{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.2, "eval_steps": 500, "global_step": 50000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 5.461854457855225, "learning_rate": 1.8e-07, "loss": 1.309, "step": 10 }, { "grad_norm": 4.280724048614502, "learning_rate": 3.8e-07, "loss": 1.2899, "step": 20 }, { "grad_norm": 2.995441198348999, "learning_rate": 5.8e-07, "loss": 1.2592, "step": 30 }, { "grad_norm": 2.0369226932525635, "learning_rate": 7.8e-07, "loss": 1.2103, "step": 40 }, { "grad_norm": 1.7288485765457153, "learning_rate": 9.8e-07, "loss": 1.1564, "step": 50 }, { "grad_norm": 4.792635917663574, "learning_rate": 1.18e-06, "loss": 1.1181, "step": 60 }, { "grad_norm": 0.9400705099105835, "learning_rate": 1.3800000000000001e-06, "loss": 1.1151, "step": 70 }, { "grad_norm": 1.0534231662750244, "learning_rate": 1.5800000000000003e-06, "loss": 1.0913, "step": 80 }, { "grad_norm": 0.8955241441726685, "learning_rate": 1.7800000000000001e-06, "loss": 1.0767, "step": 90 }, { "grad_norm": 1.005018949508667, "learning_rate": 1.98e-06, "loss": 1.0585, "step": 100 }, { "grad_norm": 0.920587956905365, "learning_rate": 2.1800000000000003e-06, "loss": 1.0681, "step": 110 }, { "grad_norm": 0.9109120965003967, "learning_rate": 2.38e-06, "loss": 1.0504, "step": 120 }, { "grad_norm": 1.1045085191726685, "learning_rate": 2.5800000000000003e-06, "loss": 1.052, "step": 130 }, { "grad_norm": 0.9113557934761047, "learning_rate": 2.78e-06, "loss": 1.0561, "step": 140 }, { "grad_norm": 0.9425560235977173, "learning_rate": 2.9800000000000003e-06, "loss": 1.0493, "step": 150 }, { "grad_norm": 0.9836668372154236, "learning_rate": 3.1800000000000005e-06, "loss": 1.046, "step": 160 }, { "grad_norm": 0.8569234013557434, "learning_rate": 3.38e-06, "loss": 1.0483, "step": 170 }, { "grad_norm": 1.0321577787399292, "learning_rate": 3.58e-06, "loss": 1.0452, "step": 180 }, { "grad_norm": 0.8527000546455383, "learning_rate": 3.7800000000000002e-06, "loss": 1.0508, "step": 190 }, { "grad_norm": 0.9701483845710754, "learning_rate": 3.98e-06, "loss": 1.0433, "step": 200 }, { "grad_norm": 0.9698002934455872, "learning_rate": 4.18e-06, "loss": 1.033, "step": 210 }, { "grad_norm": 0.985220730304718, "learning_rate": 4.38e-06, "loss": 1.0459, "step": 220 }, { "grad_norm": 0.9473835229873657, "learning_rate": 4.58e-06, "loss": 1.0395, "step": 230 }, { "grad_norm": 1.0000640153884888, "learning_rate": 4.780000000000001e-06, "loss": 1.038, "step": 240 }, { "grad_norm": 0.836566150188446, "learning_rate": 4.98e-06, "loss": 1.038, "step": 250 }, { "grad_norm": 0.885728657245636, "learning_rate": 5.18e-06, "loss": 1.0319, "step": 260 }, { "grad_norm": 0.9157370924949646, "learning_rate": 5.38e-06, "loss": 1.0233, "step": 270 }, { "grad_norm": 0.884800910949707, "learning_rate": 5.580000000000001e-06, "loss": 1.0396, "step": 280 }, { "grad_norm": 1.054032325744629, "learning_rate": 5.78e-06, "loss": 1.036, "step": 290 }, { "grad_norm": 0.8666252493858337, "learning_rate": 5.98e-06, "loss": 1.0362, "step": 300 }, { "grad_norm": 0.9244698286056519, "learning_rate": 6.18e-06, "loss": 1.0251, "step": 310 }, { "grad_norm": 1.27436363697052, "learning_rate": 6.38e-06, "loss": 1.0159, "step": 320 }, { "grad_norm": 1.138055443763733, "learning_rate": 6.58e-06, "loss": 1.004, "step": 330 }, { "grad_norm": 1.2847328186035156, "learning_rate": 6.78e-06, "loss": 0.9781, "step": 340 }, { "grad_norm": 1.3764742612838745, "learning_rate": 6.98e-06, "loss": 0.958, "step": 350 }, { "grad_norm": 1.6649106740951538, "learning_rate": 7.180000000000001e-06, "loss": 0.9411, "step": 360 }, { "grad_norm": 2.054042100906372, "learning_rate": 7.3800000000000005e-06, "loss": 0.9097, "step": 370 }, { "grad_norm": 1.6054954528808594, "learning_rate": 7.580000000000001e-06, "loss": 0.8801, "step": 380 }, { "grad_norm": 1.6744186878204346, "learning_rate": 7.78e-06, "loss": 0.825, "step": 390 }, { "grad_norm": 1.4964332580566406, "learning_rate": 7.98e-06, "loss": 0.7936, "step": 400 }, { "grad_norm": 1.553642988204956, "learning_rate": 8.18e-06, "loss": 0.7639, "step": 410 }, { "grad_norm": 1.6326075792312622, "learning_rate": 8.380000000000001e-06, "loss": 0.7103, "step": 420 }, { "grad_norm": 1.900917649269104, "learning_rate": 8.580000000000001e-06, "loss": 0.684, "step": 430 }, { "grad_norm": 2.043154716491699, "learning_rate": 8.78e-06, "loss": 0.6591, "step": 440 }, { "grad_norm": 1.7525678873062134, "learning_rate": 8.98e-06, "loss": 0.6038, "step": 450 }, { "grad_norm": 1.853177785873413, "learning_rate": 9.180000000000002e-06, "loss": 0.5873, "step": 460 }, { "grad_norm": 1.9680418968200684, "learning_rate": 9.38e-06, "loss": 0.5609, "step": 470 }, { "grad_norm": 1.9680604934692383, "learning_rate": 9.58e-06, "loss": 0.5191, "step": 480 }, { "grad_norm": 2.0379209518432617, "learning_rate": 9.78e-06, "loss": 0.503, "step": 490 }, { "grad_norm": 1.9380511045455933, "learning_rate": 9.980000000000001e-06, "loss": 0.4659, "step": 500 }, { "grad_norm": 2.3489267826080322, "learning_rate": 1.018e-05, "loss": 0.4474, "step": 510 }, { "grad_norm": 2.119300127029419, "learning_rate": 1.038e-05, "loss": 0.4172, "step": 520 }, { "grad_norm": 2.5633950233459473, "learning_rate": 1.058e-05, "loss": 0.401, "step": 530 }, { "grad_norm": 2.1650376319885254, "learning_rate": 1.0780000000000002e-05, "loss": 0.3796, "step": 540 }, { "grad_norm": 2.766003370285034, "learning_rate": 1.098e-05, "loss": 0.3841, "step": 550 }, { "grad_norm": 2.683974027633667, "learning_rate": 1.118e-05, "loss": 0.3478, "step": 560 }, { "grad_norm": 2.364849090576172, "learning_rate": 1.1380000000000001e-05, "loss": 0.3389, "step": 570 }, { "grad_norm": 2.2881429195404053, "learning_rate": 1.1580000000000001e-05, "loss": 0.3252, "step": 580 }, { "grad_norm": 1.9716356992721558, "learning_rate": 1.178e-05, "loss": 0.289, "step": 590 }, { "grad_norm": 2.6261096000671387, "learning_rate": 1.198e-05, "loss": 0.2866, "step": 600 }, { "grad_norm": 2.4541890621185303, "learning_rate": 1.2180000000000002e-05, "loss": 0.266, "step": 610 }, { "grad_norm": 2.446408271789551, "learning_rate": 1.238e-05, "loss": 0.2728, "step": 620 }, { "grad_norm": 1.991907000541687, "learning_rate": 1.258e-05, "loss": 0.2153, "step": 630 }, { "grad_norm": 2.4080991744995117, "learning_rate": 1.278e-05, "loss": 0.2243, "step": 640 }, { "grad_norm": 2.1478254795074463, "learning_rate": 1.2980000000000001e-05, "loss": 0.2002, "step": 650 }, { "grad_norm": 1.9226510524749756, "learning_rate": 1.3180000000000001e-05, "loss": 0.2115, "step": 660 }, { "grad_norm": 1.904799222946167, "learning_rate": 1.338e-05, "loss": 0.2091, "step": 670 }, { "grad_norm": 2.0946967601776123, "learning_rate": 1.358e-05, "loss": 0.1885, "step": 680 }, { "grad_norm": 2.4359583854675293, "learning_rate": 1.3780000000000002e-05, "loss": 0.1849, "step": 690 }, { "grad_norm": 1.9525563716888428, "learning_rate": 1.3980000000000002e-05, "loss": 0.1896, "step": 700 }, { "grad_norm": 2.4340641498565674, "learning_rate": 1.4180000000000001e-05, "loss": 0.196, "step": 710 }, { "grad_norm": 2.195129871368408, "learning_rate": 1.4380000000000001e-05, "loss": 0.1798, "step": 720 }, { "grad_norm": 2.118159532546997, "learning_rate": 1.4580000000000003e-05, "loss": 0.1829, "step": 730 }, { "grad_norm": 2.1260716915130615, "learning_rate": 1.4779999999999999e-05, "loss": 0.172, "step": 740 }, { "grad_norm": 2.7699837684631348, "learning_rate": 1.4979999999999999e-05, "loss": 0.1795, "step": 750 }, { "grad_norm": 2.3933463096618652, "learning_rate": 1.518e-05, "loss": 0.1588, "step": 760 }, { "grad_norm": 1.9982023239135742, "learning_rate": 1.538e-05, "loss": 0.1734, "step": 770 }, { "grad_norm": 2.6957149505615234, "learning_rate": 1.558e-05, "loss": 0.1675, "step": 780 }, { "grad_norm": 1.4379185438156128, "learning_rate": 1.578e-05, "loss": 0.1711, "step": 790 }, { "grad_norm": 1.8235595226287842, "learning_rate": 1.598e-05, "loss": 0.1688, "step": 800 }, { "grad_norm": 2.6302709579467773, "learning_rate": 1.618e-05, "loss": 0.1798, "step": 810 }, { "grad_norm": 1.7259941101074219, "learning_rate": 1.6380000000000002e-05, "loss": 0.1615, "step": 820 }, { "grad_norm": 1.7755985260009766, "learning_rate": 1.658e-05, "loss": 0.1602, "step": 830 }, { "grad_norm": 1.933347225189209, "learning_rate": 1.6780000000000002e-05, "loss": 0.1475, "step": 840 }, { "grad_norm": 1.669323444366455, "learning_rate": 1.698e-05, "loss": 0.1492, "step": 850 }, { "grad_norm": 1.8184080123901367, "learning_rate": 1.718e-05, "loss": 0.1528, "step": 860 }, { "grad_norm": 2.374333143234253, "learning_rate": 1.7380000000000003e-05, "loss": 0.151, "step": 870 }, { "grad_norm": 2.104065179824829, "learning_rate": 1.758e-05, "loss": 0.1482, "step": 880 }, { "grad_norm": 2.213362693786621, "learning_rate": 1.7780000000000003e-05, "loss": 0.1633, "step": 890 }, { "grad_norm": 1.6739733219146729, "learning_rate": 1.798e-05, "loss": 0.1635, "step": 900 }, { "grad_norm": 2.366987466812134, "learning_rate": 1.818e-05, "loss": 0.1586, "step": 910 }, { "grad_norm": 2.2367897033691406, "learning_rate": 1.838e-05, "loss": 0.1444, "step": 920 }, { "grad_norm": 1.883253812789917, "learning_rate": 1.858e-05, "loss": 0.1495, "step": 930 }, { "grad_norm": 2.2315523624420166, "learning_rate": 1.878e-05, "loss": 0.1592, "step": 940 }, { "grad_norm": 1.9257094860076904, "learning_rate": 1.898e-05, "loss": 0.1545, "step": 950 }, { "grad_norm": 2.147770643234253, "learning_rate": 1.918e-05, "loss": 0.1335, "step": 960 }, { "grad_norm": 1.6349462270736694, "learning_rate": 1.938e-05, "loss": 0.1438, "step": 970 }, { "grad_norm": 1.9289028644561768, "learning_rate": 1.9580000000000002e-05, "loss": 0.1197, "step": 980 }, { "grad_norm": 2.259413480758667, "learning_rate": 1.978e-05, "loss": 0.1188, "step": 990 }, { "grad_norm": 1.4886430501937866, "learning_rate": 1.9980000000000002e-05, "loss": 0.1408, "step": 1000 }, { "grad_norm": 1.7439996004104614, "learning_rate": 2.0180000000000003e-05, "loss": 0.1182, "step": 1010 }, { "grad_norm": 2.019468069076538, "learning_rate": 2.038e-05, "loss": 0.1385, "step": 1020 }, { "grad_norm": 1.7428712844848633, "learning_rate": 2.0580000000000003e-05, "loss": 0.134, "step": 1030 }, { "grad_norm": 1.8469316959381104, "learning_rate": 2.078e-05, "loss": 0.1232, "step": 1040 }, { "grad_norm": 1.6272313594818115, "learning_rate": 2.098e-05, "loss": 0.1194, "step": 1050 }, { "grad_norm": 1.762843132019043, "learning_rate": 2.118e-05, "loss": 0.1311, "step": 1060 }, { "grad_norm": 1.628401517868042, "learning_rate": 2.138e-05, "loss": 0.125, "step": 1070 }, { "grad_norm": 1.617635726928711, "learning_rate": 2.158e-05, "loss": 0.1429, "step": 1080 }, { "grad_norm": 2.3543941974639893, "learning_rate": 2.178e-05, "loss": 0.1374, "step": 1090 }, { "grad_norm": 1.9333502054214478, "learning_rate": 2.198e-05, "loss": 0.1139, "step": 1100 }, { "grad_norm": 2.004859685897827, "learning_rate": 2.218e-05, "loss": 0.13, "step": 1110 }, { "grad_norm": 1.3976293802261353, "learning_rate": 2.2380000000000003e-05, "loss": 0.1207, "step": 1120 }, { "grad_norm": 1.605494499206543, "learning_rate": 2.258e-05, "loss": 0.1425, "step": 1130 }, { "grad_norm": 1.6937992572784424, "learning_rate": 2.2780000000000002e-05, "loss": 0.1339, "step": 1140 }, { "grad_norm": 1.7889230251312256, "learning_rate": 2.298e-05, "loss": 0.1217, "step": 1150 }, { "grad_norm": 1.8137965202331543, "learning_rate": 2.318e-05, "loss": 0.117, "step": 1160 }, { "grad_norm": 1.47572660446167, "learning_rate": 2.3380000000000003e-05, "loss": 0.1152, "step": 1170 }, { "grad_norm": 1.3732249736785889, "learning_rate": 2.358e-05, "loss": 0.1142, "step": 1180 }, { "grad_norm": 1.5794624090194702, "learning_rate": 2.3780000000000003e-05, "loss": 0.1141, "step": 1190 }, { "grad_norm": 2.0735294818878174, "learning_rate": 2.398e-05, "loss": 0.1297, "step": 1200 }, { "grad_norm": 1.5763579607009888, "learning_rate": 2.418e-05, "loss": 0.1078, "step": 1210 }, { "grad_norm": 1.7973313331604004, "learning_rate": 2.438e-05, "loss": 0.1141, "step": 1220 }, { "grad_norm": 1.830474615097046, "learning_rate": 2.4580000000000002e-05, "loss": 0.1214, "step": 1230 }, { "grad_norm": 1.6619874238967896, "learning_rate": 2.478e-05, "loss": 0.1106, "step": 1240 }, { "grad_norm": 1.6162927150726318, "learning_rate": 2.498e-05, "loss": 0.1075, "step": 1250 }, { "grad_norm": 1.5842337608337402, "learning_rate": 2.5180000000000003e-05, "loss": 0.117, "step": 1260 }, { "grad_norm": 1.8844553232192993, "learning_rate": 2.5380000000000004e-05, "loss": 0.1142, "step": 1270 }, { "grad_norm": 1.4789831638336182, "learning_rate": 2.5580000000000002e-05, "loss": 0.1065, "step": 1280 }, { "grad_norm": 1.6017956733703613, "learning_rate": 2.5779999999999997e-05, "loss": 0.1098, "step": 1290 }, { "grad_norm": 1.6552813053131104, "learning_rate": 2.598e-05, "loss": 0.1091, "step": 1300 }, { "grad_norm": 1.2914025783538818, "learning_rate": 2.618e-05, "loss": 0.1142, "step": 1310 }, { "grad_norm": 1.4783532619476318, "learning_rate": 2.6379999999999998e-05, "loss": 0.1216, "step": 1320 }, { "grad_norm": 1.6004480123519897, "learning_rate": 2.658e-05, "loss": 0.1096, "step": 1330 }, { "grad_norm": 1.5503908395767212, "learning_rate": 2.678e-05, "loss": 0.1093, "step": 1340 }, { "grad_norm": 1.4877166748046875, "learning_rate": 2.698e-05, "loss": 0.1045, "step": 1350 }, { "grad_norm": 1.6759634017944336, "learning_rate": 2.718e-05, "loss": 0.1053, "step": 1360 }, { "grad_norm": 1.4156335592269897, "learning_rate": 2.738e-05, "loss": 0.097, "step": 1370 }, { "grad_norm": 1.5876866579055786, "learning_rate": 2.758e-05, "loss": 0.1078, "step": 1380 }, { "grad_norm": 1.429563283920288, "learning_rate": 2.778e-05, "loss": 0.1029, "step": 1390 }, { "grad_norm": 1.7056858539581299, "learning_rate": 2.798e-05, "loss": 0.101, "step": 1400 }, { "grad_norm": 1.4645469188690186, "learning_rate": 2.818e-05, "loss": 0.1159, "step": 1410 }, { "grad_norm": 1.8153839111328125, "learning_rate": 2.8380000000000003e-05, "loss": 0.1203, "step": 1420 }, { "grad_norm": 1.3237626552581787, "learning_rate": 2.858e-05, "loss": 0.1065, "step": 1430 }, { "grad_norm": 1.2180428504943848, "learning_rate": 2.8780000000000002e-05, "loss": 0.1134, "step": 1440 }, { "grad_norm": 1.7182022333145142, "learning_rate": 2.898e-05, "loss": 0.0934, "step": 1450 }, { "grad_norm": 1.5111862421035767, "learning_rate": 2.9180000000000002e-05, "loss": 0.1004, "step": 1460 }, { "grad_norm": 1.364305019378662, "learning_rate": 2.9380000000000003e-05, "loss": 0.1038, "step": 1470 }, { "grad_norm": 1.4989819526672363, "learning_rate": 2.958e-05, "loss": 0.1109, "step": 1480 }, { "grad_norm": 1.4552531242370605, "learning_rate": 2.9780000000000003e-05, "loss": 0.1077, "step": 1490 }, { "grad_norm": 1.569528579711914, "learning_rate": 2.998e-05, "loss": 0.1053, "step": 1500 }, { "grad_norm": 1.1550108194351196, "learning_rate": 3.0180000000000002e-05, "loss": 0.1126, "step": 1510 }, { "grad_norm": 1.5073966979980469, "learning_rate": 3.0380000000000004e-05, "loss": 0.1055, "step": 1520 }, { "grad_norm": 1.3160218000411987, "learning_rate": 3.058e-05, "loss": 0.1028, "step": 1530 }, { "grad_norm": 1.4165984392166138, "learning_rate": 3.078e-05, "loss": 0.0995, "step": 1540 }, { "grad_norm": 1.2122802734375, "learning_rate": 3.0980000000000005e-05, "loss": 0.1036, "step": 1550 }, { "grad_norm": 1.6234776973724365, "learning_rate": 3.118e-05, "loss": 0.1034, "step": 1560 }, { "grad_norm": 1.5230448246002197, "learning_rate": 3.138e-05, "loss": 0.108, "step": 1570 }, { "grad_norm": 1.0402367115020752, "learning_rate": 3.1580000000000006e-05, "loss": 0.103, "step": 1580 }, { "grad_norm": 1.1498876810073853, "learning_rate": 3.1780000000000004e-05, "loss": 0.0993, "step": 1590 }, { "grad_norm": 1.3644670248031616, "learning_rate": 3.198e-05, "loss": 0.1017, "step": 1600 }, { "grad_norm": 1.188448190689087, "learning_rate": 3.218e-05, "loss": 0.104, "step": 1610 }, { "grad_norm": 1.314953327178955, "learning_rate": 3.238e-05, "loss": 0.0974, "step": 1620 }, { "grad_norm": 1.3997679948806763, "learning_rate": 3.2579999999999996e-05, "loss": 0.099, "step": 1630 }, { "grad_norm": 1.455701470375061, "learning_rate": 3.278e-05, "loss": 0.0948, "step": 1640 }, { "grad_norm": 1.1160516738891602, "learning_rate": 3.298e-05, "loss": 0.0955, "step": 1650 }, { "grad_norm": 1.3265810012817383, "learning_rate": 3.318e-05, "loss": 0.1005, "step": 1660 }, { "grad_norm": 1.2919566631317139, "learning_rate": 3.338e-05, "loss": 0.1051, "step": 1670 }, { "grad_norm": 1.0544244050979614, "learning_rate": 3.358e-05, "loss": 0.0942, "step": 1680 }, { "grad_norm": 1.7509288787841797, "learning_rate": 3.378e-05, "loss": 0.1051, "step": 1690 }, { "grad_norm": 1.319934606552124, "learning_rate": 3.398e-05, "loss": 0.0995, "step": 1700 }, { "grad_norm": 1.304781198501587, "learning_rate": 3.418e-05, "loss": 0.1082, "step": 1710 }, { "grad_norm": 1.015054702758789, "learning_rate": 3.438e-05, "loss": 0.1026, "step": 1720 }, { "grad_norm": 1.097072958946228, "learning_rate": 3.4580000000000004e-05, "loss": 0.1006, "step": 1730 }, { "grad_norm": 1.4994876384735107, "learning_rate": 3.478e-05, "loss": 0.0924, "step": 1740 }, { "grad_norm": 1.058549404144287, "learning_rate": 3.498e-05, "loss": 0.0992, "step": 1750 }, { "grad_norm": 1.5893909931182861, "learning_rate": 3.518e-05, "loss": 0.1006, "step": 1760 }, { "grad_norm": 1.4650343656539917, "learning_rate": 3.5380000000000003e-05, "loss": 0.0882, "step": 1770 }, { "grad_norm": 1.1829276084899902, "learning_rate": 3.558e-05, "loss": 0.0976, "step": 1780 }, { "grad_norm": 1.0093764066696167, "learning_rate": 3.578e-05, "loss": 0.0879, "step": 1790 }, { "grad_norm": 1.2170696258544922, "learning_rate": 3.5980000000000004e-05, "loss": 0.0904, "step": 1800 }, { "grad_norm": 1.315539836883545, "learning_rate": 3.618e-05, "loss": 0.0892, "step": 1810 }, { "grad_norm": 0.9792266488075256, "learning_rate": 3.638e-05, "loss": 0.0904, "step": 1820 }, { "grad_norm": 0.9297605156898499, "learning_rate": 3.6580000000000006e-05, "loss": 0.0881, "step": 1830 }, { "grad_norm": 1.233324408531189, "learning_rate": 3.6780000000000004e-05, "loss": 0.0888, "step": 1840 }, { "grad_norm": 1.0968163013458252, "learning_rate": 3.698e-05, "loss": 0.085, "step": 1850 }, { "grad_norm": 1.294609546661377, "learning_rate": 3.7180000000000007e-05, "loss": 0.0923, "step": 1860 }, { "grad_norm": 0.9666970372200012, "learning_rate": 3.7380000000000005e-05, "loss": 0.0899, "step": 1870 }, { "grad_norm": 0.9447898864746094, "learning_rate": 3.758e-05, "loss": 0.0794, "step": 1880 }, { "grad_norm": 1.0241414308547974, "learning_rate": 3.778000000000001e-05, "loss": 0.0929, "step": 1890 }, { "grad_norm": 0.9549765586853027, "learning_rate": 3.7980000000000006e-05, "loss": 0.0858, "step": 1900 }, { "grad_norm": 1.040447473526001, "learning_rate": 3.818e-05, "loss": 0.0852, "step": 1910 }, { "grad_norm": 0.848207950592041, "learning_rate": 3.838e-05, "loss": 0.0898, "step": 1920 }, { "grad_norm": 0.922381579875946, "learning_rate": 3.858e-05, "loss": 0.0937, "step": 1930 }, { "grad_norm": 1.341477394104004, "learning_rate": 3.878e-05, "loss": 0.0945, "step": 1940 }, { "grad_norm": 1.0362097024917603, "learning_rate": 3.898e-05, "loss": 0.0865, "step": 1950 }, { "grad_norm": 1.1032885313034058, "learning_rate": 3.918e-05, "loss": 0.0937, "step": 1960 }, { "grad_norm": 0.9449819326400757, "learning_rate": 3.938e-05, "loss": 0.0993, "step": 1970 }, { "grad_norm": 1.086362600326538, "learning_rate": 3.958e-05, "loss": 0.0915, "step": 1980 }, { "grad_norm": 1.3857746124267578, "learning_rate": 3.978e-05, "loss": 0.0914, "step": 1990 }, { "grad_norm": 1.2797060012817383, "learning_rate": 3.998e-05, "loss": 0.1034, "step": 2000 }, { "grad_norm": 0.8514531850814819, "learning_rate": 4.018e-05, "loss": 0.087, "step": 2010 }, { "grad_norm": 1.0628405809402466, "learning_rate": 4.038e-05, "loss": 0.0926, "step": 2020 }, { "grad_norm": 1.1095056533813477, "learning_rate": 4.058e-05, "loss": 0.0955, "step": 2030 }, { "grad_norm": 0.9962992668151855, "learning_rate": 4.078e-05, "loss": 0.0955, "step": 2040 }, { "grad_norm": 1.043080449104309, "learning_rate": 4.0980000000000004e-05, "loss": 0.0908, "step": 2050 }, { "grad_norm": 1.097784161567688, "learning_rate": 4.118e-05, "loss": 0.102, "step": 2060 }, { "grad_norm": 1.1784327030181885, "learning_rate": 4.138e-05, "loss": 0.0956, "step": 2070 }, { "grad_norm": 1.0643510818481445, "learning_rate": 4.1580000000000005e-05, "loss": 0.0872, "step": 2080 }, { "grad_norm": 0.8668339848518372, "learning_rate": 4.178e-05, "loss": 0.0864, "step": 2090 }, { "grad_norm": 0.7604209184646606, "learning_rate": 4.198e-05, "loss": 0.0837, "step": 2100 }, { "grad_norm": 1.0055618286132812, "learning_rate": 4.2180000000000006e-05, "loss": 0.0816, "step": 2110 }, { "grad_norm": 1.1438987255096436, "learning_rate": 4.2380000000000004e-05, "loss": 0.0894, "step": 2120 }, { "grad_norm": 0.9245237112045288, "learning_rate": 4.258e-05, "loss": 0.0847, "step": 2130 }, { "grad_norm": 0.7285625338554382, "learning_rate": 4.278e-05, "loss": 0.0832, "step": 2140 }, { "grad_norm": 0.759182333946228, "learning_rate": 4.2980000000000005e-05, "loss": 0.0793, "step": 2150 }, { "grad_norm": 0.8912532925605774, "learning_rate": 4.318e-05, "loss": 0.0743, "step": 2160 }, { "grad_norm": 1.177984356880188, "learning_rate": 4.338e-05, "loss": 0.0761, "step": 2170 }, { "grad_norm": 0.674469530582428, "learning_rate": 4.3580000000000006e-05, "loss": 0.0844, "step": 2180 }, { "grad_norm": 1.0769490003585815, "learning_rate": 4.3780000000000004e-05, "loss": 0.0863, "step": 2190 }, { "grad_norm": 1.0901368856430054, "learning_rate": 4.398e-05, "loss": 0.0906, "step": 2200 }, { "grad_norm": 0.9569987058639526, "learning_rate": 4.418000000000001e-05, "loss": 0.076, "step": 2210 }, { "grad_norm": 0.8210286498069763, "learning_rate": 4.438e-05, "loss": 0.0817, "step": 2220 }, { "grad_norm": 0.9545658230781555, "learning_rate": 4.458e-05, "loss": 0.0821, "step": 2230 }, { "grad_norm": 0.7463720440864563, "learning_rate": 4.478e-05, "loss": 0.0896, "step": 2240 }, { "grad_norm": 0.7306517362594604, "learning_rate": 4.498e-05, "loss": 0.0885, "step": 2250 }, { "grad_norm": 1.0879994630813599, "learning_rate": 4.518e-05, "loss": 0.0802, "step": 2260 }, { "grad_norm": 0.9373118877410889, "learning_rate": 4.538e-05, "loss": 0.0763, "step": 2270 }, { "grad_norm": 0.9587438702583313, "learning_rate": 4.558e-05, "loss": 0.0863, "step": 2280 }, { "grad_norm": 0.6731724143028259, "learning_rate": 4.578e-05, "loss": 0.0828, "step": 2290 }, { "grad_norm": 0.8868897557258606, "learning_rate": 4.5980000000000004e-05, "loss": 0.0846, "step": 2300 }, { "grad_norm": 0.8736068606376648, "learning_rate": 4.618e-05, "loss": 0.0862, "step": 2310 }, { "grad_norm": 0.7819711565971375, "learning_rate": 4.638e-05, "loss": 0.0687, "step": 2320 }, { "grad_norm": 0.7367273569107056, "learning_rate": 4.6580000000000005e-05, "loss": 0.0814, "step": 2330 }, { "grad_norm": 0.7853010296821594, "learning_rate": 4.678e-05, "loss": 0.0862, "step": 2340 }, { "grad_norm": 0.6495829820632935, "learning_rate": 4.698e-05, "loss": 0.0822, "step": 2350 }, { "grad_norm": 0.7581531405448914, "learning_rate": 4.718e-05, "loss": 0.0785, "step": 2360 }, { "grad_norm": 1.1081537008285522, "learning_rate": 4.7380000000000004e-05, "loss": 0.082, "step": 2370 }, { "grad_norm": 0.9789213538169861, "learning_rate": 4.758e-05, "loss": 0.0819, "step": 2380 }, { "grad_norm": 0.8484362959861755, "learning_rate": 4.778e-05, "loss": 0.0909, "step": 2390 }, { "grad_norm": 0.7321324944496155, "learning_rate": 4.7980000000000005e-05, "loss": 0.0788, "step": 2400 }, { "grad_norm": 0.8728730082511902, "learning_rate": 4.818e-05, "loss": 0.0823, "step": 2410 }, { "grad_norm": 0.8971062302589417, "learning_rate": 4.838e-05, "loss": 0.0811, "step": 2420 }, { "grad_norm": 0.6480836868286133, "learning_rate": 4.8580000000000006e-05, "loss": 0.072, "step": 2430 }, { "grad_norm": 0.738477885723114, "learning_rate": 4.8780000000000004e-05, "loss": 0.0698, "step": 2440 }, { "grad_norm": 0.6887433528900146, "learning_rate": 4.898e-05, "loss": 0.0761, "step": 2450 }, { "grad_norm": 0.7615669965744019, "learning_rate": 4.918000000000001e-05, "loss": 0.0728, "step": 2460 }, { "grad_norm": 0.8043103218078613, "learning_rate": 4.9380000000000005e-05, "loss": 0.0708, "step": 2470 }, { "grad_norm": 0.8219135403633118, "learning_rate": 4.958e-05, "loss": 0.0818, "step": 2480 }, { "grad_norm": 0.8326594233512878, "learning_rate": 4.978e-05, "loss": 0.0758, "step": 2490 }, { "grad_norm": 0.6410696506500244, "learning_rate": 4.9980000000000006e-05, "loss": 0.0775, "step": 2500 }, { "grad_norm": 0.9557880163192749, "learning_rate": 5.0180000000000004e-05, "loss": 0.066, "step": 2510 }, { "grad_norm": 0.6483591794967651, "learning_rate": 5.038e-05, "loss": 0.0719, "step": 2520 }, { "grad_norm": 0.5444977283477783, "learning_rate": 5.058000000000001e-05, "loss": 0.0793, "step": 2530 }, { "grad_norm": 0.9573062062263489, "learning_rate": 5.0780000000000005e-05, "loss": 0.0865, "step": 2540 }, { "grad_norm": 0.6336964964866638, "learning_rate": 5.098e-05, "loss": 0.0781, "step": 2550 }, { "grad_norm": 1.0317317247390747, "learning_rate": 5.118000000000001e-05, "loss": 0.0788, "step": 2560 }, { "grad_norm": 0.9643881320953369, "learning_rate": 5.1380000000000006e-05, "loss": 0.08, "step": 2570 }, { "grad_norm": 0.9479894042015076, "learning_rate": 5.1580000000000004e-05, "loss": 0.0806, "step": 2580 }, { "grad_norm": 0.8187289834022522, "learning_rate": 5.178000000000001e-05, "loss": 0.0803, "step": 2590 }, { "grad_norm": 0.5568830966949463, "learning_rate": 5.198000000000001e-05, "loss": 0.0765, "step": 2600 }, { "grad_norm": 0.888601541519165, "learning_rate": 5.2180000000000005e-05, "loss": 0.0707, "step": 2610 }, { "grad_norm": 0.9046518206596375, "learning_rate": 5.238000000000001e-05, "loss": 0.0719, "step": 2620 }, { "grad_norm": 0.6892744898796082, "learning_rate": 5.258000000000001e-05, "loss": 0.0803, "step": 2630 }, { "grad_norm": 0.8026613593101501, "learning_rate": 5.2780000000000006e-05, "loss": 0.095, "step": 2640 }, { "grad_norm": 0.7978079915046692, "learning_rate": 5.2980000000000004e-05, "loss": 0.0807, "step": 2650 }, { "grad_norm": 0.70832759141922, "learning_rate": 5.318000000000001e-05, "loss": 0.0706, "step": 2660 }, { "grad_norm": 0.7207254767417908, "learning_rate": 5.338000000000001e-05, "loss": 0.0672, "step": 2670 }, { "grad_norm": 0.7340192794799805, "learning_rate": 5.3580000000000005e-05, "loss": 0.0666, "step": 2680 }, { "grad_norm": 0.8117252588272095, "learning_rate": 5.378e-05, "loss": 0.0672, "step": 2690 }, { "grad_norm": 0.6029465198516846, "learning_rate": 5.3979999999999995e-05, "loss": 0.086, "step": 2700 }, { "grad_norm": 0.6074615716934204, "learning_rate": 5.418e-05, "loss": 0.081, "step": 2710 }, { "grad_norm": 0.6762192249298096, "learning_rate": 5.438e-05, "loss": 0.0838, "step": 2720 }, { "grad_norm": 0.961165189743042, "learning_rate": 5.4579999999999996e-05, "loss": 0.0716, "step": 2730 }, { "grad_norm": 0.6749469041824341, "learning_rate": 5.478e-05, "loss": 0.072, "step": 2740 }, { "grad_norm": 0.661794126033783, "learning_rate": 5.498e-05, "loss": 0.0686, "step": 2750 }, { "grad_norm": 0.7191190123558044, "learning_rate": 5.518e-05, "loss": 0.07, "step": 2760 }, { "grad_norm": 0.8128288388252258, "learning_rate": 5.538e-05, "loss": 0.0701, "step": 2770 }, { "grad_norm": 0.6009504199028015, "learning_rate": 5.558e-05, "loss": 0.065, "step": 2780 }, { "grad_norm": 1.0121666193008423, "learning_rate": 5.578e-05, "loss": 0.0688, "step": 2790 }, { "grad_norm": 0.6706863641738892, "learning_rate": 5.5979999999999996e-05, "loss": 0.0657, "step": 2800 }, { "grad_norm": 0.7932115793228149, "learning_rate": 5.618e-05, "loss": 0.0638, "step": 2810 }, { "grad_norm": 0.7327492833137512, "learning_rate": 5.638e-05, "loss": 0.0659, "step": 2820 }, { "grad_norm": 0.8989432454109192, "learning_rate": 5.658e-05, "loss": 0.0644, "step": 2830 }, { "grad_norm": 0.7049527764320374, "learning_rate": 5.678e-05, "loss": 0.0653, "step": 2840 }, { "grad_norm": 0.653797447681427, "learning_rate": 5.698e-05, "loss": 0.0671, "step": 2850 }, { "grad_norm": 0.6398699283599854, "learning_rate": 5.718e-05, "loss": 0.0836, "step": 2860 }, { "grad_norm": 0.611112117767334, "learning_rate": 5.738e-05, "loss": 0.0656, "step": 2870 }, { "grad_norm": 0.7540693879127502, "learning_rate": 5.758e-05, "loss": 0.0719, "step": 2880 }, { "grad_norm": 0.9572031497955322, "learning_rate": 5.778e-05, "loss": 0.0704, "step": 2890 }, { "grad_norm": 0.6235620975494385, "learning_rate": 5.7980000000000004e-05, "loss": 0.062, "step": 2900 }, { "grad_norm": 0.6115180850028992, "learning_rate": 5.818e-05, "loss": 0.0746, "step": 2910 }, { "grad_norm": 0.7206747531890869, "learning_rate": 5.838e-05, "loss": 0.0667, "step": 2920 }, { "grad_norm": 0.6757913827896118, "learning_rate": 5.858e-05, "loss": 0.0684, "step": 2930 }, { "grad_norm": 0.7237628698348999, "learning_rate": 5.878e-05, "loss": 0.0601, "step": 2940 }, { "grad_norm": 0.6015501022338867, "learning_rate": 5.898e-05, "loss": 0.0644, "step": 2950 }, { "grad_norm": 0.8081467151641846, "learning_rate": 5.918e-05, "loss": 0.0675, "step": 2960 }, { "grad_norm": 0.8021536469459534, "learning_rate": 5.9380000000000004e-05, "loss": 0.0717, "step": 2970 }, { "grad_norm": 0.497595876455307, "learning_rate": 5.958e-05, "loss": 0.0752, "step": 2980 }, { "grad_norm": 0.6016698479652405, "learning_rate": 5.978e-05, "loss": 0.057, "step": 2990 }, { "grad_norm": 0.6399852633476257, "learning_rate": 5.9980000000000005e-05, "loss": 0.0754, "step": 3000 }, { "grad_norm": 0.7931088209152222, "learning_rate": 6.018e-05, "loss": 0.072, "step": 3010 }, { "grad_norm": 0.6648734211921692, "learning_rate": 6.038e-05, "loss": 0.0739, "step": 3020 }, { "grad_norm": 0.7634868621826172, "learning_rate": 6.0580000000000006e-05, "loss": 0.0717, "step": 3030 }, { "grad_norm": 0.8208682537078857, "learning_rate": 6.0780000000000004e-05, "loss": 0.0741, "step": 3040 }, { "grad_norm": 0.5452602505683899, "learning_rate": 6.098e-05, "loss": 0.073, "step": 3050 }, { "grad_norm": 0.5621604919433594, "learning_rate": 6.118000000000001e-05, "loss": 0.064, "step": 3060 }, { "grad_norm": 0.5921342968940735, "learning_rate": 6.138e-05, "loss": 0.0693, "step": 3070 }, { "grad_norm": 0.6019565463066101, "learning_rate": 6.158e-05, "loss": 0.0672, "step": 3080 }, { "grad_norm": 0.7535709738731384, "learning_rate": 6.178000000000001e-05, "loss": 0.0645, "step": 3090 }, { "grad_norm": 0.5397651195526123, "learning_rate": 6.198e-05, "loss": 0.0705, "step": 3100 }, { "grad_norm": 0.9253067970275879, "learning_rate": 6.218e-05, "loss": 0.0713, "step": 3110 }, { "grad_norm": 0.8217318058013916, "learning_rate": 6.238000000000001e-05, "loss": 0.0657, "step": 3120 }, { "grad_norm": 0.7084094285964966, "learning_rate": 6.258e-05, "loss": 0.0727, "step": 3130 }, { "grad_norm": 0.610658586025238, "learning_rate": 6.278e-05, "loss": 0.0734, "step": 3140 }, { "grad_norm": 0.7215097546577454, "learning_rate": 6.298000000000001e-05, "loss": 0.0602, "step": 3150 }, { "grad_norm": 0.6934249997138977, "learning_rate": 6.318e-05, "loss": 0.0679, "step": 3160 }, { "grad_norm": 0.6282787322998047, "learning_rate": 6.338e-05, "loss": 0.0643, "step": 3170 }, { "grad_norm": 0.8071790933609009, "learning_rate": 6.358000000000001e-05, "loss": 0.0726, "step": 3180 }, { "grad_norm": 0.6599127650260925, "learning_rate": 6.378e-05, "loss": 0.0714, "step": 3190 }, { "grad_norm": 0.6457439661026001, "learning_rate": 6.398000000000001e-05, "loss": 0.0638, "step": 3200 }, { "grad_norm": 0.5202849507331848, "learning_rate": 6.418000000000001e-05, "loss": 0.0651, "step": 3210 }, { "grad_norm": 0.5953631401062012, "learning_rate": 6.438e-05, "loss": 0.0612, "step": 3220 }, { "grad_norm": 0.9139602780342102, "learning_rate": 6.458000000000001e-05, "loss": 0.0687, "step": 3230 }, { "grad_norm": 0.669836163520813, "learning_rate": 6.478000000000001e-05, "loss": 0.0665, "step": 3240 }, { "grad_norm": 0.5023570656776428, "learning_rate": 6.498e-05, "loss": 0.0686, "step": 3250 }, { "grad_norm": 0.6526066660881042, "learning_rate": 6.518000000000001e-05, "loss": 0.0642, "step": 3260 }, { "grad_norm": 0.7484228014945984, "learning_rate": 6.538000000000001e-05, "loss": 0.0648, "step": 3270 }, { "grad_norm": 0.7181470394134521, "learning_rate": 6.558e-05, "loss": 0.0738, "step": 3280 }, { "grad_norm": 0.44841259717941284, "learning_rate": 6.578000000000001e-05, "loss": 0.0649, "step": 3290 }, { "grad_norm": 0.663599967956543, "learning_rate": 6.598e-05, "loss": 0.066, "step": 3300 }, { "grad_norm": 0.6698228120803833, "learning_rate": 6.618e-05, "loss": 0.0671, "step": 3310 }, { "grad_norm": 0.5677284598350525, "learning_rate": 6.638e-05, "loss": 0.0619, "step": 3320 }, { "grad_norm": 0.6526055335998535, "learning_rate": 6.658e-05, "loss": 0.0697, "step": 3330 }, { "grad_norm": 0.6544400453567505, "learning_rate": 6.678e-05, "loss": 0.0706, "step": 3340 }, { "grad_norm": 0.6765251755714417, "learning_rate": 6.698e-05, "loss": 0.0692, "step": 3350 }, { "grad_norm": 0.5756562352180481, "learning_rate": 6.718e-05, "loss": 0.0626, "step": 3360 }, { "grad_norm": 0.5198899507522583, "learning_rate": 6.738e-05, "loss": 0.0614, "step": 3370 }, { "grad_norm": 0.7621904611587524, "learning_rate": 6.758e-05, "loss": 0.0599, "step": 3380 }, { "grad_norm": 0.6501011848449707, "learning_rate": 6.778e-05, "loss": 0.0674, "step": 3390 }, { "grad_norm": 0.6072428226470947, "learning_rate": 6.798e-05, "loss": 0.0763, "step": 3400 }, { "grad_norm": 0.7910894751548767, "learning_rate": 6.818e-05, "loss": 0.0636, "step": 3410 }, { "grad_norm": 0.6126161217689514, "learning_rate": 6.838e-05, "loss": 0.0635, "step": 3420 }, { "grad_norm": 0.6534228324890137, "learning_rate": 6.858e-05, "loss": 0.0701, "step": 3430 }, { "grad_norm": 0.5926064252853394, "learning_rate": 6.878e-05, "loss": 0.0629, "step": 3440 }, { "grad_norm": 0.5666845440864563, "learning_rate": 6.898e-05, "loss": 0.0701, "step": 3450 }, { "grad_norm": 0.5672755837440491, "learning_rate": 6.918e-05, "loss": 0.0587, "step": 3460 }, { "grad_norm": 0.622974157333374, "learning_rate": 6.938e-05, "loss": 0.0614, "step": 3470 }, { "grad_norm": 0.5983734130859375, "learning_rate": 6.958e-05, "loss": 0.0643, "step": 3480 }, { "grad_norm": 0.6349936127662659, "learning_rate": 6.978e-05, "loss": 0.0667, "step": 3490 }, { "grad_norm": 0.6902998685836792, "learning_rate": 6.998e-05, "loss": 0.057, "step": 3500 }, { "grad_norm": 0.5985453724861145, "learning_rate": 7.018e-05, "loss": 0.0682, "step": 3510 }, { "grad_norm": 0.484272837638855, "learning_rate": 7.038e-05, "loss": 0.0551, "step": 3520 }, { "grad_norm": 0.4107840657234192, "learning_rate": 7.058e-05, "loss": 0.0577, "step": 3530 }, { "grad_norm": 0.6641446948051453, "learning_rate": 7.078e-05, "loss": 0.0829, "step": 3540 }, { "grad_norm": 0.7321299314498901, "learning_rate": 7.098e-05, "loss": 0.0639, "step": 3550 }, { "grad_norm": 0.5618053078651428, "learning_rate": 7.118e-05, "loss": 0.0638, "step": 3560 }, { "grad_norm": 0.5169888138771057, "learning_rate": 7.138e-05, "loss": 0.0725, "step": 3570 }, { "grad_norm": 0.615692138671875, "learning_rate": 7.158e-05, "loss": 0.0584, "step": 3580 }, { "grad_norm": 0.5938225388526917, "learning_rate": 7.178000000000001e-05, "loss": 0.0685, "step": 3590 }, { "grad_norm": 0.691398024559021, "learning_rate": 7.198e-05, "loss": 0.0704, "step": 3600 }, { "grad_norm": 0.7362175583839417, "learning_rate": 7.218e-05, "loss": 0.0662, "step": 3610 }, { "grad_norm": 0.5433744192123413, "learning_rate": 7.238000000000001e-05, "loss": 0.0638, "step": 3620 }, { "grad_norm": 0.6199750304222107, "learning_rate": 7.258e-05, "loss": 0.0626, "step": 3630 }, { "grad_norm": 0.5639925003051758, "learning_rate": 7.278e-05, "loss": 0.0624, "step": 3640 }, { "grad_norm": 0.6625953316688538, "learning_rate": 7.298000000000001e-05, "loss": 0.0712, "step": 3650 }, { "grad_norm": 0.4501965343952179, "learning_rate": 7.318e-05, "loss": 0.0528, "step": 3660 }, { "grad_norm": 0.5063297748565674, "learning_rate": 7.338e-05, "loss": 0.0622, "step": 3670 }, { "grad_norm": 0.5950095653533936, "learning_rate": 7.358000000000001e-05, "loss": 0.0645, "step": 3680 }, { "grad_norm": 0.6131200194358826, "learning_rate": 7.378e-05, "loss": 0.0586, "step": 3690 }, { "grad_norm": 0.6101892590522766, "learning_rate": 7.398e-05, "loss": 0.0618, "step": 3700 }, { "grad_norm": 0.6290402412414551, "learning_rate": 7.418000000000001e-05, "loss": 0.0675, "step": 3710 }, { "grad_norm": 0.7120500206947327, "learning_rate": 7.438e-05, "loss": 0.0568, "step": 3720 }, { "grad_norm": 0.5624719858169556, "learning_rate": 7.458000000000001e-05, "loss": 0.0657, "step": 3730 }, { "grad_norm": 0.7249051332473755, "learning_rate": 7.478e-05, "loss": 0.0657, "step": 3740 }, { "grad_norm": 0.4906325042247772, "learning_rate": 7.498e-05, "loss": 0.0632, "step": 3750 }, { "grad_norm": 0.6196677088737488, "learning_rate": 7.518000000000001e-05, "loss": 0.0588, "step": 3760 }, { "grad_norm": 0.5590428113937378, "learning_rate": 7.538e-05, "loss": 0.0641, "step": 3770 }, { "grad_norm": 0.5566707253456116, "learning_rate": 7.558e-05, "loss": 0.056, "step": 3780 }, { "grad_norm": 0.5099586844444275, "learning_rate": 7.578000000000001e-05, "loss": 0.0669, "step": 3790 }, { "grad_norm": 0.6393294334411621, "learning_rate": 7.598e-05, "loss": 0.0636, "step": 3800 }, { "grad_norm": 0.6185796856880188, "learning_rate": 7.618e-05, "loss": 0.0631, "step": 3810 }, { "grad_norm": 0.476358026266098, "learning_rate": 7.638000000000001e-05, "loss": 0.0615, "step": 3820 }, { "grad_norm": 0.5498641729354858, "learning_rate": 7.658e-05, "loss": 0.0587, "step": 3830 }, { "grad_norm": 0.5011379718780518, "learning_rate": 7.678000000000001e-05, "loss": 0.06, "step": 3840 }, { "grad_norm": 0.4071027636528015, "learning_rate": 7.698000000000001e-05, "loss": 0.0592, "step": 3850 }, { "grad_norm": 0.5740132927894592, "learning_rate": 7.718e-05, "loss": 0.0593, "step": 3860 }, { "grad_norm": 0.744632363319397, "learning_rate": 7.738000000000001e-05, "loss": 0.0698, "step": 3870 }, { "grad_norm": 0.6028439402580261, "learning_rate": 7.758000000000001e-05, "loss": 0.0782, "step": 3880 }, { "grad_norm": 0.49136340618133545, "learning_rate": 7.778e-05, "loss": 0.0618, "step": 3890 }, { "grad_norm": 0.5862046480178833, "learning_rate": 7.798000000000001e-05, "loss": 0.0631, "step": 3900 }, { "grad_norm": 0.6531053781509399, "learning_rate": 7.818000000000001e-05, "loss": 0.0614, "step": 3910 }, { "grad_norm": 0.6209702491760254, "learning_rate": 7.838e-05, "loss": 0.0621, "step": 3920 }, { "grad_norm": 0.5378124713897705, "learning_rate": 7.858000000000001e-05, "loss": 0.0573, "step": 3930 }, { "grad_norm": 0.5101502537727356, "learning_rate": 7.878e-05, "loss": 0.066, "step": 3940 }, { "grad_norm": 0.4943355619907379, "learning_rate": 7.897999999999999e-05, "loss": 0.0671, "step": 3950 }, { "grad_norm": 0.5018666982650757, "learning_rate": 7.918e-05, "loss": 0.0646, "step": 3960 }, { "grad_norm": 0.4508885443210602, "learning_rate": 7.938e-05, "loss": 0.0597, "step": 3970 }, { "grad_norm": 0.633209764957428, "learning_rate": 7.958e-05, "loss": 0.0629, "step": 3980 }, { "grad_norm": 0.5034310817718506, "learning_rate": 7.978e-05, "loss": 0.0578, "step": 3990 }, { "grad_norm": 0.5321434140205383, "learning_rate": 7.998e-05, "loss": 0.068, "step": 4000 }, { "grad_norm": 0.43332141637802124, "learning_rate": 8.018e-05, "loss": 0.059, "step": 4010 }, { "grad_norm": 0.4992964267730713, "learning_rate": 8.038e-05, "loss": 0.0658, "step": 4020 }, { "grad_norm": 0.5926293730735779, "learning_rate": 8.058e-05, "loss": 0.0647, "step": 4030 }, { "grad_norm": 0.5128733515739441, "learning_rate": 8.078e-05, "loss": 0.0603, "step": 4040 }, { "grad_norm": 0.48248034715652466, "learning_rate": 8.098e-05, "loss": 0.0585, "step": 4050 }, { "grad_norm": 0.624273955821991, "learning_rate": 8.118e-05, "loss": 0.0583, "step": 4060 }, { "grad_norm": 0.6895151734352112, "learning_rate": 8.138e-05, "loss": 0.0683, "step": 4070 }, { "grad_norm": 0.43657320737838745, "learning_rate": 8.158e-05, "loss": 0.0584, "step": 4080 }, { "grad_norm": 0.5000276565551758, "learning_rate": 8.178e-05, "loss": 0.0655, "step": 4090 }, { "grad_norm": 0.6020710468292236, "learning_rate": 8.198e-05, "loss": 0.0622, "step": 4100 }, { "grad_norm": 0.50217205286026, "learning_rate": 8.218e-05, "loss": 0.0639, "step": 4110 }, { "grad_norm": 0.45799577236175537, "learning_rate": 8.238000000000001e-05, "loss": 0.0609, "step": 4120 }, { "grad_norm": 0.4764360785484314, "learning_rate": 8.258e-05, "loss": 0.0627, "step": 4130 }, { "grad_norm": 0.631537139415741, "learning_rate": 8.278e-05, "loss": 0.0646, "step": 4140 }, { "grad_norm": 0.5949133634567261, "learning_rate": 8.298000000000001e-05, "loss": 0.0694, "step": 4150 }, { "grad_norm": 0.5553409457206726, "learning_rate": 8.318e-05, "loss": 0.06, "step": 4160 }, { "grad_norm": 0.6657522916793823, "learning_rate": 8.338e-05, "loss": 0.0661, "step": 4170 }, { "grad_norm": 0.5387417078018188, "learning_rate": 8.358e-05, "loss": 0.0646, "step": 4180 }, { "grad_norm": 0.49241527915000916, "learning_rate": 8.378e-05, "loss": 0.062, "step": 4190 }, { "grad_norm": 0.3781577944755554, "learning_rate": 8.398e-05, "loss": 0.0583, "step": 4200 }, { "grad_norm": 0.48474979400634766, "learning_rate": 8.418e-05, "loss": 0.05, "step": 4210 }, { "grad_norm": 0.6549875140190125, "learning_rate": 8.438e-05, "loss": 0.0664, "step": 4220 }, { "grad_norm": 0.46787330508232117, "learning_rate": 8.458e-05, "loss": 0.0588, "step": 4230 }, { "grad_norm": 0.38436517119407654, "learning_rate": 8.478e-05, "loss": 0.0565, "step": 4240 }, { "grad_norm": 0.51840740442276, "learning_rate": 8.498e-05, "loss": 0.0554, "step": 4250 }, { "grad_norm": 0.7134885787963867, "learning_rate": 8.518000000000001e-05, "loss": 0.0704, "step": 4260 }, { "grad_norm": 0.5020450949668884, "learning_rate": 8.538e-05, "loss": 0.0631, "step": 4270 }, { "grad_norm": 0.5541492104530334, "learning_rate": 8.558e-05, "loss": 0.0601, "step": 4280 }, { "grad_norm": 0.5942652821540833, "learning_rate": 8.578000000000001e-05, "loss": 0.0621, "step": 4290 }, { "grad_norm": 0.5021793842315674, "learning_rate": 8.598e-05, "loss": 0.0623, "step": 4300 }, { "grad_norm": 0.5555506944656372, "learning_rate": 8.618e-05, "loss": 0.0567, "step": 4310 }, { "grad_norm": 0.50489342212677, "learning_rate": 8.638000000000001e-05, "loss": 0.0582, "step": 4320 }, { "grad_norm": 0.4511776268482208, "learning_rate": 8.658e-05, "loss": 0.0528, "step": 4330 }, { "grad_norm": 0.512768566608429, "learning_rate": 8.678e-05, "loss": 0.0712, "step": 4340 }, { "grad_norm": 0.5370095372200012, "learning_rate": 8.698000000000001e-05, "loss": 0.0588, "step": 4350 }, { "grad_norm": 0.44296905398368835, "learning_rate": 8.718e-05, "loss": 0.0637, "step": 4360 }, { "grad_norm": 0.4213048219680786, "learning_rate": 8.738000000000001e-05, "loss": 0.0588, "step": 4370 }, { "grad_norm": 0.5483023524284363, "learning_rate": 8.758000000000001e-05, "loss": 0.0655, "step": 4380 }, { "grad_norm": 0.43625128269195557, "learning_rate": 8.778e-05, "loss": 0.0572, "step": 4390 }, { "grad_norm": 0.5052556395530701, "learning_rate": 8.798000000000001e-05, "loss": 0.0546, "step": 4400 }, { "grad_norm": 0.4453352987766266, "learning_rate": 8.818000000000001e-05, "loss": 0.0543, "step": 4410 }, { "grad_norm": 0.5833339095115662, "learning_rate": 8.838e-05, "loss": 0.0598, "step": 4420 }, { "grad_norm": 0.44214773178100586, "learning_rate": 8.858000000000001e-05, "loss": 0.0592, "step": 4430 }, { "grad_norm": 0.5236348509788513, "learning_rate": 8.878000000000001e-05, "loss": 0.0649, "step": 4440 }, { "grad_norm": 0.5635884404182434, "learning_rate": 8.898e-05, "loss": 0.0605, "step": 4450 }, { "grad_norm": 0.5376366376876831, "learning_rate": 8.918000000000001e-05, "loss": 0.064, "step": 4460 }, { "grad_norm": 0.5008459091186523, "learning_rate": 8.938e-05, "loss": 0.0636, "step": 4470 }, { "grad_norm": 0.44187575578689575, "learning_rate": 8.958e-05, "loss": 0.0552, "step": 4480 }, { "grad_norm": 0.6288156509399414, "learning_rate": 8.978000000000001e-05, "loss": 0.0705, "step": 4490 }, { "grad_norm": 0.42640694975852966, "learning_rate": 8.998e-05, "loss": 0.0621, "step": 4500 }, { "grad_norm": 0.5535826683044434, "learning_rate": 9.018000000000001e-05, "loss": 0.0719, "step": 4510 }, { "grad_norm": 0.4693029522895813, "learning_rate": 9.038000000000001e-05, "loss": 0.0583, "step": 4520 }, { "grad_norm": 0.5731720328330994, "learning_rate": 9.058e-05, "loss": 0.0635, "step": 4530 }, { "grad_norm": 0.4580608606338501, "learning_rate": 9.078000000000001e-05, "loss": 0.075, "step": 4540 }, { "grad_norm": 0.5608760118484497, "learning_rate": 9.098000000000001e-05, "loss": 0.0614, "step": 4550 }, { "grad_norm": 0.41669532656669617, "learning_rate": 9.118e-05, "loss": 0.0622, "step": 4560 }, { "grad_norm": 0.5036733150482178, "learning_rate": 9.138e-05, "loss": 0.0587, "step": 4570 }, { "grad_norm": 0.6188783049583435, "learning_rate": 9.158e-05, "loss": 0.0554, "step": 4580 }, { "grad_norm": 0.5988331437110901, "learning_rate": 9.178e-05, "loss": 0.0698, "step": 4590 }, { "grad_norm": 0.4512734115123749, "learning_rate": 9.198e-05, "loss": 0.0617, "step": 4600 }, { "grad_norm": 0.3671180307865143, "learning_rate": 9.218e-05, "loss": 0.0646, "step": 4610 }, { "grad_norm": 0.4168519079685211, "learning_rate": 9.238e-05, "loss": 0.0601, "step": 4620 }, { "grad_norm": 0.5774898529052734, "learning_rate": 9.258e-05, "loss": 0.0629, "step": 4630 }, { "grad_norm": 0.5258127450942993, "learning_rate": 9.278e-05, "loss": 0.058, "step": 4640 }, { "grad_norm": 0.47544434666633606, "learning_rate": 9.298e-05, "loss": 0.0631, "step": 4650 }, { "grad_norm": 0.6204729676246643, "learning_rate": 9.318e-05, "loss": 0.0731, "step": 4660 }, { "grad_norm": 0.5432576537132263, "learning_rate": 9.338e-05, "loss": 0.0654, "step": 4670 }, { "grad_norm": 0.42401495575904846, "learning_rate": 9.358e-05, "loss": 0.0671, "step": 4680 }, { "grad_norm": 0.4537736773490906, "learning_rate": 9.378e-05, "loss": 0.0619, "step": 4690 }, { "grad_norm": 0.5035524368286133, "learning_rate": 9.398e-05, "loss": 0.0598, "step": 4700 }, { "grad_norm": 0.4927660822868347, "learning_rate": 9.418e-05, "loss": 0.0608, "step": 4710 }, { "grad_norm": 0.5490003824234009, "learning_rate": 9.438e-05, "loss": 0.0645, "step": 4720 }, { "grad_norm": 0.4388315975666046, "learning_rate": 9.458e-05, "loss": 0.0604, "step": 4730 }, { "grad_norm": 0.563637912273407, "learning_rate": 9.478e-05, "loss": 0.0638, "step": 4740 }, { "grad_norm": 0.48013338446617126, "learning_rate": 9.498e-05, "loss": 0.0614, "step": 4750 }, { "grad_norm": 0.5169545412063599, "learning_rate": 9.518000000000001e-05, "loss": 0.0614, "step": 4760 }, { "grad_norm": 0.49003612995147705, "learning_rate": 9.538e-05, "loss": 0.0573, "step": 4770 }, { "grad_norm": 0.49287959933280945, "learning_rate": 9.558e-05, "loss": 0.0621, "step": 4780 }, { "grad_norm": 0.513694167137146, "learning_rate": 9.578000000000001e-05, "loss": 0.0562, "step": 4790 }, { "grad_norm": 0.4807327389717102, "learning_rate": 9.598e-05, "loss": 0.0652, "step": 4800 }, { "grad_norm": 0.48132795095443726, "learning_rate": 9.618e-05, "loss": 0.0561, "step": 4810 }, { "grad_norm": 0.46760690212249756, "learning_rate": 9.638000000000001e-05, "loss": 0.058, "step": 4820 }, { "grad_norm": 0.46196386218070984, "learning_rate": 9.658e-05, "loss": 0.0596, "step": 4830 }, { "grad_norm": 0.5307936668395996, "learning_rate": 9.678e-05, "loss": 0.0612, "step": 4840 }, { "grad_norm": 0.4292243421077728, "learning_rate": 9.698000000000001e-05, "loss": 0.0641, "step": 4850 }, { "grad_norm": 0.3987670838832855, "learning_rate": 9.718e-05, "loss": 0.0576, "step": 4860 }, { "grad_norm": 0.4739920496940613, "learning_rate": 9.738e-05, "loss": 0.0595, "step": 4870 }, { "grad_norm": 0.4679388701915741, "learning_rate": 9.758000000000001e-05, "loss": 0.0532, "step": 4880 }, { "grad_norm": 0.4293903112411499, "learning_rate": 9.778e-05, "loss": 0.0625, "step": 4890 }, { "grad_norm": 0.5558474659919739, "learning_rate": 9.798000000000001e-05, "loss": 0.0575, "step": 4900 }, { "grad_norm": 0.5275612473487854, "learning_rate": 9.818000000000001e-05, "loss": 0.0663, "step": 4910 }, { "grad_norm": 0.4005305767059326, "learning_rate": 9.838e-05, "loss": 0.0645, "step": 4920 }, { "grad_norm": 0.43572360277175903, "learning_rate": 9.858000000000001e-05, "loss": 0.0525, "step": 4930 }, { "grad_norm": 0.5505754351615906, "learning_rate": 9.878e-05, "loss": 0.0571, "step": 4940 }, { "grad_norm": 0.4722122550010681, "learning_rate": 9.898e-05, "loss": 0.0688, "step": 4950 }, { "grad_norm": 0.47053971886634827, "learning_rate": 9.918000000000001e-05, "loss": 0.0592, "step": 4960 }, { "grad_norm": 0.400097131729126, "learning_rate": 9.938e-05, "loss": 0.055, "step": 4970 }, { "grad_norm": 0.33890801668167114, "learning_rate": 9.958e-05, "loss": 0.0588, "step": 4980 }, { "grad_norm": 0.4984278976917267, "learning_rate": 9.978000000000001e-05, "loss": 0.0616, "step": 4990 }, { "grad_norm": 0.4999951720237732, "learning_rate": 9.998e-05, "loss": 0.0602, "step": 5000 }, { "grad_norm": 0.6281676888465881, "learning_rate": 9.999999778549045e-05, "loss": 0.0526, "step": 5010 }, { "grad_norm": 0.468824565410614, "learning_rate": 9.999999013039593e-05, "loss": 0.0711, "step": 5020 }, { "grad_norm": 0.571636438369751, "learning_rate": 9.999997700737766e-05, "loss": 0.0628, "step": 5030 }, { "grad_norm": 0.4714862108230591, "learning_rate": 9.999995841643709e-05, "loss": 0.056, "step": 5040 }, { "grad_norm": 0.38462719321250916, "learning_rate": 9.999993435757623e-05, "loss": 0.0525, "step": 5050 }, { "grad_norm": 0.46659135818481445, "learning_rate": 9.999990483079773e-05, "loss": 0.0589, "step": 5060 }, { "grad_norm": 0.496875137090683, "learning_rate": 9.999986983610481e-05, "loss": 0.0591, "step": 5070 }, { "grad_norm": 0.4752037525177002, "learning_rate": 9.99998293735013e-05, "loss": 0.0572, "step": 5080 }, { "grad_norm": 0.39057719707489014, "learning_rate": 9.999978344299161e-05, "loss": 0.0641, "step": 5090 }, { "grad_norm": 0.4905126094818115, "learning_rate": 9.99997320445808e-05, "loss": 0.0677, "step": 5100 }, { "grad_norm": 0.5188883543014526, "learning_rate": 9.999967517827444e-05, "loss": 0.0552, "step": 5110 }, { "grad_norm": 0.5375141501426697, "learning_rate": 9.999961284407879e-05, "loss": 0.0644, "step": 5120 }, { "grad_norm": 0.48214152455329895, "learning_rate": 9.999954504200067e-05, "loss": 0.0594, "step": 5130 }, { "grad_norm": 0.5270017385482788, "learning_rate": 9.999947177204744e-05, "loss": 0.0563, "step": 5140 }, { "grad_norm": 0.5173367857933044, "learning_rate": 9.999939303422718e-05, "loss": 0.0703, "step": 5150 }, { "grad_norm": 0.4473676085472107, "learning_rate": 9.999930882854847e-05, "loss": 0.0583, "step": 5160 }, { "grad_norm": 0.5820854902267456, "learning_rate": 9.999921915502051e-05, "loss": 0.0637, "step": 5170 }, { "grad_norm": 0.37926211953163147, "learning_rate": 9.99991240136531e-05, "loss": 0.0524, "step": 5180 }, { "grad_norm": 0.467831552028656, "learning_rate": 9.999902340445668e-05, "loss": 0.0614, "step": 5190 }, { "grad_norm": 0.45329171419143677, "learning_rate": 9.999891732744224e-05, "loss": 0.055, "step": 5200 }, { "grad_norm": 0.4250103235244751, "learning_rate": 9.999880578262135e-05, "loss": 0.056, "step": 5210 }, { "grad_norm": 0.4760318100452423, "learning_rate": 9.999868877000624e-05, "loss": 0.0625, "step": 5220 }, { "grad_norm": 0.39874643087387085, "learning_rate": 9.99985662896097e-05, "loss": 0.0535, "step": 5230 }, { "grad_norm": 0.4732383191585541, "learning_rate": 9.999843834144513e-05, "loss": 0.0599, "step": 5240 }, { "grad_norm": 0.49297139048576355, "learning_rate": 9.99983049255265e-05, "loss": 0.0648, "step": 5250 }, { "grad_norm": 0.4201333820819855, "learning_rate": 9.999816604186843e-05, "loss": 0.0588, "step": 5260 }, { "grad_norm": 0.43384692072868347, "learning_rate": 9.999802169048609e-05, "loss": 0.0626, "step": 5270 }, { "grad_norm": 0.4646947979927063, "learning_rate": 9.999787187139527e-05, "loss": 0.057, "step": 5280 }, { "grad_norm": 0.48651447892189026, "learning_rate": 9.999771658461234e-05, "loss": 0.0593, "step": 5290 }, { "grad_norm": 0.3482489585876465, "learning_rate": 9.999755583015431e-05, "loss": 0.0575, "step": 5300 }, { "grad_norm": 0.4375106990337372, "learning_rate": 9.999738960803874e-05, "loss": 0.0542, "step": 5310 }, { "grad_norm": 0.42200160026550293, "learning_rate": 9.99972179182838e-05, "loss": 0.0489, "step": 5320 }, { "grad_norm": 0.5561057329177856, "learning_rate": 9.99970407609083e-05, "loss": 0.0582, "step": 5330 }, { "grad_norm": 0.3855036199092865, "learning_rate": 9.999685813593159e-05, "loss": 0.0545, "step": 5340 }, { "grad_norm": 0.39202842116355896, "learning_rate": 9.999667004337362e-05, "loss": 0.0578, "step": 5350 }, { "grad_norm": 0.4747035503387451, "learning_rate": 9.9996476483255e-05, "loss": 0.0522, "step": 5360 }, { "grad_norm": 0.3810494542121887, "learning_rate": 9.999627745559688e-05, "loss": 0.0561, "step": 5370 }, { "grad_norm": 0.5186188817024231, "learning_rate": 9.999607296042101e-05, "loss": 0.0585, "step": 5380 }, { "grad_norm": 0.5118439793586731, "learning_rate": 9.99958629977498e-05, "loss": 0.0607, "step": 5390 }, { "grad_norm": 0.5760814547538757, "learning_rate": 9.999564756760615e-05, "loss": 0.0538, "step": 5400 }, { "grad_norm": 0.5757726430892944, "learning_rate": 9.999542667001366e-05, "loss": 0.0636, "step": 5410 }, { "grad_norm": 0.5080767869949341, "learning_rate": 9.999520030499647e-05, "loss": 0.063, "step": 5420 }, { "grad_norm": 0.4196869134902954, "learning_rate": 9.999496847257936e-05, "loss": 0.0567, "step": 5430 }, { "grad_norm": 0.44926735758781433, "learning_rate": 9.999473117278764e-05, "loss": 0.0545, "step": 5440 }, { "grad_norm": 0.38510891795158386, "learning_rate": 9.999448840564731e-05, "loss": 0.0554, "step": 5450 }, { "grad_norm": 0.4432104825973511, "learning_rate": 9.999424017118488e-05, "loss": 0.0608, "step": 5460 }, { "grad_norm": 0.3901810944080353, "learning_rate": 9.999398646942751e-05, "loss": 0.0589, "step": 5470 }, { "grad_norm": 0.4086199998855591, "learning_rate": 9.999372730040296e-05, "loss": 0.0578, "step": 5480 }, { "grad_norm": 0.43562743067741394, "learning_rate": 9.999346266413953e-05, "loss": 0.054, "step": 5490 }, { "grad_norm": 0.44435444474220276, "learning_rate": 9.99931925606662e-05, "loss": 0.0589, "step": 5500 }, { "grad_norm": 0.39269036054611206, "learning_rate": 9.99929169900125e-05, "loss": 0.0546, "step": 5510 }, { "grad_norm": 0.41192448139190674, "learning_rate": 9.999263595220855e-05, "loss": 0.0621, "step": 5520 }, { "grad_norm": 0.4352138936519623, "learning_rate": 9.99923494472851e-05, "loss": 0.0571, "step": 5530 }, { "grad_norm": 0.38434743881225586, "learning_rate": 9.999205747527348e-05, "loss": 0.0505, "step": 5540 }, { "grad_norm": 0.3552541434764862, "learning_rate": 9.999176003620561e-05, "loss": 0.0533, "step": 5550 }, { "grad_norm": 0.4317842423915863, "learning_rate": 9.999145713011405e-05, "loss": 0.0566, "step": 5560 }, { "grad_norm": 0.4300672709941864, "learning_rate": 9.999114875703186e-05, "loss": 0.0562, "step": 5570 }, { "grad_norm": 0.41680946946144104, "learning_rate": 9.999083491699281e-05, "loss": 0.0645, "step": 5580 }, { "grad_norm": 0.3593619167804718, "learning_rate": 9.999051561003123e-05, "loss": 0.0576, "step": 5590 }, { "grad_norm": 0.44505807757377625, "learning_rate": 9.999019083618202e-05, "loss": 0.055, "step": 5600 }, { "grad_norm": 0.39209306240081787, "learning_rate": 9.99898605954807e-05, "loss": 0.0533, "step": 5610 }, { "grad_norm": 0.49780505895614624, "learning_rate": 9.998952488796338e-05, "loss": 0.0548, "step": 5620 }, { "grad_norm": 0.4755346477031708, "learning_rate": 9.998918371366676e-05, "loss": 0.0504, "step": 5630 }, { "grad_norm": 0.3904695212841034, "learning_rate": 9.99888370726282e-05, "loss": 0.0559, "step": 5640 }, { "grad_norm": 0.4185907542705536, "learning_rate": 9.998848496488556e-05, "loss": 0.0585, "step": 5650 }, { "grad_norm": 0.3835526406764984, "learning_rate": 9.998812739047736e-05, "loss": 0.0532, "step": 5660 }, { "grad_norm": 0.45000478625297546, "learning_rate": 9.99877643494427e-05, "loss": 0.0589, "step": 5670 }, { "grad_norm": 0.3779031038284302, "learning_rate": 9.998739584182128e-05, "loss": 0.0534, "step": 5680 }, { "grad_norm": 0.34137412905693054, "learning_rate": 9.998702186765342e-05, "loss": 0.0528, "step": 5690 }, { "grad_norm": 0.3930147588253021, "learning_rate": 9.998664242698e-05, "loss": 0.0489, "step": 5700 }, { "grad_norm": 0.34228092432022095, "learning_rate": 9.998625751984251e-05, "loss": 0.0487, "step": 5710 }, { "grad_norm": 0.4306652843952179, "learning_rate": 9.998586714628307e-05, "loss": 0.0515, "step": 5720 }, { "grad_norm": 0.4027767777442932, "learning_rate": 9.998547130634432e-05, "loss": 0.0523, "step": 5730 }, { "grad_norm": 0.3810531198978424, "learning_rate": 9.99850700000696e-05, "loss": 0.0535, "step": 5740 }, { "grad_norm": 0.3578791320323944, "learning_rate": 9.998466322750278e-05, "loss": 0.0539, "step": 5750 }, { "grad_norm": 0.3967157006263733, "learning_rate": 9.998425098868834e-05, "loss": 0.0521, "step": 5760 }, { "grad_norm": 0.4962480664253235, "learning_rate": 9.998383328367136e-05, "loss": 0.0582, "step": 5770 }, { "grad_norm": 0.4166454076766968, "learning_rate": 9.99834101124975e-05, "loss": 0.0453, "step": 5780 }, { "grad_norm": 0.39478999376296997, "learning_rate": 9.998298147521309e-05, "loss": 0.0521, "step": 5790 }, { "grad_norm": 0.3759841322898865, "learning_rate": 9.998254737186496e-05, "loss": 0.0534, "step": 5800 }, { "grad_norm": 0.39819419384002686, "learning_rate": 9.99821078025006e-05, "loss": 0.0494, "step": 5810 }, { "grad_norm": 0.36043810844421387, "learning_rate": 9.998166276716807e-05, "loss": 0.0522, "step": 5820 }, { "grad_norm": 0.3737526834011078, "learning_rate": 9.998121226591606e-05, "loss": 0.0485, "step": 5830 }, { "grad_norm": 0.5193570256233215, "learning_rate": 9.998075629879382e-05, "loss": 0.0501, "step": 5840 }, { "grad_norm": 0.3492434024810791, "learning_rate": 9.99802948658512e-05, "loss": 0.0539, "step": 5850 }, { "grad_norm": 0.42981910705566406, "learning_rate": 9.99798279671387e-05, "loss": 0.0648, "step": 5860 }, { "grad_norm": 0.4155016541481018, "learning_rate": 9.997935560270734e-05, "loss": 0.0605, "step": 5870 }, { "grad_norm": 0.4104453921318054, "learning_rate": 9.997887777260879e-05, "loss": 0.0507, "step": 5880 }, { "grad_norm": 0.3627760112285614, "learning_rate": 9.997839447689532e-05, "loss": 0.0575, "step": 5890 }, { "grad_norm": 0.4042479693889618, "learning_rate": 9.997790571561978e-05, "loss": 0.0546, "step": 5900 }, { "grad_norm": 0.45203784108161926, "learning_rate": 9.99774114888356e-05, "loss": 0.0544, "step": 5910 }, { "grad_norm": 0.4834488332271576, "learning_rate": 9.997691179659684e-05, "loss": 0.0674, "step": 5920 }, { "grad_norm": 0.322923868894577, "learning_rate": 9.997640663895815e-05, "loss": 0.0509, "step": 5930 }, { "grad_norm": 0.3699667155742645, "learning_rate": 9.997589601597477e-05, "loss": 0.0553, "step": 5940 }, { "grad_norm": 0.37733134627342224, "learning_rate": 9.997537992770252e-05, "loss": 0.0503, "step": 5950 }, { "grad_norm": 0.39666423201560974, "learning_rate": 9.997485837419788e-05, "loss": 0.0512, "step": 5960 }, { "grad_norm": 0.35206153988838196, "learning_rate": 9.997433135551786e-05, "loss": 0.0492, "step": 5970 }, { "grad_norm": 0.3541375994682312, "learning_rate": 9.997379887172009e-05, "loss": 0.0473, "step": 5980 }, { "grad_norm": 0.5072792172431946, "learning_rate": 9.997326092286281e-05, "loss": 0.0528, "step": 5990 }, { "grad_norm": 0.4335786998271942, "learning_rate": 9.997271750900486e-05, "loss": 0.056, "step": 6000 }, { "grad_norm": 0.31828340888023376, "learning_rate": 9.997216863020565e-05, "loss": 0.0599, "step": 6010 }, { "grad_norm": 0.44716137647628784, "learning_rate": 9.99716142865252e-05, "loss": 0.0559, "step": 6020 }, { "grad_norm": 0.42351338267326355, "learning_rate": 9.997105447802415e-05, "loss": 0.05, "step": 6030 }, { "grad_norm": 0.3479810655117035, "learning_rate": 9.997048920476373e-05, "loss": 0.0492, "step": 6040 }, { "grad_norm": 0.4379100799560547, "learning_rate": 9.996991846680572e-05, "loss": 0.049, "step": 6050 }, { "grad_norm": 0.4359583556652069, "learning_rate": 9.996934226421257e-05, "loss": 0.0474, "step": 6060 }, { "grad_norm": 0.4007827043533325, "learning_rate": 9.996876059704726e-05, "loss": 0.0499, "step": 6070 }, { "grad_norm": 0.39124011993408203, "learning_rate": 9.996817346537343e-05, "loss": 0.0499, "step": 6080 }, { "grad_norm": 0.33709636330604553, "learning_rate": 9.996758086925526e-05, "loss": 0.0506, "step": 6090 }, { "grad_norm": 0.37401166558265686, "learning_rate": 9.996698280875759e-05, "loss": 0.0477, "step": 6100 }, { "grad_norm": 0.30686017870903015, "learning_rate": 9.99663792839458e-05, "loss": 0.0507, "step": 6110 }, { "grad_norm": 0.35302281379699707, "learning_rate": 9.99657702948859e-05, "loss": 0.0492, "step": 6120 }, { "grad_norm": 0.3049154579639435, "learning_rate": 9.996515584164448e-05, "loss": 0.0452, "step": 6130 }, { "grad_norm": 0.6002160906791687, "learning_rate": 9.996453592428873e-05, "loss": 0.0591, "step": 6140 }, { "grad_norm": 0.35909581184387207, "learning_rate": 9.996391054288646e-05, "loss": 0.0508, "step": 6150 }, { "grad_norm": 0.45249143242836, "learning_rate": 9.996327969750605e-05, "loss": 0.0548, "step": 6160 }, { "grad_norm": 0.3830428719520569, "learning_rate": 9.996264338821649e-05, "loss": 0.0558, "step": 6170 }, { "grad_norm": 0.35719865560531616, "learning_rate": 9.996200161508735e-05, "loss": 0.0501, "step": 6180 }, { "grad_norm": 0.380520761013031, "learning_rate": 9.996135437818885e-05, "loss": 0.0531, "step": 6190 }, { "grad_norm": 0.35399454832077026, "learning_rate": 9.996070167759175e-05, "loss": 0.0537, "step": 6200 }, { "grad_norm": 0.46408742666244507, "learning_rate": 9.996004351336743e-05, "loss": 0.0628, "step": 6210 }, { "grad_norm": 0.3600026071071625, "learning_rate": 9.995937988558785e-05, "loss": 0.052, "step": 6220 }, { "grad_norm": 0.4005180299282074, "learning_rate": 9.995871079432561e-05, "loss": 0.0515, "step": 6230 }, { "grad_norm": 0.42127156257629395, "learning_rate": 9.995803623965389e-05, "loss": 0.0482, "step": 6240 }, { "grad_norm": 0.4083048105239868, "learning_rate": 9.995735622164641e-05, "loss": 0.0576, "step": 6250 }, { "grad_norm": 0.4172331690788269, "learning_rate": 9.995667074037758e-05, "loss": 0.0611, "step": 6260 }, { "grad_norm": 0.4186842441558838, "learning_rate": 9.995597979592232e-05, "loss": 0.0497, "step": 6270 }, { "grad_norm": 0.45142990350723267, "learning_rate": 9.995528338835625e-05, "loss": 0.0517, "step": 6280 }, { "grad_norm": 0.4202939569950104, "learning_rate": 9.995458151775547e-05, "loss": 0.0551, "step": 6290 }, { "grad_norm": 0.3843552768230438, "learning_rate": 9.995387418419677e-05, "loss": 0.0557, "step": 6300 }, { "grad_norm": 0.3106977939605713, "learning_rate": 9.99531613877575e-05, "loss": 0.0505, "step": 6310 }, { "grad_norm": 0.3720059096813202, "learning_rate": 9.995244312851559e-05, "loss": 0.0496, "step": 6320 }, { "grad_norm": 0.30711156129837036, "learning_rate": 9.995171940654961e-05, "loss": 0.0471, "step": 6330 }, { "grad_norm": 0.35033082962036133, "learning_rate": 9.995099022193871e-05, "loss": 0.0519, "step": 6340 }, { "grad_norm": 0.46390461921691895, "learning_rate": 9.995025557476261e-05, "loss": 0.0504, "step": 6350 }, { "grad_norm": 0.4434696137905121, "learning_rate": 9.994951546510165e-05, "loss": 0.0562, "step": 6360 }, { "grad_norm": 0.45308223366737366, "learning_rate": 9.994876989303679e-05, "loss": 0.0574, "step": 6370 }, { "grad_norm": 0.35653185844421387, "learning_rate": 9.994801885864955e-05, "loss": 0.0505, "step": 6380 }, { "grad_norm": 0.4323576092720032, "learning_rate": 9.994726236202205e-05, "loss": 0.047, "step": 6390 }, { "grad_norm": 0.5076410174369812, "learning_rate": 9.994650040323704e-05, "loss": 0.0532, "step": 6400 }, { "grad_norm": 0.4264490604400635, "learning_rate": 9.994573298237784e-05, "loss": 0.0478, "step": 6410 }, { "grad_norm": 0.41723138093948364, "learning_rate": 9.994496009952837e-05, "loss": 0.0475, "step": 6420 }, { "grad_norm": 0.34701231122016907, "learning_rate": 9.994418175477316e-05, "loss": 0.0489, "step": 6430 }, { "grad_norm": 0.3824535608291626, "learning_rate": 9.994339794819733e-05, "loss": 0.0497, "step": 6440 }, { "grad_norm": 0.4102015793323517, "learning_rate": 9.994260867988658e-05, "loss": 0.062, "step": 6450 }, { "grad_norm": 0.36730965971946716, "learning_rate": 9.994181394992723e-05, "loss": 0.0474, "step": 6460 }, { "grad_norm": 0.30970028042793274, "learning_rate": 9.994101375840618e-05, "loss": 0.0483, "step": 6470 }, { "grad_norm": 0.3137117624282837, "learning_rate": 9.994020810541098e-05, "loss": 0.0555, "step": 6480 }, { "grad_norm": 0.3691087067127228, "learning_rate": 9.99393969910297e-05, "loss": 0.0514, "step": 6490 }, { "grad_norm": 0.3281560242176056, "learning_rate": 9.993858041535104e-05, "loss": 0.0488, "step": 6500 }, { "grad_norm": 0.4083360731601715, "learning_rate": 9.99377583784643e-05, "loss": 0.0497, "step": 6510 }, { "grad_norm": 0.3830919861793518, "learning_rate": 9.993693088045939e-05, "loss": 0.0525, "step": 6520 }, { "grad_norm": 0.38584765791893005, "learning_rate": 9.99360979214268e-05, "loss": 0.0489, "step": 6530 }, { "grad_norm": 0.38133686780929565, "learning_rate": 9.99352595014576e-05, "loss": 0.0526, "step": 6540 }, { "grad_norm": 0.3394782543182373, "learning_rate": 9.993441562064354e-05, "loss": 0.0472, "step": 6550 }, { "grad_norm": 0.4482647776603699, "learning_rate": 9.993356627907685e-05, "loss": 0.0456, "step": 6560 }, { "grad_norm": 0.3657776713371277, "learning_rate": 9.99327114768504e-05, "loss": 0.0535, "step": 6570 }, { "grad_norm": 0.28077515959739685, "learning_rate": 9.99318512140577e-05, "loss": 0.0528, "step": 6580 }, { "grad_norm": 0.3365846574306488, "learning_rate": 9.993098549079284e-05, "loss": 0.0509, "step": 6590 }, { "grad_norm": 0.40948399901390076, "learning_rate": 9.993011430715047e-05, "loss": 0.0541, "step": 6600 }, { "grad_norm": 0.2950868606567383, "learning_rate": 9.992923766322586e-05, "loss": 0.0444, "step": 6610 }, { "grad_norm": 0.35391905903816223, "learning_rate": 9.99283555591149e-05, "loss": 0.0535, "step": 6620 }, { "grad_norm": 0.4500798285007477, "learning_rate": 9.992746799491404e-05, "loss": 0.0527, "step": 6630 }, { "grad_norm": 0.5052061676979065, "learning_rate": 9.992657497072033e-05, "loss": 0.0564, "step": 6640 }, { "grad_norm": 0.35890108346939087, "learning_rate": 9.992567648663147e-05, "loss": 0.0499, "step": 6650 }, { "grad_norm": 0.3502295911312103, "learning_rate": 9.992477254274568e-05, "loss": 0.052, "step": 6660 }, { "grad_norm": 0.4415181577205658, "learning_rate": 9.992386313916183e-05, "loss": 0.0481, "step": 6670 }, { "grad_norm": 0.40372541546821594, "learning_rate": 9.992294827597934e-05, "loss": 0.0503, "step": 6680 }, { "grad_norm": 0.3908673822879791, "learning_rate": 9.992202795329831e-05, "loss": 0.0529, "step": 6690 }, { "grad_norm": 0.41732707619667053, "learning_rate": 9.992110217121936e-05, "loss": 0.0498, "step": 6700 }, { "grad_norm": 0.3636963665485382, "learning_rate": 9.992017092984372e-05, "loss": 0.0472, "step": 6710 }, { "grad_norm": 0.2745780050754547, "learning_rate": 9.991923422927326e-05, "loss": 0.0486, "step": 6720 }, { "grad_norm": 0.3886648416519165, "learning_rate": 9.991829206961037e-05, "loss": 0.0534, "step": 6730 }, { "grad_norm": 0.3492034375667572, "learning_rate": 9.991734445095813e-05, "loss": 0.0483, "step": 6740 }, { "grad_norm": 0.34824201464653015, "learning_rate": 9.991639137342015e-05, "loss": 0.0489, "step": 6750 }, { "grad_norm": 0.34227100014686584, "learning_rate": 9.991543283710064e-05, "loss": 0.0435, "step": 6760 }, { "grad_norm": 0.31150713562965393, "learning_rate": 9.991446884210445e-05, "loss": 0.0509, "step": 6770 }, { "grad_norm": 0.38472938537597656, "learning_rate": 9.9913499388537e-05, "loss": 0.0483, "step": 6780 }, { "grad_norm": 0.3443368375301361, "learning_rate": 9.99125244765043e-05, "loss": 0.0503, "step": 6790 }, { "grad_norm": 0.3789174258708954, "learning_rate": 9.991154410611296e-05, "loss": 0.0489, "step": 6800 }, { "grad_norm": 0.374763548374176, "learning_rate": 9.99105582774702e-05, "loss": 0.0451, "step": 6810 }, { "grad_norm": 0.3862634301185608, "learning_rate": 9.990956699068384e-05, "loss": 0.0632, "step": 6820 }, { "grad_norm": 0.4298434555530548, "learning_rate": 9.990857024586224e-05, "loss": 0.0481, "step": 6830 }, { "grad_norm": 0.35569480061531067, "learning_rate": 9.990756804311446e-05, "loss": 0.0471, "step": 6840 }, { "grad_norm": 0.34781932830810547, "learning_rate": 9.990656038255006e-05, "loss": 0.0495, "step": 6850 }, { "grad_norm": 0.4031156897544861, "learning_rate": 9.990554726427926e-05, "loss": 0.0441, "step": 6860 }, { "grad_norm": 0.40884479880332947, "learning_rate": 9.990452868841284e-05, "loss": 0.0523, "step": 6870 }, { "grad_norm": 0.37664660811424255, "learning_rate": 9.99035046550622e-05, "loss": 0.049, "step": 6880 }, { "grad_norm": 0.33401378989219666, "learning_rate": 9.99024751643393e-05, "loss": 0.0403, "step": 6890 }, { "grad_norm": 0.47062018513679504, "learning_rate": 9.990144021635677e-05, "loss": 0.0473, "step": 6900 }, { "grad_norm": 0.33362409472465515, "learning_rate": 9.990039981122775e-05, "loss": 0.0489, "step": 6910 }, { "grad_norm": 0.3974372148513794, "learning_rate": 9.989935394906602e-05, "loss": 0.0491, "step": 6920 }, { "grad_norm": 0.4030129015445709, "learning_rate": 9.989830262998598e-05, "loss": 0.0496, "step": 6930 }, { "grad_norm": 0.42734450101852417, "learning_rate": 9.989724585410259e-05, "loss": 0.0534, "step": 6940 }, { "grad_norm": 0.38304993510246277, "learning_rate": 9.989618362153139e-05, "loss": 0.0464, "step": 6950 }, { "grad_norm": 0.39348918199539185, "learning_rate": 9.989511593238859e-05, "loss": 0.0516, "step": 6960 }, { "grad_norm": 0.3849906027317047, "learning_rate": 9.98940427867909e-05, "loss": 0.0531, "step": 6970 }, { "grad_norm": 0.28995877504348755, "learning_rate": 9.989296418485573e-05, "loss": 0.0439, "step": 6980 }, { "grad_norm": 0.35554513335227966, "learning_rate": 9.989188012670101e-05, "loss": 0.0481, "step": 6990 }, { "grad_norm": 0.35811036825180054, "learning_rate": 9.989079061244528e-05, "loss": 0.0429, "step": 7000 }, { "grad_norm": 0.4287640452384949, "learning_rate": 9.988969564220769e-05, "loss": 0.0518, "step": 7010 }, { "grad_norm": 0.3962972164154053, "learning_rate": 9.988859521610801e-05, "loss": 0.0514, "step": 7020 }, { "grad_norm": 0.33407503366470337, "learning_rate": 9.988748933426656e-05, "loss": 0.0455, "step": 7030 }, { "grad_norm": 0.3635035753250122, "learning_rate": 9.988637799680428e-05, "loss": 0.0483, "step": 7040 }, { "grad_norm": 0.4053674340248108, "learning_rate": 9.98852612038427e-05, "loss": 0.0523, "step": 7050 }, { "grad_norm": 0.3443951904773712, "learning_rate": 9.988413895550397e-05, "loss": 0.0469, "step": 7060 }, { "grad_norm": 0.44912928342819214, "learning_rate": 9.98830112519108e-05, "loss": 0.0513, "step": 7070 }, { "grad_norm": 0.3183498978614807, "learning_rate": 9.98818780931865e-05, "loss": 0.0452, "step": 7080 }, { "grad_norm": 0.25777676701545715, "learning_rate": 9.988073947945502e-05, "loss": 0.0473, "step": 7090 }, { "grad_norm": 0.38706308603286743, "learning_rate": 9.987959541084087e-05, "loss": 0.0464, "step": 7100 }, { "grad_norm": 0.34786877036094666, "learning_rate": 9.987844588746915e-05, "loss": 0.048, "step": 7110 }, { "grad_norm": 0.4223873019218445, "learning_rate": 9.987729090946558e-05, "loss": 0.0481, "step": 7120 }, { "grad_norm": 0.3889995813369751, "learning_rate": 9.987613047695647e-05, "loss": 0.0462, "step": 7130 }, { "grad_norm": 0.38025906682014465, "learning_rate": 9.987496459006871e-05, "loss": 0.0543, "step": 7140 }, { "grad_norm": 0.4305076599121094, "learning_rate": 9.987379324892982e-05, "loss": 0.0513, "step": 7150 }, { "grad_norm": 0.3655429780483246, "learning_rate": 9.987261645366788e-05, "loss": 0.0494, "step": 7160 }, { "grad_norm": 0.307508260011673, "learning_rate": 9.987143420441158e-05, "loss": 0.047, "step": 7170 }, { "grad_norm": 0.36844268441200256, "learning_rate": 9.987024650129022e-05, "loss": 0.0465, "step": 7180 }, { "grad_norm": 0.43819358944892883, "learning_rate": 9.986905334443368e-05, "loss": 0.0456, "step": 7190 }, { "grad_norm": 0.2785702347755432, "learning_rate": 9.986785473397245e-05, "loss": 0.0466, "step": 7200 }, { "grad_norm": 0.33422350883483887, "learning_rate": 9.98666506700376e-05, "loss": 0.0501, "step": 7210 }, { "grad_norm": 0.38545024394989014, "learning_rate": 9.986544115276081e-05, "loss": 0.0474, "step": 7220 }, { "grad_norm": 0.3930489718914032, "learning_rate": 9.986422618227433e-05, "loss": 0.0411, "step": 7230 }, { "grad_norm": 0.3040603697299957, "learning_rate": 9.986300575871106e-05, "loss": 0.0446, "step": 7240 }, { "grad_norm": 0.32389265298843384, "learning_rate": 9.986177988220444e-05, "loss": 0.0421, "step": 7250 }, { "grad_norm": 0.3010725677013397, "learning_rate": 9.986054855288856e-05, "loss": 0.0557, "step": 7260 }, { "grad_norm": 0.39908352494239807, "learning_rate": 9.985931177089802e-05, "loss": 0.0489, "step": 7270 }, { "grad_norm": 0.3934558928012848, "learning_rate": 9.985806953636814e-05, "loss": 0.0463, "step": 7280 }, { "grad_norm": 0.38575148582458496, "learning_rate": 9.985682184943471e-05, "loss": 0.0502, "step": 7290 }, { "grad_norm": 0.46191656589508057, "learning_rate": 9.98555687102342e-05, "loss": 0.043, "step": 7300 }, { "grad_norm": 0.3106071352958679, "learning_rate": 9.985431011890367e-05, "loss": 0.0445, "step": 7310 }, { "grad_norm": 0.35007309913635254, "learning_rate": 9.985304607558075e-05, "loss": 0.0455, "step": 7320 }, { "grad_norm": 0.2999418079853058, "learning_rate": 9.985177658040364e-05, "loss": 0.0423, "step": 7330 }, { "grad_norm": 0.2839105725288391, "learning_rate": 9.985050163351119e-05, "loss": 0.0431, "step": 7340 }, { "grad_norm": 0.24146297574043274, "learning_rate": 9.984922123504286e-05, "loss": 0.0446, "step": 7350 }, { "grad_norm": 0.31203994154930115, "learning_rate": 9.984793538513862e-05, "loss": 0.0388, "step": 7360 }, { "grad_norm": 0.3312987685203552, "learning_rate": 9.984664408393912e-05, "loss": 0.0493, "step": 7370 }, { "grad_norm": 0.3741951286792755, "learning_rate": 9.984534733158556e-05, "loss": 0.0492, "step": 7380 }, { "grad_norm": 0.3329855799674988, "learning_rate": 9.984404512821977e-05, "loss": 0.0417, "step": 7390 }, { "grad_norm": 0.44041338562965393, "learning_rate": 9.984273747398411e-05, "loss": 0.0433, "step": 7400 }, { "grad_norm": 0.2848016023635864, "learning_rate": 9.984142436902165e-05, "loss": 0.0457, "step": 7410 }, { "grad_norm": 0.39160770177841187, "learning_rate": 9.984010581347596e-05, "loss": 0.0459, "step": 7420 }, { "grad_norm": 0.3703737258911133, "learning_rate": 9.983878180749121e-05, "loss": 0.0448, "step": 7430 }, { "grad_norm": 0.46928849816322327, "learning_rate": 9.983745235121222e-05, "loss": 0.0476, "step": 7440 }, { "grad_norm": 0.36430874466896057, "learning_rate": 9.983611744478438e-05, "loss": 0.0459, "step": 7450 }, { "grad_norm": 0.3394908308982849, "learning_rate": 9.983477708835365e-05, "loss": 0.0462, "step": 7460 }, { "grad_norm": 0.4746839702129364, "learning_rate": 9.983343128206664e-05, "loss": 0.0521, "step": 7470 }, { "grad_norm": 0.3493088483810425, "learning_rate": 9.983208002607049e-05, "loss": 0.0461, "step": 7480 }, { "grad_norm": 0.334138959646225, "learning_rate": 9.9830723320513e-05, "loss": 0.0472, "step": 7490 }, { "grad_norm": 0.2862480878829956, "learning_rate": 9.982936116554254e-05, "loss": 0.0455, "step": 7500 }, { "grad_norm": 0.4077260196208954, "learning_rate": 9.982799356130803e-05, "loss": 0.0446, "step": 7510 }, { "grad_norm": 0.35523301362991333, "learning_rate": 9.982662050795908e-05, "loss": 0.0514, "step": 7520 }, { "grad_norm": 0.3867308795452118, "learning_rate": 9.982524200564583e-05, "loss": 0.0408, "step": 7530 }, { "grad_norm": 0.3336535096168518, "learning_rate": 9.982385805451901e-05, "loss": 0.0454, "step": 7540 }, { "grad_norm": 0.3416626453399658, "learning_rate": 9.982246865472998e-05, "loss": 0.0502, "step": 7550 }, { "grad_norm": 0.3679117262363434, "learning_rate": 9.982107380643069e-05, "loss": 0.0473, "step": 7560 }, { "grad_norm": 0.37173712253570557, "learning_rate": 9.981967350977368e-05, "loss": 0.0485, "step": 7570 }, { "grad_norm": 0.34462541341781616, "learning_rate": 9.981826776491208e-05, "loss": 0.0398, "step": 7580 }, { "grad_norm": 0.3574542999267578, "learning_rate": 9.98168565719996e-05, "loss": 0.0528, "step": 7590 }, { "grad_norm": 0.2754696309566498, "learning_rate": 9.98154399311906e-05, "loss": 0.0419, "step": 7600 }, { "grad_norm": 0.33895835280418396, "learning_rate": 9.981401784263997e-05, "loss": 0.0508, "step": 7610 }, { "grad_norm": 0.2961021065711975, "learning_rate": 9.981259030650326e-05, "loss": 0.0499, "step": 7620 }, { "grad_norm": 0.38850244879722595, "learning_rate": 9.981115732293655e-05, "loss": 0.0514, "step": 7630 }, { "grad_norm": 0.3942968547344208, "learning_rate": 9.980971889209659e-05, "loss": 0.0573, "step": 7640 }, { "grad_norm": 0.37903475761413574, "learning_rate": 9.980827501414064e-05, "loss": 0.0451, "step": 7650 }, { "grad_norm": 0.34007176756858826, "learning_rate": 9.980682568922663e-05, "loss": 0.042, "step": 7660 }, { "grad_norm": 0.34059348702430725, "learning_rate": 9.980537091751304e-05, "loss": 0.0422, "step": 7670 }, { "grad_norm": 0.32550281286239624, "learning_rate": 9.980391069915897e-05, "loss": 0.0448, "step": 7680 }, { "grad_norm": 0.32012343406677246, "learning_rate": 9.98024450343241e-05, "loss": 0.0458, "step": 7690 }, { "grad_norm": 0.2573104500770569, "learning_rate": 9.980097392316872e-05, "loss": 0.0417, "step": 7700 }, { "grad_norm": 0.33064505457878113, "learning_rate": 9.97994973658537e-05, "loss": 0.0484, "step": 7710 }, { "grad_norm": 0.4092274606227875, "learning_rate": 9.979801536254054e-05, "loss": 0.0494, "step": 7720 }, { "grad_norm": 0.2875663936138153, "learning_rate": 9.979652791339127e-05, "loss": 0.0443, "step": 7730 }, { "grad_norm": 0.2228429913520813, "learning_rate": 9.97950350185686e-05, "loss": 0.044, "step": 7740 }, { "grad_norm": 0.2938937842845917, "learning_rate": 9.979353667823574e-05, "loss": 0.0451, "step": 7750 }, { "grad_norm": 0.34524232149124146, "learning_rate": 9.979203289255658e-05, "loss": 0.0435, "step": 7760 }, { "grad_norm": 0.2981421649456024, "learning_rate": 9.979052366169557e-05, "loss": 0.0499, "step": 7770 }, { "grad_norm": 0.31197676062583923, "learning_rate": 9.978900898581775e-05, "loss": 0.0453, "step": 7780 }, { "grad_norm": 0.36862990260124207, "learning_rate": 9.978748886508875e-05, "loss": 0.0457, "step": 7790 }, { "grad_norm": 0.34000393748283386, "learning_rate": 9.978596329967484e-05, "loss": 0.0541, "step": 7800 }, { "grad_norm": 0.3622317612171173, "learning_rate": 9.978443228974284e-05, "loss": 0.0447, "step": 7810 }, { "grad_norm": 0.41873180866241455, "learning_rate": 9.978289583546015e-05, "loss": 0.0424, "step": 7820 }, { "grad_norm": 0.34423449635505676, "learning_rate": 9.978135393699484e-05, "loss": 0.0455, "step": 7830 }, { "grad_norm": 0.33763688802719116, "learning_rate": 9.977980659451548e-05, "loss": 0.0611, "step": 7840 }, { "grad_norm": 0.2954183518886566, "learning_rate": 9.977825380819135e-05, "loss": 0.0435, "step": 7850 }, { "grad_norm": 0.3365122079849243, "learning_rate": 9.97766955781922e-05, "loss": 0.0434, "step": 7860 }, { "grad_norm": 0.33941686153411865, "learning_rate": 9.977513190468848e-05, "loss": 0.0455, "step": 7870 }, { "grad_norm": 0.3121398389339447, "learning_rate": 9.977356278785116e-05, "loss": 0.0429, "step": 7880 }, { "grad_norm": 0.3005082607269287, "learning_rate": 9.977198822785184e-05, "loss": 0.0424, "step": 7890 }, { "grad_norm": 0.309856116771698, "learning_rate": 9.977040822486273e-05, "loss": 0.0464, "step": 7900 }, { "grad_norm": 0.3545975983142853, "learning_rate": 9.97688227790566e-05, "loss": 0.0505, "step": 7910 }, { "grad_norm": 0.2862134873867035, "learning_rate": 9.976723189060684e-05, "loss": 0.0416, "step": 7920 }, { "grad_norm": 0.4244132936000824, "learning_rate": 9.976563555968742e-05, "loss": 0.0422, "step": 7930 }, { "grad_norm": 0.4086368978023529, "learning_rate": 9.976403378647292e-05, "loss": 0.0443, "step": 7940 }, { "grad_norm": 0.3564366102218628, "learning_rate": 9.97624265711385e-05, "loss": 0.043, "step": 7950 }, { "grad_norm": 0.329787015914917, "learning_rate": 9.976081391385993e-05, "loss": 0.042, "step": 7960 }, { "grad_norm": 0.2944086194038391, "learning_rate": 9.975919581481356e-05, "loss": 0.0442, "step": 7970 }, { "grad_norm": 0.28362587094306946, "learning_rate": 9.975757227417634e-05, "loss": 0.045, "step": 7980 }, { "grad_norm": 0.33706071972846985, "learning_rate": 9.975594329212586e-05, "loss": 0.0464, "step": 7990 }, { "grad_norm": 0.3625870645046234, "learning_rate": 9.97543088688402e-05, "loss": 0.0443, "step": 8000 }, { "grad_norm": 0.30680036544799805, "learning_rate": 9.975266900449814e-05, "loss": 0.0426, "step": 8010 }, { "grad_norm": 0.37274688482284546, "learning_rate": 9.975102369927898e-05, "loss": 0.044, "step": 8020 }, { "grad_norm": 0.3165898323059082, "learning_rate": 9.974937295336269e-05, "loss": 0.0429, "step": 8030 }, { "grad_norm": 0.33584561944007874, "learning_rate": 9.974771676692975e-05, "loss": 0.0437, "step": 8040 }, { "grad_norm": 0.40110862255096436, "learning_rate": 9.974605514016131e-05, "loss": 0.0502, "step": 8050 }, { "grad_norm": 0.351723849773407, "learning_rate": 9.974438807323907e-05, "loss": 0.0459, "step": 8060 }, { "grad_norm": 0.2654000520706177, "learning_rate": 9.974271556634535e-05, "loss": 0.0398, "step": 8070 }, { "grad_norm": 0.38237378001213074, "learning_rate": 9.974103761966302e-05, "loss": 0.0423, "step": 8080 }, { "grad_norm": 0.31715822219848633, "learning_rate": 9.973935423337563e-05, "loss": 0.0412, "step": 8090 }, { "grad_norm": 0.30270084738731384, "learning_rate": 9.973766540766722e-05, "loss": 0.0513, "step": 8100 }, { "grad_norm": 0.3126441240310669, "learning_rate": 9.97359711427225e-05, "loss": 0.0438, "step": 8110 }, { "grad_norm": 0.22631973028182983, "learning_rate": 9.973427143872677e-05, "loss": 0.0449, "step": 8120 }, { "grad_norm": 0.2644786834716797, "learning_rate": 9.973256629586589e-05, "loss": 0.039, "step": 8130 }, { "grad_norm": 0.35752958059310913, "learning_rate": 9.973085571432632e-05, "loss": 0.0401, "step": 8140 }, { "grad_norm": 0.3161136209964752, "learning_rate": 9.972913969429513e-05, "loss": 0.042, "step": 8150 }, { "grad_norm": 0.3712877929210663, "learning_rate": 9.972741823596e-05, "loss": 0.0427, "step": 8160 }, { "grad_norm": 0.3701988756656647, "learning_rate": 9.972569133950917e-05, "loss": 0.0452, "step": 8170 }, { "grad_norm": 0.3667009770870209, "learning_rate": 9.972395900513151e-05, "loss": 0.0446, "step": 8180 }, { "grad_norm": 0.3249559998512268, "learning_rate": 9.972222123301645e-05, "loss": 0.0402, "step": 8190 }, { "grad_norm": 0.3201614320278168, "learning_rate": 9.972047802335403e-05, "loss": 0.0417, "step": 8200 }, { "grad_norm": 0.3465367555618286, "learning_rate": 9.971872937633488e-05, "loss": 0.045, "step": 8210 }, { "grad_norm": 0.35614311695098877, "learning_rate": 9.971697529215024e-05, "loss": 0.0446, "step": 8220 }, { "grad_norm": 0.394957959651947, "learning_rate": 9.971521577099192e-05, "loss": 0.0457, "step": 8230 }, { "grad_norm": 0.37189581990242004, "learning_rate": 9.971345081305236e-05, "loss": 0.0454, "step": 8240 }, { "grad_norm": 0.298801064491272, "learning_rate": 9.971168041852456e-05, "loss": 0.0433, "step": 8250 }, { "grad_norm": 0.3227236270904541, "learning_rate": 9.970990458760215e-05, "loss": 0.0489, "step": 8260 }, { "grad_norm": 0.30732855200767517, "learning_rate": 9.970812332047929e-05, "loss": 0.0409, "step": 8270 }, { "grad_norm": 0.31049907207489014, "learning_rate": 9.97063366173508e-05, "loss": 0.0438, "step": 8280 }, { "grad_norm": 0.3790212571620941, "learning_rate": 9.970454447841207e-05, "loss": 0.0414, "step": 8290 }, { "grad_norm": 0.3430819511413574, "learning_rate": 9.970274690385909e-05, "loss": 0.0435, "step": 8300 }, { "grad_norm": 0.3483460545539856, "learning_rate": 9.970094389388844e-05, "loss": 0.0387, "step": 8310 }, { "grad_norm": 0.38088130950927734, "learning_rate": 9.969913544869728e-05, "loss": 0.0466, "step": 8320 }, { "grad_norm": 0.23976774513721466, "learning_rate": 9.96973215684834e-05, "loss": 0.0462, "step": 8330 }, { "grad_norm": 0.3099791705608368, "learning_rate": 9.969550225344513e-05, "loss": 0.0454, "step": 8340 }, { "grad_norm": 0.3557770848274231, "learning_rate": 9.969367750378147e-05, "loss": 0.0506, "step": 8350 }, { "grad_norm": 0.3513340353965759, "learning_rate": 9.969184731969194e-05, "loss": 0.0421, "step": 8360 }, { "grad_norm": 0.2761561870574951, "learning_rate": 9.96900117013767e-05, "loss": 0.0415, "step": 8370 }, { "grad_norm": 0.34934836626052856, "learning_rate": 9.96881706490365e-05, "loss": 0.0486, "step": 8380 }, { "grad_norm": 0.37592241168022156, "learning_rate": 9.968632416287265e-05, "loss": 0.0386, "step": 8390 }, { "grad_norm": 0.3632199168205261, "learning_rate": 9.96844722430871e-05, "loss": 0.0443, "step": 8400 }, { "grad_norm": 0.3301934003829956, "learning_rate": 9.968261488988235e-05, "loss": 0.0487, "step": 8410 }, { "grad_norm": 0.26254361867904663, "learning_rate": 9.968075210346155e-05, "loss": 0.0387, "step": 8420 }, { "grad_norm": 0.27894896268844604, "learning_rate": 9.967888388402839e-05, "loss": 0.0458, "step": 8430 }, { "grad_norm": 0.2564401924610138, "learning_rate": 9.967701023178717e-05, "loss": 0.0405, "step": 8440 }, { "grad_norm": 0.24118538200855255, "learning_rate": 9.967513114694282e-05, "loss": 0.0439, "step": 8450 }, { "grad_norm": 0.29890722036361694, "learning_rate": 9.967324662970079e-05, "loss": 0.0401, "step": 8460 }, { "grad_norm": 0.30617642402648926, "learning_rate": 9.96713566802672e-05, "loss": 0.0373, "step": 8470 }, { "grad_norm": 0.31956931948661804, "learning_rate": 9.966946129884873e-05, "loss": 0.0435, "step": 8480 }, { "grad_norm": 0.35663431882858276, "learning_rate": 9.966756048565265e-05, "loss": 0.045, "step": 8490 }, { "grad_norm": 0.30841606855392456, "learning_rate": 9.966565424088681e-05, "loss": 0.0397, "step": 8500 }, { "grad_norm": 0.3382907211780548, "learning_rate": 9.96637425647597e-05, "loss": 0.0441, "step": 8510 }, { "grad_norm": 0.38910001516342163, "learning_rate": 9.966182545748038e-05, "loss": 0.0425, "step": 8520 }, { "grad_norm": 0.32495036721229553, "learning_rate": 9.96599029192585e-05, "loss": 0.0415, "step": 8530 }, { "grad_norm": 0.34539246559143066, "learning_rate": 9.965797495030428e-05, "loss": 0.0426, "step": 8540 }, { "grad_norm": 0.31335124373435974, "learning_rate": 9.96560415508286e-05, "loss": 0.044, "step": 8550 }, { "grad_norm": 0.28873538970947266, "learning_rate": 9.965410272104286e-05, "loss": 0.042, "step": 8560 }, { "grad_norm": 0.33544984459877014, "learning_rate": 9.96521584611591e-05, "loss": 0.0396, "step": 8570 }, { "grad_norm": 0.27997300028800964, "learning_rate": 9.965020877138994e-05, "loss": 0.044, "step": 8580 }, { "grad_norm": 0.3328927755355835, "learning_rate": 9.964825365194861e-05, "loss": 0.0393, "step": 8590 }, { "grad_norm": 0.38286277651786804, "learning_rate": 9.96462931030489e-05, "loss": 0.0388, "step": 8600 }, { "grad_norm": 0.3613712191581726, "learning_rate": 9.96443271249052e-05, "loss": 0.0399, "step": 8610 }, { "grad_norm": 0.35003313422203064, "learning_rate": 9.964235571773255e-05, "loss": 0.0404, "step": 8620 }, { "grad_norm": 0.2749005854129791, "learning_rate": 9.96403788817465e-05, "loss": 0.0482, "step": 8630 }, { "grad_norm": 0.2717680335044861, "learning_rate": 9.963839661716325e-05, "loss": 0.0402, "step": 8640 }, { "grad_norm": 0.3392567038536072, "learning_rate": 9.963640892419958e-05, "loss": 0.038, "step": 8650 }, { "grad_norm": 0.2749521732330322, "learning_rate": 9.963441580307286e-05, "loss": 0.0386, "step": 8660 }, { "grad_norm": 0.5174877643585205, "learning_rate": 9.963241725400104e-05, "loss": 0.0472, "step": 8670 }, { "grad_norm": 0.290759801864624, "learning_rate": 9.963041327720271e-05, "loss": 0.0426, "step": 8680 }, { "grad_norm": 0.31211310625076294, "learning_rate": 9.962840387289697e-05, "loss": 0.0433, "step": 8690 }, { "grad_norm": 0.3331458270549774, "learning_rate": 9.962638904130363e-05, "loss": 0.0508, "step": 8700 }, { "grad_norm": 0.3776087164878845, "learning_rate": 9.962436878264298e-05, "loss": 0.0397, "step": 8710 }, { "grad_norm": 0.395345538854599, "learning_rate": 9.962234309713598e-05, "loss": 0.0392, "step": 8720 }, { "grad_norm": 0.36826735734939575, "learning_rate": 9.962031198500414e-05, "loss": 0.0427, "step": 8730 }, { "grad_norm": 0.34162136912345886, "learning_rate": 9.961827544646958e-05, "loss": 0.0497, "step": 8740 }, { "grad_norm": 0.3369276523590088, "learning_rate": 9.961623348175501e-05, "loss": 0.0423, "step": 8750 }, { "grad_norm": 0.3341856002807617, "learning_rate": 9.961418609108377e-05, "loss": 0.0474, "step": 8760 }, { "grad_norm": 0.323078989982605, "learning_rate": 9.961213327467971e-05, "loss": 0.0523, "step": 8770 }, { "grad_norm": 0.2716014087200165, "learning_rate": 9.961007503276736e-05, "loss": 0.0389, "step": 8780 }, { "grad_norm": 0.31872034072875977, "learning_rate": 9.960801136557179e-05, "loss": 0.0394, "step": 8790 }, { "grad_norm": 0.30677759647369385, "learning_rate": 9.960594227331866e-05, "loss": 0.0445, "step": 8800 }, { "grad_norm": 0.35119205713272095, "learning_rate": 9.960386775623429e-05, "loss": 0.0436, "step": 8810 }, { "grad_norm": 0.31789064407348633, "learning_rate": 9.96017878145455e-05, "loss": 0.0441, "step": 8820 }, { "grad_norm": 0.27407166361808777, "learning_rate": 9.959970244847977e-05, "loss": 0.0384, "step": 8830 }, { "grad_norm": 0.38659462332725525, "learning_rate": 9.959761165826518e-05, "loss": 0.0475, "step": 8840 }, { "grad_norm": 0.3593680262565613, "learning_rate": 9.959551544413033e-05, "loss": 0.0416, "step": 8850 }, { "grad_norm": 0.3252052068710327, "learning_rate": 9.959341380630448e-05, "loss": 0.0376, "step": 8860 }, { "grad_norm": 0.3545821011066437, "learning_rate": 9.959130674501746e-05, "loss": 0.0382, "step": 8870 }, { "grad_norm": 0.3660053610801697, "learning_rate": 9.958919426049968e-05, "loss": 0.0419, "step": 8880 }, { "grad_norm": 0.3457874357700348, "learning_rate": 9.958707635298219e-05, "loss": 0.0386, "step": 8890 }, { "grad_norm": 0.3626917600631714, "learning_rate": 9.958495302269657e-05, "loss": 0.0456, "step": 8900 }, { "grad_norm": 0.276231974363327, "learning_rate": 9.958282426987503e-05, "loss": 0.036, "step": 8910 }, { "grad_norm": 0.3036375939846039, "learning_rate": 9.95806900947504e-05, "loss": 0.0407, "step": 8920 }, { "grad_norm": 0.2762771546840668, "learning_rate": 9.957855049755604e-05, "loss": 0.0432, "step": 8930 }, { "grad_norm": 0.34881067276000977, "learning_rate": 9.957640547852593e-05, "loss": 0.0352, "step": 8940 }, { "grad_norm": 0.2961525022983551, "learning_rate": 9.957425503789466e-05, "loss": 0.0418, "step": 8950 }, { "grad_norm": 0.3482891619205475, "learning_rate": 9.957209917589738e-05, "loss": 0.038, "step": 8960 }, { "grad_norm": 0.3303370177745819, "learning_rate": 9.956993789276987e-05, "loss": 0.0397, "step": 8970 }, { "grad_norm": 0.2951498329639435, "learning_rate": 9.956777118874847e-05, "loss": 0.05, "step": 8980 }, { "grad_norm": 0.3262617886066437, "learning_rate": 9.956559906407016e-05, "loss": 0.0431, "step": 8990 }, { "grad_norm": 0.33903032541275024, "learning_rate": 9.956342151897245e-05, "loss": 0.0398, "step": 9000 }, { "grad_norm": 0.2506627142429352, "learning_rate": 9.956123855369346e-05, "loss": 0.0345, "step": 9010 }, { "grad_norm": 0.3417326807975769, "learning_rate": 9.955905016847196e-05, "loss": 0.0409, "step": 9020 }, { "grad_norm": 0.32897186279296875, "learning_rate": 9.955685636354723e-05, "loss": 0.0392, "step": 9030 }, { "grad_norm": 0.27261364459991455, "learning_rate": 9.95546571391592e-05, "loss": 0.0379, "step": 9040 }, { "grad_norm": 0.2902664244174957, "learning_rate": 9.955245249554837e-05, "loss": 0.0392, "step": 9050 }, { "grad_norm": 0.30898264050483704, "learning_rate": 9.955024243295582e-05, "loss": 0.0397, "step": 9060 }, { "grad_norm": 0.31254178285598755, "learning_rate": 9.954802695162328e-05, "loss": 0.044, "step": 9070 }, { "grad_norm": 0.34831714630126953, "learning_rate": 9.954580605179302e-05, "loss": 0.0349, "step": 9080 }, { "grad_norm": 0.3405020833015442, "learning_rate": 9.954357973370788e-05, "loss": 0.0362, "step": 9090 }, { "grad_norm": 0.28890979290008545, "learning_rate": 9.954134799761135e-05, "loss": 0.032, "step": 9100 }, { "grad_norm": 0.35007214546203613, "learning_rate": 9.953911084374748e-05, "loss": 0.0354, "step": 9110 }, { "grad_norm": 0.26924195885658264, "learning_rate": 9.953686827236093e-05, "loss": 0.0392, "step": 9120 }, { "grad_norm": 0.30440592765808105, "learning_rate": 9.953462028369695e-05, "loss": 0.0378, "step": 9130 }, { "grad_norm": 0.3043285608291626, "learning_rate": 9.953236687800136e-05, "loss": 0.0358, "step": 9140 }, { "grad_norm": 0.2694036364555359, "learning_rate": 9.95301080555206e-05, "loss": 0.0454, "step": 9150 }, { "grad_norm": 0.3339873254299164, "learning_rate": 9.952784381650171e-05, "loss": 0.0457, "step": 9160 }, { "grad_norm": 0.334989994764328, "learning_rate": 9.952557416119226e-05, "loss": 0.0469, "step": 9170 }, { "grad_norm": 0.3856651782989502, "learning_rate": 9.95232990898405e-05, "loss": 0.0458, "step": 9180 }, { "grad_norm": 0.2960132956504822, "learning_rate": 9.95210186026952e-05, "loss": 0.0446, "step": 9190 }, { "grad_norm": 0.285835862159729, "learning_rate": 9.951873270000576e-05, "loss": 0.0393, "step": 9200 }, { "grad_norm": 0.28047874569892883, "learning_rate": 9.951644138202216e-05, "loss": 0.0427, "step": 9210 }, { "grad_norm": 0.3339173495769501, "learning_rate": 9.951414464899498e-05, "loss": 0.0453, "step": 9220 }, { "grad_norm": 0.31642842292785645, "learning_rate": 9.951184250117538e-05, "loss": 0.0425, "step": 9230 }, { "grad_norm": 0.2661443054676056, "learning_rate": 9.950953493881513e-05, "loss": 0.0369, "step": 9240 }, { "grad_norm": 0.41367796063423157, "learning_rate": 9.950722196216658e-05, "loss": 0.0358, "step": 9250 }, { "grad_norm": 0.40667545795440674, "learning_rate": 9.950490357148265e-05, "loss": 0.0399, "step": 9260 }, { "grad_norm": 0.3225765526294708, "learning_rate": 9.950257976701692e-05, "loss": 0.0439, "step": 9270 }, { "grad_norm": 0.28053566813468933, "learning_rate": 9.950025054902348e-05, "loss": 0.0378, "step": 9280 }, { "grad_norm": 0.26401248574256897, "learning_rate": 9.949791591775706e-05, "loss": 0.0311, "step": 9290 }, { "grad_norm": 0.23211705684661865, "learning_rate": 9.949557587347298e-05, "loss": 0.0433, "step": 9300 }, { "grad_norm": 0.2413121908903122, "learning_rate": 9.949323041642713e-05, "loss": 0.038, "step": 9310 }, { "grad_norm": 0.31274309754371643, "learning_rate": 9.949087954687602e-05, "loss": 0.0416, "step": 9320 }, { "grad_norm": 0.2675059139728546, "learning_rate": 9.948852326507672e-05, "loss": 0.0341, "step": 9330 }, { "grad_norm": 0.38406431674957275, "learning_rate": 9.948616157128694e-05, "loss": 0.042, "step": 9340 }, { "grad_norm": 0.29066935181617737, "learning_rate": 9.948379446576493e-05, "loss": 0.0378, "step": 9350 }, { "grad_norm": 0.2123051881790161, "learning_rate": 9.948142194876952e-05, "loss": 0.0333, "step": 9360 }, { "grad_norm": 0.3063759207725525, "learning_rate": 9.947904402056024e-05, "loss": 0.04, "step": 9370 }, { "grad_norm": 0.28675004839897156, "learning_rate": 9.947666068139708e-05, "loss": 0.0355, "step": 9380 }, { "grad_norm": 0.29044654965400696, "learning_rate": 9.947427193154071e-05, "loss": 0.0451, "step": 9390 }, { "grad_norm": 0.35066404938697815, "learning_rate": 9.947187777125233e-05, "loss": 0.0479, "step": 9400 }, { "grad_norm": 0.3887355327606201, "learning_rate": 9.946947820079377e-05, "loss": 0.0382, "step": 9410 }, { "grad_norm": 0.31257364153862, "learning_rate": 9.946707322042747e-05, "loss": 0.0404, "step": 9420 }, { "grad_norm": 0.336124986410141, "learning_rate": 9.94646628304164e-05, "loss": 0.0339, "step": 9430 }, { "grad_norm": 0.3071836829185486, "learning_rate": 9.946224703102418e-05, "loss": 0.0369, "step": 9440 }, { "grad_norm": 0.2445797473192215, "learning_rate": 9.945982582251498e-05, "loss": 0.0386, "step": 9450 }, { "grad_norm": 0.41371241211891174, "learning_rate": 9.94573992051536e-05, "loss": 0.0358, "step": 9460 }, { "grad_norm": 0.3408028781414032, "learning_rate": 9.94549671792054e-05, "loss": 0.04, "step": 9470 }, { "grad_norm": 0.30025407671928406, "learning_rate": 9.945252974493635e-05, "loss": 0.0413, "step": 9480 }, { "grad_norm": 0.36908212304115295, "learning_rate": 9.9450086902613e-05, "loss": 0.0422, "step": 9490 }, { "grad_norm": 0.24774464964866638, "learning_rate": 9.944763865250248e-05, "loss": 0.0433, "step": 9500 }, { "grad_norm": 0.30670231580734253, "learning_rate": 9.944518499487254e-05, "loss": 0.043, "step": 9510 }, { "grad_norm": 0.3244918882846832, "learning_rate": 9.944272592999151e-05, "loss": 0.0384, "step": 9520 }, { "grad_norm": 0.3003990650177002, "learning_rate": 9.94402614581283e-05, "loss": 0.0423, "step": 9530 }, { "grad_norm": 0.3623422682285309, "learning_rate": 9.943779157955244e-05, "loss": 0.0418, "step": 9540 }, { "grad_norm": 0.40311911702156067, "learning_rate": 9.943531629453403e-05, "loss": 0.0419, "step": 9550 }, { "grad_norm": 0.3470112383365631, "learning_rate": 9.943283560334375e-05, "loss": 0.044, "step": 9560 }, { "grad_norm": 0.28404974937438965, "learning_rate": 9.943034950625288e-05, "loss": 0.0389, "step": 9570 }, { "grad_norm": 0.28359803557395935, "learning_rate": 9.942785800353332e-05, "loss": 0.0442, "step": 9580 }, { "grad_norm": 0.3119894564151764, "learning_rate": 9.942536109545751e-05, "loss": 0.0397, "step": 9590 }, { "grad_norm": 0.2728886008262634, "learning_rate": 9.942285878229853e-05, "loss": 0.0407, "step": 9600 }, { "grad_norm": 0.28049665689468384, "learning_rate": 9.942035106433001e-05, "loss": 0.0357, "step": 9610 }, { "grad_norm": 0.28084060549736023, "learning_rate": 9.94178379418262e-05, "loss": 0.0382, "step": 9620 }, { "grad_norm": 0.27910181879997253, "learning_rate": 9.941531941506194e-05, "loss": 0.0405, "step": 9630 }, { "grad_norm": 0.31066983938217163, "learning_rate": 9.941279548431263e-05, "loss": 0.0368, "step": 9640 }, { "grad_norm": 0.36518388986587524, "learning_rate": 9.941026614985431e-05, "loss": 0.0383, "step": 9650 }, { "grad_norm": 0.36640992760658264, "learning_rate": 9.940773141196357e-05, "loss": 0.0363, "step": 9660 }, { "grad_norm": 0.27404695749282837, "learning_rate": 9.94051912709176e-05, "loss": 0.0357, "step": 9670 }, { "grad_norm": 0.2777780294418335, "learning_rate": 9.940264572699421e-05, "loss": 0.038, "step": 9680 }, { "grad_norm": 0.4063309133052826, "learning_rate": 9.940009478047174e-05, "loss": 0.0418, "step": 9690 }, { "grad_norm": 0.3265550136566162, "learning_rate": 9.939753843162918e-05, "loss": 0.0375, "step": 9700 }, { "grad_norm": 0.30646926164627075, "learning_rate": 9.939497668074609e-05, "loss": 0.0326, "step": 9710 }, { "grad_norm": 0.30326345562934875, "learning_rate": 9.93924095281026e-05, "loss": 0.0401, "step": 9720 }, { "grad_norm": 0.32256752252578735, "learning_rate": 9.938983697397948e-05, "loss": 0.0431, "step": 9730 }, { "grad_norm": 0.2795185148715973, "learning_rate": 9.938725901865805e-05, "loss": 0.0416, "step": 9740 }, { "grad_norm": 0.2812884449958801, "learning_rate": 9.93846756624202e-05, "loss": 0.0419, "step": 9750 }, { "grad_norm": 0.2579319477081299, "learning_rate": 9.938208690554849e-05, "loss": 0.0366, "step": 9760 }, { "grad_norm": 0.22648178040981293, "learning_rate": 9.9379492748326e-05, "loss": 0.0412, "step": 9770 }, { "grad_norm": 0.23163215816020966, "learning_rate": 9.937689319103641e-05, "loss": 0.034, "step": 9780 }, { "grad_norm": 0.24152390658855438, "learning_rate": 9.937428823396404e-05, "loss": 0.0375, "step": 9790 }, { "grad_norm": 0.25433632731437683, "learning_rate": 9.937167787739372e-05, "loss": 0.0351, "step": 9800 }, { "grad_norm": 0.24768932163715363, "learning_rate": 9.936906212161095e-05, "loss": 0.0354, "step": 9810 }, { "grad_norm": 0.34806305170059204, "learning_rate": 9.936644096690176e-05, "loss": 0.0376, "step": 9820 }, { "grad_norm": 0.3124347925186157, "learning_rate": 9.936381441355282e-05, "loss": 0.0416, "step": 9830 }, { "grad_norm": 0.2791180908679962, "learning_rate": 9.936118246185136e-05, "loss": 0.0416, "step": 9840 }, { "grad_norm": 0.288892924785614, "learning_rate": 9.935854511208518e-05, "loss": 0.0428, "step": 9850 }, { "grad_norm": 0.26193681359291077, "learning_rate": 9.935590236454272e-05, "loss": 0.0429, "step": 9860 }, { "grad_norm": 0.3118732273578644, "learning_rate": 9.935325421951298e-05, "loss": 0.0333, "step": 9870 }, { "grad_norm": 0.27053165435791016, "learning_rate": 9.935060067728557e-05, "loss": 0.0378, "step": 9880 }, { "grad_norm": 0.4198669493198395, "learning_rate": 9.934794173815067e-05, "loss": 0.046, "step": 9890 }, { "grad_norm": 0.28397202491760254, "learning_rate": 9.934527740239906e-05, "loss": 0.0367, "step": 9900 }, { "grad_norm": 0.3555208444595337, "learning_rate": 9.934260767032209e-05, "loss": 0.0438, "step": 9910 }, { "grad_norm": 0.24680449068546295, "learning_rate": 9.933993254221172e-05, "loss": 0.0359, "step": 9920 }, { "grad_norm": 0.2775205671787262, "learning_rate": 9.933725201836053e-05, "loss": 0.0368, "step": 9930 }, { "grad_norm": 0.2665872275829315, "learning_rate": 9.933456609906162e-05, "loss": 0.033, "step": 9940 }, { "grad_norm": 0.29138848185539246, "learning_rate": 9.933187478460875e-05, "loss": 0.0454, "step": 9950 }, { "grad_norm": 0.3241884410381317, "learning_rate": 9.93291780752962e-05, "loss": 0.0374, "step": 9960 }, { "grad_norm": 0.3424249589443207, "learning_rate": 9.932647597141893e-05, "loss": 0.0367, "step": 9970 }, { "grad_norm": 0.29015541076660156, "learning_rate": 9.932376847327239e-05, "loss": 0.0429, "step": 9980 }, { "grad_norm": 0.3425186276435852, "learning_rate": 9.932105558115268e-05, "loss": 0.0372, "step": 9990 }, { "grad_norm": 0.32653483748435974, "learning_rate": 9.931833729535651e-05, "loss": 0.0374, "step": 10000 }, { "grad_norm": 0.338046669960022, "learning_rate": 9.931561361618111e-05, "loss": 0.036, "step": 10010 }, { "grad_norm": 0.2878689467906952, "learning_rate": 9.931288454392435e-05, "loss": 0.0426, "step": 10020 }, { "grad_norm": 0.3509298861026764, "learning_rate": 9.931015007888467e-05, "loss": 0.0376, "step": 10030 }, { "grad_norm": 0.308503657579422, "learning_rate": 9.930741022136112e-05, "loss": 0.0466, "step": 10040 }, { "grad_norm": 0.30734995007514954, "learning_rate": 9.930466497165333e-05, "loss": 0.0384, "step": 10050 }, { "grad_norm": 0.29953956604003906, "learning_rate": 9.93019143300615e-05, "loss": 0.0365, "step": 10060 }, { "grad_norm": 0.3714286983013153, "learning_rate": 9.929915829688644e-05, "loss": 0.0347, "step": 10070 }, { "grad_norm": 0.3724853992462158, "learning_rate": 9.929639687242955e-05, "loss": 0.038, "step": 10080 }, { "grad_norm": 0.28260573744773865, "learning_rate": 9.929363005699281e-05, "loss": 0.0338, "step": 10090 }, { "grad_norm": 0.26103684306144714, "learning_rate": 9.92908578508788e-05, "loss": 0.0409, "step": 10100 }, { "grad_norm": 0.3408873379230499, "learning_rate": 9.928808025439069e-05, "loss": 0.0459, "step": 10110 }, { "grad_norm": 0.3328457176685333, "learning_rate": 9.928529726783223e-05, "loss": 0.0412, "step": 10120 }, { "grad_norm": 0.28912538290023804, "learning_rate": 9.928250889150774e-05, "loss": 0.0424, "step": 10130 }, { "grad_norm": 0.30038216710090637, "learning_rate": 9.92797151257222e-05, "loss": 0.0354, "step": 10140 }, { "grad_norm": 0.2530566155910492, "learning_rate": 9.927691597078108e-05, "loss": 0.0374, "step": 10150 }, { "grad_norm": 0.20508994162082672, "learning_rate": 9.927411142699053e-05, "loss": 0.0364, "step": 10160 }, { "grad_norm": 0.2573155164718628, "learning_rate": 9.927130149465725e-05, "loss": 0.0385, "step": 10170 }, { "grad_norm": 0.3177042007446289, "learning_rate": 9.92684861740885e-05, "loss": 0.0393, "step": 10180 }, { "grad_norm": 0.34961721301078796, "learning_rate": 9.926566546559217e-05, "loss": 0.0408, "step": 10190 }, { "grad_norm": 0.24970099329948425, "learning_rate": 9.926283936947673e-05, "loss": 0.0394, "step": 10200 }, { "grad_norm": 0.3942928612232208, "learning_rate": 9.926000788605126e-05, "loss": 0.0402, "step": 10210 }, { "grad_norm": 0.25810423493385315, "learning_rate": 9.92571710156254e-05, "loss": 0.0364, "step": 10220 }, { "grad_norm": 0.271828293800354, "learning_rate": 9.925432875850936e-05, "loss": 0.0348, "step": 10230 }, { "grad_norm": 0.3504471182823181, "learning_rate": 9.925148111501396e-05, "loss": 0.0355, "step": 10240 }, { "grad_norm": 0.2947179973125458, "learning_rate": 9.924862808545066e-05, "loss": 0.038, "step": 10250 }, { "grad_norm": 0.28481149673461914, "learning_rate": 9.924576967013141e-05, "loss": 0.0351, "step": 10260 }, { "grad_norm": 0.2645990550518036, "learning_rate": 9.924290586936887e-05, "loss": 0.0349, "step": 10270 }, { "grad_norm": 0.29043805599212646, "learning_rate": 9.924003668347614e-05, "loss": 0.0391, "step": 10280 }, { "grad_norm": 0.23443400859832764, "learning_rate": 9.923716211276704e-05, "loss": 0.0318, "step": 10290 }, { "grad_norm": 0.26179105043411255, "learning_rate": 9.923428215755594e-05, "loss": 0.0351, "step": 10300 }, { "grad_norm": 0.2588363587856293, "learning_rate": 9.923139681815775e-05, "loss": 0.0369, "step": 10310 }, { "grad_norm": 0.26515117287635803, "learning_rate": 9.922850609488801e-05, "loss": 0.0367, "step": 10320 }, { "grad_norm": 0.3369617760181427, "learning_rate": 9.922560998806287e-05, "loss": 0.0334, "step": 10330 }, { "grad_norm": 0.3418079912662506, "learning_rate": 9.922270849799905e-05, "loss": 0.0352, "step": 10340 }, { "grad_norm": 0.2471759170293808, "learning_rate": 9.92198016250138e-05, "loss": 0.0366, "step": 10350 }, { "grad_norm": 0.25680112838745117, "learning_rate": 9.921688936942506e-05, "loss": 0.0322, "step": 10360 }, { "grad_norm": 0.3138902485370636, "learning_rate": 9.921397173155129e-05, "loss": 0.0401, "step": 10370 }, { "grad_norm": 0.31706634163856506, "learning_rate": 9.921104871171157e-05, "loss": 0.0349, "step": 10380 }, { "grad_norm": 0.21999526023864746, "learning_rate": 9.920812031022554e-05, "loss": 0.0338, "step": 10390 }, { "grad_norm": 0.3151225447654724, "learning_rate": 9.920518652741348e-05, "loss": 0.0364, "step": 10400 }, { "grad_norm": 0.247167706489563, "learning_rate": 9.920224736359618e-05, "loss": 0.037, "step": 10410 }, { "grad_norm": 0.32440727949142456, "learning_rate": 9.91993028190951e-05, "loss": 0.047, "step": 10420 }, { "grad_norm": 0.2657446563243866, "learning_rate": 9.919635289423222e-05, "loss": 0.0395, "step": 10430 }, { "grad_norm": 0.314200222492218, "learning_rate": 9.919339758933015e-05, "loss": 0.0524, "step": 10440 }, { "grad_norm": 0.32071784138679504, "learning_rate": 9.919043690471209e-05, "loss": 0.0393, "step": 10450 }, { "grad_norm": 0.23399348556995392, "learning_rate": 9.91874708407018e-05, "loss": 0.0394, "step": 10460 }, { "grad_norm": 0.33693015575408936, "learning_rate": 9.918449939762367e-05, "loss": 0.0367, "step": 10470 }, { "grad_norm": 0.341675728559494, "learning_rate": 9.91815225758026e-05, "loss": 0.0378, "step": 10480 }, { "grad_norm": 0.23338276147842407, "learning_rate": 9.917854037556419e-05, "loss": 0.0358, "step": 10490 }, { "grad_norm": 0.22320342063903809, "learning_rate": 9.917555279723454e-05, "loss": 0.0423, "step": 10500 }, { "grad_norm": 0.31569868326187134, "learning_rate": 9.917255984114036e-05, "loss": 0.0412, "step": 10510 }, { "grad_norm": 0.35966044664382935, "learning_rate": 9.916956150760896e-05, "loss": 0.037, "step": 10520 }, { "grad_norm": 0.2206423580646515, "learning_rate": 9.916655779696826e-05, "loss": 0.0366, "step": 10530 }, { "grad_norm": 0.2504793107509613, "learning_rate": 9.916354870954671e-05, "loss": 0.0392, "step": 10540 }, { "grad_norm": 0.3186176121234894, "learning_rate": 9.91605342456734e-05, "loss": 0.0393, "step": 10550 }, { "grad_norm": 0.25853437185287476, "learning_rate": 9.915751440567795e-05, "loss": 0.0352, "step": 10560 }, { "grad_norm": 0.2561294436454773, "learning_rate": 9.915448918989066e-05, "loss": 0.0349, "step": 10570 }, { "grad_norm": 0.2891503572463989, "learning_rate": 9.915145859864232e-05, "loss": 0.0434, "step": 10580 }, { "grad_norm": 0.20459219813346863, "learning_rate": 9.914842263226437e-05, "loss": 0.0398, "step": 10590 }, { "grad_norm": 0.2600446343421936, "learning_rate": 9.914538129108882e-05, "loss": 0.0444, "step": 10600 }, { "grad_norm": 0.35860171914100647, "learning_rate": 9.914233457544825e-05, "loss": 0.0459, "step": 10610 }, { "grad_norm": 0.3338770568370819, "learning_rate": 9.913928248567586e-05, "loss": 0.0451, "step": 10620 }, { "grad_norm": 0.28984054923057556, "learning_rate": 9.913622502210542e-05, "loss": 0.044, "step": 10630 }, { "grad_norm": 0.3556727468967438, "learning_rate": 9.913316218507128e-05, "loss": 0.0459, "step": 10640 }, { "grad_norm": 0.3310454487800598, "learning_rate": 9.91300939749084e-05, "loss": 0.0391, "step": 10650 }, { "grad_norm": 0.2811560034751892, "learning_rate": 9.91270203919523e-05, "loss": 0.0387, "step": 10660 }, { "grad_norm": 0.23647864162921906, "learning_rate": 9.912394143653912e-05, "loss": 0.0414, "step": 10670 }, { "grad_norm": 0.2599574029445648, "learning_rate": 9.912085710900555e-05, "loss": 0.0381, "step": 10680 }, { "grad_norm": 0.3551620543003082, "learning_rate": 9.911776740968892e-05, "loss": 0.0423, "step": 10690 }, { "grad_norm": 0.2546861469745636, "learning_rate": 9.911467233892709e-05, "loss": 0.0358, "step": 10700 }, { "grad_norm": 0.33418604731559753, "learning_rate": 9.911157189705853e-05, "loss": 0.0363, "step": 10710 }, { "grad_norm": 0.34047451615333557, "learning_rate": 9.910846608442229e-05, "loss": 0.0339, "step": 10720 }, { "grad_norm": 0.2969525456428528, "learning_rate": 9.910535490135805e-05, "loss": 0.035, "step": 10730 }, { "grad_norm": 0.2820756733417511, "learning_rate": 9.910223834820603e-05, "loss": 0.0378, "step": 10740 }, { "grad_norm": 0.3144715428352356, "learning_rate": 9.909911642530703e-05, "loss": 0.0405, "step": 10750 }, { "grad_norm": 0.2981959581375122, "learning_rate": 9.909598913300249e-05, "loss": 0.0322, "step": 10760 }, { "grad_norm": 0.343837171792984, "learning_rate": 9.909285647163438e-05, "loss": 0.0383, "step": 10770 }, { "grad_norm": 0.2608136832714081, "learning_rate": 9.908971844154531e-05, "loss": 0.0336, "step": 10780 }, { "grad_norm": 0.2751438319683075, "learning_rate": 9.908657504307843e-05, "loss": 0.033, "step": 10790 }, { "grad_norm": 0.3554334044456482, "learning_rate": 9.908342627657751e-05, "loss": 0.0341, "step": 10800 }, { "grad_norm": 0.2751835584640503, "learning_rate": 9.908027214238689e-05, "loss": 0.0361, "step": 10810 }, { "grad_norm": 0.2836107015609741, "learning_rate": 9.90771126408515e-05, "loss": 0.0338, "step": 10820 }, { "grad_norm": 0.33460086584091187, "learning_rate": 9.907394777231685e-05, "loss": 0.0379, "step": 10830 }, { "grad_norm": 0.24791830778121948, "learning_rate": 9.907077753712905e-05, "loss": 0.0378, "step": 10840 }, { "grad_norm": 0.28135332465171814, "learning_rate": 9.906760193563482e-05, "loss": 0.0355, "step": 10850 }, { "grad_norm": 0.3197508156299591, "learning_rate": 9.906442096818139e-05, "loss": 0.0387, "step": 10860 }, { "grad_norm": 0.23204900324344635, "learning_rate": 9.906123463511665e-05, "loss": 0.0365, "step": 10870 }, { "grad_norm": 0.2318378984928131, "learning_rate": 9.905804293678907e-05, "loss": 0.0321, "step": 10880 }, { "grad_norm": 0.2697940170764923, "learning_rate": 9.905484587354766e-05, "loss": 0.0368, "step": 10890 }, { "grad_norm": 0.24537189304828644, "learning_rate": 9.905164344574205e-05, "loss": 0.0379, "step": 10900 }, { "grad_norm": 0.2725086808204651, "learning_rate": 9.904843565372248e-05, "loss": 0.0387, "step": 10910 }, { "grad_norm": 0.27608540654182434, "learning_rate": 9.904522249783972e-05, "loss": 0.0368, "step": 10920 }, { "grad_norm": 0.32779523730278015, "learning_rate": 9.904200397844517e-05, "loss": 0.0377, "step": 10930 }, { "grad_norm": 0.29891571402549744, "learning_rate": 9.903878009589078e-05, "loss": 0.0379, "step": 10940 }, { "grad_norm": 0.2554378807544708, "learning_rate": 9.903555085052915e-05, "loss": 0.0404, "step": 10950 }, { "grad_norm": 0.25442343950271606, "learning_rate": 9.903231624271338e-05, "loss": 0.0371, "step": 10960 }, { "grad_norm": 0.24989387392997742, "learning_rate": 9.902907627279724e-05, "loss": 0.0384, "step": 10970 }, { "grad_norm": 0.28085193037986755, "learning_rate": 9.902583094113504e-05, "loss": 0.0372, "step": 10980 }, { "grad_norm": 0.26736003160476685, "learning_rate": 9.902258024808168e-05, "loss": 0.0347, "step": 10990 }, { "grad_norm": 0.2740558385848999, "learning_rate": 9.901932419399264e-05, "loss": 0.0317, "step": 11000 }, { "grad_norm": 0.2704288363456726, "learning_rate": 9.9016062779224e-05, "loss": 0.0346, "step": 11010 }, { "grad_norm": 0.29740211367607117, "learning_rate": 9.901279600413242e-05, "loss": 0.0401, "step": 11020 }, { "grad_norm": 0.3106135427951813, "learning_rate": 9.900952386907518e-05, "loss": 0.04, "step": 11030 }, { "grad_norm": 0.19901418685913086, "learning_rate": 9.90062463744101e-05, "loss": 0.0372, "step": 11040 }, { "grad_norm": 0.2549898028373718, "learning_rate": 9.900296352049558e-05, "loss": 0.0383, "step": 11050 }, { "grad_norm": 0.27674365043640137, "learning_rate": 9.899967530769065e-05, "loss": 0.0327, "step": 11060 }, { "grad_norm": 0.34145331382751465, "learning_rate": 9.899638173635489e-05, "loss": 0.0305, "step": 11070 }, { "grad_norm": 0.30271825194358826, "learning_rate": 9.899308280684849e-05, "loss": 0.0369, "step": 11080 }, { "grad_norm": 0.32605594396591187, "learning_rate": 9.898977851953222e-05, "loss": 0.0393, "step": 11090 }, { "grad_norm": 0.31150782108306885, "learning_rate": 9.898646887476741e-05, "loss": 0.0344, "step": 11100 }, { "grad_norm": 0.2890815734863281, "learning_rate": 9.898315387291603e-05, "loss": 0.0385, "step": 11110 }, { "grad_norm": 0.2508183717727661, "learning_rate": 9.89798335143406e-05, "loss": 0.0363, "step": 11120 }, { "grad_norm": 0.26335063576698303, "learning_rate": 9.897650779940419e-05, "loss": 0.0369, "step": 11130 }, { "grad_norm": 0.2606370449066162, "learning_rate": 9.897317672847054e-05, "loss": 0.0357, "step": 11140 }, { "grad_norm": 0.3481265604496002, "learning_rate": 9.89698403019039e-05, "loss": 0.0375, "step": 11150 }, { "grad_norm": 0.25997719168663025, "learning_rate": 9.896649852006917e-05, "loss": 0.0305, "step": 11160 }, { "grad_norm": 0.2638908624649048, "learning_rate": 9.896315138333177e-05, "loss": 0.0357, "step": 11170 }, { "grad_norm": 0.2705220580101013, "learning_rate": 9.895979889205774e-05, "loss": 0.0338, "step": 11180 }, { "grad_norm": 0.28226375579833984, "learning_rate": 9.895644104661372e-05, "loss": 0.0398, "step": 11190 }, { "grad_norm": 0.3066609501838684, "learning_rate": 9.895307784736691e-05, "loss": 0.0304, "step": 11200 }, { "grad_norm": 0.28880882263183594, "learning_rate": 9.894970929468512e-05, "loss": 0.0325, "step": 11210 }, { "grad_norm": 0.2724096179008484, "learning_rate": 9.89463353889367e-05, "loss": 0.0334, "step": 11220 }, { "grad_norm": 0.31828275322914124, "learning_rate": 9.894295613049065e-05, "loss": 0.033, "step": 11230 }, { "grad_norm": 0.2857796847820282, "learning_rate": 9.893957151971649e-05, "loss": 0.0338, "step": 11240 }, { "grad_norm": 0.20377948880195618, "learning_rate": 9.893618155698436e-05, "loss": 0.0294, "step": 11250 }, { "grad_norm": 0.32193729281425476, "learning_rate": 9.8932786242665e-05, "loss": 0.0339, "step": 11260 }, { "grad_norm": 0.23770539462566376, "learning_rate": 9.89293855771297e-05, "loss": 0.035, "step": 11270 }, { "grad_norm": 0.2801697552204132, "learning_rate": 9.892597956075036e-05, "loss": 0.033, "step": 11280 }, { "grad_norm": 0.2145938128232956, "learning_rate": 9.892256819389947e-05, "loss": 0.0341, "step": 11290 }, { "grad_norm": 0.21691451966762543, "learning_rate": 9.891915147695006e-05, "loss": 0.0349, "step": 11300 }, { "grad_norm": 0.21968045830726624, "learning_rate": 9.891572941027577e-05, "loss": 0.0404, "step": 11310 }, { "grad_norm": 0.2662462890148163, "learning_rate": 9.89123019942509e-05, "loss": 0.038, "step": 11320 }, { "grad_norm": 0.301853209733963, "learning_rate": 9.89088692292502e-05, "loss": 0.0272, "step": 11330 }, { "grad_norm": 0.22630575299263, "learning_rate": 9.89054311156491e-05, "loss": 0.0368, "step": 11340 }, { "grad_norm": 0.27717119455337524, "learning_rate": 9.890198765382357e-05, "loss": 0.0355, "step": 11350 }, { "grad_norm": 0.31197836995124817, "learning_rate": 9.889853884415021e-05, "loss": 0.0332, "step": 11360 }, { "grad_norm": 0.28417932987213135, "learning_rate": 9.889508468700614e-05, "loss": 0.0341, "step": 11370 }, { "grad_norm": 0.25183525681495667, "learning_rate": 9.889162518276915e-05, "loss": 0.0337, "step": 11380 }, { "grad_norm": 0.31927385926246643, "learning_rate": 9.888816033181752e-05, "loss": 0.0376, "step": 11390 }, { "grad_norm": 0.3580258786678314, "learning_rate": 9.888469013453018e-05, "loss": 0.0357, "step": 11400 }, { "grad_norm": 0.23879554867744446, "learning_rate": 9.888121459128663e-05, "loss": 0.0331, "step": 11410 }, { "grad_norm": 0.2634601593017578, "learning_rate": 9.887773370246693e-05, "loss": 0.0326, "step": 11420 }, { "grad_norm": 0.34269988536834717, "learning_rate": 9.887424746845177e-05, "loss": 0.0385, "step": 11430 }, { "grad_norm": 0.32689130306243896, "learning_rate": 9.887075588962239e-05, "loss": 0.0347, "step": 11440 }, { "grad_norm": 0.31771954894065857, "learning_rate": 9.88672589663606e-05, "loss": 0.0338, "step": 11450 }, { "grad_norm": 0.30273914337158203, "learning_rate": 9.886375669904886e-05, "loss": 0.0374, "step": 11460 }, { "grad_norm": 0.298831969499588, "learning_rate": 9.886024908807014e-05, "loss": 0.0341, "step": 11470 }, { "grad_norm": 0.3795946538448334, "learning_rate": 9.885673613380806e-05, "loss": 0.042, "step": 11480 }, { "grad_norm": 0.3070827126502991, "learning_rate": 9.885321783664676e-05, "loss": 0.0361, "step": 11490 }, { "grad_norm": 0.27006158232688904, "learning_rate": 9.884969419697101e-05, "loss": 0.0333, "step": 11500 }, { "grad_norm": 0.3673863112926483, "learning_rate": 9.884616521516614e-05, "loss": 0.0356, "step": 11510 }, { "grad_norm": 0.2599628269672394, "learning_rate": 9.88426308916181e-05, "loss": 0.0311, "step": 11520 }, { "grad_norm": 0.23300522565841675, "learning_rate": 9.883909122671335e-05, "loss": 0.0349, "step": 11530 }, { "grad_norm": 0.2664216458797455, "learning_rate": 9.883554622083904e-05, "loss": 0.0336, "step": 11540 }, { "grad_norm": 0.30841967463493347, "learning_rate": 9.88319958743828e-05, "loss": 0.0378, "step": 11550 }, { "grad_norm": 0.32852256298065186, "learning_rate": 9.882844018773291e-05, "loss": 0.0366, "step": 11560 }, { "grad_norm": 0.3771824836730957, "learning_rate": 9.882487916127823e-05, "loss": 0.0369, "step": 11570 }, { "grad_norm": 0.3298591673374176, "learning_rate": 9.882131279540815e-05, "loss": 0.0388, "step": 11580 }, { "grad_norm": 0.3255273997783661, "learning_rate": 9.881774109051271e-05, "loss": 0.0333, "step": 11590 }, { "grad_norm": 0.40959250926971436, "learning_rate": 9.881416404698252e-05, "loss": 0.0379, "step": 11600 }, { "grad_norm": 0.2275257408618927, "learning_rate": 9.881058166520873e-05, "loss": 0.0344, "step": 11610 }, { "grad_norm": 0.29597413539886475, "learning_rate": 9.880699394558311e-05, "loss": 0.0334, "step": 11620 }, { "grad_norm": 0.2863661050796509, "learning_rate": 9.880340088849801e-05, "loss": 0.0352, "step": 11630 }, { "grad_norm": 0.2242429107427597, "learning_rate": 9.879980249434637e-05, "loss": 0.0339, "step": 11640 }, { "grad_norm": 0.24420741200447083, "learning_rate": 9.879619876352168e-05, "loss": 0.0326, "step": 11650 }, { "grad_norm": 0.25903788208961487, "learning_rate": 9.879258969641809e-05, "loss": 0.0382, "step": 11660 }, { "grad_norm": 0.28297218680381775, "learning_rate": 9.878897529343023e-05, "loss": 0.0345, "step": 11670 }, { "grad_norm": 0.27723363041877747, "learning_rate": 9.878535555495338e-05, "loss": 0.034, "step": 11680 }, { "grad_norm": 0.22799210250377655, "learning_rate": 9.87817304813834e-05, "loss": 0.0305, "step": 11690 }, { "grad_norm": 0.3076604902744293, "learning_rate": 9.877810007311671e-05, "loss": 0.0374, "step": 11700 }, { "grad_norm": 0.2630910575389862, "learning_rate": 9.877446433055035e-05, "loss": 0.0347, "step": 11710 }, { "grad_norm": 0.2782559096813202, "learning_rate": 9.877082325408191e-05, "loss": 0.0421, "step": 11720 }, { "grad_norm": 0.3101753294467926, "learning_rate": 9.876717684410954e-05, "loss": 0.033, "step": 11730 }, { "grad_norm": 0.26600563526153564, "learning_rate": 9.876352510103204e-05, "loss": 0.0415, "step": 11740 }, { "grad_norm": 0.28065043687820435, "learning_rate": 9.875986802524875e-05, "loss": 0.0313, "step": 11750 }, { "grad_norm": 0.2775232791900635, "learning_rate": 9.87562056171596e-05, "loss": 0.0404, "step": 11760 }, { "grad_norm": 0.301629900932312, "learning_rate": 9.875253787716511e-05, "loss": 0.0349, "step": 11770 }, { "grad_norm": 0.26850447058677673, "learning_rate": 9.874886480566637e-05, "loss": 0.0336, "step": 11780 }, { "grad_norm": 0.37638840079307556, "learning_rate": 9.874518640306507e-05, "loss": 0.0348, "step": 11790 }, { "grad_norm": 0.2688868045806885, "learning_rate": 9.874150266976347e-05, "loss": 0.0385, "step": 11800 }, { "grad_norm": 0.2350042462348938, "learning_rate": 9.873781360616443e-05, "loss": 0.0362, "step": 11810 }, { "grad_norm": 0.2753790020942688, "learning_rate": 9.873411921267137e-05, "loss": 0.0358, "step": 11820 }, { "grad_norm": 0.20371095836162567, "learning_rate": 9.873041948968829e-05, "loss": 0.0263, "step": 11830 }, { "grad_norm": 0.350894570350647, "learning_rate": 9.872671443761981e-05, "loss": 0.0389, "step": 11840 }, { "grad_norm": 0.30654940009117126, "learning_rate": 9.872300405687109e-05, "loss": 0.0306, "step": 11850 }, { "grad_norm": 0.31262803077697754, "learning_rate": 9.871928834784792e-05, "loss": 0.0305, "step": 11860 }, { "grad_norm": 0.2943481504917145, "learning_rate": 9.871556731095661e-05, "loss": 0.0396, "step": 11870 }, { "grad_norm": 0.28373125195503235, "learning_rate": 9.871184094660411e-05, "loss": 0.0301, "step": 11880 }, { "grad_norm": 0.2922039330005646, "learning_rate": 9.870810925519791e-05, "loss": 0.0342, "step": 11890 }, { "grad_norm": 0.25745728611946106, "learning_rate": 9.870437223714612e-05, "loss": 0.0312, "step": 11900 }, { "grad_norm": 0.26932528614997864, "learning_rate": 9.87006298928574e-05, "loss": 0.032, "step": 11910 }, { "grad_norm": 0.27368298172950745, "learning_rate": 9.869688222274103e-05, "loss": 0.0343, "step": 11920 }, { "grad_norm": 0.2731693685054779, "learning_rate": 9.869312922720681e-05, "loss": 0.0316, "step": 11930 }, { "grad_norm": 0.26161420345306396, "learning_rate": 9.868937090666521e-05, "loss": 0.0347, "step": 11940 }, { "grad_norm": 0.2183578461408615, "learning_rate": 9.86856072615272e-05, "loss": 0.0334, "step": 11950 }, { "grad_norm": 0.3365993797779083, "learning_rate": 9.868183829220438e-05, "loss": 0.0308, "step": 11960 }, { "grad_norm": 0.20263436436653137, "learning_rate": 9.867806399910893e-05, "loss": 0.0343, "step": 11970 }, { "grad_norm": 0.3392876386642456, "learning_rate": 9.867428438265356e-05, "loss": 0.0359, "step": 11980 }, { "grad_norm": 0.25692328810691833, "learning_rate": 9.867049944325165e-05, "loss": 0.0332, "step": 11990 }, { "grad_norm": 0.1933797001838684, "learning_rate": 9.86667091813171e-05, "loss": 0.0363, "step": 12000 }, { "grad_norm": 0.241496741771698, "learning_rate": 9.866291359726438e-05, "loss": 0.0272, "step": 12010 }, { "grad_norm": 0.2364911437034607, "learning_rate": 9.865911269150861e-05, "loss": 0.0304, "step": 12020 }, { "grad_norm": 0.2841792404651642, "learning_rate": 9.865530646446544e-05, "loss": 0.0315, "step": 12030 }, { "grad_norm": 0.2783164083957672, "learning_rate": 9.86514949165511e-05, "loss": 0.0327, "step": 12040 }, { "grad_norm": 0.2877744138240814, "learning_rate": 9.864767804818243e-05, "loss": 0.0321, "step": 12050 }, { "grad_norm": 0.23386599123477936, "learning_rate": 9.86438558597768e-05, "loss": 0.03, "step": 12060 }, { "grad_norm": 0.23472800850868225, "learning_rate": 9.864002835175225e-05, "loss": 0.0344, "step": 12070 }, { "grad_norm": 0.21841870248317719, "learning_rate": 9.863619552452734e-05, "loss": 0.0313, "step": 12080 }, { "grad_norm": 0.26299747824668884, "learning_rate": 9.863235737852119e-05, "loss": 0.0331, "step": 12090 }, { "grad_norm": 0.24740305542945862, "learning_rate": 9.862851391415356e-05, "loss": 0.0362, "step": 12100 }, { "grad_norm": 0.33526501059532166, "learning_rate": 9.862466513184477e-05, "loss": 0.0314, "step": 12110 }, { "grad_norm": 0.21789544820785522, "learning_rate": 9.86208110320157e-05, "loss": 0.0321, "step": 12120 }, { "grad_norm": 0.2761966288089752, "learning_rate": 9.861695161508784e-05, "loss": 0.0341, "step": 12130 }, { "grad_norm": 0.3169209361076355, "learning_rate": 9.861308688148324e-05, "loss": 0.031, "step": 12140 }, { "grad_norm": 0.24182824790477753, "learning_rate": 9.860921683162455e-05, "loss": 0.0393, "step": 12150 }, { "grad_norm": 0.36634132266044617, "learning_rate": 9.860534146593499e-05, "loss": 0.034, "step": 12160 }, { "grad_norm": 0.25861144065856934, "learning_rate": 9.860146078483836e-05, "loss": 0.0285, "step": 12170 }, { "grad_norm": 0.20250490307807922, "learning_rate": 9.859757478875905e-05, "loss": 0.0311, "step": 12180 }, { "grad_norm": 0.29856884479522705, "learning_rate": 9.859368347812204e-05, "loss": 0.0303, "step": 12190 }, { "grad_norm": 0.31330806016921997, "learning_rate": 9.858978685335285e-05, "loss": 0.0379, "step": 12200 }, { "grad_norm": 0.2738935053348541, "learning_rate": 9.858588491487763e-05, "loss": 0.034, "step": 12210 }, { "grad_norm": 0.24579377472400665, "learning_rate": 9.858197766312308e-05, "loss": 0.0323, "step": 12220 }, { "grad_norm": 0.22780729830265045, "learning_rate": 9.857806509851649e-05, "loss": 0.0421, "step": 12230 }, { "grad_norm": 0.3313497304916382, "learning_rate": 9.857414722148574e-05, "loss": 0.0342, "step": 12240 }, { "grad_norm": 0.3543349802494049, "learning_rate": 9.857022403245928e-05, "loss": 0.0318, "step": 12250 }, { "grad_norm": 0.3098320960998535, "learning_rate": 9.856629553186615e-05, "loss": 0.0375, "step": 12260 }, { "grad_norm": 0.3002147972583771, "learning_rate": 9.856236172013595e-05, "loss": 0.0311, "step": 12270 }, { "grad_norm": 0.3102654814720154, "learning_rate": 9.85584225976989e-05, "loss": 0.0341, "step": 12280 }, { "grad_norm": 0.2599905729293823, "learning_rate": 9.855447816498575e-05, "loss": 0.0348, "step": 12290 }, { "grad_norm": 0.3131548762321472, "learning_rate": 9.855052842242787e-05, "loss": 0.0344, "step": 12300 }, { "grad_norm": 0.25383761525154114, "learning_rate": 9.85465733704572e-05, "loss": 0.0322, "step": 12310 }, { "grad_norm": 0.29719415307044983, "learning_rate": 9.854261300950624e-05, "loss": 0.0305, "step": 12320 }, { "grad_norm": 0.1838863641023636, "learning_rate": 9.853864734000813e-05, "loss": 0.0377, "step": 12330 }, { "grad_norm": 0.21797862648963928, "learning_rate": 9.85346763623965e-05, "loss": 0.0295, "step": 12340 }, { "grad_norm": 0.2682470381259918, "learning_rate": 9.853070007710564e-05, "loss": 0.0318, "step": 12350 }, { "grad_norm": 0.26940223574638367, "learning_rate": 9.85267184845704e-05, "loss": 0.0305, "step": 12360 }, { "grad_norm": 0.24222123622894287, "learning_rate": 9.852273158522616e-05, "loss": 0.0308, "step": 12370 }, { "grad_norm": 0.22448945045471191, "learning_rate": 9.851873937950896e-05, "loss": 0.0324, "step": 12380 }, { "grad_norm": 0.2040586620569229, "learning_rate": 9.851474186785537e-05, "loss": 0.0281, "step": 12390 }, { "grad_norm": 0.28002381324768066, "learning_rate": 9.851073905070254e-05, "loss": 0.0357, "step": 12400 }, { "grad_norm": 0.23538434505462646, "learning_rate": 9.850673092848824e-05, "loss": 0.0328, "step": 12410 }, { "grad_norm": 0.3594783842563629, "learning_rate": 9.850271750165077e-05, "loss": 0.0311, "step": 12420 }, { "grad_norm": 0.25435447692871094, "learning_rate": 9.849869877062902e-05, "loss": 0.0338, "step": 12430 }, { "grad_norm": 0.29509618878364563, "learning_rate": 9.849467473586252e-05, "loss": 0.0367, "step": 12440 }, { "grad_norm": 0.27786877751350403, "learning_rate": 9.849064539779127e-05, "loss": 0.0328, "step": 12450 }, { "grad_norm": 0.2732146680355072, "learning_rate": 9.848661075685594e-05, "loss": 0.0342, "step": 12460 }, { "grad_norm": 0.30103716254234314, "learning_rate": 9.848257081349778e-05, "loss": 0.0296, "step": 12470 }, { "grad_norm": 0.24376000463962555, "learning_rate": 9.847852556815856e-05, "loss": 0.0348, "step": 12480 }, { "grad_norm": 0.2542128264904022, "learning_rate": 9.847447502128067e-05, "loss": 0.0387, "step": 12490 }, { "grad_norm": 0.29013869166374207, "learning_rate": 9.847041917330708e-05, "loss": 0.0302, "step": 12500 }, { "grad_norm": 0.23488782346248627, "learning_rate": 9.846635802468132e-05, "loss": 0.0406, "step": 12510 }, { "grad_norm": 0.4013729393482208, "learning_rate": 9.84622915758475e-05, "loss": 0.0408, "step": 12520 }, { "grad_norm": 0.29871422052383423, "learning_rate": 9.845821982725034e-05, "loss": 0.0314, "step": 12530 }, { "grad_norm": 0.2738797664642334, "learning_rate": 9.845414277933514e-05, "loss": 0.0322, "step": 12540 }, { "grad_norm": 0.24473977088928223, "learning_rate": 9.845006043254771e-05, "loss": 0.0322, "step": 12550 }, { "grad_norm": 0.2185831367969513, "learning_rate": 9.844597278733451e-05, "loss": 0.0353, "step": 12560 }, { "grad_norm": 0.25380241870880127, "learning_rate": 9.844187984414259e-05, "loss": 0.0324, "step": 12570 }, { "grad_norm": 0.29273736476898193, "learning_rate": 9.84377816034195e-05, "loss": 0.0332, "step": 12580 }, { "grad_norm": 0.36607447266578674, "learning_rate": 9.843367806561345e-05, "loss": 0.0421, "step": 12590 }, { "grad_norm": 0.29144006967544556, "learning_rate": 9.842956923117317e-05, "loss": 0.0298, "step": 12600 }, { "grad_norm": 0.2613625228404999, "learning_rate": 9.842545510054802e-05, "loss": 0.0349, "step": 12610 }, { "grad_norm": 0.21849460899829865, "learning_rate": 9.842133567418792e-05, "loss": 0.0323, "step": 12620 }, { "grad_norm": 0.27134522795677185, "learning_rate": 9.841721095254333e-05, "loss": 0.0339, "step": 12630 }, { "grad_norm": 0.24544039368629456, "learning_rate": 9.841308093606537e-05, "loss": 0.0314, "step": 12640 }, { "grad_norm": 0.24163375794887543, "learning_rate": 9.840894562520565e-05, "loss": 0.0342, "step": 12650 }, { "grad_norm": 0.23987524211406708, "learning_rate": 9.840480502041642e-05, "loss": 0.0322, "step": 12660 }, { "grad_norm": 0.2983182668685913, "learning_rate": 9.840065912215049e-05, "loss": 0.0388, "step": 12670 }, { "grad_norm": 0.30116280913352966, "learning_rate": 9.839650793086124e-05, "loss": 0.029, "step": 12680 }, { "grad_norm": 0.2915896475315094, "learning_rate": 9.839235144700265e-05, "loss": 0.0373, "step": 12690 }, { "grad_norm": 0.32947883009910583, "learning_rate": 9.838818967102926e-05, "loss": 0.0376, "step": 12700 }, { "grad_norm": 0.257960706949234, "learning_rate": 9.83840226033962e-05, "loss": 0.0372, "step": 12710 }, { "grad_norm": 0.3424021601676941, "learning_rate": 9.837985024455918e-05, "loss": 0.0373, "step": 12720 }, { "grad_norm": 0.21963992714881897, "learning_rate": 9.837567259497447e-05, "loss": 0.0396, "step": 12730 }, { "grad_norm": 0.27109965682029724, "learning_rate": 9.837148965509894e-05, "loss": 0.0337, "step": 12740 }, { "grad_norm": 0.29191264510154724, "learning_rate": 9.836730142539001e-05, "loss": 0.0335, "step": 12750 }, { "grad_norm": 0.2731223702430725, "learning_rate": 9.836310790630574e-05, "loss": 0.0385, "step": 12760 }, { "grad_norm": 0.3446958363056183, "learning_rate": 9.83589090983047e-05, "loss": 0.0351, "step": 12770 }, { "grad_norm": 0.23452430963516235, "learning_rate": 9.835470500184605e-05, "loss": 0.0379, "step": 12780 }, { "grad_norm": 0.2925665080547333, "learning_rate": 9.835049561738957e-05, "loss": 0.028, "step": 12790 }, { "grad_norm": 0.28128787875175476, "learning_rate": 9.834628094539558e-05, "loss": 0.0379, "step": 12800 }, { "grad_norm": 0.23079681396484375, "learning_rate": 9.834206098632499e-05, "loss": 0.0365, "step": 12810 }, { "grad_norm": 0.26136690378189087, "learning_rate": 9.833783574063931e-05, "loss": 0.0384, "step": 12820 }, { "grad_norm": 0.33259546756744385, "learning_rate": 9.833360520880058e-05, "loss": 0.0405, "step": 12830 }, { "grad_norm": 0.28146597743034363, "learning_rate": 9.832936939127144e-05, "loss": 0.0334, "step": 12840 }, { "grad_norm": 0.29767659306526184, "learning_rate": 9.832512828851515e-05, "loss": 0.0315, "step": 12850 }, { "grad_norm": 0.2664667069911957, "learning_rate": 9.832088190099546e-05, "loss": 0.0368, "step": 12860 }, { "grad_norm": 0.3259899318218231, "learning_rate": 9.831663022917679e-05, "loss": 0.036, "step": 12870 }, { "grad_norm": 0.31001219153404236, "learning_rate": 9.831237327352407e-05, "loss": 0.0339, "step": 12880 }, { "grad_norm": 0.28746935725212097, "learning_rate": 9.830811103450286e-05, "loss": 0.0318, "step": 12890 }, { "grad_norm": 0.31499001383781433, "learning_rate": 9.830384351257924e-05, "loss": 0.0333, "step": 12900 }, { "grad_norm": 0.30042776465415955, "learning_rate": 9.829957070821993e-05, "loss": 0.0345, "step": 12910 }, { "grad_norm": 0.274446040391922, "learning_rate": 9.829529262189218e-05, "loss": 0.0306, "step": 12920 }, { "grad_norm": 0.3143242597579956, "learning_rate": 9.829100925406385e-05, "loss": 0.0315, "step": 12930 }, { "grad_norm": 0.27793583273887634, "learning_rate": 9.828672060520333e-05, "loss": 0.0333, "step": 12940 }, { "grad_norm": 0.32883015275001526, "learning_rate": 9.828242667577966e-05, "loss": 0.0354, "step": 12950 }, { "grad_norm": 0.3325413465499878, "learning_rate": 9.82781274662624e-05, "loss": 0.0292, "step": 12960 }, { "grad_norm": 0.22512109577655792, "learning_rate": 9.82738229771217e-05, "loss": 0.0298, "step": 12970 }, { "grad_norm": 0.22718560695648193, "learning_rate": 9.826951320882829e-05, "loss": 0.0338, "step": 12980 }, { "grad_norm": 0.24084265530109406, "learning_rate": 9.826519816185351e-05, "loss": 0.0302, "step": 12990 }, { "grad_norm": 0.21444281935691833, "learning_rate": 9.826087783666921e-05, "loss": 0.0333, "step": 13000 }, { "grad_norm": 0.2744850814342499, "learning_rate": 9.825655223374787e-05, "loss": 0.0294, "step": 13010 }, { "grad_norm": 0.2743355333805084, "learning_rate": 9.825222135356253e-05, "loss": 0.032, "step": 13020 }, { "grad_norm": 0.2928374111652374, "learning_rate": 9.82478851965868e-05, "loss": 0.037, "step": 13030 }, { "grad_norm": 0.31018996238708496, "learning_rate": 9.82435437632949e-05, "loss": 0.0301, "step": 13040 }, { "grad_norm": 0.328329861164093, "learning_rate": 9.823919705416158e-05, "loss": 0.0351, "step": 13050 }, { "grad_norm": 0.24361233413219452, "learning_rate": 9.82348450696622e-05, "loss": 0.0295, "step": 13060 }, { "grad_norm": 0.3842761218547821, "learning_rate": 9.823048781027268e-05, "loss": 0.0348, "step": 13070 }, { "grad_norm": 0.3283875286579132, "learning_rate": 9.822612527646953e-05, "loss": 0.0315, "step": 13080 }, { "grad_norm": 0.2717903256416321, "learning_rate": 9.822175746872984e-05, "loss": 0.027, "step": 13090 }, { "grad_norm": 0.2928403615951538, "learning_rate": 9.821738438753123e-05, "loss": 0.0429, "step": 13100 }, { "grad_norm": 0.21518556773662567, "learning_rate": 9.821300603335196e-05, "loss": 0.0306, "step": 13110 }, { "grad_norm": 0.24304984509944916, "learning_rate": 9.820862240667085e-05, "loss": 0.0336, "step": 13120 }, { "grad_norm": 0.24000240862369537, "learning_rate": 9.820423350796726e-05, "loss": 0.0338, "step": 13130 }, { "grad_norm": 0.1927068829536438, "learning_rate": 9.819983933772118e-05, "loss": 0.0383, "step": 13140 }, { "grad_norm": 0.22986316680908203, "learning_rate": 9.819543989641314e-05, "loss": 0.0311, "step": 13150 }, { "grad_norm": 0.25137531757354736, "learning_rate": 9.819103518452423e-05, "loss": 0.0294, "step": 13160 }, { "grad_norm": 0.21418829262256622, "learning_rate": 9.818662520253618e-05, "loss": 0.0326, "step": 13170 }, { "grad_norm": 0.3094789683818817, "learning_rate": 9.818220995093126e-05, "loss": 0.0327, "step": 13180 }, { "grad_norm": 0.24017980694770813, "learning_rate": 9.817778943019228e-05, "loss": 0.0352, "step": 13190 }, { "grad_norm": 0.25839120149612427, "learning_rate": 9.81733636408027e-05, "loss": 0.0351, "step": 13200 }, { "grad_norm": 0.2710009217262268, "learning_rate": 9.816893258324649e-05, "loss": 0.0295, "step": 13210 }, { "grad_norm": 0.17800387740135193, "learning_rate": 9.816449625800823e-05, "loss": 0.04, "step": 13220 }, { "grad_norm": 0.25142815709114075, "learning_rate": 9.816005466557308e-05, "loss": 0.0298, "step": 13230 }, { "grad_norm": 0.2548214793205261, "learning_rate": 9.815560780642674e-05, "loss": 0.0305, "step": 13240 }, { "grad_norm": 0.25230711698532104, "learning_rate": 9.815115568105555e-05, "loss": 0.0384, "step": 13250 }, { "grad_norm": 0.23833879828453064, "learning_rate": 9.814669828994638e-05, "loss": 0.0328, "step": 13260 }, { "grad_norm": 0.3010947108268738, "learning_rate": 9.814223563358665e-05, "loss": 0.0344, "step": 13270 }, { "grad_norm": 0.2621200382709503, "learning_rate": 9.813776771246443e-05, "loss": 0.033, "step": 13280 }, { "grad_norm": 0.24990801513195038, "learning_rate": 9.813329452706829e-05, "loss": 0.0286, "step": 13290 }, { "grad_norm": 0.23930980265140533, "learning_rate": 9.812881607788744e-05, "loss": 0.0325, "step": 13300 }, { "grad_norm": 0.2424933910369873, "learning_rate": 9.812433236541163e-05, "loss": 0.0343, "step": 13310 }, { "grad_norm": 0.2226588875055313, "learning_rate": 9.811984339013116e-05, "loss": 0.0305, "step": 13320 }, { "grad_norm": 0.2174217402935028, "learning_rate": 9.811534915253698e-05, "loss": 0.035, "step": 13330 }, { "grad_norm": 0.23460902273654938, "learning_rate": 9.811084965312056e-05, "loss": 0.0338, "step": 13340 }, { "grad_norm": 0.278913676738739, "learning_rate": 9.810634489237396e-05, "loss": 0.0346, "step": 13350 }, { "grad_norm": 0.28488603234291077, "learning_rate": 9.81018348707898e-05, "loss": 0.0324, "step": 13360 }, { "grad_norm": 0.25243353843688965, "learning_rate": 9.809731958886131e-05, "loss": 0.0298, "step": 13370 }, { "grad_norm": 0.2678108811378479, "learning_rate": 9.809279904708224e-05, "loss": 0.0344, "step": 13380 }, { "grad_norm": 0.28521955013275146, "learning_rate": 9.808827324594699e-05, "loss": 0.0344, "step": 13390 }, { "grad_norm": 0.27065083384513855, "learning_rate": 9.808374218595046e-05, "loss": 0.0324, "step": 13400 }, { "grad_norm": 0.2484390139579773, "learning_rate": 9.80792058675882e-05, "loss": 0.0367, "step": 13410 }, { "grad_norm": 0.2910666763782501, "learning_rate": 9.807466429135627e-05, "loss": 0.0352, "step": 13420 }, { "grad_norm": 0.28229624032974243, "learning_rate": 9.807011745775132e-05, "loss": 0.0343, "step": 13430 }, { "grad_norm": 0.22938241064548492, "learning_rate": 9.806556536727061e-05, "loss": 0.0308, "step": 13440 }, { "grad_norm": 0.2496262937784195, "learning_rate": 9.806100802041193e-05, "loss": 0.0314, "step": 13450 }, { "grad_norm": 0.2553529143333435, "learning_rate": 9.805644541767368e-05, "loss": 0.0304, "step": 13460 }, { "grad_norm": 0.19670526683330536, "learning_rate": 9.805187755955478e-05, "loss": 0.0307, "step": 13470 }, { "grad_norm": 0.2446461319923401, "learning_rate": 9.804730444655483e-05, "loss": 0.0344, "step": 13480 }, { "grad_norm": 0.23907285928726196, "learning_rate": 9.804272607917388e-05, "loss": 0.0279, "step": 13490 }, { "grad_norm": 0.2516063153743744, "learning_rate": 9.803814245791265e-05, "loss": 0.0341, "step": 13500 }, { "grad_norm": 0.28770318627357483, "learning_rate": 9.803355358327239e-05, "loss": 0.0305, "step": 13510 }, { "grad_norm": 0.26341649889945984, "learning_rate": 9.802895945575492e-05, "loss": 0.0314, "step": 13520 }, { "grad_norm": 0.2559368908405304, "learning_rate": 9.802436007586266e-05, "loss": 0.0323, "step": 13530 }, { "grad_norm": 0.26483285427093506, "learning_rate": 9.801975544409858e-05, "loss": 0.0282, "step": 13540 }, { "grad_norm": 0.29688572883605957, "learning_rate": 9.801514556096625e-05, "loss": 0.0287, "step": 13550 }, { "grad_norm": 0.2718730866909027, "learning_rate": 9.801053042696977e-05, "loss": 0.0349, "step": 13560 }, { "grad_norm": 0.27695390582084656, "learning_rate": 9.800591004261388e-05, "loss": 0.0285, "step": 13570 }, { "grad_norm": 0.28608235716819763, "learning_rate": 9.800128440840385e-05, "loss": 0.0331, "step": 13580 }, { "grad_norm": 0.2673387825489044, "learning_rate": 9.799665352484552e-05, "loss": 0.032, "step": 13590 }, { "grad_norm": 0.2612052261829376, "learning_rate": 9.799201739244532e-05, "loss": 0.0337, "step": 13600 }, { "grad_norm": 0.27118316292762756, "learning_rate": 9.798737601171025e-05, "loss": 0.035, "step": 13610 }, { "grad_norm": 0.2857316732406616, "learning_rate": 9.79827293831479e-05, "loss": 0.0276, "step": 13620 }, { "grad_norm": 0.27313360571861267, "learning_rate": 9.797807750726638e-05, "loss": 0.0369, "step": 13630 }, { "grad_norm": 0.28041020035743713, "learning_rate": 9.797342038457446e-05, "loss": 0.0312, "step": 13640 }, { "grad_norm": 0.1906789392232895, "learning_rate": 9.796875801558141e-05, "loss": 0.0314, "step": 13650 }, { "grad_norm": 0.23943226039409637, "learning_rate": 9.79640904007971e-05, "loss": 0.0321, "step": 13660 }, { "grad_norm": 0.23643378913402557, "learning_rate": 9.795941754073199e-05, "loss": 0.0284, "step": 13670 }, { "grad_norm": 0.22410394251346588, "learning_rate": 9.795473943589705e-05, "loss": 0.0342, "step": 13680 }, { "grad_norm": 0.19368860125541687, "learning_rate": 9.795005608680394e-05, "loss": 0.0334, "step": 13690 }, { "grad_norm": 0.28062865138053894, "learning_rate": 9.794536749396477e-05, "loss": 0.0366, "step": 13700 }, { "grad_norm": 0.2547203600406647, "learning_rate": 9.79406736578923e-05, "loss": 0.0355, "step": 13710 }, { "grad_norm": 0.24338315427303314, "learning_rate": 9.793597457909984e-05, "loss": 0.035, "step": 13720 }, { "grad_norm": 0.31960973143577576, "learning_rate": 9.793127025810127e-05, "loss": 0.0306, "step": 13730 }, { "grad_norm": 0.22054319083690643, "learning_rate": 9.792656069541104e-05, "loss": 0.0261, "step": 13740 }, { "grad_norm": 0.2876138985157013, "learning_rate": 9.79218458915442e-05, "loss": 0.0352, "step": 13750 }, { "grad_norm": 0.2630617916584015, "learning_rate": 9.791712584701634e-05, "loss": 0.0337, "step": 13760 }, { "grad_norm": 0.2682253122329712, "learning_rate": 9.791240056234364e-05, "loss": 0.0303, "step": 13770 }, { "grad_norm": 0.2162916511297226, "learning_rate": 9.790767003804283e-05, "loss": 0.0269, "step": 13780 }, { "grad_norm": 0.21535053849220276, "learning_rate": 9.790293427463126e-05, "loss": 0.0328, "step": 13790 }, { "grad_norm": 0.2806761860847473, "learning_rate": 9.789819327262684e-05, "loss": 0.0275, "step": 13800 }, { "grad_norm": 0.2498641163110733, "learning_rate": 9.7893447032548e-05, "loss": 0.0272, "step": 13810 }, { "grad_norm": 0.25317656993865967, "learning_rate": 9.78886955549138e-05, "loss": 0.0308, "step": 13820 }, { "grad_norm": 0.20677071809768677, "learning_rate": 9.788393884024387e-05, "loss": 0.0321, "step": 13830 }, { "grad_norm": 0.32149600982666016, "learning_rate": 9.787917688905836e-05, "loss": 0.0328, "step": 13840 }, { "grad_norm": 0.24392667412757874, "learning_rate": 9.787440970187807e-05, "loss": 0.0286, "step": 13850 }, { "grad_norm": 0.20881935954093933, "learning_rate": 9.786963727922429e-05, "loss": 0.0347, "step": 13860 }, { "grad_norm": 0.2315901815891266, "learning_rate": 9.786485962161897e-05, "loss": 0.0264, "step": 13870 }, { "grad_norm": 0.25668442249298096, "learning_rate": 9.786007672958455e-05, "loss": 0.0275, "step": 13880 }, { "grad_norm": 0.2199609875679016, "learning_rate": 9.78552886036441e-05, "loss": 0.0314, "step": 13890 }, { "grad_norm": 0.2555122673511505, "learning_rate": 9.785049524432124e-05, "loss": 0.0263, "step": 13900 }, { "grad_norm": 0.27922332286834717, "learning_rate": 9.784569665214016e-05, "loss": 0.0343, "step": 13910 }, { "grad_norm": 0.2876843810081482, "learning_rate": 9.784089282762563e-05, "loss": 0.0304, "step": 13920 }, { "grad_norm": 0.23776450753211975, "learning_rate": 9.7836083771303e-05, "loss": 0.0307, "step": 13930 }, { "grad_norm": 0.2280629724264145, "learning_rate": 9.783126948369817e-05, "loss": 0.0234, "step": 13940 }, { "grad_norm": 0.247774139046669, "learning_rate": 9.78264499653376e-05, "loss": 0.0324, "step": 13950 }, { "grad_norm": 0.25041189789772034, "learning_rate": 9.782162521674838e-05, "loss": 0.031, "step": 13960 }, { "grad_norm": 0.2692914605140686, "learning_rate": 9.781679523845812e-05, "loss": 0.0316, "step": 13970 }, { "grad_norm": 0.1886824369430542, "learning_rate": 9.781196003099502e-05, "loss": 0.0247, "step": 13980 }, { "grad_norm": 0.2941768169403076, "learning_rate": 9.780711959488786e-05, "loss": 0.0303, "step": 13990 }, { "grad_norm": 0.31564319133758545, "learning_rate": 9.780227393066599e-05, "loss": 0.0286, "step": 14000 }, { "grad_norm": 0.3118058741092682, "learning_rate": 9.77974230388593e-05, "loss": 0.0307, "step": 14010 }, { "grad_norm": 0.315897136926651, "learning_rate": 9.779256691999829e-05, "loss": 0.0384, "step": 14020 }, { "grad_norm": 0.17698971927165985, "learning_rate": 9.778770557461403e-05, "loss": 0.0301, "step": 14030 }, { "grad_norm": 0.21658651530742645, "learning_rate": 9.778283900323812e-05, "loss": 0.0341, "step": 14040 }, { "grad_norm": 0.25802579522132874, "learning_rate": 9.777796720640277e-05, "loss": 0.0294, "step": 14050 }, { "grad_norm": 0.2558121085166931, "learning_rate": 9.777309018464078e-05, "loss": 0.0265, "step": 14060 }, { "grad_norm": 0.28768759965896606, "learning_rate": 9.776820793848547e-05, "loss": 0.0271, "step": 14070 }, { "grad_norm": 0.24741974472999573, "learning_rate": 9.776332046847075e-05, "loss": 0.0288, "step": 14080 }, { "grad_norm": 0.36367952823638916, "learning_rate": 9.775842777513111e-05, "loss": 0.0342, "step": 14090 }, { "grad_norm": 0.24550040066242218, "learning_rate": 9.775352985900163e-05, "loss": 0.0313, "step": 14100 }, { "grad_norm": 0.3249584436416626, "learning_rate": 9.774862672061791e-05, "loss": 0.0321, "step": 14110 }, { "grad_norm": 0.27628982067108154, "learning_rate": 9.774371836051616e-05, "loss": 0.035, "step": 14120 }, { "grad_norm": 0.1976197212934494, "learning_rate": 9.773880477923315e-05, "loss": 0.0295, "step": 14130 }, { "grad_norm": 0.2353094518184662, "learning_rate": 9.773388597730623e-05, "loss": 0.0268, "step": 14140 }, { "grad_norm": 0.27937260270118713, "learning_rate": 9.77289619552733e-05, "loss": 0.0304, "step": 14150 }, { "grad_norm": 0.20700202882289886, "learning_rate": 9.772403271367285e-05, "loss": 0.031, "step": 14160 }, { "grad_norm": 0.2534429728984833, "learning_rate": 9.771909825304396e-05, "loss": 0.0334, "step": 14170 }, { "grad_norm": 0.17005540430545807, "learning_rate": 9.771415857392619e-05, "loss": 0.0305, "step": 14180 }, { "grad_norm": 0.23114578425884247, "learning_rate": 9.770921367685978e-05, "loss": 0.0274, "step": 14190 }, { "grad_norm": 0.21237371861934662, "learning_rate": 9.770426356238551e-05, "loss": 0.0299, "step": 14200 }, { "grad_norm": 0.27984195947647095, "learning_rate": 9.769930823104469e-05, "loss": 0.0337, "step": 14210 }, { "grad_norm": 0.25980696082115173, "learning_rate": 9.769434768337926e-05, "loss": 0.0304, "step": 14220 }, { "grad_norm": 0.23873242735862732, "learning_rate": 9.768938191993164e-05, "loss": 0.0326, "step": 14230 }, { "grad_norm": 0.3670525848865509, "learning_rate": 9.768441094124494e-05, "loss": 0.0316, "step": 14240 }, { "grad_norm": 0.25469255447387695, "learning_rate": 9.767943474786275e-05, "loss": 0.0284, "step": 14250 }, { "grad_norm": 0.24241161346435547, "learning_rate": 9.767445334032923e-05, "loss": 0.0285, "step": 14260 }, { "grad_norm": 0.28329983353614807, "learning_rate": 9.766946671918919e-05, "loss": 0.027, "step": 14270 }, { "grad_norm": 0.3337491750717163, "learning_rate": 9.766447488498796e-05, "loss": 0.0358, "step": 14280 }, { "grad_norm": 0.21537329256534576, "learning_rate": 9.765947783827139e-05, "loss": 0.0252, "step": 14290 }, { "grad_norm": 0.25652459263801575, "learning_rate": 9.765447557958599e-05, "loss": 0.0296, "step": 14300 }, { "grad_norm": 0.17116180062294006, "learning_rate": 9.764946810947879e-05, "loss": 0.0306, "step": 14310 }, { "grad_norm": 0.20260533690452576, "learning_rate": 9.764445542849738e-05, "loss": 0.0302, "step": 14320 }, { "grad_norm": 0.3469373881816864, "learning_rate": 9.763943753718998e-05, "loss": 0.03, "step": 14330 }, { "grad_norm": 0.23576290905475616, "learning_rate": 9.76344144361053e-05, "loss": 0.0261, "step": 14340 }, { "grad_norm": 0.2988276779651642, "learning_rate": 9.762938612579269e-05, "loss": 0.0338, "step": 14350 }, { "grad_norm": 0.24541686475276947, "learning_rate": 9.762435260680202e-05, "loss": 0.0312, "step": 14360 }, { "grad_norm": 0.31109926104545593, "learning_rate": 9.761931387968373e-05, "loss": 0.0338, "step": 14370 }, { "grad_norm": 0.23432713747024536, "learning_rate": 9.76142699449889e-05, "loss": 0.0274, "step": 14380 }, { "grad_norm": 0.24989427626132965, "learning_rate": 9.760922080326908e-05, "loss": 0.031, "step": 14390 }, { "grad_norm": 0.28324171900749207, "learning_rate": 9.760416645507644e-05, "loss": 0.0296, "step": 14400 }, { "grad_norm": 0.2358112335205078, "learning_rate": 9.759910690096375e-05, "loss": 0.0307, "step": 14410 }, { "grad_norm": 0.25238901376724243, "learning_rate": 9.759404214148429e-05, "loss": 0.0325, "step": 14420 }, { "grad_norm": 0.26203492283821106, "learning_rate": 9.758897217719191e-05, "loss": 0.0326, "step": 14430 }, { "grad_norm": 0.2116314321756363, "learning_rate": 9.758389700864113e-05, "loss": 0.0329, "step": 14440 }, { "grad_norm": 0.3613082468509674, "learning_rate": 9.757881663638688e-05, "loss": 0.0337, "step": 14450 }, { "grad_norm": 0.24982525408267975, "learning_rate": 9.757373106098478e-05, "loss": 0.0323, "step": 14460 }, { "grad_norm": 0.21670915186405182, "learning_rate": 9.756864028299097e-05, "loss": 0.0254, "step": 14470 }, { "grad_norm": 0.26957330107688904, "learning_rate": 9.75635443029622e-05, "loss": 0.0323, "step": 14480 }, { "grad_norm": 0.2605011761188507, "learning_rate": 9.755844312145572e-05, "loss": 0.0303, "step": 14490 }, { "grad_norm": 0.2640015482902527, "learning_rate": 9.755333673902941e-05, "loss": 0.0258, "step": 14500 }, { "grad_norm": 0.207247793674469, "learning_rate": 9.75482251562417e-05, "loss": 0.0299, "step": 14510 }, { "grad_norm": 0.26865750551223755, "learning_rate": 9.754310837365155e-05, "loss": 0.0295, "step": 14520 }, { "grad_norm": 0.26565632224082947, "learning_rate": 9.753798639181856e-05, "loss": 0.0279, "step": 14530 }, { "grad_norm": 0.3114386796951294, "learning_rate": 9.753285921130286e-05, "loss": 0.0306, "step": 14540 }, { "grad_norm": 0.24364826083183289, "learning_rate": 9.752772683266512e-05, "loss": 0.0285, "step": 14550 }, { "grad_norm": 0.28424447774887085, "learning_rate": 9.752258925646665e-05, "loss": 0.0258, "step": 14560 }, { "grad_norm": 0.26264306902885437, "learning_rate": 9.751744648326926e-05, "loss": 0.0336, "step": 14570 }, { "grad_norm": 0.26667729020118713, "learning_rate": 9.751229851363536e-05, "loss": 0.0254, "step": 14580 }, { "grad_norm": 0.23322738707065582, "learning_rate": 9.750714534812793e-05, "loss": 0.0318, "step": 14590 }, { "grad_norm": 0.23098525404930115, "learning_rate": 9.750198698731053e-05, "loss": 0.0288, "step": 14600 }, { "grad_norm": 0.22704237699508667, "learning_rate": 9.749682343174722e-05, "loss": 0.0261, "step": 14610 }, { "grad_norm": 0.20916180312633514, "learning_rate": 9.749165468200272e-05, "loss": 0.0282, "step": 14620 }, { "grad_norm": 0.21820728480815887, "learning_rate": 9.748648073864229e-05, "loss": 0.0288, "step": 14630 }, { "grad_norm": 0.3233199715614319, "learning_rate": 9.748130160223168e-05, "loss": 0.0289, "step": 14640 }, { "grad_norm": 0.3111107647418976, "learning_rate": 9.747611727333734e-05, "loss": 0.03, "step": 14650 }, { "grad_norm": 0.32343152165412903, "learning_rate": 9.74709277525262e-05, "loss": 0.0365, "step": 14660 }, { "grad_norm": 0.2545309364795685, "learning_rate": 9.746573304036576e-05, "loss": 0.0326, "step": 14670 }, { "grad_norm": 0.25146403908729553, "learning_rate": 9.746053313742412e-05, "loss": 0.0305, "step": 14680 }, { "grad_norm": 0.2805725932121277, "learning_rate": 9.745532804426994e-05, "loss": 0.0292, "step": 14690 }, { "grad_norm": 0.25828734040260315, "learning_rate": 9.745011776147242e-05, "loss": 0.0369, "step": 14700 }, { "grad_norm": 0.25504061579704285, "learning_rate": 9.744490228960138e-05, "loss": 0.0289, "step": 14710 }, { "grad_norm": 0.22937780618667603, "learning_rate": 9.743968162922713e-05, "loss": 0.0247, "step": 14720 }, { "grad_norm": 0.34714341163635254, "learning_rate": 9.743445578092064e-05, "loss": 0.0271, "step": 14730 }, { "grad_norm": 0.2595180571079254, "learning_rate": 9.742922474525338e-05, "loss": 0.0286, "step": 14740 }, { "grad_norm": 0.259436696767807, "learning_rate": 9.742398852279741e-05, "loss": 0.0263, "step": 14750 }, { "grad_norm": 0.2680369019508362, "learning_rate": 9.741874711412535e-05, "loss": 0.0328, "step": 14760 }, { "grad_norm": 0.23808473348617554, "learning_rate": 9.741350051981042e-05, "loss": 0.0302, "step": 14770 }, { "grad_norm": 0.2246130108833313, "learning_rate": 9.740824874042633e-05, "loss": 0.0331, "step": 14780 }, { "grad_norm": 0.22434067726135254, "learning_rate": 9.740299177654746e-05, "loss": 0.0276, "step": 14790 }, { "grad_norm": 0.27075910568237305, "learning_rate": 9.739772962874867e-05, "loss": 0.0327, "step": 14800 }, { "grad_norm": 0.24539914727210999, "learning_rate": 9.739246229760541e-05, "loss": 0.0377, "step": 14810 }, { "grad_norm": 0.27671781182289124, "learning_rate": 9.738718978369376e-05, "loss": 0.0295, "step": 14820 }, { "grad_norm": 0.24913254380226135, "learning_rate": 9.738191208759025e-05, "loss": 0.0339, "step": 14830 }, { "grad_norm": 0.19227132201194763, "learning_rate": 9.73766292098721e-05, "loss": 0.0314, "step": 14840 }, { "grad_norm": 0.2904541790485382, "learning_rate": 9.737134115111699e-05, "loss": 0.0369, "step": 14850 }, { "grad_norm": 0.33792319893836975, "learning_rate": 9.736604791190323e-05, "loss": 0.0421, "step": 14860 }, { "grad_norm": 0.31384924054145813, "learning_rate": 9.73607494928097e-05, "loss": 0.0333, "step": 14870 }, { "grad_norm": 0.2657429873943329, "learning_rate": 9.735544589441581e-05, "loss": 0.0335, "step": 14880 }, { "grad_norm": 0.3281470537185669, "learning_rate": 9.735013711730154e-05, "loss": 0.0323, "step": 14890 }, { "grad_norm": 0.2812936007976532, "learning_rate": 9.734482316204747e-05, "loss": 0.0337, "step": 14900 }, { "grad_norm": 0.2604309320449829, "learning_rate": 9.733950402923473e-05, "loss": 0.0262, "step": 14910 }, { "grad_norm": 0.2764973044395447, "learning_rate": 9.7334179719445e-05, "loss": 0.0351, "step": 14920 }, { "grad_norm": 0.3115675151348114, "learning_rate": 9.732885023326053e-05, "loss": 0.0363, "step": 14930 }, { "grad_norm": 0.3090868592262268, "learning_rate": 9.732351557126418e-05, "loss": 0.0365, "step": 14940 }, { "grad_norm": 0.25559574365615845, "learning_rate": 9.731817573403929e-05, "loss": 0.0309, "step": 14950 }, { "grad_norm": 0.21870598196983337, "learning_rate": 9.731283072216985e-05, "loss": 0.0309, "step": 14960 }, { "grad_norm": 0.25509798526763916, "learning_rate": 9.730748053624039e-05, "loss": 0.0307, "step": 14970 }, { "grad_norm": 0.2483786940574646, "learning_rate": 9.730212517683598e-05, "loss": 0.0331, "step": 14980 }, { "grad_norm": 0.2477472424507141, "learning_rate": 9.729676464454228e-05, "loss": 0.0296, "step": 14990 }, { "grad_norm": 0.18385915458202362, "learning_rate": 9.72913989399455e-05, "loss": 0.0304, "step": 15000 }, { "grad_norm": 0.23649674654006958, "learning_rate": 9.728602806363242e-05, "loss": 0.0291, "step": 15010 }, { "grad_norm": 0.272151380777359, "learning_rate": 9.728065201619043e-05, "loss": 0.0292, "step": 15020 }, { "grad_norm": 0.24274559319019318, "learning_rate": 9.727527079820742e-05, "loss": 0.0285, "step": 15030 }, { "grad_norm": 0.2424374669790268, "learning_rate": 9.726988441027186e-05, "loss": 0.0277, "step": 15040 }, { "grad_norm": 0.2272481620311737, "learning_rate": 9.726449285297281e-05, "loss": 0.0296, "step": 15050 }, { "grad_norm": 0.21790000796318054, "learning_rate": 9.72590961268999e-05, "loss": 0.0347, "step": 15060 }, { "grad_norm": 0.23486584424972534, "learning_rate": 9.725369423264328e-05, "loss": 0.0309, "step": 15070 }, { "grad_norm": 0.23672422766685486, "learning_rate": 9.72482871707937e-05, "loss": 0.0319, "step": 15080 }, { "grad_norm": 0.2368340641260147, "learning_rate": 9.724287494194247e-05, "loss": 0.0281, "step": 15090 }, { "grad_norm": 0.22656503319740295, "learning_rate": 9.723745754668147e-05, "loss": 0.0327, "step": 15100 }, { "grad_norm": 0.2122637927532196, "learning_rate": 9.723203498560313e-05, "loss": 0.028, "step": 15110 }, { "grad_norm": 0.25867119431495667, "learning_rate": 9.722660725930046e-05, "loss": 0.033, "step": 15120 }, { "grad_norm": 0.23222441971302032, "learning_rate": 9.722117436836702e-05, "loss": 0.0284, "step": 15130 }, { "grad_norm": 0.19655892252922058, "learning_rate": 9.721573631339696e-05, "loss": 0.0279, "step": 15140 }, { "grad_norm": 0.3395962119102478, "learning_rate": 9.721029309498494e-05, "loss": 0.0304, "step": 15150 }, { "grad_norm": 0.21267567574977875, "learning_rate": 9.720484471372627e-05, "loss": 0.0326, "step": 15160 }, { "grad_norm": 0.23885290324687958, "learning_rate": 9.719939117021673e-05, "loss": 0.0267, "step": 15170 }, { "grad_norm": 0.2696559429168701, "learning_rate": 9.719393246505275e-05, "loss": 0.0237, "step": 15180 }, { "grad_norm": 0.25789105892181396, "learning_rate": 9.718846859883128e-05, "loss": 0.0298, "step": 15190 }, { "grad_norm": 0.32783833146095276, "learning_rate": 9.718299957214982e-05, "loss": 0.0291, "step": 15200 }, { "grad_norm": 0.24706383049488068, "learning_rate": 9.717752538560646e-05, "loss": 0.0313, "step": 15210 }, { "grad_norm": 0.20416738092899323, "learning_rate": 9.717204603979986e-05, "loss": 0.0294, "step": 15220 }, { "grad_norm": 0.244707390666008, "learning_rate": 9.716656153532922e-05, "loss": 0.0317, "step": 15230 }, { "grad_norm": 0.22680476307868958, "learning_rate": 9.716107187279434e-05, "loss": 0.0301, "step": 15240 }, { "grad_norm": 0.20925407111644745, "learning_rate": 9.715557705279555e-05, "loss": 0.0263, "step": 15250 }, { "grad_norm": 0.2997472882270813, "learning_rate": 9.715007707593372e-05, "loss": 0.0292, "step": 15260 }, { "grad_norm": 0.30095982551574707, "learning_rate": 9.714457194281036e-05, "loss": 0.0323, "step": 15270 }, { "grad_norm": 0.317007839679718, "learning_rate": 9.713906165402751e-05, "loss": 0.0364, "step": 15280 }, { "grad_norm": 0.27419283986091614, "learning_rate": 9.713354621018774e-05, "loss": 0.0336, "step": 15290 }, { "grad_norm": 0.2510627806186676, "learning_rate": 9.712802561189422e-05, "loss": 0.0315, "step": 15300 }, { "grad_norm": 0.22140124440193176, "learning_rate": 9.712249985975069e-05, "loss": 0.027, "step": 15310 }, { "grad_norm": 0.16283151507377625, "learning_rate": 9.71169689543614e-05, "loss": 0.0248, "step": 15320 }, { "grad_norm": 0.22744446992874146, "learning_rate": 9.711143289633123e-05, "loss": 0.0291, "step": 15330 }, { "grad_norm": 0.19307489693164825, "learning_rate": 9.710589168626561e-05, "loss": 0.0253, "step": 15340 }, { "grad_norm": 0.3190961480140686, "learning_rate": 9.710034532477048e-05, "loss": 0.0276, "step": 15350 }, { "grad_norm": 0.2758296728134155, "learning_rate": 9.709479381245239e-05, "loss": 0.0324, "step": 15360 }, { "grad_norm": 0.2470536082983017, "learning_rate": 9.708923714991847e-05, "loss": 0.0325, "step": 15370 }, { "grad_norm": 0.23747606575489044, "learning_rate": 9.708367533777638e-05, "loss": 0.0279, "step": 15380 }, { "grad_norm": 0.2345106452703476, "learning_rate": 9.707810837663431e-05, "loss": 0.027, "step": 15390 }, { "grad_norm": 0.2938271760940552, "learning_rate": 9.707253626710113e-05, "loss": 0.0311, "step": 15400 }, { "grad_norm": 0.3007270395755768, "learning_rate": 9.706695900978613e-05, "loss": 0.0307, "step": 15410 }, { "grad_norm": 0.28373655676841736, "learning_rate": 9.706137660529926e-05, "loss": 0.0333, "step": 15420 }, { "grad_norm": 0.26392126083374023, "learning_rate": 9.705578905425101e-05, "loss": 0.0265, "step": 15430 }, { "grad_norm": 0.3101276457309723, "learning_rate": 9.705019635725241e-05, "loss": 0.0283, "step": 15440 }, { "grad_norm": 0.28251248598098755, "learning_rate": 9.704459851491508e-05, "loss": 0.0334, "step": 15450 }, { "grad_norm": 0.2588666081428528, "learning_rate": 9.703899552785118e-05, "loss": 0.0302, "step": 15460 }, { "grad_norm": 0.2744942307472229, "learning_rate": 9.703338739667346e-05, "loss": 0.0298, "step": 15470 }, { "grad_norm": 0.3343275785446167, "learning_rate": 9.70277741219952e-05, "loss": 0.0276, "step": 15480 }, { "grad_norm": 0.23749153316020966, "learning_rate": 9.702215570443027e-05, "loss": 0.0309, "step": 15490 }, { "grad_norm": 0.26645293831825256, "learning_rate": 9.701653214459309e-05, "loss": 0.0303, "step": 15500 }, { "grad_norm": 0.26261088252067566, "learning_rate": 9.701090344309865e-05, "loss": 0.0344, "step": 15510 }, { "grad_norm": 0.19068393111228943, "learning_rate": 9.700526960056247e-05, "loss": 0.0246, "step": 15520 }, { "grad_norm": 0.2675967514514923, "learning_rate": 9.699963061760068e-05, "loss": 0.026, "step": 15530 }, { "grad_norm": 0.3311883807182312, "learning_rate": 9.699398649482997e-05, "loss": 0.0282, "step": 15540 }, { "grad_norm": 0.2989688515663147, "learning_rate": 9.698833723286753e-05, "loss": 0.0268, "step": 15550 }, { "grad_norm": 0.31622469425201416, "learning_rate": 9.698268283233118e-05, "loss": 0.0308, "step": 15560 }, { "grad_norm": 0.2759217619895935, "learning_rate": 9.697702329383929e-05, "loss": 0.0263, "step": 15570 }, { "grad_norm": 0.2611217796802521, "learning_rate": 9.697135861801074e-05, "loss": 0.0259, "step": 15580 }, { "grad_norm": 0.2964254915714264, "learning_rate": 9.696568880546505e-05, "loss": 0.0303, "step": 15590 }, { "grad_norm": 0.2585001587867737, "learning_rate": 9.696001385682223e-05, "loss": 0.0317, "step": 15600 }, { "grad_norm": 0.3496406376361847, "learning_rate": 9.695433377270291e-05, "loss": 0.0371, "step": 15610 }, { "grad_norm": 0.2864040732383728, "learning_rate": 9.694864855372824e-05, "loss": 0.0242, "step": 15620 }, { "grad_norm": 0.16995736956596375, "learning_rate": 9.694295820051995e-05, "loss": 0.0257, "step": 15630 }, { "grad_norm": 0.26194068789482117, "learning_rate": 9.693726271370032e-05, "loss": 0.0274, "step": 15640 }, { "grad_norm": 0.2574789822101593, "learning_rate": 9.693156209389221e-05, "loss": 0.0282, "step": 15650 }, { "grad_norm": 0.19436517357826233, "learning_rate": 9.692585634171905e-05, "loss": 0.0262, "step": 15660 }, { "grad_norm": 0.3300294876098633, "learning_rate": 9.692014545780476e-05, "loss": 0.029, "step": 15670 }, { "grad_norm": 0.25011590123176575, "learning_rate": 9.691442944277393e-05, "loss": 0.0291, "step": 15680 }, { "grad_norm": 0.27996402978897095, "learning_rate": 9.690870829725162e-05, "loss": 0.0249, "step": 15690 }, { "grad_norm": 0.19064944982528687, "learning_rate": 9.69029820218635e-05, "loss": 0.0277, "step": 15700 }, { "grad_norm": 0.24751217663288116, "learning_rate": 9.689725061723579e-05, "loss": 0.0236, "step": 15710 }, { "grad_norm": 0.34346339106559753, "learning_rate": 9.689151408399527e-05, "loss": 0.0298, "step": 15720 }, { "grad_norm": 0.23127280175685883, "learning_rate": 9.688577242276924e-05, "loss": 0.027, "step": 15730 }, { "grad_norm": 0.28530362248420715, "learning_rate": 9.688002563418566e-05, "loss": 0.0246, "step": 15740 }, { "grad_norm": 0.2797820270061493, "learning_rate": 9.687427371887293e-05, "loss": 0.0305, "step": 15750 }, { "grad_norm": 0.29620373249053955, "learning_rate": 9.686851667746012e-05, "loss": 0.0327, "step": 15760 }, { "grad_norm": 0.2779466509819031, "learning_rate": 9.686275451057677e-05, "loss": 0.0314, "step": 15770 }, { "grad_norm": 0.23469862341880798, "learning_rate": 9.685698721885308e-05, "loss": 0.0256, "step": 15780 }, { "grad_norm": 0.26588675379753113, "learning_rate": 9.68512148029197e-05, "loss": 0.0316, "step": 15790 }, { "grad_norm": 0.20536908507347107, "learning_rate": 9.684543726340791e-05, "loss": 0.0281, "step": 15800 }, { "grad_norm": 0.2371443659067154, "learning_rate": 9.683965460094952e-05, "loss": 0.0299, "step": 15810 }, { "grad_norm": 0.27467021346092224, "learning_rate": 9.683386681617694e-05, "loss": 0.0363, "step": 15820 }, { "grad_norm": 0.1870955377817154, "learning_rate": 9.68280739097231e-05, "loss": 0.0266, "step": 15830 }, { "grad_norm": 0.22137147188186646, "learning_rate": 9.682227588222148e-05, "loss": 0.0275, "step": 15840 }, { "grad_norm": 0.22047056257724762, "learning_rate": 9.681647273430618e-05, "loss": 0.0273, "step": 15850 }, { "grad_norm": 0.31502124667167664, "learning_rate": 9.681066446661182e-05, "loss": 0.0335, "step": 15860 }, { "grad_norm": 0.29747939109802246, "learning_rate": 9.680485107977357e-05, "loss": 0.0295, "step": 15870 }, { "grad_norm": 0.25679728388786316, "learning_rate": 9.679903257442716e-05, "loss": 0.0306, "step": 15880 }, { "grad_norm": 0.20414267480373383, "learning_rate": 9.679320895120891e-05, "loss": 0.0246, "step": 15890 }, { "grad_norm": 0.23284950852394104, "learning_rate": 9.67873802107557e-05, "loss": 0.0258, "step": 15900 }, { "grad_norm": 0.21241459250450134, "learning_rate": 9.67815463537049e-05, "loss": 0.0263, "step": 15910 }, { "grad_norm": 0.28921401500701904, "learning_rate": 9.677570738069457e-05, "loss": 0.0294, "step": 15920 }, { "grad_norm": 0.23510584235191345, "learning_rate": 9.676986329236318e-05, "loss": 0.0294, "step": 15930 }, { "grad_norm": 0.28386759757995605, "learning_rate": 9.676401408934987e-05, "loss": 0.0243, "step": 15940 }, { "grad_norm": 0.22565732896327972, "learning_rate": 9.675815977229428e-05, "loss": 0.0339, "step": 15950 }, { "grad_norm": 0.29163485765457153, "learning_rate": 9.675230034183664e-05, "loss": 0.0243, "step": 15960 }, { "grad_norm": 0.2189633697271347, "learning_rate": 9.674643579861773e-05, "loss": 0.0265, "step": 15970 }, { "grad_norm": 0.26488274335861206, "learning_rate": 9.674056614327886e-05, "loss": 0.0278, "step": 15980 }, { "grad_norm": 0.25275614857673645, "learning_rate": 9.673469137646198e-05, "loss": 0.0286, "step": 15990 }, { "grad_norm": 0.2420332282781601, "learning_rate": 9.67288114988095e-05, "loss": 0.0244, "step": 16000 }, { "grad_norm": 0.2214280366897583, "learning_rate": 9.672292651096447e-05, "loss": 0.0312, "step": 16010 }, { "grad_norm": 0.28963056206703186, "learning_rate": 9.671703641357042e-05, "loss": 0.0285, "step": 16020 }, { "grad_norm": 0.2641667127609253, "learning_rate": 9.67111412072715e-05, "loss": 0.0311, "step": 16030 }, { "grad_norm": 0.25950029492378235, "learning_rate": 9.670524089271242e-05, "loss": 0.0281, "step": 16040 }, { "grad_norm": 0.2553952634334564, "learning_rate": 9.669933547053842e-05, "loss": 0.0266, "step": 16050 }, { "grad_norm": 0.2475549429655075, "learning_rate": 9.669342494139531e-05, "loss": 0.0301, "step": 16060 }, { "grad_norm": 0.1994388997554779, "learning_rate": 9.668750930592943e-05, "loss": 0.0263, "step": 16070 }, { "grad_norm": 0.28059691190719604, "learning_rate": 9.668158856478775e-05, "loss": 0.0333, "step": 16080 }, { "grad_norm": 0.2521807849407196, "learning_rate": 9.66756627186177e-05, "loss": 0.029, "step": 16090 }, { "grad_norm": 0.17466828227043152, "learning_rate": 9.666973176806737e-05, "loss": 0.0252, "step": 16100 }, { "grad_norm": 0.19786500930786133, "learning_rate": 9.666379571378534e-05, "loss": 0.0234, "step": 16110 }, { "grad_norm": 0.286986768245697, "learning_rate": 9.665785455642076e-05, "loss": 0.0241, "step": 16120 }, { "grad_norm": 0.22806894779205322, "learning_rate": 9.665190829662337e-05, "loss": 0.0334, "step": 16130 }, { "grad_norm": 0.29608550667762756, "learning_rate": 9.664595693504342e-05, "loss": 0.0277, "step": 16140 }, { "grad_norm": 0.24851979315280914, "learning_rate": 9.664000047233175e-05, "loss": 0.0299, "step": 16150 }, { "grad_norm": 0.23557668924331665, "learning_rate": 9.663403890913976e-05, "loss": 0.0332, "step": 16160 }, { "grad_norm": 0.2539171278476715, "learning_rate": 9.662807224611938e-05, "loss": 0.0304, "step": 16170 }, { "grad_norm": 0.2488182634115219, "learning_rate": 9.662210048392311e-05, "loss": 0.0277, "step": 16180 }, { "grad_norm": 0.23620246350765228, "learning_rate": 9.661612362320405e-05, "loss": 0.033, "step": 16190 }, { "grad_norm": 0.2421589195728302, "learning_rate": 9.661014166461579e-05, "loss": 0.0292, "step": 16200 }, { "grad_norm": 0.22989363968372345, "learning_rate": 9.66041546088125e-05, "loss": 0.0411, "step": 16210 }, { "grad_norm": 0.19711877405643463, "learning_rate": 9.659816245644895e-05, "loss": 0.026, "step": 16220 }, { "grad_norm": 0.2197689563035965, "learning_rate": 9.65921652081804e-05, "loss": 0.0292, "step": 16230 }, { "grad_norm": 0.23526936769485474, "learning_rate": 9.658616286466271e-05, "loss": 0.0286, "step": 16240 }, { "grad_norm": 0.18721705675125122, "learning_rate": 9.65801554265523e-05, "loss": 0.0298, "step": 16250 }, { "grad_norm": 0.24240605533123016, "learning_rate": 9.657414289450612e-05, "loss": 0.0263, "step": 16260 }, { "grad_norm": 0.3487049639225006, "learning_rate": 9.656812526918171e-05, "loss": 0.0302, "step": 16270 }, { "grad_norm": 0.2713465392589569, "learning_rate": 9.656210255123712e-05, "loss": 0.0316, "step": 16280 }, { "grad_norm": 0.3303852379322052, "learning_rate": 9.6556074741331e-05, "loss": 0.0285, "step": 16290 }, { "grad_norm": 0.2788407504558563, "learning_rate": 9.655004184012256e-05, "loss": 0.0305, "step": 16300 }, { "grad_norm": 0.19606848061084747, "learning_rate": 9.654400384827152e-05, "loss": 0.0324, "step": 16310 }, { "grad_norm": 0.18968890607357025, "learning_rate": 9.653796076643818e-05, "loss": 0.0279, "step": 16320 }, { "grad_norm": 0.28596925735473633, "learning_rate": 9.653191259528344e-05, "loss": 0.0273, "step": 16330 }, { "grad_norm": 0.23852747678756714, "learning_rate": 9.65258593354687e-05, "loss": 0.0275, "step": 16340 }, { "grad_norm": 0.21665416657924652, "learning_rate": 9.651980098765591e-05, "loss": 0.0301, "step": 16350 }, { "grad_norm": 0.2412339746952057, "learning_rate": 9.651373755250765e-05, "loss": 0.0305, "step": 16360 }, { "grad_norm": 0.20903773605823517, "learning_rate": 9.650766903068697e-05, "loss": 0.028, "step": 16370 }, { "grad_norm": 0.25617533922195435, "learning_rate": 9.650159542285753e-05, "loss": 0.0307, "step": 16380 }, { "grad_norm": 0.20842933654785156, "learning_rate": 9.649551672968353e-05, "loss": 0.0242, "step": 16390 }, { "grad_norm": 0.2089664191007614, "learning_rate": 9.648943295182973e-05, "loss": 0.0243, "step": 16400 }, { "grad_norm": 0.22024869918823242, "learning_rate": 9.648334408996144e-05, "loss": 0.0272, "step": 16410 }, { "grad_norm": 0.2072720229625702, "learning_rate": 9.647725014474452e-05, "loss": 0.0262, "step": 16420 }, { "grad_norm": 0.2715793550014496, "learning_rate": 9.64711511168454e-05, "loss": 0.0242, "step": 16430 }, { "grad_norm": 0.23625600337982178, "learning_rate": 9.646504700693108e-05, "loss": 0.0243, "step": 16440 }, { "grad_norm": 0.17983108758926392, "learning_rate": 9.645893781566907e-05, "loss": 0.0234, "step": 16450 }, { "grad_norm": 0.21024370193481445, "learning_rate": 9.645282354372744e-05, "loss": 0.0249, "step": 16460 }, { "grad_norm": 0.24111142754554749, "learning_rate": 9.644670419177491e-05, "loss": 0.0259, "step": 16470 }, { "grad_norm": 0.2411498874425888, "learning_rate": 9.644057976048062e-05, "loss": 0.0228, "step": 16480 }, { "grad_norm": 0.29629161953926086, "learning_rate": 9.643445025051435e-05, "loss": 0.0234, "step": 16490 }, { "grad_norm": 0.2863333225250244, "learning_rate": 9.642831566254641e-05, "loss": 0.0308, "step": 16500 }, { "grad_norm": 0.26695671677589417, "learning_rate": 9.642217599724769e-05, "loss": 0.0307, "step": 16510 }, { "grad_norm": 0.22848734259605408, "learning_rate": 9.64160312552896e-05, "loss": 0.0271, "step": 16520 }, { "grad_norm": 0.30451831221580505, "learning_rate": 9.64098814373441e-05, "loss": 0.0336, "step": 16530 }, { "grad_norm": 0.22790652513504028, "learning_rate": 9.640372654408374e-05, "loss": 0.0324, "step": 16540 }, { "grad_norm": 0.2559735178947449, "learning_rate": 9.639756657618162e-05, "loss": 0.0292, "step": 16550 }, { "grad_norm": 0.19148820638656616, "learning_rate": 9.639140153431138e-05, "loss": 0.0294, "step": 16560 }, { "grad_norm": 0.27936500310897827, "learning_rate": 9.638523141914721e-05, "loss": 0.0244, "step": 16570 }, { "grad_norm": 0.20792250335216522, "learning_rate": 9.637905623136388e-05, "loss": 0.0274, "step": 16580 }, { "grad_norm": 0.27102285623550415, "learning_rate": 9.637287597163669e-05, "loss": 0.0256, "step": 16590 }, { "grad_norm": 0.23051615059375763, "learning_rate": 9.63666906406415e-05, "loss": 0.0224, "step": 16600 }, { "grad_norm": 0.2704671621322632, "learning_rate": 9.636050023905473e-05, "loss": 0.0268, "step": 16610 }, { "grad_norm": 0.17714425921440125, "learning_rate": 9.635430476755336e-05, "loss": 0.03, "step": 16620 }, { "grad_norm": 0.3124952018260956, "learning_rate": 9.63481042268149e-05, "loss": 0.03, "step": 16630 }, { "grad_norm": 0.18953348696231842, "learning_rate": 9.634189861751745e-05, "loss": 0.0277, "step": 16640 }, { "grad_norm": 0.27382493019104004, "learning_rate": 9.633568794033967e-05, "loss": 0.0298, "step": 16650 }, { "grad_norm": 0.2180856317281723, "learning_rate": 9.63294721959607e-05, "loss": 0.025, "step": 16660 }, { "grad_norm": 0.2732153534889221, "learning_rate": 9.63232513850603e-05, "loss": 0.0276, "step": 16670 }, { "grad_norm": 0.2985450029373169, "learning_rate": 9.631702550831878e-05, "loss": 0.0254, "step": 16680 }, { "grad_norm": 0.2423521727323532, "learning_rate": 9.631079456641698e-05, "loss": 0.0265, "step": 16690 }, { "grad_norm": 0.2987765967845917, "learning_rate": 9.630455856003632e-05, "loss": 0.0271, "step": 16700 }, { "grad_norm": 0.2159416824579239, "learning_rate": 9.629831748985876e-05, "loss": 0.0267, "step": 16710 }, { "grad_norm": 0.3005499839782715, "learning_rate": 9.629207135656679e-05, "loss": 0.0225, "step": 16720 }, { "grad_norm": 0.24600602686405182, "learning_rate": 9.628582016084353e-05, "loss": 0.0251, "step": 16730 }, { "grad_norm": 0.16982261836528778, "learning_rate": 9.627956390337254e-05, "loss": 0.0269, "step": 16740 }, { "grad_norm": 0.19534516334533691, "learning_rate": 9.627330258483802e-05, "loss": 0.0214, "step": 16750 }, { "grad_norm": 0.33390969038009644, "learning_rate": 9.62670362059247e-05, "loss": 0.0338, "step": 16760 }, { "grad_norm": 0.26548513770103455, "learning_rate": 9.626076476731786e-05, "loss": 0.0278, "step": 16770 }, { "grad_norm": 0.3138889968395233, "learning_rate": 9.625448826970336e-05, "loss": 0.0286, "step": 16780 }, { "grad_norm": 0.22295236587524414, "learning_rate": 9.624820671376755e-05, "loss": 0.0257, "step": 16790 }, { "grad_norm": 0.19280529022216797, "learning_rate": 9.62419201001974e-05, "loss": 0.0261, "step": 16800 }, { "grad_norm": 0.27427050471305847, "learning_rate": 9.623562842968037e-05, "loss": 0.0261, "step": 16810 }, { "grad_norm": 0.2844151258468628, "learning_rate": 9.622933170290454e-05, "loss": 0.0276, "step": 16820 }, { "grad_norm": 0.27232685685157776, "learning_rate": 9.622302992055849e-05, "loss": 0.0266, "step": 16830 }, { "grad_norm": 0.2888261079788208, "learning_rate": 9.62167230833314e-05, "loss": 0.0314, "step": 16840 }, { "grad_norm": 0.29417094588279724, "learning_rate": 9.621041119191295e-05, "loss": 0.0279, "step": 16850 }, { "grad_norm": 0.3146347105503082, "learning_rate": 9.620409424699342e-05, "loss": 0.0313, "step": 16860 }, { "grad_norm": 0.1866532564163208, "learning_rate": 9.619777224926359e-05, "loss": 0.0269, "step": 16870 }, { "grad_norm": 0.2791719138622284, "learning_rate": 9.619144519941485e-05, "loss": 0.0338, "step": 16880 }, { "grad_norm": 0.20589232444763184, "learning_rate": 9.618511309813912e-05, "loss": 0.029, "step": 16890 }, { "grad_norm": 0.2655028700828552, "learning_rate": 9.617877594612886e-05, "loss": 0.029, "step": 16900 }, { "grad_norm": 0.25952714681625366, "learning_rate": 9.617243374407707e-05, "loss": 0.0288, "step": 16910 }, { "grad_norm": 0.21525250375270844, "learning_rate": 9.616608649267736e-05, "loss": 0.0336, "step": 16920 }, { "grad_norm": 0.19413866102695465, "learning_rate": 9.615973419262385e-05, "loss": 0.0293, "step": 16930 }, { "grad_norm": 0.17804571986198425, "learning_rate": 9.615337684461119e-05, "loss": 0.0281, "step": 16940 }, { "grad_norm": 0.2342531681060791, "learning_rate": 9.614701444933465e-05, "loss": 0.0236, "step": 16950 }, { "grad_norm": 0.3278311491012573, "learning_rate": 9.614064700748997e-05, "loss": 0.0296, "step": 16960 }, { "grad_norm": 0.33810991048812866, "learning_rate": 9.613427451977352e-05, "loss": 0.0281, "step": 16970 }, { "grad_norm": 0.2356628179550171, "learning_rate": 9.612789698688216e-05, "loss": 0.0256, "step": 16980 }, { "grad_norm": 0.20941585302352905, "learning_rate": 9.612151440951334e-05, "loss": 0.0278, "step": 16990 }, { "grad_norm": 0.2905886471271515, "learning_rate": 9.611512678836506e-05, "loss": 0.0262, "step": 17000 }, { "grad_norm": 0.20539091527462006, "learning_rate": 9.610873412413584e-05, "loss": 0.0243, "step": 17010 }, { "grad_norm": 0.2689678966999054, "learning_rate": 9.610233641752476e-05, "loss": 0.0323, "step": 17020 }, { "grad_norm": 0.2812725305557251, "learning_rate": 9.609593366923151e-05, "loss": 0.0259, "step": 17030 }, { "grad_norm": 0.28507766127586365, "learning_rate": 9.608952587995625e-05, "loss": 0.0337, "step": 17040 }, { "grad_norm": 0.3306984305381775, "learning_rate": 9.608311305039972e-05, "loss": 0.0309, "step": 17050 }, { "grad_norm": 0.2592061758041382, "learning_rate": 9.607669518126326e-05, "loss": 0.0351, "step": 17060 }, { "grad_norm": 0.31047287583351135, "learning_rate": 9.607027227324866e-05, "loss": 0.033, "step": 17070 }, { "grad_norm": 0.3612145781517029, "learning_rate": 9.606384432705837e-05, "loss": 0.0294, "step": 17080 }, { "grad_norm": 0.21687491238117218, "learning_rate": 9.60574113433953e-05, "loss": 0.0311, "step": 17090 }, { "grad_norm": 0.2952926754951477, "learning_rate": 9.6050973322963e-05, "loss": 0.0278, "step": 17100 }, { "grad_norm": 0.2538765072822571, "learning_rate": 9.604453026646547e-05, "loss": 0.0304, "step": 17110 }, { "grad_norm": 0.22493746876716614, "learning_rate": 9.603808217460735e-05, "loss": 0.0294, "step": 17120 }, { "grad_norm": 0.3082927465438843, "learning_rate": 9.603162904809377e-05, "loss": 0.0238, "step": 17130 }, { "grad_norm": 0.29541897773742676, "learning_rate": 9.602517088763045e-05, "loss": 0.0266, "step": 17140 }, { "grad_norm": 0.28364676237106323, "learning_rate": 9.601870769392365e-05, "loss": 0.0263, "step": 17150 }, { "grad_norm": 0.24177679419517517, "learning_rate": 9.601223946768017e-05, "loss": 0.0289, "step": 17160 }, { "grad_norm": 0.3415108919143677, "learning_rate": 9.600576620960734e-05, "loss": 0.0282, "step": 17170 }, { "grad_norm": 0.2535688877105713, "learning_rate": 9.599928792041308e-05, "loss": 0.0283, "step": 17180 }, { "grad_norm": 0.24630683660507202, "learning_rate": 9.599280460080587e-05, "loss": 0.0318, "step": 17190 }, { "grad_norm": 0.2559574246406555, "learning_rate": 9.59863162514947e-05, "loss": 0.0246, "step": 17200 }, { "grad_norm": 0.1985924392938614, "learning_rate": 9.597982287318911e-05, "loss": 0.0262, "step": 17210 }, { "grad_norm": 0.197652667760849, "learning_rate": 9.597332446659923e-05, "loss": 0.0272, "step": 17220 }, { "grad_norm": 0.24792151153087616, "learning_rate": 9.59668210324357e-05, "loss": 0.0313, "step": 17230 }, { "grad_norm": 0.19574329257011414, "learning_rate": 9.596031257140974e-05, "loss": 0.0264, "step": 17240 }, { "grad_norm": 0.21877871453762054, "learning_rate": 9.59537990842331e-05, "loss": 0.0237, "step": 17250 }, { "grad_norm": 0.21794071793556213, "learning_rate": 9.594728057161806e-05, "loss": 0.0253, "step": 17260 }, { "grad_norm": 0.22068051993846893, "learning_rate": 9.594075703427752e-05, "loss": 0.0253, "step": 17270 }, { "grad_norm": 0.292580246925354, "learning_rate": 9.593422847292486e-05, "loss": 0.0246, "step": 17280 }, { "grad_norm": 0.27595055103302, "learning_rate": 9.592769488827402e-05, "loss": 0.0269, "step": 17290 }, { "grad_norm": 0.26078441739082336, "learning_rate": 9.592115628103952e-05, "loss": 0.0277, "step": 17300 }, { "grad_norm": 0.2809147834777832, "learning_rate": 9.591461265193643e-05, "loss": 0.0297, "step": 17310 }, { "grad_norm": 0.2517518401145935, "learning_rate": 9.590806400168032e-05, "loss": 0.028, "step": 17320 }, { "grad_norm": 0.25367435812950134, "learning_rate": 9.590151033098735e-05, "loss": 0.0262, "step": 17330 }, { "grad_norm": 0.30810612440109253, "learning_rate": 9.589495164057423e-05, "loss": 0.0348, "step": 17340 }, { "grad_norm": 0.2486017644405365, "learning_rate": 9.58883879311582e-05, "loss": 0.0283, "step": 17350 }, { "grad_norm": 0.25343140959739685, "learning_rate": 9.588181920345705e-05, "loss": 0.0302, "step": 17360 }, { "grad_norm": 0.250172883272171, "learning_rate": 9.587524545818913e-05, "loss": 0.0311, "step": 17370 }, { "grad_norm": 0.3037523925304413, "learning_rate": 9.586866669607335e-05, "loss": 0.0309, "step": 17380 }, { "grad_norm": 0.2443382740020752, "learning_rate": 9.586208291782915e-05, "loss": 0.0236, "step": 17390 }, { "grad_norm": 0.20776863396167755, "learning_rate": 9.58554941241765e-05, "loss": 0.0313, "step": 17400 }, { "grad_norm": 0.22109976410865784, "learning_rate": 9.584890031583596e-05, "loss": 0.0295, "step": 17410 }, { "grad_norm": 0.33618301153182983, "learning_rate": 9.584230149352861e-05, "loss": 0.0216, "step": 17420 }, { "grad_norm": 0.31822991371154785, "learning_rate": 9.58356976579761e-05, "loss": 0.024, "step": 17430 }, { "grad_norm": 0.2860430181026459, "learning_rate": 9.58290888099006e-05, "loss": 0.0261, "step": 17440 }, { "grad_norm": 0.17514851689338684, "learning_rate": 9.582247495002486e-05, "loss": 0.029, "step": 17450 }, { "grad_norm": 0.2963012754917145, "learning_rate": 9.581585607907214e-05, "loss": 0.028, "step": 17460 }, { "grad_norm": 0.2715843617916107, "learning_rate": 9.580923219776628e-05, "loss": 0.0267, "step": 17470 }, { "grad_norm": 0.22341568768024445, "learning_rate": 9.580260330683167e-05, "loss": 0.0253, "step": 17480 }, { "grad_norm": 0.29974037408828735, "learning_rate": 9.579596940699322e-05, "loss": 0.027, "step": 17490 }, { "grad_norm": 0.23874230682849884, "learning_rate": 9.578933049897643e-05, "loss": 0.0322, "step": 17500 }, { "grad_norm": 0.28024211525917053, "learning_rate": 9.578268658350728e-05, "loss": 0.0254, "step": 17510 }, { "grad_norm": 0.2679814398288727, "learning_rate": 9.577603766131235e-05, "loss": 0.0255, "step": 17520 }, { "grad_norm": 0.2948351204395294, "learning_rate": 9.576938373311878e-05, "loss": 0.0264, "step": 17530 }, { "grad_norm": 0.19180934131145477, "learning_rate": 9.576272479965421e-05, "loss": 0.0228, "step": 17540 }, { "grad_norm": 0.21106252074241638, "learning_rate": 9.575606086164687e-05, "loss": 0.0229, "step": 17550 }, { "grad_norm": 0.2351865917444229, "learning_rate": 9.57493919198255e-05, "loss": 0.0277, "step": 17560 }, { "grad_norm": 0.23222555220127106, "learning_rate": 9.57427179749194e-05, "loss": 0.0256, "step": 17570 }, { "grad_norm": 0.22579820454120636, "learning_rate": 9.573603902765846e-05, "loss": 0.0245, "step": 17580 }, { "grad_norm": 0.22439447045326233, "learning_rate": 9.572935507877304e-05, "loss": 0.0285, "step": 17590 }, { "grad_norm": 0.21305695176124573, "learning_rate": 9.57226661289941e-05, "loss": 0.0331, "step": 17600 }, { "grad_norm": 0.2842496931552887, "learning_rate": 9.571597217905315e-05, "loss": 0.0266, "step": 17610 }, { "grad_norm": 0.2864820659160614, "learning_rate": 9.57092732296822e-05, "loss": 0.0286, "step": 17620 }, { "grad_norm": 0.27057400345802307, "learning_rate": 9.570256928161385e-05, "loss": 0.0321, "step": 17630 }, { "grad_norm": 0.17759084701538086, "learning_rate": 9.569586033558126e-05, "loss": 0.0283, "step": 17640 }, { "grad_norm": 0.2612975239753723, "learning_rate": 9.568914639231807e-05, "loss": 0.0356, "step": 17650 }, { "grad_norm": 0.2212100774049759, "learning_rate": 9.568242745255852e-05, "loss": 0.0274, "step": 17660 }, { "grad_norm": 0.24726347625255585, "learning_rate": 9.567570351703739e-05, "loss": 0.0264, "step": 17670 }, { "grad_norm": 0.2178552746772766, "learning_rate": 9.566897458649001e-05, "loss": 0.0331, "step": 17680 }, { "grad_norm": 0.25529876351356506, "learning_rate": 9.566224066165221e-05, "loss": 0.0287, "step": 17690 }, { "grad_norm": 0.29962679743766785, "learning_rate": 9.565550174326043e-05, "loss": 0.0373, "step": 17700 }, { "grad_norm": 0.26656606793403625, "learning_rate": 9.564875783205162e-05, "loss": 0.0276, "step": 17710 }, { "grad_norm": 0.25775623321533203, "learning_rate": 9.564200892876328e-05, "loss": 0.0325, "step": 17720 }, { "grad_norm": 0.3845066726207733, "learning_rate": 9.563525503413348e-05, "loss": 0.0352, "step": 17730 }, { "grad_norm": 0.2282642126083374, "learning_rate": 9.562849614890079e-05, "loss": 0.0243, "step": 17740 }, { "grad_norm": 0.23938991129398346, "learning_rate": 9.562173227380436e-05, "loss": 0.0251, "step": 17750 }, { "grad_norm": 0.1815536916255951, "learning_rate": 9.561496340958389e-05, "loss": 0.0266, "step": 17760 }, { "grad_norm": 0.29261788725852966, "learning_rate": 9.560818955697959e-05, "loss": 0.0292, "step": 17770 }, { "grad_norm": 0.22917509078979492, "learning_rate": 9.560141071673228e-05, "loss": 0.0285, "step": 17780 }, { "grad_norm": 0.2488020956516266, "learning_rate": 9.559462688958323e-05, "loss": 0.024, "step": 17790 }, { "grad_norm": 0.21357785165309906, "learning_rate": 9.558783807627434e-05, "loss": 0.0214, "step": 17800 }, { "grad_norm": 0.17698848247528076, "learning_rate": 9.558104427754801e-05, "loss": 0.0303, "step": 17810 }, { "grad_norm": 0.23889201879501343, "learning_rate": 9.557424549414722e-05, "loss": 0.0288, "step": 17820 }, { "grad_norm": 0.26848065853118896, "learning_rate": 9.556744172681546e-05, "loss": 0.0294, "step": 17830 }, { "grad_norm": 0.1937216967344284, "learning_rate": 9.556063297629677e-05, "loss": 0.0294, "step": 17840 }, { "grad_norm": 0.2787546217441559, "learning_rate": 9.555381924333578e-05, "loss": 0.0282, "step": 17850 }, { "grad_norm": 0.2677585482597351, "learning_rate": 9.554700052867758e-05, "loss": 0.0241, "step": 17860 }, { "grad_norm": 0.24727675318717957, "learning_rate": 9.554017683306789e-05, "loss": 0.0241, "step": 17870 }, { "grad_norm": 0.2537156939506531, "learning_rate": 9.553334815725294e-05, "loss": 0.0251, "step": 17880 }, { "grad_norm": 0.2831696569919586, "learning_rate": 9.552651450197949e-05, "loss": 0.0278, "step": 17890 }, { "grad_norm": 0.23590977489948273, "learning_rate": 9.551967586799486e-05, "loss": 0.0302, "step": 17900 }, { "grad_norm": 0.27984166145324707, "learning_rate": 9.551283225604692e-05, "loss": 0.0313, "step": 17910 }, { "grad_norm": 0.23159022629261017, "learning_rate": 9.550598366688406e-05, "loss": 0.0276, "step": 17920 }, { "grad_norm": 0.24590298533439636, "learning_rate": 9.549913010125526e-05, "loss": 0.0253, "step": 17930 }, { "grad_norm": 0.20702198147773743, "learning_rate": 9.549227155990999e-05, "loss": 0.0264, "step": 17940 }, { "grad_norm": 0.17136983573436737, "learning_rate": 9.548540804359828e-05, "loss": 0.0238, "step": 17950 }, { "grad_norm": 0.23988594114780426, "learning_rate": 9.547853955307077e-05, "loss": 0.0288, "step": 17960 }, { "grad_norm": 0.278628945350647, "learning_rate": 9.547166608907853e-05, "loss": 0.0227, "step": 17970 }, { "grad_norm": 0.2560403645038605, "learning_rate": 9.546478765237326e-05, "loss": 0.0284, "step": 17980 }, { "grad_norm": 0.22229114174842834, "learning_rate": 9.545790424370715e-05, "loss": 0.0324, "step": 17990 }, { "grad_norm": 0.19885815680027008, "learning_rate": 9.5451015863833e-05, "loss": 0.0281, "step": 18000 }, { "grad_norm": 0.18175677955150604, "learning_rate": 9.544412251350408e-05, "loss": 0.03, "step": 18010 }, { "grad_norm": 0.24063552916049957, "learning_rate": 9.543722419347422e-05, "loss": 0.0244, "step": 18020 }, { "grad_norm": 0.34514671564102173, "learning_rate": 9.543032090449788e-05, "loss": 0.0246, "step": 18030 }, { "grad_norm": 0.2329222410917282, "learning_rate": 9.542341264732992e-05, "loss": 0.0239, "step": 18040 }, { "grad_norm": 0.22679027915000916, "learning_rate": 9.541649942272585e-05, "loss": 0.0288, "step": 18050 }, { "grad_norm": 0.22646215558052063, "learning_rate": 9.54095812314417e-05, "loss": 0.03, "step": 18060 }, { "grad_norm": 0.2674712836742401, "learning_rate": 9.540265807423401e-05, "loss": 0.0261, "step": 18070 }, { "grad_norm": 0.21362079679965973, "learning_rate": 9.53957299518599e-05, "loss": 0.0283, "step": 18080 }, { "grad_norm": 0.24185912311077118, "learning_rate": 9.5388796865077e-05, "loss": 0.0206, "step": 18090 }, { "grad_norm": 0.23844754695892334, "learning_rate": 9.538185881464353e-05, "loss": 0.0291, "step": 18100 }, { "grad_norm": 0.33844754099845886, "learning_rate": 9.537491580131821e-05, "loss": 0.0289, "step": 18110 }, { "grad_norm": 0.25186941027641296, "learning_rate": 9.53679678258603e-05, "loss": 0.0229, "step": 18120 }, { "grad_norm": 0.20808227360248566, "learning_rate": 9.536101488902966e-05, "loss": 0.0281, "step": 18130 }, { "grad_norm": 0.15775886178016663, "learning_rate": 9.535405699158663e-05, "loss": 0.0257, "step": 18140 }, { "grad_norm": 0.19641219079494476, "learning_rate": 9.53470941342921e-05, "loss": 0.0272, "step": 18150 }, { "grad_norm": 0.2362169772386551, "learning_rate": 9.534012631790756e-05, "loss": 0.0265, "step": 18160 }, { "grad_norm": 0.25567662715911865, "learning_rate": 9.533315354319494e-05, "loss": 0.0293, "step": 18170 }, { "grad_norm": 0.28664278984069824, "learning_rate": 9.532617581091682e-05, "loss": 0.0276, "step": 18180 }, { "grad_norm": 0.22192052006721497, "learning_rate": 9.531919312183629e-05, "loss": 0.0284, "step": 18190 }, { "grad_norm": 0.25072038173675537, "learning_rate": 9.531220547671688e-05, "loss": 0.0237, "step": 18200 }, { "grad_norm": 0.20932862162590027, "learning_rate": 9.530521287632285e-05, "loss": 0.0327, "step": 18210 }, { "grad_norm": 0.24936236441135406, "learning_rate": 9.529821532141884e-05, "loss": 0.0235, "step": 18220 }, { "grad_norm": 0.21290118992328644, "learning_rate": 9.52912128127701e-05, "loss": 0.0264, "step": 18230 }, { "grad_norm": 0.22221042215824127, "learning_rate": 9.528420535114244e-05, "loss": 0.0271, "step": 18240 }, { "grad_norm": 0.24455100297927856, "learning_rate": 9.527719293730215e-05, "loss": 0.0259, "step": 18250 }, { "grad_norm": 0.2649688422679901, "learning_rate": 9.527017557201611e-05, "loss": 0.0271, "step": 18260 }, { "grad_norm": 0.24081097543239594, "learning_rate": 9.526315325605176e-05, "loss": 0.0253, "step": 18270 }, { "grad_norm": 0.2311069518327713, "learning_rate": 9.525612599017699e-05, "loss": 0.0292, "step": 18280 }, { "grad_norm": 0.2492189258337021, "learning_rate": 9.524909377516033e-05, "loss": 0.0216, "step": 18290 }, { "grad_norm": 0.20667359232902527, "learning_rate": 9.524205661177081e-05, "loss": 0.0262, "step": 18300 }, { "grad_norm": 0.2131081521511078, "learning_rate": 9.523501450077801e-05, "loss": 0.0259, "step": 18310 }, { "grad_norm": 0.16451312601566315, "learning_rate": 9.522796744295202e-05, "loss": 0.0254, "step": 18320 }, { "grad_norm": 0.19837437570095062, "learning_rate": 9.522091543906352e-05, "loss": 0.0301, "step": 18330 }, { "grad_norm": 0.23526325821876526, "learning_rate": 9.521385848988369e-05, "loss": 0.0264, "step": 18340 }, { "grad_norm": 0.25495997071266174, "learning_rate": 9.520679659618428e-05, "loss": 0.0291, "step": 18350 }, { "grad_norm": 0.2736593782901764, "learning_rate": 9.519972975873754e-05, "loss": 0.0297, "step": 18360 }, { "grad_norm": 0.23344263434410095, "learning_rate": 9.519265797831633e-05, "loss": 0.0271, "step": 18370 }, { "grad_norm": 0.2819019854068756, "learning_rate": 9.518558125569399e-05, "loss": 0.0277, "step": 18380 }, { "grad_norm": 0.2352612465620041, "learning_rate": 9.517849959164442e-05, "loss": 0.0211, "step": 18390 }, { "grad_norm": 0.24217970669269562, "learning_rate": 9.517141298694205e-05, "loss": 0.0289, "step": 18400 }, { "grad_norm": 0.25570645928382874, "learning_rate": 9.516432144236188e-05, "loss": 0.0316, "step": 18410 }, { "grad_norm": 0.27230092883110046, "learning_rate": 9.515722495867941e-05, "loss": 0.0239, "step": 18420 }, { "grad_norm": 0.23735788464546204, "learning_rate": 9.515012353667072e-05, "loss": 0.0312, "step": 18430 }, { "grad_norm": 0.22234860062599182, "learning_rate": 9.51430171771124e-05, "loss": 0.0261, "step": 18440 }, { "grad_norm": 0.2525511085987091, "learning_rate": 9.513590588078159e-05, "loss": 0.0258, "step": 18450 }, { "grad_norm": 0.26138216257095337, "learning_rate": 9.512878964845597e-05, "loss": 0.0253, "step": 18460 }, { "grad_norm": 0.23270748555660248, "learning_rate": 9.512166848091377e-05, "loss": 0.0264, "step": 18470 }, { "grad_norm": 0.2861996591091156, "learning_rate": 9.511454237893376e-05, "loss": 0.029, "step": 18480 }, { "grad_norm": 0.21211925148963928, "learning_rate": 9.51074113432952e-05, "loss": 0.0257, "step": 18490 }, { "grad_norm": 0.24516703188419342, "learning_rate": 9.510027537477797e-05, "loss": 0.0268, "step": 18500 }, { "grad_norm": 0.21650046110153198, "learning_rate": 9.509313447416242e-05, "loss": 0.0242, "step": 18510 }, { "grad_norm": 0.22798682749271393, "learning_rate": 9.508598864222949e-05, "loss": 0.0272, "step": 18520 }, { "grad_norm": 0.24528589844703674, "learning_rate": 9.507883787976062e-05, "loss": 0.025, "step": 18530 }, { "grad_norm": 0.22035813331604004, "learning_rate": 9.507168218753781e-05, "loss": 0.0246, "step": 18540 }, { "grad_norm": 0.25699281692504883, "learning_rate": 9.506452156634362e-05, "loss": 0.0274, "step": 18550 }, { "grad_norm": 0.20846997201442719, "learning_rate": 9.505735601696109e-05, "loss": 0.0281, "step": 18560 }, { "grad_norm": 0.16440261900424957, "learning_rate": 9.505018554017385e-05, "loss": 0.0261, "step": 18570 }, { "grad_norm": 0.26629775762557983, "learning_rate": 9.504301013676604e-05, "loss": 0.0265, "step": 18580 }, { "grad_norm": 0.2783237397670746, "learning_rate": 9.503582980752238e-05, "loss": 0.0261, "step": 18590 }, { "grad_norm": 0.2790020704269409, "learning_rate": 9.502864455322809e-05, "loss": 0.0257, "step": 18600 }, { "grad_norm": 0.18839159607887268, "learning_rate": 9.502145437466891e-05, "loss": 0.0226, "step": 18610 }, { "grad_norm": 0.2635839283466339, "learning_rate": 9.501425927263116e-05, "loss": 0.0249, "step": 18620 }, { "grad_norm": 0.24248182773590088, "learning_rate": 9.500705924790172e-05, "loss": 0.0253, "step": 18630 }, { "grad_norm": 0.19005261361598969, "learning_rate": 9.499985430126794e-05, "loss": 0.0265, "step": 18640 }, { "grad_norm": 0.2319495975971222, "learning_rate": 9.499264443351775e-05, "loss": 0.0255, "step": 18650 }, { "grad_norm": 0.23197245597839355, "learning_rate": 9.498542964543961e-05, "loss": 0.0223, "step": 18660 }, { "grad_norm": 0.18547400832176208, "learning_rate": 9.497820993782252e-05, "loss": 0.0224, "step": 18670 }, { "grad_norm": 0.2883235514163971, "learning_rate": 9.497098531145601e-05, "loss": 0.0301, "step": 18680 }, { "grad_norm": 0.2551088035106659, "learning_rate": 9.496375576713017e-05, "loss": 0.031, "step": 18690 }, { "grad_norm": 0.3160005211830139, "learning_rate": 9.49565213056356e-05, "loss": 0.0278, "step": 18700 }, { "grad_norm": 0.20244938135147095, "learning_rate": 9.494928192776342e-05, "loss": 0.0254, "step": 18710 }, { "grad_norm": 0.22933702170848846, "learning_rate": 9.494203763430538e-05, "loss": 0.0262, "step": 18720 }, { "grad_norm": 0.2266741544008255, "learning_rate": 9.493478842605366e-05, "loss": 0.0225, "step": 18730 }, { "grad_norm": 0.23379316926002502, "learning_rate": 9.492753430380105e-05, "loss": 0.0226, "step": 18740 }, { "grad_norm": 0.2624014914035797, "learning_rate": 9.492027526834083e-05, "loss": 0.0251, "step": 18750 }, { "grad_norm": 0.3104569911956787, "learning_rate": 9.491301132046684e-05, "loss": 0.0278, "step": 18760 }, { "grad_norm": 0.232793390750885, "learning_rate": 9.490574246097345e-05, "loss": 0.0259, "step": 18770 }, { "grad_norm": 0.2634981572628021, "learning_rate": 9.48984686906556e-05, "loss": 0.0227, "step": 18780 }, { "grad_norm": 0.21818485856056213, "learning_rate": 9.489119001030871e-05, "loss": 0.0228, "step": 18790 }, { "grad_norm": 0.2541278898715973, "learning_rate": 9.488390642072878e-05, "loss": 0.0276, "step": 18800 }, { "grad_norm": 0.16054020822048187, "learning_rate": 9.48766179227123e-05, "loss": 0.0173, "step": 18810 }, { "grad_norm": 0.1971156746149063, "learning_rate": 9.486932451705636e-05, "loss": 0.024, "step": 18820 }, { "grad_norm": 0.22736594080924988, "learning_rate": 9.486202620455857e-05, "loss": 0.025, "step": 18830 }, { "grad_norm": 0.2420167624950409, "learning_rate": 9.485472298601704e-05, "loss": 0.0262, "step": 18840 }, { "grad_norm": 0.32322606444358826, "learning_rate": 9.484741486223043e-05, "loss": 0.032, "step": 18850 }, { "grad_norm": 0.25095102190971375, "learning_rate": 9.484010183399797e-05, "loss": 0.0247, "step": 18860 }, { "grad_norm": 0.26396921277046204, "learning_rate": 9.483278390211938e-05, "loss": 0.0275, "step": 18870 }, { "grad_norm": 0.1765248030424118, "learning_rate": 9.482546106739496e-05, "loss": 0.0229, "step": 18880 }, { "grad_norm": 0.22369690239429474, "learning_rate": 9.48181333306255e-05, "loss": 0.0289, "step": 18890 }, { "grad_norm": 0.2722536325454712, "learning_rate": 9.481080069261237e-05, "loss": 0.0255, "step": 18900 }, { "grad_norm": 0.22309450805187225, "learning_rate": 9.480346315415745e-05, "loss": 0.0221, "step": 18910 }, { "grad_norm": 0.20231643319129944, "learning_rate": 9.479612071606314e-05, "loss": 0.0261, "step": 18920 }, { "grad_norm": 0.16192689538002014, "learning_rate": 9.478877337913244e-05, "loss": 0.0247, "step": 18930 }, { "grad_norm": 0.24366536736488342, "learning_rate": 9.478142114416881e-05, "loss": 0.0241, "step": 18940 }, { "grad_norm": 0.1954263299703598, "learning_rate": 9.47740640119763e-05, "loss": 0.0244, "step": 18950 }, { "grad_norm": 0.15682081878185272, "learning_rate": 9.476670198335947e-05, "loss": 0.024, "step": 18960 }, { "grad_norm": 0.23546159267425537, "learning_rate": 9.47593350591234e-05, "loss": 0.0292, "step": 18970 }, { "grad_norm": 0.192973330616951, "learning_rate": 9.475196324007376e-05, "loss": 0.0228, "step": 18980 }, { "grad_norm": 0.2507823407649994, "learning_rate": 9.474458652701669e-05, "loss": 0.0222, "step": 18990 }, { "grad_norm": 0.24854479730129242, "learning_rate": 9.473720492075892e-05, "loss": 0.0269, "step": 19000 }, { "grad_norm": 0.22617442905902863, "learning_rate": 9.472981842210768e-05, "loss": 0.0282, "step": 19010 }, { "grad_norm": 0.22253592312335968, "learning_rate": 9.472242703187074e-05, "loss": 0.0273, "step": 19020 }, { "grad_norm": 0.2801447808742523, "learning_rate": 9.471503075085643e-05, "loss": 0.0248, "step": 19030 }, { "grad_norm": 0.23555782437324524, "learning_rate": 9.470762957987359e-05, "loss": 0.0253, "step": 19040 }, { "grad_norm": 0.17080821096897125, "learning_rate": 9.470022351973158e-05, "loss": 0.0219, "step": 19050 }, { "grad_norm": 0.16232812404632568, "learning_rate": 9.469281257124034e-05, "loss": 0.0216, "step": 19060 }, { "grad_norm": 0.22676348686218262, "learning_rate": 9.46853967352103e-05, "loss": 0.0256, "step": 19070 }, { "grad_norm": 0.24058085680007935, "learning_rate": 9.467797601245246e-05, "loss": 0.0235, "step": 19080 }, { "grad_norm": 0.26635074615478516, "learning_rate": 9.467055040377834e-05, "loss": 0.0302, "step": 19090 }, { "grad_norm": 0.27065175771713257, "learning_rate": 9.466311990999999e-05, "loss": 0.0318, "step": 19100 }, { "grad_norm": 0.24476179480552673, "learning_rate": 9.465568453193e-05, "loss": 0.0286, "step": 19110 }, { "grad_norm": 0.24845901131629944, "learning_rate": 9.464824427038148e-05, "loss": 0.0288, "step": 19120 }, { "grad_norm": 0.18956877291202545, "learning_rate": 9.46407991261681e-05, "loss": 0.0252, "step": 19130 }, { "grad_norm": 0.20305339992046356, "learning_rate": 9.463334910010404e-05, "loss": 0.0268, "step": 19140 }, { "grad_norm": 0.30860912799835205, "learning_rate": 9.462589419300403e-05, "loss": 0.0319, "step": 19150 }, { "grad_norm": 0.2660066485404968, "learning_rate": 9.461843440568333e-05, "loss": 0.0271, "step": 19160 }, { "grad_norm": 0.18383648991584778, "learning_rate": 9.461096973895773e-05, "loss": 0.0267, "step": 19170 }, { "grad_norm": 0.27970898151397705, "learning_rate": 9.460350019364355e-05, "loss": 0.03, "step": 19180 }, { "grad_norm": 0.2445444017648697, "learning_rate": 9.459602577055764e-05, "loss": 0.0201, "step": 19190 }, { "grad_norm": 0.2165374606847763, "learning_rate": 9.45885464705174e-05, "loss": 0.0272, "step": 19200 }, { "grad_norm": 0.19110973179340363, "learning_rate": 9.458106229434076e-05, "loss": 0.0269, "step": 19210 }, { "grad_norm": 0.23075345158576965, "learning_rate": 9.457357324284617e-05, "loss": 0.0256, "step": 19220 }, { "grad_norm": 0.19564586877822876, "learning_rate": 9.456607931685262e-05, "loss": 0.022, "step": 19230 }, { "grad_norm": 0.19260062277317047, "learning_rate": 9.455858051717965e-05, "loss": 0.0205, "step": 19240 }, { "grad_norm": 0.2293093353509903, "learning_rate": 9.45510768446473e-05, "loss": 0.0212, "step": 19250 }, { "grad_norm": 0.2229546308517456, "learning_rate": 9.454356830007618e-05, "loss": 0.0229, "step": 19260 }, { "grad_norm": 0.23286961019039154, "learning_rate": 9.45360548842874e-05, "loss": 0.0292, "step": 19270 }, { "grad_norm": 0.2333688735961914, "learning_rate": 9.452853659810261e-05, "loss": 0.0237, "step": 19280 }, { "grad_norm": 0.19765803217887878, "learning_rate": 9.452101344234401e-05, "loss": 0.0267, "step": 19290 }, { "grad_norm": 0.22199749946594238, "learning_rate": 9.451348541783431e-05, "loss": 0.0274, "step": 19300 }, { "grad_norm": 0.2588273882865906, "learning_rate": 9.450595252539678e-05, "loss": 0.0304, "step": 19310 }, { "grad_norm": 0.2517019808292389, "learning_rate": 9.449841476585518e-05, "loss": 0.0244, "step": 19320 }, { "grad_norm": 0.2119835466146469, "learning_rate": 9.449087214003384e-05, "loss": 0.0226, "step": 19330 }, { "grad_norm": 0.19373676180839539, "learning_rate": 9.448332464875765e-05, "loss": 0.0247, "step": 19340 }, { "grad_norm": 0.24680911004543304, "learning_rate": 9.447577229285192e-05, "loss": 0.0234, "step": 19350 }, { "grad_norm": 0.22393818199634552, "learning_rate": 9.446821507314261e-05, "loss": 0.0211, "step": 19360 }, { "grad_norm": 0.12758684158325195, "learning_rate": 9.446065299045617e-05, "loss": 0.0228, "step": 19370 }, { "grad_norm": 0.241252601146698, "learning_rate": 9.445308604561955e-05, "loss": 0.027, "step": 19380 }, { "grad_norm": 0.22227992117404938, "learning_rate": 9.444551423946028e-05, "loss": 0.0224, "step": 19390 }, { "grad_norm": 0.264991819858551, "learning_rate": 9.443793757280638e-05, "loss": 0.03, "step": 19400 }, { "grad_norm": 0.22890813648700714, "learning_rate": 9.443035604648646e-05, "loss": 0.0245, "step": 19410 }, { "grad_norm": 0.24591590464115143, "learning_rate": 9.44227696613296e-05, "loss": 0.0292, "step": 19420 }, { "grad_norm": 0.2738024890422821, "learning_rate": 9.441517841816542e-05, "loss": 0.0272, "step": 19430 }, { "grad_norm": 0.23774631321430206, "learning_rate": 9.440758231782413e-05, "loss": 0.0239, "step": 19440 }, { "grad_norm": 0.1919962614774704, "learning_rate": 9.439998136113639e-05, "loss": 0.0243, "step": 19450 }, { "grad_norm": 0.1787751168012619, "learning_rate": 9.439237554893344e-05, "loss": 0.0229, "step": 19460 }, { "grad_norm": 0.19031554460525513, "learning_rate": 9.438476488204705e-05, "loss": 0.0209, "step": 19470 }, { "grad_norm": 0.21387015283107758, "learning_rate": 9.43771493613095e-05, "loss": 0.0254, "step": 19480 }, { "grad_norm": 0.26693183183670044, "learning_rate": 9.436952898755362e-05, "loss": 0.0225, "step": 19490 }, { "grad_norm": 0.24642986059188843, "learning_rate": 9.436190376161276e-05, "loss": 0.033, "step": 19500 }, { "grad_norm": 0.2632812559604645, "learning_rate": 9.43542736843208e-05, "loss": 0.0286, "step": 19510 }, { "grad_norm": 0.2596343457698822, "learning_rate": 9.434663875651216e-05, "loss": 0.0248, "step": 19520 }, { "grad_norm": 0.2097276747226715, "learning_rate": 9.433899897902177e-05, "loss": 0.0268, "step": 19530 }, { "grad_norm": 0.20348744094371796, "learning_rate": 9.433135435268511e-05, "loss": 0.0236, "step": 19540 }, { "grad_norm": 0.2444983273744583, "learning_rate": 9.432370487833819e-05, "loss": 0.0249, "step": 19550 }, { "grad_norm": 0.1795188933610916, "learning_rate": 9.431605055681756e-05, "loss": 0.0201, "step": 19560 }, { "grad_norm": 0.30295345187187195, "learning_rate": 9.430839138896026e-05, "loss": 0.0272, "step": 19570 }, { "grad_norm": 0.17178896069526672, "learning_rate": 9.43007273756039e-05, "loss": 0.0214, "step": 19580 }, { "grad_norm": 0.179519921541214, "learning_rate": 9.429305851758658e-05, "loss": 0.0233, "step": 19590 }, { "grad_norm": 0.1900208592414856, "learning_rate": 9.428538481574699e-05, "loss": 0.0243, "step": 19600 }, { "grad_norm": 0.14338508248329163, "learning_rate": 9.42777062709243e-05, "loss": 0.037, "step": 19610 }, { "grad_norm": 0.24311834573745728, "learning_rate": 9.427002288395821e-05, "loss": 0.0278, "step": 19620 }, { "grad_norm": 0.20739170908927917, "learning_rate": 9.426233465568898e-05, "loss": 0.0255, "step": 19630 }, { "grad_norm": 0.27275386452674866, "learning_rate": 9.42546415869574e-05, "loss": 0.0275, "step": 19640 }, { "grad_norm": 0.22165203094482422, "learning_rate": 9.424694367860473e-05, "loss": 0.0336, "step": 19650 }, { "grad_norm": 0.2574480473995209, "learning_rate": 9.423924093147284e-05, "loss": 0.0253, "step": 19660 }, { "grad_norm": 0.20907938480377197, "learning_rate": 9.423153334640407e-05, "loss": 0.0282, "step": 19670 }, { "grad_norm": 0.17890553176403046, "learning_rate": 9.42238209242413e-05, "loss": 0.028, "step": 19680 }, { "grad_norm": 0.2398984730243683, "learning_rate": 9.421610366582798e-05, "loss": 0.0301, "step": 19690 }, { "grad_norm": 0.1977805644273758, "learning_rate": 9.420838157200803e-05, "loss": 0.0275, "step": 19700 }, { "grad_norm": 0.23884764313697815, "learning_rate": 9.420065464362594e-05, "loss": 0.0263, "step": 19710 }, { "grad_norm": 0.20210780203342438, "learning_rate": 9.419292288152673e-05, "loss": 0.0266, "step": 19720 }, { "grad_norm": 0.17422430217266083, "learning_rate": 9.418518628655588e-05, "loss": 0.0222, "step": 19730 }, { "grad_norm": 0.18845798075199127, "learning_rate": 9.417744485955951e-05, "loss": 0.0211, "step": 19740 }, { "grad_norm": 0.19106893241405487, "learning_rate": 9.41696986013842e-05, "loss": 0.0225, "step": 19750 }, { "grad_norm": 0.24678775668144226, "learning_rate": 9.416194751287705e-05, "loss": 0.0224, "step": 19760 }, { "grad_norm": 0.2565496861934662, "learning_rate": 9.415419159488572e-05, "loss": 0.0245, "step": 19770 }, { "grad_norm": 0.19804997742176056, "learning_rate": 9.414643084825837e-05, "loss": 0.0225, "step": 19780 }, { "grad_norm": 0.23531927168369293, "learning_rate": 9.413866527384372e-05, "loss": 0.0229, "step": 19790 }, { "grad_norm": 0.24596700072288513, "learning_rate": 9.4130894872491e-05, "loss": 0.0233, "step": 19800 }, { "grad_norm": 0.23841485381126404, "learning_rate": 9.412311964504998e-05, "loss": 0.0257, "step": 19810 }, { "grad_norm": 0.1780909299850464, "learning_rate": 9.411533959237091e-05, "loss": 0.0238, "step": 19820 }, { "grad_norm": 0.2686213254928589, "learning_rate": 9.410755471530464e-05, "loss": 0.0249, "step": 19830 }, { "grad_norm": 0.23615296185016632, "learning_rate": 9.40997650147025e-05, "loss": 0.0234, "step": 19840 }, { "grad_norm": 0.3124179244041443, "learning_rate": 9.409197049141637e-05, "loss": 0.0244, "step": 19850 }, { "grad_norm": 0.27672716975212097, "learning_rate": 9.408417114629863e-05, "loss": 0.0232, "step": 19860 }, { "grad_norm": 0.23461824655532837, "learning_rate": 9.40763669802022e-05, "loss": 0.0181, "step": 19870 }, { "grad_norm": 0.24602575600147247, "learning_rate": 9.406855799398056e-05, "loss": 0.0227, "step": 19880 }, { "grad_norm": 0.2767868936061859, "learning_rate": 9.406074418848767e-05, "loss": 0.0311, "step": 19890 }, { "grad_norm": 0.2657608687877655, "learning_rate": 9.405292556457805e-05, "loss": 0.0256, "step": 19900 }, { "grad_norm": 0.24435141682624817, "learning_rate": 9.404510212310671e-05, "loss": 0.0223, "step": 19910 }, { "grad_norm": 0.23322080075740814, "learning_rate": 9.403727386492924e-05, "loss": 0.0271, "step": 19920 }, { "grad_norm": 0.17797090113162994, "learning_rate": 9.40294407909017e-05, "loss": 0.0257, "step": 19930 }, { "grad_norm": 0.27095937728881836, "learning_rate": 9.40216029018807e-05, "loss": 0.0211, "step": 19940 }, { "grad_norm": 0.25936248898506165, "learning_rate": 9.401376019872338e-05, "loss": 0.0232, "step": 19950 }, { "grad_norm": 0.16671521961688995, "learning_rate": 9.400591268228746e-05, "loss": 0.0213, "step": 19960 }, { "grad_norm": 0.1948556900024414, "learning_rate": 9.399806035343106e-05, "loss": 0.0219, "step": 19970 }, { "grad_norm": 0.20317348837852478, "learning_rate": 9.399020321301294e-05, "loss": 0.0235, "step": 19980 }, { "grad_norm": 0.19023798406124115, "learning_rate": 9.398234126189234e-05, "loss": 0.0236, "step": 19990 }, { "grad_norm": 0.2741576135158539, "learning_rate": 9.397447450092902e-05, "loss": 0.0244, "step": 20000 }, { "grad_norm": 0.20877182483673096, "learning_rate": 9.39666029309833e-05, "loss": 0.0227, "step": 20010 }, { "grad_norm": 0.22083449363708496, "learning_rate": 9.395872655291596e-05, "loss": 0.0223, "step": 20020 }, { "grad_norm": 0.11919509619474411, "learning_rate": 9.395084536758838e-05, "loss": 0.0223, "step": 20030 }, { "grad_norm": 0.31000441312789917, "learning_rate": 9.394295937586243e-05, "loss": 0.0291, "step": 20040 }, { "grad_norm": 0.2395665943622589, "learning_rate": 9.393506857860052e-05, "loss": 0.0243, "step": 20050 }, { "grad_norm": 0.26407933235168457, "learning_rate": 9.392717297666555e-05, "loss": 0.028, "step": 20060 }, { "grad_norm": 0.263593465089798, "learning_rate": 9.391927257092101e-05, "loss": 0.0257, "step": 20070 }, { "grad_norm": 0.20757772028446198, "learning_rate": 9.391136736223085e-05, "loss": 0.0232, "step": 20080 }, { "grad_norm": 0.19480642676353455, "learning_rate": 9.390345735145956e-05, "loss": 0.0254, "step": 20090 }, { "grad_norm": 0.23665253818035126, "learning_rate": 9.389554253947219e-05, "loss": 0.0225, "step": 20100 }, { "grad_norm": 0.22259274125099182, "learning_rate": 9.388762292713428e-05, "loss": 0.0225, "step": 20110 }, { "grad_norm": 0.22353024780750275, "learning_rate": 9.38796985153119e-05, "loss": 0.0238, "step": 20120 }, { "grad_norm": 0.2481890767812729, "learning_rate": 9.387176930487169e-05, "loss": 0.0246, "step": 20130 }, { "grad_norm": 0.2686050534248352, "learning_rate": 9.386383529668072e-05, "loss": 0.0235, "step": 20140 }, { "grad_norm": 0.23255224525928497, "learning_rate": 9.385589649160669e-05, "loss": 0.0317, "step": 20150 }, { "grad_norm": 0.24222607910633087, "learning_rate": 9.384795289051775e-05, "loss": 0.0256, "step": 20160 }, { "grad_norm": 0.18660148978233337, "learning_rate": 9.384000449428261e-05, "loss": 0.0248, "step": 20170 }, { "grad_norm": 0.19677169620990753, "learning_rate": 9.383205130377048e-05, "loss": 0.0218, "step": 20180 }, { "grad_norm": 0.1611468493938446, "learning_rate": 9.382409331985114e-05, "loss": 0.0214, "step": 20190 }, { "grad_norm": 0.19396215677261353, "learning_rate": 9.381613054339482e-05, "loss": 0.0225, "step": 20200 }, { "grad_norm": 0.21150901913642883, "learning_rate": 9.380816297527235e-05, "loss": 0.0225, "step": 20210 }, { "grad_norm": 0.16377080976963043, "learning_rate": 9.380019061635506e-05, "loss": 0.0199, "step": 20220 }, { "grad_norm": 0.16448791325092316, "learning_rate": 9.379221346751474e-05, "loss": 0.0256, "step": 20230 }, { "grad_norm": 0.24032190442085266, "learning_rate": 9.378423152962382e-05, "loss": 0.0243, "step": 20240 }, { "grad_norm": 0.2294304072856903, "learning_rate": 9.377624480355517e-05, "loss": 0.025, "step": 20250 }, { "grad_norm": 0.20350104570388794, "learning_rate": 9.376825329018219e-05, "loss": 0.0255, "step": 20260 }, { "grad_norm": 0.2988371253013611, "learning_rate": 9.376025699037884e-05, "loss": 0.0237, "step": 20270 }, { "grad_norm": 0.20386064052581787, "learning_rate": 9.37522559050196e-05, "loss": 0.0222, "step": 20280 }, { "grad_norm": 0.2819039225578308, "learning_rate": 9.37442500349794e-05, "loss": 0.0257, "step": 20290 }, { "grad_norm": 0.2643691599369049, "learning_rate": 9.373623938113381e-05, "loss": 0.023, "step": 20300 }, { "grad_norm": 0.273877888917923, "learning_rate": 9.372822394435883e-05, "loss": 0.0268, "step": 20310 }, { "grad_norm": 0.24054273962974548, "learning_rate": 9.372020372553102e-05, "loss": 0.0254, "step": 20320 }, { "grad_norm": 0.25444239377975464, "learning_rate": 9.371217872552746e-05, "loss": 0.0317, "step": 20330 }, { "grad_norm": 0.18719914555549622, "learning_rate": 9.370414894522576e-05, "loss": 0.0231, "step": 20340 }, { "grad_norm": 0.23468507826328278, "learning_rate": 9.369611438550406e-05, "loss": 0.0257, "step": 20350 }, { "grad_norm": 0.19580453634262085, "learning_rate": 9.368807504724095e-05, "loss": 0.0251, "step": 20360 }, { "grad_norm": 0.18886415660381317, "learning_rate": 9.368003093131565e-05, "loss": 0.0213, "step": 20370 }, { "grad_norm": 0.2241078019142151, "learning_rate": 9.367198203860785e-05, "loss": 0.0288, "step": 20380 }, { "grad_norm": 0.23530498147010803, "learning_rate": 9.366392836999774e-05, "loss": 0.0244, "step": 20390 }, { "grad_norm": 0.21389110386371613, "learning_rate": 9.365586992636607e-05, "loss": 0.0204, "step": 20400 }, { "grad_norm": 0.2718331813812256, "learning_rate": 9.364780670859412e-05, "loss": 0.0261, "step": 20410 }, { "grad_norm": 0.23899316787719727, "learning_rate": 9.363973871756364e-05, "loss": 0.0317, "step": 20420 }, { "grad_norm": 0.2638198137283325, "learning_rate": 9.363166595415696e-05, "loss": 0.0334, "step": 20430 }, { "grad_norm": 0.26685959100723267, "learning_rate": 9.362358841925686e-05, "loss": 0.0242, "step": 20440 }, { "grad_norm": 0.2163829654455185, "learning_rate": 9.361550611374674e-05, "loss": 0.0234, "step": 20450 }, { "grad_norm": 0.15553809702396393, "learning_rate": 9.360741903851043e-05, "loss": 0.0254, "step": 20460 }, { "grad_norm": 0.19913235306739807, "learning_rate": 9.359932719443236e-05, "loss": 0.0233, "step": 20470 }, { "grad_norm": 0.1829935610294342, "learning_rate": 9.35912305823974e-05, "loss": 0.0226, "step": 20480 }, { "grad_norm": 0.24278666079044342, "learning_rate": 9.358312920329101e-05, "loss": 0.0236, "step": 20490 }, { "grad_norm": 0.20260968804359436, "learning_rate": 9.357502305799914e-05, "loss": 0.0241, "step": 20500 }, { "grad_norm": 0.16772642731666565, "learning_rate": 9.356691214740824e-05, "loss": 0.021, "step": 20510 }, { "grad_norm": 0.2544424831867218, "learning_rate": 9.355879647240535e-05, "loss": 0.0339, "step": 20520 }, { "grad_norm": 0.22510622441768646, "learning_rate": 9.355067603387798e-05, "loss": 0.0258, "step": 20530 }, { "grad_norm": 0.20763106644153595, "learning_rate": 9.354255083271412e-05, "loss": 0.0258, "step": 20540 }, { "grad_norm": 0.3220232427120209, "learning_rate": 9.353442086980239e-05, "loss": 0.0275, "step": 20550 }, { "grad_norm": 0.2715292274951935, "learning_rate": 9.352628614603185e-05, "loss": 0.0265, "step": 20560 }, { "grad_norm": 0.2539445161819458, "learning_rate": 9.351814666229209e-05, "loss": 0.0232, "step": 20570 }, { "grad_norm": 0.20458781719207764, "learning_rate": 9.351000241947324e-05, "loss": 0.0241, "step": 20580 }, { "grad_norm": 0.20250512659549713, "learning_rate": 9.350185341846594e-05, "loss": 0.0242, "step": 20590 }, { "grad_norm": 0.204758420586586, "learning_rate": 9.349369966016134e-05, "loss": 0.0253, "step": 20600 }, { "grad_norm": 0.21545851230621338, "learning_rate": 9.348554114545117e-05, "loss": 0.0227, "step": 20610 }, { "grad_norm": 0.16833093762397766, "learning_rate": 9.347737787522758e-05, "loss": 0.0233, "step": 20620 }, { "grad_norm": 0.24895966053009033, "learning_rate": 9.346920985038332e-05, "loss": 0.0268, "step": 20630 }, { "grad_norm": 0.2609677314758301, "learning_rate": 9.346103707181162e-05, "loss": 0.0227, "step": 20640 }, { "grad_norm": 0.24862833321094513, "learning_rate": 9.345285954040626e-05, "loss": 0.0278, "step": 20650 }, { "grad_norm": 0.24075305461883545, "learning_rate": 9.34446772570615e-05, "loss": 0.0246, "step": 20660 }, { "grad_norm": 0.22158007323741913, "learning_rate": 9.343649022267214e-05, "loss": 0.0233, "step": 20670 }, { "grad_norm": 0.2275114357471466, "learning_rate": 9.342829843813353e-05, "loss": 0.0215, "step": 20680 }, { "grad_norm": 0.19022466242313385, "learning_rate": 9.342010190434149e-05, "loss": 0.0235, "step": 20690 }, { "grad_norm": 0.2080812305212021, "learning_rate": 9.34119006221924e-05, "loss": 0.0309, "step": 20700 }, { "grad_norm": 0.1917383372783661, "learning_rate": 9.340369459258313e-05, "loss": 0.0217, "step": 20710 }, { "grad_norm": 0.2353888303041458, "learning_rate": 9.339548381641106e-05, "loss": 0.0291, "step": 20720 }, { "grad_norm": 0.2038045972585678, "learning_rate": 9.338726829457413e-05, "loss": 0.0218, "step": 20730 }, { "grad_norm": 0.23404709994792938, "learning_rate": 9.337904802797078e-05, "loss": 0.0256, "step": 20740 }, { "grad_norm": 0.22182264924049377, "learning_rate": 9.337082301749993e-05, "loss": 0.0247, "step": 20750 }, { "grad_norm": 0.2473829686641693, "learning_rate": 9.336259326406109e-05, "loss": 0.0201, "step": 20760 }, { "grad_norm": 0.23375028371810913, "learning_rate": 9.335435876855427e-05, "loss": 0.021, "step": 20770 }, { "grad_norm": 0.2821841835975647, "learning_rate": 9.334611953187994e-05, "loss": 0.0264, "step": 20780 }, { "grad_norm": 0.21573428809642792, "learning_rate": 9.333787555493914e-05, "loss": 0.0244, "step": 20790 }, { "grad_norm": 0.17602524161338806, "learning_rate": 9.332962683863345e-05, "loss": 0.0212, "step": 20800 }, { "grad_norm": 0.24062491953372955, "learning_rate": 9.332137338386489e-05, "loss": 0.0229, "step": 20810 }, { "grad_norm": 0.24325379729270935, "learning_rate": 9.33131151915361e-05, "loss": 0.0277, "step": 20820 }, { "grad_norm": 0.20738258957862854, "learning_rate": 9.330485226255012e-05, "loss": 0.0206, "step": 20830 }, { "grad_norm": 0.2795798182487488, "learning_rate": 9.329658459781061e-05, "loss": 0.0194, "step": 20840 }, { "grad_norm": 0.1700015813112259, "learning_rate": 9.328831219822172e-05, "loss": 0.0226, "step": 20850 }, { "grad_norm": 0.24657417833805084, "learning_rate": 9.328003506468808e-05, "loss": 0.0253, "step": 20860 }, { "grad_norm": 0.23888808488845825, "learning_rate": 9.327175319811488e-05, "loss": 0.0221, "step": 20870 }, { "grad_norm": 0.22987724840641022, "learning_rate": 9.326346659940781e-05, "loss": 0.0235, "step": 20880 }, { "grad_norm": 0.18150202929973602, "learning_rate": 9.325517526947308e-05, "loss": 0.0229, "step": 20890 }, { "grad_norm": 0.2048075795173645, "learning_rate": 9.32468792092174e-05, "loss": 0.0226, "step": 20900 }, { "grad_norm": 0.29464128613471985, "learning_rate": 9.323857841954803e-05, "loss": 0.0232, "step": 20910 }, { "grad_norm": 0.22263434529304504, "learning_rate": 9.323027290137276e-05, "loss": 0.026, "step": 20920 }, { "grad_norm": 0.227430060505867, "learning_rate": 9.322196265559981e-05, "loss": 0.0231, "step": 20930 }, { "grad_norm": 0.3280350863933563, "learning_rate": 9.321364768313803e-05, "loss": 0.0291, "step": 20940 }, { "grad_norm": 0.21564947068691254, "learning_rate": 9.32053279848967e-05, "loss": 0.024, "step": 20950 }, { "grad_norm": 0.20297755300998688, "learning_rate": 9.319700356178567e-05, "loss": 0.027, "step": 20960 }, { "grad_norm": 0.21060636639595032, "learning_rate": 9.318867441471527e-05, "loss": 0.0207, "step": 20970 }, { "grad_norm": 0.21799932420253754, "learning_rate": 9.318034054459637e-05, "loss": 0.0224, "step": 20980 }, { "grad_norm": 0.17706532776355743, "learning_rate": 9.317200195234034e-05, "loss": 0.0235, "step": 20990 }, { "grad_norm": 0.2411198616027832, "learning_rate": 9.316365863885909e-05, "loss": 0.023, "step": 21000 }, { "grad_norm": 0.20816466212272644, "learning_rate": 9.315531060506502e-05, "loss": 0.0246, "step": 21010 }, { "grad_norm": 0.2006346583366394, "learning_rate": 9.314695785187108e-05, "loss": 0.0241, "step": 21020 }, { "grad_norm": 0.2217603176832199, "learning_rate": 9.313860038019069e-05, "loss": 0.0263, "step": 21030 }, { "grad_norm": 0.26892876625061035, "learning_rate": 9.313023819093782e-05, "loss": 0.026, "step": 21040 }, { "grad_norm": 0.20441177487373352, "learning_rate": 9.312187128502695e-05, "loss": 0.021, "step": 21050 }, { "grad_norm": 0.19344085454940796, "learning_rate": 9.311349966337307e-05, "loss": 0.0232, "step": 21060 }, { "grad_norm": 0.25464096665382385, "learning_rate": 9.310512332689169e-05, "loss": 0.0236, "step": 21070 }, { "grad_norm": 0.27307796478271484, "learning_rate": 9.309674227649883e-05, "loss": 0.0219, "step": 21080 }, { "grad_norm": 0.1940402388572693, "learning_rate": 9.308835651311103e-05, "loss": 0.0285, "step": 21090 }, { "grad_norm": 0.2536161243915558, "learning_rate": 9.307996603764533e-05, "loss": 0.027, "step": 21100 }, { "grad_norm": 0.2165067493915558, "learning_rate": 9.307157085101932e-05, "loss": 0.0235, "step": 21110 }, { "grad_norm": 0.2473997324705124, "learning_rate": 9.306317095415109e-05, "loss": 0.0224, "step": 21120 }, { "grad_norm": 0.2517998516559601, "learning_rate": 9.305476634795922e-05, "loss": 0.0266, "step": 21130 }, { "grad_norm": 0.23015081882476807, "learning_rate": 9.304635703336284e-05, "loss": 0.0217, "step": 21140 }, { "grad_norm": 0.20596002042293549, "learning_rate": 9.303794301128157e-05, "loss": 0.0223, "step": 21150 }, { "grad_norm": 0.22900664806365967, "learning_rate": 9.302952428263555e-05, "loss": 0.0245, "step": 21160 }, { "grad_norm": 0.2900446355342865, "learning_rate": 9.302110084834545e-05, "loss": 0.0271, "step": 21170 }, { "grad_norm": 0.1859627068042755, "learning_rate": 9.301267270933245e-05, "loss": 0.0258, "step": 21180 }, { "grad_norm": 0.23665158450603485, "learning_rate": 9.300423986651823e-05, "loss": 0.0295, "step": 21190 }, { "grad_norm": 0.2548101544380188, "learning_rate": 9.299580232082501e-05, "loss": 0.0251, "step": 21200 }, { "grad_norm": 0.26572299003601074, "learning_rate": 9.298736007317547e-05, "loss": 0.0299, "step": 21210 }, { "grad_norm": 0.1940212994813919, "learning_rate": 9.297891312449288e-05, "loss": 0.027, "step": 21220 }, { "grad_norm": 0.3192586898803711, "learning_rate": 9.297046147570094e-05, "loss": 0.03, "step": 21230 }, { "grad_norm": 0.24351762235164642, "learning_rate": 9.296200512772396e-05, "loss": 0.0312, "step": 21240 }, { "grad_norm": 0.2713722884654999, "learning_rate": 9.295354408148668e-05, "loss": 0.0229, "step": 21250 }, { "grad_norm": 0.2630252242088318, "learning_rate": 9.294507833791441e-05, "loss": 0.028, "step": 21260 }, { "grad_norm": 0.2835460603237152, "learning_rate": 9.293660789793295e-05, "loss": 0.0278, "step": 21270 }, { "grad_norm": 0.2506348788738251, "learning_rate": 9.292813276246858e-05, "loss": 0.0265, "step": 21280 }, { "grad_norm": 0.22674942016601562, "learning_rate": 9.291965293244816e-05, "loss": 0.0264, "step": 21290 }, { "grad_norm": 0.24228040874004364, "learning_rate": 9.291116840879904e-05, "loss": 0.0248, "step": 21300 }, { "grad_norm": 0.1864004284143448, "learning_rate": 9.290267919244904e-05, "loss": 0.0276, "step": 21310 }, { "grad_norm": 0.21696780622005463, "learning_rate": 9.289418528432655e-05, "loss": 0.0206, "step": 21320 }, { "grad_norm": 0.18801899254322052, "learning_rate": 9.288568668536045e-05, "loss": 0.021, "step": 21330 }, { "grad_norm": 0.16141298413276672, "learning_rate": 9.287718339648013e-05, "loss": 0.0194, "step": 21340 }, { "grad_norm": 0.18317022919654846, "learning_rate": 9.28686754186155e-05, "loss": 0.0199, "step": 21350 }, { "grad_norm": 0.23973654210567474, "learning_rate": 9.286016275269698e-05, "loss": 0.0185, "step": 21360 }, { "grad_norm": 0.17618796229362488, "learning_rate": 9.285164539965551e-05, "loss": 0.0213, "step": 21370 }, { "grad_norm": 0.2262418270111084, "learning_rate": 9.284312336042251e-05, "loss": 0.0227, "step": 21380 }, { "grad_norm": 0.19407542049884796, "learning_rate": 9.283459663592996e-05, "loss": 0.0228, "step": 21390 }, { "grad_norm": 0.20959709584712982, "learning_rate": 9.282606522711033e-05, "loss": 0.024, "step": 21400 }, { "grad_norm": 0.19911958277225494, "learning_rate": 9.281752913489657e-05, "loss": 0.021, "step": 21410 }, { "grad_norm": 0.21757088601589203, "learning_rate": 9.280898836022222e-05, "loss": 0.0209, "step": 21420 }, { "grad_norm": 0.18384768068790436, "learning_rate": 9.280044290402126e-05, "loss": 0.0229, "step": 21430 }, { "grad_norm": 0.20634622871875763, "learning_rate": 9.279189276722821e-05, "loss": 0.0268, "step": 21440 }, { "grad_norm": 0.2078309953212738, "learning_rate": 9.278333795077812e-05, "loss": 0.0264, "step": 21450 }, { "grad_norm": 0.23960034549236298, "learning_rate": 9.27747784556065e-05, "loss": 0.0274, "step": 21460 }, { "grad_norm": 0.21084193885326385, "learning_rate": 9.276621428264942e-05, "loss": 0.0207, "step": 21470 }, { "grad_norm": 0.25353139638900757, "learning_rate": 9.275764543284345e-05, "loss": 0.0251, "step": 21480 }, { "grad_norm": 0.2797064483165741, "learning_rate": 9.274907190712566e-05, "loss": 0.0233, "step": 21490 }, { "grad_norm": 0.22391995787620544, "learning_rate": 9.274049370643363e-05, "loss": 0.0218, "step": 21500 }, { "grad_norm": 0.23261050879955292, "learning_rate": 9.273191083170547e-05, "loss": 0.0217, "step": 21510 }, { "grad_norm": 0.24143086373806, "learning_rate": 9.27233232838798e-05, "loss": 0.0259, "step": 21520 }, { "grad_norm": 0.19296616315841675, "learning_rate": 9.27147310638957e-05, "loss": 0.0217, "step": 21530 }, { "grad_norm": 0.1867578774690628, "learning_rate": 9.270613417269286e-05, "loss": 0.0274, "step": 21540 }, { "grad_norm": 0.2118050754070282, "learning_rate": 9.269753261121138e-05, "loss": 0.021, "step": 21550 }, { "grad_norm": 0.26685798168182373, "learning_rate": 9.268892638039194e-05, "loss": 0.0213, "step": 21560 }, { "grad_norm": 0.2573513388633728, "learning_rate": 9.268031548117569e-05, "loss": 0.0276, "step": 21570 }, { "grad_norm": 0.24121862649917603, "learning_rate": 9.26716999145043e-05, "loss": 0.0242, "step": 21580 }, { "grad_norm": 0.1756524294614792, "learning_rate": 9.266307968131998e-05, "loss": 0.0187, "step": 21590 }, { "grad_norm": 0.20453861355781555, "learning_rate": 9.26544547825654e-05, "loss": 0.0257, "step": 21600 }, { "grad_norm": 0.20848669111728668, "learning_rate": 9.264582521918376e-05, "loss": 0.021, "step": 21610 }, { "grad_norm": 0.2071618288755417, "learning_rate": 9.263719099211881e-05, "loss": 0.0176, "step": 21620 }, { "grad_norm": 0.21712413430213928, "learning_rate": 9.262855210231476e-05, "loss": 0.0247, "step": 21630 }, { "grad_norm": 0.18773022294044495, "learning_rate": 9.261990855071633e-05, "loss": 0.0239, "step": 21640 }, { "grad_norm": 0.22844070196151733, "learning_rate": 9.261126033826878e-05, "loss": 0.0215, "step": 21650 }, { "grad_norm": 0.2190392166376114, "learning_rate": 9.260260746591786e-05, "loss": 0.0255, "step": 21660 }, { "grad_norm": 0.21168005466461182, "learning_rate": 9.259394993460985e-05, "loss": 0.0233, "step": 21670 }, { "grad_norm": 0.1656932532787323, "learning_rate": 9.258528774529151e-05, "loss": 0.0253, "step": 21680 }, { "grad_norm": 0.2002173364162445, "learning_rate": 9.257662089891013e-05, "loss": 0.0258, "step": 21690 }, { "grad_norm": 0.22270815074443817, "learning_rate": 9.25679493964135e-05, "loss": 0.0186, "step": 21700 }, { "grad_norm": 0.2301444560289383, "learning_rate": 9.255927323874994e-05, "loss": 0.027, "step": 21710 }, { "grad_norm": 0.2266266793012619, "learning_rate": 9.255059242686822e-05, "loss": 0.0253, "step": 21720 }, { "grad_norm": 0.16267725825309753, "learning_rate": 9.254190696171769e-05, "loss": 0.0235, "step": 21730 }, { "grad_norm": 0.24074025452136993, "learning_rate": 9.25332168442482e-05, "loss": 0.0275, "step": 21740 }, { "grad_norm": 0.23293437063694, "learning_rate": 9.252452207541004e-05, "loss": 0.0244, "step": 21750 }, { "grad_norm": 0.18818968534469604, "learning_rate": 9.251582265615409e-05, "loss": 0.0238, "step": 21760 }, { "grad_norm": 0.23535777628421783, "learning_rate": 9.250711858743169e-05, "loss": 0.0221, "step": 21770 }, { "grad_norm": 0.21267955005168915, "learning_rate": 9.24984098701947e-05, "loss": 0.0262, "step": 21780 }, { "grad_norm": 0.26789361238479614, "learning_rate": 9.248969650539552e-05, "loss": 0.018, "step": 21790 }, { "grad_norm": 0.2151847928762436, "learning_rate": 9.2480978493987e-05, "loss": 0.0234, "step": 21800 }, { "grad_norm": 0.21315160393714905, "learning_rate": 9.247225583692256e-05, "loss": 0.0259, "step": 21810 }, { "grad_norm": 0.19554615020751953, "learning_rate": 9.246352853515607e-05, "loss": 0.0238, "step": 21820 }, { "grad_norm": 0.23154202103614807, "learning_rate": 9.245479658964194e-05, "loss": 0.023, "step": 21830 }, { "grad_norm": 0.21893815696239471, "learning_rate": 9.244606000133507e-05, "loss": 0.0301, "step": 21840 }, { "grad_norm": 0.2552867829799652, "learning_rate": 9.24373187711909e-05, "loss": 0.0241, "step": 21850 }, { "grad_norm": 0.23811237514019012, "learning_rate": 9.242857290016537e-05, "loss": 0.0251, "step": 21860 }, { "grad_norm": 0.2099747359752655, "learning_rate": 9.241982238921488e-05, "loss": 0.0211, "step": 21870 }, { "grad_norm": 0.19061963260173798, "learning_rate": 9.24110672392964e-05, "loss": 0.0265, "step": 21880 }, { "grad_norm": 0.1504317820072174, "learning_rate": 9.240230745136737e-05, "loss": 0.0186, "step": 21890 }, { "grad_norm": 0.19696763157844543, "learning_rate": 9.239354302638575e-05, "loss": 0.0198, "step": 21900 }, { "grad_norm": 0.23312798142433167, "learning_rate": 9.238477396531e-05, "loss": 0.0189, "step": 21910 }, { "grad_norm": 0.2774199843406677, "learning_rate": 9.23760002690991e-05, "loss": 0.0226, "step": 21920 }, { "grad_norm": 0.21251444518566132, "learning_rate": 9.236722193871252e-05, "loss": 0.0232, "step": 21930 }, { "grad_norm": 0.21367579698562622, "learning_rate": 9.235843897511023e-05, "loss": 0.0206, "step": 21940 }, { "grad_norm": 0.20350097119808197, "learning_rate": 9.234965137925276e-05, "loss": 0.0253, "step": 21950 }, { "grad_norm": 0.23493313789367676, "learning_rate": 9.234085915210108e-05, "loss": 0.0247, "step": 21960 }, { "grad_norm": 0.2370375245809555, "learning_rate": 9.23320622946167e-05, "loss": 0.0265, "step": 21970 }, { "grad_norm": 0.28586235642433167, "learning_rate": 9.232326080776163e-05, "loss": 0.0257, "step": 21980 }, { "grad_norm": 0.2004849910736084, "learning_rate": 9.23144546924984e-05, "loss": 0.0246, "step": 21990 }, { "grad_norm": 0.22884200513362885, "learning_rate": 9.230564394979e-05, "loss": 0.0241, "step": 22000 }, { "grad_norm": 0.23117288947105408, "learning_rate": 9.22968285806e-05, "loss": 0.0223, "step": 22010 }, { "grad_norm": 0.2278052717447281, "learning_rate": 9.228800858589242e-05, "loss": 0.0225, "step": 22020 }, { "grad_norm": 0.2593560218811035, "learning_rate": 9.227918396663179e-05, "loss": 0.0212, "step": 22030 }, { "grad_norm": 0.20814068615436554, "learning_rate": 9.227035472378319e-05, "loss": 0.0233, "step": 22040 }, { "grad_norm": 0.18566186726093292, "learning_rate": 9.226152085831213e-05, "loss": 0.0226, "step": 22050 }, { "grad_norm": 0.1598251760005951, "learning_rate": 9.22526823711847e-05, "loss": 0.0229, "step": 22060 }, { "grad_norm": 0.23541651666164398, "learning_rate": 9.224383926336745e-05, "loss": 0.0239, "step": 22070 }, { "grad_norm": 0.2181917428970337, "learning_rate": 9.223499153582744e-05, "loss": 0.0211, "step": 22080 }, { "grad_norm": 0.2366449534893036, "learning_rate": 9.222613918953226e-05, "loss": 0.0231, "step": 22090 }, { "grad_norm": 0.21002228558063507, "learning_rate": 9.221728222544999e-05, "loss": 0.0241, "step": 22100 }, { "grad_norm": 0.2554378807544708, "learning_rate": 9.22084206445492e-05, "loss": 0.0271, "step": 22110 }, { "grad_norm": 0.20280075073242188, "learning_rate": 9.2199554447799e-05, "loss": 0.0241, "step": 22120 }, { "grad_norm": 0.27534908056259155, "learning_rate": 9.219068363616897e-05, "loss": 0.0257, "step": 22130 }, { "grad_norm": 0.24160924553871155, "learning_rate": 9.218180821062919e-05, "loss": 0.0259, "step": 22140 }, { "grad_norm": 0.21233628690242767, "learning_rate": 9.21729281721503e-05, "loss": 0.0252, "step": 22150 }, { "grad_norm": 0.21921870112419128, "learning_rate": 9.216404352170339e-05, "loss": 0.027, "step": 22160 }, { "grad_norm": 0.21099618077278137, "learning_rate": 9.215515426026007e-05, "loss": 0.0237, "step": 22170 }, { "grad_norm": 0.19487160444259644, "learning_rate": 9.214626038879246e-05, "loss": 0.0207, "step": 22180 }, { "grad_norm": 0.1643257886171341, "learning_rate": 9.21373619082732e-05, "loss": 0.0209, "step": 22190 }, { "grad_norm": 0.2398783564567566, "learning_rate": 9.212845881967535e-05, "loss": 0.0258, "step": 22200 }, { "grad_norm": 0.2798128128051758, "learning_rate": 9.211955112397262e-05, "loss": 0.0265, "step": 22210 }, { "grad_norm": 0.2314661294221878, "learning_rate": 9.211063882213909e-05, "loss": 0.022, "step": 22220 }, { "grad_norm": 0.22481392323970795, "learning_rate": 9.210172191514942e-05, "loss": 0.0264, "step": 22230 }, { "grad_norm": 0.23503507673740387, "learning_rate": 9.209280040397874e-05, "loss": 0.022, "step": 22240 }, { "grad_norm": 0.24293258786201477, "learning_rate": 9.208387428960268e-05, "loss": 0.0234, "step": 22250 }, { "grad_norm": 0.2440442144870758, "learning_rate": 9.20749435729974e-05, "loss": 0.0271, "step": 22260 }, { "grad_norm": 0.212848499417305, "learning_rate": 9.206600825513957e-05, "loss": 0.0267, "step": 22270 }, { "grad_norm": 0.16734808683395386, "learning_rate": 9.20570683370063e-05, "loss": 0.0197, "step": 22280 }, { "grad_norm": 0.2573005259037018, "learning_rate": 9.204812381957528e-05, "loss": 0.0258, "step": 22290 }, { "grad_norm": 0.19620902836322784, "learning_rate": 9.203917470382465e-05, "loss": 0.0231, "step": 22300 }, { "grad_norm": 0.2187337726354599, "learning_rate": 9.203022099073309e-05, "loss": 0.0203, "step": 22310 }, { "grad_norm": 0.19204598665237427, "learning_rate": 9.202126268127976e-05, "loss": 0.0198, "step": 22320 }, { "grad_norm": 0.14544573426246643, "learning_rate": 9.20122997764443e-05, "loss": 0.017, "step": 22330 }, { "grad_norm": 0.24212579429149628, "learning_rate": 9.200333227720692e-05, "loss": 0.0215, "step": 22340 }, { "grad_norm": 0.24946565926074982, "learning_rate": 9.199436018454826e-05, "loss": 0.0273, "step": 22350 }, { "grad_norm": 0.26591917872428894, "learning_rate": 9.198538349944952e-05, "loss": 0.0209, "step": 22360 }, { "grad_norm": 0.26911357045173645, "learning_rate": 9.197640222289234e-05, "loss": 0.0202, "step": 22370 }, { "grad_norm": 0.2235550433397293, "learning_rate": 9.196741635585895e-05, "loss": 0.0225, "step": 22380 }, { "grad_norm": 0.269149512052536, "learning_rate": 9.195842589933199e-05, "loss": 0.0229, "step": 22390 }, { "grad_norm": 0.22177450358867645, "learning_rate": 9.194943085429466e-05, "loss": 0.0269, "step": 22400 }, { "grad_norm": 0.26898062229156494, "learning_rate": 9.194043122173065e-05, "loss": 0.0303, "step": 22410 }, { "grad_norm": 0.24932944774627686, "learning_rate": 9.193142700262413e-05, "loss": 0.0236, "step": 22420 }, { "grad_norm": 0.2479560524225235, "learning_rate": 9.192241819795979e-05, "loss": 0.0248, "step": 22430 }, { "grad_norm": 0.20855799317359924, "learning_rate": 9.191340480872284e-05, "loss": 0.0215, "step": 22440 }, { "grad_norm": 0.21759463846683502, "learning_rate": 9.190438683589895e-05, "loss": 0.0232, "step": 22450 }, { "grad_norm": 0.22727027535438538, "learning_rate": 9.189536428047432e-05, "loss": 0.0205, "step": 22460 }, { "grad_norm": 0.21183931827545166, "learning_rate": 9.188633714343564e-05, "loss": 0.0212, "step": 22470 }, { "grad_norm": 0.21306699514389038, "learning_rate": 9.18773054257701e-05, "loss": 0.0219, "step": 22480 }, { "grad_norm": 0.21044707298278809, "learning_rate": 9.18682691284654e-05, "loss": 0.0203, "step": 22490 }, { "grad_norm": 0.15026915073394775, "learning_rate": 9.185922825250974e-05, "loss": 0.0248, "step": 22500 }, { "grad_norm": 0.2936236262321472, "learning_rate": 9.185018279889181e-05, "loss": 0.0263, "step": 22510 }, { "grad_norm": 0.2218051552772522, "learning_rate": 9.184113276860082e-05, "loss": 0.0262, "step": 22520 }, { "grad_norm": 0.19466403126716614, "learning_rate": 9.183207816262645e-05, "loss": 0.0291, "step": 22530 }, { "grad_norm": 0.26830801367759705, "learning_rate": 9.182301898195891e-05, "loss": 0.0287, "step": 22540 }, { "grad_norm": 0.2058323174715042, "learning_rate": 9.181395522758889e-05, "loss": 0.0218, "step": 22550 }, { "grad_norm": 0.23528720438480377, "learning_rate": 9.180488690050759e-05, "loss": 0.0215, "step": 22560 }, { "grad_norm": 0.24920713901519775, "learning_rate": 9.179581400170671e-05, "loss": 0.0205, "step": 22570 }, { "grad_norm": 0.18576179444789886, "learning_rate": 9.178673653217845e-05, "loss": 0.022, "step": 22580 }, { "grad_norm": 0.19547998905181885, "learning_rate": 9.177765449291551e-05, "loss": 0.0235, "step": 22590 }, { "grad_norm": 0.2546749711036682, "learning_rate": 9.176856788491109e-05, "loss": 0.021, "step": 22600 }, { "grad_norm": 0.24257659912109375, "learning_rate": 9.175947670915887e-05, "loss": 0.0259, "step": 22610 }, { "grad_norm": 0.20104388892650604, "learning_rate": 9.175038096665309e-05, "loss": 0.0287, "step": 22620 }, { "grad_norm": 0.16632425785064697, "learning_rate": 9.17412806583884e-05, "loss": 0.0212, "step": 22630 }, { "grad_norm": 0.24133867025375366, "learning_rate": 9.173217578536002e-05, "loss": 0.0198, "step": 22640 }, { "grad_norm": 0.2278628647327423, "learning_rate": 9.172306634856362e-05, "loss": 0.0254, "step": 22650 }, { "grad_norm": 0.2609823942184448, "learning_rate": 9.171395234899545e-05, "loss": 0.0231, "step": 22660 }, { "grad_norm": 0.2504703402519226, "learning_rate": 9.170483378765214e-05, "loss": 0.0259, "step": 22670 }, { "grad_norm": 0.25317081809043884, "learning_rate": 9.169571066553091e-05, "loss": 0.0232, "step": 22680 }, { "grad_norm": 0.24769432842731476, "learning_rate": 9.168658298362946e-05, "loss": 0.0238, "step": 22690 }, { "grad_norm": 0.2046700268983841, "learning_rate": 9.167745074294598e-05, "loss": 0.0251, "step": 22700 }, { "grad_norm": 0.25919628143310547, "learning_rate": 9.166831394447913e-05, "loss": 0.0206, "step": 22710 }, { "grad_norm": 0.29644864797592163, "learning_rate": 9.165917258922812e-05, "loss": 0.0224, "step": 22720 }, { "grad_norm": 0.18492035567760468, "learning_rate": 9.165002667819262e-05, "loss": 0.0206, "step": 22730 }, { "grad_norm": 0.22786752879619598, "learning_rate": 9.164087621237282e-05, "loss": 0.0195, "step": 22740 }, { "grad_norm": 0.23875534534454346, "learning_rate": 9.163172119276942e-05, "loss": 0.0207, "step": 22750 }, { "grad_norm": 0.21036431193351746, "learning_rate": 9.162256162038358e-05, "loss": 0.0262, "step": 22760 }, { "grad_norm": 0.19567669928073883, "learning_rate": 9.161339749621698e-05, "loss": 0.0273, "step": 22770 }, { "grad_norm": 0.1707673817873001, "learning_rate": 9.160422882127177e-05, "loss": 0.018, "step": 22780 }, { "grad_norm": 0.1731012761592865, "learning_rate": 9.159505559655069e-05, "loss": 0.0222, "step": 22790 }, { "grad_norm": 0.190249502658844, "learning_rate": 9.158587782305684e-05, "loss": 0.021, "step": 22800 }, { "grad_norm": 0.19711299240589142, "learning_rate": 9.157669550179391e-05, "loss": 0.0203, "step": 22810 }, { "grad_norm": 0.21316209435462952, "learning_rate": 9.156750863376609e-05, "loss": 0.0255, "step": 22820 }, { "grad_norm": 0.2002590298652649, "learning_rate": 9.155831721997801e-05, "loss": 0.0239, "step": 22830 }, { "grad_norm": 0.22198601067066193, "learning_rate": 9.154912126143484e-05, "loss": 0.0263, "step": 22840 }, { "grad_norm": 0.1655333936214447, "learning_rate": 9.153992075914224e-05, "loss": 0.0301, "step": 22850 }, { "grad_norm": 0.25850263237953186, "learning_rate": 9.153071571410635e-05, "loss": 0.0192, "step": 22860 }, { "grad_norm": 0.2459038943052292, "learning_rate": 9.152150612733384e-05, "loss": 0.0237, "step": 22870 }, { "grad_norm": 0.2736467123031616, "learning_rate": 9.151229199983184e-05, "loss": 0.0234, "step": 22880 }, { "grad_norm": 0.20249521732330322, "learning_rate": 9.150307333260802e-05, "loss": 0.021, "step": 22890 }, { "grad_norm": 0.29700952768325806, "learning_rate": 9.149385012667048e-05, "loss": 0.0237, "step": 22900 }, { "grad_norm": 0.2315143346786499, "learning_rate": 9.148462238302788e-05, "loss": 0.0261, "step": 22910 }, { "grad_norm": 0.23061618208885193, "learning_rate": 9.147539010268936e-05, "loss": 0.0218, "step": 22920 }, { "grad_norm": 0.16967181861400604, "learning_rate": 9.14661532866645e-05, "loss": 0.0221, "step": 22930 }, { "grad_norm": 0.23007267713546753, "learning_rate": 9.145691193596348e-05, "loss": 0.025, "step": 22940 }, { "grad_norm": 0.19310830533504486, "learning_rate": 9.144766605159691e-05, "loss": 0.0228, "step": 22950 }, { "grad_norm": 0.25743338465690613, "learning_rate": 9.14384156345759e-05, "loss": 0.0222, "step": 22960 }, { "grad_norm": 0.16800354421138763, "learning_rate": 9.142916068591204e-05, "loss": 0.0224, "step": 22970 }, { "grad_norm": 0.21899358928203583, "learning_rate": 9.141990120661746e-05, "loss": 0.0199, "step": 22980 }, { "grad_norm": 0.30500537157058716, "learning_rate": 9.141063719770475e-05, "loss": 0.0289, "step": 22990 }, { "grad_norm": 0.21055449545383453, "learning_rate": 9.140136866018704e-05, "loss": 0.021, "step": 23000 }, { "grad_norm": 0.1995103508234024, "learning_rate": 9.139209559507788e-05, "loss": 0.0223, "step": 23010 }, { "grad_norm": 0.19958321750164032, "learning_rate": 9.13828180033914e-05, "loss": 0.0226, "step": 23020 }, { "grad_norm": 0.26323211193084717, "learning_rate": 9.137353588614212e-05, "loss": 0.0231, "step": 23030 }, { "grad_norm": 0.3125973343849182, "learning_rate": 9.136424924434519e-05, "loss": 0.0279, "step": 23040 }, { "grad_norm": 0.24583305418491364, "learning_rate": 9.135495807901615e-05, "loss": 0.0238, "step": 23050 }, { "grad_norm": 0.2269124835729599, "learning_rate": 9.134566239117108e-05, "loss": 0.0256, "step": 23060 }, { "grad_norm": 0.1650833636522293, "learning_rate": 9.13363621818265e-05, "loss": 0.0227, "step": 23070 }, { "grad_norm": 0.2467736005783081, "learning_rate": 9.132705745199953e-05, "loss": 0.0294, "step": 23080 }, { "grad_norm": 0.2661341726779938, "learning_rate": 9.131774820270768e-05, "loss": 0.021, "step": 23090 }, { "grad_norm": 0.22302867472171783, "learning_rate": 9.130843443496901e-05, "loss": 0.0213, "step": 23100 }, { "grad_norm": 0.17695608735084534, "learning_rate": 9.129911614980206e-05, "loss": 0.0257, "step": 23110 }, { "grad_norm": 0.2579460144042969, "learning_rate": 9.128979334822584e-05, "loss": 0.0231, "step": 23120 }, { "grad_norm": 0.21343865990638733, "learning_rate": 9.128046603125992e-05, "loss": 0.0283, "step": 23130 }, { "grad_norm": 0.2898979187011719, "learning_rate": 9.12711341999243e-05, "loss": 0.0245, "step": 23140 }, { "grad_norm": 0.2045300006866455, "learning_rate": 9.12617978552395e-05, "loss": 0.0226, "step": 23150 }, { "grad_norm": 0.18316105008125305, "learning_rate": 9.12524569982265e-05, "loss": 0.0202, "step": 23160 }, { "grad_norm": 0.22353816032409668, "learning_rate": 9.124311162990684e-05, "loss": 0.0215, "step": 23170 }, { "grad_norm": 0.18153205513954163, "learning_rate": 9.12337617513025e-05, "loss": 0.02, "step": 23180 }, { "grad_norm": 0.20888502895832062, "learning_rate": 9.122440736343596e-05, "loss": 0.0174, "step": 23190 }, { "grad_norm": 0.2157958745956421, "learning_rate": 9.12150484673302e-05, "loss": 0.019, "step": 23200 }, { "grad_norm": 0.20897004008293152, "learning_rate": 9.120568506400873e-05, "loss": 0.0213, "step": 23210 }, { "grad_norm": 0.2518394887447357, "learning_rate": 9.119631715449548e-05, "loss": 0.0215, "step": 23220 }, { "grad_norm": 0.18403193354606628, "learning_rate": 9.118694473981493e-05, "loss": 0.0243, "step": 23230 }, { "grad_norm": 0.19428396224975586, "learning_rate": 9.117756782099203e-05, "loss": 0.0269, "step": 23240 }, { "grad_norm": 0.21437177062034607, "learning_rate": 9.11681863990522e-05, "loss": 0.0217, "step": 23250 }, { "grad_norm": 0.2294401377439499, "learning_rate": 9.115880047502142e-05, "loss": 0.0234, "step": 23260 }, { "grad_norm": 0.18858860433101654, "learning_rate": 9.114941004992609e-05, "loss": 0.0211, "step": 23270 }, { "grad_norm": 0.218352809548378, "learning_rate": 9.114001512479317e-05, "loss": 0.0246, "step": 23280 }, { "grad_norm": 0.2224266231060028, "learning_rate": 9.113061570065003e-05, "loss": 0.0213, "step": 23290 }, { "grad_norm": 0.26972562074661255, "learning_rate": 9.112121177852459e-05, "loss": 0.0222, "step": 23300 }, { "grad_norm": 0.22552157938480377, "learning_rate": 9.111180335944527e-05, "loss": 0.0249, "step": 23310 }, { "grad_norm": 0.18314291536808014, "learning_rate": 9.110239044444093e-05, "loss": 0.0189, "step": 23320 }, { "grad_norm": 0.2126973271369934, "learning_rate": 9.109297303454099e-05, "loss": 0.0201, "step": 23330 }, { "grad_norm": 0.21743500232696533, "learning_rate": 9.108355113077526e-05, "loss": 0.0212, "step": 23340 }, { "grad_norm": 0.21509917080402374, "learning_rate": 9.107412473417419e-05, "loss": 0.0208, "step": 23350 }, { "grad_norm": 0.2037673145532608, "learning_rate": 9.106469384576858e-05, "loss": 0.0204, "step": 23360 }, { "grad_norm": 0.19559866189956665, "learning_rate": 9.105525846658978e-05, "loss": 0.0205, "step": 23370 }, { "grad_norm": 0.22344538569450378, "learning_rate": 9.104581859766965e-05, "loss": 0.0285, "step": 23380 }, { "grad_norm": 0.18280869722366333, "learning_rate": 9.10363742400405e-05, "loss": 0.0245, "step": 23390 }, { "grad_norm": 0.22362540662288666, "learning_rate": 9.102692539473518e-05, "loss": 0.0232, "step": 23400 }, { "grad_norm": 0.20976322889328003, "learning_rate": 9.101747206278697e-05, "loss": 0.0245, "step": 23410 }, { "grad_norm": 0.18383251130580902, "learning_rate": 9.100801424522968e-05, "loss": 0.0208, "step": 23420 }, { "grad_norm": 0.16988365352153778, "learning_rate": 9.099855194309762e-05, "loss": 0.0239, "step": 23430 }, { "grad_norm": 0.25584888458251953, "learning_rate": 9.098908515742554e-05, "loss": 0.0236, "step": 23440 }, { "grad_norm": 0.22099412977695465, "learning_rate": 9.097961388924873e-05, "loss": 0.021, "step": 23450 }, { "grad_norm": 0.2274884134531021, "learning_rate": 9.097013813960298e-05, "loss": 0.0186, "step": 23460 }, { "grad_norm": 0.21623001992702484, "learning_rate": 9.09606579095245e-05, "loss": 0.0217, "step": 23470 }, { "grad_norm": 0.17716024816036224, "learning_rate": 9.095117320005008e-05, "loss": 0.0205, "step": 23480 }, { "grad_norm": 0.16419702768325806, "learning_rate": 9.094168401221691e-05, "loss": 0.0207, "step": 23490 }, { "grad_norm": 0.17754726111888885, "learning_rate": 9.093219034706273e-05, "loss": 0.0212, "step": 23500 }, { "grad_norm": 0.22699573636054993, "learning_rate": 9.092269220562577e-05, "loss": 0.0218, "step": 23510 }, { "grad_norm": 0.21633008122444153, "learning_rate": 9.09131895889447e-05, "loss": 0.0239, "step": 23520 }, { "grad_norm": 0.2023863047361374, "learning_rate": 9.090368249805873e-05, "loss": 0.0216, "step": 23530 }, { "grad_norm": 0.2416655272245407, "learning_rate": 9.089417093400754e-05, "loss": 0.0227, "step": 23540 }, { "grad_norm": 0.23319634795188904, "learning_rate": 9.088465489783131e-05, "loss": 0.022, "step": 23550 }, { "grad_norm": 0.20533324778079987, "learning_rate": 9.087513439057068e-05, "loss": 0.028, "step": 23560 }, { "grad_norm": 0.21173666417598724, "learning_rate": 9.08656094132668e-05, "loss": 0.0213, "step": 23570 }, { "grad_norm": 0.20119599997997284, "learning_rate": 9.085607996696134e-05, "loss": 0.0188, "step": 23580 }, { "grad_norm": 0.2590641677379608, "learning_rate": 9.084654605269639e-05, "loss": 0.0242, "step": 23590 }, { "grad_norm": 0.20103172957897186, "learning_rate": 9.083700767151457e-05, "loss": 0.0253, "step": 23600 }, { "grad_norm": 0.1861335188150406, "learning_rate": 9.082746482445898e-05, "loss": 0.023, "step": 23610 }, { "grad_norm": 0.25158122181892395, "learning_rate": 9.081791751257325e-05, "loss": 0.0201, "step": 23620 }, { "grad_norm": 0.2733912169933319, "learning_rate": 9.080836573690142e-05, "loss": 0.0254, "step": 23630 }, { "grad_norm": 0.20570406317710876, "learning_rate": 9.079880949848805e-05, "loss": 0.022, "step": 23640 }, { "grad_norm": 0.1983942687511444, "learning_rate": 9.078924879837822e-05, "loss": 0.0196, "step": 23650 }, { "grad_norm": 0.178458109498024, "learning_rate": 9.077968363761747e-05, "loss": 0.0244, "step": 23660 }, { "grad_norm": 0.19640779495239258, "learning_rate": 9.077011401725182e-05, "loss": 0.0195, "step": 23670 }, { "grad_norm": 0.20038914680480957, "learning_rate": 9.07605399383278e-05, "loss": 0.0244, "step": 23680 }, { "grad_norm": 0.2111203819513321, "learning_rate": 9.075096140189243e-05, "loss": 0.0247, "step": 23690 }, { "grad_norm": 0.2039097547531128, "learning_rate": 9.074137840899318e-05, "loss": 0.0197, "step": 23700 }, { "grad_norm": 0.21693706512451172, "learning_rate": 9.073179096067804e-05, "loss": 0.0252, "step": 23710 }, { "grad_norm": 0.25867167115211487, "learning_rate": 9.072219905799549e-05, "loss": 0.0209, "step": 23720 }, { "grad_norm": 0.2310677468776703, "learning_rate": 9.071260270199447e-05, "loss": 0.0276, "step": 23730 }, { "grad_norm": 0.21082252264022827, "learning_rate": 9.070300189372441e-05, "loss": 0.0219, "step": 23740 }, { "grad_norm": 0.22917906939983368, "learning_rate": 9.069339663423528e-05, "loss": 0.021, "step": 23750 }, { "grad_norm": 0.23790030181407928, "learning_rate": 9.068378692457747e-05, "loss": 0.0221, "step": 23760 }, { "grad_norm": 0.1936139613389969, "learning_rate": 9.067417276580189e-05, "loss": 0.0214, "step": 23770 }, { "grad_norm": 0.22789894044399261, "learning_rate": 9.066455415895993e-05, "loss": 0.0226, "step": 23780 }, { "grad_norm": 0.16204281151294708, "learning_rate": 9.065493110510346e-05, "loss": 0.0237, "step": 23790 }, { "grad_norm": 0.34216469526290894, "learning_rate": 9.064530360528484e-05, "loss": 0.0337, "step": 23800 }, { "grad_norm": 0.16911208629608154, "learning_rate": 9.063567166055695e-05, "loss": 0.0219, "step": 23810 }, { "grad_norm": 0.21756069362163544, "learning_rate": 9.062603527197308e-05, "loss": 0.0213, "step": 23820 }, { "grad_norm": 0.17890827357769012, "learning_rate": 9.06163944405871e-05, "loss": 0.0201, "step": 23830 }, { "grad_norm": 0.2074749916791916, "learning_rate": 9.060674916745327e-05, "loss": 0.0251, "step": 23840 }, { "grad_norm": 0.25883668661117554, "learning_rate": 9.05970994536264e-05, "loss": 0.0233, "step": 23850 }, { "grad_norm": 0.17827744781970978, "learning_rate": 9.05874453001618e-05, "loss": 0.023, "step": 23860 }, { "grad_norm": 0.20195043087005615, "learning_rate": 9.057778670811517e-05, "loss": 0.0194, "step": 23870 }, { "grad_norm": 0.21719956398010254, "learning_rate": 9.056812367854281e-05, "loss": 0.0204, "step": 23880 }, { "grad_norm": 0.24669362604618073, "learning_rate": 9.055845621250143e-05, "loss": 0.0266, "step": 23890 }, { "grad_norm": 0.2184804230928421, "learning_rate": 9.054878431104825e-05, "loss": 0.0236, "step": 23900 }, { "grad_norm": 0.3019160032272339, "learning_rate": 9.0539107975241e-05, "loss": 0.0244, "step": 23910 }, { "grad_norm": 0.2850725054740906, "learning_rate": 9.052942720613784e-05, "loss": 0.0257, "step": 23920 }, { "grad_norm": 0.23380406200885773, "learning_rate": 9.051974200479745e-05, "loss": 0.0245, "step": 23930 }, { "grad_norm": 0.26167231798171997, "learning_rate": 9.051005237227901e-05, "loss": 0.0229, "step": 23940 }, { "grad_norm": 0.16696590185165405, "learning_rate": 9.050035830964215e-05, "loss": 0.0235, "step": 23950 }, { "grad_norm": 0.20282508432865143, "learning_rate": 9.049065981794698e-05, "loss": 0.0261, "step": 23960 }, { "grad_norm": 0.26824355125427246, "learning_rate": 9.048095689825414e-05, "loss": 0.0213, "step": 23970 }, { "grad_norm": 0.2334497720003128, "learning_rate": 9.047124955162472e-05, "loss": 0.0254, "step": 23980 }, { "grad_norm": 0.22781731188297272, "learning_rate": 9.046153777912028e-05, "loss": 0.0238, "step": 23990 }, { "grad_norm": 0.20944644510746002, "learning_rate": 9.045182158180292e-05, "loss": 0.0248, "step": 24000 }, { "grad_norm": 0.19283659756183624, "learning_rate": 9.044210096073516e-05, "loss": 0.0227, "step": 24010 }, { "grad_norm": 0.23143714666366577, "learning_rate": 9.043237591698004e-05, "loss": 0.0278, "step": 24020 }, { "grad_norm": 0.26869967579841614, "learning_rate": 9.04226464516011e-05, "loss": 0.0227, "step": 24030 }, { "grad_norm": 0.19704848527908325, "learning_rate": 9.041291256566229e-05, "loss": 0.0246, "step": 24040 }, { "grad_norm": 0.20290125906467438, "learning_rate": 9.040317426022814e-05, "loss": 0.0227, "step": 24050 }, { "grad_norm": 0.24063245952129364, "learning_rate": 9.03934315363636e-05, "loss": 0.0223, "step": 24060 }, { "grad_norm": 0.18105758726596832, "learning_rate": 9.038368439513409e-05, "loss": 0.0235, "step": 24070 }, { "grad_norm": 0.18816997110843658, "learning_rate": 9.03739328376056e-05, "loss": 0.0248, "step": 24080 }, { "grad_norm": 0.22558873891830444, "learning_rate": 9.036417686484451e-05, "loss": 0.0208, "step": 24090 }, { "grad_norm": 0.2010239213705063, "learning_rate": 9.035441647791773e-05, "loss": 0.0208, "step": 24100 }, { "grad_norm": 0.33484500646591187, "learning_rate": 9.034465167789263e-05, "loss": 0.0252, "step": 24110 }, { "grad_norm": 0.1748771071434021, "learning_rate": 9.033488246583706e-05, "loss": 0.0191, "step": 24120 }, { "grad_norm": 0.19742727279663086, "learning_rate": 9.032510884281941e-05, "loss": 0.0232, "step": 24130 }, { "grad_norm": 0.2619743347167969, "learning_rate": 9.031533080990848e-05, "loss": 0.0224, "step": 24140 }, { "grad_norm": 0.16688933968544006, "learning_rate": 9.030554836817358e-05, "loss": 0.0216, "step": 24150 }, { "grad_norm": 0.2555164694786072, "learning_rate": 9.029576151868451e-05, "loss": 0.0236, "step": 24160 }, { "grad_norm": 0.20949645340442657, "learning_rate": 9.028597026251155e-05, "loss": 0.0175, "step": 24170 }, { "grad_norm": 0.231848806142807, "learning_rate": 9.027617460072547e-05, "loss": 0.0206, "step": 24180 }, { "grad_norm": 0.198483407497406, "learning_rate": 9.026637453439745e-05, "loss": 0.0268, "step": 24190 }, { "grad_norm": 0.23456408083438873, "learning_rate": 9.025657006459927e-05, "loss": 0.0174, "step": 24200 }, { "grad_norm": 0.11698855459690094, "learning_rate": 9.024676119240311e-05, "loss": 0.023, "step": 24210 }, { "grad_norm": 0.1959809511899948, "learning_rate": 9.023694791888166e-05, "loss": 0.0196, "step": 24220 }, { "grad_norm": 0.23155692219734192, "learning_rate": 9.022713024510808e-05, "loss": 0.0201, "step": 24230 }, { "grad_norm": 0.18010954558849335, "learning_rate": 9.021730817215601e-05, "loss": 0.0262, "step": 24240 }, { "grad_norm": 0.19512613117694855, "learning_rate": 9.02074817010996e-05, "loss": 0.0258, "step": 24250 }, { "grad_norm": 0.2337738275527954, "learning_rate": 9.019765083301342e-05, "loss": 0.0184, "step": 24260 }, { "grad_norm": 0.25181859731674194, "learning_rate": 9.01878155689726e-05, "loss": 0.0184, "step": 24270 }, { "grad_norm": 0.23022940754890442, "learning_rate": 9.017797591005268e-05, "loss": 0.0187, "step": 24280 }, { "grad_norm": 0.23606695234775543, "learning_rate": 9.016813185732972e-05, "loss": 0.0206, "step": 24290 }, { "grad_norm": 0.21706730127334595, "learning_rate": 9.015828341188027e-05, "loss": 0.0198, "step": 24300 }, { "grad_norm": 0.2045886367559433, "learning_rate": 9.01484305747813e-05, "loss": 0.0254, "step": 24310 }, { "grad_norm": 0.22217383980751038, "learning_rate": 9.013857334711033e-05, "loss": 0.0232, "step": 24320 }, { "grad_norm": 0.20411203801631927, "learning_rate": 9.012871172994534e-05, "loss": 0.0213, "step": 24330 }, { "grad_norm": 0.2695665657520294, "learning_rate": 9.011884572436476e-05, "loss": 0.0252, "step": 24340 }, { "grad_norm": 0.25298479199409485, "learning_rate": 9.010897533144754e-05, "loss": 0.0193, "step": 24350 }, { "grad_norm": 0.2644130289554596, "learning_rate": 9.009910055227306e-05, "loss": 0.0196, "step": 24360 }, { "grad_norm": 0.20354492962360382, "learning_rate": 9.008922138792124e-05, "loss": 0.0203, "step": 24370 }, { "grad_norm": 0.2119085043668747, "learning_rate": 9.007933783947244e-05, "loss": 0.0206, "step": 24380 }, { "grad_norm": 0.23048578202724457, "learning_rate": 9.006944990800752e-05, "loss": 0.0224, "step": 24390 }, { "grad_norm": 0.2222418487071991, "learning_rate": 9.005955759460779e-05, "loss": 0.0249, "step": 24400 }, { "grad_norm": 0.21536612510681152, "learning_rate": 9.004966090035508e-05, "loss": 0.0181, "step": 24410 }, { "grad_norm": 0.1859888881444931, "learning_rate": 9.003975982633166e-05, "loss": 0.0211, "step": 24420 }, { "grad_norm": 0.21103720366954803, "learning_rate": 9.00298543736203e-05, "loss": 0.0188, "step": 24430 }, { "grad_norm": 0.19126076996326447, "learning_rate": 9.001994454330427e-05, "loss": 0.0223, "step": 24440 }, { "grad_norm": 0.1706196367740631, "learning_rate": 9.001003033646727e-05, "loss": 0.0204, "step": 24450 }, { "grad_norm": 0.1793258637189865, "learning_rate": 9.00001117541935e-05, "loss": 0.0197, "step": 24460 }, { "grad_norm": 0.19384030997753143, "learning_rate": 8.999018879756764e-05, "loss": 0.0177, "step": 24470 }, { "grad_norm": 0.15892593562602997, "learning_rate": 8.998026146767487e-05, "loss": 0.0204, "step": 24480 }, { "grad_norm": 0.24014133214950562, "learning_rate": 8.99703297656008e-05, "loss": 0.021, "step": 24490 }, { "grad_norm": 0.173030287027359, "learning_rate": 8.996039369243156e-05, "loss": 0.024, "step": 24500 }, { "grad_norm": 0.26368969678878784, "learning_rate": 8.995045324925378e-05, "loss": 0.0217, "step": 24510 }, { "grad_norm": 0.19798849523067474, "learning_rate": 8.994050843715448e-05, "loss": 0.0201, "step": 24520 }, { "grad_norm": 0.21017304062843323, "learning_rate": 8.993055925722121e-05, "loss": 0.0269, "step": 24530 }, { "grad_norm": 0.24049361050128937, "learning_rate": 8.992060571054202e-05, "loss": 0.0236, "step": 24540 }, { "grad_norm": 0.19285278022289276, "learning_rate": 8.991064779820542e-05, "loss": 0.022, "step": 24550 }, { "grad_norm": 0.21269981563091278, "learning_rate": 8.990068552130036e-05, "loss": 0.0231, "step": 24560 }, { "grad_norm": 0.19044913351535797, "learning_rate": 8.989071888091634e-05, "loss": 0.0226, "step": 24570 }, { "grad_norm": 0.17359629273414612, "learning_rate": 8.988074787814329e-05, "loss": 0.0208, "step": 24580 }, { "grad_norm": 0.1974128633737564, "learning_rate": 8.987077251407158e-05, "loss": 0.0245, "step": 24590 }, { "grad_norm": 0.2133350670337677, "learning_rate": 8.986079278979216e-05, "loss": 0.0201, "step": 24600 }, { "grad_norm": 0.25513139367103577, "learning_rate": 8.985080870639635e-05, "loss": 0.0244, "step": 24610 }, { "grad_norm": 0.18399466574192047, "learning_rate": 8.984082026497603e-05, "loss": 0.0222, "step": 24620 }, { "grad_norm": 0.19600525498390198, "learning_rate": 8.98308274666235e-05, "loss": 0.0196, "step": 24630 }, { "grad_norm": 0.20143789052963257, "learning_rate": 8.982083031243155e-05, "loss": 0.0215, "step": 24640 }, { "grad_norm": 0.15543968975543976, "learning_rate": 8.98108288034935e-05, "loss": 0.0232, "step": 24650 }, { "grad_norm": 0.23793397843837738, "learning_rate": 8.980082294090305e-05, "loss": 0.0228, "step": 24660 }, { "grad_norm": 0.1932477355003357, "learning_rate": 8.979081272575443e-05, "loss": 0.0223, "step": 24670 }, { "grad_norm": 0.2076336145401001, "learning_rate": 8.978079815914236e-05, "loss": 0.0268, "step": 24680 }, { "grad_norm": 0.17409376800060272, "learning_rate": 8.977077924216202e-05, "loss": 0.0214, "step": 24690 }, { "grad_norm": 0.18381093442440033, "learning_rate": 8.976075597590905e-05, "loss": 0.022, "step": 24700 }, { "grad_norm": 0.18490871787071228, "learning_rate": 8.975072836147958e-05, "loss": 0.0226, "step": 24710 }, { "grad_norm": 0.28822773694992065, "learning_rate": 8.974069639997025e-05, "loss": 0.0321, "step": 24720 }, { "grad_norm": 0.13469713926315308, "learning_rate": 8.973066009247808e-05, "loss": 0.021, "step": 24730 }, { "grad_norm": 0.29228073358535767, "learning_rate": 8.972061944010066e-05, "loss": 0.0269, "step": 24740 }, { "grad_norm": 0.22267253696918488, "learning_rate": 8.971057444393603e-05, "loss": 0.0234, "step": 24750 }, { "grad_norm": 0.21943546831607819, "learning_rate": 8.970052510508268e-05, "loss": 0.0221, "step": 24760 }, { "grad_norm": 0.3297572731971741, "learning_rate": 8.969047142463959e-05, "loss": 0.022, "step": 24770 }, { "grad_norm": 0.2813566029071808, "learning_rate": 8.968041340370621e-05, "loss": 0.0245, "step": 24780 }, { "grad_norm": 0.2887701690196991, "learning_rate": 8.96703510433825e-05, "loss": 0.0251, "step": 24790 }, { "grad_norm": 0.2653029263019562, "learning_rate": 8.966028434476883e-05, "loss": 0.0252, "step": 24800 }, { "grad_norm": 0.25617536902427673, "learning_rate": 8.96502133089661e-05, "loss": 0.0225, "step": 24810 }, { "grad_norm": 0.24109096825122833, "learning_rate": 8.964013793707564e-05, "loss": 0.0207, "step": 24820 }, { "grad_norm": 0.2011227011680603, "learning_rate": 8.963005823019932e-05, "loss": 0.018, "step": 24830 }, { "grad_norm": 0.2447073757648468, "learning_rate": 8.961997418943939e-05, "loss": 0.0233, "step": 24840 }, { "grad_norm": 0.2456800639629364, "learning_rate": 8.960988581589865e-05, "loss": 0.0187, "step": 24850 }, { "grad_norm": 0.21389569342136383, "learning_rate": 8.959979311068037e-05, "loss": 0.022, "step": 24860 }, { "grad_norm": 0.20022058486938477, "learning_rate": 8.958969607488823e-05, "loss": 0.0191, "step": 24870 }, { "grad_norm": 0.22395531833171844, "learning_rate": 8.957959470962647e-05, "loss": 0.023, "step": 24880 }, { "grad_norm": 0.2612501382827759, "learning_rate": 8.956948901599971e-05, "loss": 0.0286, "step": 24890 }, { "grad_norm": 0.18218152225017548, "learning_rate": 8.955937899511315e-05, "loss": 0.0286, "step": 24900 }, { "grad_norm": 0.24649697542190552, "learning_rate": 8.954926464807238e-05, "loss": 0.02, "step": 24910 }, { "grad_norm": 0.1854904294013977, "learning_rate": 8.953914597598347e-05, "loss": 0.02, "step": 24920 }, { "grad_norm": 0.19763164222240448, "learning_rate": 8.952902297995303e-05, "loss": 0.0231, "step": 24930 }, { "grad_norm": 0.1875762641429901, "learning_rate": 8.951889566108804e-05, "loss": 0.0242, "step": 24940 }, { "grad_norm": 0.25822943449020386, "learning_rate": 8.950876402049606e-05, "loss": 0.0232, "step": 24950 }, { "grad_norm": 0.17134597897529602, "learning_rate": 8.949862805928504e-05, "loss": 0.0181, "step": 24960 }, { "grad_norm": 0.21704967319965363, "learning_rate": 8.948848777856343e-05, "loss": 0.0197, "step": 24970 }, { "grad_norm": 0.25491607189178467, "learning_rate": 8.947834317944017e-05, "loss": 0.0212, "step": 24980 }, { "grad_norm": 0.1681988537311554, "learning_rate": 8.946819426302466e-05, "loss": 0.0194, "step": 24990 }, { "grad_norm": 0.22378672659397125, "learning_rate": 8.945804103042676e-05, "loss": 0.0183, "step": 25000 }, { "grad_norm": 0.29997456073760986, "learning_rate": 8.944788348275681e-05, "loss": 0.0227, "step": 25010 }, { "grad_norm": 0.1897476762533188, "learning_rate": 8.943772162112565e-05, "loss": 0.0188, "step": 25020 }, { "grad_norm": 0.15904785692691803, "learning_rate": 8.942755544664454e-05, "loss": 0.0196, "step": 25030 }, { "grad_norm": 0.18387967348098755, "learning_rate": 8.941738496042525e-05, "loss": 0.0205, "step": 25040 }, { "grad_norm": 0.2158142775297165, "learning_rate": 8.940721016357999e-05, "loss": 0.0185, "step": 25050 }, { "grad_norm": 0.18774457275867462, "learning_rate": 8.939703105722148e-05, "loss": 0.0242, "step": 25060 }, { "grad_norm": 0.23009106516838074, "learning_rate": 8.93868476424629e-05, "loss": 0.0214, "step": 25070 }, { "grad_norm": 0.18795831501483917, "learning_rate": 8.937665992041786e-05, "loss": 0.02, "step": 25080 }, { "grad_norm": 0.18462491035461426, "learning_rate": 8.93664678922005e-05, "loss": 0.0221, "step": 25090 }, { "grad_norm": 0.2903391718864441, "learning_rate": 8.93562715589254e-05, "loss": 0.019, "step": 25100 }, { "grad_norm": 0.2501344084739685, "learning_rate": 8.934607092170762e-05, "loss": 0.0184, "step": 25110 }, { "grad_norm": 0.21352817118167877, "learning_rate": 8.933586598166266e-05, "loss": 0.0209, "step": 25120 }, { "grad_norm": 0.2484101504087448, "learning_rate": 8.932565673990655e-05, "loss": 0.022, "step": 25130 }, { "grad_norm": 0.22639408707618713, "learning_rate": 8.931544319755574e-05, "loss": 0.0218, "step": 25140 }, { "grad_norm": 0.2516220211982727, "learning_rate": 8.930522535572718e-05, "loss": 0.0212, "step": 25150 }, { "grad_norm": 0.18055060505867004, "learning_rate": 8.929500321553826e-05, "loss": 0.0205, "step": 25160 }, { "grad_norm": 0.20074135065078735, "learning_rate": 8.928477677810686e-05, "loss": 0.02, "step": 25170 }, { "grad_norm": 0.17878508567810059, "learning_rate": 8.927454604455137e-05, "loss": 0.0192, "step": 25180 }, { "grad_norm": 0.18331220746040344, "learning_rate": 8.926431101599053e-05, "loss": 0.0183, "step": 25190 }, { "grad_norm": 0.17357924580574036, "learning_rate": 8.925407169354369e-05, "loss": 0.0222, "step": 25200 }, { "grad_norm": 0.2566336691379547, "learning_rate": 8.92438280783306e-05, "loss": 0.0203, "step": 25210 }, { "grad_norm": 0.21524056792259216, "learning_rate": 8.923358017147146e-05, "loss": 0.024, "step": 25220 }, { "grad_norm": 0.19257164001464844, "learning_rate": 8.922332797408697e-05, "loss": 0.0237, "step": 25230 }, { "grad_norm": 0.31005895137786865, "learning_rate": 8.921307148729831e-05, "loss": 0.0226, "step": 25240 }, { "grad_norm": 0.22418923676013947, "learning_rate": 8.920281071222712e-05, "loss": 0.0233, "step": 25250 }, { "grad_norm": 0.19037604331970215, "learning_rate": 8.919254564999548e-05, "loss": 0.0182, "step": 25260 }, { "grad_norm": 0.20058338344097137, "learning_rate": 8.918227630172598e-05, "loss": 0.0194, "step": 25270 }, { "grad_norm": 0.2309984713792801, "learning_rate": 8.917200266854165e-05, "loss": 0.0228, "step": 25280 }, { "grad_norm": 0.17790661752223969, "learning_rate": 8.9161724751566e-05, "loss": 0.0239, "step": 25290 }, { "grad_norm": 0.19260120391845703, "learning_rate": 8.915144255192302e-05, "loss": 0.0196, "step": 25300 }, { "grad_norm": 0.18396036326885223, "learning_rate": 8.914115607073714e-05, "loss": 0.0175, "step": 25310 }, { "grad_norm": 0.24014244973659515, "learning_rate": 8.913086530913327e-05, "loss": 0.0182, "step": 25320 }, { "grad_norm": 0.21573930978775024, "learning_rate": 8.912057026823681e-05, "loss": 0.0192, "step": 25330 }, { "grad_norm": 0.21030135452747345, "learning_rate": 8.91102709491736e-05, "loss": 0.0178, "step": 25340 }, { "grad_norm": 0.23797166347503662, "learning_rate": 8.909996735306996e-05, "loss": 0.0189, "step": 25350 }, { "grad_norm": 0.21475936472415924, "learning_rate": 8.908965948105268e-05, "loss": 0.0219, "step": 25360 }, { "grad_norm": 0.2467450201511383, "learning_rate": 8.907934733424901e-05, "loss": 0.0247, "step": 25370 }, { "grad_norm": 0.22466620802879333, "learning_rate": 8.906903091378666e-05, "loss": 0.0207, "step": 25380 }, { "grad_norm": 0.23486803472042084, "learning_rate": 8.905871022079384e-05, "loss": 0.0205, "step": 25390 }, { "grad_norm": 0.2151714563369751, "learning_rate": 8.90483852563992e-05, "loss": 0.0173, "step": 25400 }, { "grad_norm": 0.2512465715408325, "learning_rate": 8.903805602173185e-05, "loss": 0.0178, "step": 25410 }, { "grad_norm": 0.1693669557571411, "learning_rate": 8.902772251792137e-05, "loss": 0.0187, "step": 25420 }, { "grad_norm": 0.20837050676345825, "learning_rate": 8.901738474609786e-05, "loss": 0.0212, "step": 25430 }, { "grad_norm": 0.20994454622268677, "learning_rate": 8.900704270739179e-05, "loss": 0.0199, "step": 25440 }, { "grad_norm": 0.1878291666507721, "learning_rate": 8.89966964029342e-05, "loss": 0.0178, "step": 25450 }, { "grad_norm": 0.2064250111579895, "learning_rate": 8.898634583385652e-05, "loss": 0.0203, "step": 25460 }, { "grad_norm": 0.22460207343101501, "learning_rate": 8.897599100129065e-05, "loss": 0.023, "step": 25470 }, { "grad_norm": 0.16293083131313324, "learning_rate": 8.896563190636903e-05, "loss": 0.0174, "step": 25480 }, { "grad_norm": 0.177289217710495, "learning_rate": 8.895526855022448e-05, "loss": 0.018, "step": 25490 }, { "grad_norm": 0.20390349626541138, "learning_rate": 8.894490093399033e-05, "loss": 0.0191, "step": 25500 }, { "grad_norm": 0.18043817579746246, "learning_rate": 8.893452905880035e-05, "loss": 0.022, "step": 25510 }, { "grad_norm": 0.19886110723018646, "learning_rate": 8.892415292578883e-05, "loss": 0.0189, "step": 25520 }, { "grad_norm": 0.2499019056558609, "learning_rate": 8.891377253609046e-05, "loss": 0.021, "step": 25530 }, { "grad_norm": 0.24286822974681854, "learning_rate": 8.890338789084043e-05, "loss": 0.0221, "step": 25540 }, { "grad_norm": 0.18176044523715973, "learning_rate": 8.88929989911744e-05, "loss": 0.0226, "step": 25550 }, { "grad_norm": 0.1961653083562851, "learning_rate": 8.888260583822847e-05, "loss": 0.0248, "step": 25560 }, { "grad_norm": 0.21135511994361877, "learning_rate": 8.887220843313921e-05, "loss": 0.0202, "step": 25570 }, { "grad_norm": 0.1896924078464508, "learning_rate": 8.88618067770437e-05, "loss": 0.0188, "step": 25580 }, { "grad_norm": 0.1423499435186386, "learning_rate": 8.885140087107942e-05, "loss": 0.0208, "step": 25590 }, { "grad_norm": 0.23885037004947662, "learning_rate": 8.884099071638436e-05, "loss": 0.0215, "step": 25600 }, { "grad_norm": 0.24280574917793274, "learning_rate": 8.883057631409695e-05, "loss": 0.0211, "step": 25610 }, { "grad_norm": 0.26537927985191345, "learning_rate": 8.882015766535608e-05, "loss": 0.0191, "step": 25620 }, { "grad_norm": 0.25200822949409485, "learning_rate": 8.880973477130115e-05, "loss": 0.0221, "step": 25630 }, { "grad_norm": 0.22180719673633575, "learning_rate": 8.879930763307197e-05, "loss": 0.0229, "step": 25640 }, { "grad_norm": 0.1745332032442093, "learning_rate": 8.878887625180884e-05, "loss": 0.0184, "step": 25650 }, { "grad_norm": 0.23576679825782776, "learning_rate": 8.877844062865253e-05, "loss": 0.0256, "step": 25660 }, { "grad_norm": 0.2854987382888794, "learning_rate": 8.876800076474424e-05, "loss": 0.0229, "step": 25670 }, { "grad_norm": 0.2516017556190491, "learning_rate": 8.875755666122568e-05, "loss": 0.0226, "step": 25680 }, { "grad_norm": 0.2444961816072464, "learning_rate": 8.8747108319239e-05, "loss": 0.021, "step": 25690 }, { "grad_norm": 0.22056730091571808, "learning_rate": 8.87366557399268e-05, "loss": 0.0238, "step": 25700 }, { "grad_norm": 0.27957600355148315, "learning_rate": 8.872619892443217e-05, "loss": 0.0252, "step": 25710 }, { "grad_norm": 0.21863918006420135, "learning_rate": 8.871573787389865e-05, "loss": 0.02, "step": 25720 }, { "grad_norm": 0.1639733612537384, "learning_rate": 8.870527258947024e-05, "loss": 0.0199, "step": 25730 }, { "grad_norm": 0.19885826110839844, "learning_rate": 8.869480307229143e-05, "loss": 0.019, "step": 25740 }, { "grad_norm": 0.22739730775356293, "learning_rate": 8.868432932350712e-05, "loss": 0.0216, "step": 25750 }, { "grad_norm": 0.16569115221500397, "learning_rate": 8.867385134426272e-05, "loss": 0.0187, "step": 25760 }, { "grad_norm": 0.15554074943065643, "learning_rate": 8.866336913570407e-05, "loss": 0.0227, "step": 25770 }, { "grad_norm": 0.20410321652889252, "learning_rate": 8.865288269897751e-05, "loss": 0.0226, "step": 25780 }, { "grad_norm": 0.23717910051345825, "learning_rate": 8.864239203522981e-05, "loss": 0.0204, "step": 25790 }, { "grad_norm": 0.24735288321971893, "learning_rate": 8.863189714560822e-05, "loss": 0.0223, "step": 25800 }, { "grad_norm": 0.2798158824443817, "learning_rate": 8.862139803126043e-05, "loss": 0.0181, "step": 25810 }, { "grad_norm": 0.19828082621097565, "learning_rate": 8.861089469333463e-05, "loss": 0.0187, "step": 25820 }, { "grad_norm": 0.22085505723953247, "learning_rate": 8.860038713297944e-05, "loss": 0.021, "step": 25830 }, { "grad_norm": 0.22279536724090576, "learning_rate": 8.858987535134394e-05, "loss": 0.0198, "step": 25840 }, { "grad_norm": 0.2654012143611908, "learning_rate": 8.857935934957769e-05, "loss": 0.0228, "step": 25850 }, { "grad_norm": 0.24435900151729584, "learning_rate": 8.856883912883071e-05, "loss": 0.0238, "step": 25860 }, { "grad_norm": 0.21216534078121185, "learning_rate": 8.855831469025346e-05, "loss": 0.0217, "step": 25870 }, { "grad_norm": 0.268400102853775, "learning_rate": 8.854778603499689e-05, "loss": 0.0263, "step": 25880 }, { "grad_norm": 0.23955823481082916, "learning_rate": 8.85372531642124e-05, "loss": 0.0187, "step": 25890 }, { "grad_norm": 0.2360231876373291, "learning_rate": 8.852671607905185e-05, "loss": 0.0205, "step": 25900 }, { "grad_norm": 0.20062732696533203, "learning_rate": 8.851617478066754e-05, "loss": 0.021, "step": 25910 }, { "grad_norm": 0.20277059078216553, "learning_rate": 8.850562927021227e-05, "loss": 0.0226, "step": 25920 }, { "grad_norm": 0.20667968690395355, "learning_rate": 8.849507954883928e-05, "loss": 0.0203, "step": 25930 }, { "grad_norm": 0.24998952448368073, "learning_rate": 8.848452561770226e-05, "loss": 0.018, "step": 25940 }, { "grad_norm": 0.2013891339302063, "learning_rate": 8.847396747795538e-05, "loss": 0.0185, "step": 25950 }, { "grad_norm": 0.22772128880023956, "learning_rate": 8.846340513075327e-05, "loss": 0.0211, "step": 25960 }, { "grad_norm": 0.21795029938220978, "learning_rate": 8.845283857725099e-05, "loss": 0.0291, "step": 25970 }, { "grad_norm": 0.22979530692100525, "learning_rate": 8.844226781860409e-05, "loss": 0.0283, "step": 25980 }, { "grad_norm": 0.18532760441303253, "learning_rate": 8.84316928559686e-05, "loss": 0.0186, "step": 25990 }, { "grad_norm": 0.24344275891780853, "learning_rate": 8.842111369050094e-05, "loss": 0.0184, "step": 26000 }, { "grad_norm": 0.18142549693584442, "learning_rate": 8.841053032335808e-05, "loss": 0.018, "step": 26010 }, { "grad_norm": 0.18360041081905365, "learning_rate": 8.839994275569735e-05, "loss": 0.0239, "step": 26020 }, { "grad_norm": 0.19323259592056274, "learning_rate": 8.838935098867662e-05, "loss": 0.0185, "step": 26030 }, { "grad_norm": 0.21199846267700195, "learning_rate": 8.837875502345418e-05, "loss": 0.0235, "step": 26040 }, { "grad_norm": 0.21873606741428375, "learning_rate": 8.83681548611888e-05, "loss": 0.0193, "step": 26050 }, { "grad_norm": 0.20520898699760437, "learning_rate": 8.835755050303969e-05, "loss": 0.0221, "step": 26060 }, { "grad_norm": 0.19156919419765472, "learning_rate": 8.834694195016653e-05, "loss": 0.0248, "step": 26070 }, { "grad_norm": 0.23434774577617645, "learning_rate": 8.833632920372942e-05, "loss": 0.0238, "step": 26080 }, { "grad_norm": 0.17917834222316742, "learning_rate": 8.832571226488903e-05, "loss": 0.0235, "step": 26090 }, { "grad_norm": 0.24688178300857544, "learning_rate": 8.831509113480634e-05, "loss": 0.0225, "step": 26100 }, { "grad_norm": 0.19300569593906403, "learning_rate": 8.83044658146429e-05, "loss": 0.0204, "step": 26110 }, { "grad_norm": 0.276560515165329, "learning_rate": 8.829383630556067e-05, "loss": 0.0175, "step": 26120 }, { "grad_norm": 0.2012842744588852, "learning_rate": 8.828320260872207e-05, "loss": 0.0173, "step": 26130 }, { "grad_norm": 0.1884298324584961, "learning_rate": 8.827256472529e-05, "loss": 0.0186, "step": 26140 }, { "grad_norm": 0.20692181587219238, "learning_rate": 8.826192265642778e-05, "loss": 0.0175, "step": 26150 }, { "grad_norm": 0.26383331418037415, "learning_rate": 8.825127640329923e-05, "loss": 0.0191, "step": 26160 }, { "grad_norm": 0.2165527641773224, "learning_rate": 8.824062596706861e-05, "loss": 0.0214, "step": 26170 }, { "grad_norm": 0.19197554886341095, "learning_rate": 8.822997134890062e-05, "loss": 0.0176, "step": 26180 }, { "grad_norm": 0.17619676887989044, "learning_rate": 8.821931254996044e-05, "loss": 0.0162, "step": 26190 }, { "grad_norm": 0.1474439799785614, "learning_rate": 8.82086495714137e-05, "loss": 0.0154, "step": 26200 }, { "grad_norm": 0.2800079584121704, "learning_rate": 8.81979824144265e-05, "loss": 0.0218, "step": 26210 }, { "grad_norm": 0.23040005564689636, "learning_rate": 8.818731108016536e-05, "loss": 0.0237, "step": 26220 }, { "grad_norm": 0.23141799867153168, "learning_rate": 8.81766355697973e-05, "loss": 0.0206, "step": 26230 }, { "grad_norm": 0.2618386447429657, "learning_rate": 8.816595588448977e-05, "loss": 0.0199, "step": 26240 }, { "grad_norm": 0.20315493643283844, "learning_rate": 8.81552720254107e-05, "loss": 0.0196, "step": 26250 }, { "grad_norm": 0.22039221227169037, "learning_rate": 8.814458399372842e-05, "loss": 0.0231, "step": 26260 }, { "grad_norm": 0.17433053255081177, "learning_rate": 8.813389179061181e-05, "loss": 0.0239, "step": 26270 }, { "grad_norm": 0.20700518786907196, "learning_rate": 8.812319541723012e-05, "loss": 0.021, "step": 26280 }, { "grad_norm": 0.23894888162612915, "learning_rate": 8.811249487475309e-05, "loss": 0.0234, "step": 26290 }, { "grad_norm": 0.22322241961956024, "learning_rate": 8.810179016435092e-05, "loss": 0.0255, "step": 26300 }, { "grad_norm": 0.1733947992324829, "learning_rate": 8.809108128719428e-05, "loss": 0.0221, "step": 26310 }, { "grad_norm": 0.24577835202217102, "learning_rate": 8.808036824445424e-05, "loss": 0.0202, "step": 26320 }, { "grad_norm": 0.266122430562973, "learning_rate": 8.806965103730238e-05, "loss": 0.0257, "step": 26330 }, { "grad_norm": 0.21655212342739105, "learning_rate": 8.805892966691074e-05, "loss": 0.0225, "step": 26340 }, { "grad_norm": 0.2550496459007263, "learning_rate": 8.804820413445175e-05, "loss": 0.0202, "step": 26350 }, { "grad_norm": 0.24568495154380798, "learning_rate": 8.803747444109837e-05, "loss": 0.0226, "step": 26360 }, { "grad_norm": 0.17633642256259918, "learning_rate": 8.802674058802399e-05, "loss": 0.0251, "step": 26370 }, { "grad_norm": 0.2709145247936249, "learning_rate": 8.801600257640241e-05, "loss": 0.0189, "step": 26380 }, { "grad_norm": 0.19318556785583496, "learning_rate": 8.800526040740795e-05, "loss": 0.0235, "step": 26390 }, { "grad_norm": 0.1939421147108078, "learning_rate": 8.799451408221535e-05, "loss": 0.0186, "step": 26400 }, { "grad_norm": 0.2528875470161438, "learning_rate": 8.798376360199982e-05, "loss": 0.0229, "step": 26410 }, { "grad_norm": 0.18805934488773346, "learning_rate": 8.797300896793701e-05, "loss": 0.0187, "step": 26420 }, { "grad_norm": 0.2129966914653778, "learning_rate": 8.796225018120302e-05, "loss": 0.023, "step": 26430 }, { "grad_norm": 0.18275824189186096, "learning_rate": 8.795148724297444e-05, "loss": 0.0209, "step": 26440 }, { "grad_norm": 0.23237724602222443, "learning_rate": 8.794072015442825e-05, "loss": 0.0258, "step": 26450 }, { "grad_norm": 0.22046419978141785, "learning_rate": 8.792994891674198e-05, "loss": 0.022, "step": 26460 }, { "grad_norm": 0.2949709892272949, "learning_rate": 8.79191735310935e-05, "loss": 0.0275, "step": 26470 }, { "grad_norm": 0.2501460313796997, "learning_rate": 8.790839399866122e-05, "loss": 0.0233, "step": 26480 }, { "grad_norm": 0.1897144764661789, "learning_rate": 8.789761032062397e-05, "loss": 0.0203, "step": 26490 }, { "grad_norm": 0.17006100714206696, "learning_rate": 8.788682249816103e-05, "loss": 0.0201, "step": 26500 }, { "grad_norm": 0.19550779461860657, "learning_rate": 8.787603053245215e-05, "loss": 0.0218, "step": 26510 }, { "grad_norm": 0.21193134784698486, "learning_rate": 8.78652344246775e-05, "loss": 0.0196, "step": 26520 }, { "grad_norm": 0.2728941738605499, "learning_rate": 8.785443417601776e-05, "loss": 0.023, "step": 26530 }, { "grad_norm": 0.15862737596035004, "learning_rate": 8.784362978765401e-05, "loss": 0.0163, "step": 26540 }, { "grad_norm": 0.20731568336486816, "learning_rate": 8.783282126076779e-05, "loss": 0.0197, "step": 26550 }, { "grad_norm": 0.18674670159816742, "learning_rate": 8.782200859654112e-05, "loss": 0.0193, "step": 26560 }, { "grad_norm": 0.2066422551870346, "learning_rate": 8.781119179615646e-05, "loss": 0.0173, "step": 26570 }, { "grad_norm": 0.16894620656967163, "learning_rate": 8.780037086079674e-05, "loss": 0.019, "step": 26580 }, { "grad_norm": 0.19030813872814178, "learning_rate": 8.778954579164527e-05, "loss": 0.0197, "step": 26590 }, { "grad_norm": 0.28153565526008606, "learning_rate": 8.777871658988588e-05, "loss": 0.0213, "step": 26600 }, { "grad_norm": 0.20110823214054108, "learning_rate": 8.776788325670285e-05, "loss": 0.0188, "step": 26610 }, { "grad_norm": 0.1880878061056137, "learning_rate": 8.775704579328089e-05, "loss": 0.0199, "step": 26620 }, { "grad_norm": 0.20449048280715942, "learning_rate": 8.774620420080517e-05, "loss": 0.0208, "step": 26630 }, { "grad_norm": 0.18156953155994415, "learning_rate": 8.773535848046131e-05, "loss": 0.0171, "step": 26640 }, { "grad_norm": 0.19818851351737976, "learning_rate": 8.772450863343538e-05, "loss": 0.0202, "step": 26650 }, { "grad_norm": 0.17284545302391052, "learning_rate": 8.77136546609139e-05, "loss": 0.0199, "step": 26660 }, { "grad_norm": 0.2843588888645172, "learning_rate": 8.770279656408385e-05, "loss": 0.0197, "step": 26670 }, { "grad_norm": 0.2336803376674652, "learning_rate": 8.769193434413265e-05, "loss": 0.0193, "step": 26680 }, { "grad_norm": 0.13939699530601501, "learning_rate": 8.76810680022482e-05, "loss": 0.0178, "step": 26690 }, { "grad_norm": 0.2073839008808136, "learning_rate": 8.767019753961878e-05, "loss": 0.02, "step": 26700 }, { "grad_norm": 0.16779255867004395, "learning_rate": 8.765932295743321e-05, "loss": 0.0196, "step": 26710 }, { "grad_norm": 0.2310784012079239, "learning_rate": 8.764844425688068e-05, "loss": 0.0219, "step": 26720 }, { "grad_norm": 0.24764756858348846, "learning_rate": 8.763756143915092e-05, "loss": 0.021, "step": 26730 }, { "grad_norm": 0.24360103905200958, "learning_rate": 8.7626674505434e-05, "loss": 0.0227, "step": 26740 }, { "grad_norm": 0.3189185559749603, "learning_rate": 8.761578345692053e-05, "loss": 0.0204, "step": 26750 }, { "grad_norm": 0.2347365766763687, "learning_rate": 8.760488829480156e-05, "loss": 0.0216, "step": 26760 }, { "grad_norm": 0.21291089057922363, "learning_rate": 8.759398902026854e-05, "loss": 0.019, "step": 26770 }, { "grad_norm": 0.2138233184814453, "learning_rate": 8.758308563451339e-05, "loss": 0.018, "step": 26780 }, { "grad_norm": 0.20559397339820862, "learning_rate": 8.75721781387285e-05, "loss": 0.022, "step": 26790 }, { "grad_norm": 0.19245582818984985, "learning_rate": 8.75612665341067e-05, "loss": 0.0203, "step": 26800 }, { "grad_norm": 0.21669617295265198, "learning_rate": 8.755035082184126e-05, "loss": 0.0198, "step": 26810 }, { "grad_norm": 0.18910348415374756, "learning_rate": 8.753943100312592e-05, "loss": 0.0157, "step": 26820 }, { "grad_norm": 0.1764003485441208, "learning_rate": 8.752850707915484e-05, "loss": 0.0175, "step": 26830 }, { "grad_norm": 0.16058075428009033, "learning_rate": 8.751757905112264e-05, "loss": 0.0182, "step": 26840 }, { "grad_norm": 0.21105018258094788, "learning_rate": 8.75066469202244e-05, "loss": 0.0214, "step": 26850 }, { "grad_norm": 0.2091876119375229, "learning_rate": 8.749571068765567e-05, "loss": 0.0216, "step": 26860 }, { "grad_norm": 0.20740091800689697, "learning_rate": 8.748477035461238e-05, "loss": 0.0203, "step": 26870 }, { "grad_norm": 0.25690412521362305, "learning_rate": 8.747382592229095e-05, "loss": 0.0269, "step": 26880 }, { "grad_norm": 0.22485250234603882, "learning_rate": 8.746287739188828e-05, "loss": 0.0192, "step": 26890 }, { "grad_norm": 0.17881932854652405, "learning_rate": 8.745192476460165e-05, "loss": 0.0214, "step": 26900 }, { "grad_norm": 0.2234371453523636, "learning_rate": 8.744096804162882e-05, "loss": 0.0227, "step": 26910 }, { "grad_norm": 0.22481317818164825, "learning_rate": 8.743000722416804e-05, "loss": 0.0203, "step": 26920 }, { "grad_norm": 0.25782325863838196, "learning_rate": 8.741904231341793e-05, "loss": 0.0257, "step": 26930 }, { "grad_norm": 0.23599083721637726, "learning_rate": 8.740807331057762e-05, "loss": 0.0227, "step": 26940 }, { "grad_norm": 0.27117055654525757, "learning_rate": 8.739710021684667e-05, "loss": 0.0244, "step": 26950 }, { "grad_norm": 0.24723565578460693, "learning_rate": 8.738612303342503e-05, "loss": 0.0262, "step": 26960 }, { "grad_norm": 0.27368220686912537, "learning_rate": 8.73751417615132e-05, "loss": 0.0215, "step": 26970 }, { "grad_norm": 0.2881929278373718, "learning_rate": 8.736415640231208e-05, "loss": 0.0203, "step": 26980 }, { "grad_norm": 0.24464651942253113, "learning_rate": 8.735316695702297e-05, "loss": 0.0227, "step": 26990 }, { "grad_norm": 0.17763490974903107, "learning_rate": 8.734217342684769e-05, "loss": 0.0165, "step": 27000 }, { "grad_norm": 0.1834113597869873, "learning_rate": 8.733117581298847e-05, "loss": 0.0245, "step": 27010 }, { "grad_norm": 0.242840975522995, "learning_rate": 8.732017411664796e-05, "loss": 0.0223, "step": 27020 }, { "grad_norm": 0.20391519367694855, "learning_rate": 8.730916833902936e-05, "loss": 0.0195, "step": 27030 }, { "grad_norm": 0.24889381229877472, "learning_rate": 8.729815848133618e-05, "loss": 0.0206, "step": 27040 }, { "grad_norm": 0.14194513857364655, "learning_rate": 8.728714454477247e-05, "loss": 0.0161, "step": 27050 }, { "grad_norm": 0.1863584816455841, "learning_rate": 8.727612653054269e-05, "loss": 0.0203, "step": 27060 }, { "grad_norm": 0.2006455659866333, "learning_rate": 8.726510443985176e-05, "loss": 0.0183, "step": 27070 }, { "grad_norm": 0.27171358466148376, "learning_rate": 8.725407827390503e-05, "loss": 0.0174, "step": 27080 }, { "grad_norm": 0.2245505005121231, "learning_rate": 8.724304803390833e-05, "loss": 0.0184, "step": 27090 }, { "grad_norm": 0.16600090265274048, "learning_rate": 8.723201372106788e-05, "loss": 0.0185, "step": 27100 }, { "grad_norm": 0.1956051141023636, "learning_rate": 8.722097533659038e-05, "loss": 0.0204, "step": 27110 }, { "grad_norm": 0.26847192645072937, "learning_rate": 8.720993288168299e-05, "loss": 0.0246, "step": 27120 }, { "grad_norm": 0.2242664396762848, "learning_rate": 8.719888635755327e-05, "loss": 0.0197, "step": 27130 }, { "grad_norm": 0.2351199835538864, "learning_rate": 8.718783576540928e-05, "loss": 0.0197, "step": 27140 }, { "grad_norm": 0.19046661257743835, "learning_rate": 8.717678110645948e-05, "loss": 0.0171, "step": 27150 }, { "grad_norm": 0.1852523535490036, "learning_rate": 8.716572238191279e-05, "loss": 0.0175, "step": 27160 }, { "grad_norm": 0.19545775651931763, "learning_rate": 8.715465959297857e-05, "loss": 0.0183, "step": 27170 }, { "grad_norm": 0.2706356644630432, "learning_rate": 8.714359274086665e-05, "loss": 0.0228, "step": 27180 }, { "grad_norm": 0.23191943764686584, "learning_rate": 8.713252182678726e-05, "loss": 0.0169, "step": 27190 }, { "grad_norm": 0.2192414551973343, "learning_rate": 8.712144685195112e-05, "loss": 0.0176, "step": 27200 }, { "grad_norm": 0.22529323399066925, "learning_rate": 8.711036781756936e-05, "loss": 0.0175, "step": 27210 }, { "grad_norm": 0.234313502907753, "learning_rate": 8.709928472485357e-05, "loss": 0.0216, "step": 27220 }, { "grad_norm": 0.16900429129600525, "learning_rate": 8.708819757501579e-05, "loss": 0.0186, "step": 27230 }, { "grad_norm": 0.16305619478225708, "learning_rate": 8.707710636926846e-05, "loss": 0.0175, "step": 27240 }, { "grad_norm": 0.22668342292308807, "learning_rate": 8.706601110882455e-05, "loss": 0.021, "step": 27250 }, { "grad_norm": 0.27829378843307495, "learning_rate": 8.705491179489738e-05, "loss": 0.0203, "step": 27260 }, { "grad_norm": 0.23675675690174103, "learning_rate": 8.704380842870077e-05, "loss": 0.0194, "step": 27270 }, { "grad_norm": 0.2028503566980362, "learning_rate": 8.703270101144895e-05, "loss": 0.02, "step": 27280 }, { "grad_norm": 0.2288213074207306, "learning_rate": 8.702158954435664e-05, "loss": 0.0188, "step": 27290 }, { "grad_norm": 0.20026056468486786, "learning_rate": 8.701047402863896e-05, "loss": 0.0196, "step": 27300 }, { "grad_norm": 0.1575300395488739, "learning_rate": 8.699935446551148e-05, "loss": 0.0179, "step": 27310 }, { "grad_norm": 0.24285930395126343, "learning_rate": 8.698823085619022e-05, "loss": 0.0184, "step": 27320 }, { "grad_norm": 0.20091448724269867, "learning_rate": 8.697710320189166e-05, "loss": 0.0168, "step": 27330 }, { "grad_norm": 0.16484642028808594, "learning_rate": 8.696597150383268e-05, "loss": 0.023, "step": 27340 }, { "grad_norm": 0.2872350513935089, "learning_rate": 8.695483576323063e-05, "loss": 0.0198, "step": 27350 }, { "grad_norm": 0.18875156342983246, "learning_rate": 8.69436959813033e-05, "loss": 0.0198, "step": 27360 }, { "grad_norm": 0.18557126820087433, "learning_rate": 8.693255215926892e-05, "loss": 0.0217, "step": 27370 }, { "grad_norm": 0.16882450878620148, "learning_rate": 8.692140429834617e-05, "loss": 0.0206, "step": 27380 }, { "grad_norm": 0.1936689019203186, "learning_rate": 8.691025239975415e-05, "loss": 0.0198, "step": 27390 }, { "grad_norm": 0.1815246343612671, "learning_rate": 8.689909646471243e-05, "loss": 0.0193, "step": 27400 }, { "grad_norm": 0.24633018672466278, "learning_rate": 8.688793649444099e-05, "loss": 0.0217, "step": 27410 }, { "grad_norm": 0.19924557209014893, "learning_rate": 8.687677249016029e-05, "loss": 0.0219, "step": 27420 }, { "grad_norm": 0.23720374703407288, "learning_rate": 8.686560445309118e-05, "loss": 0.0246, "step": 27430 }, { "grad_norm": 0.23325955867767334, "learning_rate": 8.685443238445499e-05, "loss": 0.02, "step": 27440 }, { "grad_norm": 0.23502829670906067, "learning_rate": 8.68432562854735e-05, "loss": 0.0198, "step": 27450 }, { "grad_norm": 0.22071415185928345, "learning_rate": 8.683207615736887e-05, "loss": 0.0236, "step": 27460 }, { "grad_norm": 0.3471500873565674, "learning_rate": 8.682089200136379e-05, "loss": 0.0265, "step": 27470 }, { "grad_norm": 0.20485131442546844, "learning_rate": 8.680970381868132e-05, "loss": 0.0207, "step": 27480 }, { "grad_norm": 0.31049349904060364, "learning_rate": 8.679851161054498e-05, "loss": 0.0209, "step": 27490 }, { "grad_norm": 0.19037774205207825, "learning_rate": 8.678731537817873e-05, "loss": 0.0185, "step": 27500 }, { "grad_norm": 0.18920926749706268, "learning_rate": 8.677611512280697e-05, "loss": 0.0182, "step": 27510 }, { "grad_norm": 0.18821993470191956, "learning_rate": 8.676491084565457e-05, "loss": 0.0177, "step": 27520 }, { "grad_norm": 0.2050284892320633, "learning_rate": 8.675370254794678e-05, "loss": 0.0167, "step": 27530 }, { "grad_norm": 0.16578879952430725, "learning_rate": 8.674249023090935e-05, "loss": 0.0211, "step": 27540 }, { "grad_norm": 0.21381445229053497, "learning_rate": 8.673127389576843e-05, "loss": 0.0227, "step": 27550 }, { "grad_norm": 0.18150685727596283, "learning_rate": 8.67200535437506e-05, "loss": 0.0208, "step": 27560 }, { "grad_norm": 0.18975351750850677, "learning_rate": 8.670882917608296e-05, "loss": 0.017, "step": 27570 }, { "grad_norm": 0.2178809493780136, "learning_rate": 8.669760079399292e-05, "loss": 0.02, "step": 27580 }, { "grad_norm": 0.17840376496315002, "learning_rate": 8.668636839870845e-05, "loss": 0.0188, "step": 27590 }, { "grad_norm": 0.20227886736392975, "learning_rate": 8.667513199145789e-05, "loss": 0.0222, "step": 27600 }, { "grad_norm": 0.22758500277996063, "learning_rate": 8.666389157347002e-05, "loss": 0.0186, "step": 27610 }, { "grad_norm": 0.22579003870487213, "learning_rate": 8.66526471459741e-05, "loss": 0.0203, "step": 27620 }, { "grad_norm": 0.26849329471588135, "learning_rate": 8.66413987101998e-05, "loss": 0.0193, "step": 27630 }, { "grad_norm": 0.20895837247371674, "learning_rate": 8.663014626737723e-05, "loss": 0.0219, "step": 27640 }, { "grad_norm": 0.18939976394176483, "learning_rate": 8.661888981873691e-05, "loss": 0.0185, "step": 27650 }, { "grad_norm": 0.19926218688488007, "learning_rate": 8.660762936550988e-05, "loss": 0.0176, "step": 27660 }, { "grad_norm": 0.24523933231830597, "learning_rate": 8.659636490892753e-05, "loss": 0.0192, "step": 27670 }, { "grad_norm": 0.2181890904903412, "learning_rate": 8.658509645022174e-05, "loss": 0.0205, "step": 27680 }, { "grad_norm": 0.22372551262378693, "learning_rate": 8.657382399062481e-05, "loss": 0.0201, "step": 27690 }, { "grad_norm": 0.23207108676433563, "learning_rate": 8.656254753136946e-05, "loss": 0.0224, "step": 27700 }, { "grad_norm": 0.2315693348646164, "learning_rate": 8.655126707368891e-05, "loss": 0.0208, "step": 27710 }, { "grad_norm": 0.16510158777236938, "learning_rate": 8.653998261881672e-05, "loss": 0.0211, "step": 27720 }, { "grad_norm": 0.20222735404968262, "learning_rate": 8.652869416798699e-05, "loss": 0.0239, "step": 27730 }, { "grad_norm": 0.16274823248386383, "learning_rate": 8.651740172243417e-05, "loss": 0.0163, "step": 27740 }, { "grad_norm": 0.22328457236289978, "learning_rate": 8.65061052833932e-05, "loss": 0.0211, "step": 27750 }, { "grad_norm": 0.1737731695175171, "learning_rate": 8.649480485209945e-05, "loss": 0.0197, "step": 27760 }, { "grad_norm": 0.18380357325077057, "learning_rate": 8.64835004297887e-05, "loss": 0.0174, "step": 27770 }, { "grad_norm": 0.20311400294303894, "learning_rate": 8.64721920176972e-05, "loss": 0.0195, "step": 27780 }, { "grad_norm": 0.20356422662734985, "learning_rate": 8.646087961706164e-05, "loss": 0.0209, "step": 27790 }, { "grad_norm": 0.1874639391899109, "learning_rate": 8.644956322911908e-05, "loss": 0.0186, "step": 27800 }, { "grad_norm": 0.19830290973186493, "learning_rate": 8.643824285510709e-05, "loss": 0.0154, "step": 27810 }, { "grad_norm": 0.20033445954322815, "learning_rate": 8.642691849626364e-05, "loss": 0.0153, "step": 27820 }, { "grad_norm": 0.17347848415374756, "learning_rate": 8.641559015382717e-05, "loss": 0.0152, "step": 27830 }, { "grad_norm": 0.1953282505273819, "learning_rate": 8.640425782903649e-05, "loss": 0.0194, "step": 27840 }, { "grad_norm": 0.20779691636562347, "learning_rate": 8.639292152313091e-05, "loss": 0.0174, "step": 27850 }, { "grad_norm": 0.18316985666751862, "learning_rate": 8.638158123735015e-05, "loss": 0.0233, "step": 27860 }, { "grad_norm": 0.22978943586349487, "learning_rate": 8.637023697293436e-05, "loss": 0.019, "step": 27870 }, { "grad_norm": 0.18324655294418335, "learning_rate": 8.635888873112414e-05, "loss": 0.0203, "step": 27880 }, { "grad_norm": 0.2541497051715851, "learning_rate": 8.634753651316052e-05, "loss": 0.0196, "step": 27890 }, { "grad_norm": 0.20340004563331604, "learning_rate": 8.633618032028496e-05, "loss": 0.0222, "step": 27900 }, { "grad_norm": 0.21960438787937164, "learning_rate": 8.632482015373934e-05, "loss": 0.0173, "step": 27910 }, { "grad_norm": 0.17891952395439148, "learning_rate": 8.6313456014766e-05, "loss": 0.0187, "step": 27920 }, { "grad_norm": 0.28473809361457825, "learning_rate": 8.630208790460771e-05, "loss": 0.0173, "step": 27930 }, { "grad_norm": 0.20408961176872253, "learning_rate": 8.629071582450768e-05, "loss": 0.0205, "step": 27940 }, { "grad_norm": 0.22693878412246704, "learning_rate": 8.62793397757095e-05, "loss": 0.0213, "step": 27950 }, { "grad_norm": 0.17838479578495026, "learning_rate": 8.626795975945729e-05, "loss": 0.0237, "step": 27960 }, { "grad_norm": 0.24291691184043884, "learning_rate": 8.625657577699551e-05, "loss": 0.0228, "step": 27970 }, { "grad_norm": 0.21661414206027985, "learning_rate": 8.624518782956914e-05, "loss": 0.0202, "step": 27980 }, { "grad_norm": 0.22810208797454834, "learning_rate": 8.62337959184235e-05, "loss": 0.0193, "step": 27990 }, { "grad_norm": 0.30135077238082886, "learning_rate": 8.622240004480441e-05, "loss": 0.0277, "step": 28000 }, { "grad_norm": 0.17371240258216858, "learning_rate": 8.621100020995814e-05, "loss": 0.0172, "step": 28010 }, { "grad_norm": 0.22907425463199615, "learning_rate": 8.619959641513132e-05, "loss": 0.0201, "step": 28020 }, { "grad_norm": 0.184353306889534, "learning_rate": 8.618818866157105e-05, "loss": 0.02, "step": 28030 }, { "grad_norm": 0.13857470452785492, "learning_rate": 8.617677695052487e-05, "loss": 0.0169, "step": 28040 }, { "grad_norm": 0.21612246334552765, "learning_rate": 8.616536128324078e-05, "loss": 0.0159, "step": 28050 }, { "grad_norm": 0.18421050906181335, "learning_rate": 8.615394166096712e-05, "loss": 0.0195, "step": 28060 }, { "grad_norm": 0.16495798528194427, "learning_rate": 8.614251808495279e-05, "loss": 0.0194, "step": 28070 }, { "grad_norm": 0.19451892375946045, "learning_rate": 8.6131090556447e-05, "loss": 0.0198, "step": 28080 }, { "grad_norm": 0.17197807133197784, "learning_rate": 8.611965907669947e-05, "loss": 0.0168, "step": 28090 }, { "grad_norm": 0.25055360794067383, "learning_rate": 8.610822364696034e-05, "loss": 0.0264, "step": 28100 }, { "grad_norm": 0.23205234110355377, "learning_rate": 8.609678426848015e-05, "loss": 0.0239, "step": 28110 }, { "grad_norm": 0.2857268154621124, "learning_rate": 8.60853409425099e-05, "loss": 0.0196, "step": 28120 }, { "grad_norm": 0.17833709716796875, "learning_rate": 8.607389367030104e-05, "loss": 0.0219, "step": 28130 }, { "grad_norm": 0.22342263162136078, "learning_rate": 8.606244245310538e-05, "loss": 0.0215, "step": 28140 }, { "grad_norm": 0.2729599177837372, "learning_rate": 8.605098729217525e-05, "loss": 0.0181, "step": 28150 }, { "grad_norm": 0.1745533049106598, "learning_rate": 8.603952818876335e-05, "loss": 0.0184, "step": 28160 }, { "grad_norm": 0.22840049862861633, "learning_rate": 8.602806514412281e-05, "loss": 0.0194, "step": 28170 }, { "grad_norm": 0.19720225036144257, "learning_rate": 8.601659815950726e-05, "loss": 0.0191, "step": 28180 }, { "grad_norm": 0.2473469227552414, "learning_rate": 8.600512723617067e-05, "loss": 0.0182, "step": 28190 }, { "grad_norm": 0.20981281995773315, "learning_rate": 8.59936523753675e-05, "loss": 0.0214, "step": 28200 }, { "grad_norm": 0.19054120779037476, "learning_rate": 8.598217357835264e-05, "loss": 0.0175, "step": 28210 }, { "grad_norm": 0.1955602467060089, "learning_rate": 8.597069084638135e-05, "loss": 0.0182, "step": 28220 }, { "grad_norm": 0.1497371345758438, "learning_rate": 8.595920418070939e-05, "loss": 0.0204, "step": 28230 }, { "grad_norm": 0.22335679829120636, "learning_rate": 8.594771358259295e-05, "loss": 0.0177, "step": 28240 }, { "grad_norm": 0.14514051377773285, "learning_rate": 8.593621905328858e-05, "loss": 0.0161, "step": 28250 }, { "grad_norm": 0.2354125678539276, "learning_rate": 8.592472059405333e-05, "loss": 0.0226, "step": 28260 }, { "grad_norm": 0.14934170246124268, "learning_rate": 8.591321820614464e-05, "loss": 0.016, "step": 28270 }, { "grad_norm": 0.14572182297706604, "learning_rate": 8.590171189082041e-05, "loss": 0.0184, "step": 28280 }, { "grad_norm": 0.2199043482542038, "learning_rate": 8.589020164933894e-05, "loss": 0.0183, "step": 28290 }, { "grad_norm": 0.23906028270721436, "learning_rate": 8.587868748295898e-05, "loss": 0.0201, "step": 28300 }, { "grad_norm": 0.2526264488697052, "learning_rate": 8.586716939293971e-05, "loss": 0.0238, "step": 28310 }, { "grad_norm": 0.21161095798015594, "learning_rate": 8.58556473805407e-05, "loss": 0.0183, "step": 28320 }, { "grad_norm": 0.18465229868888855, "learning_rate": 8.584412144702202e-05, "loss": 0.0195, "step": 28330 }, { "grad_norm": 0.20446999371051788, "learning_rate": 8.58325915936441e-05, "loss": 0.018, "step": 28340 }, { "grad_norm": 0.20450963079929352, "learning_rate": 8.582105782166783e-05, "loss": 0.0216, "step": 28350 }, { "grad_norm": 0.22855868935585022, "learning_rate": 8.580952013235455e-05, "loss": 0.022, "step": 28360 }, { "grad_norm": 0.20395345985889435, "learning_rate": 8.579797852696596e-05, "loss": 0.0193, "step": 28370 }, { "grad_norm": 0.1804451197385788, "learning_rate": 8.578643300676428e-05, "loss": 0.017, "step": 28380 }, { "grad_norm": 0.1519966423511505, "learning_rate": 8.577488357301209e-05, "loss": 0.0215, "step": 28390 }, { "grad_norm": 0.17901711165905, "learning_rate": 8.576333022697242e-05, "loss": 0.0223, "step": 28400 }, { "grad_norm": 0.18736043572425842, "learning_rate": 8.575177296990873e-05, "loss": 0.0137, "step": 28410 }, { "grad_norm": 0.24992401897907257, "learning_rate": 8.574021180308489e-05, "loss": 0.0157, "step": 28420 }, { "grad_norm": 0.20229469239711761, "learning_rate": 8.572864672776523e-05, "loss": 0.0179, "step": 28430 }, { "grad_norm": 0.19251646101474762, "learning_rate": 8.571707774521447e-05, "loss": 0.0166, "step": 28440 }, { "grad_norm": 0.17733867466449738, "learning_rate": 8.57055048566978e-05, "loss": 0.019, "step": 28450 }, { "grad_norm": 0.18053342401981354, "learning_rate": 8.569392806348078e-05, "loss": 0.0184, "step": 28460 }, { "grad_norm": 0.1893172264099121, "learning_rate": 8.568234736682947e-05, "loss": 0.0194, "step": 28470 }, { "grad_norm": 0.20723234117031097, "learning_rate": 8.567076276801029e-05, "loss": 0.0235, "step": 28480 }, { "grad_norm": 0.22462084889411926, "learning_rate": 8.565917426829013e-05, "loss": 0.0221, "step": 28490 }, { "grad_norm": 0.27399399876594543, "learning_rate": 8.564758186893628e-05, "loss": 0.0188, "step": 28500 }, { "grad_norm": 0.2224675565958023, "learning_rate": 8.563598557121649e-05, "loss": 0.0199, "step": 28510 }, { "grad_norm": 0.20978373289108276, "learning_rate": 8.562438537639888e-05, "loss": 0.0181, "step": 28520 }, { "grad_norm": 0.24369725584983826, "learning_rate": 8.561278128575206e-05, "loss": 0.0213, "step": 28530 }, { "grad_norm": 0.2931808829307556, "learning_rate": 8.5601173300545e-05, "loss": 0.0182, "step": 28540 }, { "grad_norm": 0.27723953127861023, "learning_rate": 8.558956142204717e-05, "loss": 0.0223, "step": 28550 }, { "grad_norm": 0.22410240769386292, "learning_rate": 8.55779456515284e-05, "loss": 0.0202, "step": 28560 }, { "grad_norm": 0.25470343232154846, "learning_rate": 8.556632599025898e-05, "loss": 0.0213, "step": 28570 }, { "grad_norm": 0.1964382380247116, "learning_rate": 8.555470243950964e-05, "loss": 0.019, "step": 28580 }, { "grad_norm": 0.2081485390663147, "learning_rate": 8.554307500055148e-05, "loss": 0.0217, "step": 28590 }, { "grad_norm": 0.19886445999145508, "learning_rate": 8.553144367465609e-05, "loss": 0.0251, "step": 28600 }, { "grad_norm": 0.18649624288082123, "learning_rate": 8.551980846309544e-05, "loss": 0.0183, "step": 28610 }, { "grad_norm": 0.1811775267124176, "learning_rate": 8.550816936714193e-05, "loss": 0.0201, "step": 28620 }, { "grad_norm": 0.229015052318573, "learning_rate": 8.549652638806841e-05, "loss": 0.0194, "step": 28630 }, { "grad_norm": 0.2585911452770233, "learning_rate": 8.548487952714812e-05, "loss": 0.0202, "step": 28640 }, { "grad_norm": 0.17194616794586182, "learning_rate": 8.547322878565478e-05, "loss": 0.0185, "step": 28650 }, { "grad_norm": 0.2510678470134735, "learning_rate": 8.546157416486245e-05, "loss": 0.022, "step": 28660 }, { "grad_norm": 0.1778154969215393, "learning_rate": 8.54499156660457e-05, "loss": 0.0184, "step": 28670 }, { "grad_norm": 0.15497499704360962, "learning_rate": 8.543825329047947e-05, "loss": 0.02, "step": 28680 }, { "grad_norm": 0.19809910655021667, "learning_rate": 8.542658703943913e-05, "loss": 0.0206, "step": 28690 }, { "grad_norm": 0.2604942321777344, "learning_rate": 8.541491691420051e-05, "loss": 0.0169, "step": 28700 }, { "grad_norm": 0.1687331646680832, "learning_rate": 8.54032429160398e-05, "loss": 0.0184, "step": 28710 }, { "grad_norm": 0.1606275588274002, "learning_rate": 8.539156504623369e-05, "loss": 0.0209, "step": 28720 }, { "grad_norm": 0.18782103061676025, "learning_rate": 8.537988330605923e-05, "loss": 0.0151, "step": 28730 }, { "grad_norm": 0.21968631446361542, "learning_rate": 8.536819769679393e-05, "loss": 0.0163, "step": 28740 }, { "grad_norm": 0.21829798817634583, "learning_rate": 8.53565082197157e-05, "loss": 0.0178, "step": 28750 }, { "grad_norm": 0.19031602144241333, "learning_rate": 8.534481487610289e-05, "loss": 0.017, "step": 28760 }, { "grad_norm": 0.17401942610740662, "learning_rate": 8.533311766723428e-05, "loss": 0.0203, "step": 28770 }, { "grad_norm": 0.21516364812850952, "learning_rate": 8.532141659438901e-05, "loss": 0.0225, "step": 28780 }, { "grad_norm": 0.1714668571949005, "learning_rate": 8.530971165884675e-05, "loss": 0.0211, "step": 28790 }, { "grad_norm": 0.17711412906646729, "learning_rate": 8.529800286188752e-05, "loss": 0.0176, "step": 28800 }, { "grad_norm": 0.22569556534290314, "learning_rate": 8.528629020479175e-05, "loss": 0.0215, "step": 28810 }, { "grad_norm": 0.22495664656162262, "learning_rate": 8.527457368884033e-05, "loss": 0.0219, "step": 28820 }, { "grad_norm": 0.1296713501214981, "learning_rate": 8.526285331531458e-05, "loss": 0.0201, "step": 28830 }, { "grad_norm": 0.18433421850204468, "learning_rate": 8.525112908549621e-05, "loss": 0.0167, "step": 28840 }, { "grad_norm": 0.2662346065044403, "learning_rate": 8.523940100066735e-05, "loss": 0.0189, "step": 28850 }, { "grad_norm": 0.14710089564323425, "learning_rate": 8.52276690621106e-05, "loss": 0.0183, "step": 28860 }, { "grad_norm": 0.1624893844127655, "learning_rate": 8.521593327110889e-05, "loss": 0.0188, "step": 28870 }, { "grad_norm": 0.15680401027202606, "learning_rate": 8.520419362894569e-05, "loss": 0.0171, "step": 28880 }, { "grad_norm": 0.21978116035461426, "learning_rate": 8.51924501369048e-05, "loss": 0.0166, "step": 28890 }, { "grad_norm": 0.22576244175434113, "learning_rate": 8.518070279627047e-05, "loss": 0.022, "step": 28900 }, { "grad_norm": 0.18659751117229462, "learning_rate": 8.516895160832737e-05, "loss": 0.0206, "step": 28910 }, { "grad_norm": 0.18289263546466827, "learning_rate": 8.515719657436061e-05, "loss": 0.0188, "step": 28920 }, { "grad_norm": 0.20036040246486664, "learning_rate": 8.514543769565568e-05, "loss": 0.0226, "step": 28930 }, { "grad_norm": 0.15454380214214325, "learning_rate": 8.513367497349853e-05, "loss": 0.0205, "step": 28940 }, { "grad_norm": 0.1579643338918686, "learning_rate": 8.51219084091755e-05, "loss": 0.0187, "step": 28950 }, { "grad_norm": 0.23684155941009521, "learning_rate": 8.511013800397338e-05, "loss": 0.0182, "step": 28960 }, { "grad_norm": 0.20078274607658386, "learning_rate": 8.509836375917937e-05, "loss": 0.0184, "step": 28970 }, { "grad_norm": 0.14252878725528717, "learning_rate": 8.508658567608104e-05, "loss": 0.0194, "step": 28980 }, { "grad_norm": 0.16473133862018585, "learning_rate": 8.507480375596647e-05, "loss": 0.0142, "step": 28990 }, { "grad_norm": 0.16153962910175323, "learning_rate": 8.506301800012408e-05, "loss": 0.0181, "step": 29000 }, { "grad_norm": 0.17077036201953888, "learning_rate": 8.505122840984278e-05, "loss": 0.0223, "step": 29010 }, { "grad_norm": 0.18465709686279297, "learning_rate": 8.503943498641182e-05, "loss": 0.0202, "step": 29020 }, { "grad_norm": 0.1747397631406784, "learning_rate": 8.502763773112095e-05, "loss": 0.0156, "step": 29030 }, { "grad_norm": 0.17188474535942078, "learning_rate": 8.501583664526026e-05, "loss": 0.0188, "step": 29040 }, { "grad_norm": 0.18848681449890137, "learning_rate": 8.500403173012032e-05, "loss": 0.0184, "step": 29050 }, { "grad_norm": 0.23803967237472534, "learning_rate": 8.499222298699211e-05, "loss": 0.0157, "step": 29060 }, { "grad_norm": 0.2101498544216156, "learning_rate": 8.498041041716701e-05, "loss": 0.0185, "step": 29070 }, { "grad_norm": 0.1869553029537201, "learning_rate": 8.496859402193681e-05, "loss": 0.0229, "step": 29080 }, { "grad_norm": 0.13847753405570984, "learning_rate": 8.495677380259374e-05, "loss": 0.0164, "step": 29090 }, { "grad_norm": 0.17952251434326172, "learning_rate": 8.494494976043045e-05, "loss": 0.0183, "step": 29100 }, { "grad_norm": 0.19857323169708252, "learning_rate": 8.493312189673998e-05, "loss": 0.0191, "step": 29110 }, { "grad_norm": 0.21433371305465698, "learning_rate": 8.492129021281584e-05, "loss": 0.0185, "step": 29120 }, { "grad_norm": 0.19495628774166107, "learning_rate": 8.490945470995188e-05, "loss": 0.0196, "step": 29130 }, { "grad_norm": 0.19970965385437012, "learning_rate": 8.489761538944247e-05, "loss": 0.0176, "step": 29140 }, { "grad_norm": 0.22742195427417755, "learning_rate": 8.48857722525823e-05, "loss": 0.0182, "step": 29150 }, { "grad_norm": 0.28606829047203064, "learning_rate": 8.487392530066652e-05, "loss": 0.0196, "step": 29160 }, { "grad_norm": 0.24411457777023315, "learning_rate": 8.486207453499069e-05, "loss": 0.0205, "step": 29170 }, { "grad_norm": 0.2243444323539734, "learning_rate": 8.485021995685082e-05, "loss": 0.0208, "step": 29180 }, { "grad_norm": 0.16897886991500854, "learning_rate": 8.483836156754328e-05, "loss": 0.0173, "step": 29190 }, { "grad_norm": 0.2070247083902359, "learning_rate": 8.482649936836491e-05, "loss": 0.0188, "step": 29200 }, { "grad_norm": 0.25397348403930664, "learning_rate": 8.481463336061293e-05, "loss": 0.02, "step": 29210 }, { "grad_norm": 0.17573972046375275, "learning_rate": 8.480276354558496e-05, "loss": 0.0167, "step": 29220 }, { "grad_norm": 0.22346407175064087, "learning_rate": 8.479088992457913e-05, "loss": 0.0189, "step": 29230 }, { "grad_norm": 0.2858067452907562, "learning_rate": 8.477901249889387e-05, "loss": 0.0229, "step": 29240 }, { "grad_norm": 0.2654416561126709, "learning_rate": 8.47671312698281e-05, "loss": 0.0203, "step": 29250 }, { "grad_norm": 0.204691082239151, "learning_rate": 8.475524623868112e-05, "loss": 0.0192, "step": 29260 }, { "grad_norm": 0.16117887198925018, "learning_rate": 8.474335740675266e-05, "loss": 0.0171, "step": 29270 }, { "grad_norm": 0.1381584256887436, "learning_rate": 8.473146477534289e-05, "loss": 0.0191, "step": 29280 }, { "grad_norm": 0.1832672506570816, "learning_rate": 8.471956834575232e-05, "loss": 0.0147, "step": 29290 }, { "grad_norm": 0.24494045972824097, "learning_rate": 8.470766811928197e-05, "loss": 0.0195, "step": 29300 }, { "grad_norm": 0.23444311320781708, "learning_rate": 8.469576409723323e-05, "loss": 0.0272, "step": 29310 }, { "grad_norm": 0.2067396491765976, "learning_rate": 8.468385628090788e-05, "loss": 0.0191, "step": 29320 }, { "grad_norm": 0.20963355898857117, "learning_rate": 8.467194467160815e-05, "loss": 0.0191, "step": 29330 }, { "grad_norm": 0.15888231992721558, "learning_rate": 8.466002927063667e-05, "loss": 0.0184, "step": 29340 }, { "grad_norm": 0.18385837972164154, "learning_rate": 8.464811007929651e-05, "loss": 0.0181, "step": 29350 }, { "grad_norm": 0.221443772315979, "learning_rate": 8.463618709889114e-05, "loss": 0.0196, "step": 29360 }, { "grad_norm": 0.17397646605968475, "learning_rate": 8.462426033072442e-05, "loss": 0.0191, "step": 29370 }, { "grad_norm": 0.16413463652133942, "learning_rate": 8.461232977610061e-05, "loss": 0.0186, "step": 29380 }, { "grad_norm": 0.1881738007068634, "learning_rate": 8.46003954363245e-05, "loss": 0.0188, "step": 29390 }, { "grad_norm": 0.2384498566389084, "learning_rate": 8.458845731270115e-05, "loss": 0.0182, "step": 29400 }, { "grad_norm": 0.16183322668075562, "learning_rate": 8.45765154065361e-05, "loss": 0.0174, "step": 29410 }, { "grad_norm": 0.13167381286621094, "learning_rate": 8.456456971913532e-05, "loss": 0.0137, "step": 29420 }, { "grad_norm": 0.21183840930461884, "learning_rate": 8.455262025180517e-05, "loss": 0.0174, "step": 29430 }, { "grad_norm": 0.15256284177303314, "learning_rate": 8.454066700585242e-05, "loss": 0.0183, "step": 29440 }, { "grad_norm": 0.17907369136810303, "learning_rate": 8.452870998258423e-05, "loss": 0.0187, "step": 29450 }, { "grad_norm": 0.212639719247818, "learning_rate": 8.451674918330825e-05, "loss": 0.0181, "step": 29460 }, { "grad_norm": 0.1764448881149292, "learning_rate": 8.450478460933246e-05, "loss": 0.0196, "step": 29470 }, { "grad_norm": 0.17995816469192505, "learning_rate": 8.449281626196532e-05, "loss": 0.025, "step": 29480 }, { "grad_norm": 0.2720528244972229, "learning_rate": 8.448084414251564e-05, "loss": 0.0184, "step": 29490 }, { "grad_norm": 0.1848011016845703, "learning_rate": 8.446886825229271e-05, "loss": 0.0192, "step": 29500 }, { "grad_norm": 0.1858597993850708, "learning_rate": 8.445688859260615e-05, "loss": 0.0169, "step": 29510 }, { "grad_norm": 0.21261587738990784, "learning_rate": 8.444490516476606e-05, "loss": 0.0192, "step": 29520 }, { "grad_norm": 0.23082730174064636, "learning_rate": 8.443291797008293e-05, "loss": 0.0183, "step": 29530 }, { "grad_norm": 0.2009122222661972, "learning_rate": 8.442092700986765e-05, "loss": 0.0181, "step": 29540 }, { "grad_norm": 0.2151041328907013, "learning_rate": 8.440893228543156e-05, "loss": 0.0187, "step": 29550 }, { "grad_norm": 0.21912506222724915, "learning_rate": 8.439693379808638e-05, "loss": 0.0178, "step": 29560 }, { "grad_norm": 0.22011037170886993, "learning_rate": 8.43849315491442e-05, "loss": 0.0192, "step": 29570 }, { "grad_norm": 0.21793368458747864, "learning_rate": 8.437292553991763e-05, "loss": 0.0209, "step": 29580 }, { "grad_norm": 0.1976400464773178, "learning_rate": 8.436091577171959e-05, "loss": 0.0151, "step": 29590 }, { "grad_norm": 0.15822672843933105, "learning_rate": 8.434890224586347e-05, "loss": 0.0201, "step": 29600 }, { "grad_norm": 0.1799924522638321, "learning_rate": 8.433688496366303e-05, "loss": 0.016, "step": 29610 }, { "grad_norm": 0.22316798567771912, "learning_rate": 8.432486392643248e-05, "loss": 0.0187, "step": 29620 }, { "grad_norm": 0.20864231884479523, "learning_rate": 8.431283913548643e-05, "loss": 0.0177, "step": 29630 }, { "grad_norm": 0.17704413831233978, "learning_rate": 8.430081059213985e-05, "loss": 0.02, "step": 29640 }, { "grad_norm": 0.24074873328208923, "learning_rate": 8.428877829770823e-05, "loss": 0.0201, "step": 29650 }, { "grad_norm": 0.2107817530632019, "learning_rate": 8.427674225350735e-05, "loss": 0.019, "step": 29660 }, { "grad_norm": 0.23028355836868286, "learning_rate": 8.426470246085347e-05, "loss": 0.0208, "step": 29670 }, { "grad_norm": 0.1235429048538208, "learning_rate": 8.425265892106324e-05, "loss": 0.0202, "step": 29680 }, { "grad_norm": 0.2363317608833313, "learning_rate": 8.424061163545374e-05, "loss": 0.0195, "step": 29690 }, { "grad_norm": 0.2037597894668579, "learning_rate": 8.422856060534243e-05, "loss": 0.0193, "step": 29700 }, { "grad_norm": 0.18824294209480286, "learning_rate": 8.421650583204718e-05, "loss": 0.0144, "step": 29710 }, { "grad_norm": 0.18654051423072815, "learning_rate": 8.420444731688633e-05, "loss": 0.0168, "step": 29720 }, { "grad_norm": 0.1480540782213211, "learning_rate": 8.419238506117852e-05, "loss": 0.0168, "step": 29730 }, { "grad_norm": 0.14388735592365265, "learning_rate": 8.418031906624289e-05, "loss": 0.0185, "step": 29740 }, { "grad_norm": 0.22635872662067413, "learning_rate": 8.416824933339898e-05, "loss": 0.0179, "step": 29750 }, { "grad_norm": 0.21827122569084167, "learning_rate": 8.415617586396667e-05, "loss": 0.0215, "step": 29760 }, { "grad_norm": 0.225932776927948, "learning_rate": 8.414409865926632e-05, "loss": 0.0195, "step": 29770 }, { "grad_norm": 0.1929963231086731, "learning_rate": 8.413201772061867e-05, "loss": 0.0201, "step": 29780 }, { "grad_norm": 0.24764302372932434, "learning_rate": 8.411993304934488e-05, "loss": 0.0194, "step": 29790 }, { "grad_norm": 0.24923168122768402, "learning_rate": 8.410784464676654e-05, "loss": 0.019, "step": 29800 }, { "grad_norm": 0.2315533459186554, "learning_rate": 8.409575251420556e-05, "loss": 0.0178, "step": 29810 }, { "grad_norm": 0.13274821639060974, "learning_rate": 8.408365665298435e-05, "loss": 0.0188, "step": 29820 }, { "grad_norm": 0.19396939873695374, "learning_rate": 8.40715570644257e-05, "loss": 0.0229, "step": 29830 }, { "grad_norm": 0.21364526450634003, "learning_rate": 8.40594537498528e-05, "loss": 0.0191, "step": 29840 }, { "grad_norm": 0.1335286647081375, "learning_rate": 8.404734671058924e-05, "loss": 0.0163, "step": 29850 }, { "grad_norm": 0.17746222019195557, "learning_rate": 8.403523594795902e-05, "loss": 0.0238, "step": 29860 }, { "grad_norm": 0.20539331436157227, "learning_rate": 8.402312146328659e-05, "loss": 0.0183, "step": 29870 }, { "grad_norm": 0.2510819733142853, "learning_rate": 8.401100325789675e-05, "loss": 0.0235, "step": 29880 }, { "grad_norm": 0.19503186643123627, "learning_rate": 8.399888133311472e-05, "loss": 0.019, "step": 29890 }, { "grad_norm": 0.20758289098739624, "learning_rate": 8.398675569026613e-05, "loss": 0.0162, "step": 29900 }, { "grad_norm": 0.22533707320690155, "learning_rate": 8.397462633067705e-05, "loss": 0.0188, "step": 29910 }, { "grad_norm": 0.19239431619644165, "learning_rate": 8.396249325567392e-05, "loss": 0.0144, "step": 29920 }, { "grad_norm": 0.30769041180610657, "learning_rate": 8.395035646658357e-05, "loss": 0.0244, "step": 29930 }, { "grad_norm": 0.18618974089622498, "learning_rate": 8.39382159647333e-05, "loss": 0.0216, "step": 29940 }, { "grad_norm": 0.23242029547691345, "learning_rate": 8.392607175145075e-05, "loss": 0.0193, "step": 29950 }, { "grad_norm": 0.20553037524223328, "learning_rate": 8.3913923828064e-05, "loss": 0.0178, "step": 29960 }, { "grad_norm": 0.18487386405467987, "learning_rate": 8.390177219590152e-05, "loss": 0.0183, "step": 29970 }, { "grad_norm": 0.20678547024726868, "learning_rate": 8.388961685629222e-05, "loss": 0.0173, "step": 29980 }, { "grad_norm": 0.23490162193775177, "learning_rate": 8.387745781056536e-05, "loss": 0.0248, "step": 29990 }, { "grad_norm": 0.21364833414554596, "learning_rate": 8.386529506005065e-05, "loss": 0.018, "step": 30000 }, { "grad_norm": 0.18126659095287323, "learning_rate": 8.38531286060782e-05, "loss": 0.0177, "step": 30010 }, { "grad_norm": 0.19605115056037903, "learning_rate": 8.384095844997849e-05, "loss": 0.0209, "step": 30020 }, { "grad_norm": 0.20939135551452637, "learning_rate": 8.382878459308245e-05, "loss": 0.0198, "step": 30030 }, { "grad_norm": 0.21955598890781403, "learning_rate": 8.381660703672138e-05, "loss": 0.0189, "step": 30040 }, { "grad_norm": 0.2517257630825043, "learning_rate": 8.380442578222702e-05, "loss": 0.0162, "step": 30050 }, { "grad_norm": 0.18426035344600677, "learning_rate": 8.379224083093146e-05, "loss": 0.0175, "step": 30060 }, { "grad_norm": 0.20536965131759644, "learning_rate": 8.378005218416727e-05, "loss": 0.0157, "step": 30070 }, { "grad_norm": 0.16685687005519867, "learning_rate": 8.376785984326735e-05, "loss": 0.0178, "step": 30080 }, { "grad_norm": 0.17038202285766602, "learning_rate": 8.375566380956506e-05, "loss": 0.0161, "step": 30090 }, { "grad_norm": 0.1679488867521286, "learning_rate": 8.374346408439411e-05, "loss": 0.0177, "step": 30100 }, { "grad_norm": 0.2220415323972702, "learning_rate": 8.373126066908868e-05, "loss": 0.0155, "step": 30110 }, { "grad_norm": 0.17567074298858643, "learning_rate": 8.371905356498326e-05, "loss": 0.0154, "step": 30120 }, { "grad_norm": 0.18284018337726593, "learning_rate": 8.370684277341288e-05, "loss": 0.0173, "step": 30130 }, { "grad_norm": 0.17312316596508026, "learning_rate": 8.369462829571282e-05, "loss": 0.0142, "step": 30140 }, { "grad_norm": 0.2272639274597168, "learning_rate": 8.36824101332189e-05, "loss": 0.018, "step": 30150 }, { "grad_norm": 0.20326854288578033, "learning_rate": 8.367018828726721e-05, "loss": 0.0172, "step": 30160 }, { "grad_norm": 0.14004464447498322, "learning_rate": 8.365796275919438e-05, "loss": 0.0161, "step": 30170 }, { "grad_norm": 0.21613731980323792, "learning_rate": 8.364573355033734e-05, "loss": 0.0177, "step": 30180 }, { "grad_norm": 0.2178744673728943, "learning_rate": 8.363350066203346e-05, "loss": 0.0165, "step": 30190 }, { "grad_norm": 0.20226092636585236, "learning_rate": 8.362126409562053e-05, "loss": 0.0175, "step": 30200 }, { "grad_norm": 0.22606170177459717, "learning_rate": 8.360902385243667e-05, "loss": 0.0149, "step": 30210 }, { "grad_norm": 0.17122001945972443, "learning_rate": 8.359677993382052e-05, "loss": 0.0203, "step": 30220 }, { "grad_norm": 0.19086132943630219, "learning_rate": 8.358453234111103e-05, "loss": 0.0139, "step": 30230 }, { "grad_norm": 0.17047831416130066, "learning_rate": 8.357228107564756e-05, "loss": 0.0205, "step": 30240 }, { "grad_norm": 0.19039016962051392, "learning_rate": 8.356002613876993e-05, "loss": 0.0198, "step": 30250 }, { "grad_norm": 0.17929644882678986, "learning_rate": 8.35477675318183e-05, "loss": 0.018, "step": 30260 }, { "grad_norm": 0.19011980295181274, "learning_rate": 8.353550525613323e-05, "loss": 0.023, "step": 30270 }, { "grad_norm": 0.1707257181406021, "learning_rate": 8.352323931305572e-05, "loss": 0.0177, "step": 30280 }, { "grad_norm": 0.17595340311527252, "learning_rate": 8.351096970392717e-05, "loss": 0.0199, "step": 30290 }, { "grad_norm": 0.1886214017868042, "learning_rate": 8.349869643008937e-05, "loss": 0.0207, "step": 30300 }, { "grad_norm": 0.16537973284721375, "learning_rate": 8.348641949288449e-05, "loss": 0.0192, "step": 30310 }, { "grad_norm": 0.2609269320964813, "learning_rate": 8.34741388936551e-05, "loss": 0.0211, "step": 30320 }, { "grad_norm": 0.3075149953365326, "learning_rate": 8.346185463374423e-05, "loss": 0.0227, "step": 30330 }, { "grad_norm": 0.178441122174263, "learning_rate": 8.344956671449524e-05, "loss": 0.0237, "step": 30340 }, { "grad_norm": 0.1947997808456421, "learning_rate": 8.343727513725192e-05, "loss": 0.0203, "step": 30350 }, { "grad_norm": 0.18361331522464752, "learning_rate": 8.342497990335847e-05, "loss": 0.0195, "step": 30360 }, { "grad_norm": 0.21785935759544373, "learning_rate": 8.341268101415946e-05, "loss": 0.0182, "step": 30370 }, { "grad_norm": 0.20580866932868958, "learning_rate": 8.34003784709999e-05, "loss": 0.0247, "step": 30380 }, { "grad_norm": 0.14943227171897888, "learning_rate": 8.338807227522516e-05, "loss": 0.0163, "step": 30390 }, { "grad_norm": 0.21705225110054016, "learning_rate": 8.337576242818103e-05, "loss": 0.0205, "step": 30400 }, { "grad_norm": 0.23283807933330536, "learning_rate": 8.336344893121372e-05, "loss": 0.0212, "step": 30410 }, { "grad_norm": 0.2369149625301361, "learning_rate": 8.335113178566977e-05, "loss": 0.0208, "step": 30420 }, { "grad_norm": 0.2020590603351593, "learning_rate": 8.33388109928962e-05, "loss": 0.0197, "step": 30430 }, { "grad_norm": 0.20966574549674988, "learning_rate": 8.33264865542404e-05, "loss": 0.0193, "step": 30440 }, { "grad_norm": 0.17452991008758545, "learning_rate": 8.331415847105013e-05, "loss": 0.0181, "step": 30450 }, { "grad_norm": 0.18311811983585358, "learning_rate": 8.330182674467357e-05, "loss": 0.0173, "step": 30460 }, { "grad_norm": 0.17538543045520782, "learning_rate": 8.32894913764593e-05, "loss": 0.0165, "step": 30470 }, { "grad_norm": 0.16304665803909302, "learning_rate": 8.327715236775633e-05, "loss": 0.0202, "step": 30480 }, { "grad_norm": 0.20797669887542725, "learning_rate": 8.326480971991398e-05, "loss": 0.0175, "step": 30490 }, { "grad_norm": 0.23385150730609894, "learning_rate": 8.325246343428206e-05, "loss": 0.0196, "step": 30500 }, { "grad_norm": 0.23089830577373505, "learning_rate": 8.324011351221072e-05, "loss": 0.0168, "step": 30510 }, { "grad_norm": 0.19048041105270386, "learning_rate": 8.322775995505057e-05, "loss": 0.0216, "step": 30520 }, { "grad_norm": 0.21349899470806122, "learning_rate": 8.321540276415254e-05, "loss": 0.0174, "step": 30530 }, { "grad_norm": 0.17587576806545258, "learning_rate": 8.320304194086798e-05, "loss": 0.0183, "step": 30540 }, { "grad_norm": 0.23387131094932556, "learning_rate": 8.31906774865487e-05, "loss": 0.0207, "step": 30550 }, { "grad_norm": 0.21324960887432098, "learning_rate": 8.317830940254682e-05, "loss": 0.0181, "step": 30560 }, { "grad_norm": 0.21472804248332977, "learning_rate": 8.316593769021491e-05, "loss": 0.0156, "step": 30570 }, { "grad_norm": 0.23477159440517426, "learning_rate": 8.315356235090592e-05, "loss": 0.0166, "step": 30580 }, { "grad_norm": 0.22442622482776642, "learning_rate": 8.314118338597319e-05, "loss": 0.0166, "step": 30590 }, { "grad_norm": 0.19905173778533936, "learning_rate": 8.312880079677048e-05, "loss": 0.0176, "step": 30600 }, { "grad_norm": 0.17853379249572754, "learning_rate": 8.311641458465191e-05, "loss": 0.0165, "step": 30610 }, { "grad_norm": 0.15276308357715607, "learning_rate": 8.310402475097205e-05, "loss": 0.0161, "step": 30620 }, { "grad_norm": 0.21675312519073486, "learning_rate": 8.309163129708581e-05, "loss": 0.0189, "step": 30630 }, { "grad_norm": 0.1813201904296875, "learning_rate": 8.307923422434852e-05, "loss": 0.0178, "step": 30640 }, { "grad_norm": 0.19611844420433044, "learning_rate": 8.30668335341159e-05, "loss": 0.0176, "step": 30650 }, { "grad_norm": 0.1631004810333252, "learning_rate": 8.305442922774408e-05, "loss": 0.0174, "step": 30660 }, { "grad_norm": 0.20400579273700714, "learning_rate": 8.304202130658959e-05, "loss": 0.0179, "step": 30670 }, { "grad_norm": 0.2125239074230194, "learning_rate": 8.302960977200931e-05, "loss": 0.0219, "step": 30680 }, { "grad_norm": 0.1848410964012146, "learning_rate": 8.301719462536058e-05, "loss": 0.0183, "step": 30690 }, { "grad_norm": 0.22395645081996918, "learning_rate": 8.300477586800108e-05, "loss": 0.0199, "step": 30700 }, { "grad_norm": 0.1690165251493454, "learning_rate": 8.299235350128892e-05, "loss": 0.0171, "step": 30710 }, { "grad_norm": 0.24046532809734344, "learning_rate": 8.297992752658258e-05, "loss": 0.0203, "step": 30720 }, { "grad_norm": 0.27862849831581116, "learning_rate": 8.296749794524094e-05, "loss": 0.0274, "step": 30730 }, { "grad_norm": 0.21387328207492828, "learning_rate": 8.295506475862332e-05, "loss": 0.0176, "step": 30740 }, { "grad_norm": 0.20986983180046082, "learning_rate": 8.294262796808933e-05, "loss": 0.0209, "step": 30750 }, { "grad_norm": 0.22753022611141205, "learning_rate": 8.293018757499909e-05, "loss": 0.0213, "step": 30760 }, { "grad_norm": 0.19435620307922363, "learning_rate": 8.291774358071305e-05, "loss": 0.0189, "step": 30770 }, { "grad_norm": 0.24379369616508484, "learning_rate": 8.290529598659205e-05, "loss": 0.0178, "step": 30780 }, { "grad_norm": 0.18410581350326538, "learning_rate": 8.289284479399738e-05, "loss": 0.0209, "step": 30790 }, { "grad_norm": 0.2254180610179901, "learning_rate": 8.288039000429064e-05, "loss": 0.0239, "step": 30800 }, { "grad_norm": 0.19257640838623047, "learning_rate": 8.286793161883388e-05, "loss": 0.0193, "step": 30810 }, { "grad_norm": 0.1585400253534317, "learning_rate": 8.285546963898954e-05, "loss": 0.0198, "step": 30820 }, { "grad_norm": 0.20780685544013977, "learning_rate": 8.284300406612044e-05, "loss": 0.0198, "step": 30830 }, { "grad_norm": 0.22000326216220856, "learning_rate": 8.283053490158978e-05, "loss": 0.0165, "step": 30840 }, { "grad_norm": 0.21340753138065338, "learning_rate": 8.28180621467612e-05, "loss": 0.0185, "step": 30850 }, { "grad_norm": 0.1757088303565979, "learning_rate": 8.280558580299867e-05, "loss": 0.0204, "step": 30860 }, { "grad_norm": 0.22215022146701813, "learning_rate": 8.279310587166661e-05, "loss": 0.0224, "step": 30870 }, { "grad_norm": 0.2692130506038666, "learning_rate": 8.278062235412978e-05, "loss": 0.0199, "step": 30880 }, { "grad_norm": 0.189572274684906, "learning_rate": 8.276813525175339e-05, "loss": 0.0184, "step": 30890 }, { "grad_norm": 0.18988002836704254, "learning_rate": 8.2755644565903e-05, "loss": 0.0178, "step": 30900 }, { "grad_norm": 0.17848196625709534, "learning_rate": 8.274315029794454e-05, "loss": 0.0208, "step": 30910 }, { "grad_norm": 0.21205946803092957, "learning_rate": 8.273065244924443e-05, "loss": 0.0141, "step": 30920 }, { "grad_norm": 0.20823155343532562, "learning_rate": 8.271815102116936e-05, "loss": 0.0191, "step": 30930 }, { "grad_norm": 0.2268962413072586, "learning_rate": 8.270564601508648e-05, "loss": 0.0215, "step": 30940 }, { "grad_norm": 0.22888211905956268, "learning_rate": 8.269313743236333e-05, "loss": 0.0202, "step": 30950 }, { "grad_norm": 0.20782431960105896, "learning_rate": 8.268062527436783e-05, "loss": 0.0194, "step": 30960 }, { "grad_norm": 0.1881447434425354, "learning_rate": 8.266810954246828e-05, "loss": 0.026, "step": 30970 }, { "grad_norm": 0.15232130885124207, "learning_rate": 8.265559023803338e-05, "loss": 0.0216, "step": 30980 }, { "grad_norm": 0.22477930784225464, "learning_rate": 8.264306736243223e-05, "loss": 0.0217, "step": 30990 }, { "grad_norm": 0.1935305893421173, "learning_rate": 8.263054091703432e-05, "loss": 0.022, "step": 31000 }, { "grad_norm": 0.21497085690498352, "learning_rate": 8.26180109032095e-05, "loss": 0.0178, "step": 31010 }, { "grad_norm": 0.19795428216457367, "learning_rate": 8.260547732232807e-05, "loss": 0.0213, "step": 31020 }, { "grad_norm": 0.2129577100276947, "learning_rate": 8.259294017576063e-05, "loss": 0.0169, "step": 31030 }, { "grad_norm": 0.1581229418516159, "learning_rate": 8.258039946487828e-05, "loss": 0.0215, "step": 31040 }, { "grad_norm": 0.18000192940235138, "learning_rate": 8.256785519105241e-05, "loss": 0.0187, "step": 31050 }, { "grad_norm": 0.20432528853416443, "learning_rate": 8.255530735565488e-05, "loss": 0.0182, "step": 31060 }, { "grad_norm": 0.17354518175125122, "learning_rate": 8.254275596005787e-05, "loss": 0.0177, "step": 31070 }, { "grad_norm": 0.17478814721107483, "learning_rate": 8.2530201005634e-05, "loss": 0.0167, "step": 31080 }, { "grad_norm": 0.21015873551368713, "learning_rate": 8.251764249375626e-05, "loss": 0.0195, "step": 31090 }, { "grad_norm": 0.2620306611061096, "learning_rate": 8.250508042579803e-05, "loss": 0.021, "step": 31100 }, { "grad_norm": 0.14711534976959229, "learning_rate": 8.249251480313307e-05, "loss": 0.0148, "step": 31110 }, { "grad_norm": 0.23087604343891144, "learning_rate": 8.247994562713555e-05, "loss": 0.0194, "step": 31120 }, { "grad_norm": 0.2363077849149704, "learning_rate": 8.246737289917998e-05, "loss": 0.0169, "step": 31130 }, { "grad_norm": 0.2101563662290573, "learning_rate": 8.245479662064135e-05, "loss": 0.0174, "step": 31140 }, { "grad_norm": 0.15941190719604492, "learning_rate": 8.244221679289496e-05, "loss": 0.0132, "step": 31150 }, { "grad_norm": 0.1705913543701172, "learning_rate": 8.242963341731652e-05, "loss": 0.0173, "step": 31160 }, { "grad_norm": 0.22632765769958496, "learning_rate": 8.24170464952821e-05, "loss": 0.0177, "step": 31170 }, { "grad_norm": 0.2571001946926117, "learning_rate": 8.240445602816824e-05, "loss": 0.0208, "step": 31180 }, { "grad_norm": 0.2137422114610672, "learning_rate": 8.239186201735179e-05, "loss": 0.0222, "step": 31190 }, { "grad_norm": 0.1510898768901825, "learning_rate": 8.237926446420998e-05, "loss": 0.0166, "step": 31200 }, { "grad_norm": 0.16438816487789154, "learning_rate": 8.236666337012052e-05, "loss": 0.0206, "step": 31210 }, { "grad_norm": 0.19864514470100403, "learning_rate": 8.23540587364614e-05, "loss": 0.0182, "step": 31220 }, { "grad_norm": 0.20598836243152618, "learning_rate": 8.234145056461107e-05, "loss": 0.0185, "step": 31230 }, { "grad_norm": 0.1697952300310135, "learning_rate": 8.232883885594832e-05, "loss": 0.0158, "step": 31240 }, { "grad_norm": 0.16195327043533325, "learning_rate": 8.231622361185236e-05, "loss": 0.0175, "step": 31250 }, { "grad_norm": 0.14259302616119385, "learning_rate": 8.230360483370278e-05, "loss": 0.0187, "step": 31260 }, { "grad_norm": 0.15084391832351685, "learning_rate": 8.229098252287953e-05, "loss": 0.0156, "step": 31270 }, { "grad_norm": 0.13488666713237762, "learning_rate": 8.2278356680763e-05, "loss": 0.0209, "step": 31280 }, { "grad_norm": 0.24893417954444885, "learning_rate": 8.22657273087339e-05, "loss": 0.0185, "step": 31290 }, { "grad_norm": 0.2233414202928543, "learning_rate": 8.225309440817336e-05, "loss": 0.0196, "step": 31300 }, { "grad_norm": 0.18540523946285248, "learning_rate": 8.224045798046293e-05, "loss": 0.0204, "step": 31310 }, { "grad_norm": 0.2508962154388428, "learning_rate": 8.22278180269845e-05, "loss": 0.0187, "step": 31320 }, { "grad_norm": 0.2437829077243805, "learning_rate": 8.221517454912031e-05, "loss": 0.0186, "step": 31330 }, { "grad_norm": 0.24894951283931732, "learning_rate": 8.220252754825308e-05, "loss": 0.0182, "step": 31340 }, { "grad_norm": 0.21141250431537628, "learning_rate": 8.218987702576586e-05, "loss": 0.0169, "step": 31350 }, { "grad_norm": 0.22580371797084808, "learning_rate": 8.217722298304207e-05, "loss": 0.0212, "step": 31360 }, { "grad_norm": 0.20387762784957886, "learning_rate": 8.216456542146557e-05, "loss": 0.016, "step": 31370 }, { "grad_norm": 0.18062745034694672, "learning_rate": 8.215190434242055e-05, "loss": 0.0185, "step": 31380 }, { "grad_norm": 0.139076828956604, "learning_rate": 8.213923974729161e-05, "loss": 0.0163, "step": 31390 }, { "grad_norm": 0.17126673460006714, "learning_rate": 8.212657163746373e-05, "loss": 0.0141, "step": 31400 }, { "grad_norm": 0.19684696197509766, "learning_rate": 8.211390001432227e-05, "loss": 0.0167, "step": 31410 }, { "grad_norm": 0.16030581295490265, "learning_rate": 8.210122487925297e-05, "loss": 0.0185, "step": 31420 }, { "grad_norm": 0.16968552768230438, "learning_rate": 8.208854623364202e-05, "loss": 0.017, "step": 31430 }, { "grad_norm": 0.23636209964752197, "learning_rate": 8.207586407887589e-05, "loss": 0.0166, "step": 31440 }, { "grad_norm": 0.20905648171901703, "learning_rate": 8.206317841634148e-05, "loss": 0.0197, "step": 31450 }, { "grad_norm": 0.24137064814567566, "learning_rate": 8.205048924742609e-05, "loss": 0.0207, "step": 31460 }, { "grad_norm": 0.2096399962902069, "learning_rate": 8.203779657351738e-05, "loss": 0.0195, "step": 31470 }, { "grad_norm": 0.176533505320549, "learning_rate": 8.20251003960034e-05, "loss": 0.0208, "step": 31480 }, { "grad_norm": 0.17365314066410065, "learning_rate": 8.201240071627258e-05, "loss": 0.0165, "step": 31490 }, { "grad_norm": 0.1922934502363205, "learning_rate": 8.199969753571377e-05, "loss": 0.0195, "step": 31500 }, { "grad_norm": 0.170059472322464, "learning_rate": 8.198699085571615e-05, "loss": 0.0144, "step": 31510 }, { "grad_norm": 0.19474516808986664, "learning_rate": 8.197428067766928e-05, "loss": 0.0228, "step": 31520 }, { "grad_norm": 0.20575274527072906, "learning_rate": 8.196156700296316e-05, "loss": 0.0173, "step": 31530 }, { "grad_norm": 0.2465728223323822, "learning_rate": 8.194884983298814e-05, "loss": 0.0229, "step": 31540 }, { "grad_norm": 0.19516612589359283, "learning_rate": 8.193612916913491e-05, "loss": 0.0165, "step": 31550 }, { "grad_norm": 0.25742754340171814, "learning_rate": 8.192340501279463e-05, "loss": 0.0207, "step": 31560 }, { "grad_norm": 0.18097059428691864, "learning_rate": 8.191067736535876e-05, "loss": 0.0173, "step": 31570 }, { "grad_norm": 0.21537645161151886, "learning_rate": 8.18979462282192e-05, "loss": 0.0184, "step": 31580 }, { "grad_norm": 0.18641090393066406, "learning_rate": 8.188521160276819e-05, "loss": 0.0193, "step": 31590 }, { "grad_norm": 0.14705799520015717, "learning_rate": 8.187247349039837e-05, "loss": 0.0154, "step": 31600 }, { "grad_norm": 0.19598110020160675, "learning_rate": 8.185973189250278e-05, "loss": 0.0158, "step": 31610 }, { "grad_norm": 0.1581810712814331, "learning_rate": 8.184698681047482e-05, "loss": 0.0163, "step": 31620 }, { "grad_norm": 0.25806301832199097, "learning_rate": 8.183423824570827e-05, "loss": 0.0261, "step": 31630 }, { "grad_norm": 0.13291172683238983, "learning_rate": 8.182148619959725e-05, "loss": 0.0155, "step": 31640 }, { "grad_norm": 0.17503105103969574, "learning_rate": 8.180873067353636e-05, "loss": 0.0154, "step": 31650 }, { "grad_norm": 0.1972164809703827, "learning_rate": 8.179597166892052e-05, "loss": 0.0146, "step": 31660 }, { "grad_norm": 0.19439521431922913, "learning_rate": 8.178320918714501e-05, "loss": 0.0201, "step": 31670 }, { "grad_norm": 0.24964267015457153, "learning_rate": 8.177044322960554e-05, "loss": 0.019, "step": 31680 }, { "grad_norm": 0.19989554584026337, "learning_rate": 8.175767379769816e-05, "loss": 0.0186, "step": 31690 }, { "grad_norm": 0.2125871628522873, "learning_rate": 8.174490089281932e-05, "loss": 0.0144, "step": 31700 }, { "grad_norm": 0.17382444441318512, "learning_rate": 8.173212451636584e-05, "loss": 0.0154, "step": 31710 }, { "grad_norm": 0.22185297310352325, "learning_rate": 8.171934466973493e-05, "loss": 0.0205, "step": 31720 }, { "grad_norm": 0.19710275530815125, "learning_rate": 8.170656135432418e-05, "loss": 0.0181, "step": 31730 }, { "grad_norm": 0.15518058836460114, "learning_rate": 8.169377457153155e-05, "loss": 0.0153, "step": 31740 }, { "grad_norm": 0.1865607053041458, "learning_rate": 8.168098432275539e-05, "loss": 0.0208, "step": 31750 }, { "grad_norm": 0.20243416726589203, "learning_rate": 8.166819060939442e-05, "loss": 0.0177, "step": 31760 }, { "grad_norm": 0.1517820507287979, "learning_rate": 8.165539343284772e-05, "loss": 0.0208, "step": 31770 }, { "grad_norm": 0.19084376096725464, "learning_rate": 8.16425927945148e-05, "loss": 0.0178, "step": 31780 }, { "grad_norm": 0.16596002876758575, "learning_rate": 8.162978869579551e-05, "loss": 0.0207, "step": 31790 }, { "grad_norm": 0.20983073115348816, "learning_rate": 8.161698113809007e-05, "loss": 0.0225, "step": 31800 }, { "grad_norm": 0.1636507660150528, "learning_rate": 8.160417012279911e-05, "loss": 0.02, "step": 31810 }, { "grad_norm": 0.1932886391878128, "learning_rate": 8.159135565132363e-05, "loss": 0.02, "step": 31820 }, { "grad_norm": 0.14040406048297882, "learning_rate": 8.157853772506498e-05, "loss": 0.0246, "step": 31830 }, { "grad_norm": 0.19682571291923523, "learning_rate": 8.156571634542494e-05, "loss": 0.0217, "step": 31840 }, { "grad_norm": 0.14845623075962067, "learning_rate": 8.15528915138056e-05, "loss": 0.0192, "step": 31850 }, { "grad_norm": 0.15094375610351562, "learning_rate": 8.154006323160949e-05, "loss": 0.0185, "step": 31860 }, { "grad_norm": 0.21630489826202393, "learning_rate": 8.152723150023949e-05, "loss": 0.0179, "step": 31870 }, { "grad_norm": 0.225356325507164, "learning_rate": 8.151439632109886e-05, "loss": 0.0189, "step": 31880 }, { "grad_norm": 0.22384315729141235, "learning_rate": 8.150155769559122e-05, "loss": 0.0213, "step": 31890 }, { "grad_norm": 0.2086501568555832, "learning_rate": 8.148871562512058e-05, "loss": 0.0236, "step": 31900 }, { "grad_norm": 0.2270362824201584, "learning_rate": 8.147587011109136e-05, "loss": 0.0215, "step": 31910 }, { "grad_norm": 0.19973312318325043, "learning_rate": 8.14630211549083e-05, "loss": 0.016, "step": 31920 }, { "grad_norm": 0.20581485331058502, "learning_rate": 8.145016875797655e-05, "loss": 0.0196, "step": 31930 }, { "grad_norm": 0.18175679445266724, "learning_rate": 8.143731292170164e-05, "loss": 0.0156, "step": 31940 }, { "grad_norm": 0.12520043551921844, "learning_rate": 8.142445364748944e-05, "loss": 0.0192, "step": 31950 }, { "grad_norm": 0.24950236082077026, "learning_rate": 8.141159093674624e-05, "loss": 0.0171, "step": 31960 }, { "grad_norm": 0.18138371407985687, "learning_rate": 8.139872479087869e-05, "loss": 0.0171, "step": 31970 }, { "grad_norm": 0.23322458565235138, "learning_rate": 8.13858552112938e-05, "loss": 0.0177, "step": 31980 }, { "grad_norm": 0.2193671315908432, "learning_rate": 8.137298219939895e-05, "loss": 0.0197, "step": 31990 }, { "grad_norm": 0.14333631098270416, "learning_rate": 8.136010575660196e-05, "loss": 0.0185, "step": 32000 }, { "grad_norm": 0.18278661370277405, "learning_rate": 8.134722588431095e-05, "loss": 0.0186, "step": 32010 }, { "grad_norm": 0.17790524661540985, "learning_rate": 8.133434258393444e-05, "loss": 0.0192, "step": 32020 }, { "grad_norm": 0.1922370344400406, "learning_rate": 8.132145585688134e-05, "loss": 0.022, "step": 32030 }, { "grad_norm": 0.12433015555143356, "learning_rate": 8.130856570456093e-05, "loss": 0.0194, "step": 32040 }, { "grad_norm": 0.19856788218021393, "learning_rate": 8.129567212838283e-05, "loss": 0.0198, "step": 32050 }, { "grad_norm": 0.19063930213451385, "learning_rate": 8.128277512975708e-05, "loss": 0.0177, "step": 32060 }, { "grad_norm": 0.18122273683547974, "learning_rate": 8.126987471009408e-05, "loss": 0.0172, "step": 32070 }, { "grad_norm": 0.14685232937335968, "learning_rate": 8.125697087080459e-05, "loss": 0.0162, "step": 32080 }, { "grad_norm": 0.18084755539894104, "learning_rate": 8.124406361329976e-05, "loss": 0.0158, "step": 32090 }, { "grad_norm": 0.22704000771045685, "learning_rate": 8.123115293899111e-05, "loss": 0.0197, "step": 32100 }, { "grad_norm": 0.20101897418498993, "learning_rate": 8.121823884929055e-05, "loss": 0.0177, "step": 32110 }, { "grad_norm": 0.22747589647769928, "learning_rate": 8.12053213456103e-05, "loss": 0.0187, "step": 32120 }, { "grad_norm": 0.14060451090335846, "learning_rate": 8.119240042936303e-05, "loss": 0.0153, "step": 32130 }, { "grad_norm": 0.1706213653087616, "learning_rate": 8.117947610196175e-05, "loss": 0.0156, "step": 32140 }, { "grad_norm": 0.19037163257598877, "learning_rate": 8.116654836481982e-05, "loss": 0.0161, "step": 32150 }, { "grad_norm": 0.19323156774044037, "learning_rate": 8.115361721935106e-05, "loss": 0.0175, "step": 32160 }, { "grad_norm": 0.15423686802387238, "learning_rate": 8.114068266696953e-05, "loss": 0.0158, "step": 32170 }, { "grad_norm": 0.20274047553539276, "learning_rate": 8.112774470908978e-05, "loss": 0.0175, "step": 32180 }, { "grad_norm": 0.17318770289421082, "learning_rate": 8.111480334712665e-05, "loss": 0.0166, "step": 32190 }, { "grad_norm": 0.19020633399486542, "learning_rate": 8.110185858249542e-05, "loss": 0.0159, "step": 32200 }, { "grad_norm": 0.21593530476093292, "learning_rate": 8.108891041661168e-05, "loss": 0.0151, "step": 32210 }, { "grad_norm": 0.15092112123966217, "learning_rate": 8.107595885089146e-05, "loss": 0.0158, "step": 32220 }, { "grad_norm": 0.19982632994651794, "learning_rate": 8.106300388675108e-05, "loss": 0.0158, "step": 32230 }, { "grad_norm": 0.19955745339393616, "learning_rate": 8.105004552560731e-05, "loss": 0.0158, "step": 32240 }, { "grad_norm": 0.22321009635925293, "learning_rate": 8.103708376887724e-05, "loss": 0.0172, "step": 32250 }, { "grad_norm": 0.1503547579050064, "learning_rate": 8.102411861797836e-05, "loss": 0.0181, "step": 32260 }, { "grad_norm": 0.202704057097435, "learning_rate": 8.101115007432851e-05, "loss": 0.0187, "step": 32270 }, { "grad_norm": 0.16256873309612274, "learning_rate": 8.09981781393459e-05, "loss": 0.0181, "step": 32280 }, { "grad_norm": 0.22492632269859314, "learning_rate": 8.098520281444915e-05, "loss": 0.0171, "step": 32290 }, { "grad_norm": 0.24491818249225616, "learning_rate": 8.09722241010572e-05, "loss": 0.017, "step": 32300 }, { "grad_norm": 0.16506798565387726, "learning_rate": 8.095924200058939e-05, "loss": 0.0175, "step": 32310 }, { "grad_norm": 0.2188122719526291, "learning_rate": 8.094625651446541e-05, "loss": 0.0173, "step": 32320 }, { "grad_norm": 0.19859057664871216, "learning_rate": 8.093326764410536e-05, "loss": 0.0168, "step": 32330 }, { "grad_norm": 0.15027163922786713, "learning_rate": 8.092027539092966e-05, "loss": 0.0176, "step": 32340 }, { "grad_norm": 0.1850799173116684, "learning_rate": 8.090727975635913e-05, "loss": 0.0196, "step": 32350 }, { "grad_norm": 0.16584913432598114, "learning_rate": 8.089428074181497e-05, "loss": 0.0181, "step": 32360 }, { "grad_norm": 0.1827545166015625, "learning_rate": 8.088127834871871e-05, "loss": 0.0166, "step": 32370 }, { "grad_norm": 0.15309156477451324, "learning_rate": 8.086827257849226e-05, "loss": 0.0176, "step": 32380 }, { "grad_norm": 0.1553041934967041, "learning_rate": 8.085526343255795e-05, "loss": 0.0164, "step": 32390 }, { "grad_norm": 0.23330995440483093, "learning_rate": 8.084225091233842e-05, "loss": 0.0182, "step": 32400 }, { "grad_norm": 0.24763910472393036, "learning_rate": 8.082923501925668e-05, "loss": 0.018, "step": 32410 }, { "grad_norm": 0.2540380656719208, "learning_rate": 8.081621575473617e-05, "loss": 0.0186, "step": 32420 }, { "grad_norm": 0.2518235146999359, "learning_rate": 8.080319312020064e-05, "loss": 0.0197, "step": 32430 }, { "grad_norm": 0.17033837735652924, "learning_rate": 8.079016711707421e-05, "loss": 0.0178, "step": 32440 }, { "grad_norm": 0.1875060796737671, "learning_rate": 8.077713774678139e-05, "loss": 0.0158, "step": 32450 }, { "grad_norm": 0.1996062994003296, "learning_rate": 8.076410501074707e-05, "loss": 0.0148, "step": 32460 }, { "grad_norm": 0.2636808454990387, "learning_rate": 8.075106891039647e-05, "loss": 0.0195, "step": 32470 }, { "grad_norm": 0.24566654860973358, "learning_rate": 8.073802944715523e-05, "loss": 0.0206, "step": 32480 }, { "grad_norm": 0.18982163071632385, "learning_rate": 8.072498662244929e-05, "loss": 0.0178, "step": 32490 }, { "grad_norm": 0.13123264908790588, "learning_rate": 8.0711940437705e-05, "loss": 0.0141, "step": 32500 }, { "grad_norm": 0.19946293532848358, "learning_rate": 8.06988908943491e-05, "loss": 0.0174, "step": 32510 }, { "grad_norm": 0.22928062081336975, "learning_rate": 8.068583799380863e-05, "loss": 0.0231, "step": 32520 }, { "grad_norm": 0.2291661649942398, "learning_rate": 8.067278173751104e-05, "loss": 0.0172, "step": 32530 }, { "grad_norm": 0.17107222974300385, "learning_rate": 8.065972212688417e-05, "loss": 0.0179, "step": 32540 }, { "grad_norm": 0.2609204947948456, "learning_rate": 8.064665916335618e-05, "loss": 0.0185, "step": 32550 }, { "grad_norm": 0.23173421621322632, "learning_rate": 8.063359284835564e-05, "loss": 0.024, "step": 32560 }, { "grad_norm": 0.20029905438423157, "learning_rate": 8.062052318331142e-05, "loss": 0.017, "step": 32570 }, { "grad_norm": 0.19514523446559906, "learning_rate": 8.060745016965283e-05, "loss": 0.0153, "step": 32580 }, { "grad_norm": 0.2753528952598572, "learning_rate": 8.059437380880952e-05, "loss": 0.0169, "step": 32590 }, { "grad_norm": 0.2503444254398346, "learning_rate": 8.058129410221146e-05, "loss": 0.0208, "step": 32600 }, { "grad_norm": 0.14537464082241058, "learning_rate": 8.056821105128908e-05, "loss": 0.0182, "step": 32610 }, { "grad_norm": 0.24136196076869965, "learning_rate": 8.05551246574731e-05, "loss": 0.0218, "step": 32620 }, { "grad_norm": 0.2247494012117386, "learning_rate": 8.05420349221946e-05, "loss": 0.0201, "step": 32630 }, { "grad_norm": 0.17778939008712769, "learning_rate": 8.05289418468851e-05, "loss": 0.0198, "step": 32640 }, { "grad_norm": 0.19179977476596832, "learning_rate": 8.051584543297642e-05, "loss": 0.0171, "step": 32650 }, { "grad_norm": 0.21325668692588806, "learning_rate": 8.050274568190074e-05, "loss": 0.0158, "step": 32660 }, { "grad_norm": 0.18425022065639496, "learning_rate": 8.048964259509067e-05, "loss": 0.0155, "step": 32670 }, { "grad_norm": 0.16545376181602478, "learning_rate": 8.047653617397914e-05, "loss": 0.0165, "step": 32680 }, { "grad_norm": 0.1715622842311859, "learning_rate": 8.046342641999941e-05, "loss": 0.0186, "step": 32690 }, { "grad_norm": 0.1955346018075943, "learning_rate": 8.045031333458517e-05, "loss": 0.0188, "step": 32700 }, { "grad_norm": 0.21135495603084564, "learning_rate": 8.043719691917047e-05, "loss": 0.0216, "step": 32710 }, { "grad_norm": 0.16906245052814484, "learning_rate": 8.042407717518966e-05, "loss": 0.0196, "step": 32720 }, { "grad_norm": 0.17433243989944458, "learning_rate": 8.041095410407751e-05, "loss": 0.0186, "step": 32730 }, { "grad_norm": 0.1511361300945282, "learning_rate": 8.039782770726913e-05, "loss": 0.0143, "step": 32740 }, { "grad_norm": 0.18242454528808594, "learning_rate": 8.038469798620004e-05, "loss": 0.0152, "step": 32750 }, { "grad_norm": 0.17797084152698517, "learning_rate": 8.037156494230604e-05, "loss": 0.0145, "step": 32760 }, { "grad_norm": 0.16283424198627472, "learning_rate": 8.035842857702338e-05, "loss": 0.0141, "step": 32770 }, { "grad_norm": 0.2065792977809906, "learning_rate": 8.03452888917886e-05, "loss": 0.0186, "step": 32780 }, { "grad_norm": 0.1922413408756256, "learning_rate": 8.033214588803866e-05, "loss": 0.0208, "step": 32790 }, { "grad_norm": 0.20545358955860138, "learning_rate": 8.031899956721083e-05, "loss": 0.0184, "step": 32800 }, { "grad_norm": 0.1760982871055603, "learning_rate": 8.030584993074282e-05, "loss": 0.0193, "step": 32810 }, { "grad_norm": 0.2588331699371338, "learning_rate": 8.02926969800726e-05, "loss": 0.0146, "step": 32820 }, { "grad_norm": 0.12440851330757141, "learning_rate": 8.027954071663859e-05, "loss": 0.0177, "step": 32830 }, { "grad_norm": 0.16183078289031982, "learning_rate": 8.026638114187954e-05, "loss": 0.0171, "step": 32840 }, { "grad_norm": 0.14690864086151123, "learning_rate": 8.025321825723456e-05, "loss": 0.015, "step": 32850 }, { "grad_norm": 0.18994668126106262, "learning_rate": 8.02400520641431e-05, "loss": 0.0166, "step": 32860 }, { "grad_norm": 0.2206994742155075, "learning_rate": 8.022688256404501e-05, "loss": 0.0182, "step": 32870 }, { "grad_norm": 0.19561515748500824, "learning_rate": 8.02137097583805e-05, "loss": 0.0173, "step": 32880 }, { "grad_norm": 0.18986760079860687, "learning_rate": 8.02005336485901e-05, "loss": 0.019, "step": 32890 }, { "grad_norm": 0.1643112152814865, "learning_rate": 8.018735423611476e-05, "loss": 0.0181, "step": 32900 }, { "grad_norm": 0.19591473042964935, "learning_rate": 8.017417152239574e-05, "loss": 0.0177, "step": 32910 }, { "grad_norm": 0.17609015107154846, "learning_rate": 8.01609855088747e-05, "loss": 0.02, "step": 32920 }, { "grad_norm": 0.1986398696899414, "learning_rate": 8.014779619699362e-05, "loss": 0.0178, "step": 32930 }, { "grad_norm": 0.15456396341323853, "learning_rate": 8.013460358819489e-05, "loss": 0.0143, "step": 32940 }, { "grad_norm": 0.20614691078662872, "learning_rate": 8.01214076839212e-05, "loss": 0.0189, "step": 32950 }, { "grad_norm": 0.1844782531261444, "learning_rate": 8.010820848561565e-05, "loss": 0.0154, "step": 32960 }, { "grad_norm": 0.20637066662311554, "learning_rate": 8.009500599472171e-05, "loss": 0.0179, "step": 32970 }, { "grad_norm": 0.2300550490617752, "learning_rate": 8.008180021268314e-05, "loss": 0.0132, "step": 32980 }, { "grad_norm": 0.21466512978076935, "learning_rate": 8.006859114094414e-05, "loss": 0.0166, "step": 32990 }, { "grad_norm": 0.13532757759094238, "learning_rate": 8.005537878094921e-05, "loss": 0.0153, "step": 33000 }, { "grad_norm": 0.17025971412658691, "learning_rate": 8.004216313414323e-05, "loss": 0.0131, "step": 33010 }, { "grad_norm": 0.18664538860321045, "learning_rate": 8.002894420197149e-05, "loss": 0.0123, "step": 33020 }, { "grad_norm": 0.15943029522895813, "learning_rate": 8.001572198587954e-05, "loss": 0.0154, "step": 33030 }, { "grad_norm": 0.2102380245923996, "learning_rate": 8.000249648731338e-05, "loss": 0.0186, "step": 33040 }, { "grad_norm": 0.1718248575925827, "learning_rate": 7.998926770771928e-05, "loss": 0.0183, "step": 33050 }, { "grad_norm": 0.170157790184021, "learning_rate": 7.997603564854397e-05, "loss": 0.0176, "step": 33060 }, { "grad_norm": 0.19558483362197876, "learning_rate": 7.996280031123448e-05, "loss": 0.0163, "step": 33070 }, { "grad_norm": 0.17531050741672516, "learning_rate": 7.994956169723818e-05, "loss": 0.0175, "step": 33080 }, { "grad_norm": 0.20135171711444855, "learning_rate": 7.993631980800285e-05, "loss": 0.0147, "step": 33090 }, { "grad_norm": 0.16800713539123535, "learning_rate": 7.992307464497659e-05, "loss": 0.0151, "step": 33100 }, { "grad_norm": 0.16249792277812958, "learning_rate": 7.990982620960787e-05, "loss": 0.0181, "step": 33110 }, { "grad_norm": 0.20432065427303314, "learning_rate": 7.989657450334554e-05, "loss": 0.0193, "step": 33120 }, { "grad_norm": 0.1843862682580948, "learning_rate": 7.988331952763877e-05, "loss": 0.0208, "step": 33130 }, { "grad_norm": 0.15774114429950714, "learning_rate": 7.987006128393709e-05, "loss": 0.0169, "step": 33140 }, { "grad_norm": 0.195957213640213, "learning_rate": 7.985679977369043e-05, "loss": 0.0175, "step": 33150 }, { "grad_norm": 0.1891787052154541, "learning_rate": 7.984353499834902e-05, "loss": 0.0141, "step": 33160 }, { "grad_norm": 0.1900712549686432, "learning_rate": 7.983026695936351e-05, "loss": 0.0165, "step": 33170 }, { "grad_norm": 0.20133842527866364, "learning_rate": 7.981699565818486e-05, "loss": 0.0171, "step": 33180 }, { "grad_norm": 0.2013472318649292, "learning_rate": 7.980372109626437e-05, "loss": 0.0182, "step": 33190 }, { "grad_norm": 0.18564818799495697, "learning_rate": 7.979044327505375e-05, "loss": 0.0152, "step": 33200 }, { "grad_norm": 0.24143795669078827, "learning_rate": 7.977716219600506e-05, "loss": 0.0158, "step": 33210 }, { "grad_norm": 0.1422279179096222, "learning_rate": 7.97638778605707e-05, "loss": 0.0158, "step": 33220 }, { "grad_norm": 0.19868104159832, "learning_rate": 7.975059027020338e-05, "loss": 0.0144, "step": 33230 }, { "grad_norm": 0.2096700370311737, "learning_rate": 7.973729942635623e-05, "loss": 0.0173, "step": 33240 }, { "grad_norm": 0.21277949213981628, "learning_rate": 7.972400533048273e-05, "loss": 0.0206, "step": 33250 }, { "grad_norm": 0.1666051745414734, "learning_rate": 7.97107079840367e-05, "loss": 0.0161, "step": 33260 }, { "grad_norm": 0.14584186673164368, "learning_rate": 7.969740738847231e-05, "loss": 0.0169, "step": 33270 }, { "grad_norm": 0.1503124088048935, "learning_rate": 7.968410354524411e-05, "loss": 0.0162, "step": 33280 }, { "grad_norm": 0.27889490127563477, "learning_rate": 7.967079645580697e-05, "loss": 0.0199, "step": 33290 }, { "grad_norm": 0.22073878347873688, "learning_rate": 7.965748612161612e-05, "loss": 0.0193, "step": 33300 }, { "grad_norm": 0.15224651992321014, "learning_rate": 7.96441725441272e-05, "loss": 0.0145, "step": 33310 }, { "grad_norm": 0.2281752973794937, "learning_rate": 7.963085572479614e-05, "loss": 0.0178, "step": 33320 }, { "grad_norm": 0.13546480238437653, "learning_rate": 7.961753566507924e-05, "loss": 0.0138, "step": 33330 }, { "grad_norm": 0.2024027407169342, "learning_rate": 7.960421236643316e-05, "loss": 0.0226, "step": 33340 }, { "grad_norm": 0.1532253473997116, "learning_rate": 7.959088583031496e-05, "loss": 0.0141, "step": 33350 }, { "grad_norm": 0.16069383919239044, "learning_rate": 7.957755605818194e-05, "loss": 0.0158, "step": 33360 }, { "grad_norm": 0.19605161249637604, "learning_rate": 7.956422305149185e-05, "loss": 0.0132, "step": 33370 }, { "grad_norm": 0.17513415217399597, "learning_rate": 7.95508868117028e-05, "loss": 0.0157, "step": 33380 }, { "grad_norm": 0.20219287276268005, "learning_rate": 7.953754734027318e-05, "loss": 0.015, "step": 33390 }, { "grad_norm": 0.12060868740081787, "learning_rate": 7.952420463866182e-05, "loss": 0.017, "step": 33400 }, { "grad_norm": 0.23385925590991974, "learning_rate": 7.951085870832782e-05, "loss": 0.0167, "step": 33410 }, { "grad_norm": 0.17265096306800842, "learning_rate": 7.949750955073067e-05, "loss": 0.0168, "step": 33420 }, { "grad_norm": 0.15895269811153412, "learning_rate": 7.948415716733022e-05, "loss": 0.0163, "step": 33430 }, { "grad_norm": 0.1591579169034958, "learning_rate": 7.947080155958669e-05, "loss": 0.0159, "step": 33440 }, { "grad_norm": 0.22047366201877594, "learning_rate": 7.94574427289606e-05, "loss": 0.0169, "step": 33450 }, { "grad_norm": 0.19670067727565765, "learning_rate": 7.944408067691284e-05, "loss": 0.0151, "step": 33460 }, { "grad_norm": 0.2687230110168457, "learning_rate": 7.943071540490473e-05, "loss": 0.0166, "step": 33470 }, { "grad_norm": 0.2322334349155426, "learning_rate": 7.94173469143978e-05, "loss": 0.0205, "step": 33480 }, { "grad_norm": 0.22819796204566956, "learning_rate": 7.940397520685406e-05, "loss": 0.021, "step": 33490 }, { "grad_norm": 0.23159341514110565, "learning_rate": 7.939060028373577e-05, "loss": 0.0167, "step": 33500 }, { "grad_norm": 0.20991484820842743, "learning_rate": 7.937722214650565e-05, "loss": 0.0182, "step": 33510 }, { "grad_norm": 0.2205730676651001, "learning_rate": 7.936384079662666e-05, "loss": 0.0187, "step": 33520 }, { "grad_norm": 0.2610173225402832, "learning_rate": 7.93504562355622e-05, "loss": 0.0214, "step": 33530 }, { "grad_norm": 0.25388118624687195, "learning_rate": 7.933706846477599e-05, "loss": 0.0184, "step": 33540 }, { "grad_norm": 0.22762151062488556, "learning_rate": 7.932367748573206e-05, "loss": 0.0208, "step": 33550 }, { "grad_norm": 0.19042301177978516, "learning_rate": 7.931028329989485e-05, "loss": 0.0158, "step": 33560 }, { "grad_norm": 0.15113654732704163, "learning_rate": 7.929688590872913e-05, "loss": 0.0153, "step": 33570 }, { "grad_norm": 0.1801440566778183, "learning_rate": 7.928348531370003e-05, "loss": 0.0144, "step": 33580 }, { "grad_norm": 0.15467001497745514, "learning_rate": 7.927008151627297e-05, "loss": 0.015, "step": 33590 }, { "grad_norm": 0.21620209515094757, "learning_rate": 7.925667451791383e-05, "loss": 0.016, "step": 33600 }, { "grad_norm": 0.20862185955047607, "learning_rate": 7.924326432008874e-05, "loss": 0.0134, "step": 33610 }, { "grad_norm": 0.18630659580230713, "learning_rate": 7.922985092426422e-05, "loss": 0.0135, "step": 33620 }, { "grad_norm": 0.17678658664226532, "learning_rate": 7.921643433190717e-05, "loss": 0.0184, "step": 33630 }, { "grad_norm": 0.1757718175649643, "learning_rate": 7.920301454448478e-05, "loss": 0.0168, "step": 33640 }, { "grad_norm": 0.13340790569782257, "learning_rate": 7.918959156346461e-05, "loss": 0.0136, "step": 33650 }, { "grad_norm": 0.19073514640331268, "learning_rate": 7.91761653903146e-05, "loss": 0.0178, "step": 33660 }, { "grad_norm": 0.244148850440979, "learning_rate": 7.916273602650302e-05, "loss": 0.0174, "step": 33670 }, { "grad_norm": 0.1552438884973526, "learning_rate": 7.914930347349847e-05, "loss": 0.0144, "step": 33680 }, { "grad_norm": 0.1506880223751068, "learning_rate": 7.913586773276992e-05, "loss": 0.0166, "step": 33690 }, { "grad_norm": 0.14898623526096344, "learning_rate": 7.912242880578667e-05, "loss": 0.0192, "step": 33700 }, { "grad_norm": 0.14053791761398315, "learning_rate": 7.910898669401839e-05, "loss": 0.0178, "step": 33710 }, { "grad_norm": 0.18680214881896973, "learning_rate": 7.909554139893511e-05, "loss": 0.0163, "step": 33720 }, { "grad_norm": 0.15235580503940582, "learning_rate": 7.908209292200715e-05, "loss": 0.0162, "step": 33730 }, { "grad_norm": 0.1641615629196167, "learning_rate": 7.906864126470523e-05, "loss": 0.0188, "step": 33740 }, { "grad_norm": 0.22946299612522125, "learning_rate": 7.905518642850041e-05, "loss": 0.0168, "step": 33750 }, { "grad_norm": 0.14217180013656616, "learning_rate": 7.904172841486409e-05, "loss": 0.017, "step": 33760 }, { "grad_norm": 0.16934485733509064, "learning_rate": 7.902826722526801e-05, "loss": 0.021, "step": 33770 }, { "grad_norm": 0.12906308472156525, "learning_rate": 7.901480286118427e-05, "loss": 0.0148, "step": 33780 }, { "grad_norm": 0.18561923503875732, "learning_rate": 7.900133532408531e-05, "loss": 0.0172, "step": 33790 }, { "grad_norm": 0.14203667640686035, "learning_rate": 7.898786461544395e-05, "loss": 0.0184, "step": 33800 }, { "grad_norm": 0.19763889908790588, "learning_rate": 7.897439073673325e-05, "loss": 0.015, "step": 33810 }, { "grad_norm": 0.1865910142660141, "learning_rate": 7.896091368942677e-05, "loss": 0.0181, "step": 33820 }, { "grad_norm": 0.20815052092075348, "learning_rate": 7.894743347499832e-05, "loss": 0.0194, "step": 33830 }, { "grad_norm": 0.2267421931028366, "learning_rate": 7.893395009492203e-05, "loss": 0.0194, "step": 33840 }, { "grad_norm": 0.17737865447998047, "learning_rate": 7.892046355067248e-05, "loss": 0.0156, "step": 33850 }, { "grad_norm": 0.17273788154125214, "learning_rate": 7.890697384372451e-05, "loss": 0.0138, "step": 33860 }, { "grad_norm": 0.14086787402629852, "learning_rate": 7.889348097555336e-05, "loss": 0.0146, "step": 33870 }, { "grad_norm": 0.16144074499607086, "learning_rate": 7.887998494763455e-05, "loss": 0.0158, "step": 33880 }, { "grad_norm": 0.24518956243991852, "learning_rate": 7.886648576144404e-05, "loss": 0.0195, "step": 33890 }, { "grad_norm": 0.169607475399971, "learning_rate": 7.885298341845802e-05, "loss": 0.0189, "step": 33900 }, { "grad_norm": 0.1319376677274704, "learning_rate": 7.883947792015311e-05, "loss": 0.0155, "step": 33910 }, { "grad_norm": 0.16523632407188416, "learning_rate": 7.882596926800628e-05, "loss": 0.0209, "step": 33920 }, { "grad_norm": 0.20529651641845703, "learning_rate": 7.881245746349477e-05, "loss": 0.0222, "step": 33930 }, { "grad_norm": 0.1914922147989273, "learning_rate": 7.879894250809623e-05, "loss": 0.0178, "step": 33940 }, { "grad_norm": 0.23422665894031525, "learning_rate": 7.878542440328865e-05, "loss": 0.0164, "step": 33950 }, { "grad_norm": 0.22735339403152466, "learning_rate": 7.877190315055031e-05, "loss": 0.0169, "step": 33960 }, { "grad_norm": 0.20582328736782074, "learning_rate": 7.875837875135991e-05, "loss": 0.0167, "step": 33970 }, { "grad_norm": 0.22857464849948883, "learning_rate": 7.874485120719646e-05, "loss": 0.0163, "step": 33980 }, { "grad_norm": 0.18098831176757812, "learning_rate": 7.873132051953928e-05, "loss": 0.015, "step": 33990 }, { "grad_norm": 0.160718634724617, "learning_rate": 7.87177866898681e-05, "loss": 0.0183, "step": 34000 }, { "grad_norm": 0.23644055426120758, "learning_rate": 7.870424971966294e-05, "loss": 0.0231, "step": 34010 }, { "grad_norm": 0.2024950087070465, "learning_rate": 7.869070961040419e-05, "loss": 0.0187, "step": 34020 }, { "grad_norm": 0.18358512222766876, "learning_rate": 7.867716636357257e-05, "loss": 0.0176, "step": 34030 }, { "grad_norm": 0.14919959008693695, "learning_rate": 7.866361998064915e-05, "loss": 0.0178, "step": 34040 }, { "grad_norm": 0.17621645331382751, "learning_rate": 7.865007046311534e-05, "loss": 0.0147, "step": 34050 }, { "grad_norm": 0.21913613379001617, "learning_rate": 7.86365178124529e-05, "loss": 0.0167, "step": 34060 }, { "grad_norm": 0.17727161943912506, "learning_rate": 7.862296203014394e-05, "loss": 0.0164, "step": 34070 }, { "grad_norm": 0.19111908972263336, "learning_rate": 7.860940311767088e-05, "loss": 0.0196, "step": 34080 }, { "grad_norm": 0.21862001717090607, "learning_rate": 7.85958410765165e-05, "loss": 0.0195, "step": 34090 }, { "grad_norm": 0.18026040494441986, "learning_rate": 7.858227590816394e-05, "loss": 0.0163, "step": 34100 }, { "grad_norm": 0.18691319227218628, "learning_rate": 7.856870761409664e-05, "loss": 0.0178, "step": 34110 }, { "grad_norm": 0.2293683886528015, "learning_rate": 7.855513619579846e-05, "loss": 0.0188, "step": 34120 }, { "grad_norm": 0.19171945750713348, "learning_rate": 7.85415616547535e-05, "loss": 0.0189, "step": 34130 }, { "grad_norm": 0.16870298981666565, "learning_rate": 7.852798399244627e-05, "loss": 0.018, "step": 34140 }, { "grad_norm": 0.15703293681144714, "learning_rate": 7.851440321036161e-05, "loss": 0.0135, "step": 34150 }, { "grad_norm": 0.11971268802881241, "learning_rate": 7.850081930998468e-05, "loss": 0.0167, "step": 34160 }, { "grad_norm": 0.21399858593940735, "learning_rate": 7.8487232292801e-05, "loss": 0.0193, "step": 34170 }, { "grad_norm": 0.21988096833229065, "learning_rate": 7.847364216029642e-05, "loss": 0.017, "step": 34180 }, { "grad_norm": 0.1118786409497261, "learning_rate": 7.846004891395716e-05, "loss": 0.0115, "step": 34190 }, { "grad_norm": 0.14991511404514313, "learning_rate": 7.844645255526972e-05, "loss": 0.0175, "step": 34200 }, { "grad_norm": 0.2592407763004303, "learning_rate": 7.843285308572101e-05, "loss": 0.0158, "step": 34210 }, { "grad_norm": 0.22201015055179596, "learning_rate": 7.841925050679823e-05, "loss": 0.0149, "step": 34220 }, { "grad_norm": 0.14965614676475525, "learning_rate": 7.840564481998895e-05, "loss": 0.0152, "step": 34230 }, { "grad_norm": 0.16868284344673157, "learning_rate": 7.839203602678105e-05, "loss": 0.0178, "step": 34240 }, { "grad_norm": 0.17149882018566132, "learning_rate": 7.837842412866279e-05, "loss": 0.0172, "step": 34250 }, { "grad_norm": 0.19536660611629486, "learning_rate": 7.836480912712272e-05, "loss": 0.015, "step": 34260 }, { "grad_norm": 0.1674199402332306, "learning_rate": 7.835119102364976e-05, "loss": 0.0205, "step": 34270 }, { "grad_norm": 0.1615225374698639, "learning_rate": 7.83375698197332e-05, "loss": 0.0145, "step": 34280 }, { "grad_norm": 0.13564032316207886, "learning_rate": 7.83239455168626e-05, "loss": 0.0154, "step": 34290 }, { "grad_norm": 0.25946205854415894, "learning_rate": 7.83103181165279e-05, "loss": 0.0179, "step": 34300 }, { "grad_norm": 0.15908609330654144, "learning_rate": 7.829668762021937e-05, "loss": 0.0193, "step": 34310 }, { "grad_norm": 0.17857132852077484, "learning_rate": 7.828305402942764e-05, "loss": 0.0162, "step": 34320 }, { "grad_norm": 0.1826576143503189, "learning_rate": 7.826941734564363e-05, "loss": 0.013, "step": 34330 }, { "grad_norm": 0.17686474323272705, "learning_rate": 7.825577757035865e-05, "loss": 0.0163, "step": 34340 }, { "grad_norm": 0.21436774730682373, "learning_rate": 7.824213470506431e-05, "loss": 0.0161, "step": 34350 }, { "grad_norm": 0.21651242673397064, "learning_rate": 7.822848875125257e-05, "loss": 0.0164, "step": 34360 }, { "grad_norm": 0.21758484840393066, "learning_rate": 7.821483971041576e-05, "loss": 0.0192, "step": 34370 }, { "grad_norm": 0.20649313926696777, "learning_rate": 7.820118758404649e-05, "loss": 0.0205, "step": 34380 }, { "grad_norm": 0.19700860977172852, "learning_rate": 7.818753237363776e-05, "loss": 0.0155, "step": 34390 }, { "grad_norm": 0.1765429675579071, "learning_rate": 7.817387408068286e-05, "loss": 0.0141, "step": 34400 }, { "grad_norm": 0.17900729179382324, "learning_rate": 7.816021270667544e-05, "loss": 0.0135, "step": 34410 }, { "grad_norm": 0.217325821518898, "learning_rate": 7.81465482531095e-05, "loss": 0.0125, "step": 34420 }, { "grad_norm": 0.19731219112873077, "learning_rate": 7.813288072147938e-05, "loss": 0.0155, "step": 34430 }, { "grad_norm": 0.21129773557186127, "learning_rate": 7.811921011327972e-05, "loss": 0.0165, "step": 34440 }, { "grad_norm": 0.23753772675991058, "learning_rate": 7.810553643000549e-05, "loss": 0.0187, "step": 34450 }, { "grad_norm": 0.17841100692749023, "learning_rate": 7.809185967315206e-05, "loss": 0.0199, "step": 34460 }, { "grad_norm": 0.187652587890625, "learning_rate": 7.80781798442151e-05, "loss": 0.0169, "step": 34470 }, { "grad_norm": 0.1822012960910797, "learning_rate": 7.806449694469061e-05, "loss": 0.0191, "step": 34480 }, { "grad_norm": 0.20260709524154663, "learning_rate": 7.805081097607492e-05, "loss": 0.0164, "step": 34490 }, { "grad_norm": 0.1813521534204483, "learning_rate": 7.803712193986474e-05, "loss": 0.0145, "step": 34500 }, { "grad_norm": 0.17280487716197968, "learning_rate": 7.802342983755702e-05, "loss": 0.0179, "step": 34510 }, { "grad_norm": 0.26338809728622437, "learning_rate": 7.800973467064918e-05, "loss": 0.0173, "step": 34520 }, { "grad_norm": 0.16106285154819489, "learning_rate": 7.799603644063884e-05, "loss": 0.0163, "step": 34530 }, { "grad_norm": 0.17159205675125122, "learning_rate": 7.798233514902405e-05, "loss": 0.014, "step": 34540 }, { "grad_norm": 0.18014615774154663, "learning_rate": 7.796863079730318e-05, "loss": 0.0176, "step": 34550 }, { "grad_norm": 0.17940719425678253, "learning_rate": 7.795492338697488e-05, "loss": 0.0138, "step": 34560 }, { "grad_norm": 0.13913026452064514, "learning_rate": 7.794121291953819e-05, "loss": 0.014, "step": 34570 }, { "grad_norm": 0.1709628701210022, "learning_rate": 7.792749939649246e-05, "loss": 0.015, "step": 34580 }, { "grad_norm": 0.18059542775154114, "learning_rate": 7.79137828193374e-05, "loss": 0.0144, "step": 34590 }, { "grad_norm": 0.1836652159690857, "learning_rate": 7.790006318957301e-05, "loss": 0.0185, "step": 34600 }, { "grad_norm": 0.162364661693573, "learning_rate": 7.788634050869965e-05, "loss": 0.0149, "step": 34610 }, { "grad_norm": 0.1777723729610443, "learning_rate": 7.787261477821803e-05, "loss": 0.0173, "step": 34620 }, { "grad_norm": 0.15170994400978088, "learning_rate": 7.785888599962916e-05, "loss": 0.0168, "step": 34630 }, { "grad_norm": 0.2202051877975464, "learning_rate": 7.784515417443439e-05, "loss": 0.0142, "step": 34640 }, { "grad_norm": 0.20931775867938995, "learning_rate": 7.783141930413545e-05, "loss": 0.0195, "step": 34650 }, { "grad_norm": 0.18222935497760773, "learning_rate": 7.78176813902343e-05, "loss": 0.0179, "step": 34660 }, { "grad_norm": 0.17771220207214355, "learning_rate": 7.780394043423336e-05, "loss": 0.0187, "step": 34670 }, { "grad_norm": 0.17031244933605194, "learning_rate": 7.77901964376353e-05, "loss": 0.0153, "step": 34680 }, { "grad_norm": 0.1870279461145401, "learning_rate": 7.777644940194316e-05, "loss": 0.0173, "step": 34690 }, { "grad_norm": 0.18794368207454681, "learning_rate": 7.776269932866023e-05, "loss": 0.0204, "step": 34700 }, { "grad_norm": 0.19232624769210815, "learning_rate": 7.774894621929026e-05, "loss": 0.0201, "step": 34710 }, { "grad_norm": 0.20736323297023773, "learning_rate": 7.773519007533725e-05, "loss": 0.0143, "step": 34720 }, { "grad_norm": 0.16758807003498077, "learning_rate": 7.772143089830556e-05, "loss": 0.0163, "step": 34730 }, { "grad_norm": 0.16972094774246216, "learning_rate": 7.770766868969985e-05, "loss": 0.014, "step": 34740 }, { "grad_norm": 0.1624872088432312, "learning_rate": 7.769390345102518e-05, "loss": 0.0172, "step": 34750 }, { "grad_norm": 0.1452779322862625, "learning_rate": 7.768013518378683e-05, "loss": 0.0127, "step": 34760 }, { "grad_norm": 0.2041352242231369, "learning_rate": 7.766636388949053e-05, "loss": 0.0187, "step": 34770 }, { "grad_norm": 0.19435305893421173, "learning_rate": 7.765258956964229e-05, "loss": 0.0157, "step": 34780 }, { "grad_norm": 0.1831212192773819, "learning_rate": 7.76388122257484e-05, "loss": 0.0164, "step": 34790 }, { "grad_norm": 0.1835515797138214, "learning_rate": 7.762503185931558e-05, "loss": 0.0207, "step": 34800 }, { "grad_norm": 0.2003551423549652, "learning_rate": 7.76112484718508e-05, "loss": 0.0147, "step": 34810 }, { "grad_norm": 0.2219925820827484, "learning_rate": 7.75974620648614e-05, "loss": 0.0185, "step": 34820 }, { "grad_norm": 0.1544821858406067, "learning_rate": 7.758367263985503e-05, "loss": 0.0163, "step": 34830 }, { "grad_norm": 0.14057134091854095, "learning_rate": 7.75698801983397e-05, "loss": 0.0152, "step": 34840 }, { "grad_norm": 0.12412522733211517, "learning_rate": 7.755608474182372e-05, "loss": 0.0118, "step": 34850 }, { "grad_norm": 0.23813927173614502, "learning_rate": 7.754228627181574e-05, "loss": 0.0187, "step": 34860 }, { "grad_norm": 0.20070931315422058, "learning_rate": 7.752848478982476e-05, "loss": 0.0167, "step": 34870 }, { "grad_norm": 0.18877220153808594, "learning_rate": 7.751468029736006e-05, "loss": 0.0166, "step": 34880 }, { "grad_norm": 0.17259541153907776, "learning_rate": 7.750087279593129e-05, "loss": 0.0141, "step": 34890 }, { "grad_norm": 0.24116401374340057, "learning_rate": 7.748706228704843e-05, "loss": 0.0206, "step": 34900 }, { "grad_norm": 0.27725929021835327, "learning_rate": 7.747324877222176e-05, "loss": 0.0157, "step": 34910 }, { "grad_norm": 0.19588559865951538, "learning_rate": 7.745943225296188e-05, "loss": 0.0172, "step": 34920 }, { "grad_norm": 0.17935794591903687, "learning_rate": 7.744561273077981e-05, "loss": 0.0156, "step": 34930 }, { "grad_norm": 0.17806588113307953, "learning_rate": 7.743179020718678e-05, "loss": 0.0205, "step": 34940 }, { "grad_norm": 0.237547367811203, "learning_rate": 7.741796468369443e-05, "loss": 0.0183, "step": 34950 }, { "grad_norm": 0.23485523462295532, "learning_rate": 7.740413616181466e-05, "loss": 0.0172, "step": 34960 }, { "grad_norm": 0.21326644718647003, "learning_rate": 7.739030464305978e-05, "loss": 0.0206, "step": 34970 }, { "grad_norm": 0.15106301009655, "learning_rate": 7.737647012894235e-05, "loss": 0.0172, "step": 34980 }, { "grad_norm": 0.18418008089065552, "learning_rate": 7.736263262097532e-05, "loss": 0.0166, "step": 34990 }, { "grad_norm": 0.19225436449050903, "learning_rate": 7.734879212067192e-05, "loss": 0.0194, "step": 35000 }, { "grad_norm": 0.14509916305541992, "learning_rate": 7.733494862954573e-05, "loss": 0.0179, "step": 35010 }, { "grad_norm": 0.2552141845226288, "learning_rate": 7.732110214911066e-05, "loss": 0.0173, "step": 35020 }, { "grad_norm": 0.17798998951911926, "learning_rate": 7.730725268088094e-05, "loss": 0.0136, "step": 35030 }, { "grad_norm": 0.12548716366291046, "learning_rate": 7.729340022637112e-05, "loss": 0.0135, "step": 35040 }, { "grad_norm": 0.19575950503349304, "learning_rate": 7.727954478709607e-05, "loss": 0.0155, "step": 35050 }, { "grad_norm": 0.15321309864521027, "learning_rate": 7.726568636457103e-05, "loss": 0.0187, "step": 35060 }, { "grad_norm": 0.18021108210086823, "learning_rate": 7.725182496031153e-05, "loss": 0.0161, "step": 35070 }, { "grad_norm": 0.17586657404899597, "learning_rate": 7.723796057583341e-05, "loss": 0.0172, "step": 35080 }, { "grad_norm": 0.23187881708145142, "learning_rate": 7.722409321265287e-05, "loss": 0.0147, "step": 35090 }, { "grad_norm": 0.1878420114517212, "learning_rate": 7.721022287228645e-05, "loss": 0.0153, "step": 35100 }, { "grad_norm": 0.16743603348731995, "learning_rate": 7.719634955625094e-05, "loss": 0.0181, "step": 35110 }, { "grad_norm": 0.1514972746372223, "learning_rate": 7.718247326606355e-05, "loss": 0.0138, "step": 35120 }, { "grad_norm": 0.19304145872592926, "learning_rate": 7.716859400324175e-05, "loss": 0.0165, "step": 35130 }, { "grad_norm": 0.21635936200618744, "learning_rate": 7.715471176930335e-05, "loss": 0.0163, "step": 35140 }, { "grad_norm": 0.25552481412887573, "learning_rate": 7.714082656576651e-05, "loss": 0.0171, "step": 35150 }, { "grad_norm": 0.21894970536231995, "learning_rate": 7.712693839414968e-05, "loss": 0.0173, "step": 35160 }, { "grad_norm": 0.21233965456485748, "learning_rate": 7.711304725597164e-05, "loss": 0.0182, "step": 35170 }, { "grad_norm": 0.1742287278175354, "learning_rate": 7.709915315275151e-05, "loss": 0.016, "step": 35180 }, { "grad_norm": 0.19331830739974976, "learning_rate": 7.708525608600876e-05, "loss": 0.0143, "step": 35190 }, { "grad_norm": 0.1886080503463745, "learning_rate": 7.707135605726311e-05, "loss": 0.0152, "step": 35200 }, { "grad_norm": 0.1939895898103714, "learning_rate": 7.705745306803466e-05, "loss": 0.0166, "step": 35210 }, { "grad_norm": 0.17002485692501068, "learning_rate": 7.704354711984383e-05, "loss": 0.0164, "step": 35220 }, { "grad_norm": 0.21666957437992096, "learning_rate": 7.702963821421133e-05, "loss": 0.0134, "step": 35230 }, { "grad_norm": 0.18372686207294464, "learning_rate": 7.701572635265826e-05, "loss": 0.0157, "step": 35240 }, { "grad_norm": 0.24029916524887085, "learning_rate": 7.700181153670596e-05, "loss": 0.0158, "step": 35250 }, { "grad_norm": 0.18385183811187744, "learning_rate": 7.698789376787616e-05, "loss": 0.0188, "step": 35260 }, { "grad_norm": 0.24105346202850342, "learning_rate": 7.697397304769085e-05, "loss": 0.0177, "step": 35270 }, { "grad_norm": 0.22675052285194397, "learning_rate": 7.696004937767241e-05, "loss": 0.0168, "step": 35280 }, { "grad_norm": 0.15866145491600037, "learning_rate": 7.694612275934352e-05, "loss": 0.0135, "step": 35290 }, { "grad_norm": 0.1614006757736206, "learning_rate": 7.693219319422714e-05, "loss": 0.0152, "step": 35300 }, { "grad_norm": 0.22118787467479706, "learning_rate": 7.69182606838466e-05, "loss": 0.014, "step": 35310 }, { "grad_norm": 0.1322404444217682, "learning_rate": 7.690432522972558e-05, "loss": 0.0141, "step": 35320 }, { "grad_norm": 0.1870598942041397, "learning_rate": 7.689038683338796e-05, "loss": 0.0185, "step": 35330 }, { "grad_norm": 0.21990695595741272, "learning_rate": 7.687644549635808e-05, "loss": 0.0143, "step": 35340 }, { "grad_norm": 0.16540788114070892, "learning_rate": 7.686250122016053e-05, "loss": 0.0182, "step": 35350 }, { "grad_norm": 0.18683765828609467, "learning_rate": 7.684855400632022e-05, "loss": 0.0186, "step": 35360 }, { "grad_norm": 0.15752795338630676, "learning_rate": 7.683460385636243e-05, "loss": 0.0136, "step": 35370 }, { "grad_norm": 0.16836659610271454, "learning_rate": 7.68206507718127e-05, "loss": 0.0175, "step": 35380 }, { "grad_norm": 0.18447381258010864, "learning_rate": 7.680669475419693e-05, "loss": 0.0153, "step": 35390 }, { "grad_norm": 0.18820367753505707, "learning_rate": 7.679273580504132e-05, "loss": 0.0197, "step": 35400 }, { "grad_norm": 0.16742382943630219, "learning_rate": 7.67787739258724e-05, "loss": 0.0141, "step": 35410 }, { "grad_norm": 0.20694588124752045, "learning_rate": 7.676480911821705e-05, "loss": 0.0152, "step": 35420 }, { "grad_norm": 0.20385362207889557, "learning_rate": 7.675084138360238e-05, "loss": 0.0156, "step": 35430 }, { "grad_norm": 0.1829616278409958, "learning_rate": 7.673687072355592e-05, "loss": 0.0204, "step": 35440 }, { "grad_norm": 0.1907791793346405, "learning_rate": 7.67228971396055e-05, "loss": 0.0165, "step": 35450 }, { "grad_norm": 0.18573573231697083, "learning_rate": 7.670892063327922e-05, "loss": 0.0153, "step": 35460 }, { "grad_norm": 0.21135956048965454, "learning_rate": 7.669494120610552e-05, "loss": 0.0212, "step": 35470 }, { "grad_norm": 0.20329785346984863, "learning_rate": 7.66809588596132e-05, "loss": 0.0172, "step": 35480 }, { "grad_norm": 0.15709999203681946, "learning_rate": 7.666697359533132e-05, "loss": 0.0151, "step": 35490 }, { "grad_norm": 0.2077692151069641, "learning_rate": 7.665298541478932e-05, "loss": 0.0147, "step": 35500 }, { "grad_norm": 0.2196311205625534, "learning_rate": 7.66389943195169e-05, "loss": 0.0194, "step": 35510 }, { "grad_norm": 0.2107093632221222, "learning_rate": 7.66250003110441e-05, "loss": 0.0264, "step": 35520 }, { "grad_norm": 0.2584936320781708, "learning_rate": 7.66110033909013e-05, "loss": 0.0198, "step": 35530 }, { "grad_norm": 0.19802971184253693, "learning_rate": 7.659700356061918e-05, "loss": 0.0162, "step": 35540 }, { "grad_norm": 0.17739540338516235, "learning_rate": 7.658300082172875e-05, "loss": 0.0178, "step": 35550 }, { "grad_norm": 0.18469123542308807, "learning_rate": 7.65689951757613e-05, "loss": 0.0167, "step": 35560 }, { "grad_norm": 0.20300790667533875, "learning_rate": 7.655498662424849e-05, "loss": 0.0148, "step": 35570 }, { "grad_norm": 0.19106024503707886, "learning_rate": 7.654097516872227e-05, "loss": 0.0199, "step": 35580 }, { "grad_norm": 0.18169169127941132, "learning_rate": 7.65269608107149e-05, "loss": 0.0148, "step": 35590 }, { "grad_norm": 0.1678583025932312, "learning_rate": 7.6512943551759e-05, "loss": 0.0181, "step": 35600 }, { "grad_norm": 0.2449265718460083, "learning_rate": 7.649892339338743e-05, "loss": 0.0191, "step": 35610 }, { "grad_norm": 0.17305606603622437, "learning_rate": 7.648490033713344e-05, "loss": 0.0202, "step": 35620 }, { "grad_norm": 0.22126047313213348, "learning_rate": 7.647087438453058e-05, "loss": 0.0182, "step": 35630 }, { "grad_norm": 0.195917010307312, "learning_rate": 7.64568455371127e-05, "loss": 0.0163, "step": 35640 }, { "grad_norm": 0.17120595276355743, "learning_rate": 7.644281379641396e-05, "loss": 0.0162, "step": 35650 }, { "grad_norm": 0.1989583969116211, "learning_rate": 7.642877916396887e-05, "loss": 0.0138, "step": 35660 }, { "grad_norm": 0.19824635982513428, "learning_rate": 7.641474164131221e-05, "loss": 0.0185, "step": 35670 }, { "grad_norm": 0.17983514070510864, "learning_rate": 7.640070122997913e-05, "loss": 0.0166, "step": 35680 }, { "grad_norm": 0.17308862507343292, "learning_rate": 7.638665793150505e-05, "loss": 0.0159, "step": 35690 }, { "grad_norm": 0.1808127909898758, "learning_rate": 7.637261174742574e-05, "loss": 0.0143, "step": 35700 }, { "grad_norm": 0.22896555066108704, "learning_rate": 7.635856267927725e-05, "loss": 0.0184, "step": 35710 }, { "grad_norm": 0.1992187798023224, "learning_rate": 7.634451072859597e-05, "loss": 0.0192, "step": 35720 }, { "grad_norm": 0.19372865557670593, "learning_rate": 7.633045589691863e-05, "loss": 0.0184, "step": 35730 }, { "grad_norm": 0.23892392218112946, "learning_rate": 7.63163981857822e-05, "loss": 0.0219, "step": 35740 }, { "grad_norm": 0.15996429324150085, "learning_rate": 7.630233759672403e-05, "loss": 0.0145, "step": 35750 }, { "grad_norm": 0.18042655289173126, "learning_rate": 7.628827413128177e-05, "loss": 0.0146, "step": 35760 }, { "grad_norm": 0.19707468152046204, "learning_rate": 7.627420779099338e-05, "loss": 0.0161, "step": 35770 }, { "grad_norm": 0.23457899689674377, "learning_rate": 7.626013857739711e-05, "loss": 0.0199, "step": 35780 }, { "grad_norm": 0.19489158689975739, "learning_rate": 7.624606649203158e-05, "loss": 0.0164, "step": 35790 }, { "grad_norm": 0.189494788646698, "learning_rate": 7.623199153643569e-05, "loss": 0.0169, "step": 35800 }, { "grad_norm": 0.18899862468242645, "learning_rate": 7.621791371214863e-05, "loss": 0.0178, "step": 35810 }, { "grad_norm": 0.1938825100660324, "learning_rate": 7.620383302070995e-05, "loss": 0.0176, "step": 35820 }, { "grad_norm": 0.15132559835910797, "learning_rate": 7.61897494636595e-05, "loss": 0.0176, "step": 35830 }, { "grad_norm": 0.1339111179113388, "learning_rate": 7.617566304253739e-05, "loss": 0.019, "step": 35840 }, { "grad_norm": 0.18164817988872528, "learning_rate": 7.616157375888416e-05, "loss": 0.0135, "step": 35850 }, { "grad_norm": 0.16640782356262207, "learning_rate": 7.614748161424053e-05, "loss": 0.0146, "step": 35860 }, { "grad_norm": 0.21270550787448883, "learning_rate": 7.613338661014763e-05, "loss": 0.0166, "step": 35870 }, { "grad_norm": 0.16206619143486023, "learning_rate": 7.611928874814686e-05, "loss": 0.0155, "step": 35880 }, { "grad_norm": 0.20122863352298737, "learning_rate": 7.610518802977996e-05, "loss": 0.0154, "step": 35890 }, { "grad_norm": 0.19027955830097198, "learning_rate": 7.609108445658893e-05, "loss": 0.0133, "step": 35900 }, { "grad_norm": 0.2118074893951416, "learning_rate": 7.607697803011612e-05, "loss": 0.0149, "step": 35910 }, { "grad_norm": 0.16588246822357178, "learning_rate": 7.606286875190421e-05, "loss": 0.0211, "step": 35920 }, { "grad_norm": 0.21498918533325195, "learning_rate": 7.604875662349617e-05, "loss": 0.0188, "step": 35930 }, { "grad_norm": 0.14980454742908478, "learning_rate": 7.603464164643525e-05, "loss": 0.0151, "step": 35940 }, { "grad_norm": 0.16660426557064056, "learning_rate": 7.602052382226507e-05, "loss": 0.018, "step": 35950 }, { "grad_norm": 0.19362974166870117, "learning_rate": 7.600640315252954e-05, "loss": 0.0144, "step": 35960 }, { "grad_norm": 0.15372022986412048, "learning_rate": 7.599227963877284e-05, "loss": 0.0145, "step": 35970 }, { "grad_norm": 0.12081833928823471, "learning_rate": 7.597815328253954e-05, "loss": 0.0161, "step": 35980 }, { "grad_norm": 0.13931764662265778, "learning_rate": 7.596402408537443e-05, "loss": 0.016, "step": 35990 }, { "grad_norm": 0.16945789754390717, "learning_rate": 7.59498920488227e-05, "loss": 0.016, "step": 36000 }, { "grad_norm": 0.15581907331943512, "learning_rate": 7.593575717442979e-05, "loss": 0.0181, "step": 36010 }, { "grad_norm": 0.19747485220432281, "learning_rate": 7.592161946374147e-05, "loss": 0.0176, "step": 36020 }, { "grad_norm": 0.1739702820777893, "learning_rate": 7.590747891830381e-05, "loss": 0.0153, "step": 36030 }, { "grad_norm": 0.16114644706249237, "learning_rate": 7.58933355396632e-05, "loss": 0.0152, "step": 36040 }, { "grad_norm": 0.13549862802028656, "learning_rate": 7.587918932936636e-05, "loss": 0.012, "step": 36050 }, { "grad_norm": 0.2011345624923706, "learning_rate": 7.586504028896028e-05, "loss": 0.0153, "step": 36060 }, { "grad_norm": 0.21395613253116608, "learning_rate": 7.585088841999228e-05, "loss": 0.0155, "step": 36070 }, { "grad_norm": 0.1971568912267685, "learning_rate": 7.583673372400999e-05, "loss": 0.0161, "step": 36080 }, { "grad_norm": 0.2128654271364212, "learning_rate": 7.582257620256134e-05, "loss": 0.0166, "step": 36090 }, { "grad_norm": 0.16690826416015625, "learning_rate": 7.580841585719458e-05, "loss": 0.0167, "step": 36100 }, { "grad_norm": 0.12381012737751007, "learning_rate": 7.579425268945825e-05, "loss": 0.0149, "step": 36110 }, { "grad_norm": 0.1577196568250656, "learning_rate": 7.578008670090127e-05, "loss": 0.0198, "step": 36120 }, { "grad_norm": 0.17788420617580414, "learning_rate": 7.576591789307272e-05, "loss": 0.0209, "step": 36130 }, { "grad_norm": 0.18385867774486542, "learning_rate": 7.575174626752216e-05, "loss": 0.0181, "step": 36140 }, { "grad_norm": 0.20965415239334106, "learning_rate": 7.573757182579934e-05, "loss": 0.0151, "step": 36150 }, { "grad_norm": 0.17772673070430756, "learning_rate": 7.572339456945435e-05, "loss": 0.0149, "step": 36160 }, { "grad_norm": 0.16300959885120392, "learning_rate": 7.570921450003762e-05, "loss": 0.016, "step": 36170 }, { "grad_norm": 0.16606980562210083, "learning_rate": 7.569503161909983e-05, "loss": 0.0156, "step": 36180 }, { "grad_norm": 0.17148204147815704, "learning_rate": 7.568084592819202e-05, "loss": 0.0151, "step": 36190 }, { "grad_norm": 0.14788655936717987, "learning_rate": 7.566665742886551e-05, "loss": 0.0162, "step": 36200 }, { "grad_norm": 0.19189944863319397, "learning_rate": 7.565246612267194e-05, "loss": 0.017, "step": 36210 }, { "grad_norm": 0.1757085770368576, "learning_rate": 7.563827201116325e-05, "loss": 0.0127, "step": 36220 }, { "grad_norm": 0.29274675250053406, "learning_rate": 7.562407509589164e-05, "loss": 0.0207, "step": 36230 }, { "grad_norm": 0.16414833068847656, "learning_rate": 7.560987537840973e-05, "loss": 0.0201, "step": 36240 }, { "grad_norm": 0.1745234876871109, "learning_rate": 7.559567286027036e-05, "loss": 0.0172, "step": 36250 }, { "grad_norm": 0.21342691779136658, "learning_rate": 7.558146754302668e-05, "loss": 0.0172, "step": 36260 }, { "grad_norm": 0.15042851865291595, "learning_rate": 7.556725942823217e-05, "loss": 0.0121, "step": 36270 }, { "grad_norm": 0.2000715434551239, "learning_rate": 7.555304851744061e-05, "loss": 0.0161, "step": 36280 }, { "grad_norm": 0.17464880645275116, "learning_rate": 7.553883481220608e-05, "loss": 0.0165, "step": 36290 }, { "grad_norm": 0.15106888115406036, "learning_rate": 7.552461831408298e-05, "loss": 0.0154, "step": 36300 }, { "grad_norm": 0.14262081682682037, "learning_rate": 7.551039902462599e-05, "loss": 0.0156, "step": 36310 }, { "grad_norm": 0.23619775474071503, "learning_rate": 7.549617694539011e-05, "loss": 0.0145, "step": 36320 }, { "grad_norm": 0.1675100177526474, "learning_rate": 7.548195207793066e-05, "loss": 0.0145, "step": 36330 }, { "grad_norm": 0.14660018682479858, "learning_rate": 7.546772442380323e-05, "loss": 0.0155, "step": 36340 }, { "grad_norm": 0.15199513733386993, "learning_rate": 7.545349398456376e-05, "loss": 0.0162, "step": 36350 }, { "grad_norm": 0.16563014686107635, "learning_rate": 7.543926076176845e-05, "loss": 0.0148, "step": 36360 }, { "grad_norm": 0.20031124353408813, "learning_rate": 7.542502475697385e-05, "loss": 0.0143, "step": 36370 }, { "grad_norm": 0.1378888189792633, "learning_rate": 7.541078597173675e-05, "loss": 0.0177, "step": 36380 }, { "grad_norm": 0.24609850347042084, "learning_rate": 7.539654440761431e-05, "loss": 0.0183, "step": 36390 }, { "grad_norm": 0.20301289856433868, "learning_rate": 7.538230006616395e-05, "loss": 0.0193, "step": 36400 }, { "grad_norm": 0.14264456927776337, "learning_rate": 7.536805294894342e-05, "loss": 0.0145, "step": 36410 }, { "grad_norm": 0.18158864974975586, "learning_rate": 7.535380305751076e-05, "loss": 0.0144, "step": 36420 }, { "grad_norm": 0.19940835237503052, "learning_rate": 7.533955039342431e-05, "loss": 0.0182, "step": 36430 }, { "grad_norm": 0.19775709509849548, "learning_rate": 7.532529495824274e-05, "loss": 0.0156, "step": 36440 }, { "grad_norm": 0.15484975278377533, "learning_rate": 7.531103675352497e-05, "loss": 0.0169, "step": 36450 }, { "grad_norm": 0.20049990713596344, "learning_rate": 7.529677578083029e-05, "loss": 0.0181, "step": 36460 }, { "grad_norm": 0.1800302118062973, "learning_rate": 7.528251204171823e-05, "loss": 0.0191, "step": 36470 }, { "grad_norm": 0.18252059817314148, "learning_rate": 7.526824553774867e-05, "loss": 0.0188, "step": 36480 }, { "grad_norm": 0.3158729672431946, "learning_rate": 7.525397627048178e-05, "loss": 0.0217, "step": 36490 }, { "grad_norm": 0.2321149706840515, "learning_rate": 7.523970424147802e-05, "loss": 0.0155, "step": 36500 }, { "grad_norm": 0.2445218712091446, "learning_rate": 7.522542945229813e-05, "loss": 0.0183, "step": 36510 }, { "grad_norm": 0.18170250952243805, "learning_rate": 7.521115190450321e-05, "loss": 0.017, "step": 36520 }, { "grad_norm": 0.19976766407489777, "learning_rate": 7.519687159965462e-05, "loss": 0.0226, "step": 36530 }, { "grad_norm": 0.19873692095279694, "learning_rate": 7.518258853931403e-05, "loss": 0.0169, "step": 36540 }, { "grad_norm": 0.2562978267669678, "learning_rate": 7.516830272504342e-05, "loss": 0.0162, "step": 36550 }, { "grad_norm": 0.17277497053146362, "learning_rate": 7.515401415840509e-05, "loss": 0.0196, "step": 36560 }, { "grad_norm": 0.17239917814731598, "learning_rate": 7.513972284096155e-05, "loss": 0.0154, "step": 36570 }, { "grad_norm": 0.16839250922203064, "learning_rate": 7.512542877427574e-05, "loss": 0.0154, "step": 36580 }, { "grad_norm": 0.20450954139232635, "learning_rate": 7.51111319599108e-05, "loss": 0.015, "step": 36590 }, { "grad_norm": 0.19810114800930023, "learning_rate": 7.509683239943024e-05, "loss": 0.0153, "step": 36600 }, { "grad_norm": 0.23711007833480835, "learning_rate": 7.50825300943978e-05, "loss": 0.0239, "step": 36610 }, { "grad_norm": 0.20012779533863068, "learning_rate": 7.50682250463776e-05, "loss": 0.0155, "step": 36620 }, { "grad_norm": 0.15969619154930115, "learning_rate": 7.5053917256934e-05, "loss": 0.0144, "step": 36630 }, { "grad_norm": 0.1644509881734848, "learning_rate": 7.503960672763166e-05, "loss": 0.0145, "step": 36640 }, { "grad_norm": 0.139770045876503, "learning_rate": 7.502529346003559e-05, "loss": 0.0161, "step": 36650 }, { "grad_norm": 0.1939622312784195, "learning_rate": 7.501097745571106e-05, "loss": 0.0199, "step": 36660 }, { "grad_norm": 0.19952023029327393, "learning_rate": 7.499665871622361e-05, "loss": 0.0164, "step": 36670 }, { "grad_norm": 0.20450101792812347, "learning_rate": 7.498233724313917e-05, "loss": 0.0167, "step": 36680 }, { "grad_norm": 0.2189088612794876, "learning_rate": 7.496801303802387e-05, "loss": 0.0182, "step": 36690 }, { "grad_norm": 0.15538784861564636, "learning_rate": 7.495368610244423e-05, "loss": 0.016, "step": 36700 }, { "grad_norm": 0.17309853434562683, "learning_rate": 7.493935643796697e-05, "loss": 0.0152, "step": 36710 }, { "grad_norm": 0.22648945450782776, "learning_rate": 7.49250240461592e-05, "loss": 0.0162, "step": 36720 }, { "grad_norm": 0.17926019430160522, "learning_rate": 7.491068892858826e-05, "loss": 0.0134, "step": 36730 }, { "grad_norm": 0.20234069228172302, "learning_rate": 7.489635108682184e-05, "loss": 0.017, "step": 36740 }, { "grad_norm": 0.18731409311294556, "learning_rate": 7.48820105224279e-05, "loss": 0.0145, "step": 36750 }, { "grad_norm": 0.21069779992103577, "learning_rate": 7.48676672369747e-05, "loss": 0.0176, "step": 36760 }, { "grad_norm": 0.1807498335838318, "learning_rate": 7.485332123203079e-05, "loss": 0.015, "step": 36770 }, { "grad_norm": 0.16700869798660278, "learning_rate": 7.483897250916504e-05, "loss": 0.0148, "step": 36780 }, { "grad_norm": 0.18253092467784882, "learning_rate": 7.48246210699466e-05, "loss": 0.0132, "step": 36790 }, { "grad_norm": 0.18501228094100952, "learning_rate": 7.481026691594492e-05, "loss": 0.0144, "step": 36800 }, { "grad_norm": 0.20331978797912598, "learning_rate": 7.479591004872975e-05, "loss": 0.0144, "step": 36810 }, { "grad_norm": 0.20505212247371674, "learning_rate": 7.478155046987114e-05, "loss": 0.0161, "step": 36820 }, { "grad_norm": 0.18823185563087463, "learning_rate": 7.476718818093944e-05, "loss": 0.0138, "step": 36830 }, { "grad_norm": 0.2203456163406372, "learning_rate": 7.475282318350527e-05, "loss": 0.0165, "step": 36840 }, { "grad_norm": 0.1992855668067932, "learning_rate": 7.473845547913957e-05, "loss": 0.0165, "step": 36850 }, { "grad_norm": 0.18188729882240295, "learning_rate": 7.472408506941357e-05, "loss": 0.0195, "step": 36860 }, { "grad_norm": 0.21033862233161926, "learning_rate": 7.470971195589881e-05, "loss": 0.0181, "step": 36870 }, { "grad_norm": 0.19514308869838715, "learning_rate": 7.46953361401671e-05, "loss": 0.0145, "step": 36880 }, { "grad_norm": 0.20951594412326813, "learning_rate": 7.468095762379055e-05, "loss": 0.0163, "step": 36890 }, { "grad_norm": 0.20466606318950653, "learning_rate": 7.466657640834158e-05, "loss": 0.0163, "step": 36900 }, { "grad_norm": 0.1682169884443283, "learning_rate": 7.46521924953929e-05, "loss": 0.0135, "step": 36910 }, { "grad_norm": 0.17820778489112854, "learning_rate": 7.463780588651752e-05, "loss": 0.0149, "step": 36920 }, { "grad_norm": 0.17437095940113068, "learning_rate": 7.462341658328873e-05, "loss": 0.0175, "step": 36930 }, { "grad_norm": 0.20256078243255615, "learning_rate": 7.460902458728012e-05, "loss": 0.014, "step": 36940 }, { "grad_norm": 0.1685517579317093, "learning_rate": 7.459462990006558e-05, "loss": 0.0144, "step": 36950 }, { "grad_norm": 0.20977775752544403, "learning_rate": 7.458023252321929e-05, "loss": 0.0133, "step": 36960 }, { "grad_norm": 0.1816856861114502, "learning_rate": 7.456583245831573e-05, "loss": 0.0167, "step": 36970 }, { "grad_norm": 0.15282757580280304, "learning_rate": 7.455142970692968e-05, "loss": 0.0168, "step": 36980 }, { "grad_norm": 0.10994868725538254, "learning_rate": 7.453702427063615e-05, "loss": 0.0157, "step": 36990 }, { "grad_norm": 0.1989336460828781, "learning_rate": 7.452261615101057e-05, "loss": 0.0162, "step": 37000 }, { "grad_norm": 0.1828743815422058, "learning_rate": 7.450820534962856e-05, "loss": 0.0173, "step": 37010 }, { "grad_norm": 0.10029217600822449, "learning_rate": 7.449379186806604e-05, "loss": 0.0144, "step": 37020 }, { "grad_norm": 0.18246039748191833, "learning_rate": 7.44793757078993e-05, "loss": 0.0151, "step": 37030 }, { "grad_norm": 0.16219162940979004, "learning_rate": 7.44649568707048e-05, "loss": 0.0162, "step": 37040 }, { "grad_norm": 0.1821316033601761, "learning_rate": 7.445053535805942e-05, "loss": 0.0149, "step": 37050 }, { "grad_norm": 0.15322616696357727, "learning_rate": 7.443611117154024e-05, "loss": 0.0145, "step": 37060 }, { "grad_norm": 0.17727328836917877, "learning_rate": 7.442168431272469e-05, "loss": 0.0139, "step": 37070 }, { "grad_norm": 0.18274462223052979, "learning_rate": 7.440725478319044e-05, "loss": 0.0187, "step": 37080 }, { "grad_norm": 0.2582640051841736, "learning_rate": 7.439282258451553e-05, "loss": 0.0167, "step": 37090 }, { "grad_norm": 0.18302516639232635, "learning_rate": 7.43783877182782e-05, "loss": 0.0173, "step": 37100 }, { "grad_norm": 0.2077675461769104, "learning_rate": 7.436395018605707e-05, "loss": 0.0138, "step": 37110 }, { "grad_norm": 0.19044597446918488, "learning_rate": 7.434950998943094e-05, "loss": 0.0162, "step": 37120 }, { "grad_norm": 0.15174616873264313, "learning_rate": 7.433506712997902e-05, "loss": 0.0147, "step": 37130 }, { "grad_norm": 0.13512301445007324, "learning_rate": 7.432062160928077e-05, "loss": 0.0159, "step": 37140 }, { "grad_norm": 0.17048367857933044, "learning_rate": 7.430617342891588e-05, "loss": 0.0133, "step": 37150 }, { "grad_norm": 0.17400027811527252, "learning_rate": 7.429172259046442e-05, "loss": 0.0147, "step": 37160 }, { "grad_norm": 0.23132458329200745, "learning_rate": 7.42772690955067e-05, "loss": 0.0123, "step": 37170 }, { "grad_norm": 0.1667497307062149, "learning_rate": 7.426281294562332e-05, "loss": 0.0168, "step": 37180 }, { "grad_norm": 0.14544790983200073, "learning_rate": 7.424835414239519e-05, "loss": 0.0139, "step": 37190 }, { "grad_norm": 0.17337176203727722, "learning_rate": 7.423389268740352e-05, "loss": 0.0143, "step": 37200 }, { "grad_norm": 0.19762873649597168, "learning_rate": 7.42194285822298e-05, "loss": 0.0175, "step": 37210 }, { "grad_norm": 0.17677390575408936, "learning_rate": 7.420496182845576e-05, "loss": 0.0171, "step": 37220 }, { "grad_norm": 0.17516110837459564, "learning_rate": 7.419049242766348e-05, "loss": 0.0179, "step": 37230 }, { "grad_norm": 0.1916445642709732, "learning_rate": 7.417602038143533e-05, "loss": 0.0177, "step": 37240 }, { "grad_norm": 0.1591489017009735, "learning_rate": 7.416154569135393e-05, "loss": 0.0166, "step": 37250 }, { "grad_norm": 0.1867416501045227, "learning_rate": 7.414706835900222e-05, "loss": 0.0147, "step": 37260 }, { "grad_norm": 0.09770629554986954, "learning_rate": 7.413258838596345e-05, "loss": 0.0138, "step": 37270 }, { "grad_norm": 0.12804438173770905, "learning_rate": 7.411810577382106e-05, "loss": 0.0126, "step": 37280 }, { "grad_norm": 0.24390535056591034, "learning_rate": 7.410362052415889e-05, "loss": 0.0142, "step": 37290 }, { "grad_norm": 0.21630392968654633, "learning_rate": 7.408913263856102e-05, "loss": 0.0136, "step": 37300 }, { "grad_norm": 0.2041902393102646, "learning_rate": 7.407464211861183e-05, "loss": 0.0153, "step": 37310 }, { "grad_norm": 0.13253271579742432, "learning_rate": 7.406014896589597e-05, "loss": 0.0126, "step": 37320 }, { "grad_norm": 0.14280207455158234, "learning_rate": 7.404565318199838e-05, "loss": 0.0148, "step": 37330 }, { "grad_norm": 0.17170089483261108, "learning_rate": 7.403115476850432e-05, "loss": 0.015, "step": 37340 }, { "grad_norm": 0.22070159018039703, "learning_rate": 7.401665372699932e-05, "loss": 0.0144, "step": 37350 }, { "grad_norm": 0.25006812810897827, "learning_rate": 7.400215005906916e-05, "loss": 0.0201, "step": 37360 }, { "grad_norm": 0.2034440040588379, "learning_rate": 7.398764376629996e-05, "loss": 0.0179, "step": 37370 }, { "grad_norm": 0.1424751579761505, "learning_rate": 7.39731348502781e-05, "loss": 0.0153, "step": 37380 }, { "grad_norm": 0.1899486929178238, "learning_rate": 7.395862331259028e-05, "loss": 0.0168, "step": 37390 }, { "grad_norm": 0.14316894114017487, "learning_rate": 7.394410915482341e-05, "loss": 0.0138, "step": 37400 }, { "grad_norm": 0.17545650899410248, "learning_rate": 7.392959237856476e-05, "loss": 0.0152, "step": 37410 }, { "grad_norm": 0.17195187509059906, "learning_rate": 7.391507298540187e-05, "loss": 0.0107, "step": 37420 }, { "grad_norm": 0.1541031002998352, "learning_rate": 7.390055097692257e-05, "loss": 0.0148, "step": 37430 }, { "grad_norm": 0.1646939069032669, "learning_rate": 7.388602635471494e-05, "loss": 0.0129, "step": 37440 }, { "grad_norm": 0.16334308683872223, "learning_rate": 7.387149912036738e-05, "loss": 0.0142, "step": 37450 }, { "grad_norm": 0.19973716139793396, "learning_rate": 7.385696927546854e-05, "loss": 0.019, "step": 37460 }, { "grad_norm": 0.264782190322876, "learning_rate": 7.384243682160743e-05, "loss": 0.0183, "step": 37470 }, { "grad_norm": 0.20608574151992798, "learning_rate": 7.382790176037329e-05, "loss": 0.0147, "step": 37480 }, { "grad_norm": 0.20533013343811035, "learning_rate": 7.381336409335562e-05, "loss": 0.017, "step": 37490 }, { "grad_norm": 0.17607815563678741, "learning_rate": 7.379882382214426e-05, "loss": 0.0141, "step": 37500 }, { "grad_norm": 0.16610954701900482, "learning_rate": 7.378428094832931e-05, "loss": 0.0118, "step": 37510 }, { "grad_norm": 0.18942837417125702, "learning_rate": 7.376973547350114e-05, "loss": 0.0132, "step": 37520 }, { "grad_norm": 0.11237017065286636, "learning_rate": 7.375518739925046e-05, "loss": 0.0134, "step": 37530 }, { "grad_norm": 0.15968602895736694, "learning_rate": 7.374063672716818e-05, "loss": 0.0117, "step": 37540 }, { "grad_norm": 0.1653088629245758, "learning_rate": 7.372608345884558e-05, "loss": 0.0146, "step": 37550 }, { "grad_norm": 0.22199146449565887, "learning_rate": 7.371152759587415e-05, "loss": 0.014, "step": 37560 }, { "grad_norm": 0.2136230766773224, "learning_rate": 7.369696913984572e-05, "loss": 0.0134, "step": 37570 }, { "grad_norm": 0.2094212770462036, "learning_rate": 7.368240809235237e-05, "loss": 0.0143, "step": 37580 }, { "grad_norm": 0.18621571362018585, "learning_rate": 7.366784445498649e-05, "loss": 0.0127, "step": 37590 }, { "grad_norm": 0.20716576278209686, "learning_rate": 7.36532782293407e-05, "loss": 0.0198, "step": 37600 }, { "grad_norm": 0.19961455464363098, "learning_rate": 7.363870941700797e-05, "loss": 0.0146, "step": 37610 }, { "grad_norm": 0.14905019104480743, "learning_rate": 7.362413801958152e-05, "loss": 0.0144, "step": 37620 }, { "grad_norm": 0.18828091025352478, "learning_rate": 7.360956403865486e-05, "loss": 0.0186, "step": 37630 }, { "grad_norm": 0.19484665989875793, "learning_rate": 7.359498747582177e-05, "loss": 0.018, "step": 37640 }, { "grad_norm": 0.13220304250717163, "learning_rate": 7.358040833267634e-05, "loss": 0.0158, "step": 37650 }, { "grad_norm": 0.19472500681877136, "learning_rate": 7.356582661081288e-05, "loss": 0.0144, "step": 37660 }, { "grad_norm": 0.13514544069766998, "learning_rate": 7.355124231182607e-05, "loss": 0.0157, "step": 37670 }, { "grad_norm": 0.23622238636016846, "learning_rate": 7.353665543731083e-05, "loss": 0.0164, "step": 37680 }, { "grad_norm": 0.18060816824436188, "learning_rate": 7.352206598886229e-05, "loss": 0.0151, "step": 37690 }, { "grad_norm": 0.1496572345495224, "learning_rate": 7.350747396807601e-05, "loss": 0.015, "step": 37700 }, { "grad_norm": 0.13895493745803833, "learning_rate": 7.349287937654772e-05, "loss": 0.0145, "step": 37710 }, { "grad_norm": 0.12536711990833282, "learning_rate": 7.347828221587345e-05, "loss": 0.0135, "step": 37720 }, { "grad_norm": 0.15703225135803223, "learning_rate": 7.346368248764955e-05, "loss": 0.0132, "step": 37730 }, { "grad_norm": 0.17761068046092987, "learning_rate": 7.34490801934726e-05, "loss": 0.0125, "step": 37740 }, { "grad_norm": 0.2063017040491104, "learning_rate": 7.343447533493947e-05, "loss": 0.0159, "step": 37750 }, { "grad_norm": 0.14726465940475464, "learning_rate": 7.341986791364739e-05, "loss": 0.0142, "step": 37760 }, { "grad_norm": 0.1585247814655304, "learning_rate": 7.340525793119374e-05, "loss": 0.0217, "step": 37770 }, { "grad_norm": 0.1986979991197586, "learning_rate": 7.33906453891763e-05, "loss": 0.014, "step": 37780 }, { "grad_norm": 0.20360977947711945, "learning_rate": 7.337603028919303e-05, "loss": 0.0136, "step": 37790 }, { "grad_norm": 0.18156129121780396, "learning_rate": 7.336141263284225e-05, "loss": 0.0172, "step": 37800 }, { "grad_norm": 0.16388946771621704, "learning_rate": 7.334679242172249e-05, "loss": 0.0135, "step": 37810 }, { "grad_norm": 0.18751879036426544, "learning_rate": 7.33321696574326e-05, "loss": 0.0157, "step": 37820 }, { "grad_norm": 0.13115744292736053, "learning_rate": 7.331754434157174e-05, "loss": 0.0155, "step": 37830 }, { "grad_norm": 0.21843309700489044, "learning_rate": 7.330291647573928e-05, "loss": 0.0156, "step": 37840 }, { "grad_norm": 0.18919190764427185, "learning_rate": 7.32882860615349e-05, "loss": 0.013, "step": 37850 }, { "grad_norm": 0.22192241251468658, "learning_rate": 7.327365310055858e-05, "loss": 0.0166, "step": 37860 }, { "grad_norm": 0.1899939477443695, "learning_rate": 7.325901759441055e-05, "loss": 0.0134, "step": 37870 }, { "grad_norm": 0.17413519322872162, "learning_rate": 7.324437954469132e-05, "loss": 0.0163, "step": 37880 }, { "grad_norm": 0.19627416133880615, "learning_rate": 7.32297389530017e-05, "loss": 0.0168, "step": 37890 }, { "grad_norm": 0.1506422758102417, "learning_rate": 7.321509582094274e-05, "loss": 0.0143, "step": 37900 }, { "grad_norm": 0.20987574756145477, "learning_rate": 7.320045015011582e-05, "loss": 0.0132, "step": 37910 }, { "grad_norm": 0.21446169912815094, "learning_rate": 7.318580194212253e-05, "loss": 0.0178, "step": 37920 }, { "grad_norm": 0.14570671319961548, "learning_rate": 7.317115119856483e-05, "loss": 0.014, "step": 37930 }, { "grad_norm": 0.19882600009441376, "learning_rate": 7.315649792104485e-05, "loss": 0.0176, "step": 37940 }, { "grad_norm": 0.15255431830883026, "learning_rate": 7.314184211116507e-05, "loss": 0.0185, "step": 37950 }, { "grad_norm": 0.18243065476417542, "learning_rate": 7.312718377052823e-05, "loss": 0.0136, "step": 37960 }, { "grad_norm": 0.14919832348823547, "learning_rate": 7.311252290073737e-05, "loss": 0.0163, "step": 37970 }, { "grad_norm": 0.15729431807994843, "learning_rate": 7.309785950339572e-05, "loss": 0.0162, "step": 37980 }, { "grad_norm": 0.18447060883045197, "learning_rate": 7.308319358010692e-05, "loss": 0.0202, "step": 37990 }, { "grad_norm": 0.1608268767595291, "learning_rate": 7.306852513247474e-05, "loss": 0.0154, "step": 38000 }, { "grad_norm": 0.2143319845199585, "learning_rate": 7.305385416210336e-05, "loss": 0.0164, "step": 38010 }, { "grad_norm": 0.1661931723356247, "learning_rate": 7.303918067059711e-05, "loss": 0.0145, "step": 38020 }, { "grad_norm": 0.17281849682331085, "learning_rate": 7.302450465956075e-05, "loss": 0.016, "step": 38030 }, { "grad_norm": 0.12415821850299835, "learning_rate": 7.300982613059914e-05, "loss": 0.0128, "step": 38040 }, { "grad_norm": 0.1627277433872223, "learning_rate": 7.299514508531757e-05, "loss": 0.0158, "step": 38050 }, { "grad_norm": 0.15406794846057892, "learning_rate": 7.298046152532148e-05, "loss": 0.0166, "step": 38060 }, { "grad_norm": 0.1374559849500656, "learning_rate": 7.296577545221668e-05, "loss": 0.0157, "step": 38070 }, { "grad_norm": 0.13518202304840088, "learning_rate": 7.29510868676092e-05, "loss": 0.014, "step": 38080 }, { "grad_norm": 0.1841762363910675, "learning_rate": 7.293639577310538e-05, "loss": 0.0137, "step": 38090 }, { "grad_norm": 0.1975388526916504, "learning_rate": 7.292170217031179e-05, "loss": 0.016, "step": 38100 }, { "grad_norm": 0.2535940408706665, "learning_rate": 7.290700606083532e-05, "loss": 0.0136, "step": 38110 }, { "grad_norm": 0.16587084531784058, "learning_rate": 7.28923074462831e-05, "loss": 0.014, "step": 38120 }, { "grad_norm": 0.20925602316856384, "learning_rate": 7.287760632826255e-05, "loss": 0.0176, "step": 38130 }, { "grad_norm": 0.19332504272460938, "learning_rate": 7.286290270838139e-05, "loss": 0.0192, "step": 38140 }, { "grad_norm": 0.19675956666469574, "learning_rate": 7.284819658824756e-05, "loss": 0.0184, "step": 38150 }, { "grad_norm": 0.20264671742916107, "learning_rate": 7.283348796946929e-05, "loss": 0.0176, "step": 38160 }, { "grad_norm": 0.1742391735315323, "learning_rate": 7.28187768536551e-05, "loss": 0.0173, "step": 38170 }, { "grad_norm": 0.164285346865654, "learning_rate": 7.280406324241379e-05, "loss": 0.0149, "step": 38180 }, { "grad_norm": 0.239922434091568, "learning_rate": 7.278934713735441e-05, "loss": 0.0187, "step": 38190 }, { "grad_norm": 0.21439336240291595, "learning_rate": 7.277462854008629e-05, "loss": 0.0144, "step": 38200 }, { "grad_norm": 0.2172052562236786, "learning_rate": 7.275990745221904e-05, "loss": 0.0132, "step": 38210 }, { "grad_norm": 0.19848139584064484, "learning_rate": 7.274518387536252e-05, "loss": 0.0177, "step": 38220 }, { "grad_norm": 0.19866158068180084, "learning_rate": 7.273045781112689e-05, "loss": 0.0178, "step": 38230 }, { "grad_norm": 0.19006110727787018, "learning_rate": 7.271572926112257e-05, "loss": 0.0171, "step": 38240 }, { "grad_norm": 0.18152116239070892, "learning_rate": 7.270099822696024e-05, "loss": 0.0133, "step": 38250 }, { "grad_norm": 0.1690383404493332, "learning_rate": 7.26862647102509e-05, "loss": 0.0127, "step": 38260 }, { "grad_norm": 0.1203392818570137, "learning_rate": 7.267152871260573e-05, "loss": 0.0133, "step": 38270 }, { "grad_norm": 0.1439013034105301, "learning_rate": 7.265679023563629e-05, "loss": 0.0149, "step": 38280 }, { "grad_norm": 0.20658664405345917, "learning_rate": 7.264204928095433e-05, "loss": 0.0143, "step": 38290 }, { "grad_norm": 0.17404094338417053, "learning_rate": 7.262730585017188e-05, "loss": 0.0127, "step": 38300 }, { "grad_norm": 0.2083403766155243, "learning_rate": 7.261255994490131e-05, "loss": 0.0136, "step": 38310 }, { "grad_norm": 0.16204164922237396, "learning_rate": 7.259781156675517e-05, "loss": 0.0134, "step": 38320 }, { "grad_norm": 0.17132557928562164, "learning_rate": 7.258306071734633e-05, "loss": 0.0135, "step": 38330 }, { "grad_norm": 0.1903504878282547, "learning_rate": 7.256830739828792e-05, "loss": 0.0149, "step": 38340 }, { "grad_norm": 0.2209458202123642, "learning_rate": 7.255355161119336e-05, "loss": 0.0172, "step": 38350 }, { "grad_norm": 0.22805674374103546, "learning_rate": 7.253879335767628e-05, "loss": 0.014, "step": 38360 }, { "grad_norm": 0.25494515895843506, "learning_rate": 7.252403263935067e-05, "loss": 0.0163, "step": 38370 }, { "grad_norm": 0.19278545677661896, "learning_rate": 7.25092694578307e-05, "loss": 0.0151, "step": 38380 }, { "grad_norm": 0.1548263430595398, "learning_rate": 7.249450381473087e-05, "loss": 0.0156, "step": 38390 }, { "grad_norm": 0.1928790807723999, "learning_rate": 7.247973571166593e-05, "loss": 0.0153, "step": 38400 }, { "grad_norm": 0.17381900548934937, "learning_rate": 7.246496515025089e-05, "loss": 0.0144, "step": 38410 }, { "grad_norm": 0.15928427875041962, "learning_rate": 7.245019213210101e-05, "loss": 0.0142, "step": 38420 }, { "grad_norm": 0.1097499430179596, "learning_rate": 7.24354166588319e-05, "loss": 0.0126, "step": 38430 }, { "grad_norm": 0.15210506319999695, "learning_rate": 7.242063873205935e-05, "loss": 0.0121, "step": 38440 }, { "grad_norm": 0.11583807319402695, "learning_rate": 7.240585835339946e-05, "loss": 0.0142, "step": 38450 }, { "grad_norm": 0.13356128334999084, "learning_rate": 7.239107552446857e-05, "loss": 0.0129, "step": 38460 }, { "grad_norm": 0.16898003220558167, "learning_rate": 7.237629024688334e-05, "loss": 0.0139, "step": 38470 }, { "grad_norm": 0.15910117328166962, "learning_rate": 7.236150252226064e-05, "loss": 0.0168, "step": 38480 }, { "grad_norm": 0.18060006201267242, "learning_rate": 7.234671235221765e-05, "loss": 0.0183, "step": 38490 }, { "grad_norm": 0.14247126877307892, "learning_rate": 7.233191973837179e-05, "loss": 0.0172, "step": 38500 }, { "grad_norm": 0.16082863509655, "learning_rate": 7.231712468234078e-05, "loss": 0.0127, "step": 38510 }, { "grad_norm": 0.19085748493671417, "learning_rate": 7.230232718574254e-05, "loss": 0.013, "step": 38520 }, { "grad_norm": 0.2092733532190323, "learning_rate": 7.228752725019535e-05, "loss": 0.0173, "step": 38530 }, { "grad_norm": 0.24911034107208252, "learning_rate": 7.227272487731769e-05, "loss": 0.0129, "step": 38540 }, { "grad_norm": 0.14356271922588348, "learning_rate": 7.225792006872831e-05, "loss": 0.0151, "step": 38550 }, { "grad_norm": 0.19413341581821442, "learning_rate": 7.224311282604628e-05, "loss": 0.0174, "step": 38560 }, { "grad_norm": 0.18822868168354034, "learning_rate": 7.222830315089085e-05, "loss": 0.0167, "step": 38570 }, { "grad_norm": 0.16477754712104797, "learning_rate": 7.22134910448816e-05, "loss": 0.0133, "step": 38580 }, { "grad_norm": 0.19383014738559723, "learning_rate": 7.219867650963839e-05, "loss": 0.0118, "step": 38590 }, { "grad_norm": 0.15795645117759705, "learning_rate": 7.21838595467813e-05, "loss": 0.015, "step": 38600 }, { "grad_norm": 0.1575373113155365, "learning_rate": 7.216904015793067e-05, "loss": 0.0126, "step": 38610 }, { "grad_norm": 0.1583859771490097, "learning_rate": 7.215421834470713e-05, "loss": 0.0128, "step": 38620 }, { "grad_norm": 0.1839040368795395, "learning_rate": 7.213939410873158e-05, "loss": 0.0159, "step": 38630 }, { "grad_norm": 0.1743564009666443, "learning_rate": 7.212456745162519e-05, "loss": 0.014, "step": 38640 }, { "grad_norm": 0.17270757257938385, "learning_rate": 7.210973837500937e-05, "loss": 0.019, "step": 38650 }, { "grad_norm": 0.1499713957309723, "learning_rate": 7.209490688050578e-05, "loss": 0.0145, "step": 38660 }, { "grad_norm": 0.18998222053050995, "learning_rate": 7.208007296973641e-05, "loss": 0.0156, "step": 38670 }, { "grad_norm": 0.1797422468662262, "learning_rate": 7.206523664432345e-05, "loss": 0.0149, "step": 38680 }, { "grad_norm": 0.1707894653081894, "learning_rate": 7.205039790588939e-05, "loss": 0.016, "step": 38690 }, { "grad_norm": 0.14697116613388062, "learning_rate": 7.203555675605697e-05, "loss": 0.0131, "step": 38700 }, { "grad_norm": 0.16296333074569702, "learning_rate": 7.202071319644917e-05, "loss": 0.0133, "step": 38710 }, { "grad_norm": 0.2281772941350937, "learning_rate": 7.200586722868932e-05, "loss": 0.0177, "step": 38720 }, { "grad_norm": 0.1656883805990219, "learning_rate": 7.199101885440088e-05, "loss": 0.0126, "step": 38730 }, { "grad_norm": 0.11252904683351517, "learning_rate": 7.197616807520771e-05, "loss": 0.0135, "step": 38740 }, { "grad_norm": 0.13600707054138184, "learning_rate": 7.196131489273381e-05, "loss": 0.0156, "step": 38750 }, { "grad_norm": 0.14279408752918243, "learning_rate": 7.194645930860355e-05, "loss": 0.0126, "step": 38760 }, { "grad_norm": 0.21315421164035797, "learning_rate": 7.19316013244415e-05, "loss": 0.0117, "step": 38770 }, { "grad_norm": 0.1454250067472458, "learning_rate": 7.191674094187248e-05, "loss": 0.018, "step": 38780 }, { "grad_norm": 0.17635302245616913, "learning_rate": 7.190187816252165e-05, "loss": 0.0167, "step": 38790 }, { "grad_norm": 0.1674346625804901, "learning_rate": 7.188701298801435e-05, "loss": 0.0136, "step": 38800 }, { "grad_norm": 0.1499079465866089, "learning_rate": 7.18721454199762e-05, "loss": 0.018, "step": 38810 }, { "grad_norm": 0.20064839720726013, "learning_rate": 7.185727546003312e-05, "loss": 0.0141, "step": 38820 }, { "grad_norm": 0.20030102133750916, "learning_rate": 7.184240310981126e-05, "loss": 0.0159, "step": 38830 }, { "grad_norm": 0.1899324357509613, "learning_rate": 7.182752837093704e-05, "loss": 0.0132, "step": 38840 }, { "grad_norm": 0.16056552529335022, "learning_rate": 7.181265124503711e-05, "loss": 0.0147, "step": 38850 }, { "grad_norm": 0.15553714334964752, "learning_rate": 7.179777173373847e-05, "loss": 0.0142, "step": 38860 }, { "grad_norm": 0.14132115244865417, "learning_rate": 7.178288983866826e-05, "loss": 0.0139, "step": 38870 }, { "grad_norm": 0.16078943014144897, "learning_rate": 7.176800556145397e-05, "loss": 0.0135, "step": 38880 }, { "grad_norm": 0.19221262633800507, "learning_rate": 7.175311890372334e-05, "loss": 0.0141, "step": 38890 }, { "grad_norm": 0.2078615128993988, "learning_rate": 7.17382298671043e-05, "loss": 0.016, "step": 38900 }, { "grad_norm": 0.16784977912902832, "learning_rate": 7.172333845322515e-05, "loss": 0.0135, "step": 38910 }, { "grad_norm": 0.1494358628988266, "learning_rate": 7.170844466371436e-05, "loss": 0.0128, "step": 38920 }, { "grad_norm": 0.17842374742031097, "learning_rate": 7.16935485002007e-05, "loss": 0.0186, "step": 38930 }, { "grad_norm": 0.16906635463237762, "learning_rate": 7.167864996431319e-05, "loss": 0.0132, "step": 38940 }, { "grad_norm": 0.1980452537536621, "learning_rate": 7.166374905768111e-05, "loss": 0.0147, "step": 38950 }, { "grad_norm": 0.16840609908103943, "learning_rate": 7.164884578193401e-05, "loss": 0.0165, "step": 38960 }, { "grad_norm": 0.13214880228042603, "learning_rate": 7.163394013870168e-05, "loss": 0.0141, "step": 38970 }, { "grad_norm": 0.18616318702697754, "learning_rate": 7.161903212961419e-05, "loss": 0.0147, "step": 38980 }, { "grad_norm": 0.16640707850456238, "learning_rate": 7.160412175630183e-05, "loss": 0.0136, "step": 38990 }, { "grad_norm": 0.15065471827983856, "learning_rate": 7.158920902039521e-05, "loss": 0.0121, "step": 39000 }, { "grad_norm": 0.15632984042167664, "learning_rate": 7.157429392352514e-05, "loss": 0.0171, "step": 39010 }, { "grad_norm": 0.1854151040315628, "learning_rate": 7.155937646732275e-05, "loss": 0.0184, "step": 39020 }, { "grad_norm": 0.14310944080352783, "learning_rate": 7.154445665341933e-05, "loss": 0.0149, "step": 39030 }, { "grad_norm": 0.18149681389331818, "learning_rate": 7.152953448344654e-05, "loss": 0.0141, "step": 39040 }, { "grad_norm": 0.1835036426782608, "learning_rate": 7.151460995903624e-05, "loss": 0.0191, "step": 39050 }, { "grad_norm": 0.22214172780513763, "learning_rate": 7.149968308182052e-05, "loss": 0.0146, "step": 39060 }, { "grad_norm": 0.12175330519676208, "learning_rate": 7.14847538534318e-05, "loss": 0.0125, "step": 39070 }, { "grad_norm": 0.13107658922672272, "learning_rate": 7.14698222755027e-05, "loss": 0.0158, "step": 39080 }, { "grad_norm": 0.15777559578418732, "learning_rate": 7.14548883496661e-05, "loss": 0.0156, "step": 39090 }, { "grad_norm": 0.18696284294128418, "learning_rate": 7.143995207755517e-05, "loss": 0.017, "step": 39100 }, { "grad_norm": 0.17344705760478973, "learning_rate": 7.142501346080333e-05, "loss": 0.0128, "step": 39110 }, { "grad_norm": 0.22537371516227722, "learning_rate": 7.141007250104421e-05, "loss": 0.0165, "step": 39120 }, { "grad_norm": 0.18866786360740662, "learning_rate": 7.139512919991176e-05, "loss": 0.0142, "step": 39130 }, { "grad_norm": 0.2069491147994995, "learning_rate": 7.138018355904014e-05, "loss": 0.0168, "step": 39140 }, { "grad_norm": 0.1383456140756607, "learning_rate": 7.13652355800638e-05, "loss": 0.013, "step": 39150 }, { "grad_norm": 0.15724262595176697, "learning_rate": 7.135028526461743e-05, "loss": 0.0113, "step": 39160 }, { "grad_norm": 0.16506922245025635, "learning_rate": 7.133533261433594e-05, "loss": 0.0155, "step": 39170 }, { "grad_norm": 0.21061097085475922, "learning_rate": 7.132037763085457e-05, "loss": 0.0185, "step": 39180 }, { "grad_norm": 0.210066556930542, "learning_rate": 7.130542031580875e-05, "loss": 0.017, "step": 39190 }, { "grad_norm": 0.19013550877571106, "learning_rate": 7.12904606708342e-05, "loss": 0.0166, "step": 39200 }, { "grad_norm": 0.1679384410381317, "learning_rate": 7.127549869756687e-05, "loss": 0.0144, "step": 39210 }, { "grad_norm": 0.13847759366035461, "learning_rate": 7.126053439764299e-05, "loss": 0.0124, "step": 39220 }, { "grad_norm": 0.19394467771053314, "learning_rate": 7.124556777269904e-05, "loss": 0.0141, "step": 39230 }, { "grad_norm": 0.17207777500152588, "learning_rate": 7.123059882437174e-05, "loss": 0.0156, "step": 39240 }, { "grad_norm": 0.2083183377981186, "learning_rate": 7.121562755429807e-05, "loss": 0.0139, "step": 39250 }, { "grad_norm": 0.21018829941749573, "learning_rate": 7.120065396411528e-05, "loss": 0.0142, "step": 39260 }, { "grad_norm": 0.17962585389614105, "learning_rate": 7.118567805546084e-05, "loss": 0.0179, "step": 39270 }, { "grad_norm": 0.1589462012052536, "learning_rate": 7.117069982997248e-05, "loss": 0.0171, "step": 39280 }, { "grad_norm": 0.1704418659210205, "learning_rate": 7.115571928928825e-05, "loss": 0.0148, "step": 39290 }, { "grad_norm": 0.18443557620048523, "learning_rate": 7.114073643504635e-05, "loss": 0.0131, "step": 39300 }, { "grad_norm": 0.16026999056339264, "learning_rate": 7.11257512688853e-05, "loss": 0.0142, "step": 39310 }, { "grad_norm": 0.21636943519115448, "learning_rate": 7.111076379244384e-05, "loss": 0.0128, "step": 39320 }, { "grad_norm": 0.21587727963924408, "learning_rate": 7.109577400736101e-05, "loss": 0.0116, "step": 39330 }, { "grad_norm": 0.20393487811088562, "learning_rate": 7.108078191527605e-05, "loss": 0.016, "step": 39340 }, { "grad_norm": 0.11079325526952744, "learning_rate": 7.106578751782847e-05, "loss": 0.0169, "step": 39350 }, { "grad_norm": 0.14248117804527283, "learning_rate": 7.105079081665803e-05, "loss": 0.0157, "step": 39360 }, { "grad_norm": 0.16553336381912231, "learning_rate": 7.103579181340476e-05, "loss": 0.0145, "step": 39370 }, { "grad_norm": 0.2398027628660202, "learning_rate": 7.102079050970893e-05, "loss": 0.0138, "step": 39380 }, { "grad_norm": 0.1889253556728363, "learning_rate": 7.100578690721105e-05, "loss": 0.0116, "step": 39390 }, { "grad_norm": 0.15213161706924438, "learning_rate": 7.09907810075519e-05, "loss": 0.0129, "step": 39400 }, { "grad_norm": 0.21060819923877716, "learning_rate": 7.097577281237249e-05, "loss": 0.0187, "step": 39410 }, { "grad_norm": 0.1824977844953537, "learning_rate": 7.09607623233141e-05, "loss": 0.0136, "step": 39420 }, { "grad_norm": 0.19484439492225647, "learning_rate": 7.094574954201827e-05, "loss": 0.013, "step": 39430 }, { "grad_norm": 0.21092356741428375, "learning_rate": 7.093073447012675e-05, "loss": 0.017, "step": 39440 }, { "grad_norm": 0.16899731755256653, "learning_rate": 7.09157171092816e-05, "loss": 0.0192, "step": 39450 }, { "grad_norm": 0.12499921768903732, "learning_rate": 7.090069746112504e-05, "loss": 0.015, "step": 39460 }, { "grad_norm": 0.14102093875408173, "learning_rate": 7.088567552729965e-05, "loss": 0.0151, "step": 39470 }, { "grad_norm": 0.17280758917331696, "learning_rate": 7.087065130944818e-05, "loss": 0.0148, "step": 39480 }, { "grad_norm": 0.2671600580215454, "learning_rate": 7.085562480921366e-05, "loss": 0.0152, "step": 39490 }, { "grad_norm": 0.16037485003471375, "learning_rate": 7.084059602823937e-05, "loss": 0.0141, "step": 39500 }, { "grad_norm": 0.21334294974803925, "learning_rate": 7.082556496816882e-05, "loss": 0.012, "step": 39510 }, { "grad_norm": 0.2092515230178833, "learning_rate": 7.081053163064582e-05, "loss": 0.0141, "step": 39520 }, { "grad_norm": 0.14979538321495056, "learning_rate": 7.079549601731434e-05, "loss": 0.0156, "step": 39530 }, { "grad_norm": 0.1913803070783615, "learning_rate": 7.07804581298187e-05, "loss": 0.0153, "step": 39540 }, { "grad_norm": 0.17155657708644867, "learning_rate": 7.07654179698034e-05, "loss": 0.0125, "step": 39550 }, { "grad_norm": 0.14481277763843536, "learning_rate": 7.075037553891321e-05, "loss": 0.0134, "step": 39560 }, { "grad_norm": 0.18830308318138123, "learning_rate": 7.073533083879315e-05, "loss": 0.0143, "step": 39570 }, { "grad_norm": 0.19792555272579193, "learning_rate": 7.072028387108849e-05, "loss": 0.0145, "step": 39580 }, { "grad_norm": 0.11505711078643799, "learning_rate": 7.070523463744472e-05, "loss": 0.014, "step": 39590 }, { "grad_norm": 0.1467093974351883, "learning_rate": 7.069018313950763e-05, "loss": 0.012, "step": 39600 }, { "grad_norm": 0.1873200684785843, "learning_rate": 7.067512937892322e-05, "loss": 0.0149, "step": 39610 }, { "grad_norm": 0.15513058006763458, "learning_rate": 7.066007335733775e-05, "loss": 0.0144, "step": 39620 }, { "grad_norm": 0.15523406863212585, "learning_rate": 7.064501507639772e-05, "loss": 0.0168, "step": 39630 }, { "grad_norm": 0.20535492897033691, "learning_rate": 7.062995453774987e-05, "loss": 0.0146, "step": 39640 }, { "grad_norm": 0.17294271290302277, "learning_rate": 7.061489174304121e-05, "loss": 0.0156, "step": 39650 }, { "grad_norm": 0.2007087618112564, "learning_rate": 7.0599826693919e-05, "loss": 0.0176, "step": 39660 }, { "grad_norm": 0.15949387848377228, "learning_rate": 7.05847593920307e-05, "loss": 0.0141, "step": 39670 }, { "grad_norm": 0.19678883254528046, "learning_rate": 7.056968983902406e-05, "loss": 0.017, "step": 39680 }, { "grad_norm": 0.14081689715385437, "learning_rate": 7.055461803654706e-05, "loss": 0.0139, "step": 39690 }, { "grad_norm": 0.21952086687088013, "learning_rate": 7.053954398624794e-05, "loss": 0.0178, "step": 39700 }, { "grad_norm": 0.20598849654197693, "learning_rate": 7.052446768977518e-05, "loss": 0.0164, "step": 39710 }, { "grad_norm": 0.22740775346755981, "learning_rate": 7.050938914877748e-05, "loss": 0.0139, "step": 39720 }, { "grad_norm": 0.17205356061458588, "learning_rate": 7.049430836490382e-05, "loss": 0.0152, "step": 39730 }, { "grad_norm": 0.13717052340507507, "learning_rate": 7.047922533980342e-05, "loss": 0.0167, "step": 39740 }, { "grad_norm": 0.15678712725639343, "learning_rate": 7.046414007512571e-05, "loss": 0.018, "step": 39750 }, { "grad_norm": 0.1899331957101822, "learning_rate": 7.044905257252042e-05, "loss": 0.0147, "step": 39760 }, { "grad_norm": 0.16768930852413177, "learning_rate": 7.04339628336375e-05, "loss": 0.012, "step": 39770 }, { "grad_norm": 0.20526674389839172, "learning_rate": 7.041887086012711e-05, "loss": 0.0177, "step": 39780 }, { "grad_norm": 0.20868735015392303, "learning_rate": 7.040377665363969e-05, "loss": 0.0155, "step": 39790 }, { "grad_norm": 0.18051567673683167, "learning_rate": 7.038868021582594e-05, "loss": 0.0153, "step": 39800 }, { "grad_norm": 0.22346101701259613, "learning_rate": 7.037358154833679e-05, "loss": 0.0141, "step": 39810 }, { "grad_norm": 0.1886085420846939, "learning_rate": 7.035848065282339e-05, "loss": 0.0226, "step": 39820 }, { "grad_norm": 0.15885291993618011, "learning_rate": 7.034337753093714e-05, "loss": 0.0162, "step": 39830 }, { "grad_norm": 0.1795164942741394, "learning_rate": 7.032827218432972e-05, "loss": 0.0143, "step": 39840 }, { "grad_norm": 0.14468465745449066, "learning_rate": 7.031316461465302e-05, "loss": 0.0127, "step": 39850 }, { "grad_norm": 0.156626358628273, "learning_rate": 7.029805482355915e-05, "loss": 0.0134, "step": 39860 }, { "grad_norm": 0.2521341145038605, "learning_rate": 7.028294281270055e-05, "loss": 0.0146, "step": 39870 }, { "grad_norm": 0.15712663531303406, "learning_rate": 7.026782858372982e-05, "loss": 0.0142, "step": 39880 }, { "grad_norm": 0.21166402101516724, "learning_rate": 7.025271213829982e-05, "loss": 0.0121, "step": 39890 }, { "grad_norm": 0.1467774510383606, "learning_rate": 7.023759347806366e-05, "loss": 0.0126, "step": 39900 }, { "grad_norm": 0.14071723818778992, "learning_rate": 7.022247260467472e-05, "loss": 0.0142, "step": 39910 }, { "grad_norm": 0.2606651782989502, "learning_rate": 7.020734951978658e-05, "loss": 0.0176, "step": 39920 }, { "grad_norm": 0.22151844203472137, "learning_rate": 7.019222422505307e-05, "loss": 0.0144, "step": 39930 }, { "grad_norm": 0.22396226227283478, "learning_rate": 7.017709672212827e-05, "loss": 0.0193, "step": 39940 }, { "grad_norm": 0.20206275582313538, "learning_rate": 7.016196701266652e-05, "loss": 0.014, "step": 39950 }, { "grad_norm": 0.18728268146514893, "learning_rate": 7.014683509832238e-05, "loss": 0.0155, "step": 39960 }, { "grad_norm": 0.19477888941764832, "learning_rate": 7.013170098075063e-05, "loss": 0.015, "step": 39970 }, { "grad_norm": 0.20235469937324524, "learning_rate": 7.011656466160632e-05, "loss": 0.0123, "step": 39980 }, { "grad_norm": 0.2001236081123352, "learning_rate": 7.010142614254475e-05, "loss": 0.017, "step": 39990 }, { "grad_norm": 0.15086741745471954, "learning_rate": 7.008628542522147e-05, "loss": 0.0126, "step": 40000 }, { "grad_norm": 0.17349347472190857, "learning_rate": 7.007114251129218e-05, "loss": 0.0192, "step": 40010 }, { "grad_norm": 0.19901171326637268, "learning_rate": 7.005599740241293e-05, "loss": 0.0144, "step": 40020 }, { "grad_norm": 0.19263747334480286, "learning_rate": 7.004085010023996e-05, "loss": 0.0144, "step": 40030 }, { "grad_norm": 0.12644152343273163, "learning_rate": 7.002570060642976e-05, "loss": 0.0162, "step": 40040 }, { "grad_norm": 0.16417977213859558, "learning_rate": 7.001054892263903e-05, "loss": 0.0175, "step": 40050 }, { "grad_norm": 0.21707285940647125, "learning_rate": 6.999539505052477e-05, "loss": 0.0146, "step": 40060 }, { "grad_norm": 0.17041409015655518, "learning_rate": 6.998023899174418e-05, "loss": 0.0149, "step": 40070 }, { "grad_norm": 0.1644403487443924, "learning_rate": 6.996508074795467e-05, "loss": 0.0161, "step": 40080 }, { "grad_norm": 0.15762285888195038, "learning_rate": 6.994992032081396e-05, "loss": 0.0129, "step": 40090 }, { "grad_norm": 0.11986733227968216, "learning_rate": 6.993475771197995e-05, "loss": 0.0129, "step": 40100 }, { "grad_norm": 0.1751406490802765, "learning_rate": 6.991959292311082e-05, "loss": 0.0139, "step": 40110 }, { "grad_norm": 0.14561443030834198, "learning_rate": 6.990442595586495e-05, "loss": 0.0156, "step": 40120 }, { "grad_norm": 0.205562025308609, "learning_rate": 6.988925681190098e-05, "loss": 0.0151, "step": 40130 }, { "grad_norm": 0.18797113001346588, "learning_rate": 6.987408549287778e-05, "loss": 0.0144, "step": 40140 }, { "grad_norm": 0.2350429743528366, "learning_rate": 6.985891200045449e-05, "loss": 0.0144, "step": 40150 }, { "grad_norm": 0.17804212868213654, "learning_rate": 6.984373633629045e-05, "loss": 0.0127, "step": 40160 }, { "grad_norm": 0.22716283798217773, "learning_rate": 6.98285585020452e-05, "loss": 0.0182, "step": 40170 }, { "grad_norm": 0.25447598099708557, "learning_rate": 6.981337849937864e-05, "loss": 0.0159, "step": 40180 }, { "grad_norm": 0.1592344343662262, "learning_rate": 6.979819632995078e-05, "loss": 0.0184, "step": 40190 }, { "grad_norm": 0.2048320323228836, "learning_rate": 6.978301199542193e-05, "loss": 0.0201, "step": 40200 }, { "grad_norm": 0.190250962972641, "learning_rate": 6.976782549745263e-05, "loss": 0.0196, "step": 40210 }, { "grad_norm": 0.19468438625335693, "learning_rate": 6.975263683770367e-05, "loss": 0.0165, "step": 40220 }, { "grad_norm": 0.1802154928445816, "learning_rate": 6.973744601783604e-05, "loss": 0.0167, "step": 40230 }, { "grad_norm": 0.20508863031864166, "learning_rate": 6.972225303951097e-05, "loss": 0.0207, "step": 40240 }, { "grad_norm": 0.23929333686828613, "learning_rate": 6.970705790438998e-05, "loss": 0.0167, "step": 40250 }, { "grad_norm": 0.2043962925672531, "learning_rate": 6.969186061413477e-05, "loss": 0.0158, "step": 40260 }, { "grad_norm": 0.17098838090896606, "learning_rate": 6.967666117040727e-05, "loss": 0.0163, "step": 40270 }, { "grad_norm": 0.18476812541484833, "learning_rate": 6.966145957486972e-05, "loss": 0.013, "step": 40280 }, { "grad_norm": 0.17559672892093658, "learning_rate": 6.964625582918449e-05, "loss": 0.0127, "step": 40290 }, { "grad_norm": 0.2139684408903122, "learning_rate": 6.963104993501425e-05, "loss": 0.0232, "step": 40300 }, { "grad_norm": 0.1610962152481079, "learning_rate": 6.961584189402192e-05, "loss": 0.0114, "step": 40310 }, { "grad_norm": 0.16231507062911987, "learning_rate": 6.96006317078706e-05, "loss": 0.0164, "step": 40320 }, { "grad_norm": 0.14778287708759308, "learning_rate": 6.958541937822367e-05, "loss": 0.0117, "step": 40330 }, { "grad_norm": 0.1803324818611145, "learning_rate": 6.957020490674473e-05, "loss": 0.015, "step": 40340 }, { "grad_norm": 0.17992335557937622, "learning_rate": 6.95549882950976e-05, "loss": 0.0132, "step": 40350 }, { "grad_norm": 0.1447915881872177, "learning_rate": 6.953976954494635e-05, "loss": 0.0138, "step": 40360 }, { "grad_norm": 0.1333901584148407, "learning_rate": 6.952454865795528e-05, "loss": 0.0149, "step": 40370 }, { "grad_norm": 0.14873266220092773, "learning_rate": 6.950932563578892e-05, "loss": 0.0134, "step": 40380 }, { "grad_norm": 0.173557847738266, "learning_rate": 6.949410048011206e-05, "loss": 0.012, "step": 40390 }, { "grad_norm": 0.13996417820453644, "learning_rate": 6.947887319258966e-05, "loss": 0.0126, "step": 40400 }, { "grad_norm": 0.15297897160053253, "learning_rate": 6.946364377488696e-05, "loss": 0.0147, "step": 40410 }, { "grad_norm": 0.17498740553855896, "learning_rate": 6.944841222866947e-05, "loss": 0.0117, "step": 40420 }, { "grad_norm": 0.12505704164505005, "learning_rate": 6.943317855560284e-05, "loss": 0.0124, "step": 40430 }, { "grad_norm": 0.16608482599258423, "learning_rate": 6.941794275735302e-05, "loss": 0.0176, "step": 40440 }, { "grad_norm": 0.20296555757522583, "learning_rate": 6.94027048355862e-05, "loss": 0.0117, "step": 40450 }, { "grad_norm": 0.11669987440109253, "learning_rate": 6.938746479196871e-05, "loss": 0.0115, "step": 40460 }, { "grad_norm": 0.14669467508792877, "learning_rate": 6.937222262816724e-05, "loss": 0.0098, "step": 40470 }, { "grad_norm": 0.21057482063770294, "learning_rate": 6.935697834584865e-05, "loss": 0.0159, "step": 40480 }, { "grad_norm": 0.21327327191829681, "learning_rate": 6.934173194667998e-05, "loss": 0.0171, "step": 40490 }, { "grad_norm": 0.1574733406305313, "learning_rate": 6.93264834323286e-05, "loss": 0.015, "step": 40500 }, { "grad_norm": 0.2690141201019287, "learning_rate": 6.931123280446206e-05, "loss": 0.0185, "step": 40510 }, { "grad_norm": 0.2122565060853958, "learning_rate": 6.929598006474811e-05, "loss": 0.0183, "step": 40520 }, { "grad_norm": 0.26415205001831055, "learning_rate": 6.92807252148548e-05, "loss": 0.0176, "step": 40530 }, { "grad_norm": 0.23080775141716003, "learning_rate": 6.926546825645037e-05, "loss": 0.0144, "step": 40540 }, { "grad_norm": 0.2438730001449585, "learning_rate": 6.92502091912033e-05, "loss": 0.0158, "step": 40550 }, { "grad_norm": 0.19292575120925903, "learning_rate": 6.923494802078229e-05, "loss": 0.0152, "step": 40560 }, { "grad_norm": 0.19795741140842438, "learning_rate": 6.921968474685628e-05, "loss": 0.0153, "step": 40570 }, { "grad_norm": 0.23410753905773163, "learning_rate": 6.920441937109445e-05, "loss": 0.0157, "step": 40580 }, { "grad_norm": 0.15332479774951935, "learning_rate": 6.91891518951662e-05, "loss": 0.0145, "step": 40590 }, { "grad_norm": 0.17377690970897675, "learning_rate": 6.917388232074114e-05, "loss": 0.0125, "step": 40600 }, { "grad_norm": 0.1592298001050949, "learning_rate": 6.915861064948914e-05, "loss": 0.0142, "step": 40610 }, { "grad_norm": 0.1480342447757721, "learning_rate": 6.914333688308025e-05, "loss": 0.0181, "step": 40620 }, { "grad_norm": 0.16118589043617249, "learning_rate": 6.912806102318487e-05, "loss": 0.0131, "step": 40630 }, { "grad_norm": 0.21114666759967804, "learning_rate": 6.911278307147347e-05, "loss": 0.0147, "step": 40640 }, { "grad_norm": 0.18598003685474396, "learning_rate": 6.909750302961684e-05, "loss": 0.0158, "step": 40650 }, { "grad_norm": 0.1833563894033432, "learning_rate": 6.9082220899286e-05, "loss": 0.0156, "step": 40660 }, { "grad_norm": 0.18414536118507385, "learning_rate": 6.906693668215219e-05, "loss": 0.0185, "step": 40670 }, { "grad_norm": 0.16063328087329865, "learning_rate": 6.905165037988683e-05, "loss": 0.0123, "step": 40680 }, { "grad_norm": 0.1318253129720688, "learning_rate": 6.903636199416164e-05, "loss": 0.015, "step": 40690 }, { "grad_norm": 0.18638145923614502, "learning_rate": 6.902107152664851e-05, "loss": 0.0182, "step": 40700 }, { "grad_norm": 0.1945021152496338, "learning_rate": 6.90057789790196e-05, "loss": 0.0178, "step": 40710 }, { "grad_norm": 0.17821229994297028, "learning_rate": 6.899048435294728e-05, "loss": 0.0138, "step": 40720 }, { "grad_norm": 0.17823725938796997, "learning_rate": 6.897518765010415e-05, "loss": 0.0149, "step": 40730 }, { "grad_norm": 0.15086062252521515, "learning_rate": 6.895988887216302e-05, "loss": 0.0159, "step": 40740 }, { "grad_norm": 0.18085452914237976, "learning_rate": 6.894458802079694e-05, "loss": 0.0136, "step": 40750 }, { "grad_norm": 0.18280398845672607, "learning_rate": 6.892928509767922e-05, "loss": 0.018, "step": 40760 }, { "grad_norm": 0.1627969592809677, "learning_rate": 6.891398010448333e-05, "loss": 0.0162, "step": 40770 }, { "grad_norm": 0.18438954651355743, "learning_rate": 6.8898673042883e-05, "loss": 0.0186, "step": 40780 }, { "grad_norm": 0.19467757642269135, "learning_rate": 6.888336391455222e-05, "loss": 0.0137, "step": 40790 }, { "grad_norm": 0.11062289029359818, "learning_rate": 6.886805272116513e-05, "loss": 0.0187, "step": 40800 }, { "grad_norm": 0.2041511982679367, "learning_rate": 6.885273946439617e-05, "loss": 0.0176, "step": 40810 }, { "grad_norm": 0.15769121050834656, "learning_rate": 6.883742414591998e-05, "loss": 0.0136, "step": 40820 }, { "grad_norm": 0.16479423642158508, "learning_rate": 6.88221067674114e-05, "loss": 0.0105, "step": 40830 }, { "grad_norm": 0.22400175034999847, "learning_rate": 6.88067873305455e-05, "loss": 0.0125, "step": 40840 }, { "grad_norm": 0.14517958462238312, "learning_rate": 6.879146583699765e-05, "loss": 0.013, "step": 40850 }, { "grad_norm": 0.1596987098455429, "learning_rate": 6.877614228844334e-05, "loss": 0.0134, "step": 40860 }, { "grad_norm": 0.1658419370651245, "learning_rate": 6.876081668655832e-05, "loss": 0.0126, "step": 40870 }, { "grad_norm": 0.16540521383285522, "learning_rate": 6.874548903301861e-05, "loss": 0.0167, "step": 40880 }, { "grad_norm": 0.18205896019935608, "learning_rate": 6.873015932950039e-05, "loss": 0.0152, "step": 40890 }, { "grad_norm": 0.17452380061149597, "learning_rate": 6.871482757768012e-05, "loss": 0.015, "step": 40900 }, { "grad_norm": 0.17336641252040863, "learning_rate": 6.869949377923442e-05, "loss": 0.0169, "step": 40910 }, { "grad_norm": 0.1564958542585373, "learning_rate": 6.868415793584022e-05, "loss": 0.0164, "step": 40920 }, { "grad_norm": 0.15620057284832, "learning_rate": 6.86688200491746e-05, "loss": 0.0151, "step": 40930 }, { "grad_norm": 0.21854878962039948, "learning_rate": 6.865348012091488e-05, "loss": 0.0153, "step": 40940 }, { "grad_norm": 0.18742038309574127, "learning_rate": 6.86381381527386e-05, "loss": 0.0189, "step": 40950 }, { "grad_norm": 0.14360050857067108, "learning_rate": 6.862279414632359e-05, "loss": 0.0136, "step": 40960 }, { "grad_norm": 0.21497592329978943, "learning_rate": 6.860744810334778e-05, "loss": 0.0199, "step": 40970 }, { "grad_norm": 0.17308935523033142, "learning_rate": 6.859210002548943e-05, "loss": 0.0156, "step": 40980 }, { "grad_norm": 0.15260010957717896, "learning_rate": 6.8576749914427e-05, "loss": 0.0139, "step": 40990 }, { "grad_norm": 0.17407065629959106, "learning_rate": 6.85613977718391e-05, "loss": 0.0144, "step": 41000 }, { "grad_norm": 0.17244485020637512, "learning_rate": 6.854604359940464e-05, "loss": 0.0141, "step": 41010 }, { "grad_norm": 0.17176847159862518, "learning_rate": 6.853068739880276e-05, "loss": 0.0139, "step": 41020 }, { "grad_norm": 0.17478971183300018, "learning_rate": 6.851532917171276e-05, "loss": 0.0205, "step": 41030 }, { "grad_norm": 0.16366605460643768, "learning_rate": 6.849996891981418e-05, "loss": 0.0132, "step": 41040 }, { "grad_norm": 0.2320614606142044, "learning_rate": 6.848460664478684e-05, "loss": 0.0209, "step": 41050 }, { "grad_norm": 0.19295604526996613, "learning_rate": 6.846924234831069e-05, "loss": 0.0138, "step": 41060 }, { "grad_norm": 0.18542161583900452, "learning_rate": 6.845387603206597e-05, "loss": 0.015, "step": 41070 }, { "grad_norm": 0.14138579368591309, "learning_rate": 6.843850769773312e-05, "loss": 0.0108, "step": 41080 }, { "grad_norm": 0.2027359902858734, "learning_rate": 6.842313734699278e-05, "loss": 0.0123, "step": 41090 }, { "grad_norm": 0.16878174245357513, "learning_rate": 6.840776498152584e-05, "loss": 0.0166, "step": 41100 }, { "grad_norm": 0.18933138251304626, "learning_rate": 6.83923906030134e-05, "loss": 0.0165, "step": 41110 }, { "grad_norm": 0.1671435534954071, "learning_rate": 6.837701421313677e-05, "loss": 0.0175, "step": 41120 }, { "grad_norm": 0.1940276175737381, "learning_rate": 6.83616358135775e-05, "loss": 0.0149, "step": 41130 }, { "grad_norm": 0.17602071166038513, "learning_rate": 6.834625540601734e-05, "loss": 0.0132, "step": 41140 }, { "grad_norm": 0.1793811023235321, "learning_rate": 6.833087299213829e-05, "loss": 0.0125, "step": 41150 }, { "grad_norm": 0.17544043064117432, "learning_rate": 6.83154885736225e-05, "loss": 0.0186, "step": 41160 }, { "grad_norm": 0.19105347990989685, "learning_rate": 6.830010215215244e-05, "loss": 0.0158, "step": 41170 }, { "grad_norm": 0.1259760856628418, "learning_rate": 6.828471372941072e-05, "loss": 0.013, "step": 41180 }, { "grad_norm": 0.14371290802955627, "learning_rate": 6.826932330708018e-05, "loss": 0.0122, "step": 41190 }, { "grad_norm": 0.15670926868915558, "learning_rate": 6.825393088684393e-05, "loss": 0.0127, "step": 41200 }, { "grad_norm": 0.18405918776988983, "learning_rate": 6.823853647038524e-05, "loss": 0.0142, "step": 41210 }, { "grad_norm": 0.13813750445842743, "learning_rate": 6.822314005938762e-05, "loss": 0.0124, "step": 41220 }, { "grad_norm": 0.17481499910354614, "learning_rate": 6.820774165553479e-05, "loss": 0.0109, "step": 41230 }, { "grad_norm": 0.14719834923744202, "learning_rate": 6.819234126051072e-05, "loss": 0.0127, "step": 41240 }, { "grad_norm": 0.14467070996761322, "learning_rate": 6.817693887599956e-05, "loss": 0.012, "step": 41250 }, { "grad_norm": 0.18441402912139893, "learning_rate": 6.816153450368568e-05, "loss": 0.016, "step": 41260 }, { "grad_norm": 0.16543202102184296, "learning_rate": 6.814612814525372e-05, "loss": 0.0149, "step": 41270 }, { "grad_norm": 0.15330158174037933, "learning_rate": 6.813071980238846e-05, "loss": 0.0135, "step": 41280 }, { "grad_norm": 0.1732095181941986, "learning_rate": 6.811530947677492e-05, "loss": 0.0142, "step": 41290 }, { "grad_norm": 0.2501441240310669, "learning_rate": 6.809989717009839e-05, "loss": 0.017, "step": 41300 }, { "grad_norm": 0.24735699594020844, "learning_rate": 6.808448288404431e-05, "loss": 0.0158, "step": 41310 }, { "grad_norm": 0.15507681667804718, "learning_rate": 6.806906662029838e-05, "loss": 0.0164, "step": 41320 }, { "grad_norm": 0.16818231344223022, "learning_rate": 6.805364838054647e-05, "loss": 0.0175, "step": 41330 }, { "grad_norm": 0.173534095287323, "learning_rate": 6.803822816647474e-05, "loss": 0.014, "step": 41340 }, { "grad_norm": 0.16916465759277344, "learning_rate": 6.802280597976949e-05, "loss": 0.0116, "step": 41350 }, { "grad_norm": 0.15100713074207306, "learning_rate": 6.800738182211727e-05, "loss": 0.0146, "step": 41360 }, { "grad_norm": 0.17315112054347992, "learning_rate": 6.799195569520485e-05, "loss": 0.0141, "step": 41370 }, { "grad_norm": 0.17867624759674072, "learning_rate": 6.79765276007192e-05, "loss": 0.0137, "step": 41380 }, { "grad_norm": 0.20646946132183075, "learning_rate": 6.796109754034753e-05, "loss": 0.0172, "step": 41390 }, { "grad_norm": 0.15104851126670837, "learning_rate": 6.794566551577724e-05, "loss": 0.0128, "step": 41400 }, { "grad_norm": 0.19687220454216003, "learning_rate": 6.793023152869594e-05, "loss": 0.0113, "step": 41410 }, { "grad_norm": 0.140904039144516, "learning_rate": 6.791479558079147e-05, "loss": 0.0149, "step": 41420 }, { "grad_norm": 0.15487602353096008, "learning_rate": 6.789935767375191e-05, "loss": 0.0162, "step": 41430 }, { "grad_norm": 0.18390026688575745, "learning_rate": 6.78839178092655e-05, "loss": 0.0144, "step": 41440 }, { "grad_norm": 0.18803836405277252, "learning_rate": 6.786847598902072e-05, "loss": 0.0167, "step": 41450 }, { "grad_norm": 0.162154883146286, "learning_rate": 6.785303221470627e-05, "loss": 0.0137, "step": 41460 }, { "grad_norm": 0.16231682896614075, "learning_rate": 6.783758648801108e-05, "loss": 0.0165, "step": 41470 }, { "grad_norm": 0.19967587292194366, "learning_rate": 6.782213881062422e-05, "loss": 0.0187, "step": 41480 }, { "grad_norm": 0.2008042335510254, "learning_rate": 6.780668918423508e-05, "loss": 0.0168, "step": 41490 }, { "grad_norm": 0.13433918356895447, "learning_rate": 6.779123761053317e-05, "loss": 0.015, "step": 41500 }, { "grad_norm": 0.17499114573001862, "learning_rate": 6.777578409120826e-05, "loss": 0.0161, "step": 41510 }, { "grad_norm": 0.17453612387180328, "learning_rate": 6.776032862795033e-05, "loss": 0.0156, "step": 41520 }, { "grad_norm": 0.14355751872062683, "learning_rate": 6.774487122244956e-05, "loss": 0.0189, "step": 41530 }, { "grad_norm": 0.14837104082107544, "learning_rate": 6.772941187639637e-05, "loss": 0.0145, "step": 41540 }, { "grad_norm": 0.16883765161037445, "learning_rate": 6.771395059148134e-05, "loss": 0.0128, "step": 41550 }, { "grad_norm": 0.21170757710933685, "learning_rate": 6.76984873693953e-05, "loss": 0.0132, "step": 41560 }, { "grad_norm": 0.21107515692710876, "learning_rate": 6.768302221182931e-05, "loss": 0.0131, "step": 41570 }, { "grad_norm": 0.23181423544883728, "learning_rate": 6.766755512047457e-05, "loss": 0.0124, "step": 41580 }, { "grad_norm": 0.1323402225971222, "learning_rate": 6.765208609702259e-05, "loss": 0.0138, "step": 41590 }, { "grad_norm": 0.1438196301460266, "learning_rate": 6.763661514316499e-05, "loss": 0.0127, "step": 41600 }, { "grad_norm": 0.11531616002321243, "learning_rate": 6.76211422605937e-05, "loss": 0.01, "step": 41610 }, { "grad_norm": 0.1383005976676941, "learning_rate": 6.760566745100076e-05, "loss": 0.0136, "step": 41620 }, { "grad_norm": 0.19660183787345886, "learning_rate": 6.75901907160785e-05, "loss": 0.0172, "step": 41630 }, { "grad_norm": 0.1254250556230545, "learning_rate": 6.757471205751943e-05, "loss": 0.0109, "step": 41640 }, { "grad_norm": 0.18743419647216797, "learning_rate": 6.75592314770163e-05, "loss": 0.0146, "step": 41650 }, { "grad_norm": 0.2012498378753662, "learning_rate": 6.7543748976262e-05, "loss": 0.0129, "step": 41660 }, { "grad_norm": 0.15501943230628967, "learning_rate": 6.752826455694968e-05, "loss": 0.0121, "step": 41670 }, { "grad_norm": 0.17986911535263062, "learning_rate": 6.751277822077271e-05, "loss": 0.0116, "step": 41680 }, { "grad_norm": 0.11069886386394501, "learning_rate": 6.749728996942465e-05, "loss": 0.0118, "step": 41690 }, { "grad_norm": 0.16890475153923035, "learning_rate": 6.748179980459924e-05, "loss": 0.0137, "step": 41700 }, { "grad_norm": 0.13141018152236938, "learning_rate": 6.746630772799052e-05, "loss": 0.0145, "step": 41710 }, { "grad_norm": 0.1866430938243866, "learning_rate": 6.745081374129262e-05, "loss": 0.0137, "step": 41720 }, { "grad_norm": 0.13935533165931702, "learning_rate": 6.743531784619998e-05, "loss": 0.0117, "step": 41730 }, { "grad_norm": 0.212330162525177, "learning_rate": 6.741982004440719e-05, "loss": 0.0155, "step": 41740 }, { "grad_norm": 0.13950881361961365, "learning_rate": 6.740432033760907e-05, "loss": 0.0124, "step": 41750 }, { "grad_norm": 0.17774143815040588, "learning_rate": 6.738881872750066e-05, "loss": 0.0171, "step": 41760 }, { "grad_norm": 0.1816575527191162, "learning_rate": 6.737331521577715e-05, "loss": 0.0122, "step": 41770 }, { "grad_norm": 0.17810475826263428, "learning_rate": 6.735780980413403e-05, "loss": 0.0145, "step": 41780 }, { "grad_norm": 0.13575562834739685, "learning_rate": 6.734230249426692e-05, "loss": 0.0131, "step": 41790 }, { "grad_norm": 0.22656236588954926, "learning_rate": 6.732679328787168e-05, "loss": 0.0142, "step": 41800 }, { "grad_norm": 0.19423967599868774, "learning_rate": 6.731128218664438e-05, "loss": 0.0169, "step": 41810 }, { "grad_norm": 0.18032649159431458, "learning_rate": 6.729576919228129e-05, "loss": 0.0119, "step": 41820 }, { "grad_norm": 0.16335754096508026, "learning_rate": 6.728025430647888e-05, "loss": 0.0144, "step": 41830 }, { "grad_norm": 0.12230943143367767, "learning_rate": 6.726473753093383e-05, "loss": 0.0126, "step": 41840 }, { "grad_norm": 0.2237766683101654, "learning_rate": 6.724921886734305e-05, "loss": 0.0131, "step": 41850 }, { "grad_norm": 0.18670576810836792, "learning_rate": 6.723369831740363e-05, "loss": 0.0139, "step": 41860 }, { "grad_norm": 0.20756036043167114, "learning_rate": 6.721817588281288e-05, "loss": 0.0183, "step": 41870 }, { "grad_norm": 0.19707746803760529, "learning_rate": 6.720265156526828e-05, "loss": 0.0157, "step": 41880 }, { "grad_norm": 0.16872930526733398, "learning_rate": 6.718712536646758e-05, "loss": 0.0145, "step": 41890 }, { "grad_norm": 0.165863037109375, "learning_rate": 6.71715972881087e-05, "loss": 0.0139, "step": 41900 }, { "grad_norm": 0.17766985297203064, "learning_rate": 6.715606733188974e-05, "loss": 0.0132, "step": 41910 }, { "grad_norm": 0.12876351177692413, "learning_rate": 6.714053549950909e-05, "loss": 0.013, "step": 41920 }, { "grad_norm": 0.11628212034702301, "learning_rate": 6.71250017926652e-05, "loss": 0.0138, "step": 41930 }, { "grad_norm": 0.17313966155052185, "learning_rate": 6.71094662130569e-05, "loss": 0.0129, "step": 41940 }, { "grad_norm": 0.14403730630874634, "learning_rate": 6.709392876238307e-05, "loss": 0.0138, "step": 41950 }, { "grad_norm": 0.16920717060565948, "learning_rate": 6.707838944234289e-05, "loss": 0.0172, "step": 41960 }, { "grad_norm": 0.1326676458120346, "learning_rate": 6.706284825463574e-05, "loss": 0.0114, "step": 41970 }, { "grad_norm": 0.1645166128873825, "learning_rate": 6.704730520096114e-05, "loss": 0.0136, "step": 41980 }, { "grad_norm": 0.1624298095703125, "learning_rate": 6.703176028301888e-05, "loss": 0.0177, "step": 41990 }, { "grad_norm": 0.15950848162174225, "learning_rate": 6.701621350250892e-05, "loss": 0.0155, "step": 42000 }, { "grad_norm": 0.1902628242969513, "learning_rate": 6.700066486113144e-05, "loss": 0.0163, "step": 42010 }, { "grad_norm": 0.166643887758255, "learning_rate": 6.69851143605868e-05, "loss": 0.0125, "step": 42020 }, { "grad_norm": 0.18044497072696686, "learning_rate": 6.696956200257561e-05, "loss": 0.0131, "step": 42030 }, { "grad_norm": 0.17195208370685577, "learning_rate": 6.695400778879863e-05, "loss": 0.0172, "step": 42040 }, { "grad_norm": 0.19143584370613098, "learning_rate": 6.693845172095683e-05, "loss": 0.0155, "step": 42050 }, { "grad_norm": 0.1697845458984375, "learning_rate": 6.692289380075142e-05, "loss": 0.0111, "step": 42060 }, { "grad_norm": 0.20250853896141052, "learning_rate": 6.690733402988379e-05, "loss": 0.0137, "step": 42070 }, { "grad_norm": 0.21864639222621918, "learning_rate": 6.689177241005553e-05, "loss": 0.0152, "step": 42080 }, { "grad_norm": 0.16370372474193573, "learning_rate": 6.687620894296844e-05, "loss": 0.0182, "step": 42090 }, { "grad_norm": 0.20474012196063995, "learning_rate": 6.686064363032451e-05, "loss": 0.0157, "step": 42100 }, { "grad_norm": 0.10736727714538574, "learning_rate": 6.684507647382596e-05, "loss": 0.0123, "step": 42110 }, { "grad_norm": 0.17746351659297943, "learning_rate": 6.682950747517513e-05, "loss": 0.0119, "step": 42120 }, { "grad_norm": 0.21383026242256165, "learning_rate": 6.68139366360747e-05, "loss": 0.0188, "step": 42130 }, { "grad_norm": 0.1617974191904068, "learning_rate": 6.679836395822744e-05, "loss": 0.0129, "step": 42140 }, { "grad_norm": 0.18308381736278534, "learning_rate": 6.678278944333633e-05, "loss": 0.0155, "step": 42150 }, { "grad_norm": 0.14290711283683777, "learning_rate": 6.676721309310462e-05, "loss": 0.0129, "step": 42160 }, { "grad_norm": 0.17924942076206207, "learning_rate": 6.67516349092357e-05, "loss": 0.0165, "step": 42170 }, { "grad_norm": 0.19592927396297455, "learning_rate": 6.673605489343315e-05, "loss": 0.0129, "step": 42180 }, { "grad_norm": 0.16456112265586853, "learning_rate": 6.672047304740081e-05, "loss": 0.0136, "step": 42190 }, { "grad_norm": 0.18023322522640228, "learning_rate": 6.670488937284268e-05, "loss": 0.0145, "step": 42200 }, { "grad_norm": 0.1468125879764557, "learning_rate": 6.668930387146294e-05, "loss": 0.0125, "step": 42210 }, { "grad_norm": 0.16366881132125854, "learning_rate": 6.667371654496605e-05, "loss": 0.0185, "step": 42220 }, { "grad_norm": 0.16209174692630768, "learning_rate": 6.665812739505659e-05, "loss": 0.0141, "step": 42230 }, { "grad_norm": 0.1429370939731598, "learning_rate": 6.664253642343935e-05, "loss": 0.0125, "step": 42240 }, { "grad_norm": 0.208765909075737, "learning_rate": 6.662694363181935e-05, "loss": 0.0175, "step": 42250 }, { "grad_norm": 0.15616777539253235, "learning_rate": 6.66113490219018e-05, "loss": 0.0152, "step": 42260 }, { "grad_norm": 0.16753405332565308, "learning_rate": 6.659575259539207e-05, "loss": 0.0114, "step": 42270 }, { "grad_norm": 0.15312767028808594, "learning_rate": 6.65801543539958e-05, "loss": 0.0188, "step": 42280 }, { "grad_norm": 0.1703231930732727, "learning_rate": 6.65645542994188e-05, "loss": 0.0164, "step": 42290 }, { "grad_norm": 0.15455424785614014, "learning_rate": 6.654895243336702e-05, "loss": 0.0129, "step": 42300 }, { "grad_norm": 0.17476794123649597, "learning_rate": 6.65333487575467e-05, "loss": 0.0155, "step": 42310 }, { "grad_norm": 0.17008742690086365, "learning_rate": 6.651774327366422e-05, "loss": 0.012, "step": 42320 }, { "grad_norm": 0.19950997829437256, "learning_rate": 6.650213598342613e-05, "loss": 0.0121, "step": 42330 }, { "grad_norm": 0.1369389444589615, "learning_rate": 6.648652688853931e-05, "loss": 0.0128, "step": 42340 }, { "grad_norm": 0.16452160477638245, "learning_rate": 6.647091599071066e-05, "loss": 0.014, "step": 42350 }, { "grad_norm": 0.1736076921224594, "learning_rate": 6.645530329164743e-05, "loss": 0.0123, "step": 42360 }, { "grad_norm": 0.15571832656860352, "learning_rate": 6.643968879305694e-05, "loss": 0.0177, "step": 42370 }, { "grad_norm": 0.21103458106517792, "learning_rate": 6.642407249664683e-05, "loss": 0.0139, "step": 42380 }, { "grad_norm": 0.16509217023849487, "learning_rate": 6.640845440412483e-05, "loss": 0.0127, "step": 42390 }, { "grad_norm": 0.15710648894309998, "learning_rate": 6.639283451719893e-05, "loss": 0.0156, "step": 42400 }, { "grad_norm": 0.13695712387561798, "learning_rate": 6.637721283757729e-05, "loss": 0.0114, "step": 42410 }, { "grad_norm": 0.16951319575309753, "learning_rate": 6.636158936696828e-05, "loss": 0.0129, "step": 42420 }, { "grad_norm": 0.17451344430446625, "learning_rate": 6.634596410708047e-05, "loss": 0.0162, "step": 42430 }, { "grad_norm": 0.14444617927074432, "learning_rate": 6.633033705962257e-05, "loss": 0.0106, "step": 42440 }, { "grad_norm": 0.17359955608844757, "learning_rate": 6.631470822630359e-05, "loss": 0.0155, "step": 42450 }, { "grad_norm": 0.1799115091562271, "learning_rate": 6.629907760883263e-05, "loss": 0.0162, "step": 42460 }, { "grad_norm": 0.16172587871551514, "learning_rate": 6.628344520891905e-05, "loss": 0.0127, "step": 42470 }, { "grad_norm": 0.21112681925296783, "learning_rate": 6.62678110282724e-05, "loss": 0.0162, "step": 42480 }, { "grad_norm": 0.1642308384180069, "learning_rate": 6.625217506860237e-05, "loss": 0.0152, "step": 42490 }, { "grad_norm": 0.12102458626031876, "learning_rate": 6.623653733161892e-05, "loss": 0.0116, "step": 42500 }, { "grad_norm": 0.2112722545862198, "learning_rate": 6.622089781903217e-05, "loss": 0.0136, "step": 42510 }, { "grad_norm": 0.1383177638053894, "learning_rate": 6.620525653255241e-05, "loss": 0.0117, "step": 42520 }, { "grad_norm": 0.14418503642082214, "learning_rate": 6.618961347389015e-05, "loss": 0.0148, "step": 42530 }, { "grad_norm": 0.2270120531320572, "learning_rate": 6.617396864475613e-05, "loss": 0.0134, "step": 42540 }, { "grad_norm": 0.14402975142002106, "learning_rate": 6.61583220468612e-05, "loss": 0.0124, "step": 42550 }, { "grad_norm": 0.1899433732032776, "learning_rate": 6.614267368191645e-05, "loss": 0.0146, "step": 42560 }, { "grad_norm": 0.1752764731645584, "learning_rate": 6.61270235516332e-05, "loss": 0.0132, "step": 42570 }, { "grad_norm": 0.15334148705005646, "learning_rate": 6.611137165772287e-05, "loss": 0.0148, "step": 42580 }, { "grad_norm": 0.15575531125068665, "learning_rate": 6.60957180018972e-05, "loss": 0.0145, "step": 42590 }, { "grad_norm": 0.16930043697357178, "learning_rate": 6.608006258586797e-05, "loss": 0.0115, "step": 42600 }, { "grad_norm": 0.16943296790122986, "learning_rate": 6.606440541134731e-05, "loss": 0.015, "step": 42610 }, { "grad_norm": 0.17758521437644958, "learning_rate": 6.60487464800474e-05, "loss": 0.0143, "step": 42620 }, { "grad_norm": 0.14147613942623138, "learning_rate": 6.603308579368071e-05, "loss": 0.0123, "step": 42630 }, { "grad_norm": 0.15331727266311646, "learning_rate": 6.601742335395987e-05, "loss": 0.0157, "step": 42640 }, { "grad_norm": 0.14697203040122986, "learning_rate": 6.600175916259769e-05, "loss": 0.012, "step": 42650 }, { "grad_norm": 0.14538811147212982, "learning_rate": 6.598609322130718e-05, "loss": 0.0092, "step": 42660 }, { "grad_norm": 0.15901651978492737, "learning_rate": 6.597042553180155e-05, "loss": 0.0158, "step": 42670 }, { "grad_norm": 0.14164966344833374, "learning_rate": 6.59547560957942e-05, "loss": 0.011, "step": 42680 }, { "grad_norm": 0.1744154989719391, "learning_rate": 6.59390849149987e-05, "loss": 0.012, "step": 42690 }, { "grad_norm": 0.20805536210536957, "learning_rate": 6.592341199112886e-05, "loss": 0.0148, "step": 42700 }, { "grad_norm": 0.1884622722864151, "learning_rate": 6.590773732589861e-05, "loss": 0.0124, "step": 42710 }, { "grad_norm": 0.22853434085845947, "learning_rate": 6.589206092102213e-05, "loss": 0.0168, "step": 42720 }, { "grad_norm": 0.17394883930683136, "learning_rate": 6.587638277821375e-05, "loss": 0.0118, "step": 42730 }, { "grad_norm": 0.17127308249473572, "learning_rate": 6.586070289918803e-05, "loss": 0.0142, "step": 42740 }, { "grad_norm": 0.16329720616340637, "learning_rate": 6.584502128565968e-05, "loss": 0.0128, "step": 42750 }, { "grad_norm": 0.15298031270503998, "learning_rate": 6.582933793934363e-05, "loss": 0.0122, "step": 42760 }, { "grad_norm": 0.13542896509170532, "learning_rate": 6.581365286195498e-05, "loss": 0.0161, "step": 42770 }, { "grad_norm": 0.17572976648807526, "learning_rate": 6.579796605520903e-05, "loss": 0.0126, "step": 42780 }, { "grad_norm": 0.139576718211174, "learning_rate": 6.578227752082127e-05, "loss": 0.0137, "step": 42790 }, { "grad_norm": 0.13735967874526978, "learning_rate": 6.576658726050735e-05, "loss": 0.0136, "step": 42800 }, { "grad_norm": 0.11181310564279556, "learning_rate": 6.575089527598316e-05, "loss": 0.0135, "step": 42810 }, { "grad_norm": 0.24590077996253967, "learning_rate": 6.573520156896476e-05, "loss": 0.0129, "step": 42820 }, { "grad_norm": 0.1548452079296112, "learning_rate": 6.571950614116835e-05, "loss": 0.0158, "step": 42830 }, { "grad_norm": 0.14194703102111816, "learning_rate": 6.570380899431039e-05, "loss": 0.0125, "step": 42840 }, { "grad_norm": 0.16231216490268707, "learning_rate": 6.568811013010749e-05, "loss": 0.0145, "step": 42850 }, { "grad_norm": 0.21236853301525116, "learning_rate": 6.567240955027645e-05, "loss": 0.0137, "step": 42860 }, { "grad_norm": 0.23325154185295105, "learning_rate": 6.565670725653427e-05, "loss": 0.0158, "step": 42870 }, { "grad_norm": 0.17644168436527252, "learning_rate": 6.564100325059812e-05, "loss": 0.014, "step": 42880 }, { "grad_norm": 0.12059630453586578, "learning_rate": 6.562529753418538e-05, "loss": 0.0154, "step": 42890 }, { "grad_norm": 0.1215037927031517, "learning_rate": 6.56095901090136e-05, "loss": 0.0168, "step": 42900 }, { "grad_norm": 0.1874144822359085, "learning_rate": 6.55938809768005e-05, "loss": 0.013, "step": 42910 }, { "grad_norm": 0.1768651008605957, "learning_rate": 6.557817013926402e-05, "loss": 0.0113, "step": 42920 }, { "grad_norm": 0.1404041200876236, "learning_rate": 6.556245759812227e-05, "loss": 0.0117, "step": 42930 }, { "grad_norm": 0.15761502087116241, "learning_rate": 6.554674335509357e-05, "loss": 0.0139, "step": 42940 }, { "grad_norm": 0.1479051411151886, "learning_rate": 6.553102741189638e-05, "loss": 0.0143, "step": 42950 }, { "grad_norm": 0.11811090260744095, "learning_rate": 6.55153097702494e-05, "loss": 0.0127, "step": 42960 }, { "grad_norm": 0.12499401718378067, "learning_rate": 6.549959043187146e-05, "loss": 0.0113, "step": 42970 }, { "grad_norm": 0.1377061903476715, "learning_rate": 6.54838693984816e-05, "loss": 0.0127, "step": 42980 }, { "grad_norm": 0.23588287830352783, "learning_rate": 6.546814667179909e-05, "loss": 0.0171, "step": 42990 }, { "grad_norm": 0.16410206258296967, "learning_rate": 6.545242225354328e-05, "loss": 0.0145, "step": 43000 }, { "grad_norm": 0.12001311033964157, "learning_rate": 6.543669614543383e-05, "loss": 0.0117, "step": 43010 }, { "grad_norm": 0.16248415410518646, "learning_rate": 6.542096834919049e-05, "loss": 0.0127, "step": 43020 }, { "grad_norm": 0.1308748573064804, "learning_rate": 6.540523886653324e-05, "loss": 0.0123, "step": 43030 }, { "grad_norm": 0.20289413630962372, "learning_rate": 6.538950769918222e-05, "loss": 0.015, "step": 43040 }, { "grad_norm": 0.19393546879291534, "learning_rate": 6.537377484885779e-05, "loss": 0.0155, "step": 43050 }, { "grad_norm": 0.20266376435756683, "learning_rate": 6.535804031728045e-05, "loss": 0.0126, "step": 43060 }, { "grad_norm": 0.17988023161888123, "learning_rate": 6.534230410617092e-05, "loss": 0.0143, "step": 43070 }, { "grad_norm": 0.1738850176334381, "learning_rate": 6.532656621725007e-05, "loss": 0.0148, "step": 43080 }, { "grad_norm": 0.18604017794132233, "learning_rate": 6.531082665223898e-05, "loss": 0.0132, "step": 43090 }, { "grad_norm": 0.15712423622608185, "learning_rate": 6.529508541285889e-05, "loss": 0.0143, "step": 43100 }, { "grad_norm": 0.20860204100608826, "learning_rate": 6.527934250083128e-05, "loss": 0.014, "step": 43110 }, { "grad_norm": 0.1837584376335144, "learning_rate": 6.526359791787775e-05, "loss": 0.0141, "step": 43120 }, { "grad_norm": 0.187745600938797, "learning_rate": 6.524785166572009e-05, "loss": 0.0129, "step": 43130 }, { "grad_norm": 0.157538041472435, "learning_rate": 6.523210374608029e-05, "loss": 0.0119, "step": 43140 }, { "grad_norm": 0.17076700925827026, "learning_rate": 6.521635416068054e-05, "loss": 0.0178, "step": 43150 }, { "grad_norm": 0.18181997537612915, "learning_rate": 6.520060291124317e-05, "loss": 0.0147, "step": 43160 }, { "grad_norm": 0.1900830715894699, "learning_rate": 6.518484999949072e-05, "loss": 0.0114, "step": 43170 }, { "grad_norm": 0.18693579733371735, "learning_rate": 6.516909542714591e-05, "loss": 0.0188, "step": 43180 }, { "grad_norm": 0.1386161595582962, "learning_rate": 6.515333919593161e-05, "loss": 0.0163, "step": 43190 }, { "grad_norm": 0.15352635085582733, "learning_rate": 6.513758130757094e-05, "loss": 0.0128, "step": 43200 }, { "grad_norm": 0.19090057909488678, "learning_rate": 6.512182176378713e-05, "loss": 0.0124, "step": 43210 }, { "grad_norm": 0.17095212638378143, "learning_rate": 6.510606056630362e-05, "loss": 0.0133, "step": 43220 }, { "grad_norm": 0.18829278647899628, "learning_rate": 6.509029771684405e-05, "loss": 0.0122, "step": 43230 }, { "grad_norm": 0.13658033311367035, "learning_rate": 6.50745332171322e-05, "loss": 0.0151, "step": 43240 }, { "grad_norm": 0.12040690332651138, "learning_rate": 6.505876706889207e-05, "loss": 0.0133, "step": 43250 }, { "grad_norm": 0.15175274014472961, "learning_rate": 6.504299927384781e-05, "loss": 0.0152, "step": 43260 }, { "grad_norm": 0.16244244575500488, "learning_rate": 6.502722983372378e-05, "loss": 0.0165, "step": 43270 }, { "grad_norm": 0.09961070865392685, "learning_rate": 6.501145875024446e-05, "loss": 0.012, "step": 43280 }, { "grad_norm": 0.17844469845294952, "learning_rate": 6.499568602513459e-05, "loss": 0.0153, "step": 43290 }, { "grad_norm": 0.1754807084798813, "learning_rate": 6.497991166011903e-05, "loss": 0.0102, "step": 43300 }, { "grad_norm": 0.13557381927967072, "learning_rate": 6.49641356569229e-05, "loss": 0.0125, "step": 43310 }, { "grad_norm": 0.2351745367050171, "learning_rate": 6.494835801727135e-05, "loss": 0.0171, "step": 43320 }, { "grad_norm": 0.211858868598938, "learning_rate": 6.493257874288986e-05, "loss": 0.0145, "step": 43330 }, { "grad_norm": 0.19700214266777039, "learning_rate": 6.491679783550399e-05, "loss": 0.0153, "step": 43340 }, { "grad_norm": 0.17404285073280334, "learning_rate": 6.490101529683957e-05, "loss": 0.0176, "step": 43350 }, { "grad_norm": 0.18619436025619507, "learning_rate": 6.488523112862248e-05, "loss": 0.0124, "step": 43360 }, { "grad_norm": 0.168466255068779, "learning_rate": 6.486944533257891e-05, "loss": 0.0124, "step": 43370 }, { "grad_norm": 0.1494259089231491, "learning_rate": 6.485365791043517e-05, "loss": 0.0165, "step": 43380 }, { "grad_norm": 0.17660579085350037, "learning_rate": 6.483786886391772e-05, "loss": 0.0132, "step": 43390 }, { "grad_norm": 0.1916263997554779, "learning_rate": 6.482207819475323e-05, "loss": 0.0132, "step": 43400 }, { "grad_norm": 0.16018816828727722, "learning_rate": 6.480628590466857e-05, "loss": 0.0172, "step": 43410 }, { "grad_norm": 0.1860579252243042, "learning_rate": 6.479049199539074e-05, "loss": 0.0135, "step": 43420 }, { "grad_norm": 0.18017198145389557, "learning_rate": 6.477469646864693e-05, "loss": 0.0185, "step": 43430 }, { "grad_norm": 0.13693074882030487, "learning_rate": 6.475889932616454e-05, "loss": 0.0204, "step": 43440 }, { "grad_norm": 0.14699386060237885, "learning_rate": 6.474310056967111e-05, "loss": 0.0165, "step": 43450 }, { "grad_norm": 0.17556828260421753, "learning_rate": 6.472730020089437e-05, "loss": 0.0203, "step": 43460 }, { "grad_norm": 0.15118975937366486, "learning_rate": 6.471149822156222e-05, "loss": 0.0135, "step": 43470 }, { "grad_norm": 0.17966845631599426, "learning_rate": 6.469569463340275e-05, "loss": 0.013, "step": 43480 }, { "grad_norm": 0.1485871821641922, "learning_rate": 6.467988943814418e-05, "loss": 0.0149, "step": 43490 }, { "grad_norm": 0.17887276411056519, "learning_rate": 6.4664082637515e-05, "loss": 0.0128, "step": 43500 }, { "grad_norm": 0.167540043592453, "learning_rate": 6.464827423324377e-05, "loss": 0.012, "step": 43510 }, { "grad_norm": 0.24057987332344055, "learning_rate": 6.463246422705931e-05, "loss": 0.0161, "step": 43520 }, { "grad_norm": 0.19907134771347046, "learning_rate": 6.461665262069055e-05, "loss": 0.0142, "step": 43530 }, { "grad_norm": 0.16501080989837646, "learning_rate": 6.460083941586665e-05, "loss": 0.0164, "step": 43540 }, { "grad_norm": 0.2036701738834381, "learning_rate": 6.45850246143169e-05, "loss": 0.012, "step": 43550 }, { "grad_norm": 0.21803797781467438, "learning_rate": 6.456920821777077e-05, "loss": 0.0142, "step": 43560 }, { "grad_norm": 0.14155061542987823, "learning_rate": 6.455339022795796e-05, "loss": 0.0141, "step": 43570 }, { "grad_norm": 0.17253845930099487, "learning_rate": 6.453757064660825e-05, "loss": 0.0119, "step": 43580 }, { "grad_norm": 0.1202184334397316, "learning_rate": 6.452174947545169e-05, "loss": 0.0125, "step": 43590 }, { "grad_norm": 0.1743212342262268, "learning_rate": 6.450592671621842e-05, "loss": 0.0143, "step": 43600 }, { "grad_norm": 0.12517155706882477, "learning_rate": 6.449010237063882e-05, "loss": 0.0144, "step": 43610 }, { "grad_norm": 0.17420867085456848, "learning_rate": 6.447427644044343e-05, "loss": 0.0119, "step": 43620 }, { "grad_norm": 0.234592467546463, "learning_rate": 6.44584489273629e-05, "loss": 0.0136, "step": 43630 }, { "grad_norm": 0.156476229429245, "learning_rate": 6.444261983312817e-05, "loss": 0.0104, "step": 43640 }, { "grad_norm": 0.12030906230211258, "learning_rate": 6.442678915947023e-05, "loss": 0.0132, "step": 43650 }, { "grad_norm": 0.17996267974376678, "learning_rate": 6.441095690812032e-05, "loss": 0.013, "step": 43660 }, { "grad_norm": 0.146706685423851, "learning_rate": 6.439512308080982e-05, "loss": 0.0146, "step": 43670 }, { "grad_norm": 0.1783820241689682, "learning_rate": 6.43792876792703e-05, "loss": 0.0138, "step": 43680 }, { "grad_norm": 0.1403113305568695, "learning_rate": 6.43634507052335e-05, "loss": 0.0159, "step": 43690 }, { "grad_norm": 0.16292673349380493, "learning_rate": 6.434761216043135e-05, "loss": 0.0147, "step": 43700 }, { "grad_norm": 0.11020573228597641, "learning_rate": 6.433177204659589e-05, "loss": 0.0134, "step": 43710 }, { "grad_norm": 0.1328207105398178, "learning_rate": 6.431593036545938e-05, "loss": 0.013, "step": 43720 }, { "grad_norm": 0.18163274228572845, "learning_rate": 6.430008711875426e-05, "loss": 0.0141, "step": 43730 }, { "grad_norm": 0.1842179149389267, "learning_rate": 6.42842423082131e-05, "loss": 0.0145, "step": 43740 }, { "grad_norm": 0.15888246893882751, "learning_rate": 6.426839593556868e-05, "loss": 0.0136, "step": 43750 }, { "grad_norm": 0.14593161642551422, "learning_rate": 6.425254800255392e-05, "loss": 0.0112, "step": 43760 }, { "grad_norm": 0.14166578650474548, "learning_rate": 6.423669851090196e-05, "loss": 0.0132, "step": 43770 }, { "grad_norm": 0.21225307881832123, "learning_rate": 6.422084746234604e-05, "loss": 0.0143, "step": 43780 }, { "grad_norm": 0.14812256395816803, "learning_rate": 6.420499485861965e-05, "loss": 0.0139, "step": 43790 }, { "grad_norm": 0.13980132341384888, "learning_rate": 6.418914070145636e-05, "loss": 0.0144, "step": 43800 }, { "grad_norm": 0.17143464088439941, "learning_rate": 6.417328499258996e-05, "loss": 0.0128, "step": 43810 }, { "grad_norm": 0.17877507209777832, "learning_rate": 6.415742773375446e-05, "loss": 0.012, "step": 43820 }, { "grad_norm": 0.16086114943027496, "learning_rate": 6.414156892668393e-05, "loss": 0.0124, "step": 43830 }, { "grad_norm": 0.1409614235162735, "learning_rate": 6.412570857311267e-05, "loss": 0.0125, "step": 43840 }, { "grad_norm": 0.21319621801376343, "learning_rate": 6.410984667477518e-05, "loss": 0.0146, "step": 43850 }, { "grad_norm": 0.13545386493206024, "learning_rate": 6.409398323340607e-05, "loss": 0.0119, "step": 43860 }, { "grad_norm": 0.1436394304037094, "learning_rate": 6.407811825074012e-05, "loss": 0.0151, "step": 43870 }, { "grad_norm": 0.1483033299446106, "learning_rate": 6.406225172851234e-05, "loss": 0.0136, "step": 43880 }, { "grad_norm": 0.17008095979690552, "learning_rate": 6.404638366845786e-05, "loss": 0.0121, "step": 43890 }, { "grad_norm": 0.15206380188465118, "learning_rate": 6.403051407231196e-05, "loss": 0.0116, "step": 43900 }, { "grad_norm": 0.16416819393634796, "learning_rate": 6.401464294181016e-05, "loss": 0.0122, "step": 43910 }, { "grad_norm": 0.12673617899417877, "learning_rate": 6.399877027868808e-05, "loss": 0.0146, "step": 43920 }, { "grad_norm": 0.17903846502304077, "learning_rate": 6.39828960846815e-05, "loss": 0.0151, "step": 43930 }, { "grad_norm": 0.12811130285263062, "learning_rate": 6.396702036152644e-05, "loss": 0.0121, "step": 43940 }, { "grad_norm": 0.194187730550766, "learning_rate": 6.395114311095905e-05, "loss": 0.0159, "step": 43950 }, { "grad_norm": 0.1895161122083664, "learning_rate": 6.393526433471559e-05, "loss": 0.0146, "step": 43960 }, { "grad_norm": 0.212679922580719, "learning_rate": 6.391938403453262e-05, "loss": 0.0124, "step": 43970 }, { "grad_norm": 0.15251559019088745, "learning_rate": 6.390350221214671e-05, "loss": 0.0141, "step": 43980 }, { "grad_norm": 0.1354236602783203, "learning_rate": 6.38876188692947e-05, "loss": 0.0111, "step": 43990 }, { "grad_norm": 0.16971366107463837, "learning_rate": 6.387173400771358e-05, "loss": 0.0127, "step": 44000 }, { "grad_norm": 0.167313814163208, "learning_rate": 6.385584762914047e-05, "loss": 0.0137, "step": 44010 }, { "grad_norm": 0.18214736878871918, "learning_rate": 6.38399597353127e-05, "loss": 0.0127, "step": 44020 }, { "grad_norm": 0.18148955702781677, "learning_rate": 6.382407032796775e-05, "loss": 0.0128, "step": 44030 }, { "grad_norm": 0.1553119271993637, "learning_rate": 6.380817940884325e-05, "loss": 0.0114, "step": 44040 }, { "grad_norm": 0.11663524806499481, "learning_rate": 6.379228697967702e-05, "loss": 0.0133, "step": 44050 }, { "grad_norm": 0.15146508812904358, "learning_rate": 6.3776393042207e-05, "loss": 0.0127, "step": 44060 }, { "grad_norm": 0.16350847482681274, "learning_rate": 6.376049759817137e-05, "loss": 0.0148, "step": 44070 }, { "grad_norm": 0.18677905201911926, "learning_rate": 6.37446006493084e-05, "loss": 0.0141, "step": 44080 }, { "grad_norm": 0.16288360953330994, "learning_rate": 6.372870219735656e-05, "loss": 0.0123, "step": 44090 }, { "grad_norm": 0.17163707315921783, "learning_rate": 6.371280224405452e-05, "loss": 0.0119, "step": 44100 }, { "grad_norm": 0.14332081377506256, "learning_rate": 6.369690079114101e-05, "loss": 0.0112, "step": 44110 }, { "grad_norm": 0.15662258863449097, "learning_rate": 6.368099784035504e-05, "loss": 0.0122, "step": 44120 }, { "grad_norm": 0.16860651969909668, "learning_rate": 6.366509339343572e-05, "loss": 0.0142, "step": 44130 }, { "grad_norm": 0.1700151413679123, "learning_rate": 6.364918745212232e-05, "loss": 0.0123, "step": 44140 }, { "grad_norm": 0.17288944125175476, "learning_rate": 6.363328001815431e-05, "loss": 0.0122, "step": 44150 }, { "grad_norm": 0.16886559128761292, "learning_rate": 6.361737109327128e-05, "loss": 0.0101, "step": 44160 }, { "grad_norm": 0.15808747708797455, "learning_rate": 6.360146067921301e-05, "loss": 0.0127, "step": 44170 }, { "grad_norm": 0.1583390235900879, "learning_rate": 6.35855487777195e-05, "loss": 0.0127, "step": 44180 }, { "grad_norm": 0.16869550943374634, "learning_rate": 6.356963539053076e-05, "loss": 0.0158, "step": 44190 }, { "grad_norm": 0.13186217844486237, "learning_rate": 6.355372051938711e-05, "loss": 0.0151, "step": 44200 }, { "grad_norm": 0.12225169688463211, "learning_rate": 6.353780416602894e-05, "loss": 0.0116, "step": 44210 }, { "grad_norm": 0.12691844999790192, "learning_rate": 6.352188633219689e-05, "loss": 0.014, "step": 44220 }, { "grad_norm": 0.19642901420593262, "learning_rate": 6.350596701963166e-05, "loss": 0.0137, "step": 44230 }, { "grad_norm": 0.19953009486198425, "learning_rate": 6.349004623007419e-05, "loss": 0.0136, "step": 44240 }, { "grad_norm": 0.21536973118782043, "learning_rate": 6.347412396526555e-05, "loss": 0.0153, "step": 44250 }, { "grad_norm": 0.2295498251914978, "learning_rate": 6.345820022694696e-05, "loss": 0.0213, "step": 44260 }, { "grad_norm": 0.17360161244869232, "learning_rate": 6.344227501685984e-05, "loss": 0.015, "step": 44270 }, { "grad_norm": 0.13732391595840454, "learning_rate": 6.342634833674572e-05, "loss": 0.0148, "step": 44280 }, { "grad_norm": 0.1604740172624588, "learning_rate": 6.341042018834635e-05, "loss": 0.0151, "step": 44290 }, { "grad_norm": 0.19792486727237701, "learning_rate": 6.339449057340359e-05, "loss": 0.0124, "step": 44300 }, { "grad_norm": 0.1511937528848648, "learning_rate": 6.337855949365945e-05, "loss": 0.011, "step": 44310 }, { "grad_norm": 0.21685899794101715, "learning_rate": 6.336262695085619e-05, "loss": 0.014, "step": 44320 }, { "grad_norm": 0.16063959896564484, "learning_rate": 6.334669294673612e-05, "loss": 0.0116, "step": 44330 }, { "grad_norm": 0.1459951549768448, "learning_rate": 6.333075748304179e-05, "loss": 0.0133, "step": 44340 }, { "grad_norm": 0.14281262457370758, "learning_rate": 6.331482056151585e-05, "loss": 0.0114, "step": 44350 }, { "grad_norm": 0.19235658645629883, "learning_rate": 6.329888218390117e-05, "loss": 0.0159, "step": 44360 }, { "grad_norm": 0.17315396666526794, "learning_rate": 6.328294235194072e-05, "loss": 0.0149, "step": 44370 }, { "grad_norm": 0.1938989758491516, "learning_rate": 6.326700106737765e-05, "loss": 0.012, "step": 44380 }, { "grad_norm": 0.19714127480983734, "learning_rate": 6.325105833195531e-05, "loss": 0.0134, "step": 44390 }, { "grad_norm": 0.1861821562051773, "learning_rate": 6.323511414741715e-05, "loss": 0.0131, "step": 44400 }, { "grad_norm": 0.1740865707397461, "learning_rate": 6.321916851550678e-05, "loss": 0.0128, "step": 44410 }, { "grad_norm": 0.13334529101848602, "learning_rate": 6.320322143796806e-05, "loss": 0.0144, "step": 44420 }, { "grad_norm": 0.17043475806713104, "learning_rate": 6.318727291654488e-05, "loss": 0.014, "step": 44430 }, { "grad_norm": 0.21808548271656036, "learning_rate": 6.317132295298134e-05, "loss": 0.0152, "step": 44440 }, { "grad_norm": 0.143747478723526, "learning_rate": 6.315537154902173e-05, "loss": 0.0125, "step": 44450 }, { "grad_norm": 0.16554296016693115, "learning_rate": 6.313941870641048e-05, "loss": 0.0124, "step": 44460 }, { "grad_norm": 0.14311841130256653, "learning_rate": 6.312346442689214e-05, "loss": 0.0125, "step": 44470 }, { "grad_norm": 0.12750029563903809, "learning_rate": 6.310750871221147e-05, "loss": 0.0131, "step": 44480 }, { "grad_norm": 0.1541389673948288, "learning_rate": 6.309155156411335e-05, "loss": 0.0132, "step": 44490 }, { "grad_norm": 0.17431023716926575, "learning_rate": 6.307559298434284e-05, "loss": 0.0185, "step": 44500 }, { "grad_norm": 0.15645748376846313, "learning_rate": 6.305963297464512e-05, "loss": 0.0166, "step": 44510 }, { "grad_norm": 0.17940981686115265, "learning_rate": 6.304367153676561e-05, "loss": 0.0108, "step": 44520 }, { "grad_norm": 0.24189865589141846, "learning_rate": 6.302770867244978e-05, "loss": 0.0132, "step": 44530 }, { "grad_norm": 0.2472502738237381, "learning_rate": 6.301174438344328e-05, "loss": 0.0118, "step": 44540 }, { "grad_norm": 0.16057045757770538, "learning_rate": 6.299577867149202e-05, "loss": 0.0113, "step": 44550 }, { "grad_norm": 0.1375546008348465, "learning_rate": 6.297981153834192e-05, "loss": 0.0143, "step": 44560 }, { "grad_norm": 0.1713705211877823, "learning_rate": 6.296384298573916e-05, "loss": 0.0126, "step": 44570 }, { "grad_norm": 0.12424535304307938, "learning_rate": 6.294787301543001e-05, "loss": 0.0112, "step": 44580 }, { "grad_norm": 0.12310139834880829, "learning_rate": 6.293190162916095e-05, "loss": 0.012, "step": 44590 }, { "grad_norm": 0.12756550312042236, "learning_rate": 6.291592882867855e-05, "loss": 0.0104, "step": 44600 }, { "grad_norm": 0.22985798120498657, "learning_rate": 6.28999546157296e-05, "loss": 0.0132, "step": 44610 }, { "grad_norm": 0.1450265347957611, "learning_rate": 6.288397899206102e-05, "loss": 0.0127, "step": 44620 }, { "grad_norm": 0.2117929905653, "learning_rate": 6.286800195941984e-05, "loss": 0.0122, "step": 44630 }, { "grad_norm": 0.1655389964580536, "learning_rate": 6.285202351955334e-05, "loss": 0.0114, "step": 44640 }, { "grad_norm": 0.15643787384033203, "learning_rate": 6.283604367420887e-05, "loss": 0.0123, "step": 44650 }, { "grad_norm": 0.13680218160152435, "learning_rate": 6.282006242513394e-05, "loss": 0.0163, "step": 44660 }, { "grad_norm": 0.15271855890750885, "learning_rate": 6.280407977407628e-05, "loss": 0.0144, "step": 44670 }, { "grad_norm": 0.16475726664066315, "learning_rate": 6.27880957227837e-05, "loss": 0.0138, "step": 44680 }, { "grad_norm": 0.15687665343284607, "learning_rate": 6.27721102730042e-05, "loss": 0.0117, "step": 44690 }, { "grad_norm": 0.14626696705818176, "learning_rate": 6.275612342648592e-05, "loss": 0.0132, "step": 44700 }, { "grad_norm": 0.09889627248048782, "learning_rate": 6.274013518497716e-05, "loss": 0.011, "step": 44710 }, { "grad_norm": 0.1777525544166565, "learning_rate": 6.272414555022636e-05, "loss": 0.0138, "step": 44720 }, { "grad_norm": 0.15688562393188477, "learning_rate": 6.270815452398215e-05, "loss": 0.0125, "step": 44730 }, { "grad_norm": 0.18539026379585266, "learning_rate": 6.269216210799326e-05, "loss": 0.0112, "step": 44740 }, { "grad_norm": 0.1796034574508667, "learning_rate": 6.26761683040086e-05, "loss": 0.012, "step": 44750 }, { "grad_norm": 0.18275262415409088, "learning_rate": 6.266017311377723e-05, "loss": 0.0122, "step": 44760 }, { "grad_norm": 0.22561199963092804, "learning_rate": 6.264417653904839e-05, "loss": 0.0162, "step": 44770 }, { "grad_norm": 0.20588558912277222, "learning_rate": 6.262817858157139e-05, "loss": 0.015, "step": 44780 }, { "grad_norm": 0.1925012320280075, "learning_rate": 6.261217924309576e-05, "loss": 0.0129, "step": 44790 }, { "grad_norm": 0.16753411293029785, "learning_rate": 6.259617852537118e-05, "loss": 0.0138, "step": 44800 }, { "grad_norm": 0.1240004152059555, "learning_rate": 6.258017643014747e-05, "loss": 0.0118, "step": 44810 }, { "grad_norm": 0.19545572996139526, "learning_rate": 6.256417295917456e-05, "loss": 0.0116, "step": 44820 }, { "grad_norm": 0.15152814984321594, "learning_rate": 6.254816811420258e-05, "loss": 0.0175, "step": 44830 }, { "grad_norm": 0.11183714121580124, "learning_rate": 6.253216189698183e-05, "loss": 0.0105, "step": 44840 }, { "grad_norm": 0.14173051714897156, "learning_rate": 6.251615430926267e-05, "loss": 0.0118, "step": 44850 }, { "grad_norm": 0.1755707859992981, "learning_rate": 6.25001453527957e-05, "loss": 0.0169, "step": 44860 }, { "grad_norm": 0.1617414951324463, "learning_rate": 6.248413502933164e-05, "loss": 0.0155, "step": 44870 }, { "grad_norm": 0.16033490002155304, "learning_rate": 6.246812334062133e-05, "loss": 0.0136, "step": 44880 }, { "grad_norm": 0.1570405513048172, "learning_rate": 6.245211028841579e-05, "loss": 0.014, "step": 44890 }, { "grad_norm": 0.14252696931362152, "learning_rate": 6.24360958744662e-05, "loss": 0.011, "step": 44900 }, { "grad_norm": 0.177836611866951, "learning_rate": 6.242008010052387e-05, "loss": 0.0106, "step": 44910 }, { "grad_norm": 0.2335115671157837, "learning_rate": 6.240406296834024e-05, "loss": 0.0131, "step": 44920 }, { "grad_norm": 0.24794287979602814, "learning_rate": 6.238804447966694e-05, "loss": 0.0127, "step": 44930 }, { "grad_norm": 0.12883099913597107, "learning_rate": 6.237202463625573e-05, "loss": 0.0135, "step": 44940 }, { "grad_norm": 0.15530382096767426, "learning_rate": 6.235600343985848e-05, "loss": 0.0102, "step": 44950 }, { "grad_norm": 0.20301271975040436, "learning_rate": 6.233998089222729e-05, "loss": 0.0125, "step": 44960 }, { "grad_norm": 0.13634605705738068, "learning_rate": 6.232395699511433e-05, "loss": 0.0106, "step": 44970 }, { "grad_norm": 0.1669865995645523, "learning_rate": 6.230793175027199e-05, "loss": 0.0128, "step": 44980 }, { "grad_norm": 0.18602266907691956, "learning_rate": 6.22919051594527e-05, "loss": 0.017, "step": 44990 }, { "grad_norm": 0.12261354923248291, "learning_rate": 6.227587722440917e-05, "loss": 0.0111, "step": 45000 }, { "grad_norm": 0.15298305451869965, "learning_rate": 6.225984794689414e-05, "loss": 0.0151, "step": 45010 }, { "grad_norm": 0.17659245431423187, "learning_rate": 6.22438173286606e-05, "loss": 0.0116, "step": 45020 }, { "grad_norm": 0.1626628041267395, "learning_rate": 6.222778537146159e-05, "loss": 0.0146, "step": 45030 }, { "grad_norm": 0.12473303079605103, "learning_rate": 6.221175207705037e-05, "loss": 0.0104, "step": 45040 }, { "grad_norm": 0.15530827641487122, "learning_rate": 6.219571744718027e-05, "loss": 0.012, "step": 45050 }, { "grad_norm": 0.17381030321121216, "learning_rate": 6.217968148360487e-05, "loss": 0.0148, "step": 45060 }, { "grad_norm": 0.1730930358171463, "learning_rate": 6.216364418807782e-05, "loss": 0.0161, "step": 45070 }, { "grad_norm": 0.14748553931713104, "learning_rate": 6.214760556235292e-05, "loss": 0.0117, "step": 45080 }, { "grad_norm": 0.2046031504869461, "learning_rate": 6.213156560818414e-05, "loss": 0.0129, "step": 45090 }, { "grad_norm": 0.18102799355983734, "learning_rate": 6.211552432732559e-05, "loss": 0.0114, "step": 45100 }, { "grad_norm": 0.18206582963466644, "learning_rate": 6.20994817215315e-05, "loss": 0.0135, "step": 45110 }, { "grad_norm": 0.1563006490468979, "learning_rate": 6.208343779255628e-05, "loss": 0.0131, "step": 45120 }, { "grad_norm": 0.17459562420845032, "learning_rate": 6.206739254215449e-05, "loss": 0.0136, "step": 45130 }, { "grad_norm": 0.21348147094249725, "learning_rate": 6.205134597208077e-05, "loss": 0.0127, "step": 45140 }, { "grad_norm": 0.18374532461166382, "learning_rate": 6.203529808409e-05, "loss": 0.0116, "step": 45150 }, { "grad_norm": 0.1290067434310913, "learning_rate": 6.201924887993712e-05, "loss": 0.0099, "step": 45160 }, { "grad_norm": 0.13796143233776093, "learning_rate": 6.200319836137725e-05, "loss": 0.0121, "step": 45170 }, { "grad_norm": 0.15422594547271729, "learning_rate": 6.198714653016565e-05, "loss": 0.012, "step": 45180 }, { "grad_norm": 0.1492772251367569, "learning_rate": 6.197109338805774e-05, "loss": 0.0108, "step": 45190 }, { "grad_norm": 0.2022426426410675, "learning_rate": 6.195503893680903e-05, "loss": 0.0129, "step": 45200 }, { "grad_norm": 0.15066790580749512, "learning_rate": 6.193898317817524e-05, "loss": 0.01, "step": 45210 }, { "grad_norm": 0.1815599650144577, "learning_rate": 6.192292611391221e-05, "loss": 0.0134, "step": 45220 }, { "grad_norm": 0.15852077305316925, "learning_rate": 6.190686774577591e-05, "loss": 0.0111, "step": 45230 }, { "grad_norm": 0.14991411566734314, "learning_rate": 6.189080807552245e-05, "loss": 0.0109, "step": 45240 }, { "grad_norm": 0.16705632209777832, "learning_rate": 6.187474710490809e-05, "loss": 0.0111, "step": 45250 }, { "grad_norm": 0.14253553748130798, "learning_rate": 6.185868483568926e-05, "loss": 0.0117, "step": 45260 }, { "grad_norm": 0.10778399556875229, "learning_rate": 6.184262126962245e-05, "loss": 0.0116, "step": 45270 }, { "grad_norm": 0.16808964312076569, "learning_rate": 6.182655640846442e-05, "loss": 0.0134, "step": 45280 }, { "grad_norm": 0.16268564760684967, "learning_rate": 6.181049025397196e-05, "loss": 0.0166, "step": 45290 }, { "grad_norm": 0.18252813816070557, "learning_rate": 6.179442280790202e-05, "loss": 0.0138, "step": 45300 }, { "grad_norm": 0.1920490860939026, "learning_rate": 6.177835407201174e-05, "loss": 0.013, "step": 45310 }, { "grad_norm": 0.16665814816951752, "learning_rate": 6.176228404805839e-05, "loss": 0.0111, "step": 45320 }, { "grad_norm": 0.16887758672237396, "learning_rate": 6.174621273779932e-05, "loss": 0.0167, "step": 45330 }, { "grad_norm": 0.20827096700668335, "learning_rate": 6.17301401429921e-05, "loss": 0.0165, "step": 45340 }, { "grad_norm": 0.13327902555465698, "learning_rate": 6.171406626539436e-05, "loss": 0.0116, "step": 45350 }, { "grad_norm": 0.2069869041442871, "learning_rate": 6.169799110676398e-05, "loss": 0.016, "step": 45360 }, { "grad_norm": 0.13061392307281494, "learning_rate": 6.168191466885885e-05, "loss": 0.0145, "step": 45370 }, { "grad_norm": 0.14482566714286804, "learning_rate": 6.166583695343711e-05, "loss": 0.0109, "step": 45380 }, { "grad_norm": 0.1710975021123886, "learning_rate": 6.164975796225698e-05, "loss": 0.0146, "step": 45390 }, { "grad_norm": 0.20803353190422058, "learning_rate": 6.163367769707683e-05, "loss": 0.0112, "step": 45400 }, { "grad_norm": 0.14713899791240692, "learning_rate": 6.161759615965519e-05, "loss": 0.0112, "step": 45410 }, { "grad_norm": 0.1228044182062149, "learning_rate": 6.16015133517507e-05, "loss": 0.0159, "step": 45420 }, { "grad_norm": 0.20463751256465912, "learning_rate": 6.158542927512214e-05, "loss": 0.0134, "step": 45430 }, { "grad_norm": 0.1521933674812317, "learning_rate": 6.156934393152846e-05, "loss": 0.0113, "step": 45440 }, { "grad_norm": 0.1576765477657318, "learning_rate": 6.15532573227287e-05, "loss": 0.012, "step": 45450 }, { "grad_norm": 0.1423977166414261, "learning_rate": 6.153716945048212e-05, "loss": 0.0119, "step": 45460 }, { "grad_norm": 0.19053256511688232, "learning_rate": 6.152108031654802e-05, "loss": 0.0123, "step": 45470 }, { "grad_norm": 0.16250072419643402, "learning_rate": 6.15049899226859e-05, "loss": 0.013, "step": 45480 }, { "grad_norm": 0.13439740240573883, "learning_rate": 6.148889827065537e-05, "loss": 0.0152, "step": 45490 }, { "grad_norm": 0.1816321164369583, "learning_rate": 6.147280536221622e-05, "loss": 0.0106, "step": 45500 }, { "grad_norm": 0.19385530054569244, "learning_rate": 6.145671119912832e-05, "loss": 0.0111, "step": 45510 }, { "grad_norm": 0.1400647908449173, "learning_rate": 6.144061578315169e-05, "loss": 0.0126, "step": 45520 }, { "grad_norm": 0.16567476093769073, "learning_rate": 6.142451911604654e-05, "loss": 0.0177, "step": 45530 }, { "grad_norm": 0.15646904706954956, "learning_rate": 6.140842119957315e-05, "loss": 0.0128, "step": 45540 }, { "grad_norm": 0.1561107635498047, "learning_rate": 6.139232203549197e-05, "loss": 0.0134, "step": 45550 }, { "grad_norm": 0.14772774279117584, "learning_rate": 6.137622162556357e-05, "loss": 0.0121, "step": 45560 }, { "grad_norm": 0.14506559073925018, "learning_rate": 6.13601199715487e-05, "loss": 0.0125, "step": 45570 }, { "grad_norm": 0.12441595643758774, "learning_rate": 6.134401707520816e-05, "loss": 0.0124, "step": 45580 }, { "grad_norm": 0.14483900368213654, "learning_rate": 6.132791293830298e-05, "loss": 0.0108, "step": 45590 }, { "grad_norm": 0.1936749815940857, "learning_rate": 6.131180756259428e-05, "loss": 0.0142, "step": 45600 }, { "grad_norm": 0.20009252429008484, "learning_rate": 6.129570094984331e-05, "loss": 0.016, "step": 45610 }, { "grad_norm": 0.20875416696071625, "learning_rate": 6.127959310181145e-05, "loss": 0.0178, "step": 45620 }, { "grad_norm": 0.18631546199321747, "learning_rate": 6.126348402026026e-05, "loss": 0.0128, "step": 45630 }, { "grad_norm": 0.21068604290485382, "learning_rate": 6.12473737069514e-05, "loss": 0.0117, "step": 45640 }, { "grad_norm": 0.1542665660381317, "learning_rate": 6.123126216364665e-05, "loss": 0.012, "step": 45650 }, { "grad_norm": 0.171253502368927, "learning_rate": 6.121514939210797e-05, "loss": 0.0159, "step": 45660 }, { "grad_norm": 0.16686159372329712, "learning_rate": 6.119903539409741e-05, "loss": 0.0123, "step": 45670 }, { "grad_norm": 0.1655169576406479, "learning_rate": 6.118292017137716e-05, "loss": 0.0117, "step": 45680 }, { "grad_norm": 0.14119014143943787, "learning_rate": 6.116680372570959e-05, "loss": 0.0135, "step": 45690 }, { "grad_norm": 0.21349124610424042, "learning_rate": 6.115068605885713e-05, "loss": 0.017, "step": 45700 }, { "grad_norm": 0.16361714899539948, "learning_rate": 6.113456717258243e-05, "loss": 0.012, "step": 45710 }, { "grad_norm": 0.15348757803440094, "learning_rate": 6.11184470686482e-05, "loss": 0.0151, "step": 45720 }, { "grad_norm": 0.15368211269378662, "learning_rate": 6.110232574881731e-05, "loss": 0.0113, "step": 45730 }, { "grad_norm": 0.1846424788236618, "learning_rate": 6.108620321485277e-05, "loss": 0.0146, "step": 45740 }, { "grad_norm": 0.18195654451847076, "learning_rate": 6.107007946851773e-05, "loss": 0.0116, "step": 45750 }, { "grad_norm": 0.16418492794036865, "learning_rate": 6.105395451157542e-05, "loss": 0.0135, "step": 45760 }, { "grad_norm": 0.16106846928596497, "learning_rate": 6.103782834578928e-05, "loss": 0.0128, "step": 45770 }, { "grad_norm": 0.20830966532230377, "learning_rate": 6.102170097292281e-05, "loss": 0.012, "step": 45780 }, { "grad_norm": 0.19370408356189728, "learning_rate": 6.1005572394739716e-05, "loss": 0.0129, "step": 45790 }, { "grad_norm": 0.1869657188653946, "learning_rate": 6.0989442613003765e-05, "loss": 0.0148, "step": 45800 }, { "grad_norm": 0.2024681568145752, "learning_rate": 6.097331162947888e-05, "loss": 0.0145, "step": 45810 }, { "grad_norm": 0.162248894572258, "learning_rate": 6.095717944592914e-05, "loss": 0.0121, "step": 45820 }, { "grad_norm": 0.15887562930583954, "learning_rate": 6.094104606411873e-05, "loss": 0.0127, "step": 45830 }, { "grad_norm": 0.15785200893878937, "learning_rate": 6.0924911485811966e-05, "loss": 0.0143, "step": 45840 }, { "grad_norm": 0.1751554161310196, "learning_rate": 6.090877571277331e-05, "loss": 0.0117, "step": 45850 }, { "grad_norm": 0.20812827348709106, "learning_rate": 6.089263874676736e-05, "loss": 0.0133, "step": 45860 }, { "grad_norm": 0.24693186581134796, "learning_rate": 6.0876500589558796e-05, "loss": 0.0144, "step": 45870 }, { "grad_norm": 0.20441633462905884, "learning_rate": 6.086036124291248e-05, "loss": 0.0128, "step": 45880 }, { "grad_norm": 0.0950908213853836, "learning_rate": 6.084422070859339e-05, "loss": 0.0132, "step": 45890 }, { "grad_norm": 0.11691734194755554, "learning_rate": 6.082807898836663e-05, "loss": 0.0148, "step": 45900 }, { "grad_norm": 0.14580444991588593, "learning_rate": 6.081193608399742e-05, "loss": 0.0153, "step": 45910 }, { "grad_norm": 0.13586868345737457, "learning_rate": 6.0795791997251164e-05, "loss": 0.0138, "step": 45920 }, { "grad_norm": 0.23669275641441345, "learning_rate": 6.0779646729893294e-05, "loss": 0.0146, "step": 45930 }, { "grad_norm": 0.16555918753147125, "learning_rate": 6.0763500283689476e-05, "loss": 0.0141, "step": 45940 }, { "grad_norm": 0.19034914672374725, "learning_rate": 6.0747352660405455e-05, "loss": 0.0126, "step": 45950 }, { "grad_norm": 0.14073997735977173, "learning_rate": 6.073120386180709e-05, "loss": 0.0119, "step": 45960 }, { "grad_norm": 0.13091795146465302, "learning_rate": 6.0715053889660425e-05, "loss": 0.0125, "step": 45970 }, { "grad_norm": 0.12720616161823273, "learning_rate": 6.069890274573157e-05, "loss": 0.0103, "step": 45980 }, { "grad_norm": 0.138410747051239, "learning_rate": 6.068275043178679e-05, "loss": 0.0156, "step": 45990 }, { "grad_norm": 0.1826154589653015, "learning_rate": 6.066659694959248e-05, "loss": 0.0125, "step": 46000 }, { "grad_norm": 0.17699337005615234, "learning_rate": 6.0650442300915176e-05, "loss": 0.0119, "step": 46010 }, { "grad_norm": 0.19577732682228088, "learning_rate": 6.063428648752152e-05, "loss": 0.0122, "step": 46020 }, { "grad_norm": 0.18537676334381104, "learning_rate": 6.0618129511178266e-05, "loss": 0.0148, "step": 46030 }, { "grad_norm": 0.1262938231229782, "learning_rate": 6.060197137365234e-05, "loss": 0.0107, "step": 46040 }, { "grad_norm": 0.159225195646286, "learning_rate": 6.058581207671077e-05, "loss": 0.011, "step": 46050 }, { "grad_norm": 0.12140247970819473, "learning_rate": 6.056965162212072e-05, "loss": 0.01, "step": 46060 }, { "grad_norm": 0.18488775193691254, "learning_rate": 6.055349001164943e-05, "loss": 0.0129, "step": 46070 }, { "grad_norm": 0.14011014997959137, "learning_rate": 6.0537327247064347e-05, "loss": 0.012, "step": 46080 }, { "grad_norm": 0.17435012757778168, "learning_rate": 6.0521163330133014e-05, "loss": 0.0112, "step": 46090 }, { "grad_norm": 0.1848168522119522, "learning_rate": 6.050499826262306e-05, "loss": 0.0125, "step": 46100 }, { "grad_norm": 0.17031869292259216, "learning_rate": 6.0488832046302294e-05, "loss": 0.0103, "step": 46110 }, { "grad_norm": 0.2529486119747162, "learning_rate": 6.0472664682938626e-05, "loss": 0.0166, "step": 46120 }, { "grad_norm": 0.18049730360507965, "learning_rate": 6.045649617430009e-05, "loss": 0.0156, "step": 46130 }, { "grad_norm": 0.19397105276584625, "learning_rate": 6.0440326522154866e-05, "loss": 0.011, "step": 46140 }, { "grad_norm": 0.14667822420597076, "learning_rate": 6.0424155728271224e-05, "loss": 0.014, "step": 46150 }, { "grad_norm": 0.1819978505373001, "learning_rate": 6.040798379441758e-05, "loss": 0.016, "step": 46160 }, { "grad_norm": 0.14671792089939117, "learning_rate": 6.0391810722362485e-05, "loss": 0.0145, "step": 46170 }, { "grad_norm": 0.11980008333921432, "learning_rate": 6.037563651387458e-05, "loss": 0.013, "step": 46180 }, { "grad_norm": 0.16477982699871063, "learning_rate": 6.0359461170722666e-05, "loss": 0.0123, "step": 46190 }, { "grad_norm": 0.1197768896818161, "learning_rate": 6.034328469467566e-05, "loss": 0.0129, "step": 46200 }, { "grad_norm": 0.16836951673030853, "learning_rate": 6.0327107087502596e-05, "loss": 0.0128, "step": 46210 }, { "grad_norm": 0.14130602777004242, "learning_rate": 6.031092835097262e-05, "loss": 0.011, "step": 46220 }, { "grad_norm": 0.1721065640449524, "learning_rate": 6.0294748486855024e-05, "loss": 0.0107, "step": 46230 }, { "grad_norm": 0.13916079699993134, "learning_rate": 6.0278567496919216e-05, "loss": 0.0117, "step": 46240 }, { "grad_norm": 0.18064646422863007, "learning_rate": 6.026238538293472e-05, "loss": 0.0129, "step": 46250 }, { "grad_norm": 0.08440041542053223, "learning_rate": 6.024620214667118e-05, "loss": 0.0116, "step": 46260 }, { "grad_norm": 0.1410927027463913, "learning_rate": 6.0230017789898384e-05, "loss": 0.0122, "step": 46270 }, { "grad_norm": 0.11374266445636749, "learning_rate": 6.021383231438622e-05, "loss": 0.0103, "step": 46280 }, { "grad_norm": 0.14316348731517792, "learning_rate": 6.0197645721904704e-05, "loss": 0.0102, "step": 46290 }, { "grad_norm": 0.14736402034759521, "learning_rate": 6.0181458014224e-05, "loss": 0.0152, "step": 46300 }, { "grad_norm": 0.14501255750656128, "learning_rate": 6.016526919311435e-05, "loss": 0.012, "step": 46310 }, { "grad_norm": 0.15199950337409973, "learning_rate": 6.014907926034613e-05, "loss": 0.0118, "step": 46320 }, { "grad_norm": 0.18816058337688446, "learning_rate": 6.0132888217689875e-05, "loss": 0.0141, "step": 46330 }, { "grad_norm": 0.12112993746995926, "learning_rate": 6.01166960669162e-05, "loss": 0.0084, "step": 46340 }, { "grad_norm": 0.1652355194091797, "learning_rate": 6.0100502809795845e-05, "loss": 0.0128, "step": 46350 }, { "grad_norm": 0.1782083511352539, "learning_rate": 6.00843084480997e-05, "loss": 0.0131, "step": 46360 }, { "grad_norm": 0.15412980318069458, "learning_rate": 6.0068112983598736e-05, "loss": 0.0148, "step": 46370 }, { "grad_norm": 0.20592670142650604, "learning_rate": 6.005191641806407e-05, "loss": 0.0143, "step": 46380 }, { "grad_norm": 0.22390295565128326, "learning_rate": 6.003571875326694e-05, "loss": 0.0172, "step": 46390 }, { "grad_norm": 0.201199471950531, "learning_rate": 6.00195199909787e-05, "loss": 0.0149, "step": 46400 }, { "grad_norm": 0.15576577186584473, "learning_rate": 6.000332013297082e-05, "loss": 0.0132, "step": 46410 }, { "grad_norm": 0.17779727280139923, "learning_rate": 5.998711918101487e-05, "loss": 0.0126, "step": 46420 }, { "grad_norm": 0.11905194818973541, "learning_rate": 5.997091713688261e-05, "loss": 0.0132, "step": 46430 }, { "grad_norm": 0.16123366355895996, "learning_rate": 5.9954714002345836e-05, "loss": 0.0098, "step": 46440 }, { "grad_norm": 0.19781407713890076, "learning_rate": 5.993850977917649e-05, "loss": 0.0127, "step": 46450 }, { "grad_norm": 0.1968468725681305, "learning_rate": 5.992230446914667e-05, "loss": 0.0126, "step": 46460 }, { "grad_norm": 0.12461142987012863, "learning_rate": 5.990609807402855e-05, "loss": 0.0124, "step": 46470 }, { "grad_norm": 0.1558467000722885, "learning_rate": 5.988989059559443e-05, "loss": 0.011, "step": 46480 }, { "grad_norm": 0.1702202707529068, "learning_rate": 5.987368203561675e-05, "loss": 0.0102, "step": 46490 }, { "grad_norm": 0.16849954426288605, "learning_rate": 5.9857472395868055e-05, "loss": 0.0109, "step": 46500 }, { "grad_norm": 0.15837034583091736, "learning_rate": 5.9841261678120983e-05, "loss": 0.0118, "step": 46510 }, { "grad_norm": 0.16827726364135742, "learning_rate": 5.982504988414834e-05, "loss": 0.0119, "step": 46520 }, { "grad_norm": 0.15936122834682465, "learning_rate": 5.9808837015723015e-05, "loss": 0.0106, "step": 46530 }, { "grad_norm": 0.1586386263370514, "learning_rate": 5.9792623074618016e-05, "loss": 0.0145, "step": 46540 }, { "grad_norm": 0.18904317915439606, "learning_rate": 5.977640806260648e-05, "loss": 0.0121, "step": 46550 }, { "grad_norm": 0.20485664904117584, "learning_rate": 5.976019198146166e-05, "loss": 0.0153, "step": 46560 }, { "grad_norm": 0.19433970749378204, "learning_rate": 5.974397483295692e-05, "loss": 0.0125, "step": 46570 }, { "grad_norm": 0.15385332703590393, "learning_rate": 5.972775661886575e-05, "loss": 0.0125, "step": 46580 }, { "grad_norm": 0.11993692815303802, "learning_rate": 5.971153734096173e-05, "loss": 0.013, "step": 46590 }, { "grad_norm": 0.15031222999095917, "learning_rate": 5.969531700101859e-05, "loss": 0.0124, "step": 46600 }, { "grad_norm": 0.1315230131149292, "learning_rate": 5.9679095600810155e-05, "loss": 0.0114, "step": 46610 }, { "grad_norm": 0.1779295653104782, "learning_rate": 5.9662873142110384e-05, "loss": 0.0128, "step": 46620 }, { "grad_norm": 0.14917965233325958, "learning_rate": 5.964664962669333e-05, "loss": 0.0116, "step": 46630 }, { "grad_norm": 0.17326323688030243, "learning_rate": 5.9630425056333186e-05, "loss": 0.0139, "step": 46640 }, { "grad_norm": 0.1548752635717392, "learning_rate": 5.961419943280422e-05, "loss": 0.0095, "step": 46650 }, { "grad_norm": 0.12368377298116684, "learning_rate": 5.959797275788087e-05, "loss": 0.0126, "step": 46660 }, { "grad_norm": 0.15363480150699615, "learning_rate": 5.958174503333765e-05, "loss": 0.0107, "step": 46670 }, { "grad_norm": 0.14791752398014069, "learning_rate": 5.9565516260949195e-05, "loss": 0.0106, "step": 46680 }, { "grad_norm": 0.18637919425964355, "learning_rate": 5.9549286442490273e-05, "loss": 0.0115, "step": 46690 }, { "grad_norm": 0.19565938413143158, "learning_rate": 5.953305557973572e-05, "loss": 0.0136, "step": 46700 }, { "grad_norm": 0.2189941555261612, "learning_rate": 5.9516823674460564e-05, "loss": 0.0144, "step": 46710 }, { "grad_norm": 0.166320338845253, "learning_rate": 5.9500590728439875e-05, "loss": 0.0146, "step": 46720 }, { "grad_norm": 0.12298783659934998, "learning_rate": 5.948435674344886e-05, "loss": 0.0144, "step": 46730 }, { "grad_norm": 0.18853797018527985, "learning_rate": 5.946812172126285e-05, "loss": 0.0133, "step": 46740 }, { "grad_norm": 0.15816351771354675, "learning_rate": 5.945188566365729e-05, "loss": 0.0153, "step": 46750 }, { "grad_norm": 0.12298301607370377, "learning_rate": 5.943564857240773e-05, "loss": 0.0127, "step": 46760 }, { "grad_norm": 0.20238101482391357, "learning_rate": 5.941941044928981e-05, "loss": 0.014, "step": 46770 }, { "grad_norm": 0.2091396450996399, "learning_rate": 5.940317129607935e-05, "loss": 0.0121, "step": 46780 }, { "grad_norm": 0.18906673789024353, "learning_rate": 5.9386931114552204e-05, "loss": 0.0127, "step": 46790 }, { "grad_norm": 0.18245142698287964, "learning_rate": 5.937068990648438e-05, "loss": 0.0131, "step": 46800 }, { "grad_norm": 0.22299443185329437, "learning_rate": 5.935444767365199e-05, "loss": 0.0155, "step": 46810 }, { "grad_norm": 0.15019120275974274, "learning_rate": 5.933820441783129e-05, "loss": 0.0115, "step": 46820 }, { "grad_norm": 0.136471688747406, "learning_rate": 5.932196014079857e-05, "loss": 0.0136, "step": 46830 }, { "grad_norm": 0.13764813542366028, "learning_rate": 5.930571484433032e-05, "loss": 0.016, "step": 46840 }, { "grad_norm": 0.21739864349365234, "learning_rate": 5.928946853020309e-05, "loss": 0.0157, "step": 46850 }, { "grad_norm": 0.17540942132472992, "learning_rate": 5.927322120019352e-05, "loss": 0.0142, "step": 46860 }, { "grad_norm": 0.17131580412387848, "learning_rate": 5.925697285607844e-05, "loss": 0.0104, "step": 46870 }, { "grad_norm": 0.16665929555892944, "learning_rate": 5.924072349963472e-05, "loss": 0.0133, "step": 46880 }, { "grad_norm": 0.2580718994140625, "learning_rate": 5.922447313263938e-05, "loss": 0.0156, "step": 46890 }, { "grad_norm": 0.1353330761194229, "learning_rate": 5.920822175686952e-05, "loss": 0.0114, "step": 46900 }, { "grad_norm": 0.13511517643928528, "learning_rate": 5.919196937410239e-05, "loss": 0.0141, "step": 46910 }, { "grad_norm": 0.1794406622648239, "learning_rate": 5.917571598611529e-05, "loss": 0.0136, "step": 46920 }, { "grad_norm": 0.1940283626317978, "learning_rate": 5.9159461594685686e-05, "loss": 0.012, "step": 46930 }, { "grad_norm": 0.1535007506608963, "learning_rate": 5.914320620159114e-05, "loss": 0.0133, "step": 46940 }, { "grad_norm": 0.12425793707370758, "learning_rate": 5.912694980860931e-05, "loss": 0.0128, "step": 46950 }, { "grad_norm": 0.12793882191181183, "learning_rate": 5.9110692417517964e-05, "loss": 0.0119, "step": 46960 }, { "grad_norm": 0.17300084233283997, "learning_rate": 5.9094434030095e-05, "loss": 0.0097, "step": 46970 }, { "grad_norm": 0.18536557257175446, "learning_rate": 5.907817464811841e-05, "loss": 0.0102, "step": 46980 }, { "grad_norm": 0.1712390035390854, "learning_rate": 5.906191427336627e-05, "loss": 0.014, "step": 46990 }, { "grad_norm": 0.12357214093208313, "learning_rate": 5.9045652907616835e-05, "loss": 0.0145, "step": 47000 }, { "grad_norm": 0.23066623508930206, "learning_rate": 5.902939055264838e-05, "loss": 0.0138, "step": 47010 }, { "grad_norm": 0.1560330092906952, "learning_rate": 5.901312721023935e-05, "loss": 0.0144, "step": 47020 }, { "grad_norm": 0.13435427844524384, "learning_rate": 5.8996862882168294e-05, "loss": 0.0131, "step": 47030 }, { "grad_norm": 0.16876110434532166, "learning_rate": 5.8980597570213826e-05, "loss": 0.0143, "step": 47040 }, { "grad_norm": 0.14004816114902496, "learning_rate": 5.896433127615471e-05, "loss": 0.013, "step": 47050 }, { "grad_norm": 0.15992207825183868, "learning_rate": 5.894806400176981e-05, "loss": 0.0152, "step": 47060 }, { "grad_norm": 0.12119881063699722, "learning_rate": 5.893179574883808e-05, "loss": 0.0179, "step": 47070 }, { "grad_norm": 0.1675572544336319, "learning_rate": 5.8915526519138585e-05, "loss": 0.0124, "step": 47080 }, { "grad_norm": 0.13206838071346283, "learning_rate": 5.889925631445053e-05, "loss": 0.0129, "step": 47090 }, { "grad_norm": 0.12389795482158661, "learning_rate": 5.888298513655318e-05, "loss": 0.0116, "step": 47100 }, { "grad_norm": 0.17888136208057404, "learning_rate": 5.886671298722592e-05, "loss": 0.0103, "step": 47110 }, { "grad_norm": 0.15239427983760834, "learning_rate": 5.885043986824828e-05, "loss": 0.012, "step": 47120 }, { "grad_norm": 0.1935834437608719, "learning_rate": 5.883416578139982e-05, "loss": 0.0114, "step": 47130 }, { "grad_norm": 0.1497461050748825, "learning_rate": 5.881789072846029e-05, "loss": 0.0122, "step": 47140 }, { "grad_norm": 0.1456627994775772, "learning_rate": 5.8801614711209474e-05, "loss": 0.0131, "step": 47150 }, { "grad_norm": 0.1583576500415802, "learning_rate": 5.878533773142732e-05, "loss": 0.0176, "step": 47160 }, { "grad_norm": 0.1455964297056198, "learning_rate": 5.8769059790893834e-05, "loss": 0.0116, "step": 47170 }, { "grad_norm": 0.1337669938802719, "learning_rate": 5.8752780891389145e-05, "loss": 0.012, "step": 47180 }, { "grad_norm": 0.1255815029144287, "learning_rate": 5.873650103469351e-05, "loss": 0.0136, "step": 47190 }, { "grad_norm": 0.19873139262199402, "learning_rate": 5.872022022258725e-05, "loss": 0.0116, "step": 47200 }, { "grad_norm": 0.16358888149261475, "learning_rate": 5.8703938456850814e-05, "loss": 0.0126, "step": 47210 }, { "grad_norm": 0.16474750638008118, "learning_rate": 5.868765573926477e-05, "loss": 0.0119, "step": 47220 }, { "grad_norm": 0.1376192271709442, "learning_rate": 5.867137207160974e-05, "loss": 0.0108, "step": 47230 }, { "grad_norm": 0.15954189002513885, "learning_rate": 5.86550874556665e-05, "loss": 0.0104, "step": 47240 }, { "grad_norm": 0.20137321949005127, "learning_rate": 5.863880189321592e-05, "loss": 0.0124, "step": 47250 }, { "grad_norm": 0.13679352402687073, "learning_rate": 5.862251538603894e-05, "loss": 0.0115, "step": 47260 }, { "grad_norm": 0.18647441267967224, "learning_rate": 5.8606227935916656e-05, "loss": 0.0147, "step": 47270 }, { "grad_norm": 0.1845760941505432, "learning_rate": 5.858993954463021e-05, "loss": 0.0113, "step": 47280 }, { "grad_norm": 0.16568653285503387, "learning_rate": 5.8573650213960896e-05, "loss": 0.0114, "step": 47290 }, { "grad_norm": 0.12682627141475677, "learning_rate": 5.8557359945690084e-05, "loss": 0.0118, "step": 47300 }, { "grad_norm": 0.18816673755645752, "learning_rate": 5.8541068741599246e-05, "loss": 0.0102, "step": 47310 }, { "grad_norm": 0.11882372945547104, "learning_rate": 5.8524776603469985e-05, "loss": 0.0111, "step": 47320 }, { "grad_norm": 0.1624467670917511, "learning_rate": 5.850848353308397e-05, "loss": 0.0117, "step": 47330 }, { "grad_norm": 0.23817236721515656, "learning_rate": 5.849218953222297e-05, "loss": 0.0143, "step": 47340 }, { "grad_norm": 0.14501258730888367, "learning_rate": 5.847589460266891e-05, "loss": 0.0101, "step": 47350 }, { "grad_norm": 0.13738703727722168, "learning_rate": 5.845959874620375e-05, "loss": 0.0097, "step": 47360 }, { "grad_norm": 0.1184304878115654, "learning_rate": 5.84433019646096e-05, "loss": 0.0103, "step": 47370 }, { "grad_norm": 0.17896269261837006, "learning_rate": 5.842700425966863e-05, "loss": 0.012, "step": 47380 }, { "grad_norm": 0.1633325219154358, "learning_rate": 5.841070563316315e-05, "loss": 0.012, "step": 47390 }, { "grad_norm": 0.16683469712734222, "learning_rate": 5.839440608687554e-05, "loss": 0.0137, "step": 47400 }, { "grad_norm": 0.17167335748672485, "learning_rate": 5.837810562258831e-05, "loss": 0.0112, "step": 47410 }, { "grad_norm": 0.13749897480010986, "learning_rate": 5.836180424208405e-05, "loss": 0.0114, "step": 47420 }, { "grad_norm": 0.16521088778972626, "learning_rate": 5.834550194714543e-05, "loss": 0.0138, "step": 47430 }, { "grad_norm": 0.1213267594575882, "learning_rate": 5.832919873955528e-05, "loss": 0.0118, "step": 47440 }, { "grad_norm": 0.21441803872585297, "learning_rate": 5.831289462109648e-05, "loss": 0.0119, "step": 47450 }, { "grad_norm": 0.13424327969551086, "learning_rate": 5.829658959355202e-05, "loss": 0.013, "step": 47460 }, { "grad_norm": 0.1373283416032791, "learning_rate": 5.8280283658704995e-05, "loss": 0.0112, "step": 47470 }, { "grad_norm": 0.15007083117961884, "learning_rate": 5.826397681833859e-05, "loss": 0.0162, "step": 47480 }, { "grad_norm": 0.17082704603672028, "learning_rate": 5.824766907423612e-05, "loss": 0.0136, "step": 47490 }, { "grad_norm": 0.15786221623420715, "learning_rate": 5.8231360428180945e-05, "loss": 0.012, "step": 47500 }, { "grad_norm": 0.12275794893503189, "learning_rate": 5.821505088195658e-05, "loss": 0.0121, "step": 47510 }, { "grad_norm": 0.17519839107990265, "learning_rate": 5.819874043734661e-05, "loss": 0.0132, "step": 47520 }, { "grad_norm": 0.26351237297058105, "learning_rate": 5.8182429096134695e-05, "loss": 0.011, "step": 47530 }, { "grad_norm": 0.21864335238933563, "learning_rate": 5.816611686010465e-05, "loss": 0.0129, "step": 47540 }, { "grad_norm": 0.17706866562366486, "learning_rate": 5.814980373104033e-05, "loss": 0.0113, "step": 47550 }, { "grad_norm": 0.1389249861240387, "learning_rate": 5.813348971072572e-05, "loss": 0.0153, "step": 47560 }, { "grad_norm": 0.1503942310810089, "learning_rate": 5.811717480094492e-05, "loss": 0.0135, "step": 47570 }, { "grad_norm": 0.15521550178527832, "learning_rate": 5.810085900348209e-05, "loss": 0.0142, "step": 47580 }, { "grad_norm": 0.19814954698085785, "learning_rate": 5.8084542320121483e-05, "loss": 0.0134, "step": 47590 }, { "grad_norm": 0.14177289605140686, "learning_rate": 5.8068224752647497e-05, "loss": 0.0132, "step": 47600 }, { "grad_norm": 0.13614186644554138, "learning_rate": 5.805190630284457e-05, "loss": 0.014, "step": 47610 }, { "grad_norm": 0.17407315969467163, "learning_rate": 5.803558697249729e-05, "loss": 0.0102, "step": 47620 }, { "grad_norm": 0.20050042867660522, "learning_rate": 5.8019266763390276e-05, "loss": 0.0133, "step": 47630 }, { "grad_norm": 0.1435883790254593, "learning_rate": 5.800294567730833e-05, "loss": 0.0138, "step": 47640 }, { "grad_norm": 0.14806853234767914, "learning_rate": 5.798662371603628e-05, "loss": 0.0126, "step": 47650 }, { "grad_norm": 0.17323654890060425, "learning_rate": 5.797030088135904e-05, "loss": 0.0101, "step": 47660 }, { "grad_norm": 0.16198544204235077, "learning_rate": 5.7953977175061705e-05, "loss": 0.0107, "step": 47670 }, { "grad_norm": 0.1753024160861969, "learning_rate": 5.793765259892938e-05, "loss": 0.0153, "step": 47680 }, { "grad_norm": 0.22689363360404968, "learning_rate": 5.792132715474729e-05, "loss": 0.0138, "step": 47690 }, { "grad_norm": 0.18094724416732788, "learning_rate": 5.790500084430078e-05, "loss": 0.0122, "step": 47700 }, { "grad_norm": 0.14624589681625366, "learning_rate": 5.7888673669375274e-05, "loss": 0.0116, "step": 47710 }, { "grad_norm": 0.15570464730262756, "learning_rate": 5.787234563175625e-05, "loss": 0.0109, "step": 47720 }, { "grad_norm": 0.12394989281892776, "learning_rate": 5.7856016733229355e-05, "loss": 0.0096, "step": 47730 }, { "grad_norm": 0.17319244146347046, "learning_rate": 5.7839686975580297e-05, "loss": 0.0123, "step": 47740 }, { "grad_norm": 0.21825093030929565, "learning_rate": 5.782335636059484e-05, "loss": 0.0122, "step": 47750 }, { "grad_norm": 0.13199803233146667, "learning_rate": 5.780702489005889e-05, "loss": 0.0125, "step": 47760 }, { "grad_norm": 0.15413188934326172, "learning_rate": 5.779069256575845e-05, "loss": 0.0143, "step": 47770 }, { "grad_norm": 0.20598575472831726, "learning_rate": 5.7774359389479574e-05, "loss": 0.0131, "step": 47780 }, { "grad_norm": 0.18158817291259766, "learning_rate": 5.775802536300845e-05, "loss": 0.0137, "step": 47790 }, { "grad_norm": 0.14280030131340027, "learning_rate": 5.774169048813134e-05, "loss": 0.0131, "step": 47800 }, { "grad_norm": 0.19191648066043854, "learning_rate": 5.77253547666346e-05, "loss": 0.0117, "step": 47810 }, { "grad_norm": 0.15009810030460358, "learning_rate": 5.770901820030465e-05, "loss": 0.0129, "step": 47820 }, { "grad_norm": 0.1293126344680786, "learning_rate": 5.769268079092809e-05, "loss": 0.0098, "step": 47830 }, { "grad_norm": 0.1681293249130249, "learning_rate": 5.767634254029151e-05, "loss": 0.0127, "step": 47840 }, { "grad_norm": 0.12269513309001923, "learning_rate": 5.7660003450181655e-05, "loss": 0.0113, "step": 47850 }, { "grad_norm": 0.1500149518251419, "learning_rate": 5.764366352238534e-05, "loss": 0.0093, "step": 47860 }, { "grad_norm": 0.11378619819879532, "learning_rate": 5.7627322758689474e-05, "loss": 0.0132, "step": 47870 }, { "grad_norm": 0.14603890478610992, "learning_rate": 5.761098116088105e-05, "loss": 0.0107, "step": 47880 }, { "grad_norm": 0.14894291758537292, "learning_rate": 5.759463873074717e-05, "loss": 0.0117, "step": 47890 }, { "grad_norm": 0.14138299226760864, "learning_rate": 5.757829547007504e-05, "loss": 0.0135, "step": 47900 }, { "grad_norm": 0.1937246024608612, "learning_rate": 5.756195138065189e-05, "loss": 0.0138, "step": 47910 }, { "grad_norm": 0.13145306706428528, "learning_rate": 5.754560646426511e-05, "loss": 0.0113, "step": 47920 }, { "grad_norm": 0.16455113887786865, "learning_rate": 5.752926072270216e-05, "loss": 0.0119, "step": 47930 }, { "grad_norm": 0.11550149321556091, "learning_rate": 5.7512914157750563e-05, "loss": 0.0092, "step": 47940 }, { "grad_norm": 0.1202947124838829, "learning_rate": 5.749656677119798e-05, "loss": 0.0105, "step": 47950 }, { "grad_norm": 0.15923361480236053, "learning_rate": 5.7480218564832125e-05, "loss": 0.0096, "step": 47960 }, { "grad_norm": 0.14687298238277435, "learning_rate": 5.746386954044082e-05, "loss": 0.0127, "step": 47970 }, { "grad_norm": 0.12775547802448273, "learning_rate": 5.744751969981195e-05, "loss": 0.0114, "step": 47980 }, { "grad_norm": 0.16214284300804138, "learning_rate": 5.7431169044733526e-05, "loss": 0.0129, "step": 47990 }, { "grad_norm": 0.18119202554225922, "learning_rate": 5.741481757699364e-05, "loss": 0.0131, "step": 48000 }, { "grad_norm": 0.1876116245985031, "learning_rate": 5.7398465298380434e-05, "loss": 0.0124, "step": 48010 }, { "grad_norm": 0.16832394897937775, "learning_rate": 5.7382112210682193e-05, "loss": 0.0085, "step": 48020 }, { "grad_norm": 0.14099258184432983, "learning_rate": 5.7365758315687266e-05, "loss": 0.0132, "step": 48030 }, { "grad_norm": 0.1815902441740036, "learning_rate": 5.734940361518407e-05, "loss": 0.0129, "step": 48040 }, { "grad_norm": 0.16761726140975952, "learning_rate": 5.733304811096116e-05, "loss": 0.0106, "step": 48050 }, { "grad_norm": 0.18984277546405792, "learning_rate": 5.731669180480713e-05, "loss": 0.0117, "step": 48060 }, { "grad_norm": 0.12436933070421219, "learning_rate": 5.730033469851067e-05, "loss": 0.0118, "step": 48070 }, { "grad_norm": 0.2107534110546112, "learning_rate": 5.7283976793860607e-05, "loss": 0.0141, "step": 48080 }, { "grad_norm": 0.2094571888446808, "learning_rate": 5.726761809264577e-05, "loss": 0.0127, "step": 48090 }, { "grad_norm": 0.15179504454135895, "learning_rate": 5.7251258596655155e-05, "loss": 0.0119, "step": 48100 }, { "grad_norm": 0.17054018378257751, "learning_rate": 5.72348983076778e-05, "loss": 0.0137, "step": 48110 }, { "grad_norm": 0.14557147026062012, "learning_rate": 5.7218537227502854e-05, "loss": 0.0144, "step": 48120 }, { "grad_norm": 0.17707881331443787, "learning_rate": 5.720217535791951e-05, "loss": 0.0123, "step": 48130 }, { "grad_norm": 0.17954497039318085, "learning_rate": 5.718581270071711e-05, "loss": 0.0115, "step": 48140 }, { "grad_norm": 0.16628973186016083, "learning_rate": 5.716944925768505e-05, "loss": 0.0102, "step": 48150 }, { "grad_norm": 0.1452990472316742, "learning_rate": 5.7153085030612786e-05, "loss": 0.0109, "step": 48160 }, { "grad_norm": 0.09975583106279373, "learning_rate": 5.713672002128989e-05, "loss": 0.0141, "step": 48170 }, { "grad_norm": 0.17155668139457703, "learning_rate": 5.712035423150602e-05, "loss": 0.0129, "step": 48180 }, { "grad_norm": 0.16623125970363617, "learning_rate": 5.710398766305094e-05, "loss": 0.0124, "step": 48190 }, { "grad_norm": 0.14416462182998657, "learning_rate": 5.708762031771442e-05, "loss": 0.0131, "step": 48200 }, { "grad_norm": 0.13082359731197357, "learning_rate": 5.707125219728642e-05, "loss": 0.0114, "step": 48210 }, { "grad_norm": 0.09435898810625076, "learning_rate": 5.7054883303556905e-05, "loss": 0.0115, "step": 48220 }, { "grad_norm": 0.14093542098999023, "learning_rate": 5.703851363831595e-05, "loss": 0.011, "step": 48230 }, { "grad_norm": 0.1500503569841385, "learning_rate": 5.702214320335374e-05, "loss": 0.015, "step": 48240 }, { "grad_norm": 0.1791047602891922, "learning_rate": 5.7005772000460514e-05, "loss": 0.0105, "step": 48250 }, { "grad_norm": 0.11564554274082184, "learning_rate": 5.698940003142656e-05, "loss": 0.0094, "step": 48260 }, { "grad_norm": 0.10723365098237991, "learning_rate": 5.697302729804236e-05, "loss": 0.0096, "step": 48270 }, { "grad_norm": 0.17553433775901794, "learning_rate": 5.695665380209837e-05, "loss": 0.0137, "step": 48280 }, { "grad_norm": 0.16419680416584015, "learning_rate": 5.6940279545385165e-05, "loss": 0.0109, "step": 48290 }, { "grad_norm": 0.12814556062221527, "learning_rate": 5.692390452969344e-05, "loss": 0.0112, "step": 48300 }, { "grad_norm": 0.16635428369045258, "learning_rate": 5.690752875681392e-05, "loss": 0.0122, "step": 48310 }, { "grad_norm": 0.16989989578723907, "learning_rate": 5.6891152228537435e-05, "loss": 0.0117, "step": 48320 }, { "grad_norm": 0.14246328175067902, "learning_rate": 5.687477494665492e-05, "loss": 0.0099, "step": 48330 }, { "grad_norm": 0.16894984245300293, "learning_rate": 5.685839691295733e-05, "loss": 0.01, "step": 48340 }, { "grad_norm": 0.27827370166778564, "learning_rate": 5.6842018129235786e-05, "loss": 0.0151, "step": 48350 }, { "grad_norm": 0.19000175595283508, "learning_rate": 5.6825638597281404e-05, "loss": 0.0149, "step": 48360 }, { "grad_norm": 0.18179179728031158, "learning_rate": 5.680925831888546e-05, "loss": 0.0192, "step": 48370 }, { "grad_norm": 0.1496829241514206, "learning_rate": 5.6792877295839274e-05, "loss": 0.0116, "step": 48380 }, { "grad_norm": 0.13570012152194977, "learning_rate": 5.6776495529934224e-05, "loss": 0.0126, "step": 48390 }, { "grad_norm": 0.1612938791513443, "learning_rate": 5.6760113022961824e-05, "loss": 0.01, "step": 48400 }, { "grad_norm": 0.14326384663581848, "learning_rate": 5.6743729776713617e-05, "loss": 0.0127, "step": 48410 }, { "grad_norm": 0.12388412654399872, "learning_rate": 5.672734579298126e-05, "loss": 0.0101, "step": 48420 }, { "grad_norm": 0.1836928129196167, "learning_rate": 5.671096107355649e-05, "loss": 0.0113, "step": 48430 }, { "grad_norm": 0.17221082746982574, "learning_rate": 5.66945756202311e-05, "loss": 0.0182, "step": 48440 }, { "grad_norm": 0.2350381463766098, "learning_rate": 5.667818943479699e-05, "loss": 0.0154, "step": 48450 }, { "grad_norm": 0.1881817877292633, "learning_rate": 5.666180251904612e-05, "loss": 0.0139, "step": 48460 }, { "grad_norm": 0.13005182147026062, "learning_rate": 5.6645414874770555e-05, "loss": 0.0126, "step": 48470 }, { "grad_norm": 0.11868749558925629, "learning_rate": 5.66290265037624e-05, "loss": 0.0127, "step": 48480 }, { "grad_norm": 0.140790656208992, "learning_rate": 5.661263740781386e-05, "loss": 0.0128, "step": 48490 }, { "grad_norm": 0.1245972290635109, "learning_rate": 5.6596247588717254e-05, "loss": 0.0122, "step": 48500 }, { "grad_norm": 0.1857285499572754, "learning_rate": 5.6579857048264926e-05, "loss": 0.0117, "step": 48510 }, { "grad_norm": 0.14993833005428314, "learning_rate": 5.6563465788249314e-05, "loss": 0.0125, "step": 48520 }, { "grad_norm": 0.22657541930675507, "learning_rate": 5.6547073810462956e-05, "loss": 0.011, "step": 48530 }, { "grad_norm": 0.15491805970668793, "learning_rate": 5.653068111669846e-05, "loss": 0.0119, "step": 48540 }, { "grad_norm": 0.12433819472789764, "learning_rate": 5.651428770874848e-05, "loss": 0.0105, "step": 48550 }, { "grad_norm": 0.14863041043281555, "learning_rate": 5.64978935884058e-05, "loss": 0.0096, "step": 48560 }, { "grad_norm": 0.11960628628730774, "learning_rate": 5.6481498757463244e-05, "loss": 0.0151, "step": 48570 }, { "grad_norm": 0.14158084988594055, "learning_rate": 5.646510321771373e-05, "loss": 0.0125, "step": 48580 }, { "grad_norm": 0.15643340349197388, "learning_rate": 5.644870697095024e-05, "loss": 0.0146, "step": 48590 }, { "grad_norm": 0.1178959384560585, "learning_rate": 5.643231001896586e-05, "loss": 0.009, "step": 48600 }, { "grad_norm": 0.1748710423707962, "learning_rate": 5.6415912363553726e-05, "loss": 0.0118, "step": 48610 }, { "grad_norm": 0.20475135743618011, "learning_rate": 5.639951400650706e-05, "loss": 0.0136, "step": 48620 }, { "grad_norm": 0.21240124106407166, "learning_rate": 5.6383114949619165e-05, "loss": 0.0109, "step": 48630 }, { "grad_norm": 0.1595284789800644, "learning_rate": 5.636671519468342e-05, "loss": 0.0136, "step": 48640 }, { "grad_norm": 0.1536012887954712, "learning_rate": 5.635031474349327e-05, "loss": 0.0103, "step": 48650 }, { "grad_norm": 0.17793098092079163, "learning_rate": 5.6333913597842246e-05, "loss": 0.013, "step": 48660 }, { "grad_norm": 0.15666411817073822, "learning_rate": 5.6317511759523955e-05, "loss": 0.0119, "step": 48670 }, { "grad_norm": 0.15148131549358368, "learning_rate": 5.630110923033207e-05, "loss": 0.0126, "step": 48680 }, { "grad_norm": 0.1389148235321045, "learning_rate": 5.628470601206036e-05, "loss": 0.0112, "step": 48690 }, { "grad_norm": 0.14397571980953217, "learning_rate": 5.626830210650266e-05, "loss": 0.0116, "step": 48700 }, { "grad_norm": 0.18307995796203613, "learning_rate": 5.625189751545285e-05, "loss": 0.014, "step": 48710 }, { "grad_norm": 0.1536867469549179, "learning_rate": 5.6235492240704936e-05, "loss": 0.012, "step": 48720 }, { "grad_norm": 0.14094310998916626, "learning_rate": 5.621908628405296e-05, "loss": 0.0103, "step": 48730 }, { "grad_norm": 0.15714140236377716, "learning_rate": 5.620267964729106e-05, "loss": 0.0163, "step": 48740 }, { "grad_norm": 0.1826382428407669, "learning_rate": 5.618627233221344e-05, "loss": 0.0136, "step": 48750 }, { "grad_norm": 0.17226527631282806, "learning_rate": 5.616986434061438e-05, "loss": 0.0121, "step": 48760 }, { "grad_norm": 0.15427540242671967, "learning_rate": 5.615345567428822e-05, "loss": 0.0118, "step": 48770 }, { "grad_norm": 0.18614529073238373, "learning_rate": 5.613704633502941e-05, "loss": 0.0121, "step": 48780 }, { "grad_norm": 0.11761874705553055, "learning_rate": 5.612063632463245e-05, "loss": 0.0115, "step": 48790 }, { "grad_norm": 0.12409191578626633, "learning_rate": 5.610422564489188e-05, "loss": 0.0095, "step": 48800 }, { "grad_norm": 0.14608249068260193, "learning_rate": 5.608781429760239e-05, "loss": 0.0111, "step": 48810 }, { "grad_norm": 0.19196252524852753, "learning_rate": 5.607140228455866e-05, "loss": 0.0135, "step": 48820 }, { "grad_norm": 0.10677681863307953, "learning_rate": 5.605498960755553e-05, "loss": 0.0122, "step": 48830 }, { "grad_norm": 0.13430410623550415, "learning_rate": 5.603857626838782e-05, "loss": 0.0108, "step": 48840 }, { "grad_norm": 0.13743910193443298, "learning_rate": 5.60221622688505e-05, "loss": 0.0129, "step": 48850 }, { "grad_norm": 0.20692497491836548, "learning_rate": 5.6005747610738565e-05, "loss": 0.0108, "step": 48860 }, { "grad_norm": 0.12920545041561127, "learning_rate": 5.59893322958471e-05, "loss": 0.0144, "step": 48870 }, { "grad_norm": 0.20993468165397644, "learning_rate": 5.5972916325971256e-05, "loss": 0.0156, "step": 48880 }, { "grad_norm": 0.20294415950775146, "learning_rate": 5.595649970290628e-05, "loss": 0.0112, "step": 48890 }, { "grad_norm": 0.15615570545196533, "learning_rate": 5.5940082428447426e-05, "loss": 0.014, "step": 48900 }, { "grad_norm": 0.13924142718315125, "learning_rate": 5.592366450439012e-05, "loss": 0.0105, "step": 48910 }, { "grad_norm": 0.17506028711795807, "learning_rate": 5.590724593252975e-05, "loss": 0.0109, "step": 48920 }, { "grad_norm": 0.14709098637104034, "learning_rate": 5.589082671466184e-05, "loss": 0.0114, "step": 48930 }, { "grad_norm": 0.15568208694458008, "learning_rate": 5.587440685258199e-05, "loss": 0.0105, "step": 48940 }, { "grad_norm": 0.1439635157585144, "learning_rate": 5.585798634808583e-05, "loss": 0.0123, "step": 48950 }, { "grad_norm": 0.09074661135673523, "learning_rate": 5.584156520296909e-05, "loss": 0.0098, "step": 48960 }, { "grad_norm": 0.1809890866279602, "learning_rate": 5.582514341902757e-05, "loss": 0.0138, "step": 48970 }, { "grad_norm": 0.17645660042762756, "learning_rate": 5.580872099805713e-05, "loss": 0.0118, "step": 48980 }, { "grad_norm": 0.18754428625106812, "learning_rate": 5.5792297941853674e-05, "loss": 0.0122, "step": 48990 }, { "grad_norm": 0.21440914273262024, "learning_rate": 5.5775874252213247e-05, "loss": 0.0165, "step": 49000 }, { "grad_norm": 0.22072254121303558, "learning_rate": 5.575944993093189e-05, "loss": 0.012, "step": 49010 }, { "grad_norm": 0.16984586417675018, "learning_rate": 5.574302497980574e-05, "loss": 0.0119, "step": 49020 }, { "grad_norm": 0.11987891048192978, "learning_rate": 5.5726599400631005e-05, "loss": 0.0116, "step": 49030 }, { "grad_norm": 0.13863982260227203, "learning_rate": 5.571017319520399e-05, "loss": 0.0123, "step": 49040 }, { "grad_norm": 0.1722998470067978, "learning_rate": 5.5693746365320984e-05, "loss": 0.0122, "step": 49050 }, { "grad_norm": 0.17215338349342346, "learning_rate": 5.567731891277848e-05, "loss": 0.0132, "step": 49060 }, { "grad_norm": 0.1625255048274994, "learning_rate": 5.566089083937288e-05, "loss": 0.0148, "step": 49070 }, { "grad_norm": 0.2068624496459961, "learning_rate": 5.564446214690079e-05, "loss": 0.0164, "step": 49080 }, { "grad_norm": 0.16798146069049835, "learning_rate": 5.56280328371588e-05, "loss": 0.0135, "step": 49090 }, { "grad_norm": 0.12129691988229752, "learning_rate": 5.5611602911943596e-05, "loss": 0.0108, "step": 49100 }, { "grad_norm": 0.14179068803787231, "learning_rate": 5.559517237305194e-05, "loss": 0.014, "step": 49110 }, { "grad_norm": 0.16152425110340118, "learning_rate": 5.557874122228064e-05, "loss": 0.0129, "step": 49120 }, { "grad_norm": 0.18081346154212952, "learning_rate": 5.55623094614266e-05, "loss": 0.0169, "step": 49130 }, { "grad_norm": 0.16698406636714935, "learning_rate": 5.5545877092286743e-05, "loss": 0.0133, "step": 49140 }, { "grad_norm": 0.15781190991401672, "learning_rate": 5.552944411665811e-05, "loss": 0.0121, "step": 49150 }, { "grad_norm": 0.16990287601947784, "learning_rate": 5.5513010536337795e-05, "loss": 0.0094, "step": 49160 }, { "grad_norm": 0.13691021502017975, "learning_rate": 5.5496576353122934e-05, "loss": 0.0122, "step": 49170 }, { "grad_norm": 0.13857890665531158, "learning_rate": 5.548014156881074e-05, "loss": 0.0091, "step": 49180 }, { "grad_norm": 0.1312515139579773, "learning_rate": 5.5463706185198494e-05, "loss": 0.0171, "step": 49190 }, { "grad_norm": 0.1995294988155365, "learning_rate": 5.544727020408358e-05, "loss": 0.0112, "step": 49200 }, { "grad_norm": 0.15317381918430328, "learning_rate": 5.5430833627263366e-05, "loss": 0.0121, "step": 49210 }, { "grad_norm": 0.16385547816753387, "learning_rate": 5.5414396456535354e-05, "loss": 0.012, "step": 49220 }, { "grad_norm": 0.1803593635559082, "learning_rate": 5.539795869369709e-05, "loss": 0.013, "step": 49230 }, { "grad_norm": 0.1584702879190445, "learning_rate": 5.538152034054618e-05, "loss": 0.0118, "step": 49240 }, { "grad_norm": 0.12262649089097977, "learning_rate": 5.536508139888028e-05, "loss": 0.0097, "step": 49250 }, { "grad_norm": 0.12022620439529419, "learning_rate": 5.534864187049716e-05, "loss": 0.0124, "step": 49260 }, { "grad_norm": 0.1272554099559784, "learning_rate": 5.5332201757194615e-05, "loss": 0.0106, "step": 49270 }, { "grad_norm": 0.16717778146266937, "learning_rate": 5.531576106077049e-05, "loss": 0.0116, "step": 49280 }, { "grad_norm": 0.16570058465003967, "learning_rate": 5.529931978302272e-05, "loss": 0.0098, "step": 49290 }, { "grad_norm": 0.12050623446702957, "learning_rate": 5.528287792574932e-05, "loss": 0.0115, "step": 49300 }, { "grad_norm": 0.20149904489517212, "learning_rate": 5.5266435490748324e-05, "loss": 0.0132, "step": 49310 }, { "grad_norm": 0.14124377071857452, "learning_rate": 5.524999247981787e-05, "loss": 0.0113, "step": 49320 }, { "grad_norm": 0.1584654599428177, "learning_rate": 5.523354889475613e-05, "loss": 0.0132, "step": 49330 }, { "grad_norm": 0.1446266919374466, "learning_rate": 5.521710473736134e-05, "loss": 0.0124, "step": 49340 }, { "grad_norm": 0.1291535198688507, "learning_rate": 5.5200660009431826e-05, "loss": 0.0141, "step": 49350 }, { "grad_norm": 0.11543772369623184, "learning_rate": 5.518421471276596e-05, "loss": 0.0128, "step": 49360 }, { "grad_norm": 0.14560140669345856, "learning_rate": 5.5167768849162174e-05, "loss": 0.015, "step": 49370 }, { "grad_norm": 0.16090917587280273, "learning_rate": 5.515132242041893e-05, "loss": 0.0113, "step": 49380 }, { "grad_norm": 0.14814302325248718, "learning_rate": 5.513487542833483e-05, "loss": 0.0146, "step": 49390 }, { "grad_norm": 0.12432004511356354, "learning_rate": 5.5118427874708466e-05, "loss": 0.011, "step": 49400 }, { "grad_norm": 0.21529927849769592, "learning_rate": 5.510197976133853e-05, "loss": 0.0107, "step": 49410 }, { "grad_norm": 0.15627504885196686, "learning_rate": 5.508553109002376e-05, "loss": 0.011, "step": 49420 }, { "grad_norm": 0.1781490594148636, "learning_rate": 5.5069081862562957e-05, "loss": 0.0132, "step": 49430 }, { "grad_norm": 0.19650660455226898, "learning_rate": 5.5052632080754965e-05, "loss": 0.016, "step": 49440 }, { "grad_norm": 0.17004097998142242, "learning_rate": 5.503618174639874e-05, "loss": 0.0113, "step": 49450 }, { "grad_norm": 0.18950538337230682, "learning_rate": 5.501973086129325e-05, "loss": 0.0121, "step": 49460 }, { "grad_norm": 0.12951058149337769, "learning_rate": 5.500327942723753e-05, "loss": 0.0114, "step": 49470 }, { "grad_norm": 0.15315119922161102, "learning_rate": 5.498682744603071e-05, "loss": 0.012, "step": 49480 }, { "grad_norm": 0.12380637228488922, "learning_rate": 5.497037491947194e-05, "loss": 0.0112, "step": 49490 }, { "grad_norm": 0.1579871028661728, "learning_rate": 5.4953921849360424e-05, "loss": 0.0114, "step": 49500 }, { "grad_norm": 0.12770795822143555, "learning_rate": 5.493746823749547e-05, "loss": 0.0114, "step": 49510 }, { "grad_norm": 0.1453772932291031, "learning_rate": 5.49210140856764e-05, "loss": 0.0169, "step": 49520 }, { "grad_norm": 0.13382573425769806, "learning_rate": 5.4904559395702647e-05, "loss": 0.0101, "step": 49530 }, { "grad_norm": 0.17808468639850616, "learning_rate": 5.488810416937364e-05, "loss": 0.0107, "step": 49540 }, { "grad_norm": 0.14567716419696808, "learning_rate": 5.4871648408488926e-05, "loss": 0.0089, "step": 49550 }, { "grad_norm": 0.1182149201631546, "learning_rate": 5.485519211484807e-05, "loss": 0.0116, "step": 49560 }, { "grad_norm": 0.18268077075481415, "learning_rate": 5.4838735290250676e-05, "loss": 0.0136, "step": 49570 }, { "grad_norm": 0.17667196691036224, "learning_rate": 5.482227793649649e-05, "loss": 0.0116, "step": 49580 }, { "grad_norm": 0.12183640152215958, "learning_rate": 5.480582005538524e-05, "loss": 0.0098, "step": 49590 }, { "grad_norm": 0.17050902545452118, "learning_rate": 5.478936164871671e-05, "loss": 0.0113, "step": 49600 }, { "grad_norm": 0.14339900016784668, "learning_rate": 5.477290271829081e-05, "loss": 0.0152, "step": 49610 }, { "grad_norm": 0.17549827694892883, "learning_rate": 5.475644326590744e-05, "loss": 0.0108, "step": 49620 }, { "grad_norm": 0.14898766577243805, "learning_rate": 5.473998329336658e-05, "loss": 0.0113, "step": 49630 }, { "grad_norm": 0.17000755667686462, "learning_rate": 5.4723522802468286e-05, "loss": 0.0088, "step": 49640 }, { "grad_norm": 0.14157159626483917, "learning_rate": 5.470706179501264e-05, "loss": 0.0096, "step": 49650 }, { "grad_norm": 0.15833304822444916, "learning_rate": 5.4690600272799776e-05, "loss": 0.0109, "step": 49660 }, { "grad_norm": 0.12297830730676651, "learning_rate": 5.467413823762993e-05, "loss": 0.01, "step": 49670 }, { "grad_norm": 0.1612798571586609, "learning_rate": 5.465767569130335e-05, "loss": 0.0112, "step": 49680 }, { "grad_norm": 0.16642890870571136, "learning_rate": 5.464121263562036e-05, "loss": 0.0098, "step": 49690 }, { "grad_norm": 0.13582035899162292, "learning_rate": 5.4624749072381343e-05, "loss": 0.0143, "step": 49700 }, { "grad_norm": 0.10354314744472504, "learning_rate": 5.460828500338672e-05, "loss": 0.0097, "step": 49710 }, { "grad_norm": 0.09557808935642242, "learning_rate": 5.459182043043698e-05, "loss": 0.0097, "step": 49720 }, { "grad_norm": 0.13805925846099854, "learning_rate": 5.457535535533265e-05, "loss": 0.0098, "step": 49730 }, { "grad_norm": 0.15327005088329315, "learning_rate": 5.4558889779874334e-05, "loss": 0.0113, "step": 49740 }, { "grad_norm": 0.1433551162481308, "learning_rate": 5.454242370586269e-05, "loss": 0.0123, "step": 49750 }, { "grad_norm": 0.12770144641399384, "learning_rate": 5.452595713509843e-05, "loss": 0.0098, "step": 49760 }, { "grad_norm": 0.10269171744585037, "learning_rate": 5.4509490069382274e-05, "loss": 0.0095, "step": 49770 }, { "grad_norm": 0.1994093954563141, "learning_rate": 5.449302251051509e-05, "loss": 0.0097, "step": 49780 }, { "grad_norm": 0.1862802952528, "learning_rate": 5.44765544602977e-05, "loss": 0.0103, "step": 49790 }, { "grad_norm": 0.1616353988647461, "learning_rate": 5.446008592053107e-05, "loss": 0.013, "step": 49800 }, { "grad_norm": 0.1583564728498459, "learning_rate": 5.4443616893016135e-05, "loss": 0.0118, "step": 49810 }, { "grad_norm": 0.14282263815402985, "learning_rate": 5.4427147379553934e-05, "loss": 0.0101, "step": 49820 }, { "grad_norm": 0.16219931840896606, "learning_rate": 5.441067738194556e-05, "loss": 0.0119, "step": 49830 }, { "grad_norm": 0.1340634822845459, "learning_rate": 5.439420690199214e-05, "loss": 0.0151, "step": 49840 }, { "grad_norm": 0.1177092120051384, "learning_rate": 5.437773594149484e-05, "loss": 0.0086, "step": 49850 }, { "grad_norm": 0.12764517962932587, "learning_rate": 5.436126450225495e-05, "loss": 0.0122, "step": 49860 }, { "grad_norm": 0.1449938267469406, "learning_rate": 5.434479258607371e-05, "loss": 0.0117, "step": 49870 }, { "grad_norm": 0.1649148166179657, "learning_rate": 5.43283201947525e-05, "loss": 0.0129, "step": 49880 }, { "grad_norm": 0.12259702384471893, "learning_rate": 5.431184733009268e-05, "loss": 0.0134, "step": 49890 }, { "grad_norm": 0.1716059297323227, "learning_rate": 5.4295373993895736e-05, "loss": 0.0119, "step": 49900 }, { "grad_norm": 0.13240832090377808, "learning_rate": 5.4278900187963157e-05, "loss": 0.0117, "step": 49910 }, { "grad_norm": 0.1513814479112625, "learning_rate": 5.426242591409646e-05, "loss": 0.0111, "step": 49920 }, { "grad_norm": 0.14883476495742798, "learning_rate": 5.42459511740973e-05, "loss": 0.0097, "step": 49930 }, { "grad_norm": 0.16087700426578522, "learning_rate": 5.422947596976729e-05, "loss": 0.0094, "step": 49940 }, { "grad_norm": 0.14907075464725494, "learning_rate": 5.4213000302908134e-05, "loss": 0.0107, "step": 49950 }, { "grad_norm": 0.15749341249465942, "learning_rate": 5.419652417532162e-05, "loss": 0.0163, "step": 49960 }, { "grad_norm": 0.20969997346401215, "learning_rate": 5.4180047588809534e-05, "loss": 0.0122, "step": 49970 }, { "grad_norm": 0.14586305618286133, "learning_rate": 5.4163570545173704e-05, "loss": 0.0116, "step": 49980 }, { "grad_norm": 0.15281137824058533, "learning_rate": 5.414709304621608e-05, "loss": 0.0101, "step": 49990 }, { "grad_norm": 0.13038457930088043, "learning_rate": 5.413061509373858e-05, "loss": 0.0122, "step": 50000 } ], "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": 24, "trial_name": null, "trial_params": null }