diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,12034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.2, + "eval_steps": 500, + "global_step": 20000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "grad_norm": 0.24323463439941406, + "learning_rate": 1.8e-07, + "loss": 1.0309, + "step": 10 + }, + { + "grad_norm": 0.34074512124061584, + "learning_rate": 3.8e-07, + "loss": 1.0393, + "step": 20 + }, + { + "grad_norm": 0.29844534397125244, + "learning_rate": 5.8e-07, + "loss": 1.012, + "step": 30 + }, + { + "grad_norm": 0.31005048751831055, + "learning_rate": 7.8e-07, + "loss": 1.0181, + "step": 40 + }, + { + "grad_norm": 0.2365904003381729, + "learning_rate": 9.8e-07, + "loss": 1.0312, + "step": 50 + }, + { + "grad_norm": 0.2460295855998993, + "learning_rate": 1.18e-06, + "loss": 1.0365, + "step": 60 + }, + { + "grad_norm": 0.3435072600841522, + "learning_rate": 1.3800000000000001e-06, + "loss": 1.0268, + "step": 70 + }, + { + "grad_norm": 0.1815941482782364, + "learning_rate": 1.5800000000000003e-06, + "loss": 1.0257, + "step": 80 + }, + { + "grad_norm": 0.19930750131607056, + "learning_rate": 1.7800000000000001e-06, + "loss": 1.0266, + "step": 90 + }, + { + "grad_norm": 0.27344945073127747, + "learning_rate": 1.98e-06, + "loss": 1.035, + "step": 100 + }, + { + "grad_norm": 0.28589773178100586, + "learning_rate": 2.1800000000000003e-06, + "loss": 1.0229, + "step": 110 + }, + { + "grad_norm": 0.2818903923034668, + "learning_rate": 2.38e-06, + "loss": 1.0171, + "step": 120 + }, + { + "grad_norm": 0.288936972618103, + "learning_rate": 2.5800000000000003e-06, + "loss": 1.0331, + "step": 130 + }, + { + "grad_norm": 0.2609023451805115, + "learning_rate": 2.78e-06, + "loss": 1.0029, + "step": 140 + }, + { + "grad_norm": 0.2229061871767044, + "learning_rate": 2.9800000000000003e-06, + "loss": 1.0233, + "step": 150 + }, + { + "grad_norm": 0.2491007298231125, + "learning_rate": 3.1800000000000005e-06, + "loss": 1.0387, + "step": 160 + }, + { + "grad_norm": 0.2276277244091034, + "learning_rate": 3.38e-06, + "loss": 1.0171, + "step": 170 + }, + { + "grad_norm": 0.30169904232025146, + "learning_rate": 3.58e-06, + "loss": 1.0272, + "step": 180 + }, + { + "grad_norm": 0.30298280715942383, + "learning_rate": 3.7800000000000002e-06, + "loss": 1.0132, + "step": 190 + }, + { + "grad_norm": 0.3040730059146881, + "learning_rate": 3.98e-06, + "loss": 1.0253, + "step": 200 + }, + { + "grad_norm": 0.32458242774009705, + "learning_rate": 4.18e-06, + "loss": 1.0278, + "step": 210 + }, + { + "grad_norm": 0.2951343357563019, + "learning_rate": 4.38e-06, + "loss": 1.011, + "step": 220 + }, + { + "grad_norm": 0.4917892515659332, + "learning_rate": 4.58e-06, + "loss": 1.0313, + "step": 230 + }, + { + "grad_norm": 0.7013012170791626, + "learning_rate": 4.780000000000001e-06, + "loss": 1.0009, + "step": 240 + }, + { + "grad_norm": 1.0409867763519287, + "learning_rate": 4.98e-06, + "loss": 0.9843, + "step": 250 + }, + { + "grad_norm": 1.3605926036834717, + "learning_rate": 5.18e-06, + "loss": 0.9535, + "step": 260 + }, + { + "grad_norm": 1.7467690706253052, + "learning_rate": 5.38e-06, + "loss": 0.9499, + "step": 270 + }, + { + "grad_norm": 1.6608549356460571, + "learning_rate": 5.580000000000001e-06, + "loss": 0.9154, + "step": 280 + }, + { + "grad_norm": 1.395630955696106, + "learning_rate": 5.78e-06, + "loss": 0.9164, + "step": 290 + }, + { + "grad_norm": 2.0484440326690674, + "learning_rate": 5.98e-06, + "loss": 0.8721, + "step": 300 + }, + { + "grad_norm": 2.3988609313964844, + "learning_rate": 6.18e-06, + "loss": 0.8632, + "step": 310 + }, + { + "grad_norm": 1.9542955160140991, + "learning_rate": 6.38e-06, + "loss": 0.7978, + "step": 320 + }, + { + "grad_norm": 2.325824022293091, + "learning_rate": 6.58e-06, + "loss": 0.7808, + "step": 330 + }, + { + "grad_norm": 2.3015785217285156, + "learning_rate": 6.78e-06, + "loss": 0.7454, + "step": 340 + }, + { + "grad_norm": 2.4168238639831543, + "learning_rate": 6.98e-06, + "loss": 0.7369, + "step": 350 + }, + { + "grad_norm": 2.097677230834961, + "learning_rate": 7.180000000000001e-06, + "loss": 0.6807, + "step": 360 + }, + { + "grad_norm": 2.358717918395996, + "learning_rate": 7.3800000000000005e-06, + "loss": 0.7152, + "step": 370 + }, + { + "grad_norm": 2.6012001037597656, + "learning_rate": 7.580000000000001e-06, + "loss": 0.6652, + "step": 380 + }, + { + "grad_norm": 2.4282021522521973, + "learning_rate": 7.78e-06, + "loss": 0.6525, + "step": 390 + }, + { + "grad_norm": 3.712512731552124, + "learning_rate": 7.98e-06, + "loss": 0.6401, + "step": 400 + }, + { + "grad_norm": 2.7904763221740723, + "learning_rate": 8.18e-06, + "loss": 0.6029, + "step": 410 + }, + { + "grad_norm": 3.173227071762085, + "learning_rate": 8.380000000000001e-06, + "loss": 0.5751, + "step": 420 + }, + { + "grad_norm": 2.756822347640991, + "learning_rate": 8.580000000000001e-06, + "loss": 0.5691, + "step": 430 + }, + { + "grad_norm": 2.6459238529205322, + "learning_rate": 8.78e-06, + "loss": 0.5309, + "step": 440 + }, + { + "grad_norm": 1.9680471420288086, + "learning_rate": 8.98e-06, + "loss": 0.4919, + "step": 450 + }, + { + "grad_norm": 3.2210946083068848, + "learning_rate": 9.180000000000002e-06, + "loss": 0.4923, + "step": 460 + }, + { + "grad_norm": 3.4105448722839355, + "learning_rate": 9.38e-06, + "loss": 0.4769, + "step": 470 + }, + { + "grad_norm": 3.3270251750946045, + "learning_rate": 9.58e-06, + "loss": 0.4766, + "step": 480 + }, + { + "grad_norm": 2.685455083847046, + "learning_rate": 9.78e-06, + "loss": 0.4554, + "step": 490 + }, + { + "grad_norm": 4.284464359283447, + "learning_rate": 9.980000000000001e-06, + "loss": 0.4061, + "step": 500 + }, + { + "grad_norm": 3.293006658554077, + "learning_rate": 1.018e-05, + "loss": 0.4232, + "step": 510 + }, + { + "grad_norm": 2.953587293624878, + "learning_rate": 1.038e-05, + "loss": 0.3862, + "step": 520 + }, + { + "grad_norm": 3.5714266300201416, + "learning_rate": 1.058e-05, + "loss": 0.3812, + "step": 530 + }, + { + "grad_norm": 2.9290807247161865, + "learning_rate": 1.0780000000000002e-05, + "loss": 0.3279, + "step": 540 + }, + { + "grad_norm": 3.8112032413482666, + "learning_rate": 1.098e-05, + "loss": 0.3573, + "step": 550 + }, + { + "grad_norm": 3.4585070610046387, + "learning_rate": 1.118e-05, + "loss": 0.3439, + "step": 560 + }, + { + "grad_norm": 3.087247610092163, + "learning_rate": 1.1380000000000001e-05, + "loss": 0.2902, + "step": 570 + }, + { + "grad_norm": 2.668964147567749, + "learning_rate": 1.1580000000000001e-05, + "loss": 0.2707, + "step": 580 + }, + { + "grad_norm": 3.6468138694763184, + "learning_rate": 1.178e-05, + "loss": 0.2743, + "step": 590 + }, + { + "grad_norm": 2.388329267501831, + "learning_rate": 1.198e-05, + "loss": 0.2539, + "step": 600 + }, + { + "grad_norm": 3.534154176712036, + "learning_rate": 1.2180000000000002e-05, + "loss": 0.2346, + "step": 610 + }, + { + "grad_norm": 3.371579170227051, + "learning_rate": 1.238e-05, + "loss": 0.2365, + "step": 620 + }, + { + "grad_norm": 3.232956647872925, + "learning_rate": 1.258e-05, + "loss": 0.2274, + "step": 630 + }, + { + "grad_norm": 4.677562713623047, + "learning_rate": 1.278e-05, + "loss": 0.2146, + "step": 640 + }, + { + "grad_norm": 3.47682523727417, + "learning_rate": 1.2980000000000001e-05, + "loss": 0.1974, + "step": 650 + }, + { + "grad_norm": 3.9820761680603027, + "learning_rate": 1.3180000000000001e-05, + "loss": 0.2201, + "step": 660 + }, + { + "grad_norm": 2.4261512756347656, + "learning_rate": 1.338e-05, + "loss": 0.1715, + "step": 670 + }, + { + "grad_norm": 4.057462692260742, + "learning_rate": 1.358e-05, + "loss": 0.1935, + "step": 680 + }, + { + "grad_norm": 3.1671414375305176, + "learning_rate": 1.3780000000000002e-05, + "loss": 0.1816, + "step": 690 + }, + { + "grad_norm": 3.836470603942871, + "learning_rate": 1.3980000000000002e-05, + "loss": 0.2009, + "step": 700 + }, + { + "grad_norm": 3.4371752738952637, + "learning_rate": 1.4180000000000001e-05, + "loss": 0.1664, + "step": 710 + }, + { + "grad_norm": 3.9317233562469482, + "learning_rate": 1.4380000000000001e-05, + "loss": 0.1795, + "step": 720 + }, + { + "grad_norm": 2.880136728286743, + "learning_rate": 1.4580000000000003e-05, + "loss": 0.1636, + "step": 730 + }, + { + "grad_norm": 1.9766147136688232, + "learning_rate": 1.4779999999999999e-05, + "loss": 0.1331, + "step": 740 + }, + { + "grad_norm": 1.6496448516845703, + "learning_rate": 1.4979999999999999e-05, + "loss": 0.1573, + "step": 750 + }, + { + "grad_norm": 2.026895761489868, + "learning_rate": 1.518e-05, + "loss": 0.1476, + "step": 760 + }, + { + "grad_norm": 2.0964386463165283, + "learning_rate": 1.538e-05, + "loss": 0.1593, + "step": 770 + }, + { + "grad_norm": 2.7636566162109375, + "learning_rate": 1.558e-05, + "loss": 0.1953, + "step": 780 + }, + { + "grad_norm": 2.352786064147949, + "learning_rate": 1.578e-05, + "loss": 0.1438, + "step": 790 + }, + { + "grad_norm": 2.7201473712921143, + "learning_rate": 1.598e-05, + "loss": 0.1386, + "step": 800 + }, + { + "grad_norm": 2.4620485305786133, + "learning_rate": 1.618e-05, + "loss": 0.1476, + "step": 810 + }, + { + "grad_norm": 2.4028594493865967, + "learning_rate": 1.6380000000000002e-05, + "loss": 0.1427, + "step": 820 + }, + { + "grad_norm": 2.453202247619629, + "learning_rate": 1.658e-05, + "loss": 0.1595, + "step": 830 + }, + { + "grad_norm": 2.199979305267334, + "learning_rate": 1.6780000000000002e-05, + "loss": 0.1543, + "step": 840 + }, + { + "grad_norm": 1.7894608974456787, + "learning_rate": 1.698e-05, + "loss": 0.1574, + "step": 850 + }, + { + "grad_norm": 2.3981385231018066, + "learning_rate": 1.718e-05, + "loss": 0.1365, + "step": 860 + }, + { + "grad_norm": 2.667057991027832, + "learning_rate": 1.7380000000000003e-05, + "loss": 0.1645, + "step": 870 + }, + { + "grad_norm": 2.543144941329956, + "learning_rate": 1.758e-05, + "loss": 0.135, + "step": 880 + }, + { + "grad_norm": 2.41051983833313, + "learning_rate": 1.7780000000000003e-05, + "loss": 0.1313, + "step": 890 + }, + { + "grad_norm": 2.010571002960205, + "learning_rate": 1.798e-05, + "loss": 0.1354, + "step": 900 + }, + { + "grad_norm": 2.220653533935547, + "learning_rate": 1.818e-05, + "loss": 0.1253, + "step": 910 + }, + { + "grad_norm": 1.8454041481018066, + "learning_rate": 1.838e-05, + "loss": 0.116, + "step": 920 + }, + { + "grad_norm": 1.979677677154541, + "learning_rate": 1.858e-05, + "loss": 0.1284, + "step": 930 + }, + { + "grad_norm": 2.5556516647338867, + "learning_rate": 1.878e-05, + "loss": 0.14, + "step": 940 + }, + { + "grad_norm": 2.128957509994507, + "learning_rate": 1.898e-05, + "loss": 0.1148, + "step": 950 + }, + { + "grad_norm": 2.0536532402038574, + "learning_rate": 1.918e-05, + "loss": 0.1233, + "step": 960 + }, + { + "grad_norm": 2.3479814529418945, + "learning_rate": 1.938e-05, + "loss": 0.1483, + "step": 970 + }, + { + "grad_norm": 1.476418137550354, + "learning_rate": 1.9580000000000002e-05, + "loss": 0.1137, + "step": 980 + }, + { + "grad_norm": 2.468423843383789, + "learning_rate": 1.978e-05, + "loss": 0.1102, + "step": 990 + }, + { + "grad_norm": 1.5869988203048706, + "learning_rate": 1.9980000000000002e-05, + "loss": 0.1133, + "step": 1000 + }, + { + "grad_norm": 1.7217639684677124, + "learning_rate": 2.0180000000000003e-05, + "loss": 0.1077, + "step": 1010 + }, + { + "grad_norm": 2.7842702865600586, + "learning_rate": 2.038e-05, + "loss": 0.1329, + "step": 1020 + }, + { + "grad_norm": 1.7710098028182983, + "learning_rate": 2.0580000000000003e-05, + "loss": 0.0992, + "step": 1030 + }, + { + "grad_norm": 1.7425968647003174, + "learning_rate": 2.078e-05, + "loss": 0.1128, + "step": 1040 + }, + { + "grad_norm": 2.5411627292633057, + "learning_rate": 2.098e-05, + "loss": 0.118, + "step": 1050 + }, + { + "grad_norm": 1.512483835220337, + "learning_rate": 2.118e-05, + "loss": 0.1274, + "step": 1060 + }, + { + "grad_norm": 2.261084794998169, + "learning_rate": 2.138e-05, + "loss": 0.1318, + "step": 1070 + }, + { + "grad_norm": 1.5960973501205444, + "learning_rate": 2.158e-05, + "loss": 0.1393, + "step": 1080 + }, + { + "grad_norm": 1.3446298837661743, + "learning_rate": 2.178e-05, + "loss": 0.1101, + "step": 1090 + }, + { + "grad_norm": 1.9963794946670532, + "learning_rate": 2.198e-05, + "loss": 0.0893, + "step": 1100 + }, + { + "grad_norm": 2.878643274307251, + "learning_rate": 2.218e-05, + "loss": 0.1342, + "step": 1110 + }, + { + "grad_norm": 1.5142366886138916, + "learning_rate": 2.2380000000000003e-05, + "loss": 0.0932, + "step": 1120 + }, + { + "grad_norm": 1.3655840158462524, + "learning_rate": 2.258e-05, + "loss": 0.1175, + "step": 1130 + }, + { + "grad_norm": 1.9727683067321777, + "learning_rate": 2.2780000000000002e-05, + "loss": 0.1521, + "step": 1140 + }, + { + "grad_norm": 1.6121673583984375, + "learning_rate": 2.298e-05, + "loss": 0.1252, + "step": 1150 + }, + { + "grad_norm": 1.2537587881088257, + "learning_rate": 2.318e-05, + "loss": 0.111, + "step": 1160 + }, + { + "grad_norm": 2.148144006729126, + "learning_rate": 2.3380000000000003e-05, + "loss": 0.108, + "step": 1170 + }, + { + "grad_norm": 1.69101881980896, + "learning_rate": 2.358e-05, + "loss": 0.0928, + "step": 1180 + }, + { + "grad_norm": 1.6988859176635742, + "learning_rate": 2.3780000000000003e-05, + "loss": 0.1105, + "step": 1190 + }, + { + "grad_norm": 4.235020160675049, + "learning_rate": 2.398e-05, + "loss": 0.1193, + "step": 1200 + }, + { + "grad_norm": 4.887176036834717, + "learning_rate": 2.418e-05, + "loss": 0.1357, + "step": 1210 + }, + { + "grad_norm": 0.813777506351471, + "learning_rate": 2.438e-05, + "loss": 0.1044, + "step": 1220 + }, + { + "grad_norm": 1.8302229642868042, + "learning_rate": 2.4580000000000002e-05, + "loss": 0.1101, + "step": 1230 + }, + { + "grad_norm": 1.3978817462921143, + "learning_rate": 2.478e-05, + "loss": 0.1184, + "step": 1240 + }, + { + "grad_norm": 2.0479328632354736, + "learning_rate": 2.498e-05, + "loss": 0.0906, + "step": 1250 + }, + { + "grad_norm": 2.38946795463562, + "learning_rate": 2.5180000000000003e-05, + "loss": 0.1043, + "step": 1260 + }, + { + "grad_norm": 1.249617338180542, + "learning_rate": 2.5380000000000004e-05, + "loss": 0.0849, + "step": 1270 + }, + { + "grad_norm": 1.517021656036377, + "learning_rate": 2.5580000000000002e-05, + "loss": 0.107, + "step": 1280 + }, + { + "grad_norm": 1.3285075426101685, + "learning_rate": 2.5779999999999997e-05, + "loss": 0.105, + "step": 1290 + }, + { + "grad_norm": 1.5297868251800537, + "learning_rate": 2.598e-05, + "loss": 0.0973, + "step": 1300 + }, + { + "grad_norm": 0.9701268672943115, + "learning_rate": 2.618e-05, + "loss": 0.0976, + "step": 1310 + }, + { + "grad_norm": 1.2006975412368774, + "learning_rate": 2.6379999999999998e-05, + "loss": 0.0994, + "step": 1320 + }, + { + "grad_norm": 1.2093645334243774, + "learning_rate": 2.658e-05, + "loss": 0.0773, + "step": 1330 + }, + { + "grad_norm": 1.6315072774887085, + "learning_rate": 2.678e-05, + "loss": 0.0838, + "step": 1340 + }, + { + "grad_norm": 0.9597839713096619, + "learning_rate": 2.698e-05, + "loss": 0.1086, + "step": 1350 + }, + { + "grad_norm": 2.636733293533325, + "learning_rate": 2.718e-05, + "loss": 0.1013, + "step": 1360 + }, + { + "grad_norm": 1.5911072492599487, + "learning_rate": 2.738e-05, + "loss": 0.0824, + "step": 1370 + }, + { + "grad_norm": 1.6432372331619263, + "learning_rate": 2.758e-05, + "loss": 0.0809, + "step": 1380 + }, + { + "grad_norm": 1.8894906044006348, + "learning_rate": 2.778e-05, + "loss": 0.0905, + "step": 1390 + }, + { + "grad_norm": 1.47123122215271, + "learning_rate": 2.798e-05, + "loss": 0.0809, + "step": 1400 + }, + { + "grad_norm": 1.0250835418701172, + "learning_rate": 2.818e-05, + "loss": 0.0923, + "step": 1410 + }, + { + "grad_norm": 1.6976494789123535, + "learning_rate": 2.8380000000000003e-05, + "loss": 0.0928, + "step": 1420 + }, + { + "grad_norm": 1.0837762355804443, + "learning_rate": 2.858e-05, + "loss": 0.0734, + "step": 1430 + }, + { + "grad_norm": 1.2509829998016357, + "learning_rate": 2.8780000000000002e-05, + "loss": 0.0987, + "step": 1440 + }, + { + "grad_norm": 1.0369926691055298, + "learning_rate": 2.898e-05, + "loss": 0.1036, + "step": 1450 + }, + { + "grad_norm": 1.3391159772872925, + "learning_rate": 2.9180000000000002e-05, + "loss": 0.0996, + "step": 1460 + }, + { + "grad_norm": 1.259562611579895, + "learning_rate": 2.9380000000000003e-05, + "loss": 0.0963, + "step": 1470 + }, + { + "grad_norm": 1.2859934568405151, + "learning_rate": 2.958e-05, + "loss": 0.0816, + "step": 1480 + }, + { + "grad_norm": 1.3876073360443115, + "learning_rate": 2.9780000000000003e-05, + "loss": 0.0902, + "step": 1490 + }, + { + "grad_norm": 1.1462910175323486, + "learning_rate": 2.998e-05, + "loss": 0.0951, + "step": 1500 + }, + { + "grad_norm": 1.2179933786392212, + "learning_rate": 3.0180000000000002e-05, + "loss": 0.0807, + "step": 1510 + }, + { + "grad_norm": 1.4895336627960205, + "learning_rate": 3.0380000000000004e-05, + "loss": 0.0893, + "step": 1520 + }, + { + "grad_norm": 1.0243799686431885, + "learning_rate": 3.058e-05, + "loss": 0.0966, + "step": 1530 + }, + { + "grad_norm": 1.0312671661376953, + "learning_rate": 3.078e-05, + "loss": 0.0845, + "step": 1540 + }, + { + "grad_norm": 1.1680856943130493, + "learning_rate": 3.0980000000000005e-05, + "loss": 0.0812, + "step": 1550 + }, + { + "grad_norm": 0.8857987523078918, + "learning_rate": 3.118e-05, + "loss": 0.1068, + "step": 1560 + }, + { + "grad_norm": 1.4180179834365845, + "learning_rate": 3.138e-05, + "loss": 0.0741, + "step": 1570 + }, + { + "grad_norm": 1.2802883386611938, + "learning_rate": 3.1580000000000006e-05, + "loss": 0.0859, + "step": 1580 + }, + { + "grad_norm": 1.434507966041565, + "learning_rate": 3.1780000000000004e-05, + "loss": 0.0923, + "step": 1590 + }, + { + "grad_norm": 1.0308754444122314, + "learning_rate": 3.198e-05, + "loss": 0.0784, + "step": 1600 + }, + { + "grad_norm": 0.8732582926750183, + "learning_rate": 3.218e-05, + "loss": 0.0672, + "step": 1610 + }, + { + "grad_norm": 0.7804297804832458, + "learning_rate": 3.238e-05, + "loss": 0.0802, + "step": 1620 + }, + { + "grad_norm": 1.3388926982879639, + "learning_rate": 3.2579999999999996e-05, + "loss": 0.1015, + "step": 1630 + }, + { + "grad_norm": 1.6440131664276123, + "learning_rate": 3.278e-05, + "loss": 0.0879, + "step": 1640 + }, + { + "grad_norm": 1.1676356792449951, + "learning_rate": 3.298e-05, + "loss": 0.0872, + "step": 1650 + }, + { + "grad_norm": 1.383884310722351, + "learning_rate": 3.318e-05, + "loss": 0.0889, + "step": 1660 + }, + { + "grad_norm": 0.8376438021659851, + "learning_rate": 3.338e-05, + "loss": 0.0666, + "step": 1670 + }, + { + "grad_norm": 1.1388325691223145, + "learning_rate": 3.358e-05, + "loss": 0.0747, + "step": 1680 + }, + { + "grad_norm": 1.247609257698059, + "learning_rate": 3.378e-05, + "loss": 0.0746, + "step": 1690 + }, + { + "grad_norm": 0.9289684295654297, + "learning_rate": 3.398e-05, + "loss": 0.0837, + "step": 1700 + }, + { + "grad_norm": 1.8538250923156738, + "learning_rate": 3.418e-05, + "loss": 0.0979, + "step": 1710 + }, + { + "grad_norm": 0.9559040665626526, + "learning_rate": 3.438e-05, + "loss": 0.0818, + "step": 1720 + }, + { + "grad_norm": 1.0027278661727905, + "learning_rate": 3.4580000000000004e-05, + "loss": 0.0782, + "step": 1730 + }, + { + "grad_norm": 1.5428392887115479, + "learning_rate": 3.478e-05, + "loss": 0.0726, + "step": 1740 + }, + { + "grad_norm": 1.6734929084777832, + "learning_rate": 3.498e-05, + "loss": 0.0868, + "step": 1750 + }, + { + "grad_norm": 2.180098295211792, + "learning_rate": 3.518e-05, + "loss": 0.0872, + "step": 1760 + }, + { + "grad_norm": 1.0983431339263916, + "learning_rate": 3.5380000000000003e-05, + "loss": 0.0743, + "step": 1770 + }, + { + "grad_norm": 0.942829966545105, + "learning_rate": 3.558e-05, + "loss": 0.0979, + "step": 1780 + }, + { + "grad_norm": 1.6690951585769653, + "learning_rate": 3.578e-05, + "loss": 0.0755, + "step": 1790 + }, + { + "grad_norm": 1.0191823244094849, + "learning_rate": 3.5980000000000004e-05, + "loss": 0.0795, + "step": 1800 + }, + { + "grad_norm": 0.9048247337341309, + "learning_rate": 3.618e-05, + "loss": 0.0683, + "step": 1810 + }, + { + "grad_norm": 1.5634185075759888, + "learning_rate": 3.638e-05, + "loss": 0.0865, + "step": 1820 + }, + { + "grad_norm": 1.554189682006836, + "learning_rate": 3.6580000000000006e-05, + "loss": 0.083, + "step": 1830 + }, + { + "grad_norm": 0.9538047909736633, + "learning_rate": 3.6780000000000004e-05, + "loss": 0.0877, + "step": 1840 + }, + { + "grad_norm": 0.9088499546051025, + "learning_rate": 3.698e-05, + "loss": 0.0988, + "step": 1850 + }, + { + "grad_norm": 1.038340449333191, + "learning_rate": 3.7180000000000007e-05, + "loss": 0.0996, + "step": 1860 + }, + { + "grad_norm": 1.010509729385376, + "learning_rate": 3.7380000000000005e-05, + "loss": 0.0771, + "step": 1870 + }, + { + "grad_norm": 0.9490908980369568, + "learning_rate": 3.758e-05, + "loss": 0.0633, + "step": 1880 + }, + { + "grad_norm": 0.8129947185516357, + "learning_rate": 3.778000000000001e-05, + "loss": 0.0784, + "step": 1890 + }, + { + "grad_norm": 0.8083765506744385, + "learning_rate": 3.7980000000000006e-05, + "loss": 0.0652, + "step": 1900 + }, + { + "grad_norm": 1.436314344406128, + "learning_rate": 3.818e-05, + "loss": 0.0988, + "step": 1910 + }, + { + "grad_norm": 0.8581476211547852, + "learning_rate": 3.838e-05, + "loss": 0.0759, + "step": 1920 + }, + { + "grad_norm": 0.699325442314148, + "learning_rate": 3.858e-05, + "loss": 0.0681, + "step": 1930 + }, + { + "grad_norm": 1.452941656112671, + "learning_rate": 3.878e-05, + "loss": 0.0802, + "step": 1940 + }, + { + "grad_norm": 0.6633671522140503, + "learning_rate": 3.898e-05, + "loss": 0.0628, + "step": 1950 + }, + { + "grad_norm": 1.7221360206604004, + "learning_rate": 3.918e-05, + "loss": 0.0941, + "step": 1960 + }, + { + "grad_norm": 0.7586915493011475, + "learning_rate": 3.938e-05, + "loss": 0.0646, + "step": 1970 + }, + { + "grad_norm": 1.2361974716186523, + "learning_rate": 3.958e-05, + "loss": 0.0867, + "step": 1980 + }, + { + "grad_norm": 0.9242485165596008, + "learning_rate": 3.978e-05, + "loss": 0.0601, + "step": 1990 + }, + { + "grad_norm": 1.0715078115463257, + "learning_rate": 3.998e-05, + "loss": 0.0782, + "step": 2000 + }, + { + "grad_norm": 1.0248526334762573, + "learning_rate": 4.018e-05, + "loss": 0.1058, + "step": 2010 + }, + { + "grad_norm": 1.2648011445999146, + "learning_rate": 4.038e-05, + "loss": 0.0801, + "step": 2020 + }, + { + "grad_norm": 1.0998585224151611, + "learning_rate": 4.058e-05, + "loss": 0.0628, + "step": 2030 + }, + { + "grad_norm": 0.7973329424858093, + "learning_rate": 4.078e-05, + "loss": 0.0833, + "step": 2040 + }, + { + "grad_norm": 0.8564066290855408, + "learning_rate": 4.0980000000000004e-05, + "loss": 0.0784, + "step": 2050 + }, + { + "grad_norm": 1.236814022064209, + "learning_rate": 4.118e-05, + "loss": 0.0703, + "step": 2060 + }, + { + "grad_norm": 0.8105146884918213, + "learning_rate": 4.138e-05, + "loss": 0.0625, + "step": 2070 + }, + { + "grad_norm": 0.7948865294456482, + "learning_rate": 4.1580000000000005e-05, + "loss": 0.0696, + "step": 2080 + }, + { + "grad_norm": 0.965408205986023, + "learning_rate": 4.178e-05, + "loss": 0.0711, + "step": 2090 + }, + { + "grad_norm": 0.9406697154045105, + "learning_rate": 4.198e-05, + "loss": 0.0727, + "step": 2100 + }, + { + "grad_norm": 0.8055232763290405, + "learning_rate": 4.2180000000000006e-05, + "loss": 0.0933, + "step": 2110 + }, + { + "grad_norm": 0.8500663042068481, + "learning_rate": 4.2380000000000004e-05, + "loss": 0.0659, + "step": 2120 + }, + { + "grad_norm": 0.8973005414009094, + "learning_rate": 4.258e-05, + "loss": 0.0707, + "step": 2130 + }, + { + "grad_norm": 0.8494951725006104, + "learning_rate": 4.278e-05, + "loss": 0.0558, + "step": 2140 + }, + { + "grad_norm": 0.8333751559257507, + "learning_rate": 4.2980000000000005e-05, + "loss": 0.0798, + "step": 2150 + }, + { + "grad_norm": 0.7871338725090027, + "learning_rate": 4.318e-05, + "loss": 0.0625, + "step": 2160 + }, + { + "grad_norm": 0.7040622234344482, + "learning_rate": 4.338e-05, + "loss": 0.0675, + "step": 2170 + }, + { + "grad_norm": 0.9401140809059143, + "learning_rate": 4.3580000000000006e-05, + "loss": 0.0773, + "step": 2180 + }, + { + "grad_norm": 0.8710242509841919, + "learning_rate": 4.3780000000000004e-05, + "loss": 0.0636, + "step": 2190 + }, + { + "grad_norm": 0.765397846698761, + "learning_rate": 4.398e-05, + "loss": 0.0613, + "step": 2200 + }, + { + "grad_norm": 0.6467282772064209, + "learning_rate": 4.418000000000001e-05, + "loss": 0.064, + "step": 2210 + }, + { + "grad_norm": 1.477567434310913, + "learning_rate": 4.438e-05, + "loss": 0.0623, + "step": 2220 + }, + { + "grad_norm": 0.9443312883377075, + "learning_rate": 4.458e-05, + "loss": 0.0668, + "step": 2230 + }, + { + "grad_norm": 0.7416368722915649, + "learning_rate": 4.478e-05, + "loss": 0.0604, + "step": 2240 + }, + { + "grad_norm": 0.7133755683898926, + "learning_rate": 4.498e-05, + "loss": 0.0737, + "step": 2250 + }, + { + "grad_norm": 0.6120943427085876, + "learning_rate": 4.518e-05, + "loss": 0.0553, + "step": 2260 + }, + { + "grad_norm": 0.8734622001647949, + "learning_rate": 4.538e-05, + "loss": 0.0544, + "step": 2270 + }, + { + "grad_norm": 0.6398209929466248, + "learning_rate": 4.558e-05, + "loss": 0.0636, + "step": 2280 + }, + { + "grad_norm": 1.380096435546875, + "learning_rate": 4.578e-05, + "loss": 0.0867, + "step": 2290 + }, + { + "grad_norm": 0.6424226760864258, + "learning_rate": 4.5980000000000004e-05, + "loss": 0.0534, + "step": 2300 + }, + { + "grad_norm": 0.5737485289573669, + "learning_rate": 4.618e-05, + "loss": 0.0591, + "step": 2310 + }, + { + "grad_norm": 0.6159701943397522, + "learning_rate": 4.638e-05, + "loss": 0.0508, + "step": 2320 + }, + { + "grad_norm": 0.6447447538375854, + "learning_rate": 4.6580000000000005e-05, + "loss": 0.0699, + "step": 2330 + }, + { + "grad_norm": 0.6198409795761108, + "learning_rate": 4.678e-05, + "loss": 0.0628, + "step": 2340 + }, + { + "grad_norm": 0.6166672110557556, + "learning_rate": 4.698e-05, + "loss": 0.0595, + "step": 2350 + }, + { + "grad_norm": 0.7590539455413818, + "learning_rate": 4.718e-05, + "loss": 0.0628, + "step": 2360 + }, + { + "grad_norm": 0.6507792472839355, + "learning_rate": 4.7380000000000004e-05, + "loss": 0.0673, + "step": 2370 + }, + { + "grad_norm": 0.766956090927124, + "learning_rate": 4.758e-05, + "loss": 0.0773, + "step": 2380 + }, + { + "grad_norm": 0.883449912071228, + "learning_rate": 4.778e-05, + "loss": 0.0778, + "step": 2390 + }, + { + "grad_norm": 0.6879324913024902, + "learning_rate": 4.7980000000000005e-05, + "loss": 0.0668, + "step": 2400 + }, + { + "grad_norm": 0.9086018204689026, + "learning_rate": 4.818e-05, + "loss": 0.0625, + "step": 2410 + }, + { + "grad_norm": 0.7436268329620361, + "learning_rate": 4.838e-05, + "loss": 0.0985, + "step": 2420 + }, + { + "grad_norm": 0.9962342381477356, + "learning_rate": 4.8580000000000006e-05, + "loss": 0.0694, + "step": 2430 + }, + { + "grad_norm": 0.8323448300361633, + "learning_rate": 4.8780000000000004e-05, + "loss": 0.0698, + "step": 2440 + }, + { + "grad_norm": 0.7347047328948975, + "learning_rate": 4.898e-05, + "loss": 0.0588, + "step": 2450 + }, + { + "grad_norm": 0.6135653853416443, + "learning_rate": 4.918000000000001e-05, + "loss": 0.0546, + "step": 2460 + }, + { + "grad_norm": 0.6037757396697998, + "learning_rate": 4.9380000000000005e-05, + "loss": 0.0519, + "step": 2470 + }, + { + "grad_norm": 0.6975889801979065, + "learning_rate": 4.958e-05, + "loss": 0.0771, + "step": 2480 + }, + { + "grad_norm": 0.7472870349884033, + "learning_rate": 4.978e-05, + "loss": 0.0598, + "step": 2490 + }, + { + "grad_norm": 0.6862678527832031, + "learning_rate": 4.9980000000000006e-05, + "loss": 0.0545, + "step": 2500 + }, + { + "grad_norm": 0.8071557283401489, + "learning_rate": 5.0180000000000004e-05, + "loss": 0.0517, + "step": 2510 + }, + { + "grad_norm": 0.6523743271827698, + "learning_rate": 5.038e-05, + "loss": 0.0621, + "step": 2520 + }, + { + "grad_norm": 0.6114144325256348, + "learning_rate": 5.058000000000001e-05, + "loss": 0.0501, + "step": 2530 + }, + { + "grad_norm": 0.6552909016609192, + "learning_rate": 5.0780000000000005e-05, + "loss": 0.0615, + "step": 2540 + }, + { + "grad_norm": 0.8600518703460693, + "learning_rate": 5.098e-05, + "loss": 0.0669, + "step": 2550 + }, + { + "grad_norm": 0.6675088405609131, + "learning_rate": 5.118000000000001e-05, + "loss": 0.0623, + "step": 2560 + }, + { + "grad_norm": 0.7757863402366638, + "learning_rate": 5.1380000000000006e-05, + "loss": 0.0623, + "step": 2570 + }, + { + "grad_norm": 0.7180933356285095, + "learning_rate": 5.1580000000000004e-05, + "loss": 0.0629, + "step": 2580 + }, + { + "grad_norm": 0.7248421311378479, + "learning_rate": 5.178000000000001e-05, + "loss": 0.059, + "step": 2590 + }, + { + "grad_norm": 0.7094396948814392, + "learning_rate": 5.198000000000001e-05, + "loss": 0.0664, + "step": 2600 + }, + { + "grad_norm": 0.7055537700653076, + "learning_rate": 5.2180000000000005e-05, + "loss": 0.0657, + "step": 2610 + }, + { + "grad_norm": 0.7061589360237122, + "learning_rate": 5.238000000000001e-05, + "loss": 0.0511, + "step": 2620 + }, + { + "grad_norm": 0.7152190208435059, + "learning_rate": 5.258000000000001e-05, + "loss": 0.0616, + "step": 2630 + }, + { + "grad_norm": 0.8776269555091858, + "learning_rate": 5.2780000000000006e-05, + "loss": 0.07, + "step": 2640 + }, + { + "grad_norm": 0.5741156935691833, + "learning_rate": 5.2980000000000004e-05, + "loss": 0.071, + "step": 2650 + }, + { + "grad_norm": 0.6530773639678955, + "learning_rate": 5.318000000000001e-05, + "loss": 0.0763, + "step": 2660 + }, + { + "grad_norm": 0.7851654887199402, + "learning_rate": 5.338000000000001e-05, + "loss": 0.0733, + "step": 2670 + }, + { + "grad_norm": 1.0083369016647339, + "learning_rate": 5.3580000000000005e-05, + "loss": 0.0827, + "step": 2680 + }, + { + "grad_norm": 0.6418405175209045, + "learning_rate": 5.378e-05, + "loss": 0.0612, + "step": 2690 + }, + { + "grad_norm": 0.6848601698875427, + "learning_rate": 5.3979999999999995e-05, + "loss": 0.0535, + "step": 2700 + }, + { + "grad_norm": 0.6649168729782104, + "learning_rate": 5.418e-05, + "loss": 0.0585, + "step": 2710 + }, + { + "grad_norm": 0.6305927038192749, + "learning_rate": 5.438e-05, + "loss": 0.0788, + "step": 2720 + }, + { + "grad_norm": 0.7496237754821777, + "learning_rate": 5.4579999999999996e-05, + "loss": 0.0567, + "step": 2730 + }, + { + "grad_norm": 0.47063612937927246, + "learning_rate": 5.478e-05, + "loss": 0.0558, + "step": 2740 + }, + { + "grad_norm": 0.5131617188453674, + "learning_rate": 5.498e-05, + "loss": 0.0615, + "step": 2750 + }, + { + "grad_norm": 0.8309383392333984, + "learning_rate": 5.518e-05, + "loss": 0.0547, + "step": 2760 + }, + { + "grad_norm": 0.6489686369895935, + "learning_rate": 5.538e-05, + "loss": 0.0559, + "step": 2770 + }, + { + "grad_norm": 0.6935626864433289, + "learning_rate": 5.558e-05, + "loss": 0.0673, + "step": 2780 + }, + { + "grad_norm": 0.6607373952865601, + "learning_rate": 5.578e-05, + "loss": 0.0611, + "step": 2790 + }, + { + "grad_norm": 0.6474405527114868, + "learning_rate": 5.5979999999999996e-05, + "loss": 0.0531, + "step": 2800 + }, + { + "grad_norm": 1.0084667205810547, + "learning_rate": 5.618e-05, + "loss": 0.0905, + "step": 2810 + }, + { + "grad_norm": 0.9654085040092468, + "learning_rate": 5.638e-05, + "loss": 0.0737, + "step": 2820 + }, + { + "grad_norm": 0.584434449672699, + "learning_rate": 5.658e-05, + "loss": 0.0585, + "step": 2830 + }, + { + "grad_norm": 0.5688425302505493, + "learning_rate": 5.678e-05, + "loss": 0.0745, + "step": 2840 + }, + { + "grad_norm": 0.5918985605239868, + "learning_rate": 5.698e-05, + "loss": 0.069, + "step": 2850 + }, + { + "grad_norm": 0.6128337979316711, + "learning_rate": 5.718e-05, + "loss": 0.0584, + "step": 2860 + }, + { + "grad_norm": 0.5219225883483887, + "learning_rate": 5.738e-05, + "loss": 0.0518, + "step": 2870 + }, + { + "grad_norm": 0.4728126525878906, + "learning_rate": 5.758e-05, + "loss": 0.0485, + "step": 2880 + }, + { + "grad_norm": 0.7647413611412048, + "learning_rate": 5.778e-05, + "loss": 0.0556, + "step": 2890 + }, + { + "grad_norm": 0.5122050642967224, + "learning_rate": 5.7980000000000004e-05, + "loss": 0.0683, + "step": 2900 + }, + { + "grad_norm": 0.7916375994682312, + "learning_rate": 5.818e-05, + "loss": 0.055, + "step": 2910 + }, + { + "grad_norm": 0.7011560797691345, + "learning_rate": 5.838e-05, + "loss": 0.0568, + "step": 2920 + }, + { + "grad_norm": 0.6700689196586609, + "learning_rate": 5.858e-05, + "loss": 0.0477, + "step": 2930 + }, + { + "grad_norm": 0.5937225818634033, + "learning_rate": 5.878e-05, + "loss": 0.0501, + "step": 2940 + }, + { + "grad_norm": 0.49369940161705017, + "learning_rate": 5.898e-05, + "loss": 0.0475, + "step": 2950 + }, + { + "grad_norm": 0.6631751656532288, + "learning_rate": 5.918e-05, + "loss": 0.0615, + "step": 2960 + }, + { + "grad_norm": 0.7893167734146118, + "learning_rate": 5.9380000000000004e-05, + "loss": 0.0477, + "step": 2970 + }, + { + "grad_norm": 0.5764897465705872, + "learning_rate": 5.958e-05, + "loss": 0.0556, + "step": 2980 + }, + { + "grad_norm": 0.5732360482215881, + "learning_rate": 5.978e-05, + "loss": 0.0627, + "step": 2990 + }, + { + "grad_norm": 0.5444154143333435, + "learning_rate": 5.9980000000000005e-05, + "loss": 0.0612, + "step": 3000 + }, + { + "grad_norm": 0.5632981657981873, + "learning_rate": 6.018e-05, + "loss": 0.0449, + "step": 3010 + }, + { + "grad_norm": 0.6956084966659546, + "learning_rate": 6.038e-05, + "loss": 0.0508, + "step": 3020 + }, + { + "grad_norm": 0.5621139407157898, + "learning_rate": 6.0580000000000006e-05, + "loss": 0.0516, + "step": 3030 + }, + { + "grad_norm": 0.6331535577774048, + "learning_rate": 6.0780000000000004e-05, + "loss": 0.055, + "step": 3040 + }, + { + "grad_norm": 0.6451631188392639, + "learning_rate": 6.098e-05, + "loss": 0.0526, + "step": 3050 + }, + { + "grad_norm": 0.4736952781677246, + "learning_rate": 6.118000000000001e-05, + "loss": 0.0548, + "step": 3060 + }, + { + "grad_norm": 0.5182963609695435, + "learning_rate": 6.138e-05, + "loss": 0.0436, + "step": 3070 + }, + { + "grad_norm": 0.7367036938667297, + "learning_rate": 6.158e-05, + "loss": 0.0555, + "step": 3080 + }, + { + "grad_norm": 1.0562509298324585, + "learning_rate": 6.178000000000001e-05, + "loss": 0.0724, + "step": 3090 + }, + { + "grad_norm": 0.7549290657043457, + "learning_rate": 6.198e-05, + "loss": 0.0526, + "step": 3100 + }, + { + "grad_norm": 0.5529669523239136, + "learning_rate": 6.218e-05, + "loss": 0.0451, + "step": 3110 + }, + { + "grad_norm": 0.5710684061050415, + "learning_rate": 6.238000000000001e-05, + "loss": 0.0469, + "step": 3120 + }, + { + "grad_norm": 0.5792014598846436, + "learning_rate": 6.258e-05, + "loss": 0.0472, + "step": 3130 + }, + { + "grad_norm": 0.6089155673980713, + "learning_rate": 6.278e-05, + "loss": 0.0512, + "step": 3140 + }, + { + "grad_norm": 0.6686075329780579, + "learning_rate": 6.298000000000001e-05, + "loss": 0.0481, + "step": 3150 + }, + { + "grad_norm": 0.48785340785980225, + "learning_rate": 6.318e-05, + "loss": 0.0534, + "step": 3160 + }, + { + "grad_norm": 0.5482088923454285, + "learning_rate": 6.338e-05, + "loss": 0.0495, + "step": 3170 + }, + { + "grad_norm": 0.5249758362770081, + "learning_rate": 6.358000000000001e-05, + "loss": 0.0668, + "step": 3180 + }, + { + "grad_norm": 0.6556891798973083, + "learning_rate": 6.378e-05, + "loss": 0.0641, + "step": 3190 + }, + { + "grad_norm": 0.6621457934379578, + "learning_rate": 6.398000000000001e-05, + "loss": 0.0615, + "step": 3200 + }, + { + "grad_norm": 0.4882360100746155, + "learning_rate": 6.418000000000001e-05, + "loss": 0.0485, + "step": 3210 + }, + { + "grad_norm": 0.9999679923057556, + "learning_rate": 6.438e-05, + "loss": 0.07, + "step": 3220 + }, + { + "grad_norm": 0.6060804128646851, + "learning_rate": 6.458000000000001e-05, + "loss": 0.0641, + "step": 3230 + }, + { + "grad_norm": 0.41649818420410156, + "learning_rate": 6.478000000000001e-05, + "loss": 0.0494, + "step": 3240 + }, + { + "grad_norm": 0.4703940749168396, + "learning_rate": 6.498e-05, + "loss": 0.0644, + "step": 3250 + }, + { + "grad_norm": 0.5934008359909058, + "learning_rate": 6.518000000000001e-05, + "loss": 0.0539, + "step": 3260 + }, + { + "grad_norm": 0.6371035575866699, + "learning_rate": 6.538000000000001e-05, + "loss": 0.0539, + "step": 3270 + }, + { + "grad_norm": 0.517002522945404, + "learning_rate": 6.558e-05, + "loss": 0.0475, + "step": 3280 + }, + { + "grad_norm": 0.49867257475852966, + "learning_rate": 6.578000000000001e-05, + "loss": 0.0479, + "step": 3290 + }, + { + "grad_norm": 0.5001471638679504, + "learning_rate": 6.598e-05, + "loss": 0.0446, + "step": 3300 + }, + { + "grad_norm": 0.5566295981407166, + "learning_rate": 6.618e-05, + "loss": 0.0595, + "step": 3310 + }, + { + "grad_norm": 0.6019293069839478, + "learning_rate": 6.638e-05, + "loss": 0.0638, + "step": 3320 + }, + { + "grad_norm": 0.5611937046051025, + "learning_rate": 6.658e-05, + "loss": 0.0535, + "step": 3330 + }, + { + "grad_norm": 0.5913999676704407, + "learning_rate": 6.678e-05, + "loss": 0.0469, + "step": 3340 + }, + { + "grad_norm": 0.6757261157035828, + "learning_rate": 6.698e-05, + "loss": 0.0491, + "step": 3350 + }, + { + "grad_norm": 1.1269656419754028, + "learning_rate": 6.718e-05, + "loss": 0.0504, + "step": 3360 + }, + { + "grad_norm": 0.6970487833023071, + "learning_rate": 6.738e-05, + "loss": 0.052, + "step": 3370 + }, + { + "grad_norm": 0.635503888130188, + "learning_rate": 6.758e-05, + "loss": 0.0887, + "step": 3380 + }, + { + "grad_norm": 0.6195010542869568, + "learning_rate": 6.778e-05, + "loss": 0.0495, + "step": 3390 + }, + { + "grad_norm": 0.6423906087875366, + "learning_rate": 6.798e-05, + "loss": 0.0576, + "step": 3400 + }, + { + "grad_norm": 0.6626360416412354, + "learning_rate": 6.818e-05, + "loss": 0.0501, + "step": 3410 + }, + { + "grad_norm": 0.5380253791809082, + "learning_rate": 6.838e-05, + "loss": 0.0497, + "step": 3420 + }, + { + "grad_norm": 0.577542781829834, + "learning_rate": 6.858e-05, + "loss": 0.0494, + "step": 3430 + }, + { + "grad_norm": 0.4449599087238312, + "learning_rate": 6.878e-05, + "loss": 0.0565, + "step": 3440 + }, + { + "grad_norm": 0.631966769695282, + "learning_rate": 6.898e-05, + "loss": 0.0466, + "step": 3450 + }, + { + "grad_norm": 0.5737388134002686, + "learning_rate": 6.918e-05, + "loss": 0.046, + "step": 3460 + }, + { + "grad_norm": 0.5362457633018494, + "learning_rate": 6.938e-05, + "loss": 0.0566, + "step": 3470 + }, + { + "grad_norm": 0.4598131775856018, + "learning_rate": 6.958e-05, + "loss": 0.0584, + "step": 3480 + }, + { + "grad_norm": 0.47502726316452026, + "learning_rate": 6.978e-05, + "loss": 0.0558, + "step": 3490 + }, + { + "grad_norm": 0.45879724621772766, + "learning_rate": 6.998e-05, + "loss": 0.042, + "step": 3500 + }, + { + "grad_norm": 0.5175849199295044, + "learning_rate": 7.018e-05, + "loss": 0.0481, + "step": 3510 + }, + { + "grad_norm": 0.5727440714836121, + "learning_rate": 7.038e-05, + "loss": 0.0485, + "step": 3520 + }, + { + "grad_norm": 0.49426883459091187, + "learning_rate": 7.058e-05, + "loss": 0.0475, + "step": 3530 + }, + { + "grad_norm": 0.47619980573654175, + "learning_rate": 7.078e-05, + "loss": 0.0477, + "step": 3540 + }, + { + "grad_norm": 0.47510600090026855, + "learning_rate": 7.098e-05, + "loss": 0.0528, + "step": 3550 + }, + { + "grad_norm": 0.5645313858985901, + "learning_rate": 7.118e-05, + "loss": 0.04, + "step": 3560 + }, + { + "grad_norm": 0.5124518275260925, + "learning_rate": 7.138e-05, + "loss": 0.0528, + "step": 3570 + }, + { + "grad_norm": 0.4331539571285248, + "learning_rate": 7.158e-05, + "loss": 0.0649, + "step": 3580 + }, + { + "grad_norm": 0.565325140953064, + "learning_rate": 7.178000000000001e-05, + "loss": 0.0641, + "step": 3590 + }, + { + "grad_norm": 0.49523457884788513, + "learning_rate": 7.198e-05, + "loss": 0.0531, + "step": 3600 + }, + { + "grad_norm": 0.40296709537506104, + "learning_rate": 7.218e-05, + "loss": 0.0507, + "step": 3610 + }, + { + "grad_norm": 0.49126020073890686, + "learning_rate": 7.238000000000001e-05, + "loss": 0.0478, + "step": 3620 + }, + { + "grad_norm": 0.5226730704307556, + "learning_rate": 7.258e-05, + "loss": 0.045, + "step": 3630 + }, + { + "grad_norm": 0.48106834292411804, + "learning_rate": 7.278e-05, + "loss": 0.0454, + "step": 3640 + }, + { + "grad_norm": 0.45139843225479126, + "learning_rate": 7.298000000000001e-05, + "loss": 0.055, + "step": 3650 + }, + { + "grad_norm": 0.5822842717170715, + "learning_rate": 7.318e-05, + "loss": 0.0498, + "step": 3660 + }, + { + "grad_norm": 0.6017340421676636, + "learning_rate": 7.338e-05, + "loss": 0.0546, + "step": 3670 + }, + { + "grad_norm": 0.6336044669151306, + "learning_rate": 7.358000000000001e-05, + "loss": 0.05, + "step": 3680 + }, + { + "grad_norm": 0.6810202598571777, + "learning_rate": 7.378e-05, + "loss": 0.0478, + "step": 3690 + }, + { + "grad_norm": 0.49058160185813904, + "learning_rate": 7.398e-05, + "loss": 0.0504, + "step": 3700 + }, + { + "grad_norm": 0.4867585003376007, + "learning_rate": 7.418000000000001e-05, + "loss": 0.0471, + "step": 3710 + }, + { + "grad_norm": 0.49752700328826904, + "learning_rate": 7.438e-05, + "loss": 0.0501, + "step": 3720 + }, + { + "grad_norm": 0.41769036650657654, + "learning_rate": 7.458000000000001e-05, + "loss": 0.0433, + "step": 3730 + }, + { + "grad_norm": 0.47804659605026245, + "learning_rate": 7.478e-05, + "loss": 0.0444, + "step": 3740 + }, + { + "grad_norm": 0.46447914838790894, + "learning_rate": 7.498e-05, + "loss": 0.0533, + "step": 3750 + }, + { + "grad_norm": 0.3227159380912781, + "learning_rate": 7.518000000000001e-05, + "loss": 0.0398, + "step": 3760 + }, + { + "grad_norm": 0.6006016731262207, + "learning_rate": 7.538e-05, + "loss": 0.0487, + "step": 3770 + }, + { + "grad_norm": 0.4678070545196533, + "learning_rate": 7.558e-05, + "loss": 0.0486, + "step": 3780 + }, + { + "grad_norm": 0.4607994258403778, + "learning_rate": 7.578000000000001e-05, + "loss": 0.044, + "step": 3790 + }, + { + "grad_norm": 0.5018726587295532, + "learning_rate": 7.598e-05, + "loss": 0.0754, + "step": 3800 + }, + { + "grad_norm": 0.4651670455932617, + "learning_rate": 7.618e-05, + "loss": 0.0438, + "step": 3810 + }, + { + "grad_norm": 0.6074767708778381, + "learning_rate": 7.638000000000001e-05, + "loss": 0.0562, + "step": 3820 + }, + { + "grad_norm": 0.4194856584072113, + "learning_rate": 7.658e-05, + "loss": 0.0438, + "step": 3830 + }, + { + "grad_norm": 0.40190210938453674, + "learning_rate": 7.678000000000001e-05, + "loss": 0.0409, + "step": 3840 + }, + { + "grad_norm": 0.5286075472831726, + "learning_rate": 7.698000000000001e-05, + "loss": 0.0606, + "step": 3850 + }, + { + "grad_norm": 0.5535075068473816, + "learning_rate": 7.718e-05, + "loss": 0.0426, + "step": 3860 + }, + { + "grad_norm": 0.450271338224411, + "learning_rate": 7.738000000000001e-05, + "loss": 0.0452, + "step": 3870 + }, + { + "grad_norm": 0.35586071014404297, + "learning_rate": 7.758000000000001e-05, + "loss": 0.0425, + "step": 3880 + }, + { + "grad_norm": 0.47037267684936523, + "learning_rate": 7.778e-05, + "loss": 0.0479, + "step": 3890 + }, + { + "grad_norm": 0.5718138217926025, + "learning_rate": 7.798000000000001e-05, + "loss": 0.0509, + "step": 3900 + }, + { + "grad_norm": 0.3333280682563782, + "learning_rate": 7.818000000000001e-05, + "loss": 0.0458, + "step": 3910 + }, + { + "grad_norm": 0.8567695617675781, + "learning_rate": 7.838e-05, + "loss": 0.0534, + "step": 3920 + }, + { + "grad_norm": 0.3658080995082855, + "learning_rate": 7.858000000000001e-05, + "loss": 0.0471, + "step": 3930 + }, + { + "grad_norm": 0.49003928899765015, + "learning_rate": 7.878e-05, + "loss": 0.0554, + "step": 3940 + }, + { + "grad_norm": 0.44335323572158813, + "learning_rate": 7.897999999999999e-05, + "loss": 0.0511, + "step": 3950 + }, + { + "grad_norm": 0.3798561990261078, + "learning_rate": 7.918e-05, + "loss": 0.0606, + "step": 3960 + }, + { + "grad_norm": 0.4345288574695587, + "learning_rate": 7.938e-05, + "loss": 0.044, + "step": 3970 + }, + { + "grad_norm": 0.49366435408592224, + "learning_rate": 7.958e-05, + "loss": 0.0507, + "step": 3980 + }, + { + "grad_norm": 0.5321271419525146, + "learning_rate": 7.978e-05, + "loss": 0.0543, + "step": 3990 + }, + { + "grad_norm": 0.6306968927383423, + "learning_rate": 7.998e-05, + "loss": 0.0509, + "step": 4000 + }, + { + "grad_norm": 0.5244865417480469, + "learning_rate": 8.018e-05, + "loss": 0.0467, + "step": 4010 + }, + { + "grad_norm": 0.459025502204895, + "learning_rate": 8.038e-05, + "loss": 0.0499, + "step": 4020 + }, + { + "grad_norm": 0.48619168996810913, + "learning_rate": 8.058e-05, + "loss": 0.0542, + "step": 4030 + }, + { + "grad_norm": 0.47001272439956665, + "learning_rate": 8.078e-05, + "loss": 0.0377, + "step": 4040 + }, + { + "grad_norm": 0.42484578490257263, + "learning_rate": 8.098e-05, + "loss": 0.0457, + "step": 4050 + }, + { + "grad_norm": 0.4668360650539398, + "learning_rate": 8.118e-05, + "loss": 0.0444, + "step": 4060 + }, + { + "grad_norm": 0.36333346366882324, + "learning_rate": 8.138e-05, + "loss": 0.0407, + "step": 4070 + }, + { + "grad_norm": 0.4790932536125183, + "learning_rate": 8.158e-05, + "loss": 0.0518, + "step": 4080 + }, + { + "grad_norm": 0.48865705728530884, + "learning_rate": 8.178e-05, + "loss": 0.0491, + "step": 4090 + }, + { + "grad_norm": 0.510968804359436, + "learning_rate": 8.198e-05, + "loss": 0.0404, + "step": 4100 + }, + { + "grad_norm": 0.5166163444519043, + "learning_rate": 8.218e-05, + "loss": 0.048, + "step": 4110 + }, + { + "grad_norm": 0.7706247568130493, + "learning_rate": 8.238000000000001e-05, + "loss": 0.0546, + "step": 4120 + }, + { + "grad_norm": 0.5259150862693787, + "learning_rate": 8.258e-05, + "loss": 0.0454, + "step": 4130 + }, + { + "grad_norm": 0.36422306299209595, + "learning_rate": 8.278e-05, + "loss": 0.0438, + "step": 4140 + }, + { + "grad_norm": 0.39548495411872864, + "learning_rate": 8.298000000000001e-05, + "loss": 0.0548, + "step": 4150 + }, + { + "grad_norm": 0.4889143109321594, + "learning_rate": 8.318e-05, + "loss": 0.0436, + "step": 4160 + }, + { + "grad_norm": 0.4294140934944153, + "learning_rate": 8.338e-05, + "loss": 0.0426, + "step": 4170 + }, + { + "grad_norm": 0.3817083239555359, + "learning_rate": 8.358e-05, + "loss": 0.0429, + "step": 4180 + }, + { + "grad_norm": 0.4154946804046631, + "learning_rate": 8.378e-05, + "loss": 0.0444, + "step": 4190 + }, + { + "grad_norm": 0.4453136920928955, + "learning_rate": 8.398e-05, + "loss": 0.0427, + "step": 4200 + }, + { + "grad_norm": 0.41065388917922974, + "learning_rate": 8.418e-05, + "loss": 0.041, + "step": 4210 + }, + { + "grad_norm": 0.41612672805786133, + "learning_rate": 8.438e-05, + "loss": 0.0543, + "step": 4220 + }, + { + "grad_norm": 0.5291852951049805, + "learning_rate": 8.458e-05, + "loss": 0.0638, + "step": 4230 + }, + { + "grad_norm": 0.4191744923591614, + "learning_rate": 8.478e-05, + "loss": 0.0544, + "step": 4240 + }, + { + "grad_norm": 0.3190278708934784, + "learning_rate": 8.498e-05, + "loss": 0.0528, + "step": 4250 + }, + { + "grad_norm": 0.43338626623153687, + "learning_rate": 8.518000000000001e-05, + "loss": 0.0507, + "step": 4260 + }, + { + "grad_norm": 0.38887307047843933, + "learning_rate": 8.538e-05, + "loss": 0.0462, + "step": 4270 + }, + { + "grad_norm": 0.47358450293540955, + "learning_rate": 8.558e-05, + "loss": 0.0433, + "step": 4280 + }, + { + "grad_norm": 0.4823721945285797, + "learning_rate": 8.578000000000001e-05, + "loss": 0.0447, + "step": 4290 + }, + { + "grad_norm": 0.3531292676925659, + "learning_rate": 8.598e-05, + "loss": 0.0396, + "step": 4300 + }, + { + "grad_norm": 0.4482009708881378, + "learning_rate": 8.618e-05, + "loss": 0.0453, + "step": 4310 + }, + { + "grad_norm": 0.5934690237045288, + "learning_rate": 8.638000000000001e-05, + "loss": 0.0711, + "step": 4320 + }, + { + "grad_norm": 0.4615998864173889, + "learning_rate": 8.658e-05, + "loss": 0.046, + "step": 4330 + }, + { + "grad_norm": 0.38178497552871704, + "learning_rate": 8.678e-05, + "loss": 0.0418, + "step": 4340 + }, + { + "grad_norm": 0.4002855122089386, + "learning_rate": 8.698000000000001e-05, + "loss": 0.0358, + "step": 4350 + }, + { + "grad_norm": 0.37864506244659424, + "learning_rate": 8.718e-05, + "loss": 0.039, + "step": 4360 + }, + { + "grad_norm": 0.4157409965991974, + "learning_rate": 8.738000000000001e-05, + "loss": 0.0458, + "step": 4370 + }, + { + "grad_norm": 0.43442434072494507, + "learning_rate": 8.758000000000001e-05, + "loss": 0.0472, + "step": 4380 + }, + { + "grad_norm": 0.48835688829421997, + "learning_rate": 8.778e-05, + "loss": 0.0398, + "step": 4390 + }, + { + "grad_norm": 0.48423928022384644, + "learning_rate": 8.798000000000001e-05, + "loss": 0.0554, + "step": 4400 + }, + { + "grad_norm": 0.47621750831604004, + "learning_rate": 8.818000000000001e-05, + "loss": 0.0448, + "step": 4410 + }, + { + "grad_norm": 0.4389837384223938, + "learning_rate": 8.838e-05, + "loss": 0.0596, + "step": 4420 + }, + { + "grad_norm": 0.3592231273651123, + "learning_rate": 8.858000000000001e-05, + "loss": 0.0501, + "step": 4430 + }, + { + "grad_norm": 0.48489102721214294, + "learning_rate": 8.878000000000001e-05, + "loss": 0.0507, + "step": 4440 + }, + { + "grad_norm": 0.3971380889415741, + "learning_rate": 8.898e-05, + "loss": 0.0587, + "step": 4450 + }, + { + "grad_norm": 0.49518850445747375, + "learning_rate": 8.918000000000001e-05, + "loss": 0.0493, + "step": 4460 + }, + { + "grad_norm": 0.4285871982574463, + "learning_rate": 8.938e-05, + "loss": 0.0463, + "step": 4470 + }, + { + "grad_norm": 0.4088277518749237, + "learning_rate": 8.958e-05, + "loss": 0.0517, + "step": 4480 + }, + { + "grad_norm": 0.4277302324771881, + "learning_rate": 8.978000000000001e-05, + "loss": 0.0506, + "step": 4490 + }, + { + "grad_norm": 0.46519720554351807, + "learning_rate": 8.998e-05, + "loss": 0.0495, + "step": 4500 + }, + { + "grad_norm": 0.4068399667739868, + "learning_rate": 9.018000000000001e-05, + "loss": 0.0687, + "step": 4510 + }, + { + "grad_norm": 0.49455416202545166, + "learning_rate": 9.038000000000001e-05, + "loss": 0.0508, + "step": 4520 + }, + { + "grad_norm": 0.4416707158088684, + "learning_rate": 9.058e-05, + "loss": 0.0481, + "step": 4530 + }, + { + "grad_norm": 0.5990498661994934, + "learning_rate": 9.078000000000001e-05, + "loss": 0.0635, + "step": 4540 + }, + { + "grad_norm": 0.5360383987426758, + "learning_rate": 9.098000000000001e-05, + "loss": 0.0612, + "step": 4550 + }, + { + "grad_norm": 0.37586185336112976, + "learning_rate": 9.118e-05, + "loss": 0.0473, + "step": 4560 + }, + { + "grad_norm": 0.40508490800857544, + "learning_rate": 9.138e-05, + "loss": 0.0533, + "step": 4570 + }, + { + "grad_norm": 0.8519114851951599, + "learning_rate": 9.158e-05, + "loss": 0.0468, + "step": 4580 + }, + { + "grad_norm": 0.4385341703891754, + "learning_rate": 9.178e-05, + "loss": 0.0473, + "step": 4590 + }, + { + "grad_norm": 0.47749242186546326, + "learning_rate": 9.198e-05, + "loss": 0.0507, + "step": 4600 + }, + { + "grad_norm": 0.4829212427139282, + "learning_rate": 9.218e-05, + "loss": 0.0499, + "step": 4610 + }, + { + "grad_norm": 0.4476567804813385, + "learning_rate": 9.238e-05, + "loss": 0.0417, + "step": 4620 + }, + { + "grad_norm": 0.5656363368034363, + "learning_rate": 9.258e-05, + "loss": 0.0546, + "step": 4630 + }, + { + "grad_norm": 0.5048646330833435, + "learning_rate": 9.278e-05, + "loss": 0.0516, + "step": 4640 + }, + { + "grad_norm": 0.540164053440094, + "learning_rate": 9.298e-05, + "loss": 0.0421, + "step": 4650 + }, + { + "grad_norm": 0.41845086216926575, + "learning_rate": 9.318e-05, + "loss": 0.0478, + "step": 4660 + }, + { + "grad_norm": 0.6415730118751526, + "learning_rate": 9.338e-05, + "loss": 0.0447, + "step": 4670 + }, + { + "grad_norm": 0.35450783371925354, + "learning_rate": 9.358e-05, + "loss": 0.0461, + "step": 4680 + }, + { + "grad_norm": 0.48354604840278625, + "learning_rate": 9.378e-05, + "loss": 0.0531, + "step": 4690 + }, + { + "grad_norm": 0.4911448061466217, + "learning_rate": 9.398e-05, + "loss": 0.0473, + "step": 4700 + }, + { + "grad_norm": 0.454332172870636, + "learning_rate": 9.418e-05, + "loss": 0.0435, + "step": 4710 + }, + { + "grad_norm": 0.42084118723869324, + "learning_rate": 9.438e-05, + "loss": 0.0426, + "step": 4720 + }, + { + "grad_norm": 0.30459684133529663, + "learning_rate": 9.458e-05, + "loss": 0.0533, + "step": 4730 + }, + { + "grad_norm": 0.3999723792076111, + "learning_rate": 9.478e-05, + "loss": 0.0519, + "step": 4740 + }, + { + "grad_norm": 0.3873731195926666, + "learning_rate": 9.498e-05, + "loss": 0.0415, + "step": 4750 + }, + { + "grad_norm": 0.3257777690887451, + "learning_rate": 9.518000000000001e-05, + "loss": 0.0458, + "step": 4760 + }, + { + "grad_norm": 0.571869969367981, + "learning_rate": 9.538e-05, + "loss": 0.0487, + "step": 4770 + }, + { + "grad_norm": 0.5172654986381531, + "learning_rate": 9.558e-05, + "loss": 0.0561, + "step": 4780 + }, + { + "grad_norm": 0.40108487010002136, + "learning_rate": 9.578000000000001e-05, + "loss": 0.0411, + "step": 4790 + }, + { + "grad_norm": 0.2834753394126892, + "learning_rate": 9.598e-05, + "loss": 0.043, + "step": 4800 + }, + { + "grad_norm": 0.3625306487083435, + "learning_rate": 9.618e-05, + "loss": 0.0444, + "step": 4810 + }, + { + "grad_norm": 0.3061979115009308, + "learning_rate": 9.638000000000001e-05, + "loss": 0.048, + "step": 4820 + }, + { + "grad_norm": 0.43342238664627075, + "learning_rate": 9.658e-05, + "loss": 0.0459, + "step": 4830 + }, + { + "grad_norm": 0.45829254388809204, + "learning_rate": 9.678e-05, + "loss": 0.0469, + "step": 4840 + }, + { + "grad_norm": 0.4165900945663452, + "learning_rate": 9.698000000000001e-05, + "loss": 0.0428, + "step": 4850 + }, + { + "grad_norm": 0.3381836712360382, + "learning_rate": 9.718e-05, + "loss": 0.0374, + "step": 4860 + }, + { + "grad_norm": 0.4262806177139282, + "learning_rate": 9.738e-05, + "loss": 0.0427, + "step": 4870 + }, + { + "grad_norm": 0.35872822999954224, + "learning_rate": 9.758000000000001e-05, + "loss": 0.0392, + "step": 4880 + }, + { + "grad_norm": 0.5044097304344177, + "learning_rate": 9.778e-05, + "loss": 0.0571, + "step": 4890 + }, + { + "grad_norm": 0.593035876750946, + "learning_rate": 9.798000000000001e-05, + "loss": 0.0391, + "step": 4900 + }, + { + "grad_norm": 0.4018821716308594, + "learning_rate": 9.818000000000001e-05, + "loss": 0.0495, + "step": 4910 + }, + { + "grad_norm": 0.36531198024749756, + "learning_rate": 9.838e-05, + "loss": 0.0461, + "step": 4920 + }, + { + "grad_norm": 0.3901344835758209, + "learning_rate": 9.858000000000001e-05, + "loss": 0.0447, + "step": 4930 + }, + { + "grad_norm": 0.3914910554885864, + "learning_rate": 9.878e-05, + "loss": 0.0417, + "step": 4940 + }, + { + "grad_norm": 0.4712642729282379, + "learning_rate": 9.898e-05, + "loss": 0.0456, + "step": 4950 + }, + { + "grad_norm": 0.3110232651233673, + "learning_rate": 9.918000000000001e-05, + "loss": 0.0429, + "step": 4960 + }, + { + "grad_norm": 0.38908815383911133, + "learning_rate": 9.938e-05, + "loss": 0.0521, + "step": 4970 + }, + { + "grad_norm": 0.350799560546875, + "learning_rate": 9.958e-05, + "loss": 0.0484, + "step": 4980 + }, + { + "grad_norm": 0.34600725769996643, + "learning_rate": 9.978000000000001e-05, + "loss": 0.0429, + "step": 4990 + }, + { + "grad_norm": 0.44985055923461914, + "learning_rate": 9.998e-05, + "loss": 0.053, + "step": 5000 + }, + { + "grad_norm": 0.4457371234893799, + "learning_rate": 9.999999778549045e-05, + "loss": 0.0444, + "step": 5010 + }, + { + "grad_norm": 0.4866776764392853, + "learning_rate": 9.999999013039593e-05, + "loss": 0.0419, + "step": 5020 + }, + { + "grad_norm": 0.4193825125694275, + "learning_rate": 9.999997700737766e-05, + "loss": 0.04, + "step": 5030 + }, + { + "grad_norm": 0.3338404595851898, + "learning_rate": 9.999995841643709e-05, + "loss": 0.0399, + "step": 5040 + }, + { + "grad_norm": 0.3945288062095642, + "learning_rate": 9.999993435757623e-05, + "loss": 0.0448, + "step": 5050 + }, + { + "grad_norm": 0.308926522731781, + "learning_rate": 9.999990483079773e-05, + "loss": 0.0346, + "step": 5060 + }, + { + "grad_norm": 0.3824428617954254, + "learning_rate": 9.999986983610481e-05, + "loss": 0.0536, + "step": 5070 + }, + { + "grad_norm": 0.32512784004211426, + "learning_rate": 9.99998293735013e-05, + "loss": 0.0498, + "step": 5080 + }, + { + "grad_norm": 0.2981336712837219, + "learning_rate": 9.999978344299161e-05, + "loss": 0.0411, + "step": 5090 + }, + { + "grad_norm": 0.36756810545921326, + "learning_rate": 9.99997320445808e-05, + "loss": 0.0417, + "step": 5100 + }, + { + "grad_norm": 0.44123566150665283, + "learning_rate": 9.999967517827444e-05, + "loss": 0.0478, + "step": 5110 + }, + { + "grad_norm": 0.4048570692539215, + "learning_rate": 9.999961284407879e-05, + "loss": 0.0558, + "step": 5120 + }, + { + "grad_norm": 0.41541656851768494, + "learning_rate": 9.999954504200067e-05, + "loss": 0.0532, + "step": 5130 + }, + { + "grad_norm": 0.370635449886322, + "learning_rate": 9.999947177204744e-05, + "loss": 0.0438, + "step": 5140 + }, + { + "grad_norm": 0.4047102928161621, + "learning_rate": 9.999939303422718e-05, + "loss": 0.039, + "step": 5150 + }, + { + "grad_norm": 0.3960151672363281, + "learning_rate": 9.999930882854847e-05, + "loss": 0.0584, + "step": 5160 + }, + { + "grad_norm": 0.35693055391311646, + "learning_rate": 9.999921915502051e-05, + "loss": 0.0462, + "step": 5170 + }, + { + "grad_norm": 0.35383063554763794, + "learning_rate": 9.99991240136531e-05, + "loss": 0.0397, + "step": 5180 + }, + { + "grad_norm": 0.38076698780059814, + "learning_rate": 9.999902340445668e-05, + "loss": 0.0598, + "step": 5190 + }, + { + "grad_norm": 0.2750827372074127, + "learning_rate": 9.999891732744224e-05, + "loss": 0.0368, + "step": 5200 + }, + { + "grad_norm": 0.4872633218765259, + "learning_rate": 9.999880578262135e-05, + "loss": 0.0456, + "step": 5210 + }, + { + "grad_norm": 0.38172680139541626, + "learning_rate": 9.999868877000624e-05, + "loss": 0.0427, + "step": 5220 + }, + { + "grad_norm": 0.33038702607154846, + "learning_rate": 9.99985662896097e-05, + "loss": 0.0417, + "step": 5230 + }, + { + "grad_norm": 0.3226548135280609, + "learning_rate": 9.999843834144513e-05, + "loss": 0.0405, + "step": 5240 + }, + { + "grad_norm": 0.3860369920730591, + "learning_rate": 9.99983049255265e-05, + "loss": 0.0513, + "step": 5250 + }, + { + "grad_norm": 0.34384334087371826, + "learning_rate": 9.999816604186843e-05, + "loss": 0.0393, + "step": 5260 + }, + { + "grad_norm": 0.38275647163391113, + "learning_rate": 9.999802169048609e-05, + "loss": 0.0418, + "step": 5270 + }, + { + "grad_norm": 0.4732551574707031, + "learning_rate": 9.999787187139527e-05, + "loss": 0.0453, + "step": 5280 + }, + { + "grad_norm": 0.34008124470710754, + "learning_rate": 9.999771658461234e-05, + "loss": 0.0468, + "step": 5290 + }, + { + "grad_norm": 0.34717661142349243, + "learning_rate": 9.999755583015431e-05, + "loss": 0.0369, + "step": 5300 + }, + { + "grad_norm": 0.3388221263885498, + "learning_rate": 9.999738960803874e-05, + "loss": 0.0344, + "step": 5310 + }, + { + "grad_norm": 0.37457239627838135, + "learning_rate": 9.99972179182838e-05, + "loss": 0.0408, + "step": 5320 + }, + { + "grad_norm": 0.46544578671455383, + "learning_rate": 9.99970407609083e-05, + "loss": 0.0561, + "step": 5330 + }, + { + "grad_norm": 0.3946884572505951, + "learning_rate": 9.999685813593159e-05, + "loss": 0.0443, + "step": 5340 + }, + { + "grad_norm": 0.2683626413345337, + "learning_rate": 9.999667004337362e-05, + "loss": 0.0374, + "step": 5350 + }, + { + "grad_norm": 0.28631749749183655, + "learning_rate": 9.9996476483255e-05, + "loss": 0.0413, + "step": 5360 + }, + { + "grad_norm": 0.36079248785972595, + "learning_rate": 9.999627745559688e-05, + "loss": 0.0356, + "step": 5370 + }, + { + "grad_norm": 0.4213111102581024, + "learning_rate": 9.999607296042101e-05, + "loss": 0.0383, + "step": 5380 + }, + { + "grad_norm": 0.31158196926116943, + "learning_rate": 9.99958629977498e-05, + "loss": 0.0375, + "step": 5390 + }, + { + "grad_norm": 0.27046331763267517, + "learning_rate": 9.999564756760615e-05, + "loss": 0.0405, + "step": 5400 + }, + { + "grad_norm": 0.27969858050346375, + "learning_rate": 9.999542667001366e-05, + "loss": 0.0333, + "step": 5410 + }, + { + "grad_norm": 0.43641960620880127, + "learning_rate": 9.999520030499647e-05, + "loss": 0.0389, + "step": 5420 + }, + { + "grad_norm": 0.3843284249305725, + "learning_rate": 9.999496847257936e-05, + "loss": 0.0377, + "step": 5430 + }, + { + "grad_norm": 0.3445203900337219, + "learning_rate": 9.999473117278764e-05, + "loss": 0.0475, + "step": 5440 + }, + { + "grad_norm": 0.3761952519416809, + "learning_rate": 9.999448840564731e-05, + "loss": 0.0322, + "step": 5450 + }, + { + "grad_norm": 0.3402634561061859, + "learning_rate": 9.999424017118488e-05, + "loss": 0.036, + "step": 5460 + }, + { + "grad_norm": 0.3042530119419098, + "learning_rate": 9.999398646942751e-05, + "loss": 0.0326, + "step": 5470 + }, + { + "grad_norm": 0.3125622570514679, + "learning_rate": 9.999372730040296e-05, + "loss": 0.0367, + "step": 5480 + }, + { + "grad_norm": 0.28172048926353455, + "learning_rate": 9.999346266413953e-05, + "loss": 0.0349, + "step": 5490 + }, + { + "grad_norm": 0.3005838394165039, + "learning_rate": 9.99931925606662e-05, + "loss": 0.037, + "step": 5500 + }, + { + "grad_norm": 0.3403024971485138, + "learning_rate": 9.99929169900125e-05, + "loss": 0.0397, + "step": 5510 + }, + { + "grad_norm": 0.2735153138637543, + "learning_rate": 9.999263595220855e-05, + "loss": 0.0353, + "step": 5520 + }, + { + "grad_norm": 0.31464219093322754, + "learning_rate": 9.99923494472851e-05, + "loss": 0.0371, + "step": 5530 + }, + { + "grad_norm": 0.3070959448814392, + "learning_rate": 9.999205747527348e-05, + "loss": 0.032, + "step": 5540 + }, + { + "grad_norm": 0.3369695544242859, + "learning_rate": 9.999176003620561e-05, + "loss": 0.035, + "step": 5550 + }, + { + "grad_norm": 0.3155800998210907, + "learning_rate": 9.999145713011405e-05, + "loss": 0.0359, + "step": 5560 + }, + { + "grad_norm": 0.36522960662841797, + "learning_rate": 9.999114875703186e-05, + "loss": 0.0381, + "step": 5570 + }, + { + "grad_norm": 0.46231067180633545, + "learning_rate": 9.999083491699281e-05, + "loss": 0.0482, + "step": 5580 + }, + { + "grad_norm": 0.43577179312705994, + "learning_rate": 9.999051561003123e-05, + "loss": 0.0361, + "step": 5590 + }, + { + "grad_norm": 0.3758983910083771, + "learning_rate": 9.999019083618202e-05, + "loss": 0.043, + "step": 5600 + }, + { + "grad_norm": 0.394493967294693, + "learning_rate": 9.99898605954807e-05, + "loss": 0.0379, + "step": 5610 + }, + { + "grad_norm": 0.2872087359428406, + "learning_rate": 9.998952488796338e-05, + "loss": 0.0367, + "step": 5620 + }, + { + "grad_norm": 0.2769279479980469, + "learning_rate": 9.998918371366676e-05, + "loss": 0.0342, + "step": 5630 + }, + { + "grad_norm": 0.2927900552749634, + "learning_rate": 9.99888370726282e-05, + "loss": 0.0358, + "step": 5640 + }, + { + "grad_norm": 0.28604426980018616, + "learning_rate": 9.998848496488556e-05, + "loss": 0.0513, + "step": 5650 + }, + { + "grad_norm": 0.29099681973457336, + "learning_rate": 9.998812739047736e-05, + "loss": 0.0377, + "step": 5660 + }, + { + "grad_norm": 0.35573118925094604, + "learning_rate": 9.99877643494427e-05, + "loss": 0.0376, + "step": 5670 + }, + { + "grad_norm": 0.3163166046142578, + "learning_rate": 9.998739584182128e-05, + "loss": 0.0447, + "step": 5680 + }, + { + "grad_norm": 0.4396672248840332, + "learning_rate": 9.998702186765342e-05, + "loss": 0.0404, + "step": 5690 + }, + { + "grad_norm": 0.3494808077812195, + "learning_rate": 9.998664242698e-05, + "loss": 0.0393, + "step": 5700 + }, + { + "grad_norm": 0.38465186953544617, + "learning_rate": 9.998625751984251e-05, + "loss": 0.0451, + "step": 5710 + }, + { + "grad_norm": 0.328311562538147, + "learning_rate": 9.998586714628307e-05, + "loss": 0.0375, + "step": 5720 + }, + { + "grad_norm": 0.29680177569389343, + "learning_rate": 9.998547130634432e-05, + "loss": 0.0377, + "step": 5730 + }, + { + "grad_norm": 0.31583890318870544, + "learning_rate": 9.99850700000696e-05, + "loss": 0.0438, + "step": 5740 + }, + { + "grad_norm": 0.26687678694725037, + "learning_rate": 9.998466322750278e-05, + "loss": 0.0348, + "step": 5750 + }, + { + "grad_norm": 0.3553796708583832, + "learning_rate": 9.998425098868834e-05, + "loss": 0.0389, + "step": 5760 + }, + { + "grad_norm": 0.38707616925239563, + "learning_rate": 9.998383328367136e-05, + "loss": 0.0395, + "step": 5770 + }, + { + "grad_norm": 0.3412174880504608, + "learning_rate": 9.99834101124975e-05, + "loss": 0.0436, + "step": 5780 + }, + { + "grad_norm": 0.2778187692165375, + "learning_rate": 9.998298147521309e-05, + "loss": 0.0373, + "step": 5790 + }, + { + "grad_norm": 0.3315787613391876, + "learning_rate": 9.998254737186496e-05, + "loss": 0.0457, + "step": 5800 + }, + { + "grad_norm": 0.32647934556007385, + "learning_rate": 9.99821078025006e-05, + "loss": 0.0431, + "step": 5810 + }, + { + "grad_norm": 0.3641613721847534, + "learning_rate": 9.998166276716807e-05, + "loss": 0.0377, + "step": 5820 + }, + { + "grad_norm": 0.3695627748966217, + "learning_rate": 9.998121226591606e-05, + "loss": 0.0407, + "step": 5830 + }, + { + "grad_norm": 0.35932645201683044, + "learning_rate": 9.998075629879382e-05, + "loss": 0.0392, + "step": 5840 + }, + { + "grad_norm": 0.3038875162601471, + "learning_rate": 9.99802948658512e-05, + "loss": 0.032, + "step": 5850 + }, + { + "grad_norm": 0.20721293985843658, + "learning_rate": 9.99798279671387e-05, + "loss": 0.042, + "step": 5860 + }, + { + "grad_norm": 0.3722662925720215, + "learning_rate": 9.997935560270734e-05, + "loss": 0.0381, + "step": 5870 + }, + { + "grad_norm": 0.4159487783908844, + "learning_rate": 9.997887777260879e-05, + "loss": 0.0471, + "step": 5880 + }, + { + "grad_norm": 0.3565455377101898, + "learning_rate": 9.997839447689532e-05, + "loss": 0.0453, + "step": 5890 + }, + { + "grad_norm": 0.3338732123374939, + "learning_rate": 9.997790571561978e-05, + "loss": 0.0601, + "step": 5900 + }, + { + "grad_norm": 0.33086299896240234, + "learning_rate": 9.99774114888356e-05, + "loss": 0.0483, + "step": 5910 + }, + { + "grad_norm": 0.3562318682670593, + "learning_rate": 9.997691179659684e-05, + "loss": 0.0383, + "step": 5920 + }, + { + "grad_norm": 0.2637366056442261, + "learning_rate": 9.997640663895815e-05, + "loss": 0.0369, + "step": 5930 + }, + { + "grad_norm": 0.2633378505706787, + "learning_rate": 9.997589601597477e-05, + "loss": 0.0412, + "step": 5940 + }, + { + "grad_norm": 0.4011734426021576, + "learning_rate": 9.997537992770252e-05, + "loss": 0.0396, + "step": 5950 + }, + { + "grad_norm": 0.362652063369751, + "learning_rate": 9.997485837419788e-05, + "loss": 0.0352, + "step": 5960 + }, + { + "grad_norm": 0.41802698373794556, + "learning_rate": 9.997433135551786e-05, + "loss": 0.0448, + "step": 5970 + }, + { + "grad_norm": 0.31231939792633057, + "learning_rate": 9.997379887172009e-05, + "loss": 0.0423, + "step": 5980 + }, + { + "grad_norm": 0.5096443295478821, + "learning_rate": 9.997326092286281e-05, + "loss": 0.0527, + "step": 5990 + }, + { + "grad_norm": 0.2742319405078888, + "learning_rate": 9.997271750900486e-05, + "loss": 0.0582, + "step": 6000 + }, + { + "grad_norm": 0.29416942596435547, + "learning_rate": 9.997216863020565e-05, + "loss": 0.0418, + "step": 6010 + }, + { + "grad_norm": 0.25156641006469727, + "learning_rate": 9.99716142865252e-05, + "loss": 0.0409, + "step": 6020 + }, + { + "grad_norm": 0.33867189288139343, + "learning_rate": 9.997105447802415e-05, + "loss": 0.0397, + "step": 6030 + }, + { + "grad_norm": 0.3623952865600586, + "learning_rate": 9.997048920476373e-05, + "loss": 0.0387, + "step": 6040 + }, + { + "grad_norm": 0.39471375942230225, + "learning_rate": 9.996991846680572e-05, + "loss": 0.0451, + "step": 6050 + }, + { + "grad_norm": 0.3980799615383148, + "learning_rate": 9.996934226421257e-05, + "loss": 0.0384, + "step": 6060 + }, + { + "grad_norm": 0.41620194911956787, + "learning_rate": 9.996876059704726e-05, + "loss": 0.0416, + "step": 6070 + }, + { + "grad_norm": 0.2997787892818451, + "learning_rate": 9.996817346537343e-05, + "loss": 0.0425, + "step": 6080 + }, + { + "grad_norm": 0.39817532896995544, + "learning_rate": 9.996758086925526e-05, + "loss": 0.0424, + "step": 6090 + }, + { + "grad_norm": 0.3052063286304474, + "learning_rate": 9.996698280875759e-05, + "loss": 0.0459, + "step": 6100 + }, + { + "grad_norm": 0.28340935707092285, + "learning_rate": 9.99663792839458e-05, + "loss": 0.0321, + "step": 6110 + }, + { + "grad_norm": 0.3321530818939209, + "learning_rate": 9.99657702948859e-05, + "loss": 0.0413, + "step": 6120 + }, + { + "grad_norm": 0.3218732178211212, + "learning_rate": 9.996515584164448e-05, + "loss": 0.0336, + "step": 6130 + }, + { + "grad_norm": 0.2716471552848816, + "learning_rate": 9.996453592428873e-05, + "loss": 0.0364, + "step": 6140 + }, + { + "grad_norm": 0.2865748703479767, + "learning_rate": 9.996391054288646e-05, + "loss": 0.0584, + "step": 6150 + }, + { + "grad_norm": 0.3972782492637634, + "learning_rate": 9.996327969750605e-05, + "loss": 0.058, + "step": 6160 + }, + { + "grad_norm": 0.36359286308288574, + "learning_rate": 9.996264338821649e-05, + "loss": 0.0507, + "step": 6170 + }, + { + "grad_norm": 0.3013223707675934, + "learning_rate": 9.996200161508735e-05, + "loss": 0.0391, + "step": 6180 + }, + { + "grad_norm": 0.2465677708387375, + "learning_rate": 9.996135437818885e-05, + "loss": 0.0485, + "step": 6190 + }, + { + "grad_norm": 0.3803187906742096, + "learning_rate": 9.996070167759175e-05, + "loss": 0.0399, + "step": 6200 + }, + { + "grad_norm": 0.34350550174713135, + "learning_rate": 9.996004351336743e-05, + "loss": 0.0414, + "step": 6210 + }, + { + "grad_norm": 0.46143826842308044, + "learning_rate": 9.995937988558785e-05, + "loss": 0.0365, + "step": 6220 + }, + { + "grad_norm": 0.31243279576301575, + "learning_rate": 9.995871079432561e-05, + "loss": 0.0402, + "step": 6230 + }, + { + "grad_norm": 0.2912810444831848, + "learning_rate": 9.995803623965389e-05, + "loss": 0.0424, + "step": 6240 + }, + { + "grad_norm": 0.3197357952594757, + "learning_rate": 9.995735622164641e-05, + "loss": 0.0447, + "step": 6250 + }, + { + "grad_norm": 0.35028401017189026, + "learning_rate": 9.995667074037758e-05, + "loss": 0.0318, + "step": 6260 + }, + { + "grad_norm": 0.3003797233104706, + "learning_rate": 9.995597979592232e-05, + "loss": 0.0405, + "step": 6270 + }, + { + "grad_norm": 0.25014162063598633, + "learning_rate": 9.995528338835625e-05, + "loss": 0.0382, + "step": 6280 + }, + { + "grad_norm": 0.2773245871067047, + "learning_rate": 9.995458151775547e-05, + "loss": 0.0357, + "step": 6290 + }, + { + "grad_norm": 0.3372359275817871, + "learning_rate": 9.995387418419677e-05, + "loss": 0.0402, + "step": 6300 + }, + { + "grad_norm": 0.3421248197555542, + "learning_rate": 9.99531613877575e-05, + "loss": 0.0348, + "step": 6310 + }, + { + "grad_norm": 0.30623844265937805, + "learning_rate": 9.995244312851559e-05, + "loss": 0.0336, + "step": 6320 + }, + { + "grad_norm": 0.3319791853427887, + "learning_rate": 9.995171940654961e-05, + "loss": 0.0351, + "step": 6330 + }, + { + "grad_norm": 0.40112826228141785, + "learning_rate": 9.995099022193871e-05, + "loss": 0.0485, + "step": 6340 + }, + { + "grad_norm": 0.518397331237793, + "learning_rate": 9.995025557476261e-05, + "loss": 0.04, + "step": 6350 + }, + { + "grad_norm": 0.37978556752204895, + "learning_rate": 9.994951546510165e-05, + "loss": 0.0399, + "step": 6360 + }, + { + "grad_norm": 0.31037217378616333, + "learning_rate": 9.994876989303679e-05, + "loss": 0.036, + "step": 6370 + }, + { + "grad_norm": 0.3502160906791687, + "learning_rate": 9.994801885864955e-05, + "loss": 0.0371, + "step": 6380 + }, + { + "grad_norm": 0.3823016583919525, + "learning_rate": 9.994726236202205e-05, + "loss": 0.0424, + "step": 6390 + }, + { + "grad_norm": 0.32516252994537354, + "learning_rate": 9.994650040323704e-05, + "loss": 0.0439, + "step": 6400 + }, + { + "grad_norm": 0.3373061418533325, + "learning_rate": 9.994573298237784e-05, + "loss": 0.0319, + "step": 6410 + }, + { + "grad_norm": 0.27572333812713623, + "learning_rate": 9.994496009952837e-05, + "loss": 0.0406, + "step": 6420 + }, + { + "grad_norm": 0.2845383584499359, + "learning_rate": 9.994418175477316e-05, + "loss": 0.0369, + "step": 6430 + }, + { + "grad_norm": 0.26869192719459534, + "learning_rate": 9.994339794819733e-05, + "loss": 0.0356, + "step": 6440 + }, + { + "grad_norm": 0.3009776473045349, + "learning_rate": 9.994260867988658e-05, + "loss": 0.0353, + "step": 6450 + }, + { + "grad_norm": 0.30609941482543945, + "learning_rate": 9.994181394992723e-05, + "loss": 0.0339, + "step": 6460 + }, + { + "grad_norm": 0.2574935853481293, + "learning_rate": 9.994101375840618e-05, + "loss": 0.0368, + "step": 6470 + }, + { + "grad_norm": 0.24934887886047363, + "learning_rate": 9.994020810541098e-05, + "loss": 0.0338, + "step": 6480 + }, + { + "grad_norm": 0.3328988254070282, + "learning_rate": 9.99393969910297e-05, + "loss": 0.0423, + "step": 6490 + }, + { + "grad_norm": 0.2232816368341446, + "learning_rate": 9.993858041535104e-05, + "loss": 0.0322, + "step": 6500 + }, + { + "grad_norm": 0.2760820686817169, + "learning_rate": 9.99377583784643e-05, + "loss": 0.0333, + "step": 6510 + }, + { + "grad_norm": 0.28666985034942627, + "learning_rate": 9.993693088045939e-05, + "loss": 0.0371, + "step": 6520 + }, + { + "grad_norm": 0.31116101145744324, + "learning_rate": 9.99360979214268e-05, + "loss": 0.0371, + "step": 6530 + }, + { + "grad_norm": 0.40249717235565186, + "learning_rate": 9.99352595014576e-05, + "loss": 0.043, + "step": 6540 + }, + { + "grad_norm": 0.4444715976715088, + "learning_rate": 9.993441562064354e-05, + "loss": 0.0354, + "step": 6550 + }, + { + "grad_norm": 0.3594377338886261, + "learning_rate": 9.993356627907685e-05, + "loss": 0.0405, + "step": 6560 + }, + { + "grad_norm": 0.44024428725242615, + "learning_rate": 9.99327114768504e-05, + "loss": 0.0443, + "step": 6570 + }, + { + "grad_norm": 0.31277918815612793, + "learning_rate": 9.99318512140577e-05, + "loss": 0.0319, + "step": 6580 + }, + { + "grad_norm": 0.26775193214416504, + "learning_rate": 9.993098549079284e-05, + "loss": 0.0449, + "step": 6590 + }, + { + "grad_norm": 0.287387877702713, + "learning_rate": 9.993011430715047e-05, + "loss": 0.0442, + "step": 6600 + }, + { + "grad_norm": 0.2790446877479553, + "learning_rate": 9.992923766322586e-05, + "loss": 0.0335, + "step": 6610 + }, + { + "grad_norm": 0.26064810156822205, + "learning_rate": 9.99283555591149e-05, + "loss": 0.0319, + "step": 6620 + }, + { + "grad_norm": 0.2634753882884979, + "learning_rate": 9.992746799491404e-05, + "loss": 0.0341, + "step": 6630 + }, + { + "grad_norm": 0.2597951889038086, + "learning_rate": 9.992657497072033e-05, + "loss": 0.0292, + "step": 6640 + }, + { + "grad_norm": 0.3425586521625519, + "learning_rate": 9.992567648663147e-05, + "loss": 0.044, + "step": 6650 + }, + { + "grad_norm": 0.3029065430164337, + "learning_rate": 9.992477254274568e-05, + "loss": 0.0321, + "step": 6660 + }, + { + "grad_norm": 0.23979894816875458, + "learning_rate": 9.992386313916183e-05, + "loss": 0.0295, + "step": 6670 + }, + { + "grad_norm": 0.2957475185394287, + "learning_rate": 9.992294827597934e-05, + "loss": 0.0404, + "step": 6680 + }, + { + "grad_norm": 0.31868672370910645, + "learning_rate": 9.992202795329831e-05, + "loss": 0.0367, + "step": 6690 + }, + { + "grad_norm": 0.37263667583465576, + "learning_rate": 9.992110217121936e-05, + "loss": 0.0377, + "step": 6700 + }, + { + "grad_norm": 0.3524804413318634, + "learning_rate": 9.992017092984372e-05, + "loss": 0.0407, + "step": 6710 + }, + { + "grad_norm": 0.31922340393066406, + "learning_rate": 9.991923422927326e-05, + "loss": 0.0473, + "step": 6720 + }, + { + "grad_norm": 0.30095764994621277, + "learning_rate": 9.991829206961037e-05, + "loss": 0.0415, + "step": 6730 + }, + { + "grad_norm": 0.3046099841594696, + "learning_rate": 9.991734445095813e-05, + "loss": 0.0623, + "step": 6740 + }, + { + "grad_norm": 0.29005441069602966, + "learning_rate": 9.991639137342015e-05, + "loss": 0.0366, + "step": 6750 + }, + { + "grad_norm": 0.28021571040153503, + "learning_rate": 9.991543283710064e-05, + "loss": 0.0335, + "step": 6760 + }, + { + "grad_norm": 0.3045392334461212, + "learning_rate": 9.991446884210445e-05, + "loss": 0.0318, + "step": 6770 + }, + { + "grad_norm": 0.2530970573425293, + "learning_rate": 9.9913499388537e-05, + "loss": 0.0376, + "step": 6780 + }, + { + "grad_norm": 0.3419201076030731, + "learning_rate": 9.99125244765043e-05, + "loss": 0.034, + "step": 6790 + }, + { + "grad_norm": 0.29189565777778625, + "learning_rate": 9.991154410611296e-05, + "loss": 0.0427, + "step": 6800 + }, + { + "grad_norm": 0.2636774480342865, + "learning_rate": 9.99105582774702e-05, + "loss": 0.033, + "step": 6810 + }, + { + "grad_norm": 0.28096434473991394, + "learning_rate": 9.990956699068384e-05, + "loss": 0.0437, + "step": 6820 + }, + { + "grad_norm": 0.3254293203353882, + "learning_rate": 9.990857024586224e-05, + "loss": 0.0354, + "step": 6830 + }, + { + "grad_norm": 0.33727508783340454, + "learning_rate": 9.990756804311446e-05, + "loss": 0.0392, + "step": 6840 + }, + { + "grad_norm": 0.26032304763793945, + "learning_rate": 9.990656038255006e-05, + "loss": 0.0345, + "step": 6850 + }, + { + "grad_norm": 0.3134416937828064, + "learning_rate": 9.990554726427926e-05, + "loss": 0.0312, + "step": 6860 + }, + { + "grad_norm": 0.33217957615852356, + "learning_rate": 9.990452868841284e-05, + "loss": 0.0479, + "step": 6870 + }, + { + "grad_norm": 0.24309852719306946, + "learning_rate": 9.99035046550622e-05, + "loss": 0.0428, + "step": 6880 + }, + { + "grad_norm": 0.2586185038089752, + "learning_rate": 9.99024751643393e-05, + "loss": 0.0382, + "step": 6890 + }, + { + "grad_norm": 0.35449615120887756, + "learning_rate": 9.990144021635677e-05, + "loss": 0.043, + "step": 6900 + }, + { + "grad_norm": 0.2781040668487549, + "learning_rate": 9.990039981122775e-05, + "loss": 0.0336, + "step": 6910 + }, + { + "grad_norm": 0.2982754409313202, + "learning_rate": 9.989935394906602e-05, + "loss": 0.0419, + "step": 6920 + }, + { + "grad_norm": 0.40027543902397156, + "learning_rate": 9.989830262998598e-05, + "loss": 0.0403, + "step": 6930 + }, + { + "grad_norm": 0.27110135555267334, + "learning_rate": 9.989724585410259e-05, + "loss": 0.033, + "step": 6940 + }, + { + "grad_norm": 0.3203417658805847, + "learning_rate": 9.989618362153139e-05, + "loss": 0.0321, + "step": 6950 + }, + { + "grad_norm": 0.3800163269042969, + "learning_rate": 9.989511593238859e-05, + "loss": 0.0343, + "step": 6960 + }, + { + "grad_norm": 0.27937543392181396, + "learning_rate": 9.98940427867909e-05, + "loss": 0.0363, + "step": 6970 + }, + { + "grad_norm": 0.2833656370639801, + "learning_rate": 9.989296418485573e-05, + "loss": 0.0438, + "step": 6980 + }, + { + "grad_norm": 0.27536827325820923, + "learning_rate": 9.989188012670101e-05, + "loss": 0.0334, + "step": 6990 + }, + { + "grad_norm": 0.3987238109111786, + "learning_rate": 9.989079061244528e-05, + "loss": 0.04, + "step": 7000 + }, + { + "grad_norm": 0.23699449002742767, + "learning_rate": 9.988969564220769e-05, + "loss": 0.0449, + "step": 7010 + }, + { + "grad_norm": 0.3830566108226776, + "learning_rate": 9.988859521610801e-05, + "loss": 0.0365, + "step": 7020 + }, + { + "grad_norm": 0.3622129261493683, + "learning_rate": 9.988748933426656e-05, + "loss": 0.0522, + "step": 7030 + }, + { + "grad_norm": 0.28122296929359436, + "learning_rate": 9.988637799680428e-05, + "loss": 0.0374, + "step": 7040 + }, + { + "grad_norm": 0.2586895525455475, + "learning_rate": 9.98852612038427e-05, + "loss": 0.0332, + "step": 7050 + }, + { + "grad_norm": 0.29525530338287354, + "learning_rate": 9.988413895550397e-05, + "loss": 0.032, + "step": 7060 + }, + { + "grad_norm": 0.2827947437763214, + "learning_rate": 9.98830112519108e-05, + "loss": 0.0363, + "step": 7070 + }, + { + "grad_norm": 0.4056331515312195, + "learning_rate": 9.98818780931865e-05, + "loss": 0.0525, + "step": 7080 + }, + { + "grad_norm": 0.3443695902824402, + "learning_rate": 9.988073947945502e-05, + "loss": 0.0371, + "step": 7090 + }, + { + "grad_norm": 0.34121641516685486, + "learning_rate": 9.987959541084087e-05, + "loss": 0.031, + "step": 7100 + }, + { + "grad_norm": 0.23405751585960388, + "learning_rate": 9.987844588746915e-05, + "loss": 0.044, + "step": 7110 + }, + { + "grad_norm": 0.29563143849372864, + "learning_rate": 9.987729090946558e-05, + "loss": 0.0304, + "step": 7120 + }, + { + "grad_norm": 0.23582780361175537, + "learning_rate": 9.987613047695647e-05, + "loss": 0.0319, + "step": 7130 + }, + { + "grad_norm": 0.22907638549804688, + "learning_rate": 9.987496459006871e-05, + "loss": 0.056, + "step": 7140 + }, + { + "grad_norm": 0.24580740928649902, + "learning_rate": 9.987379324892982e-05, + "loss": 0.0444, + "step": 7150 + }, + { + "grad_norm": 0.4440804123878479, + "learning_rate": 9.987261645366788e-05, + "loss": 0.041, + "step": 7160 + }, + { + "grad_norm": 0.28178590536117554, + "learning_rate": 9.987143420441158e-05, + "loss": 0.0391, + "step": 7170 + }, + { + "grad_norm": 0.3208364248275757, + "learning_rate": 9.987024650129022e-05, + "loss": 0.0329, + "step": 7180 + }, + { + "grad_norm": 0.2527238726615906, + "learning_rate": 9.986905334443368e-05, + "loss": 0.0413, + "step": 7190 + }, + { + "grad_norm": 0.3884478211402893, + "learning_rate": 9.986785473397245e-05, + "loss": 0.0346, + "step": 7200 + }, + { + "grad_norm": 0.26452428102493286, + "learning_rate": 9.98666506700376e-05, + "loss": 0.0393, + "step": 7210 + }, + { + "grad_norm": 0.3502344489097595, + "learning_rate": 9.986544115276081e-05, + "loss": 0.0417, + "step": 7220 + }, + { + "grad_norm": 0.31970664858818054, + "learning_rate": 9.986422618227433e-05, + "loss": 0.0335, + "step": 7230 + }, + { + "grad_norm": 0.28935691714286804, + "learning_rate": 9.986300575871106e-05, + "loss": 0.0389, + "step": 7240 + }, + { + "grad_norm": 0.30361080169677734, + "learning_rate": 9.986177988220444e-05, + "loss": 0.037, + "step": 7250 + }, + { + "grad_norm": 0.28274649381637573, + "learning_rate": 9.986054855288856e-05, + "loss": 0.0337, + "step": 7260 + }, + { + "grad_norm": 0.31100982427597046, + "learning_rate": 9.985931177089802e-05, + "loss": 0.0319, + "step": 7270 + }, + { + "grad_norm": 0.301792174577713, + "learning_rate": 9.985806953636814e-05, + "loss": 0.0324, + "step": 7280 + }, + { + "grad_norm": 0.24162042140960693, + "learning_rate": 9.985682184943471e-05, + "loss": 0.036, + "step": 7290 + }, + { + "grad_norm": 0.25785937905311584, + "learning_rate": 9.98555687102342e-05, + "loss": 0.0339, + "step": 7300 + }, + { + "grad_norm": 0.23329828679561615, + "learning_rate": 9.985431011890367e-05, + "loss": 0.035, + "step": 7310 + }, + { + "grad_norm": 0.38969555497169495, + "learning_rate": 9.985304607558075e-05, + "loss": 0.0274, + "step": 7320 + }, + { + "grad_norm": 0.2825782597064972, + "learning_rate": 9.985177658040364e-05, + "loss": 0.034, + "step": 7330 + }, + { + "grad_norm": 0.2738676965236664, + "learning_rate": 9.985050163351119e-05, + "loss": 0.0404, + "step": 7340 + }, + { + "grad_norm": 0.27064812183380127, + "learning_rate": 9.984922123504286e-05, + "loss": 0.0317, + "step": 7350 + }, + { + "grad_norm": 0.2947773337364197, + "learning_rate": 9.984793538513862e-05, + "loss": 0.0328, + "step": 7360 + }, + { + "grad_norm": 0.2917945981025696, + "learning_rate": 9.984664408393912e-05, + "loss": 0.0356, + "step": 7370 + }, + { + "grad_norm": 0.30523553490638733, + "learning_rate": 9.984534733158556e-05, + "loss": 0.032, + "step": 7380 + }, + { + "grad_norm": 0.25360366702079773, + "learning_rate": 9.984404512821977e-05, + "loss": 0.0334, + "step": 7390 + }, + { + "grad_norm": 0.33502334356307983, + "learning_rate": 9.984273747398411e-05, + "loss": 0.0341, + "step": 7400 + }, + { + "grad_norm": 0.28833675384521484, + "learning_rate": 9.984142436902165e-05, + "loss": 0.0368, + "step": 7410 + }, + { + "grad_norm": 0.3010246455669403, + "learning_rate": 9.984010581347596e-05, + "loss": 0.0408, + "step": 7420 + }, + { + "grad_norm": 0.27226006984710693, + "learning_rate": 9.983878180749121e-05, + "loss": 0.0399, + "step": 7430 + }, + { + "grad_norm": 0.3038751482963562, + "learning_rate": 9.983745235121222e-05, + "loss": 0.0358, + "step": 7440 + }, + { + "grad_norm": 0.26306214928627014, + "learning_rate": 9.983611744478438e-05, + "loss": 0.0338, + "step": 7450 + }, + { + "grad_norm": 0.2382918745279312, + "learning_rate": 9.983477708835365e-05, + "loss": 0.0342, + "step": 7460 + }, + { + "grad_norm": 0.30361855030059814, + "learning_rate": 9.983343128206664e-05, + "loss": 0.0386, + "step": 7470 + }, + { + "grad_norm": 0.3462936282157898, + "learning_rate": 9.983208002607049e-05, + "loss": 0.0423, + "step": 7480 + }, + { + "grad_norm": 0.3493572473526001, + "learning_rate": 9.9830723320513e-05, + "loss": 0.0369, + "step": 7490 + }, + { + "grad_norm": 0.3099392354488373, + "learning_rate": 9.982936116554254e-05, + "loss": 0.0321, + "step": 7500 + }, + { + "grad_norm": 0.2957872152328491, + "learning_rate": 9.982799356130803e-05, + "loss": 0.0295, + "step": 7510 + }, + { + "grad_norm": 0.2223420888185501, + "learning_rate": 9.982662050795908e-05, + "loss": 0.0299, + "step": 7520 + }, + { + "grad_norm": 0.2569833993911743, + "learning_rate": 9.982524200564583e-05, + "loss": 0.0284, + "step": 7530 + }, + { + "grad_norm": 0.27650219202041626, + "learning_rate": 9.982385805451901e-05, + "loss": 0.0356, + "step": 7540 + }, + { + "grad_norm": 0.25388774275779724, + "learning_rate": 9.982246865472998e-05, + "loss": 0.0333, + "step": 7550 + }, + { + "grad_norm": 0.21305237710475922, + "learning_rate": 9.982107380643069e-05, + "loss": 0.0332, + "step": 7560 + }, + { + "grad_norm": 0.23799386620521545, + "learning_rate": 9.981967350977368e-05, + "loss": 0.0405, + "step": 7570 + }, + { + "grad_norm": 0.3703261911869049, + "learning_rate": 9.981826776491208e-05, + "loss": 0.0419, + "step": 7580 + }, + { + "grad_norm": 0.2796971797943115, + "learning_rate": 9.98168565719996e-05, + "loss": 0.0403, + "step": 7590 + }, + { + "grad_norm": 0.23259609937667847, + "learning_rate": 9.98154399311906e-05, + "loss": 0.0481, + "step": 7600 + }, + { + "grad_norm": 0.25072360038757324, + "learning_rate": 9.981401784263997e-05, + "loss": 0.0402, + "step": 7610 + }, + { + "grad_norm": 0.2262532114982605, + "learning_rate": 9.981259030650326e-05, + "loss": 0.0395, + "step": 7620 + }, + { + "grad_norm": 0.28616514801979065, + "learning_rate": 9.981115732293655e-05, + "loss": 0.0389, + "step": 7630 + }, + { + "grad_norm": 0.3039534091949463, + "learning_rate": 9.980971889209659e-05, + "loss": 0.0368, + "step": 7640 + }, + { + "grad_norm": 0.1946086436510086, + "learning_rate": 9.980827501414064e-05, + "loss": 0.0324, + "step": 7650 + }, + { + "grad_norm": 0.25375232100486755, + "learning_rate": 9.980682568922663e-05, + "loss": 0.0316, + "step": 7660 + }, + { + "grad_norm": 0.3526664674282074, + "learning_rate": 9.980537091751304e-05, + "loss": 0.0345, + "step": 7670 + }, + { + "grad_norm": 0.3070203959941864, + "learning_rate": 9.980391069915897e-05, + "loss": 0.0357, + "step": 7680 + }, + { + "grad_norm": 0.23501381278038025, + "learning_rate": 9.98024450343241e-05, + "loss": 0.0333, + "step": 7690 + }, + { + "grad_norm": 0.2339390218257904, + "learning_rate": 9.980097392316872e-05, + "loss": 0.0283, + "step": 7700 + }, + { + "grad_norm": 0.2792384922504425, + "learning_rate": 9.97994973658537e-05, + "loss": 0.0427, + "step": 7710 + }, + { + "grad_norm": 0.26472142338752747, + "learning_rate": 9.979801536254054e-05, + "loss": 0.0382, + "step": 7720 + }, + { + "grad_norm": 0.2343011200428009, + "learning_rate": 9.979652791339127e-05, + "loss": 0.0391, + "step": 7730 + }, + { + "grad_norm": 0.3107043206691742, + "learning_rate": 9.97950350185686e-05, + "loss": 0.0351, + "step": 7740 + }, + { + "grad_norm": 0.28410080075263977, + "learning_rate": 9.979353667823574e-05, + "loss": 0.0299, + "step": 7750 + }, + { + "grad_norm": 0.22636625170707703, + "learning_rate": 9.979203289255658e-05, + "loss": 0.0335, + "step": 7760 + }, + { + "grad_norm": 0.2744749188423157, + "learning_rate": 9.979052366169557e-05, + "loss": 0.0352, + "step": 7770 + }, + { + "grad_norm": 0.2687520384788513, + "learning_rate": 9.978900898581775e-05, + "loss": 0.037, + "step": 7780 + }, + { + "grad_norm": 0.29273924231529236, + "learning_rate": 9.978748886508875e-05, + "loss": 0.0359, + "step": 7790 + }, + { + "grad_norm": 0.28796759247779846, + "learning_rate": 9.978596329967484e-05, + "loss": 0.0302, + "step": 7800 + }, + { + "grad_norm": 0.23354090750217438, + "learning_rate": 9.978443228974284e-05, + "loss": 0.0331, + "step": 7810 + }, + { + "grad_norm": 0.21528609097003937, + "learning_rate": 9.978289583546015e-05, + "loss": 0.027, + "step": 7820 + }, + { + "grad_norm": 0.297354519367218, + "learning_rate": 9.978135393699484e-05, + "loss": 0.0303, + "step": 7830 + }, + { + "grad_norm": 0.2809606194496155, + "learning_rate": 9.977980659451548e-05, + "loss": 0.0279, + "step": 7840 + }, + { + "grad_norm": 0.23026390373706818, + "learning_rate": 9.977825380819135e-05, + "loss": 0.031, + "step": 7850 + }, + { + "grad_norm": 0.3256017863750458, + "learning_rate": 9.97766955781922e-05, + "loss": 0.0365, + "step": 7860 + }, + { + "grad_norm": 0.33720579743385315, + "learning_rate": 9.977513190468848e-05, + "loss": 0.0389, + "step": 7870 + }, + { + "grad_norm": 0.35296595096588135, + "learning_rate": 9.977356278785116e-05, + "loss": 0.0371, + "step": 7880 + }, + { + "grad_norm": 0.29040560126304626, + "learning_rate": 9.977198822785184e-05, + "loss": 0.0318, + "step": 7890 + }, + { + "grad_norm": 0.33496764302253723, + "learning_rate": 9.977040822486273e-05, + "loss": 0.0474, + "step": 7900 + }, + { + "grad_norm": 0.3030688762664795, + "learning_rate": 9.97688227790566e-05, + "loss": 0.04, + "step": 7910 + }, + { + "grad_norm": 0.2527315318584442, + "learning_rate": 9.976723189060684e-05, + "loss": 0.0544, + "step": 7920 + }, + { + "grad_norm": 0.2498956322669983, + "learning_rate": 9.976563555968742e-05, + "loss": 0.034, + "step": 7930 + }, + { + "grad_norm": 0.2670711278915405, + "learning_rate": 9.976403378647292e-05, + "loss": 0.0348, + "step": 7940 + }, + { + "grad_norm": 0.2685103714466095, + "learning_rate": 9.97624265711385e-05, + "loss": 0.0328, + "step": 7950 + }, + { + "grad_norm": 0.2989571988582611, + "learning_rate": 9.976081391385993e-05, + "loss": 0.031, + "step": 7960 + }, + { + "grad_norm": 0.25616368651390076, + "learning_rate": 9.975919581481356e-05, + "loss": 0.0294, + "step": 7970 + }, + { + "grad_norm": 0.3602653443813324, + "learning_rate": 9.975757227417634e-05, + "loss": 0.0324, + "step": 7980 + }, + { + "grad_norm": 0.30340373516082764, + "learning_rate": 9.975594329212586e-05, + "loss": 0.0325, + "step": 7990 + }, + { + "grad_norm": 0.2628995180130005, + "learning_rate": 9.97543088688402e-05, + "loss": 0.0274, + "step": 8000 + }, + { + "grad_norm": 0.21238869428634644, + "learning_rate": 9.975266900449814e-05, + "loss": 0.0283, + "step": 8010 + }, + { + "grad_norm": 0.24085064232349396, + "learning_rate": 9.975102369927898e-05, + "loss": 0.029, + "step": 8020 + }, + { + "grad_norm": 0.26724568009376526, + "learning_rate": 9.974937295336269e-05, + "loss": 0.0288, + "step": 8030 + }, + { + "grad_norm": 0.2729582190513611, + "learning_rate": 9.974771676692975e-05, + "loss": 0.031, + "step": 8040 + }, + { + "grad_norm": 0.27377861738204956, + "learning_rate": 9.974605514016131e-05, + "loss": 0.0327, + "step": 8050 + }, + { + "grad_norm": 0.2355956733226776, + "learning_rate": 9.974438807323907e-05, + "loss": 0.027, + "step": 8060 + }, + { + "grad_norm": 0.26737040281295776, + "learning_rate": 9.974271556634535e-05, + "loss": 0.0286, + "step": 8070 + }, + { + "grad_norm": 0.4328813850879669, + "learning_rate": 9.974103761966302e-05, + "loss": 0.0451, + "step": 8080 + }, + { + "grad_norm": 0.2998552918434143, + "learning_rate": 9.973935423337563e-05, + "loss": 0.0368, + "step": 8090 + }, + { + "grad_norm": 0.2817609906196594, + "learning_rate": 9.973766540766722e-05, + "loss": 0.0395, + "step": 8100 + }, + { + "grad_norm": 0.2668747305870056, + "learning_rate": 9.97359711427225e-05, + "loss": 0.0419, + "step": 8110 + }, + { + "grad_norm": 0.26675042510032654, + "learning_rate": 9.973427143872677e-05, + "loss": 0.0333, + "step": 8120 + }, + { + "grad_norm": 0.3330526351928711, + "learning_rate": 9.973256629586589e-05, + "loss": 0.0465, + "step": 8130 + }, + { + "grad_norm": 0.28058531880378723, + "learning_rate": 9.973085571432632e-05, + "loss": 0.0387, + "step": 8140 + }, + { + "grad_norm": 0.2961269021034241, + "learning_rate": 9.972913969429513e-05, + "loss": 0.0315, + "step": 8150 + }, + { + "grad_norm": 0.2650843858718872, + "learning_rate": 9.972741823596e-05, + "loss": 0.0412, + "step": 8160 + }, + { + "grad_norm": 0.19450725615024567, + "learning_rate": 9.972569133950917e-05, + "loss": 0.0339, + "step": 8170 + }, + { + "grad_norm": 0.2454257607460022, + "learning_rate": 9.972395900513151e-05, + "loss": 0.0309, + "step": 8180 + }, + { + "grad_norm": 0.28682729601860046, + "learning_rate": 9.972222123301645e-05, + "loss": 0.0311, + "step": 8190 + }, + { + "grad_norm": 0.2527318298816681, + "learning_rate": 9.972047802335403e-05, + "loss": 0.0302, + "step": 8200 + }, + { + "grad_norm": 0.2567032277584076, + "learning_rate": 9.971872937633488e-05, + "loss": 0.0245, + "step": 8210 + }, + { + "grad_norm": 0.2801527976989746, + "learning_rate": 9.971697529215024e-05, + "loss": 0.0297, + "step": 8220 + }, + { + "grad_norm": 0.2584557831287384, + "learning_rate": 9.971521577099192e-05, + "loss": 0.0297, + "step": 8230 + }, + { + "grad_norm": 0.24566324055194855, + "learning_rate": 9.971345081305236e-05, + "loss": 0.0322, + "step": 8240 + }, + { + "grad_norm": 0.2597614526748657, + "learning_rate": 9.971168041852456e-05, + "loss": 0.0305, + "step": 8250 + }, + { + "grad_norm": 0.2775578200817108, + "learning_rate": 9.970990458760215e-05, + "loss": 0.0259, + "step": 8260 + }, + { + "grad_norm": 0.2215222269296646, + "learning_rate": 9.970812332047929e-05, + "loss": 0.0283, + "step": 8270 + }, + { + "grad_norm": 0.34152841567993164, + "learning_rate": 9.97063366173508e-05, + "loss": 0.0419, + "step": 8280 + }, + { + "grad_norm": 0.29298463463783264, + "learning_rate": 9.970454447841207e-05, + "loss": 0.0278, + "step": 8290 + }, + { + "grad_norm": 0.26257312297821045, + "learning_rate": 9.970274690385909e-05, + "loss": 0.0338, + "step": 8300 + }, + { + "grad_norm": 0.24145163595676422, + "learning_rate": 9.970094389388844e-05, + "loss": 0.0289, + "step": 8310 + }, + { + "grad_norm": 0.2763570249080658, + "learning_rate": 9.969913544869728e-05, + "loss": 0.0255, + "step": 8320 + }, + { + "grad_norm": 0.3095169961452484, + "learning_rate": 9.96973215684834e-05, + "loss": 0.0299, + "step": 8330 + }, + { + "grad_norm": 0.23848479986190796, + "learning_rate": 9.969550225344513e-05, + "loss": 0.0254, + "step": 8340 + }, + { + "grad_norm": 0.24103796482086182, + "learning_rate": 9.969367750378147e-05, + "loss": 0.032, + "step": 8350 + }, + { + "grad_norm": 0.21034106612205505, + "learning_rate": 9.969184731969194e-05, + "loss": 0.028, + "step": 8360 + }, + { + "grad_norm": 0.2843741178512573, + "learning_rate": 9.96900117013767e-05, + "loss": 0.0263, + "step": 8370 + }, + { + "grad_norm": 0.23153378069400787, + "learning_rate": 9.96881706490365e-05, + "loss": 0.032, + "step": 8380 + }, + { + "grad_norm": 0.20613279938697815, + "learning_rate": 9.968632416287265e-05, + "loss": 0.028, + "step": 8390 + }, + { + "grad_norm": 0.34714826941490173, + "learning_rate": 9.96844722430871e-05, + "loss": 0.0275, + "step": 8400 + }, + { + "grad_norm": 0.2388719618320465, + "learning_rate": 9.968261488988235e-05, + "loss": 0.0301, + "step": 8410 + }, + { + "grad_norm": 0.27256375551223755, + "learning_rate": 9.968075210346155e-05, + "loss": 0.0284, + "step": 8420 + }, + { + "grad_norm": 0.31579911708831787, + "learning_rate": 9.967888388402839e-05, + "loss": 0.028, + "step": 8430 + }, + { + "grad_norm": 0.20512278378009796, + "learning_rate": 9.967701023178717e-05, + "loss": 0.0297, + "step": 8440 + }, + { + "grad_norm": 0.2867315709590912, + "learning_rate": 9.967513114694282e-05, + "loss": 0.0296, + "step": 8450 + }, + { + "grad_norm": 0.280953049659729, + "learning_rate": 9.967324662970079e-05, + "loss": 0.0296, + "step": 8460 + }, + { + "grad_norm": 0.2461746484041214, + "learning_rate": 9.96713566802672e-05, + "loss": 0.0272, + "step": 8470 + }, + { + "grad_norm": 0.3342266380786896, + "learning_rate": 9.966946129884873e-05, + "loss": 0.0282, + "step": 8480 + }, + { + "grad_norm": 0.27970245480537415, + "learning_rate": 9.966756048565265e-05, + "loss": 0.0309, + "step": 8490 + }, + { + "grad_norm": 0.287237286567688, + "learning_rate": 9.966565424088681e-05, + "loss": 0.0356, + "step": 8500 + }, + { + "grad_norm": 0.32742419838905334, + "learning_rate": 9.96637425647597e-05, + "loss": 0.0293, + "step": 8510 + }, + { + "grad_norm": 0.2388700246810913, + "learning_rate": 9.966182545748038e-05, + "loss": 0.0236, + "step": 8520 + }, + { + "grad_norm": 0.38181743025779724, + "learning_rate": 9.96599029192585e-05, + "loss": 0.0336, + "step": 8530 + }, + { + "grad_norm": 0.28651759028434753, + "learning_rate": 9.965797495030428e-05, + "loss": 0.0322, + "step": 8540 + }, + { + "grad_norm": 0.20996415615081787, + "learning_rate": 9.96560415508286e-05, + "loss": 0.0319, + "step": 8550 + }, + { + "grad_norm": 0.23119144141674042, + "learning_rate": 9.965410272104286e-05, + "loss": 0.0307, + "step": 8560 + }, + { + "grad_norm": 0.31931647658348083, + "learning_rate": 9.96521584611591e-05, + "loss": 0.0528, + "step": 8570 + }, + { + "grad_norm": 0.4174204468727112, + "learning_rate": 9.965020877138994e-05, + "loss": 0.037, + "step": 8580 + }, + { + "grad_norm": 0.23783935606479645, + "learning_rate": 9.964825365194861e-05, + "loss": 0.0332, + "step": 8590 + }, + { + "grad_norm": 0.2693871557712555, + "learning_rate": 9.96462931030489e-05, + "loss": 0.0254, + "step": 8600 + }, + { + "grad_norm": 0.3077763319015503, + "learning_rate": 9.96443271249052e-05, + "loss": 0.0276, + "step": 8610 + }, + { + "grad_norm": 0.30304789543151855, + "learning_rate": 9.964235571773255e-05, + "loss": 0.0333, + "step": 8620 + }, + { + "grad_norm": 0.24959449470043182, + "learning_rate": 9.96403788817465e-05, + "loss": 0.0279, + "step": 8630 + }, + { + "grad_norm": 0.2815662622451782, + "learning_rate": 9.963839661716325e-05, + "loss": 0.0312, + "step": 8640 + }, + { + "grad_norm": 0.2716885209083557, + "learning_rate": 9.963640892419958e-05, + "loss": 0.0331, + "step": 8650 + }, + { + "grad_norm": 0.26827648282051086, + "learning_rate": 9.963441580307286e-05, + "loss": 0.0325, + "step": 8660 + }, + { + "grad_norm": 0.2859035134315491, + "learning_rate": 9.963241725400104e-05, + "loss": 0.0289, + "step": 8670 + }, + { + "grad_norm": 0.21131233870983124, + "learning_rate": 9.963041327720271e-05, + "loss": 0.025, + "step": 8680 + }, + { + "grad_norm": 0.21143798530101776, + "learning_rate": 9.962840387289697e-05, + "loss": 0.0303, + "step": 8690 + }, + { + "grad_norm": 0.21988138556480408, + "learning_rate": 9.962638904130363e-05, + "loss": 0.0293, + "step": 8700 + }, + { + "grad_norm": 0.2538502812385559, + "learning_rate": 9.962436878264298e-05, + "loss": 0.0356, + "step": 8710 + }, + { + "grad_norm": 0.419718861579895, + "learning_rate": 9.962234309713598e-05, + "loss": 0.0428, + "step": 8720 + }, + { + "grad_norm": 0.21358099579811096, + "learning_rate": 9.962031198500414e-05, + "loss": 0.0285, + "step": 8730 + }, + { + "grad_norm": 0.2663249373435974, + "learning_rate": 9.961827544646958e-05, + "loss": 0.0264, + "step": 8740 + }, + { + "grad_norm": 0.2093496471643448, + "learning_rate": 9.961623348175501e-05, + "loss": 0.0325, + "step": 8750 + }, + { + "grad_norm": 0.21907588839530945, + "learning_rate": 9.961418609108377e-05, + "loss": 0.0296, + "step": 8760 + }, + { + "grad_norm": 0.26477670669555664, + "learning_rate": 9.961213327467971e-05, + "loss": 0.0325, + "step": 8770 + }, + { + "grad_norm": 0.32057881355285645, + "learning_rate": 9.961007503276736e-05, + "loss": 0.0394, + "step": 8780 + }, + { + "grad_norm": 0.25897708535194397, + "learning_rate": 9.960801136557179e-05, + "loss": 0.0317, + "step": 8790 + }, + { + "grad_norm": 0.22486916184425354, + "learning_rate": 9.960594227331866e-05, + "loss": 0.0275, + "step": 8800 + }, + { + "grad_norm": 0.20350143313407898, + "learning_rate": 9.960386775623429e-05, + "loss": 0.0278, + "step": 8810 + }, + { + "grad_norm": 0.2133597880601883, + "learning_rate": 9.96017878145455e-05, + "loss": 0.0286, + "step": 8820 + }, + { + "grad_norm": 0.21676552295684814, + "learning_rate": 9.959970244847977e-05, + "loss": 0.0289, + "step": 8830 + }, + { + "grad_norm": 0.31035494804382324, + "learning_rate": 9.959761165826518e-05, + "loss": 0.0283, + "step": 8840 + }, + { + "grad_norm": 0.278306245803833, + "learning_rate": 9.959551544413033e-05, + "loss": 0.0262, + "step": 8850 + }, + { + "grad_norm": 0.24367694556713104, + "learning_rate": 9.959341380630448e-05, + "loss": 0.0301, + "step": 8860 + }, + { + "grad_norm": 0.3271205425262451, + "learning_rate": 9.959130674501746e-05, + "loss": 0.03, + "step": 8870 + }, + { + "grad_norm": 0.26145613193511963, + "learning_rate": 9.958919426049968e-05, + "loss": 0.0321, + "step": 8880 + }, + { + "grad_norm": 0.2227334827184677, + "learning_rate": 9.958707635298219e-05, + "loss": 0.0353, + "step": 8890 + }, + { + "grad_norm": 0.23663465678691864, + "learning_rate": 9.958495302269657e-05, + "loss": 0.0317, + "step": 8900 + }, + { + "grad_norm": 0.23389406502246857, + "learning_rate": 9.958282426987503e-05, + "loss": 0.0277, + "step": 8910 + }, + { + "grad_norm": 0.3852209150791168, + "learning_rate": 9.95806900947504e-05, + "loss": 0.0363, + "step": 8920 + }, + { + "grad_norm": 0.26318982243537903, + "learning_rate": 9.957855049755604e-05, + "loss": 0.0265, + "step": 8930 + }, + { + "grad_norm": 0.26339688897132874, + "learning_rate": 9.957640547852593e-05, + "loss": 0.0301, + "step": 8940 + }, + { + "grad_norm": 0.2245272397994995, + "learning_rate": 9.957425503789466e-05, + "loss": 0.0255, + "step": 8950 + }, + { + "grad_norm": 0.25376787781715393, + "learning_rate": 9.957209917589738e-05, + "loss": 0.0256, + "step": 8960 + }, + { + "grad_norm": 0.25946691632270813, + "learning_rate": 9.956993789276987e-05, + "loss": 0.0242, + "step": 8970 + }, + { + "grad_norm": 0.2547438144683838, + "learning_rate": 9.956777118874847e-05, + "loss": 0.0318, + "step": 8980 + }, + { + "grad_norm": 0.3102149963378906, + "learning_rate": 9.956559906407016e-05, + "loss": 0.0497, + "step": 8990 + }, + { + "grad_norm": 0.2657608389854431, + "learning_rate": 9.956342151897245e-05, + "loss": 0.0324, + "step": 9000 + }, + { + "grad_norm": 0.20772376656532288, + "learning_rate": 9.956123855369346e-05, + "loss": 0.0255, + "step": 9010 + }, + { + "grad_norm": 0.2664943337440491, + "learning_rate": 9.955905016847196e-05, + "loss": 0.0296, + "step": 9020 + }, + { + "grad_norm": 0.30040860176086426, + "learning_rate": 9.955685636354723e-05, + "loss": 0.0399, + "step": 9030 + }, + { + "grad_norm": 0.23971383273601532, + "learning_rate": 9.95546571391592e-05, + "loss": 0.0345, + "step": 9040 + }, + { + "grad_norm": 0.2846747040748596, + "learning_rate": 9.955245249554837e-05, + "loss": 0.0273, + "step": 9050 + }, + { + "grad_norm": 0.21492944657802582, + "learning_rate": 9.955024243295582e-05, + "loss": 0.0398, + "step": 9060 + }, + { + "grad_norm": 0.22139547765254974, + "learning_rate": 9.954802695162328e-05, + "loss": 0.023, + "step": 9070 + }, + { + "grad_norm": 0.2831159830093384, + "learning_rate": 9.954580605179302e-05, + "loss": 0.0261, + "step": 9080 + }, + { + "grad_norm": 0.268113911151886, + "learning_rate": 9.954357973370788e-05, + "loss": 0.0327, + "step": 9090 + }, + { + "grad_norm": 0.29760822653770447, + "learning_rate": 9.954134799761135e-05, + "loss": 0.0339, + "step": 9100 + }, + { + "grad_norm": 0.2652718424797058, + "learning_rate": 9.953911084374748e-05, + "loss": 0.0314, + "step": 9110 + }, + { + "grad_norm": 0.22531113028526306, + "learning_rate": 9.953686827236093e-05, + "loss": 0.0306, + "step": 9120 + }, + { + "grad_norm": 0.273868203163147, + "learning_rate": 9.953462028369695e-05, + "loss": 0.0372, + "step": 9130 + }, + { + "grad_norm": 0.2280065417289734, + "learning_rate": 9.953236687800136e-05, + "loss": 0.0252, + "step": 9140 + }, + { + "grad_norm": 0.2533838450908661, + "learning_rate": 9.95301080555206e-05, + "loss": 0.0348, + "step": 9150 + }, + { + "grad_norm": 0.1797395944595337, + "learning_rate": 9.952784381650171e-05, + "loss": 0.0255, + "step": 9160 + }, + { + "grad_norm": 0.22674137353897095, + "learning_rate": 9.952557416119226e-05, + "loss": 0.0265, + "step": 9170 + }, + { + "grad_norm": 0.19275228679180145, + "learning_rate": 9.95232990898405e-05, + "loss": 0.0251, + "step": 9180 + }, + { + "grad_norm": 0.2782747149467468, + "learning_rate": 9.95210186026952e-05, + "loss": 0.0236, + "step": 9190 + }, + { + "grad_norm": 0.2016344666481018, + "learning_rate": 9.951873270000576e-05, + "loss": 0.0268, + "step": 9200 + }, + { + "grad_norm": 0.19166283309459686, + "learning_rate": 9.951644138202216e-05, + "loss": 0.0395, + "step": 9210 + }, + { + "grad_norm": 0.20689427852630615, + "learning_rate": 9.951414464899498e-05, + "loss": 0.0304, + "step": 9220 + }, + { + "grad_norm": 0.24695102870464325, + "learning_rate": 9.951184250117538e-05, + "loss": 0.0279, + "step": 9230 + }, + { + "grad_norm": 0.20411330461502075, + "learning_rate": 9.950953493881513e-05, + "loss": 0.0345, + "step": 9240 + }, + { + "grad_norm": 0.20947149395942688, + "learning_rate": 9.950722196216658e-05, + "loss": 0.0235, + "step": 9250 + }, + { + "grad_norm": 0.24433375895023346, + "learning_rate": 9.950490357148265e-05, + "loss": 0.0266, + "step": 9260 + }, + { + "grad_norm": 0.23741038143634796, + "learning_rate": 9.950257976701692e-05, + "loss": 0.0312, + "step": 9270 + }, + { + "grad_norm": 0.2531261444091797, + "learning_rate": 9.950025054902348e-05, + "loss": 0.029, + "step": 9280 + }, + { + "grad_norm": 0.4013068377971649, + "learning_rate": 9.949791591775706e-05, + "loss": 0.0385, + "step": 9290 + }, + { + "grad_norm": 0.2998681366443634, + "learning_rate": 9.949557587347298e-05, + "loss": 0.0396, + "step": 9300 + }, + { + "grad_norm": 0.24046652019023895, + "learning_rate": 9.949323041642713e-05, + "loss": 0.0299, + "step": 9310 + }, + { + "grad_norm": 0.2448456734418869, + "learning_rate": 9.949087954687602e-05, + "loss": 0.0438, + "step": 9320 + }, + { + "grad_norm": 0.19716238975524902, + "learning_rate": 9.948852326507672e-05, + "loss": 0.0288, + "step": 9330 + }, + { + "grad_norm": 0.2192045897245407, + "learning_rate": 9.948616157128694e-05, + "loss": 0.0334, + "step": 9340 + }, + { + "grad_norm": 0.28484123945236206, + "learning_rate": 9.948379446576493e-05, + "loss": 0.0299, + "step": 9350 + }, + { + "grad_norm": 0.24440808594226837, + "learning_rate": 9.948142194876952e-05, + "loss": 0.026, + "step": 9360 + }, + { + "grad_norm": 0.2314007729291916, + "learning_rate": 9.947904402056024e-05, + "loss": 0.0394, + "step": 9370 + }, + { + "grad_norm": 0.1801016479730606, + "learning_rate": 9.947666068139708e-05, + "loss": 0.0266, + "step": 9380 + }, + { + "grad_norm": 0.27671197056770325, + "learning_rate": 9.947427193154071e-05, + "loss": 0.0277, + "step": 9390 + }, + { + "grad_norm": 0.27336594462394714, + "learning_rate": 9.947187777125233e-05, + "loss": 0.0374, + "step": 9400 + }, + { + "grad_norm": 0.3352520763874054, + "learning_rate": 9.946947820079377e-05, + "loss": 0.035, + "step": 9410 + }, + { + "grad_norm": 0.3108274042606354, + "learning_rate": 9.946707322042747e-05, + "loss": 0.0276, + "step": 9420 + }, + { + "grad_norm": 0.21146629750728607, + "learning_rate": 9.94646628304164e-05, + "loss": 0.0272, + "step": 9430 + }, + { + "grad_norm": 0.25181418657302856, + "learning_rate": 9.946224703102418e-05, + "loss": 0.0268, + "step": 9440 + }, + { + "grad_norm": 0.20653235912322998, + "learning_rate": 9.945982582251498e-05, + "loss": 0.0239, + "step": 9450 + }, + { + "grad_norm": 0.2121826559305191, + "learning_rate": 9.94573992051536e-05, + "loss": 0.0235, + "step": 9460 + }, + { + "grad_norm": 0.20453797280788422, + "learning_rate": 9.94549671792054e-05, + "loss": 0.0386, + "step": 9470 + }, + { + "grad_norm": 0.23002152144908905, + "learning_rate": 9.945252974493635e-05, + "loss": 0.0256, + "step": 9480 + }, + { + "grad_norm": 0.23061874508857727, + "learning_rate": 9.9450086902613e-05, + "loss": 0.0249, + "step": 9490 + }, + { + "grad_norm": 0.18634353578090668, + "learning_rate": 9.944763865250248e-05, + "loss": 0.024, + "step": 9500 + }, + { + "grad_norm": 0.15805545449256897, + "learning_rate": 9.944518499487254e-05, + "loss": 0.0243, + "step": 9510 + }, + { + "grad_norm": 0.23776651918888092, + "learning_rate": 9.944272592999151e-05, + "loss": 0.0292, + "step": 9520 + }, + { + "grad_norm": 0.21607664227485657, + "learning_rate": 9.94402614581283e-05, + "loss": 0.0265, + "step": 9530 + }, + { + "grad_norm": 0.42079275846481323, + "learning_rate": 9.943779157955244e-05, + "loss": 0.0338, + "step": 9540 + }, + { + "grad_norm": 0.2920873165130615, + "learning_rate": 9.943531629453403e-05, + "loss": 0.0292, + "step": 9550 + }, + { + "grad_norm": 0.18967023491859436, + "learning_rate": 9.943283560334375e-05, + "loss": 0.0305, + "step": 9560 + }, + { + "grad_norm": 0.18423569202423096, + "learning_rate": 9.943034950625288e-05, + "loss": 0.0409, + "step": 9570 + }, + { + "grad_norm": 0.24028638005256653, + "learning_rate": 9.942785800353332e-05, + "loss": 0.0313, + "step": 9580 + }, + { + "grad_norm": 0.2594365179538727, + "learning_rate": 9.942536109545751e-05, + "loss": 0.0257, + "step": 9590 + }, + { + "grad_norm": 0.2314215749502182, + "learning_rate": 9.942285878229853e-05, + "loss": 0.0244, + "step": 9600 + }, + { + "grad_norm": 0.22195105254650116, + "learning_rate": 9.942035106433001e-05, + "loss": 0.0251, + "step": 9610 + }, + { + "grad_norm": 0.24202179908752441, + "learning_rate": 9.94178379418262e-05, + "loss": 0.0284, + "step": 9620 + }, + { + "grad_norm": 0.21964173018932343, + "learning_rate": 9.941531941506194e-05, + "loss": 0.024, + "step": 9630 + }, + { + "grad_norm": 0.21718333661556244, + "learning_rate": 9.941279548431263e-05, + "loss": 0.0237, + "step": 9640 + }, + { + "grad_norm": 0.1686076819896698, + "learning_rate": 9.941026614985431e-05, + "loss": 0.0241, + "step": 9650 + }, + { + "grad_norm": 0.2513884902000427, + "learning_rate": 9.940773141196357e-05, + "loss": 0.0265, + "step": 9660 + }, + { + "grad_norm": 0.24969647824764252, + "learning_rate": 9.94051912709176e-05, + "loss": 0.0306, + "step": 9670 + }, + { + "grad_norm": 0.2391081154346466, + "learning_rate": 9.940264572699421e-05, + "loss": 0.0245, + "step": 9680 + }, + { + "grad_norm": 0.3072415590286255, + "learning_rate": 9.940009478047174e-05, + "loss": 0.0254, + "step": 9690 + }, + { + "grad_norm": 0.19190135598182678, + "learning_rate": 9.939753843162918e-05, + "loss": 0.0242, + "step": 9700 + }, + { + "grad_norm": 0.2722727358341217, + "learning_rate": 9.939497668074609e-05, + "loss": 0.0223, + "step": 9710 + }, + { + "grad_norm": 0.25808846950531006, + "learning_rate": 9.93924095281026e-05, + "loss": 0.0278, + "step": 9720 + }, + { + "grad_norm": 0.27858999371528625, + "learning_rate": 9.938983697397948e-05, + "loss": 0.0311, + "step": 9730 + }, + { + "grad_norm": 0.3111307919025421, + "learning_rate": 9.938725901865805e-05, + "loss": 0.0254, + "step": 9740 + }, + { + "grad_norm": 0.18427453935146332, + "learning_rate": 9.93846756624202e-05, + "loss": 0.0262, + "step": 9750 + }, + { + "grad_norm": 0.22065432369709015, + "learning_rate": 9.938208690554849e-05, + "loss": 0.0281, + "step": 9760 + }, + { + "grad_norm": 0.2541735768318176, + "learning_rate": 9.9379492748326e-05, + "loss": 0.0262, + "step": 9770 + }, + { + "grad_norm": 0.25489741563796997, + "learning_rate": 9.937689319103641e-05, + "loss": 0.0288, + "step": 9780 + }, + { + "grad_norm": 0.26940909028053284, + "learning_rate": 9.937428823396404e-05, + "loss": 0.0239, + "step": 9790 + }, + { + "grad_norm": 0.24088643491268158, + "learning_rate": 9.937167787739372e-05, + "loss": 0.0227, + "step": 9800 + }, + { + "grad_norm": 0.2515871822834015, + "learning_rate": 9.936906212161095e-05, + "loss": 0.028, + "step": 9810 + }, + { + "grad_norm": 0.20829987525939941, + "learning_rate": 9.936644096690176e-05, + "loss": 0.0289, + "step": 9820 + }, + { + "grad_norm": 0.27316397428512573, + "learning_rate": 9.936381441355282e-05, + "loss": 0.0272, + "step": 9830 + }, + { + "grad_norm": 0.19299016892910004, + "learning_rate": 9.936118246185136e-05, + "loss": 0.0362, + "step": 9840 + }, + { + "grad_norm": 0.24408580362796783, + "learning_rate": 9.935854511208518e-05, + "loss": 0.0329, + "step": 9850 + }, + { + "grad_norm": 0.17430546879768372, + "learning_rate": 9.935590236454272e-05, + "loss": 0.0219, + "step": 9860 + }, + { + "grad_norm": 0.17232699692249298, + "learning_rate": 9.935325421951298e-05, + "loss": 0.023, + "step": 9870 + }, + { + "grad_norm": 0.2845950424671173, + "learning_rate": 9.935060067728557e-05, + "loss": 0.027, + "step": 9880 + }, + { + "grad_norm": 0.15747709572315216, + "learning_rate": 9.934794173815067e-05, + "loss": 0.0203, + "step": 9890 + }, + { + "grad_norm": 0.18450337648391724, + "learning_rate": 9.934527740239906e-05, + "loss": 0.0246, + "step": 9900 + }, + { + "grad_norm": 0.18799738585948944, + "learning_rate": 9.934260767032209e-05, + "loss": 0.0259, + "step": 9910 + }, + { + "grad_norm": 0.2259480208158493, + "learning_rate": 9.933993254221172e-05, + "loss": 0.0289, + "step": 9920 + }, + { + "grad_norm": 0.2046884149312973, + "learning_rate": 9.933725201836053e-05, + "loss": 0.023, + "step": 9930 + }, + { + "grad_norm": 0.2832944989204407, + "learning_rate": 9.933456609906162e-05, + "loss": 0.0251, + "step": 9940 + }, + { + "grad_norm": 0.15886572003364563, + "learning_rate": 9.933187478460875e-05, + "loss": 0.0201, + "step": 9950 + }, + { + "grad_norm": 0.19201332330703735, + "learning_rate": 9.93291780752962e-05, + "loss": 0.0215, + "step": 9960 + }, + { + "grad_norm": 0.20374132692813873, + "learning_rate": 9.932647597141893e-05, + "loss": 0.0299, + "step": 9970 + }, + { + "grad_norm": 0.25485411286354065, + "learning_rate": 9.932376847327239e-05, + "loss": 0.0261, + "step": 9980 + }, + { + "grad_norm": 0.24204477667808533, + "learning_rate": 9.932105558115268e-05, + "loss": 0.029, + "step": 9990 + }, + { + "grad_norm": 0.30668389797210693, + "learning_rate": 9.931833729535651e-05, + "loss": 0.0308, + "step": 10000 + }, + { + "grad_norm": 0.22847627103328705, + "learning_rate": 9.931561361618111e-05, + "loss": 0.0283, + "step": 10010 + }, + { + "grad_norm": 0.19643275439739227, + "learning_rate": 9.931288454392435e-05, + "loss": 0.0294, + "step": 10020 + }, + { + "grad_norm": 0.2654076814651489, + "learning_rate": 9.931015007888467e-05, + "loss": 0.0245, + "step": 10030 + }, + { + "grad_norm": 0.2502186596393585, + "learning_rate": 9.930741022136112e-05, + "loss": 0.0311, + "step": 10040 + }, + { + "grad_norm": 0.2615163326263428, + "learning_rate": 9.930466497165333e-05, + "loss": 0.0291, + "step": 10050 + }, + { + "grad_norm": 0.25903403759002686, + "learning_rate": 9.93019143300615e-05, + "loss": 0.0322, + "step": 10060 + }, + { + "grad_norm": 0.1642703264951706, + "learning_rate": 9.929915829688644e-05, + "loss": 0.0212, + "step": 10070 + }, + { + "grad_norm": 0.18677522242069244, + "learning_rate": 9.929639687242955e-05, + "loss": 0.025, + "step": 10080 + }, + { + "grad_norm": 0.20119620859622955, + "learning_rate": 9.929363005699281e-05, + "loss": 0.0237, + "step": 10090 + }, + { + "grad_norm": 0.19396060705184937, + "learning_rate": 9.92908578508788e-05, + "loss": 0.0273, + "step": 10100 + }, + { + "grad_norm": 0.2114391028881073, + "learning_rate": 9.928808025439069e-05, + "loss": 0.0195, + "step": 10110 + }, + { + "grad_norm": 0.20611080527305603, + "learning_rate": 9.928529726783223e-05, + "loss": 0.0232, + "step": 10120 + }, + { + "grad_norm": 0.2865302264690399, + "learning_rate": 9.928250889150774e-05, + "loss": 0.0271, + "step": 10130 + }, + { + "grad_norm": 0.29436272382736206, + "learning_rate": 9.92797151257222e-05, + "loss": 0.0272, + "step": 10140 + }, + { + "grad_norm": 0.24675260484218597, + "learning_rate": 9.927691597078108e-05, + "loss": 0.032, + "step": 10150 + }, + { + "grad_norm": 0.23811346292495728, + "learning_rate": 9.927411142699053e-05, + "loss": 0.0276, + "step": 10160 + }, + { + "grad_norm": 0.2110828161239624, + "learning_rate": 9.927130149465725e-05, + "loss": 0.0263, + "step": 10170 + }, + { + "grad_norm": 0.2198689728975296, + "learning_rate": 9.92684861740885e-05, + "loss": 0.034, + "step": 10180 + }, + { + "grad_norm": 0.366068035364151, + "learning_rate": 9.926566546559217e-05, + "loss": 0.0391, + "step": 10190 + }, + { + "grad_norm": 0.2847702205181122, + "learning_rate": 9.926283936947673e-05, + "loss": 0.0318, + "step": 10200 + }, + { + "grad_norm": 0.2421785444021225, + "learning_rate": 9.926000788605126e-05, + "loss": 0.0292, + "step": 10210 + }, + { + "grad_norm": 0.22924669086933136, + "learning_rate": 9.92571710156254e-05, + "loss": 0.0254, + "step": 10220 + }, + { + "grad_norm": 0.1691027134656906, + "learning_rate": 9.925432875850936e-05, + "loss": 0.0306, + "step": 10230 + }, + { + "grad_norm": 0.1929677575826645, + "learning_rate": 9.925148111501396e-05, + "loss": 0.0266, + "step": 10240 + }, + { + "grad_norm": 0.18243707716464996, + "learning_rate": 9.924862808545066e-05, + "loss": 0.0249, + "step": 10250 + }, + { + "grad_norm": 0.2026163786649704, + "learning_rate": 9.924576967013141e-05, + "loss": 0.0204, + "step": 10260 + }, + { + "grad_norm": 0.15999555587768555, + "learning_rate": 9.924290586936887e-05, + "loss": 0.0269, + "step": 10270 + }, + { + "grad_norm": 0.2202780395746231, + "learning_rate": 9.924003668347614e-05, + "loss": 0.0305, + "step": 10280 + }, + { + "grad_norm": 0.23286528885364532, + "learning_rate": 9.923716211276704e-05, + "loss": 0.034, + "step": 10290 + }, + { + "grad_norm": 0.23497836291790009, + "learning_rate": 9.923428215755594e-05, + "loss": 0.0278, + "step": 10300 + }, + { + "grad_norm": 0.2514705955982208, + "learning_rate": 9.923139681815775e-05, + "loss": 0.0237, + "step": 10310 + }, + { + "grad_norm": 0.18826180696487427, + "learning_rate": 9.922850609488801e-05, + "loss": 0.0268, + "step": 10320 + }, + { + "grad_norm": 0.2724052667617798, + "learning_rate": 9.922560998806287e-05, + "loss": 0.0262, + "step": 10330 + }, + { + "grad_norm": 0.24208195507526398, + "learning_rate": 9.922270849799905e-05, + "loss": 0.026, + "step": 10340 + }, + { + "grad_norm": 0.27003219723701477, + "learning_rate": 9.92198016250138e-05, + "loss": 0.0289, + "step": 10350 + }, + { + "grad_norm": 0.22307470440864563, + "learning_rate": 9.921688936942506e-05, + "loss": 0.0242, + "step": 10360 + }, + { + "grad_norm": 0.18017853796482086, + "learning_rate": 9.921397173155129e-05, + "loss": 0.0223, + "step": 10370 + }, + { + "grad_norm": 0.15174813568592072, + "learning_rate": 9.921104871171157e-05, + "loss": 0.0246, + "step": 10380 + }, + { + "grad_norm": 0.15180101990699768, + "learning_rate": 9.920812031022554e-05, + "loss": 0.0223, + "step": 10390 + }, + { + "grad_norm": 0.24426856637001038, + "learning_rate": 9.920518652741348e-05, + "loss": 0.0267, + "step": 10400 + }, + { + "grad_norm": 0.16897350549697876, + "learning_rate": 9.920224736359618e-05, + "loss": 0.0242, + "step": 10410 + }, + { + "grad_norm": 0.2237374484539032, + "learning_rate": 9.91993028190951e-05, + "loss": 0.0281, + "step": 10420 + }, + { + "grad_norm": 0.23615053296089172, + "learning_rate": 9.919635289423222e-05, + "loss": 0.0207, + "step": 10430 + }, + { + "grad_norm": 0.25416046380996704, + "learning_rate": 9.919339758933015e-05, + "loss": 0.0221, + "step": 10440 + }, + { + "grad_norm": 0.21657735109329224, + "learning_rate": 9.919043690471209e-05, + "loss": 0.0284, + "step": 10450 + }, + { + "grad_norm": 0.28293582797050476, + "learning_rate": 9.91874708407018e-05, + "loss": 0.0262, + "step": 10460 + }, + { + "grad_norm": 0.26687443256378174, + "learning_rate": 9.918449939762367e-05, + "loss": 0.0247, + "step": 10470 + }, + { + "grad_norm": 0.27301326394081116, + "learning_rate": 9.91815225758026e-05, + "loss": 0.026, + "step": 10480 + }, + { + "grad_norm": 0.15694575011730194, + "learning_rate": 9.917854037556419e-05, + "loss": 0.0209, + "step": 10490 + }, + { + "grad_norm": 0.17286072671413422, + "learning_rate": 9.917555279723454e-05, + "loss": 0.0207, + "step": 10500 + }, + { + "grad_norm": 0.2426028549671173, + "learning_rate": 9.917255984114036e-05, + "loss": 0.031, + "step": 10510 + }, + { + "grad_norm": 0.18824507296085358, + "learning_rate": 9.916956150760896e-05, + "loss": 0.0211, + "step": 10520 + }, + { + "grad_norm": 0.18461254239082336, + "learning_rate": 9.916655779696826e-05, + "loss": 0.024, + "step": 10530 + }, + { + "grad_norm": 0.15825887024402618, + "learning_rate": 9.916354870954671e-05, + "loss": 0.023, + "step": 10540 + }, + { + "grad_norm": 0.21739420294761658, + "learning_rate": 9.91605342456734e-05, + "loss": 0.02, + "step": 10550 + }, + { + "grad_norm": 0.1977476328611374, + "learning_rate": 9.915751440567795e-05, + "loss": 0.0203, + "step": 10560 + }, + { + "grad_norm": 0.2597450017929077, + "learning_rate": 9.915448918989066e-05, + "loss": 0.0207, + "step": 10570 + }, + { + "grad_norm": 0.17215295135974884, + "learning_rate": 9.915145859864232e-05, + "loss": 0.0251, + "step": 10580 + }, + { + "grad_norm": 0.2348063886165619, + "learning_rate": 9.914842263226437e-05, + "loss": 0.0273, + "step": 10590 + }, + { + "grad_norm": 0.16071432828903198, + "learning_rate": 9.914538129108882e-05, + "loss": 0.0349, + "step": 10600 + }, + { + "grad_norm": 0.1594308614730835, + "learning_rate": 9.914233457544825e-05, + "loss": 0.0436, + "step": 10610 + }, + { + "grad_norm": 0.366974800825119, + "learning_rate": 9.913928248567586e-05, + "loss": 0.0393, + "step": 10620 + }, + { + "grad_norm": 0.23805655539035797, + "learning_rate": 9.913622502210542e-05, + "loss": 0.0239, + "step": 10630 + }, + { + "grad_norm": 0.20906804502010345, + "learning_rate": 9.913316218507128e-05, + "loss": 0.0325, + "step": 10640 + }, + { + "grad_norm": 0.170127272605896, + "learning_rate": 9.91300939749084e-05, + "loss": 0.0295, + "step": 10650 + }, + { + "grad_norm": 0.21453642845153809, + "learning_rate": 9.91270203919523e-05, + "loss": 0.0323, + "step": 10660 + }, + { + "grad_norm": 0.29988667368888855, + "learning_rate": 9.912394143653912e-05, + "loss": 0.0311, + "step": 10670 + }, + { + "grad_norm": 0.2286015897989273, + "learning_rate": 9.912085710900555e-05, + "loss": 0.0224, + "step": 10680 + }, + { + "grad_norm": 0.26532313227653503, + "learning_rate": 9.911776740968892e-05, + "loss": 0.0245, + "step": 10690 + }, + { + "grad_norm": 0.22534982860088348, + "learning_rate": 9.911467233892709e-05, + "loss": 0.0446, + "step": 10700 + }, + { + "grad_norm": 0.24942530691623688, + "learning_rate": 9.911157189705853e-05, + "loss": 0.0355, + "step": 10710 + }, + { + "grad_norm": 0.18762122094631195, + "learning_rate": 9.910846608442229e-05, + "loss": 0.0291, + "step": 10720 + }, + { + "grad_norm": 0.17774976789951324, + "learning_rate": 9.910535490135805e-05, + "loss": 0.0224, + "step": 10730 + }, + { + "grad_norm": 0.2681278884410858, + "learning_rate": 9.910223834820603e-05, + "loss": 0.0226, + "step": 10740 + }, + { + "grad_norm": 0.16931435465812683, + "learning_rate": 9.909911642530703e-05, + "loss": 0.0261, + "step": 10750 + }, + { + "grad_norm": 0.23430457711219788, + "learning_rate": 9.909598913300249e-05, + "loss": 0.0322, + "step": 10760 + }, + { + "grad_norm": 0.28979241847991943, + "learning_rate": 9.909285647163438e-05, + "loss": 0.0319, + "step": 10770 + }, + { + "grad_norm": 0.19815216958522797, + "learning_rate": 9.908971844154531e-05, + "loss": 0.0281, + "step": 10780 + }, + { + "grad_norm": 0.2559219002723694, + "learning_rate": 9.908657504307843e-05, + "loss": 0.0249, + "step": 10790 + }, + { + "grad_norm": 0.21558895707130432, + "learning_rate": 9.908342627657751e-05, + "loss": 0.0305, + "step": 10800 + }, + { + "grad_norm": 0.22216109931468964, + "learning_rate": 9.908027214238689e-05, + "loss": 0.0232, + "step": 10810 + }, + { + "grad_norm": 0.21503609418869019, + "learning_rate": 9.90771126408515e-05, + "loss": 0.0265, + "step": 10820 + }, + { + "grad_norm": 0.19307014346122742, + "learning_rate": 9.907394777231685e-05, + "loss": 0.0259, + "step": 10830 + }, + { + "grad_norm": 0.23382370173931122, + "learning_rate": 9.907077753712905e-05, + "loss": 0.0313, + "step": 10840 + }, + { + "grad_norm": 0.26325681805610657, + "learning_rate": 9.906760193563482e-05, + "loss": 0.0252, + "step": 10850 + }, + { + "grad_norm": 0.22320054471492767, + "learning_rate": 9.906442096818139e-05, + "loss": 0.0229, + "step": 10860 + }, + { + "grad_norm": 0.20608460903167725, + "learning_rate": 9.906123463511665e-05, + "loss": 0.0272, + "step": 10870 + }, + { + "grad_norm": 0.22655341029167175, + "learning_rate": 9.905804293678907e-05, + "loss": 0.0287, + "step": 10880 + }, + { + "grad_norm": 0.228485107421875, + "learning_rate": 9.905484587354766e-05, + "loss": 0.0303, + "step": 10890 + }, + { + "grad_norm": 0.3410513699054718, + "learning_rate": 9.905164344574205e-05, + "loss": 0.0316, + "step": 10900 + }, + { + "grad_norm": 0.2594175338745117, + "learning_rate": 9.904843565372248e-05, + "loss": 0.0276, + "step": 10910 + }, + { + "grad_norm": 0.21144439280033112, + "learning_rate": 9.904522249783972e-05, + "loss": 0.0269, + "step": 10920 + }, + { + "grad_norm": 0.17710238695144653, + "learning_rate": 9.904200397844517e-05, + "loss": 0.0354, + "step": 10930 + }, + { + "grad_norm": 0.30291834473609924, + "learning_rate": 9.903878009589078e-05, + "loss": 0.0323, + "step": 10940 + }, + { + "grad_norm": 0.29697805643081665, + "learning_rate": 9.903555085052915e-05, + "loss": 0.0272, + "step": 10950 + }, + { + "grad_norm": 0.13839080929756165, + "learning_rate": 9.903231624271338e-05, + "loss": 0.021, + "step": 10960 + }, + { + "grad_norm": 0.17306920886039734, + "learning_rate": 9.902907627279724e-05, + "loss": 0.0226, + "step": 10970 + }, + { + "grad_norm": 0.1867360770702362, + "learning_rate": 9.902583094113504e-05, + "loss": 0.0251, + "step": 10980 + }, + { + "grad_norm": 0.15005651116371155, + "learning_rate": 9.902258024808168e-05, + "loss": 0.0228, + "step": 10990 + }, + { + "grad_norm": 0.199822798371315, + "learning_rate": 9.901932419399264e-05, + "loss": 0.0235, + "step": 11000 + }, + { + "grad_norm": 0.15658538043498993, + "learning_rate": 9.9016062779224e-05, + "loss": 0.0227, + "step": 11010 + }, + { + "grad_norm": 0.1818545162677765, + "learning_rate": 9.901279600413242e-05, + "loss": 0.024, + "step": 11020 + }, + { + "grad_norm": 0.30815455317497253, + "learning_rate": 9.900952386907518e-05, + "loss": 0.0257, + "step": 11030 + }, + { + "grad_norm": 0.20397421717643738, + "learning_rate": 9.90062463744101e-05, + "loss": 0.0226, + "step": 11040 + }, + { + "grad_norm": 0.23438507318496704, + "learning_rate": 9.900296352049558e-05, + "loss": 0.0213, + "step": 11050 + }, + { + "grad_norm": 0.1967267543077469, + "learning_rate": 9.899967530769065e-05, + "loss": 0.0276, + "step": 11060 + }, + { + "grad_norm": 0.15900884568691254, + "learning_rate": 9.899638173635489e-05, + "loss": 0.022, + "step": 11070 + }, + { + "grad_norm": 0.19317078590393066, + "learning_rate": 9.899308280684849e-05, + "loss": 0.0236, + "step": 11080 + }, + { + "grad_norm": 0.2020977884531021, + "learning_rate": 9.898977851953222e-05, + "loss": 0.0227, + "step": 11090 + }, + { + "grad_norm": 0.1757260411977768, + "learning_rate": 9.898646887476741e-05, + "loss": 0.028, + "step": 11100 + }, + { + "grad_norm": 0.24692878127098083, + "learning_rate": 9.898315387291603e-05, + "loss": 0.0264, + "step": 11110 + }, + { + "grad_norm": 0.20433445274829865, + "learning_rate": 9.89798335143406e-05, + "loss": 0.0236, + "step": 11120 + }, + { + "grad_norm": 0.18330003321170807, + "learning_rate": 9.897650779940419e-05, + "loss": 0.0284, + "step": 11130 + }, + { + "grad_norm": 0.15452328324317932, + "learning_rate": 9.897317672847054e-05, + "loss": 0.0219, + "step": 11140 + }, + { + "grad_norm": 0.1887684315443039, + "learning_rate": 9.89698403019039e-05, + "loss": 0.016, + "step": 11150 + }, + { + "grad_norm": 0.23487187922000885, + "learning_rate": 9.896649852006917e-05, + "loss": 0.0221, + "step": 11160 + }, + { + "grad_norm": 0.26311561465263367, + "learning_rate": 9.896315138333177e-05, + "loss": 0.0215, + "step": 11170 + }, + { + "grad_norm": 0.20930245518684387, + "learning_rate": 9.895979889205774e-05, + "loss": 0.0247, + "step": 11180 + }, + { + "grad_norm": 0.1848008632659912, + "learning_rate": 9.895644104661372e-05, + "loss": 0.0276, + "step": 11190 + }, + { + "grad_norm": 0.17293724417686462, + "learning_rate": 9.895307784736691e-05, + "loss": 0.0253, + "step": 11200 + }, + { + "grad_norm": 0.192388653755188, + "learning_rate": 9.894970929468512e-05, + "loss": 0.0255, + "step": 11210 + }, + { + "grad_norm": 0.17037655413150787, + "learning_rate": 9.89463353889367e-05, + "loss": 0.0221, + "step": 11220 + }, + { + "grad_norm": 0.30824750661849976, + "learning_rate": 9.894295613049065e-05, + "loss": 0.0344, + "step": 11230 + }, + { + "grad_norm": 0.16636694967746735, + "learning_rate": 9.893957151971649e-05, + "loss": 0.0224, + "step": 11240 + }, + { + "grad_norm": 0.23010289669036865, + "learning_rate": 9.893618155698436e-05, + "loss": 0.0202, + "step": 11250 + }, + { + "grad_norm": 0.18815329670906067, + "learning_rate": 9.8932786242665e-05, + "loss": 0.0198, + "step": 11260 + }, + { + "grad_norm": 0.15744227170944214, + "learning_rate": 9.89293855771297e-05, + "loss": 0.0215, + "step": 11270 + }, + { + "grad_norm": 0.2967168688774109, + "learning_rate": 9.892597956075036e-05, + "loss": 0.0317, + "step": 11280 + }, + { + "grad_norm": 0.24614492058753967, + "learning_rate": 9.892256819389947e-05, + "loss": 0.025, + "step": 11290 + }, + { + "grad_norm": 0.19683314859867096, + "learning_rate": 9.891915147695006e-05, + "loss": 0.0247, + "step": 11300 + }, + { + "grad_norm": 0.2899618446826935, + "learning_rate": 9.891572941027577e-05, + "loss": 0.0257, + "step": 11310 + }, + { + "grad_norm": 0.2563789486885071, + "learning_rate": 9.89123019942509e-05, + "loss": 0.0261, + "step": 11320 + }, + { + "grad_norm": 0.15692956745624542, + "learning_rate": 9.89088692292502e-05, + "loss": 0.0239, + "step": 11330 + }, + { + "grad_norm": 0.20042259991168976, + "learning_rate": 9.89054311156491e-05, + "loss": 0.0197, + "step": 11340 + }, + { + "grad_norm": 0.13932734727859497, + "learning_rate": 9.890198765382357e-05, + "loss": 0.0395, + "step": 11350 + }, + { + "grad_norm": 0.20912913978099823, + "learning_rate": 9.889853884415021e-05, + "loss": 0.0237, + "step": 11360 + }, + { + "grad_norm": 0.22004306316375732, + "learning_rate": 9.889508468700614e-05, + "loss": 0.0204, + "step": 11370 + }, + { + "grad_norm": 0.19079676270484924, + "learning_rate": 9.889162518276915e-05, + "loss": 0.0225, + "step": 11380 + }, + { + "grad_norm": 0.22446309030056, + "learning_rate": 9.888816033181752e-05, + "loss": 0.0262, + "step": 11390 + }, + { + "grad_norm": 0.18327660858631134, + "learning_rate": 9.888469013453018e-05, + "loss": 0.0281, + "step": 11400 + }, + { + "grad_norm": 0.3433232605457306, + "learning_rate": 9.888121459128663e-05, + "loss": 0.025, + "step": 11410 + }, + { + "grad_norm": 0.21350350975990295, + "learning_rate": 9.887773370246693e-05, + "loss": 0.0365, + "step": 11420 + }, + { + "grad_norm": 0.2073519080877304, + "learning_rate": 9.887424746845177e-05, + "loss": 0.027, + "step": 11430 + }, + { + "grad_norm": 0.236676886677742, + "learning_rate": 9.887075588962239e-05, + "loss": 0.0245, + "step": 11440 + }, + { + "grad_norm": 0.18879009783267975, + "learning_rate": 9.88672589663606e-05, + "loss": 0.024, + "step": 11450 + }, + { + "grad_norm": 0.2247360348701477, + "learning_rate": 9.886375669904886e-05, + "loss": 0.0302, + "step": 11460 + }, + { + "grad_norm": 0.20852605998516083, + "learning_rate": 9.886024908807014e-05, + "loss": 0.0224, + "step": 11470 + }, + { + "grad_norm": 0.23273442685604095, + "learning_rate": 9.885673613380806e-05, + "loss": 0.0223, + "step": 11480 + }, + { + "grad_norm": 0.22791051864624023, + "learning_rate": 9.885321783664676e-05, + "loss": 0.0275, + "step": 11490 + }, + { + "grad_norm": 0.27649661898612976, + "learning_rate": 9.884969419697101e-05, + "loss": 0.0239, + "step": 11500 + }, + { + "grad_norm": 0.17674340307712555, + "learning_rate": 9.884616521516614e-05, + "loss": 0.0213, + "step": 11510 + }, + { + "grad_norm": 0.24930547177791595, + "learning_rate": 9.88426308916181e-05, + "loss": 0.0222, + "step": 11520 + }, + { + "grad_norm": 0.1840018332004547, + "learning_rate": 9.883909122671335e-05, + "loss": 0.0203, + "step": 11530 + }, + { + "grad_norm": 0.16447660326957703, + "learning_rate": 9.883554622083904e-05, + "loss": 0.0221, + "step": 11540 + }, + { + "grad_norm": 0.320231169462204, + "learning_rate": 9.88319958743828e-05, + "loss": 0.0259, + "step": 11550 + }, + { + "grad_norm": 0.1773620992898941, + "learning_rate": 9.882844018773291e-05, + "loss": 0.0223, + "step": 11560 + }, + { + "grad_norm": 0.20005793869495392, + "learning_rate": 9.882487916127823e-05, + "loss": 0.0213, + "step": 11570 + }, + { + "grad_norm": 0.24085812270641327, + "learning_rate": 9.882131279540815e-05, + "loss": 0.0274, + "step": 11580 + }, + { + "grad_norm": 0.21849165856838226, + "learning_rate": 9.881774109051271e-05, + "loss": 0.0263, + "step": 11590 + }, + { + "grad_norm": 0.20229481160640717, + "learning_rate": 9.881416404698252e-05, + "loss": 0.0216, + "step": 11600 + }, + { + "grad_norm": 0.24025803804397583, + "learning_rate": 9.881058166520873e-05, + "loss": 0.031, + "step": 11610 + }, + { + "grad_norm": 0.20828023552894592, + "learning_rate": 9.880699394558311e-05, + "loss": 0.0319, + "step": 11620 + }, + { + "grad_norm": 0.23106470704078674, + "learning_rate": 9.880340088849801e-05, + "loss": 0.0375, + "step": 11630 + }, + { + "grad_norm": 0.17633292078971863, + "learning_rate": 9.879980249434637e-05, + "loss": 0.0238, + "step": 11640 + }, + { + "grad_norm": 0.28792208433151245, + "learning_rate": 9.879619876352168e-05, + "loss": 0.0275, + "step": 11650 + }, + { + "grad_norm": 0.20732413232326508, + "learning_rate": 9.879258969641809e-05, + "loss": 0.0227, + "step": 11660 + }, + { + "grad_norm": 0.16144432127475739, + "learning_rate": 9.878897529343023e-05, + "loss": 0.0196, + "step": 11670 + }, + { + "grad_norm": 0.16630032658576965, + "learning_rate": 9.878535555495338e-05, + "loss": 0.0249, + "step": 11680 + }, + { + "grad_norm": 0.2381720244884491, + "learning_rate": 9.87817304813834e-05, + "loss": 0.0237, + "step": 11690 + }, + { + "grad_norm": 0.2279759645462036, + "learning_rate": 9.877810007311671e-05, + "loss": 0.0232, + "step": 11700 + }, + { + "grad_norm": 0.2708115577697754, + "learning_rate": 9.877446433055035e-05, + "loss": 0.0234, + "step": 11710 + }, + { + "grad_norm": 0.30552560091018677, + "learning_rate": 9.877082325408191e-05, + "loss": 0.0243, + "step": 11720 + }, + { + "grad_norm": 0.16460350155830383, + "learning_rate": 9.876717684410954e-05, + "loss": 0.0203, + "step": 11730 + }, + { + "grad_norm": 0.16834262013435364, + "learning_rate": 9.876352510103204e-05, + "loss": 0.021, + "step": 11740 + }, + { + "grad_norm": 0.1868865042924881, + "learning_rate": 9.875986802524875e-05, + "loss": 0.0288, + "step": 11750 + }, + { + "grad_norm": 0.15930218994617462, + "learning_rate": 9.87562056171596e-05, + "loss": 0.0219, + "step": 11760 + }, + { + "grad_norm": 0.17643789947032928, + "learning_rate": 9.875253787716511e-05, + "loss": 0.0233, + "step": 11770 + }, + { + "grad_norm": 0.21850240230560303, + "learning_rate": 9.874886480566637e-05, + "loss": 0.0233, + "step": 11780 + }, + { + "grad_norm": 0.1663942039012909, + "learning_rate": 9.874518640306507e-05, + "loss": 0.02, + "step": 11790 + }, + { + "grad_norm": 0.16576506197452545, + "learning_rate": 9.874150266976347e-05, + "loss": 0.0261, + "step": 11800 + }, + { + "grad_norm": 0.15494920313358307, + "learning_rate": 9.873781360616443e-05, + "loss": 0.0304, + "step": 11810 + }, + { + "grad_norm": 0.15801291167736053, + "learning_rate": 9.873411921267137e-05, + "loss": 0.0238, + "step": 11820 + }, + { + "grad_norm": 0.18863633275032043, + "learning_rate": 9.873041948968829e-05, + "loss": 0.0302, + "step": 11830 + }, + { + "grad_norm": 0.17288516461849213, + "learning_rate": 9.872671443761981e-05, + "loss": 0.0217, + "step": 11840 + }, + { + "grad_norm": 0.16020582616329193, + "learning_rate": 9.872300405687109e-05, + "loss": 0.0227, + "step": 11850 + }, + { + "grad_norm": 0.2162427455186844, + "learning_rate": 9.871928834784792e-05, + "loss": 0.0254, + "step": 11860 + }, + { + "grad_norm": 0.16531915962696075, + "learning_rate": 9.871556731095661e-05, + "loss": 0.0201, + "step": 11870 + }, + { + "grad_norm": 0.16371871531009674, + "learning_rate": 9.871184094660411e-05, + "loss": 0.0345, + "step": 11880 + }, + { + "grad_norm": 0.18528763949871063, + "learning_rate": 9.870810925519791e-05, + "loss": 0.0205, + "step": 11890 + }, + { + "grad_norm": 0.19059091806411743, + "learning_rate": 9.870437223714612e-05, + "loss": 0.022, + "step": 11900 + }, + { + "grad_norm": 0.19611909985542297, + "learning_rate": 9.87006298928574e-05, + "loss": 0.0228, + "step": 11910 + }, + { + "grad_norm": 0.16118177771568298, + "learning_rate": 9.869688222274103e-05, + "loss": 0.0205, + "step": 11920 + }, + { + "grad_norm": 0.17447493970394135, + "learning_rate": 9.869312922720681e-05, + "loss": 0.023, + "step": 11930 + }, + { + "grad_norm": 0.21070200204849243, + "learning_rate": 9.868937090666521e-05, + "loss": 0.0219, + "step": 11940 + }, + { + "grad_norm": 0.14857400953769684, + "learning_rate": 9.86856072615272e-05, + "loss": 0.0181, + "step": 11950 + }, + { + "grad_norm": 0.24491359293460846, + "learning_rate": 9.868183829220438e-05, + "loss": 0.0209, + "step": 11960 + }, + { + "grad_norm": 0.2287050485610962, + "learning_rate": 9.867806399910893e-05, + "loss": 0.0215, + "step": 11970 + }, + { + "grad_norm": 0.22646783292293549, + "learning_rate": 9.867428438265356e-05, + "loss": 0.0229, + "step": 11980 + }, + { + "grad_norm": 0.20929382741451263, + "learning_rate": 9.867049944325165e-05, + "loss": 0.0237, + "step": 11990 + }, + { + "grad_norm": 0.19067181646823883, + "learning_rate": 9.86667091813171e-05, + "loss": 0.0317, + "step": 12000 + }, + { + "grad_norm": 0.23134197294712067, + "learning_rate": 9.866291359726438e-05, + "loss": 0.0253, + "step": 12010 + }, + { + "grad_norm": 0.18194840848445892, + "learning_rate": 9.865911269150861e-05, + "loss": 0.0216, + "step": 12020 + }, + { + "grad_norm": 0.17308545112609863, + "learning_rate": 9.865530646446544e-05, + "loss": 0.0193, + "step": 12030 + }, + { + "grad_norm": 0.25795429944992065, + "learning_rate": 9.86514949165511e-05, + "loss": 0.023, + "step": 12040 + }, + { + "grad_norm": 0.23635797202587128, + "learning_rate": 9.864767804818243e-05, + "loss": 0.0258, + "step": 12050 + }, + { + "grad_norm": 0.2365824282169342, + "learning_rate": 9.86438558597768e-05, + "loss": 0.0232, + "step": 12060 + }, + { + "grad_norm": 0.16586890816688538, + "learning_rate": 9.864002835175225e-05, + "loss": 0.0336, + "step": 12070 + }, + { + "grad_norm": 0.23999793827533722, + "learning_rate": 9.863619552452734e-05, + "loss": 0.0246, + "step": 12080 + }, + { + "grad_norm": 0.19375000894069672, + "learning_rate": 9.863235737852119e-05, + "loss": 0.0196, + "step": 12090 + }, + { + "grad_norm": 0.24414919316768646, + "learning_rate": 9.862851391415356e-05, + "loss": 0.0258, + "step": 12100 + }, + { + "grad_norm": 0.18846546113491058, + "learning_rate": 9.862466513184477e-05, + "loss": 0.0234, + "step": 12110 + }, + { + "grad_norm": 0.18849043548107147, + "learning_rate": 9.86208110320157e-05, + "loss": 0.0175, + "step": 12120 + }, + { + "grad_norm": 0.15783953666687012, + "learning_rate": 9.861695161508784e-05, + "loss": 0.0233, + "step": 12130 + }, + { + "grad_norm": 0.210113987326622, + "learning_rate": 9.861308688148324e-05, + "loss": 0.0199, + "step": 12140 + }, + { + "grad_norm": 0.14968241751194, + "learning_rate": 9.860921683162455e-05, + "loss": 0.0204, + "step": 12150 + }, + { + "grad_norm": 0.16812889277935028, + "learning_rate": 9.860534146593499e-05, + "loss": 0.0185, + "step": 12160 + }, + { + "grad_norm": 0.16429181396961212, + "learning_rate": 9.860146078483836e-05, + "loss": 0.02, + "step": 12170 + }, + { + "grad_norm": 0.2724097967147827, + "learning_rate": 9.859757478875905e-05, + "loss": 0.0285, + "step": 12180 + }, + { + "grad_norm": 0.23996753990650177, + "learning_rate": 9.859368347812204e-05, + "loss": 0.0209, + "step": 12190 + }, + { + "grad_norm": 0.1860392540693283, + "learning_rate": 9.858978685335285e-05, + "loss": 0.0288, + "step": 12200 + }, + { + "grad_norm": 0.2090688943862915, + "learning_rate": 9.858588491487763e-05, + "loss": 0.038, + "step": 12210 + }, + { + "grad_norm": 0.22886458039283752, + "learning_rate": 9.858197766312308e-05, + "loss": 0.0242, + "step": 12220 + }, + { + "grad_norm": 0.19855360686779022, + "learning_rate": 9.857806509851649e-05, + "loss": 0.0249, + "step": 12230 + }, + { + "grad_norm": 0.23276641964912415, + "learning_rate": 9.857414722148574e-05, + "loss": 0.0214, + "step": 12240 + }, + { + "grad_norm": 0.21611379086971283, + "learning_rate": 9.857022403245928e-05, + "loss": 0.0297, + "step": 12250 + }, + { + "grad_norm": 0.24511033296585083, + "learning_rate": 9.856629553186615e-05, + "loss": 0.0314, + "step": 12260 + }, + { + "grad_norm": 0.1581019163131714, + "learning_rate": 9.856236172013595e-05, + "loss": 0.024, + "step": 12270 + }, + { + "grad_norm": 0.19699901342391968, + "learning_rate": 9.85584225976989e-05, + "loss": 0.0226, + "step": 12280 + }, + { + "grad_norm": 0.18282975256443024, + "learning_rate": 9.855447816498575e-05, + "loss": 0.0261, + "step": 12290 + }, + { + "grad_norm": 0.22771519422531128, + "learning_rate": 9.855052842242787e-05, + "loss": 0.0188, + "step": 12300 + }, + { + "grad_norm": 0.15357476472854614, + "learning_rate": 9.85465733704572e-05, + "loss": 0.0218, + "step": 12310 + }, + { + "grad_norm": 0.2205132395029068, + "learning_rate": 9.854261300950624e-05, + "loss": 0.026, + "step": 12320 + }, + { + "grad_norm": 0.17333334684371948, + "learning_rate": 9.853864734000813e-05, + "loss": 0.0219, + "step": 12330 + }, + { + "grad_norm": 0.17897644639015198, + "learning_rate": 9.85346763623965e-05, + "loss": 0.0247, + "step": 12340 + }, + { + "grad_norm": 0.19456519186496735, + "learning_rate": 9.853070007710564e-05, + "loss": 0.0221, + "step": 12350 + }, + { + "grad_norm": 0.19320239126682281, + "learning_rate": 9.85267184845704e-05, + "loss": 0.0232, + "step": 12360 + }, + { + "grad_norm": 0.1920768767595291, + "learning_rate": 9.852273158522616e-05, + "loss": 0.0203, + "step": 12370 + }, + { + "grad_norm": 0.19391462206840515, + "learning_rate": 9.851873937950896e-05, + "loss": 0.0223, + "step": 12380 + }, + { + "grad_norm": 0.2679179608821869, + "learning_rate": 9.851474186785537e-05, + "loss": 0.0232, + "step": 12390 + }, + { + "grad_norm": 0.2386413812637329, + "learning_rate": 9.851073905070254e-05, + "loss": 0.025, + "step": 12400 + }, + { + "grad_norm": 0.2780703008174896, + "learning_rate": 9.850673092848824e-05, + "loss": 0.0257, + "step": 12410 + }, + { + "grad_norm": 0.168027862906456, + "learning_rate": 9.850271750165077e-05, + "loss": 0.0178, + "step": 12420 + }, + { + "grad_norm": 0.32538965344429016, + "learning_rate": 9.849869877062902e-05, + "loss": 0.0328, + "step": 12430 + }, + { + "grad_norm": 0.19939382374286652, + "learning_rate": 9.849467473586252e-05, + "loss": 0.0271, + "step": 12440 + }, + { + "grad_norm": 0.21899616718292236, + "learning_rate": 9.849064539779127e-05, + "loss": 0.0224, + "step": 12450 + }, + { + "grad_norm": 0.21500538289546967, + "learning_rate": 9.848661075685594e-05, + "loss": 0.0261, + "step": 12460 + }, + { + "grad_norm": 0.14544419944286346, + "learning_rate": 9.848257081349778e-05, + "loss": 0.0234, + "step": 12470 + }, + { + "grad_norm": 0.16736792027950287, + "learning_rate": 9.847852556815856e-05, + "loss": 0.0189, + "step": 12480 + }, + { + "grad_norm": 0.16074730455875397, + "learning_rate": 9.847447502128067e-05, + "loss": 0.02, + "step": 12490 + }, + { + "grad_norm": 0.19539934396743774, + "learning_rate": 9.847041917330708e-05, + "loss": 0.0269, + "step": 12500 + }, + { + "grad_norm": 0.21927885711193085, + "learning_rate": 9.846635802468132e-05, + "loss": 0.032, + "step": 12510 + }, + { + "grad_norm": 0.22915364801883698, + "learning_rate": 9.84622915758475e-05, + "loss": 0.0256, + "step": 12520 + }, + { + "grad_norm": 0.15862935781478882, + "learning_rate": 9.845821982725034e-05, + "loss": 0.0236, + "step": 12530 + }, + { + "grad_norm": 0.23897311091423035, + "learning_rate": 9.845414277933514e-05, + "loss": 0.0296, + "step": 12540 + }, + { + "grad_norm": 0.20395700633525848, + "learning_rate": 9.845006043254771e-05, + "loss": 0.0265, + "step": 12550 + }, + { + "grad_norm": 0.23836259543895721, + "learning_rate": 9.844597278733451e-05, + "loss": 0.0227, + "step": 12560 + }, + { + "grad_norm": 0.11812478303909302, + "learning_rate": 9.844187984414259e-05, + "loss": 0.0214, + "step": 12570 + }, + { + "grad_norm": 0.13120566308498383, + "learning_rate": 9.84377816034195e-05, + "loss": 0.0218, + "step": 12580 + }, + { + "grad_norm": 0.1643330156803131, + "learning_rate": 9.843367806561345e-05, + "loss": 0.0269, + "step": 12590 + }, + { + "grad_norm": 0.18466365337371826, + "learning_rate": 9.842956923117317e-05, + "loss": 0.0219, + "step": 12600 + }, + { + "grad_norm": 0.2244631052017212, + "learning_rate": 9.842545510054802e-05, + "loss": 0.0192, + "step": 12610 + }, + { + "grad_norm": 0.20854896306991577, + "learning_rate": 9.842133567418792e-05, + "loss": 0.0196, + "step": 12620 + }, + { + "grad_norm": 0.16039612889289856, + "learning_rate": 9.841721095254333e-05, + "loss": 0.0177, + "step": 12630 + }, + { + "grad_norm": 0.22511406242847443, + "learning_rate": 9.841308093606537e-05, + "loss": 0.0271, + "step": 12640 + }, + { + "grad_norm": 0.19451841711997986, + "learning_rate": 9.840894562520565e-05, + "loss": 0.0228, + "step": 12650 + }, + { + "grad_norm": 0.20607230067253113, + "learning_rate": 9.840480502041642e-05, + "loss": 0.0269, + "step": 12660 + }, + { + "grad_norm": 0.1661987453699112, + "learning_rate": 9.840065912215049e-05, + "loss": 0.0224, + "step": 12670 + }, + { + "grad_norm": 0.13661257922649384, + "learning_rate": 9.839650793086124e-05, + "loss": 0.0317, + "step": 12680 + }, + { + "grad_norm": 0.17065025866031647, + "learning_rate": 9.839235144700265e-05, + "loss": 0.0263, + "step": 12690 + }, + { + "grad_norm": 0.26821181178092957, + "learning_rate": 9.838818967102926e-05, + "loss": 0.0226, + "step": 12700 + }, + { + "grad_norm": 0.1930248886346817, + "learning_rate": 9.83840226033962e-05, + "loss": 0.023, + "step": 12710 + }, + { + "grad_norm": 0.22751566767692566, + "learning_rate": 9.837985024455918e-05, + "loss": 0.0223, + "step": 12720 + }, + { + "grad_norm": 0.19473154842853546, + "learning_rate": 9.837567259497447e-05, + "loss": 0.0207, + "step": 12730 + }, + { + "grad_norm": 0.15432409942150116, + "learning_rate": 9.837148965509894e-05, + "loss": 0.024, + "step": 12740 + }, + { + "grad_norm": 0.17699211835861206, + "learning_rate": 9.836730142539001e-05, + "loss": 0.0185, + "step": 12750 + }, + { + "grad_norm": 0.22674058377742767, + "learning_rate": 9.836310790630574e-05, + "loss": 0.0266, + "step": 12760 + }, + { + "grad_norm": 0.21011611819267273, + "learning_rate": 9.83589090983047e-05, + "loss": 0.0273, + "step": 12770 + }, + { + "grad_norm": 0.33996444940567017, + "learning_rate": 9.835470500184605e-05, + "loss": 0.0266, + "step": 12780 + }, + { + "grad_norm": 0.21211223304271698, + "learning_rate": 9.835049561738957e-05, + "loss": 0.028, + "step": 12790 + }, + { + "grad_norm": 0.22107115387916565, + "learning_rate": 9.834628094539558e-05, + "loss": 0.0236, + "step": 12800 + }, + { + "grad_norm": 0.18725305795669556, + "learning_rate": 9.834206098632499e-05, + "loss": 0.0228, + "step": 12810 + }, + { + "grad_norm": 0.1533517688512802, + "learning_rate": 9.833783574063931e-05, + "loss": 0.0252, + "step": 12820 + }, + { + "grad_norm": 0.2761463224887848, + "learning_rate": 9.833360520880058e-05, + "loss": 0.0262, + "step": 12830 + }, + { + "grad_norm": 0.16616663336753845, + "learning_rate": 9.832936939127144e-05, + "loss": 0.0208, + "step": 12840 + }, + { + "grad_norm": 0.19194220006465912, + "learning_rate": 9.832512828851515e-05, + "loss": 0.0214, + "step": 12850 + }, + { + "grad_norm": 0.21045731008052826, + "learning_rate": 9.832088190099546e-05, + "loss": 0.0248, + "step": 12860 + }, + { + "grad_norm": 0.22156284749507904, + "learning_rate": 9.831663022917679e-05, + "loss": 0.0212, + "step": 12870 + }, + { + "grad_norm": 0.18087059259414673, + "learning_rate": 9.831237327352407e-05, + "loss": 0.018, + "step": 12880 + }, + { + "grad_norm": 0.19974391162395477, + "learning_rate": 9.830811103450286e-05, + "loss": 0.0214, + "step": 12890 + }, + { + "grad_norm": 0.1758899688720703, + "learning_rate": 9.830384351257924e-05, + "loss": 0.03, + "step": 12900 + }, + { + "grad_norm": 0.17871540784835815, + "learning_rate": 9.829957070821993e-05, + "loss": 0.0218, + "step": 12910 + }, + { + "grad_norm": 0.17277392745018005, + "learning_rate": 9.829529262189218e-05, + "loss": 0.0172, + "step": 12920 + }, + { + "grad_norm": 0.14195123314857483, + "learning_rate": 9.829100925406385e-05, + "loss": 0.0202, + "step": 12930 + }, + { + "grad_norm": 0.2592112123966217, + "learning_rate": 9.828672060520333e-05, + "loss": 0.0225, + "step": 12940 + }, + { + "grad_norm": 0.12915688753128052, + "learning_rate": 9.828242667577966e-05, + "loss": 0.0175, + "step": 12950 + }, + { + "grad_norm": 0.18046808242797852, + "learning_rate": 9.82781274662624e-05, + "loss": 0.0202, + "step": 12960 + }, + { + "grad_norm": 0.17385080456733704, + "learning_rate": 9.82738229771217e-05, + "loss": 0.0181, + "step": 12970 + }, + { + "grad_norm": 0.2113443911075592, + "learning_rate": 9.826951320882829e-05, + "loss": 0.0169, + "step": 12980 + }, + { + "grad_norm": 0.21402354538440704, + "learning_rate": 9.826519816185351e-05, + "loss": 0.0191, + "step": 12990 + }, + { + "grad_norm": 0.1944163292646408, + "learning_rate": 9.826087783666921e-05, + "loss": 0.0232, + "step": 13000 + }, + { + "grad_norm": 0.19149993360042572, + "learning_rate": 9.825655223374787e-05, + "loss": 0.038, + "step": 13010 + }, + { + "grad_norm": 0.22704754769802094, + "learning_rate": 9.825222135356253e-05, + "loss": 0.0323, + "step": 13020 + }, + { + "grad_norm": 0.15156039595603943, + "learning_rate": 9.82478851965868e-05, + "loss": 0.0319, + "step": 13030 + }, + { + "grad_norm": 0.15829913318157196, + "learning_rate": 9.82435437632949e-05, + "loss": 0.0217, + "step": 13040 + }, + { + "grad_norm": 0.15479987859725952, + "learning_rate": 9.823919705416158e-05, + "loss": 0.0196, + "step": 13050 + }, + { + "grad_norm": 0.22412778437137604, + "learning_rate": 9.82348450696622e-05, + "loss": 0.0207, + "step": 13060 + }, + { + "grad_norm": 0.23682750761508942, + "learning_rate": 9.823048781027268e-05, + "loss": 0.0316, + "step": 13070 + }, + { + "grad_norm": 0.26787540316581726, + "learning_rate": 9.822612527646953e-05, + "loss": 0.023, + "step": 13080 + }, + { + "grad_norm": 0.1484673023223877, + "learning_rate": 9.822175746872984e-05, + "loss": 0.0267, + "step": 13090 + }, + { + "grad_norm": 0.1588597595691681, + "learning_rate": 9.821738438753123e-05, + "loss": 0.0226, + "step": 13100 + }, + { + "grad_norm": 0.13903941214084625, + "learning_rate": 9.821300603335196e-05, + "loss": 0.0197, + "step": 13110 + }, + { + "grad_norm": 0.1881917119026184, + "learning_rate": 9.820862240667085e-05, + "loss": 0.0194, + "step": 13120 + }, + { + "grad_norm": 0.18558470904827118, + "learning_rate": 9.820423350796726e-05, + "loss": 0.0262, + "step": 13130 + }, + { + "grad_norm": 0.1924840658903122, + "learning_rate": 9.819983933772118e-05, + "loss": 0.024, + "step": 13140 + }, + { + "grad_norm": 0.28957125544548035, + "learning_rate": 9.819543989641314e-05, + "loss": 0.0242, + "step": 13150 + }, + { + "grad_norm": 0.22398625314235687, + "learning_rate": 9.819103518452423e-05, + "loss": 0.018, + "step": 13160 + }, + { + "grad_norm": 0.12985973060131073, + "learning_rate": 9.818662520253618e-05, + "loss": 0.022, + "step": 13170 + }, + { + "grad_norm": 0.1540331095457077, + "learning_rate": 9.818220995093126e-05, + "loss": 0.0227, + "step": 13180 + }, + { + "grad_norm": 0.23458382487297058, + "learning_rate": 9.817778943019228e-05, + "loss": 0.0179, + "step": 13190 + }, + { + "grad_norm": 0.26470619440078735, + "learning_rate": 9.81733636408027e-05, + "loss": 0.0235, + "step": 13200 + }, + { + "grad_norm": 0.15119107067584991, + "learning_rate": 9.816893258324649e-05, + "loss": 0.0162, + "step": 13210 + }, + { + "grad_norm": 0.16853472590446472, + "learning_rate": 9.816449625800823e-05, + "loss": 0.0182, + "step": 13220 + }, + { + "grad_norm": 0.1395794004201889, + "learning_rate": 9.816005466557308e-05, + "loss": 0.0223, + "step": 13230 + }, + { + "grad_norm": 0.2139284461736679, + "learning_rate": 9.815560780642674e-05, + "loss": 0.0291, + "step": 13240 + }, + { + "grad_norm": 0.18602004647254944, + "learning_rate": 9.815115568105555e-05, + "loss": 0.0219, + "step": 13250 + }, + { + "grad_norm": 0.25841423869132996, + "learning_rate": 9.814669828994638e-05, + "loss": 0.022, + "step": 13260 + }, + { + "grad_norm": 0.1906161606311798, + "learning_rate": 9.814223563358665e-05, + "loss": 0.022, + "step": 13270 + }, + { + "grad_norm": 0.2061898112297058, + "learning_rate": 9.813776771246443e-05, + "loss": 0.0196, + "step": 13280 + }, + { + "grad_norm": 0.1892378032207489, + "learning_rate": 9.813329452706829e-05, + "loss": 0.0207, + "step": 13290 + }, + { + "grad_norm": 0.18297891318798065, + "learning_rate": 9.812881607788744e-05, + "loss": 0.0182, + "step": 13300 + }, + { + "grad_norm": 0.18209515511989594, + "learning_rate": 9.812433236541163e-05, + "loss": 0.0236, + "step": 13310 + }, + { + "grad_norm": 0.2757410705089569, + "learning_rate": 9.811984339013116e-05, + "loss": 0.0259, + "step": 13320 + }, + { + "grad_norm": 0.18968841433525085, + "learning_rate": 9.811534915253698e-05, + "loss": 0.0225, + "step": 13330 + }, + { + "grad_norm": 0.18140247464179993, + "learning_rate": 9.811084965312056e-05, + "loss": 0.0211, + "step": 13340 + }, + { + "grad_norm": 0.16182762384414673, + "learning_rate": 9.810634489237396e-05, + "loss": 0.0181, + "step": 13350 + }, + { + "grad_norm": 0.18578748404979706, + "learning_rate": 9.81018348707898e-05, + "loss": 0.0209, + "step": 13360 + }, + { + "grad_norm": 0.13064348697662354, + "learning_rate": 9.809731958886131e-05, + "loss": 0.0253, + "step": 13370 + }, + { + "grad_norm": 0.3066904544830322, + "learning_rate": 9.809279904708224e-05, + "loss": 0.0258, + "step": 13380 + }, + { + "grad_norm": 0.19796819984912872, + "learning_rate": 9.808827324594699e-05, + "loss": 0.0239, + "step": 13390 + }, + { + "grad_norm": 0.2558414340019226, + "learning_rate": 9.808374218595046e-05, + "loss": 0.0301, + "step": 13400 + }, + { + "grad_norm": 0.21207885444164276, + "learning_rate": 9.80792058675882e-05, + "loss": 0.0235, + "step": 13410 + }, + { + "grad_norm": 0.14607496559619904, + "learning_rate": 9.807466429135627e-05, + "loss": 0.0248, + "step": 13420 + }, + { + "grad_norm": 0.1876640021800995, + "learning_rate": 9.807011745775132e-05, + "loss": 0.0204, + "step": 13430 + }, + { + "grad_norm": 0.19389186799526215, + "learning_rate": 9.806556536727061e-05, + "loss": 0.0216, + "step": 13440 + }, + { + "grad_norm": 0.2219102680683136, + "learning_rate": 9.806100802041193e-05, + "loss": 0.0263, + "step": 13450 + }, + { + "grad_norm": 0.17273353040218353, + "learning_rate": 9.805644541767368e-05, + "loss": 0.0345, + "step": 13460 + }, + { + "grad_norm": 0.19201411306858063, + "learning_rate": 9.805187755955478e-05, + "loss": 0.0344, + "step": 13470 + }, + { + "grad_norm": 0.1702917516231537, + "learning_rate": 9.804730444655483e-05, + "loss": 0.0326, + "step": 13480 + }, + { + "grad_norm": 0.22219261527061462, + "learning_rate": 9.804272607917388e-05, + "loss": 0.0268, + "step": 13490 + }, + { + "grad_norm": 0.2701045870780945, + "learning_rate": 9.803814245791265e-05, + "loss": 0.0176, + "step": 13500 + }, + { + "grad_norm": 0.18916724622249603, + "learning_rate": 9.803355358327239e-05, + "loss": 0.0268, + "step": 13510 + }, + { + "grad_norm": 0.20439498126506805, + "learning_rate": 9.802895945575492e-05, + "loss": 0.031, + "step": 13520 + }, + { + "grad_norm": 0.1777377426624298, + "learning_rate": 9.802436007586266e-05, + "loss": 0.0306, + "step": 13530 + }, + { + "grad_norm": 0.2089541107416153, + "learning_rate": 9.801975544409858e-05, + "loss": 0.027, + "step": 13540 + }, + { + "grad_norm": 0.1677665412425995, + "learning_rate": 9.801514556096625e-05, + "loss": 0.0208, + "step": 13550 + }, + { + "grad_norm": 0.19572888314723969, + "learning_rate": 9.801053042696977e-05, + "loss": 0.0266, + "step": 13560 + }, + { + "grad_norm": 0.15096737444400787, + "learning_rate": 9.800591004261388e-05, + "loss": 0.0215, + "step": 13570 + }, + { + "grad_norm": 0.2841437757015228, + "learning_rate": 9.800128440840385e-05, + "loss": 0.0239, + "step": 13580 + }, + { + "grad_norm": 0.1756044179201126, + "learning_rate": 9.799665352484552e-05, + "loss": 0.0329, + "step": 13590 + }, + { + "grad_norm": 0.2213645875453949, + "learning_rate": 9.799201739244532e-05, + "loss": 0.0381, + "step": 13600 + }, + { + "grad_norm": 0.18800993263721466, + "learning_rate": 9.798737601171025e-05, + "loss": 0.0213, + "step": 13610 + }, + { + "grad_norm": 0.16354207694530487, + "learning_rate": 9.79827293831479e-05, + "loss": 0.0319, + "step": 13620 + }, + { + "grad_norm": 0.2384779155254364, + "learning_rate": 9.797807750726638e-05, + "loss": 0.0226, + "step": 13630 + }, + { + "grad_norm": 0.19253763556480408, + "learning_rate": 9.797342038457446e-05, + "loss": 0.0209, + "step": 13640 + }, + { + "grad_norm": 0.20019499957561493, + "learning_rate": 9.796875801558141e-05, + "loss": 0.0265, + "step": 13650 + }, + { + "grad_norm": 0.22288911044597626, + "learning_rate": 9.79640904007971e-05, + "loss": 0.027, + "step": 13660 + }, + { + "grad_norm": 0.1982785016298294, + "learning_rate": 9.795941754073199e-05, + "loss": 0.0212, + "step": 13670 + }, + { + "grad_norm": 0.11598828434944153, + "learning_rate": 9.795473943589705e-05, + "loss": 0.0203, + "step": 13680 + }, + { + "grad_norm": 0.2284257560968399, + "learning_rate": 9.795005608680394e-05, + "loss": 0.0211, + "step": 13690 + }, + { + "grad_norm": 0.18853017687797546, + "learning_rate": 9.794536749396477e-05, + "loss": 0.0235, + "step": 13700 + }, + { + "grad_norm": 0.18426069617271423, + "learning_rate": 9.79406736578923e-05, + "loss": 0.0202, + "step": 13710 + }, + { + "grad_norm": 0.17854681611061096, + "learning_rate": 9.793597457909984e-05, + "loss": 0.0198, + "step": 13720 + }, + { + "grad_norm": 0.1685594767332077, + "learning_rate": 9.793127025810127e-05, + "loss": 0.0194, + "step": 13730 + }, + { + "grad_norm": 0.22527188062667847, + "learning_rate": 9.792656069541104e-05, + "loss": 0.0179, + "step": 13740 + }, + { + "grad_norm": 0.15930485725402832, + "learning_rate": 9.79218458915442e-05, + "loss": 0.0201, + "step": 13750 + }, + { + "grad_norm": 0.2720975875854492, + "learning_rate": 9.791712584701634e-05, + "loss": 0.0358, + "step": 13760 + }, + { + "grad_norm": 0.2056513875722885, + "learning_rate": 9.791240056234364e-05, + "loss": 0.0212, + "step": 13770 + }, + { + "grad_norm": 0.19017393887043, + "learning_rate": 9.790767003804283e-05, + "loss": 0.02, + "step": 13780 + }, + { + "grad_norm": 0.176192969083786, + "learning_rate": 9.790293427463126e-05, + "loss": 0.021, + "step": 13790 + }, + { + "grad_norm": 0.1970798522233963, + "learning_rate": 9.789819327262684e-05, + "loss": 0.0202, + "step": 13800 + }, + { + "grad_norm": 0.15416617691516876, + "learning_rate": 9.7893447032548e-05, + "loss": 0.0258, + "step": 13810 + }, + { + "grad_norm": 0.19955475628376007, + "learning_rate": 9.78886955549138e-05, + "loss": 0.0224, + "step": 13820 + }, + { + "grad_norm": 0.16509506106376648, + "learning_rate": 9.788393884024387e-05, + "loss": 0.0267, + "step": 13830 + }, + { + "grad_norm": 0.17359670996665955, + "learning_rate": 9.787917688905836e-05, + "loss": 0.0251, + "step": 13840 + }, + { + "grad_norm": 0.24574629962444305, + "learning_rate": 9.787440970187807e-05, + "loss": 0.0185, + "step": 13850 + }, + { + "grad_norm": 0.158518984913826, + "learning_rate": 9.786963727922429e-05, + "loss": 0.0222, + "step": 13860 + }, + { + "grad_norm": 0.12145864218473434, + "learning_rate": 9.786485962161897e-05, + "loss": 0.0221, + "step": 13870 + }, + { + "grad_norm": 0.15362100303173065, + "learning_rate": 9.786007672958455e-05, + "loss": 0.0178, + "step": 13880 + }, + { + "grad_norm": 0.26031774282455444, + "learning_rate": 9.78552886036441e-05, + "loss": 0.0269, + "step": 13890 + }, + { + "grad_norm": 0.14883166551589966, + "learning_rate": 9.785049524432124e-05, + "loss": 0.0194, + "step": 13900 + }, + { + "grad_norm": 0.1708686351776123, + "learning_rate": 9.784569665214016e-05, + "loss": 0.0201, + "step": 13910 + }, + { + "grad_norm": 0.20524881780147552, + "learning_rate": 9.784089282762563e-05, + "loss": 0.0224, + "step": 13920 + }, + { + "grad_norm": 0.23709675669670105, + "learning_rate": 9.7836083771303e-05, + "loss": 0.0223, + "step": 13930 + }, + { + "grad_norm": 0.16310928761959076, + "learning_rate": 9.783126948369817e-05, + "loss": 0.0196, + "step": 13940 + }, + { + "grad_norm": 0.13026170432567596, + "learning_rate": 9.78264499653376e-05, + "loss": 0.0198, + "step": 13950 + }, + { + "grad_norm": 0.14431704580783844, + "learning_rate": 9.782162521674838e-05, + "loss": 0.0207, + "step": 13960 + }, + { + "grad_norm": 0.20787157118320465, + "learning_rate": 9.781679523845812e-05, + "loss": 0.0379, + "step": 13970 + }, + { + "grad_norm": 0.1441018134355545, + "learning_rate": 9.781196003099502e-05, + "loss": 0.026, + "step": 13980 + }, + { + "grad_norm": 0.13051380217075348, + "learning_rate": 9.780711959488786e-05, + "loss": 0.0225, + "step": 13990 + }, + { + "grad_norm": 0.21909509599208832, + "learning_rate": 9.780227393066599e-05, + "loss": 0.0211, + "step": 14000 + }, + { + "grad_norm": 0.19448043406009674, + "learning_rate": 9.77974230388593e-05, + "loss": 0.0202, + "step": 14010 + }, + { + "grad_norm": 0.1597941666841507, + "learning_rate": 9.779256691999829e-05, + "loss": 0.0272, + "step": 14020 + }, + { + "grad_norm": 0.20353730022907257, + "learning_rate": 9.778770557461403e-05, + "loss": 0.0247, + "step": 14030 + }, + { + "grad_norm": 0.17038080096244812, + "learning_rate": 9.778283900323812e-05, + "loss": 0.0179, + "step": 14040 + }, + { + "grad_norm": 0.17785057425498962, + "learning_rate": 9.777796720640277e-05, + "loss": 0.0184, + "step": 14050 + }, + { + "grad_norm": 0.17627201974391937, + "learning_rate": 9.777309018464078e-05, + "loss": 0.0235, + "step": 14060 + }, + { + "grad_norm": 0.22092576324939728, + "learning_rate": 9.776820793848547e-05, + "loss": 0.0229, + "step": 14070 + }, + { + "grad_norm": 0.11341450363397598, + "learning_rate": 9.776332046847075e-05, + "loss": 0.0193, + "step": 14080 + }, + { + "grad_norm": 0.12342403829097748, + "learning_rate": 9.775842777513111e-05, + "loss": 0.0195, + "step": 14090 + }, + { + "grad_norm": 0.1861448734998703, + "learning_rate": 9.775352985900163e-05, + "loss": 0.0244, + "step": 14100 + }, + { + "grad_norm": 0.12468966096639633, + "learning_rate": 9.774862672061791e-05, + "loss": 0.0226, + "step": 14110 + }, + { + "grad_norm": 0.18342188000679016, + "learning_rate": 9.774371836051616e-05, + "loss": 0.0195, + "step": 14120 + }, + { + "grad_norm": 0.2583344578742981, + "learning_rate": 9.773880477923315e-05, + "loss": 0.028, + "step": 14130 + }, + { + "grad_norm": 0.1771797239780426, + "learning_rate": 9.773388597730623e-05, + "loss": 0.0275, + "step": 14140 + }, + { + "grad_norm": 0.20567385852336884, + "learning_rate": 9.77289619552733e-05, + "loss": 0.0207, + "step": 14150 + }, + { + "grad_norm": 0.31525325775146484, + "learning_rate": 9.772403271367285e-05, + "loss": 0.0338, + "step": 14160 + }, + { + "grad_norm": 0.24591059982776642, + "learning_rate": 9.771909825304396e-05, + "loss": 0.0235, + "step": 14170 + }, + { + "grad_norm": 0.18029507994651794, + "learning_rate": 9.771415857392619e-05, + "loss": 0.0226, + "step": 14180 + }, + { + "grad_norm": 0.22809664905071259, + "learning_rate": 9.770921367685978e-05, + "loss": 0.0252, + "step": 14190 + }, + { + "grad_norm": 0.24675194919109344, + "learning_rate": 9.770426356238551e-05, + "loss": 0.0236, + "step": 14200 + }, + { + "grad_norm": 0.24943700432777405, + "learning_rate": 9.769930823104469e-05, + "loss": 0.0223, + "step": 14210 + }, + { + "grad_norm": 0.23726001381874084, + "learning_rate": 9.769434768337926e-05, + "loss": 0.0265, + "step": 14220 + }, + { + "grad_norm": 0.19129405915737152, + "learning_rate": 9.768938191993164e-05, + "loss": 0.0224, + "step": 14230 + }, + { + "grad_norm": 0.24804796278476715, + "learning_rate": 9.768441094124494e-05, + "loss": 0.0213, + "step": 14240 + }, + { + "grad_norm": 0.18922118842601776, + "learning_rate": 9.767943474786275e-05, + "loss": 0.0241, + "step": 14250 + }, + { + "grad_norm": 0.14617517590522766, + "learning_rate": 9.767445334032923e-05, + "loss": 0.0203, + "step": 14260 + }, + { + "grad_norm": 0.19204413890838623, + "learning_rate": 9.766946671918919e-05, + "loss": 0.0269, + "step": 14270 + }, + { + "grad_norm": 0.20152969658374786, + "learning_rate": 9.766447488498796e-05, + "loss": 0.0222, + "step": 14280 + }, + { + "grad_norm": 0.17186246812343597, + "learning_rate": 9.765947783827139e-05, + "loss": 0.0191, + "step": 14290 + }, + { + "grad_norm": 0.2033412903547287, + "learning_rate": 9.765447557958599e-05, + "loss": 0.0225, + "step": 14300 + }, + { + "grad_norm": 0.22731895744800568, + "learning_rate": 9.764946810947879e-05, + "loss": 0.0195, + "step": 14310 + }, + { + "grad_norm": 0.16104164719581604, + "learning_rate": 9.764445542849738e-05, + "loss": 0.035, + "step": 14320 + }, + { + "grad_norm": 0.14508305490016937, + "learning_rate": 9.763943753718998e-05, + "loss": 0.0204, + "step": 14330 + }, + { + "grad_norm": 0.2084996998310089, + "learning_rate": 9.76344144361053e-05, + "loss": 0.0218, + "step": 14340 + }, + { + "grad_norm": 0.3200346827507019, + "learning_rate": 9.762938612579269e-05, + "loss": 0.0227, + "step": 14350 + }, + { + "grad_norm": 0.11472415179014206, + "learning_rate": 9.762435260680202e-05, + "loss": 0.0198, + "step": 14360 + }, + { + "grad_norm": 0.18148097395896912, + "learning_rate": 9.761931387968373e-05, + "loss": 0.0234, + "step": 14370 + }, + { + "grad_norm": 0.15157316625118256, + "learning_rate": 9.76142699449889e-05, + "loss": 0.02, + "step": 14380 + }, + { + "grad_norm": 0.2643571197986603, + "learning_rate": 9.760922080326908e-05, + "loss": 0.0261, + "step": 14390 + }, + { + "grad_norm": 0.18957151472568512, + "learning_rate": 9.760416645507644e-05, + "loss": 0.0228, + "step": 14400 + }, + { + "grad_norm": 0.2967056334018707, + "learning_rate": 9.759910690096375e-05, + "loss": 0.0224, + "step": 14410 + }, + { + "grad_norm": 0.20794577896595, + "learning_rate": 9.759404214148429e-05, + "loss": 0.0228, + "step": 14420 + }, + { + "grad_norm": 0.17462901771068573, + "learning_rate": 9.758897217719191e-05, + "loss": 0.0228, + "step": 14430 + }, + { + "grad_norm": 0.3013613522052765, + "learning_rate": 9.758389700864113e-05, + "loss": 0.0197, + "step": 14440 + }, + { + "grad_norm": 0.26826104521751404, + "learning_rate": 9.757881663638688e-05, + "loss": 0.023, + "step": 14450 + }, + { + "grad_norm": 0.11767105013132095, + "learning_rate": 9.757373106098478e-05, + "loss": 0.0203, + "step": 14460 + }, + { + "grad_norm": 0.17926104366779327, + "learning_rate": 9.756864028299097e-05, + "loss": 0.0228, + "step": 14470 + }, + { + "grad_norm": 0.1788896769285202, + "learning_rate": 9.75635443029622e-05, + "loss": 0.019, + "step": 14480 + }, + { + "grad_norm": 0.20952224731445312, + "learning_rate": 9.755844312145572e-05, + "loss": 0.0197, + "step": 14490 + }, + { + "grad_norm": 0.11812091618776321, + "learning_rate": 9.755333673902941e-05, + "loss": 0.026, + "step": 14500 + }, + { + "grad_norm": 0.20557008683681488, + "learning_rate": 9.75482251562417e-05, + "loss": 0.0224, + "step": 14510 + }, + { + "grad_norm": 0.15923477709293365, + "learning_rate": 9.754310837365155e-05, + "loss": 0.0178, + "step": 14520 + }, + { + "grad_norm": 0.19438670575618744, + "learning_rate": 9.753798639181856e-05, + "loss": 0.0363, + "step": 14530 + }, + { + "grad_norm": 0.19051246345043182, + "learning_rate": 9.753285921130286e-05, + "loss": 0.02, + "step": 14540 + }, + { + "grad_norm": 0.1433384120464325, + "learning_rate": 9.752772683266512e-05, + "loss": 0.0196, + "step": 14550 + }, + { + "grad_norm": 0.17908519506454468, + "learning_rate": 9.752258925646665e-05, + "loss": 0.0204, + "step": 14560 + }, + { + "grad_norm": 0.1297716647386551, + "learning_rate": 9.751744648326926e-05, + "loss": 0.0175, + "step": 14570 + }, + { + "grad_norm": 0.24698980152606964, + "learning_rate": 9.751229851363536e-05, + "loss": 0.0245, + "step": 14580 + }, + { + "grad_norm": 0.1593431830406189, + "learning_rate": 9.750714534812793e-05, + "loss": 0.034, + "step": 14590 + }, + { + "grad_norm": 0.1284456104040146, + "learning_rate": 9.750198698731053e-05, + "loss": 0.0294, + "step": 14600 + }, + { + "grad_norm": 0.16837169229984283, + "learning_rate": 9.749682343174722e-05, + "loss": 0.0184, + "step": 14610 + }, + { + "grad_norm": 0.1542179435491562, + "learning_rate": 9.749165468200272e-05, + "loss": 0.019, + "step": 14620 + }, + { + "grad_norm": 0.1604662835597992, + "learning_rate": 9.748648073864229e-05, + "loss": 0.02, + "step": 14630 + }, + { + "grad_norm": 0.2227831333875656, + "learning_rate": 9.748130160223168e-05, + "loss": 0.0207, + "step": 14640 + }, + { + "grad_norm": 0.18712344765663147, + "learning_rate": 9.747611727333734e-05, + "loss": 0.02, + "step": 14650 + }, + { + "grad_norm": 0.16670818626880646, + "learning_rate": 9.74709277525262e-05, + "loss": 0.0188, + "step": 14660 + }, + { + "grad_norm": 0.21042916178703308, + "learning_rate": 9.746573304036576e-05, + "loss": 0.0254, + "step": 14670 + }, + { + "grad_norm": 0.20920996367931366, + "learning_rate": 9.746053313742412e-05, + "loss": 0.0191, + "step": 14680 + }, + { + "grad_norm": 0.17584975063800812, + "learning_rate": 9.745532804426994e-05, + "loss": 0.0223, + "step": 14690 + }, + { + "grad_norm": 0.18833471834659576, + "learning_rate": 9.745011776147242e-05, + "loss": 0.0243, + "step": 14700 + }, + { + "grad_norm": 0.2266969531774521, + "learning_rate": 9.744490228960138e-05, + "loss": 0.025, + "step": 14710 + }, + { + "grad_norm": 0.22414347529411316, + "learning_rate": 9.743968162922713e-05, + "loss": 0.0253, + "step": 14720 + }, + { + "grad_norm": 0.17024505138397217, + "learning_rate": 9.743445578092064e-05, + "loss": 0.0235, + "step": 14730 + }, + { + "grad_norm": 0.16535751521587372, + "learning_rate": 9.742922474525338e-05, + "loss": 0.0285, + "step": 14740 + }, + { + "grad_norm": 0.17352423071861267, + "learning_rate": 9.742398852279741e-05, + "loss": 0.0271, + "step": 14750 + }, + { + "grad_norm": 0.20040717720985413, + "learning_rate": 9.741874711412535e-05, + "loss": 0.0255, + "step": 14760 + }, + { + "grad_norm": 0.18188050389289856, + "learning_rate": 9.741350051981042e-05, + "loss": 0.0215, + "step": 14770 + }, + { + "grad_norm": 0.1819373220205307, + "learning_rate": 9.740824874042633e-05, + "loss": 0.0207, + "step": 14780 + }, + { + "grad_norm": 0.16066209971904755, + "learning_rate": 9.740299177654746e-05, + "loss": 0.0178, + "step": 14790 + }, + { + "grad_norm": 0.28634339570999146, + "learning_rate": 9.739772962874867e-05, + "loss": 0.0221, + "step": 14800 + }, + { + "grad_norm": 0.13778288662433624, + "learning_rate": 9.739246229760541e-05, + "loss": 0.0184, + "step": 14810 + }, + { + "grad_norm": 0.1555616855621338, + "learning_rate": 9.738718978369376e-05, + "loss": 0.0212, + "step": 14820 + }, + { + "grad_norm": 0.1300973743200302, + "learning_rate": 9.738191208759025e-05, + "loss": 0.0298, + "step": 14830 + }, + { + "grad_norm": 0.1307215690612793, + "learning_rate": 9.73766292098721e-05, + "loss": 0.0358, + "step": 14840 + }, + { + "grad_norm": 0.24456113576889038, + "learning_rate": 9.737134115111699e-05, + "loss": 0.0388, + "step": 14850 + }, + { + "grad_norm": 0.15836361050605774, + "learning_rate": 9.736604791190323e-05, + "loss": 0.0241, + "step": 14860 + }, + { + "grad_norm": 0.18230853974819183, + "learning_rate": 9.73607494928097e-05, + "loss": 0.022, + "step": 14870 + }, + { + "grad_norm": 0.16585737466812134, + "learning_rate": 9.735544589441581e-05, + "loss": 0.0249, + "step": 14880 + }, + { + "grad_norm": 0.17573226988315582, + "learning_rate": 9.735013711730154e-05, + "loss": 0.0223, + "step": 14890 + }, + { + "grad_norm": 0.17305533587932587, + "learning_rate": 9.734482316204747e-05, + "loss": 0.0192, + "step": 14900 + }, + { + "grad_norm": 0.21089904010295868, + "learning_rate": 9.733950402923473e-05, + "loss": 0.0217, + "step": 14910 + }, + { + "grad_norm": 0.20229409635066986, + "learning_rate": 9.7334179719445e-05, + "loss": 0.0193, + "step": 14920 + }, + { + "grad_norm": 0.2923470437526703, + "learning_rate": 9.732885023326053e-05, + "loss": 0.0183, + "step": 14930 + }, + { + "grad_norm": 0.12000405043363571, + "learning_rate": 9.732351557126418e-05, + "loss": 0.0263, + "step": 14940 + }, + { + "grad_norm": 0.2534976899623871, + "learning_rate": 9.731817573403929e-05, + "loss": 0.0347, + "step": 14950 + }, + { + "grad_norm": 0.1990581750869751, + "learning_rate": 9.731283072216985e-05, + "loss": 0.0265, + "step": 14960 + }, + { + "grad_norm": 0.1513371616601944, + "learning_rate": 9.730748053624039e-05, + "loss": 0.025, + "step": 14970 + }, + { + "grad_norm": 0.17424747347831726, + "learning_rate": 9.730212517683598e-05, + "loss": 0.0272, + "step": 14980 + }, + { + "grad_norm": 0.17181776463985443, + "learning_rate": 9.729676464454228e-05, + "loss": 0.0247, + "step": 14990 + }, + { + "grad_norm": 0.24047744274139404, + "learning_rate": 9.72913989399455e-05, + "loss": 0.0262, + "step": 15000 + }, + { + "grad_norm": 0.16642619669437408, + "learning_rate": 9.728602806363242e-05, + "loss": 0.0215, + "step": 15010 + }, + { + "grad_norm": 0.24995094537734985, + "learning_rate": 9.728065201619043e-05, + "loss": 0.0212, + "step": 15020 + }, + { + "grad_norm": 0.2822050452232361, + "learning_rate": 9.727527079820742e-05, + "loss": 0.0204, + "step": 15030 + }, + { + "grad_norm": 0.13784967362880707, + "learning_rate": 9.726988441027186e-05, + "loss": 0.0307, + "step": 15040 + }, + { + "grad_norm": 0.1744374781847, + "learning_rate": 9.726449285297281e-05, + "loss": 0.0346, + "step": 15050 + }, + { + "grad_norm": 0.26359444856643677, + "learning_rate": 9.72590961268999e-05, + "loss": 0.0321, + "step": 15060 + }, + { + "grad_norm": 0.17985014617443085, + "learning_rate": 9.725369423264328e-05, + "loss": 0.0244, + "step": 15070 + }, + { + "grad_norm": 0.16373814642429352, + "learning_rate": 9.72482871707937e-05, + "loss": 0.0237, + "step": 15080 + }, + { + "grad_norm": 0.23996175825595856, + "learning_rate": 9.724287494194247e-05, + "loss": 0.0241, + "step": 15090 + }, + { + "grad_norm": 0.1508621722459793, + "learning_rate": 9.723745754668147e-05, + "loss": 0.0209, + "step": 15100 + }, + { + "grad_norm": 0.1979391872882843, + "learning_rate": 9.723203498560313e-05, + "loss": 0.0218, + "step": 15110 + }, + { + "grad_norm": 0.15029415488243103, + "learning_rate": 9.722660725930046e-05, + "loss": 0.0186, + "step": 15120 + }, + { + "grad_norm": 0.1532818078994751, + "learning_rate": 9.722117436836702e-05, + "loss": 0.0196, + "step": 15130 + }, + { + "grad_norm": 0.11257666349411011, + "learning_rate": 9.721573631339696e-05, + "loss": 0.0153, + "step": 15140 + }, + { + "grad_norm": 0.1832984983921051, + "learning_rate": 9.721029309498494e-05, + "loss": 0.0225, + "step": 15150 + }, + { + "grad_norm": 0.19913160800933838, + "learning_rate": 9.720484471372627e-05, + "loss": 0.0216, + "step": 15160 + }, + { + "grad_norm": 0.15556563436985016, + "learning_rate": 9.719939117021673e-05, + "loss": 0.0278, + "step": 15170 + }, + { + "grad_norm": 0.15178711712360382, + "learning_rate": 9.719393246505275e-05, + "loss": 0.0202, + "step": 15180 + }, + { + "grad_norm": 0.1809513419866562, + "learning_rate": 9.718846859883128e-05, + "loss": 0.0197, + "step": 15190 + }, + { + "grad_norm": 0.1799234002828598, + "learning_rate": 9.718299957214982e-05, + "loss": 0.0268, + "step": 15200 + }, + { + "grad_norm": 0.1379155069589615, + "learning_rate": 9.717752538560646e-05, + "loss": 0.0203, + "step": 15210 + }, + { + "grad_norm": 0.37399640679359436, + "learning_rate": 9.717204603979986e-05, + "loss": 0.0221, + "step": 15220 + }, + { + "grad_norm": 0.16043303906917572, + "learning_rate": 9.716656153532922e-05, + "loss": 0.0174, + "step": 15230 + }, + { + "grad_norm": 0.22133223712444305, + "learning_rate": 9.716107187279434e-05, + "loss": 0.0185, + "step": 15240 + }, + { + "grad_norm": 0.24428661167621613, + "learning_rate": 9.715557705279555e-05, + "loss": 0.0363, + "step": 15250 + }, + { + "grad_norm": 0.21249181032180786, + "learning_rate": 9.715007707593372e-05, + "loss": 0.0231, + "step": 15260 + }, + { + "grad_norm": 0.15947481989860535, + "learning_rate": 9.714457194281036e-05, + "loss": 0.0205, + "step": 15270 + }, + { + "grad_norm": 0.27906718850135803, + "learning_rate": 9.713906165402751e-05, + "loss": 0.0195, + "step": 15280 + }, + { + "grad_norm": 0.20777805149555206, + "learning_rate": 9.713354621018774e-05, + "loss": 0.0274, + "step": 15290 + }, + { + "grad_norm": 0.14752337336540222, + "learning_rate": 9.712802561189422e-05, + "loss": 0.0219, + "step": 15300 + }, + { + "grad_norm": 0.14230316877365112, + "learning_rate": 9.712249985975069e-05, + "loss": 0.0292, + "step": 15310 + }, + { + "grad_norm": 0.13026396930217743, + "learning_rate": 9.71169689543614e-05, + "loss": 0.0178, + "step": 15320 + }, + { + "grad_norm": 0.21577967703342438, + "learning_rate": 9.711143289633123e-05, + "loss": 0.0194, + "step": 15330 + }, + { + "grad_norm": 0.1306057721376419, + "learning_rate": 9.710589168626561e-05, + "loss": 0.0195, + "step": 15340 + }, + { + "grad_norm": 0.15792009234428406, + "learning_rate": 9.710034532477048e-05, + "loss": 0.0227, + "step": 15350 + }, + { + "grad_norm": 0.17315755784511566, + "learning_rate": 9.709479381245239e-05, + "loss": 0.0187, + "step": 15360 + }, + { + "grad_norm": 0.15572920441627502, + "learning_rate": 9.708923714991847e-05, + "loss": 0.022, + "step": 15370 + }, + { + "grad_norm": 0.16451194882392883, + "learning_rate": 9.708367533777638e-05, + "loss": 0.0211, + "step": 15380 + }, + { + "grad_norm": 0.19162417948246002, + "learning_rate": 9.707810837663431e-05, + "loss": 0.0201, + "step": 15390 + }, + { + "grad_norm": 0.1849055141210556, + "learning_rate": 9.707253626710113e-05, + "loss": 0.0192, + "step": 15400 + }, + { + "grad_norm": 0.16424880921840668, + "learning_rate": 9.706695900978613e-05, + "loss": 0.0176, + "step": 15410 + }, + { + "grad_norm": 0.1317879855632782, + "learning_rate": 9.706137660529926e-05, + "loss": 0.0198, + "step": 15420 + }, + { + "grad_norm": 0.20026306807994843, + "learning_rate": 9.705578905425101e-05, + "loss": 0.0321, + "step": 15430 + }, + { + "grad_norm": 0.1635892391204834, + "learning_rate": 9.705019635725241e-05, + "loss": 0.0205, + "step": 15440 + }, + { + "grad_norm": 0.12891200184822083, + "learning_rate": 9.704459851491508e-05, + "loss": 0.0266, + "step": 15450 + }, + { + "grad_norm": 0.13116709887981415, + "learning_rate": 9.703899552785118e-05, + "loss": 0.0206, + "step": 15460 + }, + { + "grad_norm": 0.1506842076778412, + "learning_rate": 9.703338739667346e-05, + "loss": 0.0182, + "step": 15470 + }, + { + "grad_norm": 0.2451898604631424, + "learning_rate": 9.70277741219952e-05, + "loss": 0.0243, + "step": 15480 + }, + { + "grad_norm": 0.43695685267448425, + "learning_rate": 9.702215570443027e-05, + "loss": 0.0242, + "step": 15490 + }, + { + "grad_norm": 0.19942164421081543, + "learning_rate": 9.701653214459309e-05, + "loss": 0.0209, + "step": 15500 + }, + { + "grad_norm": 0.19791483879089355, + "learning_rate": 9.701090344309865e-05, + "loss": 0.0258, + "step": 15510 + }, + { + "grad_norm": 0.15242278575897217, + "learning_rate": 9.700526960056247e-05, + "loss": 0.0172, + "step": 15520 + }, + { + "grad_norm": 0.12601499259471893, + "learning_rate": 9.699963061760068e-05, + "loss": 0.0181, + "step": 15530 + }, + { + "grad_norm": 0.1721322238445282, + "learning_rate": 9.699398649482997e-05, + "loss": 0.0193, + "step": 15540 + }, + { + "grad_norm": 0.17007143795490265, + "learning_rate": 9.698833723286753e-05, + "loss": 0.0196, + "step": 15550 + }, + { + "grad_norm": 0.2079053670167923, + "learning_rate": 9.698268283233118e-05, + "loss": 0.0205, + "step": 15560 + }, + { + "grad_norm": 0.2198907434940338, + "learning_rate": 9.697702329383929e-05, + "loss": 0.0249, + "step": 15570 + }, + { + "grad_norm": 0.15883125364780426, + "learning_rate": 9.697135861801074e-05, + "loss": 0.0262, + "step": 15580 + }, + { + "grad_norm": 0.1725255697965622, + "learning_rate": 9.696568880546505e-05, + "loss": 0.0209, + "step": 15590 + }, + { + "grad_norm": 0.1504587084054947, + "learning_rate": 9.696001385682223e-05, + "loss": 0.0197, + "step": 15600 + }, + { + "grad_norm": 0.19720175862312317, + "learning_rate": 9.695433377270291e-05, + "loss": 0.0236, + "step": 15610 + }, + { + "grad_norm": 0.15130719542503357, + "learning_rate": 9.694864855372824e-05, + "loss": 0.0193, + "step": 15620 + }, + { + "grad_norm": 0.16259866952896118, + "learning_rate": 9.694295820051995e-05, + "loss": 0.0195, + "step": 15630 + }, + { + "grad_norm": 0.14147628843784332, + "learning_rate": 9.693726271370032e-05, + "loss": 0.0217, + "step": 15640 + }, + { + "grad_norm": 0.14053097367286682, + "learning_rate": 9.693156209389221e-05, + "loss": 0.02, + "step": 15650 + }, + { + "grad_norm": 0.1870625615119934, + "learning_rate": 9.692585634171905e-05, + "loss": 0.0226, + "step": 15660 + }, + { + "grad_norm": 0.18313618004322052, + "learning_rate": 9.692014545780476e-05, + "loss": 0.0193, + "step": 15670 + }, + { + "grad_norm": 0.1535652130842209, + "learning_rate": 9.691442944277393e-05, + "loss": 0.017, + "step": 15680 + }, + { + "grad_norm": 0.1558559387922287, + "learning_rate": 9.690870829725162e-05, + "loss": 0.0192, + "step": 15690 + }, + { + "grad_norm": 0.2124132513999939, + "learning_rate": 9.69029820218635e-05, + "loss": 0.02, + "step": 15700 + }, + { + "grad_norm": 0.23267558217048645, + "learning_rate": 9.689725061723579e-05, + "loss": 0.0243, + "step": 15710 + }, + { + "grad_norm": 0.16114464402198792, + "learning_rate": 9.689151408399527e-05, + "loss": 0.0188, + "step": 15720 + }, + { + "grad_norm": 0.20184509456157684, + "learning_rate": 9.688577242276924e-05, + "loss": 0.0281, + "step": 15730 + }, + { + "grad_norm": 0.15289756655693054, + "learning_rate": 9.688002563418566e-05, + "loss": 0.0207, + "step": 15740 + }, + { + "grad_norm": 0.22334766387939453, + "learning_rate": 9.687427371887293e-05, + "loss": 0.026, + "step": 15750 + }, + { + "grad_norm": 0.2039691060781479, + "learning_rate": 9.686851667746012e-05, + "loss": 0.0205, + "step": 15760 + }, + { + "grad_norm": 0.25130152702331543, + "learning_rate": 9.686275451057677e-05, + "loss": 0.0234, + "step": 15770 + }, + { + "grad_norm": 0.16842985153198242, + "learning_rate": 9.685698721885308e-05, + "loss": 0.0226, + "step": 15780 + }, + { + "grad_norm": 0.2182195633649826, + "learning_rate": 9.68512148029197e-05, + "loss": 0.0251, + "step": 15790 + }, + { + "grad_norm": 0.22064755856990814, + "learning_rate": 9.684543726340791e-05, + "loss": 0.0241, + "step": 15800 + }, + { + "grad_norm": 0.20407047867774963, + "learning_rate": 9.683965460094952e-05, + "loss": 0.021, + "step": 15810 + }, + { + "grad_norm": 0.16670729219913483, + "learning_rate": 9.683386681617694e-05, + "loss": 0.0202, + "step": 15820 + }, + { + "grad_norm": 0.5364972949028015, + "learning_rate": 9.68280739097231e-05, + "loss": 0.0196, + "step": 15830 + }, + { + "grad_norm": 0.2008659690618515, + "learning_rate": 9.682227588222148e-05, + "loss": 0.0186, + "step": 15840 + }, + { + "grad_norm": 0.19648461043834686, + "learning_rate": 9.681647273430618e-05, + "loss": 0.0202, + "step": 15850 + }, + { + "grad_norm": 0.12721413373947144, + "learning_rate": 9.681066446661182e-05, + "loss": 0.0183, + "step": 15860 + }, + { + "grad_norm": 0.17354756593704224, + "learning_rate": 9.680485107977357e-05, + "loss": 0.0223, + "step": 15870 + }, + { + "grad_norm": 0.13167676329612732, + "learning_rate": 9.679903257442716e-05, + "loss": 0.0225, + "step": 15880 + }, + { + "grad_norm": 0.1604689508676529, + "learning_rate": 9.679320895120891e-05, + "loss": 0.0199, + "step": 15890 + }, + { + "grad_norm": 0.12434596568346024, + "learning_rate": 9.67873802107557e-05, + "loss": 0.0152, + "step": 15900 + }, + { + "grad_norm": 0.15131287276744843, + "learning_rate": 9.67815463537049e-05, + "loss": 0.0169, + "step": 15910 + }, + { + "grad_norm": 0.20170335471630096, + "learning_rate": 9.677570738069457e-05, + "loss": 0.0216, + "step": 15920 + }, + { + "grad_norm": 0.16980966925621033, + "learning_rate": 9.676986329236318e-05, + "loss": 0.0212, + "step": 15930 + }, + { + "grad_norm": 0.16614173352718353, + "learning_rate": 9.676401408934987e-05, + "loss": 0.0203, + "step": 15940 + }, + { + "grad_norm": 0.1673685610294342, + "learning_rate": 9.675815977229428e-05, + "loss": 0.0187, + "step": 15950 + }, + { + "grad_norm": 0.12713928520679474, + "learning_rate": 9.675230034183664e-05, + "loss": 0.0166, + "step": 15960 + }, + { + "grad_norm": 0.16778503358364105, + "learning_rate": 9.674643579861773e-05, + "loss": 0.0208, + "step": 15970 + }, + { + "grad_norm": 0.19937868416309357, + "learning_rate": 9.674056614327886e-05, + "loss": 0.016, + "step": 15980 + }, + { + "grad_norm": 0.214517280459404, + "learning_rate": 9.673469137646198e-05, + "loss": 0.023, + "step": 15990 + }, + { + "grad_norm": 0.15694685280323029, + "learning_rate": 9.67288114988095e-05, + "loss": 0.0191, + "step": 16000 + }, + { + "grad_norm": 0.19046202301979065, + "learning_rate": 9.672292651096447e-05, + "loss": 0.0193, + "step": 16010 + }, + { + "grad_norm": 0.1726117879152298, + "learning_rate": 9.671703641357042e-05, + "loss": 0.0188, + "step": 16020 + }, + { + "grad_norm": 0.1297914832830429, + "learning_rate": 9.67111412072715e-05, + "loss": 0.0195, + "step": 16030 + }, + { + "grad_norm": 0.14712730050086975, + "learning_rate": 9.670524089271242e-05, + "loss": 0.0194, + "step": 16040 + }, + { + "grad_norm": 0.12155260145664215, + "learning_rate": 9.669933547053842e-05, + "loss": 0.0178, + "step": 16050 + }, + { + "grad_norm": 0.1876615434885025, + "learning_rate": 9.669342494139531e-05, + "loss": 0.0195, + "step": 16060 + }, + { + "grad_norm": 0.22119666635990143, + "learning_rate": 9.668750930592943e-05, + "loss": 0.0173, + "step": 16070 + }, + { + "grad_norm": 0.16617971658706665, + "learning_rate": 9.668158856478775e-05, + "loss": 0.0186, + "step": 16080 + }, + { + "grad_norm": 0.16504545509815216, + "learning_rate": 9.66756627186177e-05, + "loss": 0.0203, + "step": 16090 + }, + { + "grad_norm": 0.11611940711736679, + "learning_rate": 9.666973176806737e-05, + "loss": 0.0202, + "step": 16100 + }, + { + "grad_norm": 0.22529837489128113, + "learning_rate": 9.666379571378534e-05, + "loss": 0.0211, + "step": 16110 + }, + { + "grad_norm": 0.19440312683582306, + "learning_rate": 9.665785455642076e-05, + "loss": 0.0253, + "step": 16120 + }, + { + "grad_norm": 0.26122716069221497, + "learning_rate": 9.665190829662337e-05, + "loss": 0.0217, + "step": 16130 + }, + { + "grad_norm": 0.2044735997915268, + "learning_rate": 9.664595693504342e-05, + "loss": 0.0185, + "step": 16140 + }, + { + "grad_norm": 0.16995391249656677, + "learning_rate": 9.664000047233175e-05, + "loss": 0.0256, + "step": 16150 + }, + { + "grad_norm": 0.14951781928539276, + "learning_rate": 9.663403890913976e-05, + "loss": 0.0212, + "step": 16160 + }, + { + "grad_norm": 0.17346349358558655, + "learning_rate": 9.662807224611938e-05, + "loss": 0.0212, + "step": 16170 + }, + { + "grad_norm": 0.14404243230819702, + "learning_rate": 9.662210048392311e-05, + "loss": 0.0163, + "step": 16180 + }, + { + "grad_norm": 0.20005543529987335, + "learning_rate": 9.661612362320405e-05, + "loss": 0.0184, + "step": 16190 + }, + { + "grad_norm": 0.13963541388511658, + "learning_rate": 9.661014166461579e-05, + "loss": 0.0205, + "step": 16200 + }, + { + "grad_norm": 0.15437063574790955, + "learning_rate": 9.66041546088125e-05, + "loss": 0.0269, + "step": 16210 + }, + { + "grad_norm": 0.14778704941272736, + "learning_rate": 9.659816245644895e-05, + "loss": 0.0182, + "step": 16220 + }, + { + "grad_norm": 0.13394783437252045, + "learning_rate": 9.65921652081804e-05, + "loss": 0.0241, + "step": 16230 + }, + { + "grad_norm": 0.25377240777015686, + "learning_rate": 9.658616286466271e-05, + "loss": 0.0237, + "step": 16240 + }, + { + "grad_norm": 0.23287169635295868, + "learning_rate": 9.65801554265523e-05, + "loss": 0.0195, + "step": 16250 + }, + { + "grad_norm": 0.24727307260036469, + "learning_rate": 9.657414289450612e-05, + "loss": 0.0206, + "step": 16260 + }, + { + "grad_norm": 0.1427251100540161, + "learning_rate": 9.656812526918171e-05, + "loss": 0.0206, + "step": 16270 + }, + { + "grad_norm": 0.16701585054397583, + "learning_rate": 9.656210255123712e-05, + "loss": 0.0243, + "step": 16280 + }, + { + "grad_norm": 0.34902223944664, + "learning_rate": 9.6556074741331e-05, + "loss": 0.0203, + "step": 16290 + }, + { + "grad_norm": 0.12792527675628662, + "learning_rate": 9.655004184012256e-05, + "loss": 0.017, + "step": 16300 + }, + { + "grad_norm": 0.27711719274520874, + "learning_rate": 9.654400384827152e-05, + "loss": 0.0233, + "step": 16310 + }, + { + "grad_norm": 0.1507222205400467, + "learning_rate": 9.653796076643818e-05, + "loss": 0.0184, + "step": 16320 + }, + { + "grad_norm": 0.18236646056175232, + "learning_rate": 9.653191259528344e-05, + "loss": 0.0291, + "step": 16330 + }, + { + "grad_norm": 0.1331816017627716, + "learning_rate": 9.65258593354687e-05, + "loss": 0.0184, + "step": 16340 + }, + { + "grad_norm": 0.16813215613365173, + "learning_rate": 9.651980098765591e-05, + "loss": 0.0159, + "step": 16350 + }, + { + "grad_norm": 0.20344854891300201, + "learning_rate": 9.651373755250765e-05, + "loss": 0.0203, + "step": 16360 + }, + { + "grad_norm": 0.2824576497077942, + "learning_rate": 9.650766903068697e-05, + "loss": 0.0261, + "step": 16370 + }, + { + "grad_norm": 0.1959116905927658, + "learning_rate": 9.650159542285753e-05, + "loss": 0.0209, + "step": 16380 + }, + { + "grad_norm": 0.1596689373254776, + "learning_rate": 9.649551672968353e-05, + "loss": 0.0251, + "step": 16390 + }, + { + "grad_norm": 0.17274928092956543, + "learning_rate": 9.648943295182973e-05, + "loss": 0.0203, + "step": 16400 + }, + { + "grad_norm": 0.20984774827957153, + "learning_rate": 9.648334408996144e-05, + "loss": 0.0217, + "step": 16410 + }, + { + "grad_norm": 0.14616049826145172, + "learning_rate": 9.647725014474452e-05, + "loss": 0.0193, + "step": 16420 + }, + { + "grad_norm": 0.16566751897335052, + "learning_rate": 9.64711511168454e-05, + "loss": 0.021, + "step": 16430 + }, + { + "grad_norm": 0.21652109920978546, + "learning_rate": 9.646504700693108e-05, + "loss": 0.0184, + "step": 16440 + }, + { + "grad_norm": 0.1672782450914383, + "learning_rate": 9.645893781566907e-05, + "loss": 0.0177, + "step": 16450 + }, + { + "grad_norm": 0.1438208669424057, + "learning_rate": 9.645282354372744e-05, + "loss": 0.021, + "step": 16460 + }, + { + "grad_norm": 0.14506752789020538, + "learning_rate": 9.644670419177491e-05, + "loss": 0.0174, + "step": 16470 + }, + { + "grad_norm": 0.21750515699386597, + "learning_rate": 9.644057976048062e-05, + "loss": 0.0272, + "step": 16480 + }, + { + "grad_norm": 0.1513761579990387, + "learning_rate": 9.643445025051435e-05, + "loss": 0.0223, + "step": 16490 + }, + { + "grad_norm": 0.2190374881029129, + "learning_rate": 9.642831566254641e-05, + "loss": 0.0189, + "step": 16500 + }, + { + "grad_norm": 0.17759716510772705, + "learning_rate": 9.642217599724769e-05, + "loss": 0.0188, + "step": 16510 + }, + { + "grad_norm": 0.18366079032421112, + "learning_rate": 9.64160312552896e-05, + "loss": 0.023, + "step": 16520 + }, + { + "grad_norm": 0.17398585379123688, + "learning_rate": 9.64098814373441e-05, + "loss": 0.0199, + "step": 16530 + }, + { + "grad_norm": 0.2863290011882782, + "learning_rate": 9.640372654408374e-05, + "loss": 0.0186, + "step": 16540 + }, + { + "grad_norm": 0.16058862209320068, + "learning_rate": 9.639756657618162e-05, + "loss": 0.0201, + "step": 16550 + }, + { + "grad_norm": 0.18812903761863708, + "learning_rate": 9.639140153431138e-05, + "loss": 0.0223, + "step": 16560 + }, + { + "grad_norm": 0.1370997577905655, + "learning_rate": 9.638523141914721e-05, + "loss": 0.021, + "step": 16570 + }, + { + "grad_norm": 0.17544087767601013, + "learning_rate": 9.637905623136388e-05, + "loss": 0.0248, + "step": 16580 + }, + { + "grad_norm": 0.13355325162410736, + "learning_rate": 9.637287597163669e-05, + "loss": 0.017, + "step": 16590 + }, + { + "grad_norm": 0.13413205742835999, + "learning_rate": 9.63666906406415e-05, + "loss": 0.0169, + "step": 16600 + }, + { + "grad_norm": 0.15648318827152252, + "learning_rate": 9.636050023905473e-05, + "loss": 0.0149, + "step": 16610 + }, + { + "grad_norm": 0.12259843945503235, + "learning_rate": 9.635430476755336e-05, + "loss": 0.0147, + "step": 16620 + }, + { + "grad_norm": 0.1691352128982544, + "learning_rate": 9.63481042268149e-05, + "loss": 0.0155, + "step": 16630 + }, + { + "grad_norm": 0.23620128631591797, + "learning_rate": 9.634189861751745e-05, + "loss": 0.0228, + "step": 16640 + }, + { + "grad_norm": 0.1666763424873352, + "learning_rate": 9.633568794033967e-05, + "loss": 0.0215, + "step": 16650 + }, + { + "grad_norm": 0.19014954566955566, + "learning_rate": 9.63294721959607e-05, + "loss": 0.0191, + "step": 16660 + }, + { + "grad_norm": 0.18404936790466309, + "learning_rate": 9.63232513850603e-05, + "loss": 0.0317, + "step": 16670 + }, + { + "grad_norm": 0.19489778578281403, + "learning_rate": 9.631702550831878e-05, + "loss": 0.0215, + "step": 16680 + }, + { + "grad_norm": 0.20426806807518005, + "learning_rate": 9.631079456641698e-05, + "loss": 0.0219, + "step": 16690 + }, + { + "grad_norm": 0.14335012435913086, + "learning_rate": 9.630455856003632e-05, + "loss": 0.0184, + "step": 16700 + }, + { + "grad_norm": 0.18258005380630493, + "learning_rate": 9.629831748985876e-05, + "loss": 0.0237, + "step": 16710 + }, + { + "grad_norm": 0.2831587791442871, + "learning_rate": 9.629207135656679e-05, + "loss": 0.0329, + "step": 16720 + }, + { + "grad_norm": 0.23784853518009186, + "learning_rate": 9.628582016084353e-05, + "loss": 0.0324, + "step": 16730 + }, + { + "grad_norm": 0.18631325662136078, + "learning_rate": 9.627956390337254e-05, + "loss": 0.0189, + "step": 16740 + }, + { + "grad_norm": 0.1817905157804489, + "learning_rate": 9.627330258483802e-05, + "loss": 0.0292, + "step": 16750 + }, + { + "grad_norm": 0.17063161730766296, + "learning_rate": 9.62670362059247e-05, + "loss": 0.0193, + "step": 16760 + }, + { + "grad_norm": 0.168016254901886, + "learning_rate": 9.626076476731786e-05, + "loss": 0.0191, + "step": 16770 + }, + { + "grad_norm": 0.17138415575027466, + "learning_rate": 9.625448826970336e-05, + "loss": 0.0212, + "step": 16780 + }, + { + "grad_norm": 0.13529685139656067, + "learning_rate": 9.624820671376755e-05, + "loss": 0.0172, + "step": 16790 + }, + { + "grad_norm": 0.1677829623222351, + "learning_rate": 9.62419201001974e-05, + "loss": 0.0192, + "step": 16800 + }, + { + "grad_norm": 0.15738433599472046, + "learning_rate": 9.623562842968037e-05, + "loss": 0.0169, + "step": 16810 + }, + { + "grad_norm": 0.13722850382328033, + "learning_rate": 9.622933170290454e-05, + "loss": 0.0169, + "step": 16820 + }, + { + "grad_norm": 0.12585514783859253, + "learning_rate": 9.622302992055849e-05, + "loss": 0.0193, + "step": 16830 + }, + { + "grad_norm": 0.13715484738349915, + "learning_rate": 9.62167230833314e-05, + "loss": 0.0161, + "step": 16840 + }, + { + "grad_norm": 0.1492435485124588, + "learning_rate": 9.621041119191295e-05, + "loss": 0.0185, + "step": 16850 + }, + { + "grad_norm": 0.13320568203926086, + "learning_rate": 9.620409424699342e-05, + "loss": 0.023, + "step": 16860 + }, + { + "grad_norm": 0.12619438767433167, + "learning_rate": 9.619777224926359e-05, + "loss": 0.017, + "step": 16870 + }, + { + "grad_norm": 0.19407665729522705, + "learning_rate": 9.619144519941485e-05, + "loss": 0.0173, + "step": 16880 + }, + { + "grad_norm": 0.18669100105762482, + "learning_rate": 9.618511309813912e-05, + "loss": 0.0183, + "step": 16890 + }, + { + "grad_norm": 0.1230948269367218, + "learning_rate": 9.617877594612886e-05, + "loss": 0.0176, + "step": 16900 + }, + { + "grad_norm": 0.1145622655749321, + "learning_rate": 9.617243374407707e-05, + "loss": 0.0176, + "step": 16910 + }, + { + "grad_norm": 0.11539571732282639, + "learning_rate": 9.616608649267736e-05, + "loss": 0.0172, + "step": 16920 + }, + { + "grad_norm": 0.18856383860111237, + "learning_rate": 9.615973419262385e-05, + "loss": 0.0188, + "step": 16930 + }, + { + "grad_norm": 0.1374579817056656, + "learning_rate": 9.615337684461119e-05, + "loss": 0.0179, + "step": 16940 + }, + { + "grad_norm": 0.20111244916915894, + "learning_rate": 9.614701444933465e-05, + "loss": 0.0247, + "step": 16950 + }, + { + "grad_norm": 0.20620639622211456, + "learning_rate": 9.614064700748997e-05, + "loss": 0.0196, + "step": 16960 + }, + { + "grad_norm": 0.2019098848104477, + "learning_rate": 9.613427451977352e-05, + "loss": 0.0215, + "step": 16970 + }, + { + "grad_norm": 0.14879979193210602, + "learning_rate": 9.612789698688216e-05, + "loss": 0.0194, + "step": 16980 + }, + { + "grad_norm": 0.21490399539470673, + "learning_rate": 9.612151440951334e-05, + "loss": 0.0226, + "step": 16990 + }, + { + "grad_norm": 0.23071619868278503, + "learning_rate": 9.611512678836506e-05, + "loss": 0.0218, + "step": 17000 + }, + { + "grad_norm": 0.25268909335136414, + "learning_rate": 9.610873412413584e-05, + "loss": 0.0287, + "step": 17010 + }, + { + "grad_norm": 0.23127177357673645, + "learning_rate": 9.610233641752476e-05, + "loss": 0.0284, + "step": 17020 + }, + { + "grad_norm": 0.19044871628284454, + "learning_rate": 9.609593366923151e-05, + "loss": 0.0201, + "step": 17030 + }, + { + "grad_norm": 0.18184806406497955, + "learning_rate": 9.608952587995625e-05, + "loss": 0.0187, + "step": 17040 + }, + { + "grad_norm": 0.1722000241279602, + "learning_rate": 9.608311305039972e-05, + "loss": 0.0196, + "step": 17050 + }, + { + "grad_norm": 0.33803096413612366, + "learning_rate": 9.607669518126326e-05, + "loss": 0.0268, + "step": 17060 + }, + { + "grad_norm": 0.15279702842235565, + "learning_rate": 9.607027227324866e-05, + "loss": 0.0173, + "step": 17070 + }, + { + "grad_norm": 0.19342657923698425, + "learning_rate": 9.606384432705837e-05, + "loss": 0.0166, + "step": 17080 + }, + { + "grad_norm": 0.12429074943065643, + "learning_rate": 9.60574113433953e-05, + "loss": 0.0195, + "step": 17090 + }, + { + "grad_norm": 0.11971762776374817, + "learning_rate": 9.6050973322963e-05, + "loss": 0.0172, + "step": 17100 + }, + { + "grad_norm": 0.1515258252620697, + "learning_rate": 9.604453026646547e-05, + "loss": 0.0198, + "step": 17110 + }, + { + "grad_norm": 0.19088582694530487, + "learning_rate": 9.603808217460735e-05, + "loss": 0.0176, + "step": 17120 + }, + { + "grad_norm": 0.211159348487854, + "learning_rate": 9.603162904809377e-05, + "loss": 0.0186, + "step": 17130 + }, + { + "grad_norm": 0.11558739095926285, + "learning_rate": 9.602517088763045e-05, + "loss": 0.0147, + "step": 17140 + }, + { + "grad_norm": 0.13939666748046875, + "learning_rate": 9.601870769392365e-05, + "loss": 0.0214, + "step": 17150 + }, + { + "grad_norm": 0.13686273992061615, + "learning_rate": 9.601223946768017e-05, + "loss": 0.0199, + "step": 17160 + }, + { + "grad_norm": 0.1269262135028839, + "learning_rate": 9.600576620960734e-05, + "loss": 0.0204, + "step": 17170 + }, + { + "grad_norm": 0.2078564167022705, + "learning_rate": 9.599928792041308e-05, + "loss": 0.019, + "step": 17180 + }, + { + "grad_norm": 0.20139223337173462, + "learning_rate": 9.599280460080587e-05, + "loss": 0.0174, + "step": 17190 + }, + { + "grad_norm": 0.16742077469825745, + "learning_rate": 9.59863162514947e-05, + "loss": 0.0198, + "step": 17200 + }, + { + "grad_norm": 0.1732303500175476, + "learning_rate": 9.597982287318911e-05, + "loss": 0.0179, + "step": 17210 + }, + { + "grad_norm": 0.13972607254981995, + "learning_rate": 9.597332446659923e-05, + "loss": 0.0184, + "step": 17220 + }, + { + "grad_norm": 0.15899419784545898, + "learning_rate": 9.59668210324357e-05, + "loss": 0.0178, + "step": 17230 + }, + { + "grad_norm": 0.18167293071746826, + "learning_rate": 9.596031257140974e-05, + "loss": 0.0175, + "step": 17240 + }, + { + "grad_norm": 0.23579664528369904, + "learning_rate": 9.59537990842331e-05, + "loss": 0.0276, + "step": 17250 + }, + { + "grad_norm": 0.19228799641132355, + "learning_rate": 9.594728057161806e-05, + "loss": 0.0181, + "step": 17260 + }, + { + "grad_norm": 0.17511771619319916, + "learning_rate": 9.594075703427752e-05, + "loss": 0.0188, + "step": 17270 + }, + { + "grad_norm": 0.12719599902629852, + "learning_rate": 9.593422847292486e-05, + "loss": 0.0158, + "step": 17280 + }, + { + "grad_norm": 0.15555448830127716, + "learning_rate": 9.592769488827402e-05, + "loss": 0.0194, + "step": 17290 + }, + { + "grad_norm": 0.1316969096660614, + "learning_rate": 9.592115628103952e-05, + "loss": 0.0179, + "step": 17300 + }, + { + "grad_norm": 0.08673807978630066, + "learning_rate": 9.591461265193643e-05, + "loss": 0.0169, + "step": 17310 + }, + { + "grad_norm": 0.12836283445358276, + "learning_rate": 9.590806400168032e-05, + "loss": 0.0142, + "step": 17320 + }, + { + "grad_norm": 0.12820006906986237, + "learning_rate": 9.590151033098735e-05, + "loss": 0.0159, + "step": 17330 + }, + { + "grad_norm": 0.17789334058761597, + "learning_rate": 9.589495164057423e-05, + "loss": 0.0189, + "step": 17340 + }, + { + "grad_norm": 0.12534722685813904, + "learning_rate": 9.58883879311582e-05, + "loss": 0.0165, + "step": 17350 + }, + { + "grad_norm": 0.24806447327136993, + "learning_rate": 9.588181920345705e-05, + "loss": 0.0212, + "step": 17360 + }, + { + "grad_norm": 0.19533716142177582, + "learning_rate": 9.587524545818913e-05, + "loss": 0.0176, + "step": 17370 + }, + { + "grad_norm": 0.23233434557914734, + "learning_rate": 9.586866669607335e-05, + "loss": 0.019, + "step": 17380 + }, + { + "grad_norm": 0.11792702972888947, + "learning_rate": 9.586208291782915e-05, + "loss": 0.0159, + "step": 17390 + }, + { + "grad_norm": 0.19819194078445435, + "learning_rate": 9.58554941241765e-05, + "loss": 0.0216, + "step": 17400 + }, + { + "grad_norm": 0.16610698401927948, + "learning_rate": 9.584890031583596e-05, + "loss": 0.0166, + "step": 17410 + }, + { + "grad_norm": 0.1923968642950058, + "learning_rate": 9.584230149352861e-05, + "loss": 0.019, + "step": 17420 + }, + { + "grad_norm": 0.14501355588436127, + "learning_rate": 9.58356976579761e-05, + "loss": 0.0178, + "step": 17430 + }, + { + "grad_norm": 0.19165129959583282, + "learning_rate": 9.58290888099006e-05, + "loss": 0.023, + "step": 17440 + }, + { + "grad_norm": 0.1499188244342804, + "learning_rate": 9.582247495002486e-05, + "loss": 0.0205, + "step": 17450 + }, + { + "grad_norm": 0.22566922008991241, + "learning_rate": 9.581585607907214e-05, + "loss": 0.0187, + "step": 17460 + }, + { + "grad_norm": 0.12246707081794739, + "learning_rate": 9.580923219776628e-05, + "loss": 0.0162, + "step": 17470 + }, + { + "grad_norm": 0.1772511601448059, + "learning_rate": 9.580260330683167e-05, + "loss": 0.0197, + "step": 17480 + }, + { + "grad_norm": 0.2166670709848404, + "learning_rate": 9.579596940699322e-05, + "loss": 0.0166, + "step": 17490 + }, + { + "grad_norm": 0.2035713642835617, + "learning_rate": 9.578933049897643e-05, + "loss": 0.0166, + "step": 17500 + }, + { + "grad_norm": 0.2200816571712494, + "learning_rate": 9.578268658350728e-05, + "loss": 0.0184, + "step": 17510 + }, + { + "grad_norm": 0.19936850666999817, + "learning_rate": 9.577603766131235e-05, + "loss": 0.0196, + "step": 17520 + }, + { + "grad_norm": 0.16781416535377502, + "learning_rate": 9.576938373311878e-05, + "loss": 0.0187, + "step": 17530 + }, + { + "grad_norm": 0.17356479167938232, + "learning_rate": 9.576272479965421e-05, + "loss": 0.0215, + "step": 17540 + }, + { + "grad_norm": 0.245540589094162, + "learning_rate": 9.575606086164687e-05, + "loss": 0.0192, + "step": 17550 + }, + { + "grad_norm": 0.1843673586845398, + "learning_rate": 9.57493919198255e-05, + "loss": 0.0337, + "step": 17560 + }, + { + "grad_norm": 0.21972040832042694, + "learning_rate": 9.57427179749194e-05, + "loss": 0.0274, + "step": 17570 + }, + { + "grad_norm": 0.1518850475549698, + "learning_rate": 9.573603902765846e-05, + "loss": 0.0193, + "step": 17580 + }, + { + "grad_norm": 0.14615441858768463, + "learning_rate": 9.572935507877304e-05, + "loss": 0.0237, + "step": 17590 + }, + { + "grad_norm": 0.19311341643333435, + "learning_rate": 9.57226661289941e-05, + "loss": 0.0271, + "step": 17600 + }, + { + "grad_norm": 0.18611031770706177, + "learning_rate": 9.571597217905315e-05, + "loss": 0.0245, + "step": 17610 + }, + { + "grad_norm": 0.17586012184619904, + "learning_rate": 9.57092732296822e-05, + "loss": 0.0188, + "step": 17620 + }, + { + "grad_norm": 0.12343309819698334, + "learning_rate": 9.570256928161385e-05, + "loss": 0.0164, + "step": 17630 + }, + { + "grad_norm": 0.1187998428940773, + "learning_rate": 9.569586033558126e-05, + "loss": 0.0201, + "step": 17640 + }, + { + "grad_norm": 0.14904014766216278, + "learning_rate": 9.568914639231807e-05, + "loss": 0.0176, + "step": 17650 + }, + { + "grad_norm": 0.44079622626304626, + "learning_rate": 9.568242745255852e-05, + "loss": 0.0306, + "step": 17660 + }, + { + "grad_norm": 0.1811484694480896, + "learning_rate": 9.567570351703739e-05, + "loss": 0.0237, + "step": 17670 + }, + { + "grad_norm": 0.19072183966636658, + "learning_rate": 9.566897458649001e-05, + "loss": 0.0228, + "step": 17680 + }, + { + "grad_norm": 0.13817256689071655, + "learning_rate": 9.566224066165221e-05, + "loss": 0.0225, + "step": 17690 + }, + { + "grad_norm": 0.1360534429550171, + "learning_rate": 9.565550174326043e-05, + "loss": 0.0275, + "step": 17700 + }, + { + "grad_norm": 0.2549751102924347, + "learning_rate": 9.564875783205162e-05, + "loss": 0.0233, + "step": 17710 + }, + { + "grad_norm": 0.1603786200284958, + "learning_rate": 9.564200892876328e-05, + "loss": 0.0208, + "step": 17720 + }, + { + "grad_norm": 0.12452217936515808, + "learning_rate": 9.563525503413348e-05, + "loss": 0.0184, + "step": 17730 + }, + { + "grad_norm": 0.2197364717721939, + "learning_rate": 9.562849614890079e-05, + "loss": 0.0268, + "step": 17740 + }, + { + "grad_norm": 0.2540474832057953, + "learning_rate": 9.562173227380436e-05, + "loss": 0.0316, + "step": 17750 + }, + { + "grad_norm": 0.22960379719734192, + "learning_rate": 9.561496340958389e-05, + "loss": 0.023, + "step": 17760 + }, + { + "grad_norm": 0.17902421951293945, + "learning_rate": 9.560818955697959e-05, + "loss": 0.0175, + "step": 17770 + }, + { + "grad_norm": 0.19437146186828613, + "learning_rate": 9.560141071673228e-05, + "loss": 0.0251, + "step": 17780 + }, + { + "grad_norm": 0.1188102588057518, + "learning_rate": 9.559462688958323e-05, + "loss": 0.0148, + "step": 17790 + }, + { + "grad_norm": 0.14368949830532074, + "learning_rate": 9.558783807627434e-05, + "loss": 0.016, + "step": 17800 + }, + { + "grad_norm": 0.27022475004196167, + "learning_rate": 9.558104427754801e-05, + "loss": 0.0229, + "step": 17810 + }, + { + "grad_norm": 0.239619642496109, + "learning_rate": 9.557424549414722e-05, + "loss": 0.0206, + "step": 17820 + }, + { + "grad_norm": 0.22713308036327362, + "learning_rate": 9.556744172681546e-05, + "loss": 0.0192, + "step": 17830 + }, + { + "grad_norm": 0.15671558678150177, + "learning_rate": 9.556063297629677e-05, + "loss": 0.0181, + "step": 17840 + }, + { + "grad_norm": 0.18283586204051971, + "learning_rate": 9.555381924333578e-05, + "loss": 0.0171, + "step": 17850 + }, + { + "grad_norm": 0.133639857172966, + "learning_rate": 9.554700052867758e-05, + "loss": 0.0204, + "step": 17860 + }, + { + "grad_norm": 0.1618448793888092, + "learning_rate": 9.554017683306789e-05, + "loss": 0.0184, + "step": 17870 + }, + { + "grad_norm": 0.1818198561668396, + "learning_rate": 9.553334815725294e-05, + "loss": 0.0172, + "step": 17880 + }, + { + "grad_norm": 0.21823522448539734, + "learning_rate": 9.552651450197949e-05, + "loss": 0.0194, + "step": 17890 + }, + { + "grad_norm": 0.21261535584926605, + "learning_rate": 9.551967586799486e-05, + "loss": 0.0176, + "step": 17900 + }, + { + "grad_norm": 0.22059044241905212, + "learning_rate": 9.551283225604692e-05, + "loss": 0.0192, + "step": 17910 + }, + { + "grad_norm": 0.14343805611133575, + "learning_rate": 9.550598366688406e-05, + "loss": 0.0165, + "step": 17920 + }, + { + "grad_norm": 0.17971761524677277, + "learning_rate": 9.549913010125526e-05, + "loss": 0.0168, + "step": 17930 + }, + { + "grad_norm": 0.17414745688438416, + "learning_rate": 9.549227155990999e-05, + "loss": 0.0165, + "step": 17940 + }, + { + "grad_norm": 0.14939148724079132, + "learning_rate": 9.548540804359828e-05, + "loss": 0.0236, + "step": 17950 + }, + { + "grad_norm": 0.19795101881027222, + "learning_rate": 9.547853955307077e-05, + "loss": 0.0213, + "step": 17960 + }, + { + "grad_norm": 0.1562632918357849, + "learning_rate": 9.547166608907853e-05, + "loss": 0.0147, + "step": 17970 + }, + { + "grad_norm": 0.1966799795627594, + "learning_rate": 9.546478765237326e-05, + "loss": 0.0264, + "step": 17980 + }, + { + "grad_norm": 0.19698645174503326, + "learning_rate": 9.545790424370715e-05, + "loss": 0.0168, + "step": 17990 + }, + { + "grad_norm": 0.15871930122375488, + "learning_rate": 9.5451015863833e-05, + "loss": 0.0245, + "step": 18000 + }, + { + "grad_norm": 0.22079159319400787, + "learning_rate": 9.544412251350408e-05, + "loss": 0.027, + "step": 18010 + }, + { + "grad_norm": 0.13759233057498932, + "learning_rate": 9.543722419347422e-05, + "loss": 0.0218, + "step": 18020 + }, + { + "grad_norm": 0.2363809049129486, + "learning_rate": 9.543032090449788e-05, + "loss": 0.0257, + "step": 18030 + }, + { + "grad_norm": 0.22802729904651642, + "learning_rate": 9.542341264732992e-05, + "loss": 0.0255, + "step": 18040 + }, + { + "grad_norm": 0.24015113711357117, + "learning_rate": 9.541649942272585e-05, + "loss": 0.0247, + "step": 18050 + }, + { + "grad_norm": 0.16380251944065094, + "learning_rate": 9.54095812314417e-05, + "loss": 0.0199, + "step": 18060 + }, + { + "grad_norm": 0.20321090519428253, + "learning_rate": 9.540265807423401e-05, + "loss": 0.0201, + "step": 18070 + }, + { + "grad_norm": 0.15798524022102356, + "learning_rate": 9.53957299518599e-05, + "loss": 0.0181, + "step": 18080 + }, + { + "grad_norm": 0.17707329988479614, + "learning_rate": 9.5388796865077e-05, + "loss": 0.0223, + "step": 18090 + }, + { + "grad_norm": 0.17842534184455872, + "learning_rate": 9.538185881464353e-05, + "loss": 0.0228, + "step": 18100 + }, + { + "grad_norm": 0.1735224723815918, + "learning_rate": 9.537491580131821e-05, + "loss": 0.0184, + "step": 18110 + }, + { + "grad_norm": 0.1456340104341507, + "learning_rate": 9.53679678258603e-05, + "loss": 0.0188, + "step": 18120 + }, + { + "grad_norm": 0.23411701619625092, + "learning_rate": 9.536101488902966e-05, + "loss": 0.0195, + "step": 18130 + }, + { + "grad_norm": 0.16469906270503998, + "learning_rate": 9.535405699158663e-05, + "loss": 0.0166, + "step": 18140 + }, + { + "grad_norm": 0.17826716601848602, + "learning_rate": 9.53470941342921e-05, + "loss": 0.0163, + "step": 18150 + }, + { + "grad_norm": 0.12871642410755157, + "learning_rate": 9.534012631790756e-05, + "loss": 0.019, + "step": 18160 + }, + { + "grad_norm": 0.19330409169197083, + "learning_rate": 9.533315354319494e-05, + "loss": 0.0186, + "step": 18170 + }, + { + "grad_norm": 0.14240480959415436, + "learning_rate": 9.532617581091682e-05, + "loss": 0.0156, + "step": 18180 + }, + { + "grad_norm": 0.1367618292570114, + "learning_rate": 9.531919312183629e-05, + "loss": 0.0198, + "step": 18190 + }, + { + "grad_norm": 0.14495417475700378, + "learning_rate": 9.531220547671688e-05, + "loss": 0.0154, + "step": 18200 + }, + { + "grad_norm": 0.153059184551239, + "learning_rate": 9.530521287632285e-05, + "loss": 0.0193, + "step": 18210 + }, + { + "grad_norm": 0.13408780097961426, + "learning_rate": 9.529821532141884e-05, + "loss": 0.0147, + "step": 18220 + }, + { + "grad_norm": 0.1397167295217514, + "learning_rate": 9.52912128127701e-05, + "loss": 0.017, + "step": 18230 + }, + { + "grad_norm": 0.21022088825702667, + "learning_rate": 9.528420535114244e-05, + "loss": 0.0192, + "step": 18240 + }, + { + "grad_norm": 0.20326019823551178, + "learning_rate": 9.527719293730215e-05, + "loss": 0.0242, + "step": 18250 + }, + { + "grad_norm": 0.15119020640850067, + "learning_rate": 9.527017557201611e-05, + "loss": 0.019, + "step": 18260 + }, + { + "grad_norm": 0.16186340153217316, + "learning_rate": 9.526315325605176e-05, + "loss": 0.0233, + "step": 18270 + }, + { + "grad_norm": 0.1725398451089859, + "learning_rate": 9.525612599017699e-05, + "loss": 0.0166, + "step": 18280 + }, + { + "grad_norm": 0.23013170063495636, + "learning_rate": 9.524909377516033e-05, + "loss": 0.0197, + "step": 18290 + }, + { + "grad_norm": 0.1861681342124939, + "learning_rate": 9.524205661177081e-05, + "loss": 0.0175, + "step": 18300 + }, + { + "grad_norm": 0.16925564408302307, + "learning_rate": 9.523501450077801e-05, + "loss": 0.02, + "step": 18310 + }, + { + "grad_norm": 0.21048231422901154, + "learning_rate": 9.522796744295202e-05, + "loss": 0.0221, + "step": 18320 + }, + { + "grad_norm": 0.17354194819927216, + "learning_rate": 9.522091543906352e-05, + "loss": 0.0202, + "step": 18330 + }, + { + "grad_norm": 0.14386051893234253, + "learning_rate": 9.521385848988369e-05, + "loss": 0.0188, + "step": 18340 + }, + { + "grad_norm": 0.15568502247333527, + "learning_rate": 9.520679659618428e-05, + "loss": 0.0159, + "step": 18350 + }, + { + "grad_norm": 0.16611768305301666, + "learning_rate": 9.519972975873754e-05, + "loss": 0.0193, + "step": 18360 + }, + { + "grad_norm": 0.17259441316127777, + "learning_rate": 9.519265797831633e-05, + "loss": 0.0173, + "step": 18370 + }, + { + "grad_norm": 0.1512155681848526, + "learning_rate": 9.518558125569399e-05, + "loss": 0.0192, + "step": 18380 + }, + { + "grad_norm": 0.1475355327129364, + "learning_rate": 9.517849959164442e-05, + "loss": 0.0166, + "step": 18390 + }, + { + "grad_norm": 0.23186150193214417, + "learning_rate": 9.517141298694205e-05, + "loss": 0.0236, + "step": 18400 + }, + { + "grad_norm": 0.17594440281391144, + "learning_rate": 9.516432144236188e-05, + "loss": 0.0213, + "step": 18410 + }, + { + "grad_norm": 0.15148113667964935, + "learning_rate": 9.515722495867941e-05, + "loss": 0.0217, + "step": 18420 + }, + { + "grad_norm": 0.1281021535396576, + "learning_rate": 9.515012353667072e-05, + "loss": 0.0175, + "step": 18430 + }, + { + "grad_norm": 0.1846960037946701, + "learning_rate": 9.51430171771124e-05, + "loss": 0.0154, + "step": 18440 + }, + { + "grad_norm": 0.13139033317565918, + "learning_rate": 9.513590588078159e-05, + "loss": 0.0231, + "step": 18450 + }, + { + "grad_norm": 0.19512365758419037, + "learning_rate": 9.512878964845597e-05, + "loss": 0.0191, + "step": 18460 + }, + { + "grad_norm": 0.28988271951675415, + "learning_rate": 9.512166848091377e-05, + "loss": 0.0232, + "step": 18470 + }, + { + "grad_norm": 0.12837249040603638, + "learning_rate": 9.511454237893376e-05, + "loss": 0.0177, + "step": 18480 + }, + { + "grad_norm": 0.13322801887989044, + "learning_rate": 9.51074113432952e-05, + "loss": 0.0158, + "step": 18490 + }, + { + "grad_norm": 0.15466050803661346, + "learning_rate": 9.510027537477797e-05, + "loss": 0.0239, + "step": 18500 + }, + { + "grad_norm": 0.17534281313419342, + "learning_rate": 9.509313447416242e-05, + "loss": 0.02, + "step": 18510 + }, + { + "grad_norm": 0.20654787123203278, + "learning_rate": 9.508598864222949e-05, + "loss": 0.0173, + "step": 18520 + }, + { + "grad_norm": 0.14114446938037872, + "learning_rate": 9.507883787976062e-05, + "loss": 0.021, + "step": 18530 + }, + { + "grad_norm": 0.1328234076499939, + "learning_rate": 9.507168218753781e-05, + "loss": 0.0146, + "step": 18540 + }, + { + "grad_norm": 0.15576857328414917, + "learning_rate": 9.506452156634362e-05, + "loss": 0.0175, + "step": 18550 + }, + { + "grad_norm": 0.2059352844953537, + "learning_rate": 9.505735601696109e-05, + "loss": 0.0262, + "step": 18560 + }, + { + "grad_norm": 0.18526364862918854, + "learning_rate": 9.505018554017385e-05, + "loss": 0.0199, + "step": 18570 + }, + { + "grad_norm": 0.21587447822093964, + "learning_rate": 9.504301013676604e-05, + "loss": 0.0184, + "step": 18580 + }, + { + "grad_norm": 0.18770910799503326, + "learning_rate": 9.503582980752238e-05, + "loss": 0.0171, + "step": 18590 + }, + { + "grad_norm": 0.5846869945526123, + "learning_rate": 9.502864455322809e-05, + "loss": 0.0331, + "step": 18600 + }, + { + "grad_norm": 0.14761292934417725, + "learning_rate": 9.502145437466891e-05, + "loss": 0.018, + "step": 18610 + }, + { + "grad_norm": 0.17377683520317078, + "learning_rate": 9.501425927263116e-05, + "loss": 0.033, + "step": 18620 + }, + { + "grad_norm": 0.3424476087093353, + "learning_rate": 9.500705924790172e-05, + "loss": 0.0461, + "step": 18630 + }, + { + "grad_norm": 0.22030770778656006, + "learning_rate": 9.499985430126794e-05, + "loss": 0.0234, + "step": 18640 + }, + { + "grad_norm": 0.17035162448883057, + "learning_rate": 9.499264443351775e-05, + "loss": 0.021, + "step": 18650 + }, + { + "grad_norm": 0.2081509232521057, + "learning_rate": 9.498542964543961e-05, + "loss": 0.0216, + "step": 18660 + }, + { + "grad_norm": 0.23317208886146545, + "learning_rate": 9.497820993782252e-05, + "loss": 0.0211, + "step": 18670 + }, + { + "grad_norm": 0.141694575548172, + "learning_rate": 9.497098531145601e-05, + "loss": 0.0209, + "step": 18680 + }, + { + "grad_norm": 0.14944763481616974, + "learning_rate": 9.496375576713017e-05, + "loss": 0.0185, + "step": 18690 + }, + { + "grad_norm": 0.13773611187934875, + "learning_rate": 9.49565213056356e-05, + "loss": 0.0158, + "step": 18700 + }, + { + "grad_norm": 0.12076227366924286, + "learning_rate": 9.494928192776342e-05, + "loss": 0.015, + "step": 18710 + }, + { + "grad_norm": 0.13550543785095215, + "learning_rate": 9.494203763430538e-05, + "loss": 0.0155, + "step": 18720 + }, + { + "grad_norm": 0.18993476033210754, + "learning_rate": 9.493478842605366e-05, + "loss": 0.0358, + "step": 18730 + }, + { + "grad_norm": 0.18273325264453888, + "learning_rate": 9.492753430380105e-05, + "loss": 0.016, + "step": 18740 + }, + { + "grad_norm": 0.1703435182571411, + "learning_rate": 9.492027526834083e-05, + "loss": 0.0167, + "step": 18750 + }, + { + "grad_norm": 0.19595186412334442, + "learning_rate": 9.491301132046684e-05, + "loss": 0.0183, + "step": 18760 + }, + { + "grad_norm": 0.131064772605896, + "learning_rate": 9.490574246097345e-05, + "loss": 0.0378, + "step": 18770 + }, + { + "grad_norm": 0.3044780492782593, + "learning_rate": 9.48984686906556e-05, + "loss": 0.0325, + "step": 18780 + }, + { + "grad_norm": 0.24853640794754028, + "learning_rate": 9.489119001030871e-05, + "loss": 0.0222, + "step": 18790 + }, + { + "grad_norm": 0.19803500175476074, + "learning_rate": 9.488390642072878e-05, + "loss": 0.0195, + "step": 18800 + }, + { + "grad_norm": 0.1484699696302414, + "learning_rate": 9.48766179227123e-05, + "loss": 0.0204, + "step": 18810 + }, + { + "grad_norm": 0.1853627860546112, + "learning_rate": 9.486932451705636e-05, + "loss": 0.0214, + "step": 18820 + }, + { + "grad_norm": 0.16896642744541168, + "learning_rate": 9.486202620455857e-05, + "loss": 0.0179, + "step": 18830 + }, + { + "grad_norm": 0.16422446072101593, + "learning_rate": 9.485472298601704e-05, + "loss": 0.0183, + "step": 18840 + }, + { + "grad_norm": 0.24244602024555206, + "learning_rate": 9.484741486223043e-05, + "loss": 0.0197, + "step": 18850 + }, + { + "grad_norm": 0.20923802256584167, + "learning_rate": 9.484010183399797e-05, + "loss": 0.0229, + "step": 18860 + }, + { + "grad_norm": 0.17514586448669434, + "learning_rate": 9.483278390211938e-05, + "loss": 0.0201, + "step": 18870 + }, + { + "grad_norm": 0.11981351673603058, + "learning_rate": 9.482546106739496e-05, + "loss": 0.0184, + "step": 18880 + }, + { + "grad_norm": 0.18240664899349213, + "learning_rate": 9.48181333306255e-05, + "loss": 0.017, + "step": 18890 + }, + { + "grad_norm": 0.12653538584709167, + "learning_rate": 9.481080069261237e-05, + "loss": 0.0268, + "step": 18900 + }, + { + "grad_norm": 0.16492033004760742, + "learning_rate": 9.480346315415745e-05, + "loss": 0.0164, + "step": 18910 + }, + { + "grad_norm": 0.22140853106975555, + "learning_rate": 9.479612071606314e-05, + "loss": 0.0174, + "step": 18920 + }, + { + "grad_norm": 0.1757112592458725, + "learning_rate": 9.478877337913244e-05, + "loss": 0.02, + "step": 18930 + }, + { + "grad_norm": 0.15548118948936462, + "learning_rate": 9.478142114416881e-05, + "loss": 0.0172, + "step": 18940 + }, + { + "grad_norm": 0.12196731567382812, + "learning_rate": 9.47740640119763e-05, + "loss": 0.0145, + "step": 18950 + }, + { + "grad_norm": 0.16696184873580933, + "learning_rate": 9.476670198335947e-05, + "loss": 0.0195, + "step": 18960 + }, + { + "grad_norm": 0.16144691407680511, + "learning_rate": 9.47593350591234e-05, + "loss": 0.0192, + "step": 18970 + }, + { + "grad_norm": 0.14798134565353394, + "learning_rate": 9.475196324007376e-05, + "loss": 0.0192, + "step": 18980 + }, + { + "grad_norm": 0.17827743291854858, + "learning_rate": 9.474458652701669e-05, + "loss": 0.0153, + "step": 18990 + }, + { + "grad_norm": 0.2161901444196701, + "learning_rate": 9.473720492075892e-05, + "loss": 0.0213, + "step": 19000 + }, + { + "grad_norm": 0.18233783543109894, + "learning_rate": 9.472981842210768e-05, + "loss": 0.0209, + "step": 19010 + }, + { + "grad_norm": 0.16601136326789856, + "learning_rate": 9.472242703187074e-05, + "loss": 0.0192, + "step": 19020 + }, + { + "grad_norm": 0.1236705556511879, + "learning_rate": 9.471503075085643e-05, + "loss": 0.0145, + "step": 19030 + }, + { + "grad_norm": 0.1993335336446762, + "learning_rate": 9.470762957987359e-05, + "loss": 0.021, + "step": 19040 + }, + { + "grad_norm": 0.17819608747959137, + "learning_rate": 9.470022351973158e-05, + "loss": 0.0212, + "step": 19050 + }, + { + "grad_norm": 0.20374342799186707, + "learning_rate": 9.469281257124034e-05, + "loss": 0.0173, + "step": 19060 + }, + { + "grad_norm": 0.19512693583965302, + "learning_rate": 9.46853967352103e-05, + "loss": 0.0206, + "step": 19070 + }, + { + "grad_norm": 0.15427890419960022, + "learning_rate": 9.467797601245246e-05, + "loss": 0.0161, + "step": 19080 + }, + { + "grad_norm": 0.17384685575962067, + "learning_rate": 9.467055040377834e-05, + "loss": 0.0237, + "step": 19090 + }, + { + "grad_norm": 0.16809290647506714, + "learning_rate": 9.466311990999999e-05, + "loss": 0.0238, + "step": 19100 + }, + { + "grad_norm": 0.16821721196174622, + "learning_rate": 9.465568453193e-05, + "loss": 0.0196, + "step": 19110 + }, + { + "grad_norm": 0.127919003367424, + "learning_rate": 9.464824427038148e-05, + "loss": 0.0173, + "step": 19120 + }, + { + "grad_norm": 0.11654169112443924, + "learning_rate": 9.46407991261681e-05, + "loss": 0.0273, + "step": 19130 + }, + { + "grad_norm": 0.10396882146596909, + "learning_rate": 9.463334910010404e-05, + "loss": 0.016, + "step": 19140 + }, + { + "grad_norm": 0.13910573720932007, + "learning_rate": 9.462589419300403e-05, + "loss": 0.0125, + "step": 19150 + }, + { + "grad_norm": 0.15320537984371185, + "learning_rate": 9.461843440568333e-05, + "loss": 0.0191, + "step": 19160 + }, + { + "grad_norm": 0.11951324343681335, + "learning_rate": 9.461096973895773e-05, + "loss": 0.0152, + "step": 19170 + }, + { + "grad_norm": 0.1604233831167221, + "learning_rate": 9.460350019364355e-05, + "loss": 0.0193, + "step": 19180 + }, + { + "grad_norm": 0.20535050332546234, + "learning_rate": 9.459602577055764e-05, + "loss": 0.0342, + "step": 19190 + }, + { + "grad_norm": 0.1729145348072052, + "learning_rate": 9.45885464705174e-05, + "loss": 0.0181, + "step": 19200 + }, + { + "grad_norm": 0.15082843601703644, + "learning_rate": 9.458106229434076e-05, + "loss": 0.0166, + "step": 19210 + }, + { + "grad_norm": 0.20488899946212769, + "learning_rate": 9.457357324284617e-05, + "loss": 0.0151, + "step": 19220 + }, + { + "grad_norm": 0.15571105480194092, + "learning_rate": 9.456607931685262e-05, + "loss": 0.0199, + "step": 19230 + }, + { + "grad_norm": 0.21508295834064484, + "learning_rate": 9.455858051717965e-05, + "loss": 0.0176, + "step": 19240 + }, + { + "grad_norm": 0.15895192325115204, + "learning_rate": 9.45510768446473e-05, + "loss": 0.0169, + "step": 19250 + }, + { + "grad_norm": 0.21724411845207214, + "learning_rate": 9.454356830007618e-05, + "loss": 0.0171, + "step": 19260 + }, + { + "grad_norm": 0.1266838163137436, + "learning_rate": 9.45360548842874e-05, + "loss": 0.0149, + "step": 19270 + }, + { + "grad_norm": 0.1782476305961609, + "learning_rate": 9.452853659810261e-05, + "loss": 0.019, + "step": 19280 + }, + { + "grad_norm": 0.20052076876163483, + "learning_rate": 9.452101344234401e-05, + "loss": 0.0227, + "step": 19290 + }, + { + "grad_norm": 0.13075384497642517, + "learning_rate": 9.451348541783431e-05, + "loss": 0.0188, + "step": 19300 + }, + { + "grad_norm": 0.13211585581302643, + "learning_rate": 9.450595252539678e-05, + "loss": 0.0164, + "step": 19310 + }, + { + "grad_norm": 0.13835737109184265, + "learning_rate": 9.449841476585518e-05, + "loss": 0.0243, + "step": 19320 + }, + { + "grad_norm": 0.1919039934873581, + "learning_rate": 9.449087214003384e-05, + "loss": 0.0331, + "step": 19330 + }, + { + "grad_norm": 0.2083517163991928, + "learning_rate": 9.448332464875765e-05, + "loss": 0.0337, + "step": 19340 + }, + { + "grad_norm": 0.14755715429782867, + "learning_rate": 9.447577229285192e-05, + "loss": 0.0236, + "step": 19350 + }, + { + "grad_norm": 0.1130974218249321, + "learning_rate": 9.446821507314261e-05, + "loss": 0.0148, + "step": 19360 + }, + { + "grad_norm": 0.1291045844554901, + "learning_rate": 9.446065299045617e-05, + "loss": 0.0182, + "step": 19370 + }, + { + "grad_norm": 0.18204990029335022, + "learning_rate": 9.445308604561955e-05, + "loss": 0.029, + "step": 19380 + }, + { + "grad_norm": 0.16195808351039886, + "learning_rate": 9.444551423946028e-05, + "loss": 0.021, + "step": 19390 + }, + { + "grad_norm": 0.11503028869628906, + "learning_rate": 9.443793757280638e-05, + "loss": 0.0175, + "step": 19400 + }, + { + "grad_norm": 0.13353650271892548, + "learning_rate": 9.443035604648646e-05, + "loss": 0.0158, + "step": 19410 + }, + { + "grad_norm": 0.13877806067466736, + "learning_rate": 9.44227696613296e-05, + "loss": 0.0164, + "step": 19420 + }, + { + "grad_norm": 0.27247360348701477, + "learning_rate": 9.441517841816542e-05, + "loss": 0.0297, + "step": 19430 + }, + { + "grad_norm": 0.15438807010650635, + "learning_rate": 9.440758231782413e-05, + "loss": 0.0179, + "step": 19440 + }, + { + "grad_norm": 0.15321041643619537, + "learning_rate": 9.439998136113639e-05, + "loss": 0.0166, + "step": 19450 + }, + { + "grad_norm": 0.18666008114814758, + "learning_rate": 9.439237554893344e-05, + "loss": 0.0172, + "step": 19460 + }, + { + "grad_norm": 0.13289569318294525, + "learning_rate": 9.438476488204705e-05, + "loss": 0.0224, + "step": 19470 + }, + { + "grad_norm": 0.13856227695941925, + "learning_rate": 9.43771493613095e-05, + "loss": 0.0171, + "step": 19480 + }, + { + "grad_norm": 0.19389276206493378, + "learning_rate": 9.436952898755362e-05, + "loss": 0.0279, + "step": 19490 + }, + { + "grad_norm": 0.15192767977714539, + "learning_rate": 9.436190376161276e-05, + "loss": 0.0178, + "step": 19500 + }, + { + "grad_norm": 0.1461472362279892, + "learning_rate": 9.43542736843208e-05, + "loss": 0.0287, + "step": 19510 + }, + { + "grad_norm": 0.17848698794841766, + "learning_rate": 9.434663875651216e-05, + "loss": 0.017, + "step": 19520 + }, + { + "grad_norm": 0.20570912957191467, + "learning_rate": 9.433899897902177e-05, + "loss": 0.0171, + "step": 19530 + }, + { + "grad_norm": 0.14891409873962402, + "learning_rate": 9.433135435268511e-05, + "loss": 0.0168, + "step": 19540 + }, + { + "grad_norm": 0.11298293620347977, + "learning_rate": 9.432370487833819e-05, + "loss": 0.0187, + "step": 19550 + }, + { + "grad_norm": 0.2647833526134491, + "learning_rate": 9.431605055681756e-05, + "loss": 0.0237, + "step": 19560 + }, + { + "grad_norm": 0.18187938630580902, + "learning_rate": 9.430839138896026e-05, + "loss": 0.0194, + "step": 19570 + }, + { + "grad_norm": 0.13703405857086182, + "learning_rate": 9.43007273756039e-05, + "loss": 0.0232, + "step": 19580 + }, + { + "grad_norm": 0.15095089375972748, + "learning_rate": 9.429305851758658e-05, + "loss": 0.0196, + "step": 19590 + }, + { + "grad_norm": 0.2457534223794937, + "learning_rate": 9.428538481574699e-05, + "loss": 0.0214, + "step": 19600 + }, + { + "grad_norm": 0.15796126425266266, + "learning_rate": 9.42777062709243e-05, + "loss": 0.0188, + "step": 19610 + }, + { + "grad_norm": 0.2010515034198761, + "learning_rate": 9.427002288395821e-05, + "loss": 0.0197, + "step": 19620 + }, + { + "grad_norm": 0.1911383867263794, + "learning_rate": 9.426233465568898e-05, + "loss": 0.0212, + "step": 19630 + }, + { + "grad_norm": 0.1548798829317093, + "learning_rate": 9.42546415869574e-05, + "loss": 0.0173, + "step": 19640 + }, + { + "grad_norm": 0.1870667040348053, + "learning_rate": 9.424694367860473e-05, + "loss": 0.0196, + "step": 19650 + }, + { + "grad_norm": 0.17390161752700806, + "learning_rate": 9.423924093147284e-05, + "loss": 0.0174, + "step": 19660 + }, + { + "grad_norm": 0.14238622784614563, + "learning_rate": 9.423153334640407e-05, + "loss": 0.0158, + "step": 19670 + }, + { + "grad_norm": 0.1375490427017212, + "learning_rate": 9.42238209242413e-05, + "loss": 0.0156, + "step": 19680 + }, + { + "grad_norm": 0.17629438638687134, + "learning_rate": 9.421610366582798e-05, + "loss": 0.0206, + "step": 19690 + }, + { + "grad_norm": 0.14485591650009155, + "learning_rate": 9.420838157200803e-05, + "loss": 0.017, + "step": 19700 + }, + { + "grad_norm": 0.14034612476825714, + "learning_rate": 9.420065464362594e-05, + "loss": 0.0205, + "step": 19710 + }, + { + "grad_norm": 0.18918192386627197, + "learning_rate": 9.419292288152673e-05, + "loss": 0.0201, + "step": 19720 + }, + { + "grad_norm": 0.14771851897239685, + "learning_rate": 9.418518628655588e-05, + "loss": 0.0171, + "step": 19730 + }, + { + "grad_norm": 0.12691043317317963, + "learning_rate": 9.417744485955951e-05, + "loss": 0.0159, + "step": 19740 + }, + { + "grad_norm": 0.12100371718406677, + "learning_rate": 9.41696986013842e-05, + "loss": 0.015, + "step": 19750 + }, + { + "grad_norm": 0.17419448494911194, + "learning_rate": 9.416194751287705e-05, + "loss": 0.0216, + "step": 19760 + }, + { + "grad_norm": 0.15822921693325043, + "learning_rate": 9.415419159488572e-05, + "loss": 0.0158, + "step": 19770 + }, + { + "grad_norm": 0.14083370566368103, + "learning_rate": 9.414643084825837e-05, + "loss": 0.0211, + "step": 19780 + }, + { + "grad_norm": 0.1571672409772873, + "learning_rate": 9.413866527384372e-05, + "loss": 0.0224, + "step": 19790 + }, + { + "grad_norm": 0.15686865150928497, + "learning_rate": 9.4130894872491e-05, + "loss": 0.0165, + "step": 19800 + }, + { + "grad_norm": 0.1321669965982437, + "learning_rate": 9.412311964504998e-05, + "loss": 0.0181, + "step": 19810 + }, + { + "grad_norm": 0.17046107351779938, + "learning_rate": 9.411533959237091e-05, + "loss": 0.0192, + "step": 19820 + }, + { + "grad_norm": 0.17673924565315247, + "learning_rate": 9.410755471530464e-05, + "loss": 0.0135, + "step": 19830 + }, + { + "grad_norm": 0.12554793059825897, + "learning_rate": 9.40997650147025e-05, + "loss": 0.015, + "step": 19840 + }, + { + "grad_norm": 0.2735735774040222, + "learning_rate": 9.409197049141637e-05, + "loss": 0.0257, + "step": 19850 + }, + { + "grad_norm": 0.18254448473453522, + "learning_rate": 9.408417114629863e-05, + "loss": 0.0302, + "step": 19860 + }, + { + "grad_norm": 0.12666963040828705, + "learning_rate": 9.40763669802022e-05, + "loss": 0.0164, + "step": 19870 + }, + { + "grad_norm": 0.16897419095039368, + "learning_rate": 9.406855799398056e-05, + "loss": 0.0273, + "step": 19880 + }, + { + "grad_norm": 0.1645938605070114, + "learning_rate": 9.406074418848767e-05, + "loss": 0.0219, + "step": 19890 + }, + { + "grad_norm": 0.2880553603172302, + "learning_rate": 9.405292556457805e-05, + "loss": 0.0335, + "step": 19900 + }, + { + "grad_norm": 0.2345859855413437, + "learning_rate": 9.404510212310671e-05, + "loss": 0.0204, + "step": 19910 + }, + { + "grad_norm": 0.1619919091463089, + "learning_rate": 9.403727386492924e-05, + "loss": 0.0178, + "step": 19920 + }, + { + "grad_norm": 0.18329904973506927, + "learning_rate": 9.40294407909017e-05, + "loss": 0.02, + "step": 19930 + }, + { + "grad_norm": 0.19710899889469147, + "learning_rate": 9.40216029018807e-05, + "loss": 0.0166, + "step": 19940 + }, + { + "grad_norm": 0.20877808332443237, + "learning_rate": 9.401376019872338e-05, + "loss": 0.0165, + "step": 19950 + }, + { + "grad_norm": 0.15872612595558167, + "learning_rate": 9.400591268228746e-05, + "loss": 0.0178, + "step": 19960 + }, + { + "grad_norm": 0.18141992390155792, + "learning_rate": 9.399806035343106e-05, + "loss": 0.0183, + "step": 19970 + }, + { + "grad_norm": 0.21124470233917236, + "learning_rate": 9.399020321301294e-05, + "loss": 0.0192, + "step": 19980 + }, + { + "grad_norm": 0.19280655682086945, + "learning_rate": 9.398234126189234e-05, + "loss": 0.0193, + "step": 19990 + }, + { + "grad_norm": 0.19284053146839142, + "learning_rate": 9.397447450092902e-05, + "loss": 0.0182, + "step": 20000 + } + ], + "logging_steps": 10, + "max_steps": 100000, + "num_input_tokens_seen": 0, + "num_train_epochs": 9223372036854775807, + "save_steps": 20000, + "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 +}