loss": 0.0034, "step": 53630 }, { "grad_norm": 0.07124757021665573, "learning_rate": 4.811714316251554e-05, "loss": 0.0038, "step": 53640 }, { "grad_norm": 0.09950073063373566, "learning_rate": 4.810062029534886e-05, "loss": 0.0036, "step": 53650 }, { "grad_norm": 0.07822341471910477, "learning_rate": 4.8084097635895505e-05, "loss": 0.0045, "step": 53660 }, { "grad_norm": 0.09765060245990753, "learning_rate": 4.8067575185962335e-05, "loss": 0.0043, "step": 53670 }, { "grad_norm": 0.10331033915281296, "learning_rate": 4.805105294735625e-05, "loss": 0.0047, "step": 53680 }, { "grad_norm": 0.08633862435817719, "learning_rate": 4.803453092188406e-05, "loss": 0.0033, "step": 53690 }, { "grad_norm": 0.07680127769708633, "learning_rate": 4.801800911135263e-05, "loss": 0.0038, "step": 53700 }, { "grad_norm": 0.09745801240205765, "learning_rate": 4.8001487517568744e-05, "loss": 0.0046, "step": 53710 }, { "grad_norm": 0.07895039021968842, "learning_rate": 4.798496614233918e-05, "loss": 0.0041, "step": 53720 }, { "grad_norm": 0.08986865729093552, "learning_rate": 4.796844498747069e-05, "loss": 0.0033, "step": 53730 }, { "grad_norm": 0.1193402111530304, "learning_rate": 4.795192405477e-05, "loss": 0.0043, "step": 53740 }, { "grad_norm": 0.12305216491222382, "learning_rate": 4.79354033460438e-05, "loss": 0.0041, "step": 53750 }, { "grad_norm": 0.0816611647605896, "learning_rate": 4.79188828630988e-05, "loss": 0.0037, "step": 53760 }, { "grad_norm": 0.0779176577925682, "learning_rate": 4.790236260774165e-05, "loss": 0.0037, "step": 53770 }, { "grad_norm": 0.09804166853427887, "learning_rate": 4.788584258177896e-05, "loss": 0.0043, "step": 53780 }, { "grad_norm": 0.10724659264087677, "learning_rate": 4.7869322787017356e-05, "loss": 0.0053, "step": 53790 }, { "grad_norm": 0.08318087458610535, "learning_rate": 4.78528032252634e-05, "loss": 0.004, "step": 53800 }, { "grad_norm": 0.07719352096319199, "learning_rate": 4.783628389832366e-05, "loss": 0.0043, "step": 53810 }, { "grad_norm": 0.08908591419458389, "learning_rate": 4.7819764808004644e-05, "loss": 0.0036, "step": 53820 }, { "grad_norm": 0.09701327979564667, "learning_rate": 4.780324595611289e-05, "loss": 0.0043, "step": 53830 }, { "grad_norm": 0.0875251442193985, "learning_rate": 4.7786727344454854e-05, "loss": 0.0033, "step": 53840 }, { "grad_norm": 0.07009870558977127, "learning_rate": 4.777020897483697e-05, "loss": 0.0031, "step": 53850 }, { "grad_norm": 0.10689107328653336, "learning_rate": 4.775369084906569e-05, "loss": 0.0051, "step": 53860 }, { "grad_norm": 0.06478100270032883, "learning_rate": 4.773717296894742e-05, "loss": 0.0035, "step": 53870 }, { "grad_norm": 0.09907940775156021, "learning_rate": 4.772065533628848e-05, "loss": 0.0048, "step": 53880 }, { "grad_norm": 0.08927559107542038, "learning_rate": 4.7704137952895265e-05, "loss": 0.0035, "step": 53890 }, { "grad_norm": 0.09444223344326019, "learning_rate": 4.768762082057407e-05, "loss": 0.0036, "step": 53900 }, { "grad_norm": 0.06918582320213318, "learning_rate": 4.7671103941131186e-05, "loss": 0.0031, "step": 53910 }, { "grad_norm": 0.07482510805130005, "learning_rate": 4.765458731637287e-05, "loss": 0.0033, "step": 53920 }, { "grad_norm": 0.05697180703282356, "learning_rate": 4.763807094810537e-05, "loss": 0.0035, "step": 53930 }, { "grad_norm": 0.11252898722887039, "learning_rate": 4.762155483813487e-05, "loss": 0.004, "step": 53940 }, { "grad_norm": 0.08401080965995789, "learning_rate": 4.760503898826754e-05, "loss": 0.0039, "step": 53950 }, { "grad_norm": 0.0897633507847786, "learning_rate": 4.758852340030957e-05, "loss": 0.0046, "step": 53960 }, { "grad_norm": 0.07159151881933212, "learning_rate": 4.757200807606705e-05, "loss": 0.0037, "step": 53970 }, { "grad_norm": 0.09718244522809982, "learning_rate": 4.755549301734607e-05, "loss": 0.0044, "step": 53980 }, { "grad_norm": 0.10046499222517014, "learning_rate": 4.753897822595271e-05, "loss": 0.0042, "step": 53990 }, { "grad_norm": 0.07653512805700302, "learning_rate": 4.7522463703692994e-05, "loss": 0.0048, "step": 54000 }, { "grad_norm": 0.07640454918146133, "learning_rate": 4.75059494523729e-05, "loss": 0.0039, "step": 54010 }, { "grad_norm": 0.11351823806762695, "learning_rate": 4.748943547379843e-05, "loss": 0.0035, "step": 54020 }, { "grad_norm": 0.10647325962781906, "learning_rate": 4.7472921769775535e-05, "loss": 0.0049, "step": 54030 }, { "grad_norm": 0.10131777822971344, "learning_rate": 4.745640834211011e-05, "loss": 0.0037, "step": 54040 }, { "grad_norm": 0.09456765651702881, "learning_rate": 4.7439895192608054e-05, "loss": 0.0049, "step": 54050 }, { "grad_norm": 0.08039205521345139, "learning_rate": 4.7423382323075186e-05, "loss": 0.0032, "step": 54060 }, { "grad_norm": 0.10853327065706253, "learning_rate": 4.740686973531737e-05, "loss": 0.0033, "step": 54070 }, { "grad_norm": 0.07345318794250488, "learning_rate": 4.7390357431140375e-05, "loss": 0.0042, "step": 54080 }, { "grad_norm": 0.10700001567602158, "learning_rate": 4.737384541234998e-05, "loss": 0.0043, "step": 54090 }, { "grad_norm": 0.09819003194570541, "learning_rate": 4.735733368075189e-05, "loss": 0.0039, "step": 54100 }, { "grad_norm": 0.08639954030513763, "learning_rate": 4.734082223815182e-05, "loss": 0.0037, "step": 54110 }, { "grad_norm": 0.10050681233406067, "learning_rate": 4.7324311086355436e-05, "loss": 0.0034, "step": 54120 }, { "grad_norm": 0.08409620076417923, "learning_rate": 4.730780022716836e-05, "loss": 0.0043, "step": 54130 }, { "grad_norm": 0.1079045981168747, "learning_rate": 4.729128966239619e-05, "loss": 0.0035, "step": 54140 }, { "grad_norm": 0.0893847644329071, "learning_rate": 4.7274779393844536e-05, "loss": 0.0041, "step": 54150 }, { "grad_norm": 0.08222386986017227, "learning_rate": 4.725826942331891e-05, "loss": 0.0036, "step": 54160 }, { "grad_norm": 0.08326276391744614, "learning_rate": 4.724175975262481e-05, "loss": 0.0038, "step": 54170 }, { "grad_norm": 0.07589142769575119, "learning_rate": 4.722525038356772e-05, "loss": 0.004, "step": 54180 }, { "grad_norm": 0.07217997312545776, "learning_rate": 4.7208741317953075e-05, "loss": 0.0038, "step": 54190 }, { "grad_norm": 0.08992304652929306, "learning_rate": 4.7192232557586264e-05, "loss": 0.0038, "step": 54200 }, { "grad_norm": 0.06295529007911682, "learning_rate": 4.717572410427269e-05, "loss": 0.0037, "step": 54210 }, { "grad_norm": 0.09780069440603256, "learning_rate": 4.71592159598177e-05, "loss": 0.0034, "step": 54220 }, { "grad_norm": 0.0701041966676712, "learning_rate": 4.714270812602657e-05, "loss": 0.0042, "step": 54230 }, { "grad_norm": 0.08054496347904205, "learning_rate": 4.712620060470458e-05, "loss": 0.004, "step": 54240 }, { "grad_norm": 0.07911300659179688, "learning_rate": 4.7109693397656986e-05, "loss": 0.0042, "step": 54250 }, { "grad_norm": 0.0737261176109314, "learning_rate": 4.709318650668894e-05, "loss": 0.004, "step": 54260 }, { "grad_norm": 0.10642287135124207, "learning_rate": 4.7076679933605676e-05, "loss": 0.0049, "step": 54270 }, { "grad_norm": 0.10347152501344681, "learning_rate": 4.70601736802123e-05, "loss": 0.0038, "step": 54280 }, { "grad_norm": 0.09528873860836029, "learning_rate": 4.704366774831391e-05, "loss": 0.0041, "step": 54290 }, { "grad_norm": 0.07089639455080032, "learning_rate": 4.702716213971557e-05, "loss": 0.004, "step": 54300 }, { "grad_norm": 0.08937373012304306, "learning_rate": 4.70106568562223e-05, "loss": 0.0041, "step": 54310 }, { "grad_norm": 0.09216094017028809, "learning_rate": 4.6994151899639105e-05, "loss": 0.0037, "step": 54320 }, { "grad_norm": 0.09179244190454483, "learning_rate": 4.697764727177093e-05, "loss": 0.0033, "step": 54330 }, { "grad_norm": 0.09044406563043594, "learning_rate": 4.696114297442272e-05, "loss": 0.0034, "step": 54340 }, { "grad_norm": 0.07726408541202545, "learning_rate": 4.694463900939933e-05, "loss": 0.0035, "step": 54350 }, { "grad_norm": 0.07397449016571045, "learning_rate": 4.692813537850564e-05, "loss": 0.0031, "step": 54360 }, { "grad_norm": 0.08192019164562225, "learning_rate": 4.6911632083546445e-05, "loss": 0.0048, "step": 54370 }, { "grad_norm": 0.08638506382703781, "learning_rate": 4.6895129126326525e-05, "loss": 0.0033, "step": 54380 }, { "grad_norm": 0.08432574570178986, "learning_rate": 4.687862650865059e-05, "loss": 0.0036, "step": 54390 }, { "grad_norm": 0.07160467654466629, "learning_rate": 4.686212423232339e-05, "loss": 0.0035, "step": 54400 }, { "grad_norm": 0.07233861088752747, "learning_rate": 4.6845622299149565e-05, "loss": 0.0043, "step": 54410 }, { "grad_norm": 0.07831909507513046, "learning_rate": 4.682912071093374e-05, "loss": 0.0039, "step": 54420 }, { "grad_norm": 0.09084822982549667, "learning_rate": 4.681261946948052e-05, "loss": 0.0041, "step": 54430 }, { "grad_norm": 0.0905027836561203, "learning_rate": 4.679611857659443e-05, "loss": 0.0044, "step": 54440 }, { "grad_norm": 0.14333659410476685, "learning_rate": 4.6779618034079975e-05, "loss": 0.0047, "step": 54450 }, { "grad_norm": 0.06226104125380516, "learning_rate": 4.6763117843741675e-05, "loss": 0.0039, "step": 54460 }, { "grad_norm": 0.08688328415155411, "learning_rate": 4.674661800738394e-05, "loss": 0.0044, "step": 54470 }, { "grad_norm": 0.06514216214418411, "learning_rate": 4.673011852681117e-05, "loss": 0.0037, "step": 54480 }, { "grad_norm": 0.07421384006738663, "learning_rate": 4.671361940382772e-05, "loss": 0.0035, "step": 54490 }, { "grad_norm": 0.09096456319093704, "learning_rate": 4.669712064023791e-05, "loss": 0.0043, "step": 54500 }, { "grad_norm": 0.07438311725854874, "learning_rate": 4.668062223784602e-05, "loss": 0.0035, "step": 54510 }, { "grad_norm": 0.12331154942512512, "learning_rate": 4.6664124198456274e-05, "loss": 0.0048, "step": 54520 }, { "grad_norm": 0.11117088049650192, "learning_rate": 4.6647626523872905e-05, "loss": 0.0048, "step": 54530 }, { "grad_norm": 0.08630511909723282, "learning_rate": 4.6631129215900066e-05, "loss": 0.0043, "step": 54540 }, { "grad_norm": 0.10602130740880966, "learning_rate": 4.661463227634187e-05, "loss": 0.0046, "step": 54550 }, { "grad_norm": 0.06511285156011581, "learning_rate": 4.659813570700238e-05, "loss": 0.0039, "step": 54560 }, { "grad_norm": 0.07702791690826416, "learning_rate": 4.658163950968568e-05, "loss": 0.004, "step": 54570 }, { "grad_norm": 0.08870071172714233, "learning_rate": 4.6565143686195726e-05, "loss": 0.004, "step": 54580 }, { "grad_norm": 0.09920652955770493, "learning_rate": 4.654864823833649e-05, "loss": 0.0038, "step": 54590 }, { "grad_norm": 0.10469958186149597, "learning_rate": 4.65321531679119e-05, "loss": 0.0039, "step": 54600 }, { "grad_norm": 0.07865329086780548, "learning_rate": 4.6515658476725834e-05, "loss": 0.0049, "step": 54610 }, { "grad_norm": 0.0953718051314354, "learning_rate": 4.649916416658212e-05, "loss": 0.0033, "step": 54620 }, { "grad_norm": 0.08389133214950562, "learning_rate": 4.648267023928455e-05, "loss": 0.0035, "step": 54630 }, { "grad_norm": 0.07841919362545013, "learning_rate": 4.646617669663687e-05, "loss": 0.0037, "step": 54640 }, { "grad_norm": 0.06881585717201233, "learning_rate": 4.644968354044277e-05, "loss": 0.0028, "step": 54650 }, { "grad_norm": 0.07244057953357697, "learning_rate": 4.643319077250596e-05, "loss": 0.0044, "step": 54660 }, { "grad_norm": 0.07316099107265472, "learning_rate": 4.6416698394630055e-05, "loss": 0.0037, "step": 54670 }, { "grad_norm": 0.10026867687702179, "learning_rate": 4.640020640861862e-05, "loss": 0.004, "step": 54680 }, { "grad_norm": 0.09870384633541107, "learning_rate": 4.6383714816275195e-05, "loss": 0.0041, "step": 54690 }, { "grad_norm": 0.1080564558506012, "learning_rate": 4.636722361940329e-05, "loss": 0.0039, "step": 54700 }, { "grad_norm": 0.055391184985637665, "learning_rate": 4.635073281980632e-05, "loss": 0.0034, "step": 54710 }, { "grad_norm": 0.11293882131576538, "learning_rate": 4.633424241928775e-05, "loss": 0.0039, "step": 54720 }, { "grad_norm": 0.07563518732786179, "learning_rate": 4.6317752419650914e-05, "loss": 0.0032, "step": 54730 }, { "grad_norm": 0.08300863206386566, "learning_rate": 4.630126282269914e-05, "loss": 0.0036, "step": 54740 }, { "grad_norm": 0.10482034832239151, "learning_rate": 4.628477363023569e-05, "loss": 0.0036, "step": 54750 }, { "grad_norm": 0.0830514058470726, "learning_rate": 4.626828484406383e-05, "loss": 0.0039, "step": 54760 }, { "grad_norm": 0.10221295803785324, "learning_rate": 4.625179646598669e-05, "loss": 0.0034, "step": 54770 }, { "grad_norm": 0.08004141598939896, "learning_rate": 4.6235308497807487e-05, "loss": 0.0031, "step": 54780 }, { "grad_norm": 0.06842917203903198, "learning_rate": 4.6218820941329276e-05, "loss": 0.0036, "step": 54790 }, { "grad_norm": 0.07845941185951233, "learning_rate": 4.620233379835513e-05, "loss": 0.0032, "step": 54800 }, { "grad_norm": 0.06547894328832626, "learning_rate": 4.618584707068804e-05, "loss": 0.0037, "step": 54810 }, { "grad_norm": 0.08017534017562866, "learning_rate": 4.616936076013099e-05, "loss": 0.0034, "step": 54820 }, { "grad_norm": 0.07456522434949875, "learning_rate": 4.615287486848688e-05, "loss": 0.0048, "step": 54830 }, { "grad_norm": 0.0802885964512825, "learning_rate": 4.6136389397558584e-05, "loss": 0.004, "step": 54840 }, { "grad_norm": 0.09279339015483856, "learning_rate": 4.6119904349148935e-05, "loss": 0.0043, "step": 54850 }, { "grad_norm": 0.08205332607030869, "learning_rate": 4.6103419725060726e-05, "loss": 0.0033, "step": 54860 }, { "grad_norm": 0.07220089435577393, "learning_rate": 4.608693552709667e-05, "loss": 0.0037, "step": 54870 }, { "grad_norm": 0.08612732589244843, "learning_rate": 4.607045175705947e-05, "loss": 0.0036, "step": 54880 }, { "grad_norm": 0.0964004322886467, "learning_rate": 4.605396841675176e-05, "loss": 0.0035, "step": 54890 }, { "grad_norm": 0.07947605848312378, "learning_rate": 4.6037485507976094e-05, "loss": 0.0032, "step": 54900 }, { "grad_norm": 0.09813826531171799, "learning_rate": 4.602100303253509e-05, "loss": 0.0029, "step": 54910 }, { "grad_norm": 0.09352318197488785, "learning_rate": 4.600452099223122e-05, "loss": 0.0038, "step": 54920 }, { "grad_norm": 0.08246295899152756, "learning_rate": 4.598803938886691e-05, "loss": 0.0039, "step": 54930 }, { "grad_norm": 0.1022295281291008, "learning_rate": 4.597155822424459e-05, "loss": 0.0036, "step": 54940 }, { "grad_norm": 0.07442716509103775, "learning_rate": 4.59550775001666e-05, "loss": 0.0039, "step": 54950 }, { "grad_norm": 0.10634541511535645, "learning_rate": 4.5938597218435234e-05, "loss": 0.0051, "step": 54960 }, { "grad_norm": 0.06479822844266891, "learning_rate": 4.592211738085279e-05, "loss": 0.0036, "step": 54970 }, { "grad_norm": 0.09683027118444443, "learning_rate": 4.590563798922146e-05, "loss": 0.0034, "step": 54980 }, { "grad_norm": 0.07277608662843704, "learning_rate": 4.5889159045343404e-05, "loss": 0.0039, "step": 54990 }, { "grad_norm": 0.10846611857414246, "learning_rate": 4.5872680551020734e-05, "loss": 0.0039, "step": 55000 }, { "grad_norm": 0.1180432140827179, "learning_rate": 4.585620250805552e-05, "loss": 0.0037, "step": 55010 }, { "grad_norm": 0.07871155440807343, "learning_rate": 4.583972491824975e-05, "loss": 0.0038, "step": 55020 }, { "grad_norm": 0.09187823534011841, "learning_rate": 4.582324778340541e-05, "loss": 0.0029, "step": 55030 }, { "grad_norm": 0.057269588112831116, "learning_rate": 4.5806771105324424e-05, "loss": 0.0039, "step": 55040 }, { "grad_norm": 0.06793134659528732, "learning_rate": 4.579029488580865e-05, "loss": 0.0035, "step": 55050 }, { "grad_norm": 0.08590355515480042, "learning_rate": 4.5773819126659876e-05, "loss": 0.0037, "step": 55060 }, { "grad_norm": 0.09127875417470932, "learning_rate": 4.575734382967991e-05, "loss": 0.0042, "step": 55070 }, { "grad_norm": 0.07447013258934021, "learning_rate": 4.574086899667044e-05, "loss": 0.0038, "step": 55080 }, { "grad_norm": 0.09109663963317871, "learning_rate": 4.5724394629433106e-05, "loss": 0.0037, "step": 55090 }, { "grad_norm": 0.09784015268087387, "learning_rate": 4.570792072976958e-05, "loss": 0.004, "step": 55100 }, { "grad_norm": 0.0796806737780571, "learning_rate": 4.569144729948138e-05, "loss": 0.0034, "step": 55110 }, { "grad_norm": 0.0950361043214798, "learning_rate": 4.5674974340370033e-05, "loss": 0.0041, "step": 55120 }, { "grad_norm": 0.0915859043598175, "learning_rate": 4.565850185423699e-05, "loss": 0.0038, "step": 55130 }, { "grad_norm": 0.11745433509349823, "learning_rate": 4.564202984288365e-05, "loss": 0.0038, "step": 55140 }, { "grad_norm": 0.08265961706638336, "learning_rate": 4.5625558308111356e-05, "loss": 0.0033, "step": 55150 }, { "grad_norm": 0.07985226064920425, "learning_rate": 4.560908725172145e-05, "loss": 0.0036, "step": 55160 }, { "grad_norm": 0.07133232057094574, "learning_rate": 4.559261667551516e-05, "loss": 0.0032, "step": 55170 }, { "grad_norm": 0.07987134903669357, "learning_rate": 4.5576146581293685e-05, "loss": 0.0033, "step": 55180 }, { "grad_norm": 0.07407079637050629, "learning_rate": 4.5559676970858164e-05, "loss": 0.0032, "step": 55190 }, { "grad_norm": 0.08817987143993378, "learning_rate": 4.5543207846009705e-05, "loss": 0.0045, "step": 55200 }, { "grad_norm": 0.09743481129407883, "learning_rate": 4.552673920854933e-05, "loss": 0.0035, "step": 55210 }, { "grad_norm": 0.10270975530147552, "learning_rate": 4.5510271060278006e-05, "loss": 0.0039, "step": 55220 }, { "grad_norm": 0.08526916056871414, "learning_rate": 4.5493803402996704e-05, "loss": 0.0047, "step": 55230 }, { "grad_norm": 0.09751738607883453, "learning_rate": 4.5477336238506295e-05, "loss": 0.0037, "step": 55240 }, { "grad_norm": 0.08309632539749146, "learning_rate": 4.546086956860759e-05, "loss": 0.0038, "step": 55250 }, { "grad_norm": 0.0852682888507843, "learning_rate": 4.5444403395101374e-05, "loss": 0.0036, "step": 55260 }, { "grad_norm": 0.11129888147115707, "learning_rate": 4.542793771978834e-05, "loss": 0.0051, "step": 55270 }, { "grad_norm": 0.12687718868255615, "learning_rate": 4.541147254446916e-05, "loss": 0.0041, "step": 55280 }, { "grad_norm": 0.08104901760816574, "learning_rate": 4.539500787094445e-05, "loss": 0.0039, "step": 55290 }, { "grad_norm": 0.06417567282915115, "learning_rate": 4.537854370101475e-05, "loss": 0.0034, "step": 55300 }, { "grad_norm": 0.06551583111286163, "learning_rate": 4.5362080036480566e-05, "loss": 0.0031, "step": 55310 }, { "grad_norm": 0.0818302258849144, "learning_rate": 4.534561687914235e-05, "loss": 0.0039, "step": 55320 }, { "grad_norm": 0.08849547058343887, "learning_rate": 4.5329154230800466e-05, "loss": 0.0042, "step": 55330 }, { "grad_norm": 0.07586819678544998, "learning_rate": 4.531269209325525e-05, "loss": 0.0035, "step": 55340 }, { "grad_norm": 0.09533156454563141, "learning_rate": 4.529623046830696e-05, "loss": 0.0047, "step": 55350 }, { "grad_norm": 0.09942235052585602, "learning_rate": 4.5279769357755866e-05, "loss": 0.0036, "step": 55360 }, { "grad_norm": 0.0766431912779808, "learning_rate": 4.5263308763402084e-05, "loss": 0.0032, "step": 55370 }, { "grad_norm": 0.09829381108283997, "learning_rate": 4.524684868704574e-05, "loss": 0.0039, "step": 55380 }, { "grad_norm": 0.11021061986684799, "learning_rate": 4.523038913048688e-05, "loss": 0.0039, "step": 55390 }, { "grad_norm": 0.11868810653686523, "learning_rate": 4.521393009552549e-05, "loss": 0.0047, "step": 55400 }, { "grad_norm": 0.12510967254638672, "learning_rate": 4.5197471583961484e-05, "loss": 0.0053, "step": 55410 }, { "grad_norm": 0.11482493579387665, "learning_rate": 4.51810135975948e-05, "loss": 0.0051, "step": 55420 }, { "grad_norm": 0.0766189843416214, "learning_rate": 4.516455613822521e-05, "loss": 0.0039, "step": 55430 }, { "grad_norm": 0.07864319533109665, "learning_rate": 4.514809920765249e-05, "loss": 0.0049, "step": 55440 }, { "grad_norm": 0.08471428602933884, "learning_rate": 4.5131642807676356e-05, "loss": 0.0038, "step": 55450 }, { "grad_norm": 0.0695214718580246, "learning_rate": 4.511518694009644e-05, "loss": 0.0032, "step": 55460 }, { "grad_norm": 0.09437042474746704, "learning_rate": 4.509873160671231e-05, "loss": 0.0038, "step": 55470 }, { "grad_norm": 0.07931150496006012, "learning_rate": 4.508227680932356e-05, "loss": 0.0036, "step": 55480 }, { "grad_norm": 0.08027583360671997, "learning_rate": 4.5065822549729605e-05, "loss": 0.0031, "step": 55490 }, { "grad_norm": 0.07316857576370239, "learning_rate": 4.504936882972989e-05, "loss": 0.0044, "step": 55500 }, { "grad_norm": 0.09897057712078094, "learning_rate": 4.503291565112374e-05, "loss": 0.0035, "step": 55510 }, { "grad_norm": 0.10618927329778671, "learning_rate": 4.501646301571048e-05, "loss": 0.0053, "step": 55520 }, { "grad_norm": 0.11056005954742432, "learning_rate": 4.500001092528933e-05, "loss": 0.0043, "step": 55530 }, { "grad_norm": 0.0804462656378746, "learning_rate": 4.4983559381659455e-05, "loss": 0.0047, "step": 55540 }, { "grad_norm": 0.07913728803396225, "learning_rate": 4.496710838661999e-05, "loss": 0.004, "step": 55550 }, { "grad_norm": 0.08865658193826675, "learning_rate": 4.495065794196999e-05, "loss": 0.004, "step": 55560 }, { "grad_norm": 0.09878643602132797, "learning_rate": 4.4934208049508455e-05, "loss": 0.004, "step": 55570 }, { "grad_norm": 0.09148057550191879, "learning_rate": 4.4917758711034303e-05, "loss": 0.0036, "step": 55580 }, { "grad_norm": 0.0662107840180397, "learning_rate": 4.490130992834642e-05, "loss": 0.0033, "step": 55590 }, { "grad_norm": 0.07784806936979294, "learning_rate": 4.488486170324359e-05, "loss": 0.0034, "step": 55600 }, { "grad_norm": 0.08451943099498749, "learning_rate": 4.486841403752463e-05, "loss": 0.0033, "step": 55610 }, { "grad_norm": 0.09141631424427032, "learning_rate": 4.4851966932988174e-05, "loss": 0.0037, "step": 55620 }, { "grad_norm": 0.08055353164672852, "learning_rate": 4.4835520391432894e-05, "loss": 0.004, "step": 55630 }, { "grad_norm": 0.07858437299728394, "learning_rate": 4.481907441465733e-05, "loss": 0.0041, "step": 55640 }, { "grad_norm": 0.10725604742765427, "learning_rate": 4.480262900446e-05, "loss": 0.0037, "step": 55650 }, { "grad_norm": 0.07688163220882416, "learning_rate": 4.478618416263933e-05, "loss": 0.0051, "step": 55660 }, { "grad_norm": 0.08285423368215561, "learning_rate": 4.476973989099373e-05, "loss": 0.0035, "step": 55670 }, { "grad_norm": 0.09206046909093857, "learning_rate": 4.475329619132153e-05, "loss": 0.0038, "step": 55680 }, { "grad_norm": 0.07312947511672974, "learning_rate": 4.4736853065420956e-05, "loss": 0.0032, "step": 55690 }, { "grad_norm": 0.08333349972963333, "learning_rate": 4.472041051509023e-05, "loss": 0.0034, "step": 55700 }, { "grad_norm": 0.08609087765216827, "learning_rate": 4.470396854212746e-05, "loss": 0.0055, "step": 55710 }, { "grad_norm": 0.0787021741271019, "learning_rate": 4.468752714833072e-05, "loss": 0.0036, "step": 55720 }, { "grad_norm": 0.12524445354938507, "learning_rate": 4.467108633549803e-05, "loss": 0.0034, "step": 55730 }, { "grad_norm": 0.06589355319738388, "learning_rate": 4.4654646105427335e-05, "loss": 0.0038, "step": 55740 }, { "grad_norm": 0.09359030425548553, "learning_rate": 4.463820645991651e-05, "loss": 0.0038, "step": 55750 }, { "grad_norm": 0.08140744268894196, "learning_rate": 4.462176740076334e-05, "loss": 0.0039, "step": 55760 }, { "grad_norm": 0.10071878880262375, "learning_rate": 4.4605328929765624e-05, "loss": 0.0039, "step": 55770 }, { "grad_norm": 0.070040263235569, "learning_rate": 4.458889104872102e-05, "loss": 0.0045, "step": 55780 }, { "grad_norm": 0.09099125117063522, "learning_rate": 4.457245375942715e-05, "loss": 0.0036, "step": 55790 }, { "grad_norm": 0.10612650215625763, "learning_rate": 4.455601706368157e-05, "loss": 0.005, "step": 55800 }, { "grad_norm": 0.07554437965154648, "learning_rate": 4.453958096328179e-05, "loss": 0.0031, "step": 55810 }, { "grad_norm": 0.083024762570858, "learning_rate": 4.452314546002523e-05, "loss": 0.0033, "step": 55820 }, { "grad_norm": 0.08815212547779083, "learning_rate": 4.4506710555709246e-05, "loss": 0.0033, "step": 55830 }, { "grad_norm": 0.09170600026845932, "learning_rate": 4.449027625213114e-05, "loss": 0.0037, "step": 55840 }, { "grad_norm": 0.09803982079029083, "learning_rate": 4.4473842551088135e-05, "loss": 0.0035, "step": 55850 }, { "grad_norm": 0.10444843769073486, "learning_rate": 4.445740945437738e-05, "loss": 0.0044, "step": 55860 }, { "grad_norm": 0.09404854476451874, "learning_rate": 4.444097696379602e-05, "loss": 0.0043, "step": 55870 }, { "grad_norm": 0.09439917653799057, "learning_rate": 4.442454508114106e-05, "loss": 0.0037, "step": 55880 }, { "grad_norm": 0.0834881141781807, "learning_rate": 4.4408113808209465e-05, "loss": 0.0036, "step": 55890 }, { "grad_norm": 0.07935428619384766, "learning_rate": 4.439168314679813e-05, "loss": 0.0032, "step": 55900 }, { "grad_norm": 0.0722607746720314, "learning_rate": 4.43752530987039e-05, "loss": 0.0037, "step": 55910 }, { "grad_norm": 0.05724063143134117, "learning_rate": 4.4358823665723515e-05, "loss": 0.0039, "step": 55920 }, { "grad_norm": 0.08261705935001373, "learning_rate": 4.434239484965371e-05, "loss": 0.0036, "step": 55930 }, { "grad_norm": 0.07954221963882446, "learning_rate": 4.4325966652291103e-05, "loss": 0.0032, "step": 55940 }, { "grad_norm": 0.07887262850999832, "learning_rate": 4.430953907543225e-05, "loss": 0.0032, "step": 55950 }, { "grad_norm": 0.09655016660690308, "learning_rate": 4.4293112120873645e-05, "loss": 0.0037, "step": 55960 }, { "grad_norm": 0.0778782069683075, "learning_rate": 4.427668579041172e-05, "loss": 0.0042, "step": 55970 }, { "grad_norm": 0.1357918381690979, "learning_rate": 4.4260260085842816e-05, "loss": 0.0038, "step": 55980 }, { "grad_norm": 0.0923580676317215, "learning_rate": 4.4243835008963256e-05, "loss": 0.0033, "step": 55990 }, { "grad_norm": 0.09450270980596542, "learning_rate": 4.422741056156925e-05, "loss": 0.0039, "step": 56000 }, { "grad_norm": 0.10126031935214996, "learning_rate": 4.421098674545693e-05, "loss": 0.0038, "step": 56010 }, { "grad_norm": 0.09933023154735565, "learning_rate": 4.419456356242241e-05, "loss": 0.004, "step": 56020 }, { "grad_norm": 0.09765461087226868, "learning_rate": 4.4178141014261685e-05, "loss": 0.0037, "step": 56030 }, { "grad_norm": 0.0836653783917427, "learning_rate": 4.416171910277071e-05, "loss": 0.0039, "step": 56040 }, { "grad_norm": 0.09468849003314972, "learning_rate": 4.414529782974533e-05, "loss": 0.0035, "step": 56050 }, { "grad_norm": 0.09708540886640549, "learning_rate": 4.41288771969814e-05, "loss": 0.0036, "step": 56060 }, { "grad_norm": 0.1084512323141098, "learning_rate": 4.411245720627464e-05, "loss": 0.0041, "step": 56070 }, { "grad_norm": 0.08248530328273773, "learning_rate": 4.40960378594207e-05, "loss": 0.0038, "step": 56080 }, { "grad_norm": 0.08839204907417297, "learning_rate": 4.407961915821519e-05, "loss": 0.0038, "step": 56090 }, { "grad_norm": 0.08894520252943039, "learning_rate": 4.406320110445362e-05, "loss": 0.004, "step": 56100 }, { "grad_norm": 0.07989533245563507, "learning_rate": 4.404678369993144e-05, "loss": 0.0034, "step": 56110 }, { "grad_norm": 0.06618232280015945, "learning_rate": 4.403036694644406e-05, "loss": 0.0037, "step": 56120 }, { "grad_norm": 0.07717184722423553, "learning_rate": 4.4013950845786764e-05, "loss": 0.0041, "step": 56130 }, { "grad_norm": 0.10519363731145859, "learning_rate": 4.399753539975482e-05, "loss": 0.0034, "step": 56140 }, { "grad_norm": 0.09555002301931381, "learning_rate": 4.398112061014337e-05, "loss": 0.0035, "step": 56150 }, { "grad_norm": 0.06986681371927261, "learning_rate": 4.396470647874753e-05, "loss": 0.0033, "step": 56160 }, { "grad_norm": 0.07681720703840256, "learning_rate": 4.394829300736229e-05, "loss": 0.0031, "step": 56170 }, { "grad_norm": 0.07812315225601196, "learning_rate": 4.393188019778265e-05, "loss": 0.0034, "step": 56180 }, { "grad_norm": 0.07238823920488358, "learning_rate": 4.391546805180347e-05, "loss": 0.0031, "step": 56190 }, { "grad_norm": 0.07542424649000168, "learning_rate": 4.389905657121955e-05, "loss": 0.0037, "step": 56200 }, { "grad_norm": 0.07053434103727341, "learning_rate": 4.3882645757825635e-05, "loss": 0.003, "step": 56210 }, { "grad_norm": 0.07720882445573807, "learning_rate": 4.386623561341637e-05, "loss": 0.0033, "step": 56220 }, { "grad_norm": 0.05586530268192291, "learning_rate": 4.384982613978637e-05, "loss": 0.0035, "step": 56230 }, { "grad_norm": 0.09815565496683121, "learning_rate": 4.383341733873012e-05, "loss": 0.0034, "step": 56240 }, { "grad_norm": 0.07958461344242096, "learning_rate": 4.381700921204209e-05, "loss": 0.0036, "step": 56250 }, { "grad_norm": 0.07881073653697968, "learning_rate": 4.380060176151663e-05, "loss": 0.003, "step": 56260 }, { "grad_norm": 0.07048773020505905, "learning_rate": 4.378419498894805e-05, "loss": 0.0039, "step": 56270 }, { "grad_norm": 0.09076178073883057, "learning_rate": 4.376778889613056e-05, "loss": 0.0043, "step": 56280 }, { "grad_norm": 0.06870465725660324, "learning_rate": 4.375138348485831e-05, "loss": 0.0038, "step": 56290 }, { "grad_norm": 0.08313996344804764, "learning_rate": 4.3734978756925346e-05, "loss": 0.0028, "step": 56300 }, { "grad_norm": 0.08223023265600204, "learning_rate": 4.3718574714125706e-05, "loss": 0.0034, "step": 56310 }, { "grad_norm": 0.05168125778436661, "learning_rate": 4.370217135825329e-05, "loss": 0.0027, "step": 56320 }, { "grad_norm": 0.06197969987988472, "learning_rate": 4.368576869110194e-05, "loss": 0.0033, "step": 56330 }, { "grad_norm": 0.06710612773895264, "learning_rate": 4.366936671446544e-05, "loss": 0.0038, "step": 56340 }, { "grad_norm": 0.07233352214097977, "learning_rate": 4.3652965430137474e-05, "loss": 0.0032, "step": 56350 }, { "grad_norm": 0.09646148234605789, "learning_rate": 4.3636564839911646e-05, "loss": 0.0039, "step": 56360 }, { "grad_norm": 0.09172181040048599, "learning_rate": 4.3620164945581545e-05, "loss": 0.0038, "step": 56370 }, { "grad_norm": 0.07785206288099289, "learning_rate": 4.360376574894061e-05, "loss": 0.0044, "step": 56380 }, { "grad_norm": 0.06582823395729065, "learning_rate": 4.358736725178224e-05, "loss": 0.0036, "step": 56390 }, { "grad_norm": 0.08498860150575638, "learning_rate": 4.357096945589974e-05, "loss": 0.0037, "step": 56400 }, { "grad_norm": 0.07108853757381439, "learning_rate": 4.355457236308636e-05, "loss": 0.003, "step": 56410 }, { "grad_norm": 0.07899976521730423, "learning_rate": 4.353817597513526e-05, "loss": 0.003, "step": 56420 }, { "grad_norm": 0.07486854493618011, "learning_rate": 4.352178029383948e-05, "loss": 0.0035, "step": 56430 }, { "grad_norm": 0.0729677826166153, "learning_rate": 4.3505385320992105e-05, "loss": 0.0046, "step": 56440 }, { "grad_norm": 0.10239257663488388, "learning_rate": 4.348899105838602e-05, "loss": 0.003, "step": 56450 }, { "grad_norm": 0.08741395175457001, "learning_rate": 4.3472597507814087e-05, "loss": 0.0036, "step": 56460 }, { "grad_norm": 0.08579617738723755, "learning_rate": 4.3456204671069066e-05, "loss": 0.0032, "step": 56470 }, { "grad_norm": 0.11847221106290817, "learning_rate": 4.343981254994367e-05, "loss": 0.0037, "step": 56480 }, { "grad_norm": 0.09412801265716553, "learning_rate": 4.34234211462305e-05, "loss": 0.0042, "step": 56490 }, { "grad_norm": 0.06882603466510773, "learning_rate": 4.340703046172213e-05, "loss": 0.0034, "step": 56500 }, { "grad_norm": 0.0551702156662941, "learning_rate": 4.339064049821097e-05, "loss": 0.0031, "step": 56510 }, { "grad_norm": 0.0732126533985138, "learning_rate": 4.3374251257489446e-05, "loss": 0.0036, "step": 56520 }, { "grad_norm": 0.0999356135725975, "learning_rate": 4.335786274134984e-05, "loss": 0.0033, "step": 56530 }, { "grad_norm": 0.08900068700313568, "learning_rate": 4.334147495158438e-05, "loss": 0.0034, "step": 56540 }, { "grad_norm": 0.07748387008905411, "learning_rate": 4.33250878899852e-05, "loss": 0.0035, "step": 56550 }, { "grad_norm": 0.11060640215873718, "learning_rate": 4.330870155834437e-05, "loss": 0.0037, "step": 56560 }, { "grad_norm": 0.07935784757137299, "learning_rate": 4.3292315958453884e-05, "loss": 0.005, "step": 56570 }, { "grad_norm": 0.11879908293485641, "learning_rate": 4.3275931092105646e-05, "loss": 0.0041, "step": 56580 }, { "grad_norm": 0.0998736321926117, "learning_rate": 4.325954696109147e-05, "loss": 0.0037, "step": 56590 }, { "grad_norm": 0.06172450631856918, "learning_rate": 4.32431635672031e-05, "loss": 0.0045, "step": 56600 }, { "grad_norm": 0.08207995444536209, "learning_rate": 4.3226780912232214e-05, "loss": 0.0038, "step": 56610 }, { "grad_norm": 0.07502683997154236, "learning_rate": 4.321039899797036e-05, "loss": 0.0033, "step": 56620 }, { "grad_norm": 0.07008780539035797, "learning_rate": 4.319401782620908e-05, "loss": 0.0036, "step": 56630 }, { "grad_norm": 0.10965969413518906, "learning_rate": 4.317763739873978e-05, "loss": 0.004, "step": 56640 }, { "grad_norm": 0.09053909033536911, "learning_rate": 4.31612577173538e-05, "loss": 0.0035, "step": 56650 }, { "grad_norm": 0.09137100726366043, "learning_rate": 4.314487878384239e-05, "loss": 0.0034, "step": 56660 }, { "grad_norm": 0.09037616848945618, "learning_rate": 4.3128500599996736e-05, "loss": 0.0038, "step": 56670 }, { "grad_norm": 0.09961828589439392, "learning_rate": 4.31121231676079e-05, "loss": 0.0035, "step": 56680 }, { "grad_norm": 0.07100139558315277, "learning_rate": 4.309574648846694e-05, "loss": 0.0038, "step": 56690 }, { "grad_norm": 0.10290336608886719, "learning_rate": 4.3079370564364755e-05, "loss": 0.0039, "step": 56700 }, { "grad_norm": 0.10220795124769211, "learning_rate": 4.306299539709221e-05, "loss": 0.005, "step": 56710 }, { "grad_norm": 0.12702655792236328, "learning_rate": 4.3046620988440046e-05, "loss": 0.0049, "step": 56720 }, { "grad_norm": 0.06950409710407257, "learning_rate": 4.303024734019897e-05, "loss": 0.0032, "step": 56730 }, { "grad_norm": 0.08340219408273697, "learning_rate": 4.3013874454159565e-05, "loss": 0.0033, "step": 56740 }, { "grad_norm": 0.1029047816991806, "learning_rate": 4.299750233211233e-05, "loss": 0.0041, "step": 56750 }, { "grad_norm": 0.11342155933380127, "learning_rate": 4.2981130975847715e-05, "loss": 0.0039, "step": 56760 }, { "grad_norm": 0.10382033884525299, "learning_rate": 4.2964760387156076e-05, "loss": 0.0038, "step": 56770 }, { "grad_norm": 0.1406349092721939, "learning_rate": 4.294839056782767e-05, "loss": 0.0044, "step": 56780 }, { "grad_norm": 0.10135773569345474, "learning_rate": 4.293202151965266e-05, "loss": 0.0034, "step": 56790 }, { "grad_norm": 0.1009894534945488, "learning_rate": 4.291565324442115e-05, "loss": 0.0036, "step": 56800 }, { "grad_norm": 0.0833684429526329, "learning_rate": 4.2899285743923135e-05, "loss": 0.0045, "step": 56810 }, { "grad_norm": 0.09222082048654556, "learning_rate": 4.288291901994857e-05, "loss": 0.0039, "step": 56820 }, { "grad_norm": 0.09024064987897873, "learning_rate": 4.286655307428729e-05, "loss": 0.0037, "step": 56830 }, { "grad_norm": 0.0848073735833168, "learning_rate": 4.285018790872903e-05, "loss": 0.0053, "step": 56840 }, { "grad_norm": 0.07621670514345169, "learning_rate": 4.283382352506347e-05, "loss": 0.0044, "step": 56850 }, { "grad_norm": 0.08400134742259979, "learning_rate": 4.2817459925080205e-05, "loss": 0.0035, "step": 56860 }, { "grad_norm": 0.10649720579385757, "learning_rate": 4.28010971105687e-05, "loss": 0.0038, "step": 56870 }, { "grad_norm": 0.09187941253185272, "learning_rate": 4.278473508331841e-05, "loss": 0.0034, "step": 56880 }, { "grad_norm": 0.08993922919034958, "learning_rate": 4.276837384511864e-05, "loss": 0.0048, "step": 56890 }, { "grad_norm": 0.1106434240937233, "learning_rate": 4.275201339775864e-05, "loss": 0.0042, "step": 56900 }, { "grad_norm": 0.09744936972856522, "learning_rate": 4.273565374302756e-05, "loss": 0.0039, "step": 56910 }, { "grad_norm": 0.06924288719892502, "learning_rate": 4.2719294882714454e-05, "loss": 0.0033, "step": 56920 }, { "grad_norm": 0.06249582767486572, "learning_rate": 4.270293681860831e-05, "loss": 0.0031, "step": 56930 }, { "grad_norm": 0.09471689909696579, "learning_rate": 4.268657955249802e-05, "loss": 0.0034, "step": 56940 }, { "grad_norm": 0.08066849410533905, "learning_rate": 4.267022308617241e-05, "loss": 0.0035, "step": 56950 }, { "grad_norm": 0.07529762387275696, "learning_rate": 4.265386742142018e-05, "loss": 0.0042, "step": 56960 }, { "grad_norm": 0.0741463452577591, "learning_rate": 4.263751256002996e-05, "loss": 0.004, "step": 56970 }, { "grad_norm": 0.10363619029521942, "learning_rate": 4.26211585037903e-05, "loss": 0.0039, "step": 56980 }, { "grad_norm": 0.10764097422361374, "learning_rate": 4.2604805254489656e-05, "loss": 0.0048, "step": 56990 }, { "grad_norm": 0.10038163512945175, "learning_rate": 4.258845281391638e-05, "loss": 0.0038, "step": 57000 }, { "grad_norm": 0.08399511128664017, "learning_rate": 4.257210118385876e-05, "loss": 0.0037, "step": 57010 }, { "grad_norm": 0.11670993268489838, "learning_rate": 4.2555750366104994e-05, "loss": 0.0041, "step": 57020 }, { "grad_norm": 0.08346974104642868, "learning_rate": 4.2539400362443184e-05, "loss": 0.0051, "step": 57030 }, { "grad_norm": 0.09272891283035278, "learning_rate": 4.252305117466133e-05, "loss": 0.0038, "step": 57040 }, { "grad_norm": 0.06368190050125122, "learning_rate": 4.2506702804547356e-05, "loss": 0.0037, "step": 57050 }, { "grad_norm": 0.08814819902181625, "learning_rate": 4.249035525388907e-05, "loss": 0.0036, "step": 57060 }, { "grad_norm": 0.07788627594709396, "learning_rate": 4.247400852447427e-05, "loss": 0.0039, "step": 57070 }, { "grad_norm": 0.06244895979762077, "learning_rate": 4.245766261809059e-05, "loss": 0.0032, "step": 57080 }, { "grad_norm": 0.09055781364440918, "learning_rate": 4.244131753652558e-05, "loss": 0.0036, "step": 57090 }, { "grad_norm": 0.11707891523838043, "learning_rate": 4.242497328156672e-05, "loss": 0.0033, "step": 57100 }, { "grad_norm": 0.10039696842432022, "learning_rate": 4.240862985500139e-05, "loss": 0.0037, "step": 57110 }, { "grad_norm": 0.07625966519117355, "learning_rate": 4.2392287258616894e-05, "loss": 0.0043, "step": 57120 }, { "grad_norm": 0.07142354547977448, "learning_rate": 4.2375945494200396e-05, "loss": 0.0032, "step": 57130 }, { "grad_norm": 0.11046965420246124, "learning_rate": 4.2359604563539054e-05, "loss": 0.004, "step": 57140 }, { "grad_norm": 0.10534805804491043, "learning_rate": 4.2343264468419885e-05, "loss": 0.0033, "step": 57150 }, { "grad_norm": 0.09458820521831512, "learning_rate": 4.232692521062979e-05, "loss": 0.0036, "step": 57160 }, { "grad_norm": 0.08347959816455841, "learning_rate": 4.231058679195562e-05, "loss": 0.0034, "step": 57170 }, { "grad_norm": 0.06835857033729553, "learning_rate": 4.229424921418411e-05, "loss": 0.003, "step": 57180 }, { "grad_norm": 0.0809941291809082, "learning_rate": 4.2277912479101916e-05, "loss": 0.0037, "step": 57190 }, { "grad_norm": 0.08103533089160919, "learning_rate": 4.226157658849562e-05, "loss": 0.0036, "step": 57200 }, { "grad_norm": 0.09934934973716736, "learning_rate": 4.2245241544151664e-05, "loss": 0.0054, "step": 57210 }, { "grad_norm": 0.0781053751707077, "learning_rate": 4.2228907347856425e-05, "loss": 0.0038, "step": 57220 }, { "grad_norm": 0.07709316909313202, "learning_rate": 4.2212574001396214e-05, "loss": 0.0048, "step": 57230 }, { "grad_norm": 0.0762525275349617, "learning_rate": 4.21962415065572e-05, "loss": 0.004, "step": 57240 }, { "grad_norm": 0.07936063408851624, "learning_rate": 4.217990986512548e-05, "loss": 0.004, "step": 57250 }, { "grad_norm": 0.08960862457752228, "learning_rate": 4.216357907888704e-05, "loss": 0.0042, "step": 57260 }, { "grad_norm": 0.09109481424093246, "learning_rate": 4.2147249149627824e-05, "loss": 0.0038, "step": 57270 }, { "grad_norm": 0.06848018616437912, "learning_rate": 4.213092007913364e-05, "loss": 0.0037, "step": 57280 }, { "grad_norm": 0.091092050075531, "learning_rate": 4.2114591869190205e-05, "loss": 0.0042, "step": 57290 }, { "grad_norm": 0.09452757984399796, "learning_rate": 4.209826452158315e-05, "loss": 0.0037, "step": 57300 }, { "grad_norm": 0.06907941401004791, "learning_rate": 4.2081938038098e-05, "loss": 0.0031, "step": 57310 }, { "grad_norm": 0.10025936365127563, "learning_rate": 4.2065612420520185e-05, "loss": 0.0045, "step": 57320 }, { "grad_norm": 0.09222105890512466, "learning_rate": 4.204928767063509e-05, "loss": 0.0035, "step": 57330 }, { "grad_norm": 0.09899681061506271, "learning_rate": 4.2032963790227944e-05, "loss": 0.0039, "step": 57340 }, { "grad_norm": 0.1006614938378334, "learning_rate": 4.2016640781083887e-05, "loss": 0.0044, "step": 57350 }, { "grad_norm": 0.06889523565769196, "learning_rate": 4.2000318644988004e-05, "loss": 0.0031, "step": 57360 }, { "grad_norm": 0.07218343019485474, "learning_rate": 4.198399738372524e-05, "loss": 0.0034, "step": 57370 }, { "grad_norm": 0.10028836876153946, "learning_rate": 4.196767699908045e-05, "loss": 0.0042, "step": 57380 }, { "grad_norm": 0.09435636550188065, "learning_rate": 4.195135749283845e-05, "loss": 0.0043, "step": 57390 }, { "grad_norm": 0.08220209181308746, "learning_rate": 4.1935038866783894e-05, "loss": 0.0042, "step": 57400 }, { "grad_norm": 0.0953618586063385, "learning_rate": 4.1918721122701355e-05, "loss": 0.0034, "step": 57410 }, { "grad_norm": 0.06335665285587311, "learning_rate": 4.190240426237533e-05, "loss": 0.0035, "step": 57420 }, { "grad_norm": 0.07955645769834518, "learning_rate": 4.188608828759019e-05, "loss": 0.0029, "step": 57430 }, { "grad_norm": 0.0771542340517044, "learning_rate": 4.1869773200130246e-05, "loss": 0.0038, "step": 57440 }, { "grad_norm": 0.08436021953821182, "learning_rate": 4.185345900177967e-05, "loss": 0.0035, "step": 57450 }, { "grad_norm": 0.09674631059169769, "learning_rate": 4.183714569432258e-05, "loss": 0.003, "step": 57460 }, { "grad_norm": 0.08678678423166275, "learning_rate": 4.182083327954295e-05, "loss": 0.0039, "step": 57470 }, { "grad_norm": 0.052914682775735855, "learning_rate": 4.1804521759224714e-05, "loss": 0.0037, "step": 57480 }, { "grad_norm": 0.09882428497076035, "learning_rate": 4.178821113515165e-05, "loss": 0.003, "step": 57490 }, { "grad_norm": 0.08193545788526535, "learning_rate": 4.1771901409107476e-05, "loss": 0.004, "step": 57500 }, { "grad_norm": 0.08731462061405182, "learning_rate": 4.175559258287577e-05, "loss": 0.0029, "step": 57510 }, { "grad_norm": 0.08432766795158386, "learning_rate": 4.173928465824008e-05, "loss": 0.0037, "step": 57520 }, { "grad_norm": 0.08523610234260559, "learning_rate": 4.17229776369838e-05, "loss": 0.0046, "step": 57530 }, { "grad_norm": 0.09848365187644958, "learning_rate": 4.170667152089025e-05, "loss": 0.0046, "step": 57540 }, { "grad_norm": 0.09289197623729706, "learning_rate": 4.1690366311742626e-05, "loss": 0.005, "step": 57550 }, { "grad_norm": 0.09542620182037354, "learning_rate": 4.1674062011324055e-05, "loss": 0.0033, "step": 57560 }, { "grad_norm": 0.08553573489189148, "learning_rate": 4.165775862141752e-05, "loss": 0.0031, "step": 57570 }, { "grad_norm": 0.07587627321481705, "learning_rate": 4.164145614380598e-05, "loss": 0.0035, "step": 57580 }, { "grad_norm": 0.0936349555850029, "learning_rate": 4.162515458027223e-05, "loss": 0.0042, "step": 57590 }, { "grad_norm": 0.08375055342912674, "learning_rate": 4.160885393259899e-05, "loss": 0.0038, "step": 57600 }, { "grad_norm": 0.07339252531528473, "learning_rate": 4.159255420256886e-05, "loss": 0.003, "step": 57610 }, { "grad_norm": 0.0934084877371788, "learning_rate": 4.157625539196436e-05, "loss": 0.0037, "step": 57620 }, { "grad_norm": 0.08010369539260864, "learning_rate": 4.155995750256791e-05, "loss": 0.0037, "step": 57630 }, { "grad_norm": 0.08088808506727219, "learning_rate": 4.15436605361618e-05, "loss": 0.0039, "step": 57640 }, { "grad_norm": 0.09472044557332993, "learning_rate": 4.152736449452827e-05, "loss": 0.0035, "step": 57650 }, { "grad_norm": 0.07645395398139954, "learning_rate": 4.151106937944943e-05, "loss": 0.0033, "step": 57660 }, { "grad_norm": 0.07883808761835098, "learning_rate": 4.149477519270727e-05, "loss": 0.0042, "step": 57670 }, { "grad_norm": 0.06998027861118317, "learning_rate": 4.147848193608371e-05, "loss": 0.0048, "step": 57680 }, { "grad_norm": 0.0657302588224411, "learning_rate": 4.146218961136056e-05, "loss": 0.0033, "step": 57690 }, { "grad_norm": 0.10281287133693695, "learning_rate": 4.14458982203195e-05, "loss": 0.0045, "step": 57700 }, { "grad_norm": 0.06799323111772537, "learning_rate": 4.1429607764742165e-05, "loss": 0.0043, "step": 57710 }, { "grad_norm": 0.10041530430316925, "learning_rate": 4.141331824641003e-05, "loss": 0.0046, "step": 57720 }, { "grad_norm": 0.07910066843032837, "learning_rate": 4.139702966710452e-05, "loss": 0.0031, "step": 57730 }, { "grad_norm": 0.07693018019199371, "learning_rate": 4.13807420286069e-05, "loss": 0.0035, "step": 57740 }, { "grad_norm": 0.08030074089765549, "learning_rate": 4.136445533269837e-05, "loss": 0.0031, "step": 57750 }, { "grad_norm": 0.10535610467195511, "learning_rate": 4.134816958116003e-05, "loss": 0.004, "step": 57760 }, { "grad_norm": 0.07000698149204254, "learning_rate": 4.133188477577283e-05, "loss": 0.0035, "step": 57770 }, { "grad_norm": 0.0919007733464241, "learning_rate": 4.131560091831771e-05, "loss": 0.0031, "step": 57780 }, { "grad_norm": 0.07841934263706207, "learning_rate": 4.1299318010575405e-05, "loss": 0.0041, "step": 57790 }, { "grad_norm": 0.10329439491033554, "learning_rate": 4.128303605432661e-05, "loss": 0.0041, "step": 57800 }, { "grad_norm": 0.08481249958276749, "learning_rate": 4.126675505135188e-05, "loss": 0.0035, "step": 57810 }, { "grad_norm": 0.09257574379444122, "learning_rate": 4.125047500343169e-05, "loss": 0.0032, "step": 57820 }, { "grad_norm": 0.10188530385494232, "learning_rate": 4.1234195912346376e-05, "loss": 0.0036, "step": 57830 }, { "grad_norm": 0.12044604867696762, "learning_rate": 4.1217917779876235e-05, "loss": 0.0036, "step": 57840 }, { "grad_norm": 0.08636553585529327, "learning_rate": 4.120164060780141e-05, "loss": 0.003, "step": 57850 }, { "grad_norm": 0.08612606674432755, "learning_rate": 4.1185364397901935e-05, "loss": 0.0032, "step": 57860 }, { "grad_norm": 0.09830431640148163, "learning_rate": 4.116908915195776e-05, "loss": 0.0036, "step": 57870 }, { "grad_norm": 0.06685566902160645, "learning_rate": 4.115281487174872e-05, "loss": 0.0041, "step": 57880 }, { "grad_norm": 0.07370919734239578, "learning_rate": 4.113654155905452e-05, "loss": 0.0033, "step": 57890 }, { "grad_norm": 0.09533438086509705, "learning_rate": 4.112026921565483e-05, "loss": 0.0051, "step": 57900 }, { "grad_norm": 0.07898341864347458, "learning_rate": 4.1103997843329147e-05, "loss": 0.0035, "step": 57910 }, { "grad_norm": 0.07587289810180664, "learning_rate": 4.108772744385689e-05, "loss": 0.003, "step": 57920 }, { "grad_norm": 0.12105610221624374, "learning_rate": 4.1071458019017346e-05, "loss": 0.0034, "step": 57930 }, { "grad_norm": 0.08667679131031036, "learning_rate": 4.1055189570589734e-05, "loss": 0.0046, "step": 57940 }, { "grad_norm": 0.07976721972227097, "learning_rate": 4.103892210035315e-05, "loss": 0.0041, "step": 57950 }, { "grad_norm": 0.09301341325044632, "learning_rate": 4.102265561008656e-05, "loss": 0.0036, "step": 57960 }, { "grad_norm": 0.07291140407323837, "learning_rate": 4.100639010156885e-05, "loss": 0.0028, "step": 57970 }, { "grad_norm": 0.07633375376462936, "learning_rate": 4.099012557657882e-05, "loss": 0.0035, "step": 57980 }, { "grad_norm": 0.0779988020658493, "learning_rate": 4.09738620368951e-05, "loss": 0.0038, "step": 57990 }, { "grad_norm": 0.08937787264585495, "learning_rate": 4.095759948429626e-05, "loss": 0.0033, "step": 58000 }, { "grad_norm": 0.07298244535923004, "learning_rate": 4.094133792056075e-05, "loss": 0.0044, "step": 58010 }, { "grad_norm": 0.11226065456867218, "learning_rate": 4.092507734746688e-05, "loss": 0.0036, "step": 58020 }, { "grad_norm": 0.08568040281534195, "learning_rate": 4.090881776679293e-05, "loss": 0.0032, "step": 58030 }, { "grad_norm": 0.0931430384516716, "learning_rate": 4.0892559180316994e-05, "loss": 0.0046, "step": 58040 }, { "grad_norm": 0.11013757437467575, "learning_rate": 4.08763015898171e-05, "loss": 0.0037, "step": 58050 }, { "grad_norm": 0.09895595163106918, "learning_rate": 4.0860044997071145e-05, "loss": 0.0035, "step": 58060 }, { "grad_norm": 0.0631137490272522, "learning_rate": 4.084378940385693e-05, "loss": 0.0039, "step": 58070 }, { "grad_norm": 0.1274215430021286, "learning_rate": 4.082753481195211e-05, "loss": 0.0035, "step": 58080 }, { "grad_norm": 0.08829212188720703, "learning_rate": 4.0811281223134324e-05, "loss": 0.0039, "step": 58090 }, { "grad_norm": 0.10780180245637894, "learning_rate": 4.0795028639181e-05, "loss": 0.004, "step": 58100 }, { "grad_norm": 0.07185806334018707, "learning_rate": 4.077877706186951e-05, "loss": 0.0039, "step": 58110 }, { "grad_norm": 0.08651009947061539, "learning_rate": 4.076252649297709e-05, "loss": 0.0034, "step": 58120 }, { "grad_norm": 0.0713714063167572, "learning_rate": 4.07462769342809e-05, "loss": 0.0027, "step": 58130 }, { "grad_norm": 0.07700909674167633, "learning_rate": 4.073002838755793e-05, "loss": 0.0035, "step": 58140 }, { "grad_norm": 0.0801481157541275, "learning_rate": 4.0713780854585115e-05, "loss": 0.0026, "step": 58150 }, { "grad_norm": 0.07151385396718979, "learning_rate": 4.069753433713927e-05, "loss": 0.0028, "step": 58160 }, { "grad_norm": 0.07377336174249649, "learning_rate": 4.068128883699709e-05, "loss": 0.0034, "step": 58170 }, { "grad_norm": 0.06580205261707306, "learning_rate": 4.0665044355935155e-05, "loss": 0.0034, "step": 58180 }, { "grad_norm": 0.10508006066083908, "learning_rate": 4.064880089572993e-05, "loss": 0.004, "step": 58190 }, { "grad_norm": 0.09843476116657257, "learning_rate": 4.063255845815779e-05, "loss": 0.0032, "step": 58200 }, { "grad_norm": 0.061519354581832886, "learning_rate": 4.0616317044994945e-05, "loss": 0.0031, "step": 58210 }, { "grad_norm": 0.0746728777885437, "learning_rate": 4.0600076658017585e-05, "loss": 0.0037, "step": 58220 }, { "grad_norm": 0.0936744287610054, "learning_rate": 4.0583837299001706e-05, "loss": 0.0034, "step": 58230 }, { "grad_norm": 0.07072034478187561, "learning_rate": 4.056759896972322e-05, "loss": 0.0035, "step": 58240 }, { "grad_norm": 0.07600240409374237, "learning_rate": 4.055136167195793e-05, "loss": 0.0033, "step": 58250 }, { "grad_norm": 0.0823509693145752, "learning_rate": 4.053512540748152e-05, "loss": 0.003, "step": 58260 }, { "grad_norm": 0.07089713215827942, "learning_rate": 4.051889017806955e-05, "loss": 0.0036, "step": 58270 }, { "grad_norm": 0.10772385448217392, "learning_rate": 4.0502655985497516e-05, "loss": 0.0035, "step": 58280 }, { "grad_norm": 0.07756500691175461, "learning_rate": 4.048642283154074e-05, "loss": 0.0039, "step": 58290 }, { "grad_norm": 0.08728501945734024, "learning_rate": 4.0470190717974464e-05, "loss": 0.0046, "step": 58300 }, { "grad_norm": 0.04921119660139084, "learning_rate": 4.045395964657381e-05, "loss": 0.003, "step": 58310 }, { "grad_norm": 0.06230512261390686, "learning_rate": 4.043772961911376e-05, "loss": 0.0027, "step": 58320 }, { "grad_norm": 0.09222838282585144, "learning_rate": 4.042150063736924e-05, "loss": 0.0031, "step": 58330 }, { "grad_norm": 0.07150889933109283, "learning_rate": 4.0405272703114976e-05, "loss": 0.0033, "step": 58340 }, { "grad_norm": 0.08709102869033813, "learning_rate": 4.038904581812569e-05, "loss": 0.0048, "step": 58350 }, { "grad_norm": 0.08079089969396591, "learning_rate": 4.03728199841759e-05, "loss": 0.0029, "step": 58360 }, { "grad_norm": 0.06801334768533707, "learning_rate": 4.0356595203040044e-05, "loss": 0.003, "step": 58370 }, { "grad_norm": 0.08362573385238647, "learning_rate": 4.0340371476492426e-05, "loss": 0.0032, "step": 58380 }, { "grad_norm": 0.08536825329065323, "learning_rate": 4.032414880630727e-05, "loss": 0.0028, "step": 58390 }, { "grad_norm": 0.076679527759552, "learning_rate": 4.0307927194258646e-05, "loss": 0.0039, "step": 58400 }, { "grad_norm": 0.10312729328870773, "learning_rate": 4.029170664212054e-05, "loss": 0.0033, "step": 58410 }, { "grad_norm": 0.10804728418588638, "learning_rate": 4.0275487151666794e-05, "loss": 0.0029, "step": 58420 }, { "grad_norm": 0.08802998065948486, "learning_rate": 4.025926872467116e-05, "loss": 0.0039, "step": 58430 }, { "grad_norm": 0.08327410370111465, "learning_rate": 4.024305136290725e-05, "loss": 0.004, "step": 58440 }, { "grad_norm": 0.08879310637712479, "learning_rate": 4.022683506814858e-05, "loss": 0.0031, "step": 58450 }, { "grad_norm": 0.08159051835536957, "learning_rate": 4.021061984216853e-05, "loss": 0.0038, "step": 58460 }, { "grad_norm": 0.08548089861869812, "learning_rate": 4.019440568674035e-05, "loss": 0.0029, "step": 58470 }, { "grad_norm": 0.09273149073123932, "learning_rate": 4.017819260363724e-05, "loss": 0.0035, "step": 58480 }, { "grad_norm": 0.10349368304014206, "learning_rate": 4.016198059463223e-05, "loss": 0.0036, "step": 58490 }, { "grad_norm": 0.08721016347408295, "learning_rate": 4.0145769661498233e-05, "loss": 0.0035, "step": 58500 }, { "grad_norm": 0.08983748406171799, "learning_rate": 4.012955980600804e-05, "loss": 0.0038, "step": 58510 }, { "grad_norm": 0.09293725341558456, "learning_rate": 4.0113351029934355e-05, "loss": 0.0036, "step": 58520 }, { "grad_norm": 0.11599914729595184, "learning_rate": 4.0097143335049714e-05, "loss": 0.0043, "step": 58530 }, { "grad_norm": 0.07245860993862152, "learning_rate": 4.0080936723126606e-05, "loss": 0.0031, "step": 58540 }, { "grad_norm": 0.08492404222488403, "learning_rate": 4.006473119593735e-05, "loss": 0.003, "step": 58550 }, { "grad_norm": 0.07083405554294586, "learning_rate": 4.004852675525415e-05, "loss": 0.0031, "step": 58560 }, { "grad_norm": 0.09019500017166138, "learning_rate": 4.00323234028491e-05, "loss": 0.0044, "step": 58570 }, { "grad_norm": 0.09931345283985138, "learning_rate": 4.0016121140494186e-05, "loss": 0.0036, "step": 58580 }, { "grad_norm": 0.08698663860559464, "learning_rate": 3.999991996996122e-05, "loss": 0.0031, "step": 58590 }, { "grad_norm": 0.07993793487548828, "learning_rate": 3.9983719893022e-05, "loss": 0.0035, "step": 58600 }, { "grad_norm": 0.08056613802909851, "learning_rate": 3.996752091144812e-05, "loss": 0.0035, "step": 58610 }, { "grad_norm": 0.07098864018917084, "learning_rate": 3.9951323027011055e-05, "loss": 0.003, "step": 58620 }, { "grad_norm": 0.06711789965629578, "learning_rate": 3.993512624148219e-05, "loss": 0.0034, "step": 58630 }, { "grad_norm": 0.07837193459272385, "learning_rate": 3.99189305566328e-05, "loss": 0.0031, "step": 58640 }, { "grad_norm": 0.07074619084596634, "learning_rate": 3.990273597423401e-05, "loss": 0.0039, "step": 58650 }, { "grad_norm": 0.1164625957608223, "learning_rate": 3.988654249605681e-05, "loss": 0.0045, "step": 58660 }, { "grad_norm": 0.10879531502723694, "learning_rate": 3.987035012387213e-05, "loss": 0.0037, "step": 58670 }, { "grad_norm": 0.08306138962507248, "learning_rate": 3.9854158859450735e-05, "loss": 0.0032, "step": 58680 }, { "grad_norm": 0.07170221209526062, "learning_rate": 3.983796870456328e-05, "loss": 0.0033, "step": 58690 }, { "grad_norm": 0.07783384621143341, "learning_rate": 3.982177966098028e-05, "loss": 0.0039, "step": 58700 }, { "grad_norm": 0.07360498607158661, "learning_rate": 3.980559173047216e-05, "loss": 0.0036, "step": 58710 }, { "grad_norm": 0.10003930330276489, "learning_rate": 3.9789404914809176e-05, "loss": 0.0039, "step": 58720 }, { "grad_norm": 0.08022746443748474, "learning_rate": 3.9773219215761544e-05, "loss": 0.0039, "step": 58730 }, { "grad_norm": 0.09142697602510452, "learning_rate": 3.975703463509928e-05, "loss": 0.0034, "step": 58740 }, { "grad_norm": 0.07827746123075485, "learning_rate": 3.974085117459231e-05, "loss": 0.003, "step": 58750 }, { "grad_norm": 0.08705619722604752, "learning_rate": 3.972466883601043e-05, "loss": 0.0042, "step": 58760 }, { "grad_norm": 0.10592952370643616, "learning_rate": 3.9708487621123316e-05, "loss": 0.0036, "step": 58770 }, { "grad_norm": 0.08858661353588104, "learning_rate": 3.969230753170051e-05, "loss": 0.0032, "step": 58780 }, { "grad_norm": 0.09509039670228958, "learning_rate": 3.967612856951146e-05, "loss": 0.0044, "step": 58790 }, { "grad_norm": 0.08694383502006531, "learning_rate": 3.9659950736325476e-05, "loss": 0.0038, "step": 58800 }, { "grad_norm": 0.07656620442867279, "learning_rate": 3.964377403391174e-05, "loss": 0.003, "step": 58810 }, { "grad_norm": 0.0789785236120224, "learning_rate": 3.9627598464039303e-05, "loss": 0.0039, "step": 58820 }, { "grad_norm": 0.09552274644374847, "learning_rate": 3.9611424028477096e-05, "loss": 0.0039, "step": 58830 }, { "grad_norm": 0.05709819123148918, "learning_rate": 3.959525072899394e-05, "loss": 0.0044, "step": 58840 }, { "grad_norm": 0.0776544064283371, "learning_rate": 3.9579078567358506e-05, "loss": 0.0029, "step": 58850 }, { "grad_norm": 0.06445219367742538, "learning_rate": 3.956290754533939e-05, "loss": 0.0032, "step": 58860 }, { "grad_norm": 0.06689225137233734, "learning_rate": 3.954673766470501e-05, "loss": 0.0031, "step": 58870 }, { "grad_norm": 0.09418769180774689, "learning_rate": 3.9530568927223674e-05, "loss": 0.0036, "step": 58880 }, { "grad_norm": 0.09692582488059998, "learning_rate": 3.95144013346636e-05, "loss": 0.0036, "step": 58890 }, { "grad_norm": 0.08494085818529129, "learning_rate": 3.949823488879282e-05, "loss": 0.0035, "step": 58900 }, { "grad_norm": 0.07653439790010452, "learning_rate": 3.9482069591379276e-05, "loss": 0.0028, "step": 58910 }, { "grad_norm": 0.07367721945047379, "learning_rate": 3.9465905444190784e-05, "loss": 0.0028, "step": 58920 }, { "grad_norm": 0.09624603390693665, "learning_rate": 3.944974244899505e-05, "loss": 0.0033, "step": 58930 }, { "grad_norm": 0.08687976002693176, "learning_rate": 3.9433580607559616e-05, "loss": 0.0032, "step": 58940 }, { "grad_norm": 0.12719641625881195, "learning_rate": 3.941741992165193e-05, "loss": 0.0035, "step": 58950 }, { "grad_norm": 0.09162414073944092, "learning_rate": 3.940126039303928e-05, "loss": 0.0036, "step": 58960 }, { "grad_norm": 0.07873454689979553, "learning_rate": 3.938510202348886e-05, "loss": 0.0039, "step": 58970 }, { "grad_norm": 0.06701270490884781, "learning_rate": 3.93689448147677e-05, "loss": 0.0032, "step": 58980 }, { "grad_norm": 0.11679007858037949, "learning_rate": 3.935278876864278e-05, "loss": 0.0035, "step": 58990 }, { "grad_norm": 0.11626604199409485, "learning_rate": 3.933663388688087e-05, "loss": 0.0044, "step": 59000 }, { "grad_norm": 0.08463343977928162, "learning_rate": 3.9320480171248645e-05, "loss": 0.0036, "step": 59010 }, { "grad_norm": 0.07571269571781158, "learning_rate": 3.930432762351265e-05, "loss": 0.003, "step": 59020 }, { "grad_norm": 0.08093293756246567, "learning_rate": 3.928817624543931e-05, "loss": 0.0039, "step": 59030 }, { "grad_norm": 0.08428289741277695, "learning_rate": 3.9272026038794885e-05, "loss": 0.0034, "step": 59040 }, { "grad_norm": 0.07237762212753296, "learning_rate": 3.9255877005345586e-05, "loss": 0.0028, "step": 59050 }, { "grad_norm": 0.08783337473869324, "learning_rate": 3.923972914685743e-05, "loss": 0.0029, "step": 59060 }, { "grad_norm": 0.07393727451562881, "learning_rate": 3.922358246509631e-05, "loss": 0.0031, "step": 59070 }, { "grad_norm": 0.08714951574802399, "learning_rate": 3.920743696182801e-05, "loss": 0.0029, "step": 59080 }, { "grad_norm": 0.09970152378082275, "learning_rate": 3.9191292638818164e-05, "loss": 0.0037, "step": 59090 }, { "grad_norm": 0.09850611537694931, "learning_rate": 3.9175149497832316e-05, "loss": 0.0045, "step": 59100 }, { "grad_norm": 0.06412645429372787, "learning_rate": 3.915900754063584e-05, "loss": 0.0041, "step": 59110 }, { "grad_norm": 0.08118822425603867, "learning_rate": 3.9142866768994014e-05, "loss": 0.003, "step": 59120 }, { "grad_norm": 0.10360510647296906, "learning_rate": 3.912672718467193e-05, "loss": 0.0041, "step": 59130 }, { "grad_norm": 0.07588206231594086, "learning_rate": 3.911058878943463e-05, "loss": 0.0031, "step": 59140 }, { "grad_norm": 0.08636581152677536, "learning_rate": 3.9094451585046974e-05, "loss": 0.0031, "step": 59150 }, { "grad_norm": 0.08532586693763733, "learning_rate": 3.90783155732737e-05, "loss": 0.0032, "step": 59160 }, { "grad_norm": 0.07399067282676697, "learning_rate": 3.9062180755879373e-05, "loss": 0.0028, "step": 59170 }, { "grad_norm": 0.08305728435516357, "learning_rate": 3.904604713462855e-05, "loss": 0.0049, "step": 59180 }, { "grad_norm": 0.09659472852945328, "learning_rate": 3.902991471128554e-05, "loss": 0.0036, "step": 59190 }, { "grad_norm": 0.12181113660335541, "learning_rate": 3.901378348761456e-05, "loss": 0.0036, "step": 59200 }, { "grad_norm": 0.08735824376344681, "learning_rate": 3.899765346537971e-05, "loss": 0.0035, "step": 59210 }, { "grad_norm": 0.08333611488342285, "learning_rate": 3.898152464634493e-05, "loss": 0.0035, "step": 59220 }, { "grad_norm": 0.12087668478488922, "learning_rate": 3.896539703227402e-05, "loss": 0.0052, "step": 59230 }, { "grad_norm": 0.09167470037937164, "learning_rate": 3.894927062493073e-05, "loss": 0.0035, "step": 59240 }, { "grad_norm": 0.10151921212673187, "learning_rate": 3.893314542607859e-05, "loss": 0.0043, "step": 59250 }, { "grad_norm": 0.06975199282169342, "learning_rate": 3.891702143748103e-05, "loss": 0.0031, "step": 59260 }, { "grad_norm": 0.09144353121519089, "learning_rate": 3.890089866090133e-05, "loss": 0.0034, "step": 59270 }, { "grad_norm": 0.08933620154857635, "learning_rate": 3.888477709810267e-05, "loss": 0.003, "step": 59280 }, { "grad_norm": 0.0699104368686676, "learning_rate": 3.886865675084806e-05, "loss": 0.0044, "step": 59290 }, { "grad_norm": 0.10919585824012756, "learning_rate": 3.885253762090043e-05, "loss": 0.0034, "step": 59300 }, { "grad_norm": 0.08664034307003021, "learning_rate": 3.8836419710022534e-05, "loss": 0.0043, "step": 59310 }, { "grad_norm": 0.0937667191028595, "learning_rate": 3.882030301997698e-05, "loss": 0.0035, "step": 59320 }, { "grad_norm": 0.09236972033977509, "learning_rate": 3.880418755252629e-05, "loss": 0.004, "step": 59330 }, { "grad_norm": 0.0701519250869751, "learning_rate": 3.87880733094328e-05, "loss": 0.0026, "step": 59340 }, { "grad_norm": 0.08408821374177933, "learning_rate": 3.877196029245877e-05, "loss": 0.0029, "step": 59350 }, { "grad_norm": 0.08009960502386093, "learning_rate": 3.875584850336627e-05, "loss": 0.004, "step": 59360 }, { "grad_norm": 0.06123427301645279, "learning_rate": 3.873973794391728e-05, "loss": 0.0034, "step": 59370 }, { "grad_norm": 0.08178939670324326, "learning_rate": 3.872362861587361e-05, "loss": 0.0031, "step": 59380 }, { "grad_norm": 0.0994231253862381, "learning_rate": 3.870752052099698e-05, "loss": 0.0034, "step": 59390 }, { "grad_norm": 0.08751550316810608, "learning_rate": 3.8691413661048916e-05, "loss": 0.0031, "step": 59400 }, { "grad_norm": 0.08009553700685501, "learning_rate": 3.867530803779085e-05, "loss": 0.0034, "step": 59410 }, { "grad_norm": 0.09538479894399643, "learning_rate": 3.8659203652984054e-05, "loss": 0.0037, "step": 59420 }, { "grad_norm": 0.07948645204305649, "learning_rate": 3.864310050838972e-05, "loss": 0.0031, "step": 59430 }, { "grad_norm": 0.08957983553409576, "learning_rate": 3.8626998605768835e-05, "loss": 0.0035, "step": 59440 }, { "grad_norm": 0.08702249825000763, "learning_rate": 3.8610897946882294e-05, "loss": 0.0028, "step": 59450 }, { "grad_norm": 0.08464992046356201, "learning_rate": 3.859479853349082e-05, "loss": 0.004, "step": 59460 }, { "grad_norm": 0.08801723271608353, "learning_rate": 3.857870036735503e-05, "loss": 0.0034, "step": 59470 }, { "grad_norm": 0.0775417611002922, "learning_rate": 3.856260345023538e-05, "loss": 0.0033, "step": 59480 }, { "grad_norm": 0.10044167935848236, "learning_rate": 3.854650778389224e-05, "loss": 0.004, "step": 59490 }, { "grad_norm": 0.07793110609054565, "learning_rate": 3.8530413370085786e-05, "loss": 0.0027, "step": 59500 }, { "grad_norm": 0.07759585231542587, "learning_rate": 3.851432021057608e-05, "loss": 0.0029, "step": 59510 }, { "grad_norm": 0.09512144327163696, "learning_rate": 3.849822830712306e-05, "loss": 0.0031, "step": 59520 }, { "grad_norm": 0.09327305108308792, "learning_rate": 3.848213766148649e-05, "loss": 0.003, "step": 59530 }, { "grad_norm": 0.07963917404413223, "learning_rate": 3.846604827542602e-05, "loss": 0.0034, "step": 59540 }, { "grad_norm": 0.08687444776296616, "learning_rate": 3.844996015070117e-05, "loss": 0.0035, "step": 59550 }, { "grad_norm": 0.09256397187709808, "learning_rate": 3.843387328907132e-05, "loss": 0.0034, "step": 59560 }, { "grad_norm": 0.07962992042303085, "learning_rate": 3.8417787692295696e-05, "loss": 0.0053, "step": 59570 }, { "grad_norm": 0.10762570798397064, "learning_rate": 3.84017033621334e-05, "loss": 0.003, "step": 59580 }, { "grad_norm": 0.09251861274242401, "learning_rate": 3.838562030034338e-05, "loss": 0.0034, "step": 59590 }, { "grad_norm": 0.08029194921255112, "learning_rate": 3.836953850868447e-05, "loss": 0.0035, "step": 59600 }, { "grad_norm": 0.08475454151630402, "learning_rate": 3.835345798891533e-05, "loss": 0.0034, "step": 59610 }, { "grad_norm": 0.05742347240447998, "learning_rate": 3.833737874279452e-05, "loss": 0.0038, "step": 59620 }, { "grad_norm": 0.0863957554101944, "learning_rate": 3.8321300772080433e-05, "loss": 0.0032, "step": 59630 }, { "grad_norm": 0.06486199051141739, "learning_rate": 3.8305224078531345e-05, "loss": 0.0029, "step": 59640 }, { "grad_norm": 0.08323030918836594, "learning_rate": 3.828914866390537e-05, "loss": 0.0036, "step": 59650 }, { "grad_norm": 0.07175807654857635, "learning_rate": 3.827307452996048e-05, "loss": 0.003, "step": 59660 }, { "grad_norm": 0.10542234778404236, "learning_rate": 3.825700167845454e-05, "loss": 0.0037, "step": 59670 }, { "grad_norm": 0.07510074228048325, "learning_rate": 3.8240930111145214e-05, "loss": 0.0029, "step": 59680 }, { "grad_norm": 0.12464327365159988, "learning_rate": 3.8224859829790116e-05, "loss": 0.0043, "step": 59690 }, { "grad_norm": 0.10414893925189972, "learning_rate": 3.820879083614664e-05, "loss": 0.0031, "step": 59700 }, { "grad_norm": 0.08984380215406418, "learning_rate": 3.819272313197208e-05, "loss": 0.0033, "step": 59710 }, { "grad_norm": 0.08391398191452026, "learning_rate": 3.817665671902355e-05, "loss": 0.0032, "step": 59720 }, { "grad_norm": 0.08176839351654053, "learning_rate": 3.816059159905808e-05, "loss": 0.0042, "step": 59730 }, { "grad_norm": 0.11409042775630951, "learning_rate": 3.8144527773832476e-05, "loss": 0.0038, "step": 59740 }, { "grad_norm": 0.08450039476156235, "learning_rate": 3.812846524510352e-05, "loss": 0.0031, "step": 59750 }, { "grad_norm": 0.11093487590551376, "learning_rate": 3.811240401462775e-05, "loss": 0.0036, "step": 59760 }, { "grad_norm": 0.07063011080026627, "learning_rate": 3.809634408416162e-05, "loss": 0.0035, "step": 59770 }, { "grad_norm": 0.10786901414394379, "learning_rate": 3.8080285455461394e-05, "loss": 0.0035, "step": 59780 }, { "grad_norm": 0.10755554586648941, "learning_rate": 3.806422813028323e-05, "loss": 0.0035, "step": 59790 }, { "grad_norm": 0.06534074991941452, "learning_rate": 3.8048172110383114e-05, "loss": 0.0031, "step": 59800 }, { "grad_norm": 0.08734267950057983, "learning_rate": 3.8032117397516944e-05, "loss": 0.0041, "step": 59810 }, { "grad_norm": 0.09111757576465607, "learning_rate": 3.801606399344042e-05, "loss": 0.0034, "step": 59820 }, { "grad_norm": 0.07129298895597458, "learning_rate": 3.800001189990913e-05, "loss": 0.003, "step": 59830 }, { "grad_norm": 0.09590810537338257, "learning_rate": 3.798396111867847e-05, "loss": 0.0045, "step": 59840 }, { "grad_norm": 0.07938768714666367, "learning_rate": 3.796791165150378e-05, "loss": 0.0028, "step": 59850 }, { "grad_norm": 0.09468654543161392, "learning_rate": 3.795186350014018e-05, "loss": 0.0041, "step": 59860 }, { "grad_norm": 0.1069919690489769, "learning_rate": 3.793581666634266e-05, "loss": 0.0041, "step": 59870 }, { "grad_norm": 0.08739949017763138, "learning_rate": 3.791977115186609e-05, "loss": 0.0038, "step": 59880 }, { "grad_norm": 0.11805036664009094, "learning_rate": 3.79037269584652e-05, "loss": 0.0031, "step": 59890 }, { "grad_norm": 0.08794800937175751, "learning_rate": 3.7887684087894545e-05, "loss": 0.003, "step": 59900 }, { "grad_norm": 0.10160800069570541, "learning_rate": 3.787164254190854e-05, "loss": 0.0034, "step": 59910 }, { "grad_norm": 0.08573426306247711, "learning_rate": 3.7855602322261473e-05, "loss": 0.0031, "step": 59920 }, { "grad_norm": 0.0770532563328743, "learning_rate": 3.783956343070746e-05, "loss": 0.0035, "step": 59930 }, { "grad_norm": 0.08864621818065643, "learning_rate": 3.7823525869000526e-05, "loss": 0.0033, "step": 59940 }, { "grad_norm": 0.09202350676059723, "learning_rate": 3.78074896388945e-05, "loss": 0.0033, "step": 59950 }, { "grad_norm": 0.10006053745746613, "learning_rate": 3.7791454742143075e-05, "loss": 0.0034, "step": 59960 }, { "grad_norm": 0.09227731823921204, "learning_rate": 3.7775421180499795e-05, "loss": 0.0035, "step": 59970 }, { "grad_norm": 0.0974658876657486, "learning_rate": 3.7759388955718086e-05, "loss": 0.0036, "step": 59980 }, { "grad_norm": 0.07734023779630661, "learning_rate": 3.774335806955117e-05, "loss": 0.0031, "step": 59990 }, { "grad_norm": 0.08346842229366302, "learning_rate": 3.7727328523752215e-05, "loss": 0.0032, "step": 60000 } ], "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 }