diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,6034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.3333333333333333, + "eval_steps": 500, + "global_step": 10000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "grad_norm": 1.3314505815505981, + "learning_rate": 6.000000000000001e-07, + "loss": 1.2179, + "step": 10 + }, + { + "grad_norm": 1.5507653951644897, + "learning_rate": 1.2666666666666667e-06, + "loss": 1.2173, + "step": 20 + }, + { + "grad_norm": 1.178682565689087, + "learning_rate": 1.9333333333333336e-06, + "loss": 1.212, + "step": 30 + }, + { + "grad_norm": 0.68263179063797, + "learning_rate": 2.6e-06, + "loss": 1.1905, + "step": 40 + }, + { + "grad_norm": 0.5855809450149536, + "learning_rate": 3.2666666666666666e-06, + "loss": 1.1888, + "step": 50 + }, + { + "grad_norm": 0.3312073051929474, + "learning_rate": 3.9333333333333335e-06, + "loss": 1.1345, + "step": 60 + }, + { + "grad_norm": 0.3775692582130432, + "learning_rate": 4.6e-06, + "loss": 1.1437, + "step": 70 + }, + { + "grad_norm": 0.35685569047927856, + "learning_rate": 5.266666666666667e-06, + "loss": 1.1419, + "step": 80 + }, + { + "grad_norm": 0.4233785569667816, + "learning_rate": 5.933333333333334e-06, + "loss": 1.1079, + "step": 90 + }, + { + "grad_norm": 0.45181751251220703, + "learning_rate": 6.6e-06, + "loss": 1.1288, + "step": 100 + }, + { + "grad_norm": 0.4978986084461212, + "learning_rate": 7.266666666666668e-06, + "loss": 1.1431, + "step": 110 + }, + { + "grad_norm": 0.4479588270187378, + "learning_rate": 7.933333333333334e-06, + "loss": 1.1471, + "step": 120 + }, + { + "grad_norm": 0.5029391050338745, + "learning_rate": 8.599999999999999e-06, + "loss": 1.117, + "step": 130 + }, + { + "grad_norm": 0.8853879570960999, + "learning_rate": 9.266666666666667e-06, + "loss": 1.1064, + "step": 140 + }, + { + "grad_norm": 0.5811502933502197, + "learning_rate": 9.933333333333334e-06, + "loss": 1.1036, + "step": 150 + }, + { + "grad_norm": 0.7939465045928955, + "learning_rate": 1.06e-05, + "loss": 1.0916, + "step": 160 + }, + { + "grad_norm": 0.6251223087310791, + "learning_rate": 1.1266666666666667e-05, + "loss": 1.0927, + "step": 170 + }, + { + "grad_norm": 0.7339620590209961, + "learning_rate": 1.1933333333333333e-05, + "loss": 1.0813, + "step": 180 + }, + { + "grad_norm": 0.6614446640014648, + "learning_rate": 1.2600000000000001e-05, + "loss": 1.0855, + "step": 190 + }, + { + "grad_norm": 0.501140832901001, + "learning_rate": 1.3266666666666666e-05, + "loss": 1.0701, + "step": 200 + }, + { + "grad_norm": 0.7906708717346191, + "learning_rate": 1.3933333333333334e-05, + "loss": 1.0887, + "step": 210 + }, + { + "grad_norm": 0.6906638145446777, + "learning_rate": 1.4599999999999999e-05, + "loss": 1.0612, + "step": 220 + }, + { + "grad_norm": 0.7390074133872986, + "learning_rate": 1.5266666666666667e-05, + "loss": 1.0774, + "step": 230 + }, + { + "grad_norm": 0.5626471638679504, + "learning_rate": 1.5933333333333332e-05, + "loss": 1.0671, + "step": 240 + }, + { + "grad_norm": 0.7262792587280273, + "learning_rate": 1.66e-05, + "loss": 1.0679, + "step": 250 + }, + { + "grad_norm": 0.5234604477882385, + "learning_rate": 1.726666666666667e-05, + "loss": 1.0684, + "step": 260 + }, + { + "grad_norm": 0.5092823505401611, + "learning_rate": 1.7933333333333337e-05, + "loss": 1.0641, + "step": 270 + }, + { + "grad_norm": 0.6354775428771973, + "learning_rate": 1.86e-05, + "loss": 1.0725, + "step": 280 + }, + { + "grad_norm": 0.5659180283546448, + "learning_rate": 1.926666666666667e-05, + "loss": 1.0633, + "step": 290 + }, + { + "grad_norm": 0.8407434225082397, + "learning_rate": 1.9933333333333334e-05, + "loss": 1.0568, + "step": 300 + }, + { + "grad_norm": 0.7417938113212585, + "learning_rate": 2.06e-05, + "loss": 1.0474, + "step": 310 + }, + { + "grad_norm": 0.7800794243812561, + "learning_rate": 2.1266666666666667e-05, + "loss": 1.041, + "step": 320 + }, + { + "grad_norm": 0.7304247617721558, + "learning_rate": 2.1933333333333332e-05, + "loss": 1.052, + "step": 330 + }, + { + "grad_norm": 0.9249001741409302, + "learning_rate": 2.26e-05, + "loss": 1.027, + "step": 340 + }, + { + "grad_norm": 0.7380666732788086, + "learning_rate": 2.326666666666667e-05, + "loss": 1.0311, + "step": 350 + }, + { + "grad_norm": 0.7331956624984741, + "learning_rate": 2.3933333333333337e-05, + "loss": 1.0275, + "step": 360 + }, + { + "grad_norm": 0.9304872155189514, + "learning_rate": 2.46e-05, + "loss": 1.0111, + "step": 370 + }, + { + "grad_norm": 0.8601149916648865, + "learning_rate": 2.5266666666666666e-05, + "loss": 0.997, + "step": 380 + }, + { + "grad_norm": 0.9442141652107239, + "learning_rate": 2.5933333333333338e-05, + "loss": 0.9944, + "step": 390 + }, + { + "grad_norm": 0.8515417575836182, + "learning_rate": 2.6600000000000003e-05, + "loss": 0.9864, + "step": 400 + }, + { + "grad_norm": 0.9509739875793457, + "learning_rate": 2.7266666666666668e-05, + "loss": 0.955, + "step": 410 + }, + { + "grad_norm": 1.084962248802185, + "learning_rate": 2.7933333333333332e-05, + "loss": 0.9272, + "step": 420 + }, + { + "grad_norm": 0.9800599813461304, + "learning_rate": 2.86e-05, + "loss": 0.9069, + "step": 430 + }, + { + "grad_norm": 1.0645278692245483, + "learning_rate": 2.926666666666667e-05, + "loss": 0.8653, + "step": 440 + }, + { + "grad_norm": 1.049734354019165, + "learning_rate": 2.9933333333333337e-05, + "loss": 0.8439, + "step": 450 + }, + { + "grad_norm": 1.1224632263183594, + "learning_rate": 3.06e-05, + "loss": 0.8167, + "step": 460 + }, + { + "grad_norm": 1.0559662580490112, + "learning_rate": 3.126666666666666e-05, + "loss": 0.7786, + "step": 470 + }, + { + "grad_norm": 1.1540216207504272, + "learning_rate": 3.1933333333333335e-05, + "loss": 0.7453, + "step": 480 + }, + { + "grad_norm": 1.0124099254608154, + "learning_rate": 3.26e-05, + "loss": 0.7176, + "step": 490 + }, + { + "grad_norm": 1.3634672164916992, + "learning_rate": 3.326666666666667e-05, + "loss": 0.6796, + "step": 500 + }, + { + "grad_norm": 1.2447317838668823, + "learning_rate": 3.3933333333333336e-05, + "loss": 0.6478, + "step": 510 + }, + { + "grad_norm": 1.3635326623916626, + "learning_rate": 3.46e-05, + "loss": 0.6181, + "step": 520 + }, + { + "grad_norm": 1.189446210861206, + "learning_rate": 3.526666666666667e-05, + "loss": 0.5902, + "step": 530 + }, + { + "grad_norm": 0.9477949738502502, + "learning_rate": 3.593333333333334e-05, + "loss": 0.5247, + "step": 540 + }, + { + "grad_norm": 1.3037503957748413, + "learning_rate": 3.66e-05, + "loss": 0.5328, + "step": 550 + }, + { + "grad_norm": 1.5641331672668457, + "learning_rate": 3.726666666666667e-05, + "loss": 0.5011, + "step": 560 + }, + { + "grad_norm": 1.127152442932129, + "learning_rate": 3.793333333333334e-05, + "loss": 0.4387, + "step": 570 + }, + { + "grad_norm": 1.2061371803283691, + "learning_rate": 3.86e-05, + "loss": 0.407, + "step": 580 + }, + { + "grad_norm": 1.273613452911377, + "learning_rate": 3.926666666666667e-05, + "loss": 0.3972, + "step": 590 + }, + { + "grad_norm": 1.1315085887908936, + "learning_rate": 3.993333333333333e-05, + "loss": 0.3814, + "step": 600 + }, + { + "grad_norm": 1.5654405355453491, + "learning_rate": 4.0600000000000004e-05, + "loss": 0.3573, + "step": 610 + }, + { + "grad_norm": 1.298802375793457, + "learning_rate": 4.126666666666667e-05, + "loss": 0.335, + "step": 620 + }, + { + "grad_norm": 1.101986050605774, + "learning_rate": 4.1933333333333334e-05, + "loss": 0.324, + "step": 630 + }, + { + "grad_norm": 1.480060338973999, + "learning_rate": 4.26e-05, + "loss": 0.3052, + "step": 640 + }, + { + "grad_norm": 1.656790018081665, + "learning_rate": 4.3266666666666664e-05, + "loss": 0.2794, + "step": 650 + }, + { + "grad_norm": 1.57819664478302, + "learning_rate": 4.3933333333333335e-05, + "loss": 0.2932, + "step": 660 + }, + { + "grad_norm": 1.248097538948059, + "learning_rate": 4.46e-05, + "loss": 0.2431, + "step": 670 + }, + { + "grad_norm": 1.4449474811553955, + "learning_rate": 4.526666666666667e-05, + "loss": 0.2639, + "step": 680 + }, + { + "grad_norm": 1.2288621664047241, + "learning_rate": 4.5933333333333336e-05, + "loss": 0.2493, + "step": 690 + }, + { + "grad_norm": 1.2860236167907715, + "learning_rate": 4.660000000000001e-05, + "loss": 0.2454, + "step": 700 + }, + { + "grad_norm": 1.2656257152557373, + "learning_rate": 4.726666666666667e-05, + "loss": 0.2165, + "step": 710 + }, + { + "grad_norm": 1.2987090349197388, + "learning_rate": 4.793333333333334e-05, + "loss": 0.2253, + "step": 720 + }, + { + "grad_norm": 1.233296513557434, + "learning_rate": 4.86e-05, + "loss": 0.1972, + "step": 730 + }, + { + "grad_norm": 1.1606056690216064, + "learning_rate": 4.926666666666667e-05, + "loss": 0.1532, + "step": 740 + }, + { + "grad_norm": 0.9332922697067261, + "learning_rate": 4.993333333333334e-05, + "loss": 0.1803, + "step": 750 + }, + { + "grad_norm": 1.067075490951538, + "learning_rate": 5.0600000000000003e-05, + "loss": 0.1656, + "step": 760 + }, + { + "grad_norm": 1.1522564888000488, + "learning_rate": 5.1266666666666675e-05, + "loss": 0.1762, + "step": 770 + }, + { + "grad_norm": 0.9858033061027527, + "learning_rate": 5.193333333333333e-05, + "loss": 0.2118, + "step": 780 + }, + { + "grad_norm": 0.9792870879173279, + "learning_rate": 5.2600000000000005e-05, + "loss": 0.1497, + "step": 790 + }, + { + "grad_norm": 1.120789647102356, + "learning_rate": 5.326666666666666e-05, + "loss": 0.1549, + "step": 800 + }, + { + "grad_norm": 0.9506876468658447, + "learning_rate": 5.3933333333333334e-05, + "loss": 0.165, + "step": 810 + }, + { + "grad_norm": 0.9334282279014587, + "learning_rate": 5.4600000000000006e-05, + "loss": 0.1629, + "step": 820 + }, + { + "grad_norm": 1.1798714399337769, + "learning_rate": 5.5266666666666664e-05, + "loss": 0.1924, + "step": 830 + }, + { + "grad_norm": 1.1720198392868042, + "learning_rate": 5.5933333333333335e-05, + "loss": 0.1665, + "step": 840 + }, + { + "grad_norm": 0.849250316619873, + "learning_rate": 5.66e-05, + "loss": 0.1652, + "step": 850 + }, + { + "grad_norm": 0.9155775904655457, + "learning_rate": 5.726666666666667e-05, + "loss": 0.1528, + "step": 860 + }, + { + "grad_norm": 1.171196699142456, + "learning_rate": 5.7933333333333337e-05, + "loss": 0.1722, + "step": 870 + }, + { + "grad_norm": 0.9858519434928894, + "learning_rate": 5.86e-05, + "loss": 0.1412, + "step": 880 + }, + { + "grad_norm": 1.081385850906372, + "learning_rate": 5.926666666666667e-05, + "loss": 0.1561, + "step": 890 + }, + { + "grad_norm": 0.8932467699050903, + "learning_rate": 5.9933333333333345e-05, + "loss": 0.1579, + "step": 900 + }, + { + "grad_norm": 0.8548662662506104, + "learning_rate": 6.06e-05, + "loss": 0.1505, + "step": 910 + }, + { + "grad_norm": 0.7572005391120911, + "learning_rate": 6.126666666666667e-05, + "loss": 0.1391, + "step": 920 + }, + { + "grad_norm": 0.9849821925163269, + "learning_rate": 6.193333333333333e-05, + "loss": 0.1444, + "step": 930 + }, + { + "grad_norm": 0.9562931656837463, + "learning_rate": 6.26e-05, + "loss": 0.1566, + "step": 940 + }, + { + "grad_norm": 1.2492797374725342, + "learning_rate": 6.326666666666667e-05, + "loss": 0.1419, + "step": 950 + }, + { + "grad_norm": 0.9293448328971863, + "learning_rate": 6.393333333333333e-05, + "loss": 0.144, + "step": 960 + }, + { + "grad_norm": 0.858593761920929, + "learning_rate": 6.460000000000001e-05, + "loss": 0.1649, + "step": 970 + }, + { + "grad_norm": 0.8152620196342468, + "learning_rate": 6.526666666666666e-05, + "loss": 0.1397, + "step": 980 + }, + { + "grad_norm": 0.824725866317749, + "learning_rate": 6.593333333333334e-05, + "loss": 0.1409, + "step": 990 + }, + { + "grad_norm": 0.834763765335083, + "learning_rate": 6.66e-05, + "loss": 0.1463, + "step": 1000 + }, + { + "grad_norm": 0.7459161877632141, + "learning_rate": 6.726666666666667e-05, + "loss": 0.1465, + "step": 1010 + }, + { + "grad_norm": 0.9462399482727051, + "learning_rate": 6.793333333333334e-05, + "loss": 0.1646, + "step": 1020 + }, + { + "grad_norm": 0.9641590714454651, + "learning_rate": 6.860000000000001e-05, + "loss": 0.138, + "step": 1030 + }, + { + "grad_norm": 0.8934517502784729, + "learning_rate": 6.926666666666667e-05, + "loss": 0.144, + "step": 1040 + }, + { + "grad_norm": 0.8419318199157715, + "learning_rate": 6.993333333333334e-05, + "loss": 0.1457, + "step": 1050 + }, + { + "grad_norm": 0.7819538116455078, + "learning_rate": 7.06e-05, + "loss": 0.1535, + "step": 1060 + }, + { + "grad_norm": 0.8466137051582336, + "learning_rate": 7.126666666666667e-05, + "loss": 0.151, + "step": 1070 + }, + { + "grad_norm": 1.0024020671844482, + "learning_rate": 7.193333333333334e-05, + "loss": 0.1749, + "step": 1080 + }, + { + "grad_norm": 0.8885055780410767, + "learning_rate": 7.26e-05, + "loss": 0.1457, + "step": 1090 + }, + { + "grad_norm": 0.8589239716529846, + "learning_rate": 7.326666666666667e-05, + "loss": 0.1313, + "step": 1100 + }, + { + "grad_norm": 0.8235303163528442, + "learning_rate": 7.393333333333333e-05, + "loss": 0.1657, + "step": 1110 + }, + { + "grad_norm": 0.8818825483322144, + "learning_rate": 7.46e-05, + "loss": 0.1293, + "step": 1120 + }, + { + "grad_norm": 0.7876054644584656, + "learning_rate": 7.526666666666668e-05, + "loss": 0.1478, + "step": 1130 + }, + { + "grad_norm": 0.7659526467323303, + "learning_rate": 7.593333333333334e-05, + "loss": 0.1765, + "step": 1140 + }, + { + "grad_norm": 0.7604926228523254, + "learning_rate": 7.66e-05, + "loss": 0.1656, + "step": 1150 + }, + { + "grad_norm": 0.8470979332923889, + "learning_rate": 7.726666666666667e-05, + "loss": 0.1483, + "step": 1160 + }, + { + "grad_norm": 0.6591774225234985, + "learning_rate": 7.793333333333333e-05, + "loss": 0.1422, + "step": 1170 + }, + { + "grad_norm": 0.7418103218078613, + "learning_rate": 7.860000000000001e-05, + "loss": 0.1253, + "step": 1180 + }, + { + "grad_norm": 0.8148030042648315, + "learning_rate": 7.926666666666666e-05, + "loss": 0.1569, + "step": 1190 + }, + { + "grad_norm": 1.147688865661621, + "learning_rate": 7.993333333333334e-05, + "loss": 0.1548, + "step": 1200 + }, + { + "grad_norm": 1.0288184881210327, + "learning_rate": 8.060000000000001e-05, + "loss": 0.1759, + "step": 1210 + }, + { + "grad_norm": 0.7353565096855164, + "learning_rate": 8.126666666666667e-05, + "loss": 0.1465, + "step": 1220 + }, + { + "grad_norm": 0.7904092669487, + "learning_rate": 8.193333333333334e-05, + "loss": 0.1369, + "step": 1230 + }, + { + "grad_norm": 0.6588637232780457, + "learning_rate": 8.26e-05, + "loss": 0.1438, + "step": 1240 + }, + { + "grad_norm": 0.9510173201560974, + "learning_rate": 8.326666666666667e-05, + "loss": 0.1328, + "step": 1250 + }, + { + "grad_norm": 0.7137603759765625, + "learning_rate": 8.393333333333335e-05, + "loss": 0.1355, + "step": 1260 + }, + { + "grad_norm": 0.5818139314651489, + "learning_rate": 8.46e-05, + "loss": 0.1161, + "step": 1270 + }, + { + "grad_norm": 0.9060401320457458, + "learning_rate": 8.526666666666667e-05, + "loss": 0.1396, + "step": 1280 + }, + { + "grad_norm": 0.7194551825523376, + "learning_rate": 8.593333333333333e-05, + "loss": 0.1404, + "step": 1290 + }, + { + "grad_norm": 0.5931652784347534, + "learning_rate": 8.66e-05, + "loss": 0.1346, + "step": 1300 + }, + { + "grad_norm": 0.7356128692626953, + "learning_rate": 8.726666666666667e-05, + "loss": 0.1328, + "step": 1310 + }, + { + "grad_norm": 0.6496874094009399, + "learning_rate": 8.793333333333333e-05, + "loss": 0.1323, + "step": 1320 + }, + { + "grad_norm": 0.7630678415298462, + "learning_rate": 8.86e-05, + "loss": 0.121, + "step": 1330 + }, + { + "grad_norm": 0.6775751113891602, + "learning_rate": 8.926666666666668e-05, + "loss": 0.1141, + "step": 1340 + }, + { + "grad_norm": 0.6647273302078247, + "learning_rate": 8.993333333333334e-05, + "loss": 0.1472, + "step": 1350 + }, + { + "grad_norm": 0.7508830428123474, + "learning_rate": 9.06e-05, + "loss": 0.1329, + "step": 1360 + }, + { + "grad_norm": 0.6608383655548096, + "learning_rate": 9.126666666666667e-05, + "loss": 0.122, + "step": 1370 + }, + { + "grad_norm": 0.6588468551635742, + "learning_rate": 9.193333333333334e-05, + "loss": 0.1148, + "step": 1380 + }, + { + "grad_norm": 0.5878859162330627, + "learning_rate": 9.260000000000001e-05, + "loss": 0.1255, + "step": 1390 + }, + { + "grad_norm": 0.5673328042030334, + "learning_rate": 9.326666666666667e-05, + "loss": 0.1165, + "step": 1400 + }, + { + "grad_norm": 0.7046443223953247, + "learning_rate": 9.393333333333334e-05, + "loss": 0.1376, + "step": 1410 + }, + { + "grad_norm": 0.6791790127754211, + "learning_rate": 9.46e-05, + "loss": 0.1376, + "step": 1420 + }, + { + "grad_norm": 0.5826202630996704, + "learning_rate": 9.526666666666667e-05, + "loss": 0.1105, + "step": 1430 + }, + { + "grad_norm": 0.6816211938858032, + "learning_rate": 9.593333333333334e-05, + "loss": 0.12, + "step": 1440 + }, + { + "grad_norm": 0.7346307635307312, + "learning_rate": 9.66e-05, + "loss": 0.1457, + "step": 1450 + }, + { + "grad_norm": 0.7110230922698975, + "learning_rate": 9.726666666666667e-05, + "loss": 0.1356, + "step": 1460 + }, + { + "grad_norm": 0.6250454187393188, + "learning_rate": 9.793333333333333e-05, + "loss": 0.1378, + "step": 1470 + }, + { + "grad_norm": 0.7070907354354858, + "learning_rate": 9.86e-05, + "loss": 0.1106, + "step": 1480 + }, + { + "grad_norm": 0.541174590587616, + "learning_rate": 9.926666666666668e-05, + "loss": 0.1152, + "step": 1490 + }, + { + "grad_norm": 0.46901485323905945, + "learning_rate": 9.993333333333334e-05, + "loss": 0.125, + "step": 1500 + }, + { + "grad_norm": 0.43468305468559265, + "learning_rate": 9.999997539434007e-05, + "loss": 0.1087, + "step": 1510 + }, + { + "grad_norm": 0.7124031782150269, + "learning_rate": 9.999989033776898e-05, + "loss": 0.1236, + "step": 1520 + }, + { + "grad_norm": 0.5573038458824158, + "learning_rate": 9.999974452661641e-05, + "loss": 0.132, + "step": 1530 + }, + { + "grad_norm": 0.5823051929473877, + "learning_rate": 9.999953796105959e-05, + "loss": 0.1186, + "step": 1540 + }, + { + "grad_norm": 0.6794359683990479, + "learning_rate": 9.999927064134949e-05, + "loss": 0.1214, + "step": 1550 + }, + { + "grad_norm": 0.725846529006958, + "learning_rate": 9.999894256781095e-05, + "loss": 0.1408, + "step": 1560 + }, + { + "grad_norm": 0.598950982093811, + "learning_rate": 9.99985537408426e-05, + "loss": 0.1115, + "step": 1570 + }, + { + "grad_norm": 0.7087956666946411, + "learning_rate": 9.999810416091688e-05, + "loss": 0.1331, + "step": 1580 + }, + { + "grad_norm": 0.6899130344390869, + "learning_rate": 9.99975938285801e-05, + "loss": 0.1427, + "step": 1590 + }, + { + "grad_norm": 0.599158525466919, + "learning_rate": 9.999702274445236e-05, + "loss": 0.1306, + "step": 1600 + }, + { + "grad_norm": 0.7044597268104553, + "learning_rate": 9.999639090922756e-05, + "loss": 0.1057, + "step": 1610 + }, + { + "grad_norm": 0.6268491744995117, + "learning_rate": 9.999569832367346e-05, + "loss": 0.1255, + "step": 1620 + }, + { + "grad_norm": 0.5614187717437744, + "learning_rate": 9.999494498863162e-05, + "loss": 0.1545, + "step": 1630 + }, + { + "grad_norm": 0.5501335263252258, + "learning_rate": 9.99941309050174e-05, + "loss": 0.1217, + "step": 1640 + }, + { + "grad_norm": 0.5499849319458008, + "learning_rate": 9.999325607381999e-05, + "loss": 0.1267, + "step": 1650 + }, + { + "grad_norm": 0.44044461846351624, + "learning_rate": 9.999232049610238e-05, + "loss": 0.126, + "step": 1660 + }, + { + "grad_norm": 0.49347344040870667, + "learning_rate": 9.999132417300142e-05, + "loss": 0.1058, + "step": 1670 + }, + { + "grad_norm": 0.6333240866661072, + "learning_rate": 9.99902671057277e-05, + "loss": 0.1163, + "step": 1680 + }, + { + "grad_norm": 0.5511918663978577, + "learning_rate": 9.998914929556569e-05, + "loss": 0.1168, + "step": 1690 + }, + { + "grad_norm": 0.574537992477417, + "learning_rate": 9.998797074387361e-05, + "loss": 0.1187, + "step": 1700 + }, + { + "grad_norm": 0.6028810143470764, + "learning_rate": 9.99867314520835e-05, + "loss": 0.1316, + "step": 1710 + }, + { + "grad_norm": 0.4513114392757416, + "learning_rate": 9.998543142170126e-05, + "loss": 0.1211, + "step": 1720 + }, + { + "grad_norm": 0.4847932457923889, + "learning_rate": 9.99840706543065e-05, + "loss": 0.1147, + "step": 1730 + }, + { + "grad_norm": 0.4815286695957184, + "learning_rate": 9.998264915155274e-05, + "loss": 0.11, + "step": 1740 + }, + { + "grad_norm": 0.6546589136123657, + "learning_rate": 9.998116691516718e-05, + "loss": 0.1305, + "step": 1750 + }, + { + "grad_norm": 0.6325458884239197, + "learning_rate": 9.997962394695091e-05, + "loss": 0.1249, + "step": 1760 + }, + { + "grad_norm": 0.5429052114486694, + "learning_rate": 9.997802024877875e-05, + "loss": 0.1129, + "step": 1770 + }, + { + "grad_norm": 0.5844834446907043, + "learning_rate": 9.99763558225994e-05, + "loss": 0.1456, + "step": 1780 + }, + { + "grad_norm": 0.650963306427002, + "learning_rate": 9.997463067043526e-05, + "loss": 0.1163, + "step": 1790 + }, + { + "grad_norm": 0.5388178825378418, + "learning_rate": 9.997284479438253e-05, + "loss": 0.1189, + "step": 1800 + }, + { + "grad_norm": 0.43301138281822205, + "learning_rate": 9.997099819661127e-05, + "loss": 0.1067, + "step": 1810 + }, + { + "grad_norm": 0.5625163912773132, + "learning_rate": 9.996909087936524e-05, + "loss": 0.1218, + "step": 1820 + }, + { + "grad_norm": 0.6838710904121399, + "learning_rate": 9.996712284496201e-05, + "loss": 0.1181, + "step": 1830 + }, + { + "grad_norm": 0.7385373711585999, + "learning_rate": 9.996509409579293e-05, + "loss": 0.1291, + "step": 1840 + }, + { + "grad_norm": 0.6341670155525208, + "learning_rate": 9.996300463432312e-05, + "loss": 0.1467, + "step": 1850 + }, + { + "grad_norm": 0.6742323637008667, + "learning_rate": 9.996085446309148e-05, + "loss": 0.1402, + "step": 1860 + }, + { + "grad_norm": 0.3937931954860687, + "learning_rate": 9.995864358471066e-05, + "loss": 0.1098, + "step": 1870 + }, + { + "grad_norm": 0.5168664455413818, + "learning_rate": 9.99563720018671e-05, + "loss": 0.0913, + "step": 1880 + }, + { + "grad_norm": 0.6387127041816711, + "learning_rate": 9.995403971732098e-05, + "loss": 0.1066, + "step": 1890 + }, + { + "grad_norm": 0.5926902890205383, + "learning_rate": 9.995164673390625e-05, + "loss": 0.1004, + "step": 1900 + }, + { + "grad_norm": 0.568058431148529, + "learning_rate": 9.994919305453059e-05, + "loss": 0.1259, + "step": 1910 + }, + { + "grad_norm": 0.6348229050636292, + "learning_rate": 9.994667868217548e-05, + "loss": 0.1107, + "step": 1920 + }, + { + "grad_norm": 0.4053514301776886, + "learning_rate": 9.99441036198961e-05, + "loss": 0.1035, + "step": 1930 + }, + { + "grad_norm": 0.5806289911270142, + "learning_rate": 9.99414678708214e-05, + "loss": 0.111, + "step": 1940 + }, + { + "grad_norm": 0.521151065826416, + "learning_rate": 9.993877143815407e-05, + "loss": 0.0986, + "step": 1950 + }, + { + "grad_norm": 0.6184947490692139, + "learning_rate": 9.993601432517053e-05, + "loss": 0.1161, + "step": 1960 + }, + { + "grad_norm": 0.5990853905677795, + "learning_rate": 9.993319653522091e-05, + "loss": 0.0941, + "step": 1970 + }, + { + "grad_norm": 0.5944622159004211, + "learning_rate": 9.993031807172911e-05, + "loss": 0.1273, + "step": 1980 + }, + { + "grad_norm": 0.5887544751167297, + "learning_rate": 9.992737893819273e-05, + "loss": 0.1019, + "step": 1990 + }, + { + "grad_norm": 0.49060699343681335, + "learning_rate": 9.992437913818312e-05, + "loss": 0.126, + "step": 2000 + }, + { + "grad_norm": 0.38806021213531494, + "learning_rate": 9.992131867534526e-05, + "loss": 0.1398, + "step": 2010 + }, + { + "grad_norm": 0.6121862530708313, + "learning_rate": 9.991819755339796e-05, + "loss": 0.1115, + "step": 2020 + }, + { + "grad_norm": 0.45789533853530884, + "learning_rate": 9.991501577613365e-05, + "loss": 0.0937, + "step": 2030 + }, + { + "grad_norm": 0.4618423283100128, + "learning_rate": 9.99117733474185e-05, + "loss": 0.1158, + "step": 2040 + }, + { + "grad_norm": 0.5058461427688599, + "learning_rate": 9.990847027119234e-05, + "loss": 0.1087, + "step": 2050 + }, + { + "grad_norm": 0.5403944253921509, + "learning_rate": 9.990510655146877e-05, + "loss": 0.1089, + "step": 2060 + }, + { + "grad_norm": 0.6129620671272278, + "learning_rate": 9.990168219233496e-05, + "loss": 0.101, + "step": 2070 + }, + { + "grad_norm": 0.49735042452812195, + "learning_rate": 9.989819719795188e-05, + "loss": 0.0979, + "step": 2080 + }, + { + "grad_norm": 0.5445982217788696, + "learning_rate": 9.989465157255412e-05, + "loss": 0.1028, + "step": 2090 + }, + { + "grad_norm": 0.44239258766174316, + "learning_rate": 9.989104532044994e-05, + "loss": 0.1015, + "step": 2100 + }, + { + "grad_norm": 0.403062641620636, + "learning_rate": 9.988737844602128e-05, + "loss": 0.1296, + "step": 2110 + }, + { + "grad_norm": 0.4222145080566406, + "learning_rate": 9.988365095372372e-05, + "loss": 0.0996, + "step": 2120 + }, + { + "grad_norm": 0.4066031575202942, + "learning_rate": 9.987986284808654e-05, + "loss": 0.1031, + "step": 2130 + }, + { + "grad_norm": 0.5209201574325562, + "learning_rate": 9.987601413371264e-05, + "loss": 0.0835, + "step": 2140 + }, + { + "grad_norm": 0.37327784299850464, + "learning_rate": 9.987210481527855e-05, + "loss": 0.1194, + "step": 2150 + }, + { + "grad_norm": 0.36498987674713135, + "learning_rate": 9.98681348975345e-05, + "loss": 0.0953, + "step": 2160 + }, + { + "grad_norm": 0.4885857105255127, + "learning_rate": 9.986410438530427e-05, + "loss": 0.0983, + "step": 2170 + }, + { + "grad_norm": 0.3909528851509094, + "learning_rate": 9.986001328348534e-05, + "loss": 0.1147, + "step": 2180 + }, + { + "grad_norm": 0.4662936329841614, + "learning_rate": 9.985586159704878e-05, + "loss": 0.0994, + "step": 2190 + }, + { + "grad_norm": 0.5220146775245667, + "learning_rate": 9.985164933103929e-05, + "loss": 0.0929, + "step": 2200 + }, + { + "grad_norm": 0.43521198630332947, + "learning_rate": 9.984737649057513e-05, + "loss": 0.1034, + "step": 2210 + }, + { + "grad_norm": 0.49455365538597107, + "learning_rate": 9.984304308084827e-05, + "loss": 0.0955, + "step": 2220 + }, + { + "grad_norm": 0.3833504915237427, + "learning_rate": 9.983864910712416e-05, + "loss": 0.0991, + "step": 2230 + }, + { + "grad_norm": 0.5373802781105042, + "learning_rate": 9.98341945747419e-05, + "loss": 0.1002, + "step": 2240 + }, + { + "grad_norm": 0.5095442533493042, + "learning_rate": 9.98296794891142e-05, + "loss": 0.1142, + "step": 2250 + }, + { + "grad_norm": 0.3834283947944641, + "learning_rate": 9.982510385572725e-05, + "loss": 0.0883, + "step": 2260 + }, + { + "grad_norm": 0.44791552424430847, + "learning_rate": 9.982046768014094e-05, + "loss": 0.0911, + "step": 2270 + }, + { + "grad_norm": 0.47189849615097046, + "learning_rate": 9.981577096798863e-05, + "loss": 0.0856, + "step": 2280 + }, + { + "grad_norm": 0.5494545102119446, + "learning_rate": 9.981101372497727e-05, + "loss": 0.1238, + "step": 2290 + }, + { + "grad_norm": 0.4184468686580658, + "learning_rate": 9.980619595688737e-05, + "loss": 0.0866, + "step": 2300 + }, + { + "grad_norm": 0.5321471691131592, + "learning_rate": 9.980131766957295e-05, + "loss": 0.1032, + "step": 2310 + }, + { + "grad_norm": 0.358763724565506, + "learning_rate": 9.979637886896163e-05, + "loss": 0.0858, + "step": 2320 + }, + { + "grad_norm": 0.4319607615470886, + "learning_rate": 9.979137956105447e-05, + "loss": 0.1104, + "step": 2330 + }, + { + "grad_norm": 0.41809988021850586, + "learning_rate": 9.978631975192613e-05, + "loss": 0.0977, + "step": 2340 + }, + { + "grad_norm": 0.4701225459575653, + "learning_rate": 9.978119944772475e-05, + "loss": 0.0986, + "step": 2350 + }, + { + "grad_norm": 0.3694179356098175, + "learning_rate": 9.977601865467197e-05, + "loss": 0.1003, + "step": 2360 + }, + { + "grad_norm": 0.4124026298522949, + "learning_rate": 9.977077737906297e-05, + "loss": 0.1002, + "step": 2370 + }, + { + "grad_norm": 0.4824465215206146, + "learning_rate": 9.976547562726636e-05, + "loss": 0.1003, + "step": 2380 + }, + { + "grad_norm": 0.4390423595905304, + "learning_rate": 9.976011340572429e-05, + "loss": 0.117, + "step": 2390 + }, + { + "grad_norm": 0.42644989490509033, + "learning_rate": 9.975469072095237e-05, + "loss": 0.0992, + "step": 2400 + }, + { + "grad_norm": 0.4766586720943451, + "learning_rate": 9.974920757953965e-05, + "loss": 0.0902, + "step": 2410 + }, + { + "grad_norm": 0.49637940526008606, + "learning_rate": 9.97436639881487e-05, + "loss": 0.1299, + "step": 2420 + }, + { + "grad_norm": 0.39479148387908936, + "learning_rate": 9.973805995351545e-05, + "loss": 0.1039, + "step": 2430 + }, + { + "grad_norm": 0.41260361671447754, + "learning_rate": 9.973239548244939e-05, + "loss": 0.1082, + "step": 2440 + }, + { + "grad_norm": 0.50091952085495, + "learning_rate": 9.972667058183333e-05, + "loss": 0.0938, + "step": 2450 + }, + { + "grad_norm": 0.3772203028202057, + "learning_rate": 9.972088525862362e-05, + "loss": 0.0894, + "step": 2460 + }, + { + "grad_norm": 0.41473454236984253, + "learning_rate": 9.971503951984995e-05, + "loss": 0.0972, + "step": 2470 + }, + { + "grad_norm": 0.48513537645339966, + "learning_rate": 9.970913337261543e-05, + "loss": 0.1148, + "step": 2480 + }, + { + "grad_norm": 0.42165645956993103, + "learning_rate": 9.97031668240966e-05, + "loss": 0.0958, + "step": 2490 + }, + { + "grad_norm": 0.49460962414741516, + "learning_rate": 9.969713988154339e-05, + "loss": 0.0924, + "step": 2500 + }, + { + "grad_norm": 0.41141465306282043, + "learning_rate": 9.969105255227906e-05, + "loss": 0.0846, + "step": 2510 + }, + { + "grad_norm": 0.3961196541786194, + "learning_rate": 9.968490484370035e-05, + "loss": 0.0951, + "step": 2520 + }, + { + "grad_norm": 0.4131496250629425, + "learning_rate": 9.967869676327726e-05, + "loss": 0.078, + "step": 2530 + }, + { + "grad_norm": 0.5037717223167419, + "learning_rate": 9.96724283185532e-05, + "loss": 0.0893, + "step": 2540 + }, + { + "grad_norm": 0.3475835621356964, + "learning_rate": 9.966609951714494e-05, + "loss": 0.1065, + "step": 2550 + }, + { + "grad_norm": 0.37139299511909485, + "learning_rate": 9.965971036674255e-05, + "loss": 0.1043, + "step": 2560 + }, + { + "grad_norm": 0.4540638029575348, + "learning_rate": 9.965326087510947e-05, + "loss": 0.0989, + "step": 2570 + }, + { + "grad_norm": 0.5618516802787781, + "learning_rate": 9.964675105008244e-05, + "loss": 0.1035, + "step": 2580 + }, + { + "grad_norm": 0.49820294976234436, + "learning_rate": 9.964018089957147e-05, + "loss": 0.0875, + "step": 2590 + }, + { + "grad_norm": 0.5172275900840759, + "learning_rate": 9.963355043155996e-05, + "loss": 0.107, + "step": 2600 + }, + { + "grad_norm": 0.5463851094245911, + "learning_rate": 9.962685965410455e-05, + "loss": 0.0989, + "step": 2610 + }, + { + "grad_norm": 0.44594430923461914, + "learning_rate": 9.962010857533514e-05, + "loss": 0.0856, + "step": 2620 + }, + { + "grad_norm": 0.5318399667739868, + "learning_rate": 9.961329720345493e-05, + "loss": 0.0916, + "step": 2630 + }, + { + "grad_norm": 0.40706491470336914, + "learning_rate": 9.96064255467404e-05, + "loss": 0.1134, + "step": 2640 + }, + { + "grad_norm": 0.4159218668937683, + "learning_rate": 9.959949361354126e-05, + "loss": 0.1084, + "step": 2650 + }, + { + "grad_norm": 0.5552453398704529, + "learning_rate": 9.959250141228045e-05, + "loss": 0.1167, + "step": 2660 + }, + { + "grad_norm": 0.5203363299369812, + "learning_rate": 9.958544895145414e-05, + "loss": 0.1168, + "step": 2670 + }, + { + "grad_norm": 0.47239911556243896, + "learning_rate": 9.957833623963177e-05, + "loss": 0.1268, + "step": 2680 + }, + { + "grad_norm": 0.33993610739707947, + "learning_rate": 9.957116328545593e-05, + "loss": 0.0983, + "step": 2690 + }, + { + "grad_norm": 0.4728415608406067, + "learning_rate": 9.956393009764244e-05, + "loss": 0.0821, + "step": 2700 + }, + { + "grad_norm": 0.37891140580177307, + "learning_rate": 9.955663668498032e-05, + "loss": 0.098, + "step": 2710 + }, + { + "grad_norm": 0.4417562484741211, + "learning_rate": 9.954928305633173e-05, + "loss": 0.1134, + "step": 2720 + }, + { + "grad_norm": 0.3930230736732483, + "learning_rate": 9.954186922063204e-05, + "loss": 0.0938, + "step": 2730 + }, + { + "grad_norm": 0.36004993319511414, + "learning_rate": 9.953439518688974e-05, + "loss": 0.0893, + "step": 2740 + }, + { + "grad_norm": 0.3530312478542328, + "learning_rate": 9.952686096418652e-05, + "loss": 0.0972, + "step": 2750 + }, + { + "grad_norm": 0.5520556569099426, + "learning_rate": 9.951926656167714e-05, + "loss": 0.0903, + "step": 2760 + }, + { + "grad_norm": 0.4239886403083801, + "learning_rate": 9.951161198858953e-05, + "loss": 0.0956, + "step": 2770 + }, + { + "grad_norm": 0.35769134759902954, + "learning_rate": 9.950389725422471e-05, + "loss": 0.0962, + "step": 2780 + }, + { + "grad_norm": 0.45160043239593506, + "learning_rate": 9.949612236795682e-05, + "loss": 0.0994, + "step": 2790 + }, + { + "grad_norm": 0.4155345559120178, + "learning_rate": 9.948828733923305e-05, + "loss": 0.0932, + "step": 2800 + }, + { + "grad_norm": 0.46791842579841614, + "learning_rate": 9.948039217757374e-05, + "loss": 0.1243, + "step": 2810 + }, + { + "grad_norm": 0.4612118899822235, + "learning_rate": 9.947243689257225e-05, + "loss": 0.0978, + "step": 2820 + }, + { + "grad_norm": 0.398109495639801, + "learning_rate": 9.946442149389497e-05, + "loss": 0.0863, + "step": 2830 + }, + { + "grad_norm": 0.5814754962921143, + "learning_rate": 9.945634599128139e-05, + "loss": 0.1136, + "step": 2840 + }, + { + "grad_norm": 0.46510931849479675, + "learning_rate": 9.944821039454402e-05, + "loss": 0.1119, + "step": 2850 + }, + { + "grad_norm": 0.4345492124557495, + "learning_rate": 9.944001471356835e-05, + "loss": 0.0982, + "step": 2860 + }, + { + "grad_norm": 0.391938716173172, + "learning_rate": 9.94317589583129e-05, + "loss": 0.0901, + "step": 2870 + }, + { + "grad_norm": 0.43730536103248596, + "learning_rate": 9.942344313880922e-05, + "loss": 0.089, + "step": 2880 + }, + { + "grad_norm": 0.4900282919406891, + "learning_rate": 9.941506726516179e-05, + "loss": 0.0984, + "step": 2890 + }, + { + "grad_norm": 0.45530518889427185, + "learning_rate": 9.94066313475481e-05, + "loss": 0.1065, + "step": 2900 + }, + { + "grad_norm": 0.4501086473464966, + "learning_rate": 9.939813539621857e-05, + "loss": 0.0916, + "step": 2910 + }, + { + "grad_norm": 0.3313567340373993, + "learning_rate": 9.93895794214966e-05, + "loss": 0.0912, + "step": 2920 + }, + { + "grad_norm": 0.32529062032699585, + "learning_rate": 9.938096343377852e-05, + "loss": 0.0761, + "step": 2930 + }, + { + "grad_norm": 0.3651159703731537, + "learning_rate": 9.937228744353353e-05, + "loss": 0.0731, + "step": 2940 + }, + { + "grad_norm": 0.40185627341270447, + "learning_rate": 9.936355146130379e-05, + "loss": 0.0799, + "step": 2950 + }, + { + "grad_norm": 0.38098153471946716, + "learning_rate": 9.935475549770435e-05, + "loss": 0.0846, + "step": 2960 + }, + { + "grad_norm": 0.4270680248737335, + "learning_rate": 9.934589956342315e-05, + "loss": 0.0842, + "step": 2970 + }, + { + "grad_norm": 0.5225918889045715, + "learning_rate": 9.933698366922093e-05, + "loss": 0.0891, + "step": 2980 + }, + { + "grad_norm": 0.3626577854156494, + "learning_rate": 9.93280078259314e-05, + "loss": 0.0968, + "step": 2990 + }, + { + "grad_norm": 0.4718676507472992, + "learning_rate": 9.931897204446104e-05, + "loss": 0.0976, + "step": 3000 + }, + { + "grad_norm": 0.34026843309402466, + "learning_rate": 9.930987633578915e-05, + "loss": 0.0789, + "step": 3010 + }, + { + "grad_norm": 0.39102548360824585, + "learning_rate": 9.93007207109679e-05, + "loss": 0.0832, + "step": 3020 + }, + { + "grad_norm": 0.3950934410095215, + "learning_rate": 9.929150518112224e-05, + "loss": 0.0884, + "step": 3030 + }, + { + "grad_norm": 0.40357714891433716, + "learning_rate": 9.928222975744991e-05, + "loss": 0.0872, + "step": 3040 + }, + { + "grad_norm": 0.4253603518009186, + "learning_rate": 9.92728944512214e-05, + "loss": 0.0904, + "step": 3050 + }, + { + "grad_norm": 0.37781596183776855, + "learning_rate": 9.926349927378001e-05, + "loss": 0.0906, + "step": 3060 + }, + { + "grad_norm": 0.42643070220947266, + "learning_rate": 9.925404423654174e-05, + "loss": 0.0718, + "step": 3070 + }, + { + "grad_norm": 0.47125905752182007, + "learning_rate": 9.924452935099537e-05, + "loss": 0.0918, + "step": 3080 + }, + { + "grad_norm": 0.5146117210388184, + "learning_rate": 9.92349546287024e-05, + "loss": 0.1131, + "step": 3090 + }, + { + "grad_norm": 0.41173461079597473, + "learning_rate": 9.9225320081297e-05, + "loss": 0.0903, + "step": 3100 + }, + { + "grad_norm": 0.36617806553840637, + "learning_rate": 9.921562572048606e-05, + "loss": 0.0822, + "step": 3110 + }, + { + "grad_norm": 0.4469146430492401, + "learning_rate": 9.920587155804913e-05, + "loss": 0.0779, + "step": 3120 + }, + { + "grad_norm": 0.4347069263458252, + "learning_rate": 9.919605760583845e-05, + "loss": 0.0837, + "step": 3130 + }, + { + "grad_norm": 0.5212476849555969, + "learning_rate": 9.91861838757789e-05, + "loss": 0.086, + "step": 3140 + }, + { + "grad_norm": 0.37679511308670044, + "learning_rate": 9.917625037986798e-05, + "loss": 0.0768, + "step": 3150 + }, + { + "grad_norm": 0.5342316627502441, + "learning_rate": 9.916625713017583e-05, + "loss": 0.0892, + "step": 3160 + }, + { + "grad_norm": 0.4032861590385437, + "learning_rate": 9.915620413884519e-05, + "loss": 0.0767, + "step": 3170 + }, + { + "grad_norm": 0.4224550127983093, + "learning_rate": 9.914609141809139e-05, + "loss": 0.0998, + "step": 3180 + }, + { + "grad_norm": 0.39286476373672485, + "learning_rate": 9.913591898020235e-05, + "loss": 0.0966, + "step": 3190 + }, + { + "grad_norm": 0.40394288301467896, + "learning_rate": 9.912568683753853e-05, + "loss": 0.0881, + "step": 3200 + }, + { + "grad_norm": 0.3444503843784332, + "learning_rate": 9.911539500253295e-05, + "loss": 0.0775, + "step": 3210 + }, + { + "grad_norm": 0.41828569769859314, + "learning_rate": 9.910504348769118e-05, + "loss": 0.1134, + "step": 3220 + }, + { + "grad_norm": 0.405741810798645, + "learning_rate": 9.909463230559127e-05, + "loss": 0.0991, + "step": 3230 + }, + { + "grad_norm": 0.37217283248901367, + "learning_rate": 9.908416146888376e-05, + "loss": 0.0795, + "step": 3240 + }, + { + "grad_norm": 0.5055076479911804, + "learning_rate": 9.907363099029175e-05, + "loss": 0.0948, + "step": 3250 + }, + { + "grad_norm": 0.4393453896045685, + "learning_rate": 9.906304088261073e-05, + "loss": 0.0898, + "step": 3260 + }, + { + "grad_norm": 0.3707945942878723, + "learning_rate": 9.905239115870872e-05, + "loss": 0.091, + "step": 3270 + }, + { + "grad_norm": 0.4471019208431244, + "learning_rate": 9.90416818315261e-05, + "loss": 0.0814, + "step": 3280 + }, + { + "grad_norm": 0.4025392234325409, + "learning_rate": 9.903091291407573e-05, + "loss": 0.079, + "step": 3290 + }, + { + "grad_norm": 0.2657795548439026, + "learning_rate": 9.902008441944286e-05, + "loss": 0.0782, + "step": 3300 + }, + { + "grad_norm": 0.40615376830101013, + "learning_rate": 9.900919636078512e-05, + "loss": 0.0953, + "step": 3310 + }, + { + "grad_norm": 0.35301724076271057, + "learning_rate": 9.899824875133255e-05, + "loss": 0.1031, + "step": 3320 + }, + { + "grad_norm": 0.42564162611961365, + "learning_rate": 9.898724160438749e-05, + "loss": 0.0911, + "step": 3330 + }, + { + "grad_norm": 0.46213772892951965, + "learning_rate": 9.89761749333247e-05, + "loss": 0.0886, + "step": 3340 + }, + { + "grad_norm": 0.39841514825820923, + "learning_rate": 9.896504875159122e-05, + "loss": 0.0885, + "step": 3350 + }, + { + "grad_norm": 0.4602282643318176, + "learning_rate": 9.89538630727064e-05, + "loss": 0.0878, + "step": 3360 + }, + { + "grad_norm": 0.4212428629398346, + "learning_rate": 9.894261791026189e-05, + "loss": 0.0887, + "step": 3370 + }, + { + "grad_norm": 0.3691689670085907, + "learning_rate": 9.893131327792165e-05, + "loss": 0.1267, + "step": 3380 + }, + { + "grad_norm": 0.3867242932319641, + "learning_rate": 9.891994918942182e-05, + "loss": 0.0811, + "step": 3390 + }, + { + "grad_norm": 0.3816985487937927, + "learning_rate": 9.890852565857092e-05, + "loss": 0.0897, + "step": 3400 + }, + { + "grad_norm": 0.3411717712879181, + "learning_rate": 9.889704269924954e-05, + "loss": 0.0887, + "step": 3410 + }, + { + "grad_norm": 0.47457781434059143, + "learning_rate": 9.888550032541059e-05, + "loss": 0.0859, + "step": 3420 + }, + { + "grad_norm": 0.3714228868484497, + "learning_rate": 9.887389855107916e-05, + "loss": 0.0904, + "step": 3430 + }, + { + "grad_norm": 0.40049436688423157, + "learning_rate": 9.886223739035248e-05, + "loss": 0.1005, + "step": 3440 + }, + { + "grad_norm": 0.4535018503665924, + "learning_rate": 9.885051685739997e-05, + "loss": 0.0846, + "step": 3450 + }, + { + "grad_norm": 0.44700679183006287, + "learning_rate": 9.883873696646316e-05, + "loss": 0.0779, + "step": 3460 + }, + { + "grad_norm": 0.4428093135356903, + "learning_rate": 9.882689773185575e-05, + "loss": 0.0926, + "step": 3470 + }, + { + "grad_norm": 0.5374898910522461, + "learning_rate": 9.881499916796353e-05, + "loss": 0.098, + "step": 3480 + }, + { + "grad_norm": 0.38435623049736023, + "learning_rate": 9.880304128924434e-05, + "loss": 0.0839, + "step": 3490 + }, + { + "grad_norm": 0.40651312470436096, + "learning_rate": 9.879102411022817e-05, + "loss": 0.0767, + "step": 3500 + }, + { + "grad_norm": 0.35287466645240784, + "learning_rate": 9.877894764551703e-05, + "loss": 0.0835, + "step": 3510 + }, + { + "grad_norm": 0.40496355295181274, + "learning_rate": 9.876681190978494e-05, + "loss": 0.0835, + "step": 3520 + }, + { + "grad_norm": 0.40663811564445496, + "learning_rate": 9.875461691777797e-05, + "loss": 0.0835, + "step": 3530 + }, + { + "grad_norm": 0.37282705307006836, + "learning_rate": 9.874236268431417e-05, + "loss": 0.0844, + "step": 3540 + }, + { + "grad_norm": 0.36780479550361633, + "learning_rate": 9.873004922428361e-05, + "loss": 0.0843, + "step": 3550 + }, + { + "grad_norm": 0.4676786959171295, + "learning_rate": 9.871767655264829e-05, + "loss": 0.0711, + "step": 3560 + }, + { + "grad_norm": 0.3080819249153137, + "learning_rate": 9.87052446844422e-05, + "loss": 0.0878, + "step": 3570 + }, + { + "grad_norm": 0.4025127589702606, + "learning_rate": 9.869275363477122e-05, + "loss": 0.0974, + "step": 3580 + }, + { + "grad_norm": 0.4986024796962738, + "learning_rate": 9.868020341881312e-05, + "loss": 0.1015, + "step": 3590 + }, + { + "grad_norm": 0.4008992612361908, + "learning_rate": 9.866759405181765e-05, + "loss": 0.091, + "step": 3600 + }, + { + "grad_norm": 0.3436967432498932, + "learning_rate": 9.865492554910633e-05, + "loss": 0.0812, + "step": 3610 + }, + { + "grad_norm": 0.33228084444999695, + "learning_rate": 9.864219792607262e-05, + "loss": 0.0755, + "step": 3620 + }, + { + "grad_norm": 0.38205280900001526, + "learning_rate": 9.862941119818177e-05, + "loss": 0.0686, + "step": 3630 + }, + { + "grad_norm": 0.463189959526062, + "learning_rate": 9.861656538097086e-05, + "loss": 0.083, + "step": 3640 + }, + { + "grad_norm": 0.4077150523662567, + "learning_rate": 9.860366049004877e-05, + "loss": 0.0854, + "step": 3650 + }, + { + "grad_norm": 0.44698309898376465, + "learning_rate": 9.859069654109615e-05, + "loss": 0.0795, + "step": 3660 + }, + { + "grad_norm": 0.4371110200881958, + "learning_rate": 9.857767354986545e-05, + "loss": 0.0868, + "step": 3670 + }, + { + "grad_norm": 0.4124850630760193, + "learning_rate": 9.856459153218078e-05, + "loss": 0.0805, + "step": 3680 + }, + { + "grad_norm": 0.34927934408187866, + "learning_rate": 9.855145050393808e-05, + "loss": 0.0749, + "step": 3690 + }, + { + "grad_norm": 0.320297509431839, + "learning_rate": 9.85382504811049e-05, + "loss": 0.07, + "step": 3700 + }, + { + "grad_norm": 0.38276198506355286, + "learning_rate": 9.852499147972054e-05, + "loss": 0.0782, + "step": 3710 + }, + { + "grad_norm": 0.41985082626342773, + "learning_rate": 9.851167351589592e-05, + "loss": 0.0834, + "step": 3720 + }, + { + "grad_norm": 0.39600902795791626, + "learning_rate": 9.849829660581363e-05, + "loss": 0.0765, + "step": 3730 + }, + { + "grad_norm": 0.4088449478149414, + "learning_rate": 9.848486076572787e-05, + "loss": 0.0785, + "step": 3740 + }, + { + "grad_norm": 0.38219979405403137, + "learning_rate": 9.847136601196446e-05, + "loss": 0.0883, + "step": 3750 + }, + { + "grad_norm": 0.3892247974872589, + "learning_rate": 9.845781236092078e-05, + "loss": 0.0683, + "step": 3760 + }, + { + "grad_norm": 0.39979979395866394, + "learning_rate": 9.844419982906583e-05, + "loss": 0.0891, + "step": 3770 + }, + { + "grad_norm": 0.36324378848075867, + "learning_rate": 9.843052843294008e-05, + "loss": 0.0816, + "step": 3780 + }, + { + "grad_norm": 0.37790122628211975, + "learning_rate": 9.841679818915559e-05, + "loss": 0.0744, + "step": 3790 + }, + { + "grad_norm": 0.4262799620628357, + "learning_rate": 9.840300911439591e-05, + "loss": 0.1055, + "step": 3800 + }, + { + "grad_norm": 0.320901095867157, + "learning_rate": 9.838916122541603e-05, + "loss": 0.078, + "step": 3810 + }, + { + "grad_norm": 0.3061879575252533, + "learning_rate": 9.837525453904246e-05, + "loss": 0.0856, + "step": 3820 + }, + { + "grad_norm": 0.31549859046936035, + "learning_rate": 9.836128907217314e-05, + "loss": 0.0692, + "step": 3830 + }, + { + "grad_norm": 0.35183101892471313, + "learning_rate": 9.834726484177743e-05, + "loss": 0.0726, + "step": 3840 + }, + { + "grad_norm": 0.3567470908164978, + "learning_rate": 9.833318186489609e-05, + "loss": 0.0921, + "step": 3850 + }, + { + "grad_norm": 0.42611056566238403, + "learning_rate": 9.831904015864126e-05, + "loss": 0.0771, + "step": 3860 + }, + { + "grad_norm": 0.41698014736175537, + "learning_rate": 9.830483974019645e-05, + "loss": 0.0745, + "step": 3870 + }, + { + "grad_norm": 0.3562908172607422, + "learning_rate": 9.82905806268165e-05, + "loss": 0.0713, + "step": 3880 + }, + { + "grad_norm": 0.32529765367507935, + "learning_rate": 9.82762628358276e-05, + "loss": 0.0848, + "step": 3890 + }, + { + "grad_norm": 0.3613531291484833, + "learning_rate": 9.826188638462718e-05, + "loss": 0.0833, + "step": 3900 + }, + { + "grad_norm": 0.4655057191848755, + "learning_rate": 9.824745129068402e-05, + "loss": 0.0775, + "step": 3910 + }, + { + "grad_norm": 0.3808934986591339, + "learning_rate": 9.82329575715381e-05, + "loss": 0.0875, + "step": 3920 + }, + { + "grad_norm": 0.3568260669708252, + "learning_rate": 9.821840524480066e-05, + "loss": 0.0768, + "step": 3930 + }, + { + "grad_norm": 0.31418269872665405, + "learning_rate": 9.820379432815414e-05, + "loss": 0.0878, + "step": 3940 + }, + { + "grad_norm": 0.443363755941391, + "learning_rate": 9.81891248393522e-05, + "loss": 0.0882, + "step": 3950 + }, + { + "grad_norm": 0.44039997458457947, + "learning_rate": 9.817439679621963e-05, + "loss": 0.0911, + "step": 3960 + }, + { + "grad_norm": 0.4505123496055603, + "learning_rate": 9.815961021665243e-05, + "loss": 0.0708, + "step": 3970 + }, + { + "grad_norm": 0.3179085850715637, + "learning_rate": 9.814476511861763e-05, + "loss": 0.0883, + "step": 3980 + }, + { + "grad_norm": 0.40261784195899963, + "learning_rate": 9.812986152015348e-05, + "loss": 0.0902, + "step": 3990 + }, + { + "grad_norm": 0.4452002942562103, + "learning_rate": 9.811489943936922e-05, + "loss": 0.0801, + "step": 4000 + }, + { + "grad_norm": 0.3613608777523041, + "learning_rate": 9.809987889444522e-05, + "loss": 0.076, + "step": 4010 + }, + { + "grad_norm": 0.3730508089065552, + "learning_rate": 9.808479990363282e-05, + "loss": 0.0843, + "step": 4020 + }, + { + "grad_norm": 0.3806791305541992, + "learning_rate": 9.806966248525445e-05, + "loss": 0.0895, + "step": 4030 + }, + { + "grad_norm": 0.3366662263870239, + "learning_rate": 9.805446665770348e-05, + "loss": 0.0712, + "step": 4040 + }, + { + "grad_norm": 0.4037311375141144, + "learning_rate": 9.803921243944429e-05, + "loss": 0.0846, + "step": 4050 + }, + { + "grad_norm": 0.2969924211502075, + "learning_rate": 9.802389984901218e-05, + "loss": 0.0752, + "step": 4060 + }, + { + "grad_norm": 0.32146286964416504, + "learning_rate": 9.80085289050134e-05, + "loss": 0.0772, + "step": 4070 + }, + { + "grad_norm": 0.26851075887680054, + "learning_rate": 9.799309962612508e-05, + "loss": 0.088, + "step": 4080 + }, + { + "grad_norm": 0.3639536499977112, + "learning_rate": 9.797761203109527e-05, + "loss": 0.0779, + "step": 4090 + }, + { + "grad_norm": 0.3490807116031647, + "learning_rate": 9.796206613874283e-05, + "loss": 0.0701, + "step": 4100 + }, + { + "grad_norm": 0.373683899641037, + "learning_rate": 9.794646196795754e-05, + "loss": 0.0788, + "step": 4110 + }, + { + "grad_norm": 0.6021084785461426, + "learning_rate": 9.793079953769987e-05, + "loss": 0.0861, + "step": 4120 + }, + { + "grad_norm": 0.3044678568840027, + "learning_rate": 9.79150788670012e-05, + "loss": 0.0855, + "step": 4130 + }, + { + "grad_norm": 0.38936150074005127, + "learning_rate": 9.78992999749636e-05, + "loss": 0.0803, + "step": 4140 + }, + { + "grad_norm": 0.38170117139816284, + "learning_rate": 9.788346288075994e-05, + "loss": 0.0902, + "step": 4150 + }, + { + "grad_norm": 0.3732798993587494, + "learning_rate": 9.786756760363373e-05, + "loss": 0.0782, + "step": 4160 + }, + { + "grad_norm": 0.3583327829837799, + "learning_rate": 9.78516141628993e-05, + "loss": 0.078, + "step": 4170 + }, + { + "grad_norm": 0.35289281606674194, + "learning_rate": 9.783560257794154e-05, + "loss": 0.0737, + "step": 4180 + }, + { + "grad_norm": 0.35022667050361633, + "learning_rate": 9.781953286821603e-05, + "loss": 0.0795, + "step": 4190 + }, + { + "grad_norm": 0.36590373516082764, + "learning_rate": 9.780340505324901e-05, + "loss": 0.076, + "step": 4200 + }, + { + "grad_norm": 0.33465808629989624, + "learning_rate": 9.778721915263727e-05, + "loss": 0.0773, + "step": 4210 + }, + { + "grad_norm": 0.31979843974113464, + "learning_rate": 9.777097518604824e-05, + "loss": 0.0934, + "step": 4220 + }, + { + "grad_norm": 0.350436806678772, + "learning_rate": 9.775467317321984e-05, + "loss": 0.0928, + "step": 4230 + }, + { + "grad_norm": 0.32108908891677856, + "learning_rate": 9.773831313396055e-05, + "loss": 0.0897, + "step": 4240 + }, + { + "grad_norm": 0.37086448073387146, + "learning_rate": 9.77218950881494e-05, + "loss": 0.0846, + "step": 4250 + }, + { + "grad_norm": 0.3528153896331787, + "learning_rate": 9.770541905573583e-05, + "loss": 0.0838, + "step": 4260 + }, + { + "grad_norm": 0.3567950129508972, + "learning_rate": 9.768888505673976e-05, + "loss": 0.0786, + "step": 4270 + }, + { + "grad_norm": 0.3640773594379425, + "learning_rate": 9.767229311125162e-05, + "loss": 0.0694, + "step": 4280 + }, + { + "grad_norm": 0.4295765161514282, + "learning_rate": 9.765564323943211e-05, + "loss": 0.0766, + "step": 4290 + }, + { + "grad_norm": 0.36064642667770386, + "learning_rate": 9.763893546151244e-05, + "loss": 0.0654, + "step": 4300 + }, + { + "grad_norm": 0.37981274724006653, + "learning_rate": 9.762216979779412e-05, + "loss": 0.0803, + "step": 4310 + }, + { + "grad_norm": 0.4492497146129608, + "learning_rate": 9.760534626864902e-05, + "loss": 0.0998, + "step": 4320 + }, + { + "grad_norm": 0.43422701954841614, + "learning_rate": 9.758846489451931e-05, + "loss": 0.075, + "step": 4330 + }, + { + "grad_norm": 0.36232230067253113, + "learning_rate": 9.757152569591748e-05, + "loss": 0.0675, + "step": 4340 + }, + { + "grad_norm": 0.2780834436416626, + "learning_rate": 9.75545286934262e-05, + "loss": 0.0607, + "step": 4350 + }, + { + "grad_norm": 0.346218079328537, + "learning_rate": 9.753747390769847e-05, + "loss": 0.0713, + "step": 4360 + }, + { + "grad_norm": 0.39921265840530396, + "learning_rate": 9.752036135945744e-05, + "loss": 0.0779, + "step": 4370 + }, + { + "grad_norm": 0.3122653663158417, + "learning_rate": 9.750319106949649e-05, + "loss": 0.0804, + "step": 4380 + }, + { + "grad_norm": 0.39125514030456543, + "learning_rate": 9.748596305867913e-05, + "loss": 0.0685, + "step": 4390 + }, + { + "grad_norm": 0.39845985174179077, + "learning_rate": 9.746867734793903e-05, + "loss": 0.0887, + "step": 4400 + }, + { + "grad_norm": 0.39573806524276733, + "learning_rate": 9.745133395827993e-05, + "loss": 0.0816, + "step": 4410 + }, + { + "grad_norm": 0.43776023387908936, + "learning_rate": 9.743393291077572e-05, + "loss": 0.0947, + "step": 4420 + }, + { + "grad_norm": 0.4129426181316376, + "learning_rate": 9.741647422657028e-05, + "loss": 0.0846, + "step": 4430 + }, + { + "grad_norm": 0.29558053612709045, + "learning_rate": 9.739895792687758e-05, + "loss": 0.0814, + "step": 4440 + }, + { + "grad_norm": 0.3221525251865387, + "learning_rate": 9.738138403298157e-05, + "loss": 0.0884, + "step": 4450 + }, + { + "grad_norm": 0.3943980634212494, + "learning_rate": 9.736375256623619e-05, + "loss": 0.0757, + "step": 4460 + }, + { + "grad_norm": 0.34983253479003906, + "learning_rate": 9.734606354806533e-05, + "loss": 0.081, + "step": 4470 + }, + { + "grad_norm": 0.3389854431152344, + "learning_rate": 9.73283169999628e-05, + "loss": 0.0873, + "step": 4480 + }, + { + "grad_norm": 0.36907777190208435, + "learning_rate": 9.731051294349238e-05, + "loss": 0.0872, + "step": 4490 + }, + { + "grad_norm": 0.3549933135509491, + "learning_rate": 9.729265140028762e-05, + "loss": 0.0822, + "step": 4500 + }, + { + "grad_norm": 0.39877042174339294, + "learning_rate": 9.727473239205201e-05, + "loss": 0.105, + "step": 4510 + }, + { + "grad_norm": 0.3073529303073883, + "learning_rate": 9.725675594055883e-05, + "loss": 0.0772, + "step": 4520 + }, + { + "grad_norm": 0.33667460083961487, + "learning_rate": 9.723872206765116e-05, + "loss": 0.0788, + "step": 4530 + }, + { + "grad_norm": 0.4809662401676178, + "learning_rate": 9.722063079524185e-05, + "loss": 0.0945, + "step": 4540 + }, + { + "grad_norm": 0.42365628480911255, + "learning_rate": 9.720248214531351e-05, + "loss": 0.0917, + "step": 4550 + }, + { + "grad_norm": 0.3855436146259308, + "learning_rate": 9.718427613991848e-05, + "loss": 0.0739, + "step": 4560 + }, + { + "grad_norm": 0.38019341230392456, + "learning_rate": 9.716601280117873e-05, + "loss": 0.0911, + "step": 4570 + }, + { + "grad_norm": 0.39614251255989075, + "learning_rate": 9.714769215128596e-05, + "loss": 0.0846, + "step": 4580 + }, + { + "grad_norm": 0.3134308457374573, + "learning_rate": 9.712931421250152e-05, + "loss": 0.0771, + "step": 4590 + }, + { + "grad_norm": 0.3715668320655823, + "learning_rate": 9.711087900715627e-05, + "loss": 0.0793, + "step": 4600 + }, + { + "grad_norm": 0.34114009141921997, + "learning_rate": 9.709238655765078e-05, + "loss": 0.0796, + "step": 4610 + }, + { + "grad_norm": 0.34276294708251953, + "learning_rate": 9.707383688645511e-05, + "loss": 0.0688, + "step": 4620 + }, + { + "grad_norm": 0.4090898633003235, + "learning_rate": 9.705523001610883e-05, + "loss": 0.0817, + "step": 4630 + }, + { + "grad_norm": 0.3592730760574341, + "learning_rate": 9.703656596922107e-05, + "loss": 0.0761, + "step": 4640 + }, + { + "grad_norm": 0.3821835219860077, + "learning_rate": 9.70178447684704e-05, + "loss": 0.0688, + "step": 4650 + }, + { + "grad_norm": 0.4046807885169983, + "learning_rate": 9.699906643660483e-05, + "loss": 0.0818, + "step": 4660 + }, + { + "grad_norm": 0.4721336364746094, + "learning_rate": 9.698023099644185e-05, + "loss": 0.0747, + "step": 4670 + }, + { + "grad_norm": 0.4179655909538269, + "learning_rate": 9.696133847086823e-05, + "loss": 0.0794, + "step": 4680 + }, + { + "grad_norm": 0.3553186357021332, + "learning_rate": 9.694238888284022e-05, + "loss": 0.0991, + "step": 4690 + }, + { + "grad_norm": 0.3423902988433838, + "learning_rate": 9.692338225538333e-05, + "loss": 0.0799, + "step": 4700 + }, + { + "grad_norm": 0.41327622532844543, + "learning_rate": 9.690431861159241e-05, + "loss": 0.0872, + "step": 4710 + }, + { + "grad_norm": 0.40878844261169434, + "learning_rate": 9.688519797463161e-05, + "loss": 0.0738, + "step": 4720 + }, + { + "grad_norm": 0.34178587794303894, + "learning_rate": 9.686602036773426e-05, + "loss": 0.0934, + "step": 4730 + }, + { + "grad_norm": 0.3249698579311371, + "learning_rate": 9.684678581420302e-05, + "loss": 0.0818, + "step": 4740 + }, + { + "grad_norm": 0.30668920278549194, + "learning_rate": 9.682749433740962e-05, + "loss": 0.0709, + "step": 4750 + }, + { + "grad_norm": 0.3032350540161133, + "learning_rate": 9.680814596079507e-05, + "loss": 0.0781, + "step": 4760 + }, + { + "grad_norm": 0.3527970314025879, + "learning_rate": 9.678874070786945e-05, + "loss": 0.085, + "step": 4770 + }, + { + "grad_norm": 0.4545914828777313, + "learning_rate": 9.676927860221199e-05, + "loss": 0.0889, + "step": 4780 + }, + { + "grad_norm": 0.32243612408638, + "learning_rate": 9.674975966747097e-05, + "loss": 0.0734, + "step": 4790 + }, + { + "grad_norm": 0.3319143056869507, + "learning_rate": 9.673018392736374e-05, + "loss": 0.075, + "step": 4800 + }, + { + "grad_norm": 0.30873703956604004, + "learning_rate": 9.671055140567667e-05, + "loss": 0.0811, + "step": 4810 + }, + { + "grad_norm": 0.3133257031440735, + "learning_rate": 9.669086212626511e-05, + "loss": 0.0848, + "step": 4820 + }, + { + "grad_norm": 0.38964468240737915, + "learning_rate": 9.667111611305341e-05, + "loss": 0.0831, + "step": 4830 + }, + { + "grad_norm": 0.3691699802875519, + "learning_rate": 9.665131339003486e-05, + "loss": 0.0802, + "step": 4840 + }, + { + "grad_norm": 0.3550565540790558, + "learning_rate": 9.663145398127158e-05, + "loss": 0.0811, + "step": 4850 + }, + { + "grad_norm": 0.34546104073524475, + "learning_rate": 9.661153791089467e-05, + "loss": 0.0657, + "step": 4860 + }, + { + "grad_norm": 0.38052675127983093, + "learning_rate": 9.659156520310402e-05, + "loss": 0.0751, + "step": 4870 + }, + { + "grad_norm": 0.31614619493484497, + "learning_rate": 9.657153588216835e-05, + "loss": 0.0701, + "step": 4880 + }, + { + "grad_norm": 0.3901197910308838, + "learning_rate": 9.655144997242516e-05, + "loss": 0.0884, + "step": 4890 + }, + { + "grad_norm": 0.2958650588989258, + "learning_rate": 9.653130749828075e-05, + "loss": 0.0676, + "step": 4900 + }, + { + "grad_norm": 0.5070604681968689, + "learning_rate": 9.65111084842101e-05, + "loss": 0.079, + "step": 4910 + }, + { + "grad_norm": 0.3388635516166687, + "learning_rate": 9.649085295475695e-05, + "loss": 0.0811, + "step": 4920 + }, + { + "grad_norm": 0.373856782913208, + "learning_rate": 9.647054093453365e-05, + "loss": 0.0796, + "step": 4930 + }, + { + "grad_norm": 0.3570548892021179, + "learning_rate": 9.645017244822123e-05, + "loss": 0.0673, + "step": 4940 + }, + { + "grad_norm": 0.3147738575935364, + "learning_rate": 9.642974752056931e-05, + "loss": 0.0811, + "step": 4950 + }, + { + "grad_norm": 0.3751422166824341, + "learning_rate": 9.640926617639613e-05, + "loss": 0.0767, + "step": 4960 + }, + { + "grad_norm": 0.35979461669921875, + "learning_rate": 9.638872844058843e-05, + "loss": 0.0851, + "step": 4970 + }, + { + "grad_norm": 0.3602491319179535, + "learning_rate": 9.63681343381015e-05, + "loss": 0.0772, + "step": 4980 + }, + { + "grad_norm": 0.21682311594486237, + "learning_rate": 9.634748389395914e-05, + "loss": 0.0681, + "step": 4990 + }, + { + "grad_norm": 0.36068013310432434, + "learning_rate": 9.632677713325353e-05, + "loss": 0.0817, + "step": 5000 + }, + { + "grad_norm": 0.4300607144832611, + "learning_rate": 9.63060140811454e-05, + "loss": 0.0802, + "step": 5010 + }, + { + "grad_norm": 0.39744478464126587, + "learning_rate": 9.628519476286379e-05, + "loss": 0.0717, + "step": 5020 + }, + { + "grad_norm": 0.26179471611976624, + "learning_rate": 9.626431920370612e-05, + "loss": 0.0609, + "step": 5030 + }, + { + "grad_norm": 0.39819440245628357, + "learning_rate": 9.624338742903819e-05, + "loss": 0.071, + "step": 5040 + }, + { + "grad_norm": 0.32530081272125244, + "learning_rate": 9.622239946429406e-05, + "loss": 0.0837, + "step": 5050 + }, + { + "grad_norm": 0.2880935072898865, + "learning_rate": 9.620135533497609e-05, + "loss": 0.0616, + "step": 5060 + }, + { + "grad_norm": 0.29914334416389465, + "learning_rate": 9.61802550666549e-05, + "loss": 0.0982, + "step": 5070 + }, + { + "grad_norm": 0.3077535927295685, + "learning_rate": 9.615909868496928e-05, + "loss": 0.0862, + "step": 5080 + }, + { + "grad_norm": 0.3008662760257721, + "learning_rate": 9.613788621562622e-05, + "loss": 0.0677, + "step": 5090 + }, + { + "grad_norm": 0.37613266706466675, + "learning_rate": 9.611661768440093e-05, + "loss": 0.0706, + "step": 5100 + }, + { + "grad_norm": 0.37448614835739136, + "learning_rate": 9.609529311713661e-05, + "loss": 0.0834, + "step": 5110 + }, + { + "grad_norm": 0.30840733647346497, + "learning_rate": 9.607391253974466e-05, + "loss": 0.0906, + "step": 5120 + }, + { + "grad_norm": 0.32020097970962524, + "learning_rate": 9.605247597820448e-05, + "loss": 0.0806, + "step": 5130 + }, + { + "grad_norm": 0.3742756247520447, + "learning_rate": 9.603098345856354e-05, + "loss": 0.0823, + "step": 5140 + }, + { + "grad_norm": 0.34984713792800903, + "learning_rate": 9.600943500693725e-05, + "loss": 0.0685, + "step": 5150 + }, + { + "grad_norm": 0.31098803877830505, + "learning_rate": 9.598783064950902e-05, + "loss": 0.0925, + "step": 5160 + }, + { + "grad_norm": 0.3337903916835785, + "learning_rate": 9.596617041253018e-05, + "loss": 0.0747, + "step": 5170 + }, + { + "grad_norm": 0.2817566394805908, + "learning_rate": 9.594445432231996e-05, + "loss": 0.0679, + "step": 5180 + }, + { + "grad_norm": 0.31515994668006897, + "learning_rate": 9.592268240526547e-05, + "loss": 0.0961, + "step": 5190 + }, + { + "grad_norm": 0.32990148663520813, + "learning_rate": 9.590085468782162e-05, + "loss": 0.0628, + "step": 5200 + }, + { + "grad_norm": 0.3763536810874939, + "learning_rate": 9.587897119651116e-05, + "loss": 0.0746, + "step": 5210 + }, + { + "grad_norm": 0.3864741623401642, + "learning_rate": 9.585703195792459e-05, + "loss": 0.0715, + "step": 5220 + }, + { + "grad_norm": 0.40956375002861023, + "learning_rate": 9.583503699872016e-05, + "loss": 0.0705, + "step": 5230 + }, + { + "grad_norm": 0.3632172644138336, + "learning_rate": 9.581298634562381e-05, + "loss": 0.0698, + "step": 5240 + }, + { + "grad_norm": 0.37006452679634094, + "learning_rate": 9.579088002542917e-05, + "loss": 0.0897, + "step": 5250 + }, + { + "grad_norm": 0.3521057069301605, + "learning_rate": 9.57687180649975e-05, + "loss": 0.069, + "step": 5260 + }, + { + "grad_norm": 0.3896416425704956, + "learning_rate": 9.574650049125768e-05, + "loss": 0.0713, + "step": 5270 + }, + { + "grad_norm": 0.34124037623405457, + "learning_rate": 9.572422733120614e-05, + "loss": 0.0737, + "step": 5280 + }, + { + "grad_norm": 0.2962722182273865, + "learning_rate": 9.570189861190689e-05, + "loss": 0.0742, + "step": 5290 + }, + { + "grad_norm": 0.3290007710456848, + "learning_rate": 9.56795143604914e-05, + "loss": 0.0651, + "step": 5300 + }, + { + "grad_norm": 0.2914034426212311, + "learning_rate": 9.565707460415869e-05, + "loss": 0.0578, + "step": 5310 + }, + { + "grad_norm": 0.27696359157562256, + "learning_rate": 9.563457937017515e-05, + "loss": 0.0739, + "step": 5320 + }, + { + "grad_norm": 0.4333876073360443, + "learning_rate": 9.56120286858746e-05, + "loss": 0.0865, + "step": 5330 + }, + { + "grad_norm": 0.37735089659690857, + "learning_rate": 9.558942257865829e-05, + "loss": 0.0712, + "step": 5340 + }, + { + "grad_norm": 0.2352084219455719, + "learning_rate": 9.556676107599472e-05, + "loss": 0.0631, + "step": 5350 + }, + { + "grad_norm": 0.435627281665802, + "learning_rate": 9.554404420541978e-05, + "loss": 0.0716, + "step": 5360 + }, + { + "grad_norm": 0.3229515254497528, + "learning_rate": 9.55212719945366e-05, + "loss": 0.0584, + "step": 5370 + }, + { + "grad_norm": 0.3254145085811615, + "learning_rate": 9.549844447101559e-05, + "loss": 0.0613, + "step": 5380 + }, + { + "grad_norm": 0.2995050847530365, + "learning_rate": 9.547556166259433e-05, + "loss": 0.0666, + "step": 5390 + }, + { + "grad_norm": 0.35769715905189514, + "learning_rate": 9.545262359707756e-05, + "loss": 0.0756, + "step": 5400 + }, + { + "grad_norm": 0.2691922187805176, + "learning_rate": 9.542963030233724e-05, + "loss": 0.0667, + "step": 5410 + }, + { + "grad_norm": 0.37167802453041077, + "learning_rate": 9.540658180631237e-05, + "loss": 0.0684, + "step": 5420 + }, + { + "grad_norm": 0.3279150128364563, + "learning_rate": 9.538347813700904e-05, + "loss": 0.0662, + "step": 5430 + }, + { + "grad_norm": 0.30230098962783813, + "learning_rate": 9.536031932250036e-05, + "loss": 0.083, + "step": 5440 + }, + { + "grad_norm": 0.3736446499824524, + "learning_rate": 9.533710539092653e-05, + "loss": 0.0546, + "step": 5450 + }, + { + "grad_norm": 0.2874501049518585, + "learning_rate": 9.531383637049464e-05, + "loss": 0.0614, + "step": 5460 + }, + { + "grad_norm": 0.33796513080596924, + "learning_rate": 9.529051228947875e-05, + "loss": 0.0636, + "step": 5470 + }, + { + "grad_norm": 0.30049335956573486, + "learning_rate": 9.52671331762198e-05, + "loss": 0.0684, + "step": 5480 + }, + { + "grad_norm": 0.2800152599811554, + "learning_rate": 9.524369905912565e-05, + "loss": 0.0676, + "step": 5490 + }, + { + "grad_norm": 0.34990885853767395, + "learning_rate": 9.522020996667092e-05, + "loss": 0.0741, + "step": 5500 + }, + { + "grad_norm": 0.2770771384239197, + "learning_rate": 9.519666592739709e-05, + "loss": 0.0693, + "step": 5510 + }, + { + "grad_norm": 0.2811853289604187, + "learning_rate": 9.517306696991241e-05, + "loss": 0.0619, + "step": 5520 + }, + { + "grad_norm": 0.30597004294395447, + "learning_rate": 9.51494131228918e-05, + "loss": 0.0685, + "step": 5530 + }, + { + "grad_norm": 0.3272843360900879, + "learning_rate": 9.512570441507695e-05, + "loss": 0.0652, + "step": 5540 + }, + { + "grad_norm": 0.36188942193984985, + "learning_rate": 9.510194087527615e-05, + "loss": 0.0653, + "step": 5550 + }, + { + "grad_norm": 0.2715345025062561, + "learning_rate": 9.507812253236435e-05, + "loss": 0.0633, + "step": 5560 + }, + { + "grad_norm": 0.36622563004493713, + "learning_rate": 9.505424941528309e-05, + "loss": 0.0638, + "step": 5570 + }, + { + "grad_norm": 0.343437135219574, + "learning_rate": 9.503032155304046e-05, + "loss": 0.089, + "step": 5580 + }, + { + "grad_norm": 0.3444344103336334, + "learning_rate": 9.500633897471106e-05, + "loss": 0.062, + "step": 5590 + }, + { + "grad_norm": 0.3655736744403839, + "learning_rate": 9.498230170943596e-05, + "loss": 0.0756, + "step": 5600 + }, + { + "grad_norm": 0.32791855931282043, + "learning_rate": 9.495820978642275e-05, + "loss": 0.0676, + "step": 5610 + }, + { + "grad_norm": 0.35262367129325867, + "learning_rate": 9.493406323494535e-05, + "loss": 0.0634, + "step": 5620 + }, + { + "grad_norm": 0.2358086109161377, + "learning_rate": 9.490986208434413e-05, + "loss": 0.0553, + "step": 5630 + }, + { + "grad_norm": 0.4025478661060333, + "learning_rate": 9.488560636402577e-05, + "loss": 0.0614, + "step": 5640 + }, + { + "grad_norm": 0.4370748698711395, + "learning_rate": 9.486129610346321e-05, + "loss": 0.0818, + "step": 5650 + }, + { + "grad_norm": 0.34372904896736145, + "learning_rate": 9.483693133219576e-05, + "loss": 0.0664, + "step": 5660 + }, + { + "grad_norm": 0.3874727487564087, + "learning_rate": 9.481251207982888e-05, + "loss": 0.0674, + "step": 5670 + }, + { + "grad_norm": 0.3849292993545532, + "learning_rate": 9.47880383760343e-05, + "loss": 0.0712, + "step": 5680 + }, + { + "grad_norm": 0.37215951085090637, + "learning_rate": 9.476351025054983e-05, + "loss": 0.0728, + "step": 5690 + }, + { + "grad_norm": 0.34720635414123535, + "learning_rate": 9.473892773317952e-05, + "loss": 0.0631, + "step": 5700 + }, + { + "grad_norm": 0.29277053475379944, + "learning_rate": 9.471429085379338e-05, + "loss": 0.0778, + "step": 5710 + }, + { + "grad_norm": 0.34614256024360657, + "learning_rate": 9.468959964232757e-05, + "loss": 0.0635, + "step": 5720 + }, + { + "grad_norm": 0.32548248767852783, + "learning_rate": 9.466485412878425e-05, + "loss": 0.0678, + "step": 5730 + }, + { + "grad_norm": 0.30909767746925354, + "learning_rate": 9.464005434323154e-05, + "loss": 0.0765, + "step": 5740 + }, + { + "grad_norm": 0.319947212934494, + "learning_rate": 9.461520031580352e-05, + "loss": 0.0647, + "step": 5750 + }, + { + "grad_norm": 0.32429876923561096, + "learning_rate": 9.459029207670019e-05, + "loss": 0.0685, + "step": 5760 + }, + { + "grad_norm": 0.23448695242404938, + "learning_rate": 9.456532965618737e-05, + "loss": 0.0641, + "step": 5770 + }, + { + "grad_norm": 0.3355346620082855, + "learning_rate": 9.454031308459681e-05, + "loss": 0.0825, + "step": 5780 + }, + { + "grad_norm": 0.25151774287223816, + "learning_rate": 9.451524239232595e-05, + "loss": 0.0603, + "step": 5790 + }, + { + "grad_norm": 0.3149898052215576, + "learning_rate": 9.449011760983809e-05, + "loss": 0.0812, + "step": 5800 + }, + { + "grad_norm": 0.3684616982936859, + "learning_rate": 9.446493876766218e-05, + "loss": 0.0709, + "step": 5810 + }, + { + "grad_norm": 0.2988879680633545, + "learning_rate": 9.44397058963929e-05, + "loss": 0.0665, + "step": 5820 + }, + { + "grad_norm": 0.35528990626335144, + "learning_rate": 9.441441902669056e-05, + "loss": 0.0709, + "step": 5830 + }, + { + "grad_norm": 0.3166639804840088, + "learning_rate": 9.43890781892811e-05, + "loss": 0.0672, + "step": 5840 + }, + { + "grad_norm": 0.3343195617198944, + "learning_rate": 9.436368341495603e-05, + "loss": 0.0595, + "step": 5850 + }, + { + "grad_norm": 0.23363587260246277, + "learning_rate": 9.43382347345724e-05, + "loss": 0.0716, + "step": 5860 + }, + { + "grad_norm": 0.3526093065738678, + "learning_rate": 9.431273217905273e-05, + "loss": 0.0752, + "step": 5870 + }, + { + "grad_norm": 0.3657768964767456, + "learning_rate": 9.428717577938504e-05, + "loss": 0.08, + "step": 5880 + }, + { + "grad_norm": 0.3492029905319214, + "learning_rate": 9.426156556662276e-05, + "loss": 0.0671, + "step": 5890 + }, + { + "grad_norm": 0.33571958541870117, + "learning_rate": 9.423590157188474e-05, + "loss": 0.0846, + "step": 5900 + }, + { + "grad_norm": 0.22962675988674164, + "learning_rate": 9.421018382635513e-05, + "loss": 0.0699, + "step": 5910 + }, + { + "grad_norm": 0.34968316555023193, + "learning_rate": 9.418441236128343e-05, + "loss": 0.0637, + "step": 5920 + }, + { + "grad_norm": 0.2607312798500061, + "learning_rate": 9.41585872079844e-05, + "loss": 0.0605, + "step": 5930 + }, + { + "grad_norm": 0.2509673833847046, + "learning_rate": 9.413270839783802e-05, + "loss": 0.0659, + "step": 5940 + }, + { + "grad_norm": 0.27132976055145264, + "learning_rate": 9.41067759622895e-05, + "loss": 0.0575, + "step": 5950 + }, + { + "grad_norm": 0.2866787910461426, + "learning_rate": 9.408078993284917e-05, + "loss": 0.0595, + "step": 5960 + }, + { + "grad_norm": 0.3039873540401459, + "learning_rate": 9.405475034109254e-05, + "loss": 0.0645, + "step": 5970 + }, + { + "grad_norm": 0.41472241282463074, + "learning_rate": 9.402865721866015e-05, + "loss": 0.0668, + "step": 5980 + }, + { + "grad_norm": 0.2999403178691864, + "learning_rate": 9.400251059725762e-05, + "loss": 0.0791, + "step": 5990 + }, + { + "grad_norm": 0.37193819880485535, + "learning_rate": 9.397631050865554e-05, + "loss": 0.092, + "step": 6000 + }, + { + "grad_norm": 0.31223922967910767, + "learning_rate": 9.395005698468949e-05, + "loss": 0.0642, + "step": 6010 + }, + { + "grad_norm": 0.3581950068473816, + "learning_rate": 9.392375005726e-05, + "loss": 0.0635, + "step": 6020 + }, + { + "grad_norm": 0.24546028673648834, + "learning_rate": 9.389738975833242e-05, + "loss": 0.0675, + "step": 6030 + }, + { + "grad_norm": 0.3020484149456024, + "learning_rate": 9.387097611993707e-05, + "loss": 0.0619, + "step": 6040 + }, + { + "grad_norm": 0.39207983016967773, + "learning_rate": 9.384450917416894e-05, + "loss": 0.0714, + "step": 6050 + }, + { + "grad_norm": 0.35458531975746155, + "learning_rate": 9.381798895318792e-05, + "loss": 0.0608, + "step": 6060 + }, + { + "grad_norm": 0.35320064425468445, + "learning_rate": 9.379141548921853e-05, + "loss": 0.0679, + "step": 6070 + }, + { + "grad_norm": 0.3492647409439087, + "learning_rate": 9.376478881455009e-05, + "loss": 0.0737, + "step": 6080 + }, + { + "grad_norm": 0.26932772994041443, + "learning_rate": 9.373810896153647e-05, + "loss": 0.068, + "step": 6090 + }, + { + "grad_norm": 0.4199330806732178, + "learning_rate": 9.371137596259623e-05, + "loss": 0.0775, + "step": 6100 + }, + { + "grad_norm": 0.24818776547908783, + "learning_rate": 9.368458985021248e-05, + "loss": 0.0587, + "step": 6110 + }, + { + "grad_norm": 0.30468839406967163, + "learning_rate": 9.365775065693287e-05, + "loss": 0.0673, + "step": 6120 + }, + { + "grad_norm": 0.2986454367637634, + "learning_rate": 9.363085841536957e-05, + "loss": 0.0573, + "step": 6130 + }, + { + "grad_norm": 0.31049644947052, + "learning_rate": 9.360391315819917e-05, + "loss": 0.0683, + "step": 6140 + }, + { + "grad_norm": 0.3427204191684723, + "learning_rate": 9.357691491816269e-05, + "loss": 0.091, + "step": 6150 + }, + { + "grad_norm": 0.3594108521938324, + "learning_rate": 9.354986372806557e-05, + "loss": 0.0875, + "step": 6160 + }, + { + "grad_norm": 0.27852752804756165, + "learning_rate": 9.352275962077752e-05, + "loss": 0.0776, + "step": 6170 + }, + { + "grad_norm": 0.36947229504585266, + "learning_rate": 9.349560262923262e-05, + "loss": 0.0568, + "step": 6180 + }, + { + "grad_norm": 0.3593618869781494, + "learning_rate": 9.346839278642913e-05, + "loss": 0.0768, + "step": 6190 + }, + { + "grad_norm": 0.29034745693206787, + "learning_rate": 9.344113012542963e-05, + "loss": 0.062, + "step": 6200 + }, + { + "grad_norm": 0.29143935441970825, + "learning_rate": 9.341381467936079e-05, + "loss": 0.0711, + "step": 6210 + }, + { + "grad_norm": 0.39614546298980713, + "learning_rate": 9.338644648141346e-05, + "loss": 0.0572, + "step": 6220 + }, + { + "grad_norm": 0.297868549823761, + "learning_rate": 9.335902556484257e-05, + "loss": 0.0689, + "step": 6230 + }, + { + "grad_norm": 0.3448638617992401, + "learning_rate": 9.333155196296712e-05, + "loss": 0.0695, + "step": 6240 + }, + { + "grad_norm": 0.28475135564804077, + "learning_rate": 9.330402570917016e-05, + "loss": 0.0679, + "step": 6250 + }, + { + "grad_norm": 0.25220218300819397, + "learning_rate": 9.327644683689865e-05, + "loss": 0.0562, + "step": 6260 + }, + { + "grad_norm": 0.28449541330337524, + "learning_rate": 9.324881537966354e-05, + "loss": 0.0592, + "step": 6270 + }, + { + "grad_norm": 0.4012209177017212, + "learning_rate": 9.322113137103964e-05, + "loss": 0.0594, + "step": 6280 + }, + { + "grad_norm": 0.36757007241249084, + "learning_rate": 9.319339484466564e-05, + "loss": 0.0666, + "step": 6290 + }, + { + "grad_norm": 0.22058990597724915, + "learning_rate": 9.316560583424403e-05, + "loss": 0.0592, + "step": 6300 + }, + { + "grad_norm": 0.3151439428329468, + "learning_rate": 9.313776437354108e-05, + "loss": 0.0555, + "step": 6310 + }, + { + "grad_norm": 0.3250259459018707, + "learning_rate": 9.31098704963868e-05, + "loss": 0.0588, + "step": 6320 + }, + { + "grad_norm": 0.26238471269607544, + "learning_rate": 9.308192423667485e-05, + "loss": 0.0589, + "step": 6330 + }, + { + "grad_norm": 0.3021993339061737, + "learning_rate": 9.305392562836262e-05, + "loss": 0.0718, + "step": 6340 + }, + { + "grad_norm": 0.3117139935493469, + "learning_rate": 9.3025874705471e-05, + "loss": 0.0638, + "step": 6350 + }, + { + "grad_norm": 0.349623441696167, + "learning_rate": 9.299777150208457e-05, + "loss": 0.0661, + "step": 6360 + }, + { + "grad_norm": 0.3632104992866516, + "learning_rate": 9.296961605235133e-05, + "loss": 0.0581, + "step": 6370 + }, + { + "grad_norm": 0.3572457432746887, + "learning_rate": 9.29414083904828e-05, + "loss": 0.0593, + "step": 6380 + }, + { + "grad_norm": 0.3871201276779175, + "learning_rate": 9.291314855075397e-05, + "loss": 0.0696, + "step": 6390 + }, + { + "grad_norm": 0.27238932251930237, + "learning_rate": 9.288483656750322e-05, + "loss": 0.0662, + "step": 6400 + }, + { + "grad_norm": 0.2907124161720276, + "learning_rate": 9.285647247513225e-05, + "loss": 0.0535, + "step": 6410 + }, + { + "grad_norm": 0.28582894802093506, + "learning_rate": 9.282805630810614e-05, + "loss": 0.0683, + "step": 6420 + }, + { + "grad_norm": 0.3173106908798218, + "learning_rate": 9.279958810095317e-05, + "loss": 0.0633, + "step": 6430 + }, + { + "grad_norm": 0.31924968957901, + "learning_rate": 9.277106788826494e-05, + "loss": 0.0677, + "step": 6440 + }, + { + "grad_norm": 0.3618471622467041, + "learning_rate": 9.274249570469618e-05, + "loss": 0.0585, + "step": 6450 + }, + { + "grad_norm": 0.3017713129520416, + "learning_rate": 9.271387158496476e-05, + "loss": 0.0611, + "step": 6460 + }, + { + "grad_norm": 0.27444908022880554, + "learning_rate": 9.268519556385173e-05, + "loss": 0.0649, + "step": 6470 + }, + { + "grad_norm": 0.2426140308380127, + "learning_rate": 9.265646767620113e-05, + "loss": 0.0539, + "step": 6480 + }, + { + "grad_norm": 0.3995105028152466, + "learning_rate": 9.262768795692006e-05, + "loss": 0.0729, + "step": 6490 + }, + { + "grad_norm": 0.33021706342697144, + "learning_rate": 9.259885644097861e-05, + "loss": 0.0519, + "step": 6500 + }, + { + "grad_norm": 0.38711899518966675, + "learning_rate": 9.256997316340976e-05, + "loss": 0.0534, + "step": 6510 + }, + { + "grad_norm": 0.3509060740470886, + "learning_rate": 9.254103815930943e-05, + "loss": 0.061, + "step": 6520 + }, + { + "grad_norm": 0.2589935064315796, + "learning_rate": 9.251205146383636e-05, + "loss": 0.0583, + "step": 6530 + }, + { + "grad_norm": 0.2807754576206207, + "learning_rate": 9.248301311221216e-05, + "loss": 0.0724, + "step": 6540 + }, + { + "grad_norm": 0.3476012647151947, + "learning_rate": 9.245392313972115e-05, + "loss": 0.0537, + "step": 6550 + }, + { + "grad_norm": 0.2692490816116333, + "learning_rate": 9.242478158171037e-05, + "loss": 0.0651, + "step": 6560 + }, + { + "grad_norm": 0.4631505608558655, + "learning_rate": 9.239558847358957e-05, + "loss": 0.0736, + "step": 6570 + }, + { + "grad_norm": 0.2911108434200287, + "learning_rate": 9.236634385083114e-05, + "loss": 0.0576, + "step": 6580 + }, + { + "grad_norm": 0.4063659608364105, + "learning_rate": 9.233704774897005e-05, + "loss": 0.0807, + "step": 6590 + }, + { + "grad_norm": 0.3191191256046295, + "learning_rate": 9.230770020360382e-05, + "loss": 0.0775, + "step": 6600 + }, + { + "grad_norm": 0.2895166575908661, + "learning_rate": 9.227830125039247e-05, + "loss": 0.0553, + "step": 6610 + }, + { + "grad_norm": 0.23409207165241241, + "learning_rate": 9.224885092505853e-05, + "loss": 0.052, + "step": 6620 + }, + { + "grad_norm": 0.30995839834213257, + "learning_rate": 9.221934926338691e-05, + "loss": 0.06, + "step": 6630 + }, + { + "grad_norm": 0.2553870975971222, + "learning_rate": 9.21897963012249e-05, + "loss": 0.0483, + "step": 6640 + }, + { + "grad_norm": 0.2702620029449463, + "learning_rate": 9.216019207448217e-05, + "loss": 0.074, + "step": 6650 + }, + { + "grad_norm": 0.29379966855049133, + "learning_rate": 9.21305366191306e-05, + "loss": 0.0529, + "step": 6660 + }, + { + "grad_norm": 0.32158833742141724, + "learning_rate": 9.21008299712044e-05, + "loss": 0.0521, + "step": 6670 + }, + { + "grad_norm": 0.2277429848909378, + "learning_rate": 9.207107216679993e-05, + "loss": 0.0636, + "step": 6680 + }, + { + "grad_norm": 0.2905857563018799, + "learning_rate": 9.204126324207574e-05, + "loss": 0.0621, + "step": 6690 + }, + { + "grad_norm": 0.3196415305137634, + "learning_rate": 9.201140323325248e-05, + "loss": 0.0617, + "step": 6700 + }, + { + "grad_norm": 0.2943153977394104, + "learning_rate": 9.198149217661287e-05, + "loss": 0.0625, + "step": 6710 + }, + { + "grad_norm": 0.2521283030509949, + "learning_rate": 9.195153010850167e-05, + "loss": 0.0688, + "step": 6720 + }, + { + "grad_norm": 0.28602156043052673, + "learning_rate": 9.192151706532562e-05, + "loss": 0.0618, + "step": 6730 + }, + { + "grad_norm": 0.2941585183143616, + "learning_rate": 9.189145308355337e-05, + "loss": 0.0855, + "step": 6740 + }, + { + "grad_norm": 0.28725162148475647, + "learning_rate": 9.186133819971555e-05, + "loss": 0.0534, + "step": 6750 + }, + { + "grad_norm": 0.2897113263607025, + "learning_rate": 9.183117245040455e-05, + "loss": 0.0576, + "step": 6760 + }, + { + "grad_norm": 0.32031962275505066, + "learning_rate": 9.180095587227458e-05, + "loss": 0.0506, + "step": 6770 + }, + { + "grad_norm": 0.3022593557834625, + "learning_rate": 9.177068850204167e-05, + "loss": 0.065, + "step": 6780 + }, + { + "grad_norm": 0.3544381260871887, + "learning_rate": 9.174037037648351e-05, + "loss": 0.0564, + "step": 6790 + }, + { + "grad_norm": 0.29177233576774597, + "learning_rate": 9.171000153243947e-05, + "loss": 0.0706, + "step": 6800 + }, + { + "grad_norm": 0.29167279601097107, + "learning_rate": 9.167958200681057e-05, + "loss": 0.049, + "step": 6810 + }, + { + "grad_norm": 0.3325619399547577, + "learning_rate": 9.164911183655943e-05, + "loss": 0.0682, + "step": 6820 + }, + { + "grad_norm": 0.28959158062934875, + "learning_rate": 9.161859105871013e-05, + "loss": 0.0648, + "step": 6830 + }, + { + "grad_norm": 0.36841556429862976, + "learning_rate": 9.158801971034832e-05, + "loss": 0.0677, + "step": 6840 + }, + { + "grad_norm": 0.2530005872249603, + "learning_rate": 9.155739782862107e-05, + "loss": 0.0625, + "step": 6850 + }, + { + "grad_norm": 0.31044507026672363, + "learning_rate": 9.152672545073687e-05, + "loss": 0.0535, + "step": 6860 + }, + { + "grad_norm": 0.30744463205337524, + "learning_rate": 9.149600261396551e-05, + "loss": 0.0863, + "step": 6870 + }, + { + "grad_norm": 0.2991034984588623, + "learning_rate": 9.146522935563817e-05, + "loss": 0.0728, + "step": 6880 + }, + { + "grad_norm": 0.30861696600914, + "learning_rate": 9.143440571314723e-05, + "loss": 0.0593, + "step": 6890 + }, + { + "grad_norm": 0.3867635130882263, + "learning_rate": 9.140353172394636e-05, + "loss": 0.0697, + "step": 6900 + }, + { + "grad_norm": 0.3190160095691681, + "learning_rate": 9.137260742555032e-05, + "loss": 0.0562, + "step": 6910 + }, + { + "grad_norm": 0.3696102201938629, + "learning_rate": 9.13416328555351e-05, + "loss": 0.0707, + "step": 6920 + }, + { + "grad_norm": 0.3417053520679474, + "learning_rate": 9.131060805153769e-05, + "loss": 0.0716, + "step": 6930 + }, + { + "grad_norm": 0.3282773494720459, + "learning_rate": 9.127953305125618e-05, + "loss": 0.0553, + "step": 6940 + }, + { + "grad_norm": 0.37075749039649963, + "learning_rate": 9.124840789244958e-05, + "loss": 0.0564, + "step": 6950 + }, + { + "grad_norm": 0.236242413520813, + "learning_rate": 9.121723261293794e-05, + "loss": 0.0622, + "step": 6960 + }, + { + "grad_norm": 0.3080188035964966, + "learning_rate": 9.118600725060214e-05, + "loss": 0.0634, + "step": 6970 + }, + { + "grad_norm": 0.28138449788093567, + "learning_rate": 9.115473184338392e-05, + "loss": 0.0693, + "step": 6980 + }, + { + "grad_norm": 0.21389149129390717, + "learning_rate": 9.11234064292859e-05, + "loss": 0.0526, + "step": 6990 + }, + { + "grad_norm": 0.35704877972602844, + "learning_rate": 9.109203104637137e-05, + "loss": 0.0671, + "step": 7000 + }, + { + "grad_norm": 0.3446478545665741, + "learning_rate": 9.10606057327644e-05, + "loss": 0.0685, + "step": 7010 + }, + { + "grad_norm": 0.3182772397994995, + "learning_rate": 9.10291305266497e-05, + "loss": 0.0585, + "step": 7020 + }, + { + "grad_norm": 0.3335762619972229, + "learning_rate": 9.099760546627261e-05, + "loss": 0.0833, + "step": 7030 + }, + { + "grad_norm": 0.2443947046995163, + "learning_rate": 9.096603058993907e-05, + "loss": 0.0555, + "step": 7040 + }, + { + "grad_norm": 0.3097827434539795, + "learning_rate": 9.093440593601552e-05, + "loss": 0.053, + "step": 7050 + }, + { + "grad_norm": 0.28366947174072266, + "learning_rate": 9.090273154292889e-05, + "loss": 0.0532, + "step": 7060 + }, + { + "grad_norm": 0.27883103489875793, + "learning_rate": 9.087100744916656e-05, + "loss": 0.0628, + "step": 7070 + }, + { + "grad_norm": 0.33641988039016724, + "learning_rate": 9.08392336932763e-05, + "loss": 0.0773, + "step": 7080 + }, + { + "grad_norm": 0.3267173171043396, + "learning_rate": 9.080741031386619e-05, + "loss": 0.068, + "step": 7090 + }, + { + "grad_norm": 0.250838041305542, + "learning_rate": 9.077553734960468e-05, + "loss": 0.0512, + "step": 7100 + }, + { + "grad_norm": 0.29570168256759644, + "learning_rate": 9.074361483922041e-05, + "loss": 0.0666, + "step": 7110 + }, + { + "grad_norm": 0.33114975690841675, + "learning_rate": 9.071164282150224e-05, + "loss": 0.0518, + "step": 7120 + }, + { + "grad_norm": 0.26472166180610657, + "learning_rate": 9.067962133529919e-05, + "loss": 0.0541, + "step": 7130 + }, + { + "grad_norm": 0.26712125539779663, + "learning_rate": 9.064755041952036e-05, + "loss": 0.0736, + "step": 7140 + }, + { + "grad_norm": 0.34713247418403625, + "learning_rate": 9.061543011313497e-05, + "loss": 0.0754, + "step": 7150 + }, + { + "grad_norm": 0.3616321086883545, + "learning_rate": 9.05832604551722e-05, + "loss": 0.0699, + "step": 7160 + }, + { + "grad_norm": 0.32782942056655884, + "learning_rate": 9.055104148472124e-05, + "loss": 0.0681, + "step": 7170 + }, + { + "grad_norm": 0.3218350112438202, + "learning_rate": 9.051877324093114e-05, + "loss": 0.056, + "step": 7180 + }, + { + "grad_norm": 0.26620590686798096, + "learning_rate": 9.04864557630109e-05, + "loss": 0.0652, + "step": 7190 + }, + { + "grad_norm": 0.34265294671058655, + "learning_rate": 9.045408909022928e-05, + "loss": 0.058, + "step": 7200 + }, + { + "grad_norm": 0.2833801209926605, + "learning_rate": 9.042167326191484e-05, + "loss": 0.0665, + "step": 7210 + }, + { + "grad_norm": 0.28618761897087097, + "learning_rate": 9.038920831745587e-05, + "loss": 0.063, + "step": 7220 + }, + { + "grad_norm": 0.2561211884021759, + "learning_rate": 9.035669429630036e-05, + "loss": 0.0561, + "step": 7230 + }, + { + "grad_norm": 0.24540375173091888, + "learning_rate": 9.032413123795588e-05, + "loss": 0.0628, + "step": 7240 + }, + { + "grad_norm": 0.3402605652809143, + "learning_rate": 9.029151918198962e-05, + "loss": 0.068, + "step": 7250 + }, + { + "grad_norm": 0.26309555768966675, + "learning_rate": 9.025885816802833e-05, + "loss": 0.0539, + "step": 7260 + }, + { + "grad_norm": 0.3143633008003235, + "learning_rate": 9.02261482357582e-05, + "loss": 0.0495, + "step": 7270 + }, + { + "grad_norm": 0.30007120966911316, + "learning_rate": 9.019338942492485e-05, + "loss": 0.0509, + "step": 7280 + }, + { + "grad_norm": 0.33782634139060974, + "learning_rate": 9.01605817753334e-05, + "loss": 0.0583, + "step": 7290 + }, + { + "grad_norm": 0.4020177125930786, + "learning_rate": 9.012772532684818e-05, + "loss": 0.056, + "step": 7300 + }, + { + "grad_norm": 0.357867032289505, + "learning_rate": 9.00948201193929e-05, + "loss": 0.0595, + "step": 7310 + }, + { + "grad_norm": 0.3049239218235016, + "learning_rate": 9.006186619295047e-05, + "loss": 0.0478, + "step": 7320 + }, + { + "grad_norm": 0.2606947422027588, + "learning_rate": 9.002886358756305e-05, + "loss": 0.0611, + "step": 7330 + }, + { + "grad_norm": 0.30634990334510803, + "learning_rate": 8.999581234333188e-05, + "loss": 0.0599, + "step": 7340 + }, + { + "grad_norm": 0.29688361287117004, + "learning_rate": 8.996271250041734e-05, + "loss": 0.0505, + "step": 7350 + }, + { + "grad_norm": 0.31640371680259705, + "learning_rate": 8.99295640990389e-05, + "loss": 0.0548, + "step": 7360 + }, + { + "grad_norm": 0.21952670812606812, + "learning_rate": 8.989636717947494e-05, + "loss": 0.0614, + "step": 7370 + }, + { + "grad_norm": 0.21746519207954407, + "learning_rate": 8.986312178206291e-05, + "loss": 0.0473, + "step": 7380 + }, + { + "grad_norm": 0.23434697091579437, + "learning_rate": 8.982982794719904e-05, + "loss": 0.0602, + "step": 7390 + }, + { + "grad_norm": 0.25415071845054626, + "learning_rate": 8.979648571533852e-05, + "loss": 0.0587, + "step": 7400 + }, + { + "grad_norm": 0.3684408366680145, + "learning_rate": 8.976309512699528e-05, + "loss": 0.061, + "step": 7410 + }, + { + "grad_norm": 0.253826767206192, + "learning_rate": 8.972965622274205e-05, + "loss": 0.0685, + "step": 7420 + }, + { + "grad_norm": 0.29526665806770325, + "learning_rate": 8.969616904321026e-05, + "loss": 0.0674, + "step": 7430 + }, + { + "grad_norm": 0.35492202639579773, + "learning_rate": 8.966263362908998e-05, + "loss": 0.0612, + "step": 7440 + }, + { + "grad_norm": 0.34254756569862366, + "learning_rate": 8.962905002112989e-05, + "loss": 0.0528, + "step": 7450 + }, + { + "grad_norm": 0.23975923657417297, + "learning_rate": 8.959541826013724e-05, + "loss": 0.0637, + "step": 7460 + }, + { + "grad_norm": 0.2604082524776459, + "learning_rate": 8.956173838697781e-05, + "loss": 0.0635, + "step": 7470 + }, + { + "grad_norm": 0.31560009717941284, + "learning_rate": 8.952801044257581e-05, + "loss": 0.0698, + "step": 7480 + }, + { + "grad_norm": 0.3092935383319855, + "learning_rate": 8.949423446791387e-05, + "loss": 0.0574, + "step": 7490 + }, + { + "grad_norm": 0.2915799021720886, + "learning_rate": 8.9460410504033e-05, + "loss": 0.0525, + "step": 7500 + }, + { + "grad_norm": 0.29869532585144043, + "learning_rate": 8.942653859203248e-05, + "loss": 0.0482, + "step": 7510 + }, + { + "grad_norm": 0.369875431060791, + "learning_rate": 8.93926187730699e-05, + "loss": 0.047, + "step": 7520 + }, + { + "grad_norm": 0.24332013726234436, + "learning_rate": 8.935865108836102e-05, + "loss": 0.0496, + "step": 7530 + }, + { + "grad_norm": 0.26538342237472534, + "learning_rate": 8.932463557917981e-05, + "loss": 0.0546, + "step": 7540 + }, + { + "grad_norm": 0.29162678122520447, + "learning_rate": 8.929057228685829e-05, + "loss": 0.0541, + "step": 7550 + }, + { + "grad_norm": 0.28678497672080994, + "learning_rate": 8.925646125278658e-05, + "loss": 0.0516, + "step": 7560 + }, + { + "grad_norm": 0.35382333397865295, + "learning_rate": 8.92223025184128e-05, + "loss": 0.063, + "step": 7570 + }, + { + "grad_norm": 0.2611851692199707, + "learning_rate": 8.918809612524303e-05, + "loss": 0.0614, + "step": 7580 + }, + { + "grad_norm": 0.248627707362175, + "learning_rate": 8.91538421148413e-05, + "loss": 0.0621, + "step": 7590 + }, + { + "grad_norm": 0.3188674747943878, + "learning_rate": 8.91195405288294e-05, + "loss": 0.0727, + "step": 7600 + }, + { + "grad_norm": 0.2575676441192627, + "learning_rate": 8.908519140888705e-05, + "loss": 0.0581, + "step": 7610 + }, + { + "grad_norm": 0.35355809330940247, + "learning_rate": 8.905079479675164e-05, + "loss": 0.0564, + "step": 7620 + }, + { + "grad_norm": 0.28264912962913513, + "learning_rate": 8.901635073421831e-05, + "loss": 0.0605, + "step": 7630 + }, + { + "grad_norm": 0.27571675181388855, + "learning_rate": 8.898185926313983e-05, + "loss": 0.0618, + "step": 7640 + }, + { + "grad_norm": 0.3539575934410095, + "learning_rate": 8.894732042542659e-05, + "loss": 0.053, + "step": 7650 + }, + { + "grad_norm": 0.32171306014060974, + "learning_rate": 8.891273426304654e-05, + "loss": 0.0523, + "step": 7660 + }, + { + "grad_norm": 0.2886122763156891, + "learning_rate": 8.887810081802515e-05, + "loss": 0.0633, + "step": 7670 + }, + { + "grad_norm": 0.3257535994052887, + "learning_rate": 8.884342013244529e-05, + "loss": 0.0629, + "step": 7680 + }, + { + "grad_norm": 0.31563088297843933, + "learning_rate": 8.880869224844725e-05, + "loss": 0.0561, + "step": 7690 + }, + { + "grad_norm": 0.3266989588737488, + "learning_rate": 8.877391720822875e-05, + "loss": 0.0503, + "step": 7700 + }, + { + "grad_norm": 0.333683580160141, + "learning_rate": 8.873909505404468e-05, + "loss": 0.0697, + "step": 7710 + }, + { + "grad_norm": 0.3123229146003723, + "learning_rate": 8.870422582820726e-05, + "loss": 0.0612, + "step": 7720 + }, + { + "grad_norm": 0.29147085547447205, + "learning_rate": 8.866930957308588e-05, + "loss": 0.058, + "step": 7730 + }, + { + "grad_norm": 0.2849343717098236, + "learning_rate": 8.86343463311071e-05, + "loss": 0.0599, + "step": 7740 + }, + { + "grad_norm": 0.26260074973106384, + "learning_rate": 8.859933614475452e-05, + "loss": 0.048, + "step": 7750 + }, + { + "grad_norm": 0.27800849080085754, + "learning_rate": 8.85642790565689e-05, + "loss": 0.0596, + "step": 7760 + }, + { + "grad_norm": 0.2939211428165436, + "learning_rate": 8.852917510914782e-05, + "loss": 0.061, + "step": 7770 + }, + { + "grad_norm": 0.22196416556835175, + "learning_rate": 8.84940243451459e-05, + "loss": 0.0618, + "step": 7780 + }, + { + "grad_norm": 0.28265464305877686, + "learning_rate": 8.845882680727469e-05, + "loss": 0.0546, + "step": 7790 + }, + { + "grad_norm": 0.24199794232845306, + "learning_rate": 8.842358253830246e-05, + "loss": 0.051, + "step": 7800 + }, + { + "grad_norm": 0.2887401580810547, + "learning_rate": 8.838829158105434e-05, + "loss": 0.0557, + "step": 7810 + }, + { + "grad_norm": 0.22729307413101196, + "learning_rate": 8.835295397841217e-05, + "loss": 0.0444, + "step": 7820 + }, + { + "grad_norm": 0.2723986506462097, + "learning_rate": 8.831756977331447e-05, + "loss": 0.0497, + "step": 7830 + }, + { + "grad_norm": 0.29254522919654846, + "learning_rate": 8.828213900875638e-05, + "loss": 0.044, + "step": 7840 + }, + { + "grad_norm": 0.232318714261055, + "learning_rate": 8.824666172778964e-05, + "loss": 0.0487, + "step": 7850 + }, + { + "grad_norm": 0.3090560734272003, + "learning_rate": 8.821113797352245e-05, + "loss": 0.0645, + "step": 7860 + }, + { + "grad_norm": 0.27316391468048096, + "learning_rate": 8.817556778911956e-05, + "loss": 0.0659, + "step": 7870 + }, + { + "grad_norm": 0.33472269773483276, + "learning_rate": 8.81399512178021e-05, + "loss": 0.062, + "step": 7880 + }, + { + "grad_norm": 0.27600476145744324, + "learning_rate": 8.810428830284752e-05, + "loss": 0.0518, + "step": 7890 + }, + { + "grad_norm": 0.292817622423172, + "learning_rate": 8.806857908758967e-05, + "loss": 0.0693, + "step": 7900 + }, + { + "grad_norm": 0.2777009904384613, + "learning_rate": 8.80328236154186e-05, + "loss": 0.0642, + "step": 7910 + }, + { + "grad_norm": 0.37165528535842896, + "learning_rate": 8.799702192978056e-05, + "loss": 0.0854, + "step": 7920 + }, + { + "grad_norm": 0.2624307870864868, + "learning_rate": 8.7961174074178e-05, + "loss": 0.0556, + "step": 7930 + }, + { + "grad_norm": 0.30395254492759705, + "learning_rate": 8.792528009216942e-05, + "loss": 0.063, + "step": 7940 + }, + { + "grad_norm": 0.39645254611968994, + "learning_rate": 8.788934002736943e-05, + "loss": 0.0542, + "step": 7950 + }, + { + "grad_norm": 0.27897000312805176, + "learning_rate": 8.785335392344857e-05, + "loss": 0.0532, + "step": 7960 + }, + { + "grad_norm": 0.299923300743103, + "learning_rate": 8.781732182413336e-05, + "loss": 0.0517, + "step": 7970 + }, + { + "grad_norm": 0.30059191584587097, + "learning_rate": 8.778124377320618e-05, + "loss": 0.0532, + "step": 7980 + }, + { + "grad_norm": 0.32196223735809326, + "learning_rate": 8.774511981450529e-05, + "loss": 0.0472, + "step": 7990 + }, + { + "grad_norm": 0.3021567165851593, + "learning_rate": 8.770894999192467e-05, + "loss": 0.0473, + "step": 8000 + }, + { + "grad_norm": 0.28544217348098755, + "learning_rate": 8.767273434941412e-05, + "loss": 0.0494, + "step": 8010 + }, + { + "grad_norm": 0.3216398358345032, + "learning_rate": 8.7636472930979e-05, + "loss": 0.0563, + "step": 8020 + }, + { + "grad_norm": 0.2400844395160675, + "learning_rate": 8.76001657806804e-05, + "loss": 0.0494, + "step": 8030 + }, + { + "grad_norm": 0.28355246782302856, + "learning_rate": 8.756381294263491e-05, + "loss": 0.0483, + "step": 8040 + }, + { + "grad_norm": 0.28866615891456604, + "learning_rate": 8.752741446101464e-05, + "loss": 0.0564, + "step": 8050 + }, + { + "grad_norm": 0.24818548560142517, + "learning_rate": 8.749097038004722e-05, + "loss": 0.0444, + "step": 8060 + }, + { + "grad_norm": 0.27400869131088257, + "learning_rate": 8.745448074401561e-05, + "loss": 0.0488, + "step": 8070 + }, + { + "grad_norm": 0.27506697177886963, + "learning_rate": 8.741794559725818e-05, + "loss": 0.0683, + "step": 8080 + }, + { + "grad_norm": 0.2994530200958252, + "learning_rate": 8.738136498416856e-05, + "loss": 0.0567, + "step": 8090 + }, + { + "grad_norm": 0.27273520827293396, + "learning_rate": 8.734473894919564e-05, + "loss": 0.0627, + "step": 8100 + }, + { + "grad_norm": 0.35509568452835083, + "learning_rate": 8.730806753684353e-05, + "loss": 0.0608, + "step": 8110 + }, + { + "grad_norm": 0.21389254927635193, + "learning_rate": 8.727135079167144e-05, + "loss": 0.0546, + "step": 8120 + }, + { + "grad_norm": 0.28771549463272095, + "learning_rate": 8.723458875829368e-05, + "loss": 0.0663, + "step": 8130 + }, + { + "grad_norm": 0.2096043974161148, + "learning_rate": 8.719778148137958e-05, + "loss": 0.0602, + "step": 8140 + }, + { + "grad_norm": 0.264454185962677, + "learning_rate": 8.716092900565346e-05, + "loss": 0.0495, + "step": 8150 + }, + { + "grad_norm": 0.3526304066181183, + "learning_rate": 8.712403137589456e-05, + "loss": 0.0676, + "step": 8160 + }, + { + "grad_norm": 0.2823895812034607, + "learning_rate": 8.708708863693697e-05, + "loss": 0.0569, + "step": 8170 + }, + { + "grad_norm": 0.24215340614318848, + "learning_rate": 8.70501008336696e-05, + "loss": 0.0516, + "step": 8180 + }, + { + "grad_norm": 0.30370357632637024, + "learning_rate": 8.701306801103611e-05, + "loss": 0.0547, + "step": 8190 + }, + { + "grad_norm": 0.2672850787639618, + "learning_rate": 8.697599021403489e-05, + "loss": 0.05, + "step": 8200 + }, + { + "grad_norm": 0.24326245486736298, + "learning_rate": 8.693886748771896e-05, + "loss": 0.0415, + "step": 8210 + }, + { + "grad_norm": 0.21991650760173798, + "learning_rate": 8.690169987719593e-05, + "loss": 0.0454, + "step": 8220 + }, + { + "grad_norm": 0.28604015707969666, + "learning_rate": 8.686448742762791e-05, + "loss": 0.0494, + "step": 8230 + }, + { + "grad_norm": 0.24010229110717773, + "learning_rate": 8.682723018423161e-05, + "loss": 0.0535, + "step": 8240 + }, + { + "grad_norm": 0.3204074800014496, + "learning_rate": 8.678992819227804e-05, + "loss": 0.0486, + "step": 8250 + }, + { + "grad_norm": 0.26610270142555237, + "learning_rate": 8.675258149709265e-05, + "loss": 0.0409, + "step": 8260 + }, + { + "grad_norm": 0.2511579692363739, + "learning_rate": 8.67151901440552e-05, + "loss": 0.0431, + "step": 8270 + }, + { + "grad_norm": 0.3090936839580536, + "learning_rate": 8.667775417859971e-05, + "loss": 0.0599, + "step": 8280 + }, + { + "grad_norm": 0.310330331325531, + "learning_rate": 8.664027364621441e-05, + "loss": 0.0439, + "step": 8290 + }, + { + "grad_norm": 0.3341936469078064, + "learning_rate": 8.660274859244168e-05, + "loss": 0.0596, + "step": 8300 + }, + { + "grad_norm": 0.3190235197544098, + "learning_rate": 8.656517906287799e-05, + "loss": 0.0485, + "step": 8310 + }, + { + "grad_norm": 0.25204184651374817, + "learning_rate": 8.652756510317387e-05, + "loss": 0.0416, + "step": 8320 + }, + { + "grad_norm": 0.30357614159584045, + "learning_rate": 8.648990675903382e-05, + "loss": 0.0529, + "step": 8330 + }, + { + "grad_norm": 0.29918432235717773, + "learning_rate": 8.64522040762163e-05, + "loss": 0.0429, + "step": 8340 + }, + { + "grad_norm": 0.2640821933746338, + "learning_rate": 8.64144571005336e-05, + "loss": 0.0614, + "step": 8350 + }, + { + "grad_norm": 0.280055969953537, + "learning_rate": 8.637666587785184e-05, + "loss": 0.0516, + "step": 8360 + }, + { + "grad_norm": 0.290903240442276, + "learning_rate": 8.633883045409096e-05, + "loss": 0.0452, + "step": 8370 + }, + { + "grad_norm": 0.2710651755332947, + "learning_rate": 8.630095087522457e-05, + "loss": 0.0535, + "step": 8380 + }, + { + "grad_norm": 0.2425202876329422, + "learning_rate": 8.62630271872799e-05, + "loss": 0.0469, + "step": 8390 + }, + { + "grad_norm": 0.24323712289333344, + "learning_rate": 8.62250594363378e-05, + "loss": 0.0484, + "step": 8400 + }, + { + "grad_norm": 0.29924124479293823, + "learning_rate": 8.61870476685327e-05, + "loss": 0.054, + "step": 8410 + }, + { + "grad_norm": 0.22607800364494324, + "learning_rate": 8.614899193005248e-05, + "loss": 0.0446, + "step": 8420 + }, + { + "grad_norm": 0.3191208839416504, + "learning_rate": 8.611089226713842e-05, + "loss": 0.0509, + "step": 8430 + }, + { + "grad_norm": 0.32413366436958313, + "learning_rate": 8.607274872608521e-05, + "loss": 0.0502, + "step": 8440 + }, + { + "grad_norm": 0.28586545586586, + "learning_rate": 8.603456135324088e-05, + "loss": 0.0464, + "step": 8450 + }, + { + "grad_norm": 0.23136994242668152, + "learning_rate": 8.599633019500665e-05, + "loss": 0.0468, + "step": 8460 + }, + { + "grad_norm": 0.25228628516197205, + "learning_rate": 8.595805529783702e-05, + "loss": 0.044, + "step": 8470 + }, + { + "grad_norm": 0.2524213492870331, + "learning_rate": 8.591973670823958e-05, + "loss": 0.0459, + "step": 8480 + }, + { + "grad_norm": 0.29968252778053284, + "learning_rate": 8.588137447277503e-05, + "loss": 0.0531, + "step": 8490 + }, + { + "grad_norm": 0.3116902709007263, + "learning_rate": 8.584296863805709e-05, + "loss": 0.0563, + "step": 8500 + }, + { + "grad_norm": 0.3455532491207123, + "learning_rate": 8.580451925075248e-05, + "loss": 0.0496, + "step": 8510 + }, + { + "grad_norm": 0.2749980390071869, + "learning_rate": 8.576602635758085e-05, + "loss": 0.0364, + "step": 8520 + }, + { + "grad_norm": 0.22466090321540833, + "learning_rate": 8.572749000531469e-05, + "loss": 0.0461, + "step": 8530 + }, + { + "grad_norm": 0.23747560381889343, + "learning_rate": 8.568891024077926e-05, + "loss": 0.0495, + "step": 8540 + }, + { + "grad_norm": 0.31052207946777344, + "learning_rate": 8.565028711085265e-05, + "loss": 0.0525, + "step": 8550 + }, + { + "grad_norm": 0.27240148186683655, + "learning_rate": 8.561162066246561e-05, + "loss": 0.0512, + "step": 8560 + }, + { + "grad_norm": 0.2537265121936798, + "learning_rate": 8.557291094260152e-05, + "loss": 0.0775, + "step": 8570 + }, + { + "grad_norm": 0.3093377947807312, + "learning_rate": 8.553415799829629e-05, + "loss": 0.0597, + "step": 8580 + }, + { + "grad_norm": 0.26136863231658936, + "learning_rate": 8.549536187663847e-05, + "loss": 0.0532, + "step": 8590 + }, + { + "grad_norm": 0.22629521787166595, + "learning_rate": 8.545652262476897e-05, + "loss": 0.0402, + "step": 8600 + }, + { + "grad_norm": 0.3351418972015381, + "learning_rate": 8.541764028988115e-05, + "loss": 0.0461, + "step": 8610 + }, + { + "grad_norm": 0.29206499457359314, + "learning_rate": 8.537871491922071e-05, + "loss": 0.0562, + "step": 8620 + }, + { + "grad_norm": 0.26203903555870056, + "learning_rate": 8.533974656008566e-05, + "loss": 0.0454, + "step": 8630 + }, + { + "grad_norm": 0.3431650698184967, + "learning_rate": 8.530073525982621e-05, + "loss": 0.0516, + "step": 8640 + }, + { + "grad_norm": 0.2787887156009674, + "learning_rate": 8.526168106584476e-05, + "loss": 0.0528, + "step": 8650 + }, + { + "grad_norm": 0.23733307421207428, + "learning_rate": 8.522258402559586e-05, + "loss": 0.0478, + "step": 8660 + }, + { + "grad_norm": 0.30951014161109924, + "learning_rate": 8.518344418658613e-05, + "loss": 0.0468, + "step": 8670 + }, + { + "grad_norm": 0.25972819328308105, + "learning_rate": 8.51442615963741e-05, + "loss": 0.0407, + "step": 8680 + }, + { + "grad_norm": 0.27463382482528687, + "learning_rate": 8.510503630257034e-05, + "loss": 0.0454, + "step": 8690 + }, + { + "grad_norm": 0.31636977195739746, + "learning_rate": 8.50657683528373e-05, + "loss": 0.0461, + "step": 8700 + }, + { + "grad_norm": 0.2419833391904831, + "learning_rate": 8.502645779488922e-05, + "loss": 0.0546, + "step": 8710 + }, + { + "grad_norm": 0.3163353204727173, + "learning_rate": 8.498710467649213e-05, + "loss": 0.0693, + "step": 8720 + }, + { + "grad_norm": 0.232834130525589, + "learning_rate": 8.494770904546381e-05, + "loss": 0.0475, + "step": 8730 + }, + { + "grad_norm": 0.24939203262329102, + "learning_rate": 8.490827094967363e-05, + "loss": 0.0459, + "step": 8740 + }, + { + "grad_norm": 0.25879377126693726, + "learning_rate": 8.486879043704263e-05, + "loss": 0.0556, + "step": 8750 + }, + { + "grad_norm": 0.2087334543466568, + "learning_rate": 8.482926755554332e-05, + "loss": 0.0442, + "step": 8760 + }, + { + "grad_norm": 0.3335881233215332, + "learning_rate": 8.478970235319975e-05, + "loss": 0.0517, + "step": 8770 + }, + { + "grad_norm": 0.26189175248146057, + "learning_rate": 8.475009487808737e-05, + "loss": 0.0522, + "step": 8780 + }, + { + "grad_norm": 0.2205239236354828, + "learning_rate": 8.471044517833298e-05, + "loss": 0.0415, + "step": 8790 + }, + { + "grad_norm": 0.272456556558609, + "learning_rate": 8.467075330211474e-05, + "loss": 0.0386, + "step": 8800 + }, + { + "grad_norm": 0.2636466920375824, + "learning_rate": 8.463101929766198e-05, + "loss": 0.0505, + "step": 8810 + }, + { + "grad_norm": 0.2145378589630127, + "learning_rate": 8.459124321325528e-05, + "loss": 0.0447, + "step": 8820 + }, + { + "grad_norm": 0.22128890454769135, + "learning_rate": 8.455142509722634e-05, + "loss": 0.0465, + "step": 8830 + }, + { + "grad_norm": 0.2453867346048355, + "learning_rate": 8.451156499795792e-05, + "loss": 0.0453, + "step": 8840 + }, + { + "grad_norm": 0.2712414860725403, + "learning_rate": 8.447166296388381e-05, + "loss": 0.0397, + "step": 8850 + }, + { + "grad_norm": 0.26136285066604614, + "learning_rate": 8.443171904348872e-05, + "loss": 0.0414, + "step": 8860 + }, + { + "grad_norm": 0.29648956656455994, + "learning_rate": 8.43917332853083e-05, + "loss": 0.0467, + "step": 8870 + }, + { + "grad_norm": 0.2118912935256958, + "learning_rate": 8.435170573792901e-05, + "loss": 0.0476, + "step": 8880 + }, + { + "grad_norm": 0.30057474970817566, + "learning_rate": 8.431163644998808e-05, + "loss": 0.0508, + "step": 8890 + }, + { + "grad_norm": 0.2780839502811432, + "learning_rate": 8.42715254701735e-05, + "loss": 0.0476, + "step": 8900 + }, + { + "grad_norm": 0.28566572070121765, + "learning_rate": 8.42313728472239e-05, + "loss": 0.0434, + "step": 8910 + }, + { + "grad_norm": 0.36405450105667114, + "learning_rate": 8.419117862992844e-05, + "loss": 0.0573, + "step": 8920 + }, + { + "grad_norm": 0.2788868546485901, + "learning_rate": 8.415094286712694e-05, + "loss": 0.0409, + "step": 8930 + }, + { + "grad_norm": 0.3000909090042114, + "learning_rate": 8.411066560770965e-05, + "loss": 0.0494, + "step": 8940 + }, + { + "grad_norm": 0.2516811788082123, + "learning_rate": 8.407034690061721e-05, + "loss": 0.045, + "step": 8950 + }, + { + "grad_norm": 0.24953649938106537, + "learning_rate": 8.402998679484067e-05, + "loss": 0.0397, + "step": 8960 + }, + { + "grad_norm": 0.24034424126148224, + "learning_rate": 8.398958533942135e-05, + "loss": 0.0376, + "step": 8970 + }, + { + "grad_norm": 0.30077075958251953, + "learning_rate": 8.394914258345083e-05, + "loss": 0.0495, + "step": 8980 + }, + { + "grad_norm": 0.3470993638038635, + "learning_rate": 8.390865857607089e-05, + "loss": 0.069, + "step": 8990 + }, + { + "grad_norm": 0.2332250475883484, + "learning_rate": 8.386813336647339e-05, + "loss": 0.0456, + "step": 9000 + }, + { + "grad_norm": 0.2492009848356247, + "learning_rate": 8.38275670039003e-05, + "loss": 0.0376, + "step": 9010 + }, + { + "grad_norm": 0.2934356927871704, + "learning_rate": 8.378695953764356e-05, + "loss": 0.0456, + "step": 9020 + }, + { + "grad_norm": 0.2467339038848877, + "learning_rate": 8.374631101704509e-05, + "loss": 0.0608, + "step": 9030 + }, + { + "grad_norm": 0.23218746483325958, + "learning_rate": 8.370562149149665e-05, + "loss": 0.0553, + "step": 9040 + }, + { + "grad_norm": 0.2821784019470215, + "learning_rate": 8.366489101043989e-05, + "loss": 0.0452, + "step": 9050 + }, + { + "grad_norm": 0.22763745486736298, + "learning_rate": 8.362411962336614e-05, + "loss": 0.0532, + "step": 9060 + }, + { + "grad_norm": 0.23225706815719604, + "learning_rate": 8.358330737981651e-05, + "loss": 0.0371, + "step": 9070 + }, + { + "grad_norm": 0.21868430078029633, + "learning_rate": 8.35424543293817e-05, + "loss": 0.0442, + "step": 9080 + }, + { + "grad_norm": 0.21281558275222778, + "learning_rate": 8.350156052170206e-05, + "loss": 0.0552, + "step": 9090 + }, + { + "grad_norm": 0.3598162531852722, + "learning_rate": 8.346062600646738e-05, + "loss": 0.0579, + "step": 9100 + }, + { + "grad_norm": 0.2622242271900177, + "learning_rate": 8.341965083341697e-05, + "loss": 0.0475, + "step": 9110 + }, + { + "grad_norm": 0.2924858629703522, + "learning_rate": 8.337863505233953e-05, + "loss": 0.0507, + "step": 9120 + }, + { + "grad_norm": 0.2943890690803528, + "learning_rate": 8.33375787130731e-05, + "loss": 0.0538, + "step": 9130 + }, + { + "grad_norm": 0.30198705196380615, + "learning_rate": 8.3296481865505e-05, + "loss": 0.0393, + "step": 9140 + }, + { + "grad_norm": 0.3284444212913513, + "learning_rate": 8.325534455957181e-05, + "loss": 0.0578, + "step": 9150 + }, + { + "grad_norm": 0.3026992678642273, + "learning_rate": 8.321416684525917e-05, + "loss": 0.0368, + "step": 9160 + }, + { + "grad_norm": 0.24541352689266205, + "learning_rate": 8.317294877260193e-05, + "loss": 0.042, + "step": 9170 + }, + { + "grad_norm": 0.34712836146354675, + "learning_rate": 8.313169039168394e-05, + "loss": 0.0454, + "step": 9180 + }, + { + "grad_norm": 0.22341030836105347, + "learning_rate": 8.3090391752638e-05, + "loss": 0.036, + "step": 9190 + }, + { + "grad_norm": 0.16605167090892792, + "learning_rate": 8.304905290564586e-05, + "loss": 0.0481, + "step": 9200 + }, + { + "grad_norm": 0.24110513925552368, + "learning_rate": 8.300767390093813e-05, + "loss": 0.061, + "step": 9210 + }, + { + "grad_norm": 0.248176708817482, + "learning_rate": 8.296625478879417e-05, + "loss": 0.0574, + "step": 9220 + }, + { + "grad_norm": 0.2800736725330353, + "learning_rate": 8.292479561954214e-05, + "loss": 0.0446, + "step": 9230 + }, + { + "grad_norm": 0.27925363183021545, + "learning_rate": 8.288329644355883e-05, + "loss": 0.0524, + "step": 9240 + }, + { + "grad_norm": 0.1991814225912094, + "learning_rate": 8.284175731126965e-05, + "loss": 0.0381, + "step": 9250 + }, + { + "grad_norm": 0.26170802116394043, + "learning_rate": 8.280017827314854e-05, + "loss": 0.0432, + "step": 9260 + }, + { + "grad_norm": 0.32718122005462646, + "learning_rate": 8.2758559379718e-05, + "loss": 0.0466, + "step": 9270 + }, + { + "grad_norm": 0.2426115870475769, + "learning_rate": 8.271690068154887e-05, + "loss": 0.0453, + "step": 9280 + }, + { + "grad_norm": 0.2564951777458191, + "learning_rate": 8.26752022292604e-05, + "loss": 0.0585, + "step": 9290 + }, + { + "grad_norm": 0.2689344584941864, + "learning_rate": 8.263346407352018e-05, + "loss": 0.0505, + "step": 9300 + }, + { + "grad_norm": 0.30103445053100586, + "learning_rate": 8.259168626504395e-05, + "loss": 0.0454, + "step": 9310 + }, + { + "grad_norm": 0.2145436853170395, + "learning_rate": 8.254986885459572e-05, + "loss": 0.0628, + "step": 9320 + }, + { + "grad_norm": 0.2973059415817261, + "learning_rate": 8.250801189298757e-05, + "loss": 0.045, + "step": 9330 + }, + { + "grad_norm": 0.2533753216266632, + "learning_rate": 8.246611543107967e-05, + "loss": 0.0463, + "step": 9340 + }, + { + "grad_norm": 0.18445678055286407, + "learning_rate": 8.242417951978018e-05, + "loss": 0.042, + "step": 9350 + }, + { + "grad_norm": 0.246917724609375, + "learning_rate": 8.238220421004517e-05, + "loss": 0.0404, + "step": 9360 + }, + { + "grad_norm": 0.19765189290046692, + "learning_rate": 8.234018955287859e-05, + "loss": 0.0551, + "step": 9370 + }, + { + "grad_norm": 0.3127460777759552, + "learning_rate": 8.229813559933224e-05, + "loss": 0.0489, + "step": 9380 + }, + { + "grad_norm": 0.28748321533203125, + "learning_rate": 8.22560424005056e-05, + "loss": 0.0485, + "step": 9390 + }, + { + "grad_norm": 0.3034312427043915, + "learning_rate": 8.22139100075459e-05, + "loss": 0.0563, + "step": 9400 + }, + { + "grad_norm": 0.22079165279865265, + "learning_rate": 8.217173847164799e-05, + "loss": 0.0555, + "step": 9410 + }, + { + "grad_norm": 0.32796576619148254, + "learning_rate": 8.212952784405422e-05, + "loss": 0.0479, + "step": 9420 + }, + { + "grad_norm": 0.3491758704185486, + "learning_rate": 8.208727817605451e-05, + "loss": 0.0455, + "step": 9430 + }, + { + "grad_norm": 0.258953720331192, + "learning_rate": 8.204498951898618e-05, + "loss": 0.0497, + "step": 9440 + }, + { + "grad_norm": 0.2378096729516983, + "learning_rate": 8.200266192423395e-05, + "loss": 0.0437, + "step": 9450 + }, + { + "grad_norm": 0.2638334035873413, + "learning_rate": 8.196029544322981e-05, + "loss": 0.0402, + "step": 9460 + }, + { + "grad_norm": 0.25130248069763184, + "learning_rate": 8.191789012745305e-05, + "loss": 0.063, + "step": 9470 + }, + { + "grad_norm": 0.2027653306722641, + "learning_rate": 8.187544602843014e-05, + "loss": 0.0422, + "step": 9480 + }, + { + "grad_norm": 0.2631757855415344, + "learning_rate": 8.183296319773466e-05, + "loss": 0.0452, + "step": 9490 + }, + { + "grad_norm": 0.21387606859207153, + "learning_rate": 8.179044168698721e-05, + "loss": 0.0376, + "step": 9500 + }, + { + "grad_norm": 0.22583898901939392, + "learning_rate": 8.174788154785548e-05, + "loss": 0.0397, + "step": 9510 + }, + { + "grad_norm": 0.2272488921880722, + "learning_rate": 8.170528283205403e-05, + "loss": 0.0492, + "step": 9520 + }, + { + "grad_norm": 0.21835656464099884, + "learning_rate": 8.166264559134435e-05, + "loss": 0.0486, + "step": 9530 + }, + { + "grad_norm": 0.3125370144844055, + "learning_rate": 8.161996987753467e-05, + "loss": 0.0608, + "step": 9540 + }, + { + "grad_norm": 0.2734662890434265, + "learning_rate": 8.157725574248e-05, + "loss": 0.0501, + "step": 9550 + }, + { + "grad_norm": 0.27803707122802734, + "learning_rate": 8.153450323808209e-05, + "loss": 0.0539, + "step": 9560 + }, + { + "grad_norm": 0.375032514333725, + "learning_rate": 8.149171241628923e-05, + "loss": 0.0643, + "step": 9570 + }, + { + "grad_norm": 0.2398683726787567, + "learning_rate": 8.144888332909631e-05, + "loss": 0.0498, + "step": 9580 + }, + { + "grad_norm": 0.25993067026138306, + "learning_rate": 8.14060160285447e-05, + "loss": 0.0502, + "step": 9590 + }, + { + "grad_norm": 0.32848912477493286, + "learning_rate": 8.136311056672223e-05, + "loss": 0.0408, + "step": 9600 + }, + { + "grad_norm": 0.26502951979637146, + "learning_rate": 8.132016699576308e-05, + "loss": 0.0454, + "step": 9610 + }, + { + "grad_norm": 0.21933794021606445, + "learning_rate": 8.127718536784771e-05, + "loss": 0.0485, + "step": 9620 + }, + { + "grad_norm": 0.3090316951274872, + "learning_rate": 8.12341657352029e-05, + "loss": 0.0442, + "step": 9630 + }, + { + "grad_norm": 0.2781892418861389, + "learning_rate": 8.119110815010151e-05, + "loss": 0.0447, + "step": 9640 + }, + { + "grad_norm": 0.30298709869384766, + "learning_rate": 8.114801266486261e-05, + "loss": 0.0499, + "step": 9650 + }, + { + "grad_norm": 0.22665993869304657, + "learning_rate": 8.110487933185122e-05, + "loss": 0.05, + "step": 9660 + }, + { + "grad_norm": 0.2654498517513275, + "learning_rate": 8.106170820347847e-05, + "loss": 0.0536, + "step": 9670 + }, + { + "grad_norm": 0.2829221487045288, + "learning_rate": 8.101849933220132e-05, + "loss": 0.0375, + "step": 9680 + }, + { + "grad_norm": 0.2343592345714569, + "learning_rate": 8.097525277052264e-05, + "loss": 0.0442, + "step": 9690 + }, + { + "grad_norm": 0.1920456886291504, + "learning_rate": 8.093196857099106e-05, + "loss": 0.0389, + "step": 9700 + }, + { + "grad_norm": 0.3087690472602844, + "learning_rate": 8.088864678620096e-05, + "loss": 0.0426, + "step": 9710 + }, + { + "grad_norm": 0.2967287003993988, + "learning_rate": 8.084528746879243e-05, + "loss": 0.0476, + "step": 9720 + }, + { + "grad_norm": 0.2517305910587311, + "learning_rate": 8.080189067145108e-05, + "loss": 0.0563, + "step": 9730 + }, + { + "grad_norm": 0.2537696957588196, + "learning_rate": 8.075845644690813e-05, + "loss": 0.0429, + "step": 9740 + }, + { + "grad_norm": 0.266324907541275, + "learning_rate": 8.07149848479403e-05, + "loss": 0.0435, + "step": 9750 + }, + { + "grad_norm": 0.24241960048675537, + "learning_rate": 8.067147592736962e-05, + "loss": 0.0482, + "step": 9760 + }, + { + "grad_norm": 0.24011065065860748, + "learning_rate": 8.062792973806357e-05, + "loss": 0.0543, + "step": 9770 + }, + { + "grad_norm": 0.2036685347557068, + "learning_rate": 8.058434633293485e-05, + "loss": 0.0472, + "step": 9780 + }, + { + "grad_norm": 0.2262846827507019, + "learning_rate": 8.054072576494143e-05, + "loss": 0.047, + "step": 9790 + }, + { + "grad_norm": 0.2544848918914795, + "learning_rate": 8.04970680870864e-05, + "loss": 0.0404, + "step": 9800 + }, + { + "grad_norm": 0.29485446214675903, + "learning_rate": 8.045337335241793e-05, + "loss": 0.0475, + "step": 9810 + }, + { + "grad_norm": 0.29331570863723755, + "learning_rate": 8.040964161402932e-05, + "loss": 0.0529, + "step": 9820 + }, + { + "grad_norm": 0.25500836968421936, + "learning_rate": 8.036587292505869e-05, + "loss": 0.0468, + "step": 9830 + }, + { + "grad_norm": 0.18651726841926575, + "learning_rate": 8.032206733868912e-05, + "loss": 0.0558, + "step": 9840 + }, + { + "grad_norm": 0.233231320977211, + "learning_rate": 8.027822490814858e-05, + "loss": 0.055, + "step": 9850 + }, + { + "grad_norm": 0.19117946922779083, + "learning_rate": 8.023434568670972e-05, + "loss": 0.039, + "step": 9860 + }, + { + "grad_norm": 0.24533501267433167, + "learning_rate": 8.019042972768992e-05, + "loss": 0.0388, + "step": 9870 + }, + { + "grad_norm": 0.2981744408607483, + "learning_rate": 8.014647708445124e-05, + "loss": 0.0508, + "step": 9880 + }, + { + "grad_norm": 0.24683181941509247, + "learning_rate": 8.010248781040027e-05, + "loss": 0.0383, + "step": 9890 + }, + { + "grad_norm": 0.2710823118686676, + "learning_rate": 8.005846195898814e-05, + "loss": 0.0493, + "step": 9900 + }, + { + "grad_norm": 0.2593339681625366, + "learning_rate": 8.00143995837104e-05, + "loss": 0.0466, + "step": 9910 + }, + { + "grad_norm": 0.23342756927013397, + "learning_rate": 7.997030073810698e-05, + "loss": 0.0539, + "step": 9920 + }, + { + "grad_norm": 0.2977638840675354, + "learning_rate": 7.992616547576217e-05, + "loss": 0.0417, + "step": 9930 + }, + { + "grad_norm": 0.2937262952327728, + "learning_rate": 7.988199385030445e-05, + "loss": 0.0412, + "step": 9940 + }, + { + "grad_norm": 0.26820459961891174, + "learning_rate": 7.98377859154065e-05, + "loss": 0.035, + "step": 9950 + }, + { + "grad_norm": 0.2686287760734558, + "learning_rate": 7.979354172478516e-05, + "loss": 0.0374, + "step": 9960 + }, + { + "grad_norm": 0.2354012131690979, + "learning_rate": 7.974926133220127e-05, + "loss": 0.0475, + "step": 9970 + }, + { + "grad_norm": 0.2203461080789566, + "learning_rate": 7.970494479145967e-05, + "loss": 0.0436, + "step": 9980 + }, + { + "grad_norm": 0.290770024061203, + "learning_rate": 7.966059215640919e-05, + "loss": 0.0463, + "step": 9990 + }, + { + "grad_norm": 0.3050641417503357, + "learning_rate": 7.961620348094241e-05, + "loss": 0.0541, + "step": 10000 + } + ], + "logging_steps": 10, + "max_steps": 30000, + "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": 8, + "trial_name": null, + "trial_params": null +}