{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 100000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 2.694854736328125, "learning_rate": 1.8e-07, "loss": 1.3541, "step": 10 }, { "grad_norm": 2.596203565597534, "learning_rate": 3.8e-07, "loss": 1.3666, "step": 20 }, { "grad_norm": 2.4758846759796143, "learning_rate": 5.8e-07, "loss": 1.3502, "step": 30 }, { "grad_norm": 2.126450777053833, "learning_rate": 7.8e-07, "loss": 1.3425, "step": 40 }, { "grad_norm": 1.9595330953598022, "learning_rate": 9.8e-07, "loss": 1.317, "step": 50 }, { "grad_norm": 1.2296674251556396, "learning_rate": 1.18e-06, "loss": 1.291, "step": 60 }, { "grad_norm": 1.0623071193695068, "learning_rate": 1.3800000000000001e-06, "loss": 1.2437, "step": 70 }, { "grad_norm": 0.7917822599411011, "learning_rate": 1.5800000000000003e-06, "loss": 1.2121, "step": 80 }, { "grad_norm": 0.7399148941040039, "learning_rate": 1.7800000000000001e-06, "loss": 1.183, "step": 90 }, { "grad_norm": 0.532645583152771, "learning_rate": 1.98e-06, "loss": 1.1685, "step": 100 }, { "grad_norm": 0.41063618659973145, "learning_rate": 2.1800000000000003e-06, "loss": 1.141, "step": 110 }, { "grad_norm": 0.3665008842945099, "learning_rate": 2.38e-06, "loss": 1.138, "step": 120 }, { "grad_norm": 0.4364762008190155, "learning_rate": 2.5800000000000003e-06, "loss": 1.139, "step": 130 }, { "grad_norm": 0.3728153109550476, "learning_rate": 2.78e-06, "loss": 1.1512, "step": 140 }, { "grad_norm": 0.2814396023750305, "learning_rate": 2.9800000000000003e-06, "loss": 1.1468, "step": 150 }, { "grad_norm": 0.3334779739379883, "learning_rate": 3.1800000000000005e-06, "loss": 1.13, "step": 160 }, { "grad_norm": 0.33682262897491455, "learning_rate": 3.38e-06, "loss": 1.1235, "step": 170 }, { "grad_norm": 0.4340789020061493, "learning_rate": 3.58e-06, "loss": 1.1094, "step": 180 }, { "grad_norm": 0.49096783995628357, "learning_rate": 3.7800000000000002e-06, "loss": 1.0899, "step": 190 }, { "grad_norm": 0.4652177691459656, "learning_rate": 3.98e-06, "loss": 1.0794, "step": 200 }, { "grad_norm": 0.42950764298439026, "learning_rate": 4.18e-06, "loss": 1.0767, "step": 210 }, { "grad_norm": 0.46818482875823975, "learning_rate": 4.38e-06, "loss": 1.0644, "step": 220 }, { "grad_norm": 0.4194670617580414, "learning_rate": 4.58e-06, "loss": 1.0622, "step": 230 }, { "grad_norm": 0.43106621503829956, "learning_rate": 4.780000000000001e-06, "loss": 1.0552, "step": 240 }, { "grad_norm": 0.4176578223705292, "learning_rate": 4.98e-06, "loss": 1.0512, "step": 250 }, { "grad_norm": 0.4569450318813324, "learning_rate": 5.18e-06, "loss": 1.0579, "step": 260 }, { "grad_norm": 0.43798592686653137, "learning_rate": 5.38e-06, "loss": 1.0483, "step": 270 }, { "grad_norm": 0.40002256631851196, "learning_rate": 5.580000000000001e-06, "loss": 1.0471, "step": 280 }, { "grad_norm": 0.4324214458465576, "learning_rate": 5.78e-06, "loss": 1.0364, "step": 290 }, { "grad_norm": 0.5187270045280457, "learning_rate": 5.98e-06, "loss": 1.0339, "step": 300 }, { "grad_norm": 0.4738016426563263, "learning_rate": 6.18e-06, "loss": 1.034, "step": 310 }, { "grad_norm": 0.5641390085220337, "learning_rate": 6.38e-06, "loss": 1.0354, "step": 320 }, { "grad_norm": 0.741642415523529, "learning_rate": 6.58e-06, "loss": 1.02, "step": 330 }, { "grad_norm": 0.8953841924667358, "learning_rate": 6.78e-06, "loss": 1.0039, "step": 340 }, { "grad_norm": 0.6173702478408813, "learning_rate": 6.98e-06, "loss": 0.9954, "step": 350 }, { "grad_norm": 1.0601762533187866, "learning_rate": 7.180000000000001e-06, "loss": 0.9691, "step": 360 }, { "grad_norm": 0.8148952126502991, "learning_rate": 7.3800000000000005e-06, "loss": 0.9538, "step": 370 }, { "grad_norm": 0.7504807114601135, "learning_rate": 7.580000000000001e-06, "loss": 0.9418, "step": 380 }, { "grad_norm": 0.9909512996673584, "learning_rate": 7.78e-06, "loss": 0.9305, "step": 390 }, { "grad_norm": 1.1595311164855957, "learning_rate": 7.98e-06, "loss": 0.9108, "step": 400 }, { "grad_norm": 1.134146809577942, "learning_rate": 8.18e-06, "loss": 0.8912, "step": 410 }, { "grad_norm": 0.9221531748771667, "learning_rate": 8.380000000000001e-06, "loss": 0.8698, "step": 420 }, { "grad_norm": 1.581253170967102, "learning_rate": 8.580000000000001e-06, "loss": 0.8585, "step": 430 }, { "grad_norm": 1.462759256362915, "learning_rate": 8.78e-06, "loss": 0.8364, "step": 440 }, { "grad_norm": 1.457959771156311, "learning_rate": 8.98e-06, "loss": 0.8136, "step": 450 }, { "grad_norm": 1.309896469116211, "learning_rate": 9.180000000000002e-06, "loss": 0.7868, "step": 460 }, { "grad_norm": 1.655473232269287, "learning_rate": 9.38e-06, "loss": 0.7715, "step": 470 }, { "grad_norm": 1.5691627264022827, "learning_rate": 9.58e-06, "loss": 0.7446, "step": 480 }, { "grad_norm": 1.6667720079421997, "learning_rate": 9.78e-06, "loss": 0.7094, "step": 490 }, { "grad_norm": 1.6918824911117554, "learning_rate": 9.980000000000001e-06, "loss": 0.6825, "step": 500 }, { "grad_norm": 1.7891321182250977, "learning_rate": 1.018e-05, "loss": 0.6462, "step": 510 }, { "grad_norm": 1.9962478876113892, "learning_rate": 1.038e-05, "loss": 0.6235, "step": 520 }, { "grad_norm": 1.8309444189071655, "learning_rate": 1.058e-05, "loss": 0.5979, "step": 530 }, { "grad_norm": 1.9304026365280151, "learning_rate": 1.0780000000000002e-05, "loss": 0.5733, "step": 540 }, { "grad_norm": 1.8346161842346191, "learning_rate": 1.098e-05, "loss": 0.5408, "step": 550 }, { "grad_norm": 1.7506791353225708, "learning_rate": 1.118e-05, "loss": 0.5271, "step": 560 }, { "grad_norm": 2.037799835205078, "learning_rate": 1.1380000000000001e-05, "loss": 0.5175, "step": 570 }, { "grad_norm": 1.9981427192687988, "learning_rate": 1.1580000000000001e-05, "loss": 0.4968, "step": 580 }, { "grad_norm": 2.0727975368499756, "learning_rate": 1.178e-05, "loss": 0.4632, "step": 590 }, { "grad_norm": 2.056100845336914, "learning_rate": 1.198e-05, "loss": 0.4561, "step": 600 }, { "grad_norm": 1.8760777711868286, "learning_rate": 1.2180000000000002e-05, "loss": 0.424, "step": 610 }, { "grad_norm": 1.9546928405761719, "learning_rate": 1.238e-05, "loss": 0.4169, "step": 620 }, { "grad_norm": 2.0427405834198, "learning_rate": 1.258e-05, "loss": 0.3869, "step": 630 }, { "grad_norm": 2.2207467555999756, "learning_rate": 1.278e-05, "loss": 0.3633, "step": 640 }, { "grad_norm": 2.4976277351379395, "learning_rate": 1.2980000000000001e-05, "loss": 0.3613, "step": 650 }, { "grad_norm": 1.99053156375885, "learning_rate": 1.3180000000000001e-05, "loss": 0.3479, "step": 660 }, { "grad_norm": 1.6395918130874634, "learning_rate": 1.338e-05, "loss": 0.3271, "step": 670 }, { "grad_norm": 2.307037115097046, "learning_rate": 1.358e-05, "loss": 0.3163, "step": 680 }, { "grad_norm": 2.4509775638580322, "learning_rate": 1.3780000000000002e-05, "loss": 0.3046, "step": 690 }, { "grad_norm": 1.748297929763794, "learning_rate": 1.3980000000000002e-05, "loss": 0.2872, "step": 700 }, { "grad_norm": 2.4859049320220947, "learning_rate": 1.4180000000000001e-05, "loss": 0.2986, "step": 710 }, { "grad_norm": 2.4076526165008545, "learning_rate": 1.4380000000000001e-05, "loss": 0.2751, "step": 720 }, { "grad_norm": 2.5215964317321777, "learning_rate": 1.4580000000000003e-05, "loss": 0.2507, "step": 730 }, { "grad_norm": 2.3426661491394043, "learning_rate": 1.4779999999999999e-05, "loss": 0.2513, "step": 740 }, { "grad_norm": 2.1209638118743896, "learning_rate": 1.4979999999999999e-05, "loss": 0.241, "step": 750 }, { "grad_norm": 2.2149646282196045, "learning_rate": 1.518e-05, "loss": 0.2387, "step": 760 }, { "grad_norm": 2.047320604324341, "learning_rate": 1.538e-05, "loss": 0.2232, "step": 770 }, { "grad_norm": 1.9111849069595337, "learning_rate": 1.558e-05, "loss": 0.2176, "step": 780 }, { "grad_norm": 2.281914710998535, "learning_rate": 1.578e-05, "loss": 0.2042, "step": 790 }, { "grad_norm": 2.033903121948242, "learning_rate": 1.598e-05, "loss": 0.1982, "step": 800 }, { "grad_norm": 2.1687207221984863, "learning_rate": 1.618e-05, "loss": 0.1827, "step": 810 }, { "grad_norm": 2.872370958328247, "learning_rate": 1.6380000000000002e-05, "loss": 0.1746, "step": 820 }, { "grad_norm": 1.9542465209960938, "learning_rate": 1.658e-05, "loss": 0.1694, "step": 830 }, { "grad_norm": 1.724130392074585, "learning_rate": 1.6780000000000002e-05, "loss": 0.1604, "step": 840 }, { "grad_norm": 2.3694534301757812, "learning_rate": 1.698e-05, "loss": 0.1598, "step": 850 }, { "grad_norm": 2.3520636558532715, "learning_rate": 1.718e-05, "loss": 0.1671, "step": 860 }, { "grad_norm": 2.3977627754211426, "learning_rate": 1.7380000000000003e-05, "loss": 0.154, "step": 870 }, { "grad_norm": 2.4218649864196777, "learning_rate": 1.758e-05, "loss": 0.1536, "step": 880 }, { "grad_norm": 1.860933780670166, "learning_rate": 1.7780000000000003e-05, "loss": 0.1459, "step": 890 }, { "grad_norm": 2.1696035861968994, "learning_rate": 1.798e-05, "loss": 0.135, "step": 900 }, { "grad_norm": 2.153006076812744, "learning_rate": 1.818e-05, "loss": 0.134, "step": 910 }, { "grad_norm": 1.8260118961334229, "learning_rate": 1.838e-05, "loss": 0.1345, "step": 920 }, { "grad_norm": 1.909118413925171, "learning_rate": 1.858e-05, "loss": 0.1388, "step": 930 }, { "grad_norm": 2.0129892826080322, "learning_rate": 1.878e-05, "loss": 0.128, "step": 940 }, { "grad_norm": 1.913816213607788, "learning_rate": 1.898e-05, "loss": 0.1342, "step": 950 }, { "grad_norm": 1.8119494915008545, "learning_rate": 1.918e-05, "loss": 0.1324, "step": 960 }, { "grad_norm": 2.2811315059661865, "learning_rate": 1.938e-05, "loss": 0.1299, "step": 970 }, { "grad_norm": 1.8016881942749023, "learning_rate": 1.9580000000000002e-05, "loss": 0.1353, "step": 980 }, { "grad_norm": 2.339780569076538, "learning_rate": 1.978e-05, "loss": 0.129, "step": 990 }, { "grad_norm": 1.8016932010650635, "learning_rate": 1.9980000000000002e-05, "loss": 0.1329, "step": 1000 }, { "grad_norm": 1.7011120319366455, "learning_rate": 2.0180000000000003e-05, "loss": 0.125, "step": 1010 }, { "grad_norm": 1.878551721572876, "learning_rate": 2.038e-05, "loss": 0.1308, "step": 1020 }, { "grad_norm": 1.8353983163833618, "learning_rate": 2.0580000000000003e-05, "loss": 0.13, "step": 1030 }, { "grad_norm": 1.7689043283462524, "learning_rate": 2.078e-05, "loss": 0.1213, "step": 1040 }, { "grad_norm": 1.7843544483184814, "learning_rate": 2.098e-05, "loss": 0.1139, "step": 1050 }, { "grad_norm": 1.5686310529708862, "learning_rate": 2.118e-05, "loss": 0.1108, "step": 1060 }, { "grad_norm": 1.9034448862075806, "learning_rate": 2.138e-05, "loss": 0.1159, "step": 1070 }, { "grad_norm": 1.902099847793579, "learning_rate": 2.158e-05, "loss": 0.1084, "step": 1080 }, { "grad_norm": 2.1763083934783936, "learning_rate": 2.178e-05, "loss": 0.1118, "step": 1090 }, { "grad_norm": 2.058079481124878, "learning_rate": 2.198e-05, "loss": 0.1227, "step": 1100 }, { "grad_norm": 1.8606290817260742, "learning_rate": 2.218e-05, "loss": 0.105, "step": 1110 }, { "grad_norm": 1.6709848642349243, "learning_rate": 2.2380000000000003e-05, "loss": 0.1142, "step": 1120 }, { "grad_norm": 1.7351598739624023, "learning_rate": 2.258e-05, "loss": 0.1125, "step": 1130 }, { "grad_norm": 2.0031752586364746, "learning_rate": 2.2780000000000002e-05, "loss": 0.1109, "step": 1140 }, { "grad_norm": 1.6995254755020142, "learning_rate": 2.298e-05, "loss": 0.1138, "step": 1150 }, { "grad_norm": 1.6779791116714478, "learning_rate": 2.318e-05, "loss": 0.1076, "step": 1160 }, { "grad_norm": 1.9552077054977417, "learning_rate": 2.3380000000000003e-05, "loss": 0.1122, "step": 1170 }, { "grad_norm": 2.059473752975464, "learning_rate": 2.358e-05, "loss": 0.113, "step": 1180 }, { "grad_norm": 1.655415654182434, "learning_rate": 2.3780000000000003e-05, "loss": 0.1041, "step": 1190 }, { "grad_norm": 1.5405687093734741, "learning_rate": 2.398e-05, "loss": 0.112, "step": 1200 }, { "grad_norm": 1.5950849056243896, "learning_rate": 2.418e-05, "loss": 0.1037, "step": 1210 }, { "grad_norm": 1.4484301805496216, "learning_rate": 2.438e-05, "loss": 0.096, "step": 1220 }, { "grad_norm": 1.5433027744293213, "learning_rate": 2.4580000000000002e-05, "loss": 0.1019, "step": 1230 }, { "grad_norm": 1.5371448993682861, "learning_rate": 2.478e-05, "loss": 0.1055, "step": 1240 }, { "grad_norm": 1.4507646560668945, "learning_rate": 2.498e-05, "loss": 0.1048, "step": 1250 }, { "grad_norm": 1.4765233993530273, "learning_rate": 2.5180000000000003e-05, "loss": 0.0891, "step": 1260 }, { "grad_norm": 1.2069228887557983, "learning_rate": 2.5380000000000004e-05, "loss": 0.1074, "step": 1270 }, { "grad_norm": 1.9308463335037231, "learning_rate": 2.5580000000000002e-05, "loss": 0.1013, "step": 1280 }, { "grad_norm": 1.773505449295044, "learning_rate": 2.5779999999999997e-05, "loss": 0.1004, "step": 1290 }, { "grad_norm": 1.8108407258987427, "learning_rate": 2.598e-05, "loss": 0.0988, "step": 1300 }, { "grad_norm": 1.7801239490509033, "learning_rate": 2.618e-05, "loss": 0.096, "step": 1310 }, { "grad_norm": 1.5893381834030151, "learning_rate": 2.6379999999999998e-05, "loss": 0.1098, "step": 1320 }, { "grad_norm": 1.4967936277389526, "learning_rate": 2.658e-05, "loss": 0.0941, "step": 1330 }, { "grad_norm": 1.5417097806930542, "learning_rate": 2.678e-05, "loss": 0.0878, "step": 1340 }, { "grad_norm": 1.5704185962677002, "learning_rate": 2.698e-05, "loss": 0.1009, "step": 1350 }, { "grad_norm": 1.455478549003601, "learning_rate": 2.718e-05, "loss": 0.1029, "step": 1360 }, { "grad_norm": 1.667087197303772, "learning_rate": 2.738e-05, "loss": 0.0923, "step": 1370 }, { "grad_norm": 1.583909511566162, "learning_rate": 2.758e-05, "loss": 0.0894, "step": 1380 }, { "grad_norm": 1.6111894845962524, "learning_rate": 2.778e-05, "loss": 0.0837, "step": 1390 }, { "grad_norm": 1.426738977432251, "learning_rate": 2.798e-05, "loss": 0.0877, "step": 1400 }, { "grad_norm": 1.4552030563354492, "learning_rate": 2.818e-05, "loss": 0.087, "step": 1410 }, { "grad_norm": 1.548185110092163, "learning_rate": 2.8380000000000003e-05, "loss": 0.0885, "step": 1420 }, { "grad_norm": 1.5078884363174438, "learning_rate": 2.858e-05, "loss": 0.1035, "step": 1430 }, { "grad_norm": 1.5525906085968018, "learning_rate": 2.8780000000000002e-05, "loss": 0.0864, "step": 1440 }, { "grad_norm": 1.4639451503753662, "learning_rate": 2.898e-05, "loss": 0.0807, "step": 1450 }, { "grad_norm": 1.3646188974380493, "learning_rate": 2.9180000000000002e-05, "loss": 0.092, "step": 1460 }, { "grad_norm": 1.311794400215149, "learning_rate": 2.9380000000000003e-05, "loss": 0.0864, "step": 1470 }, { "grad_norm": 1.2510517835617065, "learning_rate": 2.958e-05, "loss": 0.0847, "step": 1480 }, { "grad_norm": 1.3302819728851318, "learning_rate": 2.9780000000000003e-05, "loss": 0.095, "step": 1490 }, { "grad_norm": 1.685748815536499, "learning_rate": 2.998e-05, "loss": 0.0854, "step": 1500 }, { "grad_norm": 1.5429292917251587, "learning_rate": 3.0180000000000002e-05, "loss": 0.0901, "step": 1510 }, { "grad_norm": 1.6838124990463257, "learning_rate": 3.0380000000000004e-05, "loss": 0.09, "step": 1520 }, { "grad_norm": 1.2915279865264893, "learning_rate": 3.058e-05, "loss": 0.0797, "step": 1530 }, { "grad_norm": 1.3355578184127808, "learning_rate": 3.078e-05, "loss": 0.0853, "step": 1540 }, { "grad_norm": 1.3953392505645752, "learning_rate": 3.0980000000000005e-05, "loss": 0.0863, "step": 1550 }, { "grad_norm": 1.7329026460647583, "learning_rate": 3.118e-05, "loss": 0.0862, "step": 1560 }, { "grad_norm": 1.5056416988372803, "learning_rate": 3.138e-05, "loss": 0.088, "step": 1570 }, { "grad_norm": 1.362551212310791, "learning_rate": 3.1580000000000006e-05, "loss": 0.0819, "step": 1580 }, { "grad_norm": 1.3808388710021973, "learning_rate": 3.1780000000000004e-05, "loss": 0.0793, "step": 1590 }, { "grad_norm": 1.3953951597213745, "learning_rate": 3.198e-05, "loss": 0.0897, "step": 1600 }, { "grad_norm": 0.9060927033424377, "learning_rate": 3.218e-05, "loss": 0.0798, "step": 1610 }, { "grad_norm": 1.406683087348938, "learning_rate": 3.238e-05, "loss": 0.0786, "step": 1620 }, { "grad_norm": 1.385572910308838, "learning_rate": 3.2579999999999996e-05, "loss": 0.0886, "step": 1630 }, { "grad_norm": 1.415607213973999, "learning_rate": 3.278e-05, "loss": 0.0862, "step": 1640 }, { "grad_norm": 1.208055019378662, "learning_rate": 3.298e-05, "loss": 0.0786, "step": 1650 }, { "grad_norm": 1.1099950075149536, "learning_rate": 3.318e-05, "loss": 0.0772, "step": 1660 }, { "grad_norm": 1.6103872060775757, "learning_rate": 3.338e-05, "loss": 0.0882, "step": 1670 }, { "grad_norm": 1.4322764873504639, "learning_rate": 3.358e-05, "loss": 0.0799, "step": 1680 }, { "grad_norm": 1.3305655717849731, "learning_rate": 3.378e-05, "loss": 0.0757, "step": 1690 }, { "grad_norm": 1.5972950458526611, "learning_rate": 3.398e-05, "loss": 0.0924, "step": 1700 }, { "grad_norm": 1.3676496744155884, "learning_rate": 3.418e-05, "loss": 0.0813, "step": 1710 }, { "grad_norm": 1.5617046356201172, "learning_rate": 3.438e-05, "loss": 0.0824, "step": 1720 }, { "grad_norm": 1.076871633529663, "learning_rate": 3.4580000000000004e-05, "loss": 0.0814, "step": 1730 }, { "grad_norm": 1.4556300640106201, "learning_rate": 3.478e-05, "loss": 0.0765, "step": 1740 }, { "grad_norm": 1.017491102218628, "learning_rate": 3.498e-05, "loss": 0.0741, "step": 1750 }, { "grad_norm": 0.9546638131141663, "learning_rate": 3.518e-05, "loss": 0.0719, "step": 1760 }, { "grad_norm": 1.30952787399292, "learning_rate": 3.5380000000000003e-05, "loss": 0.0928, "step": 1770 }, { "grad_norm": 1.0513885021209717, "learning_rate": 3.558e-05, "loss": 0.0831, "step": 1780 }, { "grad_norm": 1.3774968385696411, "learning_rate": 3.578e-05, "loss": 0.077, "step": 1790 }, { "grad_norm": 1.2923922538757324, "learning_rate": 3.5980000000000004e-05, "loss": 0.086, "step": 1800 }, { "grad_norm": 0.9718403220176697, "learning_rate": 3.618e-05, "loss": 0.0777, "step": 1810 }, { "grad_norm": 1.239272117614746, "learning_rate": 3.638e-05, "loss": 0.0806, "step": 1820 }, { "grad_norm": 1.260016679763794, "learning_rate": 3.6580000000000006e-05, "loss": 0.0759, "step": 1830 }, { "grad_norm": 1.1806528568267822, "learning_rate": 3.6780000000000004e-05, "loss": 0.0746, "step": 1840 }, { "grad_norm": 1.021004319190979, "learning_rate": 3.698e-05, "loss": 0.074, "step": 1850 }, { "grad_norm": 1.156205415725708, "learning_rate": 3.7180000000000007e-05, "loss": 0.0726, "step": 1860 }, { "grad_norm": 1.1397563219070435, "learning_rate": 3.7380000000000005e-05, "loss": 0.076, "step": 1870 }, { "grad_norm": 1.049697756767273, "learning_rate": 3.758e-05, "loss": 0.0736, "step": 1880 }, { "grad_norm": 1.0322024822235107, "learning_rate": 3.778000000000001e-05, "loss": 0.0692, "step": 1890 }, { "grad_norm": 1.2455365657806396, "learning_rate": 3.7980000000000006e-05, "loss": 0.0772, "step": 1900 }, { "grad_norm": 1.4081645011901855, "learning_rate": 3.818e-05, "loss": 0.0792, "step": 1910 }, { "grad_norm": 1.040506362915039, "learning_rate": 3.838e-05, "loss": 0.0703, "step": 1920 }, { "grad_norm": 0.9524482488632202, "learning_rate": 3.858e-05, "loss": 0.0761, "step": 1930 }, { "grad_norm": 1.0899049043655396, "learning_rate": 3.878e-05, "loss": 0.0863, "step": 1940 }, { "grad_norm": 0.9092950820922852, "learning_rate": 3.898e-05, "loss": 0.0736, "step": 1950 }, { "grad_norm": 1.2211405038833618, "learning_rate": 3.918e-05, "loss": 0.0706, "step": 1960 }, { "grad_norm": 0.9409964680671692, "learning_rate": 3.938e-05, "loss": 0.0764, "step": 1970 }, { "grad_norm": 1.0157418251037598, "learning_rate": 3.958e-05, "loss": 0.0768, "step": 1980 }, { "grad_norm": 0.8890381455421448, "learning_rate": 3.978e-05, "loss": 0.0706, "step": 1990 }, { "grad_norm": 0.9611288905143738, "learning_rate": 3.998e-05, "loss": 0.0718, "step": 2000 }, { "grad_norm": 1.2166037559509277, "learning_rate": 4.018e-05, "loss": 0.0731, "step": 2010 }, { "grad_norm": 0.9212486147880554, "learning_rate": 4.038e-05, "loss": 0.0735, "step": 2020 }, { "grad_norm": 1.0661203861236572, "learning_rate": 4.058e-05, "loss": 0.0674, "step": 2030 }, { "grad_norm": 1.0995800495147705, "learning_rate": 4.078e-05, "loss": 0.0744, "step": 2040 }, { "grad_norm": 1.454452395439148, "learning_rate": 4.0980000000000004e-05, "loss": 0.0766, "step": 2050 }, { "grad_norm": 1.1840869188308716, "learning_rate": 4.118e-05, "loss": 0.0772, "step": 2060 }, { "grad_norm": 1.1620875597000122, "learning_rate": 4.138e-05, "loss": 0.0744, "step": 2070 }, { "grad_norm": 0.9451385736465454, "learning_rate": 4.1580000000000005e-05, "loss": 0.0707, "step": 2080 }, { "grad_norm": 1.0929653644561768, "learning_rate": 4.178e-05, "loss": 0.0747, "step": 2090 }, { "grad_norm": 1.048553228378296, "learning_rate": 4.198e-05, "loss": 0.0723, "step": 2100 }, { "grad_norm": 1.2365347146987915, "learning_rate": 4.2180000000000006e-05, "loss": 0.0727, "step": 2110 }, { "grad_norm": 1.2353882789611816, "learning_rate": 4.2380000000000004e-05, "loss": 0.0672, "step": 2120 }, { "grad_norm": 0.9020118713378906, "learning_rate": 4.258e-05, "loss": 0.0751, "step": 2130 }, { "grad_norm": 1.0503695011138916, "learning_rate": 4.278e-05, "loss": 0.0741, "step": 2140 }, { "grad_norm": 1.526049256324768, "learning_rate": 4.2980000000000005e-05, "loss": 0.075, "step": 2150 }, { "grad_norm": 1.022071123123169, "learning_rate": 4.318e-05, "loss": 0.062, "step": 2160 }, { "grad_norm": 1.2062560319900513, "learning_rate": 4.338e-05, "loss": 0.0718, "step": 2170 }, { "grad_norm": 1.0171078443527222, "learning_rate": 4.3580000000000006e-05, "loss": 0.0724, "step": 2180 }, { "grad_norm": 0.8487358689308167, "learning_rate": 4.3780000000000004e-05, "loss": 0.0702, "step": 2190 }, { "grad_norm": 0.987600564956665, "learning_rate": 4.398e-05, "loss": 0.0618, "step": 2200 }, { "grad_norm": 1.0413403511047363, "learning_rate": 4.418000000000001e-05, "loss": 0.0655, "step": 2210 }, { "grad_norm": 1.0169041156768799, "learning_rate": 4.438e-05, "loss": 0.0659, "step": 2220 }, { "grad_norm": 1.0092949867248535, "learning_rate": 4.458e-05, "loss": 0.066, "step": 2230 }, { "grad_norm": 0.9649950265884399, "learning_rate": 4.478e-05, "loss": 0.0698, "step": 2240 }, { "grad_norm": 0.9927655458450317, "learning_rate": 4.498e-05, "loss": 0.0763, "step": 2250 }, { "grad_norm": 1.0967907905578613, "learning_rate": 4.518e-05, "loss": 0.0778, "step": 2260 }, { "grad_norm": 1.1020543575286865, "learning_rate": 4.538e-05, "loss": 0.0754, "step": 2270 }, { "grad_norm": 0.7802571058273315, "learning_rate": 4.558e-05, "loss": 0.0681, "step": 2280 }, { "grad_norm": 0.9696395993232727, "learning_rate": 4.578e-05, "loss": 0.0615, "step": 2290 }, { "grad_norm": 0.6801761984825134, "learning_rate": 4.5980000000000004e-05, "loss": 0.0658, "step": 2300 }, { "grad_norm": 0.940011203289032, "learning_rate": 4.618e-05, "loss": 0.0683, "step": 2310 }, { "grad_norm": 0.7646051049232483, "learning_rate": 4.638e-05, "loss": 0.0671, "step": 2320 }, { "grad_norm": 1.0434030294418335, "learning_rate": 4.6580000000000005e-05, "loss": 0.0709, "step": 2330 }, { "grad_norm": 0.8395400047302246, "learning_rate": 4.678e-05, "loss": 0.0685, "step": 2340 }, { "grad_norm": 1.0456112623214722, "learning_rate": 4.698e-05, "loss": 0.0658, "step": 2350 }, { "grad_norm": 0.9458516240119934, "learning_rate": 4.718e-05, "loss": 0.072, "step": 2360 }, { "grad_norm": 0.9685782194137573, "learning_rate": 4.7380000000000004e-05, "loss": 0.0675, "step": 2370 }, { "grad_norm": 1.03807532787323, "learning_rate": 4.758e-05, "loss": 0.0726, "step": 2380 }, { "grad_norm": 0.8847252130508423, "learning_rate": 4.778e-05, "loss": 0.064, "step": 2390 }, { "grad_norm": 0.7209271192550659, "learning_rate": 4.7980000000000005e-05, "loss": 0.0672, "step": 2400 }, { "grad_norm": 1.0068143606185913, "learning_rate": 4.818e-05, "loss": 0.064, "step": 2410 }, { "grad_norm": 0.763004720211029, "learning_rate": 4.838e-05, "loss": 0.0692, "step": 2420 }, { "grad_norm": 0.9391849040985107, "learning_rate": 4.8580000000000006e-05, "loss": 0.067, "step": 2430 }, { "grad_norm": 1.1383466720581055, "learning_rate": 4.8780000000000004e-05, "loss": 0.0674, "step": 2440 }, { "grad_norm": 0.7036862969398499, "learning_rate": 4.898e-05, "loss": 0.0656, "step": 2450 }, { "grad_norm": 1.0651841163635254, "learning_rate": 4.918000000000001e-05, "loss": 0.0742, "step": 2460 }, { "grad_norm": 1.064895510673523, "learning_rate": 4.9380000000000005e-05, "loss": 0.0634, "step": 2470 }, { "grad_norm": 0.8207208514213562, "learning_rate": 4.958e-05, "loss": 0.0703, "step": 2480 }, { "grad_norm": 0.8498119115829468, "learning_rate": 4.978e-05, "loss": 0.0636, "step": 2490 }, { "grad_norm": 0.8440943956375122, "learning_rate": 4.9980000000000006e-05, "loss": 0.066, "step": 2500 }, { "grad_norm": 0.8588313460350037, "learning_rate": 5.0180000000000004e-05, "loss": 0.0695, "step": 2510 }, { "grad_norm": 0.7617659568786621, "learning_rate": 5.038e-05, "loss": 0.0631, "step": 2520 }, { "grad_norm": 1.0904728174209595, "learning_rate": 5.058000000000001e-05, "loss": 0.0612, "step": 2530 }, { "grad_norm": 0.5877794027328491, "learning_rate": 5.0780000000000005e-05, "loss": 0.0584, "step": 2540 }, { "grad_norm": 1.230660080909729, "learning_rate": 5.098e-05, "loss": 0.0678, "step": 2550 }, { "grad_norm": 0.6841692328453064, "learning_rate": 5.118000000000001e-05, "loss": 0.061, "step": 2560 }, { "grad_norm": 0.8808116912841797, "learning_rate": 5.1380000000000006e-05, "loss": 0.0617, "step": 2570 }, { "grad_norm": 1.1295050382614136, "learning_rate": 5.1580000000000004e-05, "loss": 0.0712, "step": 2580 }, { "grad_norm": 0.9928704500198364, "learning_rate": 5.178000000000001e-05, "loss": 0.0621, "step": 2590 }, { "grad_norm": 0.9991413354873657, "learning_rate": 5.198000000000001e-05, "loss": 0.0653, "step": 2600 }, { "grad_norm": 0.9295065999031067, "learning_rate": 5.2180000000000005e-05, "loss": 0.0664, "step": 2610 }, { "grad_norm": 1.0966871976852417, "learning_rate": 5.238000000000001e-05, "loss": 0.0617, "step": 2620 }, { "grad_norm": 0.841902494430542, "learning_rate": 5.258000000000001e-05, "loss": 0.0697, "step": 2630 }, { "grad_norm": 0.6770904660224915, "learning_rate": 5.2780000000000006e-05, "loss": 0.0647, "step": 2640 }, { "grad_norm": 0.9994043707847595, "learning_rate": 5.2980000000000004e-05, "loss": 0.0649, "step": 2650 }, { "grad_norm": 0.7011931538581848, "learning_rate": 5.318000000000001e-05, "loss": 0.062, "step": 2660 }, { "grad_norm": 0.7652668356895447, "learning_rate": 5.338000000000001e-05, "loss": 0.0615, "step": 2670 }, { "grad_norm": 0.8984162211418152, "learning_rate": 5.3580000000000005e-05, "loss": 0.0565, "step": 2680 }, { "grad_norm": 0.8604142069816589, "learning_rate": 5.378e-05, "loss": 0.0646, "step": 2690 }, { "grad_norm": 0.7709184885025024, "learning_rate": 5.3979999999999995e-05, "loss": 0.0641, "step": 2700 }, { "grad_norm": 0.9449307918548584, "learning_rate": 5.418e-05, "loss": 0.0675, "step": 2710 }, { "grad_norm": 0.8418259620666504, "learning_rate": 5.438e-05, "loss": 0.0584, "step": 2720 }, { "grad_norm": 0.8688193559646606, "learning_rate": 5.4579999999999996e-05, "loss": 0.0649, "step": 2730 }, { "grad_norm": 0.7081212401390076, "learning_rate": 5.478e-05, "loss": 0.064, "step": 2740 }, { "grad_norm": 0.9140228629112244, "learning_rate": 5.498e-05, "loss": 0.0618, "step": 2750 }, { "grad_norm": 0.9443033933639526, "learning_rate": 5.518e-05, "loss": 0.0647, "step": 2760 }, { "grad_norm": 0.6639568209648132, "learning_rate": 5.538e-05, "loss": 0.0582, "step": 2770 }, { "grad_norm": 0.5913040041923523, "learning_rate": 5.558e-05, "loss": 0.0597, "step": 2780 }, { "grad_norm": 0.8690099716186523, "learning_rate": 5.578e-05, "loss": 0.0601, "step": 2790 }, { "grad_norm": 0.7661793828010559, "learning_rate": 5.5979999999999996e-05, "loss": 0.0608, "step": 2800 }, { "grad_norm": 0.7647290229797363, "learning_rate": 5.618e-05, "loss": 0.0577, "step": 2810 }, { "grad_norm": 0.7679623365402222, "learning_rate": 5.638e-05, "loss": 0.0512, "step": 2820 }, { "grad_norm": 0.7960045337677002, "learning_rate": 5.658e-05, "loss": 0.0619, "step": 2830 }, { "grad_norm": 0.7950032353401184, "learning_rate": 5.678e-05, "loss": 0.062, "step": 2840 }, { "grad_norm": 0.6819756031036377, "learning_rate": 5.698e-05, "loss": 0.0569, "step": 2850 }, { "grad_norm": 0.8518178462982178, "learning_rate": 5.718e-05, "loss": 0.0565, "step": 2860 }, { "grad_norm": 0.7512794137001038, "learning_rate": 5.738e-05, "loss": 0.0604, "step": 2870 }, { "grad_norm": 0.5904653668403625, "learning_rate": 5.758e-05, "loss": 0.0544, "step": 2880 }, { "grad_norm": 0.42944639921188354, "learning_rate": 5.778e-05, "loss": 0.0545, "step": 2890 }, { "grad_norm": 0.9649351835250854, "learning_rate": 5.7980000000000004e-05, "loss": 0.0569, "step": 2900 }, { "grad_norm": 0.8394961357116699, "learning_rate": 5.818e-05, "loss": 0.0564, "step": 2910 }, { "grad_norm": 0.7755361795425415, "learning_rate": 5.838e-05, "loss": 0.0529, "step": 2920 }, { "grad_norm": 0.6657257676124573, "learning_rate": 5.858e-05, "loss": 0.0651, "step": 2930 }, { "grad_norm": 0.6956703662872314, "learning_rate": 5.878e-05, "loss": 0.0649, "step": 2940 }, { "grad_norm": 0.7149800062179565, "learning_rate": 5.898e-05, "loss": 0.0619, "step": 2950 }, { "grad_norm": 0.8453929424285889, "learning_rate": 5.918e-05, "loss": 0.0629, "step": 2960 }, { "grad_norm": 0.7426488995552063, "learning_rate": 5.9380000000000004e-05, "loss": 0.0578, "step": 2970 }, { "grad_norm": 0.9025492072105408, "learning_rate": 5.958e-05, "loss": 0.0566, "step": 2980 }, { "grad_norm": 0.8643544912338257, "learning_rate": 5.978e-05, "loss": 0.0566, "step": 2990 }, { "grad_norm": 0.682625412940979, "learning_rate": 5.9980000000000005e-05, "loss": 0.0569, "step": 3000 }, { "grad_norm": 0.754028856754303, "learning_rate": 6.018e-05, "loss": 0.0651, "step": 3010 }, { "grad_norm": 0.9288793206214905, "learning_rate": 6.038e-05, "loss": 0.061, "step": 3020 }, { "grad_norm": 0.7358014583587646, "learning_rate": 6.0580000000000006e-05, "loss": 0.057, "step": 3030 }, { "grad_norm": 0.6921884417533875, "learning_rate": 6.0780000000000004e-05, "loss": 0.0557, "step": 3040 }, { "grad_norm": 0.6683744192123413, "learning_rate": 6.098e-05, "loss": 0.0587, "step": 3050 }, { "grad_norm": 0.8776196241378784, "learning_rate": 6.118000000000001e-05, "loss": 0.0593, "step": 3060 }, { "grad_norm": 0.8616880774497986, "learning_rate": 6.138e-05, "loss": 0.0604, "step": 3070 }, { "grad_norm": 0.6801915764808655, "learning_rate": 6.158e-05, "loss": 0.0577, "step": 3080 }, { "grad_norm": 0.7711888551712036, "learning_rate": 6.178000000000001e-05, "loss": 0.0566, "step": 3090 }, { "grad_norm": 0.7812537550926208, "learning_rate": 6.198e-05, "loss": 0.0591, "step": 3100 }, { "grad_norm": 0.7505896687507629, "learning_rate": 6.218e-05, "loss": 0.0658, "step": 3110 }, { "grad_norm": 0.6621214151382446, "learning_rate": 6.238000000000001e-05, "loss": 0.0515, "step": 3120 }, { "grad_norm": 0.706023633480072, "learning_rate": 6.258e-05, "loss": 0.0616, "step": 3130 }, { "grad_norm": 0.7778743505477905, "learning_rate": 6.278e-05, "loss": 0.0529, "step": 3140 }, { "grad_norm": 0.8823725581169128, "learning_rate": 6.298000000000001e-05, "loss": 0.0608, "step": 3150 }, { "grad_norm": 0.6907094120979309, "learning_rate": 6.318e-05, "loss": 0.0518, "step": 3160 }, { "grad_norm": 0.5903347730636597, "learning_rate": 6.338e-05, "loss": 0.0571, "step": 3170 }, { "grad_norm": 0.7423139810562134, "learning_rate": 6.358000000000001e-05, "loss": 0.058, "step": 3180 }, { "grad_norm": 0.7213485240936279, "learning_rate": 6.378e-05, "loss": 0.0605, "step": 3190 }, { "grad_norm": 0.7129867076873779, "learning_rate": 6.398000000000001e-05, "loss": 0.0533, "step": 3200 }, { "grad_norm": 0.8217514753341675, "learning_rate": 6.418000000000001e-05, "loss": 0.0522, "step": 3210 }, { "grad_norm": 0.829096794128418, "learning_rate": 6.438e-05, "loss": 0.0572, "step": 3220 }, { "grad_norm": 0.7662721276283264, "learning_rate": 6.458000000000001e-05, "loss": 0.0643, "step": 3230 }, { "grad_norm": 0.6620997190475464, "learning_rate": 6.478000000000001e-05, "loss": 0.0577, "step": 3240 }, { "grad_norm": 0.685779333114624, "learning_rate": 6.498e-05, "loss": 0.0583, "step": 3250 }, { "grad_norm": 0.6641210913658142, "learning_rate": 6.518000000000001e-05, "loss": 0.0575, "step": 3260 }, { "grad_norm": 0.8922712802886963, "learning_rate": 6.538000000000001e-05, "loss": 0.0595, "step": 3270 }, { "grad_norm": 0.6155965328216553, "learning_rate": 6.558e-05, "loss": 0.0519, "step": 3280 }, { "grad_norm": 0.6449828743934631, "learning_rate": 6.578000000000001e-05, "loss": 0.0561, "step": 3290 }, { "grad_norm": 0.8047433495521545, "learning_rate": 6.598e-05, "loss": 0.0562, "step": 3300 }, { "grad_norm": 0.5262227654457092, "learning_rate": 6.618e-05, "loss": 0.0555, "step": 3310 }, { "grad_norm": 0.632722795009613, "learning_rate": 6.638e-05, "loss": 0.0573, "step": 3320 }, { "grad_norm": 0.7328101992607117, "learning_rate": 6.658e-05, "loss": 0.0541, "step": 3330 }, { "grad_norm": 0.6979526281356812, "learning_rate": 6.678e-05, "loss": 0.0576, "step": 3340 }, { "grad_norm": 0.8734522461891174, "learning_rate": 6.698e-05, "loss": 0.0591, "step": 3350 }, { "grad_norm": 0.6581406593322754, "learning_rate": 6.718e-05, "loss": 0.0544, "step": 3360 }, { "grad_norm": 0.5466970801353455, "learning_rate": 6.738e-05, "loss": 0.0544, "step": 3370 }, { "grad_norm": 0.6768118739128113, "learning_rate": 6.758e-05, "loss": 0.0528, "step": 3380 }, { "grad_norm": 0.6351532936096191, "learning_rate": 6.778e-05, "loss": 0.0567, "step": 3390 }, { "grad_norm": 0.6079491376876831, "learning_rate": 6.798e-05, "loss": 0.057, "step": 3400 }, { "grad_norm": 0.6835048794746399, "learning_rate": 6.818e-05, "loss": 0.0622, "step": 3410 }, { "grad_norm": 0.6320196390151978, "learning_rate": 6.838e-05, "loss": 0.0537, "step": 3420 }, { "grad_norm": 0.6056031584739685, "learning_rate": 6.858e-05, "loss": 0.0508, "step": 3430 }, { "grad_norm": 0.7936848998069763, "learning_rate": 6.878e-05, "loss": 0.0581, "step": 3440 }, { "grad_norm": 0.8324952125549316, "learning_rate": 6.898e-05, "loss": 0.0485, "step": 3450 }, { "grad_norm": 0.6675503253936768, "learning_rate": 6.918e-05, "loss": 0.0537, "step": 3460 }, { "grad_norm": 0.7332781553268433, "learning_rate": 6.938e-05, "loss": 0.0601, "step": 3470 }, { "grad_norm": 0.563319206237793, "learning_rate": 6.958e-05, "loss": 0.056, "step": 3480 }, { "grad_norm": 0.691288948059082, "learning_rate": 6.978e-05, "loss": 0.0553, "step": 3490 }, { "grad_norm": 0.6328545808792114, "learning_rate": 6.998e-05, "loss": 0.0516, "step": 3500 }, { "grad_norm": 0.8291574120521545, "learning_rate": 7.018e-05, "loss": 0.0589, "step": 3510 }, { "grad_norm": 0.5662462711334229, "learning_rate": 7.038e-05, "loss": 0.0525, "step": 3520 }, { "grad_norm": 0.6067913770675659, "learning_rate": 7.058e-05, "loss": 0.0534, "step": 3530 }, { "grad_norm": 0.6333366632461548, "learning_rate": 7.078e-05, "loss": 0.0539, "step": 3540 }, { "grad_norm": 0.8065729737281799, "learning_rate": 7.098e-05, "loss": 0.053, "step": 3550 }, { "grad_norm": 0.6301604509353638, "learning_rate": 7.118e-05, "loss": 0.0548, "step": 3560 }, { "grad_norm": 0.7065660357475281, "learning_rate": 7.138e-05, "loss": 0.0565, "step": 3570 }, { "grad_norm": 0.6226935982704163, "learning_rate": 7.158e-05, "loss": 0.0569, "step": 3580 }, { "grad_norm": 0.655795156955719, "learning_rate": 7.178000000000001e-05, "loss": 0.0502, "step": 3590 }, { "grad_norm": 0.6377870440483093, "learning_rate": 7.198e-05, "loss": 0.052, "step": 3600 }, { "grad_norm": 0.7083209156990051, "learning_rate": 7.218e-05, "loss": 0.0523, "step": 3610 }, { "grad_norm": 0.6695072054862976, "learning_rate": 7.238000000000001e-05, "loss": 0.0502, "step": 3620 }, { "grad_norm": 0.5858951210975647, "learning_rate": 7.258e-05, "loss": 0.0523, "step": 3630 }, { "grad_norm": 0.6276965141296387, "learning_rate": 7.278e-05, "loss": 0.0475, "step": 3640 }, { "grad_norm": 0.5498921275138855, "learning_rate": 7.298000000000001e-05, "loss": 0.0477, "step": 3650 }, { "grad_norm": 0.7147716283798218, "learning_rate": 7.318e-05, "loss": 0.0512, "step": 3660 }, { "grad_norm": 0.560576319694519, "learning_rate": 7.338e-05, "loss": 0.0496, "step": 3670 }, { "grad_norm": 0.6377435326576233, "learning_rate": 7.358000000000001e-05, "loss": 0.0469, "step": 3680 }, { "grad_norm": 0.5697843432426453, "learning_rate": 7.378e-05, "loss": 0.057, "step": 3690 }, { "grad_norm": 0.79034423828125, "learning_rate": 7.398e-05, "loss": 0.0488, "step": 3700 }, { "grad_norm": 0.5773378610610962, "learning_rate": 7.418000000000001e-05, "loss": 0.0507, "step": 3710 }, { "grad_norm": 0.4400804340839386, "learning_rate": 7.438e-05, "loss": 0.0519, "step": 3720 }, { "grad_norm": 0.5750156044960022, "learning_rate": 7.458000000000001e-05, "loss": 0.0545, "step": 3730 }, { "grad_norm": 0.6422926187515259, "learning_rate": 7.478e-05, "loss": 0.054, "step": 3740 }, { "grad_norm": 0.6804867386817932, "learning_rate": 7.498e-05, "loss": 0.052, "step": 3750 }, { "grad_norm": 0.4712640345096588, "learning_rate": 7.518000000000001e-05, "loss": 0.0533, "step": 3760 }, { "grad_norm": 0.6688985824584961, "learning_rate": 7.538e-05, "loss": 0.0465, "step": 3770 }, { "grad_norm": 0.6465003490447998, "learning_rate": 7.558e-05, "loss": 0.0524, "step": 3780 }, { "grad_norm": 0.6730716824531555, "learning_rate": 7.578000000000001e-05, "loss": 0.0502, "step": 3790 }, { "grad_norm": 0.610188901424408, "learning_rate": 7.598e-05, "loss": 0.0538, "step": 3800 }, { "grad_norm": 0.6906505227088928, "learning_rate": 7.618e-05, "loss": 0.0579, "step": 3810 }, { "grad_norm": 0.7066339254379272, "learning_rate": 7.638000000000001e-05, "loss": 0.0621, "step": 3820 }, { "grad_norm": 0.47428205609321594, "learning_rate": 7.658e-05, "loss": 0.0501, "step": 3830 }, { "grad_norm": 0.571767270565033, "learning_rate": 7.678000000000001e-05, "loss": 0.0499, "step": 3840 }, { "grad_norm": 0.6406869888305664, "learning_rate": 7.698000000000001e-05, "loss": 0.0529, "step": 3850 }, { "grad_norm": 0.5015462636947632, "learning_rate": 7.718e-05, "loss": 0.0515, "step": 3860 }, { "grad_norm": 0.5495204329490662, "learning_rate": 7.738000000000001e-05, "loss": 0.0474, "step": 3870 }, { "grad_norm": 0.584661066532135, "learning_rate": 7.758000000000001e-05, "loss": 0.0502, "step": 3880 }, { "grad_norm": 0.5329467058181763, "learning_rate": 7.778e-05, "loss": 0.054, "step": 3890 }, { "grad_norm": 0.4563973844051361, "learning_rate": 7.798000000000001e-05, "loss": 0.0559, "step": 3900 }, { "grad_norm": 0.6789462566375732, "learning_rate": 7.818000000000001e-05, "loss": 0.0522, "step": 3910 }, { "grad_norm": 0.6281589865684509, "learning_rate": 7.838e-05, "loss": 0.0627, "step": 3920 }, { "grad_norm": 0.5551241636276245, "learning_rate": 7.858000000000001e-05, "loss": 0.048, "step": 3930 }, { "grad_norm": 0.7618772387504578, "learning_rate": 7.878e-05, "loss": 0.0517, "step": 3940 }, { "grad_norm": 0.6502484679222107, "learning_rate": 7.897999999999999e-05, "loss": 0.0514, "step": 3950 }, { "grad_norm": 0.6113073825836182, "learning_rate": 7.918e-05, "loss": 0.0492, "step": 3960 }, { "grad_norm": 0.6507073044776917, "learning_rate": 7.938e-05, "loss": 0.048, "step": 3970 }, { "grad_norm": 0.506917417049408, "learning_rate": 7.958e-05, "loss": 0.0463, "step": 3980 }, { "grad_norm": 0.5822247862815857, "learning_rate": 7.978e-05, "loss": 0.0489, "step": 3990 }, { "grad_norm": 0.6221882700920105, "learning_rate": 7.998e-05, "loss": 0.0539, "step": 4000 }, { "grad_norm": 0.49984636902809143, "learning_rate": 8.018e-05, "loss": 0.0486, "step": 4010 }, { "grad_norm": 0.5334073305130005, "learning_rate": 8.038e-05, "loss": 0.048, "step": 4020 }, { "grad_norm": 0.5979265570640564, "learning_rate": 8.058e-05, "loss": 0.0529, "step": 4030 }, { "grad_norm": 0.5348339080810547, "learning_rate": 8.078e-05, "loss": 0.0465, "step": 4040 }, { "grad_norm": 0.48906242847442627, "learning_rate": 8.098e-05, "loss": 0.0514, "step": 4050 }, { "grad_norm": 0.46153944730758667, "learning_rate": 8.118e-05, "loss": 0.0508, "step": 4060 }, { "grad_norm": 0.5767505168914795, "learning_rate": 8.138e-05, "loss": 0.046, "step": 4070 }, { "grad_norm": 0.548282265663147, "learning_rate": 8.158e-05, "loss": 0.05, "step": 4080 }, { "grad_norm": 0.48629969358444214, "learning_rate": 8.178e-05, "loss": 0.0487, "step": 4090 }, { "grad_norm": 0.5570286512374878, "learning_rate": 8.198e-05, "loss": 0.0457, "step": 4100 }, { "grad_norm": 0.5384827256202698, "learning_rate": 8.218e-05, "loss": 0.0525, "step": 4110 }, { "grad_norm": 0.48497098684310913, "learning_rate": 8.238000000000001e-05, "loss": 0.0526, "step": 4120 }, { "grad_norm": 0.6552274823188782, "learning_rate": 8.258e-05, "loss": 0.0559, "step": 4130 }, { "grad_norm": 0.5505346655845642, "learning_rate": 8.278e-05, "loss": 0.05, "step": 4140 }, { "grad_norm": 0.6627893447875977, "learning_rate": 8.298000000000001e-05, "loss": 0.0521, "step": 4150 }, { "grad_norm": 0.46908432245254517, "learning_rate": 8.318e-05, "loss": 0.0482, "step": 4160 }, { "grad_norm": 0.4426194727420807, "learning_rate": 8.338e-05, "loss": 0.0514, "step": 4170 }, { "grad_norm": 0.5676274299621582, "learning_rate": 8.358e-05, "loss": 0.051, "step": 4180 }, { "grad_norm": 0.5120558142662048, "learning_rate": 8.378e-05, "loss": 0.0539, "step": 4190 }, { "grad_norm": 0.7255561947822571, "learning_rate": 8.398e-05, "loss": 0.0467, "step": 4200 }, { "grad_norm": 0.6107543706893921, "learning_rate": 8.418e-05, "loss": 0.0483, "step": 4210 }, { "grad_norm": 0.7340674996376038, "learning_rate": 8.438e-05, "loss": 0.0501, "step": 4220 }, { "grad_norm": 0.4928207993507385, "learning_rate": 8.458e-05, "loss": 0.0491, "step": 4230 }, { "grad_norm": 0.6419109106063843, "learning_rate": 8.478e-05, "loss": 0.0495, "step": 4240 }, { "grad_norm": 0.6129544377326965, "learning_rate": 8.498e-05, "loss": 0.0478, "step": 4250 }, { "grad_norm": 0.6506026983261108, "learning_rate": 8.518000000000001e-05, "loss": 0.0514, "step": 4260 }, { "grad_norm": 0.5956999063491821, "learning_rate": 8.538e-05, "loss": 0.0489, "step": 4270 }, { "grad_norm": 0.5743863582611084, "learning_rate": 8.558e-05, "loss": 0.0518, "step": 4280 }, { "grad_norm": 0.4685247540473938, "learning_rate": 8.578000000000001e-05, "loss": 0.0464, "step": 4290 }, { "grad_norm": 0.5139967799186707, "learning_rate": 8.598e-05, "loss": 0.0434, "step": 4300 }, { "grad_norm": 0.6676983833312988, "learning_rate": 8.618e-05, "loss": 0.0464, "step": 4310 }, { "grad_norm": 0.43759000301361084, "learning_rate": 8.638000000000001e-05, "loss": 0.0503, "step": 4320 }, { "grad_norm": 0.5922346115112305, "learning_rate": 8.658e-05, "loss": 0.048, "step": 4330 }, { "grad_norm": 0.5554268956184387, "learning_rate": 8.678e-05, "loss": 0.049, "step": 4340 }, { "grad_norm": 0.44343939423561096, "learning_rate": 8.698000000000001e-05, "loss": 0.0537, "step": 4350 }, { "grad_norm": 0.43446454405784607, "learning_rate": 8.718e-05, "loss": 0.0444, "step": 4360 }, { "grad_norm": 0.4966972768306732, "learning_rate": 8.738000000000001e-05, "loss": 0.0495, "step": 4370 }, { "grad_norm": 0.6118032336235046, "learning_rate": 8.758000000000001e-05, "loss": 0.0513, "step": 4380 }, { "grad_norm": 0.6128403544425964, "learning_rate": 8.778e-05, "loss": 0.0577, "step": 4390 }, { "grad_norm": 0.6486364006996155, "learning_rate": 8.798000000000001e-05, "loss": 0.049, "step": 4400 }, { "grad_norm": 0.5506572127342224, "learning_rate": 8.818000000000001e-05, "loss": 0.0503, "step": 4410 }, { "grad_norm": 0.558600127696991, "learning_rate": 8.838e-05, "loss": 0.0486, "step": 4420 }, { "grad_norm": 0.6284540891647339, "learning_rate": 8.858000000000001e-05, "loss": 0.0469, "step": 4430 }, { "grad_norm": 0.5778012871742249, "learning_rate": 8.878000000000001e-05, "loss": 0.0485, "step": 4440 }, { "grad_norm": 0.5883769989013672, "learning_rate": 8.898e-05, "loss": 0.0491, "step": 4450 }, { "grad_norm": 0.7078779339790344, "learning_rate": 8.918000000000001e-05, "loss": 0.0473, "step": 4460 }, { "grad_norm": 0.44417819380760193, "learning_rate": 8.938e-05, "loss": 0.0459, "step": 4470 }, { "grad_norm": 0.6385723948478699, "learning_rate": 8.958e-05, "loss": 0.0489, "step": 4480 }, { "grad_norm": 0.6533469557762146, "learning_rate": 8.978000000000001e-05, "loss": 0.0545, "step": 4490 }, { "grad_norm": 0.4954751133918762, "learning_rate": 8.998e-05, "loss": 0.0473, "step": 4500 }, { "grad_norm": 0.4292506277561188, "learning_rate": 9.018000000000001e-05, "loss": 0.0436, "step": 4510 }, { "grad_norm": 0.5257400870323181, "learning_rate": 9.038000000000001e-05, "loss": 0.045, "step": 4520 }, { "grad_norm": 0.6952337026596069, "learning_rate": 9.058e-05, "loss": 0.0507, "step": 4530 }, { "grad_norm": 0.5642014145851135, "learning_rate": 9.078000000000001e-05, "loss": 0.0494, "step": 4540 }, { "grad_norm": 0.3371942341327667, "learning_rate": 9.098000000000001e-05, "loss": 0.0437, "step": 4550 }, { "grad_norm": 0.36483561992645264, "learning_rate": 9.118e-05, "loss": 0.0441, "step": 4560 }, { "grad_norm": 0.393741637468338, "learning_rate": 9.138e-05, "loss": 0.0448, "step": 4570 }, { "grad_norm": 0.5485238432884216, "learning_rate": 9.158e-05, "loss": 0.054, "step": 4580 }, { "grad_norm": 0.4129837453365326, "learning_rate": 9.178e-05, "loss": 0.0521, "step": 4590 }, { "grad_norm": 0.3965202867984772, "learning_rate": 9.198e-05, "loss": 0.0455, "step": 4600 }, { "grad_norm": 0.48870721459388733, "learning_rate": 9.218e-05, "loss": 0.0515, "step": 4610 }, { "grad_norm": 0.45476090908050537, "learning_rate": 9.238e-05, "loss": 0.0482, "step": 4620 }, { "grad_norm": 0.6202360987663269, "learning_rate": 9.258e-05, "loss": 0.0485, "step": 4630 }, { "grad_norm": 0.4131334125995636, "learning_rate": 9.278e-05, "loss": 0.0488, "step": 4640 }, { "grad_norm": 0.5266363620758057, "learning_rate": 9.298e-05, "loss": 0.0441, "step": 4650 }, { "grad_norm": 0.4210203289985657, "learning_rate": 9.318e-05, "loss": 0.0484, "step": 4660 }, { "grad_norm": 0.40804803371429443, "learning_rate": 9.338e-05, "loss": 0.0448, "step": 4670 }, { "grad_norm": 0.47611290216445923, "learning_rate": 9.358e-05, "loss": 0.0423, "step": 4680 }, { "grad_norm": 0.49742504954338074, "learning_rate": 9.378e-05, "loss": 0.0463, "step": 4690 }, { "grad_norm": 0.4997280240058899, "learning_rate": 9.398e-05, "loss": 0.0547, "step": 4700 }, { "grad_norm": 0.4569966197013855, "learning_rate": 9.418e-05, "loss": 0.0497, "step": 4710 }, { "grad_norm": 0.5915447473526001, "learning_rate": 9.438e-05, "loss": 0.0433, "step": 4720 }, { "grad_norm": 0.5859195590019226, "learning_rate": 9.458e-05, "loss": 0.0465, "step": 4730 }, { "grad_norm": 0.601357638835907, "learning_rate": 9.478e-05, "loss": 0.0469, "step": 4740 }, { "grad_norm": 0.5418255925178528, "learning_rate": 9.498e-05, "loss": 0.0454, "step": 4750 }, { "grad_norm": 0.5514786839485168, "learning_rate": 9.518000000000001e-05, "loss": 0.0514, "step": 4760 }, { "grad_norm": 0.4372907876968384, "learning_rate": 9.538e-05, "loss": 0.0551, "step": 4770 }, { "grad_norm": 0.4601995050907135, "learning_rate": 9.558e-05, "loss": 0.0478, "step": 4780 }, { "grad_norm": 0.45894548296928406, "learning_rate": 9.578000000000001e-05, "loss": 0.0499, "step": 4790 }, { "grad_norm": 0.5423482060432434, "learning_rate": 9.598e-05, "loss": 0.0494, "step": 4800 }, { "grad_norm": 0.4714123606681824, "learning_rate": 9.618e-05, "loss": 0.0453, "step": 4810 }, { "grad_norm": 0.49344268441200256, "learning_rate": 9.638000000000001e-05, "loss": 0.0458, "step": 4820 }, { "grad_norm": 0.41661375761032104, "learning_rate": 9.658e-05, "loss": 0.0445, "step": 4830 }, { "grad_norm": 0.4327770471572876, "learning_rate": 9.678e-05, "loss": 0.0447, "step": 4840 }, { "grad_norm": 0.41495978832244873, "learning_rate": 9.698000000000001e-05, "loss": 0.0473, "step": 4850 }, { "grad_norm": 0.47467562556266785, "learning_rate": 9.718e-05, "loss": 0.0429, "step": 4860 }, { "grad_norm": 0.5802949070930481, "learning_rate": 9.738e-05, "loss": 0.0503, "step": 4870 }, { "grad_norm": 0.39776504039764404, "learning_rate": 9.758000000000001e-05, "loss": 0.044, "step": 4880 }, { "grad_norm": 0.4538721740245819, "learning_rate": 9.778e-05, "loss": 0.0462, "step": 4890 }, { "grad_norm": 0.4459095597267151, "learning_rate": 9.798000000000001e-05, "loss": 0.0457, "step": 4900 }, { "grad_norm": 0.3916681110858917, "learning_rate": 9.818000000000001e-05, "loss": 0.043, "step": 4910 }, { "grad_norm": 0.532321572303772, "learning_rate": 9.838e-05, "loss": 0.0492, "step": 4920 }, { "grad_norm": 0.6300221681594849, "learning_rate": 9.858000000000001e-05, "loss": 0.0492, "step": 4930 }, { "grad_norm": 0.5011418461799622, "learning_rate": 9.878e-05, "loss": 0.0419, "step": 4940 }, { "grad_norm": 0.5453963875770569, "learning_rate": 9.898e-05, "loss": 0.0474, "step": 4950 }, { "grad_norm": 0.5667087435722351, "learning_rate": 9.918000000000001e-05, "loss": 0.0476, "step": 4960 }, { "grad_norm": 0.5009118318557739, "learning_rate": 9.938e-05, "loss": 0.0404, "step": 4970 }, { "grad_norm": 0.43450728058815, "learning_rate": 9.958e-05, "loss": 0.0491, "step": 4980 }, { "grad_norm": 0.6190786361694336, "learning_rate": 9.978000000000001e-05, "loss": 0.0494, "step": 4990 }, { "grad_norm": 0.5346346497535706, "learning_rate": 9.998e-05, "loss": 0.0474, "step": 5000 }, { "grad_norm": 0.4930327832698822, "learning_rate": 9.999999778549045e-05, "loss": 0.0459, "step": 5010 }, { "grad_norm": 0.4875353276729584, "learning_rate": 9.999999013039593e-05, "loss": 0.0498, "step": 5020 }, { "grad_norm": 0.4928962290287018, "learning_rate": 9.999997700737766e-05, "loss": 0.0445, "step": 5030 }, { "grad_norm": 0.5196244716644287, "learning_rate": 9.999995841643709e-05, "loss": 0.0449, "step": 5040 }, { "grad_norm": 0.4910435378551483, "learning_rate": 9.999993435757623e-05, "loss": 0.0463, "step": 5050 }, { "grad_norm": 0.373821884393692, "learning_rate": 9.999990483079773e-05, "loss": 0.0494, "step": 5060 }, { "grad_norm": 0.5007010102272034, "learning_rate": 9.999986983610481e-05, "loss": 0.0451, "step": 5070 }, { "grad_norm": 0.496733158826828, "learning_rate": 9.99998293735013e-05, "loss": 0.0438, "step": 5080 }, { "grad_norm": 0.5700880885124207, "learning_rate": 9.999978344299161e-05, "loss": 0.0476, "step": 5090 }, { "grad_norm": 0.4850098788738251, "learning_rate": 9.99997320445808e-05, "loss": 0.0472, "step": 5100 }, { "grad_norm": 0.4947509169578552, "learning_rate": 9.999967517827444e-05, "loss": 0.0471, "step": 5110 }, { "grad_norm": 0.4341823160648346, "learning_rate": 9.999961284407879e-05, "loss": 0.0443, "step": 5120 }, { "grad_norm": 0.4102516174316406, "learning_rate": 9.999954504200067e-05, "loss": 0.0458, "step": 5130 }, { "grad_norm": 0.4239036440849304, "learning_rate": 9.999947177204744e-05, "loss": 0.0474, "step": 5140 }, { "grad_norm": 0.418055921792984, "learning_rate": 9.999939303422718e-05, "loss": 0.0464, "step": 5150 }, { "grad_norm": 0.5444912314414978, "learning_rate": 9.999930882854847e-05, "loss": 0.0449, "step": 5160 }, { "grad_norm": 0.4048571288585663, "learning_rate": 9.999921915502051e-05, "loss": 0.0421, "step": 5170 }, { "grad_norm": 0.35428309440612793, "learning_rate": 9.99991240136531e-05, "loss": 0.04, "step": 5180 }, { "grad_norm": 0.3673541843891144, "learning_rate": 9.999902340445668e-05, "loss": 0.0443, "step": 5190 }, { "grad_norm": 0.4074915051460266, "learning_rate": 9.999891732744224e-05, "loss": 0.0427, "step": 5200 }, { "grad_norm": 0.6064218282699585, "learning_rate": 9.999880578262135e-05, "loss": 0.0503, "step": 5210 }, { "grad_norm": 0.4345485270023346, "learning_rate": 9.999868877000624e-05, "loss": 0.052, "step": 5220 }, { "grad_norm": 0.4451915919780731, "learning_rate": 9.99985662896097e-05, "loss": 0.0474, "step": 5230 }, { "grad_norm": 0.35080599784851074, "learning_rate": 9.999843834144513e-05, "loss": 0.0393, "step": 5240 }, { "grad_norm": 0.35985806584358215, "learning_rate": 9.99983049255265e-05, "loss": 0.0431, "step": 5250 }, { "grad_norm": 0.45487740635871887, "learning_rate": 9.999816604186843e-05, "loss": 0.0424, "step": 5260 }, { "grad_norm": 0.5120032429695129, "learning_rate": 9.999802169048609e-05, "loss": 0.0449, "step": 5270 }, { "grad_norm": 0.44850367307662964, "learning_rate": 9.999787187139527e-05, "loss": 0.0457, "step": 5280 }, { "grad_norm": 0.3980175256729126, "learning_rate": 9.999771658461234e-05, "loss": 0.0454, "step": 5290 }, { "grad_norm": 0.3866479694843292, "learning_rate": 9.999755583015431e-05, "loss": 0.0486, "step": 5300 }, { "grad_norm": 0.42455771565437317, "learning_rate": 9.999738960803874e-05, "loss": 0.0529, "step": 5310 }, { "grad_norm": 0.5284426212310791, "learning_rate": 9.99972179182838e-05, "loss": 0.0533, "step": 5320 }, { "grad_norm": 0.44234246015548706, "learning_rate": 9.99970407609083e-05, "loss": 0.0442, "step": 5330 }, { "grad_norm": 0.47789064049720764, "learning_rate": 9.999685813593159e-05, "loss": 0.0483, "step": 5340 }, { "grad_norm": 0.3398672640323639, "learning_rate": 9.999667004337362e-05, "loss": 0.0449, "step": 5350 }, { "grad_norm": 0.4055880606174469, "learning_rate": 9.9996476483255e-05, "loss": 0.0398, "step": 5360 }, { "grad_norm": 0.4442795515060425, "learning_rate": 9.999627745559688e-05, "loss": 0.0422, "step": 5370 }, { "grad_norm": 0.42009562253952026, "learning_rate": 9.999607296042101e-05, "loss": 0.0414, "step": 5380 }, { "grad_norm": 0.4025697410106659, "learning_rate": 9.99958629977498e-05, "loss": 0.0421, "step": 5390 }, { "grad_norm": 0.3789048194885254, "learning_rate": 9.999564756760615e-05, "loss": 0.0392, "step": 5400 }, { "grad_norm": 0.4683997929096222, "learning_rate": 9.999542667001366e-05, "loss": 0.0424, "step": 5410 }, { "grad_norm": 0.441446989774704, "learning_rate": 9.999520030499647e-05, "loss": 0.042, "step": 5420 }, { "grad_norm": 0.4049890339374542, "learning_rate": 9.999496847257936e-05, "loss": 0.0433, "step": 5430 }, { "grad_norm": 0.47054633498191833, "learning_rate": 9.999473117278764e-05, "loss": 0.0433, "step": 5440 }, { "grad_norm": 0.45301109552383423, "learning_rate": 9.999448840564731e-05, "loss": 0.0436, "step": 5450 }, { "grad_norm": 0.4558650553226471, "learning_rate": 9.999424017118488e-05, "loss": 0.0417, "step": 5460 }, { "grad_norm": 0.43057873845100403, "learning_rate": 9.999398646942751e-05, "loss": 0.0416, "step": 5470 }, { "grad_norm": 0.32680800557136536, "learning_rate": 9.999372730040296e-05, "loss": 0.0426, "step": 5480 }, { "grad_norm": 0.38931429386138916, "learning_rate": 9.999346266413953e-05, "loss": 0.0421, "step": 5490 }, { "grad_norm": 0.3554435074329376, "learning_rate": 9.99931925606662e-05, "loss": 0.0385, "step": 5500 }, { "grad_norm": 0.43945953249931335, "learning_rate": 9.99929169900125e-05, "loss": 0.0446, "step": 5510 }, { "grad_norm": 0.44710662961006165, "learning_rate": 9.999263595220855e-05, "loss": 0.0458, "step": 5520 }, { "grad_norm": 0.33693447709083557, "learning_rate": 9.99923494472851e-05, "loss": 0.0435, "step": 5530 }, { "grad_norm": 0.4048866033554077, "learning_rate": 9.999205747527348e-05, "loss": 0.0407, "step": 5540 }, { "grad_norm": 0.44547393918037415, "learning_rate": 9.999176003620561e-05, "loss": 0.0425, "step": 5550 }, { "grad_norm": 0.4358760714530945, "learning_rate": 9.999145713011405e-05, "loss": 0.0433, "step": 5560 }, { "grad_norm": 0.439212441444397, "learning_rate": 9.999114875703186e-05, "loss": 0.0433, "step": 5570 }, { "grad_norm": 0.40211671590805054, "learning_rate": 9.999083491699281e-05, "loss": 0.0388, "step": 5580 }, { "grad_norm": 0.4133226275444031, "learning_rate": 9.999051561003123e-05, "loss": 0.0429, "step": 5590 }, { "grad_norm": 0.33257409930229187, "learning_rate": 9.999019083618202e-05, "loss": 0.0386, "step": 5600 }, { "grad_norm": 0.3818873465061188, "learning_rate": 9.99898605954807e-05, "loss": 0.0396, "step": 5610 }, { "grad_norm": 0.46465322375297546, "learning_rate": 9.998952488796338e-05, "loss": 0.0404, "step": 5620 }, { "grad_norm": 0.43858662247657776, "learning_rate": 9.998918371366676e-05, "loss": 0.0392, "step": 5630 }, { "grad_norm": 0.40395018458366394, "learning_rate": 9.99888370726282e-05, "loss": 0.0401, "step": 5640 }, { "grad_norm": 0.40830549597740173, "learning_rate": 9.998848496488556e-05, "loss": 0.0395, "step": 5650 }, { "grad_norm": 0.4195409417152405, "learning_rate": 9.998812739047736e-05, "loss": 0.0481, "step": 5660 }, { "grad_norm": 0.42938339710235596, "learning_rate": 9.99877643494427e-05, "loss": 0.0405, "step": 5670 }, { "grad_norm": 0.43323856592178345, "learning_rate": 9.998739584182128e-05, "loss": 0.043, "step": 5680 }, { "grad_norm": 0.33740895986557007, "learning_rate": 9.998702186765342e-05, "loss": 0.0378, "step": 5690 }, { "grad_norm": 0.3088549077510834, "learning_rate": 9.998664242698e-05, "loss": 0.0425, "step": 5700 }, { "grad_norm": 0.481258362531662, "learning_rate": 9.998625751984251e-05, "loss": 0.0392, "step": 5710 }, { "grad_norm": 0.3325827717781067, "learning_rate": 9.998586714628307e-05, "loss": 0.0407, "step": 5720 }, { "grad_norm": 0.31501758098602295, "learning_rate": 9.998547130634432e-05, "loss": 0.0407, "step": 5730 }, { "grad_norm": 0.4683568775653839, "learning_rate": 9.99850700000696e-05, "loss": 0.0435, "step": 5740 }, { "grad_norm": 0.3902255594730377, "learning_rate": 9.998466322750278e-05, "loss": 0.0418, "step": 5750 }, { "grad_norm": 0.372846782207489, "learning_rate": 9.998425098868834e-05, "loss": 0.0382, "step": 5760 }, { "grad_norm": 0.33819255232810974, "learning_rate": 9.998383328367136e-05, "loss": 0.0418, "step": 5770 }, { "grad_norm": 0.3943864405155182, "learning_rate": 9.99834101124975e-05, "loss": 0.0398, "step": 5780 }, { "grad_norm": 0.459538996219635, "learning_rate": 9.998298147521309e-05, "loss": 0.0433, "step": 5790 }, { "grad_norm": 0.3495081961154938, "learning_rate": 9.998254737186496e-05, "loss": 0.0417, "step": 5800 }, { "grad_norm": 0.3771171569824219, "learning_rate": 9.99821078025006e-05, "loss": 0.0388, "step": 5810 }, { "grad_norm": 0.4046260118484497, "learning_rate": 9.998166276716807e-05, "loss": 0.0423, "step": 5820 }, { "grad_norm": 0.39279067516326904, "learning_rate": 9.998121226591606e-05, "loss": 0.0427, "step": 5830 }, { "grad_norm": 0.39388391375541687, "learning_rate": 9.998075629879382e-05, "loss": 0.0419, "step": 5840 }, { "grad_norm": 0.3225713074207306, "learning_rate": 9.99802948658512e-05, "loss": 0.0429, "step": 5850 }, { "grad_norm": 0.3767167627811432, "learning_rate": 9.99798279671387e-05, "loss": 0.045, "step": 5860 }, { "grad_norm": 0.60139399766922, "learning_rate": 9.997935560270734e-05, "loss": 0.0468, "step": 5870 }, { "grad_norm": 0.4788617193698883, "learning_rate": 9.997887777260879e-05, "loss": 0.0449, "step": 5880 }, { "grad_norm": 0.4124008119106293, "learning_rate": 9.997839447689532e-05, "loss": 0.0402, "step": 5890 }, { "grad_norm": 0.531157374382019, "learning_rate": 9.997790571561978e-05, "loss": 0.0418, "step": 5900 }, { "grad_norm": 0.435441255569458, "learning_rate": 9.99774114888356e-05, "loss": 0.0438, "step": 5910 }, { "grad_norm": 0.3423534035682678, "learning_rate": 9.997691179659684e-05, "loss": 0.0388, "step": 5920 }, { "grad_norm": 0.5740451812744141, "learning_rate": 9.997640663895815e-05, "loss": 0.0369, "step": 5930 }, { "grad_norm": 0.41063669323921204, "learning_rate": 9.997589601597477e-05, "loss": 0.0424, "step": 5940 }, { "grad_norm": 0.42003268003463745, "learning_rate": 9.997537992770252e-05, "loss": 0.04, "step": 5950 }, { "grad_norm": 0.40149134397506714, "learning_rate": 9.997485837419788e-05, "loss": 0.0379, "step": 5960 }, { "grad_norm": 0.3464455008506775, "learning_rate": 9.997433135551786e-05, "loss": 0.0421, "step": 5970 }, { "grad_norm": 0.43339696526527405, "learning_rate": 9.997379887172009e-05, "loss": 0.0401, "step": 5980 }, { "grad_norm": 0.4286225140094757, "learning_rate": 9.997326092286281e-05, "loss": 0.0438, "step": 5990 }, { "grad_norm": 0.39775678515434265, "learning_rate": 9.997271750900486e-05, "loss": 0.0434, "step": 6000 }, { "grad_norm": 0.3500625491142273, "learning_rate": 9.997216863020565e-05, "loss": 0.0378, "step": 6010 }, { "grad_norm": 0.40559929609298706, "learning_rate": 9.99716142865252e-05, "loss": 0.0381, "step": 6020 }, { "grad_norm": 0.3403685986995697, "learning_rate": 9.997105447802415e-05, "loss": 0.0364, "step": 6030 }, { "grad_norm": 0.32061222195625305, "learning_rate": 9.997048920476373e-05, "loss": 0.0398, "step": 6040 }, { "grad_norm": 0.4058338403701782, "learning_rate": 9.996991846680572e-05, "loss": 0.0406, "step": 6050 }, { "grad_norm": 0.37529128789901733, "learning_rate": 9.996934226421257e-05, "loss": 0.0375, "step": 6060 }, { "grad_norm": 0.37936121225357056, "learning_rate": 9.996876059704726e-05, "loss": 0.0399, "step": 6070 }, { "grad_norm": 0.3423641622066498, "learning_rate": 9.996817346537343e-05, "loss": 0.0441, "step": 6080 }, { "grad_norm": 0.378364622592926, "learning_rate": 9.996758086925526e-05, "loss": 0.0359, "step": 6090 }, { "grad_norm": 0.35229554772377014, "learning_rate": 9.996698280875759e-05, "loss": 0.0429, "step": 6100 }, { "grad_norm": 0.4545978307723999, "learning_rate": 9.99663792839458e-05, "loss": 0.0387, "step": 6110 }, { "grad_norm": 0.34073251485824585, "learning_rate": 9.99657702948859e-05, "loss": 0.0467, "step": 6120 }, { "grad_norm": 0.36163485050201416, "learning_rate": 9.996515584164448e-05, "loss": 0.0399, "step": 6130 }, { "grad_norm": 0.5435682535171509, "learning_rate": 9.996453592428873e-05, "loss": 0.0401, "step": 6140 }, { "grad_norm": 0.478726327419281, "learning_rate": 9.996391054288646e-05, "loss": 0.0434, "step": 6150 }, { "grad_norm": 0.4724079370498657, "learning_rate": 9.996327969750605e-05, "loss": 0.0405, "step": 6160 }, { "grad_norm": 0.45997118949890137, "learning_rate": 9.996264338821649e-05, "loss": 0.0437, "step": 6170 }, { "grad_norm": 0.3638051748275757, "learning_rate": 9.996200161508735e-05, "loss": 0.0395, "step": 6180 }, { "grad_norm": 0.38347068428993225, "learning_rate": 9.996135437818885e-05, "loss": 0.037, "step": 6190 }, { "grad_norm": 0.426827073097229, "learning_rate": 9.996070167759175e-05, "loss": 0.0392, "step": 6200 }, { "grad_norm": 0.47191163897514343, "learning_rate": 9.996004351336743e-05, "loss": 0.0387, "step": 6210 }, { "grad_norm": 0.4606572985649109, "learning_rate": 9.995937988558785e-05, "loss": 0.0426, "step": 6220 }, { "grad_norm": 0.4656868577003479, "learning_rate": 9.995871079432561e-05, "loss": 0.0412, "step": 6230 }, { "grad_norm": 0.46210038661956787, "learning_rate": 9.995803623965389e-05, "loss": 0.0397, "step": 6240 }, { "grad_norm": 0.3348703682422638, "learning_rate": 9.995735622164641e-05, "loss": 0.0438, "step": 6250 }, { "grad_norm": 0.4392736554145813, "learning_rate": 9.995667074037758e-05, "loss": 0.0421, "step": 6260 }, { "grad_norm": 0.33882978558540344, "learning_rate": 9.995597979592232e-05, "loss": 0.0356, "step": 6270 }, { "grad_norm": 0.44773566722869873, "learning_rate": 9.995528338835625e-05, "loss": 0.0399, "step": 6280 }, { "grad_norm": 0.28475889563560486, "learning_rate": 9.995458151775547e-05, "loss": 0.0362, "step": 6290 }, { "grad_norm": 0.2619144320487976, "learning_rate": 9.995387418419677e-05, "loss": 0.0445, "step": 6300 }, { "grad_norm": 0.39204832911491394, "learning_rate": 9.99531613877575e-05, "loss": 0.0402, "step": 6310 }, { "grad_norm": 0.4141026437282562, "learning_rate": 9.995244312851559e-05, "loss": 0.0385, "step": 6320 }, { "grad_norm": 0.3730993866920471, "learning_rate": 9.995171940654961e-05, "loss": 0.0404, "step": 6330 }, { "grad_norm": 0.3500424921512604, "learning_rate": 9.995099022193871e-05, "loss": 0.0403, "step": 6340 }, { "grad_norm": 0.4876438081264496, "learning_rate": 9.995025557476261e-05, "loss": 0.0416, "step": 6350 }, { "grad_norm": 0.3441062271595001, "learning_rate": 9.994951546510165e-05, "loss": 0.0444, "step": 6360 }, { "grad_norm": 0.391104131937027, "learning_rate": 9.994876989303679e-05, "loss": 0.042, "step": 6370 }, { "grad_norm": 0.42519569396972656, "learning_rate": 9.994801885864955e-05, "loss": 0.0443, "step": 6380 }, { "grad_norm": 0.36418384313583374, "learning_rate": 9.994726236202205e-05, "loss": 0.0408, "step": 6390 }, { "grad_norm": 0.5083003640174866, "learning_rate": 9.994650040323704e-05, "loss": 0.0372, "step": 6400 }, { "grad_norm": 0.4411623775959015, "learning_rate": 9.994573298237784e-05, "loss": 0.0437, "step": 6410 }, { "grad_norm": 0.30335134267807007, "learning_rate": 9.994496009952837e-05, "loss": 0.0426, "step": 6420 }, { "grad_norm": 0.40352779626846313, "learning_rate": 9.994418175477316e-05, "loss": 0.0383, "step": 6430 }, { "grad_norm": 0.4274291396141052, "learning_rate": 9.994339794819733e-05, "loss": 0.0371, "step": 6440 }, { "grad_norm": 0.35014501214027405, "learning_rate": 9.994260867988658e-05, "loss": 0.0392, "step": 6450 }, { "grad_norm": 0.41186729073524475, "learning_rate": 9.994181394992723e-05, "loss": 0.0379, "step": 6460 }, { "grad_norm": 0.35636138916015625, "learning_rate": 9.994101375840618e-05, "loss": 0.0379, "step": 6470 }, { "grad_norm": 0.44967782497406006, "learning_rate": 9.994020810541098e-05, "loss": 0.0399, "step": 6480 }, { "grad_norm": 0.48709210753440857, "learning_rate": 9.99393969910297e-05, "loss": 0.0435, "step": 6490 }, { "grad_norm": 0.39032021164894104, "learning_rate": 9.993858041535104e-05, "loss": 0.0382, "step": 6500 }, { "grad_norm": 0.360129714012146, "learning_rate": 9.99377583784643e-05, "loss": 0.0382, "step": 6510 }, { "grad_norm": 0.43630000948905945, "learning_rate": 9.993693088045939e-05, "loss": 0.0379, "step": 6520 }, { "grad_norm": 0.3194243907928467, "learning_rate": 9.99360979214268e-05, "loss": 0.0372, "step": 6530 }, { "grad_norm": 0.4715670049190521, "learning_rate": 9.99352595014576e-05, "loss": 0.0386, "step": 6540 }, { "grad_norm": 0.3671007752418518, "learning_rate": 9.993441562064354e-05, "loss": 0.04, "step": 6550 }, { "grad_norm": 0.3149530291557312, "learning_rate": 9.993356627907685e-05, "loss": 0.0403, "step": 6560 }, { "grad_norm": 0.3117694556713104, "learning_rate": 9.99327114768504e-05, "loss": 0.0422, "step": 6570 }, { "grad_norm": 0.3923572897911072, "learning_rate": 9.99318512140577e-05, "loss": 0.0356, "step": 6580 }, { "grad_norm": 0.46033963561058044, "learning_rate": 9.993098549079284e-05, "loss": 0.0396, "step": 6590 }, { "grad_norm": 0.3365720808506012, "learning_rate": 9.993011430715047e-05, "loss": 0.0435, "step": 6600 }, { "grad_norm": 0.32677215337753296, "learning_rate": 9.992923766322586e-05, "loss": 0.0417, "step": 6610 }, { "grad_norm": 0.377174973487854, "learning_rate": 9.99283555591149e-05, "loss": 0.0336, "step": 6620 }, { "grad_norm": 0.4022127687931061, "learning_rate": 9.992746799491404e-05, "loss": 0.0436, "step": 6630 }, { "grad_norm": 0.33939996361732483, "learning_rate": 9.992657497072033e-05, "loss": 0.0411, "step": 6640 }, { "grad_norm": 0.32048681378364563, "learning_rate": 9.992567648663147e-05, "loss": 0.037, "step": 6650 }, { "grad_norm": 0.2725292444229126, "learning_rate": 9.992477254274568e-05, "loss": 0.0413, "step": 6660 }, { "grad_norm": 0.4074423313140869, "learning_rate": 9.992386313916183e-05, "loss": 0.0419, "step": 6670 }, { "grad_norm": 0.36262941360473633, "learning_rate": 9.992294827597934e-05, "loss": 0.0396, "step": 6680 }, { "grad_norm": 0.41331154108047485, "learning_rate": 9.992202795329831e-05, "loss": 0.0397, "step": 6690 }, { "grad_norm": 0.38172775506973267, "learning_rate": 9.992110217121936e-05, "loss": 0.0399, "step": 6700 }, { "grad_norm": 0.30377447605133057, "learning_rate": 9.992017092984372e-05, "loss": 0.0395, "step": 6710 }, { "grad_norm": 0.3743003308773041, "learning_rate": 9.991923422927326e-05, "loss": 0.039, "step": 6720 }, { "grad_norm": 0.44245147705078125, "learning_rate": 9.991829206961037e-05, "loss": 0.0365, "step": 6730 }, { "grad_norm": 0.43337076902389526, "learning_rate": 9.991734445095813e-05, "loss": 0.0396, "step": 6740 }, { "grad_norm": 0.3833915889263153, "learning_rate": 9.991639137342015e-05, "loss": 0.0366, "step": 6750 }, { "grad_norm": 0.2941930294036865, "learning_rate": 9.991543283710064e-05, "loss": 0.0363, "step": 6760 }, { "grad_norm": 0.3855396509170532, "learning_rate": 9.991446884210445e-05, "loss": 0.0354, "step": 6770 }, { "grad_norm": 0.31651532649993896, "learning_rate": 9.9913499388537e-05, "loss": 0.0358, "step": 6780 }, { "grad_norm": 0.3719916343688965, "learning_rate": 9.99125244765043e-05, "loss": 0.0372, "step": 6790 }, { "grad_norm": 0.4044415056705475, "learning_rate": 9.991154410611296e-05, "loss": 0.0368, "step": 6800 }, { "grad_norm": 0.37101346254348755, "learning_rate": 9.99105582774702e-05, "loss": 0.0366, "step": 6810 }, { "grad_norm": 0.35378119349479675, "learning_rate": 9.990956699068384e-05, "loss": 0.0396, "step": 6820 }, { "grad_norm": 0.3601917028427124, "learning_rate": 9.990857024586224e-05, "loss": 0.0364, "step": 6830 }, { "grad_norm": 0.3924209475517273, "learning_rate": 9.990756804311446e-05, "loss": 0.0381, "step": 6840 }, { "grad_norm": 0.37697118520736694, "learning_rate": 9.990656038255006e-05, "loss": 0.0397, "step": 6850 }, { "grad_norm": 0.3961713910102844, "learning_rate": 9.990554726427926e-05, "loss": 0.0419, "step": 6860 }, { "grad_norm": 0.3668981194496155, "learning_rate": 9.990452868841284e-05, "loss": 0.0354, "step": 6870 }, { "grad_norm": 0.3997190594673157, "learning_rate": 9.99035046550622e-05, "loss": 0.0413, "step": 6880 }, { "grad_norm": 0.3694162368774414, "learning_rate": 9.99024751643393e-05, "loss": 0.0349, "step": 6890 }, { "grad_norm": 0.40198996663093567, "learning_rate": 9.990144021635677e-05, "loss": 0.0355, "step": 6900 }, { "grad_norm": 0.49718379974365234, "learning_rate": 9.990039981122775e-05, "loss": 0.035, "step": 6910 }, { "grad_norm": 0.34196123480796814, "learning_rate": 9.989935394906602e-05, "loss": 0.0385, "step": 6920 }, { "grad_norm": 0.3929988145828247, "learning_rate": 9.989830262998598e-05, "loss": 0.0375, "step": 6930 }, { "grad_norm": 0.29127129912376404, "learning_rate": 9.989724585410259e-05, "loss": 0.0359, "step": 6940 }, { "grad_norm": 0.493431955575943, "learning_rate": 9.989618362153139e-05, "loss": 0.035, "step": 6950 }, { "grad_norm": 0.3368125855922699, "learning_rate": 9.989511593238859e-05, "loss": 0.0388, "step": 6960 }, { "grad_norm": 0.3266507685184479, "learning_rate": 9.98940427867909e-05, "loss": 0.0358, "step": 6970 }, { "grad_norm": 0.42519840598106384, "learning_rate": 9.989296418485573e-05, "loss": 0.0377, "step": 6980 }, { "grad_norm": 0.26322102546691895, "learning_rate": 9.989188012670101e-05, "loss": 0.0356, "step": 6990 }, { "grad_norm": 0.3158705234527588, "learning_rate": 9.989079061244528e-05, "loss": 0.0357, "step": 7000 }, { "grad_norm": 0.34771960973739624, "learning_rate": 9.988969564220769e-05, "loss": 0.0406, "step": 7010 }, { "grad_norm": 0.32467061281204224, "learning_rate": 9.988859521610801e-05, "loss": 0.0388, "step": 7020 }, { "grad_norm": 0.3916394114494324, "learning_rate": 9.988748933426656e-05, "loss": 0.0335, "step": 7030 }, { "grad_norm": 0.45270511507987976, "learning_rate": 9.988637799680428e-05, "loss": 0.0363, "step": 7040 }, { "grad_norm": 0.3786257803440094, "learning_rate": 9.98852612038427e-05, "loss": 0.0358, "step": 7050 }, { "grad_norm": 0.3271924555301666, "learning_rate": 9.988413895550397e-05, "loss": 0.0383, "step": 7060 }, { "grad_norm": 0.3049079179763794, "learning_rate": 9.98830112519108e-05, "loss": 0.0343, "step": 7070 }, { "grad_norm": 0.38160425424575806, "learning_rate": 9.98818780931865e-05, "loss": 0.0355, "step": 7080 }, { "grad_norm": 0.3022402822971344, "learning_rate": 9.988073947945502e-05, "loss": 0.0362, "step": 7090 }, { "grad_norm": 0.26094329357147217, "learning_rate": 9.987959541084087e-05, "loss": 0.0306, "step": 7100 }, { "grad_norm": 0.34336718916893005, "learning_rate": 9.987844588746915e-05, "loss": 0.0357, "step": 7110 }, { "grad_norm": 0.33350008726119995, "learning_rate": 9.987729090946558e-05, "loss": 0.0365, "step": 7120 }, { "grad_norm": 0.31124237179756165, "learning_rate": 9.987613047695647e-05, "loss": 0.0328, "step": 7130 }, { "grad_norm": 0.47525182366371155, "learning_rate": 9.987496459006871e-05, "loss": 0.0381, "step": 7140 }, { "grad_norm": 0.38305824995040894, "learning_rate": 9.987379324892982e-05, "loss": 0.0327, "step": 7150 }, { "grad_norm": 0.5045137405395508, "learning_rate": 9.987261645366788e-05, "loss": 0.0341, "step": 7160 }, { "grad_norm": 0.33429062366485596, "learning_rate": 9.987143420441158e-05, "loss": 0.0326, "step": 7170 }, { "grad_norm": 0.32856181263923645, "learning_rate": 9.987024650129022e-05, "loss": 0.0383, "step": 7180 }, { "grad_norm": 0.3527155816555023, "learning_rate": 9.986905334443368e-05, "loss": 0.0345, "step": 7190 }, { "grad_norm": 0.28530430793762207, "learning_rate": 9.986785473397245e-05, "loss": 0.0363, "step": 7200 }, { "grad_norm": 0.3587827980518341, "learning_rate": 9.98666506700376e-05, "loss": 0.0371, "step": 7210 }, { "grad_norm": 0.36360588669776917, "learning_rate": 9.986544115276081e-05, "loss": 0.0365, "step": 7220 }, { "grad_norm": 0.3008158504962921, "learning_rate": 9.986422618227433e-05, "loss": 0.0379, "step": 7230 }, { "grad_norm": 0.33833494782447815, "learning_rate": 9.986300575871106e-05, "loss": 0.0319, "step": 7240 }, { "grad_norm": 0.43487548828125, "learning_rate": 9.986177988220444e-05, "loss": 0.0336, "step": 7250 }, { "grad_norm": 0.26329508423805237, "learning_rate": 9.986054855288856e-05, "loss": 0.0345, "step": 7260 }, { "grad_norm": 0.4136829376220703, "learning_rate": 9.985931177089802e-05, "loss": 0.0347, "step": 7270 }, { "grad_norm": 0.31765422224998474, "learning_rate": 9.985806953636814e-05, "loss": 0.0311, "step": 7280 }, { "grad_norm": 0.3140242397785187, "learning_rate": 9.985682184943471e-05, "loss": 0.0367, "step": 7290 }, { "grad_norm": 0.36305975914001465, "learning_rate": 9.98555687102342e-05, "loss": 0.0348, "step": 7300 }, { "grad_norm": 0.3200128972530365, "learning_rate": 9.985431011890367e-05, "loss": 0.0314, "step": 7310 }, { "grad_norm": 0.3417735993862152, "learning_rate": 9.985304607558075e-05, "loss": 0.0338, "step": 7320 }, { "grad_norm": 0.3888949751853943, "learning_rate": 9.985177658040364e-05, "loss": 0.0382, "step": 7330 }, { "grad_norm": 0.3978913724422455, "learning_rate": 9.985050163351119e-05, "loss": 0.0366, "step": 7340 }, { "grad_norm": 0.36755380034446716, "learning_rate": 9.984922123504286e-05, "loss": 0.04, "step": 7350 }, { "grad_norm": 0.3289949297904968, "learning_rate": 9.984793538513862e-05, "loss": 0.0328, "step": 7360 }, { "grad_norm": 0.34456706047058105, "learning_rate": 9.984664408393912e-05, "loss": 0.0335, "step": 7370 }, { "grad_norm": 0.3439664840698242, "learning_rate": 9.984534733158556e-05, "loss": 0.0351, "step": 7380 }, { "grad_norm": 0.3847636580467224, "learning_rate": 9.984404512821977e-05, "loss": 0.0373, "step": 7390 }, { "grad_norm": 0.4000604450702667, "learning_rate": 9.984273747398411e-05, "loss": 0.0348, "step": 7400 }, { "grad_norm": 0.3546028435230255, "learning_rate": 9.984142436902165e-05, "loss": 0.0386, "step": 7410 }, { "grad_norm": 0.37977683544158936, "learning_rate": 9.984010581347596e-05, "loss": 0.0357, "step": 7420 }, { "grad_norm": 0.3237621486186981, "learning_rate": 9.983878180749121e-05, "loss": 0.0433, "step": 7430 }, { "grad_norm": 0.37641721963882446, "learning_rate": 9.983745235121222e-05, "loss": 0.0397, "step": 7440 }, { "grad_norm": 0.4008927643299103, "learning_rate": 9.983611744478438e-05, "loss": 0.0389, "step": 7450 }, { "grad_norm": 0.2718869149684906, "learning_rate": 9.983477708835365e-05, "loss": 0.0382, "step": 7460 }, { "grad_norm": 0.33694788813591003, "learning_rate": 9.983343128206664e-05, "loss": 0.0411, "step": 7470 }, { "grad_norm": 0.3830762505531311, "learning_rate": 9.983208002607049e-05, "loss": 0.0364, "step": 7480 }, { "grad_norm": 0.3180420398712158, "learning_rate": 9.9830723320513e-05, "loss": 0.0359, "step": 7490 }, { "grad_norm": 0.28142935037612915, "learning_rate": 9.982936116554254e-05, "loss": 0.0366, "step": 7500 }, { "grad_norm": 0.3129305839538574, "learning_rate": 9.982799356130803e-05, "loss": 0.0324, "step": 7510 }, { "grad_norm": 0.28153160214424133, "learning_rate": 9.982662050795908e-05, "loss": 0.0318, "step": 7520 }, { "grad_norm": 0.37088027596473694, "learning_rate": 9.982524200564583e-05, "loss": 0.036, "step": 7530 }, { "grad_norm": 0.3173096179962158, "learning_rate": 9.982385805451901e-05, "loss": 0.0356, "step": 7540 }, { "grad_norm": 0.3478301763534546, "learning_rate": 9.982246865472998e-05, "loss": 0.0353, "step": 7550 }, { "grad_norm": 0.3376436233520508, "learning_rate": 9.982107380643069e-05, "loss": 0.0364, "step": 7560 }, { "grad_norm": 0.37455132603645325, "learning_rate": 9.981967350977368e-05, "loss": 0.0312, "step": 7570 }, { "grad_norm": 0.28738152980804443, "learning_rate": 9.981826776491208e-05, "loss": 0.0361, "step": 7580 }, { "grad_norm": 0.2919321656227112, "learning_rate": 9.98168565719996e-05, "loss": 0.0301, "step": 7590 }, { "grad_norm": 0.30625462532043457, "learning_rate": 9.98154399311906e-05, "loss": 0.0319, "step": 7600 }, { "grad_norm": 0.39850911498069763, "learning_rate": 9.981401784263997e-05, "loss": 0.0337, "step": 7610 }, { "grad_norm": 0.3740253150463104, "learning_rate": 9.981259030650326e-05, "loss": 0.0374, "step": 7620 }, { "grad_norm": 0.3526107668876648, "learning_rate": 9.981115732293655e-05, "loss": 0.0331, "step": 7630 }, { "grad_norm": 0.3700868785381317, "learning_rate": 9.980971889209659e-05, "loss": 0.0419, "step": 7640 }, { "grad_norm": 0.32445764541625977, "learning_rate": 9.980827501414064e-05, "loss": 0.039, "step": 7650 }, { "grad_norm": 0.3552091121673584, "learning_rate": 9.980682568922663e-05, "loss": 0.0301, "step": 7660 }, { "grad_norm": 0.2666378915309906, "learning_rate": 9.980537091751304e-05, "loss": 0.0327, "step": 7670 }, { "grad_norm": 0.36999034881591797, "learning_rate": 9.980391069915897e-05, "loss": 0.0346, "step": 7680 }, { "grad_norm": 0.36037665605545044, "learning_rate": 9.98024450343241e-05, "loss": 0.0354, "step": 7690 }, { "grad_norm": 0.2999962568283081, "learning_rate": 9.980097392316872e-05, "loss": 0.0364, "step": 7700 }, { "grad_norm": 0.3751548230648041, "learning_rate": 9.97994973658537e-05, "loss": 0.0365, "step": 7710 }, { "grad_norm": 0.3970993459224701, "learning_rate": 9.979801536254054e-05, "loss": 0.0293, "step": 7720 }, { "grad_norm": 0.31749871373176575, "learning_rate": 9.979652791339127e-05, "loss": 0.0347, "step": 7730 }, { "grad_norm": 0.3478972613811493, "learning_rate": 9.97950350185686e-05, "loss": 0.0328, "step": 7740 }, { "grad_norm": 0.2966483235359192, "learning_rate": 9.979353667823574e-05, "loss": 0.0336, "step": 7750 }, { "grad_norm": 0.28110307455062866, "learning_rate": 9.979203289255658e-05, "loss": 0.0342, "step": 7760 }, { "grad_norm": 0.367398738861084, "learning_rate": 9.979052366169557e-05, "loss": 0.0328, "step": 7770 }, { "grad_norm": 0.33405688405036926, "learning_rate": 9.978900898581775e-05, "loss": 0.0352, "step": 7780 }, { "grad_norm": 0.3873628079891205, "learning_rate": 9.978748886508875e-05, "loss": 0.0314, "step": 7790 }, { "grad_norm": 0.462948739528656, "learning_rate": 9.978596329967484e-05, "loss": 0.0377, "step": 7800 }, { "grad_norm": 0.2758557200431824, "learning_rate": 9.978443228974284e-05, "loss": 0.0364, "step": 7810 }, { "grad_norm": 0.43551963567733765, "learning_rate": 9.978289583546015e-05, "loss": 0.0339, "step": 7820 }, { "grad_norm": 0.4961867034435272, "learning_rate": 9.978135393699484e-05, "loss": 0.0325, "step": 7830 }, { "grad_norm": 0.4545512795448303, "learning_rate": 9.977980659451548e-05, "loss": 0.0354, "step": 7840 }, { "grad_norm": 0.2672812044620514, "learning_rate": 9.977825380819135e-05, "loss": 0.0309, "step": 7850 }, { "grad_norm": 0.2879508137702942, "learning_rate": 9.97766955781922e-05, "loss": 0.0327, "step": 7860 }, { "grad_norm": 0.3311128616333008, "learning_rate": 9.977513190468848e-05, "loss": 0.0313, "step": 7870 }, { "grad_norm": 0.29748162627220154, "learning_rate": 9.977356278785116e-05, "loss": 0.0355, "step": 7880 }, { "grad_norm": 0.24684199690818787, "learning_rate": 9.977198822785184e-05, "loss": 0.0305, "step": 7890 }, { "grad_norm": 0.27950310707092285, "learning_rate": 9.977040822486273e-05, "loss": 0.0326, "step": 7900 }, { "grad_norm": 0.3411608338356018, "learning_rate": 9.97688227790566e-05, "loss": 0.0379, "step": 7910 }, { "grad_norm": 0.3641958236694336, "learning_rate": 9.976723189060684e-05, "loss": 0.0332, "step": 7920 }, { "grad_norm": 0.3471604287624359, "learning_rate": 9.976563555968742e-05, "loss": 0.0349, "step": 7930 }, { "grad_norm": 0.2584208846092224, "learning_rate": 9.976403378647292e-05, "loss": 0.0363, "step": 7940 }, { "grad_norm": 0.28630489110946655, "learning_rate": 9.97624265711385e-05, "loss": 0.0308, "step": 7950 }, { "grad_norm": 0.3058238923549652, "learning_rate": 9.976081391385993e-05, "loss": 0.0339, "step": 7960 }, { "grad_norm": 0.28713420033454895, "learning_rate": 9.975919581481356e-05, "loss": 0.0324, "step": 7970 }, { "grad_norm": 0.27407026290893555, "learning_rate": 9.975757227417634e-05, "loss": 0.0341, "step": 7980 }, { "grad_norm": 0.3867538869380951, "learning_rate": 9.975594329212586e-05, "loss": 0.033, "step": 7990 }, { "grad_norm": 0.3059749901294708, "learning_rate": 9.97543088688402e-05, "loss": 0.0303, "step": 8000 }, { "grad_norm": 0.4260646402835846, "learning_rate": 9.975266900449814e-05, "loss": 0.0342, "step": 8010 }, { "grad_norm": 0.2574023902416229, "learning_rate": 9.975102369927898e-05, "loss": 0.0364, "step": 8020 }, { "grad_norm": 0.3035416305065155, "learning_rate": 9.974937295336269e-05, "loss": 0.0312, "step": 8030 }, { "grad_norm": 0.302504301071167, "learning_rate": 9.974771676692975e-05, "loss": 0.0361, "step": 8040 }, { "grad_norm": 0.33358192443847656, "learning_rate": 9.974605514016131e-05, "loss": 0.0317, "step": 8050 }, { "grad_norm": 0.3499114215373993, "learning_rate": 9.974438807323907e-05, "loss": 0.0308, "step": 8060 }, { "grad_norm": 0.2801569700241089, "learning_rate": 9.974271556634535e-05, "loss": 0.0324, "step": 8070 }, { "grad_norm": 0.3852558135986328, "learning_rate": 9.974103761966302e-05, "loss": 0.0331, "step": 8080 }, { "grad_norm": 0.24588720500469208, "learning_rate": 9.973935423337563e-05, "loss": 0.0305, "step": 8090 }, { "grad_norm": 0.40208661556243896, "learning_rate": 9.973766540766722e-05, "loss": 0.0332, "step": 8100 }, { "grad_norm": 0.36214593052864075, "learning_rate": 9.97359711427225e-05, "loss": 0.0366, "step": 8110 }, { "grad_norm": 0.3282177150249481, "learning_rate": 9.973427143872677e-05, "loss": 0.0353, "step": 8120 }, { "grad_norm": 0.3221787214279175, "learning_rate": 9.973256629586589e-05, "loss": 0.0321, "step": 8130 }, { "grad_norm": 0.291110098361969, "learning_rate": 9.973085571432632e-05, "loss": 0.0352, "step": 8140 }, { "grad_norm": 0.31691083312034607, "learning_rate": 9.972913969429513e-05, "loss": 0.0337, "step": 8150 }, { "grad_norm": 0.3722432851791382, "learning_rate": 9.972741823596e-05, "loss": 0.0331, "step": 8160 }, { "grad_norm": 0.4376160204410553, "learning_rate": 9.972569133950917e-05, "loss": 0.0339, "step": 8170 }, { "grad_norm": 0.313141793012619, "learning_rate": 9.972395900513151e-05, "loss": 0.0341, "step": 8180 }, { "grad_norm": 0.33736491203308105, "learning_rate": 9.972222123301645e-05, "loss": 0.0335, "step": 8190 }, { "grad_norm": 0.29269540309906006, "learning_rate": 9.972047802335403e-05, "loss": 0.0301, "step": 8200 }, { "grad_norm": 0.29566678404808044, "learning_rate": 9.971872937633488e-05, "loss": 0.0332, "step": 8210 }, { "grad_norm": 0.32220539450645447, "learning_rate": 9.971697529215024e-05, "loss": 0.0308, "step": 8220 }, { "grad_norm": 0.38573694229125977, "learning_rate": 9.971521577099192e-05, "loss": 0.0309, "step": 8230 }, { "grad_norm": 0.25597402453422546, "learning_rate": 9.971345081305236e-05, "loss": 0.0292, "step": 8240 }, { "grad_norm": 0.3390074670314789, "learning_rate": 9.971168041852456e-05, "loss": 0.0306, "step": 8250 }, { "grad_norm": 0.3350459635257721, "learning_rate": 9.970990458760215e-05, "loss": 0.0333, "step": 8260 }, { "grad_norm": 0.32454001903533936, "learning_rate": 9.970812332047929e-05, "loss": 0.0352, "step": 8270 }, { "grad_norm": 0.3020656108856201, "learning_rate": 9.97063366173508e-05, "loss": 0.035, "step": 8280 }, { "grad_norm": 0.3617379069328308, "learning_rate": 9.970454447841207e-05, "loss": 0.0327, "step": 8290 }, { "grad_norm": 0.2750944197177887, "learning_rate": 9.970274690385909e-05, "loss": 0.029, "step": 8300 }, { "grad_norm": 0.33323410153388977, "learning_rate": 9.970094389388844e-05, "loss": 0.0345, "step": 8310 }, { "grad_norm": 0.2702880799770355, "learning_rate": 9.969913544869728e-05, "loss": 0.0321, "step": 8320 }, { "grad_norm": 0.3264695703983307, "learning_rate": 9.96973215684834e-05, "loss": 0.0305, "step": 8330 }, { "grad_norm": 0.3327026665210724, "learning_rate": 9.969550225344513e-05, "loss": 0.0316, "step": 8340 }, { "grad_norm": 0.3515046536922455, "learning_rate": 9.969367750378147e-05, "loss": 0.0308, "step": 8350 }, { "grad_norm": 0.30750247836112976, "learning_rate": 9.969184731969194e-05, "loss": 0.03, "step": 8360 }, { "grad_norm": 0.4311336576938629, "learning_rate": 9.96900117013767e-05, "loss": 0.0327, "step": 8370 }, { "grad_norm": 0.3493780791759491, "learning_rate": 9.96881706490365e-05, "loss": 0.0326, "step": 8380 }, { "grad_norm": 0.4028374254703522, "learning_rate": 9.968632416287265e-05, "loss": 0.0321, "step": 8390 }, { "grad_norm": 0.33931592106819153, "learning_rate": 9.96844722430871e-05, "loss": 0.0301, "step": 8400 }, { "grad_norm": 0.3174893856048584, "learning_rate": 9.968261488988235e-05, "loss": 0.0298, "step": 8410 }, { "grad_norm": 0.36120331287384033, "learning_rate": 9.968075210346155e-05, "loss": 0.0292, "step": 8420 }, { "grad_norm": 0.40763089060783386, "learning_rate": 9.967888388402839e-05, "loss": 0.0287, "step": 8430 }, { "grad_norm": 0.2809947431087494, "learning_rate": 9.967701023178717e-05, "loss": 0.0306, "step": 8440 }, { "grad_norm": 0.3263746201992035, "learning_rate": 9.967513114694282e-05, "loss": 0.0312, "step": 8450 }, { "grad_norm": 0.3364204466342926, "learning_rate": 9.967324662970079e-05, "loss": 0.0328, "step": 8460 }, { "grad_norm": 0.23519229888916016, "learning_rate": 9.96713566802672e-05, "loss": 0.033, "step": 8470 }, { "grad_norm": 0.3672114908695221, "learning_rate": 9.966946129884873e-05, "loss": 0.0301, "step": 8480 }, { "grad_norm": 0.3036722242832184, "learning_rate": 9.966756048565265e-05, "loss": 0.028, "step": 8490 }, { "grad_norm": 0.3382765054702759, "learning_rate": 9.966565424088681e-05, "loss": 0.0312, "step": 8500 }, { "grad_norm": 0.3519648313522339, "learning_rate": 9.96637425647597e-05, "loss": 0.0314, "step": 8510 }, { "grad_norm": 0.2947901487350464, "learning_rate": 9.966182545748038e-05, "loss": 0.0337, "step": 8520 }, { "grad_norm": 0.23833970725536346, "learning_rate": 9.96599029192585e-05, "loss": 0.0346, "step": 8530 }, { "grad_norm": 0.3534085750579834, "learning_rate": 9.965797495030428e-05, "loss": 0.0355, "step": 8540 }, { "grad_norm": 0.37775918841362, "learning_rate": 9.96560415508286e-05, "loss": 0.0326, "step": 8550 }, { "grad_norm": 0.40145859122276306, "learning_rate": 9.965410272104286e-05, "loss": 0.0341, "step": 8560 }, { "grad_norm": 0.42571374773979187, "learning_rate": 9.96521584611591e-05, "loss": 0.0283, "step": 8570 }, { "grad_norm": 0.33074501156806946, "learning_rate": 9.965020877138994e-05, "loss": 0.0327, "step": 8580 }, { "grad_norm": 0.33666807413101196, "learning_rate": 9.964825365194861e-05, "loss": 0.0311, "step": 8590 }, { "grad_norm": 0.25594690442085266, "learning_rate": 9.96462931030489e-05, "loss": 0.0301, "step": 8600 }, { "grad_norm": 0.27394917607307434, "learning_rate": 9.96443271249052e-05, "loss": 0.031, "step": 8610 }, { "grad_norm": 0.2996005415916443, "learning_rate": 9.964235571773255e-05, "loss": 0.0326, "step": 8620 }, { "grad_norm": 0.29888710379600525, "learning_rate": 9.96403788817465e-05, "loss": 0.0275, "step": 8630 }, { "grad_norm": 0.23718048632144928, "learning_rate": 9.963839661716325e-05, "loss": 0.028, "step": 8640 }, { "grad_norm": 0.20107188820838928, "learning_rate": 9.963640892419958e-05, "loss": 0.0296, "step": 8650 }, { "grad_norm": 0.21695196628570557, "learning_rate": 9.963441580307286e-05, "loss": 0.0321, "step": 8660 }, { "grad_norm": 0.2928410470485687, "learning_rate": 9.963241725400104e-05, "loss": 0.0324, "step": 8670 }, { "grad_norm": 0.31239569187164307, "learning_rate": 9.963041327720271e-05, "loss": 0.0324, "step": 8680 }, { "grad_norm": 0.33916884660720825, "learning_rate": 9.962840387289697e-05, "loss": 0.0335, "step": 8690 }, { "grad_norm": 0.3334940969944, "learning_rate": 9.962638904130363e-05, "loss": 0.03, "step": 8700 }, { "grad_norm": 0.2681860327720642, "learning_rate": 9.962436878264298e-05, "loss": 0.0287, "step": 8710 }, { "grad_norm": 0.35445165634155273, "learning_rate": 9.962234309713598e-05, "loss": 0.0263, "step": 8720 }, { "grad_norm": 0.3605536222457886, "learning_rate": 9.962031198500414e-05, "loss": 0.0328, "step": 8730 }, { "grad_norm": 0.26948586106300354, "learning_rate": 9.961827544646958e-05, "loss": 0.0289, "step": 8740 }, { "grad_norm": 0.2935613989830017, "learning_rate": 9.961623348175501e-05, "loss": 0.0335, "step": 8750 }, { "grad_norm": 0.27722516655921936, "learning_rate": 9.961418609108377e-05, "loss": 0.0291, "step": 8760 }, { "grad_norm": 0.2013476938009262, "learning_rate": 9.961213327467971e-05, "loss": 0.0255, "step": 8770 }, { "grad_norm": 0.263334184885025, "learning_rate": 9.961007503276736e-05, "loss": 0.0275, "step": 8780 }, { "grad_norm": 0.3383256196975708, "learning_rate": 9.960801136557179e-05, "loss": 0.027, "step": 8790 }, { "grad_norm": 0.3436921536922455, "learning_rate": 9.960594227331866e-05, "loss": 0.0335, "step": 8800 }, { "grad_norm": 0.3877090811729431, "learning_rate": 9.960386775623429e-05, "loss": 0.0322, "step": 8810 }, { "grad_norm": 0.28788241744041443, "learning_rate": 9.96017878145455e-05, "loss": 0.0332, "step": 8820 }, { "grad_norm": 0.3457981050014496, "learning_rate": 9.959970244847977e-05, "loss": 0.0354, "step": 8830 }, { "grad_norm": 0.35030800104141235, "learning_rate": 9.959761165826518e-05, "loss": 0.0318, "step": 8840 }, { "grad_norm": 0.2755729854106903, "learning_rate": 9.959551544413033e-05, "loss": 0.0368, "step": 8850 }, { "grad_norm": 0.2596593201160431, "learning_rate": 9.959341380630448e-05, "loss": 0.032, "step": 8860 }, { "grad_norm": 0.26276150345802307, "learning_rate": 9.959130674501746e-05, "loss": 0.0336, "step": 8870 }, { "grad_norm": 0.30674105882644653, "learning_rate": 9.958919426049968e-05, "loss": 0.0298, "step": 8880 }, { "grad_norm": 0.3518267571926117, "learning_rate": 9.958707635298219e-05, "loss": 0.0326, "step": 8890 }, { "grad_norm": 0.3697209358215332, "learning_rate": 9.958495302269657e-05, "loss": 0.0278, "step": 8900 }, { "grad_norm": 0.29225510358810425, "learning_rate": 9.958282426987503e-05, "loss": 0.0306, "step": 8910 }, { "grad_norm": 0.3523034453392029, "learning_rate": 9.95806900947504e-05, "loss": 0.0289, "step": 8920 }, { "grad_norm": 0.27972233295440674, "learning_rate": 9.957855049755604e-05, "loss": 0.0284, "step": 8930 }, { "grad_norm": 0.30516260862350464, "learning_rate": 9.957640547852593e-05, "loss": 0.0278, "step": 8940 }, { "grad_norm": 0.4170140027999878, "learning_rate": 9.957425503789466e-05, "loss": 0.0316, "step": 8950 }, { "grad_norm": 0.3201404809951782, "learning_rate": 9.957209917589738e-05, "loss": 0.0341, "step": 8960 }, { "grad_norm": 0.2869085371494293, "learning_rate": 9.956993789276987e-05, "loss": 0.0336, "step": 8970 }, { "grad_norm": 0.3325463831424713, "learning_rate": 9.956777118874847e-05, "loss": 0.0337, "step": 8980 }, { "grad_norm": 0.3580181896686554, "learning_rate": 9.956559906407016e-05, "loss": 0.0316, "step": 8990 }, { "grad_norm": 0.25695306062698364, "learning_rate": 9.956342151897245e-05, "loss": 0.035, "step": 9000 }, { "grad_norm": 0.28897130489349365, "learning_rate": 9.956123855369346e-05, "loss": 0.0332, "step": 9010 }, { "grad_norm": 0.34872496128082275, "learning_rate": 9.955905016847196e-05, "loss": 0.0263, "step": 9020 }, { "grad_norm": 0.26400429010391235, "learning_rate": 9.955685636354723e-05, "loss": 0.032, "step": 9030 }, { "grad_norm": 0.38816338777542114, "learning_rate": 9.95546571391592e-05, "loss": 0.0312, "step": 9040 }, { "grad_norm": 0.47426077723503113, "learning_rate": 9.955245249554837e-05, "loss": 0.0279, "step": 9050 }, { "grad_norm": 0.28619474172592163, "learning_rate": 9.955024243295582e-05, "loss": 0.0289, "step": 9060 }, { "grad_norm": 0.35920003056526184, "learning_rate": 9.954802695162328e-05, "loss": 0.0295, "step": 9070 }, { "grad_norm": 0.24488767981529236, "learning_rate": 9.954580605179302e-05, "loss": 0.0315, "step": 9080 }, { "grad_norm": 0.34547120332717896, "learning_rate": 9.954357973370788e-05, "loss": 0.0311, "step": 9090 }, { "grad_norm": 0.27109187841415405, "learning_rate": 9.954134799761135e-05, "loss": 0.0291, "step": 9100 }, { "grad_norm": 0.34335753321647644, "learning_rate": 9.953911084374748e-05, "loss": 0.0316, "step": 9110 }, { "grad_norm": 0.2949541211128235, "learning_rate": 9.953686827236093e-05, "loss": 0.0299, "step": 9120 }, { "grad_norm": 0.3339429795742035, "learning_rate": 9.953462028369695e-05, "loss": 0.0288, "step": 9130 }, { "grad_norm": 0.2625051736831665, "learning_rate": 9.953236687800136e-05, "loss": 0.03, "step": 9140 }, { "grad_norm": 0.2255660742521286, "learning_rate": 9.95301080555206e-05, "loss": 0.0265, "step": 9150 }, { "grad_norm": 0.2938251793384552, "learning_rate": 9.952784381650171e-05, "loss": 0.0291, "step": 9160 }, { "grad_norm": 0.3320755958557129, "learning_rate": 9.952557416119226e-05, "loss": 0.0302, "step": 9170 }, { "grad_norm": 0.3082626163959503, "learning_rate": 9.95232990898405e-05, "loss": 0.0271, "step": 9180 }, { "grad_norm": 0.33914345502853394, "learning_rate": 9.95210186026952e-05, "loss": 0.029, "step": 9190 }, { "grad_norm": 0.21286895871162415, "learning_rate": 9.951873270000576e-05, "loss": 0.0302, "step": 9200 }, { "grad_norm": 0.28832244873046875, "learning_rate": 9.951644138202216e-05, "loss": 0.03, "step": 9210 }, { "grad_norm": 0.3731837868690491, "learning_rate": 9.951414464899498e-05, "loss": 0.0316, "step": 9220 }, { "grad_norm": 0.3126441538333893, "learning_rate": 9.951184250117538e-05, "loss": 0.0274, "step": 9230 }, { "grad_norm": 0.2931117117404938, "learning_rate": 9.950953493881513e-05, "loss": 0.0313, "step": 9240 }, { "grad_norm": 0.3054898679256439, "learning_rate": 9.950722196216658e-05, "loss": 0.0264, "step": 9250 }, { "grad_norm": 0.2946455776691437, "learning_rate": 9.950490357148265e-05, "loss": 0.0283, "step": 9260 }, { "grad_norm": 0.377108097076416, "learning_rate": 9.950257976701692e-05, "loss": 0.0282, "step": 9270 }, { "grad_norm": 0.3991256356239319, "learning_rate": 9.950025054902348e-05, "loss": 0.0307, "step": 9280 }, { "grad_norm": 0.3567090332508087, "learning_rate": 9.949791591775706e-05, "loss": 0.0288, "step": 9290 }, { "grad_norm": 0.2704122066497803, "learning_rate": 9.949557587347298e-05, "loss": 0.0305, "step": 9300 }, { "grad_norm": 0.3434557020664215, "learning_rate": 9.949323041642713e-05, "loss": 0.026, "step": 9310 }, { "grad_norm": 0.33574220538139343, "learning_rate": 9.949087954687602e-05, "loss": 0.0275, "step": 9320 }, { "grad_norm": 0.2995290458202362, "learning_rate": 9.948852326507672e-05, "loss": 0.0271, "step": 9330 }, { "grad_norm": 0.2849792242050171, "learning_rate": 9.948616157128694e-05, "loss": 0.026, "step": 9340 }, { "grad_norm": 0.21644915640354156, "learning_rate": 9.948379446576493e-05, "loss": 0.0292, "step": 9350 }, { "grad_norm": 0.3060958683490753, "learning_rate": 9.948142194876952e-05, "loss": 0.027, "step": 9360 }, { "grad_norm": 0.31803515553474426, "learning_rate": 9.947904402056024e-05, "loss": 0.0261, "step": 9370 }, { "grad_norm": 0.3257351219654083, "learning_rate": 9.947666068139708e-05, "loss": 0.0293, "step": 9380 }, { "grad_norm": 0.328621506690979, "learning_rate": 9.947427193154071e-05, "loss": 0.0243, "step": 9390 }, { "grad_norm": 0.28858068585395813, "learning_rate": 9.947187777125233e-05, "loss": 0.0278, "step": 9400 }, { "grad_norm": 0.32951605319976807, "learning_rate": 9.946947820079377e-05, "loss": 0.0239, "step": 9410 }, { "grad_norm": 0.2995976507663727, "learning_rate": 9.946707322042747e-05, "loss": 0.0278, "step": 9420 }, { "grad_norm": 0.2935679256916046, "learning_rate": 9.94646628304164e-05, "loss": 0.0292, "step": 9430 }, { "grad_norm": 0.3527924716472626, "learning_rate": 9.946224703102418e-05, "loss": 0.027, "step": 9440 }, { "grad_norm": 0.2547304332256317, "learning_rate": 9.945982582251498e-05, "loss": 0.0293, "step": 9450 }, { "grad_norm": 0.3033081591129303, "learning_rate": 9.94573992051536e-05, "loss": 0.0257, "step": 9460 }, { "grad_norm": 0.26306161284446716, "learning_rate": 9.94549671792054e-05, "loss": 0.027, "step": 9470 }, { "grad_norm": 0.21488405764102936, "learning_rate": 9.945252974493635e-05, "loss": 0.0284, "step": 9480 }, { "grad_norm": 0.3070753216743469, "learning_rate": 9.9450086902613e-05, "loss": 0.0269, "step": 9490 }, { "grad_norm": 0.29755935072898865, "learning_rate": 9.944763865250248e-05, "loss": 0.0285, "step": 9500 }, { "grad_norm": 0.2902873754501343, "learning_rate": 9.944518499487254e-05, "loss": 0.0295, "step": 9510 }, { "grad_norm": 0.31989142298698425, "learning_rate": 9.944272592999151e-05, "loss": 0.0251, "step": 9520 }, { "grad_norm": 0.26676812767982483, "learning_rate": 9.94402614581283e-05, "loss": 0.0224, "step": 9530 }, { "grad_norm": 0.24307133257389069, "learning_rate": 9.943779157955244e-05, "loss": 0.0242, "step": 9540 }, { "grad_norm": 0.2640295922756195, "learning_rate": 9.943531629453403e-05, "loss": 0.0327, "step": 9550 }, { "grad_norm": 0.30696532130241394, "learning_rate": 9.943283560334375e-05, "loss": 0.0288, "step": 9560 }, { "grad_norm": 0.29403677582740784, "learning_rate": 9.943034950625288e-05, "loss": 0.0289, "step": 9570 }, { "grad_norm": 0.28468403220176697, "learning_rate": 9.942785800353332e-05, "loss": 0.0312, "step": 9580 }, { "grad_norm": 0.30309033393859863, "learning_rate": 9.942536109545751e-05, "loss": 0.031, "step": 9590 }, { "grad_norm": 0.3457804322242737, "learning_rate": 9.942285878229853e-05, "loss": 0.0251, "step": 9600 }, { "grad_norm": 0.33485618233680725, "learning_rate": 9.942035106433001e-05, "loss": 0.0287, "step": 9610 }, { "grad_norm": 0.3395652174949646, "learning_rate": 9.94178379418262e-05, "loss": 0.0244, "step": 9620 }, { "grad_norm": 0.3575223386287689, "learning_rate": 9.941531941506194e-05, "loss": 0.0263, "step": 9630 }, { "grad_norm": 0.21980895102024078, "learning_rate": 9.941279548431263e-05, "loss": 0.0307, "step": 9640 }, { "grad_norm": 0.2864498496055603, "learning_rate": 9.941026614985431e-05, "loss": 0.0299, "step": 9650 }, { "grad_norm": 0.3097398579120636, "learning_rate": 9.940773141196357e-05, "loss": 0.0284, "step": 9660 }, { "grad_norm": 0.4302430748939514, "learning_rate": 9.94051912709176e-05, "loss": 0.0252, "step": 9670 }, { "grad_norm": 0.25386926531791687, "learning_rate": 9.940264572699421e-05, "loss": 0.0257, "step": 9680 }, { "grad_norm": 0.28089526295661926, "learning_rate": 9.940009478047174e-05, "loss": 0.0298, "step": 9690 }, { "grad_norm": 0.35615453124046326, "learning_rate": 9.939753843162918e-05, "loss": 0.0296, "step": 9700 }, { "grad_norm": 0.29906824231147766, "learning_rate": 9.939497668074609e-05, "loss": 0.029, "step": 9710 }, { "grad_norm": 0.31690770387649536, "learning_rate": 9.93924095281026e-05, "loss": 0.0276, "step": 9720 }, { "grad_norm": 0.26124635338783264, "learning_rate": 9.938983697397948e-05, "loss": 0.0254, "step": 9730 }, { "grad_norm": 0.25643888115882874, "learning_rate": 9.938725901865805e-05, "loss": 0.0278, "step": 9740 }, { "grad_norm": 0.3927459120750427, "learning_rate": 9.93846756624202e-05, "loss": 0.0294, "step": 9750 }, { "grad_norm": 0.2964530885219574, "learning_rate": 9.938208690554849e-05, "loss": 0.027, "step": 9760 }, { "grad_norm": 0.24057352542877197, "learning_rate": 9.9379492748326e-05, "loss": 0.0248, "step": 9770 }, { "grad_norm": 0.23927243053913116, "learning_rate": 9.937689319103641e-05, "loss": 0.0235, "step": 9780 }, { "grad_norm": 0.25426119565963745, "learning_rate": 9.937428823396404e-05, "loss": 0.0255, "step": 9790 }, { "grad_norm": 0.3275435268878937, "learning_rate": 9.937167787739372e-05, "loss": 0.0261, "step": 9800 }, { "grad_norm": 0.41283363103866577, "learning_rate": 9.936906212161095e-05, "loss": 0.0356, "step": 9810 }, { "grad_norm": 0.35788559913635254, "learning_rate": 9.936644096690176e-05, "loss": 0.029, "step": 9820 }, { "grad_norm": 0.22190624475479126, "learning_rate": 9.936381441355282e-05, "loss": 0.0301, "step": 9830 }, { "grad_norm": 0.25758275389671326, "learning_rate": 9.936118246185136e-05, "loss": 0.0283, "step": 9840 }, { "grad_norm": 0.2470431923866272, "learning_rate": 9.935854511208518e-05, "loss": 0.0312, "step": 9850 }, { "grad_norm": 0.24931728839874268, "learning_rate": 9.935590236454272e-05, "loss": 0.0266, "step": 9860 }, { "grad_norm": 0.2761193811893463, "learning_rate": 9.935325421951298e-05, "loss": 0.0231, "step": 9870 }, { "grad_norm": 0.2806534767150879, "learning_rate": 9.935060067728557e-05, "loss": 0.0248, "step": 9880 }, { "grad_norm": 0.28225451707839966, "learning_rate": 9.934794173815067e-05, "loss": 0.026, "step": 9890 }, { "grad_norm": 0.27776414155960083, "learning_rate": 9.934527740239906e-05, "loss": 0.024, "step": 9900 }, { "grad_norm": 0.28179535269737244, "learning_rate": 9.934260767032209e-05, "loss": 0.0273, "step": 9910 }, { "grad_norm": 0.38073545694351196, "learning_rate": 9.933993254221172e-05, "loss": 0.0253, "step": 9920 }, { "grad_norm": 0.3258437216281891, "learning_rate": 9.933725201836053e-05, "loss": 0.0258, "step": 9930 }, { "grad_norm": 0.32897549867630005, "learning_rate": 9.933456609906162e-05, "loss": 0.0234, "step": 9940 }, { "grad_norm": 0.2640579044818878, "learning_rate": 9.933187478460875e-05, "loss": 0.0317, "step": 9950 }, { "grad_norm": 0.28458544611930847, "learning_rate": 9.93291780752962e-05, "loss": 0.0272, "step": 9960 }, { "grad_norm": 0.22889871895313263, "learning_rate": 9.932647597141893e-05, "loss": 0.0258, "step": 9970 }, { "grad_norm": 0.25231775641441345, "learning_rate": 9.932376847327239e-05, "loss": 0.0217, "step": 9980 }, { "grad_norm": 0.2928237020969391, "learning_rate": 9.932105558115268e-05, "loss": 0.0258, "step": 9990 }, { "grad_norm": 0.2952015697956085, "learning_rate": 9.931833729535651e-05, "loss": 0.0269, "step": 10000 }, { "grad_norm": 0.21283864974975586, "learning_rate": 9.931561361618111e-05, "loss": 0.0223, "step": 10010 }, { "grad_norm": 0.3013679087162018, "learning_rate": 9.931288454392435e-05, "loss": 0.0256, "step": 10020 }, { "grad_norm": 0.27382946014404297, "learning_rate": 9.931015007888467e-05, "loss": 0.0274, "step": 10030 }, { "grad_norm": 0.35699722170829773, "learning_rate": 9.930741022136112e-05, "loss": 0.0247, "step": 10040 }, { "grad_norm": 0.2667570114135742, "learning_rate": 9.930466497165333e-05, "loss": 0.0245, "step": 10050 }, { "grad_norm": 0.3091995120048523, "learning_rate": 9.93019143300615e-05, "loss": 0.0232, "step": 10060 }, { "grad_norm": 0.29420626163482666, "learning_rate": 9.929915829688644e-05, "loss": 0.0236, "step": 10070 }, { "grad_norm": 0.2574520707130432, "learning_rate": 9.929639687242955e-05, "loss": 0.0324, "step": 10080 }, { "grad_norm": 0.28785479068756104, "learning_rate": 9.929363005699281e-05, "loss": 0.0262, "step": 10090 }, { "grad_norm": 0.1985587626695633, "learning_rate": 9.92908578508788e-05, "loss": 0.0243, "step": 10100 }, { "grad_norm": 0.19719573855400085, "learning_rate": 9.928808025439069e-05, "loss": 0.0216, "step": 10110 }, { "grad_norm": 0.2850008010864258, "learning_rate": 9.928529726783223e-05, "loss": 0.0245, "step": 10120 }, { "grad_norm": 0.27583178877830505, "learning_rate": 9.928250889150774e-05, "loss": 0.0277, "step": 10130 }, { "grad_norm": 0.2608410120010376, "learning_rate": 9.92797151257222e-05, "loss": 0.0249, "step": 10140 }, { "grad_norm": 0.24976098537445068, "learning_rate": 9.927691597078108e-05, "loss": 0.0243, "step": 10150 }, { "grad_norm": 0.303570032119751, "learning_rate": 9.927411142699053e-05, "loss": 0.0263, "step": 10160 }, { "grad_norm": 0.35601675510406494, "learning_rate": 9.927130149465725e-05, "loss": 0.0277, "step": 10170 }, { "grad_norm": 0.3444821536540985, "learning_rate": 9.92684861740885e-05, "loss": 0.0283, "step": 10180 }, { "grad_norm": 0.3393697738647461, "learning_rate": 9.926566546559217e-05, "loss": 0.0273, "step": 10190 }, { "grad_norm": 0.2598492503166199, "learning_rate": 9.926283936947673e-05, "loss": 0.0274, "step": 10200 }, { "grad_norm": 0.3327743411064148, "learning_rate": 9.926000788605126e-05, "loss": 0.0311, "step": 10210 }, { "grad_norm": 0.3420042395591736, "learning_rate": 9.92571710156254e-05, "loss": 0.0317, "step": 10220 }, { "grad_norm": 0.23707368969917297, "learning_rate": 9.925432875850936e-05, "loss": 0.0272, "step": 10230 }, { "grad_norm": 0.2470540851354599, "learning_rate": 9.925148111501396e-05, "loss": 0.0231, "step": 10240 }, { "grad_norm": 0.3102260231971741, "learning_rate": 9.924862808545066e-05, "loss": 0.023, "step": 10250 }, { "grad_norm": 0.2965361773967743, "learning_rate": 9.924576967013141e-05, "loss": 0.0282, "step": 10260 }, { "grad_norm": 0.22562885284423828, "learning_rate": 9.924290586936887e-05, "loss": 0.0284, "step": 10270 }, { "grad_norm": 0.33088427782058716, "learning_rate": 9.924003668347614e-05, "loss": 0.0291, "step": 10280 }, { "grad_norm": 0.2820895314216614, "learning_rate": 9.923716211276704e-05, "loss": 0.0231, "step": 10290 }, { "grad_norm": 0.25154680013656616, "learning_rate": 9.923428215755594e-05, "loss": 0.0255, "step": 10300 }, { "grad_norm": 0.32103604078292847, "learning_rate": 9.923139681815775e-05, "loss": 0.0238, "step": 10310 }, { "grad_norm": 0.3798871636390686, "learning_rate": 9.922850609488801e-05, "loss": 0.0233, "step": 10320 }, { "grad_norm": 0.30821019411087036, "learning_rate": 9.922560998806287e-05, "loss": 0.0258, "step": 10330 }, { "grad_norm": 0.2732117176055908, "learning_rate": 9.922270849799905e-05, "loss": 0.0235, "step": 10340 }, { "grad_norm": 0.3023161292076111, "learning_rate": 9.92198016250138e-05, "loss": 0.0306, "step": 10350 }, { "grad_norm": 0.3830276429653168, "learning_rate": 9.921688936942506e-05, "loss": 0.0291, "step": 10360 }, { "grad_norm": 0.26624783873558044, "learning_rate": 9.921397173155129e-05, "loss": 0.025, "step": 10370 }, { "grad_norm": 0.26772618293762207, "learning_rate": 9.921104871171157e-05, "loss": 0.0241, "step": 10380 }, { "grad_norm": 0.2818860709667206, "learning_rate": 9.920812031022554e-05, "loss": 0.0244, "step": 10390 }, { "grad_norm": 0.28586944937705994, "learning_rate": 9.920518652741348e-05, "loss": 0.0248, "step": 10400 }, { "grad_norm": 0.40559640526771545, "learning_rate": 9.920224736359618e-05, "loss": 0.0242, "step": 10410 }, { "grad_norm": 0.2124626338481903, "learning_rate": 9.91993028190951e-05, "loss": 0.0216, "step": 10420 }, { "grad_norm": 0.2610033452510834, "learning_rate": 9.919635289423222e-05, "loss": 0.0257, "step": 10430 }, { "grad_norm": 0.2604069113731384, "learning_rate": 9.919339758933015e-05, "loss": 0.0245, "step": 10440 }, { "grad_norm": 0.22789981961250305, "learning_rate": 9.919043690471209e-05, "loss": 0.0221, "step": 10450 }, { "grad_norm": 0.20010752975940704, "learning_rate": 9.91874708407018e-05, "loss": 0.0237, "step": 10460 }, { "grad_norm": 0.2762945592403412, "learning_rate": 9.918449939762367e-05, "loss": 0.0252, "step": 10470 }, { "grad_norm": 0.29559651017189026, "learning_rate": 9.91815225758026e-05, "loss": 0.0252, "step": 10480 }, { "grad_norm": 0.23727981746196747, "learning_rate": 9.917854037556419e-05, "loss": 0.0238, "step": 10490 }, { "grad_norm": 0.2677132487297058, "learning_rate": 9.917555279723454e-05, "loss": 0.025, "step": 10500 }, { "grad_norm": 0.2760778069496155, "learning_rate": 9.917255984114036e-05, "loss": 0.0263, "step": 10510 }, { "grad_norm": 0.2983883023262024, "learning_rate": 9.916956150760896e-05, "loss": 0.027, "step": 10520 }, { "grad_norm": 0.29794764518737793, "learning_rate": 9.916655779696826e-05, "loss": 0.0255, "step": 10530 }, { "grad_norm": 0.2039802074432373, "learning_rate": 9.916354870954671e-05, "loss": 0.0252, "step": 10540 }, { "grad_norm": 0.34671056270599365, "learning_rate": 9.91605342456734e-05, "loss": 0.0285, "step": 10550 }, { "grad_norm": 0.3047863841056824, "learning_rate": 9.915751440567795e-05, "loss": 0.0234, "step": 10560 }, { "grad_norm": 0.2701346278190613, "learning_rate": 9.915448918989066e-05, "loss": 0.0251, "step": 10570 }, { "grad_norm": 0.22930243611335754, "learning_rate": 9.915145859864232e-05, "loss": 0.0213, "step": 10580 }, { "grad_norm": 0.33309486508369446, "learning_rate": 9.914842263226437e-05, "loss": 0.0277, "step": 10590 }, { "grad_norm": 0.295423299074173, "learning_rate": 9.914538129108882e-05, "loss": 0.0269, "step": 10600 }, { "grad_norm": 0.28827425837516785, "learning_rate": 9.914233457544825e-05, "loss": 0.0284, "step": 10610 }, { "grad_norm": 0.34188905358314514, "learning_rate": 9.913928248567586e-05, "loss": 0.0317, "step": 10620 }, { "grad_norm": 0.20793549716472626, "learning_rate": 9.913622502210542e-05, "loss": 0.0247, "step": 10630 }, { "grad_norm": 0.3311805725097656, "learning_rate": 9.913316218507128e-05, "loss": 0.0282, "step": 10640 }, { "grad_norm": 0.23764312267303467, "learning_rate": 9.91300939749084e-05, "loss": 0.0255, "step": 10650 }, { "grad_norm": 0.3185916244983673, "learning_rate": 9.91270203919523e-05, "loss": 0.0252, "step": 10660 }, { "grad_norm": 0.3001890778541565, "learning_rate": 9.912394143653912e-05, "loss": 0.0227, "step": 10670 }, { "grad_norm": 0.286185085773468, "learning_rate": 9.912085710900555e-05, "loss": 0.0205, "step": 10680 }, { "grad_norm": 0.35318347811698914, "learning_rate": 9.911776740968892e-05, "loss": 0.0228, "step": 10690 }, { "grad_norm": 0.228633314371109, "learning_rate": 9.911467233892709e-05, "loss": 0.0248, "step": 10700 }, { "grad_norm": 0.34315547347068787, "learning_rate": 9.911157189705853e-05, "loss": 0.0218, "step": 10710 }, { "grad_norm": 0.3366950750350952, "learning_rate": 9.910846608442229e-05, "loss": 0.0269, "step": 10720 }, { "grad_norm": 0.27488356828689575, "learning_rate": 9.910535490135805e-05, "loss": 0.0269, "step": 10730 }, { "grad_norm": 0.25907015800476074, "learning_rate": 9.910223834820603e-05, "loss": 0.0222, "step": 10740 }, { "grad_norm": 0.26961129903793335, "learning_rate": 9.909911642530703e-05, "loss": 0.0217, "step": 10750 }, { "grad_norm": 0.261339396238327, "learning_rate": 9.909598913300249e-05, "loss": 0.0238, "step": 10760 }, { "grad_norm": 0.2954864203929901, "learning_rate": 9.909285647163438e-05, "loss": 0.0251, "step": 10770 }, { "grad_norm": 0.2895660996437073, "learning_rate": 9.908971844154531e-05, "loss": 0.0225, "step": 10780 }, { "grad_norm": 0.3033091127872467, "learning_rate": 9.908657504307843e-05, "loss": 0.0286, "step": 10790 }, { "grad_norm": 0.33022329211235046, "learning_rate": 9.908342627657751e-05, "loss": 0.0254, "step": 10800 }, { "grad_norm": 0.29992759227752686, "learning_rate": 9.908027214238689e-05, "loss": 0.0208, "step": 10810 }, { "grad_norm": 0.3454894721508026, "learning_rate": 9.90771126408515e-05, "loss": 0.0278, "step": 10820 }, { "grad_norm": 0.22863928973674774, "learning_rate": 9.907394777231685e-05, "loss": 0.0277, "step": 10830 }, { "grad_norm": 0.3218694031238556, "learning_rate": 9.907077753712905e-05, "loss": 0.0247, "step": 10840 }, { "grad_norm": 0.3159656822681427, "learning_rate": 9.906760193563482e-05, "loss": 0.0218, "step": 10850 }, { "grad_norm": 0.26197150349617004, "learning_rate": 9.906442096818139e-05, "loss": 0.0277, "step": 10860 }, { "grad_norm": 0.2081240862607956, "learning_rate": 9.906123463511665e-05, "loss": 0.0256, "step": 10870 }, { "grad_norm": 0.2640310227870941, "learning_rate": 9.905804293678907e-05, "loss": 0.0233, "step": 10880 }, { "grad_norm": 0.31526416540145874, "learning_rate": 9.905484587354766e-05, "loss": 0.0269, "step": 10890 }, { "grad_norm": 0.25556978583335876, "learning_rate": 9.905164344574205e-05, "loss": 0.022, "step": 10900 }, { "grad_norm": 0.28371402621269226, "learning_rate": 9.904843565372248e-05, "loss": 0.0244, "step": 10910 }, { "grad_norm": 0.29057902097702026, "learning_rate": 9.904522249783972e-05, "loss": 0.0265, "step": 10920 }, { "grad_norm": 0.25361019372940063, "learning_rate": 9.904200397844517e-05, "loss": 0.0258, "step": 10930 }, { "grad_norm": 0.2010461390018463, "learning_rate": 9.903878009589078e-05, "loss": 0.0245, "step": 10940 }, { "grad_norm": 0.21480441093444824, "learning_rate": 9.903555085052915e-05, "loss": 0.0198, "step": 10950 }, { "grad_norm": 0.23259590566158295, "learning_rate": 9.903231624271338e-05, "loss": 0.022, "step": 10960 }, { "grad_norm": 0.23372700810432434, "learning_rate": 9.902907627279724e-05, "loss": 0.0217, "step": 10970 }, { "grad_norm": 0.23598091304302216, "learning_rate": 9.902583094113504e-05, "loss": 0.0244, "step": 10980 }, { "grad_norm": 0.27919620275497437, "learning_rate": 9.902258024808168e-05, "loss": 0.028, "step": 10990 }, { "grad_norm": 0.2452593892812729, "learning_rate": 9.901932419399264e-05, "loss": 0.0244, "step": 11000 }, { "grad_norm": 0.32388609647750854, "learning_rate": 9.9016062779224e-05, "loss": 0.0228, "step": 11010 }, { "grad_norm": 0.2612384855747223, "learning_rate": 9.901279600413242e-05, "loss": 0.0233, "step": 11020 }, { "grad_norm": 0.30981433391571045, "learning_rate": 9.900952386907518e-05, "loss": 0.0257, "step": 11030 }, { "grad_norm": 0.2621666491031647, "learning_rate": 9.90062463744101e-05, "loss": 0.0229, "step": 11040 }, { "grad_norm": 0.21797432005405426, "learning_rate": 9.900296352049558e-05, "loss": 0.0237, "step": 11050 }, { "grad_norm": 0.22369533777236938, "learning_rate": 9.899967530769065e-05, "loss": 0.0224, "step": 11060 }, { "grad_norm": 0.296770304441452, "learning_rate": 9.899638173635489e-05, "loss": 0.0238, "step": 11070 }, { "grad_norm": 0.2934802770614624, "learning_rate": 9.899308280684849e-05, "loss": 0.024, "step": 11080 }, { "grad_norm": 0.329357385635376, "learning_rate": 9.898977851953222e-05, "loss": 0.0227, "step": 11090 }, { "grad_norm": 0.24012060463428497, "learning_rate": 9.898646887476741e-05, "loss": 0.024, "step": 11100 }, { "grad_norm": 0.23919691145420074, "learning_rate": 9.898315387291603e-05, "loss": 0.0231, "step": 11110 }, { "grad_norm": 0.3419879376888275, "learning_rate": 9.89798335143406e-05, "loss": 0.0229, "step": 11120 }, { "grad_norm": 0.2539970576763153, "learning_rate": 9.897650779940419e-05, "loss": 0.0236, "step": 11130 }, { "grad_norm": 0.244145005941391, "learning_rate": 9.897317672847054e-05, "loss": 0.025, "step": 11140 }, { "grad_norm": 0.25768378376960754, "learning_rate": 9.89698403019039e-05, "loss": 0.0256, "step": 11150 }, { "grad_norm": 0.255776971578598, "learning_rate": 9.896649852006917e-05, "loss": 0.0194, "step": 11160 }, { "grad_norm": 0.2543421685695648, "learning_rate": 9.896315138333177e-05, "loss": 0.0204, "step": 11170 }, { "grad_norm": 0.31103289127349854, "learning_rate": 9.895979889205774e-05, "loss": 0.0246, "step": 11180 }, { "grad_norm": 0.308844655752182, "learning_rate": 9.895644104661372e-05, "loss": 0.0237, "step": 11190 }, { "grad_norm": 0.2519705295562744, "learning_rate": 9.895307784736691e-05, "loss": 0.0229, "step": 11200 }, { "grad_norm": 0.23869603872299194, "learning_rate": 9.894970929468512e-05, "loss": 0.0266, "step": 11210 }, { "grad_norm": 0.31376826763153076, "learning_rate": 9.89463353889367e-05, "loss": 0.0215, "step": 11220 }, { "grad_norm": 0.3101905584335327, "learning_rate": 9.894295613049065e-05, "loss": 0.0232, "step": 11230 }, { "grad_norm": 0.24736951291561127, "learning_rate": 9.893957151971649e-05, "loss": 0.0205, "step": 11240 }, { "grad_norm": 0.2330443263053894, "learning_rate": 9.893618155698436e-05, "loss": 0.0231, "step": 11250 }, { "grad_norm": 0.3940250277519226, "learning_rate": 9.8932786242665e-05, "loss": 0.0247, "step": 11260 }, { "grad_norm": 0.350640207529068, "learning_rate": 9.89293855771297e-05, "loss": 0.0294, "step": 11270 }, { "grad_norm": 0.22059747576713562, "learning_rate": 9.892597956075036e-05, "loss": 0.0223, "step": 11280 }, { "grad_norm": 0.3025301396846771, "learning_rate": 9.892256819389947e-05, "loss": 0.02, "step": 11290 }, { "grad_norm": 0.29218819737434387, "learning_rate": 9.891915147695006e-05, "loss": 0.0226, "step": 11300 }, { "grad_norm": 0.2694501280784607, "learning_rate": 9.891572941027577e-05, "loss": 0.0273, "step": 11310 }, { "grad_norm": 0.33248379826545715, "learning_rate": 9.89123019942509e-05, "loss": 0.0211, "step": 11320 }, { "grad_norm": 0.2471710741519928, "learning_rate": 9.89088692292502e-05, "loss": 0.0231, "step": 11330 }, { "grad_norm": 0.29077890515327454, "learning_rate": 9.89054311156491e-05, "loss": 0.0269, "step": 11340 }, { "grad_norm": 0.2409762591123581, "learning_rate": 9.890198765382357e-05, "loss": 0.0257, "step": 11350 }, { "grad_norm": 0.25097963213920593, "learning_rate": 9.889853884415021e-05, "loss": 0.0202, "step": 11360 }, { "grad_norm": 0.3252369165420532, "learning_rate": 9.889508468700614e-05, "loss": 0.0217, "step": 11370 }, { "grad_norm": 0.3216778337955475, "learning_rate": 9.889162518276915e-05, "loss": 0.0246, "step": 11380 }, { "grad_norm": 0.20798926055431366, "learning_rate": 9.888816033181752e-05, "loss": 0.0233, "step": 11390 }, { "grad_norm": 0.23768934607505798, "learning_rate": 9.888469013453018e-05, "loss": 0.0217, "step": 11400 }, { "grad_norm": 0.307638019323349, "learning_rate": 9.888121459128663e-05, "loss": 0.0251, "step": 11410 }, { "grad_norm": 0.2642775774002075, "learning_rate": 9.887773370246693e-05, "loss": 0.028, "step": 11420 }, { "grad_norm": 0.22991275787353516, "learning_rate": 9.887424746845177e-05, "loss": 0.026, "step": 11430 }, { "grad_norm": 0.3193686604499817, "learning_rate": 9.887075588962239e-05, "loss": 0.0249, "step": 11440 }, { "grad_norm": 0.42937299609184265, "learning_rate": 9.88672589663606e-05, "loss": 0.0222, "step": 11450 }, { "grad_norm": 0.26680299639701843, "learning_rate": 9.886375669904886e-05, "loss": 0.024, "step": 11460 }, { "grad_norm": 0.23135526478290558, "learning_rate": 9.886024908807014e-05, "loss": 0.0248, "step": 11470 }, { "grad_norm": 0.30259355902671814, "learning_rate": 9.885673613380806e-05, "loss": 0.0244, "step": 11480 }, { "grad_norm": 0.22876392304897308, "learning_rate": 9.885321783664676e-05, "loss": 0.0215, "step": 11490 }, { "grad_norm": 0.26357391476631165, "learning_rate": 9.884969419697101e-05, "loss": 0.0262, "step": 11500 }, { "grad_norm": 0.25656038522720337, "learning_rate": 9.884616521516614e-05, "loss": 0.0233, "step": 11510 }, { "grad_norm": 0.244365394115448, "learning_rate": 9.88426308916181e-05, "loss": 0.0249, "step": 11520 }, { "grad_norm": 0.3121228814125061, "learning_rate": 9.883909122671335e-05, "loss": 0.0249, "step": 11530 }, { "grad_norm": 0.3170926868915558, "learning_rate": 9.883554622083904e-05, "loss": 0.0263, "step": 11540 }, { "grad_norm": 0.2843755781650543, "learning_rate": 9.88319958743828e-05, "loss": 0.0201, "step": 11550 }, { "grad_norm": 0.20699062943458557, "learning_rate": 9.882844018773291e-05, "loss": 0.0255, "step": 11560 }, { "grad_norm": 0.330938458442688, "learning_rate": 9.882487916127823e-05, "loss": 0.0248, "step": 11570 }, { "grad_norm": 0.21523968875408173, "learning_rate": 9.882131279540815e-05, "loss": 0.0206, "step": 11580 }, { "grad_norm": 0.21866245567798615, "learning_rate": 9.881774109051271e-05, "loss": 0.0208, "step": 11590 }, { "grad_norm": 0.20247900485992432, "learning_rate": 9.881416404698252e-05, "loss": 0.0189, "step": 11600 }, { "grad_norm": 0.27960237860679626, "learning_rate": 9.881058166520873e-05, "loss": 0.0222, "step": 11610 }, { "grad_norm": 0.2910289764404297, "learning_rate": 9.880699394558311e-05, "loss": 0.0261, "step": 11620 }, { "grad_norm": 0.24055489897727966, "learning_rate": 9.880340088849801e-05, "loss": 0.0212, "step": 11630 }, { "grad_norm": 0.28091558814048767, "learning_rate": 9.879980249434637e-05, "loss": 0.0226, "step": 11640 }, { "grad_norm": 0.3232271671295166, "learning_rate": 9.879619876352168e-05, "loss": 0.0234, "step": 11650 }, { "grad_norm": 0.2346329689025879, "learning_rate": 9.879258969641809e-05, "loss": 0.0223, "step": 11660 }, { "grad_norm": 0.24545595049858093, "learning_rate": 9.878897529343023e-05, "loss": 0.0195, "step": 11670 }, { "grad_norm": 0.22734272480010986, "learning_rate": 9.878535555495338e-05, "loss": 0.0197, "step": 11680 }, { "grad_norm": 0.3510306179523468, "learning_rate": 9.87817304813834e-05, "loss": 0.0261, "step": 11690 }, { "grad_norm": 0.3501336872577667, "learning_rate": 9.877810007311671e-05, "loss": 0.0249, "step": 11700 }, { "grad_norm": 0.26687169075012207, "learning_rate": 9.877446433055035e-05, "loss": 0.0231, "step": 11710 }, { "grad_norm": 0.22978761792182922, "learning_rate": 9.877082325408191e-05, "loss": 0.0255, "step": 11720 }, { "grad_norm": 0.3118702471256256, "learning_rate": 9.876717684410954e-05, "loss": 0.0189, "step": 11730 }, { "grad_norm": 0.1848238706588745, "learning_rate": 9.876352510103204e-05, "loss": 0.0183, "step": 11740 }, { "grad_norm": 0.24512244760990143, "learning_rate": 9.875986802524875e-05, "loss": 0.0218, "step": 11750 }, { "grad_norm": 0.25814494490623474, "learning_rate": 9.87562056171596e-05, "loss": 0.0237, "step": 11760 }, { "grad_norm": 0.2834542393684387, "learning_rate": 9.875253787716511e-05, "loss": 0.022, "step": 11770 }, { "grad_norm": 0.2805335223674774, "learning_rate": 9.874886480566637e-05, "loss": 0.0252, "step": 11780 }, { "grad_norm": 0.23261035978794098, "learning_rate": 9.874518640306507e-05, "loss": 0.0201, "step": 11790 }, { "grad_norm": 0.32043975591659546, "learning_rate": 9.874150266976347e-05, "loss": 0.0267, "step": 11800 }, { "grad_norm": 0.25006526708602905, "learning_rate": 9.873781360616443e-05, "loss": 0.0215, "step": 11810 }, { "grad_norm": 0.26879894733428955, "learning_rate": 9.873411921267137e-05, "loss": 0.0255, "step": 11820 }, { "grad_norm": 0.27718019485473633, "learning_rate": 9.873041948968829e-05, "loss": 0.0248, "step": 11830 }, { "grad_norm": 0.26912203431129456, "learning_rate": 9.872671443761981e-05, "loss": 0.0211, "step": 11840 }, { "grad_norm": 0.17525094747543335, "learning_rate": 9.872300405687109e-05, "loss": 0.0242, "step": 11850 }, { "grad_norm": 0.23014827072620392, "learning_rate": 9.871928834784792e-05, "loss": 0.0252, "step": 11860 }, { "grad_norm": 0.25287312269210815, "learning_rate": 9.871556731095661e-05, "loss": 0.0255, "step": 11870 }, { "grad_norm": 0.2844031751155853, "learning_rate": 9.871184094660411e-05, "loss": 0.0224, "step": 11880 }, { "grad_norm": 0.20610839128494263, "learning_rate": 9.870810925519791e-05, "loss": 0.0219, "step": 11890 }, { "grad_norm": 0.23625729978084564, "learning_rate": 9.870437223714612e-05, "loss": 0.0297, "step": 11900 }, { "grad_norm": 0.27197960019111633, "learning_rate": 9.87006298928574e-05, "loss": 0.0215, "step": 11910 }, { "grad_norm": 0.2782038748264313, "learning_rate": 9.869688222274103e-05, "loss": 0.0225, "step": 11920 }, { "grad_norm": 0.2502128481864929, "learning_rate": 9.869312922720681e-05, "loss": 0.0216, "step": 11930 }, { "grad_norm": 0.3059728741645813, "learning_rate": 9.868937090666521e-05, "loss": 0.0195, "step": 11940 }, { "grad_norm": 0.22201763093471527, "learning_rate": 9.86856072615272e-05, "loss": 0.0257, "step": 11950 }, { "grad_norm": 0.20225651562213898, "learning_rate": 9.868183829220438e-05, "loss": 0.0227, "step": 11960 }, { "grad_norm": 0.30711817741394043, "learning_rate": 9.867806399910893e-05, "loss": 0.0227, "step": 11970 }, { "grad_norm": 0.27177780866622925, "learning_rate": 9.867428438265356e-05, "loss": 0.0261, "step": 11980 }, { "grad_norm": 0.28120747208595276, "learning_rate": 9.867049944325165e-05, "loss": 0.0237, "step": 11990 }, { "grad_norm": 0.32355010509490967, "learning_rate": 9.86667091813171e-05, "loss": 0.0235, "step": 12000 }, { "grad_norm": 0.21877162158489227, "learning_rate": 9.866291359726438e-05, "loss": 0.025, "step": 12010 }, { "grad_norm": 0.27454638481140137, "learning_rate": 9.865911269150861e-05, "loss": 0.0233, "step": 12020 }, { "grad_norm": 0.2871120870113373, "learning_rate": 9.865530646446544e-05, "loss": 0.0243, "step": 12030 }, { "grad_norm": 0.28550055623054504, "learning_rate": 9.86514949165511e-05, "loss": 0.0252, "step": 12040 }, { "grad_norm": 0.35289475321769714, "learning_rate": 9.864767804818243e-05, "loss": 0.0213, "step": 12050 }, { "grad_norm": 0.20835968852043152, "learning_rate": 9.86438558597768e-05, "loss": 0.0224, "step": 12060 }, { "grad_norm": 0.2738994359970093, "learning_rate": 9.864002835175225e-05, "loss": 0.0221, "step": 12070 }, { "grad_norm": 0.22752241790294647, "learning_rate": 9.863619552452734e-05, "loss": 0.0219, "step": 12080 }, { "grad_norm": 0.25358912348747253, "learning_rate": 9.863235737852119e-05, "loss": 0.0198, "step": 12090 }, { "grad_norm": 0.2789964973926544, "learning_rate": 9.862851391415356e-05, "loss": 0.0231, "step": 12100 }, { "grad_norm": 0.21768678724765778, "learning_rate": 9.862466513184477e-05, "loss": 0.0219, "step": 12110 }, { "grad_norm": 0.23688369989395142, "learning_rate": 9.86208110320157e-05, "loss": 0.0258, "step": 12120 }, { "grad_norm": 0.2449105829000473, "learning_rate": 9.861695161508784e-05, "loss": 0.021, "step": 12130 }, { "grad_norm": 0.2201952487230301, "learning_rate": 9.861308688148324e-05, "loss": 0.0186, "step": 12140 }, { "grad_norm": 0.2729302644729614, "learning_rate": 9.860921683162455e-05, "loss": 0.0209, "step": 12150 }, { "grad_norm": 0.26216697692871094, "learning_rate": 9.860534146593499e-05, "loss": 0.0234, "step": 12160 }, { "grad_norm": 0.268252432346344, "learning_rate": 9.860146078483836e-05, "loss": 0.0241, "step": 12170 }, { "grad_norm": 0.26440170407295227, "learning_rate": 9.859757478875905e-05, "loss": 0.0209, "step": 12180 }, { "grad_norm": 0.18826064467430115, "learning_rate": 9.859368347812204e-05, "loss": 0.02, "step": 12190 }, { "grad_norm": 0.29943522810935974, "learning_rate": 9.858978685335285e-05, "loss": 0.0259, "step": 12200 }, { "grad_norm": 0.23681668937206268, "learning_rate": 9.858588491487763e-05, "loss": 0.0204, "step": 12210 }, { "grad_norm": 0.3067672550678253, "learning_rate": 9.858197766312308e-05, "loss": 0.0211, "step": 12220 }, { "grad_norm": 0.2609763741493225, "learning_rate": 9.857806509851649e-05, "loss": 0.0211, "step": 12230 }, { "grad_norm": 0.3405957818031311, "learning_rate": 9.857414722148574e-05, "loss": 0.0207, "step": 12240 }, { "grad_norm": 0.2804301381111145, "learning_rate": 9.857022403245928e-05, "loss": 0.0233, "step": 12250 }, { "grad_norm": 0.27672919631004333, "learning_rate": 9.856629553186615e-05, "loss": 0.022, "step": 12260 }, { "grad_norm": 0.22047314047813416, "learning_rate": 9.856236172013595e-05, "loss": 0.0254, "step": 12270 }, { "grad_norm": 0.30933666229248047, "learning_rate": 9.85584225976989e-05, "loss": 0.0225, "step": 12280 }, { "grad_norm": 0.312703013420105, "learning_rate": 9.855447816498575e-05, "loss": 0.0219, "step": 12290 }, { "grad_norm": 0.2647002339363098, "learning_rate": 9.855052842242787e-05, "loss": 0.023, "step": 12300 }, { "grad_norm": 0.33635249733924866, "learning_rate": 9.85465733704572e-05, "loss": 0.0232, "step": 12310 }, { "grad_norm": 0.3065815269947052, "learning_rate": 9.854261300950624e-05, "loss": 0.019, "step": 12320 }, { "grad_norm": 0.23920400440692902, "learning_rate": 9.853864734000813e-05, "loss": 0.0218, "step": 12330 }, { "grad_norm": 0.29539206624031067, "learning_rate": 9.85346763623965e-05, "loss": 0.0249, "step": 12340 }, { "grad_norm": 0.15778590738773346, "learning_rate": 9.853070007710564e-05, "loss": 0.0216, "step": 12350 }, { "grad_norm": 0.2642677426338196, "learning_rate": 9.85267184845704e-05, "loss": 0.0267, "step": 12360 }, { "grad_norm": 0.2731860280036926, "learning_rate": 9.852273158522616e-05, "loss": 0.0233, "step": 12370 }, { "grad_norm": 0.1984437257051468, "learning_rate": 9.851873937950896e-05, "loss": 0.0221, "step": 12380 }, { "grad_norm": 0.18918856978416443, "learning_rate": 9.851474186785537e-05, "loss": 0.0204, "step": 12390 }, { "grad_norm": 0.23903943598270416, "learning_rate": 9.851073905070254e-05, "loss": 0.0246, "step": 12400 }, { "grad_norm": 0.1990455985069275, "learning_rate": 9.850673092848824e-05, "loss": 0.0214, "step": 12410 }, { "grad_norm": 0.24615438282489777, "learning_rate": 9.850271750165077e-05, "loss": 0.0207, "step": 12420 }, { "grad_norm": 0.19024641811847687, "learning_rate": 9.849869877062902e-05, "loss": 0.018, "step": 12430 }, { "grad_norm": 0.28118711709976196, "learning_rate": 9.849467473586252e-05, "loss": 0.0198, "step": 12440 }, { "grad_norm": 0.2695680856704712, "learning_rate": 9.849064539779127e-05, "loss": 0.0287, "step": 12450 }, { "grad_norm": 0.25126031041145325, "learning_rate": 9.848661075685594e-05, "loss": 0.0203, "step": 12460 }, { "grad_norm": 0.28631702065467834, "learning_rate": 9.848257081349778e-05, "loss": 0.0253, "step": 12470 }, { "grad_norm": 0.20070147514343262, "learning_rate": 9.847852556815856e-05, "loss": 0.0191, "step": 12480 }, { "grad_norm": 0.22661885619163513, "learning_rate": 9.847447502128067e-05, "loss": 0.0184, "step": 12490 }, { "grad_norm": 0.2804342806339264, "learning_rate": 9.847041917330708e-05, "loss": 0.0195, "step": 12500 }, { "grad_norm": 0.18012970685958862, "learning_rate": 9.846635802468132e-05, "loss": 0.0192, "step": 12510 }, { "grad_norm": 0.23947007954120636, "learning_rate": 9.84622915758475e-05, "loss": 0.0206, "step": 12520 }, { "grad_norm": 0.2311568409204483, "learning_rate": 9.845821982725034e-05, "loss": 0.0228, "step": 12530 }, { "grad_norm": 0.2722040116786957, "learning_rate": 9.845414277933514e-05, "loss": 0.021, "step": 12540 }, { "grad_norm": 0.28177428245544434, "learning_rate": 9.845006043254771e-05, "loss": 0.0219, "step": 12550 }, { "grad_norm": 0.29597392678260803, "learning_rate": 9.844597278733451e-05, "loss": 0.0196, "step": 12560 }, { "grad_norm": 0.23198792338371277, "learning_rate": 9.844187984414259e-05, "loss": 0.0191, "step": 12570 }, { "grad_norm": 0.2716493308544159, "learning_rate": 9.84377816034195e-05, "loss": 0.019, "step": 12580 }, { "grad_norm": 0.26693448424339294, "learning_rate": 9.843367806561345e-05, "loss": 0.0202, "step": 12590 }, { "grad_norm": 0.23816223442554474, "learning_rate": 9.842956923117317e-05, "loss": 0.0222, "step": 12600 }, { "grad_norm": 0.2675732374191284, "learning_rate": 9.842545510054802e-05, "loss": 0.0206, "step": 12610 }, { "grad_norm": 0.23231814801692963, "learning_rate": 9.842133567418792e-05, "loss": 0.0233, "step": 12620 }, { "grad_norm": 0.2167866975069046, "learning_rate": 9.841721095254333e-05, "loss": 0.0197, "step": 12630 }, { "grad_norm": 0.24713562428951263, "learning_rate": 9.841308093606537e-05, "loss": 0.0202, "step": 12640 }, { "grad_norm": 0.3359953463077545, "learning_rate": 9.840894562520565e-05, "loss": 0.0214, "step": 12650 }, { "grad_norm": 0.2917778789997101, "learning_rate": 9.840480502041642e-05, "loss": 0.0179, "step": 12660 }, { "grad_norm": 0.3016885817050934, "learning_rate": 9.840065912215049e-05, "loss": 0.0188, "step": 12670 }, { "grad_norm": 0.24210116267204285, "learning_rate": 9.839650793086124e-05, "loss": 0.0217, "step": 12680 }, { "grad_norm": 0.25592732429504395, "learning_rate": 9.839235144700265e-05, "loss": 0.0214, "step": 12690 }, { "grad_norm": 0.18980273604393005, "learning_rate": 9.838818967102926e-05, "loss": 0.0193, "step": 12700 }, { "grad_norm": 0.2610267102718353, "learning_rate": 9.83840226033962e-05, "loss": 0.0185, "step": 12710 }, { "grad_norm": 0.2708145081996918, "learning_rate": 9.837985024455918e-05, "loss": 0.0214, "step": 12720 }, { "grad_norm": 0.21397873759269714, "learning_rate": 9.837567259497447e-05, "loss": 0.0199, "step": 12730 }, { "grad_norm": 0.286615252494812, "learning_rate": 9.837148965509894e-05, "loss": 0.0211, "step": 12740 }, { "grad_norm": 0.22650088369846344, "learning_rate": 9.836730142539001e-05, "loss": 0.0207, "step": 12750 }, { "grad_norm": 0.24372655153274536, "learning_rate": 9.836310790630574e-05, "loss": 0.0205, "step": 12760 }, { "grad_norm": 0.24980786442756653, "learning_rate": 9.83589090983047e-05, "loss": 0.0217, "step": 12770 }, { "grad_norm": 0.30023011565208435, "learning_rate": 9.835470500184605e-05, "loss": 0.0244, "step": 12780 }, { "grad_norm": 0.2653786838054657, "learning_rate": 9.835049561738957e-05, "loss": 0.0193, "step": 12790 }, { "grad_norm": 0.21681749820709229, "learning_rate": 9.834628094539558e-05, "loss": 0.0201, "step": 12800 }, { "grad_norm": 0.24572238326072693, "learning_rate": 9.834206098632499e-05, "loss": 0.0183, "step": 12810 }, { "grad_norm": 0.3031686842441559, "learning_rate": 9.833783574063931e-05, "loss": 0.0222, "step": 12820 }, { "grad_norm": 0.26436275243759155, "learning_rate": 9.833360520880058e-05, "loss": 0.0218, "step": 12830 }, { "grad_norm": 0.1955804079771042, "learning_rate": 9.832936939127144e-05, "loss": 0.0214, "step": 12840 }, { "grad_norm": 0.25059112906455994, "learning_rate": 9.832512828851515e-05, "loss": 0.0206, "step": 12850 }, { "grad_norm": 0.3285377621650696, "learning_rate": 9.832088190099546e-05, "loss": 0.0227, "step": 12860 }, { "grad_norm": 0.2352757602930069, "learning_rate": 9.831663022917679e-05, "loss": 0.0201, "step": 12870 }, { "grad_norm": 0.1720447838306427, "learning_rate": 9.831237327352407e-05, "loss": 0.02, "step": 12880 }, { "grad_norm": 0.20262828469276428, "learning_rate": 9.830811103450286e-05, "loss": 0.0212, "step": 12890 }, { "grad_norm": 0.2769779860973358, "learning_rate": 9.830384351257924e-05, "loss": 0.0209, "step": 12900 }, { "grad_norm": 0.19927597045898438, "learning_rate": 9.829957070821993e-05, "loss": 0.0194, "step": 12910 }, { "grad_norm": 0.24212227761745453, "learning_rate": 9.829529262189218e-05, "loss": 0.0208, "step": 12920 }, { "grad_norm": 0.23476473987102509, "learning_rate": 9.829100925406385e-05, "loss": 0.021, "step": 12930 }, { "grad_norm": 0.2888120412826538, "learning_rate": 9.828672060520333e-05, "loss": 0.0227, "step": 12940 }, { "grad_norm": 0.23532040417194366, "learning_rate": 9.828242667577966e-05, "loss": 0.0187, "step": 12950 }, { "grad_norm": 0.23376965522766113, "learning_rate": 9.82781274662624e-05, "loss": 0.0218, "step": 12960 }, { "grad_norm": 0.2679961323738098, "learning_rate": 9.82738229771217e-05, "loss": 0.0189, "step": 12970 }, { "grad_norm": 0.25345826148986816, "learning_rate": 9.826951320882829e-05, "loss": 0.0225, "step": 12980 }, { "grad_norm": 0.3481816351413727, "learning_rate": 9.826519816185351e-05, "loss": 0.029, "step": 12990 }, { "grad_norm": 0.2643929719924927, "learning_rate": 9.826087783666921e-05, "loss": 0.0193, "step": 13000 }, { "grad_norm": 0.2053724229335785, "learning_rate": 9.825655223374787e-05, "loss": 0.0208, "step": 13010 }, { "grad_norm": 0.26535147428512573, "learning_rate": 9.825222135356253e-05, "loss": 0.0204, "step": 13020 }, { "grad_norm": 0.2728377878665924, "learning_rate": 9.82478851965868e-05, "loss": 0.0185, "step": 13030 }, { "grad_norm": 0.262076199054718, "learning_rate": 9.82435437632949e-05, "loss": 0.0245, "step": 13040 }, { "grad_norm": 0.25271686911582947, "learning_rate": 9.823919705416158e-05, "loss": 0.0235, "step": 13050 }, { "grad_norm": 0.28325945138931274, "learning_rate": 9.82348450696622e-05, "loss": 0.0186, "step": 13060 }, { "grad_norm": 0.2694768011569977, "learning_rate": 9.823048781027268e-05, "loss": 0.0188, "step": 13070 }, { "grad_norm": 0.3054232597351074, "learning_rate": 9.822612527646953e-05, "loss": 0.0187, "step": 13080 }, { "grad_norm": 0.3511613607406616, "learning_rate": 9.822175746872984e-05, "loss": 0.0203, "step": 13090 }, { "grad_norm": 0.30798935890197754, "learning_rate": 9.821738438753123e-05, "loss": 0.0175, "step": 13100 }, { "grad_norm": 0.2893470227718353, "learning_rate": 9.821300603335196e-05, "loss": 0.0226, "step": 13110 }, { "grad_norm": 0.21271806955337524, "learning_rate": 9.820862240667085e-05, "loss": 0.0187, "step": 13120 }, { "grad_norm": 0.19545002281665802, "learning_rate": 9.820423350796726e-05, "loss": 0.0215, "step": 13130 }, { "grad_norm": 0.17865002155303955, "learning_rate": 9.819983933772118e-05, "loss": 0.0219, "step": 13140 }, { "grad_norm": 0.27591779828071594, "learning_rate": 9.819543989641314e-05, "loss": 0.0235, "step": 13150 }, { "grad_norm": 0.25769495964050293, "learning_rate": 9.819103518452423e-05, "loss": 0.0209, "step": 13160 }, { "grad_norm": 0.22148357331752777, "learning_rate": 9.818662520253618e-05, "loss": 0.0241, "step": 13170 }, { "grad_norm": 0.20417667925357819, "learning_rate": 9.818220995093126e-05, "loss": 0.0204, "step": 13180 }, { "grad_norm": 0.2124776542186737, "learning_rate": 9.817778943019228e-05, "loss": 0.0215, "step": 13190 }, { "grad_norm": 0.2973294258117676, "learning_rate": 9.81733636408027e-05, "loss": 0.0207, "step": 13200 }, { "grad_norm": 0.20329798758029938, "learning_rate": 9.816893258324649e-05, "loss": 0.0196, "step": 13210 }, { "grad_norm": 0.19214850664138794, "learning_rate": 9.816449625800823e-05, "loss": 0.0236, "step": 13220 }, { "grad_norm": 0.22116144001483917, "learning_rate": 9.816005466557308e-05, "loss": 0.0242, "step": 13230 }, { "grad_norm": 0.31948819756507874, "learning_rate": 9.815560780642674e-05, "loss": 0.0265, "step": 13240 }, { "grad_norm": 0.24899984896183014, "learning_rate": 9.815115568105555e-05, "loss": 0.0211, "step": 13250 }, { "grad_norm": 0.3081030249595642, "learning_rate": 9.814669828994638e-05, "loss": 0.0213, "step": 13260 }, { "grad_norm": 0.2350628525018692, "learning_rate": 9.814223563358665e-05, "loss": 0.0188, "step": 13270 }, { "grad_norm": 0.288864403963089, "learning_rate": 9.813776771246443e-05, "loss": 0.021, "step": 13280 }, { "grad_norm": 0.2553951144218445, "learning_rate": 9.813329452706829e-05, "loss": 0.0165, "step": 13290 }, { "grad_norm": 0.21150854229927063, "learning_rate": 9.812881607788744e-05, "loss": 0.0217, "step": 13300 }, { "grad_norm": 0.17057934403419495, "learning_rate": 9.812433236541163e-05, "loss": 0.0199, "step": 13310 }, { "grad_norm": 0.23496250808238983, "learning_rate": 9.811984339013116e-05, "loss": 0.0196, "step": 13320 }, { "grad_norm": 0.23484045267105103, "learning_rate": 9.811534915253698e-05, "loss": 0.0195, "step": 13330 }, { "grad_norm": 0.2709386944770813, "learning_rate": 9.811084965312056e-05, "loss": 0.0197, "step": 13340 }, { "grad_norm": 0.24185693264007568, "learning_rate": 9.810634489237396e-05, "loss": 0.0199, "step": 13350 }, { "grad_norm": 0.20582552254199982, "learning_rate": 9.81018348707898e-05, "loss": 0.0208, "step": 13360 }, { "grad_norm": 0.2505855858325958, "learning_rate": 9.809731958886131e-05, "loss": 0.0208, "step": 13370 }, { "grad_norm": 0.21128837764263153, "learning_rate": 9.809279904708224e-05, "loss": 0.0196, "step": 13380 }, { "grad_norm": 0.2605488896369934, "learning_rate": 9.808827324594699e-05, "loss": 0.018, "step": 13390 }, { "grad_norm": 0.20113353431224823, "learning_rate": 9.808374218595046e-05, "loss": 0.02, "step": 13400 }, { "grad_norm": 0.2221902459859848, "learning_rate": 9.80792058675882e-05, "loss": 0.0167, "step": 13410 }, { "grad_norm": 0.2543381154537201, "learning_rate": 9.807466429135627e-05, "loss": 0.0179, "step": 13420 }, { "grad_norm": 0.2155766487121582, "learning_rate": 9.807011745775132e-05, "loss": 0.021, "step": 13430 }, { "grad_norm": 0.26871925592422485, "learning_rate": 9.806556536727061e-05, "loss": 0.0242, "step": 13440 }, { "grad_norm": 0.23470467329025269, "learning_rate": 9.806100802041193e-05, "loss": 0.0197, "step": 13450 }, { "grad_norm": 0.2873106598854065, "learning_rate": 9.805644541767368e-05, "loss": 0.0217, "step": 13460 }, { "grad_norm": 0.24040386080741882, "learning_rate": 9.805187755955478e-05, "loss": 0.0252, "step": 13470 }, { "grad_norm": 0.26461976766586304, "learning_rate": 9.804730444655483e-05, "loss": 0.0209, "step": 13480 }, { "grad_norm": 0.2518714666366577, "learning_rate": 9.804272607917388e-05, "loss": 0.0178, "step": 13490 }, { "grad_norm": 0.16133886575698853, "learning_rate": 9.803814245791265e-05, "loss": 0.0182, "step": 13500 }, { "grad_norm": 0.19913044571876526, "learning_rate": 9.803355358327239e-05, "loss": 0.0218, "step": 13510 }, { "grad_norm": 0.2705351412296295, "learning_rate": 9.802895945575492e-05, "loss": 0.0181, "step": 13520 }, { "grad_norm": 0.25114721059799194, "learning_rate": 9.802436007586266e-05, "loss": 0.018, "step": 13530 }, { "grad_norm": 0.3484266698360443, "learning_rate": 9.801975544409858e-05, "loss": 0.0185, "step": 13540 }, { "grad_norm": 0.2464236170053482, "learning_rate": 9.801514556096625e-05, "loss": 0.0189, "step": 13550 }, { "grad_norm": 0.27290645241737366, "learning_rate": 9.801053042696977e-05, "loss": 0.025, "step": 13560 }, { "grad_norm": 0.2305658757686615, "learning_rate": 9.800591004261388e-05, "loss": 0.0196, "step": 13570 }, { "grad_norm": 0.2679183781147003, "learning_rate": 9.800128440840385e-05, "loss": 0.0183, "step": 13580 }, { "grad_norm": 0.20849177241325378, "learning_rate": 9.799665352484552e-05, "loss": 0.0189, "step": 13590 }, { "grad_norm": 0.17487715184688568, "learning_rate": 9.799201739244532e-05, "loss": 0.0161, "step": 13600 }, { "grad_norm": 0.20202922821044922, "learning_rate": 9.798737601171025e-05, "loss": 0.0198, "step": 13610 }, { "grad_norm": 0.24780048429965973, "learning_rate": 9.79827293831479e-05, "loss": 0.0194, "step": 13620 }, { "grad_norm": 0.17304813861846924, "learning_rate": 9.797807750726638e-05, "loss": 0.0192, "step": 13630 }, { "grad_norm": 0.25308284163475037, "learning_rate": 9.797342038457446e-05, "loss": 0.0192, "step": 13640 }, { "grad_norm": 0.2832927107810974, "learning_rate": 9.796875801558141e-05, "loss": 0.0198, "step": 13650 }, { "grad_norm": 0.27717408537864685, "learning_rate": 9.79640904007971e-05, "loss": 0.018, "step": 13660 }, { "grad_norm": 0.2513796091079712, "learning_rate": 9.795941754073199e-05, "loss": 0.02, "step": 13670 }, { "grad_norm": 0.21330180764198303, "learning_rate": 9.795473943589705e-05, "loss": 0.0192, "step": 13680 }, { "grad_norm": 0.23491515219211578, "learning_rate": 9.795005608680394e-05, "loss": 0.0198, "step": 13690 }, { "grad_norm": 0.21500493586063385, "learning_rate": 9.794536749396477e-05, "loss": 0.0196, "step": 13700 }, { "grad_norm": 0.21020355820655823, "learning_rate": 9.79406736578923e-05, "loss": 0.0212, "step": 13710 }, { "grad_norm": 0.19653312861919403, "learning_rate": 9.793597457909984e-05, "loss": 0.0222, "step": 13720 }, { "grad_norm": 0.22475318610668182, "learning_rate": 9.793127025810127e-05, "loss": 0.0192, "step": 13730 }, { "grad_norm": 0.25407981872558594, "learning_rate": 9.792656069541104e-05, "loss": 0.0226, "step": 13740 }, { "grad_norm": 0.25471770763397217, "learning_rate": 9.79218458915442e-05, "loss": 0.018, "step": 13750 }, { "grad_norm": 0.24554140865802765, "learning_rate": 9.791712584701634e-05, "loss": 0.0179, "step": 13760 }, { "grad_norm": 0.20786407589912415, "learning_rate": 9.791240056234364e-05, "loss": 0.0215, "step": 13770 }, { "grad_norm": 0.20885415375232697, "learning_rate": 9.790767003804283e-05, "loss": 0.0223, "step": 13780 }, { "grad_norm": 0.2042102962732315, "learning_rate": 9.790293427463126e-05, "loss": 0.019, "step": 13790 }, { "grad_norm": 0.24414102733135223, "learning_rate": 9.789819327262684e-05, "loss": 0.0189, "step": 13800 }, { "grad_norm": 0.2774653732776642, "learning_rate": 9.7893447032548e-05, "loss": 0.0192, "step": 13810 }, { "grad_norm": 0.28308388590812683, "learning_rate": 9.78886955549138e-05, "loss": 0.0213, "step": 13820 }, { "grad_norm": 0.27352091670036316, "learning_rate": 9.788393884024387e-05, "loss": 0.0191, "step": 13830 }, { "grad_norm": 0.2246207743883133, "learning_rate": 9.787917688905836e-05, "loss": 0.017, "step": 13840 }, { "grad_norm": 0.2341911792755127, "learning_rate": 9.787440970187807e-05, "loss": 0.0195, "step": 13850 }, { "grad_norm": 0.2803623676300049, "learning_rate": 9.786963727922429e-05, "loss": 0.0203, "step": 13860 }, { "grad_norm": 0.21551601588726044, "learning_rate": 9.786485962161897e-05, "loss": 0.0195, "step": 13870 }, { "grad_norm": 0.16166403889656067, "learning_rate": 9.786007672958455e-05, "loss": 0.0208, "step": 13880 }, { "grad_norm": 0.22905834019184113, "learning_rate": 9.78552886036441e-05, "loss": 0.0188, "step": 13890 }, { "grad_norm": 0.2263936549425125, "learning_rate": 9.785049524432124e-05, "loss": 0.0218, "step": 13900 }, { "grad_norm": 0.31739047169685364, "learning_rate": 9.784569665214016e-05, "loss": 0.0168, "step": 13910 }, { "grad_norm": 0.28167271614074707, "learning_rate": 9.784089282762563e-05, "loss": 0.0176, "step": 13920 }, { "grad_norm": 0.23850417137145996, "learning_rate": 9.7836083771303e-05, "loss": 0.0181, "step": 13930 }, { "grad_norm": 0.2687508463859558, "learning_rate": 9.783126948369817e-05, "loss": 0.0193, "step": 13940 }, { "grad_norm": 0.2427755892276764, "learning_rate": 9.78264499653376e-05, "loss": 0.0219, "step": 13950 }, { "grad_norm": 0.34835106134414673, "learning_rate": 9.782162521674838e-05, "loss": 0.0197, "step": 13960 }, { "grad_norm": 0.24484239518642426, "learning_rate": 9.781679523845812e-05, "loss": 0.0213, "step": 13970 }, { "grad_norm": 0.2206116020679474, "learning_rate": 9.781196003099502e-05, "loss": 0.0212, "step": 13980 }, { "grad_norm": 0.25434938073158264, "learning_rate": 9.780711959488786e-05, "loss": 0.0233, "step": 13990 }, { "grad_norm": 0.2518887221813202, "learning_rate": 9.780227393066599e-05, "loss": 0.0251, "step": 14000 }, { "grad_norm": 0.25208938121795654, "learning_rate": 9.77974230388593e-05, "loss": 0.0166, "step": 14010 }, { "grad_norm": 0.2679097652435303, "learning_rate": 9.779256691999829e-05, "loss": 0.0203, "step": 14020 }, { "grad_norm": 0.24796968698501587, "learning_rate": 9.778770557461403e-05, "loss": 0.0193, "step": 14030 }, { "grad_norm": 0.26484954357147217, "learning_rate": 9.778283900323812e-05, "loss": 0.0215, "step": 14040 }, { "grad_norm": 0.2785414755344391, "learning_rate": 9.777796720640277e-05, "loss": 0.0204, "step": 14050 }, { "grad_norm": 0.23328721523284912, "learning_rate": 9.777309018464078e-05, "loss": 0.0197, "step": 14060 }, { "grad_norm": 0.2259647697210312, "learning_rate": 9.776820793848547e-05, "loss": 0.0205, "step": 14070 }, { "grad_norm": 0.2838175296783447, "learning_rate": 9.776332046847075e-05, "loss": 0.0194, "step": 14080 }, { "grad_norm": 0.1981555074453354, "learning_rate": 9.775842777513111e-05, "loss": 0.02, "step": 14090 }, { "grad_norm": 0.2236587256193161, "learning_rate": 9.775352985900163e-05, "loss": 0.0189, "step": 14100 }, { "grad_norm": 0.1797345131635666, "learning_rate": 9.774862672061791e-05, "loss": 0.0167, "step": 14110 }, { "grad_norm": 0.20993562042713165, "learning_rate": 9.774371836051616e-05, "loss": 0.0203, "step": 14120 }, { "grad_norm": 0.2714230716228485, "learning_rate": 9.773880477923315e-05, "loss": 0.0177, "step": 14130 }, { "grad_norm": 0.28030312061309814, "learning_rate": 9.773388597730623e-05, "loss": 0.0167, "step": 14140 }, { "grad_norm": 0.31121963262557983, "learning_rate": 9.77289619552733e-05, "loss": 0.0203, "step": 14150 }, { "grad_norm": 0.24385212361812592, "learning_rate": 9.772403271367285e-05, "loss": 0.02, "step": 14160 }, { "grad_norm": 0.2939782738685608, "learning_rate": 9.771909825304396e-05, "loss": 0.0198, "step": 14170 }, { "grad_norm": 0.279916375875473, "learning_rate": 9.771415857392619e-05, "loss": 0.0234, "step": 14180 }, { "grad_norm": 0.21019144356250763, "learning_rate": 9.770921367685978e-05, "loss": 0.0197, "step": 14190 }, { "grad_norm": 0.2015158087015152, "learning_rate": 9.770426356238551e-05, "loss": 0.0218, "step": 14200 }, { "grad_norm": 0.23465414345264435, "learning_rate": 9.769930823104469e-05, "loss": 0.0151, "step": 14210 }, { "grad_norm": 0.20254525542259216, "learning_rate": 9.769434768337926e-05, "loss": 0.018, "step": 14220 }, { "grad_norm": 0.25063636898994446, "learning_rate": 9.768938191993164e-05, "loss": 0.0183, "step": 14230 }, { "grad_norm": 0.23957915604114532, "learning_rate": 9.768441094124494e-05, "loss": 0.0217, "step": 14240 }, { "grad_norm": 0.3151913285255432, "learning_rate": 9.767943474786275e-05, "loss": 0.0212, "step": 14250 }, { "grad_norm": 0.20440125465393066, "learning_rate": 9.767445334032923e-05, "loss": 0.0191, "step": 14260 }, { "grad_norm": 0.23182551562786102, "learning_rate": 9.766946671918919e-05, "loss": 0.0211, "step": 14270 }, { "grad_norm": 0.24628357589244843, "learning_rate": 9.766447488498796e-05, "loss": 0.0196, "step": 14280 }, { "grad_norm": 0.39356938004493713, "learning_rate": 9.765947783827139e-05, "loss": 0.0201, "step": 14290 }, { "grad_norm": 0.25205352902412415, "learning_rate": 9.765447557958599e-05, "loss": 0.023, "step": 14300 }, { "grad_norm": 0.22405850887298584, "learning_rate": 9.764946810947879e-05, "loss": 0.0192, "step": 14310 }, { "grad_norm": 0.26363590359687805, "learning_rate": 9.764445542849738e-05, "loss": 0.0205, "step": 14320 }, { "grad_norm": 0.2104509174823761, "learning_rate": 9.763943753718998e-05, "loss": 0.02, "step": 14330 }, { "grad_norm": 0.24097664654254913, "learning_rate": 9.76344144361053e-05, "loss": 0.0183, "step": 14340 }, { "grad_norm": 0.2199251651763916, "learning_rate": 9.762938612579269e-05, "loss": 0.0193, "step": 14350 }, { "grad_norm": 0.18792477250099182, "learning_rate": 9.762435260680202e-05, "loss": 0.018, "step": 14360 }, { "grad_norm": 0.26086390018463135, "learning_rate": 9.761931387968373e-05, "loss": 0.0162, "step": 14370 }, { "grad_norm": 0.1941196322441101, "learning_rate": 9.76142699449889e-05, "loss": 0.0175, "step": 14380 }, { "grad_norm": 0.22449330985546112, "learning_rate": 9.760922080326908e-05, "loss": 0.0192, "step": 14390 }, { "grad_norm": 0.24040652811527252, "learning_rate": 9.760416645507644e-05, "loss": 0.0189, "step": 14400 }, { "grad_norm": 0.2737739384174347, "learning_rate": 9.759910690096375e-05, "loss": 0.0203, "step": 14410 }, { "grad_norm": 0.2480490356683731, "learning_rate": 9.759404214148429e-05, "loss": 0.0186, "step": 14420 }, { "grad_norm": 0.22791168093681335, "learning_rate": 9.758897217719191e-05, "loss": 0.017, "step": 14430 }, { "grad_norm": 0.2674652934074402, "learning_rate": 9.758389700864113e-05, "loss": 0.0174, "step": 14440 }, { "grad_norm": 0.21882611513137817, "learning_rate": 9.757881663638688e-05, "loss": 0.0178, "step": 14450 }, { "grad_norm": 0.2930465042591095, "learning_rate": 9.757373106098478e-05, "loss": 0.0191, "step": 14460 }, { "grad_norm": 0.19465486705303192, "learning_rate": 9.756864028299097e-05, "loss": 0.0198, "step": 14470 }, { "grad_norm": 0.23605073988437653, "learning_rate": 9.75635443029622e-05, "loss": 0.018, "step": 14480 }, { "grad_norm": 0.17747509479522705, "learning_rate": 9.755844312145572e-05, "loss": 0.0184, "step": 14490 }, { "grad_norm": 0.2934539020061493, "learning_rate": 9.755333673902941e-05, "loss": 0.0175, "step": 14500 }, { "grad_norm": 0.2929013669490814, "learning_rate": 9.75482251562417e-05, "loss": 0.0201, "step": 14510 }, { "grad_norm": 0.2624059021472931, "learning_rate": 9.754310837365155e-05, "loss": 0.0177, "step": 14520 }, { "grad_norm": 0.25129857659339905, "learning_rate": 9.753798639181856e-05, "loss": 0.0183, "step": 14530 }, { "grad_norm": 0.32529589533805847, "learning_rate": 9.753285921130286e-05, "loss": 0.0216, "step": 14540 }, { "grad_norm": 0.2064739614725113, "learning_rate": 9.752772683266512e-05, "loss": 0.0173, "step": 14550 }, { "grad_norm": 0.25926846265792847, "learning_rate": 9.752258925646665e-05, "loss": 0.0189, "step": 14560 }, { "grad_norm": 0.28970417380332947, "learning_rate": 9.751744648326926e-05, "loss": 0.0197, "step": 14570 }, { "grad_norm": 0.22389845550060272, "learning_rate": 9.751229851363536e-05, "loss": 0.0191, "step": 14580 }, { "grad_norm": 0.2963545024394989, "learning_rate": 9.750714534812793e-05, "loss": 0.0188, "step": 14590 }, { "grad_norm": 0.2363317906856537, "learning_rate": 9.750198698731053e-05, "loss": 0.0179, "step": 14600 }, { "grad_norm": 0.2360946387052536, "learning_rate": 9.749682343174722e-05, "loss": 0.0189, "step": 14610 }, { "grad_norm": 0.24903824925422668, "learning_rate": 9.749165468200272e-05, "loss": 0.0192, "step": 14620 }, { "grad_norm": 0.2664070129394531, "learning_rate": 9.748648073864229e-05, "loss": 0.0214, "step": 14630 }, { "grad_norm": 0.19837586581707, "learning_rate": 9.748130160223168e-05, "loss": 0.0178, "step": 14640 }, { "grad_norm": 0.29618510603904724, "learning_rate": 9.747611727333734e-05, "loss": 0.0162, "step": 14650 }, { "grad_norm": 0.19197195768356323, "learning_rate": 9.74709277525262e-05, "loss": 0.0231, "step": 14660 }, { "grad_norm": 0.22253884375095367, "learning_rate": 9.746573304036576e-05, "loss": 0.0185, "step": 14670 }, { "grad_norm": 0.19250760972499847, "learning_rate": 9.746053313742412e-05, "loss": 0.0192, "step": 14680 }, { "grad_norm": 0.21338926255702972, "learning_rate": 9.745532804426994e-05, "loss": 0.0192, "step": 14690 }, { "grad_norm": 0.2426111251115799, "learning_rate": 9.745011776147242e-05, "loss": 0.0192, "step": 14700 }, { "grad_norm": 0.1984439194202423, "learning_rate": 9.744490228960138e-05, "loss": 0.0165, "step": 14710 }, { "grad_norm": 0.20076748728752136, "learning_rate": 9.743968162922713e-05, "loss": 0.0196, "step": 14720 }, { "grad_norm": 0.20022568106651306, "learning_rate": 9.743445578092064e-05, "loss": 0.0192, "step": 14730 }, { "grad_norm": 0.32933324575424194, "learning_rate": 9.742922474525338e-05, "loss": 0.0234, "step": 14740 }, { "grad_norm": 0.22511069476604462, "learning_rate": 9.742398852279741e-05, "loss": 0.0187, "step": 14750 }, { "grad_norm": 0.24998340010643005, "learning_rate": 9.741874711412535e-05, "loss": 0.0174, "step": 14760 }, { "grad_norm": 0.21976150572299957, "learning_rate": 9.741350051981042e-05, "loss": 0.0199, "step": 14770 }, { "grad_norm": 0.2420385628938675, "learning_rate": 9.740824874042633e-05, "loss": 0.0187, "step": 14780 }, { "grad_norm": 0.21586838364601135, "learning_rate": 9.740299177654746e-05, "loss": 0.0201, "step": 14790 }, { "grad_norm": 0.21942536532878876, "learning_rate": 9.739772962874867e-05, "loss": 0.0183, "step": 14800 }, { "grad_norm": 0.25409555435180664, "learning_rate": 9.739246229760541e-05, "loss": 0.018, "step": 14810 }, { "grad_norm": 0.23139415681362152, "learning_rate": 9.738718978369376e-05, "loss": 0.0198, "step": 14820 }, { "grad_norm": 0.20480534434318542, "learning_rate": 9.738191208759025e-05, "loss": 0.02, "step": 14830 }, { "grad_norm": 0.2528938055038452, "learning_rate": 9.73766292098721e-05, "loss": 0.0205, "step": 14840 }, { "grad_norm": 0.28415197134017944, "learning_rate": 9.737134115111699e-05, "loss": 0.02, "step": 14850 }, { "grad_norm": 0.21562644839286804, "learning_rate": 9.736604791190323e-05, "loss": 0.0163, "step": 14860 }, { "grad_norm": 0.25017449259757996, "learning_rate": 9.73607494928097e-05, "loss": 0.0172, "step": 14870 }, { "grad_norm": 0.23218074440956116, "learning_rate": 9.735544589441581e-05, "loss": 0.0196, "step": 14880 }, { "grad_norm": 0.2281000316143036, "learning_rate": 9.735013711730154e-05, "loss": 0.0191, "step": 14890 }, { "grad_norm": 0.16498741507530212, "learning_rate": 9.734482316204747e-05, "loss": 0.0184, "step": 14900 }, { "grad_norm": 0.23017451167106628, "learning_rate": 9.733950402923473e-05, "loss": 0.0162, "step": 14910 }, { "grad_norm": 0.180561825633049, "learning_rate": 9.7334179719445e-05, "loss": 0.0204, "step": 14920 }, { "grad_norm": 0.28598877787590027, "learning_rate": 9.732885023326053e-05, "loss": 0.0185, "step": 14930 }, { "grad_norm": 0.23784442245960236, "learning_rate": 9.732351557126418e-05, "loss": 0.0211, "step": 14940 }, { "grad_norm": 0.2670738995075226, "learning_rate": 9.731817573403929e-05, "loss": 0.0178, "step": 14950 }, { "grad_norm": 0.2752399444580078, "learning_rate": 9.731283072216985e-05, "loss": 0.0168, "step": 14960 }, { "grad_norm": 0.257798433303833, "learning_rate": 9.730748053624039e-05, "loss": 0.0239, "step": 14970 }, { "grad_norm": 0.1911325603723526, "learning_rate": 9.730212517683598e-05, "loss": 0.0178, "step": 14980 }, { "grad_norm": 0.18320420384407043, "learning_rate": 9.729676464454228e-05, "loss": 0.0162, "step": 14990 }, { "grad_norm": 0.21195635199546814, "learning_rate": 9.72913989399455e-05, "loss": 0.0194, "step": 15000 }, { "grad_norm": 0.2273433804512024, "learning_rate": 9.728602806363242e-05, "loss": 0.0166, "step": 15010 }, { "grad_norm": 0.20955424010753632, "learning_rate": 9.728065201619043e-05, "loss": 0.0193, "step": 15020 }, { "grad_norm": 0.23621095716953278, "learning_rate": 9.727527079820742e-05, "loss": 0.0167, "step": 15030 }, { "grad_norm": 0.25110313296318054, "learning_rate": 9.726988441027186e-05, "loss": 0.0147, "step": 15040 }, { "grad_norm": 0.1489604413509369, "learning_rate": 9.726449285297281e-05, "loss": 0.0169, "step": 15050 }, { "grad_norm": 0.20689058303833008, "learning_rate": 9.72590961268999e-05, "loss": 0.016, "step": 15060 }, { "grad_norm": 0.2475339025259018, "learning_rate": 9.725369423264328e-05, "loss": 0.0165, "step": 15070 }, { "grad_norm": 0.23765183985233307, "learning_rate": 9.72482871707937e-05, "loss": 0.0186, "step": 15080 }, { "grad_norm": 0.2699069082736969, "learning_rate": 9.724287494194247e-05, "loss": 0.0172, "step": 15090 }, { "grad_norm": 0.26932406425476074, "learning_rate": 9.723745754668147e-05, "loss": 0.0189, "step": 15100 }, { "grad_norm": 0.2890337109565735, "learning_rate": 9.723203498560313e-05, "loss": 0.0193, "step": 15110 }, { "grad_norm": 0.23724760115146637, "learning_rate": 9.722660725930046e-05, "loss": 0.019, "step": 15120 }, { "grad_norm": 0.2354573756456375, "learning_rate": 9.722117436836702e-05, "loss": 0.0187, "step": 15130 }, { "grad_norm": 0.22294853627681732, "learning_rate": 9.721573631339696e-05, "loss": 0.0213, "step": 15140 }, { "grad_norm": 0.18640656769275665, "learning_rate": 9.721029309498494e-05, "loss": 0.0202, "step": 15150 }, { "grad_norm": 0.2656531035900116, "learning_rate": 9.720484471372627e-05, "loss": 0.021, "step": 15160 }, { "grad_norm": 0.40046921372413635, "learning_rate": 9.719939117021673e-05, "loss": 0.02, "step": 15170 }, { "grad_norm": 0.3300333619117737, "learning_rate": 9.719393246505275e-05, "loss": 0.0202, "step": 15180 }, { "grad_norm": 0.2712109386920929, "learning_rate": 9.718846859883128e-05, "loss": 0.0223, "step": 15190 }, { "grad_norm": 0.25776368379592896, "learning_rate": 9.718299957214982e-05, "loss": 0.0199, "step": 15200 }, { "grad_norm": 0.3186577558517456, "learning_rate": 9.717752538560646e-05, "loss": 0.0197, "step": 15210 }, { "grad_norm": 0.31239065527915955, "learning_rate": 9.717204603979986e-05, "loss": 0.0184, "step": 15220 }, { "grad_norm": 0.25294017791748047, "learning_rate": 9.716656153532922e-05, "loss": 0.0163, "step": 15230 }, { "grad_norm": 0.2268655151128769, "learning_rate": 9.716107187279434e-05, "loss": 0.0192, "step": 15240 }, { "grad_norm": 0.25151771306991577, "learning_rate": 9.715557705279555e-05, "loss": 0.0168, "step": 15250 }, { "grad_norm": 0.19223053753376007, "learning_rate": 9.715007707593372e-05, "loss": 0.0193, "step": 15260 }, { "grad_norm": 0.19994351267814636, "learning_rate": 9.714457194281036e-05, "loss": 0.0177, "step": 15270 }, { "grad_norm": 0.2644948959350586, "learning_rate": 9.713906165402751e-05, "loss": 0.0194, "step": 15280 }, { "grad_norm": 0.22304460406303406, "learning_rate": 9.713354621018774e-05, "loss": 0.0177, "step": 15290 }, { "grad_norm": 0.23888207972049713, "learning_rate": 9.712802561189422e-05, "loss": 0.0165, "step": 15300 }, { "grad_norm": 0.2404317855834961, "learning_rate": 9.712249985975069e-05, "loss": 0.0167, "step": 15310 }, { "grad_norm": 0.3131415843963623, "learning_rate": 9.71169689543614e-05, "loss": 0.0204, "step": 15320 }, { "grad_norm": 0.2071506381034851, "learning_rate": 9.711143289633123e-05, "loss": 0.017, "step": 15330 }, { "grad_norm": 0.2236786037683487, "learning_rate": 9.710589168626561e-05, "loss": 0.0192, "step": 15340 }, { "grad_norm": 0.2382163256406784, "learning_rate": 9.710034532477048e-05, "loss": 0.0178, "step": 15350 }, { "grad_norm": 0.23613156378269196, "learning_rate": 9.709479381245239e-05, "loss": 0.0196, "step": 15360 }, { "grad_norm": 0.25308072566986084, "learning_rate": 9.708923714991847e-05, "loss": 0.0237, "step": 15370 }, { "grad_norm": 0.21178027987480164, "learning_rate": 9.708367533777638e-05, "loss": 0.0222, "step": 15380 }, { "grad_norm": 0.1903669834136963, "learning_rate": 9.707810837663431e-05, "loss": 0.0189, "step": 15390 }, { "grad_norm": 0.21109549701213837, "learning_rate": 9.707253626710113e-05, "loss": 0.0211, "step": 15400 }, { "grad_norm": 0.20496642589569092, "learning_rate": 9.706695900978613e-05, "loss": 0.0194, "step": 15410 }, { "grad_norm": 0.19703637063503265, "learning_rate": 9.706137660529926e-05, "loss": 0.0185, "step": 15420 }, { "grad_norm": 0.1867263913154602, "learning_rate": 9.705578905425101e-05, "loss": 0.0195, "step": 15430 }, { "grad_norm": 0.17305907607078552, "learning_rate": 9.705019635725241e-05, "loss": 0.0202, "step": 15440 }, { "grad_norm": 0.3548576533794403, "learning_rate": 9.704459851491508e-05, "loss": 0.0177, "step": 15450 }, { "grad_norm": 0.24438324570655823, "learning_rate": 9.703899552785118e-05, "loss": 0.0191, "step": 15460 }, { "grad_norm": 0.22638779878616333, "learning_rate": 9.703338739667346e-05, "loss": 0.0215, "step": 15470 }, { "grad_norm": 0.2070508748292923, "learning_rate": 9.70277741219952e-05, "loss": 0.0198, "step": 15480 }, { "grad_norm": 0.2055637389421463, "learning_rate": 9.702215570443027e-05, "loss": 0.0249, "step": 15490 }, { "grad_norm": 0.3137443959712982, "learning_rate": 9.701653214459309e-05, "loss": 0.023, "step": 15500 }, { "grad_norm": 0.2984990179538727, "learning_rate": 9.701090344309865e-05, "loss": 0.0206, "step": 15510 }, { "grad_norm": 0.2588450610637665, "learning_rate": 9.700526960056247e-05, "loss": 0.0205, "step": 15520 }, { "grad_norm": 0.2962475121021271, "learning_rate": 9.699963061760068e-05, "loss": 0.0193, "step": 15530 }, { "grad_norm": 0.14946980774402618, "learning_rate": 9.699398649482997e-05, "loss": 0.0181, "step": 15540 }, { "grad_norm": 0.292804092168808, "learning_rate": 9.698833723286753e-05, "loss": 0.0188, "step": 15550 }, { "grad_norm": 0.23699240386486053, "learning_rate": 9.698268283233118e-05, "loss": 0.0179, "step": 15560 }, { "grad_norm": 0.27240660786628723, "learning_rate": 9.697702329383929e-05, "loss": 0.0177, "step": 15570 }, { "grad_norm": 0.25485101342201233, "learning_rate": 9.697135861801074e-05, "loss": 0.0151, "step": 15580 }, { "grad_norm": 0.2662970721721649, "learning_rate": 9.696568880546505e-05, "loss": 0.0184, "step": 15590 }, { "grad_norm": 0.18846474587917328, "learning_rate": 9.696001385682223e-05, "loss": 0.0197, "step": 15600 }, { "grad_norm": 0.18403780460357666, "learning_rate": 9.695433377270291e-05, "loss": 0.0179, "step": 15610 }, { "grad_norm": 0.18065325915813446, "learning_rate": 9.694864855372824e-05, "loss": 0.0164, "step": 15620 }, { "grad_norm": 0.19837354123592377, "learning_rate": 9.694295820051995e-05, "loss": 0.0163, "step": 15630 }, { "grad_norm": 0.2582974433898926, "learning_rate": 9.693726271370032e-05, "loss": 0.019, "step": 15640 }, { "grad_norm": 0.20642410218715668, "learning_rate": 9.693156209389221e-05, "loss": 0.0187, "step": 15650 }, { "grad_norm": 0.2757168710231781, "learning_rate": 9.692585634171905e-05, "loss": 0.0207, "step": 15660 }, { "grad_norm": 0.2989133894443512, "learning_rate": 9.692014545780476e-05, "loss": 0.0197, "step": 15670 }, { "grad_norm": 0.2628140449523926, "learning_rate": 9.691442944277393e-05, "loss": 0.02, "step": 15680 }, { "grad_norm": 0.24691957235336304, "learning_rate": 9.690870829725162e-05, "loss": 0.0164, "step": 15690 }, { "grad_norm": 0.21299654245376587, "learning_rate": 9.69029820218635e-05, "loss": 0.015, "step": 15700 }, { "grad_norm": 0.23913165926933289, "learning_rate": 9.689725061723579e-05, "loss": 0.0178, "step": 15710 }, { "grad_norm": 0.24879802763462067, "learning_rate": 9.689151408399527e-05, "loss": 0.0188, "step": 15720 }, { "grad_norm": 0.30879539251327515, "learning_rate": 9.688577242276924e-05, "loss": 0.0205, "step": 15730 }, { "grad_norm": 0.19415371119976044, "learning_rate": 9.688002563418566e-05, "loss": 0.0188, "step": 15740 }, { "grad_norm": 0.2483912855386734, "learning_rate": 9.687427371887293e-05, "loss": 0.0184, "step": 15750 }, { "grad_norm": 0.2782413363456726, "learning_rate": 9.686851667746012e-05, "loss": 0.0208, "step": 15760 }, { "grad_norm": 0.16630974411964417, "learning_rate": 9.686275451057677e-05, "loss": 0.0198, "step": 15770 }, { "grad_norm": 0.3061389625072479, "learning_rate": 9.685698721885308e-05, "loss": 0.0184, "step": 15780 }, { "grad_norm": 0.2253051996231079, "learning_rate": 9.68512148029197e-05, "loss": 0.018, "step": 15790 }, { "grad_norm": 0.208334818482399, "learning_rate": 9.684543726340791e-05, "loss": 0.0189, "step": 15800 }, { "grad_norm": 0.2350933849811554, "learning_rate": 9.683965460094952e-05, "loss": 0.0167, "step": 15810 }, { "grad_norm": 0.20684881508350372, "learning_rate": 9.683386681617694e-05, "loss": 0.0203, "step": 15820 }, { "grad_norm": 0.24168388545513153, "learning_rate": 9.68280739097231e-05, "loss": 0.0164, "step": 15830 }, { "grad_norm": 0.26697590947151184, "learning_rate": 9.682227588222148e-05, "loss": 0.022, "step": 15840 }, { "grad_norm": 0.18913869559764862, "learning_rate": 9.681647273430618e-05, "loss": 0.0171, "step": 15850 }, { "grad_norm": 0.25664788484573364, "learning_rate": 9.681066446661182e-05, "loss": 0.0198, "step": 15860 }, { "grad_norm": 0.1858476847410202, "learning_rate": 9.680485107977357e-05, "loss": 0.0165, "step": 15870 }, { "grad_norm": 0.21210715174674988, "learning_rate": 9.679903257442716e-05, "loss": 0.0197, "step": 15880 }, { "grad_norm": 0.2298712134361267, "learning_rate": 9.679320895120891e-05, "loss": 0.0187, "step": 15890 }, { "grad_norm": 0.21098913252353668, "learning_rate": 9.67873802107557e-05, "loss": 0.0213, "step": 15900 }, { "grad_norm": 0.19973057508468628, "learning_rate": 9.67815463537049e-05, "loss": 0.0196, "step": 15910 }, { "grad_norm": 0.22202305495738983, "learning_rate": 9.677570738069457e-05, "loss": 0.0233, "step": 15920 }, { "grad_norm": 0.25181034207344055, "learning_rate": 9.676986329236318e-05, "loss": 0.0185, "step": 15930 }, { "grad_norm": 0.24693556129932404, "learning_rate": 9.676401408934987e-05, "loss": 0.019, "step": 15940 }, { "grad_norm": 0.2489759773015976, "learning_rate": 9.675815977229428e-05, "loss": 0.0221, "step": 15950 }, { "grad_norm": 0.24107781052589417, "learning_rate": 9.675230034183664e-05, "loss": 0.0187, "step": 15960 }, { "grad_norm": 0.17176386713981628, "learning_rate": 9.674643579861773e-05, "loss": 0.0179, "step": 15970 }, { "grad_norm": 0.2019897997379303, "learning_rate": 9.674056614327886e-05, "loss": 0.0169, "step": 15980 }, { "grad_norm": 0.16684889793395996, "learning_rate": 9.673469137646198e-05, "loss": 0.016, "step": 15990 }, { "grad_norm": 0.24870023131370544, "learning_rate": 9.67288114988095e-05, "loss": 0.0216, "step": 16000 }, { "grad_norm": 0.28548723459243774, "learning_rate": 9.672292651096447e-05, "loss": 0.0192, "step": 16010 }, { "grad_norm": 0.22416341304779053, "learning_rate": 9.671703641357042e-05, "loss": 0.0204, "step": 16020 }, { "grad_norm": 0.19322705268859863, "learning_rate": 9.67111412072715e-05, "loss": 0.0181, "step": 16030 }, { "grad_norm": 0.22695937752723694, "learning_rate": 9.670524089271242e-05, "loss": 0.017, "step": 16040 }, { "grad_norm": 0.2162391096353531, "learning_rate": 9.669933547053842e-05, "loss": 0.0184, "step": 16050 }, { "grad_norm": 0.3097580671310425, "learning_rate": 9.669342494139531e-05, "loss": 0.0179, "step": 16060 }, { "grad_norm": 0.19233477115631104, "learning_rate": 9.668750930592943e-05, "loss": 0.0162, "step": 16070 }, { "grad_norm": 0.20032547414302826, "learning_rate": 9.668158856478775e-05, "loss": 0.0177, "step": 16080 }, { "grad_norm": 0.2236863225698471, "learning_rate": 9.66756627186177e-05, "loss": 0.0163, "step": 16090 }, { "grad_norm": 0.27315574884414673, "learning_rate": 9.666973176806737e-05, "loss": 0.0165, "step": 16100 }, { "grad_norm": 0.26202699542045593, "learning_rate": 9.666379571378534e-05, "loss": 0.0174, "step": 16110 }, { "grad_norm": 0.1937863677740097, "learning_rate": 9.665785455642076e-05, "loss": 0.0171, "step": 16120 }, { "grad_norm": 0.22323665022850037, "learning_rate": 9.665190829662337e-05, "loss": 0.0188, "step": 16130 }, { "grad_norm": 0.2690662443637848, "learning_rate": 9.664595693504342e-05, "loss": 0.0187, "step": 16140 }, { "grad_norm": 0.28733378648757935, "learning_rate": 9.664000047233175e-05, "loss": 0.0187, "step": 16150 }, { "grad_norm": 0.2871973514556885, "learning_rate": 9.663403890913976e-05, "loss": 0.0198, "step": 16160 }, { "grad_norm": 0.1865316778421402, "learning_rate": 9.662807224611938e-05, "loss": 0.0175, "step": 16170 }, { "grad_norm": 0.22111932933330536, "learning_rate": 9.662210048392311e-05, "loss": 0.0181, "step": 16180 }, { "grad_norm": 0.18446962535381317, "learning_rate": 9.661612362320405e-05, "loss": 0.0168, "step": 16190 }, { "grad_norm": 0.3004204332828522, "learning_rate": 9.661014166461579e-05, "loss": 0.0168, "step": 16200 }, { "grad_norm": 0.25319960713386536, "learning_rate": 9.66041546088125e-05, "loss": 0.0173, "step": 16210 }, { "grad_norm": 0.22596386075019836, "learning_rate": 9.659816245644895e-05, "loss": 0.0194, "step": 16220 }, { "grad_norm": 0.19223058223724365, "learning_rate": 9.65921652081804e-05, "loss": 0.0168, "step": 16230 }, { "grad_norm": 0.2782018780708313, "learning_rate": 9.658616286466271e-05, "loss": 0.0195, "step": 16240 }, { "grad_norm": 0.22007909417152405, "learning_rate": 9.65801554265523e-05, "loss": 0.0153, "step": 16250 }, { "grad_norm": 0.23331479728221893, "learning_rate": 9.657414289450612e-05, "loss": 0.0222, "step": 16260 }, { "grad_norm": 0.30209261178970337, "learning_rate": 9.656812526918171e-05, "loss": 0.0179, "step": 16270 }, { "grad_norm": 0.19492562115192413, "learning_rate": 9.656210255123712e-05, "loss": 0.0212, "step": 16280 }, { "grad_norm": 0.16627874970436096, "learning_rate": 9.6556074741331e-05, "loss": 0.015, "step": 16290 }, { "grad_norm": 0.2555200457572937, "learning_rate": 9.655004184012256e-05, "loss": 0.0196, "step": 16300 }, { "grad_norm": 0.22734369337558746, "learning_rate": 9.654400384827152e-05, "loss": 0.0187, "step": 16310 }, { "grad_norm": 0.25163406133651733, "learning_rate": 9.653796076643818e-05, "loss": 0.0185, "step": 16320 }, { "grad_norm": 0.27270692586898804, "learning_rate": 9.653191259528344e-05, "loss": 0.0163, "step": 16330 }, { "grad_norm": 0.2532217800617218, "learning_rate": 9.65258593354687e-05, "loss": 0.0169, "step": 16340 }, { "grad_norm": 0.32540151476860046, "learning_rate": 9.651980098765591e-05, "loss": 0.0197, "step": 16350 }, { "grad_norm": 0.21328486502170563, "learning_rate": 9.651373755250765e-05, "loss": 0.0211, "step": 16360 }, { "grad_norm": 0.15313532948493958, "learning_rate": 9.650766903068697e-05, "loss": 0.015, "step": 16370 }, { "grad_norm": 0.17251364886760712, "learning_rate": 9.650159542285753e-05, "loss": 0.0148, "step": 16380 }, { "grad_norm": 0.2306550145149231, "learning_rate": 9.649551672968353e-05, "loss": 0.0171, "step": 16390 }, { "grad_norm": 0.19853927195072174, "learning_rate": 9.648943295182973e-05, "loss": 0.021, "step": 16400 }, { "grad_norm": 0.24364575743675232, "learning_rate": 9.648334408996144e-05, "loss": 0.0188, "step": 16410 }, { "grad_norm": 0.19130511581897736, "learning_rate": 9.647725014474452e-05, "loss": 0.0195, "step": 16420 }, { "grad_norm": 0.24095939099788666, "learning_rate": 9.64711511168454e-05, "loss": 0.0162, "step": 16430 }, { "grad_norm": 0.23611345887184143, "learning_rate": 9.646504700693108e-05, "loss": 0.0161, "step": 16440 }, { "grad_norm": 0.23858371376991272, "learning_rate": 9.645893781566907e-05, "loss": 0.0192, "step": 16450 }, { "grad_norm": 0.259247362613678, "learning_rate": 9.645282354372744e-05, "loss": 0.0189, "step": 16460 }, { "grad_norm": 0.2198825180530548, "learning_rate": 9.644670419177491e-05, "loss": 0.0161, "step": 16470 }, { "grad_norm": 0.2003561556339264, "learning_rate": 9.644057976048062e-05, "loss": 0.0182, "step": 16480 }, { "grad_norm": 0.18261343240737915, "learning_rate": 9.643445025051435e-05, "loss": 0.0167, "step": 16490 }, { "grad_norm": 0.1412617266178131, "learning_rate": 9.642831566254641e-05, "loss": 0.0177, "step": 16500 }, { "grad_norm": 0.21510356664657593, "learning_rate": 9.642217599724769e-05, "loss": 0.0127, "step": 16510 }, { "grad_norm": 0.21736837923526764, "learning_rate": 9.64160312552896e-05, "loss": 0.0168, "step": 16520 }, { "grad_norm": 0.1895725578069687, "learning_rate": 9.64098814373441e-05, "loss": 0.0181, "step": 16530 }, { "grad_norm": 0.2633892595767975, "learning_rate": 9.640372654408374e-05, "loss": 0.0163, "step": 16540 }, { "grad_norm": 0.20585060119628906, "learning_rate": 9.639756657618162e-05, "loss": 0.0148, "step": 16550 }, { "grad_norm": 0.19495542347431183, "learning_rate": 9.639140153431138e-05, "loss": 0.0192, "step": 16560 }, { "grad_norm": 0.20450176298618317, "learning_rate": 9.638523141914721e-05, "loss": 0.0201, "step": 16570 }, { "grad_norm": 0.21701449155807495, "learning_rate": 9.637905623136388e-05, "loss": 0.0164, "step": 16580 }, { "grad_norm": 0.17691414058208466, "learning_rate": 9.637287597163669e-05, "loss": 0.0176, "step": 16590 }, { "grad_norm": 0.22074982523918152, "learning_rate": 9.63666906406415e-05, "loss": 0.0161, "step": 16600 }, { "grad_norm": 0.18536122143268585, "learning_rate": 9.636050023905473e-05, "loss": 0.0149, "step": 16610 }, { "grad_norm": 0.21514461934566498, "learning_rate": 9.635430476755336e-05, "loss": 0.017, "step": 16620 }, { "grad_norm": 0.21478553116321564, "learning_rate": 9.63481042268149e-05, "loss": 0.0196, "step": 16630 }, { "grad_norm": 0.27715858817100525, "learning_rate": 9.634189861751745e-05, "loss": 0.0167, "step": 16640 }, { "grad_norm": 0.22627702355384827, "learning_rate": 9.633568794033967e-05, "loss": 0.0231, "step": 16650 }, { "grad_norm": 0.18545715510845184, "learning_rate": 9.63294721959607e-05, "loss": 0.0162, "step": 16660 }, { "grad_norm": 0.20432135462760925, "learning_rate": 9.63232513850603e-05, "loss": 0.0196, "step": 16670 }, { "grad_norm": 0.225392147898674, "learning_rate": 9.631702550831878e-05, "loss": 0.0167, "step": 16680 }, { "grad_norm": 0.20848077535629272, "learning_rate": 9.631079456641698e-05, "loss": 0.0151, "step": 16690 }, { "grad_norm": 0.2306084781885147, "learning_rate": 9.630455856003632e-05, "loss": 0.0161, "step": 16700 }, { "grad_norm": 0.17233020067214966, "learning_rate": 9.629831748985876e-05, "loss": 0.0178, "step": 16710 }, { "grad_norm": 0.2605937719345093, "learning_rate": 9.629207135656679e-05, "loss": 0.0148, "step": 16720 }, { "grad_norm": 0.31103241443634033, "learning_rate": 9.628582016084353e-05, "loss": 0.0165, "step": 16730 }, { "grad_norm": 0.2806118130683899, "learning_rate": 9.627956390337254e-05, "loss": 0.0193, "step": 16740 }, { "grad_norm": 0.22293277084827423, "learning_rate": 9.627330258483802e-05, "loss": 0.018, "step": 16750 }, { "grad_norm": 0.2688979208469391, "learning_rate": 9.62670362059247e-05, "loss": 0.0187, "step": 16760 }, { "grad_norm": 0.3305911421775818, "learning_rate": 9.626076476731786e-05, "loss": 0.0195, "step": 16770 }, { "grad_norm": 0.24513816833496094, "learning_rate": 9.625448826970336e-05, "loss": 0.0176, "step": 16780 }, { "grad_norm": 0.19799001514911652, "learning_rate": 9.624820671376755e-05, "loss": 0.0152, "step": 16790 }, { "grad_norm": 0.21636264026165009, "learning_rate": 9.62419201001974e-05, "loss": 0.016, "step": 16800 }, { "grad_norm": 0.2144276648759842, "learning_rate": 9.623562842968037e-05, "loss": 0.0136, "step": 16810 }, { "grad_norm": 0.29770979285240173, "learning_rate": 9.622933170290454e-05, "loss": 0.0164, "step": 16820 }, { "grad_norm": 0.2525472044944763, "learning_rate": 9.622302992055849e-05, "loss": 0.017, "step": 16830 }, { "grad_norm": 0.19956479966640472, "learning_rate": 9.62167230833314e-05, "loss": 0.0156, "step": 16840 }, { "grad_norm": 0.2765447497367859, "learning_rate": 9.621041119191295e-05, "loss": 0.019, "step": 16850 }, { "grad_norm": 0.25838616490364075, "learning_rate": 9.620409424699342e-05, "loss": 0.0163, "step": 16860 }, { "grad_norm": 0.3392568528652191, "learning_rate": 9.619777224926359e-05, "loss": 0.0174, "step": 16870 }, { "grad_norm": 0.27198275923728943, "learning_rate": 9.619144519941485e-05, "loss": 0.017, "step": 16880 }, { "grad_norm": 0.2395709753036499, "learning_rate": 9.618511309813912e-05, "loss": 0.0175, "step": 16890 }, { "grad_norm": 0.22025811672210693, "learning_rate": 9.617877594612886e-05, "loss": 0.0186, "step": 16900 }, { "grad_norm": 0.17234596610069275, "learning_rate": 9.617243374407707e-05, "loss": 0.0164, "step": 16910 }, { "grad_norm": 0.25928282737731934, "learning_rate": 9.616608649267736e-05, "loss": 0.0199, "step": 16920 }, { "grad_norm": 0.21347233653068542, "learning_rate": 9.615973419262385e-05, "loss": 0.0154, "step": 16930 }, { "grad_norm": 0.20096708834171295, "learning_rate": 9.615337684461119e-05, "loss": 0.0169, "step": 16940 }, { "grad_norm": 0.29008498787879944, "learning_rate": 9.614701444933465e-05, "loss": 0.0196, "step": 16950 }, { "grad_norm": 0.1745767742395401, "learning_rate": 9.614064700748997e-05, "loss": 0.0207, "step": 16960 }, { "grad_norm": 0.2339542955160141, "learning_rate": 9.613427451977352e-05, "loss": 0.021, "step": 16970 }, { "grad_norm": 0.2315192073583603, "learning_rate": 9.612789698688216e-05, "loss": 0.0169, "step": 16980 }, { "grad_norm": 0.18440251052379608, "learning_rate": 9.612151440951334e-05, "loss": 0.0143, "step": 16990 }, { "grad_norm": 0.19225835800170898, "learning_rate": 9.611512678836506e-05, "loss": 0.0183, "step": 17000 }, { "grad_norm": 0.24110615253448486, "learning_rate": 9.610873412413584e-05, "loss": 0.0232, "step": 17010 }, { "grad_norm": 0.2444438636302948, "learning_rate": 9.610233641752476e-05, "loss": 0.0201, "step": 17020 }, { "grad_norm": 0.2314738929271698, "learning_rate": 9.609593366923151e-05, "loss": 0.0181, "step": 17030 }, { "grad_norm": 0.26004305481910706, "learning_rate": 9.608952587995625e-05, "loss": 0.0192, "step": 17040 }, { "grad_norm": 0.20738676190376282, "learning_rate": 9.608311305039972e-05, "loss": 0.0198, "step": 17050 }, { "grad_norm": 0.14916974306106567, "learning_rate": 9.607669518126326e-05, "loss": 0.0177, "step": 17060 }, { "grad_norm": 0.20310278236865997, "learning_rate": 9.607027227324866e-05, "loss": 0.0186, "step": 17070 }, { "grad_norm": 0.20561285316944122, "learning_rate": 9.606384432705837e-05, "loss": 0.0153, "step": 17080 }, { "grad_norm": 0.1984809935092926, "learning_rate": 9.60574113433953e-05, "loss": 0.0189, "step": 17090 }, { "grad_norm": 0.25751596689224243, "learning_rate": 9.6050973322963e-05, "loss": 0.0153, "step": 17100 }, { "grad_norm": 0.30238303542137146, "learning_rate": 9.604453026646547e-05, "loss": 0.0162, "step": 17110 }, { "grad_norm": 0.24119111895561218, "learning_rate": 9.603808217460735e-05, "loss": 0.017, "step": 17120 }, { "grad_norm": 0.24318204820156097, "learning_rate": 9.603162904809377e-05, "loss": 0.0173, "step": 17130 }, { "grad_norm": 0.18383049964904785, "learning_rate": 9.602517088763045e-05, "loss": 0.0183, "step": 17140 }, { "grad_norm": 0.25545281171798706, "learning_rate": 9.601870769392365e-05, "loss": 0.0162, "step": 17150 }, { "grad_norm": 0.2306479811668396, "learning_rate": 9.601223946768017e-05, "loss": 0.0177, "step": 17160 }, { "grad_norm": 0.19132913649082184, "learning_rate": 9.600576620960734e-05, "loss": 0.0155, "step": 17170 }, { "grad_norm": 0.18730118870735168, "learning_rate": 9.599928792041308e-05, "loss": 0.0158, "step": 17180 }, { "grad_norm": 0.2748897671699524, "learning_rate": 9.599280460080587e-05, "loss": 0.0206, "step": 17190 }, { "grad_norm": 0.2627412974834442, "learning_rate": 9.59863162514947e-05, "loss": 0.0157, "step": 17200 }, { "grad_norm": 0.23091202974319458, "learning_rate": 9.597982287318911e-05, "loss": 0.0137, "step": 17210 }, { "grad_norm": 0.21098028123378754, "learning_rate": 9.597332446659923e-05, "loss": 0.017, "step": 17220 }, { "grad_norm": 0.2130013257265091, "learning_rate": 9.59668210324357e-05, "loss": 0.0198, "step": 17230 }, { "grad_norm": 0.24603047966957092, "learning_rate": 9.596031257140974e-05, "loss": 0.0188, "step": 17240 }, { "grad_norm": 0.3133513629436493, "learning_rate": 9.59537990842331e-05, "loss": 0.0162, "step": 17250 }, { "grad_norm": 0.22422030568122864, "learning_rate": 9.594728057161806e-05, "loss": 0.0191, "step": 17260 }, { "grad_norm": 0.21449986100196838, "learning_rate": 9.594075703427752e-05, "loss": 0.0164, "step": 17270 }, { "grad_norm": 0.1403755396604538, "learning_rate": 9.593422847292486e-05, "loss": 0.0168, "step": 17280 }, { "grad_norm": 0.2125183641910553, "learning_rate": 9.592769488827402e-05, "loss": 0.0153, "step": 17290 }, { "grad_norm": 0.20763233304023743, "learning_rate": 9.592115628103952e-05, "loss": 0.0172, "step": 17300 }, { "grad_norm": 0.22702524065971375, "learning_rate": 9.591461265193643e-05, "loss": 0.0159, "step": 17310 }, { "grad_norm": 0.160615012049675, "learning_rate": 9.590806400168032e-05, "loss": 0.0166, "step": 17320 }, { "grad_norm": 0.19173403084278107, "learning_rate": 9.590151033098735e-05, "loss": 0.0189, "step": 17330 }, { "grad_norm": 0.23532021045684814, "learning_rate": 9.589495164057423e-05, "loss": 0.0163, "step": 17340 }, { "grad_norm": 0.2838195264339447, "learning_rate": 9.58883879311582e-05, "loss": 0.0189, "step": 17350 }, { "grad_norm": 0.19104242324829102, "learning_rate": 9.588181920345705e-05, "loss": 0.0167, "step": 17360 }, { "grad_norm": 0.20199482142925262, "learning_rate": 9.587524545818913e-05, "loss": 0.0147, "step": 17370 }, { "grad_norm": 0.23055019974708557, "learning_rate": 9.586866669607335e-05, "loss": 0.0159, "step": 17380 }, { "grad_norm": 0.19228534400463104, "learning_rate": 9.586208291782915e-05, "loss": 0.0175, "step": 17390 }, { "grad_norm": 0.23962320387363434, "learning_rate": 9.58554941241765e-05, "loss": 0.0191, "step": 17400 }, { "grad_norm": 0.20090635120868683, "learning_rate": 9.584890031583596e-05, "loss": 0.0172, "step": 17410 }, { "grad_norm": 0.2766587436199188, "learning_rate": 9.584230149352861e-05, "loss": 0.0165, "step": 17420 }, { "grad_norm": 0.23822492361068726, "learning_rate": 9.58356976579761e-05, "loss": 0.0161, "step": 17430 }, { "grad_norm": 0.23660501837730408, "learning_rate": 9.58290888099006e-05, "loss": 0.0169, "step": 17440 }, { "grad_norm": 0.2155815064907074, "learning_rate": 9.582247495002486e-05, "loss": 0.0179, "step": 17450 }, { "grad_norm": 0.19209228456020355, "learning_rate": 9.581585607907214e-05, "loss": 0.0157, "step": 17460 }, { "grad_norm": 0.1874765008687973, "learning_rate": 9.580923219776628e-05, "loss": 0.0185, "step": 17470 }, { "grad_norm": 0.2557556927204132, "learning_rate": 9.580260330683167e-05, "loss": 0.0174, "step": 17480 }, { "grad_norm": 0.26812225580215454, "learning_rate": 9.579596940699322e-05, "loss": 0.0166, "step": 17490 }, { "grad_norm": 0.16675162315368652, "learning_rate": 9.578933049897643e-05, "loss": 0.0172, "step": 17500 }, { "grad_norm": 0.1950807124376297, "learning_rate": 9.578268658350728e-05, "loss": 0.0167, "step": 17510 }, { "grad_norm": 0.1877523511648178, "learning_rate": 9.577603766131235e-05, "loss": 0.0139, "step": 17520 }, { "grad_norm": 0.22977957129478455, "learning_rate": 9.576938373311878e-05, "loss": 0.0179, "step": 17530 }, { "grad_norm": 0.22567792236804962, "learning_rate": 9.576272479965421e-05, "loss": 0.0167, "step": 17540 }, { "grad_norm": 0.20277749001979828, "learning_rate": 9.575606086164687e-05, "loss": 0.0159, "step": 17550 }, { "grad_norm": 0.18969689309597015, "learning_rate": 9.57493919198255e-05, "loss": 0.0168, "step": 17560 }, { "grad_norm": 0.2616947293281555, "learning_rate": 9.57427179749194e-05, "loss": 0.0179, "step": 17570 }, { "grad_norm": 0.1700340360403061, "learning_rate": 9.573603902765846e-05, "loss": 0.0173, "step": 17580 }, { "grad_norm": 0.18502207100391388, "learning_rate": 9.572935507877304e-05, "loss": 0.0171, "step": 17590 }, { "grad_norm": 0.2221074104309082, "learning_rate": 9.57226661289941e-05, "loss": 0.0144, "step": 17600 }, { "grad_norm": 0.17950530350208282, "learning_rate": 9.571597217905315e-05, "loss": 0.0164, "step": 17610 }, { "grad_norm": 0.16716626286506653, "learning_rate": 9.57092732296822e-05, "loss": 0.0151, "step": 17620 }, { "grad_norm": 0.24112018942832947, "learning_rate": 9.570256928161385e-05, "loss": 0.0168, "step": 17630 }, { "grad_norm": 0.18545861542224884, "learning_rate": 9.569586033558126e-05, "loss": 0.0162, "step": 17640 }, { "grad_norm": 0.15963897109031677, "learning_rate": 9.568914639231807e-05, "loss": 0.013, "step": 17650 }, { "grad_norm": 0.2552218437194824, "learning_rate": 9.568242745255852e-05, "loss": 0.0141, "step": 17660 }, { "grad_norm": 0.15597037971019745, "learning_rate": 9.567570351703739e-05, "loss": 0.0166, "step": 17670 }, { "grad_norm": 0.20033545792102814, "learning_rate": 9.566897458649001e-05, "loss": 0.0206, "step": 17680 }, { "grad_norm": 0.24034996330738068, "learning_rate": 9.566224066165221e-05, "loss": 0.0151, "step": 17690 }, { "grad_norm": 0.27775871753692627, "learning_rate": 9.565550174326043e-05, "loss": 0.018, "step": 17700 }, { "grad_norm": 0.16614799201488495, "learning_rate": 9.564875783205162e-05, "loss": 0.016, "step": 17710 }, { "grad_norm": 0.15449222922325134, "learning_rate": 9.564200892876328e-05, "loss": 0.0186, "step": 17720 }, { "grad_norm": 0.19649669528007507, "learning_rate": 9.563525503413348e-05, "loss": 0.0156, "step": 17730 }, { "grad_norm": 0.20681892335414886, "learning_rate": 9.562849614890079e-05, "loss": 0.019, "step": 17740 }, { "grad_norm": 0.24867214262485504, "learning_rate": 9.562173227380436e-05, "loss": 0.0135, "step": 17750 }, { "grad_norm": 0.19534456729888916, "learning_rate": 9.561496340958389e-05, "loss": 0.0227, "step": 17760 }, { "grad_norm": 0.19656559824943542, "learning_rate": 9.560818955697959e-05, "loss": 0.0173, "step": 17770 }, { "grad_norm": 0.22119058668613434, "learning_rate": 9.560141071673228e-05, "loss": 0.0154, "step": 17780 }, { "grad_norm": 0.22350183129310608, "learning_rate": 9.559462688958323e-05, "loss": 0.0174, "step": 17790 }, { "grad_norm": 0.2594278156757355, "learning_rate": 9.558783807627434e-05, "loss": 0.0161, "step": 17800 }, { "grad_norm": 0.18888922035694122, "learning_rate": 9.558104427754801e-05, "loss": 0.0199, "step": 17810 }, { "grad_norm": 0.20429685711860657, "learning_rate": 9.557424549414722e-05, "loss": 0.0164, "step": 17820 }, { "grad_norm": 0.22368460893630981, "learning_rate": 9.556744172681546e-05, "loss": 0.0145, "step": 17830 }, { "grad_norm": 0.17436495423316956, "learning_rate": 9.556063297629677e-05, "loss": 0.0164, "step": 17840 }, { "grad_norm": 0.1889360398054123, "learning_rate": 9.555381924333578e-05, "loss": 0.0154, "step": 17850 }, { "grad_norm": 0.1872352659702301, "learning_rate": 9.554700052867758e-05, "loss": 0.0162, "step": 17860 }, { "grad_norm": 0.2772279977798462, "learning_rate": 9.554017683306789e-05, "loss": 0.0202, "step": 17870 }, { "grad_norm": 0.22168686985969543, "learning_rate": 9.553334815725294e-05, "loss": 0.0153, "step": 17880 }, { "grad_norm": 0.27190110087394714, "learning_rate": 9.552651450197949e-05, "loss": 0.0184, "step": 17890 }, { "grad_norm": 0.23202817142009735, "learning_rate": 9.551967586799486e-05, "loss": 0.0161, "step": 17900 }, { "grad_norm": 0.18934252858161926, "learning_rate": 9.551283225604692e-05, "loss": 0.0186, "step": 17910 }, { "grad_norm": 0.21593748033046722, "learning_rate": 9.550598366688406e-05, "loss": 0.0187, "step": 17920 }, { "grad_norm": 0.1996971070766449, "learning_rate": 9.549913010125526e-05, "loss": 0.015, "step": 17930 }, { "grad_norm": 0.2126038819551468, "learning_rate": 9.549227155990999e-05, "loss": 0.017, "step": 17940 }, { "grad_norm": 0.16797807812690735, "learning_rate": 9.548540804359828e-05, "loss": 0.0135, "step": 17950 }, { "grad_norm": 0.2333494871854782, "learning_rate": 9.547853955307077e-05, "loss": 0.0193, "step": 17960 }, { "grad_norm": 0.2139659821987152, "learning_rate": 9.547166608907853e-05, "loss": 0.0173, "step": 17970 }, { "grad_norm": 0.2433938831090927, "learning_rate": 9.546478765237326e-05, "loss": 0.0157, "step": 17980 }, { "grad_norm": 0.2255486696958542, "learning_rate": 9.545790424370715e-05, "loss": 0.0179, "step": 17990 }, { "grad_norm": 0.19993619620800018, "learning_rate": 9.5451015863833e-05, "loss": 0.0155, "step": 18000 }, { "grad_norm": 0.2615847885608673, "learning_rate": 9.544412251350408e-05, "loss": 0.0187, "step": 18010 }, { "grad_norm": 0.2172890156507492, "learning_rate": 9.543722419347422e-05, "loss": 0.015, "step": 18020 }, { "grad_norm": 0.2116304486989975, "learning_rate": 9.543032090449788e-05, "loss": 0.0142, "step": 18030 }, { "grad_norm": 0.17011810839176178, "learning_rate": 9.542341264732992e-05, "loss": 0.0151, "step": 18040 }, { "grad_norm": 0.18334297835826874, "learning_rate": 9.541649942272585e-05, "loss": 0.0167, "step": 18050 }, { "grad_norm": 0.23310072720050812, "learning_rate": 9.54095812314417e-05, "loss": 0.0195, "step": 18060 }, { "grad_norm": 0.2265799194574356, "learning_rate": 9.540265807423401e-05, "loss": 0.0178, "step": 18070 }, { "grad_norm": 0.20345690846443176, "learning_rate": 9.53957299518599e-05, "loss": 0.0152, "step": 18080 }, { "grad_norm": 0.2680896520614624, "learning_rate": 9.5388796865077e-05, "loss": 0.0176, "step": 18090 }, { "grad_norm": 0.18680095672607422, "learning_rate": 9.538185881464353e-05, "loss": 0.0176, "step": 18100 }, { "grad_norm": 0.23559193313121796, "learning_rate": 9.537491580131821e-05, "loss": 0.0156, "step": 18110 }, { "grad_norm": 0.2078387439250946, "learning_rate": 9.53679678258603e-05, "loss": 0.0206, "step": 18120 }, { "grad_norm": 0.2192363142967224, "learning_rate": 9.536101488902966e-05, "loss": 0.0171, "step": 18130 }, { "grad_norm": 0.20516036450862885, "learning_rate": 9.535405699158663e-05, "loss": 0.014, "step": 18140 }, { "grad_norm": 0.1857737898826599, "learning_rate": 9.53470941342921e-05, "loss": 0.0131, "step": 18150 }, { "grad_norm": 0.19288784265518188, "learning_rate": 9.534012631790756e-05, "loss": 0.0151, "step": 18160 }, { "grad_norm": 0.18574278056621552, "learning_rate": 9.533315354319494e-05, "loss": 0.0141, "step": 18170 }, { "grad_norm": 0.24296866357326508, "learning_rate": 9.532617581091682e-05, "loss": 0.0155, "step": 18180 }, { "grad_norm": 0.2170851230621338, "learning_rate": 9.531919312183629e-05, "loss": 0.0173, "step": 18190 }, { "grad_norm": 0.1971776783466339, "learning_rate": 9.531220547671688e-05, "loss": 0.0183, "step": 18200 }, { "grad_norm": 0.19252392649650574, "learning_rate": 9.530521287632285e-05, "loss": 0.0153, "step": 18210 }, { "grad_norm": 0.22748300433158875, "learning_rate": 9.529821532141884e-05, "loss": 0.0163, "step": 18220 }, { "grad_norm": 0.22164738178253174, "learning_rate": 9.52912128127701e-05, "loss": 0.0192, "step": 18230 }, { "grad_norm": 0.22258062660694122, "learning_rate": 9.528420535114244e-05, "loss": 0.0202, "step": 18240 }, { "grad_norm": 0.2542295455932617, "learning_rate": 9.527719293730215e-05, "loss": 0.0183, "step": 18250 }, { "grad_norm": 0.22104863822460175, "learning_rate": 9.527017557201611e-05, "loss": 0.0173, "step": 18260 }, { "grad_norm": 0.19266727566719055, "learning_rate": 9.526315325605176e-05, "loss": 0.0188, "step": 18270 }, { "grad_norm": 0.23316746950149536, "learning_rate": 9.525612599017699e-05, "loss": 0.0179, "step": 18280 }, { "grad_norm": 0.21351258456707, "learning_rate": 9.524909377516033e-05, "loss": 0.0157, "step": 18290 }, { "grad_norm": 0.2008112221956253, "learning_rate": 9.524205661177081e-05, "loss": 0.0142, "step": 18300 }, { "grad_norm": 0.22888167202472687, "learning_rate": 9.523501450077801e-05, "loss": 0.0203, "step": 18310 }, { "grad_norm": 0.2037426084280014, "learning_rate": 9.522796744295202e-05, "loss": 0.0152, "step": 18320 }, { "grad_norm": 0.22750163078308105, "learning_rate": 9.522091543906352e-05, "loss": 0.0214, "step": 18330 }, { "grad_norm": 0.2916030287742615, "learning_rate": 9.521385848988369e-05, "loss": 0.017, "step": 18340 }, { "grad_norm": 0.2910306453704834, "learning_rate": 9.520679659618428e-05, "loss": 0.017, "step": 18350 }, { "grad_norm": 0.21370837092399597, "learning_rate": 9.519972975873754e-05, "loss": 0.0154, "step": 18360 }, { "grad_norm": 0.2161671370267868, "learning_rate": 9.519265797831633e-05, "loss": 0.0149, "step": 18370 }, { "grad_norm": 0.1868269443511963, "learning_rate": 9.518558125569399e-05, "loss": 0.016, "step": 18380 }, { "grad_norm": 0.22981759905815125, "learning_rate": 9.517849959164442e-05, "loss": 0.0136, "step": 18390 }, { "grad_norm": 0.19947344064712524, "learning_rate": 9.517141298694205e-05, "loss": 0.0143, "step": 18400 }, { "grad_norm": 0.282819926738739, "learning_rate": 9.516432144236188e-05, "loss": 0.0163, "step": 18410 }, { "grad_norm": 0.17747211456298828, "learning_rate": 9.515722495867941e-05, "loss": 0.0153, "step": 18420 }, { "grad_norm": 0.2135978639125824, "learning_rate": 9.515012353667072e-05, "loss": 0.0178, "step": 18430 }, { "grad_norm": 0.2361542284488678, "learning_rate": 9.51430171771124e-05, "loss": 0.0186, "step": 18440 }, { "grad_norm": 0.21878091990947723, "learning_rate": 9.513590588078159e-05, "loss": 0.0179, "step": 18450 }, { "grad_norm": 0.20214253664016724, "learning_rate": 9.512878964845597e-05, "loss": 0.0143, "step": 18460 }, { "grad_norm": 0.22224675118923187, "learning_rate": 9.512166848091377e-05, "loss": 0.0142, "step": 18470 }, { "grad_norm": 0.18366849422454834, "learning_rate": 9.511454237893376e-05, "loss": 0.0153, "step": 18480 }, { "grad_norm": 0.23773689568042755, "learning_rate": 9.51074113432952e-05, "loss": 0.02, "step": 18490 }, { "grad_norm": 0.30152764916419983, "learning_rate": 9.510027537477797e-05, "loss": 0.0167, "step": 18500 }, { "grad_norm": 0.21317671239376068, "learning_rate": 9.509313447416242e-05, "loss": 0.0167, "step": 18510 }, { "grad_norm": 0.24111266434192657, "learning_rate": 9.508598864222949e-05, "loss": 0.0174, "step": 18520 }, { "grad_norm": 0.18121054768562317, "learning_rate": 9.507883787976062e-05, "loss": 0.0163, "step": 18530 }, { "grad_norm": 0.26030927896499634, "learning_rate": 9.507168218753781e-05, "loss": 0.018, "step": 18540 }, { "grad_norm": 0.20954324305057526, "learning_rate": 9.506452156634362e-05, "loss": 0.0176, "step": 18550 }, { "grad_norm": 0.20720745623111725, "learning_rate": 9.505735601696109e-05, "loss": 0.0156, "step": 18560 }, { "grad_norm": 0.1853705793619156, "learning_rate": 9.505018554017385e-05, "loss": 0.0143, "step": 18570 }, { "grad_norm": 0.1481938511133194, "learning_rate": 9.504301013676604e-05, "loss": 0.0136, "step": 18580 }, { "grad_norm": 0.2699800729751587, "learning_rate": 9.503582980752238e-05, "loss": 0.0151, "step": 18590 }, { "grad_norm": 0.2529743015766144, "learning_rate": 9.502864455322809e-05, "loss": 0.0161, "step": 18600 }, { "grad_norm": 0.18350544571876526, "learning_rate": 9.502145437466891e-05, "loss": 0.0179, "step": 18610 }, { "grad_norm": 0.24194180965423584, "learning_rate": 9.501425927263116e-05, "loss": 0.0192, "step": 18620 }, { "grad_norm": 0.23897700011730194, "learning_rate": 9.500705924790172e-05, "loss": 0.0182, "step": 18630 }, { "grad_norm": 0.182654470205307, "learning_rate": 9.499985430126794e-05, "loss": 0.0143, "step": 18640 }, { "grad_norm": 0.18895353376865387, "learning_rate": 9.499264443351775e-05, "loss": 0.0152, "step": 18650 }, { "grad_norm": 0.22687247395515442, "learning_rate": 9.498542964543961e-05, "loss": 0.016, "step": 18660 }, { "grad_norm": 0.16348782181739807, "learning_rate": 9.497820993782252e-05, "loss": 0.0147, "step": 18670 }, { "grad_norm": 0.22019819915294647, "learning_rate": 9.497098531145601e-05, "loss": 0.0179, "step": 18680 }, { "grad_norm": 0.2363109141588211, "learning_rate": 9.496375576713017e-05, "loss": 0.0158, "step": 18690 }, { "grad_norm": 0.225972980260849, "learning_rate": 9.49565213056356e-05, "loss": 0.0159, "step": 18700 }, { "grad_norm": 0.19201330840587616, "learning_rate": 9.494928192776342e-05, "loss": 0.0133, "step": 18710 }, { "grad_norm": 0.18816930055618286, "learning_rate": 9.494203763430538e-05, "loss": 0.0139, "step": 18720 }, { "grad_norm": 0.19977417588233948, "learning_rate": 9.493478842605366e-05, "loss": 0.0176, "step": 18730 }, { "grad_norm": 0.23868748545646667, "learning_rate": 9.492753430380105e-05, "loss": 0.0159, "step": 18740 }, { "grad_norm": 0.23248443007469177, "learning_rate": 9.492027526834083e-05, "loss": 0.0155, "step": 18750 }, { "grad_norm": 0.19983698427677155, "learning_rate": 9.491301132046684e-05, "loss": 0.0124, "step": 18760 }, { "grad_norm": 0.16191749274730682, "learning_rate": 9.490574246097345e-05, "loss": 0.0135, "step": 18770 }, { "grad_norm": 0.18005695939064026, "learning_rate": 9.48984686906556e-05, "loss": 0.0153, "step": 18780 }, { "grad_norm": 0.16862139105796814, "learning_rate": 9.489119001030871e-05, "loss": 0.0135, "step": 18790 }, { "grad_norm": 0.21963873505592346, "learning_rate": 9.488390642072878e-05, "loss": 0.0156, "step": 18800 }, { "grad_norm": 0.20107601583003998, "learning_rate": 9.48766179227123e-05, "loss": 0.0146, "step": 18810 }, { "grad_norm": 0.2350926548242569, "learning_rate": 9.486932451705636e-05, "loss": 0.0204, "step": 18820 }, { "grad_norm": 0.20641720294952393, "learning_rate": 9.486202620455857e-05, "loss": 0.0164, "step": 18830 }, { "grad_norm": 0.21837429702281952, "learning_rate": 9.485472298601704e-05, "loss": 0.0162, "step": 18840 }, { "grad_norm": 0.20050758123397827, "learning_rate": 9.484741486223043e-05, "loss": 0.0165, "step": 18850 }, { "grad_norm": 0.16409048438072205, "learning_rate": 9.484010183399797e-05, "loss": 0.014, "step": 18860 }, { "grad_norm": 0.20396959781646729, "learning_rate": 9.483278390211938e-05, "loss": 0.014, "step": 18870 }, { "grad_norm": 0.15880604088306427, "learning_rate": 9.482546106739496e-05, "loss": 0.0164, "step": 18880 }, { "grad_norm": 0.18916305899620056, "learning_rate": 9.48181333306255e-05, "loss": 0.0188, "step": 18890 }, { "grad_norm": 0.22724679112434387, "learning_rate": 9.481080069261237e-05, "loss": 0.015, "step": 18900 }, { "grad_norm": 0.2393253594636917, "learning_rate": 9.480346315415745e-05, "loss": 0.0173, "step": 18910 }, { "grad_norm": 0.19321545958518982, "learning_rate": 9.479612071606314e-05, "loss": 0.0159, "step": 18920 }, { "grad_norm": 0.19760742783546448, "learning_rate": 9.478877337913244e-05, "loss": 0.0178, "step": 18930 }, { "grad_norm": 0.19400519132614136, "learning_rate": 9.478142114416881e-05, "loss": 0.0162, "step": 18940 }, { "grad_norm": 0.1889536827802658, "learning_rate": 9.47740640119763e-05, "loss": 0.0173, "step": 18950 }, { "grad_norm": 0.2211039513349533, "learning_rate": 9.476670198335947e-05, "loss": 0.0186, "step": 18960 }, { "grad_norm": 0.21054589748382568, "learning_rate": 9.47593350591234e-05, "loss": 0.0159, "step": 18970 }, { "grad_norm": 0.20402823388576508, "learning_rate": 9.475196324007376e-05, "loss": 0.0158, "step": 18980 }, { "grad_norm": 0.19837036728858948, "learning_rate": 9.474458652701669e-05, "loss": 0.0174, "step": 18990 }, { "grad_norm": 0.16113927960395813, "learning_rate": 9.473720492075892e-05, "loss": 0.0171, "step": 19000 }, { "grad_norm": 0.21207502484321594, "learning_rate": 9.472981842210768e-05, "loss": 0.0154, "step": 19010 }, { "grad_norm": 0.17790986597537994, "learning_rate": 9.472242703187074e-05, "loss": 0.0155, "step": 19020 }, { "grad_norm": 0.2080545723438263, "learning_rate": 9.471503075085643e-05, "loss": 0.0178, "step": 19030 }, { "grad_norm": 0.22169654071331024, "learning_rate": 9.470762957987359e-05, "loss": 0.0142, "step": 19040 }, { "grad_norm": 0.22068770229816437, "learning_rate": 9.470022351973158e-05, "loss": 0.0157, "step": 19050 }, { "grad_norm": 0.1869010031223297, "learning_rate": 9.469281257124034e-05, "loss": 0.0144, "step": 19060 }, { "grad_norm": 0.2188190519809723, "learning_rate": 9.46853967352103e-05, "loss": 0.0166, "step": 19070 }, { "grad_norm": 0.18516574800014496, "learning_rate": 9.467797601245246e-05, "loss": 0.0197, "step": 19080 }, { "grad_norm": 0.19830253720283508, "learning_rate": 9.467055040377834e-05, "loss": 0.0173, "step": 19090 }, { "grad_norm": 0.19412194192409515, "learning_rate": 9.466311990999999e-05, "loss": 0.0157, "step": 19100 }, { "grad_norm": 0.16117912530899048, "learning_rate": 9.465568453193e-05, "loss": 0.0156, "step": 19110 }, { "grad_norm": 0.22441208362579346, "learning_rate": 9.464824427038148e-05, "loss": 0.0174, "step": 19120 }, { "grad_norm": 0.19907067716121674, "learning_rate": 9.46407991261681e-05, "loss": 0.0147, "step": 19130 }, { "grad_norm": 0.15545716881752014, "learning_rate": 9.463334910010404e-05, "loss": 0.0146, "step": 19140 }, { "grad_norm": 0.1945619136095047, "learning_rate": 9.462589419300403e-05, "loss": 0.0158, "step": 19150 }, { "grad_norm": 0.18326033651828766, "learning_rate": 9.461843440568333e-05, "loss": 0.0135, "step": 19160 }, { "grad_norm": 0.21615061163902283, "learning_rate": 9.461096973895773e-05, "loss": 0.0152, "step": 19170 }, { "grad_norm": 0.20368073880672455, "learning_rate": 9.460350019364355e-05, "loss": 0.0189, "step": 19180 }, { "grad_norm": 0.1861567199230194, "learning_rate": 9.459602577055764e-05, "loss": 0.0172, "step": 19190 }, { "grad_norm": 0.23974572122097015, "learning_rate": 9.45885464705174e-05, "loss": 0.0156, "step": 19200 }, { "grad_norm": 0.20649509131908417, "learning_rate": 9.458106229434076e-05, "loss": 0.0138, "step": 19210 }, { "grad_norm": 0.17852559685707092, "learning_rate": 9.457357324284617e-05, "loss": 0.0131, "step": 19220 }, { "grad_norm": 0.19758810102939606, "learning_rate": 9.456607931685262e-05, "loss": 0.0148, "step": 19230 }, { "grad_norm": 0.22621938586235046, "learning_rate": 9.455858051717965e-05, "loss": 0.0143, "step": 19240 }, { "grad_norm": 0.18433977663516998, "learning_rate": 9.45510768446473e-05, "loss": 0.018, "step": 19250 }, { "grad_norm": 0.18131697177886963, "learning_rate": 9.454356830007618e-05, "loss": 0.0132, "step": 19260 }, { "grad_norm": 0.21862757205963135, "learning_rate": 9.45360548842874e-05, "loss": 0.0147, "step": 19270 }, { "grad_norm": 0.22827917337417603, "learning_rate": 9.452853659810261e-05, "loss": 0.0169, "step": 19280 }, { "grad_norm": 0.19332288205623627, "learning_rate": 9.452101344234401e-05, "loss": 0.0162, "step": 19290 }, { "grad_norm": 0.19470186531543732, "learning_rate": 9.451348541783431e-05, "loss": 0.0122, "step": 19300 }, { "grad_norm": 0.1577935814857483, "learning_rate": 9.450595252539678e-05, "loss": 0.0132, "step": 19310 }, { "grad_norm": 0.17375445365905762, "learning_rate": 9.449841476585518e-05, "loss": 0.015, "step": 19320 }, { "grad_norm": 0.16900065541267395, "learning_rate": 9.449087214003384e-05, "loss": 0.0179, "step": 19330 }, { "grad_norm": 0.15793628990650177, "learning_rate": 9.448332464875765e-05, "loss": 0.0166, "step": 19340 }, { "grad_norm": 0.15997256338596344, "learning_rate": 9.447577229285192e-05, "loss": 0.0138, "step": 19350 }, { "grad_norm": 0.29404953122138977, "learning_rate": 9.446821507314261e-05, "loss": 0.0156, "step": 19360 }, { "grad_norm": 0.31125929951667786, "learning_rate": 9.446065299045617e-05, "loss": 0.0149, "step": 19370 }, { "grad_norm": 0.22690996527671814, "learning_rate": 9.445308604561955e-05, "loss": 0.0157, "step": 19380 }, { "grad_norm": 0.18413610756397247, "learning_rate": 9.444551423946028e-05, "loss": 0.0181, "step": 19390 }, { "grad_norm": 0.20654913783073425, "learning_rate": 9.443793757280638e-05, "loss": 0.0168, "step": 19400 }, { "grad_norm": 0.19988474249839783, "learning_rate": 9.443035604648646e-05, "loss": 0.0184, "step": 19410 }, { "grad_norm": 0.18763349950313568, "learning_rate": 9.44227696613296e-05, "loss": 0.0136, "step": 19420 }, { "grad_norm": 0.20153270661830902, "learning_rate": 9.441517841816542e-05, "loss": 0.0146, "step": 19430 }, { "grad_norm": 0.20342783629894257, "learning_rate": 9.440758231782413e-05, "loss": 0.0147, "step": 19440 }, { "grad_norm": 0.2456694096326828, "learning_rate": 9.439998136113639e-05, "loss": 0.0157, "step": 19450 }, { "grad_norm": 0.2628188729286194, "learning_rate": 9.439237554893344e-05, "loss": 0.0138, "step": 19460 }, { "grad_norm": 0.1973292976617813, "learning_rate": 9.438476488204705e-05, "loss": 0.0183, "step": 19470 }, { "grad_norm": 0.19474337995052338, "learning_rate": 9.43771493613095e-05, "loss": 0.017, "step": 19480 }, { "grad_norm": 0.17243431508541107, "learning_rate": 9.436952898755362e-05, "loss": 0.0142, "step": 19490 }, { "grad_norm": 0.21091721951961517, "learning_rate": 9.436190376161276e-05, "loss": 0.0123, "step": 19500 }, { "grad_norm": 0.20655761659145355, "learning_rate": 9.43542736843208e-05, "loss": 0.0183, "step": 19510 }, { "grad_norm": 0.22108334302902222, "learning_rate": 9.434663875651216e-05, "loss": 0.0144, "step": 19520 }, { "grad_norm": 0.21188423037528992, "learning_rate": 9.433899897902177e-05, "loss": 0.0158, "step": 19530 }, { "grad_norm": 0.19831833243370056, "learning_rate": 9.433135435268511e-05, "loss": 0.0137, "step": 19540 }, { "grad_norm": 0.2434650957584381, "learning_rate": 9.432370487833819e-05, "loss": 0.017, "step": 19550 }, { "grad_norm": 0.16799944639205933, "learning_rate": 9.431605055681756e-05, "loss": 0.0153, "step": 19560 }, { "grad_norm": 0.1691613346338272, "learning_rate": 9.430839138896026e-05, "loss": 0.0167, "step": 19570 }, { "grad_norm": 0.19002310931682587, "learning_rate": 9.43007273756039e-05, "loss": 0.0125, "step": 19580 }, { "grad_norm": 0.24150164425373077, "learning_rate": 9.429305851758658e-05, "loss": 0.0169, "step": 19590 }, { "grad_norm": 0.2684189975261688, "learning_rate": 9.428538481574699e-05, "loss": 0.0188, "step": 19600 }, { "grad_norm": 0.1981664001941681, "learning_rate": 9.42777062709243e-05, "loss": 0.0155, "step": 19610 }, { "grad_norm": 0.2320268154144287, "learning_rate": 9.427002288395821e-05, "loss": 0.0158, "step": 19620 }, { "grad_norm": 0.24814385175704956, "learning_rate": 9.426233465568898e-05, "loss": 0.0138, "step": 19630 }, { "grad_norm": 0.20348390936851501, "learning_rate": 9.42546415869574e-05, "loss": 0.0173, "step": 19640 }, { "grad_norm": 0.19342976808547974, "learning_rate": 9.424694367860473e-05, "loss": 0.0147, "step": 19650 }, { "grad_norm": 0.19766613841056824, "learning_rate": 9.423924093147284e-05, "loss": 0.014, "step": 19660 }, { "grad_norm": 0.2004372477531433, "learning_rate": 9.423153334640407e-05, "loss": 0.0149, "step": 19670 }, { "grad_norm": 0.18934035301208496, "learning_rate": 9.42238209242413e-05, "loss": 0.0167, "step": 19680 }, { "grad_norm": 0.18607228994369507, "learning_rate": 9.421610366582798e-05, "loss": 0.0139, "step": 19690 }, { "grad_norm": 0.24335505068302155, "learning_rate": 9.420838157200803e-05, "loss": 0.0128, "step": 19700 }, { "grad_norm": 0.17790640890598297, "learning_rate": 9.420065464362594e-05, "loss": 0.0163, "step": 19710 }, { "grad_norm": 0.1934637725353241, "learning_rate": 9.419292288152673e-05, "loss": 0.0149, "step": 19720 }, { "grad_norm": 0.17496702075004578, "learning_rate": 9.418518628655588e-05, "loss": 0.018, "step": 19730 }, { "grad_norm": 0.19295130670070648, "learning_rate": 9.417744485955951e-05, "loss": 0.0163, "step": 19740 }, { "grad_norm": 0.2125600278377533, "learning_rate": 9.41696986013842e-05, "loss": 0.0141, "step": 19750 }, { "grad_norm": 0.19057458639144897, "learning_rate": 9.416194751287705e-05, "loss": 0.0131, "step": 19760 }, { "grad_norm": 0.2559678554534912, "learning_rate": 9.415419159488572e-05, "loss": 0.0168, "step": 19770 }, { "grad_norm": 0.2723550796508789, "learning_rate": 9.414643084825837e-05, "loss": 0.0148, "step": 19780 }, { "grad_norm": 0.1745573729276657, "learning_rate": 9.413866527384372e-05, "loss": 0.0145, "step": 19790 }, { "grad_norm": 0.19071006774902344, "learning_rate": 9.4130894872491e-05, "loss": 0.0121, "step": 19800 }, { "grad_norm": 0.21287497878074646, "learning_rate": 9.412311964504998e-05, "loss": 0.0144, "step": 19810 }, { "grad_norm": 0.21783508360385895, "learning_rate": 9.411533959237091e-05, "loss": 0.0154, "step": 19820 }, { "grad_norm": 0.22634436190128326, "learning_rate": 9.410755471530464e-05, "loss": 0.0167, "step": 19830 }, { "grad_norm": 0.19493378698825836, "learning_rate": 9.40997650147025e-05, "loss": 0.0155, "step": 19840 }, { "grad_norm": 0.1476636379957199, "learning_rate": 9.409197049141637e-05, "loss": 0.0135, "step": 19850 }, { "grad_norm": 0.2119777649641037, "learning_rate": 9.408417114629863e-05, "loss": 0.0171, "step": 19860 }, { "grad_norm": 0.18118540942668915, "learning_rate": 9.40763669802022e-05, "loss": 0.0181, "step": 19870 }, { "grad_norm": 0.2417001575231552, "learning_rate": 9.406855799398056e-05, "loss": 0.0176, "step": 19880 }, { "grad_norm": 0.19013667106628418, "learning_rate": 9.406074418848767e-05, "loss": 0.0156, "step": 19890 }, { "grad_norm": 0.17778362333774567, "learning_rate": 9.405292556457805e-05, "loss": 0.0163, "step": 19900 }, { "grad_norm": 0.23389503359794617, "learning_rate": 9.404510212310671e-05, "loss": 0.0188, "step": 19910 }, { "grad_norm": 0.25047048926353455, "learning_rate": 9.403727386492924e-05, "loss": 0.0155, "step": 19920 }, { "grad_norm": 0.21785365045070648, "learning_rate": 9.40294407909017e-05, "loss": 0.0149, "step": 19930 }, { "grad_norm": 0.211991086602211, "learning_rate": 9.40216029018807e-05, "loss": 0.0163, "step": 19940 }, { "grad_norm": 0.18550288677215576, "learning_rate": 9.401376019872338e-05, "loss": 0.0128, "step": 19950 }, { "grad_norm": 0.19449831545352936, "learning_rate": 9.400591268228746e-05, "loss": 0.015, "step": 19960 }, { "grad_norm": 0.19786128401756287, "learning_rate": 9.399806035343106e-05, "loss": 0.0193, "step": 19970 }, { "grad_norm": 0.180292010307312, "learning_rate": 9.399020321301294e-05, "loss": 0.0143, "step": 19980 }, { "grad_norm": 0.18164955079555511, "learning_rate": 9.398234126189234e-05, "loss": 0.0142, "step": 19990 }, { "grad_norm": 0.18356813490390778, "learning_rate": 9.397447450092902e-05, "loss": 0.0164, "step": 20000 }, { "grad_norm": 0.19531887769699097, "learning_rate": 9.39666029309833e-05, "loss": 0.0189, "step": 20010 }, { "grad_norm": 0.19588615000247955, "learning_rate": 9.395872655291596e-05, "loss": 0.0155, "step": 20020 }, { "grad_norm": 0.19638456404209137, "learning_rate": 9.395084536758838e-05, "loss": 0.0151, "step": 20030 }, { "grad_norm": 0.20911385118961334, "learning_rate": 9.394295937586243e-05, "loss": 0.0144, "step": 20040 }, { "grad_norm": 0.21155162155628204, "learning_rate": 9.393506857860052e-05, "loss": 0.014, "step": 20050 }, { "grad_norm": 0.18010038137435913, "learning_rate": 9.392717297666555e-05, "loss": 0.0152, "step": 20060 }, { "grad_norm": 0.24968458712100983, "learning_rate": 9.391927257092101e-05, "loss": 0.0175, "step": 20070 }, { "grad_norm": 0.2549588084220886, "learning_rate": 9.391136736223085e-05, "loss": 0.0142, "step": 20080 }, { "grad_norm": 0.205298513174057, "learning_rate": 9.390345735145956e-05, "loss": 0.0174, "step": 20090 }, { "grad_norm": 0.2705083191394806, "learning_rate": 9.389554253947219e-05, "loss": 0.0193, "step": 20100 }, { "grad_norm": 0.25074630975723267, "learning_rate": 9.388762292713428e-05, "loss": 0.0201, "step": 20110 }, { "grad_norm": 0.24715974926948547, "learning_rate": 9.38796985153119e-05, "loss": 0.0191, "step": 20120 }, { "grad_norm": 0.23295433819293976, "learning_rate": 9.387176930487169e-05, "loss": 0.0179, "step": 20130 }, { "grad_norm": 0.15685079991817474, "learning_rate": 9.386383529668072e-05, "loss": 0.0155, "step": 20140 }, { "grad_norm": 0.23175013065338135, "learning_rate": 9.385589649160669e-05, "loss": 0.0182, "step": 20150 }, { "grad_norm": 0.24832825362682343, "learning_rate": 9.384795289051775e-05, "loss": 0.0144, "step": 20160 }, { "grad_norm": 0.17265936732292175, "learning_rate": 9.384000449428261e-05, "loss": 0.013, "step": 20170 }, { "grad_norm": 0.18453805148601532, "learning_rate": 9.383205130377048e-05, "loss": 0.0141, "step": 20180 }, { "grad_norm": 0.1785581111907959, "learning_rate": 9.382409331985114e-05, "loss": 0.0164, "step": 20190 }, { "grad_norm": 0.1490326225757599, "learning_rate": 9.381613054339482e-05, "loss": 0.0135, "step": 20200 }, { "grad_norm": 0.18255308270454407, "learning_rate": 9.380816297527235e-05, "loss": 0.015, "step": 20210 }, { "grad_norm": 0.22930574417114258, "learning_rate": 9.380019061635506e-05, "loss": 0.0185, "step": 20220 }, { "grad_norm": 0.2228304147720337, "learning_rate": 9.379221346751474e-05, "loss": 0.013, "step": 20230 }, { "grad_norm": 0.24514560401439667, "learning_rate": 9.378423152962382e-05, "loss": 0.0146, "step": 20240 }, { "grad_norm": 0.20312680304050446, "learning_rate": 9.377624480355517e-05, "loss": 0.0171, "step": 20250 }, { "grad_norm": 0.23038558661937714, "learning_rate": 9.376825329018219e-05, "loss": 0.0176, "step": 20260 }, { "grad_norm": 0.2041732221841812, "learning_rate": 9.376025699037884e-05, "loss": 0.0139, "step": 20270 }, { "grad_norm": 0.22920949757099152, "learning_rate": 9.37522559050196e-05, "loss": 0.0148, "step": 20280 }, { "grad_norm": 0.18200257420539856, "learning_rate": 9.37442500349794e-05, "loss": 0.0155, "step": 20290 }, { "grad_norm": 0.22074410319328308, "learning_rate": 9.373623938113381e-05, "loss": 0.0147, "step": 20300 }, { "grad_norm": 0.2000265270471573, "learning_rate": 9.372822394435883e-05, "loss": 0.0191, "step": 20310 }, { "grad_norm": 0.20545731484889984, "learning_rate": 9.372020372553102e-05, "loss": 0.0158, "step": 20320 }, { "grad_norm": 0.18711687624454498, "learning_rate": 9.371217872552746e-05, "loss": 0.0154, "step": 20330 }, { "grad_norm": 0.21185661852359772, "learning_rate": 9.370414894522576e-05, "loss": 0.0147, "step": 20340 }, { "grad_norm": 0.21273744106292725, "learning_rate": 9.369611438550406e-05, "loss": 0.0172, "step": 20350 }, { "grad_norm": 0.268685519695282, "learning_rate": 9.368807504724095e-05, "loss": 0.0158, "step": 20360 }, { "grad_norm": 0.22099606692790985, "learning_rate": 9.368003093131565e-05, "loss": 0.0155, "step": 20370 }, { "grad_norm": 0.20086461305618286, "learning_rate": 9.367198203860785e-05, "loss": 0.0177, "step": 20380 }, { "grad_norm": 0.19656991958618164, "learning_rate": 9.366392836999774e-05, "loss": 0.0171, "step": 20390 }, { "grad_norm": 0.16772672533988953, "learning_rate": 9.365586992636607e-05, "loss": 0.0162, "step": 20400 }, { "grad_norm": 0.2347378432750702, "learning_rate": 9.364780670859412e-05, "loss": 0.0156, "step": 20410 }, { "grad_norm": 0.17372795939445496, "learning_rate": 9.363973871756364e-05, "loss": 0.0114, "step": 20420 }, { "grad_norm": 0.20153731107711792, "learning_rate": 9.363166595415696e-05, "loss": 0.0116, "step": 20430 }, { "grad_norm": 0.16642212867736816, "learning_rate": 9.362358841925686e-05, "loss": 0.0148, "step": 20440 }, { "grad_norm": 0.21389207243919373, "learning_rate": 9.361550611374674e-05, "loss": 0.0134, "step": 20450 }, { "grad_norm": 0.18293426930904388, "learning_rate": 9.360741903851043e-05, "loss": 0.0155, "step": 20460 }, { "grad_norm": 0.23167620599269867, "learning_rate": 9.359932719443236e-05, "loss": 0.0151, "step": 20470 }, { "grad_norm": 0.19713075459003448, "learning_rate": 9.35912305823974e-05, "loss": 0.016, "step": 20480 }, { "grad_norm": 0.22636091709136963, "learning_rate": 9.358312920329101e-05, "loss": 0.0186, "step": 20490 }, { "grad_norm": 0.17579159140586853, "learning_rate": 9.357502305799914e-05, "loss": 0.0159, "step": 20500 }, { "grad_norm": 0.19942916929721832, "learning_rate": 9.356691214740824e-05, "loss": 0.0129, "step": 20510 }, { "grad_norm": 0.21241244673728943, "learning_rate": 9.355879647240535e-05, "loss": 0.0189, "step": 20520 }, { "grad_norm": 0.17079728841781616, "learning_rate": 9.355067603387798e-05, "loss": 0.0163, "step": 20530 }, { "grad_norm": 0.1638360619544983, "learning_rate": 9.354255083271412e-05, "loss": 0.0123, "step": 20540 }, { "grad_norm": 0.16089729964733124, "learning_rate": 9.353442086980239e-05, "loss": 0.015, "step": 20550 }, { "grad_norm": 0.2573931813240051, "learning_rate": 9.352628614603185e-05, "loss": 0.0177, "step": 20560 }, { "grad_norm": 0.1977410465478897, "learning_rate": 9.351814666229209e-05, "loss": 0.0154, "step": 20570 }, { "grad_norm": 0.24254480004310608, "learning_rate": 9.351000241947324e-05, "loss": 0.0162, "step": 20580 }, { "grad_norm": 0.2581315338611603, "learning_rate": 9.350185341846594e-05, "loss": 0.0156, "step": 20590 }, { "grad_norm": 0.16455020010471344, "learning_rate": 9.349369966016134e-05, "loss": 0.0128, "step": 20600 }, { "grad_norm": 0.17147524654865265, "learning_rate": 9.348554114545117e-05, "loss": 0.0146, "step": 20610 }, { "grad_norm": 0.21410810947418213, "learning_rate": 9.347737787522758e-05, "loss": 0.0125, "step": 20620 }, { "grad_norm": 0.20137302577495575, "learning_rate": 9.346920985038332e-05, "loss": 0.0158, "step": 20630 }, { "grad_norm": 0.1905655413866043, "learning_rate": 9.346103707181162e-05, "loss": 0.015, "step": 20640 }, { "grad_norm": 0.24339328706264496, "learning_rate": 9.345285954040626e-05, "loss": 0.0172, "step": 20650 }, { "grad_norm": 0.21797725558280945, "learning_rate": 9.34446772570615e-05, "loss": 0.0178, "step": 20660 }, { "grad_norm": 0.21224215626716614, "learning_rate": 9.343649022267214e-05, "loss": 0.0167, "step": 20670 }, { "grad_norm": 0.2046021819114685, "learning_rate": 9.342829843813353e-05, "loss": 0.0159, "step": 20680 }, { "grad_norm": 0.23217086493968964, "learning_rate": 9.342010190434149e-05, "loss": 0.0153, "step": 20690 }, { "grad_norm": 0.19698898494243622, "learning_rate": 9.34119006221924e-05, "loss": 0.0163, "step": 20700 }, { "grad_norm": 0.22827297449111938, "learning_rate": 9.340369459258313e-05, "loss": 0.0152, "step": 20710 }, { "grad_norm": 0.2262742966413498, "learning_rate": 9.339548381641106e-05, "loss": 0.0154, "step": 20720 }, { "grad_norm": 0.2375992238521576, "learning_rate": 9.338726829457413e-05, "loss": 0.0165, "step": 20730 }, { "grad_norm": 0.21336597204208374, "learning_rate": 9.337904802797078e-05, "loss": 0.0185, "step": 20740 }, { "grad_norm": 0.24220816791057587, "learning_rate": 9.337082301749993e-05, "loss": 0.0155, "step": 20750 }, { "grad_norm": 0.16928242146968842, "learning_rate": 9.336259326406109e-05, "loss": 0.0168, "step": 20760 }, { "grad_norm": 0.2963846027851105, "learning_rate": 9.335435876855427e-05, "loss": 0.0188, "step": 20770 }, { "grad_norm": 0.1616017073392868, "learning_rate": 9.334611953187994e-05, "loss": 0.0136, "step": 20780 }, { "grad_norm": 0.22472447156906128, "learning_rate": 9.333787555493914e-05, "loss": 0.0151, "step": 20790 }, { "grad_norm": 0.2534206211566925, "learning_rate": 9.332962683863345e-05, "loss": 0.0148, "step": 20800 }, { "grad_norm": 0.2196665108203888, "learning_rate": 9.332137338386489e-05, "loss": 0.0139, "step": 20810 }, { "grad_norm": 0.2135997712612152, "learning_rate": 9.33131151915361e-05, "loss": 0.0145, "step": 20820 }, { "grad_norm": 0.21457506716251373, "learning_rate": 9.330485226255012e-05, "loss": 0.0157, "step": 20830 }, { "grad_norm": 0.1505342423915863, "learning_rate": 9.329658459781061e-05, "loss": 0.0134, "step": 20840 }, { "grad_norm": 0.18944811820983887, "learning_rate": 9.328831219822172e-05, "loss": 0.0141, "step": 20850 }, { "grad_norm": 0.19646477699279785, "learning_rate": 9.328003506468808e-05, "loss": 0.0153, "step": 20860 }, { "grad_norm": 0.17291903495788574, "learning_rate": 9.327175319811488e-05, "loss": 0.0134, "step": 20870 }, { "grad_norm": 0.22395895421504974, "learning_rate": 9.326346659940781e-05, "loss": 0.015, "step": 20880 }, { "grad_norm": 0.1860131323337555, "learning_rate": 9.325517526947308e-05, "loss": 0.0134, "step": 20890 }, { "grad_norm": 0.2096748948097229, "learning_rate": 9.32468792092174e-05, "loss": 0.0118, "step": 20900 }, { "grad_norm": 0.2094639390707016, "learning_rate": 9.323857841954803e-05, "loss": 0.0144, "step": 20910 }, { "grad_norm": 0.19979602098464966, "learning_rate": 9.323027290137276e-05, "loss": 0.0139, "step": 20920 }, { "grad_norm": 0.18379861116409302, "learning_rate": 9.322196265559981e-05, "loss": 0.014, "step": 20930 }, { "grad_norm": 0.2610354423522949, "learning_rate": 9.321364768313803e-05, "loss": 0.0172, "step": 20940 }, { "grad_norm": 0.24854528903961182, "learning_rate": 9.32053279848967e-05, "loss": 0.0126, "step": 20950 }, { "grad_norm": 0.23370905220508575, "learning_rate": 9.319700356178567e-05, "loss": 0.0142, "step": 20960 }, { "grad_norm": 0.17265941202640533, "learning_rate": 9.318867441471527e-05, "loss": 0.0137, "step": 20970 }, { "grad_norm": 0.17694184184074402, "learning_rate": 9.318034054459637e-05, "loss": 0.0146, "step": 20980 }, { "grad_norm": 0.2054661363363266, "learning_rate": 9.317200195234034e-05, "loss": 0.0136, "step": 20990 }, { "grad_norm": 0.17624884843826294, "learning_rate": 9.316365863885909e-05, "loss": 0.0173, "step": 21000 }, { "grad_norm": 0.1857154816389084, "learning_rate": 9.315531060506502e-05, "loss": 0.0142, "step": 21010 }, { "grad_norm": 0.16938196122646332, "learning_rate": 9.314695785187108e-05, "loss": 0.0162, "step": 21020 }, { "grad_norm": 0.1530320942401886, "learning_rate": 9.313860038019069e-05, "loss": 0.0137, "step": 21030 }, { "grad_norm": 0.22349661588668823, "learning_rate": 9.313023819093782e-05, "loss": 0.0162, "step": 21040 }, { "grad_norm": 0.24431730806827545, "learning_rate": 9.312187128502695e-05, "loss": 0.0189, "step": 21050 }, { "grad_norm": 0.19086843729019165, "learning_rate": 9.311349966337307e-05, "loss": 0.0136, "step": 21060 }, { "grad_norm": 0.19689813256263733, "learning_rate": 9.310512332689169e-05, "loss": 0.0178, "step": 21070 }, { "grad_norm": 0.1752699464559555, "learning_rate": 9.309674227649883e-05, "loss": 0.0154, "step": 21080 }, { "grad_norm": 0.19587618112564087, "learning_rate": 9.308835651311103e-05, "loss": 0.0153, "step": 21090 }, { "grad_norm": 0.24794311821460724, "learning_rate": 9.307996603764533e-05, "loss": 0.0143, "step": 21100 }, { "grad_norm": 0.18624120950698853, "learning_rate": 9.307157085101932e-05, "loss": 0.0177, "step": 21110 }, { "grad_norm": 0.20856277644634247, "learning_rate": 9.306317095415109e-05, "loss": 0.0138, "step": 21120 }, { "grad_norm": 0.2428787797689438, "learning_rate": 9.305476634795922e-05, "loss": 0.0167, "step": 21130 }, { "grad_norm": 0.21069100499153137, "learning_rate": 9.304635703336284e-05, "loss": 0.0164, "step": 21140 }, { "grad_norm": 0.18209083378314972, "learning_rate": 9.303794301128157e-05, "loss": 0.0161, "step": 21150 }, { "grad_norm": 0.2355855256319046, "learning_rate": 9.302952428263555e-05, "loss": 0.0155, "step": 21160 }, { "grad_norm": 0.18905839323997498, "learning_rate": 9.302110084834545e-05, "loss": 0.0142, "step": 21170 }, { "grad_norm": 0.2240324318408966, "learning_rate": 9.301267270933245e-05, "loss": 0.016, "step": 21180 }, { "grad_norm": 0.2059956192970276, "learning_rate": 9.300423986651823e-05, "loss": 0.0186, "step": 21190 }, { "grad_norm": 0.19397512078285217, "learning_rate": 9.299580232082501e-05, "loss": 0.0156, "step": 21200 }, { "grad_norm": 0.21256372332572937, "learning_rate": 9.298736007317547e-05, "loss": 0.0136, "step": 21210 }, { "grad_norm": 0.23693779110908508, "learning_rate": 9.297891312449288e-05, "loss": 0.0157, "step": 21220 }, { "grad_norm": 0.20214630663394928, "learning_rate": 9.297046147570094e-05, "loss": 0.0154, "step": 21230 }, { "grad_norm": 0.20266342163085938, "learning_rate": 9.296200512772396e-05, "loss": 0.0152, "step": 21240 }, { "grad_norm": 0.21090462803840637, "learning_rate": 9.295354408148668e-05, "loss": 0.0134, "step": 21250 }, { "grad_norm": 0.1739768534898758, "learning_rate": 9.294507833791441e-05, "loss": 0.0144, "step": 21260 }, { "grad_norm": 0.21435919404029846, "learning_rate": 9.293660789793295e-05, "loss": 0.0164, "step": 21270 }, { "grad_norm": 0.20803195238113403, "learning_rate": 9.292813276246858e-05, "loss": 0.0152, "step": 21280 }, { "grad_norm": 0.22595597803592682, "learning_rate": 9.291965293244816e-05, "loss": 0.016, "step": 21290 }, { "grad_norm": 0.2367512434720993, "learning_rate": 9.291116840879904e-05, "loss": 0.014, "step": 21300 }, { "grad_norm": 0.19994334876537323, "learning_rate": 9.290267919244904e-05, "loss": 0.0179, "step": 21310 }, { "grad_norm": 0.1555165946483612, "learning_rate": 9.289418528432655e-05, "loss": 0.0156, "step": 21320 }, { "grad_norm": 0.18754661083221436, "learning_rate": 9.288568668536045e-05, "loss": 0.0122, "step": 21330 }, { "grad_norm": 0.1828911006450653, "learning_rate": 9.287718339648013e-05, "loss": 0.016, "step": 21340 }, { "grad_norm": 0.2362218052148819, "learning_rate": 9.28686754186155e-05, "loss": 0.0159, "step": 21350 }, { "grad_norm": 0.18860049545764923, "learning_rate": 9.286016275269698e-05, "loss": 0.0142, "step": 21360 }, { "grad_norm": 0.2040890008211136, "learning_rate": 9.285164539965551e-05, "loss": 0.0134, "step": 21370 }, { "grad_norm": 0.2860497534275055, "learning_rate": 9.284312336042251e-05, "loss": 0.0126, "step": 21380 }, { "grad_norm": 0.23619651794433594, "learning_rate": 9.283459663592996e-05, "loss": 0.0163, "step": 21390 }, { "grad_norm": 0.21958675980567932, "learning_rate": 9.282606522711033e-05, "loss": 0.0139, "step": 21400 }, { "grad_norm": 0.19290131330490112, "learning_rate": 9.281752913489657e-05, "loss": 0.0138, "step": 21410 }, { "grad_norm": 0.17522117495536804, "learning_rate": 9.280898836022222e-05, "loss": 0.0145, "step": 21420 }, { "grad_norm": 0.20237737894058228, "learning_rate": 9.280044290402126e-05, "loss": 0.015, "step": 21430 }, { "grad_norm": 0.19449496269226074, "learning_rate": 9.279189276722821e-05, "loss": 0.0167, "step": 21440 }, { "grad_norm": 0.20160919427871704, "learning_rate": 9.278333795077812e-05, "loss": 0.0191, "step": 21450 }, { "grad_norm": 0.17019177973270416, "learning_rate": 9.27747784556065e-05, "loss": 0.0127, "step": 21460 }, { "grad_norm": 0.18113021552562714, "learning_rate": 9.276621428264942e-05, "loss": 0.0122, "step": 21470 }, { "grad_norm": 0.18218807876110077, "learning_rate": 9.275764543284345e-05, "loss": 0.0141, "step": 21480 }, { "grad_norm": 0.17536427080631256, "learning_rate": 9.274907190712566e-05, "loss": 0.0144, "step": 21490 }, { "grad_norm": 0.16895531117916107, "learning_rate": 9.274049370643363e-05, "loss": 0.0155, "step": 21500 }, { "grad_norm": 0.1592005342245102, "learning_rate": 9.273191083170547e-05, "loss": 0.0119, "step": 21510 }, { "grad_norm": 0.21609938144683838, "learning_rate": 9.27233232838798e-05, "loss": 0.0152, "step": 21520 }, { "grad_norm": 0.15201692283153534, "learning_rate": 9.27147310638957e-05, "loss": 0.015, "step": 21530 }, { "grad_norm": 0.20355278253555298, "learning_rate": 9.270613417269286e-05, "loss": 0.0162, "step": 21540 }, { "grad_norm": 0.20280133187770844, "learning_rate": 9.269753261121138e-05, "loss": 0.0142, "step": 21550 }, { "grad_norm": 0.20409053564071655, "learning_rate": 9.268892638039194e-05, "loss": 0.0148, "step": 21560 }, { "grad_norm": 0.18224972486495972, "learning_rate": 9.268031548117569e-05, "loss": 0.0134, "step": 21570 }, { "grad_norm": 0.15948013961315155, "learning_rate": 9.26716999145043e-05, "loss": 0.0175, "step": 21580 }, { "grad_norm": 0.2168949842453003, "learning_rate": 9.266307968131998e-05, "loss": 0.016, "step": 21590 }, { "grad_norm": 0.204420804977417, "learning_rate": 9.26544547825654e-05, "loss": 0.0164, "step": 21600 }, { "grad_norm": 0.21193057298660278, "learning_rate": 9.264582521918376e-05, "loss": 0.0144, "step": 21610 }, { "grad_norm": 0.1996489018201828, "learning_rate": 9.263719099211881e-05, "loss": 0.016, "step": 21620 }, { "grad_norm": 0.16653349995613098, "learning_rate": 9.262855210231476e-05, "loss": 0.0163, "step": 21630 }, { "grad_norm": 0.13833194971084595, "learning_rate": 9.261990855071633e-05, "loss": 0.0151, "step": 21640 }, { "grad_norm": 0.1796727031469345, "learning_rate": 9.261126033826878e-05, "loss": 0.0133, "step": 21650 }, { "grad_norm": 0.22248409688472748, "learning_rate": 9.260260746591786e-05, "loss": 0.0148, "step": 21660 }, { "grad_norm": 0.21603278815746307, "learning_rate": 9.259394993460985e-05, "loss": 0.0155, "step": 21670 }, { "grad_norm": 0.19268816709518433, "learning_rate": 9.258528774529151e-05, "loss": 0.0158, "step": 21680 }, { "grad_norm": 0.174570232629776, "learning_rate": 9.257662089891013e-05, "loss": 0.0158, "step": 21690 }, { "grad_norm": 0.19629022479057312, "learning_rate": 9.25679493964135e-05, "loss": 0.0155, "step": 21700 }, { "grad_norm": 0.1691417694091797, "learning_rate": 9.255927323874994e-05, "loss": 0.0158, "step": 21710 }, { "grad_norm": 0.22220371663570404, "learning_rate": 9.255059242686822e-05, "loss": 0.0207, "step": 21720 }, { "grad_norm": 0.2406976819038391, "learning_rate": 9.254190696171769e-05, "loss": 0.0182, "step": 21730 }, { "grad_norm": 0.25125160813331604, "learning_rate": 9.25332168442482e-05, "loss": 0.0145, "step": 21740 }, { "grad_norm": 0.17512507736682892, "learning_rate": 9.252452207541004e-05, "loss": 0.0144, "step": 21750 }, { "grad_norm": 0.18050198256969452, "learning_rate": 9.251582265615409e-05, "loss": 0.0149, "step": 21760 }, { "grad_norm": 0.17306315898895264, "learning_rate": 9.250711858743169e-05, "loss": 0.0178, "step": 21770 }, { "grad_norm": 0.2820708453655243, "learning_rate": 9.24984098701947e-05, "loss": 0.0147, "step": 21780 }, { "grad_norm": 0.17780010402202606, "learning_rate": 9.248969650539552e-05, "loss": 0.0133, "step": 21790 }, { "grad_norm": 0.25607702136039734, "learning_rate": 9.2480978493987e-05, "loss": 0.014, "step": 21800 }, { "grad_norm": 0.20545744895935059, "learning_rate": 9.247225583692256e-05, "loss": 0.0147, "step": 21810 }, { "grad_norm": 0.24689055979251862, "learning_rate": 9.246352853515607e-05, "loss": 0.0127, "step": 21820 }, { "grad_norm": 0.24009335041046143, "learning_rate": 9.245479658964194e-05, "loss": 0.0141, "step": 21830 }, { "grad_norm": 0.1808711737394333, "learning_rate": 9.244606000133507e-05, "loss": 0.0144, "step": 21840 }, { "grad_norm": 0.20611116290092468, "learning_rate": 9.24373187711909e-05, "loss": 0.0162, "step": 21850 }, { "grad_norm": 0.1742132306098938, "learning_rate": 9.242857290016537e-05, "loss": 0.0144, "step": 21860 }, { "grad_norm": 0.1890120506286621, "learning_rate": 9.241982238921488e-05, "loss": 0.0155, "step": 21870 }, { "grad_norm": 0.2176998406648636, "learning_rate": 9.24110672392964e-05, "loss": 0.0138, "step": 21880 }, { "grad_norm": 0.1679946482181549, "learning_rate": 9.240230745136737e-05, "loss": 0.015, "step": 21890 }, { "grad_norm": 0.21390338242053986, "learning_rate": 9.239354302638575e-05, "loss": 0.0153, "step": 21900 }, { "grad_norm": 0.2027822583913803, "learning_rate": 9.238477396531e-05, "loss": 0.0142, "step": 21910 }, { "grad_norm": 0.18235117197036743, "learning_rate": 9.23760002690991e-05, "loss": 0.0142, "step": 21920 }, { "grad_norm": 0.1750355362892151, "learning_rate": 9.236722193871252e-05, "loss": 0.014, "step": 21930 }, { "grad_norm": 0.18254826962947845, "learning_rate": 9.235843897511023e-05, "loss": 0.0144, "step": 21940 }, { "grad_norm": 0.18127337098121643, "learning_rate": 9.234965137925276e-05, "loss": 0.0121, "step": 21950 }, { "grad_norm": 0.14597991108894348, "learning_rate": 9.234085915210108e-05, "loss": 0.0131, "step": 21960 }, { "grad_norm": 0.18327923119068146, "learning_rate": 9.23320622946167e-05, "loss": 0.0167, "step": 21970 }, { "grad_norm": 0.22284559905529022, "learning_rate": 9.232326080776163e-05, "loss": 0.017, "step": 21980 }, { "grad_norm": 0.17206434905529022, "learning_rate": 9.23144546924984e-05, "loss": 0.0163, "step": 21990 }, { "grad_norm": 0.2701965272426605, "learning_rate": 9.230564394979e-05, "loss": 0.0125, "step": 22000 }, { "grad_norm": 0.21865223348140717, "learning_rate": 9.22968285806e-05, "loss": 0.0143, "step": 22010 }, { "grad_norm": 0.16129112243652344, "learning_rate": 9.228800858589242e-05, "loss": 0.0136, "step": 22020 }, { "grad_norm": 0.14999327063560486, "learning_rate": 9.227918396663179e-05, "loss": 0.0124, "step": 22030 }, { "grad_norm": 0.19997403025627136, "learning_rate": 9.227035472378319e-05, "loss": 0.0142, "step": 22040 }, { "grad_norm": 0.17474998533725739, "learning_rate": 9.226152085831213e-05, "loss": 0.0122, "step": 22050 }, { "grad_norm": 0.17824895679950714, "learning_rate": 9.22526823711847e-05, "loss": 0.0124, "step": 22060 }, { "grad_norm": 0.16592907905578613, "learning_rate": 9.224383926336745e-05, "loss": 0.0148, "step": 22070 }, { "grad_norm": 0.15698660910129547, "learning_rate": 9.223499153582744e-05, "loss": 0.0131, "step": 22080 }, { "grad_norm": 0.1930922120809555, "learning_rate": 9.222613918953226e-05, "loss": 0.0129, "step": 22090 }, { "grad_norm": 0.1886027604341507, "learning_rate": 9.221728222544999e-05, "loss": 0.017, "step": 22100 }, { "grad_norm": 0.23879215121269226, "learning_rate": 9.22084206445492e-05, "loss": 0.0125, "step": 22110 }, { "grad_norm": 0.2269887924194336, "learning_rate": 9.2199554447799e-05, "loss": 0.0145, "step": 22120 }, { "grad_norm": 0.19532684981822968, "learning_rate": 9.219068363616897e-05, "loss": 0.018, "step": 22130 }, { "grad_norm": 0.1273169070482254, "learning_rate": 9.218180821062919e-05, "loss": 0.014, "step": 22140 }, { "grad_norm": 0.16206340491771698, "learning_rate": 9.21729281721503e-05, "loss": 0.0136, "step": 22150 }, { "grad_norm": 0.1865641325712204, "learning_rate": 9.216404352170339e-05, "loss": 0.0136, "step": 22160 }, { "grad_norm": 0.2240491658449173, "learning_rate": 9.215515426026007e-05, "loss": 0.0142, "step": 22170 }, { "grad_norm": 0.22914744913578033, "learning_rate": 9.214626038879246e-05, "loss": 0.0147, "step": 22180 }, { "grad_norm": 0.1625586450099945, "learning_rate": 9.21373619082732e-05, "loss": 0.0156, "step": 22190 }, { "grad_norm": 0.21820929646492004, "learning_rate": 9.212845881967535e-05, "loss": 0.0161, "step": 22200 }, { "grad_norm": 0.15866835415363312, "learning_rate": 9.211955112397262e-05, "loss": 0.0147, "step": 22210 }, { "grad_norm": 0.19606928527355194, "learning_rate": 9.211063882213909e-05, "loss": 0.0161, "step": 22220 }, { "grad_norm": 0.18008647859096527, "learning_rate": 9.210172191514942e-05, "loss": 0.0144, "step": 22230 }, { "grad_norm": 0.19143535196781158, "learning_rate": 9.209280040397874e-05, "loss": 0.0152, "step": 22240 }, { "grad_norm": 0.23474213480949402, "learning_rate": 9.208387428960268e-05, "loss": 0.0194, "step": 22250 }, { "grad_norm": 0.24880555272102356, "learning_rate": 9.20749435729974e-05, "loss": 0.0171, "step": 22260 }, { "grad_norm": 0.1741355061531067, "learning_rate": 9.206600825513957e-05, "loss": 0.0137, "step": 22270 }, { "grad_norm": 0.17262478172779083, "learning_rate": 9.20570683370063e-05, "loss": 0.0142, "step": 22280 }, { "grad_norm": 0.1518886387348175, "learning_rate": 9.204812381957528e-05, "loss": 0.0138, "step": 22290 }, { "grad_norm": 0.16712774336338043, "learning_rate": 9.203917470382465e-05, "loss": 0.0117, "step": 22300 }, { "grad_norm": 0.23551325500011444, "learning_rate": 9.203022099073309e-05, "loss": 0.0139, "step": 22310 }, { "grad_norm": 0.2145652025938034, "learning_rate": 9.202126268127976e-05, "loss": 0.0134, "step": 22320 }, { "grad_norm": 0.20092186331748962, "learning_rate": 9.20122997764443e-05, "loss": 0.0174, "step": 22330 }, { "grad_norm": 0.22837598621845245, "learning_rate": 9.200333227720692e-05, "loss": 0.0158, "step": 22340 }, { "grad_norm": 0.15929846465587616, "learning_rate": 9.199436018454826e-05, "loss": 0.014, "step": 22350 }, { "grad_norm": 0.22181697189807892, "learning_rate": 9.198538349944952e-05, "loss": 0.0136, "step": 22360 }, { "grad_norm": 0.18189416825771332, "learning_rate": 9.197640222289234e-05, "loss": 0.0147, "step": 22370 }, { "grad_norm": 0.18257778882980347, "learning_rate": 9.196741635585895e-05, "loss": 0.0126, "step": 22380 }, { "grad_norm": 0.203364759683609, "learning_rate": 9.195842589933199e-05, "loss": 0.0173, "step": 22390 }, { "grad_norm": 0.1993296593427658, "learning_rate": 9.194943085429466e-05, "loss": 0.0164, "step": 22400 }, { "grad_norm": 0.23538455367088318, "learning_rate": 9.194043122173065e-05, "loss": 0.0135, "step": 22410 }, { "grad_norm": 0.16775605082511902, "learning_rate": 9.193142700262413e-05, "loss": 0.0173, "step": 22420 }, { "grad_norm": 0.20025067031383514, "learning_rate": 9.192241819795979e-05, "loss": 0.0159, "step": 22430 }, { "grad_norm": 0.20080801844596863, "learning_rate": 9.191340480872284e-05, "loss": 0.016, "step": 22440 }, { "grad_norm": 0.19646243751049042, "learning_rate": 9.190438683589895e-05, "loss": 0.0144, "step": 22450 }, { "grad_norm": 0.20725007355213165, "learning_rate": 9.189536428047432e-05, "loss": 0.0134, "step": 22460 }, { "grad_norm": 0.19579613208770752, "learning_rate": 9.188633714343564e-05, "loss": 0.0141, "step": 22470 }, { "grad_norm": 0.2477605640888214, "learning_rate": 9.18773054257701e-05, "loss": 0.0129, "step": 22480 }, { "grad_norm": 0.1874033808708191, "learning_rate": 9.18682691284654e-05, "loss": 0.0127, "step": 22490 }, { "grad_norm": 0.1978670060634613, "learning_rate": 9.185922825250974e-05, "loss": 0.0142, "step": 22500 }, { "grad_norm": 0.21049946546554565, "learning_rate": 9.185018279889181e-05, "loss": 0.0145, "step": 22510 }, { "grad_norm": 0.19895488023757935, "learning_rate": 9.184113276860082e-05, "loss": 0.0132, "step": 22520 }, { "grad_norm": 0.18387296795845032, "learning_rate": 9.183207816262645e-05, "loss": 0.0148, "step": 22530 }, { "grad_norm": 0.2633856534957886, "learning_rate": 9.182301898195891e-05, "loss": 0.0159, "step": 22540 }, { "grad_norm": 0.17818941175937653, "learning_rate": 9.181395522758889e-05, "loss": 0.0134, "step": 22550 }, { "grad_norm": 0.1726434975862503, "learning_rate": 9.180488690050759e-05, "loss": 0.0151, "step": 22560 }, { "grad_norm": 0.2328484207391739, "learning_rate": 9.179581400170671e-05, "loss": 0.0135, "step": 22570 }, { "grad_norm": 0.22125305235385895, "learning_rate": 9.178673653217845e-05, "loss": 0.0117, "step": 22580 }, { "grad_norm": 0.18342380225658417, "learning_rate": 9.177765449291551e-05, "loss": 0.0132, "step": 22590 }, { "grad_norm": 0.14506664872169495, "learning_rate": 9.176856788491109e-05, "loss": 0.0146, "step": 22600 }, { "grad_norm": 0.15208350121974945, "learning_rate": 9.175947670915887e-05, "loss": 0.0133, "step": 22610 }, { "grad_norm": 0.1756669580936432, "learning_rate": 9.175038096665309e-05, "loss": 0.0112, "step": 22620 }, { "grad_norm": 0.1454291045665741, "learning_rate": 9.17412806583884e-05, "loss": 0.0135, "step": 22630 }, { "grad_norm": 0.2701079249382019, "learning_rate": 9.173217578536002e-05, "loss": 0.0137, "step": 22640 }, { "grad_norm": 0.2173529416322708, "learning_rate": 9.172306634856362e-05, "loss": 0.0166, "step": 22650 }, { "grad_norm": 0.15783634781837463, "learning_rate": 9.171395234899545e-05, "loss": 0.0139, "step": 22660 }, { "grad_norm": 0.1950979232788086, "learning_rate": 9.170483378765214e-05, "loss": 0.019, "step": 22670 }, { "grad_norm": 0.19414390623569489, "learning_rate": 9.169571066553091e-05, "loss": 0.0147, "step": 22680 }, { "grad_norm": 0.15959341824054718, "learning_rate": 9.168658298362946e-05, "loss": 0.0145, "step": 22690 }, { "grad_norm": 0.1607877016067505, "learning_rate": 9.167745074294598e-05, "loss": 0.0139, "step": 22700 }, { "grad_norm": 0.18204639852046967, "learning_rate": 9.166831394447913e-05, "loss": 0.0154, "step": 22710 }, { "grad_norm": 0.19542460143566132, "learning_rate": 9.165917258922812e-05, "loss": 0.013, "step": 22720 }, { "grad_norm": 0.17105495929718018, "learning_rate": 9.165002667819262e-05, "loss": 0.0133, "step": 22730 }, { "grad_norm": 0.21982093155384064, "learning_rate": 9.164087621237282e-05, "loss": 0.0144, "step": 22740 }, { "grad_norm": 0.16044984757900238, "learning_rate": 9.163172119276942e-05, "loss": 0.0139, "step": 22750 }, { "grad_norm": 0.16910907626152039, "learning_rate": 9.162256162038358e-05, "loss": 0.0128, "step": 22760 }, { "grad_norm": 0.16099341213703156, "learning_rate": 9.161339749621698e-05, "loss": 0.0157, "step": 22770 }, { "grad_norm": 0.14582465589046478, "learning_rate": 9.160422882127177e-05, "loss": 0.0135, "step": 22780 }, { "grad_norm": 0.2041366994380951, "learning_rate": 9.159505559655069e-05, "loss": 0.0139, "step": 22790 }, { "grad_norm": 0.19715900719165802, "learning_rate": 9.158587782305684e-05, "loss": 0.0143, "step": 22800 }, { "grad_norm": 0.1901339739561081, "learning_rate": 9.157669550179391e-05, "loss": 0.0164, "step": 22810 }, { "grad_norm": 0.16475248336791992, "learning_rate": 9.156750863376609e-05, "loss": 0.0111, "step": 22820 }, { "grad_norm": 0.32444044947624207, "learning_rate": 9.155831721997801e-05, "loss": 0.0155, "step": 22830 }, { "grad_norm": 0.16951529681682587, "learning_rate": 9.154912126143484e-05, "loss": 0.0129, "step": 22840 }, { "grad_norm": 0.27977102994918823, "learning_rate": 9.153992075914224e-05, "loss": 0.0158, "step": 22850 }, { "grad_norm": 0.21274621784687042, "learning_rate": 9.153071571410635e-05, "loss": 0.014, "step": 22860 }, { "grad_norm": 0.19473186135292053, "learning_rate": 9.152150612733384e-05, "loss": 0.013, "step": 22870 }, { "grad_norm": 0.18514996767044067, "learning_rate": 9.151229199983184e-05, "loss": 0.0141, "step": 22880 }, { "grad_norm": 0.19141137599945068, "learning_rate": 9.150307333260802e-05, "loss": 0.0132, "step": 22890 }, { "grad_norm": 0.1946808397769928, "learning_rate": 9.149385012667048e-05, "loss": 0.0156, "step": 22900 }, { "grad_norm": 0.21053126454353333, "learning_rate": 9.148462238302788e-05, "loss": 0.015, "step": 22910 }, { "grad_norm": 0.17560558021068573, "learning_rate": 9.147539010268936e-05, "loss": 0.0153, "step": 22920 }, { "grad_norm": 0.20128123462200165, "learning_rate": 9.14661532866645e-05, "loss": 0.0139, "step": 22930 }, { "grad_norm": 0.1753312647342682, "learning_rate": 9.145691193596348e-05, "loss": 0.0145, "step": 22940 }, { "grad_norm": 0.18623462319374084, "learning_rate": 9.144766605159691e-05, "loss": 0.0149, "step": 22950 }, { "grad_norm": 0.21088488399982452, "learning_rate": 9.14384156345759e-05, "loss": 0.015, "step": 22960 }, { "grad_norm": 0.23446334898471832, "learning_rate": 9.142916068591204e-05, "loss": 0.0158, "step": 22970 }, { "grad_norm": 0.2326941192150116, "learning_rate": 9.141990120661746e-05, "loss": 0.0142, "step": 22980 }, { "grad_norm": 0.18893904983997345, "learning_rate": 9.141063719770475e-05, "loss": 0.0122, "step": 22990 }, { "grad_norm": 0.2209293395280838, "learning_rate": 9.140136866018704e-05, "loss": 0.0158, "step": 23000 }, { "grad_norm": 0.1879260390996933, "learning_rate": 9.139209559507788e-05, "loss": 0.0143, "step": 23010 }, { "grad_norm": 0.1820458173751831, "learning_rate": 9.13828180033914e-05, "loss": 0.0131, "step": 23020 }, { "grad_norm": 0.1506364941596985, "learning_rate": 9.137353588614212e-05, "loss": 0.0116, "step": 23030 }, { "grad_norm": 0.1475699543952942, "learning_rate": 9.136424924434519e-05, "loss": 0.0144, "step": 23040 }, { "grad_norm": 0.1821097433567047, "learning_rate": 9.135495807901615e-05, "loss": 0.0124, "step": 23050 }, { "grad_norm": 0.22350358963012695, "learning_rate": 9.134566239117108e-05, "loss": 0.0139, "step": 23060 }, { "grad_norm": 0.16873721778392792, "learning_rate": 9.13363621818265e-05, "loss": 0.014, "step": 23070 }, { "grad_norm": 0.24246738851070404, "learning_rate": 9.132705745199953e-05, "loss": 0.0156, "step": 23080 }, { "grad_norm": 0.1951896846294403, "learning_rate": 9.131774820270768e-05, "loss": 0.0155, "step": 23090 }, { "grad_norm": 0.1857994645833969, "learning_rate": 9.130843443496901e-05, "loss": 0.0161, "step": 23100 }, { "grad_norm": 0.16752180457115173, "learning_rate": 9.129911614980206e-05, "loss": 0.0127, "step": 23110 }, { "grad_norm": 0.1568581610918045, "learning_rate": 9.128979334822584e-05, "loss": 0.0122, "step": 23120 }, { "grad_norm": 0.21348780393600464, "learning_rate": 9.128046603125992e-05, "loss": 0.0127, "step": 23130 }, { "grad_norm": 0.20745138823986053, "learning_rate": 9.12711341999243e-05, "loss": 0.0122, "step": 23140 }, { "grad_norm": 0.17473304271697998, "learning_rate": 9.12617978552395e-05, "loss": 0.0142, "step": 23150 }, { "grad_norm": 0.2018103450536728, "learning_rate": 9.12524569982265e-05, "loss": 0.0137, "step": 23160 }, { "grad_norm": 0.1688775271177292, "learning_rate": 9.124311162990684e-05, "loss": 0.0133, "step": 23170 }, { "grad_norm": 0.2487609088420868, "learning_rate": 9.12337617513025e-05, "loss": 0.0139, "step": 23180 }, { "grad_norm": 0.2045906037092209, "learning_rate": 9.122440736343596e-05, "loss": 0.0174, "step": 23190 }, { "grad_norm": 0.16671490669250488, "learning_rate": 9.12150484673302e-05, "loss": 0.0181, "step": 23200 }, { "grad_norm": 0.16812296211719513, "learning_rate": 9.120568506400873e-05, "loss": 0.013, "step": 23210 }, { "grad_norm": 0.150957390666008, "learning_rate": 9.119631715449548e-05, "loss": 0.0114, "step": 23220 }, { "grad_norm": 0.21648240089416504, "learning_rate": 9.118694473981493e-05, "loss": 0.0138, "step": 23230 }, { "grad_norm": 0.18126694858074188, "learning_rate": 9.117756782099203e-05, "loss": 0.0139, "step": 23240 }, { "grad_norm": 0.19866658747196198, "learning_rate": 9.11681863990522e-05, "loss": 0.0111, "step": 23250 }, { "grad_norm": 0.17126090824604034, "learning_rate": 9.115880047502142e-05, "loss": 0.0122, "step": 23260 }, { "grad_norm": 0.1864488124847412, "learning_rate": 9.114941004992609e-05, "loss": 0.013, "step": 23270 }, { "grad_norm": 0.25654807686805725, "learning_rate": 9.114001512479317e-05, "loss": 0.0147, "step": 23280 }, { "grad_norm": 0.20216181874275208, "learning_rate": 9.113061570065003e-05, "loss": 0.0126, "step": 23290 }, { "grad_norm": 0.15994037687778473, "learning_rate": 9.112121177852459e-05, "loss": 0.0132, "step": 23300 }, { "grad_norm": 0.21002879738807678, "learning_rate": 9.111180335944527e-05, "loss": 0.0126, "step": 23310 }, { "grad_norm": 0.1547642946243286, "learning_rate": 9.110239044444093e-05, "loss": 0.0145, "step": 23320 }, { "grad_norm": 0.1661802977323532, "learning_rate": 9.109297303454099e-05, "loss": 0.0125, "step": 23330 }, { "grad_norm": 0.16914597153663635, "learning_rate": 9.108355113077526e-05, "loss": 0.0116, "step": 23340 }, { "grad_norm": 0.17308777570724487, "learning_rate": 9.107412473417419e-05, "loss": 0.013, "step": 23350 }, { "grad_norm": 0.16358745098114014, "learning_rate": 9.106469384576858e-05, "loss": 0.0145, "step": 23360 }, { "grad_norm": 0.12418806552886963, "learning_rate": 9.105525846658978e-05, "loss": 0.0141, "step": 23370 }, { "grad_norm": 0.2540501058101654, "learning_rate": 9.104581859766965e-05, "loss": 0.0172, "step": 23380 }, { "grad_norm": 0.23612475395202637, "learning_rate": 9.10363742400405e-05, "loss": 0.0148, "step": 23390 }, { "grad_norm": 0.17769350111484528, "learning_rate": 9.102692539473518e-05, "loss": 0.0143, "step": 23400 }, { "grad_norm": 0.20720572769641876, "learning_rate": 9.101747206278697e-05, "loss": 0.0148, "step": 23410 }, { "grad_norm": 0.22612901031970978, "learning_rate": 9.100801424522968e-05, "loss": 0.0167, "step": 23420 }, { "grad_norm": 0.23151539266109467, "learning_rate": 9.099855194309762e-05, "loss": 0.0161, "step": 23430 }, { "grad_norm": 0.19694086909294128, "learning_rate": 9.098908515742554e-05, "loss": 0.0125, "step": 23440 }, { "grad_norm": 0.22731053829193115, "learning_rate": 9.097961388924873e-05, "loss": 0.0148, "step": 23450 }, { "grad_norm": 0.17639155685901642, "learning_rate": 9.097013813960298e-05, "loss": 0.0154, "step": 23460 }, { "grad_norm": 0.2621200680732727, "learning_rate": 9.09606579095245e-05, "loss": 0.0149, "step": 23470 }, { "grad_norm": 0.2077457159757614, "learning_rate": 9.095117320005008e-05, "loss": 0.0133, "step": 23480 }, { "grad_norm": 0.17855028808116913, "learning_rate": 9.094168401221691e-05, "loss": 0.0107, "step": 23490 }, { "grad_norm": 0.21619366109371185, "learning_rate": 9.093219034706273e-05, "loss": 0.0145, "step": 23500 }, { "grad_norm": 0.1861942708492279, "learning_rate": 9.092269220562577e-05, "loss": 0.0157, "step": 23510 }, { "grad_norm": 0.22576206922531128, "learning_rate": 9.09131895889447e-05, "loss": 0.0152, "step": 23520 }, { "grad_norm": 0.18485130369663239, "learning_rate": 9.090368249805873e-05, "loss": 0.0157, "step": 23530 }, { "grad_norm": 0.20125995576381683, "learning_rate": 9.089417093400754e-05, "loss": 0.0161, "step": 23540 }, { "grad_norm": 0.1880805939435959, "learning_rate": 9.088465489783131e-05, "loss": 0.0136, "step": 23550 }, { "grad_norm": 0.22410212457180023, "learning_rate": 9.087513439057068e-05, "loss": 0.0114, "step": 23560 }, { "grad_norm": 0.18563330173492432, "learning_rate": 9.08656094132668e-05, "loss": 0.0126, "step": 23570 }, { "grad_norm": 0.18965651094913483, "learning_rate": 9.085607996696134e-05, "loss": 0.018, "step": 23580 }, { "grad_norm": 0.16005998849868774, "learning_rate": 9.084654605269639e-05, "loss": 0.0138, "step": 23590 }, { "grad_norm": 0.2741300165653229, "learning_rate": 9.083700767151457e-05, "loss": 0.0137, "step": 23600 }, { "grad_norm": 0.16237464547157288, "learning_rate": 9.082746482445898e-05, "loss": 0.0127, "step": 23610 }, { "grad_norm": 0.13066178560256958, "learning_rate": 9.081791751257325e-05, "loss": 0.0128, "step": 23620 }, { "grad_norm": 0.2223547250032425, "learning_rate": 9.080836573690142e-05, "loss": 0.0137, "step": 23630 }, { "grad_norm": 0.20824269950389862, "learning_rate": 9.079880949848805e-05, "loss": 0.0132, "step": 23640 }, { "grad_norm": 0.1773054003715515, "learning_rate": 9.078924879837822e-05, "loss": 0.0112, "step": 23650 }, { "grad_norm": 0.17374740540981293, "learning_rate": 9.077968363761747e-05, "loss": 0.0124, "step": 23660 }, { "grad_norm": 0.1537875533103943, "learning_rate": 9.077011401725182e-05, "loss": 0.0141, "step": 23670 }, { "grad_norm": 0.17224346101284027, "learning_rate": 9.07605399383278e-05, "loss": 0.0117, "step": 23680 }, { "grad_norm": 0.2315269112586975, "learning_rate": 9.075096140189243e-05, "loss": 0.0121, "step": 23690 }, { "grad_norm": 0.20412009954452515, "learning_rate": 9.074137840899318e-05, "loss": 0.0137, "step": 23700 }, { "grad_norm": 0.2086557000875473, "learning_rate": 9.073179096067804e-05, "loss": 0.0122, "step": 23710 }, { "grad_norm": 0.18799857795238495, "learning_rate": 9.072219905799549e-05, "loss": 0.0151, "step": 23720 }, { "grad_norm": 0.1587083488702774, "learning_rate": 9.071260270199447e-05, "loss": 0.0128, "step": 23730 }, { "grad_norm": 0.20421265065670013, "learning_rate": 9.070300189372441e-05, "loss": 0.0148, "step": 23740 }, { "grad_norm": 0.1996258944272995, "learning_rate": 9.069339663423528e-05, "loss": 0.0129, "step": 23750 }, { "grad_norm": 0.16230294108390808, "learning_rate": 9.068378692457747e-05, "loss": 0.0142, "step": 23760 }, { "grad_norm": 0.1988297551870346, "learning_rate": 9.067417276580189e-05, "loss": 0.0133, "step": 23770 }, { "grad_norm": 0.2083693891763687, "learning_rate": 9.066455415895993e-05, "loss": 0.0135, "step": 23780 }, { "grad_norm": 0.18406611680984497, "learning_rate": 9.065493110510346e-05, "loss": 0.0128, "step": 23790 }, { "grad_norm": 0.17064645886421204, "learning_rate": 9.064530360528484e-05, "loss": 0.0141, "step": 23800 }, { "grad_norm": 0.26106420159339905, "learning_rate": 9.063567166055695e-05, "loss": 0.0136, "step": 23810 }, { "grad_norm": 0.22536739706993103, "learning_rate": 9.062603527197308e-05, "loss": 0.0169, "step": 23820 }, { "grad_norm": 0.22252939641475677, "learning_rate": 9.06163944405871e-05, "loss": 0.0119, "step": 23830 }, { "grad_norm": 0.1754385381937027, "learning_rate": 9.060674916745327e-05, "loss": 0.0137, "step": 23840 }, { "grad_norm": 0.16937655210494995, "learning_rate": 9.05970994536264e-05, "loss": 0.0129, "step": 23850 }, { "grad_norm": 0.19148722290992737, "learning_rate": 9.05874453001618e-05, "loss": 0.0127, "step": 23860 }, { "grad_norm": 0.1617276668548584, "learning_rate": 9.057778670811517e-05, "loss": 0.0132, "step": 23870 }, { "grad_norm": 0.14512422680854797, "learning_rate": 9.056812367854281e-05, "loss": 0.0125, "step": 23880 }, { "grad_norm": 0.18607541918754578, "learning_rate": 9.055845621250143e-05, "loss": 0.0148, "step": 23890 }, { "grad_norm": 0.18467602133750916, "learning_rate": 9.054878431104825e-05, "loss": 0.0146, "step": 23900 }, { "grad_norm": 0.18069535493850708, "learning_rate": 9.0539107975241e-05, "loss": 0.0129, "step": 23910 }, { "grad_norm": 0.17463482916355133, "learning_rate": 9.052942720613784e-05, "loss": 0.0127, "step": 23920 }, { "grad_norm": 0.17424772679805756, "learning_rate": 9.051974200479745e-05, "loss": 0.0116, "step": 23930 }, { "grad_norm": 0.20681899785995483, "learning_rate": 9.051005237227901e-05, "loss": 0.0138, "step": 23940 }, { "grad_norm": 0.22906869649887085, "learning_rate": 9.050035830964215e-05, "loss": 0.0138, "step": 23950 }, { "grad_norm": 0.1764940619468689, "learning_rate": 9.049065981794698e-05, "loss": 0.0157, "step": 23960 }, { "grad_norm": 0.18284966051578522, "learning_rate": 9.048095689825414e-05, "loss": 0.0111, "step": 23970 }, { "grad_norm": 0.19760847091674805, "learning_rate": 9.047124955162472e-05, "loss": 0.0123, "step": 23980 }, { "grad_norm": 0.18020464479923248, "learning_rate": 9.046153777912028e-05, "loss": 0.013, "step": 23990 }, { "grad_norm": 0.18729747831821442, "learning_rate": 9.045182158180292e-05, "loss": 0.0131, "step": 24000 }, { "grad_norm": 0.21491290628910065, "learning_rate": 9.044210096073516e-05, "loss": 0.0129, "step": 24010 }, { "grad_norm": 0.14622865617275238, "learning_rate": 9.043237591698004e-05, "loss": 0.0141, "step": 24020 }, { "grad_norm": 0.18023575842380524, "learning_rate": 9.04226464516011e-05, "loss": 0.0127, "step": 24030 }, { "grad_norm": 0.2124701887369156, "learning_rate": 9.041291256566229e-05, "loss": 0.0147, "step": 24040 }, { "grad_norm": 0.22396758198738098, "learning_rate": 9.040317426022814e-05, "loss": 0.0152, "step": 24050 }, { "grad_norm": 0.22996024787425995, "learning_rate": 9.03934315363636e-05, "loss": 0.0164, "step": 24060 }, { "grad_norm": 0.16324064135551453, "learning_rate": 9.038368439513409e-05, "loss": 0.0134, "step": 24070 }, { "grad_norm": 0.1615142673254013, "learning_rate": 9.03739328376056e-05, "loss": 0.0116, "step": 24080 }, { "grad_norm": 0.1858004629611969, "learning_rate": 9.036417686484451e-05, "loss": 0.0151, "step": 24090 }, { "grad_norm": 0.21046839654445648, "learning_rate": 9.035441647791773e-05, "loss": 0.0134, "step": 24100 }, { "grad_norm": 0.1651192307472229, "learning_rate": 9.034465167789263e-05, "loss": 0.0126, "step": 24110 }, { "grad_norm": 0.19996723532676697, "learning_rate": 9.033488246583706e-05, "loss": 0.0138, "step": 24120 }, { "grad_norm": 0.2324429601430893, "learning_rate": 9.032510884281941e-05, "loss": 0.0119, "step": 24130 }, { "grad_norm": 0.19369323551654816, "learning_rate": 9.031533080990848e-05, "loss": 0.0127, "step": 24140 }, { "grad_norm": 0.16205868124961853, "learning_rate": 9.030554836817358e-05, "loss": 0.0124, "step": 24150 }, { "grad_norm": 0.13563553988933563, "learning_rate": 9.029576151868451e-05, "loss": 0.0131, "step": 24160 }, { "grad_norm": 0.1850966364145279, "learning_rate": 9.028597026251155e-05, "loss": 0.0109, "step": 24170 }, { "grad_norm": 0.24279119074344635, "learning_rate": 9.027617460072547e-05, "loss": 0.016, "step": 24180 }, { "grad_norm": 0.16006587445735931, "learning_rate": 9.026637453439745e-05, "loss": 0.0166, "step": 24190 }, { "grad_norm": 0.2095530778169632, "learning_rate": 9.025657006459927e-05, "loss": 0.0129, "step": 24200 }, { "grad_norm": 0.3007853031158447, "learning_rate": 9.024676119240311e-05, "loss": 0.014, "step": 24210 }, { "grad_norm": 0.19886881113052368, "learning_rate": 9.023694791888166e-05, "loss": 0.0165, "step": 24220 }, { "grad_norm": 0.1658303290605545, "learning_rate": 9.022713024510808e-05, "loss": 0.0142, "step": 24230 }, { "grad_norm": 0.20987501740455627, "learning_rate": 9.021730817215601e-05, "loss": 0.0126, "step": 24240 }, { "grad_norm": 0.1842327117919922, "learning_rate": 9.02074817010996e-05, "loss": 0.0143, "step": 24250 }, { "grad_norm": 0.24087655544281006, "learning_rate": 9.019765083301342e-05, "loss": 0.0132, "step": 24260 }, { "grad_norm": 0.16316771507263184, "learning_rate": 9.01878155689726e-05, "loss": 0.0118, "step": 24270 }, { "grad_norm": 0.18406403064727783, "learning_rate": 9.017797591005268e-05, "loss": 0.0127, "step": 24280 }, { "grad_norm": 0.20493465662002563, "learning_rate": 9.016813185732972e-05, "loss": 0.0167, "step": 24290 }, { "grad_norm": 0.14221425354480743, "learning_rate": 9.015828341188027e-05, "loss": 0.0136, "step": 24300 }, { "grad_norm": 0.19436633586883545, "learning_rate": 9.01484305747813e-05, "loss": 0.0141, "step": 24310 }, { "grad_norm": 0.2078462541103363, "learning_rate": 9.013857334711033e-05, "loss": 0.0127, "step": 24320 }, { "grad_norm": 0.12844078242778778, "learning_rate": 9.012871172994534e-05, "loss": 0.0132, "step": 24330 }, { "grad_norm": 0.1644139438867569, "learning_rate": 9.011884572436476e-05, "loss": 0.0145, "step": 24340 }, { "grad_norm": 0.15801170468330383, "learning_rate": 9.010897533144754e-05, "loss": 0.0122, "step": 24350 }, { "grad_norm": 0.17734327912330627, "learning_rate": 9.009910055227306e-05, "loss": 0.0144, "step": 24360 }, { "grad_norm": 0.2012241929769516, "learning_rate": 9.008922138792124e-05, "loss": 0.0144, "step": 24370 }, { "grad_norm": 0.20198063552379608, "learning_rate": 9.007933783947244e-05, "loss": 0.0132, "step": 24380 }, { "grad_norm": 0.195786714553833, "learning_rate": 9.006944990800752e-05, "loss": 0.0142, "step": 24390 }, { "grad_norm": 0.17990052700042725, "learning_rate": 9.005955759460779e-05, "loss": 0.0137, "step": 24400 }, { "grad_norm": 0.16798342764377594, "learning_rate": 9.004966090035508e-05, "loss": 0.0141, "step": 24410 }, { "grad_norm": 0.2183929979801178, "learning_rate": 9.003975982633166e-05, "loss": 0.0128, "step": 24420 }, { "grad_norm": 0.2140144407749176, "learning_rate": 9.00298543736203e-05, "loss": 0.0153, "step": 24430 }, { "grad_norm": 0.17506416141986847, "learning_rate": 9.001994454330427e-05, "loss": 0.0162, "step": 24440 }, { "grad_norm": 0.2119501233100891, "learning_rate": 9.001003033646727e-05, "loss": 0.0147, "step": 24450 }, { "grad_norm": 0.14297392964363098, "learning_rate": 9.00001117541935e-05, "loss": 0.014, "step": 24460 }, { "grad_norm": 0.21404708921909332, "learning_rate": 8.999018879756764e-05, "loss": 0.0127, "step": 24470 }, { "grad_norm": 0.1482110470533371, "learning_rate": 8.998026146767487e-05, "loss": 0.0126, "step": 24480 }, { "grad_norm": 0.2301146239042282, "learning_rate": 8.99703297656008e-05, "loss": 0.0157, "step": 24490 }, { "grad_norm": 0.2654757499694824, "learning_rate": 8.996039369243156e-05, "loss": 0.017, "step": 24500 }, { "grad_norm": 0.20895595848560333, "learning_rate": 8.995045324925378e-05, "loss": 0.0139, "step": 24510 }, { "grad_norm": 0.1800021231174469, "learning_rate": 8.994050843715448e-05, "loss": 0.0135, "step": 24520 }, { "grad_norm": 0.14947514235973358, "learning_rate": 8.993055925722121e-05, "loss": 0.0144, "step": 24530 }, { "grad_norm": 0.19799265265464783, "learning_rate": 8.992060571054202e-05, "loss": 0.0169, "step": 24540 }, { "grad_norm": 0.21262411773204803, "learning_rate": 8.991064779820542e-05, "loss": 0.0134, "step": 24550 }, { "grad_norm": 0.19679954648017883, "learning_rate": 8.990068552130036e-05, "loss": 0.0144, "step": 24560 }, { "grad_norm": 0.2106962949037552, "learning_rate": 8.989071888091634e-05, "loss": 0.0165, "step": 24570 }, { "grad_norm": 0.2160099744796753, "learning_rate": 8.988074787814329e-05, "loss": 0.0121, "step": 24580 }, { "grad_norm": 0.21301431953907013, "learning_rate": 8.987077251407158e-05, "loss": 0.0113, "step": 24590 }, { "grad_norm": 0.20605546236038208, "learning_rate": 8.986079278979216e-05, "loss": 0.0158, "step": 24600 }, { "grad_norm": 0.19499704241752625, "learning_rate": 8.985080870639635e-05, "loss": 0.0137, "step": 24610 }, { "grad_norm": 0.24643474817276, "learning_rate": 8.984082026497603e-05, "loss": 0.0145, "step": 24620 }, { "grad_norm": 0.1719214916229248, "learning_rate": 8.98308274666235e-05, "loss": 0.0116, "step": 24630 }, { "grad_norm": 0.2153373658657074, "learning_rate": 8.982083031243155e-05, "loss": 0.0118, "step": 24640 }, { "grad_norm": 0.1471537947654724, "learning_rate": 8.98108288034935e-05, "loss": 0.0119, "step": 24650 }, { "grad_norm": 0.20173698663711548, "learning_rate": 8.980082294090305e-05, "loss": 0.0141, "step": 24660 }, { "grad_norm": 0.23772764205932617, "learning_rate": 8.979081272575443e-05, "loss": 0.0145, "step": 24670 }, { "grad_norm": 0.18586629629135132, "learning_rate": 8.978079815914236e-05, "loss": 0.0144, "step": 24680 }, { "grad_norm": 0.19816085696220398, "learning_rate": 8.977077924216202e-05, "loss": 0.0155, "step": 24690 }, { "grad_norm": 0.2016010582447052, "learning_rate": 8.976075597590905e-05, "loss": 0.0128, "step": 24700 }, { "grad_norm": 0.21539665758609772, "learning_rate": 8.975072836147958e-05, "loss": 0.0123, "step": 24710 }, { "grad_norm": 0.1961664855480194, "learning_rate": 8.974069639997025e-05, "loss": 0.0165, "step": 24720 }, { "grad_norm": 0.2580021321773529, "learning_rate": 8.973066009247808e-05, "loss": 0.0131, "step": 24730 }, { "grad_norm": 0.13507291674613953, "learning_rate": 8.972061944010066e-05, "loss": 0.0133, "step": 24740 }, { "grad_norm": 0.17641925811767578, "learning_rate": 8.971057444393603e-05, "loss": 0.0121, "step": 24750 }, { "grad_norm": 0.22395718097686768, "learning_rate": 8.970052510508268e-05, "loss": 0.0175, "step": 24760 }, { "grad_norm": 0.13897475600242615, "learning_rate": 8.969047142463959e-05, "loss": 0.0115, "step": 24770 }, { "grad_norm": 0.19934864342212677, "learning_rate": 8.968041340370621e-05, "loss": 0.0123, "step": 24780 }, { "grad_norm": 0.17633776366710663, "learning_rate": 8.96703510433825e-05, "loss": 0.0147, "step": 24790 }, { "grad_norm": 0.21618838608264923, "learning_rate": 8.966028434476883e-05, "loss": 0.0131, "step": 24800 }, { "grad_norm": 0.29347750544548035, "learning_rate": 8.96502133089661e-05, "loss": 0.0149, "step": 24810 }, { "grad_norm": 0.2277935892343521, "learning_rate": 8.964013793707564e-05, "loss": 0.012, "step": 24820 }, { "grad_norm": 0.17642952501773834, "learning_rate": 8.963005823019932e-05, "loss": 0.0123, "step": 24830 }, { "grad_norm": 0.18943725526332855, "learning_rate": 8.961997418943939e-05, "loss": 0.0118, "step": 24840 }, { "grad_norm": 0.1540994644165039, "learning_rate": 8.960988581589865e-05, "loss": 0.0128, "step": 24850 }, { "grad_norm": 0.14246435463428497, "learning_rate": 8.959979311068037e-05, "loss": 0.0104, "step": 24860 }, { "grad_norm": 0.18764404952526093, "learning_rate": 8.958969607488823e-05, "loss": 0.011, "step": 24870 }, { "grad_norm": 0.21704979240894318, "learning_rate": 8.957959470962647e-05, "loss": 0.0128, "step": 24880 }, { "grad_norm": 0.2385474592447281, "learning_rate": 8.956948901599971e-05, "loss": 0.0109, "step": 24890 }, { "grad_norm": 0.20033539831638336, "learning_rate": 8.955937899511315e-05, "loss": 0.013, "step": 24900 }, { "grad_norm": 0.16943170130252838, "learning_rate": 8.954926464807238e-05, "loss": 0.0122, "step": 24910 }, { "grad_norm": 0.15855255722999573, "learning_rate": 8.953914597598347e-05, "loss": 0.0155, "step": 24920 }, { "grad_norm": 0.19768494367599487, "learning_rate": 8.952902297995303e-05, "loss": 0.0133, "step": 24930 }, { "grad_norm": 0.16241057217121124, "learning_rate": 8.951889566108804e-05, "loss": 0.0116, "step": 24940 }, { "grad_norm": 0.1513100117444992, "learning_rate": 8.950876402049606e-05, "loss": 0.0119, "step": 24950 }, { "grad_norm": 0.14452563226222992, "learning_rate": 8.949862805928504e-05, "loss": 0.0139, "step": 24960 }, { "grad_norm": 0.21417489647865295, "learning_rate": 8.948848777856343e-05, "loss": 0.0112, "step": 24970 }, { "grad_norm": 0.2079704999923706, "learning_rate": 8.947834317944017e-05, "loss": 0.016, "step": 24980 }, { "grad_norm": 0.13944889605045319, "learning_rate": 8.946819426302466e-05, "loss": 0.0139, "step": 24990 }, { "grad_norm": 0.22804416716098785, "learning_rate": 8.945804103042676e-05, "loss": 0.0141, "step": 25000 }, { "grad_norm": 0.13791663944721222, "learning_rate": 8.944788348275681e-05, "loss": 0.0129, "step": 25010 }, { "grad_norm": 0.19554199278354645, "learning_rate": 8.943772162112565e-05, "loss": 0.0124, "step": 25020 }, { "grad_norm": 0.18653205037117004, "learning_rate": 8.942755544664454e-05, "loss": 0.0143, "step": 25030 }, { "grad_norm": 0.185913547873497, "learning_rate": 8.941738496042525e-05, "loss": 0.0119, "step": 25040 }, { "grad_norm": 0.15879835188388824, "learning_rate": 8.940721016357999e-05, "loss": 0.0134, "step": 25050 }, { "grad_norm": 0.14020977914333344, "learning_rate": 8.939703105722148e-05, "loss": 0.0136, "step": 25060 }, { "grad_norm": 0.1842188984155655, "learning_rate": 8.93868476424629e-05, "loss": 0.0166, "step": 25070 }, { "grad_norm": 0.16002054512500763, "learning_rate": 8.937665992041786e-05, "loss": 0.0122, "step": 25080 }, { "grad_norm": 0.17530089616775513, "learning_rate": 8.93664678922005e-05, "loss": 0.013, "step": 25090 }, { "grad_norm": 0.14398688077926636, "learning_rate": 8.93562715589254e-05, "loss": 0.0143, "step": 25100 }, { "grad_norm": 0.1390765756368637, "learning_rate": 8.934607092170762e-05, "loss": 0.0126, "step": 25110 }, { "grad_norm": 0.1907324492931366, "learning_rate": 8.933586598166266e-05, "loss": 0.0146, "step": 25120 }, { "grad_norm": 0.19296851754188538, "learning_rate": 8.932565673990655e-05, "loss": 0.0137, "step": 25130 }, { "grad_norm": 0.1827576607465744, "learning_rate": 8.931544319755574e-05, "loss": 0.0142, "step": 25140 }, { "grad_norm": 0.192660391330719, "learning_rate": 8.930522535572718e-05, "loss": 0.0131, "step": 25150 }, { "grad_norm": 0.15328280627727509, "learning_rate": 8.929500321553826e-05, "loss": 0.0115, "step": 25160 }, { "grad_norm": 0.16832326352596283, "learning_rate": 8.928477677810686e-05, "loss": 0.0118, "step": 25170 }, { "grad_norm": 0.17460127174854279, "learning_rate": 8.927454604455137e-05, "loss": 0.0128, "step": 25180 }, { "grad_norm": 0.21444889903068542, "learning_rate": 8.926431101599053e-05, "loss": 0.0134, "step": 25190 }, { "grad_norm": 0.1763206124305725, "learning_rate": 8.925407169354369e-05, "loss": 0.0108, "step": 25200 }, { "grad_norm": 0.1320035457611084, "learning_rate": 8.92438280783306e-05, "loss": 0.0101, "step": 25210 }, { "grad_norm": 0.17438222467899323, "learning_rate": 8.923358017147146e-05, "loss": 0.0148, "step": 25220 }, { "grad_norm": 0.2104928195476532, "learning_rate": 8.922332797408697e-05, "loss": 0.0135, "step": 25230 }, { "grad_norm": 0.2746444642543793, "learning_rate": 8.921307148729831e-05, "loss": 0.0129, "step": 25240 }, { "grad_norm": 0.17727157473564148, "learning_rate": 8.920281071222712e-05, "loss": 0.0113, "step": 25250 }, { "grad_norm": 0.15170201659202576, "learning_rate": 8.919254564999548e-05, "loss": 0.0116, "step": 25260 }, { "grad_norm": 0.17757190763950348, "learning_rate": 8.918227630172598e-05, "loss": 0.0139, "step": 25270 }, { "grad_norm": 0.15836475789546967, "learning_rate": 8.917200266854165e-05, "loss": 0.0125, "step": 25280 }, { "grad_norm": 0.19395394623279572, "learning_rate": 8.9161724751566e-05, "loss": 0.0152, "step": 25290 }, { "grad_norm": 0.1707041710615158, "learning_rate": 8.915144255192302e-05, "loss": 0.0133, "step": 25300 }, { "grad_norm": 0.17847658693790436, "learning_rate": 8.914115607073714e-05, "loss": 0.016, "step": 25310 }, { "grad_norm": 0.16537976264953613, "learning_rate": 8.913086530913327e-05, "loss": 0.0149, "step": 25320 }, { "grad_norm": 0.1338975876569748, "learning_rate": 8.912057026823681e-05, "loss": 0.0111, "step": 25330 }, { "grad_norm": 0.2068447321653366, "learning_rate": 8.91102709491736e-05, "loss": 0.0125, "step": 25340 }, { "grad_norm": 0.224120631814003, "learning_rate": 8.909996735306996e-05, "loss": 0.0111, "step": 25350 }, { "grad_norm": 0.24302910268306732, "learning_rate": 8.908965948105268e-05, "loss": 0.0127, "step": 25360 }, { "grad_norm": 0.19416041672229767, "learning_rate": 8.907934733424901e-05, "loss": 0.0148, "step": 25370 }, { "grad_norm": 0.18797488510608673, "learning_rate": 8.906903091378666e-05, "loss": 0.0133, "step": 25380 }, { "grad_norm": 0.1643301397562027, "learning_rate": 8.905871022079384e-05, "loss": 0.0139, "step": 25390 }, { "grad_norm": 0.1289331018924713, "learning_rate": 8.90483852563992e-05, "loss": 0.0137, "step": 25400 }, { "grad_norm": 0.12658986449241638, "learning_rate": 8.903805602173185e-05, "loss": 0.0168, "step": 25410 }, { "grad_norm": 0.14321275055408478, "learning_rate": 8.902772251792137e-05, "loss": 0.0144, "step": 25420 }, { "grad_norm": 0.19377011060714722, "learning_rate": 8.901738474609786e-05, "loss": 0.0122, "step": 25430 }, { "grad_norm": 0.22257426381111145, "learning_rate": 8.900704270739179e-05, "loss": 0.0154, "step": 25440 }, { "grad_norm": 0.21675224602222443, "learning_rate": 8.89966964029342e-05, "loss": 0.0108, "step": 25450 }, { "grad_norm": 0.24248936772346497, "learning_rate": 8.898634583385652e-05, "loss": 0.0155, "step": 25460 }, { "grad_norm": 0.20866306126117706, "learning_rate": 8.897599100129065e-05, "loss": 0.0174, "step": 25470 }, { "grad_norm": 0.1580723524093628, "learning_rate": 8.896563190636903e-05, "loss": 0.0157, "step": 25480 }, { "grad_norm": 0.15259769558906555, "learning_rate": 8.895526855022448e-05, "loss": 0.0145, "step": 25490 }, { "grad_norm": 0.23918873071670532, "learning_rate": 8.894490093399033e-05, "loss": 0.0134, "step": 25500 }, { "grad_norm": 0.13061527907848358, "learning_rate": 8.893452905880035e-05, "loss": 0.0137, "step": 25510 }, { "grad_norm": 0.1920856386423111, "learning_rate": 8.892415292578883e-05, "loss": 0.0169, "step": 25520 }, { "grad_norm": 0.18107008934020996, "learning_rate": 8.891377253609046e-05, "loss": 0.016, "step": 25530 }, { "grad_norm": 0.22509871423244476, "learning_rate": 8.890338789084043e-05, "loss": 0.0149, "step": 25540 }, { "grad_norm": 0.18140901625156403, "learning_rate": 8.88929989911744e-05, "loss": 0.0104, "step": 25550 }, { "grad_norm": 0.20011284947395325, "learning_rate": 8.888260583822847e-05, "loss": 0.0118, "step": 25560 }, { "grad_norm": 0.1640671044588089, "learning_rate": 8.887220843313921e-05, "loss": 0.013, "step": 25570 }, { "grad_norm": 0.22715744376182556, "learning_rate": 8.88618067770437e-05, "loss": 0.0148, "step": 25580 }, { "grad_norm": 0.18572910130023956, "learning_rate": 8.885140087107942e-05, "loss": 0.0152, "step": 25590 }, { "grad_norm": 0.17885176837444305, "learning_rate": 8.884099071638436e-05, "loss": 0.0142, "step": 25600 }, { "grad_norm": 0.21417374908924103, "learning_rate": 8.883057631409695e-05, "loss": 0.012, "step": 25610 }, { "grad_norm": 0.2049722820520401, "learning_rate": 8.882015766535608e-05, "loss": 0.0131, "step": 25620 }, { "grad_norm": 0.16861063241958618, "learning_rate": 8.880973477130115e-05, "loss": 0.0147, "step": 25630 }, { "grad_norm": 0.16143208742141724, "learning_rate": 8.879930763307197e-05, "loss": 0.0123, "step": 25640 }, { "grad_norm": 0.14631883800029755, "learning_rate": 8.878887625180884e-05, "loss": 0.0127, "step": 25650 }, { "grad_norm": 0.23450133204460144, "learning_rate": 8.877844062865253e-05, "loss": 0.0132, "step": 25660 }, { "grad_norm": 0.20436803996562958, "learning_rate": 8.876800076474424e-05, "loss": 0.018, "step": 25670 }, { "grad_norm": 0.14414970576763153, "learning_rate": 8.875755666122568e-05, "loss": 0.0139, "step": 25680 }, { "grad_norm": 0.17940959334373474, "learning_rate": 8.8747108319239e-05, "loss": 0.0105, "step": 25690 }, { "grad_norm": 0.1526135504245758, "learning_rate": 8.87366557399268e-05, "loss": 0.0143, "step": 25700 }, { "grad_norm": 0.16448132693767548, "learning_rate": 8.872619892443217e-05, "loss": 0.0144, "step": 25710 }, { "grad_norm": 0.15997529029846191, "learning_rate": 8.871573787389865e-05, "loss": 0.0116, "step": 25720 }, { "grad_norm": 0.17132891714572906, "learning_rate": 8.870527258947024e-05, "loss": 0.0138, "step": 25730 }, { "grad_norm": 0.2002885341644287, "learning_rate": 8.869480307229143e-05, "loss": 0.015, "step": 25740 }, { "grad_norm": 0.2241247594356537, "learning_rate": 8.868432932350712e-05, "loss": 0.0144, "step": 25750 }, { "grad_norm": 0.24645985662937164, "learning_rate": 8.867385134426272e-05, "loss": 0.0142, "step": 25760 }, { "grad_norm": 0.20086583495140076, "learning_rate": 8.866336913570407e-05, "loss": 0.0119, "step": 25770 }, { "grad_norm": 0.1671452522277832, "learning_rate": 8.865288269897751e-05, "loss": 0.0145, "step": 25780 }, { "grad_norm": 0.20030660927295685, "learning_rate": 8.864239203522981e-05, "loss": 0.0139, "step": 25790 }, { "grad_norm": 0.17258858680725098, "learning_rate": 8.863189714560822e-05, "loss": 0.0111, "step": 25800 }, { "grad_norm": 0.2101529985666275, "learning_rate": 8.862139803126043e-05, "loss": 0.013, "step": 25810 }, { "grad_norm": 0.18513555824756622, "learning_rate": 8.861089469333463e-05, "loss": 0.0135, "step": 25820 }, { "grad_norm": 0.1947997510433197, "learning_rate": 8.860038713297944e-05, "loss": 0.0147, "step": 25830 }, { "grad_norm": 0.1515681892633438, "learning_rate": 8.858987535134394e-05, "loss": 0.0132, "step": 25840 }, { "grad_norm": 0.12720175087451935, "learning_rate": 8.857935934957769e-05, "loss": 0.0128, "step": 25850 }, { "grad_norm": 0.22535455226898193, "learning_rate": 8.856883912883071e-05, "loss": 0.017, "step": 25860 }, { "grad_norm": 0.18896424770355225, "learning_rate": 8.855831469025346e-05, "loss": 0.0116, "step": 25870 }, { "grad_norm": 0.18304644525051117, "learning_rate": 8.854778603499689e-05, "loss": 0.0161, "step": 25880 }, { "grad_norm": 0.2087920904159546, "learning_rate": 8.85372531642124e-05, "loss": 0.0129, "step": 25890 }, { "grad_norm": 0.17395718395709991, "learning_rate": 8.852671607905185e-05, "loss": 0.0125, "step": 25900 }, { "grad_norm": 0.2155085653066635, "learning_rate": 8.851617478066754e-05, "loss": 0.0153, "step": 25910 }, { "grad_norm": 0.17688144743442535, "learning_rate": 8.850562927021227e-05, "loss": 0.0142, "step": 25920 }, { "grad_norm": 0.16710306704044342, "learning_rate": 8.849507954883928e-05, "loss": 0.0104, "step": 25930 }, { "grad_norm": 0.23663170635700226, "learning_rate": 8.848452561770226e-05, "loss": 0.0109, "step": 25940 }, { "grad_norm": 0.17030687630176544, "learning_rate": 8.847396747795538e-05, "loss": 0.0133, "step": 25950 }, { "grad_norm": 0.1644994616508484, "learning_rate": 8.846340513075327e-05, "loss": 0.0115, "step": 25960 }, { "grad_norm": 0.1537155658006668, "learning_rate": 8.845283857725099e-05, "loss": 0.01, "step": 25970 }, { "grad_norm": 0.17097675800323486, "learning_rate": 8.844226781860409e-05, "loss": 0.0122, "step": 25980 }, { "grad_norm": 0.16593217849731445, "learning_rate": 8.84316928559686e-05, "loss": 0.0121, "step": 25990 }, { "grad_norm": 0.1734667718410492, "learning_rate": 8.842111369050094e-05, "loss": 0.0107, "step": 26000 }, { "grad_norm": 0.2407815009355545, "learning_rate": 8.841053032335808e-05, "loss": 0.0174, "step": 26010 }, { "grad_norm": 0.1927841752767563, "learning_rate": 8.839994275569735e-05, "loss": 0.0119, "step": 26020 }, { "grad_norm": 0.21736566722393036, "learning_rate": 8.838935098867662e-05, "loss": 0.0156, "step": 26030 }, { "grad_norm": 0.22867707908153534, "learning_rate": 8.837875502345418e-05, "loss": 0.0122, "step": 26040 }, { "grad_norm": 0.16643336415290833, "learning_rate": 8.83681548611888e-05, "loss": 0.0108, "step": 26050 }, { "grad_norm": 0.18738418817520142, "learning_rate": 8.835755050303969e-05, "loss": 0.0126, "step": 26060 }, { "grad_norm": 0.19185394048690796, "learning_rate": 8.834694195016653e-05, "loss": 0.0148, "step": 26070 }, { "grad_norm": 0.15775811672210693, "learning_rate": 8.833632920372942e-05, "loss": 0.0131, "step": 26080 }, { "grad_norm": 0.1726207137107849, "learning_rate": 8.832571226488903e-05, "loss": 0.0148, "step": 26090 }, { "grad_norm": 0.23216570913791656, "learning_rate": 8.831509113480634e-05, "loss": 0.0144, "step": 26100 }, { "grad_norm": 0.17126275599002838, "learning_rate": 8.83044658146429e-05, "loss": 0.0125, "step": 26110 }, { "grad_norm": 0.21137309074401855, "learning_rate": 8.829383630556067e-05, "loss": 0.0118, "step": 26120 }, { "grad_norm": 0.19306710362434387, "learning_rate": 8.828320260872207e-05, "loss": 0.0134, "step": 26130 }, { "grad_norm": 0.2722136080265045, "learning_rate": 8.827256472529e-05, "loss": 0.0125, "step": 26140 }, { "grad_norm": 0.1697443723678589, "learning_rate": 8.826192265642778e-05, "loss": 0.0151, "step": 26150 }, { "grad_norm": 0.18902552127838135, "learning_rate": 8.825127640329923e-05, "loss": 0.0113, "step": 26160 }, { "grad_norm": 0.21966679394245148, "learning_rate": 8.824062596706861e-05, "loss": 0.0157, "step": 26170 }, { "grad_norm": 0.18207710981369019, "learning_rate": 8.822997134890062e-05, "loss": 0.011, "step": 26180 }, { "grad_norm": 0.14683079719543457, "learning_rate": 8.821931254996044e-05, "loss": 0.0114, "step": 26190 }, { "grad_norm": 0.19647565484046936, "learning_rate": 8.82086495714137e-05, "loss": 0.014, "step": 26200 }, { "grad_norm": 0.15263009071350098, "learning_rate": 8.81979824144265e-05, "loss": 0.0151, "step": 26210 }, { "grad_norm": 0.1997985988855362, "learning_rate": 8.818731108016536e-05, "loss": 0.0129, "step": 26220 }, { "grad_norm": 0.19391880929470062, "learning_rate": 8.81766355697973e-05, "loss": 0.0158, "step": 26230 }, { "grad_norm": 0.17095254361629486, "learning_rate": 8.816595588448977e-05, "loss": 0.0118, "step": 26240 }, { "grad_norm": 0.1997332125902176, "learning_rate": 8.81552720254107e-05, "loss": 0.0104, "step": 26250 }, { "grad_norm": 0.1834573745727539, "learning_rate": 8.814458399372842e-05, "loss": 0.0128, "step": 26260 }, { "grad_norm": 0.21168822050094604, "learning_rate": 8.813389179061181e-05, "loss": 0.0118, "step": 26270 }, { "grad_norm": 0.13471367955207825, "learning_rate": 8.812319541723012e-05, "loss": 0.0095, "step": 26280 }, { "grad_norm": 0.12961967289447784, "learning_rate": 8.811249487475309e-05, "loss": 0.0112, "step": 26290 }, { "grad_norm": 0.17902199923992157, "learning_rate": 8.810179016435092e-05, "loss": 0.0138, "step": 26300 }, { "grad_norm": 0.16644296050071716, "learning_rate": 8.809108128719428e-05, "loss": 0.0113, "step": 26310 }, { "grad_norm": 0.1988699734210968, "learning_rate": 8.808036824445424e-05, "loss": 0.0138, "step": 26320 }, { "grad_norm": 0.1575334668159485, "learning_rate": 8.806965103730238e-05, "loss": 0.0141, "step": 26330 }, { "grad_norm": 0.180674746632576, "learning_rate": 8.805892966691074e-05, "loss": 0.0134, "step": 26340 }, { "grad_norm": 0.1789298802614212, "learning_rate": 8.804820413445175e-05, "loss": 0.0156, "step": 26350 }, { "grad_norm": 0.21250484883785248, "learning_rate": 8.803747444109837e-05, "loss": 0.0132, "step": 26360 }, { "grad_norm": 0.18500444293022156, "learning_rate": 8.802674058802399e-05, "loss": 0.0119, "step": 26370 }, { "grad_norm": 0.20353342592716217, "learning_rate": 8.801600257640241e-05, "loss": 0.015, "step": 26380 }, { "grad_norm": 0.15889817476272583, "learning_rate": 8.800526040740795e-05, "loss": 0.0145, "step": 26390 }, { "grad_norm": 0.13079635798931122, "learning_rate": 8.799451408221535e-05, "loss": 0.0117, "step": 26400 }, { "grad_norm": 0.14566010236740112, "learning_rate": 8.798376360199982e-05, "loss": 0.0122, "step": 26410 }, { "grad_norm": 0.18524178862571716, "learning_rate": 8.797300896793701e-05, "loss": 0.0135, "step": 26420 }, { "grad_norm": 0.14233501255512238, "learning_rate": 8.796225018120302e-05, "loss": 0.0123, "step": 26430 }, { "grad_norm": 0.12872624397277832, "learning_rate": 8.795148724297444e-05, "loss": 0.0135, "step": 26440 }, { "grad_norm": 0.13489225506782532, "learning_rate": 8.794072015442825e-05, "loss": 0.0116, "step": 26450 }, { "grad_norm": 0.27789273858070374, "learning_rate": 8.792994891674198e-05, "loss": 0.0125, "step": 26460 }, { "grad_norm": 0.203682541847229, "learning_rate": 8.79191735310935e-05, "loss": 0.0124, "step": 26470 }, { "grad_norm": 0.19364796578884125, "learning_rate": 8.790839399866122e-05, "loss": 0.0133, "step": 26480 }, { "grad_norm": 0.19795116782188416, "learning_rate": 8.789761032062397e-05, "loss": 0.0137, "step": 26490 }, { "grad_norm": 0.1976529061794281, "learning_rate": 8.788682249816103e-05, "loss": 0.0113, "step": 26500 }, { "grad_norm": 0.17445603013038635, "learning_rate": 8.787603053245215e-05, "loss": 0.013, "step": 26510 }, { "grad_norm": 0.20874619483947754, "learning_rate": 8.78652344246775e-05, "loss": 0.0126, "step": 26520 }, { "grad_norm": 0.1705693006515503, "learning_rate": 8.785443417601776e-05, "loss": 0.0129, "step": 26530 }, { "grad_norm": 0.15904337167739868, "learning_rate": 8.784362978765401e-05, "loss": 0.0147, "step": 26540 }, { "grad_norm": 0.18971800804138184, "learning_rate": 8.783282126076779e-05, "loss": 0.0132, "step": 26550 }, { "grad_norm": 0.1252216249704361, "learning_rate": 8.782200859654112e-05, "loss": 0.0125, "step": 26560 }, { "grad_norm": 0.16951178014278412, "learning_rate": 8.781119179615646e-05, "loss": 0.0155, "step": 26570 }, { "grad_norm": 0.18087883293628693, "learning_rate": 8.780037086079674e-05, "loss": 0.0155, "step": 26580 }, { "grad_norm": 0.21687844395637512, "learning_rate": 8.778954579164527e-05, "loss": 0.0109, "step": 26590 }, { "grad_norm": 0.1491355001926422, "learning_rate": 8.777871658988588e-05, "loss": 0.0123, "step": 26600 }, { "grad_norm": 0.25356850028038025, "learning_rate": 8.776788325670285e-05, "loss": 0.0156, "step": 26610 }, { "grad_norm": 0.20993250608444214, "learning_rate": 8.775704579328089e-05, "loss": 0.0157, "step": 26620 }, { "grad_norm": 0.2832415699958801, "learning_rate": 8.774620420080517e-05, "loss": 0.0137, "step": 26630 }, { "grad_norm": 0.14419615268707275, "learning_rate": 8.773535848046131e-05, "loss": 0.0119, "step": 26640 }, { "grad_norm": 0.15914610028266907, "learning_rate": 8.772450863343538e-05, "loss": 0.0115, "step": 26650 }, { "grad_norm": 0.14849618077278137, "learning_rate": 8.77136546609139e-05, "loss": 0.0125, "step": 26660 }, { "grad_norm": 0.18260769546031952, "learning_rate": 8.770279656408385e-05, "loss": 0.0124, "step": 26670 }, { "grad_norm": 0.13248638808727264, "learning_rate": 8.769193434413265e-05, "loss": 0.0114, "step": 26680 }, { "grad_norm": 0.16249892115592957, "learning_rate": 8.76810680022482e-05, "loss": 0.0122, "step": 26690 }, { "grad_norm": 0.17050254344940186, "learning_rate": 8.767019753961878e-05, "loss": 0.0113, "step": 26700 }, { "grad_norm": 0.17459098994731903, "learning_rate": 8.765932295743321e-05, "loss": 0.0137, "step": 26710 }, { "grad_norm": 0.19303956627845764, "learning_rate": 8.764844425688068e-05, "loss": 0.0124, "step": 26720 }, { "grad_norm": 0.16301977634429932, "learning_rate": 8.763756143915092e-05, "loss": 0.0114, "step": 26730 }, { "grad_norm": 0.15516838431358337, "learning_rate": 8.7626674505434e-05, "loss": 0.0101, "step": 26740 }, { "grad_norm": 0.13784906268119812, "learning_rate": 8.761578345692053e-05, "loss": 0.0118, "step": 26750 }, { "grad_norm": 0.1764765828847885, "learning_rate": 8.760488829480156e-05, "loss": 0.0117, "step": 26760 }, { "grad_norm": 0.223662331700325, "learning_rate": 8.759398902026854e-05, "loss": 0.0123, "step": 26770 }, { "grad_norm": 0.20481404662132263, "learning_rate": 8.758308563451339e-05, "loss": 0.0149, "step": 26780 }, { "grad_norm": 0.2032148838043213, "learning_rate": 8.75721781387285e-05, "loss": 0.0096, "step": 26790 }, { "grad_norm": 0.2210971862077713, "learning_rate": 8.75612665341067e-05, "loss": 0.0134, "step": 26800 }, { "grad_norm": 0.1115608811378479, "learning_rate": 8.755035082184126e-05, "loss": 0.0127, "step": 26810 }, { "grad_norm": 0.17584377527236938, "learning_rate": 8.753943100312592e-05, "loss": 0.0119, "step": 26820 }, { "grad_norm": 0.17810989916324615, "learning_rate": 8.752850707915484e-05, "loss": 0.0138, "step": 26830 }, { "grad_norm": 0.15762169659137726, "learning_rate": 8.751757905112264e-05, "loss": 0.0122, "step": 26840 }, { "grad_norm": 0.1602567732334137, "learning_rate": 8.75066469202244e-05, "loss": 0.0119, "step": 26850 }, { "grad_norm": 0.2160285860300064, "learning_rate": 8.749571068765567e-05, "loss": 0.0108, "step": 26860 }, { "grad_norm": 0.2398572713136673, "learning_rate": 8.748477035461238e-05, "loss": 0.0135, "step": 26870 }, { "grad_norm": 0.23486751317977905, "learning_rate": 8.747382592229095e-05, "loss": 0.0133, "step": 26880 }, { "grad_norm": 0.17561957240104675, "learning_rate": 8.746287739188828e-05, "loss": 0.0153, "step": 26890 }, { "grad_norm": 0.15576685965061188, "learning_rate": 8.745192476460165e-05, "loss": 0.0117, "step": 26900 }, { "grad_norm": 0.1526782512664795, "learning_rate": 8.744096804162882e-05, "loss": 0.0125, "step": 26910 }, { "grad_norm": 0.1642904281616211, "learning_rate": 8.743000722416804e-05, "loss": 0.0125, "step": 26920 }, { "grad_norm": 0.23566602170467377, "learning_rate": 8.741904231341793e-05, "loss": 0.0125, "step": 26930 }, { "grad_norm": 0.148307666182518, "learning_rate": 8.740807331057762e-05, "loss": 0.0135, "step": 26940 }, { "grad_norm": 0.20725038647651672, "learning_rate": 8.739710021684667e-05, "loss": 0.0154, "step": 26950 }, { "grad_norm": 0.20534279942512512, "learning_rate": 8.738612303342503e-05, "loss": 0.0145, "step": 26960 }, { "grad_norm": 0.2302570790052414, "learning_rate": 8.73751417615132e-05, "loss": 0.0133, "step": 26970 }, { "grad_norm": 0.16863156855106354, "learning_rate": 8.736415640231208e-05, "loss": 0.0129, "step": 26980 }, { "grad_norm": 0.13675202429294586, "learning_rate": 8.735316695702297e-05, "loss": 0.0141, "step": 26990 }, { "grad_norm": 0.15841196477413177, "learning_rate": 8.734217342684769e-05, "loss": 0.0119, "step": 27000 }, { "grad_norm": 0.18140125274658203, "learning_rate": 8.733117581298847e-05, "loss": 0.0139, "step": 27010 }, { "grad_norm": 0.20932038128376007, "learning_rate": 8.732017411664796e-05, "loss": 0.0121, "step": 27020 }, { "grad_norm": 0.18537753820419312, "learning_rate": 8.730916833902936e-05, "loss": 0.0127, "step": 27030 }, { "grad_norm": 0.1480608433485031, "learning_rate": 8.729815848133618e-05, "loss": 0.0111, "step": 27040 }, { "grad_norm": 0.15883097052574158, "learning_rate": 8.728714454477247e-05, "loss": 0.0114, "step": 27050 }, { "grad_norm": 0.214971125125885, "learning_rate": 8.727612653054269e-05, "loss": 0.0145, "step": 27060 }, { "grad_norm": 0.20520029962062836, "learning_rate": 8.726510443985176e-05, "loss": 0.0123, "step": 27070 }, { "grad_norm": 0.24233901500701904, "learning_rate": 8.725407827390503e-05, "loss": 0.0145, "step": 27080 }, { "grad_norm": 0.15501591563224792, "learning_rate": 8.724304803390833e-05, "loss": 0.0104, "step": 27090 }, { "grad_norm": 0.16494347155094147, "learning_rate": 8.723201372106788e-05, "loss": 0.0116, "step": 27100 }, { "grad_norm": 0.14153525233268738, "learning_rate": 8.722097533659038e-05, "loss": 0.0111, "step": 27110 }, { "grad_norm": 0.15733009576797485, "learning_rate": 8.720993288168299e-05, "loss": 0.0129, "step": 27120 }, { "grad_norm": 0.20046205818653107, "learning_rate": 8.719888635755327e-05, "loss": 0.0122, "step": 27130 }, { "grad_norm": 0.1586742103099823, "learning_rate": 8.718783576540928e-05, "loss": 0.0108, "step": 27140 }, { "grad_norm": 0.19700618088245392, "learning_rate": 8.717678110645948e-05, "loss": 0.0149, "step": 27150 }, { "grad_norm": 0.1552387773990631, "learning_rate": 8.716572238191279e-05, "loss": 0.0155, "step": 27160 }, { "grad_norm": 0.24084992706775665, "learning_rate": 8.715465959297857e-05, "loss": 0.0161, "step": 27170 }, { "grad_norm": 0.15820126235485077, "learning_rate": 8.714359274086665e-05, "loss": 0.0128, "step": 27180 }, { "grad_norm": 0.22936886548995972, "learning_rate": 8.713252182678726e-05, "loss": 0.0155, "step": 27190 }, { "grad_norm": 0.1417044848203659, "learning_rate": 8.712144685195112e-05, "loss": 0.0119, "step": 27200 }, { "grad_norm": 0.19473199546337128, "learning_rate": 8.711036781756936e-05, "loss": 0.0132, "step": 27210 }, { "grad_norm": 0.1608353853225708, "learning_rate": 8.709928472485357e-05, "loss": 0.0103, "step": 27220 }, { "grad_norm": 0.19978705048561096, "learning_rate": 8.708819757501579e-05, "loss": 0.0117, "step": 27230 }, { "grad_norm": 0.2092844694852829, "learning_rate": 8.707710636926846e-05, "loss": 0.0121, "step": 27240 }, { "grad_norm": 0.21097397804260254, "learning_rate": 8.706601110882455e-05, "loss": 0.0135, "step": 27250 }, { "grad_norm": 0.1970789134502411, "learning_rate": 8.705491179489738e-05, "loss": 0.0128, "step": 27260 }, { "grad_norm": 0.1459755301475525, "learning_rate": 8.704380842870077e-05, "loss": 0.0119, "step": 27270 }, { "grad_norm": 0.1355750858783722, "learning_rate": 8.703270101144895e-05, "loss": 0.0114, "step": 27280 }, { "grad_norm": 0.14056865870952606, "learning_rate": 8.702158954435664e-05, "loss": 0.0125, "step": 27290 }, { "grad_norm": 0.18718481063842773, "learning_rate": 8.701047402863896e-05, "loss": 0.0151, "step": 27300 }, { "grad_norm": 0.1740855723619461, "learning_rate": 8.699935446551148e-05, "loss": 0.0135, "step": 27310 }, { "grad_norm": 0.1915011703968048, "learning_rate": 8.698823085619022e-05, "loss": 0.0125, "step": 27320 }, { "grad_norm": 0.1701381653547287, "learning_rate": 8.697710320189166e-05, "loss": 0.0105, "step": 27330 }, { "grad_norm": 0.1833663135766983, "learning_rate": 8.696597150383268e-05, "loss": 0.0109, "step": 27340 }, { "grad_norm": 0.19623830914497375, "learning_rate": 8.695483576323063e-05, "loss": 0.0114, "step": 27350 }, { "grad_norm": 0.18595965206623077, "learning_rate": 8.69436959813033e-05, "loss": 0.0108, "step": 27360 }, { "grad_norm": 0.17437423765659332, "learning_rate": 8.693255215926892e-05, "loss": 0.0128, "step": 27370 }, { "grad_norm": 0.22513309121131897, "learning_rate": 8.692140429834617e-05, "loss": 0.0138, "step": 27380 }, { "grad_norm": 0.15185967087745667, "learning_rate": 8.691025239975415e-05, "loss": 0.0122, "step": 27390 }, { "grad_norm": 0.17652985453605652, "learning_rate": 8.689909646471243e-05, "loss": 0.0107, "step": 27400 }, { "grad_norm": 0.1534751057624817, "learning_rate": 8.688793649444099e-05, "loss": 0.0119, "step": 27410 }, { "grad_norm": 0.21219758689403534, "learning_rate": 8.687677249016029e-05, "loss": 0.0122, "step": 27420 }, { "grad_norm": 0.24449951946735382, "learning_rate": 8.686560445309118e-05, "loss": 0.0144, "step": 27430 }, { "grad_norm": 0.21869267523288727, "learning_rate": 8.685443238445499e-05, "loss": 0.0124, "step": 27440 }, { "grad_norm": 0.1713731586933136, "learning_rate": 8.68432562854735e-05, "loss": 0.0137, "step": 27450 }, { "grad_norm": 0.1686079055070877, "learning_rate": 8.683207615736887e-05, "loss": 0.0105, "step": 27460 }, { "grad_norm": 0.1416277289390564, "learning_rate": 8.682089200136379e-05, "loss": 0.0138, "step": 27470 }, { "grad_norm": 0.1817556917667389, "learning_rate": 8.680970381868132e-05, "loss": 0.0127, "step": 27480 }, { "grad_norm": 0.18767882883548737, "learning_rate": 8.679851161054498e-05, "loss": 0.0107, "step": 27490 }, { "grad_norm": 0.16310997307300568, "learning_rate": 8.678731537817873e-05, "loss": 0.0119, "step": 27500 }, { "grad_norm": 0.12362325936555862, "learning_rate": 8.677611512280697e-05, "loss": 0.0108, "step": 27510 }, { "grad_norm": 0.14531071484088898, "learning_rate": 8.676491084565457e-05, "loss": 0.0119, "step": 27520 }, { "grad_norm": 0.15874192118644714, "learning_rate": 8.675370254794678e-05, "loss": 0.0097, "step": 27530 }, { "grad_norm": 0.16462978720664978, "learning_rate": 8.674249023090935e-05, "loss": 0.0126, "step": 27540 }, { "grad_norm": 0.16724243760108948, "learning_rate": 8.673127389576843e-05, "loss": 0.0124, "step": 27550 }, { "grad_norm": 0.19021695852279663, "learning_rate": 8.67200535437506e-05, "loss": 0.0109, "step": 27560 }, { "grad_norm": 0.1720370650291443, "learning_rate": 8.670882917608296e-05, "loss": 0.0141, "step": 27570 }, { "grad_norm": 0.17378249764442444, "learning_rate": 8.669760079399292e-05, "loss": 0.0121, "step": 27580 }, { "grad_norm": 0.1788606494665146, "learning_rate": 8.668636839870845e-05, "loss": 0.0151, "step": 27590 }, { "grad_norm": 0.15805238485336304, "learning_rate": 8.667513199145789e-05, "loss": 0.0134, "step": 27600 }, { "grad_norm": 0.20434682071208954, "learning_rate": 8.666389157347002e-05, "loss": 0.0137, "step": 27610 }, { "grad_norm": 0.21333545446395874, "learning_rate": 8.66526471459741e-05, "loss": 0.0173, "step": 27620 }, { "grad_norm": 0.15115885436534882, "learning_rate": 8.66413987101998e-05, "loss": 0.0158, "step": 27630 }, { "grad_norm": 0.21360550820827484, "learning_rate": 8.663014626737723e-05, "loss": 0.0127, "step": 27640 }, { "grad_norm": 0.16974292695522308, "learning_rate": 8.661888981873691e-05, "loss": 0.0124, "step": 27650 }, { "grad_norm": 0.19048155844211578, "learning_rate": 8.660762936550988e-05, "loss": 0.0118, "step": 27660 }, { "grad_norm": 0.17271068692207336, "learning_rate": 8.659636490892753e-05, "loss": 0.0111, "step": 27670 }, { "grad_norm": 0.22066934406757355, "learning_rate": 8.658509645022174e-05, "loss": 0.0101, "step": 27680 }, { "grad_norm": 0.2056097686290741, "learning_rate": 8.657382399062481e-05, "loss": 0.0125, "step": 27690 }, { "grad_norm": 0.24425670504570007, "learning_rate": 8.656254753136946e-05, "loss": 0.0145, "step": 27700 }, { "grad_norm": 0.17407816648483276, "learning_rate": 8.655126707368891e-05, "loss": 0.0122, "step": 27710 }, { "grad_norm": 0.16981010138988495, "learning_rate": 8.653998261881672e-05, "loss": 0.0102, "step": 27720 }, { "grad_norm": 0.21127402782440186, "learning_rate": 8.652869416798699e-05, "loss": 0.0101, "step": 27730 }, { "grad_norm": 0.16775979101657867, "learning_rate": 8.651740172243417e-05, "loss": 0.0122, "step": 27740 }, { "grad_norm": 0.15852561593055725, "learning_rate": 8.65061052833932e-05, "loss": 0.0124, "step": 27750 }, { "grad_norm": 0.16645704209804535, "learning_rate": 8.649480485209945e-05, "loss": 0.0118, "step": 27760 }, { "grad_norm": 0.182280495762825, "learning_rate": 8.64835004297887e-05, "loss": 0.0137, "step": 27770 }, { "grad_norm": 0.16793939471244812, "learning_rate": 8.64721920176972e-05, "loss": 0.0103, "step": 27780 }, { "grad_norm": 0.16052477061748505, "learning_rate": 8.646087961706164e-05, "loss": 0.0128, "step": 27790 }, { "grad_norm": 0.17697101831436157, "learning_rate": 8.644956322911908e-05, "loss": 0.0109, "step": 27800 }, { "grad_norm": 0.21743819117546082, "learning_rate": 8.643824285510709e-05, "loss": 0.0113, "step": 27810 }, { "grad_norm": 0.19984635710716248, "learning_rate": 8.642691849626364e-05, "loss": 0.0133, "step": 27820 }, { "grad_norm": 0.20479539036750793, "learning_rate": 8.641559015382717e-05, "loss": 0.0111, "step": 27830 }, { "grad_norm": 0.15477047860622406, "learning_rate": 8.640425782903649e-05, "loss": 0.011, "step": 27840 }, { "grad_norm": 0.15573233366012573, "learning_rate": 8.639292152313091e-05, "loss": 0.0117, "step": 27850 }, { "grad_norm": 0.1722409874200821, "learning_rate": 8.638158123735015e-05, "loss": 0.0119, "step": 27860 }, { "grad_norm": 0.169804185628891, "learning_rate": 8.637023697293436e-05, "loss": 0.013, "step": 27870 }, { "grad_norm": 0.12297607213258743, "learning_rate": 8.635888873112414e-05, "loss": 0.0121, "step": 27880 }, { "grad_norm": 0.18549078702926636, "learning_rate": 8.634753651316052e-05, "loss": 0.0137, "step": 27890 }, { "grad_norm": 0.17054550349712372, "learning_rate": 8.633618032028496e-05, "loss": 0.011, "step": 27900 }, { "grad_norm": 0.22351476550102234, "learning_rate": 8.632482015373934e-05, "loss": 0.0147, "step": 27910 }, { "grad_norm": 0.16741596162319183, "learning_rate": 8.6313456014766e-05, "loss": 0.0104, "step": 27920 }, { "grad_norm": 0.1739085614681244, "learning_rate": 8.630208790460771e-05, "loss": 0.0139, "step": 27930 }, { "grad_norm": 0.23412632942199707, "learning_rate": 8.629071582450768e-05, "loss": 0.0117, "step": 27940 }, { "grad_norm": 0.21590717136859894, "learning_rate": 8.62793397757095e-05, "loss": 0.0169, "step": 27950 }, { "grad_norm": 0.15180888772010803, "learning_rate": 8.626795975945729e-05, "loss": 0.0119, "step": 27960 }, { "grad_norm": 0.17749372124671936, "learning_rate": 8.625657577699551e-05, "loss": 0.0138, "step": 27970 }, { "grad_norm": 0.15637080371379852, "learning_rate": 8.624518782956914e-05, "loss": 0.0117, "step": 27980 }, { "grad_norm": 0.18491248786449432, "learning_rate": 8.62337959184235e-05, "loss": 0.013, "step": 27990 }, { "grad_norm": 0.19460029900074005, "learning_rate": 8.622240004480441e-05, "loss": 0.0123, "step": 28000 }, { "grad_norm": 0.17954260110855103, "learning_rate": 8.621100020995814e-05, "loss": 0.0138, "step": 28010 }, { "grad_norm": 0.1741485446691513, "learning_rate": 8.619959641513132e-05, "loss": 0.0128, "step": 28020 }, { "grad_norm": 0.1633070558309555, "learning_rate": 8.618818866157105e-05, "loss": 0.0122, "step": 28030 }, { "grad_norm": 0.14224250614643097, "learning_rate": 8.617677695052487e-05, "loss": 0.0127, "step": 28040 }, { "grad_norm": 0.17247523367404938, "learning_rate": 8.616536128324078e-05, "loss": 0.0161, "step": 28050 }, { "grad_norm": 0.20270779728889465, "learning_rate": 8.615394166096712e-05, "loss": 0.0128, "step": 28060 }, { "grad_norm": 0.1919977068901062, "learning_rate": 8.614251808495279e-05, "loss": 0.0116, "step": 28070 }, { "grad_norm": 0.18650144338607788, "learning_rate": 8.6131090556447e-05, "loss": 0.0115, "step": 28080 }, { "grad_norm": 0.14465400576591492, "learning_rate": 8.611965907669947e-05, "loss": 0.01, "step": 28090 }, { "grad_norm": 0.23658542335033417, "learning_rate": 8.610822364696034e-05, "loss": 0.0145, "step": 28100 }, { "grad_norm": 0.19222794473171234, "learning_rate": 8.609678426848015e-05, "loss": 0.0106, "step": 28110 }, { "grad_norm": 0.20535853505134583, "learning_rate": 8.60853409425099e-05, "loss": 0.0117, "step": 28120 }, { "grad_norm": 0.14301535487174988, "learning_rate": 8.607389367030104e-05, "loss": 0.0113, "step": 28130 }, { "grad_norm": 0.18798382580280304, "learning_rate": 8.606244245310538e-05, "loss": 0.0126, "step": 28140 }, { "grad_norm": 0.18158289790153503, "learning_rate": 8.605098729217525e-05, "loss": 0.0101, "step": 28150 }, { "grad_norm": 0.12315775454044342, "learning_rate": 8.603952818876335e-05, "loss": 0.0129, "step": 28160 }, { "grad_norm": 0.14829644560813904, "learning_rate": 8.602806514412281e-05, "loss": 0.0144, "step": 28170 }, { "grad_norm": 0.1487744301557541, "learning_rate": 8.601659815950726e-05, "loss": 0.0113, "step": 28180 }, { "grad_norm": 0.1868462711572647, "learning_rate": 8.600512723617067e-05, "loss": 0.0139, "step": 28190 }, { "grad_norm": 0.1575184166431427, "learning_rate": 8.59936523753675e-05, "loss": 0.0132, "step": 28200 }, { "grad_norm": 0.13255220651626587, "learning_rate": 8.598217357835264e-05, "loss": 0.0118, "step": 28210 }, { "grad_norm": 0.17698588967323303, "learning_rate": 8.597069084638135e-05, "loss": 0.0137, "step": 28220 }, { "grad_norm": 0.18847377598285675, "learning_rate": 8.595920418070939e-05, "loss": 0.0112, "step": 28230 }, { "grad_norm": 0.2126147747039795, "learning_rate": 8.594771358259295e-05, "loss": 0.0136, "step": 28240 }, { "grad_norm": 0.2776000499725342, "learning_rate": 8.593621905328858e-05, "loss": 0.0117, "step": 28250 }, { "grad_norm": 0.18482354283332825, "learning_rate": 8.592472059405333e-05, "loss": 0.0128, "step": 28260 }, { "grad_norm": 0.16425025463104248, "learning_rate": 8.591321820614464e-05, "loss": 0.0102, "step": 28270 }, { "grad_norm": 0.13288474082946777, "learning_rate": 8.590171189082041e-05, "loss": 0.0121, "step": 28280 }, { "grad_norm": 0.16538110375404358, "learning_rate": 8.589020164933894e-05, "loss": 0.0102, "step": 28290 }, { "grad_norm": 0.17050686478614807, "learning_rate": 8.587868748295898e-05, "loss": 0.0118, "step": 28300 }, { "grad_norm": 0.12109918147325516, "learning_rate": 8.586716939293971e-05, "loss": 0.0134, "step": 28310 }, { "grad_norm": 0.19437478482723236, "learning_rate": 8.58556473805407e-05, "loss": 0.0132, "step": 28320 }, { "grad_norm": 0.11668483912944794, "learning_rate": 8.584412144702202e-05, "loss": 0.0103, "step": 28330 }, { "grad_norm": 0.13127432763576508, "learning_rate": 8.58325915936441e-05, "loss": 0.0107, "step": 28340 }, { "grad_norm": 0.13560450077056885, "learning_rate": 8.582105782166783e-05, "loss": 0.0105, "step": 28350 }, { "grad_norm": 0.20630750060081482, "learning_rate": 8.580952013235455e-05, "loss": 0.0098, "step": 28360 }, { "grad_norm": 0.25271347165107727, "learning_rate": 8.579797852696596e-05, "loss": 0.0107, "step": 28370 }, { "grad_norm": 0.13363461196422577, "learning_rate": 8.578643300676428e-05, "loss": 0.0127, "step": 28380 }, { "grad_norm": 0.16512177884578705, "learning_rate": 8.577488357301209e-05, "loss": 0.0131, "step": 28390 }, { "grad_norm": 0.1329052597284317, "learning_rate": 8.576333022697242e-05, "loss": 0.0097, "step": 28400 }, { "grad_norm": 0.15590634942054749, "learning_rate": 8.575177296990873e-05, "loss": 0.0153, "step": 28410 }, { "grad_norm": 0.16486866772174835, "learning_rate": 8.574021180308489e-05, "loss": 0.0112, "step": 28420 }, { "grad_norm": 0.17914453148841858, "learning_rate": 8.572864672776523e-05, "loss": 0.012, "step": 28430 }, { "grad_norm": 0.17339877784252167, "learning_rate": 8.571707774521447e-05, "loss": 0.0118, "step": 28440 }, { "grad_norm": 0.21132338047027588, "learning_rate": 8.57055048566978e-05, "loss": 0.0125, "step": 28450 }, { "grad_norm": 0.19764839112758636, "learning_rate": 8.569392806348078e-05, "loss": 0.0119, "step": 28460 }, { "grad_norm": 0.19443993270397186, "learning_rate": 8.568234736682947e-05, "loss": 0.014, "step": 28470 }, { "grad_norm": 0.19632206857204437, "learning_rate": 8.567076276801029e-05, "loss": 0.0134, "step": 28480 }, { "grad_norm": 0.14459513127803802, "learning_rate": 8.565917426829013e-05, "loss": 0.0123, "step": 28490 }, { "grad_norm": 0.195736363530159, "learning_rate": 8.564758186893628e-05, "loss": 0.0128, "step": 28500 }, { "grad_norm": 0.14520540833473206, "learning_rate": 8.563598557121649e-05, "loss": 0.0097, "step": 28510 }, { "grad_norm": 0.17589855194091797, "learning_rate": 8.562438537639888e-05, "loss": 0.0116, "step": 28520 }, { "grad_norm": 0.16953393816947937, "learning_rate": 8.561278128575206e-05, "loss": 0.013, "step": 28530 }, { "grad_norm": 0.18467073142528534, "learning_rate": 8.5601173300545e-05, "loss": 0.0122, "step": 28540 }, { "grad_norm": 0.15035229921340942, "learning_rate": 8.558956142204717e-05, "loss": 0.0132, "step": 28550 }, { "grad_norm": 0.1311822086572647, "learning_rate": 8.55779456515284e-05, "loss": 0.0132, "step": 28560 }, { "grad_norm": 0.10665658116340637, "learning_rate": 8.556632599025898e-05, "loss": 0.0117, "step": 28570 }, { "grad_norm": 0.38170239329338074, "learning_rate": 8.555470243950964e-05, "loss": 0.0122, "step": 28580 }, { "grad_norm": 0.1989929974079132, "learning_rate": 8.554307500055148e-05, "loss": 0.0117, "step": 28590 }, { "grad_norm": 0.14548656344413757, "learning_rate": 8.553144367465609e-05, "loss": 0.0115, "step": 28600 }, { "grad_norm": 0.18556725978851318, "learning_rate": 8.551980846309544e-05, "loss": 0.0117, "step": 28610 }, { "grad_norm": 0.16977763175964355, "learning_rate": 8.550816936714193e-05, "loss": 0.0116, "step": 28620 }, { "grad_norm": 0.18181151151657104, "learning_rate": 8.549652638806841e-05, "loss": 0.0144, "step": 28630 }, { "grad_norm": 0.1218993067741394, "learning_rate": 8.548487952714812e-05, "loss": 0.0106, "step": 28640 }, { "grad_norm": 0.19740773737430573, "learning_rate": 8.547322878565478e-05, "loss": 0.0127, "step": 28650 }, { "grad_norm": 0.1727212518453598, "learning_rate": 8.546157416486245e-05, "loss": 0.01, "step": 28660 }, { "grad_norm": 0.1448899209499359, "learning_rate": 8.54499156660457e-05, "loss": 0.0109, "step": 28670 }, { "grad_norm": 0.2599235475063324, "learning_rate": 8.543825329047947e-05, "loss": 0.0135, "step": 28680 }, { "grad_norm": 0.17839248478412628, "learning_rate": 8.542658703943913e-05, "loss": 0.0126, "step": 28690 }, { "grad_norm": 0.17869208753108978, "learning_rate": 8.541491691420051e-05, "loss": 0.0135, "step": 28700 }, { "grad_norm": 0.14606435596942902, "learning_rate": 8.54032429160398e-05, "loss": 0.013, "step": 28710 }, { "grad_norm": 0.16835461556911469, "learning_rate": 8.539156504623369e-05, "loss": 0.0109, "step": 28720 }, { "grad_norm": 0.1743796318769455, "learning_rate": 8.537988330605923e-05, "loss": 0.0126, "step": 28730 }, { "grad_norm": 0.15522488951683044, "learning_rate": 8.536819769679393e-05, "loss": 0.0112, "step": 28740 }, { "grad_norm": 0.16332951188087463, "learning_rate": 8.53565082197157e-05, "loss": 0.013, "step": 28750 }, { "grad_norm": 0.17973071336746216, "learning_rate": 8.534481487610289e-05, "loss": 0.0123, "step": 28760 }, { "grad_norm": 0.18072037398815155, "learning_rate": 8.533311766723428e-05, "loss": 0.0133, "step": 28770 }, { "grad_norm": 0.17038904130458832, "learning_rate": 8.532141659438901e-05, "loss": 0.0131, "step": 28780 }, { "grad_norm": 0.1795405000448227, "learning_rate": 8.530971165884675e-05, "loss": 0.0106, "step": 28790 }, { "grad_norm": 0.21770162880420685, "learning_rate": 8.529800286188752e-05, "loss": 0.0108, "step": 28800 }, { "grad_norm": 0.1951054036617279, "learning_rate": 8.528629020479175e-05, "loss": 0.0108, "step": 28810 }, { "grad_norm": 0.17203909158706665, "learning_rate": 8.527457368884033e-05, "loss": 0.0141, "step": 28820 }, { "grad_norm": 0.14099010825157166, "learning_rate": 8.526285331531458e-05, "loss": 0.0104, "step": 28830 }, { "grad_norm": 0.16297869384288788, "learning_rate": 8.525112908549621e-05, "loss": 0.0127, "step": 28840 }, { "grad_norm": 0.17621545493602753, "learning_rate": 8.523940100066735e-05, "loss": 0.011, "step": 28850 }, { "grad_norm": 0.1343572735786438, "learning_rate": 8.52276690621106e-05, "loss": 0.0116, "step": 28860 }, { "grad_norm": 0.21175235509872437, "learning_rate": 8.521593327110889e-05, "loss": 0.0121, "step": 28870 }, { "grad_norm": 0.18981371819972992, "learning_rate": 8.520419362894569e-05, "loss": 0.0115, "step": 28880 }, { "grad_norm": 0.14723853766918182, "learning_rate": 8.51924501369048e-05, "loss": 0.0105, "step": 28890 }, { "grad_norm": 0.16561101377010345, "learning_rate": 8.518070279627047e-05, "loss": 0.012, "step": 28900 }, { "grad_norm": 0.1924734264612198, "learning_rate": 8.516895160832737e-05, "loss": 0.0105, "step": 28910 }, { "grad_norm": 0.1606590449810028, "learning_rate": 8.515719657436061e-05, "loss": 0.0096, "step": 28920 }, { "grad_norm": 0.1661025583744049, "learning_rate": 8.514543769565568e-05, "loss": 0.0136, "step": 28930 }, { "grad_norm": 0.09772367775440216, "learning_rate": 8.513367497349853e-05, "loss": 0.0097, "step": 28940 }, { "grad_norm": 0.17950855195522308, "learning_rate": 8.51219084091755e-05, "loss": 0.0113, "step": 28950 }, { "grad_norm": 0.16356463730335236, "learning_rate": 8.511013800397338e-05, "loss": 0.0142, "step": 28960 }, { "grad_norm": 0.17962303757667542, "learning_rate": 8.509836375917937e-05, "loss": 0.0114, "step": 28970 }, { "grad_norm": 0.204096257686615, "learning_rate": 8.508658567608104e-05, "loss": 0.0116, "step": 28980 }, { "grad_norm": 0.23898816108703613, "learning_rate": 8.507480375596647e-05, "loss": 0.0139, "step": 28990 }, { "grad_norm": 0.16967670619487762, "learning_rate": 8.506301800012408e-05, "loss": 0.0122, "step": 29000 }, { "grad_norm": 0.15201731026172638, "learning_rate": 8.505122840984278e-05, "loss": 0.0111, "step": 29010 }, { "grad_norm": 0.1477261632680893, "learning_rate": 8.503943498641182e-05, "loss": 0.0122, "step": 29020 }, { "grad_norm": 0.17595252394676208, "learning_rate": 8.502763773112095e-05, "loss": 0.0108, "step": 29030 }, { "grad_norm": 0.22302208840847015, "learning_rate": 8.501583664526026e-05, "loss": 0.0149, "step": 29040 }, { "grad_norm": 0.1656947284936905, "learning_rate": 8.500403173012032e-05, "loss": 0.0089, "step": 29050 }, { "grad_norm": 0.1467975676059723, "learning_rate": 8.499222298699211e-05, "loss": 0.0113, "step": 29060 }, { "grad_norm": 0.17187000811100006, "learning_rate": 8.498041041716701e-05, "loss": 0.0123, "step": 29070 }, { "grad_norm": 0.15208874642848969, "learning_rate": 8.496859402193681e-05, "loss": 0.0118, "step": 29080 }, { "grad_norm": 0.17004595696926117, "learning_rate": 8.495677380259374e-05, "loss": 0.0108, "step": 29090 }, { "grad_norm": 0.21248628199100494, "learning_rate": 8.494494976043045e-05, "loss": 0.0125, "step": 29100 }, { "grad_norm": 0.15106740593910217, "learning_rate": 8.493312189673998e-05, "loss": 0.0106, "step": 29110 }, { "grad_norm": 0.19516794383525848, "learning_rate": 8.492129021281584e-05, "loss": 0.0127, "step": 29120 }, { "grad_norm": 0.20614638924598694, "learning_rate": 8.490945470995188e-05, "loss": 0.0124, "step": 29130 }, { "grad_norm": 0.13970761001110077, "learning_rate": 8.489761538944247e-05, "loss": 0.0141, "step": 29140 }, { "grad_norm": 0.19784139096736908, "learning_rate": 8.48857722525823e-05, "loss": 0.013, "step": 29150 }, { "grad_norm": 0.18204480409622192, "learning_rate": 8.487392530066652e-05, "loss": 0.0141, "step": 29160 }, { "grad_norm": 0.16426275670528412, "learning_rate": 8.486207453499069e-05, "loss": 0.0122, "step": 29170 }, { "grad_norm": 0.12492232024669647, "learning_rate": 8.485021995685082e-05, "loss": 0.012, "step": 29180 }, { "grad_norm": 0.16010722517967224, "learning_rate": 8.483836156754328e-05, "loss": 0.012, "step": 29190 }, { "grad_norm": 0.13992910087108612, "learning_rate": 8.482649936836491e-05, "loss": 0.0095, "step": 29200 }, { "grad_norm": 0.17694757878780365, "learning_rate": 8.481463336061293e-05, "loss": 0.011, "step": 29210 }, { "grad_norm": 0.15848799049854279, "learning_rate": 8.480276354558496e-05, "loss": 0.0114, "step": 29220 }, { "grad_norm": 0.2123221904039383, "learning_rate": 8.479088992457913e-05, "loss": 0.0128, "step": 29230 }, { "grad_norm": 0.16417750716209412, "learning_rate": 8.477901249889387e-05, "loss": 0.0153, "step": 29240 }, { "grad_norm": 0.18548166751861572, "learning_rate": 8.47671312698281e-05, "loss": 0.0128, "step": 29250 }, { "grad_norm": 0.14207723736763, "learning_rate": 8.475524623868112e-05, "loss": 0.0109, "step": 29260 }, { "grad_norm": 0.1190791130065918, "learning_rate": 8.474335740675266e-05, "loss": 0.0109, "step": 29270 }, { "grad_norm": 0.1561785489320755, "learning_rate": 8.473146477534289e-05, "loss": 0.0108, "step": 29280 }, { "grad_norm": 0.17165908217430115, "learning_rate": 8.471956834575232e-05, "loss": 0.0124, "step": 29290 }, { "grad_norm": 0.15586775541305542, "learning_rate": 8.470766811928197e-05, "loss": 0.0129, "step": 29300 }, { "grad_norm": 0.14129386842250824, "learning_rate": 8.469576409723323e-05, "loss": 0.011, "step": 29310 }, { "grad_norm": 0.16319586336612701, "learning_rate": 8.468385628090788e-05, "loss": 0.0116, "step": 29320 }, { "grad_norm": 0.16050194203853607, "learning_rate": 8.467194467160815e-05, "loss": 0.0129, "step": 29330 }, { "grad_norm": 0.17872023582458496, "learning_rate": 8.466002927063667e-05, "loss": 0.0102, "step": 29340 }, { "grad_norm": 0.15570642054080963, "learning_rate": 8.464811007929651e-05, "loss": 0.012, "step": 29350 }, { "grad_norm": 0.18738198280334473, "learning_rate": 8.463618709889114e-05, "loss": 0.0125, "step": 29360 }, { "grad_norm": 0.20736807584762573, "learning_rate": 8.462426033072442e-05, "loss": 0.0123, "step": 29370 }, { "grad_norm": 0.19192080199718475, "learning_rate": 8.461232977610061e-05, "loss": 0.0112, "step": 29380 }, { "grad_norm": 0.17545169591903687, "learning_rate": 8.46003954363245e-05, "loss": 0.0113, "step": 29390 }, { "grad_norm": 0.2008998543024063, "learning_rate": 8.458845731270115e-05, "loss": 0.0112, "step": 29400 }, { "grad_norm": 0.14887477457523346, "learning_rate": 8.45765154065361e-05, "loss": 0.0093, "step": 29410 }, { "grad_norm": 0.17135284841060638, "learning_rate": 8.456456971913532e-05, "loss": 0.0091, "step": 29420 }, { "grad_norm": 0.1844160407781601, "learning_rate": 8.455262025180517e-05, "loss": 0.0116, "step": 29430 }, { "grad_norm": 0.148481085896492, "learning_rate": 8.454066700585242e-05, "loss": 0.0111, "step": 29440 }, { "grad_norm": 0.12438088655471802, "learning_rate": 8.452870998258423e-05, "loss": 0.0112, "step": 29450 }, { "grad_norm": 0.1381150782108307, "learning_rate": 8.451674918330825e-05, "loss": 0.0123, "step": 29460 }, { "grad_norm": 0.17775753140449524, "learning_rate": 8.450478460933246e-05, "loss": 0.0099, "step": 29470 }, { "grad_norm": 0.12229477614164352, "learning_rate": 8.449281626196532e-05, "loss": 0.0117, "step": 29480 }, { "grad_norm": 0.16393204033374786, "learning_rate": 8.448084414251564e-05, "loss": 0.0098, "step": 29490 }, { "grad_norm": 0.17003987729549408, "learning_rate": 8.446886825229271e-05, "loss": 0.0116, "step": 29500 }, { "grad_norm": 0.09769846498966217, "learning_rate": 8.445688859260615e-05, "loss": 0.0122, "step": 29510 }, { "grad_norm": 0.21751521527767181, "learning_rate": 8.444490516476606e-05, "loss": 0.013, "step": 29520 }, { "grad_norm": 0.1707538664340973, "learning_rate": 8.443291797008293e-05, "loss": 0.0099, "step": 29530 }, { "grad_norm": 0.19275642931461334, "learning_rate": 8.442092700986765e-05, "loss": 0.0148, "step": 29540 }, { "grad_norm": 0.17040853202342987, "learning_rate": 8.440893228543156e-05, "loss": 0.0114, "step": 29550 }, { "grad_norm": 0.20632171630859375, "learning_rate": 8.439693379808638e-05, "loss": 0.0126, "step": 29560 }, { "grad_norm": 0.1835964322090149, "learning_rate": 8.43849315491442e-05, "loss": 0.0117, "step": 29570 }, { "grad_norm": 0.14387109875679016, "learning_rate": 8.437292553991763e-05, "loss": 0.011, "step": 29580 }, { "grad_norm": 0.13408495485782623, "learning_rate": 8.436091577171959e-05, "loss": 0.0105, "step": 29590 }, { "grad_norm": 0.21738548576831818, "learning_rate": 8.434890224586347e-05, "loss": 0.0136, "step": 29600 }, { "grad_norm": 0.2255699634552002, "learning_rate": 8.433688496366303e-05, "loss": 0.0144, "step": 29610 }, { "grad_norm": 0.19527935981750488, "learning_rate": 8.432486392643248e-05, "loss": 0.0125, "step": 29620 }, { "grad_norm": 0.16935454308986664, "learning_rate": 8.431283913548643e-05, "loss": 0.0104, "step": 29630 }, { "grad_norm": 0.18069447576999664, "learning_rate": 8.430081059213985e-05, "loss": 0.0108, "step": 29640 }, { "grad_norm": 0.1425824612379074, "learning_rate": 8.428877829770823e-05, "loss": 0.0108, "step": 29650 }, { "grad_norm": 0.16050319373607635, "learning_rate": 8.427674225350735e-05, "loss": 0.0113, "step": 29660 }, { "grad_norm": 0.1720908284187317, "learning_rate": 8.426470246085347e-05, "loss": 0.0096, "step": 29670 }, { "grad_norm": 0.17172478139400482, "learning_rate": 8.425265892106324e-05, "loss": 0.0099, "step": 29680 }, { "grad_norm": 0.1525202989578247, "learning_rate": 8.424061163545374e-05, "loss": 0.0112, "step": 29690 }, { "grad_norm": 0.1769842952489853, "learning_rate": 8.422856060534243e-05, "loss": 0.0114, "step": 29700 }, { "grad_norm": 0.15653394162654877, "learning_rate": 8.421650583204718e-05, "loss": 0.0114, "step": 29710 }, { "grad_norm": 0.18709728121757507, "learning_rate": 8.420444731688633e-05, "loss": 0.0108, "step": 29720 }, { "grad_norm": 0.15012957155704498, "learning_rate": 8.419238506117852e-05, "loss": 0.0148, "step": 29730 }, { "grad_norm": 0.18739144504070282, "learning_rate": 8.418031906624289e-05, "loss": 0.0101, "step": 29740 }, { "grad_norm": 0.18141710758209229, "learning_rate": 8.416824933339898e-05, "loss": 0.013, "step": 29750 }, { "grad_norm": 0.16306209564208984, "learning_rate": 8.415617586396667e-05, "loss": 0.012, "step": 29760 }, { "grad_norm": 0.17036955058574677, "learning_rate": 8.414409865926632e-05, "loss": 0.0136, "step": 29770 }, { "grad_norm": 0.1481000930070877, "learning_rate": 8.413201772061867e-05, "loss": 0.0116, "step": 29780 }, { "grad_norm": 0.2316170036792755, "learning_rate": 8.411993304934488e-05, "loss": 0.0123, "step": 29790 }, { "grad_norm": 0.20307937264442444, "learning_rate": 8.410784464676654e-05, "loss": 0.0147, "step": 29800 }, { "grad_norm": 0.1772179752588272, "learning_rate": 8.409575251420556e-05, "loss": 0.0102, "step": 29810 }, { "grad_norm": 0.17002877593040466, "learning_rate": 8.408365665298435e-05, "loss": 0.0109, "step": 29820 }, { "grad_norm": 0.1918412148952484, "learning_rate": 8.40715570644257e-05, "loss": 0.0128, "step": 29830 }, { "grad_norm": 0.17440727353096008, "learning_rate": 8.40594537498528e-05, "loss": 0.0104, "step": 29840 }, { "grad_norm": 0.21916908025741577, "learning_rate": 8.404734671058924e-05, "loss": 0.0133, "step": 29850 }, { "grad_norm": 0.16520462930202484, "learning_rate": 8.403523594795902e-05, "loss": 0.0129, "step": 29860 }, { "grad_norm": 0.14790117740631104, "learning_rate": 8.402312146328659e-05, "loss": 0.0126, "step": 29870 }, { "grad_norm": 0.15735222399234772, "learning_rate": 8.401100325789675e-05, "loss": 0.0134, "step": 29880 }, { "grad_norm": 0.1701342910528183, "learning_rate": 8.399888133311472e-05, "loss": 0.0144, "step": 29890 }, { "grad_norm": 0.19763332605361938, "learning_rate": 8.398675569026613e-05, "loss": 0.0114, "step": 29900 }, { "grad_norm": 0.17457716166973114, "learning_rate": 8.397462633067705e-05, "loss": 0.0124, "step": 29910 }, { "grad_norm": 0.18451185524463654, "learning_rate": 8.396249325567392e-05, "loss": 0.012, "step": 29920 }, { "grad_norm": 0.17761950194835663, "learning_rate": 8.395035646658357e-05, "loss": 0.0149, "step": 29930 }, { "grad_norm": 0.1463920772075653, "learning_rate": 8.39382159647333e-05, "loss": 0.0136, "step": 29940 }, { "grad_norm": 0.20429755747318268, "learning_rate": 8.392607175145075e-05, "loss": 0.0124, "step": 29950 }, { "grad_norm": 0.1292044073343277, "learning_rate": 8.3913923828064e-05, "loss": 0.01, "step": 29960 }, { "grad_norm": 0.21315643191337585, "learning_rate": 8.390177219590152e-05, "loss": 0.0102, "step": 29970 }, { "grad_norm": 0.17801544070243835, "learning_rate": 8.388961685629222e-05, "loss": 0.0105, "step": 29980 }, { "grad_norm": 0.1731099635362625, "learning_rate": 8.387745781056536e-05, "loss": 0.0104, "step": 29990 }, { "grad_norm": 0.17658856511116028, "learning_rate": 8.386529506005065e-05, "loss": 0.0114, "step": 30000 }, { "grad_norm": 0.16284911334514618, "learning_rate": 8.38531286060782e-05, "loss": 0.011, "step": 30010 }, { "grad_norm": 0.12833981215953827, "learning_rate": 8.384095844997849e-05, "loss": 0.0157, "step": 30020 }, { "grad_norm": 0.19952109456062317, "learning_rate": 8.382878459308245e-05, "loss": 0.0123, "step": 30030 }, { "grad_norm": 0.1401919275522232, "learning_rate": 8.381660703672138e-05, "loss": 0.0102, "step": 30040 }, { "grad_norm": 0.20719318091869354, "learning_rate": 8.380442578222702e-05, "loss": 0.0139, "step": 30050 }, { "grad_norm": 0.2078249603509903, "learning_rate": 8.379224083093146e-05, "loss": 0.0141, "step": 30060 }, { "grad_norm": 0.17776423692703247, "learning_rate": 8.378005218416727e-05, "loss": 0.0113, "step": 30070 }, { "grad_norm": 0.1238759383559227, "learning_rate": 8.376785984326735e-05, "loss": 0.0108, "step": 30080 }, { "grad_norm": 0.20825818181037903, "learning_rate": 8.375566380956506e-05, "loss": 0.0133, "step": 30090 }, { "grad_norm": 0.13826337456703186, "learning_rate": 8.374346408439411e-05, "loss": 0.0127, "step": 30100 }, { "grad_norm": 0.1415441930294037, "learning_rate": 8.373126066908868e-05, "loss": 0.0122, "step": 30110 }, { "grad_norm": 0.16312243044376373, "learning_rate": 8.371905356498326e-05, "loss": 0.0119, "step": 30120 }, { "grad_norm": 0.20534829795360565, "learning_rate": 8.370684277341288e-05, "loss": 0.011, "step": 30130 }, { "grad_norm": 0.1493532508611679, "learning_rate": 8.369462829571282e-05, "loss": 0.0155, "step": 30140 }, { "grad_norm": 0.19579029083251953, "learning_rate": 8.36824101332189e-05, "loss": 0.0121, "step": 30150 }, { "grad_norm": 0.14722706377506256, "learning_rate": 8.367018828726721e-05, "loss": 0.012, "step": 30160 }, { "grad_norm": 0.16191276907920837, "learning_rate": 8.365796275919438e-05, "loss": 0.0133, "step": 30170 }, { "grad_norm": 0.19122502207756042, "learning_rate": 8.364573355033734e-05, "loss": 0.0122, "step": 30180 }, { "grad_norm": 0.21535080671310425, "learning_rate": 8.363350066203346e-05, "loss": 0.0113, "step": 30190 }, { "grad_norm": 0.1791376918554306, "learning_rate": 8.362126409562053e-05, "loss": 0.0129, "step": 30200 }, { "grad_norm": 0.12393040955066681, "learning_rate": 8.360902385243667e-05, "loss": 0.0123, "step": 30210 }, { "grad_norm": 0.2274404913187027, "learning_rate": 8.359677993382052e-05, "loss": 0.0133, "step": 30220 }, { "grad_norm": 0.14308124780654907, "learning_rate": 8.358453234111103e-05, "loss": 0.01, "step": 30230 }, { "grad_norm": 0.18786615133285522, "learning_rate": 8.357228107564756e-05, "loss": 0.0116, "step": 30240 }, { "grad_norm": 0.1910385638475418, "learning_rate": 8.356002613876993e-05, "loss": 0.0146, "step": 30250 }, { "grad_norm": 0.1820971518754959, "learning_rate": 8.35477675318183e-05, "loss": 0.0112, "step": 30260 }, { "grad_norm": 0.15807482600212097, "learning_rate": 8.353550525613323e-05, "loss": 0.0124, "step": 30270 }, { "grad_norm": 0.2627162039279938, "learning_rate": 8.352323931305572e-05, "loss": 0.0142, "step": 30280 }, { "grad_norm": 0.14509856700897217, "learning_rate": 8.351096970392717e-05, "loss": 0.0118, "step": 30290 }, { "grad_norm": 0.17036105692386627, "learning_rate": 8.349869643008937e-05, "loss": 0.0115, "step": 30300 }, { "grad_norm": 0.14783498644828796, "learning_rate": 8.348641949288449e-05, "loss": 0.0095, "step": 30310 }, { "grad_norm": 0.1515408456325531, "learning_rate": 8.34741388936551e-05, "loss": 0.0121, "step": 30320 }, { "grad_norm": 0.15499845147132874, "learning_rate": 8.346185463374423e-05, "loss": 0.0115, "step": 30330 }, { "grad_norm": 0.14851966500282288, "learning_rate": 8.344956671449524e-05, "loss": 0.0133, "step": 30340 }, { "grad_norm": 0.19983455538749695, "learning_rate": 8.343727513725192e-05, "loss": 0.012, "step": 30350 }, { "grad_norm": 0.19327788054943085, "learning_rate": 8.342497990335847e-05, "loss": 0.0135, "step": 30360 }, { "grad_norm": 0.19469581544399261, "learning_rate": 8.341268101415946e-05, "loss": 0.0133, "step": 30370 }, { "grad_norm": 0.17912380397319794, "learning_rate": 8.34003784709999e-05, "loss": 0.0118, "step": 30380 }, { "grad_norm": 0.25623783469200134, "learning_rate": 8.338807227522516e-05, "loss": 0.0124, "step": 30390 }, { "grad_norm": 0.13751763105392456, "learning_rate": 8.337576242818103e-05, "loss": 0.0095, "step": 30400 }, { "grad_norm": 0.20502609014511108, "learning_rate": 8.336344893121372e-05, "loss": 0.0112, "step": 30410 }, { "grad_norm": 0.14384731650352478, "learning_rate": 8.335113178566977e-05, "loss": 0.0122, "step": 30420 }, { "grad_norm": 0.18697811663150787, "learning_rate": 8.33388109928962e-05, "loss": 0.0081, "step": 30430 }, { "grad_norm": 0.1380903124809265, "learning_rate": 8.33264865542404e-05, "loss": 0.0096, "step": 30440 }, { "grad_norm": 0.20769648253917694, "learning_rate": 8.331415847105013e-05, "loss": 0.0103, "step": 30450 }, { "grad_norm": 0.14248624444007874, "learning_rate": 8.330182674467357e-05, "loss": 0.0096, "step": 30460 }, { "grad_norm": 0.18709422647953033, "learning_rate": 8.32894913764593e-05, "loss": 0.0106, "step": 30470 }, { "grad_norm": 0.13909444212913513, "learning_rate": 8.327715236775633e-05, "loss": 0.0097, "step": 30480 }, { "grad_norm": 0.2019360512495041, "learning_rate": 8.326480971991398e-05, "loss": 0.0113, "step": 30490 }, { "grad_norm": 0.1747889518737793, "learning_rate": 8.325246343428206e-05, "loss": 0.0117, "step": 30500 }, { "grad_norm": 0.19714151322841644, "learning_rate": 8.324011351221072e-05, "loss": 0.012, "step": 30510 }, { "grad_norm": 0.1512061357498169, "learning_rate": 8.322775995505057e-05, "loss": 0.0114, "step": 30520 }, { "grad_norm": 0.1648336499929428, "learning_rate": 8.321540276415254e-05, "loss": 0.0138, "step": 30530 }, { "grad_norm": 0.15370625257492065, "learning_rate": 8.320304194086798e-05, "loss": 0.0109, "step": 30540 }, { "grad_norm": 0.2059975266456604, "learning_rate": 8.31906774865487e-05, "loss": 0.0106, "step": 30550 }, { "grad_norm": 0.14444327354431152, "learning_rate": 8.317830940254682e-05, "loss": 0.0128, "step": 30560 }, { "grad_norm": 0.14828014373779297, "learning_rate": 8.316593769021491e-05, "loss": 0.0139, "step": 30570 }, { "grad_norm": 0.18851538002490997, "learning_rate": 8.315356235090592e-05, "loss": 0.0137, "step": 30580 }, { "grad_norm": 0.14907442033290863, "learning_rate": 8.314118338597319e-05, "loss": 0.01, "step": 30590 }, { "grad_norm": 0.1709534078836441, "learning_rate": 8.312880079677048e-05, "loss": 0.0111, "step": 30600 }, { "grad_norm": 0.14873069524765015, "learning_rate": 8.311641458465191e-05, "loss": 0.0123, "step": 30610 }, { "grad_norm": 0.17279043793678284, "learning_rate": 8.310402475097205e-05, "loss": 0.009, "step": 30620 }, { "grad_norm": 0.17472191154956818, "learning_rate": 8.309163129708581e-05, "loss": 0.0127, "step": 30630 }, { "grad_norm": 0.22383065521717072, "learning_rate": 8.307923422434852e-05, "loss": 0.0125, "step": 30640 }, { "grad_norm": 0.1681080460548401, "learning_rate": 8.30668335341159e-05, "loss": 0.0111, "step": 30650 }, { "grad_norm": 0.17451970279216766, "learning_rate": 8.305442922774408e-05, "loss": 0.0143, "step": 30660 }, { "grad_norm": 0.13952012360095978, "learning_rate": 8.304202130658959e-05, "loss": 0.0106, "step": 30670 }, { "grad_norm": 0.16824544966220856, "learning_rate": 8.302960977200931e-05, "loss": 0.0109, "step": 30680 }, { "grad_norm": 0.17056305706501007, "learning_rate": 8.301719462536058e-05, "loss": 0.0109, "step": 30690 }, { "grad_norm": 0.1945553570985794, "learning_rate": 8.300477586800108e-05, "loss": 0.0111, "step": 30700 }, { "grad_norm": 0.14413034915924072, "learning_rate": 8.299235350128892e-05, "loss": 0.0112, "step": 30710 }, { "grad_norm": 0.09310799837112427, "learning_rate": 8.297992752658258e-05, "loss": 0.0112, "step": 30720 }, { "grad_norm": 0.1671232134103775, "learning_rate": 8.296749794524094e-05, "loss": 0.0131, "step": 30730 }, { "grad_norm": 0.14963537454605103, "learning_rate": 8.295506475862332e-05, "loss": 0.0121, "step": 30740 }, { "grad_norm": 0.15311303734779358, "learning_rate": 8.294262796808933e-05, "loss": 0.0094, "step": 30750 }, { "grad_norm": 0.1562008112668991, "learning_rate": 8.293018757499909e-05, "loss": 0.012, "step": 30760 }, { "grad_norm": 0.1883082538843155, "learning_rate": 8.291774358071305e-05, "loss": 0.0114, "step": 30770 }, { "grad_norm": 0.15789330005645752, "learning_rate": 8.290529598659205e-05, "loss": 0.0139, "step": 30780 }, { "grad_norm": 0.23003488779067993, "learning_rate": 8.289284479399738e-05, "loss": 0.01, "step": 30790 }, { "grad_norm": 0.16367650032043457, "learning_rate": 8.288039000429064e-05, "loss": 0.0133, "step": 30800 }, { "grad_norm": 0.17887915670871735, "learning_rate": 8.286793161883388e-05, "loss": 0.0126, "step": 30810 }, { "grad_norm": 0.23720109462738037, "learning_rate": 8.285546963898954e-05, "loss": 0.0127, "step": 30820 }, { "grad_norm": 0.11482082307338715, "learning_rate": 8.284300406612044e-05, "loss": 0.0118, "step": 30830 }, { "grad_norm": 0.15039807558059692, "learning_rate": 8.283053490158978e-05, "loss": 0.01, "step": 30840 }, { "grad_norm": 0.16659851372241974, "learning_rate": 8.28180621467612e-05, "loss": 0.0107, "step": 30850 }, { "grad_norm": 0.1728159338235855, "learning_rate": 8.280558580299867e-05, "loss": 0.0127, "step": 30860 }, { "grad_norm": 0.209575816988945, "learning_rate": 8.279310587166661e-05, "loss": 0.0155, "step": 30870 }, { "grad_norm": 0.14325089752674103, "learning_rate": 8.278062235412978e-05, "loss": 0.0125, "step": 30880 }, { "grad_norm": 0.16817325353622437, "learning_rate": 8.276813525175339e-05, "loss": 0.0109, "step": 30890 }, { "grad_norm": 0.17048397660255432, "learning_rate": 8.2755644565903e-05, "loss": 0.0119, "step": 30900 }, { "grad_norm": 0.17710116505622864, "learning_rate": 8.274315029794454e-05, "loss": 0.0122, "step": 30910 }, { "grad_norm": 0.16648030281066895, "learning_rate": 8.273065244924443e-05, "loss": 0.0123, "step": 30920 }, { "grad_norm": 0.13608789443969727, "learning_rate": 8.271815102116936e-05, "loss": 0.0107, "step": 30930 }, { "grad_norm": 0.16670839488506317, "learning_rate": 8.270564601508648e-05, "loss": 0.0096, "step": 30940 }, { "grad_norm": 0.14403244853019714, "learning_rate": 8.269313743236333e-05, "loss": 0.0101, "step": 30950 }, { "grad_norm": 0.18100017309188843, "learning_rate": 8.268062527436783e-05, "loss": 0.0116, "step": 30960 }, { "grad_norm": 0.1471172571182251, "learning_rate": 8.266810954246828e-05, "loss": 0.0104, "step": 30970 }, { "grad_norm": 0.13852675259113312, "learning_rate": 8.265559023803338e-05, "loss": 0.0099, "step": 30980 }, { "grad_norm": 0.17191816866397858, "learning_rate": 8.264306736243223e-05, "loss": 0.0136, "step": 30990 }, { "grad_norm": 0.16483375430107117, "learning_rate": 8.263054091703432e-05, "loss": 0.0106, "step": 31000 }, { "grad_norm": 0.1870371401309967, "learning_rate": 8.26180109032095e-05, "loss": 0.0122, "step": 31010 }, { "grad_norm": 0.16621100902557373, "learning_rate": 8.260547732232807e-05, "loss": 0.012, "step": 31020 }, { "grad_norm": 0.1433776170015335, "learning_rate": 8.259294017576063e-05, "loss": 0.0094, "step": 31030 }, { "grad_norm": 0.17946605384349823, "learning_rate": 8.258039946487828e-05, "loss": 0.0135, "step": 31040 }, { "grad_norm": 0.1393696367740631, "learning_rate": 8.256785519105241e-05, "loss": 0.0107, "step": 31050 }, { "grad_norm": 0.15625332295894623, "learning_rate": 8.255530735565488e-05, "loss": 0.0099, "step": 31060 }, { "grad_norm": 0.1121085062623024, "learning_rate": 8.254275596005787e-05, "loss": 0.0109, "step": 31070 }, { "grad_norm": 0.1631680130958557, "learning_rate": 8.2530201005634e-05, "loss": 0.0123, "step": 31080 }, { "grad_norm": 0.15714587271213531, "learning_rate": 8.251764249375626e-05, "loss": 0.0102, "step": 31090 }, { "grad_norm": 0.16845384240150452, "learning_rate": 8.250508042579803e-05, "loss": 0.0111, "step": 31100 }, { "grad_norm": 0.20554091036319733, "learning_rate": 8.249251480313307e-05, "loss": 0.0112, "step": 31110 }, { "grad_norm": 0.24574922025203705, "learning_rate": 8.247994562713555e-05, "loss": 0.0126, "step": 31120 }, { "grad_norm": 0.20159241557121277, "learning_rate": 8.246737289917998e-05, "loss": 0.0112, "step": 31130 }, { "grad_norm": 0.1439375877380371, "learning_rate": 8.245479662064135e-05, "loss": 0.011, "step": 31140 }, { "grad_norm": 0.2162359207868576, "learning_rate": 8.244221679289496e-05, "loss": 0.0116, "step": 31150 }, { "grad_norm": 0.21443413197994232, "learning_rate": 8.242963341731652e-05, "loss": 0.0131, "step": 31160 }, { "grad_norm": 0.2526591122150421, "learning_rate": 8.24170464952821e-05, "loss": 0.0124, "step": 31170 }, { "grad_norm": 0.21152061223983765, "learning_rate": 8.240445602816824e-05, "loss": 0.0122, "step": 31180 }, { "grad_norm": 0.17783144116401672, "learning_rate": 8.239186201735179e-05, "loss": 0.0109, "step": 31190 }, { "grad_norm": 0.23331956565380096, "learning_rate": 8.237926446420998e-05, "loss": 0.0108, "step": 31200 }, { "grad_norm": 0.16477681696414948, "learning_rate": 8.236666337012052e-05, "loss": 0.0116, "step": 31210 }, { "grad_norm": 0.1692584753036499, "learning_rate": 8.23540587364614e-05, "loss": 0.0133, "step": 31220 }, { "grad_norm": 0.12888875603675842, "learning_rate": 8.234145056461107e-05, "loss": 0.0114, "step": 31230 }, { "grad_norm": 0.14272065460681915, "learning_rate": 8.232883885594832e-05, "loss": 0.0139, "step": 31240 }, { "grad_norm": 0.14896206557750702, "learning_rate": 8.231622361185236e-05, "loss": 0.0107, "step": 31250 }, { "grad_norm": 0.13190950453281403, "learning_rate": 8.230360483370278e-05, "loss": 0.0091, "step": 31260 }, { "grad_norm": 0.16405300796031952, "learning_rate": 8.229098252287953e-05, "loss": 0.0114, "step": 31270 }, { "grad_norm": 0.18598522245883942, "learning_rate": 8.2278356680763e-05, "loss": 0.0145, "step": 31280 }, { "grad_norm": 0.15844255685806274, "learning_rate": 8.22657273087339e-05, "loss": 0.0141, "step": 31290 }, { "grad_norm": 0.32998865842819214, "learning_rate": 8.225309440817336e-05, "loss": 0.0158, "step": 31300 }, { "grad_norm": 0.14356106519699097, "learning_rate": 8.224045798046293e-05, "loss": 0.0117, "step": 31310 }, { "grad_norm": 0.14149750769138336, "learning_rate": 8.22278180269845e-05, "loss": 0.0123, "step": 31320 }, { "grad_norm": 0.22008207440376282, "learning_rate": 8.221517454912031e-05, "loss": 0.0133, "step": 31330 }, { "grad_norm": 0.13435246050357819, "learning_rate": 8.220252754825308e-05, "loss": 0.0125, "step": 31340 }, { "grad_norm": 0.13217249512672424, "learning_rate": 8.218987702576586e-05, "loss": 0.0138, "step": 31350 }, { "grad_norm": 0.17786382138729095, "learning_rate": 8.217722298304207e-05, "loss": 0.0148, "step": 31360 }, { "grad_norm": 0.1275610774755478, "learning_rate": 8.216456542146557e-05, "loss": 0.0137, "step": 31370 }, { "grad_norm": 0.1721896529197693, "learning_rate": 8.215190434242055e-05, "loss": 0.0132, "step": 31380 }, { "grad_norm": 0.20661509037017822, "learning_rate": 8.213923974729161e-05, "loss": 0.0151, "step": 31390 }, { "grad_norm": 0.16551810503005981, "learning_rate": 8.212657163746373e-05, "loss": 0.0124, "step": 31400 }, { "grad_norm": 0.19486947357654572, "learning_rate": 8.211390001432227e-05, "loss": 0.013, "step": 31410 }, { "grad_norm": 0.17405501008033752, "learning_rate": 8.210122487925297e-05, "loss": 0.0137, "step": 31420 }, { "grad_norm": 0.138968825340271, "learning_rate": 8.208854623364202e-05, "loss": 0.0101, "step": 31430 }, { "grad_norm": 0.18855740129947662, "learning_rate": 8.207586407887589e-05, "loss": 0.0113, "step": 31440 }, { "grad_norm": 0.2092055231332779, "learning_rate": 8.206317841634148e-05, "loss": 0.0115, "step": 31450 }, { "grad_norm": 0.2572079598903656, "learning_rate": 8.205048924742609e-05, "loss": 0.0126, "step": 31460 }, { "grad_norm": 0.12432383000850677, "learning_rate": 8.203779657351738e-05, "loss": 0.0108, "step": 31470 }, { "grad_norm": 0.1583569049835205, "learning_rate": 8.20251003960034e-05, "loss": 0.0111, "step": 31480 }, { "grad_norm": 0.17932140827178955, "learning_rate": 8.201240071627258e-05, "loss": 0.0123, "step": 31490 }, { "grad_norm": 0.16318771243095398, "learning_rate": 8.199969753571377e-05, "loss": 0.0124, "step": 31500 }, { "grad_norm": 0.12898188829421997, "learning_rate": 8.198699085571615e-05, "loss": 0.0104, "step": 31510 }, { "grad_norm": 0.1316218078136444, "learning_rate": 8.197428067766928e-05, "loss": 0.0097, "step": 31520 }, { "grad_norm": 0.16190597414970398, "learning_rate": 8.196156700296316e-05, "loss": 0.0109, "step": 31530 }, { "grad_norm": 0.20025824010372162, "learning_rate": 8.194884983298814e-05, "loss": 0.0138, "step": 31540 }, { "grad_norm": 0.14898276329040527, "learning_rate": 8.193612916913491e-05, "loss": 0.0147, "step": 31550 }, { "grad_norm": 0.2485109567642212, "learning_rate": 8.192340501279463e-05, "loss": 0.0104, "step": 31560 }, { "grad_norm": 0.18954558670520782, "learning_rate": 8.191067736535876e-05, "loss": 0.01, "step": 31570 }, { "grad_norm": 0.1770269274711609, "learning_rate": 8.18979462282192e-05, "loss": 0.0094, "step": 31580 }, { "grad_norm": 0.14027591049671173, "learning_rate": 8.188521160276819e-05, "loss": 0.0104, "step": 31590 }, { "grad_norm": 0.17740949988365173, "learning_rate": 8.187247349039837e-05, "loss": 0.0107, "step": 31600 }, { "grad_norm": 0.1769450306892395, "learning_rate": 8.185973189250278e-05, "loss": 0.0104, "step": 31610 }, { "grad_norm": 0.16254636645317078, "learning_rate": 8.184698681047482e-05, "loss": 0.0142, "step": 31620 }, { "grad_norm": 0.2238789051771164, "learning_rate": 8.183423824570827e-05, "loss": 0.0083, "step": 31630 }, { "grad_norm": 0.13675041496753693, "learning_rate": 8.182148619959725e-05, "loss": 0.0122, "step": 31640 }, { "grad_norm": 0.17855286598205566, "learning_rate": 8.180873067353636e-05, "loss": 0.0115, "step": 31650 }, { "grad_norm": 0.1895774006843567, "learning_rate": 8.179597166892052e-05, "loss": 0.013, "step": 31660 }, { "grad_norm": 0.1528840959072113, "learning_rate": 8.178320918714501e-05, "loss": 0.0119, "step": 31670 }, { "grad_norm": 0.12990613281726837, "learning_rate": 8.177044322960554e-05, "loss": 0.0111, "step": 31680 }, { "grad_norm": 0.19235476851463318, "learning_rate": 8.175767379769816e-05, "loss": 0.0094, "step": 31690 }, { "grad_norm": 0.17422935366630554, "learning_rate": 8.174490089281932e-05, "loss": 0.0114, "step": 31700 }, { "grad_norm": 0.21025606989860535, "learning_rate": 8.173212451636584e-05, "loss": 0.0105, "step": 31710 }, { "grad_norm": 0.16235187649726868, "learning_rate": 8.171934466973493e-05, "loss": 0.0103, "step": 31720 }, { "grad_norm": 0.1775379627943039, "learning_rate": 8.170656135432418e-05, "loss": 0.0112, "step": 31730 }, { "grad_norm": 0.22785325348377228, "learning_rate": 8.169377457153155e-05, "loss": 0.0118, "step": 31740 }, { "grad_norm": 0.15772190690040588, "learning_rate": 8.168098432275539e-05, "loss": 0.0125, "step": 31750 }, { "grad_norm": 0.18942968547344208, "learning_rate": 8.166819060939442e-05, "loss": 0.0107, "step": 31760 }, { "grad_norm": 0.18919049203395844, "learning_rate": 8.165539343284772e-05, "loss": 0.0115, "step": 31770 }, { "grad_norm": 0.19994975626468658, "learning_rate": 8.16425927945148e-05, "loss": 0.0101, "step": 31780 }, { "grad_norm": 0.18709130585193634, "learning_rate": 8.162978869579551e-05, "loss": 0.0121, "step": 31790 }, { "grad_norm": 0.20482459664344788, "learning_rate": 8.161698113809007e-05, "loss": 0.012, "step": 31800 }, { "grad_norm": 0.21851298213005066, "learning_rate": 8.160417012279911e-05, "loss": 0.0121, "step": 31810 }, { "grad_norm": 0.14174728095531464, "learning_rate": 8.159135565132363e-05, "loss": 0.0099, "step": 31820 }, { "grad_norm": 0.10365364700555801, "learning_rate": 8.157853772506498e-05, "loss": 0.0108, "step": 31830 }, { "grad_norm": 0.20173436403274536, "learning_rate": 8.156571634542494e-05, "loss": 0.0095, "step": 31840 }, { "grad_norm": 0.18232622742652893, "learning_rate": 8.15528915138056e-05, "loss": 0.0108, "step": 31850 }, { "grad_norm": 0.16960027813911438, "learning_rate": 8.154006323160949e-05, "loss": 0.0118, "step": 31860 }, { "grad_norm": 0.18328174948692322, "learning_rate": 8.152723150023949e-05, "loss": 0.0118, "step": 31870 }, { "grad_norm": 0.11797115951776505, "learning_rate": 8.151439632109886e-05, "loss": 0.0104, "step": 31880 }, { "grad_norm": 0.14105218648910522, "learning_rate": 8.150155769559122e-05, "loss": 0.009, "step": 31890 }, { "grad_norm": 0.12084587663412094, "learning_rate": 8.148871562512058e-05, "loss": 0.01, "step": 31900 }, { "grad_norm": 0.16787837445735931, "learning_rate": 8.147587011109136e-05, "loss": 0.0118, "step": 31910 }, { "grad_norm": 0.20486916601657867, "learning_rate": 8.14630211549083e-05, "loss": 0.0099, "step": 31920 }, { "grad_norm": 0.11771903932094574, "learning_rate": 8.145016875797655e-05, "loss": 0.0107, "step": 31930 }, { "grad_norm": 0.1823834627866745, "learning_rate": 8.143731292170164e-05, "loss": 0.0136, "step": 31940 }, { "grad_norm": 0.2239546924829483, "learning_rate": 8.142445364748944e-05, "loss": 0.0116, "step": 31950 }, { "grad_norm": 0.1798936426639557, "learning_rate": 8.141159093674624e-05, "loss": 0.0149, "step": 31960 }, { "grad_norm": 0.17717371881008148, "learning_rate": 8.139872479087869e-05, "loss": 0.0107, "step": 31970 }, { "grad_norm": 0.15675051510334015, "learning_rate": 8.13858552112938e-05, "loss": 0.013, "step": 31980 }, { "grad_norm": 0.1126430481672287, "learning_rate": 8.137298219939895e-05, "loss": 0.0097, "step": 31990 }, { "grad_norm": 0.14657215774059296, "learning_rate": 8.136010575660196e-05, "loss": 0.011, "step": 32000 }, { "grad_norm": 0.1527908742427826, "learning_rate": 8.134722588431095e-05, "loss": 0.0125, "step": 32010 }, { "grad_norm": 0.21076145768165588, "learning_rate": 8.133434258393444e-05, "loss": 0.0114, "step": 32020 }, { "grad_norm": 0.19497676193714142, "learning_rate": 8.132145585688134e-05, "loss": 0.0103, "step": 32030 }, { "grad_norm": 0.12740468978881836, "learning_rate": 8.130856570456093e-05, "loss": 0.0091, "step": 32040 }, { "grad_norm": 0.16234508156776428, "learning_rate": 8.129567212838283e-05, "loss": 0.0095, "step": 32050 }, { "grad_norm": 0.15371714532375336, "learning_rate": 8.128277512975708e-05, "loss": 0.011, "step": 32060 }, { "grad_norm": 0.21531322598457336, "learning_rate": 8.126987471009408e-05, "loss": 0.0086, "step": 32070 }, { "grad_norm": 0.22014960646629333, "learning_rate": 8.125697087080459e-05, "loss": 0.0123, "step": 32080 }, { "grad_norm": 0.21418894827365875, "learning_rate": 8.124406361329976e-05, "loss": 0.012, "step": 32090 }, { "grad_norm": 0.17652128636837006, "learning_rate": 8.123115293899111e-05, "loss": 0.0106, "step": 32100 }, { "grad_norm": 0.14925552904605865, "learning_rate": 8.121823884929055e-05, "loss": 0.0102, "step": 32110 }, { "grad_norm": 0.1199040338397026, "learning_rate": 8.12053213456103e-05, "loss": 0.0115, "step": 32120 }, { "grad_norm": 0.1498444676399231, "learning_rate": 8.119240042936303e-05, "loss": 0.0115, "step": 32130 }, { "grad_norm": 0.11704257130622864, "learning_rate": 8.117947610196175e-05, "loss": 0.0106, "step": 32140 }, { "grad_norm": 0.1487962007522583, "learning_rate": 8.116654836481982e-05, "loss": 0.0115, "step": 32150 }, { "grad_norm": 0.16843925416469574, "learning_rate": 8.115361721935106e-05, "loss": 0.0096, "step": 32160 }, { "grad_norm": 0.24213379621505737, "learning_rate": 8.114068266696953e-05, "loss": 0.0106, "step": 32170 }, { "grad_norm": 0.14697152376174927, "learning_rate": 8.112774470908978e-05, "loss": 0.0111, "step": 32180 }, { "grad_norm": 0.16144540905952454, "learning_rate": 8.111480334712665e-05, "loss": 0.0088, "step": 32190 }, { "grad_norm": 0.12261418998241425, "learning_rate": 8.110185858249542e-05, "loss": 0.0092, "step": 32200 }, { "grad_norm": 0.1906021237373352, "learning_rate": 8.108891041661168e-05, "loss": 0.0139, "step": 32210 }, { "grad_norm": 0.1744023859500885, "learning_rate": 8.107595885089146e-05, "loss": 0.0111, "step": 32220 }, { "grad_norm": 0.22068937122821808, "learning_rate": 8.106300388675108e-05, "loss": 0.0118, "step": 32230 }, { "grad_norm": 0.26434266567230225, "learning_rate": 8.105004552560731e-05, "loss": 0.0121, "step": 32240 }, { "grad_norm": 0.16995686292648315, "learning_rate": 8.103708376887724e-05, "loss": 0.0091, "step": 32250 }, { "grad_norm": 0.12548288702964783, "learning_rate": 8.102411861797836e-05, "loss": 0.0132, "step": 32260 }, { "grad_norm": 0.16611579060554504, "learning_rate": 8.101115007432851e-05, "loss": 0.0107, "step": 32270 }, { "grad_norm": 0.1868477314710617, "learning_rate": 8.09981781393459e-05, "loss": 0.0123, "step": 32280 }, { "grad_norm": 0.19905981421470642, "learning_rate": 8.098520281444915e-05, "loss": 0.0124, "step": 32290 }, { "grad_norm": 0.19284072518348694, "learning_rate": 8.09722241010572e-05, "loss": 0.0115, "step": 32300 }, { "grad_norm": 0.1737644225358963, "learning_rate": 8.095924200058939e-05, "loss": 0.0109, "step": 32310 }, { "grad_norm": 0.14159667491912842, "learning_rate": 8.094625651446541e-05, "loss": 0.0083, "step": 32320 }, { "grad_norm": 0.18970105051994324, "learning_rate": 8.093326764410536e-05, "loss": 0.0113, "step": 32330 }, { "grad_norm": 0.1602163165807724, "learning_rate": 8.092027539092966e-05, "loss": 0.0098, "step": 32340 }, { "grad_norm": 0.15799453854560852, "learning_rate": 8.090727975635913e-05, "loss": 0.01, "step": 32350 }, { "grad_norm": 0.12989342212677002, "learning_rate": 8.089428074181497e-05, "loss": 0.011, "step": 32360 }, { "grad_norm": 0.1517866849899292, "learning_rate": 8.088127834871871e-05, "loss": 0.0112, "step": 32370 }, { "grad_norm": 0.1499134749174118, "learning_rate": 8.086827257849226e-05, "loss": 0.0102, "step": 32380 }, { "grad_norm": 0.17651207745075226, "learning_rate": 8.085526343255795e-05, "loss": 0.0137, "step": 32390 }, { "grad_norm": 0.2259342074394226, "learning_rate": 8.084225091233842e-05, "loss": 0.0127, "step": 32400 }, { "grad_norm": 0.17867201566696167, "learning_rate": 8.082923501925668e-05, "loss": 0.0145, "step": 32410 }, { "grad_norm": 0.16540005803108215, "learning_rate": 8.081621575473617e-05, "loss": 0.0118, "step": 32420 }, { "grad_norm": 0.1668899804353714, "learning_rate": 8.080319312020064e-05, "loss": 0.0106, "step": 32430 }, { "grad_norm": 0.1762693077325821, "learning_rate": 8.079016711707421e-05, "loss": 0.0101, "step": 32440 }, { "grad_norm": 0.149547278881073, "learning_rate": 8.077713774678139e-05, "loss": 0.0111, "step": 32450 }, { "grad_norm": 0.162734255194664, "learning_rate": 8.076410501074707e-05, "loss": 0.012, "step": 32460 }, { "grad_norm": 0.14501650631427765, "learning_rate": 8.075106891039647e-05, "loss": 0.0121, "step": 32470 }, { "grad_norm": 0.16206417977809906, "learning_rate": 8.073802944715523e-05, "loss": 0.0119, "step": 32480 }, { "grad_norm": 0.15934807062149048, "learning_rate": 8.072498662244929e-05, "loss": 0.011, "step": 32490 }, { "grad_norm": 0.1640813946723938, "learning_rate": 8.0711940437705e-05, "loss": 0.0098, "step": 32500 }, { "grad_norm": 0.24064436554908752, "learning_rate": 8.06988908943491e-05, "loss": 0.0103, "step": 32510 }, { "grad_norm": 0.1620214432477951, "learning_rate": 8.068583799380863e-05, "loss": 0.0119, "step": 32520 }, { "grad_norm": 0.18303807079792023, "learning_rate": 8.067278173751104e-05, "loss": 0.0106, "step": 32530 }, { "grad_norm": 0.20225486159324646, "learning_rate": 8.065972212688417e-05, "loss": 0.0167, "step": 32540 }, { "grad_norm": 0.13214685022830963, "learning_rate": 8.064665916335618e-05, "loss": 0.0103, "step": 32550 }, { "grad_norm": 0.14057889580726624, "learning_rate": 8.063359284835564e-05, "loss": 0.01, "step": 32560 }, { "grad_norm": 0.12480136007070541, "learning_rate": 8.062052318331142e-05, "loss": 0.0118, "step": 32570 }, { "grad_norm": 0.3318080008029938, "learning_rate": 8.060745016965283e-05, "loss": 0.0174, "step": 32580 }, { "grad_norm": 0.22359749674797058, "learning_rate": 8.059437380880952e-05, "loss": 0.0166, "step": 32590 }, { "grad_norm": 0.17949041724205017, "learning_rate": 8.058129410221146e-05, "loss": 0.0178, "step": 32600 }, { "grad_norm": 0.1347218006849289, "learning_rate": 8.056821105128908e-05, "loss": 0.0106, "step": 32610 }, { "grad_norm": 0.15194523334503174, "learning_rate": 8.05551246574731e-05, "loss": 0.0095, "step": 32620 }, { "grad_norm": 0.14026330411434174, "learning_rate": 8.05420349221946e-05, "loss": 0.0129, "step": 32630 }, { "grad_norm": 0.1771097183227539, "learning_rate": 8.05289418468851e-05, "loss": 0.0105, "step": 32640 }, { "grad_norm": 0.1821604073047638, "learning_rate": 8.051584543297642e-05, "loss": 0.0098, "step": 32650 }, { "grad_norm": 0.16759654879570007, "learning_rate": 8.050274568190074e-05, "loss": 0.0128, "step": 32660 }, { "grad_norm": 0.14941728115081787, "learning_rate": 8.048964259509067e-05, "loss": 0.0116, "step": 32670 }, { "grad_norm": 0.19090673327445984, "learning_rate": 8.047653617397914e-05, "loss": 0.0143, "step": 32680 }, { "grad_norm": 0.1890869140625, "learning_rate": 8.046342641999941e-05, "loss": 0.0118, "step": 32690 }, { "grad_norm": 0.18066224455833435, "learning_rate": 8.045031333458517e-05, "loss": 0.0122, "step": 32700 }, { "grad_norm": 0.14231155812740326, "learning_rate": 8.043719691917047e-05, "loss": 0.0093, "step": 32710 }, { "grad_norm": 0.1660681515932083, "learning_rate": 8.042407717518966e-05, "loss": 0.0103, "step": 32720 }, { "grad_norm": 0.19362299144268036, "learning_rate": 8.041095410407751e-05, "loss": 0.0128, "step": 32730 }, { "grad_norm": 0.1875160187482834, "learning_rate": 8.039782770726913e-05, "loss": 0.0088, "step": 32740 }, { "grad_norm": 0.15861967206001282, "learning_rate": 8.038469798620004e-05, "loss": 0.0121, "step": 32750 }, { "grad_norm": 0.18014423549175262, "learning_rate": 8.037156494230604e-05, "loss": 0.0115, "step": 32760 }, { "grad_norm": 0.15270566940307617, "learning_rate": 8.035842857702338e-05, "loss": 0.0126, "step": 32770 }, { "grad_norm": 0.1752544641494751, "learning_rate": 8.03452888917886e-05, "loss": 0.0113, "step": 32780 }, { "grad_norm": 0.18249443173408508, "learning_rate": 8.033214588803866e-05, "loss": 0.0098, "step": 32790 }, { "grad_norm": 0.10832995176315308, "learning_rate": 8.031899956721083e-05, "loss": 0.0109, "step": 32800 }, { "grad_norm": 0.16725751757621765, "learning_rate": 8.030584993074282e-05, "loss": 0.0102, "step": 32810 }, { "grad_norm": 0.15581142902374268, "learning_rate": 8.02926969800726e-05, "loss": 0.0094, "step": 32820 }, { "grad_norm": 0.2008056789636612, "learning_rate": 8.027954071663859e-05, "loss": 0.0136, "step": 32830 }, { "grad_norm": 0.14697346091270447, "learning_rate": 8.026638114187954e-05, "loss": 0.0098, "step": 32840 }, { "grad_norm": 0.15172132849693298, "learning_rate": 8.025321825723456e-05, "loss": 0.0102, "step": 32850 }, { "grad_norm": 0.15902329981327057, "learning_rate": 8.02400520641431e-05, "loss": 0.01, "step": 32860 }, { "grad_norm": 0.14340633153915405, "learning_rate": 8.022688256404501e-05, "loss": 0.01, "step": 32870 }, { "grad_norm": 0.19720135629177094, "learning_rate": 8.02137097583805e-05, "loss": 0.0129, "step": 32880 }, { "grad_norm": 0.11878041177988052, "learning_rate": 8.02005336485901e-05, "loss": 0.0089, "step": 32890 }, { "grad_norm": 0.18437352776527405, "learning_rate": 8.018735423611476e-05, "loss": 0.011, "step": 32900 }, { "grad_norm": 0.1548442244529724, "learning_rate": 8.017417152239574e-05, "loss": 0.0084, "step": 32910 }, { "grad_norm": 0.14198429882526398, "learning_rate": 8.01609855088747e-05, "loss": 0.0108, "step": 32920 }, { "grad_norm": 0.12783066928386688, "learning_rate": 8.014779619699362e-05, "loss": 0.0094, "step": 32930 }, { "grad_norm": 0.17837724089622498, "learning_rate": 8.013460358819489e-05, "loss": 0.0144, "step": 32940 }, { "grad_norm": 0.12802931666374207, "learning_rate": 8.01214076839212e-05, "loss": 0.0117, "step": 32950 }, { "grad_norm": 0.18283823132514954, "learning_rate": 8.010820848561565e-05, "loss": 0.0098, "step": 32960 }, { "grad_norm": 0.14566439390182495, "learning_rate": 8.009500599472171e-05, "loss": 0.0099, "step": 32970 }, { "grad_norm": 0.1607653796672821, "learning_rate": 8.008180021268314e-05, "loss": 0.0099, "step": 32980 }, { "grad_norm": 0.17421714961528778, "learning_rate": 8.006859114094414e-05, "loss": 0.012, "step": 32990 }, { "grad_norm": 0.16497699916362762, "learning_rate": 8.005537878094921e-05, "loss": 0.011, "step": 33000 }, { "grad_norm": 0.14803782105445862, "learning_rate": 8.004216313414323e-05, "loss": 0.0101, "step": 33010 }, { "grad_norm": 0.17875410616397858, "learning_rate": 8.002894420197149e-05, "loss": 0.0108, "step": 33020 }, { "grad_norm": 0.11397300660610199, "learning_rate": 8.001572198587954e-05, "loss": 0.0133, "step": 33030 }, { "grad_norm": 0.16570353507995605, "learning_rate": 8.000249648731338e-05, "loss": 0.0113, "step": 33040 }, { "grad_norm": 0.18634332716464996, "learning_rate": 7.998926770771928e-05, "loss": 0.011, "step": 33050 }, { "grad_norm": 0.17855563759803772, "learning_rate": 7.997603564854397e-05, "loss": 0.0114, "step": 33060 }, { "grad_norm": 0.16370996832847595, "learning_rate": 7.996280031123448e-05, "loss": 0.0107, "step": 33070 }, { "grad_norm": 0.17322051525115967, "learning_rate": 7.994956169723818e-05, "loss": 0.0124, "step": 33080 }, { "grad_norm": 0.15569505095481873, "learning_rate": 7.993631980800285e-05, "loss": 0.0127, "step": 33090 }, { "grad_norm": 0.1384337693452835, "learning_rate": 7.992307464497659e-05, "loss": 0.011, "step": 33100 }, { "grad_norm": 0.1877748817205429, "learning_rate": 7.990982620960787e-05, "loss": 0.0106, "step": 33110 }, { "grad_norm": 0.19714923202991486, "learning_rate": 7.989657450334554e-05, "loss": 0.0102, "step": 33120 }, { "grad_norm": 0.14449886977672577, "learning_rate": 7.988331952763877e-05, "loss": 0.0099, "step": 33130 }, { "grad_norm": 0.24928414821624756, "learning_rate": 7.987006128393709e-05, "loss": 0.0158, "step": 33140 }, { "grad_norm": 0.20106519758701324, "learning_rate": 7.985679977369043e-05, "loss": 0.0125, "step": 33150 }, { "grad_norm": 0.13125865161418915, "learning_rate": 7.984353499834902e-05, "loss": 0.0108, "step": 33160 }, { "grad_norm": 0.16160543262958527, "learning_rate": 7.983026695936351e-05, "loss": 0.0124, "step": 33170 }, { "grad_norm": 0.15251274406909943, "learning_rate": 7.981699565818486e-05, "loss": 0.0123, "step": 33180 }, { "grad_norm": 0.1873120218515396, "learning_rate": 7.980372109626437e-05, "loss": 0.0099, "step": 33190 }, { "grad_norm": 0.13734224438667297, "learning_rate": 7.979044327505375e-05, "loss": 0.0095, "step": 33200 }, { "grad_norm": 0.21144317090511322, "learning_rate": 7.977716219600506e-05, "loss": 0.0137, "step": 33210 }, { "grad_norm": 0.1265891194343567, "learning_rate": 7.97638778605707e-05, "loss": 0.0107, "step": 33220 }, { "grad_norm": 0.1520155966281891, "learning_rate": 7.975059027020338e-05, "loss": 0.0111, "step": 33230 }, { "grad_norm": 0.16373179852962494, "learning_rate": 7.973729942635623e-05, "loss": 0.01, "step": 33240 }, { "grad_norm": 0.12275949120521545, "learning_rate": 7.972400533048273e-05, "loss": 0.0104, "step": 33250 }, { "grad_norm": 0.18363375961780548, "learning_rate": 7.97107079840367e-05, "loss": 0.0139, "step": 33260 }, { "grad_norm": 0.1612112671136856, "learning_rate": 7.969740738847231e-05, "loss": 0.0106, "step": 33270 }, { "grad_norm": 0.1382022649049759, "learning_rate": 7.968410354524411e-05, "loss": 0.0104, "step": 33280 }, { "grad_norm": 0.18052485585212708, "learning_rate": 7.967079645580697e-05, "loss": 0.0116, "step": 33290 }, { "grad_norm": 0.18034590780735016, "learning_rate": 7.965748612161612e-05, "loss": 0.0098, "step": 33300 }, { "grad_norm": 0.15725816786289215, "learning_rate": 7.96441725441272e-05, "loss": 0.0111, "step": 33310 }, { "grad_norm": 0.13398034870624542, "learning_rate": 7.963085572479614e-05, "loss": 0.0092, "step": 33320 }, { "grad_norm": 0.2192578762769699, "learning_rate": 7.961753566507924e-05, "loss": 0.0146, "step": 33330 }, { "grad_norm": 0.18032988905906677, "learning_rate": 7.960421236643316e-05, "loss": 0.0099, "step": 33340 }, { "grad_norm": 0.17271152138710022, "learning_rate": 7.959088583031496e-05, "loss": 0.0087, "step": 33350 }, { "grad_norm": 0.13878761231899261, "learning_rate": 7.957755605818194e-05, "loss": 0.0099, "step": 33360 }, { "grad_norm": 0.1385239213705063, "learning_rate": 7.956422305149185e-05, "loss": 0.0106, "step": 33370 }, { "grad_norm": 0.1551564335823059, "learning_rate": 7.95508868117028e-05, "loss": 0.0105, "step": 33380 }, { "grad_norm": 0.1489994376897812, "learning_rate": 7.953754734027318e-05, "loss": 0.0114, "step": 33390 }, { "grad_norm": 0.19577382504940033, "learning_rate": 7.952420463866182e-05, "loss": 0.0105, "step": 33400 }, { "grad_norm": 0.14249354600906372, "learning_rate": 7.951085870832782e-05, "loss": 0.0117, "step": 33410 }, { "grad_norm": 0.22422753274440765, "learning_rate": 7.949750955073067e-05, "loss": 0.0101, "step": 33420 }, { "grad_norm": 0.18001043796539307, "learning_rate": 7.948415716733022e-05, "loss": 0.0101, "step": 33430 }, { "grad_norm": 0.22839663922786713, "learning_rate": 7.947080155958669e-05, "loss": 0.015, "step": 33440 }, { "grad_norm": 0.1403205245733261, "learning_rate": 7.94574427289606e-05, "loss": 0.0098, "step": 33450 }, { "grad_norm": 0.1416887640953064, "learning_rate": 7.944408067691284e-05, "loss": 0.0083, "step": 33460 }, { "grad_norm": 0.15070433914661407, "learning_rate": 7.943071540490473e-05, "loss": 0.0114, "step": 33470 }, { "grad_norm": 0.18982195854187012, "learning_rate": 7.94173469143978e-05, "loss": 0.0127, "step": 33480 }, { "grad_norm": 0.13496610522270203, "learning_rate": 7.940397520685406e-05, "loss": 0.0092, "step": 33490 }, { "grad_norm": 0.16417323052883148, "learning_rate": 7.939060028373577e-05, "loss": 0.0093, "step": 33500 }, { "grad_norm": 0.18032526969909668, "learning_rate": 7.937722214650565e-05, "loss": 0.0132, "step": 33510 }, { "grad_norm": 0.16613802313804626, "learning_rate": 7.936384079662666e-05, "loss": 0.0094, "step": 33520 }, { "grad_norm": 0.17007479071617126, "learning_rate": 7.93504562355622e-05, "loss": 0.0122, "step": 33530 }, { "grad_norm": 0.1611342877149582, "learning_rate": 7.933706846477599e-05, "loss": 0.0089, "step": 33540 }, { "grad_norm": 0.15859150886535645, "learning_rate": 7.932367748573206e-05, "loss": 0.0111, "step": 33550 }, { "grad_norm": 0.13494829833507538, "learning_rate": 7.931028329989485e-05, "loss": 0.0103, "step": 33560 }, { "grad_norm": 0.19312603771686554, "learning_rate": 7.929688590872913e-05, "loss": 0.0098, "step": 33570 }, { "grad_norm": 0.15858541429042816, "learning_rate": 7.928348531370003e-05, "loss": 0.0114, "step": 33580 }, { "grad_norm": 0.14621849358081818, "learning_rate": 7.927008151627297e-05, "loss": 0.0115, "step": 33590 }, { "grad_norm": 0.12649521231651306, "learning_rate": 7.925667451791383e-05, "loss": 0.0094, "step": 33600 }, { "grad_norm": 0.16888469457626343, "learning_rate": 7.924326432008874e-05, "loss": 0.0114, "step": 33610 }, { "grad_norm": 0.14017023146152496, "learning_rate": 7.922985092426422e-05, "loss": 0.0096, "step": 33620 }, { "grad_norm": 0.15803633630275726, "learning_rate": 7.921643433190717e-05, "loss": 0.0108, "step": 33630 }, { "grad_norm": 0.1483568251132965, "learning_rate": 7.920301454448478e-05, "loss": 0.0103, "step": 33640 }, { "grad_norm": 0.17487157881259918, "learning_rate": 7.918959156346461e-05, "loss": 0.0112, "step": 33650 }, { "grad_norm": 0.17919208109378815, "learning_rate": 7.91761653903146e-05, "loss": 0.0106, "step": 33660 }, { "grad_norm": 0.1214420273900032, "learning_rate": 7.916273602650302e-05, "loss": 0.0106, "step": 33670 }, { "grad_norm": 0.17466068267822266, "learning_rate": 7.914930347349847e-05, "loss": 0.0094, "step": 33680 }, { "grad_norm": 0.18680262565612793, "learning_rate": 7.913586773276992e-05, "loss": 0.0123, "step": 33690 }, { "grad_norm": 0.15809059143066406, "learning_rate": 7.912242880578667e-05, "loss": 0.0095, "step": 33700 }, { "grad_norm": 0.12275751680135727, "learning_rate": 7.910898669401839e-05, "loss": 0.0115, "step": 33710 }, { "grad_norm": 0.1208077073097229, "learning_rate": 7.909554139893511e-05, "loss": 0.0115, "step": 33720 }, { "grad_norm": 0.14300328493118286, "learning_rate": 7.908209292200715e-05, "loss": 0.0097, "step": 33730 }, { "grad_norm": 0.12208807468414307, "learning_rate": 7.906864126470523e-05, "loss": 0.0085, "step": 33740 }, { "grad_norm": 0.11382871121168137, "learning_rate": 7.905518642850041e-05, "loss": 0.0112, "step": 33750 }, { "grad_norm": 0.12615597248077393, "learning_rate": 7.904172841486409e-05, "loss": 0.0101, "step": 33760 }, { "grad_norm": 0.1217949166893959, "learning_rate": 7.902826722526801e-05, "loss": 0.0109, "step": 33770 }, { "grad_norm": 0.1992114633321762, "learning_rate": 7.901480286118427e-05, "loss": 0.0116, "step": 33780 }, { "grad_norm": 0.1582263708114624, "learning_rate": 7.900133532408531e-05, "loss": 0.0113, "step": 33790 }, { "grad_norm": 0.11619328707456589, "learning_rate": 7.898786461544395e-05, "loss": 0.0119, "step": 33800 }, { "grad_norm": 0.1368795931339264, "learning_rate": 7.897439073673325e-05, "loss": 0.011, "step": 33810 }, { "grad_norm": 0.12163395434617996, "learning_rate": 7.896091368942677e-05, "loss": 0.0078, "step": 33820 }, { "grad_norm": 0.15837496519088745, "learning_rate": 7.894743347499832e-05, "loss": 0.0121, "step": 33830 }, { "grad_norm": 0.15969035029411316, "learning_rate": 7.893395009492203e-05, "loss": 0.01, "step": 33840 }, { "grad_norm": 0.17169144749641418, "learning_rate": 7.892046355067248e-05, "loss": 0.0113, "step": 33850 }, { "grad_norm": 0.13166910409927368, "learning_rate": 7.890697384372451e-05, "loss": 0.0082, "step": 33860 }, { "grad_norm": 0.12704645097255707, "learning_rate": 7.889348097555336e-05, "loss": 0.0092, "step": 33870 }, { "grad_norm": 0.13032826781272888, "learning_rate": 7.887998494763455e-05, "loss": 0.0116, "step": 33880 }, { "grad_norm": 0.1523837298154831, "learning_rate": 7.886648576144404e-05, "loss": 0.0117, "step": 33890 }, { "grad_norm": 0.20636886358261108, "learning_rate": 7.885298341845802e-05, "loss": 0.0132, "step": 33900 }, { "grad_norm": 0.14611072838306427, "learning_rate": 7.883947792015311e-05, "loss": 0.0091, "step": 33910 }, { "grad_norm": 0.13653726875782013, "learning_rate": 7.882596926800628e-05, "loss": 0.0107, "step": 33920 }, { "grad_norm": 0.18662357330322266, "learning_rate": 7.881245746349477e-05, "loss": 0.0143, "step": 33930 }, { "grad_norm": 0.1765502393245697, "learning_rate": 7.879894250809623e-05, "loss": 0.01, "step": 33940 }, { "grad_norm": 0.13629448413848877, "learning_rate": 7.878542440328865e-05, "loss": 0.01, "step": 33950 }, { "grad_norm": 0.15482832491397858, "learning_rate": 7.877190315055031e-05, "loss": 0.012, "step": 33960 }, { "grad_norm": 0.1652509719133377, "learning_rate": 7.875837875135991e-05, "loss": 0.0117, "step": 33970 }, { "grad_norm": 0.17753855884075165, "learning_rate": 7.874485120719646e-05, "loss": 0.0133, "step": 33980 }, { "grad_norm": 0.18711867928504944, "learning_rate": 7.873132051953928e-05, "loss": 0.0102, "step": 33990 }, { "grad_norm": 0.12485774606466293, "learning_rate": 7.87177866898681e-05, "loss": 0.0127, "step": 34000 }, { "grad_norm": 0.16020850837230682, "learning_rate": 7.870424971966294e-05, "loss": 0.0097, "step": 34010 }, { "grad_norm": 0.13149192929267883, "learning_rate": 7.869070961040419e-05, "loss": 0.0111, "step": 34020 }, { "grad_norm": 0.13317549228668213, "learning_rate": 7.867716636357257e-05, "loss": 0.013, "step": 34030 }, { "grad_norm": 0.11714976280927658, "learning_rate": 7.866361998064915e-05, "loss": 0.0104, "step": 34040 }, { "grad_norm": 0.1718059927225113, "learning_rate": 7.865007046311534e-05, "loss": 0.0097, "step": 34050 }, { "grad_norm": 0.11418232321739197, "learning_rate": 7.86365178124529e-05, "loss": 0.0139, "step": 34060 }, { "grad_norm": 0.19368992745876312, "learning_rate": 7.862296203014394e-05, "loss": 0.0105, "step": 34070 }, { "grad_norm": 0.12495610117912292, "learning_rate": 7.860940311767088e-05, "loss": 0.0103, "step": 34080 }, { "grad_norm": 0.18625394999980927, "learning_rate": 7.85958410765165e-05, "loss": 0.0129, "step": 34090 }, { "grad_norm": 0.17971491813659668, "learning_rate": 7.858227590816394e-05, "loss": 0.0099, "step": 34100 }, { "grad_norm": 0.13938505947589874, "learning_rate": 7.856870761409664e-05, "loss": 0.0127, "step": 34110 }, { "grad_norm": 0.13832399249076843, "learning_rate": 7.855513619579846e-05, "loss": 0.0119, "step": 34120 }, { "grad_norm": 0.19118960201740265, "learning_rate": 7.85415616547535e-05, "loss": 0.0124, "step": 34130 }, { "grad_norm": 0.14174339175224304, "learning_rate": 7.852798399244627e-05, "loss": 0.0089, "step": 34140 }, { "grad_norm": 0.12486440688371658, "learning_rate": 7.851440321036161e-05, "loss": 0.0112, "step": 34150 }, { "grad_norm": 0.1848648637533188, "learning_rate": 7.850081930998468e-05, "loss": 0.0098, "step": 34160 }, { "grad_norm": 0.16845737397670746, "learning_rate": 7.8487232292801e-05, "loss": 0.0102, "step": 34170 }, { "grad_norm": 0.14195166528224945, "learning_rate": 7.847364216029642e-05, "loss": 0.0115, "step": 34180 }, { "grad_norm": 0.1896403431892395, "learning_rate": 7.846004891395716e-05, "loss": 0.0096, "step": 34190 }, { "grad_norm": 0.18123386800289154, "learning_rate": 7.844645255526972e-05, "loss": 0.0091, "step": 34200 }, { "grad_norm": 0.18261034786701202, "learning_rate": 7.843285308572101e-05, "loss": 0.0097, "step": 34210 }, { "grad_norm": 0.22764535248279572, "learning_rate": 7.841925050679823e-05, "loss": 0.0105, "step": 34220 }, { "grad_norm": 0.18870632350444794, "learning_rate": 7.840564481998895e-05, "loss": 0.011, "step": 34230 }, { "grad_norm": 0.13560935854911804, "learning_rate": 7.839203602678105e-05, "loss": 0.0115, "step": 34240 }, { "grad_norm": 0.09893140196800232, "learning_rate": 7.837842412866279e-05, "loss": 0.0108, "step": 34250 }, { "grad_norm": 0.1613190621137619, "learning_rate": 7.836480912712272e-05, "loss": 0.0119, "step": 34260 }, { "grad_norm": 0.1552504003047943, "learning_rate": 7.835119102364976e-05, "loss": 0.0104, "step": 34270 }, { "grad_norm": 0.13531285524368286, "learning_rate": 7.83375698197332e-05, "loss": 0.012, "step": 34280 }, { "grad_norm": 0.17811231315135956, "learning_rate": 7.83239455168626e-05, "loss": 0.0107, "step": 34290 }, { "grad_norm": 0.1384430080652237, "learning_rate": 7.83103181165279e-05, "loss": 0.0091, "step": 34300 }, { "grad_norm": 0.1433325707912445, "learning_rate": 7.829668762021937e-05, "loss": 0.0084, "step": 34310 }, { "grad_norm": 0.13860100507736206, "learning_rate": 7.828305402942764e-05, "loss": 0.0118, "step": 34320 }, { "grad_norm": 0.1414739042520523, "learning_rate": 7.826941734564363e-05, "loss": 0.0104, "step": 34330 }, { "grad_norm": 0.1362542062997818, "learning_rate": 7.825577757035865e-05, "loss": 0.0091, "step": 34340 }, { "grad_norm": 0.13609617948532104, "learning_rate": 7.824213470506431e-05, "loss": 0.0106, "step": 34350 }, { "grad_norm": 0.1616816371679306, "learning_rate": 7.822848875125257e-05, "loss": 0.0111, "step": 34360 }, { "grad_norm": 0.1895401030778885, "learning_rate": 7.821483971041576e-05, "loss": 0.0098, "step": 34370 }, { "grad_norm": 0.12974193692207336, "learning_rate": 7.820118758404649e-05, "loss": 0.0105, "step": 34380 }, { "grad_norm": 0.22765901684761047, "learning_rate": 7.818753237363776e-05, "loss": 0.0146, "step": 34390 }, { "grad_norm": 0.19754068553447723, "learning_rate": 7.817387408068286e-05, "loss": 0.0098, "step": 34400 }, { "grad_norm": 0.12827982008457184, "learning_rate": 7.816021270667544e-05, "loss": 0.0108, "step": 34410 }, { "grad_norm": 0.14191338419914246, "learning_rate": 7.81465482531095e-05, "loss": 0.0093, "step": 34420 }, { "grad_norm": 0.154013529419899, "learning_rate": 7.813288072147938e-05, "loss": 0.0116, "step": 34430 }, { "grad_norm": 0.18138234317302704, "learning_rate": 7.811921011327972e-05, "loss": 0.0095, "step": 34440 }, { "grad_norm": 0.22247569262981415, "learning_rate": 7.810553643000549e-05, "loss": 0.0128, "step": 34450 }, { "grad_norm": 0.13818404078483582, "learning_rate": 7.809185967315206e-05, "loss": 0.0114, "step": 34460 }, { "grad_norm": 0.14382578432559967, "learning_rate": 7.80781798442151e-05, "loss": 0.0107, "step": 34470 }, { "grad_norm": 0.1691904217004776, "learning_rate": 7.806449694469061e-05, "loss": 0.0114, "step": 34480 }, { "grad_norm": 0.16924747824668884, "learning_rate": 7.805081097607492e-05, "loss": 0.0095, "step": 34490 }, { "grad_norm": 0.1354944258928299, "learning_rate": 7.803712193986474e-05, "loss": 0.0109, "step": 34500 }, { "grad_norm": 0.13285648822784424, "learning_rate": 7.802342983755702e-05, "loss": 0.0112, "step": 34510 }, { "grad_norm": 0.17798155546188354, "learning_rate": 7.800973467064918e-05, "loss": 0.0105, "step": 34520 }, { "grad_norm": 0.17552968859672546, "learning_rate": 7.799603644063884e-05, "loss": 0.011, "step": 34530 }, { "grad_norm": 0.20486462116241455, "learning_rate": 7.798233514902405e-05, "loss": 0.0104, "step": 34540 }, { "grad_norm": 0.15237107872962952, "learning_rate": 7.796863079730318e-05, "loss": 0.0105, "step": 34550 }, { "grad_norm": 0.1752522587776184, "learning_rate": 7.795492338697488e-05, "loss": 0.0099, "step": 34560 }, { "grad_norm": 0.16531109809875488, "learning_rate": 7.794121291953819e-05, "loss": 0.0106, "step": 34570 }, { "grad_norm": 0.12701530754566193, "learning_rate": 7.792749939649246e-05, "loss": 0.0117, "step": 34580 }, { "grad_norm": 0.14944025874137878, "learning_rate": 7.79137828193374e-05, "loss": 0.0098, "step": 34590 }, { "grad_norm": 0.13518275320529938, "learning_rate": 7.790006318957301e-05, "loss": 0.009, "step": 34600 }, { "grad_norm": 0.11842802911996841, "learning_rate": 7.788634050869965e-05, "loss": 0.0096, "step": 34610 }, { "grad_norm": 0.2309107929468155, "learning_rate": 7.787261477821803e-05, "loss": 0.0115, "step": 34620 }, { "grad_norm": 0.134303480386734, "learning_rate": 7.785888599962916e-05, "loss": 0.0122, "step": 34630 }, { "grad_norm": 0.18803633749485016, "learning_rate": 7.784515417443439e-05, "loss": 0.0123, "step": 34640 }, { "grad_norm": 0.2034211903810501, "learning_rate": 7.783141930413545e-05, "loss": 0.0142, "step": 34650 }, { "grad_norm": 0.13831284642219543, "learning_rate": 7.78176813902343e-05, "loss": 0.0093, "step": 34660 }, { "grad_norm": 0.1393681764602661, "learning_rate": 7.780394043423336e-05, "loss": 0.0096, "step": 34670 }, { "grad_norm": 0.13947436213493347, "learning_rate": 7.77901964376353e-05, "loss": 0.0114, "step": 34680 }, { "grad_norm": 0.17292682826519012, "learning_rate": 7.777644940194316e-05, "loss": 0.0103, "step": 34690 }, { "grad_norm": 0.17638397216796875, "learning_rate": 7.776269932866023e-05, "loss": 0.0123, "step": 34700 }, { "grad_norm": 0.1607959121465683, "learning_rate": 7.774894621929026e-05, "loss": 0.0112, "step": 34710 }, { "grad_norm": 0.18672236800193787, "learning_rate": 7.773519007533725e-05, "loss": 0.0111, "step": 34720 }, { "grad_norm": 0.17281043529510498, "learning_rate": 7.772143089830556e-05, "loss": 0.0099, "step": 34730 }, { "grad_norm": 0.13363169133663177, "learning_rate": 7.770766868969985e-05, "loss": 0.011, "step": 34740 }, { "grad_norm": 0.16986322402954102, "learning_rate": 7.769390345102518e-05, "loss": 0.0121, "step": 34750 }, { "grad_norm": 0.15378914773464203, "learning_rate": 7.768013518378683e-05, "loss": 0.0095, "step": 34760 }, { "grad_norm": 0.2345481514930725, "learning_rate": 7.766636388949053e-05, "loss": 0.0109, "step": 34770 }, { "grad_norm": 0.12784750759601593, "learning_rate": 7.765258956964229e-05, "loss": 0.0095, "step": 34780 }, { "grad_norm": 0.1895691454410553, "learning_rate": 7.76388122257484e-05, "loss": 0.0109, "step": 34790 }, { "grad_norm": 0.13022233545780182, "learning_rate": 7.762503185931558e-05, "loss": 0.0102, "step": 34800 }, { "grad_norm": 0.18987587094306946, "learning_rate": 7.76112484718508e-05, "loss": 0.0136, "step": 34810 }, { "grad_norm": 0.16454580426216125, "learning_rate": 7.75974620648614e-05, "loss": 0.0105, "step": 34820 }, { "grad_norm": 0.14640448987483978, "learning_rate": 7.758367263985503e-05, "loss": 0.0122, "step": 34830 }, { "grad_norm": 0.2046220749616623, "learning_rate": 7.75698801983397e-05, "loss": 0.0095, "step": 34840 }, { "grad_norm": 0.19003598392009735, "learning_rate": 7.755608474182372e-05, "loss": 0.0097, "step": 34850 }, { "grad_norm": 0.16346368193626404, "learning_rate": 7.754228627181574e-05, "loss": 0.0089, "step": 34860 }, { "grad_norm": 0.13288144767284393, "learning_rate": 7.752848478982476e-05, "loss": 0.0122, "step": 34870 }, { "grad_norm": 0.12283780425786972, "learning_rate": 7.751468029736006e-05, "loss": 0.0103, "step": 34880 }, { "grad_norm": 0.15071620047092438, "learning_rate": 7.750087279593129e-05, "loss": 0.0107, "step": 34890 }, { "grad_norm": 0.16264402866363525, "learning_rate": 7.748706228704843e-05, "loss": 0.009, "step": 34900 }, { "grad_norm": 0.14137867093086243, "learning_rate": 7.747324877222176e-05, "loss": 0.0091, "step": 34910 }, { "grad_norm": 0.15521180629730225, "learning_rate": 7.745943225296188e-05, "loss": 0.0093, "step": 34920 }, { "grad_norm": 0.21185408532619476, "learning_rate": 7.744561273077981e-05, "loss": 0.0098, "step": 34930 }, { "grad_norm": 0.12844207882881165, "learning_rate": 7.743179020718678e-05, "loss": 0.0111, "step": 34940 }, { "grad_norm": 0.1250641644001007, "learning_rate": 7.741796468369443e-05, "loss": 0.0093, "step": 34950 }, { "grad_norm": 0.15588518977165222, "learning_rate": 7.740413616181466e-05, "loss": 0.0103, "step": 34960 }, { "grad_norm": 0.11919965595006943, "learning_rate": 7.739030464305978e-05, "loss": 0.0097, "step": 34970 }, { "grad_norm": 0.12350868433713913, "learning_rate": 7.737647012894235e-05, "loss": 0.0136, "step": 34980 }, { "grad_norm": 0.1551479548215866, "learning_rate": 7.736263262097532e-05, "loss": 0.0105, "step": 34990 }, { "grad_norm": 0.17444635927677155, "learning_rate": 7.734879212067192e-05, "loss": 0.01, "step": 35000 }, { "grad_norm": 0.17935119569301605, "learning_rate": 7.733494862954573e-05, "loss": 0.0152, "step": 35010 }, { "grad_norm": 0.1927412897348404, "learning_rate": 7.732110214911066e-05, "loss": 0.0108, "step": 35020 }, { "grad_norm": 0.15679341554641724, "learning_rate": 7.730725268088094e-05, "loss": 0.0123, "step": 35030 }, { "grad_norm": 0.1725023090839386, "learning_rate": 7.729340022637112e-05, "loss": 0.01, "step": 35040 }, { "grad_norm": 0.11826899647712708, "learning_rate": 7.727954478709607e-05, "loss": 0.0091, "step": 35050 }, { "grad_norm": 0.1578342467546463, "learning_rate": 7.726568636457103e-05, "loss": 0.0086, "step": 35060 }, { "grad_norm": 0.2088523805141449, "learning_rate": 7.725182496031153e-05, "loss": 0.0107, "step": 35070 }, { "grad_norm": 0.12473296374082565, "learning_rate": 7.723796057583341e-05, "loss": 0.0146, "step": 35080 }, { "grad_norm": 0.19043797254562378, "learning_rate": 7.722409321265287e-05, "loss": 0.0114, "step": 35090 }, { "grad_norm": 0.1650310456752777, "learning_rate": 7.721022287228645e-05, "loss": 0.011, "step": 35100 }, { "grad_norm": 0.16487641632556915, "learning_rate": 7.719634955625094e-05, "loss": 0.0137, "step": 35110 }, { "grad_norm": 0.20447880029678345, "learning_rate": 7.718247326606355e-05, "loss": 0.0124, "step": 35120 }, { "grad_norm": 0.13261856138706207, "learning_rate": 7.716859400324175e-05, "loss": 0.0108, "step": 35130 }, { "grad_norm": 0.19203545153141022, "learning_rate": 7.715471176930335e-05, "loss": 0.0093, "step": 35140 }, { "grad_norm": 0.14651848375797272, "learning_rate": 7.714082656576651e-05, "loss": 0.0094, "step": 35150 }, { "grad_norm": 0.182736337184906, "learning_rate": 7.712693839414968e-05, "loss": 0.0101, "step": 35160 }, { "grad_norm": 0.1386726051568985, "learning_rate": 7.711304725597164e-05, "loss": 0.0102, "step": 35170 }, { "grad_norm": 0.1366732120513916, "learning_rate": 7.709915315275151e-05, "loss": 0.0138, "step": 35180 }, { "grad_norm": 0.2187080681324005, "learning_rate": 7.708525608600876e-05, "loss": 0.0129, "step": 35190 }, { "grad_norm": 0.18227866291999817, "learning_rate": 7.707135605726311e-05, "loss": 0.0137, "step": 35200 }, { "grad_norm": 0.16089054942131042, "learning_rate": 7.705745306803466e-05, "loss": 0.0102, "step": 35210 }, { "grad_norm": 0.1985083818435669, "learning_rate": 7.704354711984383e-05, "loss": 0.0104, "step": 35220 }, { "grad_norm": 0.17233657836914062, "learning_rate": 7.702963821421133e-05, "loss": 0.0107, "step": 35230 }, { "grad_norm": 0.14789915084838867, "learning_rate": 7.701572635265826e-05, "loss": 0.0109, "step": 35240 }, { "grad_norm": 0.16896866261959076, "learning_rate": 7.700181153670596e-05, "loss": 0.01, "step": 35250 }, { "grad_norm": 0.2008851021528244, "learning_rate": 7.698789376787616e-05, "loss": 0.011, "step": 35260 }, { "grad_norm": 0.17455126345157623, "learning_rate": 7.697397304769085e-05, "loss": 0.0093, "step": 35270 }, { "grad_norm": 0.14008089900016785, "learning_rate": 7.696004937767241e-05, "loss": 0.008, "step": 35280 }, { "grad_norm": 0.13794638216495514, "learning_rate": 7.694612275934352e-05, "loss": 0.0107, "step": 35290 }, { "grad_norm": 0.14313605427742004, "learning_rate": 7.693219319422714e-05, "loss": 0.0104, "step": 35300 }, { "grad_norm": 0.19383175671100616, "learning_rate": 7.69182606838466e-05, "loss": 0.0101, "step": 35310 }, { "grad_norm": 0.14815180003643036, "learning_rate": 7.690432522972558e-05, "loss": 0.0111, "step": 35320 }, { "grad_norm": 0.19683507084846497, "learning_rate": 7.689038683338796e-05, "loss": 0.0125, "step": 35330 }, { "grad_norm": 0.1639845371246338, "learning_rate": 7.687644549635808e-05, "loss": 0.0115, "step": 35340 }, { "grad_norm": 0.1454838067293167, "learning_rate": 7.686250122016053e-05, "loss": 0.0126, "step": 35350 }, { "grad_norm": 0.1304146647453308, "learning_rate": 7.684855400632022e-05, "loss": 0.0091, "step": 35360 }, { "grad_norm": 0.1250068098306656, "learning_rate": 7.683460385636243e-05, "loss": 0.0134, "step": 35370 }, { "grad_norm": 0.13228768110275269, "learning_rate": 7.68206507718127e-05, "loss": 0.0082, "step": 35380 }, { "grad_norm": 0.14810900390148163, "learning_rate": 7.680669475419693e-05, "loss": 0.0108, "step": 35390 }, { "grad_norm": 0.20245034992694855, "learning_rate": 7.679273580504132e-05, "loss": 0.011, "step": 35400 }, { "grad_norm": 0.11860544234514236, "learning_rate": 7.67787739258724e-05, "loss": 0.0105, "step": 35410 }, { "grad_norm": 0.13976261019706726, "learning_rate": 7.676480911821705e-05, "loss": 0.0133, "step": 35420 }, { "grad_norm": 0.11989108473062515, "learning_rate": 7.675084138360238e-05, "loss": 0.0118, "step": 35430 }, { "grad_norm": 0.18142473697662354, "learning_rate": 7.673687072355592e-05, "loss": 0.0139, "step": 35440 }, { "grad_norm": 0.16000503301620483, "learning_rate": 7.67228971396055e-05, "loss": 0.0105, "step": 35450 }, { "grad_norm": 0.17377276718616486, "learning_rate": 7.670892063327922e-05, "loss": 0.0117, "step": 35460 }, { "grad_norm": 0.18830284476280212, "learning_rate": 7.669494120610552e-05, "loss": 0.0119, "step": 35470 }, { "grad_norm": 0.17218568921089172, "learning_rate": 7.66809588596132e-05, "loss": 0.0106, "step": 35480 }, { "grad_norm": 0.14075207710266113, "learning_rate": 7.666697359533132e-05, "loss": 0.0069, "step": 35490 }, { "grad_norm": 0.13055776059627533, "learning_rate": 7.665298541478932e-05, "loss": 0.0113, "step": 35500 }, { "grad_norm": 0.12980274856090546, "learning_rate": 7.66389943195169e-05, "loss": 0.0138, "step": 35510 }, { "grad_norm": 0.16095983982086182, "learning_rate": 7.66250003110441e-05, "loss": 0.0112, "step": 35520 }, { "grad_norm": 0.14510154724121094, "learning_rate": 7.66110033909013e-05, "loss": 0.011, "step": 35530 }, { "grad_norm": 0.2529091238975525, "learning_rate": 7.659700356061918e-05, "loss": 0.0119, "step": 35540 }, { "grad_norm": 0.15680082142353058, "learning_rate": 7.658300082172875e-05, "loss": 0.0107, "step": 35550 }, { "grad_norm": 0.14470088481903076, "learning_rate": 7.65689951757613e-05, "loss": 0.0109, "step": 35560 }, { "grad_norm": 0.16723673045635223, "learning_rate": 7.655498662424849e-05, "loss": 0.0117, "step": 35570 }, { "grad_norm": 0.14200852811336517, "learning_rate": 7.654097516872227e-05, "loss": 0.0093, "step": 35580 }, { "grad_norm": 0.1872544139623642, "learning_rate": 7.65269608107149e-05, "loss": 0.011, "step": 35590 }, { "grad_norm": 0.17756296694278717, "learning_rate": 7.6512943551759e-05, "loss": 0.0103, "step": 35600 }, { "grad_norm": 0.14094868302345276, "learning_rate": 7.649892339338743e-05, "loss": 0.0103, "step": 35610 }, { "grad_norm": 0.17923696339130402, "learning_rate": 7.648490033713344e-05, "loss": 0.0118, "step": 35620 }, { "grad_norm": 0.20711944997310638, "learning_rate": 7.647087438453058e-05, "loss": 0.0119, "step": 35630 }, { "grad_norm": 0.18733656406402588, "learning_rate": 7.64568455371127e-05, "loss": 0.0142, "step": 35640 }, { "grad_norm": 0.1829269975423813, "learning_rate": 7.644281379641396e-05, "loss": 0.0102, "step": 35650 }, { "grad_norm": 0.10185018926858902, "learning_rate": 7.642877916396887e-05, "loss": 0.0103, "step": 35660 }, { "grad_norm": 0.18054191768169403, "learning_rate": 7.641474164131221e-05, "loss": 0.0098, "step": 35670 }, { "grad_norm": 0.19305294752120972, "learning_rate": 7.640070122997913e-05, "loss": 0.0108, "step": 35680 }, { "grad_norm": 0.13555413484573364, "learning_rate": 7.638665793150505e-05, "loss": 0.0121, "step": 35690 }, { "grad_norm": 0.14083439111709595, "learning_rate": 7.637261174742574e-05, "loss": 0.0111, "step": 35700 }, { "grad_norm": 0.17049379646778107, "learning_rate": 7.635856267927725e-05, "loss": 0.0092, "step": 35710 }, { "grad_norm": 0.14800025522708893, "learning_rate": 7.634451072859597e-05, "loss": 0.0136, "step": 35720 }, { "grad_norm": 0.15042664110660553, "learning_rate": 7.633045589691863e-05, "loss": 0.0118, "step": 35730 }, { "grad_norm": 0.15267862379550934, "learning_rate": 7.63163981857822e-05, "loss": 0.01, "step": 35740 }, { "grad_norm": 0.1528867930173874, "learning_rate": 7.630233759672403e-05, "loss": 0.0094, "step": 35750 }, { "grad_norm": 0.1429154872894287, "learning_rate": 7.628827413128177e-05, "loss": 0.0094, "step": 35760 }, { "grad_norm": 0.1613544523715973, "learning_rate": 7.627420779099338e-05, "loss": 0.0114, "step": 35770 }, { "grad_norm": 0.15918874740600586, "learning_rate": 7.626013857739711e-05, "loss": 0.008, "step": 35780 }, { "grad_norm": 0.18339039385318756, "learning_rate": 7.624606649203158e-05, "loss": 0.0088, "step": 35790 }, { "grad_norm": 0.1839008629322052, "learning_rate": 7.623199153643569e-05, "loss": 0.0119, "step": 35800 }, { "grad_norm": 0.17141148447990417, "learning_rate": 7.621791371214863e-05, "loss": 0.0122, "step": 35810 }, { "grad_norm": 0.18022038042545319, "learning_rate": 7.620383302070995e-05, "loss": 0.0091, "step": 35820 }, { "grad_norm": 0.14446473121643066, "learning_rate": 7.61897494636595e-05, "loss": 0.0097, "step": 35830 }, { "grad_norm": 0.17917171120643616, "learning_rate": 7.617566304253739e-05, "loss": 0.01, "step": 35840 }, { "grad_norm": 0.1259559839963913, "learning_rate": 7.616157375888416e-05, "loss": 0.0101, "step": 35850 }, { "grad_norm": 0.1380169540643692, "learning_rate": 7.614748161424053e-05, "loss": 0.0081, "step": 35860 }, { "grad_norm": 0.15841546654701233, "learning_rate": 7.613338661014763e-05, "loss": 0.0136, "step": 35870 }, { "grad_norm": 0.23341187834739685, "learning_rate": 7.611928874814686e-05, "loss": 0.0089, "step": 35880 }, { "grad_norm": 0.1788533478975296, "learning_rate": 7.610518802977996e-05, "loss": 0.0095, "step": 35890 }, { "grad_norm": 0.1473398506641388, "learning_rate": 7.609108445658893e-05, "loss": 0.0101, "step": 35900 }, { "grad_norm": 0.11985140293836594, "learning_rate": 7.607697803011612e-05, "loss": 0.0133, "step": 35910 }, { "grad_norm": 0.17331431806087494, "learning_rate": 7.606286875190421e-05, "loss": 0.0085, "step": 35920 }, { "grad_norm": 0.12774580717086792, "learning_rate": 7.604875662349617e-05, "loss": 0.009, "step": 35930 }, { "grad_norm": 0.13934946060180664, "learning_rate": 7.603464164643525e-05, "loss": 0.009, "step": 35940 }, { "grad_norm": 0.16205061972141266, "learning_rate": 7.602052382226507e-05, "loss": 0.0098, "step": 35950 }, { "grad_norm": 0.13597813248634338, "learning_rate": 7.600640315252954e-05, "loss": 0.0085, "step": 35960 }, { "grad_norm": 0.1455027163028717, "learning_rate": 7.599227963877284e-05, "loss": 0.0102, "step": 35970 }, { "grad_norm": 0.1619662642478943, "learning_rate": 7.597815328253954e-05, "loss": 0.0106, "step": 35980 }, { "grad_norm": 0.20819097757339478, "learning_rate": 7.596402408537443e-05, "loss": 0.0114, "step": 35990 }, { "grad_norm": 0.16784615814685822, "learning_rate": 7.59498920488227e-05, "loss": 0.0106, "step": 36000 }, { "grad_norm": 0.13069024682044983, "learning_rate": 7.593575717442979e-05, "loss": 0.0095, "step": 36010 }, { "grad_norm": 0.1409541815519333, "learning_rate": 7.592161946374147e-05, "loss": 0.0093, "step": 36020 }, { "grad_norm": 0.1493159383535385, "learning_rate": 7.590747891830381e-05, "loss": 0.013, "step": 36030 }, { "grad_norm": 0.10312427580356598, "learning_rate": 7.58933355396632e-05, "loss": 0.01, "step": 36040 }, { "grad_norm": 0.1364608108997345, "learning_rate": 7.587918932936636e-05, "loss": 0.0095, "step": 36050 }, { "grad_norm": 0.14377503097057343, "learning_rate": 7.586504028896028e-05, "loss": 0.0105, "step": 36060 }, { "grad_norm": 0.1814510077238083, "learning_rate": 7.585088841999228e-05, "loss": 0.0106, "step": 36070 }, { "grad_norm": 0.20724011957645416, "learning_rate": 7.583673372400999e-05, "loss": 0.0115, "step": 36080 }, { "grad_norm": 0.16847936809062958, "learning_rate": 7.582257620256134e-05, "loss": 0.0116, "step": 36090 }, { "grad_norm": 0.11707240343093872, "learning_rate": 7.580841585719458e-05, "loss": 0.0125, "step": 36100 }, { "grad_norm": 0.1721201241016388, "learning_rate": 7.579425268945825e-05, "loss": 0.0108, "step": 36110 }, { "grad_norm": 0.1995989829301834, "learning_rate": 7.578008670090127e-05, "loss": 0.0101, "step": 36120 }, { "grad_norm": 0.15694652497768402, "learning_rate": 7.576591789307272e-05, "loss": 0.0105, "step": 36130 }, { "grad_norm": 0.13284635543823242, "learning_rate": 7.575174626752216e-05, "loss": 0.0085, "step": 36140 }, { "grad_norm": 0.18072408437728882, "learning_rate": 7.573757182579934e-05, "loss": 0.01, "step": 36150 }, { "grad_norm": 0.1668103039264679, "learning_rate": 7.572339456945435e-05, "loss": 0.009, "step": 36160 }, { "grad_norm": 0.15875528752803802, "learning_rate": 7.570921450003762e-05, "loss": 0.0122, "step": 36170 }, { "grad_norm": 0.12910139560699463, "learning_rate": 7.569503161909983e-05, "loss": 0.0093, "step": 36180 }, { "grad_norm": 0.16872967779636383, "learning_rate": 7.568084592819202e-05, "loss": 0.0091, "step": 36190 }, { "grad_norm": 0.19298312067985535, "learning_rate": 7.566665742886551e-05, "loss": 0.0083, "step": 36200 }, { "grad_norm": 0.15899373590946198, "learning_rate": 7.565246612267194e-05, "loss": 0.0086, "step": 36210 }, { "grad_norm": 0.20735247433185577, "learning_rate": 7.563827201116325e-05, "loss": 0.009, "step": 36220 }, { "grad_norm": 0.1600172370672226, "learning_rate": 7.562407509589164e-05, "loss": 0.0118, "step": 36230 }, { "grad_norm": 0.20057213306427002, "learning_rate": 7.560987537840973e-05, "loss": 0.0107, "step": 36240 }, { "grad_norm": 0.14049223065376282, "learning_rate": 7.559567286027036e-05, "loss": 0.0102, "step": 36250 }, { "grad_norm": 0.19297274947166443, "learning_rate": 7.558146754302668e-05, "loss": 0.0102, "step": 36260 }, { "grad_norm": 0.14502184092998505, "learning_rate": 7.556725942823217e-05, "loss": 0.0102, "step": 36270 }, { "grad_norm": 0.14085526764392853, "learning_rate": 7.555304851744061e-05, "loss": 0.0111, "step": 36280 }, { "grad_norm": 0.15654197335243225, "learning_rate": 7.553883481220608e-05, "loss": 0.0102, "step": 36290 }, { "grad_norm": 0.12676197290420532, "learning_rate": 7.552461831408298e-05, "loss": 0.0092, "step": 36300 }, { "grad_norm": 0.1619689166545868, "learning_rate": 7.551039902462599e-05, "loss": 0.0098, "step": 36310 }, { "grad_norm": 0.1190667375922203, "learning_rate": 7.549617694539011e-05, "loss": 0.0096, "step": 36320 }, { "grad_norm": 0.15282553434371948, "learning_rate": 7.548195207793066e-05, "loss": 0.0108, "step": 36330 }, { "grad_norm": 0.17915695905685425, "learning_rate": 7.546772442380323e-05, "loss": 0.0127, "step": 36340 }, { "grad_norm": 0.1344687044620514, "learning_rate": 7.545349398456376e-05, "loss": 0.0102, "step": 36350 }, { "grad_norm": 0.15396007895469666, "learning_rate": 7.543926076176845e-05, "loss": 0.0082, "step": 36360 }, { "grad_norm": 0.16489282250404358, "learning_rate": 7.542502475697385e-05, "loss": 0.0098, "step": 36370 }, { "grad_norm": 0.16518032550811768, "learning_rate": 7.541078597173675e-05, "loss": 0.0108, "step": 36380 }, { "grad_norm": 0.16420269012451172, "learning_rate": 7.539654440761431e-05, "loss": 0.0122, "step": 36390 }, { "grad_norm": 0.11869291961193085, "learning_rate": 7.538230006616395e-05, "loss": 0.0092, "step": 36400 }, { "grad_norm": 0.16284410655498505, "learning_rate": 7.536805294894342e-05, "loss": 0.0093, "step": 36410 }, { "grad_norm": 0.13150501251220703, "learning_rate": 7.535380305751076e-05, "loss": 0.009, "step": 36420 }, { "grad_norm": 0.1581917554140091, "learning_rate": 7.533955039342431e-05, "loss": 0.0107, "step": 36430 }, { "grad_norm": 0.12613700330257416, "learning_rate": 7.532529495824274e-05, "loss": 0.0102, "step": 36440 }, { "grad_norm": 0.16588427126407623, "learning_rate": 7.531103675352497e-05, "loss": 0.01, "step": 36450 }, { "grad_norm": 0.16191759705543518, "learning_rate": 7.529677578083029e-05, "loss": 0.009, "step": 36460 }, { "grad_norm": 0.1977425068616867, "learning_rate": 7.528251204171823e-05, "loss": 0.0097, "step": 36470 }, { "grad_norm": 0.11552368849515915, "learning_rate": 7.526824553774867e-05, "loss": 0.0096, "step": 36480 }, { "grad_norm": 0.1758790910243988, "learning_rate": 7.525397627048178e-05, "loss": 0.0097, "step": 36490 }, { "grad_norm": 0.14657993614673615, "learning_rate": 7.523970424147802e-05, "loss": 0.009, "step": 36500 }, { "grad_norm": 0.18173345923423767, "learning_rate": 7.522542945229813e-05, "loss": 0.008, "step": 36510 }, { "grad_norm": 0.1304994374513626, "learning_rate": 7.521115190450321e-05, "loss": 0.0077, "step": 36520 }, { "grad_norm": 0.136684849858284, "learning_rate": 7.519687159965462e-05, "loss": 0.0084, "step": 36530 }, { "grad_norm": 0.19104832410812378, "learning_rate": 7.518258853931403e-05, "loss": 0.0119, "step": 36540 }, { "grad_norm": 0.16715064644813538, "learning_rate": 7.516830272504342e-05, "loss": 0.0089, "step": 36550 }, { "grad_norm": 0.14990602433681488, "learning_rate": 7.515401415840509e-05, "loss": 0.0082, "step": 36560 }, { "grad_norm": 0.11322885751724243, "learning_rate": 7.513972284096155e-05, "loss": 0.0088, "step": 36570 }, { "grad_norm": 0.14712604880332947, "learning_rate": 7.512542877427574e-05, "loss": 0.0126, "step": 36580 }, { "grad_norm": 0.13143061101436615, "learning_rate": 7.51111319599108e-05, "loss": 0.0118, "step": 36590 }, { "grad_norm": 0.14842170476913452, "learning_rate": 7.509683239943024e-05, "loss": 0.0112, "step": 36600 }, { "grad_norm": 0.12364630401134491, "learning_rate": 7.50825300943978e-05, "loss": 0.0113, "step": 36610 }, { "grad_norm": 0.09264295548200607, "learning_rate": 7.50682250463776e-05, "loss": 0.0107, "step": 36620 }, { "grad_norm": 0.11583953350782394, "learning_rate": 7.5053917256934e-05, "loss": 0.0086, "step": 36630 }, { "grad_norm": 0.18073104321956635, "learning_rate": 7.503960672763166e-05, "loss": 0.009, "step": 36640 }, { "grad_norm": 0.1626461297273636, "learning_rate": 7.502529346003559e-05, "loss": 0.0117, "step": 36650 }, { "grad_norm": 0.12854041159152985, "learning_rate": 7.501097745571106e-05, "loss": 0.0085, "step": 36660 }, { "grad_norm": 0.14304126799106598, "learning_rate": 7.499665871622361e-05, "loss": 0.0092, "step": 36670 }, { "grad_norm": 0.15703722834587097, "learning_rate": 7.498233724313917e-05, "loss": 0.0102, "step": 36680 }, { "grad_norm": 0.1628221720457077, "learning_rate": 7.496801303802387e-05, "loss": 0.0121, "step": 36690 }, { "grad_norm": 0.14603106677532196, "learning_rate": 7.495368610244423e-05, "loss": 0.0102, "step": 36700 }, { "grad_norm": 0.12660516798496246, "learning_rate": 7.493935643796697e-05, "loss": 0.0102, "step": 36710 }, { "grad_norm": 0.12559127807617188, "learning_rate": 7.49250240461592e-05, "loss": 0.01, "step": 36720 }, { "grad_norm": 0.12069360911846161, "learning_rate": 7.491068892858826e-05, "loss": 0.0089, "step": 36730 }, { "grad_norm": 0.12035617977380753, "learning_rate": 7.489635108682184e-05, "loss": 0.0092, "step": 36740 }, { "grad_norm": 0.08566692471504211, "learning_rate": 7.48820105224279e-05, "loss": 0.0122, "step": 36750 }, { "grad_norm": 0.207138791680336, "learning_rate": 7.48676672369747e-05, "loss": 0.0121, "step": 36760 }, { "grad_norm": 0.15938487648963928, "learning_rate": 7.485332123203079e-05, "loss": 0.0078, "step": 36770 }, { "grad_norm": 0.18470044434070587, "learning_rate": 7.483897250916504e-05, "loss": 0.0094, "step": 36780 }, { "grad_norm": 0.16684424877166748, "learning_rate": 7.48246210699466e-05, "loss": 0.0101, "step": 36790 }, { "grad_norm": 0.18027374148368835, "learning_rate": 7.481026691594492e-05, "loss": 0.0114, "step": 36800 }, { "grad_norm": 0.15301184356212616, "learning_rate": 7.479591004872975e-05, "loss": 0.0089, "step": 36810 }, { "grad_norm": 0.17665007710456848, "learning_rate": 7.478155046987114e-05, "loss": 0.0113, "step": 36820 }, { "grad_norm": 0.202742800116539, "learning_rate": 7.476718818093944e-05, "loss": 0.0125, "step": 36830 }, { "grad_norm": 0.16571518778800964, "learning_rate": 7.475282318350527e-05, "loss": 0.01, "step": 36840 }, { "grad_norm": 0.12074892967939377, "learning_rate": 7.473845547913957e-05, "loss": 0.0105, "step": 36850 }, { "grad_norm": 0.15557964146137238, "learning_rate": 7.472408506941357e-05, "loss": 0.0106, "step": 36860 }, { "grad_norm": 0.1840595304965973, "learning_rate": 7.470971195589881e-05, "loss": 0.0092, "step": 36870 }, { "grad_norm": 0.12269306182861328, "learning_rate": 7.46953361401671e-05, "loss": 0.0108, "step": 36880 }, { "grad_norm": 0.1904594451189041, "learning_rate": 7.468095762379055e-05, "loss": 0.0083, "step": 36890 }, { "grad_norm": 0.14237456023693085, "learning_rate": 7.466657640834158e-05, "loss": 0.0098, "step": 36900 }, { "grad_norm": 0.1397530883550644, "learning_rate": 7.46521924953929e-05, "loss": 0.0127, "step": 36910 }, { "grad_norm": 0.11008864641189575, "learning_rate": 7.463780588651752e-05, "loss": 0.009, "step": 36920 }, { "grad_norm": 0.12065346539020538, "learning_rate": 7.462341658328873e-05, "loss": 0.0079, "step": 36930 }, { "grad_norm": 0.17628321051597595, "learning_rate": 7.460902458728012e-05, "loss": 0.0098, "step": 36940 }, { "grad_norm": 0.17526520788669586, "learning_rate": 7.459462990006558e-05, "loss": 0.0125, "step": 36950 }, { "grad_norm": 0.1244019865989685, "learning_rate": 7.458023252321929e-05, "loss": 0.0145, "step": 36960 }, { "grad_norm": 0.1314639002084732, "learning_rate": 7.456583245831573e-05, "loss": 0.0102, "step": 36970 }, { "grad_norm": 0.14155054092407227, "learning_rate": 7.455142970692968e-05, "loss": 0.0099, "step": 36980 }, { "grad_norm": 0.18374936282634735, "learning_rate": 7.453702427063615e-05, "loss": 0.0094, "step": 36990 }, { "grad_norm": 0.13701818883419037, "learning_rate": 7.452261615101057e-05, "loss": 0.0124, "step": 37000 }, { "grad_norm": 0.2163773477077484, "learning_rate": 7.450820534962856e-05, "loss": 0.0111, "step": 37010 }, { "grad_norm": 0.13942661881446838, "learning_rate": 7.449379186806604e-05, "loss": 0.0087, "step": 37020 }, { "grad_norm": 0.13056381046772003, "learning_rate": 7.44793757078993e-05, "loss": 0.0094, "step": 37030 }, { "grad_norm": 0.1652272343635559, "learning_rate": 7.44649568707048e-05, "loss": 0.0118, "step": 37040 }, { "grad_norm": 0.1457093209028244, "learning_rate": 7.445053535805942e-05, "loss": 0.0105, "step": 37050 }, { "grad_norm": 0.17097584903240204, "learning_rate": 7.443611117154024e-05, "loss": 0.0128, "step": 37060 }, { "grad_norm": 0.1533176302909851, "learning_rate": 7.442168431272469e-05, "loss": 0.0097, "step": 37070 }, { "grad_norm": 0.16428637504577637, "learning_rate": 7.440725478319044e-05, "loss": 0.0093, "step": 37080 }, { "grad_norm": 0.19722531735897064, "learning_rate": 7.439282258451553e-05, "loss": 0.0118, "step": 37090 }, { "grad_norm": 0.17073415219783783, "learning_rate": 7.43783877182782e-05, "loss": 0.0098, "step": 37100 }, { "grad_norm": 0.15979433059692383, "learning_rate": 7.436395018605707e-05, "loss": 0.0092, "step": 37110 }, { "grad_norm": 0.21174240112304688, "learning_rate": 7.434950998943094e-05, "loss": 0.0078, "step": 37120 }, { "grad_norm": 0.11877777427434921, "learning_rate": 7.433506712997902e-05, "loss": 0.0091, "step": 37130 }, { "grad_norm": 0.12786681950092316, "learning_rate": 7.432062160928077e-05, "loss": 0.008, "step": 37140 }, { "grad_norm": 0.1897336095571518, "learning_rate": 7.430617342891588e-05, "loss": 0.0104, "step": 37150 }, { "grad_norm": 0.12900584936141968, "learning_rate": 7.429172259046442e-05, "loss": 0.009, "step": 37160 }, { "grad_norm": 0.18657879531383514, "learning_rate": 7.42772690955067e-05, "loss": 0.0119, "step": 37170 }, { "grad_norm": 0.1482619047164917, "learning_rate": 7.426281294562332e-05, "loss": 0.0104, "step": 37180 }, { "grad_norm": 0.16444815695285797, "learning_rate": 7.424835414239519e-05, "loss": 0.0144, "step": 37190 }, { "grad_norm": 0.1481471210718155, "learning_rate": 7.423389268740352e-05, "loss": 0.0104, "step": 37200 }, { "grad_norm": 0.12693184614181519, "learning_rate": 7.42194285822298e-05, "loss": 0.0128, "step": 37210 }, { "grad_norm": 0.1463911533355713, "learning_rate": 7.420496182845576e-05, "loss": 0.0111, "step": 37220 }, { "grad_norm": 0.12833529710769653, "learning_rate": 7.419049242766348e-05, "loss": 0.0134, "step": 37230 }, { "grad_norm": 0.14030276238918304, "learning_rate": 7.417602038143533e-05, "loss": 0.0084, "step": 37240 }, { "grad_norm": 0.14418642222881317, "learning_rate": 7.416154569135393e-05, "loss": 0.0107, "step": 37250 }, { "grad_norm": 0.1515274941921234, "learning_rate": 7.414706835900222e-05, "loss": 0.0105, "step": 37260 }, { "grad_norm": 0.2345154881477356, "learning_rate": 7.413258838596345e-05, "loss": 0.0096, "step": 37270 }, { "grad_norm": 0.18323254585266113, "learning_rate": 7.411810577382106e-05, "loss": 0.012, "step": 37280 }, { "grad_norm": 0.18575991690158844, "learning_rate": 7.410362052415889e-05, "loss": 0.0087, "step": 37290 }, { "grad_norm": 0.17531763017177582, "learning_rate": 7.408913263856102e-05, "loss": 0.0105, "step": 37300 }, { "grad_norm": 0.1258610635995865, "learning_rate": 7.407464211861183e-05, "loss": 0.0088, "step": 37310 }, { "grad_norm": 0.15529774129390717, "learning_rate": 7.406014896589597e-05, "loss": 0.0132, "step": 37320 }, { "grad_norm": 0.23477715253829956, "learning_rate": 7.404565318199838e-05, "loss": 0.0123, "step": 37330 }, { "grad_norm": 0.17997238039970398, "learning_rate": 7.403115476850432e-05, "loss": 0.0114, "step": 37340 }, { "grad_norm": 0.16167905926704407, "learning_rate": 7.401665372699932e-05, "loss": 0.0111, "step": 37350 }, { "grad_norm": 0.1658868044614792, "learning_rate": 7.400215005906916e-05, "loss": 0.0117, "step": 37360 }, { "grad_norm": 0.16010580956935883, "learning_rate": 7.398764376629996e-05, "loss": 0.0084, "step": 37370 }, { "grad_norm": 0.14563453197479248, "learning_rate": 7.39731348502781e-05, "loss": 0.0103, "step": 37380 }, { "grad_norm": 0.149265855550766, "learning_rate": 7.395862331259028e-05, "loss": 0.0121, "step": 37390 }, { "grad_norm": 0.14945614337921143, "learning_rate": 7.394410915482341e-05, "loss": 0.0113, "step": 37400 }, { "grad_norm": 0.13068515062332153, "learning_rate": 7.392959237856476e-05, "loss": 0.0128, "step": 37410 }, { "grad_norm": 0.13648921251296997, "learning_rate": 7.391507298540187e-05, "loss": 0.0093, "step": 37420 }, { "grad_norm": 0.15794464945793152, "learning_rate": 7.390055097692257e-05, "loss": 0.0099, "step": 37430 }, { "grad_norm": 0.14963242411613464, "learning_rate": 7.388602635471494e-05, "loss": 0.0125, "step": 37440 }, { "grad_norm": 0.18697336316108704, "learning_rate": 7.387149912036738e-05, "loss": 0.0101, "step": 37450 }, { "grad_norm": 0.17511968314647675, "learning_rate": 7.385696927546854e-05, "loss": 0.0133, "step": 37460 }, { "grad_norm": 0.1903676986694336, "learning_rate": 7.384243682160743e-05, "loss": 0.0114, "step": 37470 }, { "grad_norm": 0.1924910694360733, "learning_rate": 7.382790176037329e-05, "loss": 0.0101, "step": 37480 }, { "grad_norm": 0.20567072927951813, "learning_rate": 7.381336409335562e-05, "loss": 0.0103, "step": 37490 }, { "grad_norm": 0.13005085289478302, "learning_rate": 7.379882382214426e-05, "loss": 0.0108, "step": 37500 }, { "grad_norm": 0.1395324319601059, "learning_rate": 7.378428094832931e-05, "loss": 0.0104, "step": 37510 }, { "grad_norm": 0.12385425716638565, "learning_rate": 7.376973547350114e-05, "loss": 0.0123, "step": 37520 }, { "grad_norm": 0.19293756783008575, "learning_rate": 7.375518739925046e-05, "loss": 0.0108, "step": 37530 }, { "grad_norm": 0.17471656203269958, "learning_rate": 7.374063672716818e-05, "loss": 0.0109, "step": 37540 }, { "grad_norm": 0.13944312930107117, "learning_rate": 7.372608345884558e-05, "loss": 0.0101, "step": 37550 }, { "grad_norm": 0.1372007429599762, "learning_rate": 7.371152759587415e-05, "loss": 0.0103, "step": 37560 }, { "grad_norm": 0.14581149816513062, "learning_rate": 7.369696913984572e-05, "loss": 0.0132, "step": 37570 }, { "grad_norm": 0.1604749709367752, "learning_rate": 7.368240809235237e-05, "loss": 0.01, "step": 37580 }, { "grad_norm": 0.1472436487674713, "learning_rate": 7.366784445498649e-05, "loss": 0.0096, "step": 37590 }, { "grad_norm": 0.11964692175388336, "learning_rate": 7.36532782293407e-05, "loss": 0.0069, "step": 37600 }, { "grad_norm": 0.15423956513404846, "learning_rate": 7.363870941700797e-05, "loss": 0.0095, "step": 37610 }, { "grad_norm": 0.11322619765996933, "learning_rate": 7.362413801958152e-05, "loss": 0.0079, "step": 37620 }, { "grad_norm": 0.16641631722450256, "learning_rate": 7.360956403865486e-05, "loss": 0.0087, "step": 37630 }, { "grad_norm": 0.21218569576740265, "learning_rate": 7.359498747582177e-05, "loss": 0.0087, "step": 37640 }, { "grad_norm": 0.20526807010173798, "learning_rate": 7.358040833267634e-05, "loss": 0.0106, "step": 37650 }, { "grad_norm": 0.1351737529039383, "learning_rate": 7.356582661081288e-05, "loss": 0.0105, "step": 37660 }, { "grad_norm": 0.1524611860513687, "learning_rate": 7.355124231182607e-05, "loss": 0.0112, "step": 37670 }, { "grad_norm": 0.14079692959785461, "learning_rate": 7.353665543731083e-05, "loss": 0.0136, "step": 37680 }, { "grad_norm": 0.11322113126516342, "learning_rate": 7.352206598886229e-05, "loss": 0.0111, "step": 37690 }, { "grad_norm": 0.14006777107715607, "learning_rate": 7.350747396807601e-05, "loss": 0.0082, "step": 37700 }, { "grad_norm": 0.1167503148317337, "learning_rate": 7.349287937654772e-05, "loss": 0.0094, "step": 37710 }, { "grad_norm": 0.15083591639995575, "learning_rate": 7.347828221587345e-05, "loss": 0.0092, "step": 37720 }, { "grad_norm": 0.2160237729549408, "learning_rate": 7.346368248764955e-05, "loss": 0.0092, "step": 37730 }, { "grad_norm": 0.2125367522239685, "learning_rate": 7.34490801934726e-05, "loss": 0.0105, "step": 37740 }, { "grad_norm": 0.1268092542886734, "learning_rate": 7.343447533493947e-05, "loss": 0.0105, "step": 37750 }, { "grad_norm": 0.15254457294940948, "learning_rate": 7.341986791364739e-05, "loss": 0.0094, "step": 37760 }, { "grad_norm": 0.1204846054315567, "learning_rate": 7.340525793119374e-05, "loss": 0.0089, "step": 37770 }, { "grad_norm": 0.12935851514339447, "learning_rate": 7.33906453891763e-05, "loss": 0.0117, "step": 37780 }, { "grad_norm": 0.17682738602161407, "learning_rate": 7.337603028919303e-05, "loss": 0.0082, "step": 37790 }, { "grad_norm": 0.1345185786485672, "learning_rate": 7.336141263284225e-05, "loss": 0.0091, "step": 37800 }, { "grad_norm": 0.12530089914798737, "learning_rate": 7.334679242172249e-05, "loss": 0.0102, "step": 37810 }, { "grad_norm": 0.15504124760627747, "learning_rate": 7.33321696574326e-05, "loss": 0.0105, "step": 37820 }, { "grad_norm": 0.15439003705978394, "learning_rate": 7.331754434157174e-05, "loss": 0.0081, "step": 37830 }, { "grad_norm": 0.18265590071678162, "learning_rate": 7.330291647573928e-05, "loss": 0.0101, "step": 37840 }, { "grad_norm": 0.127482607960701, "learning_rate": 7.32882860615349e-05, "loss": 0.0119, "step": 37850 }, { "grad_norm": 0.1129908338189125, "learning_rate": 7.327365310055858e-05, "loss": 0.0085, "step": 37860 }, { "grad_norm": 0.13261406123638153, "learning_rate": 7.325901759441055e-05, "loss": 0.0083, "step": 37870 }, { "grad_norm": 0.1263551265001297, "learning_rate": 7.324437954469132e-05, "loss": 0.009, "step": 37880 }, { "grad_norm": 0.1584750860929489, "learning_rate": 7.32297389530017e-05, "loss": 0.0089, "step": 37890 }, { "grad_norm": 0.14217303693294525, "learning_rate": 7.321509582094274e-05, "loss": 0.0079, "step": 37900 }, { "grad_norm": 0.13818404078483582, "learning_rate": 7.320045015011582e-05, "loss": 0.0085, "step": 37910 }, { "grad_norm": 0.1561713069677353, "learning_rate": 7.318580194212253e-05, "loss": 0.0086, "step": 37920 }, { "grad_norm": 0.16968697309494019, "learning_rate": 7.317115119856483e-05, "loss": 0.0094, "step": 37930 }, { "grad_norm": 0.12805649638175964, "learning_rate": 7.315649792104485e-05, "loss": 0.0092, "step": 37940 }, { "grad_norm": 0.15448281168937683, "learning_rate": 7.314184211116507e-05, "loss": 0.0085, "step": 37950 }, { "grad_norm": 0.14284645020961761, "learning_rate": 7.312718377052823e-05, "loss": 0.0098, "step": 37960 }, { "grad_norm": 0.13763613998889923, "learning_rate": 7.311252290073737e-05, "loss": 0.0104, "step": 37970 }, { "grad_norm": 0.20204973220825195, "learning_rate": 7.309785950339572e-05, "loss": 0.0103, "step": 37980 }, { "grad_norm": 0.11680524796247482, "learning_rate": 7.308319358010692e-05, "loss": 0.0078, "step": 37990 }, { "grad_norm": 0.17153464257717133, "learning_rate": 7.306852513247474e-05, "loss": 0.0128, "step": 38000 }, { "grad_norm": 0.18193860352039337, "learning_rate": 7.305385416210336e-05, "loss": 0.0102, "step": 38010 }, { "grad_norm": 0.16474583745002747, "learning_rate": 7.303918067059711e-05, "loss": 0.0095, "step": 38020 }, { "grad_norm": 0.14388291537761688, "learning_rate": 7.302450465956075e-05, "loss": 0.0078, "step": 38030 }, { "grad_norm": 0.10511767864227295, "learning_rate": 7.300982613059914e-05, "loss": 0.0077, "step": 38040 }, { "grad_norm": 0.19091275334358215, "learning_rate": 7.299514508531757e-05, "loss": 0.0103, "step": 38050 }, { "grad_norm": 0.12573300302028656, "learning_rate": 7.298046152532148e-05, "loss": 0.0093, "step": 38060 }, { "grad_norm": 0.13779251277446747, "learning_rate": 7.296577545221668e-05, "loss": 0.0075, "step": 38070 }, { "grad_norm": 0.19811725616455078, "learning_rate": 7.29510868676092e-05, "loss": 0.0098, "step": 38080 }, { "grad_norm": 0.11901479214429855, "learning_rate": 7.293639577310538e-05, "loss": 0.0119, "step": 38090 }, { "grad_norm": 0.1915784776210785, "learning_rate": 7.292170217031179e-05, "loss": 0.0119, "step": 38100 }, { "grad_norm": 0.1312355399131775, "learning_rate": 7.290700606083532e-05, "loss": 0.0098, "step": 38110 }, { "grad_norm": 0.21250097453594208, "learning_rate": 7.28923074462831e-05, "loss": 0.0104, "step": 38120 }, { "grad_norm": 0.17391039431095123, "learning_rate": 7.287760632826255e-05, "loss": 0.0122, "step": 38130 }, { "grad_norm": 0.12138700485229492, "learning_rate": 7.286290270838139e-05, "loss": 0.0082, "step": 38140 }, { "grad_norm": 0.09251000732183456, "learning_rate": 7.284819658824756e-05, "loss": 0.0084, "step": 38150 }, { "grad_norm": 0.15786288678646088, "learning_rate": 7.283348796946929e-05, "loss": 0.0104, "step": 38160 }, { "grad_norm": 0.13050362467765808, "learning_rate": 7.28187768536551e-05, "loss": 0.0113, "step": 38170 }, { "grad_norm": 0.11770743131637573, "learning_rate": 7.280406324241379e-05, "loss": 0.0078, "step": 38180 }, { "grad_norm": 0.11879673600196838, "learning_rate": 7.278934713735441e-05, "loss": 0.0099, "step": 38190 }, { "grad_norm": 0.1524057239294052, "learning_rate": 7.277462854008629e-05, "loss": 0.009, "step": 38200 }, { "grad_norm": 0.18828202784061432, "learning_rate": 7.275990745221904e-05, "loss": 0.0105, "step": 38210 }, { "grad_norm": 0.1564924269914627, "learning_rate": 7.274518387536252e-05, "loss": 0.0099, "step": 38220 }, { "grad_norm": 0.1714199334383011, "learning_rate": 7.273045781112689e-05, "loss": 0.0118, "step": 38230 }, { "grad_norm": 0.15973195433616638, "learning_rate": 7.271572926112257e-05, "loss": 0.0093, "step": 38240 }, { "grad_norm": 0.17829973995685577, "learning_rate": 7.270099822696024e-05, "loss": 0.0111, "step": 38250 }, { "grad_norm": 0.16311635076999664, "learning_rate": 7.26862647102509e-05, "loss": 0.0104, "step": 38260 }, { "grad_norm": 0.1595093458890915, "learning_rate": 7.267152871260573e-05, "loss": 0.0094, "step": 38270 }, { "grad_norm": 0.1639014333486557, "learning_rate": 7.265679023563629e-05, "loss": 0.0098, "step": 38280 }, { "grad_norm": 0.15091834962368011, "learning_rate": 7.264204928095433e-05, "loss": 0.0098, "step": 38290 }, { "grad_norm": 0.13302525877952576, "learning_rate": 7.262730585017188e-05, "loss": 0.009, "step": 38300 }, { "grad_norm": 0.11552393436431885, "learning_rate": 7.261255994490131e-05, "loss": 0.0099, "step": 38310 }, { "grad_norm": 0.11343579739332199, "learning_rate": 7.259781156675517e-05, "loss": 0.0084, "step": 38320 }, { "grad_norm": 0.14899468421936035, "learning_rate": 7.258306071734633e-05, "loss": 0.0111, "step": 38330 }, { "grad_norm": 0.13763512670993805, "learning_rate": 7.256830739828792e-05, "loss": 0.0133, "step": 38340 }, { "grad_norm": 0.1568269431591034, "learning_rate": 7.255355161119336e-05, "loss": 0.0099, "step": 38350 }, { "grad_norm": 0.1550169289112091, "learning_rate": 7.253879335767628e-05, "loss": 0.009, "step": 38360 }, { "grad_norm": 0.2553246021270752, "learning_rate": 7.252403263935067e-05, "loss": 0.0116, "step": 38370 }, { "grad_norm": 0.2692388594150543, "learning_rate": 7.25092694578307e-05, "loss": 0.0106, "step": 38380 }, { "grad_norm": 0.12495012581348419, "learning_rate": 7.249450381473087e-05, "loss": 0.0085, "step": 38390 }, { "grad_norm": 0.1349116861820221, "learning_rate": 7.247973571166593e-05, "loss": 0.0093, "step": 38400 }, { "grad_norm": 0.1589568704366684, "learning_rate": 7.246496515025089e-05, "loss": 0.012, "step": 38410 }, { "grad_norm": 0.19381563365459442, "learning_rate": 7.245019213210101e-05, "loss": 0.0109, "step": 38420 }, { "grad_norm": 0.14663870632648468, "learning_rate": 7.24354166588319e-05, "loss": 0.0104, "step": 38430 }, { "grad_norm": 0.13882645964622498, "learning_rate": 7.242063873205935e-05, "loss": 0.0088, "step": 38440 }, { "grad_norm": 0.10016459226608276, "learning_rate": 7.240585835339946e-05, "loss": 0.0095, "step": 38450 }, { "grad_norm": 0.12836433947086334, "learning_rate": 7.239107552446857e-05, "loss": 0.0115, "step": 38460 }, { "grad_norm": 0.1462632417678833, "learning_rate": 7.237629024688334e-05, "loss": 0.0107, "step": 38470 }, { "grad_norm": 0.19154959917068481, "learning_rate": 7.236150252226064e-05, "loss": 0.0101, "step": 38480 }, { "grad_norm": 0.19381995499134064, "learning_rate": 7.234671235221765e-05, "loss": 0.011, "step": 38490 }, { "grad_norm": 0.1522667407989502, "learning_rate": 7.233191973837179e-05, "loss": 0.0119, "step": 38500 }, { "grad_norm": 0.20659707486629486, "learning_rate": 7.231712468234078e-05, "loss": 0.0116, "step": 38510 }, { "grad_norm": 0.18431946635246277, "learning_rate": 7.230232718574254e-05, "loss": 0.0089, "step": 38520 }, { "grad_norm": 0.14083078503608704, "learning_rate": 7.228752725019535e-05, "loss": 0.0096, "step": 38530 }, { "grad_norm": 0.13114194571971893, "learning_rate": 7.227272487731769e-05, "loss": 0.0116, "step": 38540 }, { "grad_norm": 0.18323631584644318, "learning_rate": 7.225792006872831e-05, "loss": 0.0125, "step": 38550 }, { "grad_norm": 0.12505362927913666, "learning_rate": 7.224311282604628e-05, "loss": 0.0088, "step": 38560 }, { "grad_norm": 0.1007862314581871, "learning_rate": 7.222830315089085e-05, "loss": 0.0088, "step": 38570 }, { "grad_norm": 0.15559574961662292, "learning_rate": 7.22134910448816e-05, "loss": 0.0102, "step": 38580 }, { "grad_norm": 0.15453219413757324, "learning_rate": 7.219867650963839e-05, "loss": 0.0107, "step": 38590 }, { "grad_norm": 0.12035142630338669, "learning_rate": 7.21838595467813e-05, "loss": 0.0108, "step": 38600 }, { "grad_norm": 0.14532938599586487, "learning_rate": 7.216904015793067e-05, "loss": 0.0105, "step": 38610 }, { "grad_norm": 0.1692330539226532, "learning_rate": 7.215421834470713e-05, "loss": 0.0139, "step": 38620 }, { "grad_norm": 0.1154639720916748, "learning_rate": 7.213939410873158e-05, "loss": 0.0094, "step": 38630 }, { "grad_norm": 0.15959390997886658, "learning_rate": 7.212456745162519e-05, "loss": 0.0103, "step": 38640 }, { "grad_norm": 0.1465870440006256, "learning_rate": 7.210973837500937e-05, "loss": 0.0097, "step": 38650 }, { "grad_norm": 0.1703789234161377, "learning_rate": 7.209490688050578e-05, "loss": 0.0106, "step": 38660 }, { "grad_norm": 0.1278703361749649, "learning_rate": 7.208007296973641e-05, "loss": 0.0088, "step": 38670 }, { "grad_norm": 0.15710856020450592, "learning_rate": 7.206523664432345e-05, "loss": 0.0101, "step": 38680 }, { "grad_norm": 0.17986176908016205, "learning_rate": 7.205039790588939e-05, "loss": 0.0102, "step": 38690 }, { "grad_norm": 0.1502174437046051, "learning_rate": 7.203555675605697e-05, "loss": 0.0089, "step": 38700 }, { "grad_norm": 0.18616440892219543, "learning_rate": 7.202071319644917e-05, "loss": 0.0097, "step": 38710 }, { "grad_norm": 0.12820303440093994, "learning_rate": 7.200586722868932e-05, "loss": 0.01, "step": 38720 }, { "grad_norm": 0.16166695952415466, "learning_rate": 7.199101885440088e-05, "loss": 0.0111, "step": 38730 }, { "grad_norm": 0.1328854113817215, "learning_rate": 7.197616807520771e-05, "loss": 0.0111, "step": 38740 }, { "grad_norm": 0.154733806848526, "learning_rate": 7.196131489273381e-05, "loss": 0.0124, "step": 38750 }, { "grad_norm": 0.13331468403339386, "learning_rate": 7.194645930860355e-05, "loss": 0.0091, "step": 38760 }, { "grad_norm": 0.15501157939434052, "learning_rate": 7.19316013244415e-05, "loss": 0.0111, "step": 38770 }, { "grad_norm": 0.15584491193294525, "learning_rate": 7.191674094187248e-05, "loss": 0.0082, "step": 38780 }, { "grad_norm": 0.1802246868610382, "learning_rate": 7.190187816252165e-05, "loss": 0.0086, "step": 38790 }, { "grad_norm": 0.11818312108516693, "learning_rate": 7.188701298801435e-05, "loss": 0.0092, "step": 38800 }, { "grad_norm": 0.1554499864578247, "learning_rate": 7.18721454199762e-05, "loss": 0.0098, "step": 38810 }, { "grad_norm": 0.11128924041986465, "learning_rate": 7.185727546003312e-05, "loss": 0.0099, "step": 38820 }, { "grad_norm": 0.21665890514850616, "learning_rate": 7.184240310981126e-05, "loss": 0.0103, "step": 38830 }, { "grad_norm": 0.1487477570772171, "learning_rate": 7.182752837093704e-05, "loss": 0.0095, "step": 38840 }, { "grad_norm": 0.13783876597881317, "learning_rate": 7.181265124503711e-05, "loss": 0.0104, "step": 38850 }, { "grad_norm": 0.10208401829004288, "learning_rate": 7.179777173373847e-05, "loss": 0.0132, "step": 38860 }, { "grad_norm": 0.15792766213417053, "learning_rate": 7.178288983866826e-05, "loss": 0.0099, "step": 38870 }, { "grad_norm": 0.16642095148563385, "learning_rate": 7.176800556145397e-05, "loss": 0.0104, "step": 38880 }, { "grad_norm": 0.15093688666820526, "learning_rate": 7.175311890372334e-05, "loss": 0.01, "step": 38890 }, { "grad_norm": 0.13611267507076263, "learning_rate": 7.17382298671043e-05, "loss": 0.0118, "step": 38900 }, { "grad_norm": 0.16307926177978516, "learning_rate": 7.172333845322515e-05, "loss": 0.0108, "step": 38910 }, { "grad_norm": 0.13893766701221466, "learning_rate": 7.170844466371436e-05, "loss": 0.0089, "step": 38920 }, { "grad_norm": 0.11289000511169434, "learning_rate": 7.16935485002007e-05, "loss": 0.0074, "step": 38930 }, { "grad_norm": 0.15126435458660126, "learning_rate": 7.167864996431319e-05, "loss": 0.0087, "step": 38940 }, { "grad_norm": 0.15489766001701355, "learning_rate": 7.166374905768111e-05, "loss": 0.0111, "step": 38950 }, { "grad_norm": 0.12534834444522858, "learning_rate": 7.164884578193401e-05, "loss": 0.0085, "step": 38960 }, { "grad_norm": 0.18813259899616241, "learning_rate": 7.163394013870168e-05, "loss": 0.009, "step": 38970 }, { "grad_norm": 0.1857166588306427, "learning_rate": 7.161903212961419e-05, "loss": 0.0094, "step": 38980 }, { "grad_norm": 0.1342872679233551, "learning_rate": 7.160412175630183e-05, "loss": 0.0099, "step": 38990 }, { "grad_norm": 0.17135576903820038, "learning_rate": 7.158920902039521e-05, "loss": 0.0089, "step": 39000 }, { "grad_norm": 0.1659494787454605, "learning_rate": 7.157429392352514e-05, "loss": 0.0114, "step": 39010 }, { "grad_norm": 0.16136376559734344, "learning_rate": 7.155937646732275e-05, "loss": 0.0089, "step": 39020 }, { "grad_norm": 0.1510789841413498, "learning_rate": 7.154445665341933e-05, "loss": 0.0107, "step": 39030 }, { "grad_norm": 0.1541212648153305, "learning_rate": 7.152953448344654e-05, "loss": 0.0114, "step": 39040 }, { "grad_norm": 0.14154528081417084, "learning_rate": 7.151460995903624e-05, "loss": 0.0075, "step": 39050 }, { "grad_norm": 0.14560121297836304, "learning_rate": 7.149968308182052e-05, "loss": 0.0115, "step": 39060 }, { "grad_norm": 0.18355652689933777, "learning_rate": 7.14847538534318e-05, "loss": 0.0106, "step": 39070 }, { "grad_norm": 0.1783284693956375, "learning_rate": 7.14698222755027e-05, "loss": 0.0084, "step": 39080 }, { "grad_norm": 0.16681717336177826, "learning_rate": 7.14548883496661e-05, "loss": 0.0103, "step": 39090 }, { "grad_norm": 0.10463259369134903, "learning_rate": 7.143995207755517e-05, "loss": 0.0103, "step": 39100 }, { "grad_norm": 0.11946147680282593, "learning_rate": 7.142501346080333e-05, "loss": 0.0083, "step": 39110 }, { "grad_norm": 0.1643315702676773, "learning_rate": 7.141007250104421e-05, "loss": 0.0129, "step": 39120 }, { "grad_norm": 0.14170852303504944, "learning_rate": 7.139512919991176e-05, "loss": 0.013, "step": 39130 }, { "grad_norm": 0.16397061944007874, "learning_rate": 7.138018355904014e-05, "loss": 0.0102, "step": 39140 }, { "grad_norm": 0.14576461911201477, "learning_rate": 7.13652355800638e-05, "loss": 0.008, "step": 39150 }, { "grad_norm": 0.20991970598697662, "learning_rate": 7.135028526461743e-05, "loss": 0.0092, "step": 39160 }, { "grad_norm": 0.12466856092214584, "learning_rate": 7.133533261433594e-05, "loss": 0.0082, "step": 39170 }, { "grad_norm": 0.18265043199062347, "learning_rate": 7.132037763085457e-05, "loss": 0.0085, "step": 39180 }, { "grad_norm": 0.19036392867565155, "learning_rate": 7.130542031580875e-05, "loss": 0.0104, "step": 39190 }, { "grad_norm": 0.1932864934206009, "learning_rate": 7.12904606708342e-05, "loss": 0.009, "step": 39200 }, { "grad_norm": 0.17372708022594452, "learning_rate": 7.127549869756687e-05, "loss": 0.0097, "step": 39210 }, { "grad_norm": 0.1505516618490219, "learning_rate": 7.126053439764299e-05, "loss": 0.0094, "step": 39220 }, { "grad_norm": 0.15872274339199066, "learning_rate": 7.124556777269904e-05, "loss": 0.0096, "step": 39230 }, { "grad_norm": 0.18488313257694244, "learning_rate": 7.123059882437174e-05, "loss": 0.0136, "step": 39240 }, { "grad_norm": 0.1359449326992035, "learning_rate": 7.121562755429807e-05, "loss": 0.0069, "step": 39250 }, { "grad_norm": 0.15684574842453003, "learning_rate": 7.120065396411528e-05, "loss": 0.0075, "step": 39260 }, { "grad_norm": 0.1551939845085144, "learning_rate": 7.118567805546084e-05, "loss": 0.0077, "step": 39270 }, { "grad_norm": 0.17677420377731323, "learning_rate": 7.117069982997248e-05, "loss": 0.0116, "step": 39280 }, { "grad_norm": 0.15946044027805328, "learning_rate": 7.115571928928825e-05, "loss": 0.0112, "step": 39290 }, { "grad_norm": 0.1990761011838913, "learning_rate": 7.114073643504635e-05, "loss": 0.0094, "step": 39300 }, { "grad_norm": 0.14874553680419922, "learning_rate": 7.11257512688853e-05, "loss": 0.0097, "step": 39310 }, { "grad_norm": 0.15524309873580933, "learning_rate": 7.111076379244384e-05, "loss": 0.0106, "step": 39320 }, { "grad_norm": 0.11557763814926147, "learning_rate": 7.109577400736101e-05, "loss": 0.0107, "step": 39330 }, { "grad_norm": 0.13304463028907776, "learning_rate": 7.108078191527605e-05, "loss": 0.0076, "step": 39340 }, { "grad_norm": 0.19452232122421265, "learning_rate": 7.106578751782847e-05, "loss": 0.0125, "step": 39350 }, { "grad_norm": 0.1589631736278534, "learning_rate": 7.105079081665803e-05, "loss": 0.0098, "step": 39360 }, { "grad_norm": 0.13693943619728088, "learning_rate": 7.103579181340476e-05, "loss": 0.0081, "step": 39370 }, { "grad_norm": 0.17842575907707214, "learning_rate": 7.102079050970893e-05, "loss": 0.0089, "step": 39380 }, { "grad_norm": 0.15867799520492554, "learning_rate": 7.100578690721105e-05, "loss": 0.0097, "step": 39390 }, { "grad_norm": 0.18000204861164093, "learning_rate": 7.09907810075519e-05, "loss": 0.0085, "step": 39400 }, { "grad_norm": 0.14861446619033813, "learning_rate": 7.097577281237249e-05, "loss": 0.0094, "step": 39410 }, { "grad_norm": 0.12313400954008102, "learning_rate": 7.09607623233141e-05, "loss": 0.0099, "step": 39420 }, { "grad_norm": 0.16341732442378998, "learning_rate": 7.094574954201827e-05, "loss": 0.011, "step": 39430 }, { "grad_norm": 0.14168253540992737, "learning_rate": 7.093073447012675e-05, "loss": 0.0112, "step": 39440 }, { "grad_norm": 0.12950293719768524, "learning_rate": 7.09157171092816e-05, "loss": 0.0093, "step": 39450 }, { "grad_norm": 0.1274385303258896, "learning_rate": 7.090069746112504e-05, "loss": 0.0092, "step": 39460 }, { "grad_norm": 0.1738554835319519, "learning_rate": 7.088567552729965e-05, "loss": 0.0096, "step": 39470 }, { "grad_norm": 0.1292485147714615, "learning_rate": 7.087065130944818e-05, "loss": 0.0089, "step": 39480 }, { "grad_norm": 0.1433672457933426, "learning_rate": 7.085562480921366e-05, "loss": 0.0102, "step": 39490 }, { "grad_norm": 0.12623639404773712, "learning_rate": 7.084059602823937e-05, "loss": 0.0083, "step": 39500 }, { "grad_norm": 0.17514564096927643, "learning_rate": 7.082556496816882e-05, "loss": 0.0138, "step": 39510 }, { "grad_norm": 0.1323332041501999, "learning_rate": 7.081053163064582e-05, "loss": 0.0108, "step": 39520 }, { "grad_norm": 0.13377894461154938, "learning_rate": 7.079549601731434e-05, "loss": 0.0079, "step": 39530 }, { "grad_norm": 0.16568177938461304, "learning_rate": 7.07804581298187e-05, "loss": 0.0091, "step": 39540 }, { "grad_norm": 0.12535975873470306, "learning_rate": 7.07654179698034e-05, "loss": 0.009, "step": 39550 }, { "grad_norm": 0.1249440386891365, "learning_rate": 7.075037553891321e-05, "loss": 0.0078, "step": 39560 }, { "grad_norm": 0.11791206151247025, "learning_rate": 7.073533083879315e-05, "loss": 0.009, "step": 39570 }, { "grad_norm": 0.18186770379543304, "learning_rate": 7.072028387108849e-05, "loss": 0.0109, "step": 39580 }, { "grad_norm": 0.1599419116973877, "learning_rate": 7.070523463744472e-05, "loss": 0.01, "step": 39590 }, { "grad_norm": 0.13453426957130432, "learning_rate": 7.069018313950763e-05, "loss": 0.0085, "step": 39600 }, { "grad_norm": 0.11767565459012985, "learning_rate": 7.067512937892322e-05, "loss": 0.0106, "step": 39610 }, { "grad_norm": 0.15629807114601135, "learning_rate": 7.066007335733775e-05, "loss": 0.0073, "step": 39620 }, { "grad_norm": 0.12586264312267303, "learning_rate": 7.064501507639772e-05, "loss": 0.0082, "step": 39630 }, { "grad_norm": 0.12610405683517456, "learning_rate": 7.062995453774987e-05, "loss": 0.0087, "step": 39640 }, { "grad_norm": 0.13309824466705322, "learning_rate": 7.061489174304121e-05, "loss": 0.0072, "step": 39650 }, { "grad_norm": 0.16413751244544983, "learning_rate": 7.0599826693919e-05, "loss": 0.0093, "step": 39660 }, { "grad_norm": 0.16103285551071167, "learning_rate": 7.05847593920307e-05, "loss": 0.0092, "step": 39670 }, { "grad_norm": 0.16166317462921143, "learning_rate": 7.056968983902406e-05, "loss": 0.0097, "step": 39680 }, { "grad_norm": 0.149245023727417, "learning_rate": 7.055461803654706e-05, "loss": 0.0102, "step": 39690 }, { "grad_norm": 0.14345471560955048, "learning_rate": 7.053954398624794e-05, "loss": 0.0104, "step": 39700 }, { "grad_norm": 0.14546635746955872, "learning_rate": 7.052446768977518e-05, "loss": 0.0117, "step": 39710 }, { "grad_norm": 0.13496659696102142, "learning_rate": 7.050938914877748e-05, "loss": 0.009, "step": 39720 }, { "grad_norm": 0.17919160425662994, "learning_rate": 7.049430836490382e-05, "loss": 0.0118, "step": 39730 }, { "grad_norm": 0.14836330711841583, "learning_rate": 7.047922533980342e-05, "loss": 0.0131, "step": 39740 }, { "grad_norm": 0.19711989164352417, "learning_rate": 7.046414007512571e-05, "loss": 0.0081, "step": 39750 }, { "grad_norm": 0.11115194112062454, "learning_rate": 7.044905257252042e-05, "loss": 0.009, "step": 39760 }, { "grad_norm": 0.1358633190393448, "learning_rate": 7.04339628336375e-05, "loss": 0.0094, "step": 39770 }, { "grad_norm": 0.09535624086856842, "learning_rate": 7.041887086012711e-05, "loss": 0.0087, "step": 39780 }, { "grad_norm": 0.11545128375291824, "learning_rate": 7.040377665363969e-05, "loss": 0.01, "step": 39790 }, { "grad_norm": 0.12278378754854202, "learning_rate": 7.038868021582594e-05, "loss": 0.0095, "step": 39800 }, { "grad_norm": 0.16455623507499695, "learning_rate": 7.037358154833679e-05, "loss": 0.0104, "step": 39810 }, { "grad_norm": 0.1549268364906311, "learning_rate": 7.035848065282339e-05, "loss": 0.0108, "step": 39820 }, { "grad_norm": 0.1503935158252716, "learning_rate": 7.034337753093714e-05, "loss": 0.0089, "step": 39830 }, { "grad_norm": 0.17825765907764435, "learning_rate": 7.032827218432972e-05, "loss": 0.0102, "step": 39840 }, { "grad_norm": 0.15251420438289642, "learning_rate": 7.031316461465302e-05, "loss": 0.0109, "step": 39850 }, { "grad_norm": 0.1521739810705185, "learning_rate": 7.029805482355915e-05, "loss": 0.0097, "step": 39860 }, { "grad_norm": 0.1255362182855606, "learning_rate": 7.028294281270055e-05, "loss": 0.0081, "step": 39870 }, { "grad_norm": 0.17176654934883118, "learning_rate": 7.026782858372982e-05, "loss": 0.0096, "step": 39880 }, { "grad_norm": 0.1553684026002884, "learning_rate": 7.025271213829982e-05, "loss": 0.0124, "step": 39890 }, { "grad_norm": 0.15914416313171387, "learning_rate": 7.023759347806366e-05, "loss": 0.0116, "step": 39900 }, { "grad_norm": 0.16892001032829285, "learning_rate": 7.022247260467472e-05, "loss": 0.0095, "step": 39910 }, { "grad_norm": 0.1431811898946762, "learning_rate": 7.020734951978658e-05, "loss": 0.0111, "step": 39920 }, { "grad_norm": 0.14982850849628448, "learning_rate": 7.019222422505307e-05, "loss": 0.0128, "step": 39930 }, { "grad_norm": 0.15452127158641815, "learning_rate": 7.017709672212827e-05, "loss": 0.0088, "step": 39940 }, { "grad_norm": 0.1897956132888794, "learning_rate": 7.016196701266652e-05, "loss": 0.0078, "step": 39950 }, { "grad_norm": 0.18089401721954346, "learning_rate": 7.014683509832238e-05, "loss": 0.0107, "step": 39960 }, { "grad_norm": 0.13647791743278503, "learning_rate": 7.013170098075063e-05, "loss": 0.0073, "step": 39970 }, { "grad_norm": 0.17481105029582977, "learning_rate": 7.011656466160632e-05, "loss": 0.0098, "step": 39980 }, { "grad_norm": 0.15680725872516632, "learning_rate": 7.010142614254475e-05, "loss": 0.0081, "step": 39990 }, { "grad_norm": 0.14165689051151276, "learning_rate": 7.008628542522147e-05, "loss": 0.008, "step": 40000 }, { "grad_norm": 0.18425410985946655, "learning_rate": 7.007114251129218e-05, "loss": 0.0103, "step": 40010 }, { "grad_norm": 0.2432059794664383, "learning_rate": 7.005599740241293e-05, "loss": 0.0116, "step": 40020 }, { "grad_norm": 0.1283426284790039, "learning_rate": 7.004085010023996e-05, "loss": 0.0104, "step": 40030 }, { "grad_norm": 0.12693002820014954, "learning_rate": 7.002570060642976e-05, "loss": 0.0084, "step": 40040 }, { "grad_norm": 0.11457281559705734, "learning_rate": 7.001054892263903e-05, "loss": 0.011, "step": 40050 }, { "grad_norm": 0.16460439562797546, "learning_rate": 6.999539505052477e-05, "loss": 0.01, "step": 40060 }, { "grad_norm": 0.19025300443172455, "learning_rate": 6.998023899174418e-05, "loss": 0.0092, "step": 40070 }, { "grad_norm": 0.16210472583770752, "learning_rate": 6.996508074795467e-05, "loss": 0.0091, "step": 40080 }, { "grad_norm": 0.13706475496292114, "learning_rate": 6.994992032081396e-05, "loss": 0.0105, "step": 40090 }, { "grad_norm": 0.16124415397644043, "learning_rate": 6.993475771197995e-05, "loss": 0.0116, "step": 40100 }, { "grad_norm": 0.14201821386814117, "learning_rate": 6.991959292311082e-05, "loss": 0.0092, "step": 40110 }, { "grad_norm": 0.10674703121185303, "learning_rate": 6.990442595586495e-05, "loss": 0.01, "step": 40120 }, { "grad_norm": 0.1350528597831726, "learning_rate": 6.988925681190098e-05, "loss": 0.0131, "step": 40130 }, { "grad_norm": 0.14008742570877075, "learning_rate": 6.987408549287778e-05, "loss": 0.0118, "step": 40140 }, { "grad_norm": 0.10440048575401306, "learning_rate": 6.985891200045449e-05, "loss": 0.0084, "step": 40150 }, { "grad_norm": 0.16832862794399261, "learning_rate": 6.984373633629045e-05, "loss": 0.0087, "step": 40160 }, { "grad_norm": 0.15820655226707458, "learning_rate": 6.98285585020452e-05, "loss": 0.0111, "step": 40170 }, { "grad_norm": 0.139517679810524, "learning_rate": 6.981337849937864e-05, "loss": 0.01, "step": 40180 }, { "grad_norm": 0.19840964674949646, "learning_rate": 6.979819632995078e-05, "loss": 0.0097, "step": 40190 }, { "grad_norm": 0.15752436220645905, "learning_rate": 6.978301199542193e-05, "loss": 0.0103, "step": 40200 }, { "grad_norm": 0.11758612841367722, "learning_rate": 6.976782549745263e-05, "loss": 0.0083, "step": 40210 }, { "grad_norm": 0.1203770712018013, "learning_rate": 6.975263683770367e-05, "loss": 0.0087, "step": 40220 }, { "grad_norm": 0.21542228758335114, "learning_rate": 6.973744601783604e-05, "loss": 0.01, "step": 40230 }, { "grad_norm": 0.13602851331233978, "learning_rate": 6.972225303951097e-05, "loss": 0.0088, "step": 40240 }, { "grad_norm": 0.13068248331546783, "learning_rate": 6.970705790438998e-05, "loss": 0.0075, "step": 40250 }, { "grad_norm": 0.1520204097032547, "learning_rate": 6.969186061413477e-05, "loss": 0.011, "step": 40260 }, { "grad_norm": 0.13069643080234528, "learning_rate": 6.967666117040727e-05, "loss": 0.008, "step": 40270 }, { "grad_norm": 0.12191604822874069, "learning_rate": 6.966145957486972e-05, "loss": 0.0109, "step": 40280 }, { "grad_norm": 0.1816605180501938, "learning_rate": 6.964625582918449e-05, "loss": 0.0081, "step": 40290 }, { "grad_norm": 0.18616853654384613, "learning_rate": 6.963104993501425e-05, "loss": 0.0073, "step": 40300 }, { "grad_norm": 0.11120900511741638, "learning_rate": 6.961584189402192e-05, "loss": 0.0105, "step": 40310 }, { "grad_norm": 0.15692375600337982, "learning_rate": 6.96006317078706e-05, "loss": 0.0111, "step": 40320 }, { "grad_norm": 0.11793197691440582, "learning_rate": 6.958541937822367e-05, "loss": 0.0129, "step": 40330 }, { "grad_norm": 0.19096876680850983, "learning_rate": 6.957020490674473e-05, "loss": 0.0088, "step": 40340 }, { "grad_norm": 0.1317945271730423, "learning_rate": 6.95549882950976e-05, "loss": 0.0083, "step": 40350 }, { "grad_norm": 0.13571524620056152, "learning_rate": 6.953976954494635e-05, "loss": 0.0138, "step": 40360 }, { "grad_norm": 0.1419856995344162, "learning_rate": 6.952454865795528e-05, "loss": 0.0091, "step": 40370 }, { "grad_norm": 0.1408054530620575, "learning_rate": 6.950932563578892e-05, "loss": 0.0089, "step": 40380 }, { "grad_norm": 0.17495699226856232, "learning_rate": 6.949410048011206e-05, "loss": 0.0083, "step": 40390 }, { "grad_norm": 0.20758667588233948, "learning_rate": 6.947887319258966e-05, "loss": 0.0122, "step": 40400 }, { "grad_norm": 0.1888323575258255, "learning_rate": 6.946364377488696e-05, "loss": 0.0094, "step": 40410 }, { "grad_norm": 0.13406850397586823, "learning_rate": 6.944841222866947e-05, "loss": 0.0085, "step": 40420 }, { "grad_norm": 0.18618588149547577, "learning_rate": 6.943317855560284e-05, "loss": 0.0102, "step": 40430 }, { "grad_norm": 0.11431611329317093, "learning_rate": 6.941794275735302e-05, "loss": 0.0114, "step": 40440 }, { "grad_norm": 0.13382424414157867, "learning_rate": 6.94027048355862e-05, "loss": 0.009, "step": 40450 }, { "grad_norm": 0.10172290354967117, "learning_rate": 6.938746479196871e-05, "loss": 0.0125, "step": 40460 }, { "grad_norm": 0.11159460246562958, "learning_rate": 6.937222262816724e-05, "loss": 0.0081, "step": 40470 }, { "grad_norm": 0.12927551567554474, "learning_rate": 6.935697834584865e-05, "loss": 0.0093, "step": 40480 }, { "grad_norm": 0.12683354318141937, "learning_rate": 6.934173194667998e-05, "loss": 0.008, "step": 40490 }, { "grad_norm": 0.15893521904945374, "learning_rate": 6.93264834323286e-05, "loss": 0.0088, "step": 40500 }, { "grad_norm": 0.20505036413669586, "learning_rate": 6.931123280446206e-05, "loss": 0.0095, "step": 40510 }, { "grad_norm": 0.13172243535518646, "learning_rate": 6.929598006474811e-05, "loss": 0.0091, "step": 40520 }, { "grad_norm": 0.15208576619625092, "learning_rate": 6.92807252148548e-05, "loss": 0.0109, "step": 40530 }, { "grad_norm": 0.16199345886707306, "learning_rate": 6.926546825645037e-05, "loss": 0.0132, "step": 40540 }, { "grad_norm": 0.13558238744735718, "learning_rate": 6.92502091912033e-05, "loss": 0.0109, "step": 40550 }, { "grad_norm": 0.13854120671749115, "learning_rate": 6.923494802078229e-05, "loss": 0.0104, "step": 40560 }, { "grad_norm": 0.10061612725257874, "learning_rate": 6.921968474685628e-05, "loss": 0.0072, "step": 40570 }, { "grad_norm": 0.13385526835918427, "learning_rate": 6.920441937109445e-05, "loss": 0.0076, "step": 40580 }, { "grad_norm": 0.11970826983451843, "learning_rate": 6.91891518951662e-05, "loss": 0.0071, "step": 40590 }, { "grad_norm": 0.1388300210237503, "learning_rate": 6.917388232074114e-05, "loss": 0.0093, "step": 40600 }, { "grad_norm": 0.16630569100379944, "learning_rate": 6.915861064948914e-05, "loss": 0.009, "step": 40610 }, { "grad_norm": 0.143556147813797, "learning_rate": 6.914333688308025e-05, "loss": 0.007, "step": 40620 }, { "grad_norm": 0.13639390468597412, "learning_rate": 6.912806102318487e-05, "loss": 0.008, "step": 40630 }, { "grad_norm": 0.14720068871974945, "learning_rate": 6.911278307147347e-05, "loss": 0.0074, "step": 40640 }, { "grad_norm": 0.10323093831539154, "learning_rate": 6.909750302961684e-05, "loss": 0.008, "step": 40650 }, { "grad_norm": 0.10724202543497086, "learning_rate": 6.9082220899286e-05, "loss": 0.0073, "step": 40660 }, { "grad_norm": 0.15161578357219696, "learning_rate": 6.906693668215219e-05, "loss": 0.0062, "step": 40670 }, { "grad_norm": 0.14742755889892578, "learning_rate": 6.905165037988683e-05, "loss": 0.0074, "step": 40680 }, { "grad_norm": 0.15120695531368256, "learning_rate": 6.903636199416164e-05, "loss": 0.0111, "step": 40690 }, { "grad_norm": 0.13642072677612305, "learning_rate": 6.902107152664851e-05, "loss": 0.007, "step": 40700 }, { "grad_norm": 0.162401482462883, "learning_rate": 6.90057789790196e-05, "loss": 0.0093, "step": 40710 }, { "grad_norm": 0.1616695672273636, "learning_rate": 6.899048435294728e-05, "loss": 0.0114, "step": 40720 }, { "grad_norm": 0.12697818875312805, "learning_rate": 6.897518765010415e-05, "loss": 0.0076, "step": 40730 }, { "grad_norm": 0.15900251269340515, "learning_rate": 6.895988887216302e-05, "loss": 0.0117, "step": 40740 }, { "grad_norm": 0.14424987137317657, "learning_rate": 6.894458802079694e-05, "loss": 0.0086, "step": 40750 }, { "grad_norm": 0.11632620543241501, "learning_rate": 6.892928509767922e-05, "loss": 0.01, "step": 40760 }, { "grad_norm": 0.12692897021770477, "learning_rate": 6.891398010448333e-05, "loss": 0.0056, "step": 40770 }, { "grad_norm": 0.17229007184505463, "learning_rate": 6.8898673042883e-05, "loss": 0.0094, "step": 40780 }, { "grad_norm": 0.1828107237815857, "learning_rate": 6.888336391455222e-05, "loss": 0.0096, "step": 40790 }, { "grad_norm": 0.16594593226909637, "learning_rate": 6.886805272116513e-05, "loss": 0.0097, "step": 40800 }, { "grad_norm": 0.14985263347625732, "learning_rate": 6.885273946439617e-05, "loss": 0.0104, "step": 40810 }, { "grad_norm": 0.11721502244472504, "learning_rate": 6.883742414591998e-05, "loss": 0.0086, "step": 40820 }, { "grad_norm": 0.14971095323562622, "learning_rate": 6.88221067674114e-05, "loss": 0.0088, "step": 40830 }, { "grad_norm": 0.14248274266719818, "learning_rate": 6.88067873305455e-05, "loss": 0.0086, "step": 40840 }, { "grad_norm": 0.17192243039608002, "learning_rate": 6.879146583699765e-05, "loss": 0.0077, "step": 40850 }, { "grad_norm": 0.13405124843120575, "learning_rate": 6.877614228844334e-05, "loss": 0.0072, "step": 40860 }, { "grad_norm": 0.20555835962295532, "learning_rate": 6.876081668655832e-05, "loss": 0.008, "step": 40870 }, { "grad_norm": 0.14264439046382904, "learning_rate": 6.874548903301861e-05, "loss": 0.0095, "step": 40880 }, { "grad_norm": 0.12762391567230225, "learning_rate": 6.873015932950039e-05, "loss": 0.0094, "step": 40890 }, { "grad_norm": 0.13779513537883759, "learning_rate": 6.871482757768012e-05, "loss": 0.0104, "step": 40900 }, { "grad_norm": 0.09600787609815598, "learning_rate": 6.869949377923442e-05, "loss": 0.0105, "step": 40910 }, { "grad_norm": 0.12518273293972015, "learning_rate": 6.868415793584022e-05, "loss": 0.0077, "step": 40920 }, { "grad_norm": 0.10886268317699432, "learning_rate": 6.86688200491746e-05, "loss": 0.0087, "step": 40930 }, { "grad_norm": 0.11862324178218842, "learning_rate": 6.865348012091488e-05, "loss": 0.0094, "step": 40940 }, { "grad_norm": 0.13196654617786407, "learning_rate": 6.86381381527386e-05, "loss": 0.0105, "step": 40950 }, { "grad_norm": 0.16348373889923096, "learning_rate": 6.862279414632359e-05, "loss": 0.0089, "step": 40960 }, { "grad_norm": 0.16792593896389008, "learning_rate": 6.860744810334778e-05, "loss": 0.0119, "step": 40970 }, { "grad_norm": 0.13558189570903778, "learning_rate": 6.859210002548943e-05, "loss": 0.008, "step": 40980 }, { "grad_norm": 0.12450007349252701, "learning_rate": 6.8576749914427e-05, "loss": 0.0111, "step": 40990 }, { "grad_norm": 0.14867526292800903, "learning_rate": 6.85613977718391e-05, "loss": 0.0092, "step": 41000 }, { "grad_norm": 0.15177291631698608, "learning_rate": 6.854604359940464e-05, "loss": 0.011, "step": 41010 }, { "grad_norm": 0.1816917508840561, "learning_rate": 6.853068739880276e-05, "loss": 0.0083, "step": 41020 }, { "grad_norm": 0.11701415479183197, "learning_rate": 6.851532917171276e-05, "loss": 0.0089, "step": 41030 }, { "grad_norm": 0.16538679599761963, "learning_rate": 6.849996891981418e-05, "loss": 0.0076, "step": 41040 }, { "grad_norm": 0.16859127581119537, "learning_rate": 6.848460664478684e-05, "loss": 0.0087, "step": 41050 }, { "grad_norm": 0.16824275255203247, "learning_rate": 6.846924234831069e-05, "loss": 0.0085, "step": 41060 }, { "grad_norm": 0.12508484721183777, "learning_rate": 6.845387603206597e-05, "loss": 0.0085, "step": 41070 }, { "grad_norm": 0.12810097634792328, "learning_rate": 6.843850769773312e-05, "loss": 0.009, "step": 41080 }, { "grad_norm": 0.12803740799427032, "learning_rate": 6.842313734699278e-05, "loss": 0.0095, "step": 41090 }, { "grad_norm": 0.13319525122642517, "learning_rate": 6.840776498152584e-05, "loss": 0.0093, "step": 41100 }, { "grad_norm": 0.17052431404590607, "learning_rate": 6.83923906030134e-05, "loss": 0.0088, "step": 41110 }, { "grad_norm": 0.13585151731967926, "learning_rate": 6.837701421313677e-05, "loss": 0.0079, "step": 41120 }, { "grad_norm": 0.12295249849557877, "learning_rate": 6.83616358135775e-05, "loss": 0.0075, "step": 41130 }, { "grad_norm": 0.15752163529396057, "learning_rate": 6.834625540601734e-05, "loss": 0.0088, "step": 41140 }, { "grad_norm": 0.14568908512592316, "learning_rate": 6.833087299213829e-05, "loss": 0.0081, "step": 41150 }, { "grad_norm": 0.10779447853565216, "learning_rate": 6.83154885736225e-05, "loss": 0.0073, "step": 41160 }, { "grad_norm": 0.17596685886383057, "learning_rate": 6.830010215215244e-05, "loss": 0.0087, "step": 41170 }, { "grad_norm": 0.15873514115810394, "learning_rate": 6.828471372941072e-05, "loss": 0.0088, "step": 41180 }, { "grad_norm": 0.17752474546432495, "learning_rate": 6.826932330708018e-05, "loss": 0.0114, "step": 41190 }, { "grad_norm": 0.1995047926902771, "learning_rate": 6.825393088684393e-05, "loss": 0.0132, "step": 41200 }, { "grad_norm": 0.1625913381576538, "learning_rate": 6.823853647038524e-05, "loss": 0.0084, "step": 41210 }, { "grad_norm": 0.15401889383792877, "learning_rate": 6.822314005938762e-05, "loss": 0.0082, "step": 41220 }, { "grad_norm": 0.13922826945781708, "learning_rate": 6.820774165553479e-05, "loss": 0.0067, "step": 41230 }, { "grad_norm": 0.10325939953327179, "learning_rate": 6.819234126051072e-05, "loss": 0.0083, "step": 41240 }, { "grad_norm": 0.14192046225070953, "learning_rate": 6.817693887599956e-05, "loss": 0.011, "step": 41250 }, { "grad_norm": 0.17007657885551453, "learning_rate": 6.816153450368568e-05, "loss": 0.0089, "step": 41260 }, { "grad_norm": 0.18697014451026917, "learning_rate": 6.814612814525372e-05, "loss": 0.0112, "step": 41270 }, { "grad_norm": 0.15920600295066833, "learning_rate": 6.813071980238846e-05, "loss": 0.011, "step": 41280 }, { "grad_norm": 0.08003930747509003, "learning_rate": 6.811530947677492e-05, "loss": 0.0059, "step": 41290 }, { "grad_norm": 0.18588288128376007, "learning_rate": 6.809989717009839e-05, "loss": 0.0081, "step": 41300 }, { "grad_norm": 0.17299281060695648, "learning_rate": 6.808448288404431e-05, "loss": 0.0081, "step": 41310 }, { "grad_norm": 0.15651193261146545, "learning_rate": 6.806906662029838e-05, "loss": 0.0094, "step": 41320 }, { "grad_norm": 0.10533513128757477, "learning_rate": 6.805364838054647e-05, "loss": 0.0094, "step": 41330 }, { "grad_norm": 0.18430384993553162, "learning_rate": 6.803822816647474e-05, "loss": 0.0089, "step": 41340 }, { "grad_norm": 0.12861725687980652, "learning_rate": 6.802280597976949e-05, "loss": 0.0096, "step": 41350 }, { "grad_norm": 0.13432949781417847, "learning_rate": 6.800738182211727e-05, "loss": 0.0086, "step": 41360 }, { "grad_norm": 0.13804911077022552, "learning_rate": 6.799195569520485e-05, "loss": 0.0077, "step": 41370 }, { "grad_norm": 0.1366790533065796, "learning_rate": 6.79765276007192e-05, "loss": 0.0088, "step": 41380 }, { "grad_norm": 0.1294831931591034, "learning_rate": 6.796109754034753e-05, "loss": 0.0121, "step": 41390 }, { "grad_norm": 0.11029079556465149, "learning_rate": 6.794566551577724e-05, "loss": 0.0073, "step": 41400 }, { "grad_norm": 0.15710116922855377, "learning_rate": 6.793023152869594e-05, "loss": 0.0097, "step": 41410 }, { "grad_norm": 0.1352638602256775, "learning_rate": 6.791479558079147e-05, "loss": 0.0094, "step": 41420 }, { "grad_norm": 0.117304727435112, "learning_rate": 6.789935767375191e-05, "loss": 0.0082, "step": 41430 }, { "grad_norm": 0.127314493060112, "learning_rate": 6.78839178092655e-05, "loss": 0.0098, "step": 41440 }, { "grad_norm": 0.1590372920036316, "learning_rate": 6.786847598902072e-05, "loss": 0.0082, "step": 41450 }, { "grad_norm": 0.11422865837812424, "learning_rate": 6.785303221470627e-05, "loss": 0.0078, "step": 41460 }, { "grad_norm": 0.13724836707115173, "learning_rate": 6.783758648801108e-05, "loss": 0.0094, "step": 41470 }, { "grad_norm": 0.14800243079662323, "learning_rate": 6.782213881062422e-05, "loss": 0.0095, "step": 41480 }, { "grad_norm": 0.18907910585403442, "learning_rate": 6.780668918423508e-05, "loss": 0.0096, "step": 41490 }, { "grad_norm": 0.15454313158988953, "learning_rate": 6.779123761053317e-05, "loss": 0.0093, "step": 41500 }, { "grad_norm": 0.14049634337425232, "learning_rate": 6.777578409120826e-05, "loss": 0.0091, "step": 41510 }, { "grad_norm": 0.14048932492733002, "learning_rate": 6.776032862795033e-05, "loss": 0.0083, "step": 41520 }, { "grad_norm": 0.1356852650642395, "learning_rate": 6.774487122244956e-05, "loss": 0.0077, "step": 41530 }, { "grad_norm": 0.13148309290409088, "learning_rate": 6.772941187639637e-05, "loss": 0.0082, "step": 41540 }, { "grad_norm": 0.14491935074329376, "learning_rate": 6.771395059148134e-05, "loss": 0.0093, "step": 41550 }, { "grad_norm": 0.168482705950737, "learning_rate": 6.76984873693953e-05, "loss": 0.0079, "step": 41560 }, { "grad_norm": 0.1176966205239296, "learning_rate": 6.768302221182931e-05, "loss": 0.0104, "step": 41570 }, { "grad_norm": 0.1594305783510208, "learning_rate": 6.766755512047457e-05, "loss": 0.0096, "step": 41580 }, { "grad_norm": 0.15919333696365356, "learning_rate": 6.765208609702259e-05, "loss": 0.0126, "step": 41590 }, { "grad_norm": 0.14265567064285278, "learning_rate": 6.763661514316499e-05, "loss": 0.0091, "step": 41600 }, { "grad_norm": 0.15219171345233917, "learning_rate": 6.76211422605937e-05, "loss": 0.0082, "step": 41610 }, { "grad_norm": 0.16644513607025146, "learning_rate": 6.760566745100076e-05, "loss": 0.0114, "step": 41620 }, { "grad_norm": 0.16529962420463562, "learning_rate": 6.75901907160785e-05, "loss": 0.0077, "step": 41630 }, { "grad_norm": 0.158138707280159, "learning_rate": 6.757471205751943e-05, "loss": 0.0077, "step": 41640 }, { "grad_norm": 0.192947655916214, "learning_rate": 6.75592314770163e-05, "loss": 0.008, "step": 41650 }, { "grad_norm": 0.13595150411128998, "learning_rate": 6.7543748976262e-05, "loss": 0.0075, "step": 41660 }, { "grad_norm": 0.14154426753520966, "learning_rate": 6.752826455694968e-05, "loss": 0.011, "step": 41670 }, { "grad_norm": 0.12283337861299515, "learning_rate": 6.751277822077271e-05, "loss": 0.0081, "step": 41680 }, { "grad_norm": 0.16898952424526215, "learning_rate": 6.749728996942465e-05, "loss": 0.0096, "step": 41690 }, { "grad_norm": 0.1350351721048355, "learning_rate": 6.748179980459924e-05, "loss": 0.0089, "step": 41700 }, { "grad_norm": 0.18823373317718506, "learning_rate": 6.746630772799052e-05, "loss": 0.0093, "step": 41710 }, { "grad_norm": 0.16310754418373108, "learning_rate": 6.745081374129262e-05, "loss": 0.0095, "step": 41720 }, { "grad_norm": 0.10783931612968445, "learning_rate": 6.743531784619998e-05, "loss": 0.0085, "step": 41730 }, { "grad_norm": 0.19073157012462616, "learning_rate": 6.741982004440719e-05, "loss": 0.0101, "step": 41740 }, { "grad_norm": 0.11626777052879333, "learning_rate": 6.740432033760907e-05, "loss": 0.0081, "step": 41750 }, { "grad_norm": 0.1221085861325264, "learning_rate": 6.738881872750066e-05, "loss": 0.0078, "step": 41760 }, { "grad_norm": 0.1302037090063095, "learning_rate": 6.737331521577715e-05, "loss": 0.0123, "step": 41770 }, { "grad_norm": 0.12562739849090576, "learning_rate": 6.735780980413403e-05, "loss": 0.0095, "step": 41780 }, { "grad_norm": 0.17356827855110168, "learning_rate": 6.734230249426692e-05, "loss": 0.0097, "step": 41790 }, { "grad_norm": 0.14938868582248688, "learning_rate": 6.732679328787168e-05, "loss": 0.009, "step": 41800 }, { "grad_norm": 0.1765953153371811, "learning_rate": 6.731128218664438e-05, "loss": 0.0087, "step": 41810 }, { "grad_norm": 0.18301166594028473, "learning_rate": 6.729576919228129e-05, "loss": 0.0094, "step": 41820 }, { "grad_norm": 0.18138477206230164, "learning_rate": 6.728025430647888e-05, "loss": 0.0108, "step": 41830 }, { "grad_norm": 0.10386641323566437, "learning_rate": 6.726473753093383e-05, "loss": 0.0074, "step": 41840 }, { "grad_norm": 0.1365969032049179, "learning_rate": 6.724921886734305e-05, "loss": 0.0114, "step": 41850 }, { "grad_norm": 0.18086902797222137, "learning_rate": 6.723369831740363e-05, "loss": 0.0094, "step": 41860 }, { "grad_norm": 0.11178980767726898, "learning_rate": 6.721817588281288e-05, "loss": 0.0077, "step": 41870 }, { "grad_norm": 0.11985199898481369, "learning_rate": 6.720265156526828e-05, "loss": 0.0121, "step": 41880 }, { "grad_norm": 0.13678793609142303, "learning_rate": 6.718712536646758e-05, "loss": 0.008, "step": 41890 }, { "grad_norm": 0.1481180340051651, "learning_rate": 6.71715972881087e-05, "loss": 0.0079, "step": 41900 }, { "grad_norm": 0.1534864604473114, "learning_rate": 6.715606733188974e-05, "loss": 0.0104, "step": 41910 }, { "grad_norm": 0.13958941400051117, "learning_rate": 6.714053549950909e-05, "loss": 0.0081, "step": 41920 }, { "grad_norm": 0.11397716403007507, "learning_rate": 6.71250017926652e-05, "loss": 0.0088, "step": 41930 }, { "grad_norm": 0.1192726120352745, "learning_rate": 6.71094662130569e-05, "loss": 0.0069, "step": 41940 }, { "grad_norm": 0.1893436461687088, "learning_rate": 6.709392876238307e-05, "loss": 0.0074, "step": 41950 }, { "grad_norm": 0.22119170427322388, "learning_rate": 6.707838944234289e-05, "loss": 0.0084, "step": 41960 }, { "grad_norm": 0.17126178741455078, "learning_rate": 6.706284825463574e-05, "loss": 0.0091, "step": 41970 }, { "grad_norm": 0.12957236170768738, "learning_rate": 6.704730520096114e-05, "loss": 0.0092, "step": 41980 }, { "grad_norm": 0.1729816049337387, "learning_rate": 6.703176028301888e-05, "loss": 0.0091, "step": 41990 }, { "grad_norm": 0.09819487482309341, "learning_rate": 6.701621350250892e-05, "loss": 0.0069, "step": 42000 }, { "grad_norm": 0.15048567950725555, "learning_rate": 6.700066486113144e-05, "loss": 0.0093, "step": 42010 }, { "grad_norm": 0.14755043387413025, "learning_rate": 6.69851143605868e-05, "loss": 0.0079, "step": 42020 }, { "grad_norm": 0.13916607201099396, "learning_rate": 6.696956200257561e-05, "loss": 0.0098, "step": 42030 }, { "grad_norm": 0.16234229505062103, "learning_rate": 6.695400778879863e-05, "loss": 0.0087, "step": 42040 }, { "grad_norm": 0.15736135840415955, "learning_rate": 6.693845172095683e-05, "loss": 0.0078, "step": 42050 }, { "grad_norm": 0.08786483854055405, "learning_rate": 6.692289380075142e-05, "loss": 0.0107, "step": 42060 }, { "grad_norm": 0.14284250140190125, "learning_rate": 6.690733402988379e-05, "loss": 0.0092, "step": 42070 }, { "grad_norm": 0.11945761740207672, "learning_rate": 6.689177241005553e-05, "loss": 0.0085, "step": 42080 }, { "grad_norm": 0.23437970876693726, "learning_rate": 6.687620894296844e-05, "loss": 0.0085, "step": 42090 }, { "grad_norm": 0.1525203287601471, "learning_rate": 6.686064363032451e-05, "loss": 0.0098, "step": 42100 }, { "grad_norm": 0.146973118185997, "learning_rate": 6.684507647382596e-05, "loss": 0.0082, "step": 42110 }, { "grad_norm": 0.16107143461704254, "learning_rate": 6.682950747517513e-05, "loss": 0.0077, "step": 42120 }, { "grad_norm": 0.1734536737203598, "learning_rate": 6.68139366360747e-05, "loss": 0.0082, "step": 42130 }, { "grad_norm": 0.1071779802441597, "learning_rate": 6.679836395822744e-05, "loss": 0.0071, "step": 42140 }, { "grad_norm": 0.14377108216285706, "learning_rate": 6.678278944333633e-05, "loss": 0.0086, "step": 42150 }, { "grad_norm": 0.18256911635398865, "learning_rate": 6.676721309310462e-05, "loss": 0.0097, "step": 42160 }, { "grad_norm": 0.19895526766777039, "learning_rate": 6.67516349092357e-05, "loss": 0.0092, "step": 42170 }, { "grad_norm": 0.13954447209835052, "learning_rate": 6.673605489343315e-05, "loss": 0.0092, "step": 42180 }, { "grad_norm": 0.12809070944786072, "learning_rate": 6.672047304740081e-05, "loss": 0.0082, "step": 42190 }, { "grad_norm": 0.13379837572574615, "learning_rate": 6.670488937284268e-05, "loss": 0.0088, "step": 42200 }, { "grad_norm": 0.14226755499839783, "learning_rate": 6.668930387146294e-05, "loss": 0.0074, "step": 42210 }, { "grad_norm": 0.155796617269516, "learning_rate": 6.667371654496605e-05, "loss": 0.0075, "step": 42220 }, { "grad_norm": 0.1504117101430893, "learning_rate": 6.665812739505659e-05, "loss": 0.0092, "step": 42230 }, { "grad_norm": 0.1464841663837433, "learning_rate": 6.664253642343935e-05, "loss": 0.0091, "step": 42240 }, { "grad_norm": 0.15445758402347565, "learning_rate": 6.662694363181935e-05, "loss": 0.0085, "step": 42250 }, { "grad_norm": 0.1252889782190323, "learning_rate": 6.66113490219018e-05, "loss": 0.0085, "step": 42260 }, { "grad_norm": 0.14228323101997375, "learning_rate": 6.659575259539207e-05, "loss": 0.0095, "step": 42270 }, { "grad_norm": 0.10243680328130722, "learning_rate": 6.65801543539958e-05, "loss": 0.0076, "step": 42280 }, { "grad_norm": 0.09900281578302383, "learning_rate": 6.65645542994188e-05, "loss": 0.0077, "step": 42290 }, { "grad_norm": 0.11286811530590057, "learning_rate": 6.654895243336702e-05, "loss": 0.0092, "step": 42300 }, { "grad_norm": 0.1245792806148529, "learning_rate": 6.65333487575467e-05, "loss": 0.0079, "step": 42310 }, { "grad_norm": 0.13903024792671204, "learning_rate": 6.651774327366422e-05, "loss": 0.0113, "step": 42320 }, { "grad_norm": 0.14788351953029633, "learning_rate": 6.650213598342613e-05, "loss": 0.0099, "step": 42330 }, { "grad_norm": 0.18313734233379364, "learning_rate": 6.648652688853931e-05, "loss": 0.0073, "step": 42340 }, { "grad_norm": 0.12887808680534363, "learning_rate": 6.647091599071066e-05, "loss": 0.0091, "step": 42350 }, { "grad_norm": 0.16513730585575104, "learning_rate": 6.645530329164743e-05, "loss": 0.0086, "step": 42360 }, { "grad_norm": 0.16047324240207672, "learning_rate": 6.643968879305694e-05, "loss": 0.0119, "step": 42370 }, { "grad_norm": 0.18107834458351135, "learning_rate": 6.642407249664683e-05, "loss": 0.0116, "step": 42380 }, { "grad_norm": 0.12686973810195923, "learning_rate": 6.640845440412483e-05, "loss": 0.0113, "step": 42390 }, { "grad_norm": 0.11912813782691956, "learning_rate": 6.639283451719893e-05, "loss": 0.0071, "step": 42400 }, { "grad_norm": 0.13285501301288605, "learning_rate": 6.637721283757729e-05, "loss": 0.0085, "step": 42410 }, { "grad_norm": 0.14018666744232178, "learning_rate": 6.636158936696828e-05, "loss": 0.0102, "step": 42420 }, { "grad_norm": 0.14379358291625977, "learning_rate": 6.634596410708047e-05, "loss": 0.0115, "step": 42430 }, { "grad_norm": 0.11993358284235, "learning_rate": 6.633033705962257e-05, "loss": 0.0076, "step": 42440 }, { "grad_norm": 0.13388565182685852, "learning_rate": 6.631470822630359e-05, "loss": 0.0091, "step": 42450 }, { "grad_norm": 0.18851879239082336, "learning_rate": 6.629907760883263e-05, "loss": 0.0097, "step": 42460 }, { "grad_norm": 0.1874988079071045, "learning_rate": 6.628344520891905e-05, "loss": 0.0114, "step": 42470 }, { "grad_norm": 0.12106040865182877, "learning_rate": 6.62678110282724e-05, "loss": 0.0103, "step": 42480 }, { "grad_norm": 0.16280381381511688, "learning_rate": 6.625217506860237e-05, "loss": 0.0087, "step": 42490 }, { "grad_norm": 0.17653828859329224, "learning_rate": 6.623653733161892e-05, "loss": 0.01, "step": 42500 }, { "grad_norm": 0.18063043057918549, "learning_rate": 6.622089781903217e-05, "loss": 0.0093, "step": 42510 }, { "grad_norm": 0.1098223403096199, "learning_rate": 6.620525653255241e-05, "loss": 0.0117, "step": 42520 }, { "grad_norm": 0.2200961858034134, "learning_rate": 6.618961347389015e-05, "loss": 0.0104, "step": 42530 }, { "grad_norm": 0.16207082569599152, "learning_rate": 6.617396864475613e-05, "loss": 0.0095, "step": 42540 }, { "grad_norm": 0.18690700829029083, "learning_rate": 6.61583220468612e-05, "loss": 0.0091, "step": 42550 }, { "grad_norm": 0.13470354676246643, "learning_rate": 6.614267368191645e-05, "loss": 0.0072, "step": 42560 }, { "grad_norm": 0.1294795125722885, "learning_rate": 6.61270235516332e-05, "loss": 0.0076, "step": 42570 }, { "grad_norm": 0.11935687065124512, "learning_rate": 6.611137165772287e-05, "loss": 0.0096, "step": 42580 }, { "grad_norm": 0.14698797464370728, "learning_rate": 6.60957180018972e-05, "loss": 0.0081, "step": 42590 }, { "grad_norm": 0.16762742400169373, "learning_rate": 6.608006258586797e-05, "loss": 0.014, "step": 42600 }, { "grad_norm": 0.13671982288360596, "learning_rate": 6.606440541134731e-05, "loss": 0.0076, "step": 42610 }, { "grad_norm": 0.15766394138336182, "learning_rate": 6.60487464800474e-05, "loss": 0.0091, "step": 42620 }, { "grad_norm": 0.1403007209300995, "learning_rate": 6.603308579368071e-05, "loss": 0.0117, "step": 42630 }, { "grad_norm": 0.1788547933101654, "learning_rate": 6.601742335395987e-05, "loss": 0.0089, "step": 42640 }, { "grad_norm": 0.13945752382278442, "learning_rate": 6.600175916259769e-05, "loss": 0.0095, "step": 42650 }, { "grad_norm": 0.10458947718143463, "learning_rate": 6.598609322130718e-05, "loss": 0.0075, "step": 42660 }, { "grad_norm": 0.12287236750125885, "learning_rate": 6.597042553180155e-05, "loss": 0.0086, "step": 42670 }, { "grad_norm": 0.14527349174022675, "learning_rate": 6.59547560957942e-05, "loss": 0.0091, "step": 42680 }, { "grad_norm": 0.13194318115711212, "learning_rate": 6.59390849149987e-05, "loss": 0.0099, "step": 42690 }, { "grad_norm": 0.22407139837741852, "learning_rate": 6.592341199112886e-05, "loss": 0.0118, "step": 42700 }, { "grad_norm": 0.1538165807723999, "learning_rate": 6.590773732589861e-05, "loss": 0.0112, "step": 42710 }, { "grad_norm": 0.16535019874572754, "learning_rate": 6.589206092102213e-05, "loss": 0.0093, "step": 42720 }, { "grad_norm": 0.20790943503379822, "learning_rate": 6.587638277821375e-05, "loss": 0.0096, "step": 42730 }, { "grad_norm": 0.1675291508436203, "learning_rate": 6.586070289918803e-05, "loss": 0.0098, "step": 42740 }, { "grad_norm": 0.11390415579080582, "learning_rate": 6.584502128565968e-05, "loss": 0.0075, "step": 42750 }, { "grad_norm": 0.20594432950019836, "learning_rate": 6.582933793934363e-05, "loss": 0.0094, "step": 42760 }, { "grad_norm": 0.16301904618740082, "learning_rate": 6.581365286195498e-05, "loss": 0.0078, "step": 42770 }, { "grad_norm": 0.1139618307352066, "learning_rate": 6.579796605520903e-05, "loss": 0.0101, "step": 42780 }, { "grad_norm": 0.12635797262191772, "learning_rate": 6.578227752082127e-05, "loss": 0.0087, "step": 42790 }, { "grad_norm": 0.17661558091640472, "learning_rate": 6.576658726050735e-05, "loss": 0.011, "step": 42800 }, { "grad_norm": 0.2020920068025589, "learning_rate": 6.575089527598316e-05, "loss": 0.0072, "step": 42810 }, { "grad_norm": 0.12633192539215088, "learning_rate": 6.573520156896476e-05, "loss": 0.0118, "step": 42820 }, { "grad_norm": 0.14599667489528656, "learning_rate": 6.571950614116835e-05, "loss": 0.0085, "step": 42830 }, { "grad_norm": 0.10714384913444519, "learning_rate": 6.570380899431039e-05, "loss": 0.0101, "step": 42840 }, { "grad_norm": 0.18302956223487854, "learning_rate": 6.568811013010749e-05, "loss": 0.0091, "step": 42850 }, { "grad_norm": 0.17184311151504517, "learning_rate": 6.567240955027645e-05, "loss": 0.0095, "step": 42860 }, { "grad_norm": 0.23229461908340454, "learning_rate": 6.565670725653427e-05, "loss": 0.0079, "step": 42870 }, { "grad_norm": 0.1653539091348648, "learning_rate": 6.564100325059812e-05, "loss": 0.0106, "step": 42880 }, { "grad_norm": 0.13297316431999207, "learning_rate": 6.562529753418538e-05, "loss": 0.0078, "step": 42890 }, { "grad_norm": 0.15332072973251343, "learning_rate": 6.56095901090136e-05, "loss": 0.0099, "step": 42900 }, { "grad_norm": 0.16877520084381104, "learning_rate": 6.55938809768005e-05, "loss": 0.0083, "step": 42910 }, { "grad_norm": 0.1047544926404953, "learning_rate": 6.557817013926402e-05, "loss": 0.0083, "step": 42920 }, { "grad_norm": 0.13653188943862915, "learning_rate": 6.556245759812227e-05, "loss": 0.0093, "step": 42930 }, { "grad_norm": 0.18020521104335785, "learning_rate": 6.554674335509357e-05, "loss": 0.0077, "step": 42940 }, { "grad_norm": 0.14852182567119598, "learning_rate": 6.553102741189638e-05, "loss": 0.008, "step": 42950 }, { "grad_norm": 0.14902783930301666, "learning_rate": 6.55153097702494e-05, "loss": 0.0092, "step": 42960 }, { "grad_norm": 0.1668926477432251, "learning_rate": 6.549959043187146e-05, "loss": 0.0087, "step": 42970 }, { "grad_norm": 0.1559496819972992, "learning_rate": 6.54838693984816e-05, "loss": 0.0105, "step": 42980 }, { "grad_norm": 0.11060509830713272, "learning_rate": 6.546814667179909e-05, "loss": 0.0072, "step": 42990 }, { "grad_norm": 0.21604740619659424, "learning_rate": 6.545242225354328e-05, "loss": 0.0095, "step": 43000 }, { "grad_norm": 0.1781473606824875, "learning_rate": 6.543669614543383e-05, "loss": 0.0098, "step": 43010 }, { "grad_norm": 0.20766295492649078, "learning_rate": 6.542096834919049e-05, "loss": 0.0088, "step": 43020 }, { "grad_norm": 0.1380918025970459, "learning_rate": 6.540523886653324e-05, "loss": 0.0077, "step": 43030 }, { "grad_norm": 0.15486983954906464, "learning_rate": 6.538950769918222e-05, "loss": 0.0088, "step": 43040 }, { "grad_norm": 0.1462591141462326, "learning_rate": 6.537377484885779e-05, "loss": 0.0095, "step": 43050 }, { "grad_norm": 0.1832704395055771, "learning_rate": 6.535804031728045e-05, "loss": 0.0086, "step": 43060 }, { "grad_norm": 0.1513615995645523, "learning_rate": 6.534230410617092e-05, "loss": 0.0122, "step": 43070 }, { "grad_norm": 0.1466236412525177, "learning_rate": 6.532656621725007e-05, "loss": 0.0078, "step": 43080 }, { "grad_norm": 0.16813601553440094, "learning_rate": 6.531082665223898e-05, "loss": 0.0091, "step": 43090 }, { "grad_norm": 0.12033572047948837, "learning_rate": 6.529508541285889e-05, "loss": 0.0074, "step": 43100 }, { "grad_norm": 0.15197376906871796, "learning_rate": 6.527934250083128e-05, "loss": 0.009, "step": 43110 }, { "grad_norm": 0.19567440450191498, "learning_rate": 6.526359791787775e-05, "loss": 0.0107, "step": 43120 }, { "grad_norm": 0.13540509343147278, "learning_rate": 6.524785166572009e-05, "loss": 0.0102, "step": 43130 }, { "grad_norm": 0.1141989603638649, "learning_rate": 6.523210374608029e-05, "loss": 0.007, "step": 43140 }, { "grad_norm": 0.16398470103740692, "learning_rate": 6.521635416068054e-05, "loss": 0.0099, "step": 43150 }, { "grad_norm": 0.15727144479751587, "learning_rate": 6.520060291124317e-05, "loss": 0.0074, "step": 43160 }, { "grad_norm": 0.1847059428691864, "learning_rate": 6.518484999949072e-05, "loss": 0.011, "step": 43170 }, { "grad_norm": 0.1494075059890747, "learning_rate": 6.516909542714591e-05, "loss": 0.0109, "step": 43180 }, { "grad_norm": 0.14232146739959717, "learning_rate": 6.515333919593161e-05, "loss": 0.0091, "step": 43190 }, { "grad_norm": 0.2140941321849823, "learning_rate": 6.513758130757094e-05, "loss": 0.0086, "step": 43200 }, { "grad_norm": 0.13176369667053223, "learning_rate": 6.512182176378713e-05, "loss": 0.0091, "step": 43210 }, { "grad_norm": 0.17844681441783905, "learning_rate": 6.510606056630362e-05, "loss": 0.0117, "step": 43220 }, { "grad_norm": 0.15043607354164124, "learning_rate": 6.509029771684405e-05, "loss": 0.0127, "step": 43230 }, { "grad_norm": 0.1307327151298523, "learning_rate": 6.50745332171322e-05, "loss": 0.0102, "step": 43240 }, { "grad_norm": 0.12351435422897339, "learning_rate": 6.505876706889207e-05, "loss": 0.0096, "step": 43250 }, { "grad_norm": 0.1255827397108078, "learning_rate": 6.504299927384781e-05, "loss": 0.0096, "step": 43260 }, { "grad_norm": 0.14150969684123993, "learning_rate": 6.502722983372378e-05, "loss": 0.008, "step": 43270 }, { "grad_norm": 0.10364075750112534, "learning_rate": 6.501145875024446e-05, "loss": 0.0072, "step": 43280 }, { "grad_norm": 0.14032578468322754, "learning_rate": 6.499568602513459e-05, "loss": 0.0081, "step": 43290 }, { "grad_norm": 0.12465931475162506, "learning_rate": 6.497991166011903e-05, "loss": 0.0088, "step": 43300 }, { "grad_norm": 0.13402783870697021, "learning_rate": 6.49641356569229e-05, "loss": 0.0085, "step": 43310 }, { "grad_norm": 0.11872529983520508, "learning_rate": 6.494835801727135e-05, "loss": 0.0078, "step": 43320 }, { "grad_norm": 0.23699569702148438, "learning_rate": 6.493257874288986e-05, "loss": 0.0133, "step": 43330 }, { "grad_norm": 0.13874545693397522, "learning_rate": 6.491679783550399e-05, "loss": 0.0088, "step": 43340 }, { "grad_norm": 0.13072045147418976, "learning_rate": 6.490101529683957e-05, "loss": 0.0095, "step": 43350 }, { "grad_norm": 0.1489279419183731, "learning_rate": 6.488523112862248e-05, "loss": 0.0087, "step": 43360 }, { "grad_norm": 0.10309316962957382, "learning_rate": 6.486944533257891e-05, "loss": 0.0089, "step": 43370 }, { "grad_norm": 0.1470382660627365, "learning_rate": 6.485365791043517e-05, "loss": 0.0086, "step": 43380 }, { "grad_norm": 0.14473679661750793, "learning_rate": 6.483786886391772e-05, "loss": 0.0096, "step": 43390 }, { "grad_norm": 0.10132993757724762, "learning_rate": 6.482207819475323e-05, "loss": 0.0084, "step": 43400 }, { "grad_norm": 0.13478976488113403, "learning_rate": 6.480628590466857e-05, "loss": 0.0092, "step": 43410 }, { "grad_norm": 0.1682950109243393, "learning_rate": 6.479049199539074e-05, "loss": 0.0116, "step": 43420 }, { "grad_norm": 0.09450078755617142, "learning_rate": 6.477469646864693e-05, "loss": 0.0116, "step": 43430 }, { "grad_norm": 0.15819914638996124, "learning_rate": 6.475889932616454e-05, "loss": 0.0082, "step": 43440 }, { "grad_norm": 0.14831189811229706, "learning_rate": 6.474310056967111e-05, "loss": 0.0083, "step": 43450 }, { "grad_norm": 0.21684058010578156, "learning_rate": 6.472730020089437e-05, "loss": 0.0087, "step": 43460 }, { "grad_norm": 0.16308286786079407, "learning_rate": 6.471149822156222e-05, "loss": 0.0082, "step": 43470 }, { "grad_norm": 0.12946705520153046, "learning_rate": 6.469569463340275e-05, "loss": 0.0106, "step": 43480 }, { "grad_norm": 0.1545572131872177, "learning_rate": 6.467988943814418e-05, "loss": 0.0095, "step": 43490 }, { "grad_norm": 0.16563789546489716, "learning_rate": 6.4664082637515e-05, "loss": 0.0097, "step": 43500 }, { "grad_norm": 0.11117444187402725, "learning_rate": 6.464827423324377e-05, "loss": 0.0067, "step": 43510 }, { "grad_norm": 0.15592806041240692, "learning_rate": 6.463246422705931e-05, "loss": 0.0086, "step": 43520 }, { "grad_norm": 0.1336999088525772, "learning_rate": 6.461665262069055e-05, "loss": 0.0091, "step": 43530 }, { "grad_norm": 0.15271487832069397, "learning_rate": 6.460083941586665e-05, "loss": 0.0076, "step": 43540 }, { "grad_norm": 0.1135052815079689, "learning_rate": 6.45850246143169e-05, "loss": 0.0081, "step": 43550 }, { "grad_norm": 0.13772709667682648, "learning_rate": 6.456920821777077e-05, "loss": 0.0083, "step": 43560 }, { "grad_norm": 0.19519264996051788, "learning_rate": 6.455339022795796e-05, "loss": 0.0091, "step": 43570 }, { "grad_norm": 0.14963635802268982, "learning_rate": 6.453757064660825e-05, "loss": 0.0079, "step": 43580 }, { "grad_norm": 0.1519165188074112, "learning_rate": 6.452174947545169e-05, "loss": 0.0075, "step": 43590 }, { "grad_norm": 0.18318291008472443, "learning_rate": 6.450592671621842e-05, "loss": 0.0102, "step": 43600 }, { "grad_norm": 0.14897306263446808, "learning_rate": 6.449010237063882e-05, "loss": 0.0093, "step": 43610 }, { "grad_norm": 0.11019061505794525, "learning_rate": 6.447427644044343e-05, "loss": 0.0081, "step": 43620 }, { "grad_norm": 0.11938909441232681, "learning_rate": 6.44584489273629e-05, "loss": 0.0072, "step": 43630 }, { "grad_norm": 0.1587497740983963, "learning_rate": 6.444261983312817e-05, "loss": 0.0098, "step": 43640 }, { "grad_norm": 0.15201880037784576, "learning_rate": 6.442678915947023e-05, "loss": 0.0094, "step": 43650 }, { "grad_norm": 0.1480167657136917, "learning_rate": 6.441095690812032e-05, "loss": 0.0084, "step": 43660 }, { "grad_norm": 0.11486655473709106, "learning_rate": 6.439512308080982e-05, "loss": 0.0078, "step": 43670 }, { "grad_norm": 0.11044253408908844, "learning_rate": 6.43792876792703e-05, "loss": 0.0098, "step": 43680 }, { "grad_norm": 0.18291974067687988, "learning_rate": 6.43634507052335e-05, "loss": 0.0107, "step": 43690 }, { "grad_norm": 0.16705043613910675, "learning_rate": 6.434761216043135e-05, "loss": 0.0111, "step": 43700 }, { "grad_norm": 0.13757270574569702, "learning_rate": 6.433177204659589e-05, "loss": 0.0097, "step": 43710 }, { "grad_norm": 0.1497400403022766, "learning_rate": 6.431593036545938e-05, "loss": 0.0105, "step": 43720 }, { "grad_norm": 0.1339922398328781, "learning_rate": 6.430008711875426e-05, "loss": 0.0076, "step": 43730 }, { "grad_norm": 0.148899644613266, "learning_rate": 6.42842423082131e-05, "loss": 0.0091, "step": 43740 }, { "grad_norm": 0.1551208198070526, "learning_rate": 6.426839593556868e-05, "loss": 0.0087, "step": 43750 }, { "grad_norm": 0.1474055051803589, "learning_rate": 6.425254800255392e-05, "loss": 0.0084, "step": 43760 }, { "grad_norm": 0.10007350146770477, "learning_rate": 6.423669851090196e-05, "loss": 0.007, "step": 43770 }, { "grad_norm": 0.14367032051086426, "learning_rate": 6.422084746234604e-05, "loss": 0.009, "step": 43780 }, { "grad_norm": 0.14107145369052887, "learning_rate": 6.420499485861965e-05, "loss": 0.0101, "step": 43790 }, { "grad_norm": 0.10873478651046753, "learning_rate": 6.418914070145636e-05, "loss": 0.0106, "step": 43800 }, { "grad_norm": 0.11814986914396286, "learning_rate": 6.417328499258996e-05, "loss": 0.0069, "step": 43810 }, { "grad_norm": 0.09544062614440918, "learning_rate": 6.415742773375446e-05, "loss": 0.0103, "step": 43820 }, { "grad_norm": 0.10562364012002945, "learning_rate": 6.414156892668393e-05, "loss": 0.009, "step": 43830 }, { "grad_norm": 0.15414975583553314, "learning_rate": 6.412570857311267e-05, "loss": 0.009, "step": 43840 }, { "grad_norm": 0.17856736481189728, "learning_rate": 6.410984667477518e-05, "loss": 0.0101, "step": 43850 }, { "grad_norm": 0.14858607947826385, "learning_rate": 6.409398323340607e-05, "loss": 0.0099, "step": 43860 }, { "grad_norm": 0.1407947987318039, "learning_rate": 6.407811825074012e-05, "loss": 0.01, "step": 43870 }, { "grad_norm": 0.1043509691953659, "learning_rate": 6.406225172851234e-05, "loss": 0.0096, "step": 43880 }, { "grad_norm": 0.14841027557849884, "learning_rate": 6.404638366845786e-05, "loss": 0.007, "step": 43890 }, { "grad_norm": 0.14292173087596893, "learning_rate": 6.403051407231196e-05, "loss": 0.0086, "step": 43900 }, { "grad_norm": 0.14405131340026855, "learning_rate": 6.401464294181016e-05, "loss": 0.0083, "step": 43910 }, { "grad_norm": 0.2688356935977936, "learning_rate": 6.399877027868808e-05, "loss": 0.0072, "step": 43920 }, { "grad_norm": 0.14595700800418854, "learning_rate": 6.39828960846815e-05, "loss": 0.007, "step": 43930 }, { "grad_norm": 0.10824459791183472, "learning_rate": 6.396702036152644e-05, "loss": 0.0087, "step": 43940 }, { "grad_norm": 0.13305041193962097, "learning_rate": 6.395114311095905e-05, "loss": 0.0132, "step": 43950 }, { "grad_norm": 0.15637823939323425, "learning_rate": 6.393526433471559e-05, "loss": 0.0081, "step": 43960 }, { "grad_norm": 0.13111703097820282, "learning_rate": 6.391938403453262e-05, "loss": 0.0068, "step": 43970 }, { "grad_norm": 0.14438453316688538, "learning_rate": 6.390350221214671e-05, "loss": 0.0091, "step": 43980 }, { "grad_norm": 0.16550707817077637, "learning_rate": 6.38876188692947e-05, "loss": 0.0094, "step": 43990 }, { "grad_norm": 0.14131927490234375, "learning_rate": 6.387173400771358e-05, "loss": 0.0078, "step": 44000 }, { "grad_norm": 0.14201118052005768, "learning_rate": 6.385584762914047e-05, "loss": 0.0082, "step": 44010 }, { "grad_norm": 0.14954456686973572, "learning_rate": 6.38399597353127e-05, "loss": 0.0085, "step": 44020 }, { "grad_norm": 0.1398138403892517, "learning_rate": 6.382407032796775e-05, "loss": 0.0085, "step": 44030 }, { "grad_norm": 0.17786434292793274, "learning_rate": 6.380817940884325e-05, "loss": 0.0087, "step": 44040 }, { "grad_norm": 0.165127232670784, "learning_rate": 6.379228697967702e-05, "loss": 0.0068, "step": 44050 }, { "grad_norm": 0.09992137551307678, "learning_rate": 6.3776393042207e-05, "loss": 0.0089, "step": 44060 }, { "grad_norm": 0.1479111611843109, "learning_rate": 6.376049759817137e-05, "loss": 0.0089, "step": 44070 }, { "grad_norm": 0.14022564888000488, "learning_rate": 6.37446006493084e-05, "loss": 0.0081, "step": 44080 }, { "grad_norm": 0.11592023074626923, "learning_rate": 6.372870219735656e-05, "loss": 0.007, "step": 44090 }, { "grad_norm": 0.1671343296766281, "learning_rate": 6.371280224405452e-05, "loss": 0.0088, "step": 44100 }, { "grad_norm": 0.10071511566638947, "learning_rate": 6.369690079114101e-05, "loss": 0.0098, "step": 44110 }, { "grad_norm": 0.1094026267528534, "learning_rate": 6.368099784035504e-05, "loss": 0.0093, "step": 44120 }, { "grad_norm": 0.12752443552017212, "learning_rate": 6.366509339343572e-05, "loss": 0.0082, "step": 44130 }, { "grad_norm": 0.11943061649799347, "learning_rate": 6.364918745212232e-05, "loss": 0.0083, "step": 44140 }, { "grad_norm": 0.12281709909439087, "learning_rate": 6.363328001815431e-05, "loss": 0.0076, "step": 44150 }, { "grad_norm": 0.15138784050941467, "learning_rate": 6.361737109327128e-05, "loss": 0.0072, "step": 44160 }, { "grad_norm": 0.12232646346092224, "learning_rate": 6.360146067921301e-05, "loss": 0.0099, "step": 44170 }, { "grad_norm": 0.16153684258460999, "learning_rate": 6.35855487777195e-05, "loss": 0.0066, "step": 44180 }, { "grad_norm": 0.16745774447917938, "learning_rate": 6.356963539053076e-05, "loss": 0.0077, "step": 44190 }, { "grad_norm": 0.12573789060115814, "learning_rate": 6.355372051938711e-05, "loss": 0.0094, "step": 44200 }, { "grad_norm": 0.11123122274875641, "learning_rate": 6.353780416602894e-05, "loss": 0.0073, "step": 44210 }, { "grad_norm": 0.13699398934841156, "learning_rate": 6.352188633219689e-05, "loss": 0.0075, "step": 44220 }, { "grad_norm": 0.13578833639621735, "learning_rate": 6.350596701963166e-05, "loss": 0.0106, "step": 44230 }, { "grad_norm": 0.16872122883796692, "learning_rate": 6.349004623007419e-05, "loss": 0.0081, "step": 44240 }, { "grad_norm": 0.15659727156162262, "learning_rate": 6.347412396526555e-05, "loss": 0.0092, "step": 44250 }, { "grad_norm": 0.13165591657161713, "learning_rate": 6.345820022694696e-05, "loss": 0.0082, "step": 44260 }, { "grad_norm": 0.2393989861011505, "learning_rate": 6.344227501685984e-05, "loss": 0.0107, "step": 44270 }, { "grad_norm": 0.13739445805549622, "learning_rate": 6.342634833674572e-05, "loss": 0.0096, "step": 44280 }, { "grad_norm": 0.14374248683452606, "learning_rate": 6.341042018834635e-05, "loss": 0.01, "step": 44290 }, { "grad_norm": 0.13289517164230347, "learning_rate": 6.339449057340359e-05, "loss": 0.0081, "step": 44300 }, { "grad_norm": 0.14688174426555634, "learning_rate": 6.337855949365945e-05, "loss": 0.0083, "step": 44310 }, { "grad_norm": 0.14681659638881683, "learning_rate": 6.336262695085619e-05, "loss": 0.0105, "step": 44320 }, { "grad_norm": 0.161886528134346, "learning_rate": 6.334669294673612e-05, "loss": 0.0101, "step": 44330 }, { "grad_norm": 0.13302849233150482, "learning_rate": 6.333075748304179e-05, "loss": 0.0098, "step": 44340 }, { "grad_norm": 0.14786654710769653, "learning_rate": 6.331482056151585e-05, "loss": 0.0089, "step": 44350 }, { "grad_norm": 0.11029578000307083, "learning_rate": 6.329888218390117e-05, "loss": 0.0078, "step": 44360 }, { "grad_norm": 0.12457606196403503, "learning_rate": 6.328294235194072e-05, "loss": 0.006, "step": 44370 }, { "grad_norm": 0.09772173315286636, "learning_rate": 6.326700106737765e-05, "loss": 0.0084, "step": 44380 }, { "grad_norm": 0.13378606736660004, "learning_rate": 6.325105833195531e-05, "loss": 0.0102, "step": 44390 }, { "grad_norm": 0.2041173130273819, "learning_rate": 6.323511414741715e-05, "loss": 0.0071, "step": 44400 }, { "grad_norm": 0.11898983269929886, "learning_rate": 6.321916851550678e-05, "loss": 0.0091, "step": 44410 }, { "grad_norm": 0.1482560783624649, "learning_rate": 6.320322143796806e-05, "loss": 0.009, "step": 44420 }, { "grad_norm": 0.14468273520469666, "learning_rate": 6.318727291654488e-05, "loss": 0.0092, "step": 44430 }, { "grad_norm": 0.11379671841859818, "learning_rate": 6.317132295298134e-05, "loss": 0.0084, "step": 44440 }, { "grad_norm": 0.14324623346328735, "learning_rate": 6.315537154902173e-05, "loss": 0.0093, "step": 44450 }, { "grad_norm": 0.13047806918621063, "learning_rate": 6.313941870641048e-05, "loss": 0.0091, "step": 44460 }, { "grad_norm": 0.14219538867473602, "learning_rate": 6.312346442689214e-05, "loss": 0.0096, "step": 44470 }, { "grad_norm": 0.15076234936714172, "learning_rate": 6.310750871221147e-05, "loss": 0.0092, "step": 44480 }, { "grad_norm": 0.14217452704906464, "learning_rate": 6.309155156411335e-05, "loss": 0.0064, "step": 44490 }, { "grad_norm": 0.15951888263225555, "learning_rate": 6.307559298434284e-05, "loss": 0.008, "step": 44500 }, { "grad_norm": 0.12564648687839508, "learning_rate": 6.305963297464512e-05, "loss": 0.0071, "step": 44510 }, { "grad_norm": 0.11184273660182953, "learning_rate": 6.304367153676561e-05, "loss": 0.0073, "step": 44520 }, { "grad_norm": 0.16154775023460388, "learning_rate": 6.302770867244978e-05, "loss": 0.0067, "step": 44530 }, { "grad_norm": 0.1435869038105011, "learning_rate": 6.301174438344328e-05, "loss": 0.0097, "step": 44540 }, { "grad_norm": 0.10890448093414307, "learning_rate": 6.299577867149202e-05, "loss": 0.0079, "step": 44550 }, { "grad_norm": 0.19289541244506836, "learning_rate": 6.297981153834192e-05, "loss": 0.0082, "step": 44560 }, { "grad_norm": 0.15092749893665314, "learning_rate": 6.296384298573916e-05, "loss": 0.0086, "step": 44570 }, { "grad_norm": 0.15408842265605927, "learning_rate": 6.294787301543001e-05, "loss": 0.0109, "step": 44580 }, { "grad_norm": 0.10098302364349365, "learning_rate": 6.293190162916095e-05, "loss": 0.0094, "step": 44590 }, { "grad_norm": 0.11759831011295319, "learning_rate": 6.291592882867855e-05, "loss": 0.0085, "step": 44600 }, { "grad_norm": 0.12487073242664337, "learning_rate": 6.28999546157296e-05, "loss": 0.0094, "step": 44610 }, { "grad_norm": 0.18183255195617676, "learning_rate": 6.288397899206102e-05, "loss": 0.0098, "step": 44620 }, { "grad_norm": 0.19494350254535675, "learning_rate": 6.286800195941984e-05, "loss": 0.0086, "step": 44630 }, { "grad_norm": 0.13987988233566284, "learning_rate": 6.285202351955334e-05, "loss": 0.0093, "step": 44640 }, { "grad_norm": 0.137060284614563, "learning_rate": 6.283604367420887e-05, "loss": 0.0087, "step": 44650 }, { "grad_norm": 0.1408076286315918, "learning_rate": 6.282006242513394e-05, "loss": 0.0093, "step": 44660 }, { "grad_norm": 0.14770467579364777, "learning_rate": 6.280407977407628e-05, "loss": 0.0079, "step": 44670 }, { "grad_norm": 0.15006902813911438, "learning_rate": 6.27880957227837e-05, "loss": 0.009, "step": 44680 }, { "grad_norm": 0.12717069685459137, "learning_rate": 6.27721102730042e-05, "loss": 0.0081, "step": 44690 }, { "grad_norm": 0.17477184534072876, "learning_rate": 6.275612342648592e-05, "loss": 0.0096, "step": 44700 }, { "grad_norm": 0.14287438988685608, "learning_rate": 6.274013518497716e-05, "loss": 0.0103, "step": 44710 }, { "grad_norm": 0.12127934396266937, "learning_rate": 6.272414555022636e-05, "loss": 0.0085, "step": 44720 }, { "grad_norm": 0.14067496359348297, "learning_rate": 6.270815452398215e-05, "loss": 0.0082, "step": 44730 }, { "grad_norm": 0.15528884530067444, "learning_rate": 6.269216210799326e-05, "loss": 0.0098, "step": 44740 }, { "grad_norm": 0.15944626927375793, "learning_rate": 6.26761683040086e-05, "loss": 0.0092, "step": 44750 }, { "grad_norm": 0.09891590476036072, "learning_rate": 6.266017311377723e-05, "loss": 0.0068, "step": 44760 }, { "grad_norm": 0.13845692574977875, "learning_rate": 6.264417653904839e-05, "loss": 0.0074, "step": 44770 }, { "grad_norm": 0.1518610715866089, "learning_rate": 6.262817858157139e-05, "loss": 0.0108, "step": 44780 }, { "grad_norm": 0.11428794264793396, "learning_rate": 6.261217924309576e-05, "loss": 0.0071, "step": 44790 }, { "grad_norm": 0.1572670191526413, "learning_rate": 6.259617852537118e-05, "loss": 0.0078, "step": 44800 }, { "grad_norm": 0.16817212104797363, "learning_rate": 6.258017643014747e-05, "loss": 0.0072, "step": 44810 }, { "grad_norm": 0.10205160081386566, "learning_rate": 6.256417295917456e-05, "loss": 0.009, "step": 44820 }, { "grad_norm": 0.13022659718990326, "learning_rate": 6.254816811420258e-05, "loss": 0.0109, "step": 44830 }, { "grad_norm": 0.16392424702644348, "learning_rate": 6.253216189698183e-05, "loss": 0.0096, "step": 44840 }, { "grad_norm": 0.1480032205581665, "learning_rate": 6.251615430926267e-05, "loss": 0.0084, "step": 44850 }, { "grad_norm": 0.20425792038440704, "learning_rate": 6.25001453527957e-05, "loss": 0.009, "step": 44860 }, { "grad_norm": 0.29410144686698914, "learning_rate": 6.248413502933164e-05, "loss": 0.0089, "step": 44870 }, { "grad_norm": 0.14268611371517181, "learning_rate": 6.246812334062133e-05, "loss": 0.0084, "step": 44880 }, { "grad_norm": 0.15109287202358246, "learning_rate": 6.245211028841579e-05, "loss": 0.007, "step": 44890 }, { "grad_norm": 0.13455669581890106, "learning_rate": 6.24360958744662e-05, "loss": 0.0084, "step": 44900 }, { "grad_norm": 0.17674314975738525, "learning_rate": 6.242008010052387e-05, "loss": 0.0081, "step": 44910 }, { "grad_norm": 0.08194762468338013, "learning_rate": 6.240406296834024e-05, "loss": 0.0071, "step": 44920 }, { "grad_norm": 0.16181620955467224, "learning_rate": 6.238804447966694e-05, "loss": 0.0137, "step": 44930 }, { "grad_norm": 0.1041453629732132, "learning_rate": 6.237202463625573e-05, "loss": 0.0078, "step": 44940 }, { "grad_norm": 0.22375273704528809, "learning_rate": 6.235600343985848e-05, "loss": 0.0076, "step": 44950 }, { "grad_norm": 0.16010989248752594, "learning_rate": 6.233998089222729e-05, "loss": 0.0082, "step": 44960 }, { "grad_norm": 0.09542428702116013, "learning_rate": 6.232395699511433e-05, "loss": 0.0066, "step": 44970 }, { "grad_norm": 0.1414061188697815, "learning_rate": 6.230793175027199e-05, "loss": 0.0083, "step": 44980 }, { "grad_norm": 0.1479421705007553, "learning_rate": 6.22919051594527e-05, "loss": 0.0108, "step": 44990 }, { "grad_norm": 0.14794176816940308, "learning_rate": 6.227587722440917e-05, "loss": 0.0074, "step": 45000 }, { "grad_norm": 0.10927410423755646, "learning_rate": 6.225984794689414e-05, "loss": 0.0077, "step": 45010 }, { "grad_norm": 0.15535537898540497, "learning_rate": 6.22438173286606e-05, "loss": 0.0088, "step": 45020 }, { "grad_norm": 0.12585505843162537, "learning_rate": 6.222778537146159e-05, "loss": 0.0078, "step": 45030 }, { "grad_norm": 0.14933590590953827, "learning_rate": 6.221175207705037e-05, "loss": 0.0096, "step": 45040 }, { "grad_norm": 0.15742552280426025, "learning_rate": 6.219571744718027e-05, "loss": 0.0099, "step": 45050 }, { "grad_norm": 0.17550306022167206, "learning_rate": 6.217968148360487e-05, "loss": 0.0087, "step": 45060 }, { "grad_norm": 0.17989486455917358, "learning_rate": 6.216364418807782e-05, "loss": 0.0078, "step": 45070 }, { "grad_norm": 0.13798263669013977, "learning_rate": 6.214760556235292e-05, "loss": 0.0088, "step": 45080 }, { "grad_norm": 0.16269125044345856, "learning_rate": 6.213156560818414e-05, "loss": 0.0066, "step": 45090 }, { "grad_norm": 0.1141849085688591, "learning_rate": 6.211552432732559e-05, "loss": 0.0088, "step": 45100 }, { "grad_norm": 0.13578447699546814, "learning_rate": 6.20994817215315e-05, "loss": 0.009, "step": 45110 }, { "grad_norm": 0.15116499364376068, "learning_rate": 6.208343779255628e-05, "loss": 0.0082, "step": 45120 }, { "grad_norm": 0.14304323494434357, "learning_rate": 6.206739254215449e-05, "loss": 0.0081, "step": 45130 }, { "grad_norm": 0.20245075225830078, "learning_rate": 6.205134597208077e-05, "loss": 0.0091, "step": 45140 }, { "grad_norm": 0.13314950466156006, "learning_rate": 6.203529808409e-05, "loss": 0.0086, "step": 45150 }, { "grad_norm": 0.16731904447078705, "learning_rate": 6.201924887993712e-05, "loss": 0.0094, "step": 45160 }, { "grad_norm": 0.1547771841287613, "learning_rate": 6.200319836137725e-05, "loss": 0.0082, "step": 45170 }, { "grad_norm": 0.09847455471754074, "learning_rate": 6.198714653016565e-05, "loss": 0.0098, "step": 45180 }, { "grad_norm": 0.23215532302856445, "learning_rate": 6.197109338805774e-05, "loss": 0.0095, "step": 45190 }, { "grad_norm": 0.1251964420080185, "learning_rate": 6.195503893680903e-05, "loss": 0.0106, "step": 45200 }, { "grad_norm": 0.1684102565050125, "learning_rate": 6.193898317817524e-05, "loss": 0.0079, "step": 45210 }, { "grad_norm": 0.13730712234973907, "learning_rate": 6.192292611391221e-05, "loss": 0.0068, "step": 45220 }, { "grad_norm": 0.1295456886291504, "learning_rate": 6.190686774577591e-05, "loss": 0.008, "step": 45230 }, { "grad_norm": 0.11558122932910919, "learning_rate": 6.189080807552245e-05, "loss": 0.0087, "step": 45240 }, { "grad_norm": 0.0980461984872818, "learning_rate": 6.187474710490809e-05, "loss": 0.011, "step": 45250 }, { "grad_norm": 0.1494695544242859, "learning_rate": 6.185868483568926e-05, "loss": 0.009, "step": 45260 }, { "grad_norm": 0.10586129873991013, "learning_rate": 6.184262126962245e-05, "loss": 0.0092, "step": 45270 }, { "grad_norm": 0.13365502655506134, "learning_rate": 6.182655640846442e-05, "loss": 0.0081, "step": 45280 }, { "grad_norm": 0.08448431640863419, "learning_rate": 6.181049025397196e-05, "loss": 0.0086, "step": 45290 }, { "grad_norm": 0.17646266520023346, "learning_rate": 6.179442280790202e-05, "loss": 0.0103, "step": 45300 }, { "grad_norm": 0.13150733709335327, "learning_rate": 6.177835407201174e-05, "loss": 0.0074, "step": 45310 }, { "grad_norm": 0.14285418391227722, "learning_rate": 6.176228404805839e-05, "loss": 0.0071, "step": 45320 }, { "grad_norm": 0.09063268452882767, "learning_rate": 6.174621273779932e-05, "loss": 0.0078, "step": 45330 }, { "grad_norm": 0.16990867257118225, "learning_rate": 6.17301401429921e-05, "loss": 0.0088, "step": 45340 }, { "grad_norm": 0.13643746078014374, "learning_rate": 6.171406626539436e-05, "loss": 0.0073, "step": 45350 }, { "grad_norm": 0.10947076231241226, "learning_rate": 6.169799110676398e-05, "loss": 0.0072, "step": 45360 }, { "grad_norm": 0.16311430931091309, "learning_rate": 6.168191466885885e-05, "loss": 0.0094, "step": 45370 }, { "grad_norm": 0.18678320944309235, "learning_rate": 6.166583695343711e-05, "loss": 0.0089, "step": 45380 }, { "grad_norm": 0.18299295008182526, "learning_rate": 6.164975796225698e-05, "loss": 0.0097, "step": 45390 }, { "grad_norm": 0.20399267971515656, "learning_rate": 6.163367769707683e-05, "loss": 0.0112, "step": 45400 }, { "grad_norm": 0.10977555811405182, "learning_rate": 6.161759615965519e-05, "loss": 0.0077, "step": 45410 }, { "grad_norm": 0.1156286671757698, "learning_rate": 6.16015133517507e-05, "loss": 0.0073, "step": 45420 }, { "grad_norm": 0.1198887899518013, "learning_rate": 6.158542927512214e-05, "loss": 0.0066, "step": 45430 }, { "grad_norm": 0.16967980563640594, "learning_rate": 6.156934393152846e-05, "loss": 0.0087, "step": 45440 }, { "grad_norm": 0.11374549567699432, "learning_rate": 6.15532573227287e-05, "loss": 0.0106, "step": 45450 }, { "grad_norm": 0.09918329119682312, "learning_rate": 6.153716945048212e-05, "loss": 0.0067, "step": 45460 }, { "grad_norm": 0.1546909511089325, "learning_rate": 6.152108031654802e-05, "loss": 0.0094, "step": 45470 }, { "grad_norm": 0.1321554183959961, "learning_rate": 6.15049899226859e-05, "loss": 0.0084, "step": 45480 }, { "grad_norm": 0.17648008465766907, "learning_rate": 6.148889827065537e-05, "loss": 0.008, "step": 45490 }, { "grad_norm": 0.16410568356513977, "learning_rate": 6.147280536221622e-05, "loss": 0.0068, "step": 45500 }, { "grad_norm": 0.1399376094341278, "learning_rate": 6.145671119912832e-05, "loss": 0.0082, "step": 45510 }, { "grad_norm": 0.2032223492860794, "learning_rate": 6.144061578315169e-05, "loss": 0.0113, "step": 45520 }, { "grad_norm": 0.21912814676761627, "learning_rate": 6.142451911604654e-05, "loss": 0.0077, "step": 45530 }, { "grad_norm": 0.15499460697174072, "learning_rate": 6.140842119957315e-05, "loss": 0.009, "step": 45540 }, { "grad_norm": 0.12229763716459274, "learning_rate": 6.139232203549197e-05, "loss": 0.007, "step": 45550 }, { "grad_norm": 0.1309020072221756, "learning_rate": 6.137622162556357e-05, "loss": 0.0083, "step": 45560 }, { "grad_norm": 0.10162603110074997, "learning_rate": 6.13601199715487e-05, "loss": 0.0077, "step": 45570 }, { "grad_norm": 0.12024009972810745, "learning_rate": 6.134401707520816e-05, "loss": 0.0087, "step": 45580 }, { "grad_norm": 0.18467143177986145, "learning_rate": 6.132791293830298e-05, "loss": 0.0089, "step": 45590 }, { "grad_norm": 0.1328478753566742, "learning_rate": 6.131180756259428e-05, "loss": 0.0108, "step": 45600 }, { "grad_norm": 0.09616959095001221, "learning_rate": 6.129570094984331e-05, "loss": 0.0083, "step": 45610 }, { "grad_norm": 0.12692688405513763, "learning_rate": 6.127959310181145e-05, "loss": 0.0087, "step": 45620 }, { "grad_norm": 0.11590571701526642, "learning_rate": 6.126348402026026e-05, "loss": 0.0066, "step": 45630 }, { "grad_norm": 0.18662306666374207, "learning_rate": 6.12473737069514e-05, "loss": 0.0098, "step": 45640 }, { "grad_norm": 0.1706753671169281, "learning_rate": 6.123126216364665e-05, "loss": 0.0076, "step": 45650 }, { "grad_norm": 0.13880521059036255, "learning_rate": 6.121514939210797e-05, "loss": 0.0065, "step": 45660 }, { "grad_norm": 0.09890437126159668, "learning_rate": 6.119903539409741e-05, "loss": 0.0073, "step": 45670 }, { "grad_norm": 0.12946151196956635, "learning_rate": 6.118292017137716e-05, "loss": 0.0089, "step": 45680 }, { "grad_norm": 0.15852093696594238, "learning_rate": 6.116680372570959e-05, "loss": 0.0089, "step": 45690 }, { "grad_norm": 0.143085777759552, "learning_rate": 6.115068605885713e-05, "loss": 0.0075, "step": 45700 }, { "grad_norm": 0.12088757008314133, "learning_rate": 6.113456717258243e-05, "loss": 0.0101, "step": 45710 }, { "grad_norm": 0.16015516221523285, "learning_rate": 6.11184470686482e-05, "loss": 0.008, "step": 45720 }, { "grad_norm": 0.1119459792971611, "learning_rate": 6.110232574881731e-05, "loss": 0.0101, "step": 45730 }, { "grad_norm": 0.12734447419643402, "learning_rate": 6.108620321485277e-05, "loss": 0.008, "step": 45740 }, { "grad_norm": 0.12854361534118652, "learning_rate": 6.107007946851773e-05, "loss": 0.012, "step": 45750 }, { "grad_norm": 0.1672356277704239, "learning_rate": 6.105395451157542e-05, "loss": 0.0097, "step": 45760 }, { "grad_norm": 0.16298826038837433, "learning_rate": 6.103782834578928e-05, "loss": 0.0124, "step": 45770 }, { "grad_norm": 0.1772921234369278, "learning_rate": 6.102170097292281e-05, "loss": 0.0111, "step": 45780 }, { "grad_norm": 0.16843535006046295, "learning_rate": 6.1005572394739716e-05, "loss": 0.0094, "step": 45790 }, { "grad_norm": 0.17780092358589172, "learning_rate": 6.0989442613003765e-05, "loss": 0.0126, "step": 45800 }, { "grad_norm": 0.13131839036941528, "learning_rate": 6.097331162947888e-05, "loss": 0.0079, "step": 45810 }, { "grad_norm": 0.1482413411140442, "learning_rate": 6.095717944592914e-05, "loss": 0.0077, "step": 45820 }, { "grad_norm": 0.13680407404899597, "learning_rate": 6.094104606411873e-05, "loss": 0.0085, "step": 45830 }, { "grad_norm": 0.17691096663475037, "learning_rate": 6.0924911485811966e-05, "loss": 0.0096, "step": 45840 }, { "grad_norm": 0.14542384445667267, "learning_rate": 6.090877571277331e-05, "loss": 0.0088, "step": 45850 }, { "grad_norm": 0.1544668972492218, "learning_rate": 6.089263874676736e-05, "loss": 0.0084, "step": 45860 }, { "grad_norm": 0.15971174836158752, "learning_rate": 6.0876500589558796e-05, "loss": 0.011, "step": 45870 }, { "grad_norm": 0.13394233584403992, "learning_rate": 6.086036124291248e-05, "loss": 0.0101, "step": 45880 }, { "grad_norm": 0.16303633153438568, "learning_rate": 6.084422070859339e-05, "loss": 0.0073, "step": 45890 }, { "grad_norm": 0.11280432343482971, "learning_rate": 6.082807898836663e-05, "loss": 0.0069, "step": 45900 }, { "grad_norm": 0.10865559428930283, "learning_rate": 6.081193608399742e-05, "loss": 0.0069, "step": 45910 }, { "grad_norm": 0.167584627866745, "learning_rate": 6.0795791997251164e-05, "loss": 0.0095, "step": 45920 }, { "grad_norm": 0.12929455935955048, "learning_rate": 6.0779646729893294e-05, "loss": 0.0073, "step": 45930 }, { "grad_norm": 0.14698295295238495, "learning_rate": 6.0763500283689476e-05, "loss": 0.0073, "step": 45940 }, { "grad_norm": 0.127268984913826, "learning_rate": 6.0747352660405455e-05, "loss": 0.0075, "step": 45950 }, { "grad_norm": 0.13995321094989777, "learning_rate": 6.073120386180709e-05, "loss": 0.0076, "step": 45960 }, { "grad_norm": 0.1496225893497467, "learning_rate": 6.0715053889660425e-05, "loss": 0.0085, "step": 45970 }, { "grad_norm": 0.1323724389076233, "learning_rate": 6.069890274573157e-05, "loss": 0.0083, "step": 45980 }, { "grad_norm": 0.15009288489818573, "learning_rate": 6.068275043178679e-05, "loss": 0.0086, "step": 45990 }, { "grad_norm": 0.17315451800823212, "learning_rate": 6.066659694959248e-05, "loss": 0.011, "step": 46000 }, { "grad_norm": 0.14495696127414703, "learning_rate": 6.0650442300915176e-05, "loss": 0.009, "step": 46010 }, { "grad_norm": 0.1250389814376831, "learning_rate": 6.063428648752152e-05, "loss": 0.0075, "step": 46020 }, { "grad_norm": 0.14872407913208008, "learning_rate": 6.0618129511178266e-05, "loss": 0.007, "step": 46030 }, { "grad_norm": 0.15620730817317963, "learning_rate": 6.060197137365234e-05, "loss": 0.0103, "step": 46040 }, { "grad_norm": 0.1408640742301941, "learning_rate": 6.058581207671077e-05, "loss": 0.0113, "step": 46050 }, { "grad_norm": 0.10325264930725098, "learning_rate": 6.056965162212072e-05, "loss": 0.0102, "step": 46060 }, { "grad_norm": 0.1661839336156845, "learning_rate": 6.055349001164943e-05, "loss": 0.0076, "step": 46070 }, { "grad_norm": 0.18801972270011902, "learning_rate": 6.0537327247064347e-05, "loss": 0.0102, "step": 46080 }, { "grad_norm": 0.151511549949646, "learning_rate": 6.0521163330133014e-05, "loss": 0.008, "step": 46090 }, { "grad_norm": 0.15363608300685883, "learning_rate": 6.050499826262306e-05, "loss": 0.0088, "step": 46100 }, { "grad_norm": 0.2198340743780136, "learning_rate": 6.0488832046302294e-05, "loss": 0.0087, "step": 46110 }, { "grad_norm": 0.1996387243270874, "learning_rate": 6.0472664682938626e-05, "loss": 0.0114, "step": 46120 }, { "grad_norm": 0.15285181999206543, "learning_rate": 6.045649617430009e-05, "loss": 0.0098, "step": 46130 }, { "grad_norm": 0.10391556471586227, "learning_rate": 6.0440326522154866e-05, "loss": 0.0086, "step": 46140 }, { "grad_norm": 0.13860569894313812, "learning_rate": 6.0424155728271224e-05, "loss": 0.0074, "step": 46150 }, { "grad_norm": 0.1227497085928917, "learning_rate": 6.040798379441758e-05, "loss": 0.0076, "step": 46160 }, { "grad_norm": 0.10449395328760147, "learning_rate": 6.0391810722362485e-05, "loss": 0.0072, "step": 46170 }, { "grad_norm": 0.14042958617210388, "learning_rate": 6.037563651387458e-05, "loss": 0.0069, "step": 46180 }, { "grad_norm": 0.11865068972110748, "learning_rate": 6.0359461170722666e-05, "loss": 0.0098, "step": 46190 }, { "grad_norm": 0.12589474022388458, "learning_rate": 6.034328469467566e-05, "loss": 0.0084, "step": 46200 }, { "grad_norm": 0.13201725482940674, "learning_rate": 6.0327107087502596e-05, "loss": 0.0079, "step": 46210 }, { "grad_norm": 0.11459247767925262, "learning_rate": 6.031092835097262e-05, "loss": 0.0067, "step": 46220 }, { "grad_norm": 0.17069704830646515, "learning_rate": 6.0294748486855024e-05, "loss": 0.0091, "step": 46230 }, { "grad_norm": 0.17700286209583282, "learning_rate": 6.0278567496919216e-05, "loss": 0.0073, "step": 46240 }, { "grad_norm": 0.14817498624324799, "learning_rate": 6.026238538293472e-05, "loss": 0.0072, "step": 46250 }, { "grad_norm": 0.10540226846933365, "learning_rate": 6.024620214667118e-05, "loss": 0.0068, "step": 46260 }, { "grad_norm": 0.1371375471353531, "learning_rate": 6.0230017789898384e-05, "loss": 0.0073, "step": 46270 }, { "grad_norm": 0.1395271271467209, "learning_rate": 6.021383231438622e-05, "loss": 0.0074, "step": 46280 }, { "grad_norm": 0.1602579653263092, "learning_rate": 6.0197645721904704e-05, "loss": 0.0071, "step": 46290 }, { "grad_norm": 0.1460113227367401, "learning_rate": 6.0181458014224e-05, "loss": 0.0075, "step": 46300 }, { "grad_norm": 0.1506108194589615, "learning_rate": 6.016526919311435e-05, "loss": 0.0093, "step": 46310 }, { "grad_norm": 0.11616382002830505, "learning_rate": 6.014907926034613e-05, "loss": 0.0092, "step": 46320 }, { "grad_norm": 0.17805646359920502, "learning_rate": 6.0132888217689875e-05, "loss": 0.0091, "step": 46330 }, { "grad_norm": 0.13011416792869568, "learning_rate": 6.01166960669162e-05, "loss": 0.0108, "step": 46340 }, { "grad_norm": 0.16332052648067474, "learning_rate": 6.0100502809795845e-05, "loss": 0.0092, "step": 46350 }, { "grad_norm": 0.11490874737501144, "learning_rate": 6.00843084480997e-05, "loss": 0.0088, "step": 46360 }, { "grad_norm": 0.15347816050052643, "learning_rate": 6.0068112983598736e-05, "loss": 0.0087, "step": 46370 }, { "grad_norm": 0.12799987196922302, "learning_rate": 6.005191641806407e-05, "loss": 0.008, "step": 46380 }, { "grad_norm": 0.2055705338716507, "learning_rate": 6.003571875326694e-05, "loss": 0.0101, "step": 46390 }, { "grad_norm": 0.12685975432395935, "learning_rate": 6.00195199909787e-05, "loss": 0.0083, "step": 46400 }, { "grad_norm": 0.09559415280818939, "learning_rate": 6.000332013297082e-05, "loss": 0.011, "step": 46410 }, { "grad_norm": 0.1830577254295349, "learning_rate": 5.998711918101487e-05, "loss": 0.0084, "step": 46420 }, { "grad_norm": 0.11048440635204315, "learning_rate": 5.997091713688261e-05, "loss": 0.0088, "step": 46430 }, { "grad_norm": 0.1437585949897766, "learning_rate": 5.9954714002345836e-05, "loss": 0.0082, "step": 46440 }, { "grad_norm": 0.14489832520484924, "learning_rate": 5.993850977917649e-05, "loss": 0.0083, "step": 46450 }, { "grad_norm": 0.1380816102027893, "learning_rate": 5.992230446914667e-05, "loss": 0.0088, "step": 46460 }, { "grad_norm": 0.14697976410388947, "learning_rate": 5.990609807402855e-05, "loss": 0.0082, "step": 46470 }, { "grad_norm": 0.1515607386827469, "learning_rate": 5.988989059559443e-05, "loss": 0.0101, "step": 46480 }, { "grad_norm": 0.11632689833641052, "learning_rate": 5.987368203561675e-05, "loss": 0.0077, "step": 46490 }, { "grad_norm": 0.13194702565670013, "learning_rate": 5.9857472395868055e-05, "loss": 0.0065, "step": 46500 }, { "grad_norm": 0.17947113513946533, "learning_rate": 5.9841261678120983e-05, "loss": 0.0093, "step": 46510 }, { "grad_norm": 0.1166321262717247, "learning_rate": 5.982504988414834e-05, "loss": 0.0074, "step": 46520 }, { "grad_norm": 0.12315540760755539, "learning_rate": 5.9808837015723015e-05, "loss": 0.0077, "step": 46530 }, { "grad_norm": 0.14642328023910522, "learning_rate": 5.9792623074618016e-05, "loss": 0.0064, "step": 46540 }, { "grad_norm": 0.2072887122631073, "learning_rate": 5.977640806260648e-05, "loss": 0.01, "step": 46550 }, { "grad_norm": 0.2134414166212082, "learning_rate": 5.976019198146166e-05, "loss": 0.0081, "step": 46560 }, { "grad_norm": 0.12397775053977966, "learning_rate": 5.974397483295692e-05, "loss": 0.0081, "step": 46570 }, { "grad_norm": 0.1404530555009842, "learning_rate": 5.972775661886575e-05, "loss": 0.0095, "step": 46580 }, { "grad_norm": 0.11097496002912521, "learning_rate": 5.971153734096173e-05, "loss": 0.0075, "step": 46590 }, { "grad_norm": 0.10133321583271027, "learning_rate": 5.969531700101859e-05, "loss": 0.007, "step": 46600 }, { "grad_norm": 0.10083981603384018, "learning_rate": 5.9679095600810155e-05, "loss": 0.0077, "step": 46610 }, { "grad_norm": 0.14351488649845123, "learning_rate": 5.9662873142110384e-05, "loss": 0.0091, "step": 46620 }, { "grad_norm": 0.1299135684967041, "learning_rate": 5.964664962669333e-05, "loss": 0.009, "step": 46630 }, { "grad_norm": 0.10128579288721085, "learning_rate": 5.9630425056333186e-05, "loss": 0.007, "step": 46640 }, { "grad_norm": 0.11791660636663437, "learning_rate": 5.961419943280422e-05, "loss": 0.0081, "step": 46650 }, { "grad_norm": 0.14556989073753357, "learning_rate": 5.959797275788087e-05, "loss": 0.0076, "step": 46660 }, { "grad_norm": 0.10177737474441528, "learning_rate": 5.958174503333765e-05, "loss": 0.0077, "step": 46670 }, { "grad_norm": 0.10767778009176254, "learning_rate": 5.9565516260949195e-05, "loss": 0.0073, "step": 46680 }, { "grad_norm": 0.10952945798635483, "learning_rate": 5.9549286442490273e-05, "loss": 0.0078, "step": 46690 }, { "grad_norm": 0.14007382094860077, "learning_rate": 5.953305557973572e-05, "loss": 0.0107, "step": 46700 }, { "grad_norm": 0.13774502277374268, "learning_rate": 5.9516823674460564e-05, "loss": 0.0072, "step": 46710 }, { "grad_norm": 0.15770553052425385, "learning_rate": 5.9500590728439875e-05, "loss": 0.0082, "step": 46720 }, { "grad_norm": 0.1622246950864792, "learning_rate": 5.948435674344886e-05, "loss": 0.0093, "step": 46730 }, { "grad_norm": 0.16091899573802948, "learning_rate": 5.946812172126285e-05, "loss": 0.0083, "step": 46740 }, { "grad_norm": 0.12470352649688721, "learning_rate": 5.945188566365729e-05, "loss": 0.0087, "step": 46750 }, { "grad_norm": 0.10847002267837524, "learning_rate": 5.943564857240773e-05, "loss": 0.0077, "step": 46760 }, { "grad_norm": 0.13683338463306427, "learning_rate": 5.941941044928981e-05, "loss": 0.0079, "step": 46770 }, { "grad_norm": 0.10924641042947769, "learning_rate": 5.940317129607935e-05, "loss": 0.0072, "step": 46780 }, { "grad_norm": 0.12890100479125977, "learning_rate": 5.9386931114552204e-05, "loss": 0.0086, "step": 46790 }, { "grad_norm": 0.143100306391716, "learning_rate": 5.937068990648438e-05, "loss": 0.0095, "step": 46800 }, { "grad_norm": 0.12614388763904572, "learning_rate": 5.935444767365199e-05, "loss": 0.0071, "step": 46810 }, { "grad_norm": 0.12684443593025208, "learning_rate": 5.933820441783129e-05, "loss": 0.0079, "step": 46820 }, { "grad_norm": 0.16587664186954498, "learning_rate": 5.932196014079857e-05, "loss": 0.0071, "step": 46830 }, { "grad_norm": 0.12278135120868683, "learning_rate": 5.930571484433032e-05, "loss": 0.007, "step": 46840 }, { "grad_norm": 0.12557388842105865, "learning_rate": 5.928946853020309e-05, "loss": 0.006, "step": 46850 }, { "grad_norm": 0.12222062051296234, "learning_rate": 5.927322120019352e-05, "loss": 0.0074, "step": 46860 }, { "grad_norm": 0.139552041888237, "learning_rate": 5.925697285607844e-05, "loss": 0.0064, "step": 46870 }, { "grad_norm": 0.11826049536466599, "learning_rate": 5.924072349963472e-05, "loss": 0.0092, "step": 46880 }, { "grad_norm": 0.14614252746105194, "learning_rate": 5.922447313263938e-05, "loss": 0.0059, "step": 46890 }, { "grad_norm": 0.12995728850364685, "learning_rate": 5.920822175686952e-05, "loss": 0.0065, "step": 46900 }, { "grad_norm": 0.1534838229417801, "learning_rate": 5.919196937410239e-05, "loss": 0.0072, "step": 46910 }, { "grad_norm": 0.11998327076435089, "learning_rate": 5.917571598611529e-05, "loss": 0.0059, "step": 46920 }, { "grad_norm": 0.1268114149570465, "learning_rate": 5.9159461594685686e-05, "loss": 0.0084, "step": 46930 }, { "grad_norm": 0.16671571135520935, "learning_rate": 5.914320620159114e-05, "loss": 0.0091, "step": 46940 }, { "grad_norm": 0.10882103443145752, "learning_rate": 5.912694980860931e-05, "loss": 0.0093, "step": 46950 }, { "grad_norm": 0.1380501240491867, "learning_rate": 5.9110692417517964e-05, "loss": 0.0083, "step": 46960 }, { "grad_norm": 0.12075664103031158, "learning_rate": 5.9094434030095e-05, "loss": 0.0082, "step": 46970 }, { "grad_norm": 0.1651306450366974, "learning_rate": 5.907817464811841e-05, "loss": 0.0114, "step": 46980 }, { "grad_norm": 0.12022750079631805, "learning_rate": 5.906191427336627e-05, "loss": 0.0067, "step": 46990 }, { "grad_norm": 0.11650066822767258, "learning_rate": 5.9045652907616835e-05, "loss": 0.0057, "step": 47000 }, { "grad_norm": 0.16756553947925568, "learning_rate": 5.902939055264838e-05, "loss": 0.0073, "step": 47010 }, { "grad_norm": 0.11295709758996964, "learning_rate": 5.901312721023935e-05, "loss": 0.006, "step": 47020 }, { "grad_norm": 0.11842741072177887, "learning_rate": 5.8996862882168294e-05, "loss": 0.0078, "step": 47030 }, { "grad_norm": 0.10863760858774185, "learning_rate": 5.8980597570213826e-05, "loss": 0.0104, "step": 47040 }, { "grad_norm": 0.14566682279109955, "learning_rate": 5.896433127615471e-05, "loss": 0.0078, "step": 47050 }, { "grad_norm": 0.1483490914106369, "learning_rate": 5.894806400176981e-05, "loss": 0.0079, "step": 47060 }, { "grad_norm": 0.12169234454631805, "learning_rate": 5.893179574883808e-05, "loss": 0.008, "step": 47070 }, { "grad_norm": 0.11870867013931274, "learning_rate": 5.8915526519138585e-05, "loss": 0.0071, "step": 47080 }, { "grad_norm": 0.10985637456178665, "learning_rate": 5.889925631445053e-05, "loss": 0.0085, "step": 47090 }, { "grad_norm": 0.13948635756969452, "learning_rate": 5.888298513655318e-05, "loss": 0.0091, "step": 47100 }, { "grad_norm": 0.09934923052787781, "learning_rate": 5.886671298722592e-05, "loss": 0.0091, "step": 47110 }, { "grad_norm": 0.13587835431098938, "learning_rate": 5.885043986824828e-05, "loss": 0.0082, "step": 47120 }, { "grad_norm": 0.18127012252807617, "learning_rate": 5.883416578139982e-05, "loss": 0.0085, "step": 47130 }, { "grad_norm": 0.14290057122707367, "learning_rate": 5.881789072846029e-05, "loss": 0.0081, "step": 47140 }, { "grad_norm": 0.10222680866718292, "learning_rate": 5.8801614711209474e-05, "loss": 0.0098, "step": 47150 }, { "grad_norm": 0.1291130632162094, "learning_rate": 5.878533773142732e-05, "loss": 0.0093, "step": 47160 }, { "grad_norm": 0.1219773218035698, "learning_rate": 5.8769059790893834e-05, "loss": 0.007, "step": 47170 }, { "grad_norm": 0.1844453513622284, "learning_rate": 5.8752780891389145e-05, "loss": 0.0098, "step": 47180 }, { "grad_norm": 0.10954050719738007, "learning_rate": 5.873650103469351e-05, "loss": 0.0077, "step": 47190 }, { "grad_norm": 0.11700577288866043, "learning_rate": 5.872022022258725e-05, "loss": 0.0103, "step": 47200 }, { "grad_norm": 0.12281350046396255, "learning_rate": 5.8703938456850814e-05, "loss": 0.0093, "step": 47210 }, { "grad_norm": 0.18102240562438965, "learning_rate": 5.868765573926477e-05, "loss": 0.0096, "step": 47220 }, { "grad_norm": 0.1385345757007599, "learning_rate": 5.867137207160974e-05, "loss": 0.0083, "step": 47230 }, { "grad_norm": 0.11454480141401291, "learning_rate": 5.86550874556665e-05, "loss": 0.009, "step": 47240 }, { "grad_norm": 0.1419447660446167, "learning_rate": 5.863880189321592e-05, "loss": 0.0087, "step": 47250 }, { "grad_norm": 0.1263938546180725, "learning_rate": 5.862251538603894e-05, "loss": 0.0079, "step": 47260 }, { "grad_norm": 0.17035621404647827, "learning_rate": 5.8606227935916656e-05, "loss": 0.0097, "step": 47270 }, { "grad_norm": 0.13831809163093567, "learning_rate": 5.858993954463021e-05, "loss": 0.0076, "step": 47280 }, { "grad_norm": 0.11083220690488815, "learning_rate": 5.8573650213960896e-05, "loss": 0.0102, "step": 47290 }, { "grad_norm": 0.11718304455280304, "learning_rate": 5.8557359945690084e-05, "loss": 0.0054, "step": 47300 }, { "grad_norm": 0.1393827497959137, "learning_rate": 5.8541068741599246e-05, "loss": 0.0066, "step": 47310 }, { "grad_norm": 0.10526565462350845, "learning_rate": 5.8524776603469985e-05, "loss": 0.0082, "step": 47320 }, { "grad_norm": 0.17042717337608337, "learning_rate": 5.850848353308397e-05, "loss": 0.0071, "step": 47330 }, { "grad_norm": 0.1655944138765335, "learning_rate": 5.849218953222297e-05, "loss": 0.0077, "step": 47340 }, { "grad_norm": 0.12319321185350418, "learning_rate": 5.847589460266891e-05, "loss": 0.0094, "step": 47350 }, { "grad_norm": 0.1476910412311554, "learning_rate": 5.845959874620375e-05, "loss": 0.0102, "step": 47360 }, { "grad_norm": 0.11738580465316772, "learning_rate": 5.84433019646096e-05, "loss": 0.0084, "step": 47370 }, { "grad_norm": 0.11854327470064163, "learning_rate": 5.842700425966863e-05, "loss": 0.0085, "step": 47380 }, { "grad_norm": 0.11499853432178497, "learning_rate": 5.841070563316315e-05, "loss": 0.0086, "step": 47390 }, { "grad_norm": 0.1540999859571457, "learning_rate": 5.839440608687554e-05, "loss": 0.0083, "step": 47400 }, { "grad_norm": 0.15135471522808075, "learning_rate": 5.837810562258831e-05, "loss": 0.0095, "step": 47410 }, { "grad_norm": 0.13744886219501495, "learning_rate": 5.836180424208405e-05, "loss": 0.0077, "step": 47420 }, { "grad_norm": 0.1775258332490921, "learning_rate": 5.834550194714543e-05, "loss": 0.0079, "step": 47430 }, { "grad_norm": 0.11425654590129852, "learning_rate": 5.832919873955528e-05, "loss": 0.0074, "step": 47440 }, { "grad_norm": 0.13699883222579956, "learning_rate": 5.831289462109648e-05, "loss": 0.0075, "step": 47450 }, { "grad_norm": 0.16743329167366028, "learning_rate": 5.829658959355202e-05, "loss": 0.0099, "step": 47460 }, { "grad_norm": 0.15168018639087677, "learning_rate": 5.8280283658704995e-05, "loss": 0.0062, "step": 47470 }, { "grad_norm": 0.16320551931858063, "learning_rate": 5.826397681833859e-05, "loss": 0.0095, "step": 47480 }, { "grad_norm": 0.12863954901695251, "learning_rate": 5.824766907423612e-05, "loss": 0.0072, "step": 47490 }, { "grad_norm": 0.11709883064031601, "learning_rate": 5.8231360428180945e-05, "loss": 0.0069, "step": 47500 }, { "grad_norm": 0.14123030006885529, "learning_rate": 5.821505088195658e-05, "loss": 0.0088, "step": 47510 }, { "grad_norm": 0.11584483832120895, "learning_rate": 5.819874043734661e-05, "loss": 0.0066, "step": 47520 }, { "grad_norm": 0.11637228727340698, "learning_rate": 5.8182429096134695e-05, "loss": 0.0071, "step": 47530 }, { "grad_norm": 0.13115206360816956, "learning_rate": 5.816611686010465e-05, "loss": 0.0089, "step": 47540 }, { "grad_norm": 0.14249993860721588, "learning_rate": 5.814980373104033e-05, "loss": 0.0086, "step": 47550 }, { "grad_norm": 0.12519057095050812, "learning_rate": 5.813348971072572e-05, "loss": 0.0084, "step": 47560 }, { "grad_norm": 0.17093269526958466, "learning_rate": 5.811717480094492e-05, "loss": 0.0082, "step": 47570 }, { "grad_norm": 0.15836161375045776, "learning_rate": 5.810085900348209e-05, "loss": 0.0097, "step": 47580 }, { "grad_norm": 0.16431009769439697, "learning_rate": 5.8084542320121483e-05, "loss": 0.008, "step": 47590 }, { "grad_norm": 0.1142321452498436, "learning_rate": 5.8068224752647497e-05, "loss": 0.0057, "step": 47600 }, { "grad_norm": 0.10619568824768066, "learning_rate": 5.805190630284457e-05, "loss": 0.0101, "step": 47610 }, { "grad_norm": 0.10105240345001221, "learning_rate": 5.803558697249729e-05, "loss": 0.0083, "step": 47620 }, { "grad_norm": 0.11870919167995453, "learning_rate": 5.8019266763390276e-05, "loss": 0.0059, "step": 47630 }, { "grad_norm": 0.12563031911849976, "learning_rate": 5.800294567730833e-05, "loss": 0.0067, "step": 47640 }, { "grad_norm": 0.1508501023054123, "learning_rate": 5.798662371603628e-05, "loss": 0.0086, "step": 47650 }, { "grad_norm": 0.10784189403057098, "learning_rate": 5.797030088135904e-05, "loss": 0.0068, "step": 47660 }, { "grad_norm": 0.11985315382480621, "learning_rate": 5.7953977175061705e-05, "loss": 0.0074, "step": 47670 }, { "grad_norm": 0.147902250289917, "learning_rate": 5.793765259892938e-05, "loss": 0.0069, "step": 47680 }, { "grad_norm": 0.2783259153366089, "learning_rate": 5.792132715474729e-05, "loss": 0.0077, "step": 47690 }, { "grad_norm": 0.09688164293766022, "learning_rate": 5.790500084430078e-05, "loss": 0.0066, "step": 47700 }, { "grad_norm": 0.16879573464393616, "learning_rate": 5.7888673669375274e-05, "loss": 0.0079, "step": 47710 }, { "grad_norm": 0.11796946078538895, "learning_rate": 5.787234563175625e-05, "loss": 0.0076, "step": 47720 }, { "grad_norm": 0.13472704589366913, "learning_rate": 5.7856016733229355e-05, "loss": 0.0075, "step": 47730 }, { "grad_norm": 0.16122421622276306, "learning_rate": 5.7839686975580297e-05, "loss": 0.0094, "step": 47740 }, { "grad_norm": 0.09625690430402756, "learning_rate": 5.782335636059484e-05, "loss": 0.0095, "step": 47750 }, { "grad_norm": 0.15136824548244476, "learning_rate": 5.780702489005889e-05, "loss": 0.0071, "step": 47760 }, { "grad_norm": 0.13745072484016418, "learning_rate": 5.779069256575845e-05, "loss": 0.0067, "step": 47770 }, { "grad_norm": 0.08911745250225067, "learning_rate": 5.7774359389479574e-05, "loss": 0.0074, "step": 47780 }, { "grad_norm": 0.15346825122833252, "learning_rate": 5.775802536300845e-05, "loss": 0.0074, "step": 47790 }, { "grad_norm": 0.1680728942155838, "learning_rate": 5.774169048813134e-05, "loss": 0.0084, "step": 47800 }, { "grad_norm": 0.18483249843120575, "learning_rate": 5.77253547666346e-05, "loss": 0.0079, "step": 47810 }, { "grad_norm": 0.14870211482048035, "learning_rate": 5.770901820030465e-05, "loss": 0.0084, "step": 47820 }, { "grad_norm": 0.14689548313617706, "learning_rate": 5.769268079092809e-05, "loss": 0.0079, "step": 47830 }, { "grad_norm": 0.1197199895977974, "learning_rate": 5.767634254029151e-05, "loss": 0.0062, "step": 47840 }, { "grad_norm": 0.2974281907081604, "learning_rate": 5.7660003450181655e-05, "loss": 0.008, "step": 47850 }, { "grad_norm": 0.12000668048858643, "learning_rate": 5.764366352238534e-05, "loss": 0.0077, "step": 47860 }, { "grad_norm": 0.12652994692325592, "learning_rate": 5.7627322758689474e-05, "loss": 0.008, "step": 47870 }, { "grad_norm": 0.15328006446361542, "learning_rate": 5.761098116088105e-05, "loss": 0.0099, "step": 47880 }, { "grad_norm": 0.15727096796035767, "learning_rate": 5.759463873074717e-05, "loss": 0.0095, "step": 47890 }, { "grad_norm": 0.15957792103290558, "learning_rate": 5.757829547007504e-05, "loss": 0.0071, "step": 47900 }, { "grad_norm": 0.13990376889705658, "learning_rate": 5.756195138065189e-05, "loss": 0.0076, "step": 47910 }, { "grad_norm": 0.13638630509376526, "learning_rate": 5.754560646426511e-05, "loss": 0.0079, "step": 47920 }, { "grad_norm": 0.13333368301391602, "learning_rate": 5.752926072270216e-05, "loss": 0.0084, "step": 47930 }, { "grad_norm": 0.15128947794437408, "learning_rate": 5.7512914157750563e-05, "loss": 0.0069, "step": 47940 }, { "grad_norm": 0.10438726097345352, "learning_rate": 5.749656677119798e-05, "loss": 0.0077, "step": 47950 }, { "grad_norm": 0.1555228978395462, "learning_rate": 5.7480218564832125e-05, "loss": 0.0091, "step": 47960 }, { "grad_norm": 0.2002280205488205, "learning_rate": 5.746386954044082e-05, "loss": 0.0108, "step": 47970 }, { "grad_norm": 0.1234133318066597, "learning_rate": 5.744751969981195e-05, "loss": 0.0079, "step": 47980 }, { "grad_norm": 0.14029479026794434, "learning_rate": 5.7431169044733526e-05, "loss": 0.0086, "step": 47990 }, { "grad_norm": 0.1520002782344818, "learning_rate": 5.741481757699364e-05, "loss": 0.0093, "step": 48000 }, { "grad_norm": 0.16255363821983337, "learning_rate": 5.7398465298380434e-05, "loss": 0.0067, "step": 48010 }, { "grad_norm": 0.1213901937007904, "learning_rate": 5.7382112210682193e-05, "loss": 0.0077, "step": 48020 }, { "grad_norm": 0.10520928353071213, "learning_rate": 5.7365758315687266e-05, "loss": 0.0074, "step": 48030 }, { "grad_norm": 0.11705996841192245, "learning_rate": 5.734940361518407e-05, "loss": 0.0078, "step": 48040 }, { "grad_norm": 0.1285971999168396, "learning_rate": 5.733304811096116e-05, "loss": 0.0075, "step": 48050 }, { "grad_norm": 0.11597500741481781, "learning_rate": 5.731669180480713e-05, "loss": 0.0076, "step": 48060 }, { "grad_norm": 0.13734164834022522, "learning_rate": 5.730033469851067e-05, "loss": 0.008, "step": 48070 }, { "grad_norm": 0.09367206692695618, "learning_rate": 5.7283976793860607e-05, "loss": 0.0062, "step": 48080 }, { "grad_norm": 0.13136498630046844, "learning_rate": 5.726761809264577e-05, "loss": 0.0087, "step": 48090 }, { "grad_norm": 0.177344411611557, "learning_rate": 5.7251258596655155e-05, "loss": 0.0091, "step": 48100 }, { "grad_norm": 0.14558497071266174, "learning_rate": 5.72348983076778e-05, "loss": 0.0079, "step": 48110 }, { "grad_norm": 0.08361127227544785, "learning_rate": 5.7218537227502854e-05, "loss": 0.0076, "step": 48120 }, { "grad_norm": 0.10338626056909561, "learning_rate": 5.720217535791951e-05, "loss": 0.0078, "step": 48130 }, { "grad_norm": 0.1347842961549759, "learning_rate": 5.718581270071711e-05, "loss": 0.0067, "step": 48140 }, { "grad_norm": 0.1183113232254982, "learning_rate": 5.716944925768505e-05, "loss": 0.0087, "step": 48150 }, { "grad_norm": 0.13653859496116638, "learning_rate": 5.7153085030612786e-05, "loss": 0.007, "step": 48160 }, { "grad_norm": 0.12444246560335159, "learning_rate": 5.713672002128989e-05, "loss": 0.0072, "step": 48170 }, { "grad_norm": 0.10321735590696335, "learning_rate": 5.712035423150602e-05, "loss": 0.0063, "step": 48180 }, { "grad_norm": 0.11798357963562012, "learning_rate": 5.710398766305094e-05, "loss": 0.0064, "step": 48190 }, { "grad_norm": 0.18389403820037842, "learning_rate": 5.708762031771442e-05, "loss": 0.0079, "step": 48200 }, { "grad_norm": 0.13970839977264404, "learning_rate": 5.707125219728642e-05, "loss": 0.0068, "step": 48210 }, { "grad_norm": 0.11294014751911163, "learning_rate": 5.7054883303556905e-05, "loss": 0.0064, "step": 48220 }, { "grad_norm": 0.1610022336244583, "learning_rate": 5.703851363831595e-05, "loss": 0.0088, "step": 48230 }, { "grad_norm": 0.16459989547729492, "learning_rate": 5.702214320335374e-05, "loss": 0.0091, "step": 48240 }, { "grad_norm": 0.14041918516159058, "learning_rate": 5.7005772000460514e-05, "loss": 0.0079, "step": 48250 }, { "grad_norm": 0.12791946530342102, "learning_rate": 5.698940003142656e-05, "loss": 0.0079, "step": 48260 }, { "grad_norm": 0.08657008409500122, "learning_rate": 5.697302729804236e-05, "loss": 0.0085, "step": 48270 }, { "grad_norm": 0.1399514377117157, "learning_rate": 5.695665380209837e-05, "loss": 0.0075, "step": 48280 }, { "grad_norm": 0.14542868733406067, "learning_rate": 5.6940279545385165e-05, "loss": 0.0074, "step": 48290 }, { "grad_norm": 0.14683844149112701, "learning_rate": 5.692390452969344e-05, "loss": 0.0087, "step": 48300 }, { "grad_norm": 0.19479534029960632, "learning_rate": 5.690752875681392e-05, "loss": 0.0095, "step": 48310 }, { "grad_norm": 0.08798415213823318, "learning_rate": 5.6891152228537435e-05, "loss": 0.0094, "step": 48320 }, { "grad_norm": 0.1580623835325241, "learning_rate": 5.687477494665492e-05, "loss": 0.0068, "step": 48330 }, { "grad_norm": 0.1161065474152565, "learning_rate": 5.685839691295733e-05, "loss": 0.0074, "step": 48340 }, { "grad_norm": 0.1431746631860733, "learning_rate": 5.6842018129235786e-05, "loss": 0.0075, "step": 48350 }, { "grad_norm": 0.2393990457057953, "learning_rate": 5.6825638597281404e-05, "loss": 0.0105, "step": 48360 }, { "grad_norm": 0.11816010624170303, "learning_rate": 5.680925831888546e-05, "loss": 0.0098, "step": 48370 }, { "grad_norm": 0.13504943251609802, "learning_rate": 5.6792877295839274e-05, "loss": 0.0092, "step": 48380 }, { "grad_norm": 0.1408810019493103, "learning_rate": 5.6776495529934224e-05, "loss": 0.009, "step": 48390 }, { "grad_norm": 0.1017107143998146, "learning_rate": 5.6760113022961824e-05, "loss": 0.0091, "step": 48400 }, { "grad_norm": 0.11120456457138062, "learning_rate": 5.6743729776713617e-05, "loss": 0.0086, "step": 48410 }, { "grad_norm": 0.09694799035787582, "learning_rate": 5.672734579298126e-05, "loss": 0.0075, "step": 48420 }, { "grad_norm": 0.13537107408046722, "learning_rate": 5.671096107355649e-05, "loss": 0.0078, "step": 48430 }, { "grad_norm": 0.10873930901288986, "learning_rate": 5.66945756202311e-05, "loss": 0.0093, "step": 48440 }, { "grad_norm": 0.11325094103813171, "learning_rate": 5.667818943479699e-05, "loss": 0.0071, "step": 48450 }, { "grad_norm": 0.13697248697280884, "learning_rate": 5.666180251904612e-05, "loss": 0.0095, "step": 48460 }, { "grad_norm": 0.14612683653831482, "learning_rate": 5.6645414874770555e-05, "loss": 0.007, "step": 48470 }, { "grad_norm": 0.13570517301559448, "learning_rate": 5.66290265037624e-05, "loss": 0.0072, "step": 48480 }, { "grad_norm": 0.13090302050113678, "learning_rate": 5.661263740781386e-05, "loss": 0.0098, "step": 48490 }, { "grad_norm": 0.1456175595521927, "learning_rate": 5.6596247588717254e-05, "loss": 0.0083, "step": 48500 }, { "grad_norm": 0.12177885323762894, "learning_rate": 5.6579857048264926e-05, "loss": 0.0099, "step": 48510 }, { "grad_norm": 0.1389061063528061, "learning_rate": 5.6563465788249314e-05, "loss": 0.0113, "step": 48520 }, { "grad_norm": 0.14968153834342957, "learning_rate": 5.6547073810462956e-05, "loss": 0.011, "step": 48530 }, { "grad_norm": 0.13711552321910858, "learning_rate": 5.653068111669846e-05, "loss": 0.009, "step": 48540 }, { "grad_norm": 0.14449109137058258, "learning_rate": 5.651428770874848e-05, "loss": 0.0079, "step": 48550 }, { "grad_norm": 0.17959173023700714, "learning_rate": 5.64978935884058e-05, "loss": 0.0091, "step": 48560 }, { "grad_norm": 0.12983082234859467, "learning_rate": 5.6481498757463244e-05, "loss": 0.0068, "step": 48570 }, { "grad_norm": 0.1331292986869812, "learning_rate": 5.646510321771373e-05, "loss": 0.008, "step": 48580 }, { "grad_norm": 0.14351382851600647, "learning_rate": 5.644870697095024e-05, "loss": 0.0086, "step": 48590 }, { "grad_norm": 0.14076079428195953, "learning_rate": 5.643231001896586e-05, "loss": 0.007, "step": 48600 }, { "grad_norm": 0.15885160863399506, "learning_rate": 5.6415912363553726e-05, "loss": 0.0108, "step": 48610 }, { "grad_norm": 0.2455519735813141, "learning_rate": 5.639951400650706e-05, "loss": 0.0069, "step": 48620 }, { "grad_norm": 0.16898925602436066, "learning_rate": 5.6383114949619165e-05, "loss": 0.007, "step": 48630 }, { "grad_norm": 0.17230021953582764, "learning_rate": 5.636671519468342e-05, "loss": 0.007, "step": 48640 }, { "grad_norm": 0.17508722841739655, "learning_rate": 5.635031474349327e-05, "loss": 0.0114, "step": 48650 }, { "grad_norm": 0.13008365035057068, "learning_rate": 5.6333913597842246e-05, "loss": 0.008, "step": 48660 }, { "grad_norm": 0.1770496368408203, "learning_rate": 5.6317511759523955e-05, "loss": 0.0081, "step": 48670 }, { "grad_norm": 0.11134500056505203, "learning_rate": 5.630110923033207e-05, "loss": 0.0068, "step": 48680 }, { "grad_norm": 0.11768247932195663, "learning_rate": 5.628470601206036e-05, "loss": 0.0065, "step": 48690 }, { "grad_norm": 0.23032218217849731, "learning_rate": 5.626830210650266e-05, "loss": 0.01, "step": 48700 }, { "grad_norm": 0.11536434292793274, "learning_rate": 5.625189751545285e-05, "loss": 0.0085, "step": 48710 }, { "grad_norm": 0.13631989061832428, "learning_rate": 5.6235492240704936e-05, "loss": 0.0063, "step": 48720 }, { "grad_norm": 0.147165447473526, "learning_rate": 5.621908628405296e-05, "loss": 0.0089, "step": 48730 }, { "grad_norm": 0.13608621060848236, "learning_rate": 5.620267964729106e-05, "loss": 0.0066, "step": 48740 }, { "grad_norm": 0.13422150909900665, "learning_rate": 5.618627233221344e-05, "loss": 0.0077, "step": 48750 }, { "grad_norm": 0.14387944340705872, "learning_rate": 5.616986434061438e-05, "loss": 0.0069, "step": 48760 }, { "grad_norm": 0.1693633794784546, "learning_rate": 5.615345567428822e-05, "loss": 0.0085, "step": 48770 }, { "grad_norm": 0.15290264785289764, "learning_rate": 5.613704633502941e-05, "loss": 0.0094, "step": 48780 }, { "grad_norm": 0.1294773519039154, "learning_rate": 5.612063632463245e-05, "loss": 0.0081, "step": 48790 }, { "grad_norm": 0.18044130504131317, "learning_rate": 5.610422564489188e-05, "loss": 0.0068, "step": 48800 }, { "grad_norm": 0.098409004509449, "learning_rate": 5.608781429760239e-05, "loss": 0.0063, "step": 48810 }, { "grad_norm": 0.14194165170192719, "learning_rate": 5.607140228455866e-05, "loss": 0.0079, "step": 48820 }, { "grad_norm": 0.1595250517129898, "learning_rate": 5.605498960755553e-05, "loss": 0.0087, "step": 48830 }, { "grad_norm": 0.15124164521694183, "learning_rate": 5.603857626838782e-05, "loss": 0.0103, "step": 48840 }, { "grad_norm": 0.149464949965477, "learning_rate": 5.60221622688505e-05, "loss": 0.0096, "step": 48850 }, { "grad_norm": 0.15296436846256256, "learning_rate": 5.6005747610738565e-05, "loss": 0.0087, "step": 48860 }, { "grad_norm": 0.1402866542339325, "learning_rate": 5.59893322958471e-05, "loss": 0.0091, "step": 48870 }, { "grad_norm": 0.18853864073753357, "learning_rate": 5.5972916325971256e-05, "loss": 0.0084, "step": 48880 }, { "grad_norm": 0.15924377739429474, "learning_rate": 5.595649970290628e-05, "loss": 0.0067, "step": 48890 }, { "grad_norm": 0.14069430530071259, "learning_rate": 5.5940082428447426e-05, "loss": 0.0085, "step": 48900 }, { "grad_norm": 0.13863159716129303, "learning_rate": 5.592366450439012e-05, "loss": 0.0065, "step": 48910 }, { "grad_norm": 0.13364318013191223, "learning_rate": 5.590724593252975e-05, "loss": 0.0072, "step": 48920 }, { "grad_norm": 0.15786024928092957, "learning_rate": 5.589082671466184e-05, "loss": 0.0065, "step": 48930 }, { "grad_norm": 0.14246118068695068, "learning_rate": 5.587440685258199e-05, "loss": 0.0109, "step": 48940 }, { "grad_norm": 0.12783083319664001, "learning_rate": 5.585798634808583e-05, "loss": 0.0089, "step": 48950 }, { "grad_norm": 0.13199728727340698, "learning_rate": 5.584156520296909e-05, "loss": 0.0063, "step": 48960 }, { "grad_norm": 0.1543487012386322, "learning_rate": 5.582514341902757e-05, "loss": 0.0079, "step": 48970 }, { "grad_norm": 0.1936434805393219, "learning_rate": 5.580872099805713e-05, "loss": 0.0092, "step": 48980 }, { "grad_norm": 0.10973899811506271, "learning_rate": 5.5792297941853674e-05, "loss": 0.0064, "step": 48990 }, { "grad_norm": 0.15086379647254944, "learning_rate": 5.5775874252213247e-05, "loss": 0.009, "step": 49000 }, { "grad_norm": 0.12336145341396332, "learning_rate": 5.575944993093189e-05, "loss": 0.0065, "step": 49010 }, { "grad_norm": 0.12140180170536041, "learning_rate": 5.574302497980574e-05, "loss": 0.0066, "step": 49020 }, { "grad_norm": 0.10274042934179306, "learning_rate": 5.5726599400631005e-05, "loss": 0.0084, "step": 49030 }, { "grad_norm": 0.13790978491306305, "learning_rate": 5.571017319520399e-05, "loss": 0.0077, "step": 49040 }, { "grad_norm": 0.13302181661128998, "learning_rate": 5.5693746365320984e-05, "loss": 0.0084, "step": 49050 }, { "grad_norm": 0.10485538840293884, "learning_rate": 5.567731891277848e-05, "loss": 0.0062, "step": 49060 }, { "grad_norm": 0.08739285171031952, "learning_rate": 5.566089083937288e-05, "loss": 0.006, "step": 49070 }, { "grad_norm": 0.13646069169044495, "learning_rate": 5.564446214690079e-05, "loss": 0.0119, "step": 49080 }, { "grad_norm": 0.12434166669845581, "learning_rate": 5.56280328371588e-05, "loss": 0.0056, "step": 49090 }, { "grad_norm": 0.10250921547412872, "learning_rate": 5.5611602911943596e-05, "loss": 0.0067, "step": 49100 }, { "grad_norm": 0.19019058346748352, "learning_rate": 5.559517237305194e-05, "loss": 0.0115, "step": 49110 }, { "grad_norm": 0.14020022749900818, "learning_rate": 5.557874122228064e-05, "loss": 0.0055, "step": 49120 }, { "grad_norm": 0.13448020815849304, "learning_rate": 5.55623094614266e-05, "loss": 0.0063, "step": 49130 }, { "grad_norm": 0.1207888051867485, "learning_rate": 5.5545877092286743e-05, "loss": 0.0092, "step": 49140 }, { "grad_norm": 0.13810458779335022, "learning_rate": 5.552944411665811e-05, "loss": 0.0071, "step": 49150 }, { "grad_norm": 0.15845996141433716, "learning_rate": 5.5513010536337795e-05, "loss": 0.0088, "step": 49160 }, { "grad_norm": 0.17364303767681122, "learning_rate": 5.5496576353122934e-05, "loss": 0.0073, "step": 49170 }, { "grad_norm": 0.09325256943702698, "learning_rate": 5.548014156881074e-05, "loss": 0.0065, "step": 49180 }, { "grad_norm": 0.3301217555999756, "learning_rate": 5.5463706185198494e-05, "loss": 0.009, "step": 49190 }, { "grad_norm": 0.1280476152896881, "learning_rate": 5.544727020408358e-05, "loss": 0.0091, "step": 49200 }, { "grad_norm": 0.13560669124126434, "learning_rate": 5.5430833627263366e-05, "loss": 0.0073, "step": 49210 }, { "grad_norm": 0.11818040907382965, "learning_rate": 5.5414396456535354e-05, "loss": 0.0071, "step": 49220 }, { "grad_norm": 0.10240400582551956, "learning_rate": 5.539795869369709e-05, "loss": 0.008, "step": 49230 }, { "grad_norm": 0.1644240915775299, "learning_rate": 5.538152034054618e-05, "loss": 0.0087, "step": 49240 }, { "grad_norm": 0.11562691628932953, "learning_rate": 5.536508139888028e-05, "loss": 0.0092, "step": 49250 }, { "grad_norm": 0.1533733308315277, "learning_rate": 5.534864187049716e-05, "loss": 0.0074, "step": 49260 }, { "grad_norm": 0.13711734116077423, "learning_rate": 5.5332201757194615e-05, "loss": 0.0085, "step": 49270 }, { "grad_norm": 0.13288514316082, "learning_rate": 5.531576106077049e-05, "loss": 0.0068, "step": 49280 }, { "grad_norm": 0.10344425588846207, "learning_rate": 5.529931978302272e-05, "loss": 0.0068, "step": 49290 }, { "grad_norm": 0.1258476972579956, "learning_rate": 5.528287792574932e-05, "loss": 0.0092, "step": 49300 }, { "grad_norm": 0.1841341257095337, "learning_rate": 5.5266435490748324e-05, "loss": 0.0075, "step": 49310 }, { "grad_norm": 0.09431194514036179, "learning_rate": 5.524999247981787e-05, "loss": 0.0062, "step": 49320 }, { "grad_norm": 0.14913338422775269, "learning_rate": 5.523354889475613e-05, "loss": 0.0067, "step": 49330 }, { "grad_norm": 0.1269586831331253, "learning_rate": 5.521710473736134e-05, "loss": 0.0086, "step": 49340 }, { "grad_norm": 0.09348920732736588, "learning_rate": 5.5200660009431826e-05, "loss": 0.0068, "step": 49350 }, { "grad_norm": 0.15746167302131653, "learning_rate": 5.518421471276596e-05, "loss": 0.0068, "step": 49360 }, { "grad_norm": 0.1243584156036377, "learning_rate": 5.5167768849162174e-05, "loss": 0.0078, "step": 49370 }, { "grad_norm": 0.16071906685829163, "learning_rate": 5.515132242041893e-05, "loss": 0.0091, "step": 49380 }, { "grad_norm": 0.22199921309947968, "learning_rate": 5.513487542833483e-05, "loss": 0.0097, "step": 49390 }, { "grad_norm": 0.1614035815000534, "learning_rate": 5.5118427874708466e-05, "loss": 0.0068, "step": 49400 }, { "grad_norm": 0.13974787294864655, "learning_rate": 5.510197976133853e-05, "loss": 0.009, "step": 49410 }, { "grad_norm": 0.10843642055988312, "learning_rate": 5.508553109002376e-05, "loss": 0.0103, "step": 49420 }, { "grad_norm": 0.11055710911750793, "learning_rate": 5.5069081862562957e-05, "loss": 0.0075, "step": 49430 }, { "grad_norm": 0.14514102041721344, "learning_rate": 5.5052632080754965e-05, "loss": 0.0069, "step": 49440 }, { "grad_norm": 0.1367277055978775, "learning_rate": 5.503618174639874e-05, "loss": 0.0063, "step": 49450 }, { "grad_norm": 0.22532965242862701, "learning_rate": 5.501973086129325e-05, "loss": 0.0085, "step": 49460 }, { "grad_norm": 0.11550179868936539, "learning_rate": 5.500327942723753e-05, "loss": 0.0067, "step": 49470 }, { "grad_norm": 0.12590357661247253, "learning_rate": 5.498682744603071e-05, "loss": 0.0075, "step": 49480 }, { "grad_norm": 0.13845817744731903, "learning_rate": 5.497037491947194e-05, "loss": 0.0086, "step": 49490 }, { "grad_norm": 0.16062308847904205, "learning_rate": 5.4953921849360424e-05, "loss": 0.0092, "step": 49500 }, { "grad_norm": 0.14877212047576904, "learning_rate": 5.493746823749547e-05, "loss": 0.009, "step": 49510 }, { "grad_norm": 0.13001511991024017, "learning_rate": 5.49210140856764e-05, "loss": 0.0061, "step": 49520 }, { "grad_norm": 0.11846704035997391, "learning_rate": 5.4904559395702647e-05, "loss": 0.0068, "step": 49530 }, { "grad_norm": 0.10533619672060013, "learning_rate": 5.488810416937364e-05, "loss": 0.0082, "step": 49540 }, { "grad_norm": 0.18694822490215302, "learning_rate": 5.4871648408488926e-05, "loss": 0.008, "step": 49550 }, { "grad_norm": 0.08400654792785645, "learning_rate": 5.485519211484807e-05, "loss": 0.007, "step": 49560 }, { "grad_norm": 0.13223223388195038, "learning_rate": 5.4838735290250676e-05, "loss": 0.0078, "step": 49570 }, { "grad_norm": 0.11413610726594925, "learning_rate": 5.482227793649649e-05, "loss": 0.0076, "step": 49580 }, { "grad_norm": 0.16219833493232727, "learning_rate": 5.480582005538524e-05, "loss": 0.0129, "step": 49590 }, { "grad_norm": 0.11744369566440582, "learning_rate": 5.478936164871671e-05, "loss": 0.0078, "step": 49600 }, { "grad_norm": 0.1493758261203766, "learning_rate": 5.477290271829081e-05, "loss": 0.0079, "step": 49610 }, { "grad_norm": 0.15844105184078217, "learning_rate": 5.475644326590744e-05, "loss": 0.0065, "step": 49620 }, { "grad_norm": 0.13556799292564392, "learning_rate": 5.473998329336658e-05, "loss": 0.0078, "step": 49630 }, { "grad_norm": 0.17883405089378357, "learning_rate": 5.4723522802468286e-05, "loss": 0.0079, "step": 49640 }, { "grad_norm": 0.18987588584423065, "learning_rate": 5.470706179501264e-05, "loss": 0.009, "step": 49650 }, { "grad_norm": 0.1286962777376175, "learning_rate": 5.4690600272799776e-05, "loss": 0.0067, "step": 49660 }, { "grad_norm": 0.11576374620199203, "learning_rate": 5.467413823762993e-05, "loss": 0.0089, "step": 49670 }, { "grad_norm": 0.12357192486524582, "learning_rate": 5.465767569130335e-05, "loss": 0.0079, "step": 49680 }, { "grad_norm": 0.12966622412204742, "learning_rate": 5.464121263562036e-05, "loss": 0.009, "step": 49690 }, { "grad_norm": 0.1253865361213684, "learning_rate": 5.4624749072381343e-05, "loss": 0.0076, "step": 49700 }, { "grad_norm": 0.13437895476818085, "learning_rate": 5.460828500338672e-05, "loss": 0.0083, "step": 49710 }, { "grad_norm": 0.12513461709022522, "learning_rate": 5.459182043043698e-05, "loss": 0.0096, "step": 49720 }, { "grad_norm": 0.12043800950050354, "learning_rate": 5.457535535533265e-05, "loss": 0.0084, "step": 49730 }, { "grad_norm": 0.12335190922021866, "learning_rate": 5.4558889779874334e-05, "loss": 0.0107, "step": 49740 }, { "grad_norm": 0.08606653660535812, "learning_rate": 5.454242370586269e-05, "loss": 0.0077, "step": 49750 }, { "grad_norm": 0.09195342659950256, "learning_rate": 5.452595713509843e-05, "loss": 0.0072, "step": 49760 }, { "grad_norm": 0.16654904186725616, "learning_rate": 5.4509490069382274e-05, "loss": 0.0077, "step": 49770 }, { "grad_norm": 0.14298366010189056, "learning_rate": 5.449302251051509e-05, "loss": 0.007, "step": 49780 }, { "grad_norm": 0.16170912981033325, "learning_rate": 5.44765544602977e-05, "loss": 0.0078, "step": 49790 }, { "grad_norm": 0.10981157422065735, "learning_rate": 5.446008592053107e-05, "loss": 0.0078, "step": 49800 }, { "grad_norm": 0.1623876690864563, "learning_rate": 5.4443616893016135e-05, "loss": 0.0067, "step": 49810 }, { "grad_norm": 0.1393021047115326, "learning_rate": 5.4427147379553934e-05, "loss": 0.0075, "step": 49820 }, { "grad_norm": 0.15846838057041168, "learning_rate": 5.441067738194556e-05, "loss": 0.0073, "step": 49830 }, { "grad_norm": 0.1882946789264679, "learning_rate": 5.439420690199214e-05, "loss": 0.0078, "step": 49840 }, { "grad_norm": 0.1685374528169632, "learning_rate": 5.437773594149484e-05, "loss": 0.0084, "step": 49850 }, { "grad_norm": 0.11916321516036987, "learning_rate": 5.436126450225495e-05, "loss": 0.0062, "step": 49860 }, { "grad_norm": 0.1793537735939026, "learning_rate": 5.434479258607371e-05, "loss": 0.0074, "step": 49870 }, { "grad_norm": 0.13555143773555756, "learning_rate": 5.43283201947525e-05, "loss": 0.0077, "step": 49880 }, { "grad_norm": 0.14161013066768646, "learning_rate": 5.431184733009268e-05, "loss": 0.008, "step": 49890 }, { "grad_norm": 0.15202036499977112, "learning_rate": 5.4295373993895736e-05, "loss": 0.0073, "step": 49900 }, { "grad_norm": 0.11969731003046036, "learning_rate": 5.4278900187963157e-05, "loss": 0.0065, "step": 49910 }, { "grad_norm": 0.10433089733123779, "learning_rate": 5.426242591409646e-05, "loss": 0.0077, "step": 49920 }, { "grad_norm": 0.13523219525814056, "learning_rate": 5.42459511740973e-05, "loss": 0.01, "step": 49930 }, { "grad_norm": 0.16541604697704315, "learning_rate": 5.422947596976729e-05, "loss": 0.0084, "step": 49940 }, { "grad_norm": 0.16542917490005493, "learning_rate": 5.4213000302908134e-05, "loss": 0.0068, "step": 49950 }, { "grad_norm": 0.13731445372104645, "learning_rate": 5.419652417532162e-05, "loss": 0.0069, "step": 49960 }, { "grad_norm": 0.11380593478679657, "learning_rate": 5.4180047588809534e-05, "loss": 0.0059, "step": 49970 }, { "grad_norm": 0.11901726573705673, "learning_rate": 5.4163570545173704e-05, "loss": 0.0075, "step": 49980 }, { "grad_norm": 0.12406362593173981, "learning_rate": 5.414709304621608e-05, "loss": 0.0088, "step": 49990 }, { "grad_norm": 0.1607252061367035, "learning_rate": 5.413061509373858e-05, "loss": 0.0083, "step": 50000 }, { "grad_norm": 0.19111205637454987, "learning_rate": 5.411413668954324e-05, "loss": 0.0128, "step": 50010 }, { "grad_norm": 0.09582927078008652, "learning_rate": 5.4097657835432083e-05, "loss": 0.008, "step": 50020 }, { "grad_norm": 0.20442558825016022, "learning_rate": 5.408117853320723e-05, "loss": 0.0104, "step": 50030 }, { "grad_norm": 0.16016699373722076, "learning_rate": 5.406469878467084e-05, "loss": 0.0083, "step": 50040 }, { "grad_norm": 0.17320694029331207, "learning_rate": 5.404821859162509e-05, "loss": 0.0103, "step": 50050 }, { "grad_norm": 0.12481150776147842, "learning_rate": 5.403173795587225e-05, "loss": 0.0079, "step": 50060 }, { "grad_norm": 0.15211163461208344, "learning_rate": 5.4015256879214606e-05, "loss": 0.007, "step": 50070 }, { "grad_norm": 0.13335540890693665, "learning_rate": 5.39987753634545e-05, "loss": 0.0084, "step": 50080 }, { "grad_norm": 0.1720762848854065, "learning_rate": 5.3982293410394335e-05, "loss": 0.0096, "step": 50090 }, { "grad_norm": 0.08572977781295776, "learning_rate": 5.396581102183655e-05, "loss": 0.0074, "step": 50100 }, { "grad_norm": 0.11642687022686005, "learning_rate": 5.3949328199583634e-05, "loss": 0.0076, "step": 50110 }, { "grad_norm": 0.1394183188676834, "learning_rate": 5.393284494543813e-05, "loss": 0.0104, "step": 50120 }, { "grad_norm": 0.18146444857120514, "learning_rate": 5.391636126120262e-05, "loss": 0.0116, "step": 50130 }, { "grad_norm": 0.14429689943790436, "learning_rate": 5.389987714867971e-05, "loss": 0.0072, "step": 50140 }, { "grad_norm": 0.14535965025424957, "learning_rate": 5.3883392609672114e-05, "loss": 0.0081, "step": 50150 }, { "grad_norm": 0.08366277813911438, "learning_rate": 5.3866907645982545e-05, "loss": 0.0066, "step": 50160 }, { "grad_norm": 0.11458059400320053, "learning_rate": 5.385042225941375e-05, "loss": 0.0079, "step": 50170 }, { "grad_norm": 0.0935371071100235, "learning_rate": 5.3833936451768585e-05, "loss": 0.0072, "step": 50180 }, { "grad_norm": 0.16562412679195404, "learning_rate": 5.3817450224849884e-05, "loss": 0.008, "step": 50190 }, { "grad_norm": 0.168731689453125, "learning_rate": 5.380096358046056e-05, "loss": 0.0076, "step": 50200 }, { "grad_norm": 0.10629768669605255, "learning_rate": 5.378447652040359e-05, "loss": 0.0076, "step": 50210 }, { "grad_norm": 0.13037213683128357, "learning_rate": 5.376798904648195e-05, "loss": 0.0069, "step": 50220 }, { "grad_norm": 0.13407577574253082, "learning_rate": 5.375150116049869e-05, "loss": 0.0066, "step": 50230 }, { "grad_norm": 0.11460094153881073, "learning_rate": 5.373501286425691e-05, "loss": 0.0067, "step": 50240 }, { "grad_norm": 0.15787912905216217, "learning_rate": 5.3718524159559726e-05, "loss": 0.0062, "step": 50250 }, { "grad_norm": 0.17442671954631805, "learning_rate": 5.370203504821034e-05, "loss": 0.0085, "step": 50260 }, { "grad_norm": 0.12892156839370728, "learning_rate": 5.368554553201196e-05, "loss": 0.0084, "step": 50270 }, { "grad_norm": 0.15299850702285767, "learning_rate": 5.3669055612767874e-05, "loss": 0.0086, "step": 50280 }, { "grad_norm": 0.13882391154766083, "learning_rate": 5.3652565292281376e-05, "loss": 0.0067, "step": 50290 }, { "grad_norm": 0.16278403997421265, "learning_rate": 5.3636074572355824e-05, "loss": 0.0075, "step": 50300 }, { "grad_norm": 0.1616508960723877, "learning_rate": 5.361958345479463e-05, "loss": 0.0085, "step": 50310 }, { "grad_norm": 0.1527259200811386, "learning_rate": 5.360309194140123e-05, "loss": 0.0109, "step": 50320 }, { "grad_norm": 0.12498839944601059, "learning_rate": 5.3586600033979106e-05, "loss": 0.0074, "step": 50330 }, { "grad_norm": 0.11105518043041229, "learning_rate": 5.357010773433181e-05, "loss": 0.0076, "step": 50340 }, { "grad_norm": 0.12634681165218353, "learning_rate": 5.3553615044262895e-05, "loss": 0.0064, "step": 50350 }, { "grad_norm": 0.1173352301120758, "learning_rate": 5.3537121965575964e-05, "loss": 0.0084, "step": 50360 }, { "grad_norm": 0.18535147607326508, "learning_rate": 5.352062850007471e-05, "loss": 0.0078, "step": 50370 }, { "grad_norm": 0.09850887209177017, "learning_rate": 5.350413464956282e-05, "loss": 0.0068, "step": 50380 }, { "grad_norm": 0.10853113979101181, "learning_rate": 5.348764041584403e-05, "loss": 0.0068, "step": 50390 }, { "grad_norm": 0.2847137451171875, "learning_rate": 5.347114580072212e-05, "loss": 0.0067, "step": 50400 }, { "grad_norm": 0.12158408015966415, "learning_rate": 5.345465080600093e-05, "loss": 0.0084, "step": 50410 }, { "grad_norm": 0.17801469564437866, "learning_rate": 5.3438155433484326e-05, "loss": 0.0076, "step": 50420 }, { "grad_norm": 0.12571610510349274, "learning_rate": 5.3421659684976197e-05, "loss": 0.0068, "step": 50430 }, { "grad_norm": 0.1538165807723999, "learning_rate": 5.340516356228052e-05, "loss": 0.0104, "step": 50440 }, { "grad_norm": 0.14363306760787964, "learning_rate": 5.338866706720128e-05, "loss": 0.008, "step": 50450 }, { "grad_norm": 0.15092028677463531, "learning_rate": 5.337217020154249e-05, "loss": 0.0072, "step": 50460 }, { "grad_norm": 0.13613218069076538, "learning_rate": 5.335567296710825e-05, "loss": 0.0085, "step": 50470 }, { "grad_norm": 0.11222653836011887, "learning_rate": 5.333917536570265e-05, "loss": 0.0067, "step": 50480 }, { "grad_norm": 0.15562015771865845, "learning_rate": 5.332267739912986e-05, "loss": 0.0079, "step": 50490 }, { "grad_norm": 0.12290617823600769, "learning_rate": 5.330617906919405e-05, "loss": 0.0091, "step": 50500 }, { "grad_norm": 0.09643223881721497, "learning_rate": 5.328968037769949e-05, "loss": 0.0081, "step": 50510 }, { "grad_norm": 0.12186959385871887, "learning_rate": 5.327318132645042e-05, "loss": 0.0057, "step": 50520 }, { "grad_norm": 0.16146555542945862, "learning_rate": 5.325668191725116e-05, "loss": 0.008, "step": 50530 }, { "grad_norm": 0.12948296964168549, "learning_rate": 5.324018215190606e-05, "loss": 0.0064, "step": 50540 }, { "grad_norm": 0.16642239689826965, "learning_rate": 5.3223682032219515e-05, "loss": 0.0078, "step": 50550 }, { "grad_norm": 0.18031810224056244, "learning_rate": 5.320718155999595e-05, "loss": 0.0094, "step": 50560 }, { "grad_norm": 0.16707487404346466, "learning_rate": 5.3190680737039835e-05, "loss": 0.0087, "step": 50570 }, { "grad_norm": 0.11294641345739365, "learning_rate": 5.317417956515567e-05, "loss": 0.008, "step": 50580 }, { "grad_norm": 0.11393539607524872, "learning_rate": 5.315767804614798e-05, "loss": 0.006, "step": 50590 }, { "grad_norm": 0.11719507724046707, "learning_rate": 5.3141176181821395e-05, "loss": 0.008, "step": 50600 }, { "grad_norm": 0.10510504990816116, "learning_rate": 5.31246739739805e-05, "loss": 0.0081, "step": 50610 }, { "grad_norm": 0.12081212550401688, "learning_rate": 5.310817142442994e-05, "loss": 0.0071, "step": 50620 }, { "grad_norm": 0.10618510842323303, "learning_rate": 5.309166853497445e-05, "loss": 0.0083, "step": 50630 }, { "grad_norm": 0.09084063768386841, "learning_rate": 5.307516530741873e-05, "loss": 0.0072, "step": 50640 }, { "grad_norm": 0.11122597754001617, "learning_rate": 5.305866174356754e-05, "loss": 0.0057, "step": 50650 }, { "grad_norm": 0.12371170520782471, "learning_rate": 5.304215784522571e-05, "loss": 0.0062, "step": 50660 }, { "grad_norm": 0.09902053326368332, "learning_rate": 5.302565361419808e-05, "loss": 0.0056, "step": 50670 }, { "grad_norm": 0.10702725499868393, "learning_rate": 5.3009149052289507e-05, "loss": 0.0063, "step": 50680 }, { "grad_norm": 0.09293480217456818, "learning_rate": 5.299264416130493e-05, "loss": 0.0059, "step": 50690 }, { "grad_norm": 0.15293392539024353, "learning_rate": 5.297613894304928e-05, "loss": 0.008, "step": 50700 }, { "grad_norm": 0.13466905057430267, "learning_rate": 5.2959633399327534e-05, "loss": 0.0067, "step": 50710 }, { "grad_norm": 0.1619490087032318, "learning_rate": 5.294312753194476e-05, "loss": 0.0067, "step": 50720 }, { "grad_norm": 0.1251765340566635, "learning_rate": 5.292662134270596e-05, "loss": 0.007, "step": 50730 }, { "grad_norm": 0.10766062885522842, "learning_rate": 5.291011483341626e-05, "loss": 0.0072, "step": 50740 }, { "grad_norm": 0.13106904923915863, "learning_rate": 5.2893608005880767e-05, "loss": 0.0079, "step": 50750 }, { "grad_norm": 0.11639998853206635, "learning_rate": 5.287710086190467e-05, "loss": 0.0081, "step": 50760 }, { "grad_norm": 0.11858316510915756, "learning_rate": 5.286059340329314e-05, "loss": 0.0097, "step": 50770 }, { "grad_norm": 0.1349981278181076, "learning_rate": 5.284408563185141e-05, "loss": 0.0077, "step": 50780 }, { "grad_norm": 0.08584948629140854, "learning_rate": 5.282757754938476e-05, "loss": 0.0067, "step": 50790 }, { "grad_norm": 0.12380402535200119, "learning_rate": 5.281106915769849e-05, "loss": 0.0061, "step": 50800 }, { "grad_norm": 0.13120736181735992, "learning_rate": 5.2794560458597897e-05, "loss": 0.0067, "step": 50810 }, { "grad_norm": 0.08812259882688522, "learning_rate": 5.27780514538884e-05, "loss": 0.0068, "step": 50820 }, { "grad_norm": 0.17139814794063568, "learning_rate": 5.2761542145375365e-05, "loss": 0.0095, "step": 50830 }, { "grad_norm": 0.11508510261774063, "learning_rate": 5.274503253486421e-05, "loss": 0.0074, "step": 50840 }, { "grad_norm": 0.15348775684833527, "learning_rate": 5.272852262416046e-05, "loss": 0.0061, "step": 50850 }, { "grad_norm": 0.13238660991191864, "learning_rate": 5.2712012415069555e-05, "loss": 0.009, "step": 50860 }, { "grad_norm": 0.11560345441102982, "learning_rate": 5.269550190939705e-05, "loss": 0.0076, "step": 50870 }, { "grad_norm": 0.13561707735061646, "learning_rate": 5.267899110894852e-05, "loss": 0.0086, "step": 50880 }, { "grad_norm": 0.15508417785167694, "learning_rate": 5.266248001552955e-05, "loss": 0.0068, "step": 50890 }, { "grad_norm": 0.1322047859430313, "learning_rate": 5.264596863094575e-05, "loss": 0.0086, "step": 50900 }, { "grad_norm": 0.09490656852722168, "learning_rate": 5.262945695700282e-05, "loss": 0.008, "step": 50910 }, { "grad_norm": 0.15099871158599854, "learning_rate": 5.261294499550643e-05, "loss": 0.0082, "step": 50920 }, { "grad_norm": 0.11885476112365723, "learning_rate": 5.25964327482623e-05, "loss": 0.0062, "step": 50930 }, { "grad_norm": 0.11039038747549057, "learning_rate": 5.257992021707617e-05, "loss": 0.0082, "step": 50940 }, { "grad_norm": 0.09268426150083542, "learning_rate": 5.256340740375387e-05, "loss": 0.0074, "step": 50950 }, { "grad_norm": 0.10086830705404282, "learning_rate": 5.254689431010117e-05, "loss": 0.0063, "step": 50960 }, { "grad_norm": 0.1117882952094078, "learning_rate": 5.253038093792395e-05, "loss": 0.0065, "step": 50970 }, { "grad_norm": 0.09813911467790604, "learning_rate": 5.251386728902806e-05, "loss": 0.0061, "step": 50980 }, { "grad_norm": 0.10499216616153717, "learning_rate": 5.2497353365219446e-05, "loss": 0.0063, "step": 50990 }, { "grad_norm": 0.11781337857246399, "learning_rate": 5.2480839168304e-05, "loss": 0.0063, "step": 51000 }, { "grad_norm": 0.12439651042222977, "learning_rate": 5.246432470008773e-05, "loss": 0.0058, "step": 51010 }, { "grad_norm": 0.1481752097606659, "learning_rate": 5.2447809962376616e-05, "loss": 0.0087, "step": 51020 }, { "grad_norm": 0.14972452819347382, "learning_rate": 5.243129495697666e-05, "loss": 0.0074, "step": 51030 }, { "grad_norm": 0.15592354536056519, "learning_rate": 5.241477968569397e-05, "loss": 0.0063, "step": 51040 }, { "grad_norm": 0.0925660952925682, "learning_rate": 5.23982641503346e-05, "loss": 0.0071, "step": 51050 }, { "grad_norm": 0.11609230935573578, "learning_rate": 5.238174835270465e-05, "loss": 0.0088, "step": 51060 }, { "grad_norm": 0.21076710522174835, "learning_rate": 5.23652322946103e-05, "loss": 0.0094, "step": 51070 }, { "grad_norm": 0.1404542624950409, "learning_rate": 5.234871597785771e-05, "loss": 0.0081, "step": 51080 }, { "grad_norm": 0.18875306844711304, "learning_rate": 5.233219940425306e-05, "loss": 0.0059, "step": 51090 }, { "grad_norm": 0.13851258158683777, "learning_rate": 5.2315682575602576e-05, "loss": 0.0103, "step": 51100 }, { "grad_norm": 0.11896408349275589, "learning_rate": 5.229916549371254e-05, "loss": 0.0079, "step": 51110 }, { "grad_norm": 0.1798940747976303, "learning_rate": 5.228264816038924e-05, "loss": 0.0081, "step": 51120 }, { "grad_norm": 0.1173955649137497, "learning_rate": 5.2266130577438935e-05, "loss": 0.0071, "step": 51130 }, { "grad_norm": 0.12474408000707626, "learning_rate": 5.224961274666802e-05, "loss": 0.0067, "step": 51140 }, { "grad_norm": 0.13538521528244019, "learning_rate": 5.223309466988283e-05, "loss": 0.0064, "step": 51150 }, { "grad_norm": 0.07585019618272781, "learning_rate": 5.221657634888976e-05, "loss": 0.0072, "step": 51160 }, { "grad_norm": 0.08323429524898529, "learning_rate": 5.2200057785495235e-05, "loss": 0.0052, "step": 51170 }, { "grad_norm": 0.11919739097356796, "learning_rate": 5.2183538981505706e-05, "loss": 0.0077, "step": 51180 }, { "grad_norm": 0.1479857712984085, "learning_rate": 5.216701993872762e-05, "loss": 0.0084, "step": 51190 }, { "grad_norm": 0.12426641583442688, "learning_rate": 5.215050065896752e-05, "loss": 0.0082, "step": 51200 }, { "grad_norm": 0.11378533393144608, "learning_rate": 5.2133981144031875e-05, "loss": 0.0067, "step": 51210 }, { "grad_norm": 0.1105336919426918, "learning_rate": 5.211746139572727e-05, "loss": 0.0066, "step": 51220 }, { "grad_norm": 0.1525164693593979, "learning_rate": 5.210094141586026e-05, "loss": 0.007, "step": 51230 }, { "grad_norm": 0.18888063728809357, "learning_rate": 5.2084421206237476e-05, "loss": 0.0096, "step": 51240 }, { "grad_norm": 0.13236011564731598, "learning_rate": 5.20679007686655e-05, "loss": 0.0069, "step": 51250 }, { "grad_norm": 0.1409851312637329, "learning_rate": 5.205138010495102e-05, "loss": 0.0083, "step": 51260 }, { "grad_norm": 0.16110263764858246, "learning_rate": 5.20348592169007e-05, "loss": 0.0086, "step": 51270 }, { "grad_norm": 0.13646899163722992, "learning_rate": 5.201833810632123e-05, "loss": 0.0082, "step": 51280 }, { "grad_norm": 0.15424203872680664, "learning_rate": 5.200181677501933e-05, "loss": 0.0062, "step": 51290 }, { "grad_norm": 0.14109091460704803, "learning_rate": 5.1985295224801777e-05, "loss": 0.0074, "step": 51300 }, { "grad_norm": 0.1373022496700287, "learning_rate": 5.196877345747531e-05, "loss": 0.0078, "step": 51310 }, { "grad_norm": 0.1191810667514801, "learning_rate": 5.195225147484674e-05, "loss": 0.0072, "step": 51320 }, { "grad_norm": 0.10785619914531708, "learning_rate": 5.193572927872288e-05, "loss": 0.0064, "step": 51330 }, { "grad_norm": 0.1439443975687027, "learning_rate": 5.191920687091059e-05, "loss": 0.008, "step": 51340 }, { "grad_norm": 0.15259265899658203, "learning_rate": 5.1902684253216695e-05, "loss": 0.0082, "step": 51350 }, { "grad_norm": 0.1340363323688507, "learning_rate": 5.1886161427448134e-05, "loss": 0.0071, "step": 51360 }, { "grad_norm": 0.1570436954498291, "learning_rate": 5.186963839541179e-05, "loss": 0.0078, "step": 51370 }, { "grad_norm": 0.1400882452726364, "learning_rate": 5.185311515891459e-05, "loss": 0.0064, "step": 51380 }, { "grad_norm": 0.09836286306381226, "learning_rate": 5.1836591719763516e-05, "loss": 0.0068, "step": 51390 }, { "grad_norm": 0.12620697915554047, "learning_rate": 5.182006807976554e-05, "loss": 0.007, "step": 51400 }, { "grad_norm": 0.14062468707561493, "learning_rate": 5.180354424072763e-05, "loss": 0.0085, "step": 51410 }, { "grad_norm": 0.17187215387821198, "learning_rate": 5.178702020445686e-05, "loss": 0.0077, "step": 51420 }, { "grad_norm": 0.11201104521751404, "learning_rate": 5.177049597276024e-05, "loss": 0.008, "step": 51430 }, { "grad_norm": 0.1117190569639206, "learning_rate": 5.175397154744482e-05, "loss": 0.0074, "step": 51440 }, { "grad_norm": 0.2359316200017929, "learning_rate": 5.1737446930317736e-05, "loss": 0.0085, "step": 51450 }, { "grad_norm": 0.13730721175670624, "learning_rate": 5.172092212318604e-05, "loss": 0.0069, "step": 51460 }, { "grad_norm": 0.17934125661849976, "learning_rate": 5.170439712785692e-05, "loss": 0.0088, "step": 51470 }, { "grad_norm": 0.11282529681921005, "learning_rate": 5.1687871946137467e-05, "loss": 0.0092, "step": 51480 }, { "grad_norm": 0.11979042738676071, "learning_rate": 5.16713465798349e-05, "loss": 0.0057, "step": 51490 }, { "grad_norm": 0.11983758211135864, "learning_rate": 5.165482103075637e-05, "loss": 0.0076, "step": 51500 }, { "grad_norm": 0.1342083364725113, "learning_rate": 5.163829530070909e-05, "loss": 0.0057, "step": 51510 }, { "grad_norm": 0.17743074893951416, "learning_rate": 5.162176939150032e-05, "loss": 0.0069, "step": 51520 }, { "grad_norm": 0.1362251192331314, "learning_rate": 5.160524330493728e-05, "loss": 0.0082, "step": 51530 }, { "grad_norm": 0.10711490362882614, "learning_rate": 5.158871704282724e-05, "loss": 0.0075, "step": 51540 }, { "grad_norm": 0.13989099860191345, "learning_rate": 5.15721906069775e-05, "loss": 0.0061, "step": 51550 }, { "grad_norm": 0.10045725852251053, "learning_rate": 5.155566399919536e-05, "loss": 0.0078, "step": 51560 }, { "grad_norm": 0.19893397390842438, "learning_rate": 5.153913722128813e-05, "loss": 0.0097, "step": 51570 }, { "grad_norm": 0.16294534504413605, "learning_rate": 5.1522610275063176e-05, "loss": 0.0065, "step": 51580 }, { "grad_norm": 0.16176961362361908, "learning_rate": 5.150608316232786e-05, "loss": 0.0075, "step": 51590 }, { "grad_norm": 0.14382316172122955, "learning_rate": 5.148955588488955e-05, "loss": 0.0097, "step": 51600 }, { "grad_norm": 0.1243729442358017, "learning_rate": 5.147302844455564e-05, "loss": 0.006, "step": 51610 }, { "grad_norm": 0.12493537366390228, "learning_rate": 5.145650084313356e-05, "loss": 0.0069, "step": 51620 }, { "grad_norm": 0.14651349186897278, "learning_rate": 5.1439973082430734e-05, "loss": 0.0073, "step": 51630 }, { "grad_norm": 0.16484902799129486, "learning_rate": 5.142344516425461e-05, "loss": 0.0088, "step": 51640 }, { "grad_norm": 0.10978595912456512, "learning_rate": 5.1406917090412673e-05, "loss": 0.0076, "step": 51650 }, { "grad_norm": 0.1656716763973236, "learning_rate": 5.1390388862712404e-05, "loss": 0.008, "step": 51660 }, { "grad_norm": 0.14643026888370514, "learning_rate": 5.137386048296129e-05, "loss": 0.0071, "step": 51670 }, { "grad_norm": 0.1113663986325264, "learning_rate": 5.135733195296686e-05, "loss": 0.0091, "step": 51680 }, { "grad_norm": 0.11988091468811035, "learning_rate": 5.134080327453664e-05, "loss": 0.0072, "step": 51690 }, { "grad_norm": 0.1469922512769699, "learning_rate": 5.1324274449478204e-05, "loss": 0.0054, "step": 51700 }, { "grad_norm": 0.10783061385154724, "learning_rate": 5.130774547959909e-05, "loss": 0.0061, "step": 51710 }, { "grad_norm": 0.11583391577005386, "learning_rate": 5.129121636670692e-05, "loss": 0.0071, "step": 51720 }, { "grad_norm": 0.13624979555606842, "learning_rate": 5.1274687112609246e-05, "loss": 0.0052, "step": 51730 }, { "grad_norm": 0.12196013331413269, "learning_rate": 5.125815771911372e-05, "loss": 0.008, "step": 51740 }, { "grad_norm": 0.12953370809555054, "learning_rate": 5.1241628188027956e-05, "loss": 0.0077, "step": 51750 }, { "grad_norm": 0.12345239520072937, "learning_rate": 5.1225098521159585e-05, "loss": 0.0073, "step": 51760 }, { "grad_norm": 0.1971556693315506, "learning_rate": 5.12085687203163e-05, "loss": 0.0079, "step": 51770 }, { "grad_norm": 0.1160244420170784, "learning_rate": 5.119203878730575e-05, "loss": 0.0075, "step": 51780 }, { "grad_norm": 0.12442266941070557, "learning_rate": 5.117550872393564e-05, "loss": 0.0061, "step": 51790 }, { "grad_norm": 0.10817144066095352, "learning_rate": 5.1158978532013645e-05, "loss": 0.0098, "step": 51800 }, { "grad_norm": 0.1594957411289215, "learning_rate": 5.114244821334751e-05, "loss": 0.0055, "step": 51810 }, { "grad_norm": 0.1260361522436142, "learning_rate": 5.112591776974497e-05, "loss": 0.009, "step": 51820 }, { "grad_norm": 0.13309071958065033, "learning_rate": 5.110938720301374e-05, "loss": 0.0067, "step": 51830 }, { "grad_norm": 0.11585328727960587, "learning_rate": 5.1092856514961595e-05, "loss": 0.0074, "step": 51840 }, { "grad_norm": 0.26332753896713257, "learning_rate": 5.107632570739631e-05, "loss": 0.0075, "step": 51850 }, { "grad_norm": 0.13771305978298187, "learning_rate": 5.1059794782125645e-05, "loss": 0.0078, "step": 51860 }, { "grad_norm": 0.09979111701250076, "learning_rate": 5.104326374095744e-05, "loss": 0.0062, "step": 51870 }, { "grad_norm": 0.14035268127918243, "learning_rate": 5.102673258569947e-05, "loss": 0.0075, "step": 51880 }, { "grad_norm": 0.12565426528453827, "learning_rate": 5.101020131815956e-05, "loss": 0.0064, "step": 51890 }, { "grad_norm": 0.13395343720912933, "learning_rate": 5.099366994014556e-05, "loss": 0.0072, "step": 51900 }, { "grad_norm": 0.1083076000213623, "learning_rate": 5.097713845346532e-05, "loss": 0.0076, "step": 51910 }, { "grad_norm": 0.10234357416629791, "learning_rate": 5.096060685992668e-05, "loss": 0.0077, "step": 51920 }, { "grad_norm": 0.1084291860461235, "learning_rate": 5.0944075161337514e-05, "loss": 0.0064, "step": 51930 }, { "grad_norm": 0.10647878050804138, "learning_rate": 5.09275433595057e-05, "loss": 0.006, "step": 51940 }, { "grad_norm": 0.09234023094177246, "learning_rate": 5.0911011456239157e-05, "loss": 0.0064, "step": 51950 }, { "grad_norm": 0.11979459971189499, "learning_rate": 5.089447945334574e-05, "loss": 0.0051, "step": 51960 }, { "grad_norm": 0.12508657574653625, "learning_rate": 5.087794735263343e-05, "loss": 0.0055, "step": 51970 }, { "grad_norm": 0.19940689206123352, "learning_rate": 5.086141515591011e-05, "loss": 0.0063, "step": 51980 }, { "grad_norm": 0.11354939639568329, "learning_rate": 5.084488286498371e-05, "loss": 0.0111, "step": 51990 }, { "grad_norm": 0.17137067019939423, "learning_rate": 5.08283504816622e-05, "loss": 0.009, "step": 52000 }, { "grad_norm": 0.12635396420955658, "learning_rate": 5.081181800775353e-05, "loss": 0.0062, "step": 52010 }, { "grad_norm": 0.10646749287843704, "learning_rate": 5.079528544506565e-05, "loss": 0.0082, "step": 52020 }, { "grad_norm": 0.14527466893196106, "learning_rate": 5.077875279540657e-05, "loss": 0.0069, "step": 52030 }, { "grad_norm": 0.11330903321504593, "learning_rate": 5.076222006058424e-05, "loss": 0.0078, "step": 52040 }, { "grad_norm": 0.11947813630104065, "learning_rate": 5.074568724240667e-05, "loss": 0.0062, "step": 52050 }, { "grad_norm": 0.18170705437660217, "learning_rate": 5.072915434268187e-05, "loss": 0.0082, "step": 52060 }, { "grad_norm": 0.12365123629570007, "learning_rate": 5.071262136321785e-05, "loss": 0.0068, "step": 52070 }, { "grad_norm": 0.11417708545923233, "learning_rate": 5.069608830582262e-05, "loss": 0.0062, "step": 52080 }, { "grad_norm": 0.11069494485855103, "learning_rate": 5.067955517230423e-05, "loss": 0.0072, "step": 52090 }, { "grad_norm": 0.13461185991764069, "learning_rate": 5.06630219644707e-05, "loss": 0.007, "step": 52100 }, { "grad_norm": 0.2587152123451233, "learning_rate": 5.064648868413008e-05, "loss": 0.0067, "step": 52110 }, { "grad_norm": 0.11797726154327393, "learning_rate": 5.062995533309044e-05, "loss": 0.0062, "step": 52120 }, { "grad_norm": 0.15416964888572693, "learning_rate": 5.061342191315983e-05, "loss": 0.0071, "step": 52130 }, { "grad_norm": 0.1146247535943985, "learning_rate": 5.0596888426146325e-05, "loss": 0.0063, "step": 52140 }, { "grad_norm": 0.12219896167516708, "learning_rate": 5.058035487385798e-05, "loss": 0.0078, "step": 52150 }, { "grad_norm": 0.15757176280021667, "learning_rate": 5.0563821258102915e-05, "loss": 0.007, "step": 52160 }, { "grad_norm": 0.13464868068695068, "learning_rate": 5.05472875806892e-05, "loss": 0.0062, "step": 52170 }, { "grad_norm": 0.14284932613372803, "learning_rate": 5.053075384342494e-05, "loss": 0.0062, "step": 52180 }, { "grad_norm": 0.17225907742977142, "learning_rate": 5.0514220048118234e-05, "loss": 0.0091, "step": 52190 }, { "grad_norm": 0.13868452608585358, "learning_rate": 5.049768619657721e-05, "loss": 0.008, "step": 52200 }, { "grad_norm": 0.13467513024806976, "learning_rate": 5.048115229060996e-05, "loss": 0.0064, "step": 52210 }, { "grad_norm": 0.16344603896141052, "learning_rate": 5.046461833202462e-05, "loss": 0.0079, "step": 52220 }, { "grad_norm": 0.1676369607448578, "learning_rate": 5.044808432262932e-05, "loss": 0.0078, "step": 52230 }, { "grad_norm": 0.13939815759658813, "learning_rate": 5.043155026423219e-05, "loss": 0.0077, "step": 52240 }, { "grad_norm": 0.11197083443403244, "learning_rate": 5.041501615864137e-05, "loss": 0.0064, "step": 52250 }, { "grad_norm": 0.1169017031788826, "learning_rate": 5.039848200766502e-05, "loss": 0.0063, "step": 52260 }, { "grad_norm": 0.13209956884384155, "learning_rate": 5.038194781311125e-05, "loss": 0.0073, "step": 52270 }, { "grad_norm": 0.10442677140235901, "learning_rate": 5.0365413576788264e-05, "loss": 0.0078, "step": 52280 }, { "grad_norm": 0.13377584517002106, "learning_rate": 5.0348879300504194e-05, "loss": 0.0069, "step": 52290 }, { "grad_norm": 0.17124654352664948, "learning_rate": 5.033234498606721e-05, "loss": 0.0092, "step": 52300 }, { "grad_norm": 0.1091776192188263, "learning_rate": 5.0315810635285476e-05, "loss": 0.0067, "step": 52310 }, { "grad_norm": 0.13357208669185638, "learning_rate": 5.0299276249967164e-05, "loss": 0.0068, "step": 52320 }, { "grad_norm": 0.13185112178325653, "learning_rate": 5.0282741831920454e-05, "loss": 0.0073, "step": 52330 }, { "grad_norm": 0.1350032538175583, "learning_rate": 5.026620738295351e-05, "loss": 0.0075, "step": 52340 }, { "grad_norm": 0.09269687533378601, "learning_rate": 5.024967290487455e-05, "loss": 0.007, "step": 52350 }, { "grad_norm": 0.09848898649215698, "learning_rate": 5.023313839949172e-05, "loss": 0.0095, "step": 52360 }, { "grad_norm": 0.133492112159729, "learning_rate": 5.0216603868613224e-05, "loss": 0.0072, "step": 52370 }, { "grad_norm": 0.08752767741680145, "learning_rate": 5.020006931404726e-05, "loss": 0.0069, "step": 52380 }, { "grad_norm": 0.13083282113075256, "learning_rate": 5.0183534737602014e-05, "loss": 0.0054, "step": 52390 }, { "grad_norm": 0.08860697597265244, "learning_rate": 5.0167000141085686e-05, "loss": 0.0056, "step": 52400 }, { "grad_norm": 0.09714443236589432, "learning_rate": 5.015046552630647e-05, "loss": 0.0123, "step": 52410 }, { "grad_norm": 0.10933004319667816, "learning_rate": 5.013393089507258e-05, "loss": 0.0089, "step": 52420 }, { "grad_norm": 0.13603171706199646, "learning_rate": 5.011739624919219e-05, "loss": 0.0082, "step": 52430 }, { "grad_norm": 0.13074623048305511, "learning_rate": 5.0100861590473545e-05, "loss": 0.007, "step": 52440 }, { "grad_norm": 0.08546943962574005, "learning_rate": 5.008432692072482e-05, "loss": 0.0091, "step": 52450 }, { "grad_norm": 0.10538817197084427, "learning_rate": 5.006779224175423e-05, "loss": 0.0082, "step": 52460 }, { "grad_norm": 0.1229468509554863, "learning_rate": 5.005125755536998e-05, "loss": 0.0086, "step": 52470 }, { "grad_norm": 0.12458381801843643, "learning_rate": 5.003472286338029e-05, "loss": 0.0079, "step": 52480 }, { "grad_norm": 0.1840001344680786, "learning_rate": 5.001818816759335e-05, "loss": 0.0091, "step": 52490 }, { "grad_norm": 0.1786855012178421, "learning_rate": 5.000165346981738e-05, "loss": 0.0083, "step": 52500 }, { "grad_norm": 0.1578131765127182, "learning_rate": 4.9985118771860595e-05, "loss": 0.0082, "step": 52510 }, { "grad_norm": 0.14339007437229156, "learning_rate": 4.996858407553119e-05, "loss": 0.0069, "step": 52520 }, { "grad_norm": 0.11695558577775955, "learning_rate": 4.99520493826374e-05, "loss": 0.0072, "step": 52530 }, { "grad_norm": 0.14411018788814545, "learning_rate": 4.9935514694987384e-05, "loss": 0.0062, "step": 52540 }, { "grad_norm": 0.12516486644744873, "learning_rate": 4.991898001438941e-05, "loss": 0.0078, "step": 52550 }, { "grad_norm": 0.1671990007162094, "learning_rate": 4.990244534265166e-05, "loss": 0.0087, "step": 52560 }, { "grad_norm": 0.10938788950443268, "learning_rate": 4.988591068158234e-05, "loss": 0.0065, "step": 52570 }, { "grad_norm": 0.13258646428585052, "learning_rate": 4.986937603298965e-05, "loss": 0.0071, "step": 52580 }, { "grad_norm": 0.12366943806409836, "learning_rate": 4.985284139868181e-05, "loss": 0.0065, "step": 52590 }, { "grad_norm": 0.12117046862840652, "learning_rate": 4.9836306780467e-05, "loss": 0.0082, "step": 52600 }, { "grad_norm": 0.1042405366897583, "learning_rate": 4.981977218015343e-05, "loss": 0.0085, "step": 52610 }, { "grad_norm": 0.10620473325252533, "learning_rate": 4.980323759954932e-05, "loss": 0.0063, "step": 52620 }, { "grad_norm": 0.1400187909603119, "learning_rate": 4.978670304046284e-05, "loss": 0.0065, "step": 52630 }, { "grad_norm": 0.12237117439508438, "learning_rate": 4.9770168504702195e-05, "loss": 0.0058, "step": 52640 }, { "grad_norm": 0.07922912389039993, "learning_rate": 4.975363399407559e-05, "loss": 0.0055, "step": 52650 }, { "grad_norm": 0.11901216208934784, "learning_rate": 4.9737099510391194e-05, "loss": 0.005, "step": 52660 }, { "grad_norm": 0.11293372511863708, "learning_rate": 4.972056505545719e-05, "loss": 0.0055, "step": 52670 }, { "grad_norm": 0.09692545980215073, "learning_rate": 4.9704030631081775e-05, "loss": 0.0079, "step": 52680 }, { "grad_norm": 0.09737340360879898, "learning_rate": 4.968749623907313e-05, "loss": 0.005, "step": 52690 }, { "grad_norm": 0.1148303672671318, "learning_rate": 4.967096188123943e-05, "loss": 0.0066, "step": 52700 }, { "grad_norm": 0.14067129790782928, "learning_rate": 4.965442755938884e-05, "loss": 0.0077, "step": 52710 }, { "grad_norm": 0.13307301700115204, "learning_rate": 4.963789327532954e-05, "loss": 0.0086, "step": 52720 }, { "grad_norm": 0.12121331691741943, "learning_rate": 4.962135903086968e-05, "loss": 0.0056, "step": 52730 }, { "grad_norm": 0.1063932254910469, "learning_rate": 4.960482482781741e-05, "loss": 0.0083, "step": 52740 }, { "grad_norm": 0.17805582284927368, "learning_rate": 4.958829066798092e-05, "loss": 0.0065, "step": 52750 }, { "grad_norm": 0.1447032392024994, "learning_rate": 4.957175655316835e-05, "loss": 0.0113, "step": 52760 }, { "grad_norm": 0.18256881833076477, "learning_rate": 4.9555222485187834e-05, "loss": 0.0078, "step": 52770 }, { "grad_norm": 0.09333159774541855, "learning_rate": 4.953868846584752e-05, "loss": 0.0055, "step": 52780 }, { "grad_norm": 0.13689900934696198, "learning_rate": 4.9522154496955544e-05, "loss": 0.0052, "step": 52790 }, { "grad_norm": 0.1479947865009308, "learning_rate": 4.950562058032001e-05, "loss": 0.0088, "step": 52800 }, { "grad_norm": 0.11049128323793411, "learning_rate": 4.948908671774909e-05, "loss": 0.0074, "step": 52810 }, { "grad_norm": 0.11134988069534302, "learning_rate": 4.947255291105087e-05, "loss": 0.0075, "step": 52820 }, { "grad_norm": 0.08634460717439651, "learning_rate": 4.945601916203348e-05, "loss": 0.0087, "step": 52830 }, { "grad_norm": 0.12881718575954437, "learning_rate": 4.9439485472505016e-05, "loss": 0.0066, "step": 52840 }, { "grad_norm": 0.09066595882177353, "learning_rate": 4.942295184427356e-05, "loss": 0.0071, "step": 52850 }, { "grad_norm": 0.15608565509319305, "learning_rate": 4.9406418279147234e-05, "loss": 0.0063, "step": 52860 }, { "grad_norm": 0.1262647658586502, "learning_rate": 4.9389884778934116e-05, "loss": 0.0075, "step": 52870 }, { "grad_norm": 0.10713599622249603, "learning_rate": 4.937335134544228e-05, "loss": 0.0072, "step": 52880 }, { "grad_norm": 0.1075330302119255, "learning_rate": 4.935681798047979e-05, "loss": 0.0075, "step": 52890 }, { "grad_norm": 0.10291044414043427, "learning_rate": 4.934028468585473e-05, "loss": 0.0051, "step": 52900 }, { "grad_norm": 0.13555660843849182, "learning_rate": 4.932375146337514e-05, "loss": 0.01, "step": 52910 }, { "grad_norm": 0.11151960492134094, "learning_rate": 4.930721831484908e-05, "loss": 0.0096, "step": 52920 }, { "grad_norm": 0.19518722593784332, "learning_rate": 4.929068524208456e-05, "loss": 0.0091, "step": 52930 }, { "grad_norm": 0.1385420262813568, "learning_rate": 4.927415224688965e-05, "loss": 0.0071, "step": 52940 }, { "grad_norm": 0.11385916918516159, "learning_rate": 4.9257619331072375e-05, "loss": 0.0078, "step": 52950 }, { "grad_norm": 0.11113055795431137, "learning_rate": 4.924108649644072e-05, "loss": 0.0058, "step": 52960 }, { "grad_norm": 0.16233979165554047, "learning_rate": 4.922455374480271e-05, "loss": 0.0075, "step": 52970 }, { "grad_norm": 0.09627413749694824, "learning_rate": 4.920802107796634e-05, "loss": 0.0075, "step": 52980 }, { "grad_norm": 0.16240708529949188, "learning_rate": 4.9191488497739564e-05, "loss": 0.0075, "step": 52990 }, { "grad_norm": 0.10715142637491226, "learning_rate": 4.917495600593042e-05, "loss": 0.006, "step": 53000 }, { "grad_norm": 0.12826228141784668, "learning_rate": 4.915842360434684e-05, "loss": 0.0075, "step": 53010 }, { "grad_norm": 0.1316259503364563, "learning_rate": 4.914189129479679e-05, "loss": 0.0077, "step": 53020 }, { "grad_norm": 0.13060465455055237, "learning_rate": 4.912535907908822e-05, "loss": 0.0069, "step": 53030 }, { "grad_norm": 0.14142903685569763, "learning_rate": 4.910882695902907e-05, "loss": 0.0089, "step": 53040 }, { "grad_norm": 0.13490213453769684, "learning_rate": 4.9092294936427234e-05, "loss": 0.0077, "step": 53050 }, { "grad_norm": 0.1611090749502182, "learning_rate": 4.9075763013090685e-05, "loss": 0.0094, "step": 53060 }, { "grad_norm": 0.10399552434682846, "learning_rate": 4.9059231190827305e-05, "loss": 0.0081, "step": 53070 }, { "grad_norm": 0.1613408327102661, "learning_rate": 4.904269947144499e-05, "loss": 0.0064, "step": 53080 }, { "grad_norm": 0.11831830441951752, "learning_rate": 4.902616785675161e-05, "loss": 0.0135, "step": 53090 }, { "grad_norm": 0.1724509745836258, "learning_rate": 4.900963634855505e-05, "loss": 0.0074, "step": 53100 }, { "grad_norm": 0.16576571762561798, "learning_rate": 4.8993104948663173e-05, "loss": 0.0073, "step": 53110 }, { "grad_norm": 0.11214526742696762, "learning_rate": 4.8976573658883823e-05, "loss": 0.0067, "step": 53120 }, { "grad_norm": 0.1801733374595642, "learning_rate": 4.896004248102484e-05, "loss": 0.0089, "step": 53130 }, { "grad_norm": 0.1760759949684143, "learning_rate": 4.8943511416894045e-05, "loss": 0.0069, "step": 53140 }, { "grad_norm": 0.0977073460817337, "learning_rate": 4.892698046829926e-05, "loss": 0.0061, "step": 53150 }, { "grad_norm": 0.10747169703245163, "learning_rate": 4.8910449637048275e-05, "loss": 0.0067, "step": 53160 }, { "grad_norm": 0.1378851681947708, "learning_rate": 4.8893918924948884e-05, "loss": 0.0087, "step": 53170 }, { "grad_norm": 0.17493726313114166, "learning_rate": 4.8877388333808835e-05, "loss": 0.007, "step": 53180 }, { "grad_norm": 0.11504347622394562, "learning_rate": 4.8860857865435925e-05, "loss": 0.008, "step": 53190 }, { "grad_norm": 0.1370222270488739, "learning_rate": 4.884432752163789e-05, "loss": 0.007, "step": 53200 }, { "grad_norm": 0.19050757586956024, "learning_rate": 4.8827797304222466e-05, "loss": 0.0074, "step": 53210 }, { "grad_norm": 0.12378186732530594, "learning_rate": 4.8811267214997366e-05, "loss": 0.0067, "step": 53220 }, { "grad_norm": 0.09707044810056686, "learning_rate": 4.879473725577029e-05, "loss": 0.0055, "step": 53230 }, { "grad_norm": 0.1400015503168106, "learning_rate": 4.877820742834891e-05, "loss": 0.0074, "step": 53240 }, { "grad_norm": 0.11454357206821442, "learning_rate": 4.8761677734540956e-05, "loss": 0.0049, "step": 53250 }, { "grad_norm": 0.12145312130451202, "learning_rate": 4.874514817615407e-05, "loss": 0.0066, "step": 53260 }, { "grad_norm": 0.1679210513830185, "learning_rate": 4.872861875499588e-05, "loss": 0.0071, "step": 53270 }, { "grad_norm": 0.13445423543453217, "learning_rate": 4.871208947287404e-05, "loss": 0.0064, "step": 53280 }, { "grad_norm": 0.15966767072677612, "learning_rate": 4.869556033159615e-05, "loss": 0.0078, "step": 53290 }, { "grad_norm": 0.10208921879529953, "learning_rate": 4.8679031332969835e-05, "loss": 0.0079, "step": 53300 }, { "grad_norm": 0.15081234276294708, "learning_rate": 4.8662502478802635e-05, "loss": 0.0079, "step": 53310 }, { "grad_norm": 0.13973331451416016, "learning_rate": 4.8645973770902176e-05, "loss": 0.0074, "step": 53320 }, { "grad_norm": 0.13564322888851166, "learning_rate": 4.862944521107599e-05, "loss": 0.0095, "step": 53330 }, { "grad_norm": 0.1257297247648239, "learning_rate": 4.861291680113162e-05, "loss": 0.0082, "step": 53340 }, { "grad_norm": 0.13361774384975433, "learning_rate": 4.8596388542876563e-05, "loss": 0.0078, "step": 53350 }, { "grad_norm": 0.15987682342529297, "learning_rate": 4.857986043811836e-05, "loss": 0.0074, "step": 53360 }, { "grad_norm": 0.12697061896324158, "learning_rate": 4.856333248866448e-05, "loss": 0.009, "step": 53370 }, { "grad_norm": 0.1054743304848671, "learning_rate": 4.85468046963224e-05, "loss": 0.0078, "step": 53380 }, { "grad_norm": 0.12172853201627731, "learning_rate": 4.853027706289957e-05, "loss": 0.0075, "step": 53390 }, { "grad_norm": 0.11279202252626419, "learning_rate": 4.851374959020344e-05, "loss": 0.0073, "step": 53400 }, { "grad_norm": 0.11679842323064804, "learning_rate": 4.849722228004141e-05, "loss": 0.0068, "step": 53410 }, { "grad_norm": 0.10288651287555695, "learning_rate": 4.848069513422091e-05, "loss": 0.0086, "step": 53420 }, { "grad_norm": 0.2064804583787918, "learning_rate": 4.8464168154549294e-05, "loss": 0.0092, "step": 53430 }, { "grad_norm": 0.23722675442695618, "learning_rate": 4.844764134283392e-05, "loss": 0.0114, "step": 53440 }, { "grad_norm": 0.1318385750055313, "learning_rate": 4.843111470088218e-05, "loss": 0.0076, "step": 53450 }, { "grad_norm": 0.1660161167383194, "learning_rate": 4.841458823050137e-05, "loss": 0.0073, "step": 53460 }, { "grad_norm": 0.14358660578727722, "learning_rate": 4.8398061933498816e-05, "loss": 0.0078, "step": 53470 }, { "grad_norm": 0.09126182645559311, "learning_rate": 4.83815358116818e-05, "loss": 0.007, "step": 53480 }, { "grad_norm": 0.11463213711977005, "learning_rate": 4.8365009866857584e-05, "loss": 0.0056, "step": 53490 }, { "grad_norm": 0.0958278551697731, "learning_rate": 4.834848410083342e-05, "loss": 0.0069, "step": 53500 }, { "grad_norm": 0.14327451586723328, "learning_rate": 4.833195851541657e-05, "loss": 0.0073, "step": 53510 }, { "grad_norm": 0.1509966105222702, "learning_rate": 4.831543311241423e-05, "loss": 0.0065, "step": 53520 }, { "grad_norm": 0.16637857258319855, "learning_rate": 4.829890789363359e-05, "loss": 0.0082, "step": 53530 }, { "grad_norm": 0.1735903024673462, "learning_rate": 4.828238286088182e-05, "loss": 0.0066, "step": 53540 }, { "grad_norm": 0.12929068505764008, "learning_rate": 4.826585801596609e-05, "loss": 0.0097, "step": 53550 }, { "grad_norm": 0.17391465604305267, "learning_rate": 4.8249333360693486e-05, "loss": 0.0087, "step": 53560 }, { "grad_norm": 0.10034499317407608, "learning_rate": 4.8232808896871185e-05, "loss": 0.0047, "step": 53570 }, { "grad_norm": 0.13791514933109283, "learning_rate": 4.821628462630624e-05, "loss": 0.0072, "step": 53580 }, { "grad_norm": 0.09358048439025879, "learning_rate": 4.8199760550805735e-05, "loss": 0.0066, "step": 53590 }, { "grad_norm": 0.12265826761722565, "learning_rate": 4.818323667217669e-05, "loss": 0.0062, "step": 53600 }, { "grad_norm": 0.1364860087633133, "learning_rate": 4.8166712992226165e-05, "loss": 0.0063, "step": 53610 }, { "grad_norm": 0.11773181706666946, "learning_rate": 4.8150189512761154e-05, "loss": 0.0072, "step": 53620 }, { "grad_norm": 0.15646864473819733, "learning_rate": 4.813366623558862e-05, "loss": 0.0077, "step": 53630 }, { "grad_norm": 0.09519819915294647, "learning_rate": 4.811714316251554e-05, "loss": 0.0053, "step": 53640 }, { "grad_norm": 0.07692355662584305, "learning_rate": 4.810062029534886e-05, "loss": 0.0077, "step": 53650 }, { "grad_norm": 0.12991270422935486, "learning_rate": 4.8084097635895505e-05, "loss": 0.0111, "step": 53660 }, { "grad_norm": 0.12423983216285706, "learning_rate": 4.8067575185962335e-05, "loss": 0.0079, "step": 53670 }, { "grad_norm": 0.12478872388601303, "learning_rate": 4.805105294735625e-05, "loss": 0.007, "step": 53680 }, { "grad_norm": 0.11986833065748215, "learning_rate": 4.803453092188406e-05, "loss": 0.0056, "step": 53690 }, { "grad_norm": 0.11346438527107239, "learning_rate": 4.801800911135263e-05, "loss": 0.0066, "step": 53700 }, { "grad_norm": 0.18205782771110535, "learning_rate": 4.8001487517568744e-05, "loss": 0.0065, "step": 53710 }, { "grad_norm": 0.14526955783367157, "learning_rate": 4.798496614233918e-05, "loss": 0.0082, "step": 53720 }, { "grad_norm": 0.18978995084762573, "learning_rate": 4.796844498747069e-05, "loss": 0.01, "step": 53730 }, { "grad_norm": 0.13871893286705017, "learning_rate": 4.795192405477e-05, "loss": 0.0104, "step": 53740 }, { "grad_norm": 0.1183917298913002, "learning_rate": 4.79354033460438e-05, "loss": 0.0058, "step": 53750 }, { "grad_norm": 0.10408677160739899, "learning_rate": 4.79188828630988e-05, "loss": 0.006, "step": 53760 }, { "grad_norm": 0.12181583791971207, "learning_rate": 4.790236260774165e-05, "loss": 0.0071, "step": 53770 }, { "grad_norm": 0.13740800321102142, "learning_rate": 4.788584258177896e-05, "loss": 0.0071, "step": 53780 }, { "grad_norm": 0.14255107939243317, "learning_rate": 4.7869322787017356e-05, "loss": 0.0063, "step": 53790 }, { "grad_norm": 0.169157475233078, "learning_rate": 4.78528032252634e-05, "loss": 0.0071, "step": 53800 }, { "grad_norm": 0.15670707821846008, "learning_rate": 4.783628389832366e-05, "loss": 0.0077, "step": 53810 }, { "grad_norm": 0.23784461617469788, "learning_rate": 4.7819764808004644e-05, "loss": 0.0082, "step": 53820 }, { "grad_norm": 0.1348322480916977, "learning_rate": 4.780324595611289e-05, "loss": 0.007, "step": 53830 }, { "grad_norm": 0.11521773785352707, "learning_rate": 4.7786727344454854e-05, "loss": 0.0063, "step": 53840 }, { "grad_norm": 0.1190447062253952, "learning_rate": 4.777020897483697e-05, "loss": 0.0074, "step": 53850 }, { "grad_norm": 0.12102323025465012, "learning_rate": 4.775369084906569e-05, "loss": 0.0055, "step": 53860 }, { "grad_norm": 0.14333541691303253, "learning_rate": 4.773717296894742e-05, "loss": 0.0068, "step": 53870 }, { "grad_norm": 0.11432825773954391, "learning_rate": 4.772065533628848e-05, "loss": 0.0072, "step": 53880 }, { "grad_norm": 0.13287389278411865, "learning_rate": 4.7704137952895265e-05, "loss": 0.0063, "step": 53890 }, { "grad_norm": 0.13957537710666656, "learning_rate": 4.768762082057407e-05, "loss": 0.0098, "step": 53900 }, { "grad_norm": 0.13183194398880005, "learning_rate": 4.7671103941131186e-05, "loss": 0.0062, "step": 53910 }, { "grad_norm": 0.1393086314201355, "learning_rate": 4.765458731637287e-05, "loss": 0.0084, "step": 53920 }, { "grad_norm": 0.14046300947666168, "learning_rate": 4.763807094810537e-05, "loss": 0.0056, "step": 53930 }, { "grad_norm": 0.12010595202445984, "learning_rate": 4.762155483813487e-05, "loss": 0.0048, "step": 53940 }, { "grad_norm": 0.1294916272163391, "learning_rate": 4.760503898826754e-05, "loss": 0.007, "step": 53950 }, { "grad_norm": 0.16496455669403076, "learning_rate": 4.758852340030957e-05, "loss": 0.0079, "step": 53960 }, { "grad_norm": 0.09206386655569077, "learning_rate": 4.757200807606705e-05, "loss": 0.0061, "step": 53970 }, { "grad_norm": 0.13558271527290344, "learning_rate": 4.755549301734607e-05, "loss": 0.008, "step": 53980 }, { "grad_norm": 0.11946938186883926, "learning_rate": 4.753897822595271e-05, "loss": 0.0063, "step": 53990 }, { "grad_norm": 0.11375442892313004, "learning_rate": 4.7522463703692994e-05, "loss": 0.0058, "step": 54000 }, { "grad_norm": 0.08750802278518677, "learning_rate": 4.75059494523729e-05, "loss": 0.0071, "step": 54010 }, { "grad_norm": 0.17202876508235931, "learning_rate": 4.748943547379843e-05, "loss": 0.007, "step": 54020 }, { "grad_norm": 0.09914718568325043, "learning_rate": 4.7472921769775535e-05, "loss": 0.0062, "step": 54030 }, { "grad_norm": 0.12416479736566544, "learning_rate": 4.745640834211011e-05, "loss": 0.0052, "step": 54040 }, { "grad_norm": 0.09234107285737991, "learning_rate": 4.7439895192608054e-05, "loss": 0.008, "step": 54050 }, { "grad_norm": 0.1176227480173111, "learning_rate": 4.7423382323075186e-05, "loss": 0.0058, "step": 54060 }, { "grad_norm": 0.10817384719848633, "learning_rate": 4.740686973531737e-05, "loss": 0.006, "step": 54070 }, { "grad_norm": 0.14694122970104218, "learning_rate": 4.7390357431140375e-05, "loss": 0.0075, "step": 54080 }, { "grad_norm": 0.09951192140579224, "learning_rate": 4.737384541234998e-05, "loss": 0.007, "step": 54090 }, { "grad_norm": 0.12672829627990723, "learning_rate": 4.735733368075189e-05, "loss": 0.0061, "step": 54100 }, { "grad_norm": 0.11128879338502884, "learning_rate": 4.734082223815182e-05, "loss": 0.0073, "step": 54110 }, { "grad_norm": 0.090122289955616, "learning_rate": 4.7324311086355436e-05, "loss": 0.006, "step": 54120 }, { "grad_norm": 0.14672377705574036, "learning_rate": 4.730780022716836e-05, "loss": 0.0066, "step": 54130 }, { "grad_norm": 0.12785866856575012, "learning_rate": 4.729128966239619e-05, "loss": 0.0066, "step": 54140 }, { "grad_norm": 0.24347704648971558, "learning_rate": 4.7274779393844536e-05, "loss": 0.0091, "step": 54150 }, { "grad_norm": 0.11557506024837494, "learning_rate": 4.725826942331891e-05, "loss": 0.008, "step": 54160 }, { "grad_norm": 0.15680387616157532, "learning_rate": 4.724175975262481e-05, "loss": 0.0061, "step": 54170 }, { "grad_norm": 0.09526464343070984, "learning_rate": 4.722525038356772e-05, "loss": 0.0067, "step": 54180 }, { "grad_norm": 0.20313046872615814, "learning_rate": 4.7208741317953075e-05, "loss": 0.0058, "step": 54190 }, { "grad_norm": 0.09219371527433395, "learning_rate": 4.7192232557586264e-05, "loss": 0.0051, "step": 54200 }, { "grad_norm": 0.12377218902111053, "learning_rate": 4.717572410427269e-05, "loss": 0.0073, "step": 54210 }, { "grad_norm": 0.12116695940494537, "learning_rate": 4.71592159598177e-05, "loss": 0.0062, "step": 54220 }, { "grad_norm": 0.12214083969593048, "learning_rate": 4.714270812602657e-05, "loss": 0.0065, "step": 54230 }, { "grad_norm": 0.17843063175678253, "learning_rate": 4.712620060470458e-05, "loss": 0.0076, "step": 54240 }, { "grad_norm": 0.13311335444450378, "learning_rate": 4.7109693397656986e-05, "loss": 0.0067, "step": 54250 }, { "grad_norm": 0.11317678540945053, "learning_rate": 4.709318650668894e-05, "loss": 0.0061, "step": 54260 }, { "grad_norm": 0.1406688392162323, "learning_rate": 4.7076679933605676e-05, "loss": 0.0081, "step": 54270 }, { "grad_norm": 0.19480472803115845, "learning_rate": 4.70601736802123e-05, "loss": 0.0078, "step": 54280 }, { "grad_norm": 0.1917324662208557, "learning_rate": 4.704366774831391e-05, "loss": 0.0078, "step": 54290 }, { "grad_norm": 0.10955515503883362, "learning_rate": 4.702716213971557e-05, "loss": 0.0084, "step": 54300 }, { "grad_norm": 0.11272890865802765, "learning_rate": 4.70106568562223e-05, "loss": 0.0079, "step": 54310 }, { "grad_norm": 0.15063625574111938, "learning_rate": 4.6994151899639105e-05, "loss": 0.0077, "step": 54320 }, { "grad_norm": 0.14105038344860077, "learning_rate": 4.697764727177093e-05, "loss": 0.0054, "step": 54330 }, { "grad_norm": 0.11628986895084381, "learning_rate": 4.696114297442272e-05, "loss": 0.0069, "step": 54340 }, { "grad_norm": 0.15891161561012268, "learning_rate": 4.694463900939933e-05, "loss": 0.0087, "step": 54350 }, { "grad_norm": 0.14643998444080353, "learning_rate": 4.692813537850564e-05, "loss": 0.0072, "step": 54360 }, { "grad_norm": 0.15334174036979675, "learning_rate": 4.6911632083546445e-05, "loss": 0.0065, "step": 54370 }, { "grad_norm": 0.0895380973815918, "learning_rate": 4.6895129126326525e-05, "loss": 0.0059, "step": 54380 }, { "grad_norm": 0.09333202242851257, "learning_rate": 4.687862650865059e-05, "loss": 0.0061, "step": 54390 }, { "grad_norm": 0.10127297788858414, "learning_rate": 4.686212423232339e-05, "loss": 0.0069, "step": 54400 }, { "grad_norm": 0.12061833590269089, "learning_rate": 4.6845622299149565e-05, "loss": 0.0058, "step": 54410 }, { "grad_norm": 0.12554576992988586, "learning_rate": 4.682912071093374e-05, "loss": 0.0077, "step": 54420 }, { "grad_norm": 0.11010555177927017, "learning_rate": 4.681261946948052e-05, "loss": 0.0082, "step": 54430 }, { "grad_norm": 0.12610052525997162, "learning_rate": 4.679611857659443e-05, "loss": 0.0074, "step": 54440 }, { "grad_norm": 0.1274353414773941, "learning_rate": 4.6779618034079975e-05, "loss": 0.0062, "step": 54450 }, { "grad_norm": 0.16982875764369965, "learning_rate": 4.6763117843741675e-05, "loss": 0.0062, "step": 54460 }, { "grad_norm": 0.13985049724578857, "learning_rate": 4.674661800738394e-05, "loss": 0.0056, "step": 54470 }, { "grad_norm": 0.1496385633945465, "learning_rate": 4.673011852681117e-05, "loss": 0.0071, "step": 54480 }, { "grad_norm": 0.1380639523267746, "learning_rate": 4.671361940382772e-05, "loss": 0.0072, "step": 54490 }, { "grad_norm": 0.24049438536167145, "learning_rate": 4.669712064023791e-05, "loss": 0.0105, "step": 54500 }, { "grad_norm": 0.11681806296110153, "learning_rate": 4.668062223784602e-05, "loss": 0.0073, "step": 54510 }, { "grad_norm": 0.12602612376213074, "learning_rate": 4.6664124198456274e-05, "loss": 0.0076, "step": 54520 }, { "grad_norm": 0.12356489896774292, "learning_rate": 4.6647626523872905e-05, "loss": 0.0071, "step": 54530 }, { "grad_norm": 0.14899294078350067, "learning_rate": 4.6631129215900066e-05, "loss": 0.0072, "step": 54540 }, { "grad_norm": 0.1161317229270935, "learning_rate": 4.661463227634187e-05, "loss": 0.007, "step": 54550 }, { "grad_norm": 0.14141583442687988, "learning_rate": 4.659813570700238e-05, "loss": 0.0063, "step": 54560 }, { "grad_norm": 0.135675311088562, "learning_rate": 4.658163950968568e-05, "loss": 0.0062, "step": 54570 }, { "grad_norm": 0.0867476612329483, "learning_rate": 4.6565143686195726e-05, "loss": 0.0073, "step": 54580 }, { "grad_norm": 0.10968931764364243, "learning_rate": 4.654864823833649e-05, "loss": 0.0093, "step": 54590 }, { "grad_norm": 0.1292683333158493, "learning_rate": 4.65321531679119e-05, "loss": 0.0058, "step": 54600 }, { "grad_norm": 0.11361725628376007, "learning_rate": 4.6515658476725834e-05, "loss": 0.0058, "step": 54610 }, { "grad_norm": 0.13076071441173553, "learning_rate": 4.649916416658212e-05, "loss": 0.0079, "step": 54620 }, { "grad_norm": 0.13787713646888733, "learning_rate": 4.648267023928455e-05, "loss": 0.0066, "step": 54630 }, { "grad_norm": 0.17983129620552063, "learning_rate": 4.646617669663687e-05, "loss": 0.0078, "step": 54640 }, { "grad_norm": 0.1525692194700241, "learning_rate": 4.644968354044277e-05, "loss": 0.005, "step": 54650 }, { "grad_norm": 0.09733931720256805, "learning_rate": 4.643319077250596e-05, "loss": 0.0084, "step": 54660 }, { "grad_norm": 0.09365759044885635, "learning_rate": 4.6416698394630055e-05, "loss": 0.006, "step": 54670 }, { "grad_norm": 0.120855413377285, "learning_rate": 4.640020640861862e-05, "loss": 0.0069, "step": 54680 }, { "grad_norm": 0.15019354224205017, "learning_rate": 4.6383714816275195e-05, "loss": 0.0058, "step": 54690 }, { "grad_norm": 0.1213843896985054, "learning_rate": 4.636722361940329e-05, "loss": 0.0075, "step": 54700 }, { "grad_norm": 0.10204347968101501, "learning_rate": 4.635073281980632e-05, "loss": 0.0083, "step": 54710 }, { "grad_norm": 0.08365689218044281, "learning_rate": 4.633424241928775e-05, "loss": 0.0072, "step": 54720 }, { "grad_norm": 0.12330715358257294, "learning_rate": 4.6317752419650914e-05, "loss": 0.0073, "step": 54730 }, { "grad_norm": 0.15971983969211578, "learning_rate": 4.630126282269914e-05, "loss": 0.0054, "step": 54740 }, { "grad_norm": 0.0959428995847702, "learning_rate": 4.628477363023569e-05, "loss": 0.0065, "step": 54750 }, { "grad_norm": 0.10515028983354568, "learning_rate": 4.626828484406383e-05, "loss": 0.0059, "step": 54760 }, { "grad_norm": 0.145382821559906, "learning_rate": 4.625179646598669e-05, "loss": 0.0056, "step": 54770 }, { "grad_norm": 0.15841856598854065, "learning_rate": 4.6235308497807487e-05, "loss": 0.0068, "step": 54780 }, { "grad_norm": 0.0825120136141777, "learning_rate": 4.6218820941329276e-05, "loss": 0.0068, "step": 54790 }, { "grad_norm": 0.09817365556955338, "learning_rate": 4.620233379835513e-05, "loss": 0.0063, "step": 54800 }, { "grad_norm": 0.10745887458324432, "learning_rate": 4.618584707068804e-05, "loss": 0.0103, "step": 54810 }, { "grad_norm": 0.1502522975206375, "learning_rate": 4.616936076013099e-05, "loss": 0.0072, "step": 54820 }, { "grad_norm": 0.11828140914440155, "learning_rate": 4.615287486848688e-05, "loss": 0.0076, "step": 54830 }, { "grad_norm": 0.12236309796571732, "learning_rate": 4.6136389397558584e-05, "loss": 0.0063, "step": 54840 }, { "grad_norm": 0.1714177280664444, "learning_rate": 4.6119904349148935e-05, "loss": 0.0082, "step": 54850 }, { "grad_norm": 0.12156657129526138, "learning_rate": 4.6103419725060726e-05, "loss": 0.0064, "step": 54860 }, { "grad_norm": 0.1363583654165268, "learning_rate": 4.608693552709667e-05, "loss": 0.006, "step": 54870 }, { "grad_norm": 0.128513902425766, "learning_rate": 4.607045175705947e-05, "loss": 0.006, "step": 54880 }, { "grad_norm": 0.1262054294347763, "learning_rate": 4.605396841675176e-05, "loss": 0.0068, "step": 54890 }, { "grad_norm": 0.11512008309364319, "learning_rate": 4.6037485507976094e-05, "loss": 0.006, "step": 54900 }, { "grad_norm": 0.09428904205560684, "learning_rate": 4.602100303253509e-05, "loss": 0.0125, "step": 54910 }, { "grad_norm": 0.1070825457572937, "learning_rate": 4.600452099223122e-05, "loss": 0.0082, "step": 54920 }, { "grad_norm": 0.12184716761112213, "learning_rate": 4.598803938886691e-05, "loss": 0.006, "step": 54930 }, { "grad_norm": 0.08948118984699249, "learning_rate": 4.597155822424459e-05, "loss": 0.0068, "step": 54940 }, { "grad_norm": 0.11399612575769424, "learning_rate": 4.59550775001666e-05, "loss": 0.0059, "step": 54950 }, { "grad_norm": 0.09969639778137207, "learning_rate": 4.5938597218435234e-05, "loss": 0.0072, "step": 54960 }, { "grad_norm": 0.2185727208852768, "learning_rate": 4.592211738085279e-05, "loss": 0.0072, "step": 54970 }, { "grad_norm": 0.17522075772285461, "learning_rate": 4.590563798922146e-05, "loss": 0.0068, "step": 54980 }, { "grad_norm": 0.1200457364320755, "learning_rate": 4.5889159045343404e-05, "loss": 0.0059, "step": 54990 }, { "grad_norm": 0.11363644152879715, "learning_rate": 4.5872680551020734e-05, "loss": 0.0067, "step": 55000 }, { "grad_norm": 0.13410785794258118, "learning_rate": 4.585620250805552e-05, "loss": 0.0078, "step": 55010 }, { "grad_norm": 0.12250758707523346, "learning_rate": 4.583972491824975e-05, "loss": 0.0076, "step": 55020 }, { "grad_norm": 0.1261807680130005, "learning_rate": 4.582324778340541e-05, "loss": 0.0063, "step": 55030 }, { "grad_norm": 0.09458144009113312, "learning_rate": 4.5806771105324424e-05, "loss": 0.0072, "step": 55040 }, { "grad_norm": 0.1121341660618782, "learning_rate": 4.579029488580865e-05, "loss": 0.0073, "step": 55050 }, { "grad_norm": 0.12591853737831116, "learning_rate": 4.5773819126659876e-05, "loss": 0.0083, "step": 55060 }, { "grad_norm": 0.171354278922081, "learning_rate": 4.575734382967991e-05, "loss": 0.0067, "step": 55070 }, { "grad_norm": 0.1607343703508377, "learning_rate": 4.574086899667044e-05, "loss": 0.0075, "step": 55080 }, { "grad_norm": 0.18184278905391693, "learning_rate": 4.5724394629433106e-05, "loss": 0.0063, "step": 55090 }, { "grad_norm": 0.14119569957256317, "learning_rate": 4.570792072976958e-05, "loss": 0.0086, "step": 55100 }, { "grad_norm": 0.11237815022468567, "learning_rate": 4.569144729948138e-05, "loss": 0.008, "step": 55110 }, { "grad_norm": 0.1467788964509964, "learning_rate": 4.5674974340370033e-05, "loss": 0.0067, "step": 55120 }, { "grad_norm": 0.08993035554885864, "learning_rate": 4.565850185423699e-05, "loss": 0.0076, "step": 55130 }, { "grad_norm": 0.12709173560142517, "learning_rate": 4.564202984288365e-05, "loss": 0.0064, "step": 55140 }, { "grad_norm": 0.11586527526378632, "learning_rate": 4.5625558308111356e-05, "loss": 0.0064, "step": 55150 }, { "grad_norm": 0.09509899467229843, "learning_rate": 4.560908725172145e-05, "loss": 0.0074, "step": 55160 }, { "grad_norm": 0.1225828230381012, "learning_rate": 4.559261667551516e-05, "loss": 0.0073, "step": 55170 }, { "grad_norm": 0.15563656389713287, "learning_rate": 4.5576146581293685e-05, "loss": 0.0087, "step": 55180 }, { "grad_norm": 0.12346313893795013, "learning_rate": 4.5559676970858164e-05, "loss": 0.0073, "step": 55190 }, { "grad_norm": 0.111387237906456, "learning_rate": 4.5543207846009705e-05, "loss": 0.0086, "step": 55200 }, { "grad_norm": 0.09585835039615631, "learning_rate": 4.552673920854933e-05, "loss": 0.0061, "step": 55210 }, { "grad_norm": 0.0912935733795166, "learning_rate": 4.5510271060278006e-05, "loss": 0.0055, "step": 55220 }, { "grad_norm": 0.13602277636528015, "learning_rate": 4.5493803402996704e-05, "loss": 0.0053, "step": 55230 }, { "grad_norm": 0.11951883137226105, "learning_rate": 4.5477336238506295e-05, "loss": 0.0059, "step": 55240 }, { "grad_norm": 0.15593643486499786, "learning_rate": 4.546086956860759e-05, "loss": 0.0075, "step": 55250 }, { "grad_norm": 0.1567264348268509, "learning_rate": 4.5444403395101374e-05, "loss": 0.0074, "step": 55260 }, { "grad_norm": 0.08955591917037964, "learning_rate": 4.542793771978834e-05, "loss": 0.0053, "step": 55270 }, { "grad_norm": 0.12059862911701202, "learning_rate": 4.541147254446916e-05, "loss": 0.0063, "step": 55280 }, { "grad_norm": 0.12662631273269653, "learning_rate": 4.539500787094445e-05, "loss": 0.0089, "step": 55290 }, { "grad_norm": 0.12040850520133972, "learning_rate": 4.537854370101475e-05, "loss": 0.0069, "step": 55300 }, { "grad_norm": 0.13198548555374146, "learning_rate": 4.5362080036480566e-05, "loss": 0.0068, "step": 55310 }, { "grad_norm": 0.1565338522195816, "learning_rate": 4.534561687914235e-05, "loss": 0.0066, "step": 55320 }, { "grad_norm": 0.12194429337978363, "learning_rate": 4.5329154230800466e-05, "loss": 0.006, "step": 55330 }, { "grad_norm": 0.08878416568040848, "learning_rate": 4.531269209325525e-05, "loss": 0.0056, "step": 55340 }, { "grad_norm": 0.14387424290180206, "learning_rate": 4.529623046830696e-05, "loss": 0.0088, "step": 55350 }, { "grad_norm": 0.0879441499710083, "learning_rate": 4.5279769357755866e-05, "loss": 0.0054, "step": 55360 }, { "grad_norm": 0.13820914924144745, "learning_rate": 4.5263308763402084e-05, "loss": 0.0064, "step": 55370 }, { "grad_norm": 0.07377644628286362, "learning_rate": 4.524684868704574e-05, "loss": 0.0051, "step": 55380 }, { "grad_norm": 0.10442475974559784, "learning_rate": 4.523038913048688e-05, "loss": 0.0057, "step": 55390 }, { "grad_norm": 0.09409096091985703, "learning_rate": 4.521393009552549e-05, "loss": 0.0075, "step": 55400 }, { "grad_norm": 0.14935754239559174, "learning_rate": 4.5197471583961484e-05, "loss": 0.007, "step": 55410 }, { "grad_norm": 0.1023506447672844, "learning_rate": 4.51810135975948e-05, "loss": 0.0055, "step": 55420 }, { "grad_norm": 0.14161498844623566, "learning_rate": 4.516455613822521e-05, "loss": 0.0067, "step": 55430 }, { "grad_norm": 0.08443975448608398, "learning_rate": 4.514809920765249e-05, "loss": 0.0068, "step": 55440 }, { "grad_norm": 0.1419650912284851, "learning_rate": 4.5131642807676356e-05, "loss": 0.0067, "step": 55450 }, { "grad_norm": 0.12506918609142303, "learning_rate": 4.511518694009644e-05, "loss": 0.0058, "step": 55460 }, { "grad_norm": 0.1151534765958786, "learning_rate": 4.509873160671231e-05, "loss": 0.0059, "step": 55470 }, { "grad_norm": 0.13346423208713531, "learning_rate": 4.508227680932356e-05, "loss": 0.0056, "step": 55480 }, { "grad_norm": 0.1591210812330246, "learning_rate": 4.5065822549729605e-05, "loss": 0.0063, "step": 55490 }, { "grad_norm": 0.1305425465106964, "learning_rate": 4.504936882972989e-05, "loss": 0.0094, "step": 55500 }, { "grad_norm": 0.1441287100315094, "learning_rate": 4.503291565112374e-05, "loss": 0.0061, "step": 55510 }, { "grad_norm": 0.15444877743721008, "learning_rate": 4.501646301571048e-05, "loss": 0.0071, "step": 55520 }, { "grad_norm": 0.12318233400583267, "learning_rate": 4.500001092528933e-05, "loss": 0.0055, "step": 55530 }, { "grad_norm": 0.11888053268194199, "learning_rate": 4.4983559381659455e-05, "loss": 0.009, "step": 55540 }, { "grad_norm": 0.1363670527935028, "learning_rate": 4.496710838661999e-05, "loss": 0.0059, "step": 55550 }, { "grad_norm": 0.1602737307548523, "learning_rate": 4.495065794196999e-05, "loss": 0.0078, "step": 55560 }, { "grad_norm": 0.08742687106132507, "learning_rate": 4.4934208049508455e-05, "loss": 0.0053, "step": 55570 }, { "grad_norm": 0.13520628213882446, "learning_rate": 4.4917758711034303e-05, "loss": 0.0063, "step": 55580 }, { "grad_norm": 0.13159812986850739, "learning_rate": 4.490130992834642e-05, "loss": 0.0066, "step": 55590 }, { "grad_norm": 0.16954903304576874, "learning_rate": 4.488486170324359e-05, "loss": 0.0082, "step": 55600 }, { "grad_norm": 0.12970003485679626, "learning_rate": 4.486841403752463e-05, "loss": 0.0066, "step": 55610 }, { "grad_norm": 0.09363836795091629, "learning_rate": 4.4851966932988174e-05, "loss": 0.0058, "step": 55620 }, { "grad_norm": 0.11280863732099533, "learning_rate": 4.4835520391432894e-05, "loss": 0.0063, "step": 55630 }, { "grad_norm": 0.1140005961060524, "learning_rate": 4.481907441465733e-05, "loss": 0.0074, "step": 55640 }, { "grad_norm": 0.11910563707351685, "learning_rate": 4.480262900446e-05, "loss": 0.0076, "step": 55650 }, { "grad_norm": 0.11951173096895218, "learning_rate": 4.478618416263933e-05, "loss": 0.005, "step": 55660 }, { "grad_norm": 0.14871682226657867, "learning_rate": 4.476973989099373e-05, "loss": 0.0061, "step": 55670 }, { "grad_norm": 0.13782437145709991, "learning_rate": 4.475329619132153e-05, "loss": 0.0073, "step": 55680 }, { "grad_norm": 0.16188475489616394, "learning_rate": 4.4736853065420956e-05, "loss": 0.0074, "step": 55690 }, { "grad_norm": 0.13946959376335144, "learning_rate": 4.472041051509023e-05, "loss": 0.0084, "step": 55700 }, { "grad_norm": 0.12333479523658752, "learning_rate": 4.470396854212746e-05, "loss": 0.0073, "step": 55710 }, { "grad_norm": 0.09784813225269318, "learning_rate": 4.468752714833072e-05, "loss": 0.0069, "step": 55720 }, { "grad_norm": 0.12318436801433563, "learning_rate": 4.467108633549803e-05, "loss": 0.006, "step": 55730 }, { "grad_norm": 0.07701879739761353, "learning_rate": 4.4654646105427335e-05, "loss": 0.0055, "step": 55740 }, { "grad_norm": 0.0854106992483139, "learning_rate": 4.463820645991651e-05, "loss": 0.0051, "step": 55750 }, { "grad_norm": 0.13999412953853607, "learning_rate": 4.462176740076334e-05, "loss": 0.0063, "step": 55760 }, { "grad_norm": 0.11436349898576736, "learning_rate": 4.4605328929765624e-05, "loss": 0.0059, "step": 55770 }, { "grad_norm": 0.12033013254404068, "learning_rate": 4.458889104872102e-05, "loss": 0.0098, "step": 55780 }, { "grad_norm": 0.10894913971424103, "learning_rate": 4.457245375942715e-05, "loss": 0.0075, "step": 55790 }, { "grad_norm": 0.14148513972759247, "learning_rate": 4.455601706368157e-05, "loss": 0.0068, "step": 55800 }, { "grad_norm": 0.10788346827030182, "learning_rate": 4.453958096328179e-05, "loss": 0.0073, "step": 55810 }, { "grad_norm": 0.11733000725507736, "learning_rate": 4.452314546002523e-05, "loss": 0.0054, "step": 55820 }, { "grad_norm": 0.14870624244213104, "learning_rate": 4.4506710555709246e-05, "loss": 0.0072, "step": 55830 }, { "grad_norm": 0.12037333846092224, "learning_rate": 4.449027625213114e-05, "loss": 0.0074, "step": 55840 }, { "grad_norm": 0.12842540442943573, "learning_rate": 4.4473842551088135e-05, "loss": 0.0065, "step": 55850 }, { "grad_norm": 0.1478033810853958, "learning_rate": 4.445740945437738e-05, "loss": 0.007, "step": 55860 }, { "grad_norm": 0.12037457525730133, "learning_rate": 4.444097696379602e-05, "loss": 0.008, "step": 55870 }, { "grad_norm": 0.11077938228845596, "learning_rate": 4.442454508114106e-05, "loss": 0.0072, "step": 55880 }, { "grad_norm": 0.11451497673988342, "learning_rate": 4.4408113808209465e-05, "loss": 0.0057, "step": 55890 }, { "grad_norm": 0.1236770823597908, "learning_rate": 4.439168314679813e-05, "loss": 0.0064, "step": 55900 }, { "grad_norm": 0.14483310282230377, "learning_rate": 4.43752530987039e-05, "loss": 0.0078, "step": 55910 }, { "grad_norm": 0.12516117095947266, "learning_rate": 4.4358823665723515e-05, "loss": 0.0063, "step": 55920 }, { "grad_norm": 0.1342700570821762, "learning_rate": 4.434239484965371e-05, "loss": 0.0065, "step": 55930 }, { "grad_norm": 0.13397963345050812, "learning_rate": 4.4325966652291103e-05, "loss": 0.0072, "step": 55940 }, { "grad_norm": 0.12940943241119385, "learning_rate": 4.430953907543225e-05, "loss": 0.0079, "step": 55950 }, { "grad_norm": 0.13297834992408752, "learning_rate": 4.4293112120873645e-05, "loss": 0.0091, "step": 55960 }, { "grad_norm": 0.10890454053878784, "learning_rate": 4.427668579041172e-05, "loss": 0.0065, "step": 55970 }, { "grad_norm": 0.1327180713415146, "learning_rate": 4.4260260085842816e-05, "loss": 0.0048, "step": 55980 }, { "grad_norm": 0.12301377952098846, "learning_rate": 4.4243835008963256e-05, "loss": 0.0063, "step": 55990 }, { "grad_norm": 0.11652856320142746, "learning_rate": 4.422741056156925e-05, "loss": 0.0048, "step": 56000 }, { "grad_norm": 0.09968530386686325, "learning_rate": 4.421098674545693e-05, "loss": 0.0079, "step": 56010 }, { "grad_norm": 0.12524378299713135, "learning_rate": 4.419456356242241e-05, "loss": 0.0078, "step": 56020 }, { "grad_norm": 0.10762656480073929, "learning_rate": 4.4178141014261685e-05, "loss": 0.0068, "step": 56030 }, { "grad_norm": 0.11284178495407104, "learning_rate": 4.416171910277071e-05, "loss": 0.0056, "step": 56040 }, { "grad_norm": 0.11290041357278824, "learning_rate": 4.414529782974533e-05, "loss": 0.0057, "step": 56050 }, { "grad_norm": 0.11407545953989029, "learning_rate": 4.41288771969814e-05, "loss": 0.005, "step": 56060 }, { "grad_norm": 0.15581992268562317, "learning_rate": 4.411245720627464e-05, "loss": 0.0097, "step": 56070 }, { "grad_norm": 0.14203529059886932, "learning_rate": 4.40960378594207e-05, "loss": 0.0076, "step": 56080 }, { "grad_norm": 0.15121698379516602, "learning_rate": 4.407961915821519e-05, "loss": 0.0117, "step": 56090 }, { "grad_norm": 0.13385216891765594, "learning_rate": 4.406320110445362e-05, "loss": 0.0072, "step": 56100 }, { "grad_norm": 0.12708882987499237, "learning_rate": 4.404678369993144e-05, "loss": 0.0073, "step": 56110 }, { "grad_norm": 0.12009242922067642, "learning_rate": 4.403036694644406e-05, "loss": 0.0059, "step": 56120 }, { "grad_norm": 0.1204388216137886, "learning_rate": 4.4013950845786764e-05, "loss": 0.0053, "step": 56130 }, { "grad_norm": 0.11360874027013779, "learning_rate": 4.399753539975482e-05, "loss": 0.0058, "step": 56140 }, { "grad_norm": 0.14880000054836273, "learning_rate": 4.398112061014337e-05, "loss": 0.008, "step": 56150 }, { "grad_norm": 0.09699366241693497, "learning_rate": 4.396470647874753e-05, "loss": 0.0056, "step": 56160 }, { "grad_norm": 0.08032996952533722, "learning_rate": 4.394829300736229e-05, "loss": 0.0048, "step": 56170 }, { "grad_norm": 0.22926253080368042, "learning_rate": 4.393188019778265e-05, "loss": 0.0071, "step": 56180 }, { "grad_norm": 0.1109914630651474, "learning_rate": 4.391546805180347e-05, "loss": 0.0068, "step": 56190 }, { "grad_norm": 0.09824188798666, "learning_rate": 4.389905657121955e-05, "loss": 0.0076, "step": 56200 }, { "grad_norm": 0.1545938402414322, "learning_rate": 4.3882645757825635e-05, "loss": 0.0073, "step": 56210 }, { "grad_norm": 0.08576811850070953, "learning_rate": 4.386623561341637e-05, "loss": 0.007, "step": 56220 }, { "grad_norm": 0.09271478652954102, "learning_rate": 4.384982613978637e-05, "loss": 0.0081, "step": 56230 }, { "grad_norm": 0.12901929020881653, "learning_rate": 4.383341733873012e-05, "loss": 0.006, "step": 56240 }, { "grad_norm": 0.0927257239818573, "learning_rate": 4.381700921204209e-05, "loss": 0.0073, "step": 56250 }, { "grad_norm": 0.17764201760292053, "learning_rate": 4.380060176151663e-05, "loss": 0.006, "step": 56260 }, { "grad_norm": 0.1624291092157364, "learning_rate": 4.378419498894805e-05, "loss": 0.0075, "step": 56270 }, { "grad_norm": 0.12720496952533722, "learning_rate": 4.376778889613056e-05, "loss": 0.0064, "step": 56280 }, { "grad_norm": 0.14778019487857819, "learning_rate": 4.375138348485831e-05, "loss": 0.0071, "step": 56290 }, { "grad_norm": 0.12109880149364471, "learning_rate": 4.3734978756925346e-05, "loss": 0.0062, "step": 56300 }, { "grad_norm": 0.11491234600543976, "learning_rate": 4.3718574714125706e-05, "loss": 0.0051, "step": 56310 }, { "grad_norm": 0.1295991837978363, "learning_rate": 4.370217135825329e-05, "loss": 0.0066, "step": 56320 }, { "grad_norm": 0.12928229570388794, "learning_rate": 4.368576869110194e-05, "loss": 0.0062, "step": 56330 }, { "grad_norm": 0.11884888261556625, "learning_rate": 4.366936671446544e-05, "loss": 0.007, "step": 56340 }, { "grad_norm": 0.12316581606864929, "learning_rate": 4.3652965430137474e-05, "loss": 0.0065, "step": 56350 }, { "grad_norm": 0.07586947083473206, "learning_rate": 4.3636564839911646e-05, "loss": 0.0065, "step": 56360 }, { "grad_norm": 0.17357368767261505, "learning_rate": 4.3620164945581545e-05, "loss": 0.0063, "step": 56370 }, { "grad_norm": 0.1162395104765892, "learning_rate": 4.360376574894061e-05, "loss": 0.0067, "step": 56380 }, { "grad_norm": 0.21952468156814575, "learning_rate": 4.358736725178224e-05, "loss": 0.0065, "step": 56390 }, { "grad_norm": 0.12787485122680664, "learning_rate": 4.357096945589974e-05, "loss": 0.0059, "step": 56400 }, { "grad_norm": 0.1233164593577385, "learning_rate": 4.355457236308636e-05, "loss": 0.0071, "step": 56410 }, { "grad_norm": 0.19679513573646545, "learning_rate": 4.353817597513526e-05, "loss": 0.0084, "step": 56420 }, { "grad_norm": 0.15866732597351074, "learning_rate": 4.352178029383948e-05, "loss": 0.0074, "step": 56430 }, { "grad_norm": 0.15567880868911743, "learning_rate": 4.3505385320992105e-05, "loss": 0.0072, "step": 56440 }, { "grad_norm": 0.12307562679052353, "learning_rate": 4.348899105838602e-05, "loss": 0.0057, "step": 56450 }, { "grad_norm": 0.09849046915769577, "learning_rate": 4.3472597507814087e-05, "loss": 0.0057, "step": 56460 }, { "grad_norm": 0.08099552243947983, "learning_rate": 4.3456204671069066e-05, "loss": 0.0067, "step": 56470 }, { "grad_norm": 0.10233364999294281, "learning_rate": 4.343981254994367e-05, "loss": 0.0055, "step": 56480 }, { "grad_norm": 0.1200234591960907, "learning_rate": 4.34234211462305e-05, "loss": 0.0089, "step": 56490 }, { "grad_norm": 0.13569523394107819, "learning_rate": 4.340703046172213e-05, "loss": 0.0089, "step": 56500 }, { "grad_norm": 0.12163767963647842, "learning_rate": 4.339064049821097e-05, "loss": 0.0063, "step": 56510 }, { "grad_norm": 0.1680813729763031, "learning_rate": 4.3374251257489446e-05, "loss": 0.0067, "step": 56520 }, { "grad_norm": 0.1629902720451355, "learning_rate": 4.335786274134984e-05, "loss": 0.0056, "step": 56530 }, { "grad_norm": 0.13733278214931488, "learning_rate": 4.334147495158438e-05, "loss": 0.0074, "step": 56540 }, { "grad_norm": 0.12472109496593475, "learning_rate": 4.33250878899852e-05, "loss": 0.0067, "step": 56550 }, { "grad_norm": 0.13270671665668488, "learning_rate": 4.330870155834437e-05, "loss": 0.0077, "step": 56560 }, { "grad_norm": 0.11284130066633224, "learning_rate": 4.3292315958453884e-05, "loss": 0.0092, "step": 56570 }, { "grad_norm": 0.15237516164779663, "learning_rate": 4.3275931092105646e-05, "loss": 0.0079, "step": 56580 }, { "grad_norm": 0.11196565628051758, "learning_rate": 4.325954696109147e-05, "loss": 0.007, "step": 56590 }, { "grad_norm": 0.10026060789823532, "learning_rate": 4.32431635672031e-05, "loss": 0.008, "step": 56600 }, { "grad_norm": 0.12503038346767426, "learning_rate": 4.3226780912232214e-05, "loss": 0.0053, "step": 56610 }, { "grad_norm": 0.15825608372688293, "learning_rate": 4.321039899797036e-05, "loss": 0.0075, "step": 56620 }, { "grad_norm": 0.21473589539527893, "learning_rate": 4.319401782620908e-05, "loss": 0.0074, "step": 56630 }, { "grad_norm": 0.12371716648340225, "learning_rate": 4.317763739873978e-05, "loss": 0.0066, "step": 56640 }, { "grad_norm": 0.09543471783399582, "learning_rate": 4.31612577173538e-05, "loss": 0.007, "step": 56650 }, { "grad_norm": 0.13088549673557281, "learning_rate": 4.314487878384239e-05, "loss": 0.0086, "step": 56660 }, { "grad_norm": 0.11172778159379959, "learning_rate": 4.3128500599996736e-05, "loss": 0.0058, "step": 56670 }, { "grad_norm": 0.14701181650161743, "learning_rate": 4.31121231676079e-05, "loss": 0.0057, "step": 56680 }, { "grad_norm": 0.1671552062034607, "learning_rate": 4.309574648846694e-05, "loss": 0.0082, "step": 56690 }, { "grad_norm": 0.15775923430919647, "learning_rate": 4.3079370564364755e-05, "loss": 0.0073, "step": 56700 }, { "grad_norm": 0.14702822268009186, "learning_rate": 4.306299539709221e-05, "loss": 0.0072, "step": 56710 }, { "grad_norm": 0.13683032989501953, "learning_rate": 4.3046620988440046e-05, "loss": 0.008, "step": 56720 }, { "grad_norm": 0.13947395980358124, "learning_rate": 4.303024734019897e-05, "loss": 0.0086, "step": 56730 }, { "grad_norm": 0.10893470793962479, "learning_rate": 4.3013874454159565e-05, "loss": 0.0065, "step": 56740 }, { "grad_norm": 0.1392446905374527, "learning_rate": 4.299750233211233e-05, "loss": 0.0076, "step": 56750 }, { "grad_norm": 0.10985320806503296, "learning_rate": 4.2981130975847715e-05, "loss": 0.0059, "step": 56760 }, { "grad_norm": 0.1328398734331131, "learning_rate": 4.2964760387156076e-05, "loss": 0.0055, "step": 56770 }, { "grad_norm": 0.15247824788093567, "learning_rate": 4.294839056782767e-05, "loss": 0.0079, "step": 56780 }, { "grad_norm": 0.10727833956480026, "learning_rate": 4.293202151965266e-05, "loss": 0.0063, "step": 56790 }, { "grad_norm": 0.12970802187919617, "learning_rate": 4.291565324442115e-05, "loss": 0.0064, "step": 56800 }, { "grad_norm": 0.12524384260177612, "learning_rate": 4.2899285743923135e-05, "loss": 0.0065, "step": 56810 }, { "grad_norm": 0.12616974115371704, "learning_rate": 4.288291901994857e-05, "loss": 0.0057, "step": 56820 }, { "grad_norm": 0.12752829492092133, "learning_rate": 4.286655307428729e-05, "loss": 0.0071, "step": 56830 }, { "grad_norm": 0.11183291673660278, "learning_rate": 4.285018790872903e-05, "loss": 0.0076, "step": 56840 }, { "grad_norm": 0.115350142121315, "learning_rate": 4.283382352506347e-05, "loss": 0.0071, "step": 56850 }, { "grad_norm": 0.1576695740222931, "learning_rate": 4.2817459925080205e-05, "loss": 0.0068, "step": 56860 }, { "grad_norm": 0.10298418253660202, "learning_rate": 4.28010971105687e-05, "loss": 0.0061, "step": 56870 }, { "grad_norm": 0.15746329724788666, "learning_rate": 4.278473508331841e-05, "loss": 0.0081, "step": 56880 }, { "grad_norm": 0.13310588896274567, "learning_rate": 4.276837384511864e-05, "loss": 0.0066, "step": 56890 }, { "grad_norm": 0.11118058115243912, "learning_rate": 4.275201339775864e-05, "loss": 0.0071, "step": 56900 }, { "grad_norm": 0.10370785742998123, "learning_rate": 4.273565374302756e-05, "loss": 0.0079, "step": 56910 }, { "grad_norm": 0.10570455342531204, "learning_rate": 4.2719294882714454e-05, "loss": 0.0069, "step": 56920 }, { "grad_norm": 0.10539944469928741, "learning_rate": 4.270293681860831e-05, "loss": 0.0061, "step": 56930 }, { "grad_norm": 0.11388864368200302, "learning_rate": 4.268657955249802e-05, "loss": 0.0073, "step": 56940 }, { "grad_norm": 0.14796678721904755, "learning_rate": 4.267022308617241e-05, "loss": 0.0058, "step": 56950 }, { "grad_norm": 0.11216778308153152, "learning_rate": 4.265386742142018e-05, "loss": 0.0064, "step": 56960 }, { "grad_norm": 0.1053631603717804, "learning_rate": 4.263751256002996e-05, "loss": 0.0068, "step": 56970 }, { "grad_norm": 0.12249606847763062, "learning_rate": 4.26211585037903e-05, "loss": 0.0059, "step": 56980 }, { "grad_norm": 0.16406334936618805, "learning_rate": 4.2604805254489656e-05, "loss": 0.0062, "step": 56990 }, { "grad_norm": 0.14640550315380096, "learning_rate": 4.258845281391638e-05, "loss": 0.0081, "step": 57000 }, { "grad_norm": 0.09293985366821289, "learning_rate": 4.257210118385876e-05, "loss": 0.0049, "step": 57010 }, { "grad_norm": 0.0943465605378151, "learning_rate": 4.2555750366104994e-05, "loss": 0.0039, "step": 57020 }, { "grad_norm": 0.1392676681280136, "learning_rate": 4.2539400362443184e-05, "loss": 0.0093, "step": 57030 }, { "grad_norm": 0.11403381079435349, "learning_rate": 4.252305117466133e-05, "loss": 0.0068, "step": 57040 }, { "grad_norm": 0.16208235919475555, "learning_rate": 4.2506702804547356e-05, "loss": 0.0067, "step": 57050 }, { "grad_norm": 0.11147434264421463, "learning_rate": 4.249035525388907e-05, "loss": 0.0056, "step": 57060 }, { "grad_norm": 0.11658385396003723, "learning_rate": 4.247400852447427e-05, "loss": 0.0057, "step": 57070 }, { "grad_norm": 0.1584496796131134, "learning_rate": 4.245766261809059e-05, "loss": 0.0058, "step": 57080 }, { "grad_norm": 0.331478089094162, "learning_rate": 4.244131753652558e-05, "loss": 0.006, "step": 57090 }, { "grad_norm": 0.11523038148880005, "learning_rate": 4.242497328156672e-05, "loss": 0.0057, "step": 57100 }, { "grad_norm": 0.11410827189683914, "learning_rate": 4.240862985500139e-05, "loss": 0.0062, "step": 57110 }, { "grad_norm": 0.1170959621667862, "learning_rate": 4.2392287258616894e-05, "loss": 0.0071, "step": 57120 }, { "grad_norm": 0.12042615562677383, "learning_rate": 4.2375945494200396e-05, "loss": 0.0095, "step": 57130 }, { "grad_norm": 0.11074237525463104, "learning_rate": 4.2359604563539054e-05, "loss": 0.0056, "step": 57140 }, { "grad_norm": 0.12152789533138275, "learning_rate": 4.2343264468419885e-05, "loss": 0.0058, "step": 57150 }, { "grad_norm": 0.09754668921232224, "learning_rate": 4.232692521062979e-05, "loss": 0.0062, "step": 57160 }, { "grad_norm": 0.1363610327243805, "learning_rate": 4.231058679195562e-05, "loss": 0.006, "step": 57170 }, { "grad_norm": 0.1465170979499817, "learning_rate": 4.229424921418411e-05, "loss": 0.0055, "step": 57180 }, { "grad_norm": 0.12341859936714172, "learning_rate": 4.2277912479101916e-05, "loss": 0.0083, "step": 57190 }, { "grad_norm": 0.13904425501823425, "learning_rate": 4.226157658849562e-05, "loss": 0.0069, "step": 57200 }, { "grad_norm": 0.08940917253494263, "learning_rate": 4.2245241544151664e-05, "loss": 0.0053, "step": 57210 }, { "grad_norm": 0.11038455367088318, "learning_rate": 4.2228907347856425e-05, "loss": 0.0114, "step": 57220 }, { "grad_norm": 0.07742302119731903, "learning_rate": 4.2212574001396214e-05, "loss": 0.0055, "step": 57230 }, { "grad_norm": 0.13034096360206604, "learning_rate": 4.21962415065572e-05, "loss": 0.0052, "step": 57240 }, { "grad_norm": 0.11609198153018951, "learning_rate": 4.217990986512548e-05, "loss": 0.0057, "step": 57250 }, { "grad_norm": 0.10848946124315262, "learning_rate": 4.216357907888704e-05, "loss": 0.0055, "step": 57260 }, { "grad_norm": 0.15010331571102142, "learning_rate": 4.2147249149627824e-05, "loss": 0.0057, "step": 57270 }, { "grad_norm": 0.13319502770900726, "learning_rate": 4.213092007913364e-05, "loss": 0.0069, "step": 57280 }, { "grad_norm": 0.13491085171699524, "learning_rate": 4.2114591869190205e-05, "loss": 0.0075, "step": 57290 }, { "grad_norm": 0.13075631856918335, "learning_rate": 4.209826452158315e-05, "loss": 0.0057, "step": 57300 }, { "grad_norm": 0.09010060876607895, "learning_rate": 4.2081938038098e-05, "loss": 0.0051, "step": 57310 }, { "grad_norm": 0.17874573171138763, "learning_rate": 4.2065612420520185e-05, "loss": 0.0071, "step": 57320 }, { "grad_norm": 0.12800894677639008, "learning_rate": 4.204928767063509e-05, "loss": 0.0077, "step": 57330 }, { "grad_norm": 0.1588289737701416, "learning_rate": 4.2032963790227944e-05, "loss": 0.0054, "step": 57340 }, { "grad_norm": 0.12483081221580505, "learning_rate": 4.2016640781083887e-05, "loss": 0.0083, "step": 57350 }, { "grad_norm": 0.11669939011335373, "learning_rate": 4.2000318644988004e-05, "loss": 0.0068, "step": 57360 }, { "grad_norm": 0.10766125470399857, "learning_rate": 4.198399738372524e-05, "loss": 0.0054, "step": 57370 }, { "grad_norm": 0.10907246917486191, "learning_rate": 4.196767699908045e-05, "loss": 0.0077, "step": 57380 }, { "grad_norm": 0.13983650505542755, "learning_rate": 4.195135749283845e-05, "loss": 0.0074, "step": 57390 }, { "grad_norm": 0.13453084230422974, "learning_rate": 4.1935038866783894e-05, "loss": 0.0063, "step": 57400 }, { "grad_norm": 0.11815927177667618, "learning_rate": 4.1918721122701355e-05, "loss": 0.0059, "step": 57410 }, { "grad_norm": 0.11197543889284134, "learning_rate": 4.190240426237533e-05, "loss": 0.0075, "step": 57420 }, { "grad_norm": 0.07831054925918579, "learning_rate": 4.188608828759019e-05, "loss": 0.0076, "step": 57430 }, { "grad_norm": 0.11982768774032593, "learning_rate": 4.1869773200130246e-05, "loss": 0.0068, "step": 57440 }, { "grad_norm": 0.11510886251926422, "learning_rate": 4.185345900177967e-05, "loss": 0.0083, "step": 57450 }, { "grad_norm": 0.177919402718544, "learning_rate": 4.183714569432258e-05, "loss": 0.0064, "step": 57460 }, { "grad_norm": 0.1300252079963684, "learning_rate": 4.182083327954295e-05, "loss": 0.0066, "step": 57470 }, { "grad_norm": 0.1204061433672905, "learning_rate": 4.1804521759224714e-05, "loss": 0.0053, "step": 57480 }, { "grad_norm": 0.1446349173784256, "learning_rate": 4.178821113515165e-05, "loss": 0.0069, "step": 57490 }, { "grad_norm": 0.13810324668884277, "learning_rate": 4.1771901409107476e-05, "loss": 0.0068, "step": 57500 }, { "grad_norm": 0.09874290227890015, "learning_rate": 4.175559258287577e-05, "loss": 0.0063, "step": 57510 }, { "grad_norm": 0.09426404535770416, "learning_rate": 4.173928465824008e-05, "loss": 0.0068, "step": 57520 }, { "grad_norm": 0.13389825820922852, "learning_rate": 4.17229776369838e-05, "loss": 0.0067, "step": 57530 }, { "grad_norm": 0.13315187394618988, "learning_rate": 4.170667152089025e-05, "loss": 0.0075, "step": 57540 }, { "grad_norm": 0.10760392993688583, "learning_rate": 4.1690366311742626e-05, "loss": 0.0079, "step": 57550 }, { "grad_norm": 0.14796091616153717, "learning_rate": 4.1674062011324055e-05, "loss": 0.0105, "step": 57560 }, { "grad_norm": 0.09715337306261063, "learning_rate": 4.165775862141752e-05, "loss": 0.0057, "step": 57570 }, { "grad_norm": 0.19485566020011902, "learning_rate": 4.164145614380598e-05, "loss": 0.0062, "step": 57580 }, { "grad_norm": 0.12872347235679626, "learning_rate": 4.162515458027223e-05, "loss": 0.0063, "step": 57590 }, { "grad_norm": 0.09773395955562592, "learning_rate": 4.160885393259899e-05, "loss": 0.0082, "step": 57600 }, { "grad_norm": 0.11326436698436737, "learning_rate": 4.159255420256886e-05, "loss": 0.0046, "step": 57610 }, { "grad_norm": 0.12611739337444305, "learning_rate": 4.157625539196436e-05, "loss": 0.0099, "step": 57620 }, { "grad_norm": 0.17656221985816956, "learning_rate": 4.155995750256791e-05, "loss": 0.0065, "step": 57630 }, { "grad_norm": 0.12214788794517517, "learning_rate": 4.15436605361618e-05, "loss": 0.0094, "step": 57640 }, { "grad_norm": 0.11186548322439194, "learning_rate": 4.152736449452827e-05, "loss": 0.0072, "step": 57650 }, { "grad_norm": 0.16589075326919556, "learning_rate": 4.151106937944943e-05, "loss": 0.0063, "step": 57660 }, { "grad_norm": 0.18462036550045013, "learning_rate": 4.149477519270727e-05, "loss": 0.0068, "step": 57670 }, { "grad_norm": 0.15737789869308472, "learning_rate": 4.147848193608371e-05, "loss": 0.0065, "step": 57680 }, { "grad_norm": 0.200541153550148, "learning_rate": 4.146218961136056e-05, "loss": 0.0091, "step": 57690 }, { "grad_norm": 0.14417091012001038, "learning_rate": 4.14458982203195e-05, "loss": 0.0073, "step": 57700 }, { "grad_norm": 0.09534519910812378, "learning_rate": 4.1429607764742165e-05, "loss": 0.0085, "step": 57710 }, { "grad_norm": 0.11062347888946533, "learning_rate": 4.141331824641003e-05, "loss": 0.0053, "step": 57720 }, { "grad_norm": 0.13395161926746368, "learning_rate": 4.139702966710452e-05, "loss": 0.0062, "step": 57730 }, { "grad_norm": 0.20653891563415527, "learning_rate": 4.13807420286069e-05, "loss": 0.0071, "step": 57740 }, { "grad_norm": 0.15712974965572357, "learning_rate": 4.136445533269837e-05, "loss": 0.0065, "step": 57750 }, { "grad_norm": 0.11231247335672379, "learning_rate": 4.134816958116003e-05, "loss": 0.0081, "step": 57760 }, { "grad_norm": 0.11430549621582031, "learning_rate": 4.133188477577283e-05, "loss": 0.007, "step": 57770 }, { "grad_norm": 0.15638019144535065, "learning_rate": 4.131560091831771e-05, "loss": 0.0061, "step": 57780 }, { "grad_norm": 0.09183097630739212, "learning_rate": 4.1299318010575405e-05, "loss": 0.0068, "step": 57790 }, { "grad_norm": 0.0691462829709053, "learning_rate": 4.128303605432661e-05, "loss": 0.0072, "step": 57800 }, { "grad_norm": 0.12127190083265305, "learning_rate": 4.126675505135188e-05, "loss": 0.0056, "step": 57810 }, { "grad_norm": 0.1607084721326828, "learning_rate": 4.125047500343169e-05, "loss": 0.0083, "step": 57820 }, { "grad_norm": 0.09661979228258133, "learning_rate": 4.1234195912346376e-05, "loss": 0.0056, "step": 57830 }, { "grad_norm": 0.14292632043361664, "learning_rate": 4.1217917779876235e-05, "loss": 0.0086, "step": 57840 }, { "grad_norm": 0.1790495365858078, "learning_rate": 4.120164060780141e-05, "loss": 0.0056, "step": 57850 }, { "grad_norm": 0.1121848002076149, "learning_rate": 4.1185364397901935e-05, "loss": 0.0053, "step": 57860 }, { "grad_norm": 0.15874381363391876, "learning_rate": 4.116908915195776e-05, "loss": 0.0068, "step": 57870 }, { "grad_norm": 0.08258801698684692, "learning_rate": 4.115281487174872e-05, "loss": 0.0052, "step": 57880 }, { "grad_norm": 0.12421686202287674, "learning_rate": 4.113654155905452e-05, "loss": 0.0065, "step": 57890 }, { "grad_norm": 0.09564720094203949, "learning_rate": 4.112026921565483e-05, "loss": 0.0047, "step": 57900 }, { "grad_norm": 0.08376026153564453, "learning_rate": 4.1103997843329147e-05, "loss": 0.0058, "step": 57910 }, { "grad_norm": 0.1591099500656128, "learning_rate": 4.108772744385689e-05, "loss": 0.0055, "step": 57920 }, { "grad_norm": 0.08648394793272018, "learning_rate": 4.1071458019017346e-05, "loss": 0.0069, "step": 57930 }, { "grad_norm": 0.10243793576955795, "learning_rate": 4.1055189570589734e-05, "loss": 0.0052, "step": 57940 }, { "grad_norm": 0.10962341725826263, "learning_rate": 4.103892210035315e-05, "loss": 0.0051, "step": 57950 }, { "grad_norm": 0.08514393866062164, "learning_rate": 4.102265561008656e-05, "loss": 0.0058, "step": 57960 }, { "grad_norm": 0.15392838418483734, "learning_rate": 4.100639010156885e-05, "loss": 0.0076, "step": 57970 }, { "grad_norm": 0.12021007388830185, "learning_rate": 4.099012557657882e-05, "loss": 0.0047, "step": 57980 }, { "grad_norm": 0.09752213954925537, "learning_rate": 4.09738620368951e-05, "loss": 0.0078, "step": 57990 }, { "grad_norm": 0.13423104584217072, "learning_rate": 4.095759948429626e-05, "loss": 0.0071, "step": 58000 }, { "grad_norm": 0.10474158078432083, "learning_rate": 4.094133792056075e-05, "loss": 0.0076, "step": 58010 }, { "grad_norm": 0.08807393908500671, "learning_rate": 4.092507734746688e-05, "loss": 0.0051, "step": 58020 }, { "grad_norm": 0.11888016015291214, "learning_rate": 4.090881776679293e-05, "loss": 0.0057, "step": 58030 }, { "grad_norm": 0.12214174121618271, "learning_rate": 4.0892559180316994e-05, "loss": 0.0066, "step": 58040 }, { "grad_norm": 0.13373510539531708, "learning_rate": 4.08763015898171e-05, "loss": 0.007, "step": 58050 }, { "grad_norm": 0.14500677585601807, "learning_rate": 4.0860044997071145e-05, "loss": 0.007, "step": 58060 }, { "grad_norm": 0.09704746305942535, "learning_rate": 4.084378940385693e-05, "loss": 0.0084, "step": 58070 }, { "grad_norm": 0.1959729641675949, "learning_rate": 4.082753481195211e-05, "loss": 0.0098, "step": 58080 }, { "grad_norm": 0.1657520979642868, "learning_rate": 4.0811281223134324e-05, "loss": 0.0057, "step": 58090 }, { "grad_norm": 0.09260405600070953, "learning_rate": 4.0795028639181e-05, "loss": 0.0073, "step": 58100 }, { "grad_norm": 0.11782171577215195, "learning_rate": 4.077877706186951e-05, "loss": 0.0074, "step": 58110 }, { "grad_norm": 0.1147746816277504, "learning_rate": 4.076252649297709e-05, "loss": 0.0079, "step": 58120 }, { "grad_norm": 0.11983578652143478, "learning_rate": 4.07462769342809e-05, "loss": 0.0078, "step": 58130 }, { "grad_norm": 0.1335531622171402, "learning_rate": 4.073002838755793e-05, "loss": 0.0079, "step": 58140 }, { "grad_norm": 0.13360729813575745, "learning_rate": 4.0713780854585115e-05, "loss": 0.0057, "step": 58150 }, { "grad_norm": 0.124890998005867, "learning_rate": 4.069753433713927e-05, "loss": 0.0074, "step": 58160 }, { "grad_norm": 0.14961253106594086, "learning_rate": 4.068128883699709e-05, "loss": 0.0069, "step": 58170 }, { "grad_norm": 0.16201166808605194, "learning_rate": 4.0665044355935155e-05, "loss": 0.0067, "step": 58180 }, { "grad_norm": 0.2644840180873871, "learning_rate": 4.064880089572993e-05, "loss": 0.0066, "step": 58190 }, { "grad_norm": 0.17250023782253265, "learning_rate": 4.063255845815779e-05, "loss": 0.0069, "step": 58200 }, { "grad_norm": 0.09958498924970627, "learning_rate": 4.0616317044994945e-05, "loss": 0.007, "step": 58210 }, { "grad_norm": 0.10391553491353989, "learning_rate": 4.0600076658017585e-05, "loss": 0.0055, "step": 58220 }, { "grad_norm": 0.1485825628042221, "learning_rate": 4.0583837299001706e-05, "loss": 0.005, "step": 58230 }, { "grad_norm": 0.09428347647190094, "learning_rate": 4.056759896972322e-05, "loss": 0.0064, "step": 58240 }, { "grad_norm": 0.07999923080205917, "learning_rate": 4.055136167195793e-05, "loss": 0.0059, "step": 58250 }, { "grad_norm": 0.10664070397615433, "learning_rate": 4.053512540748152e-05, "loss": 0.0073, "step": 58260 }, { "grad_norm": 0.09558174014091492, "learning_rate": 4.051889017806955e-05, "loss": 0.0069, "step": 58270 }, { "grad_norm": 0.09785564243793488, "learning_rate": 4.0502655985497516e-05, "loss": 0.0063, "step": 58280 }, { "grad_norm": 0.11597879976034164, "learning_rate": 4.048642283154074e-05, "loss": 0.0063, "step": 58290 }, { "grad_norm": 0.11387576907873154, "learning_rate": 4.0470190717974464e-05, "loss": 0.006, "step": 58300 }, { "grad_norm": 0.09626719355583191, "learning_rate": 4.045395964657381e-05, "loss": 0.0073, "step": 58310 }, { "grad_norm": 0.11123599857091904, "learning_rate": 4.043772961911376e-05, "loss": 0.0054, "step": 58320 }, { "grad_norm": 0.11520525068044662, "learning_rate": 4.042150063736924e-05, "loss": 0.0071, "step": 58330 }, { "grad_norm": 0.12349722534418106, "learning_rate": 4.0405272703114976e-05, "loss": 0.0074, "step": 58340 }, { "grad_norm": 0.14925162494182587, "learning_rate": 4.038904581812569e-05, "loss": 0.0073, "step": 58350 }, { "grad_norm": 0.1363936960697174, "learning_rate": 4.03728199841759e-05, "loss": 0.006, "step": 58360 }, { "grad_norm": 0.13614438474178314, "learning_rate": 4.0356595203040044e-05, "loss": 0.0067, "step": 58370 }, { "grad_norm": 0.08936632424592972, "learning_rate": 4.0340371476492426e-05, "loss": 0.0048, "step": 58380 }, { "grad_norm": 0.10004864633083344, "learning_rate": 4.032414880630727e-05, "loss": 0.0061, "step": 58390 }, { "grad_norm": 0.10031197220087051, "learning_rate": 4.0307927194258646e-05, "loss": 0.0103, "step": 58400 }, { "grad_norm": 0.1710497885942459, "learning_rate": 4.029170664212054e-05, "loss": 0.0073, "step": 58410 }, { "grad_norm": 0.14776083827018738, "learning_rate": 4.0275487151666794e-05, "loss": 0.0062, "step": 58420 }, { "grad_norm": 0.15802858769893646, "learning_rate": 4.025926872467116e-05, "loss": 0.0051, "step": 58430 }, { "grad_norm": 0.11488331854343414, "learning_rate": 4.024305136290725e-05, "loss": 0.0064, "step": 58440 }, { "grad_norm": 0.17618033289909363, "learning_rate": 4.022683506814858e-05, "loss": 0.007, "step": 58450 }, { "grad_norm": 0.11805538833141327, "learning_rate": 4.021061984216853e-05, "loss": 0.0087, "step": 58460 }, { "grad_norm": 0.13920506834983826, "learning_rate": 4.019440568674035e-05, "loss": 0.0066, "step": 58470 }, { "grad_norm": 0.1900320202112198, "learning_rate": 4.017819260363724e-05, "loss": 0.0059, "step": 58480 }, { "grad_norm": 0.11148320883512497, "learning_rate": 4.016198059463223e-05, "loss": 0.0066, "step": 58490 }, { "grad_norm": 0.13364467024803162, "learning_rate": 4.0145769661498233e-05, "loss": 0.0076, "step": 58500 }, { "grad_norm": 0.12491337954998016, "learning_rate": 4.012955980600804e-05, "loss": 0.0058, "step": 58510 }, { "grad_norm": 0.10675006359815598, "learning_rate": 4.0113351029934355e-05, "loss": 0.0086, "step": 58520 }, { "grad_norm": 0.1749853491783142, "learning_rate": 4.0097143335049714e-05, "loss": 0.0059, "step": 58530 }, { "grad_norm": 0.16064167022705078, "learning_rate": 4.0080936723126606e-05, "loss": 0.0093, "step": 58540 }, { "grad_norm": 0.22764118015766144, "learning_rate": 4.006473119593735e-05, "loss": 0.0074, "step": 58550 }, { "grad_norm": 0.1223105862736702, "learning_rate": 4.004852675525415e-05, "loss": 0.0064, "step": 58560 }, { "grad_norm": 0.1440538913011551, "learning_rate": 4.00323234028491e-05, "loss": 0.0068, "step": 58570 }, { "grad_norm": 0.13173122704029083, "learning_rate": 4.0016121140494186e-05, "loss": 0.0072, "step": 58580 }, { "grad_norm": 0.11718092113733292, "learning_rate": 3.999991996996122e-05, "loss": 0.0061, "step": 58590 }, { "grad_norm": 0.08851552754640579, "learning_rate": 3.9983719893022e-05, "loss": 0.0053, "step": 58600 }, { "grad_norm": 0.12421390414237976, "learning_rate": 3.996752091144812e-05, "loss": 0.0065, "step": 58610 }, { "grad_norm": 0.13406580686569214, "learning_rate": 3.9951323027011055e-05, "loss": 0.0062, "step": 58620 }, { "grad_norm": 0.11924286931753159, "learning_rate": 3.993512624148219e-05, "loss": 0.0073, "step": 58630 }, { "grad_norm": 0.1085299700498581, "learning_rate": 3.99189305566328e-05, "loss": 0.0073, "step": 58640 }, { "grad_norm": 0.1430048644542694, "learning_rate": 3.990273597423401e-05, "loss": 0.0059, "step": 58650 }, { "grad_norm": 0.11944271624088287, "learning_rate": 3.988654249605681e-05, "loss": 0.0075, "step": 58660 }, { "grad_norm": 0.14621716737747192, "learning_rate": 3.987035012387213e-05, "loss": 0.0047, "step": 58670 }, { "grad_norm": 0.08661412447690964, "learning_rate": 3.9854158859450735e-05, "loss": 0.0044, "step": 58680 }, { "grad_norm": 0.08551695197820663, "learning_rate": 3.983796870456328e-05, "loss": 0.0061, "step": 58690 }, { "grad_norm": 0.15154515206813812, "learning_rate": 3.982177966098028e-05, "loss": 0.0067, "step": 58700 }, { "grad_norm": 0.11012198776006699, "learning_rate": 3.980559173047216e-05, "loss": 0.0052, "step": 58710 }, { "grad_norm": 0.1235063374042511, "learning_rate": 3.9789404914809176e-05, "loss": 0.0055, "step": 58720 }, { "grad_norm": 0.106600821018219, "learning_rate": 3.9773219215761544e-05, "loss": 0.0067, "step": 58730 }, { "grad_norm": 0.10944601148366928, "learning_rate": 3.975703463509928e-05, "loss": 0.0064, "step": 58740 }, { "grad_norm": 0.1324702352285385, "learning_rate": 3.974085117459231e-05, "loss": 0.0058, "step": 58750 }, { "grad_norm": 0.11792008578777313, "learning_rate": 3.972466883601043e-05, "loss": 0.0063, "step": 58760 }, { "grad_norm": 0.17879725992679596, "learning_rate": 3.9708487621123316e-05, "loss": 0.007, "step": 58770 }, { "grad_norm": 0.13517038524150848, "learning_rate": 3.969230753170051e-05, "loss": 0.0076, "step": 58780 }, { "grad_norm": 0.10819555073976517, "learning_rate": 3.967612856951146e-05, "loss": 0.0049, "step": 58790 }, { "grad_norm": 0.08781200647354126, "learning_rate": 3.9659950736325476e-05, "loss": 0.0082, "step": 58800 }, { "grad_norm": 0.13876888155937195, "learning_rate": 3.964377403391174e-05, "loss": 0.0087, "step": 58810 }, { "grad_norm": 0.13445959985256195, "learning_rate": 3.9627598464039303e-05, "loss": 0.0051, "step": 58820 }, { "grad_norm": 0.170535147190094, "learning_rate": 3.9611424028477096e-05, "loss": 0.0074, "step": 58830 }, { "grad_norm": 0.10584019869565964, "learning_rate": 3.959525072899394e-05, "loss": 0.0055, "step": 58840 }, { "grad_norm": 0.14411960542201996, "learning_rate": 3.9579078567358506e-05, "loss": 0.006, "step": 58850 }, { "grad_norm": 0.1742650270462036, "learning_rate": 3.956290754533939e-05, "loss": 0.0078, "step": 58860 }, { "grad_norm": 0.15710721909999847, "learning_rate": 3.954673766470501e-05, "loss": 0.0091, "step": 58870 }, { "grad_norm": 0.1532592922449112, "learning_rate": 3.9530568927223674e-05, "loss": 0.0081, "step": 58880 }, { "grad_norm": 0.10136507451534271, "learning_rate": 3.95144013346636e-05, "loss": 0.0084, "step": 58890 }, { "grad_norm": 0.11937779933214188, "learning_rate": 3.949823488879282e-05, "loss": 0.0059, "step": 58900 }, { "grad_norm": 0.11405239254236221, "learning_rate": 3.9482069591379276e-05, "loss": 0.006, "step": 58910 }, { "grad_norm": 0.09067899733781815, "learning_rate": 3.9465905444190784e-05, "loss": 0.0048, "step": 58920 }, { "grad_norm": 0.12973931431770325, "learning_rate": 3.944974244899505e-05, "loss": 0.0067, "step": 58930 }, { "grad_norm": 0.13173037767410278, "learning_rate": 3.9433580607559616e-05, "loss": 0.0051, "step": 58940 }, { "grad_norm": 0.13716082274913788, "learning_rate": 3.941741992165193e-05, "loss": 0.0085, "step": 58950 }, { "grad_norm": 0.09667752683162689, "learning_rate": 3.940126039303928e-05, "loss": 0.0073, "step": 58960 }, { "grad_norm": 0.12494947761297226, "learning_rate": 3.938510202348886e-05, "loss": 0.0068, "step": 58970 }, { "grad_norm": 0.12913307547569275, "learning_rate": 3.93689448147677e-05, "loss": 0.0083, "step": 58980 }, { "grad_norm": 0.15591807663440704, "learning_rate": 3.935278876864278e-05, "loss": 0.0102, "step": 58990 }, { "grad_norm": 0.10676116496324539, "learning_rate": 3.933663388688087e-05, "loss": 0.0068, "step": 59000 }, { "grad_norm": 0.15512560307979584, "learning_rate": 3.9320480171248645e-05, "loss": 0.0073, "step": 59010 }, { "grad_norm": 0.10944211483001709, "learning_rate": 3.930432762351265e-05, "loss": 0.0075, "step": 59020 }, { "grad_norm": 0.09062949568033218, "learning_rate": 3.928817624543931e-05, "loss": 0.0061, "step": 59030 }, { "grad_norm": 0.14989255368709564, "learning_rate": 3.9272026038794885e-05, "loss": 0.0073, "step": 59040 }, { "grad_norm": 0.11128377169370651, "learning_rate": 3.9255877005345586e-05, "loss": 0.0064, "step": 59050 }, { "grad_norm": 0.13626639544963837, "learning_rate": 3.923972914685743e-05, "loss": 0.0056, "step": 59060 }, { "grad_norm": 0.1255524605512619, "learning_rate": 3.922358246509631e-05, "loss": 0.0049, "step": 59070 }, { "grad_norm": 0.11338762193918228, "learning_rate": 3.920743696182801e-05, "loss": 0.0047, "step": 59080 }, { "grad_norm": 0.11215103417634964, "learning_rate": 3.9191292638818164e-05, "loss": 0.006, "step": 59090 }, { "grad_norm": 0.12496239691972733, "learning_rate": 3.9175149497832316e-05, "loss": 0.0065, "step": 59100 }, { "grad_norm": 0.16122518479824066, "learning_rate": 3.915900754063584e-05, "loss": 0.0067, "step": 59110 }, { "grad_norm": 0.10041507333517075, "learning_rate": 3.9142866768994014e-05, "loss": 0.0053, "step": 59120 }, { "grad_norm": 0.18534517288208008, "learning_rate": 3.912672718467193e-05, "loss": 0.0055, "step": 59130 }, { "grad_norm": 0.09756584465503693, "learning_rate": 3.911058878943463e-05, "loss": 0.0074, "step": 59140 }, { "grad_norm": 0.11346521228551865, "learning_rate": 3.9094451585046974e-05, "loss": 0.0063, "step": 59150 }, { "grad_norm": 0.1241532489657402, "learning_rate": 3.90783155732737e-05, "loss": 0.0054, "step": 59160 }, { "grad_norm": 0.11950608342885971, "learning_rate": 3.9062180755879373e-05, "loss": 0.0058, "step": 59170 }, { "grad_norm": 0.0936231017112732, "learning_rate": 3.904604713462855e-05, "loss": 0.0048, "step": 59180 }, { "grad_norm": 0.12514245510101318, "learning_rate": 3.902991471128554e-05, "loss": 0.0057, "step": 59190 }, { "grad_norm": 0.0887029767036438, "learning_rate": 3.901378348761456e-05, "loss": 0.0055, "step": 59200 }, { "grad_norm": 0.08082146942615509, "learning_rate": 3.899765346537971e-05, "loss": 0.0053, "step": 59210 }, { "grad_norm": 0.0905381366610527, "learning_rate": 3.898152464634493e-05, "loss": 0.0078, "step": 59220 }, { "grad_norm": 0.13937941193580627, "learning_rate": 3.896539703227402e-05, "loss": 0.0066, "step": 59230 }, { "grad_norm": 0.1065593957901001, "learning_rate": 3.894927062493073e-05, "loss": 0.0052, "step": 59240 }, { "grad_norm": 0.12492584437131882, "learning_rate": 3.893314542607859e-05, "loss": 0.0055, "step": 59250 }, { "grad_norm": 0.11823965609073639, "learning_rate": 3.891702143748103e-05, "loss": 0.0065, "step": 59260 }, { "grad_norm": 0.13292379677295685, "learning_rate": 3.890089866090133e-05, "loss": 0.0068, "step": 59270 }, { "grad_norm": 0.12315236777067184, "learning_rate": 3.888477709810267e-05, "loss": 0.0058, "step": 59280 }, { "grad_norm": 0.14367614686489105, "learning_rate": 3.886865675084806e-05, "loss": 0.0059, "step": 59290 }, { "grad_norm": 0.08883925527334213, "learning_rate": 3.885253762090043e-05, "loss": 0.0061, "step": 59300 }, { "grad_norm": 0.10377298295497894, "learning_rate": 3.8836419710022534e-05, "loss": 0.0055, "step": 59310 }, { "grad_norm": 0.15158315002918243, "learning_rate": 3.882030301997698e-05, "loss": 0.0052, "step": 59320 }, { "grad_norm": 0.15584227442741394, "learning_rate": 3.880418755252629e-05, "loss": 0.0074, "step": 59330 }, { "grad_norm": 0.1756119728088379, "learning_rate": 3.87880733094328e-05, "loss": 0.0059, "step": 59340 }, { "grad_norm": 0.10885696858167648, "learning_rate": 3.877196029245877e-05, "loss": 0.0063, "step": 59350 }, { "grad_norm": 0.11949575692415237, "learning_rate": 3.875584850336627e-05, "loss": 0.0049, "step": 59360 }, { "grad_norm": 0.09257914870977402, "learning_rate": 3.873973794391728e-05, "loss": 0.0089, "step": 59370 }, { "grad_norm": 0.14359413087368011, "learning_rate": 3.872362861587361e-05, "loss": 0.0068, "step": 59380 }, { "grad_norm": 0.12407132983207703, "learning_rate": 3.870752052099698e-05, "loss": 0.008, "step": 59390 }, { "grad_norm": 0.18984811007976532, "learning_rate": 3.8691413661048916e-05, "loss": 0.0074, "step": 59400 }, { "grad_norm": 0.14583788812160492, "learning_rate": 3.867530803779085e-05, "loss": 0.006, "step": 59410 }, { "grad_norm": 0.1483808159828186, "learning_rate": 3.8659203652984054e-05, "loss": 0.0077, "step": 59420 }, { "grad_norm": 0.1043081283569336, "learning_rate": 3.864310050838972e-05, "loss": 0.0083, "step": 59430 }, { "grad_norm": 0.10302407294511795, "learning_rate": 3.8626998605768835e-05, "loss": 0.0067, "step": 59440 }, { "grad_norm": 0.0997040644288063, "learning_rate": 3.8610897946882294e-05, "loss": 0.0052, "step": 59450 }, { "grad_norm": 0.12124191224575043, "learning_rate": 3.859479853349082e-05, "loss": 0.0047, "step": 59460 }, { "grad_norm": 0.09312205761671066, "learning_rate": 3.857870036735503e-05, "loss": 0.0048, "step": 59470 }, { "grad_norm": 0.12624135613441467, "learning_rate": 3.856260345023538e-05, "loss": 0.0071, "step": 59480 }, { "grad_norm": 0.09260203689336777, "learning_rate": 3.854650778389224e-05, "loss": 0.0053, "step": 59490 }, { "grad_norm": 0.09838856011629105, "learning_rate": 3.8530413370085786e-05, "loss": 0.005, "step": 59500 }, { "grad_norm": 0.10244248807430267, "learning_rate": 3.851432021057608e-05, "loss": 0.0053, "step": 59510 }, { "grad_norm": 0.11839284747838974, "learning_rate": 3.849822830712306e-05, "loss": 0.0052, "step": 59520 }, { "grad_norm": 0.15181423723697662, "learning_rate": 3.848213766148649e-05, "loss": 0.0069, "step": 59530 }, { "grad_norm": 0.12943360209465027, "learning_rate": 3.846604827542602e-05, "loss": 0.0046, "step": 59540 }, { "grad_norm": 0.08202537149190903, "learning_rate": 3.844996015070117e-05, "loss": 0.0061, "step": 59550 }, { "grad_norm": 0.14136549830436707, "learning_rate": 3.843387328907132e-05, "loss": 0.0056, "step": 59560 }, { "grad_norm": 0.1080055832862854, "learning_rate": 3.8417787692295696e-05, "loss": 0.0101, "step": 59570 }, { "grad_norm": 0.1832386702299118, "learning_rate": 3.84017033621334e-05, "loss": 0.01, "step": 59580 }, { "grad_norm": 0.15220797061920166, "learning_rate": 3.838562030034338e-05, "loss": 0.0059, "step": 59590 }, { "grad_norm": 0.13350237905979156, "learning_rate": 3.836953850868447e-05, "loss": 0.0074, "step": 59600 }, { "grad_norm": 0.17869932949543, "learning_rate": 3.835345798891533e-05, "loss": 0.0053, "step": 59610 }, { "grad_norm": 0.11459322273731232, "learning_rate": 3.833737874279452e-05, "loss": 0.0076, "step": 59620 }, { "grad_norm": 0.11995068192481995, "learning_rate": 3.8321300772080433e-05, "loss": 0.0085, "step": 59630 }, { "grad_norm": 0.14943189918994904, "learning_rate": 3.8305224078531345e-05, "loss": 0.0052, "step": 59640 }, { "grad_norm": 0.13764819502830505, "learning_rate": 3.828914866390537e-05, "loss": 0.0047, "step": 59650 }, { "grad_norm": 0.10584290325641632, "learning_rate": 3.827307452996048e-05, "loss": 0.0056, "step": 59660 }, { "grad_norm": 0.16773894429206848, "learning_rate": 3.825700167845454e-05, "loss": 0.0062, "step": 59670 }, { "grad_norm": 0.15772302448749542, "learning_rate": 3.8240930111145214e-05, "loss": 0.0045, "step": 59680 }, { "grad_norm": 0.14773865044116974, "learning_rate": 3.8224859829790116e-05, "loss": 0.0058, "step": 59690 }, { "grad_norm": 0.14923103153705597, "learning_rate": 3.820879083614664e-05, "loss": 0.0056, "step": 59700 }, { "grad_norm": 0.1455094814300537, "learning_rate": 3.819272313197208e-05, "loss": 0.0053, "step": 59710 }, { "grad_norm": 0.18721315264701843, "learning_rate": 3.817665671902355e-05, "loss": 0.0048, "step": 59720 }, { "grad_norm": 0.16868680715560913, "learning_rate": 3.816059159905808e-05, "loss": 0.0054, "step": 59730 }, { "grad_norm": 0.165322944521904, "learning_rate": 3.8144527773832476e-05, "loss": 0.0066, "step": 59740 }, { "grad_norm": 0.11221658438444138, "learning_rate": 3.812846524510352e-05, "loss": 0.007, "step": 59750 }, { "grad_norm": 0.14128844439983368, "learning_rate": 3.811240401462775e-05, "loss": 0.006, "step": 59760 }, { "grad_norm": 0.14490896463394165, "learning_rate": 3.809634408416162e-05, "loss": 0.0058, "step": 59770 }, { "grad_norm": 0.09063790738582611, "learning_rate": 3.8080285455461394e-05, "loss": 0.0072, "step": 59780 }, { "grad_norm": 0.16327476501464844, "learning_rate": 3.806422813028323e-05, "loss": 0.0057, "step": 59790 }, { "grad_norm": 0.12814617156982422, "learning_rate": 3.8048172110383114e-05, "loss": 0.0058, "step": 59800 }, { "grad_norm": 0.15569911897182465, "learning_rate": 3.8032117397516944e-05, "loss": 0.006, "step": 59810 }, { "grad_norm": 0.119504414498806, "learning_rate": 3.801606399344042e-05, "loss": 0.0045, "step": 59820 }, { "grad_norm": 0.14169123768806458, "learning_rate": 3.800001189990913e-05, "loss": 0.005, "step": 59830 }, { "grad_norm": 0.10580454766750336, "learning_rate": 3.798396111867847e-05, "loss": 0.0061, "step": 59840 }, { "grad_norm": 0.1287897676229477, "learning_rate": 3.796791165150378e-05, "loss": 0.0069, "step": 59850 }, { "grad_norm": 0.19206345081329346, "learning_rate": 3.795186350014018e-05, "loss": 0.0074, "step": 59860 }, { "grad_norm": 0.0847475603222847, "learning_rate": 3.793581666634266e-05, "loss": 0.0058, "step": 59870 }, { "grad_norm": 0.115706667304039, "learning_rate": 3.791977115186609e-05, "loss": 0.0076, "step": 59880 }, { "grad_norm": 0.1050005853176117, "learning_rate": 3.79037269584652e-05, "loss": 0.0082, "step": 59890 }, { "grad_norm": 0.10894601047039032, "learning_rate": 3.7887684087894545e-05, "loss": 0.007, "step": 59900 }, { "grad_norm": 0.11310026794672012, "learning_rate": 3.787164254190854e-05, "loss": 0.0062, "step": 59910 }, { "grad_norm": 0.12842288613319397, "learning_rate": 3.7855602322261473e-05, "loss": 0.0074, "step": 59920 }, { "grad_norm": 0.1683162897825241, "learning_rate": 3.783956343070746e-05, "loss": 0.006, "step": 59930 }, { "grad_norm": 0.08378086239099503, "learning_rate": 3.7823525869000526e-05, "loss": 0.0058, "step": 59940 }, { "grad_norm": 0.0818273201584816, "learning_rate": 3.78074896388945e-05, "loss": 0.0043, "step": 59950 }, { "grad_norm": 0.12803031504154205, "learning_rate": 3.7791454742143075e-05, "loss": 0.0049, "step": 59960 }, { "grad_norm": 0.16279785335063934, "learning_rate": 3.7775421180499795e-05, "loss": 0.005, "step": 59970 }, { "grad_norm": 0.14834867417812347, "learning_rate": 3.7759388955718086e-05, "loss": 0.0076, "step": 59980 }, { "grad_norm": 0.1755141019821167, "learning_rate": 3.774335806955117e-05, "loss": 0.0053, "step": 59990 }, { "grad_norm": 0.15283603966236115, "learning_rate": 3.7727328523752215e-05, "loss": 0.0076, "step": 60000 }, { "grad_norm": 0.13229615986347198, "learning_rate": 3.771130032007415e-05, "loss": 0.0067, "step": 60010 }, { "grad_norm": 0.1418183594942093, "learning_rate": 3.769527346026982e-05, "loss": 0.0103, "step": 60020 }, { "grad_norm": 0.08155514299869537, "learning_rate": 3.767924794609188e-05, "loss": 0.006, "step": 60030 }, { "grad_norm": 0.09983082860708237, "learning_rate": 3.7663223779292854e-05, "loss": 0.0067, "step": 60040 }, { "grad_norm": 0.14099669456481934, "learning_rate": 3.764720096162512e-05, "loss": 0.006, "step": 60050 }, { "grad_norm": 0.1271630823612213, "learning_rate": 3.763117949484092e-05, "loss": 0.0086, "step": 60060 }, { "grad_norm": 0.16197654604911804, "learning_rate": 3.7615159380692344e-05, "loss": 0.0064, "step": 60070 }, { "grad_norm": 0.09969203174114227, "learning_rate": 3.759914062093132e-05, "loss": 0.0059, "step": 60080 }, { "grad_norm": 0.11568230390548706, "learning_rate": 3.7583123217309615e-05, "loss": 0.0087, "step": 60090 }, { "grad_norm": 0.16810062527656555, "learning_rate": 3.7567107171578904e-05, "loss": 0.0054, "step": 60100 }, { "grad_norm": 0.12395858019590378, "learning_rate": 3.755109248549066e-05, "loss": 0.0058, "step": 60110 }, { "grad_norm": 0.13242290914058685, "learning_rate": 3.75350791607962e-05, "loss": 0.0062, "step": 60120 }, { "grad_norm": 0.08572085946798325, "learning_rate": 3.751906719924676e-05, "loss": 0.0065, "step": 60130 }, { "grad_norm": 0.10457800328731537, "learning_rate": 3.750305660259337e-05, "loss": 0.007, "step": 60140 }, { "grad_norm": 0.10260425508022308, "learning_rate": 3.748704737258693e-05, "loss": 0.006, "step": 60150 }, { "grad_norm": 0.1059797927737236, "learning_rate": 3.747103951097816e-05, "loss": 0.007, "step": 60160 }, { "grad_norm": 0.11372777819633484, "learning_rate": 3.745503301951768e-05, "loss": 0.0063, "step": 60170 }, { "grad_norm": 0.12455368041992188, "learning_rate": 3.743902789995592e-05, "loss": 0.0074, "step": 60180 }, { "grad_norm": 0.12766076624393463, "learning_rate": 3.7423024154043183e-05, "loss": 0.005, "step": 60190 }, { "grad_norm": 0.09084492921829224, "learning_rate": 3.740702178352963e-05, "loss": 0.0052, "step": 60200 }, { "grad_norm": 0.12762250006198883, "learning_rate": 3.739102079016523e-05, "loss": 0.0058, "step": 60210 }, { "grad_norm": 0.10442392528057098, "learning_rate": 3.7375021175699846e-05, "loss": 0.0048, "step": 60220 }, { "grad_norm": 0.10496868938207626, "learning_rate": 3.735902294188317e-05, "loss": 0.0059, "step": 60230 }, { "grad_norm": 0.12298816442489624, "learning_rate": 3.734302609046473e-05, "loss": 0.0044, "step": 60240 }, { "grad_norm": 0.15012846887111664, "learning_rate": 3.732703062319391e-05, "loss": 0.0054, "step": 60250 }, { "grad_norm": 0.0922180637717247, "learning_rate": 3.731103654181999e-05, "loss": 0.0062, "step": 60260 }, { "grad_norm": 0.13218259811401367, "learning_rate": 3.729504384809203e-05, "loss": 0.0076, "step": 60270 }, { "grad_norm": 0.12705126404762268, "learning_rate": 3.727905254375898e-05, "loss": 0.0069, "step": 60280 }, { "grad_norm": 0.10110469162464142, "learning_rate": 3.7263062630569614e-05, "loss": 0.0065, "step": 60290 }, { "grad_norm": 0.11901326477527618, "learning_rate": 3.724707411027256e-05, "loss": 0.0049, "step": 60300 }, { "grad_norm": 0.10349886864423752, "learning_rate": 3.723108698461631e-05, "loss": 0.0049, "step": 60310 }, { "grad_norm": 0.12606629729270935, "learning_rate": 3.72151012553492e-05, "loss": 0.006, "step": 60320 }, { "grad_norm": 0.13853757083415985, "learning_rate": 3.719911692421939e-05, "loss": 0.0079, "step": 60330 }, { "grad_norm": 0.13641208410263062, "learning_rate": 3.71831339929749e-05, "loss": 0.0061, "step": 60340 }, { "grad_norm": 0.1512281894683838, "learning_rate": 3.716715246336362e-05, "loss": 0.0067, "step": 60350 }, { "grad_norm": 0.15800268948078156, "learning_rate": 3.715117233713324e-05, "loss": 0.0076, "step": 60360 }, { "grad_norm": 0.08898501098155975, "learning_rate": 3.713519361603135e-05, "loss": 0.0066, "step": 60370 }, { "grad_norm": 0.14261238276958466, "learning_rate": 3.711921630180532e-05, "loss": 0.0065, "step": 60380 }, { "grad_norm": 0.161880761384964, "learning_rate": 3.710324039620245e-05, "loss": 0.0083, "step": 60390 }, { "grad_norm": 0.08033808320760727, "learning_rate": 3.708726590096982e-05, "loss": 0.0054, "step": 60400 }, { "grad_norm": 0.14147880673408508, "learning_rate": 3.707129281785437e-05, "loss": 0.0086, "step": 60410 }, { "grad_norm": 0.11501006782054901, "learning_rate": 3.705532114860291e-05, "loss": 0.007, "step": 60420 }, { "grad_norm": 0.11653851717710495, "learning_rate": 3.703935089496207e-05, "loss": 0.0038, "step": 60430 }, { "grad_norm": 0.0725301131606102, "learning_rate": 3.702338205867829e-05, "loss": 0.0057, "step": 60440 }, { "grad_norm": 0.15634499490261078, "learning_rate": 3.7007414641497976e-05, "loss": 0.0064, "step": 60450 }, { "grad_norm": 0.09855052828788757, "learning_rate": 3.699144864516726e-05, "loss": 0.0071, "step": 60460 }, { "grad_norm": 0.08443493396043777, "learning_rate": 3.6975484071432153e-05, "loss": 0.0051, "step": 60470 }, { "grad_norm": 0.19218258559703827, "learning_rate": 3.695952092203854e-05, "loss": 0.0048, "step": 60480 }, { "grad_norm": 0.13548819720745087, "learning_rate": 3.69435591987321e-05, "loss": 0.0051, "step": 60490 }, { "grad_norm": 0.10379159450531006, "learning_rate": 3.6927598903258374e-05, "loss": 0.0055, "step": 60500 }, { "grad_norm": 0.099985770881176, "learning_rate": 3.691164003736281e-05, "loss": 0.0045, "step": 60510 }, { "grad_norm": 0.1251041293144226, "learning_rate": 3.689568260279061e-05, "loss": 0.006, "step": 60520 }, { "grad_norm": 0.11226349323987961, "learning_rate": 3.687972660128686e-05, "loss": 0.0051, "step": 60530 }, { "grad_norm": 0.11926886439323425, "learning_rate": 3.686377203459648e-05, "loss": 0.0059, "step": 60540 }, { "grad_norm": 0.10717789083719254, "learning_rate": 3.6847818904464226e-05, "loss": 0.0056, "step": 60550 }, { "grad_norm": 0.16887077689170837, "learning_rate": 3.683186721263474e-05, "loss": 0.0063, "step": 60560 }, { "grad_norm": 0.09982556104660034, "learning_rate": 3.6815916960852435e-05, "loss": 0.0061, "step": 60570 }, { "grad_norm": 0.16725040972232819, "learning_rate": 3.679996815086165e-05, "loss": 0.0053, "step": 60580 }, { "grad_norm": 0.1589452177286148, "learning_rate": 3.678402078440649e-05, "loss": 0.0055, "step": 60590 }, { "grad_norm": 0.2084205150604248, "learning_rate": 3.676807486323096e-05, "loss": 0.0077, "step": 60600 }, { "grad_norm": 0.12448908388614655, "learning_rate": 3.6752130389078864e-05, "loss": 0.008, "step": 60610 }, { "grad_norm": 0.10360414534807205, "learning_rate": 3.673618736369388e-05, "loss": 0.0081, "step": 60620 }, { "grad_norm": 0.08132769912481308, "learning_rate": 3.6720245788819474e-05, "loss": 0.0063, "step": 60630 }, { "grad_norm": 0.16571789979934692, "learning_rate": 3.6704305666199044e-05, "loss": 0.0065, "step": 60640 }, { "grad_norm": 0.1421719640493393, "learning_rate": 3.668836699757576e-05, "loss": 0.0061, "step": 60650 }, { "grad_norm": 0.1583578884601593, "learning_rate": 3.667242978469265e-05, "loss": 0.0066, "step": 60660 }, { "grad_norm": 0.21347762644290924, "learning_rate": 3.6656494029292575e-05, "loss": 0.0062, "step": 60670 }, { "grad_norm": 0.10488148778676987, "learning_rate": 3.664055973311825e-05, "loss": 0.0056, "step": 60680 }, { "grad_norm": 0.1572737991809845, "learning_rate": 3.662462689791221e-05, "loss": 0.007, "step": 60690 }, { "grad_norm": 0.10262294858694077, "learning_rate": 3.660869552541689e-05, "loss": 0.0075, "step": 60700 }, { "grad_norm": 0.2023920863866806, "learning_rate": 3.659276561737448e-05, "loss": 0.0064, "step": 60710 }, { "grad_norm": 0.07633671909570694, "learning_rate": 3.657683717552708e-05, "loss": 0.0052, "step": 60720 }, { "grad_norm": 0.13373111188411713, "learning_rate": 3.6560910201616574e-05, "loss": 0.0087, "step": 60730 }, { "grad_norm": 0.1122470423579216, "learning_rate": 3.654498469738472e-05, "loss": 0.0065, "step": 60740 }, { "grad_norm": 0.12331818044185638, "learning_rate": 3.6529060664573126e-05, "loss": 0.0056, "step": 60750 }, { "grad_norm": 0.1060282364487648, "learning_rate": 3.6513138104923176e-05, "loss": 0.0069, "step": 60760 }, { "grad_norm": 0.12184758484363556, "learning_rate": 3.6497217020176176e-05, "loss": 0.0053, "step": 60770 }, { "grad_norm": 0.11135238409042358, "learning_rate": 3.648129741207323e-05, "loss": 0.005, "step": 60780 }, { "grad_norm": 0.07531139999628067, "learning_rate": 3.646537928235527e-05, "loss": 0.0052, "step": 60790 }, { "grad_norm": 0.0974794402718544, "learning_rate": 3.6449462632763075e-05, "loss": 0.0058, "step": 60800 }, { "grad_norm": 0.11491678655147552, "learning_rate": 3.643354746503729e-05, "loss": 0.0056, "step": 60810 }, { "grad_norm": 0.1301293671131134, "learning_rate": 3.6417633780918345e-05, "loss": 0.0051, "step": 60820 }, { "grad_norm": 0.16447338461875916, "learning_rate": 3.640172158214655e-05, "loss": 0.0051, "step": 60830 }, { "grad_norm": 0.21749253571033478, "learning_rate": 3.6385810870462035e-05, "loss": 0.007, "step": 60840 }, { "grad_norm": 0.12267079949378967, "learning_rate": 3.636990164760479e-05, "loss": 0.0062, "step": 60850 }, { "grad_norm": 0.10249996930360794, "learning_rate": 3.6353993915314596e-05, "loss": 0.0085, "step": 60860 }, { "grad_norm": 0.14089880883693695, "learning_rate": 3.6338087675331124e-05, "loss": 0.0059, "step": 60870 }, { "grad_norm": 0.14774052798748016, "learning_rate": 3.632218292939383e-05, "loss": 0.0073, "step": 60880 }, { "grad_norm": 0.21216526627540588, "learning_rate": 3.6306279679242033e-05, "loss": 0.0048, "step": 60890 }, { "grad_norm": 0.08296418935060501, "learning_rate": 3.629037792661492e-05, "loss": 0.0058, "step": 60900 }, { "grad_norm": 0.07484036684036255, "learning_rate": 3.627447767325147e-05, "loss": 0.0062, "step": 60910 }, { "grad_norm": 0.1277722865343094, "learning_rate": 3.62585789208905e-05, "loss": 0.005, "step": 60920 }, { "grad_norm": 0.1312277466058731, "learning_rate": 3.624268167127068e-05, "loss": 0.0049, "step": 60930 }, { "grad_norm": 0.17445504665374756, "learning_rate": 3.622678592613051e-05, "loss": 0.0071, "step": 60940 }, { "grad_norm": 0.10343536734580994, "learning_rate": 3.6210891687208304e-05, "loss": 0.0067, "step": 60950 }, { "grad_norm": 0.09344594180583954, "learning_rate": 3.619499895624228e-05, "loss": 0.0047, "step": 60960 }, { "grad_norm": 0.1042521670460701, "learning_rate": 3.61791077349704e-05, "loss": 0.0048, "step": 60970 }, { "grad_norm": 0.13780345022678375, "learning_rate": 3.616321802513053e-05, "loss": 0.0055, "step": 60980 }, { "grad_norm": 0.11301539093255997, "learning_rate": 3.6147329828460325e-05, "loss": 0.0048, "step": 60990 }, { "grad_norm": 0.12215906381607056, "learning_rate": 3.6131443146697306e-05, "loss": 0.0045, "step": 61000 }, { "grad_norm": 0.08873448520898819, "learning_rate": 3.611555798157879e-05, "loss": 0.0098, "step": 61010 }, { "grad_norm": 0.09506304562091827, "learning_rate": 3.6099674334842e-05, "loss": 0.0049, "step": 61020 }, { "grad_norm": 0.16273070871829987, "learning_rate": 3.6083792208223934e-05, "loss": 0.0074, "step": 61030 }, { "grad_norm": 0.11917969584465027, "learning_rate": 3.606791160346142e-05, "loss": 0.0058, "step": 61040 }, { "grad_norm": 0.15397268533706665, "learning_rate": 3.6052032522291134e-05, "loss": 0.0058, "step": 61050 }, { "grad_norm": 0.0807265192270279, "learning_rate": 3.603615496644962e-05, "loss": 0.0049, "step": 61060 }, { "grad_norm": 0.11912711709737778, "learning_rate": 3.60202789376732e-05, "loss": 0.0048, "step": 61070 }, { "grad_norm": 0.0764232724905014, "learning_rate": 3.6004404437698036e-05, "loss": 0.0069, "step": 61080 }, { "grad_norm": 0.14249195158481598, "learning_rate": 3.5988531468260184e-05, "loss": 0.0055, "step": 61090 }, { "grad_norm": 0.07327976822853088, "learning_rate": 3.597266003109546e-05, "loss": 0.0054, "step": 61100 }, { "grad_norm": 0.13233479857444763, "learning_rate": 3.5956790127939557e-05, "loss": 0.0056, "step": 61110 }, { "grad_norm": 0.08169447630643845, "learning_rate": 3.594092176052796e-05, "loss": 0.0054, "step": 61120 }, { "grad_norm": 0.08329003304243088, "learning_rate": 3.592505493059603e-05, "loss": 0.0055, "step": 61130 }, { "grad_norm": 0.10376084595918655, "learning_rate": 3.5909189639878905e-05, "loss": 0.0062, "step": 61140 }, { "grad_norm": 0.09744235128164291, "learning_rate": 3.5893325890111626e-05, "loss": 0.0045, "step": 61150 }, { "grad_norm": 0.10214725881814957, "learning_rate": 3.5877463683029034e-05, "loss": 0.0061, "step": 61160 }, { "grad_norm": 0.14581215381622314, "learning_rate": 3.586160302036578e-05, "loss": 0.006, "step": 61170 }, { "grad_norm": 0.1492418348789215, "learning_rate": 3.584574390385636e-05, "loss": 0.0065, "step": 61180 }, { "grad_norm": 0.1494307965040207, "learning_rate": 3.582988633523511e-05, "loss": 0.0053, "step": 61190 }, { "grad_norm": 0.11501973867416382, "learning_rate": 3.581403031623616e-05, "loss": 0.0075, "step": 61200 }, { "grad_norm": 0.11113297939300537, "learning_rate": 3.579817584859354e-05, "loss": 0.0065, "step": 61210 }, { "grad_norm": 0.15837638080120087, "learning_rate": 3.578232293404107e-05, "loss": 0.006, "step": 61220 }, { "grad_norm": 0.10940618067979813, "learning_rate": 3.5766471574312385e-05, "loss": 0.0071, "step": 61230 }, { "grad_norm": 0.11425189673900604, "learning_rate": 3.5750621771140964e-05, "loss": 0.0053, "step": 61240 }, { "grad_norm": 0.13798809051513672, "learning_rate": 3.5734773526260115e-05, "loss": 0.0062, "step": 61250 }, { "grad_norm": 0.09950363636016846, "learning_rate": 3.5718926841402996e-05, "loss": 0.0052, "step": 61260 }, { "grad_norm": 0.14700311422348022, "learning_rate": 3.5703081718302555e-05, "loss": 0.0055, "step": 61270 }, { "grad_norm": 0.10061334073543549, "learning_rate": 3.568723815869162e-05, "loss": 0.0053, "step": 61280 }, { "grad_norm": 0.1182139664888382, "learning_rate": 3.5671396164302777e-05, "loss": 0.0068, "step": 61290 }, { "grad_norm": 0.18437890708446503, "learning_rate": 3.565555573686853e-05, "loss": 0.0091, "step": 61300 }, { "grad_norm": 0.09770362824201584, "learning_rate": 3.563971687812113e-05, "loss": 0.0078, "step": 61310 }, { "grad_norm": 0.11564457416534424, "learning_rate": 3.5623879589792706e-05, "loss": 0.0054, "step": 61320 }, { "grad_norm": 0.12186261266469955, "learning_rate": 3.5608043873615185e-05, "loss": 0.0055, "step": 61330 }, { "grad_norm": 0.11108900606632233, "learning_rate": 3.559220973132035e-05, "loss": 0.0066, "step": 61340 }, { "grad_norm": 0.11661601066589355, "learning_rate": 3.557637716463981e-05, "loss": 0.0078, "step": 61350 }, { "grad_norm": 0.12075775861740112, "learning_rate": 3.556054617530497e-05, "loss": 0.0063, "step": 61360 }, { "grad_norm": 0.13956108689308167, "learning_rate": 3.5544716765047106e-05, "loss": 0.0055, "step": 61370 }, { "grad_norm": 0.1194257065653801, "learning_rate": 3.5528888935597274e-05, "loss": 0.0082, "step": 61380 }, { "grad_norm": 0.11516223102807999, "learning_rate": 3.551306268868636e-05, "loss": 0.0073, "step": 61390 }, { "grad_norm": 0.1136527732014656, "learning_rate": 3.5497238026045166e-05, "loss": 0.0081, "step": 61400 }, { "grad_norm": 0.1676129698753357, "learning_rate": 3.548141494940421e-05, "loss": 0.0054, "step": 61410 }, { "grad_norm": 0.1172952800989151, "learning_rate": 3.546559346049388e-05, "loss": 0.0062, "step": 61420 }, { "grad_norm": 0.11571570485830307, "learning_rate": 3.544977356104441e-05, "loss": 0.0057, "step": 61430 }, { "grad_norm": 0.13084129989147186, "learning_rate": 3.543395525278582e-05, "loss": 0.0067, "step": 61440 }, { "grad_norm": 0.10268957167863846, "learning_rate": 3.5418138537447985e-05, "loss": 0.0048, "step": 61450 }, { "grad_norm": 0.09384115785360336, "learning_rate": 3.5402323416760574e-05, "loss": 0.0041, "step": 61460 }, { "grad_norm": 0.12065331637859344, "learning_rate": 3.538650989245316e-05, "loss": 0.0063, "step": 61470 }, { "grad_norm": 0.1398148089647293, "learning_rate": 3.537069796625504e-05, "loss": 0.0061, "step": 61480 }, { "grad_norm": 0.10738114267587662, "learning_rate": 3.53548876398954e-05, "loss": 0.0054, "step": 61490 }, { "grad_norm": 0.09088965505361557, "learning_rate": 3.5339078915103214e-05, "loss": 0.0053, "step": 61500 }, { "grad_norm": 0.1573333889245987, "learning_rate": 3.532327179360733e-05, "loss": 0.0041, "step": 61510 }, { "grad_norm": 0.1424860805273056, "learning_rate": 3.530746627713636e-05, "loss": 0.005, "step": 61520 }, { "grad_norm": 0.1492898017168045, "learning_rate": 3.5291662367418805e-05, "loss": 0.007, "step": 61530 }, { "grad_norm": 0.10272236168384552, "learning_rate": 3.527586006618293e-05, "loss": 0.0062, "step": 61540 }, { "grad_norm": 0.11252737790346146, "learning_rate": 3.526005937515687e-05, "loss": 0.0065, "step": 61550 }, { "grad_norm": 0.090447798371315, "learning_rate": 3.524426029606856e-05, "loss": 0.0088, "step": 61560 }, { "grad_norm": 0.14857779443264008, "learning_rate": 3.5228462830645746e-05, "loss": 0.0061, "step": 61570 }, { "grad_norm": 0.07515240460634232, "learning_rate": 3.5212666980616036e-05, "loss": 0.0044, "step": 61580 }, { "grad_norm": 0.10003922134637833, "learning_rate": 3.519687274770681e-05, "loss": 0.0053, "step": 61590 }, { "grad_norm": 0.1268211305141449, "learning_rate": 3.518108013364535e-05, "loss": 0.0056, "step": 61600 }, { "grad_norm": 0.18744949996471405, "learning_rate": 3.516528914015868e-05, "loss": 0.0048, "step": 61610 }, { "grad_norm": 0.10444993525743484, "learning_rate": 3.5149499768973694e-05, "loss": 0.005, "step": 61620 }, { "grad_norm": 0.14151398837566376, "learning_rate": 3.513371202181708e-05, "loss": 0.0046, "step": 61630 }, { "grad_norm": 0.09608228504657745, "learning_rate": 3.511792590041537e-05, "loss": 0.0057, "step": 61640 }, { "grad_norm": 0.1508546769618988, "learning_rate": 3.510214140649488e-05, "loss": 0.0062, "step": 61650 }, { "grad_norm": 0.1464724987745285, "learning_rate": 3.508635854178184e-05, "loss": 0.0056, "step": 61660 }, { "grad_norm": 0.11048596352338791, "learning_rate": 3.50705773080022e-05, "loss": 0.005, "step": 61670 }, { "grad_norm": 0.15096336603164673, "learning_rate": 3.505479770688178e-05, "loss": 0.0054, "step": 61680 }, { "grad_norm": 0.10389725118875504, "learning_rate": 3.503901974014622e-05, "loss": 0.0048, "step": 61690 }, { "grad_norm": 0.13495752215385437, "learning_rate": 3.502324340952095e-05, "loss": 0.0064, "step": 61700 }, { "grad_norm": 0.08745406568050385, "learning_rate": 3.500746871673126e-05, "loss": 0.0047, "step": 61710 }, { "grad_norm": 0.11950106173753738, "learning_rate": 3.499169566350227e-05, "loss": 0.0055, "step": 61720 }, { "grad_norm": 0.09461450576782227, "learning_rate": 3.4975924251558876e-05, "loss": 0.01, "step": 61730 }, { "grad_norm": 0.07052811235189438, "learning_rate": 3.496015448262582e-05, "loss": 0.005, "step": 61740 }, { "grad_norm": 0.15974196791648865, "learning_rate": 3.494438635842764e-05, "loss": 0.0051, "step": 61750 }, { "grad_norm": 0.1286388486623764, "learning_rate": 3.492861988068874e-05, "loss": 0.0063, "step": 61760 }, { "grad_norm": 0.11055786907672882, "learning_rate": 3.491285505113332e-05, "loss": 0.0057, "step": 61770 }, { "grad_norm": 0.13736115396022797, "learning_rate": 3.4897091871485376e-05, "loss": 0.0055, "step": 61780 }, { "grad_norm": 0.13599473237991333, "learning_rate": 3.488133034346875e-05, "loss": 0.0057, "step": 61790 }, { "grad_norm": 0.13223427534103394, "learning_rate": 3.486557046880713e-05, "loss": 0.0045, "step": 61800 }, { "grad_norm": 0.13397985696792603, "learning_rate": 3.484981224922395e-05, "loss": 0.0054, "step": 61810 }, { "grad_norm": 0.166352316737175, "learning_rate": 3.4834055686442544e-05, "loss": 0.0069, "step": 61820 }, { "grad_norm": 0.1674892157316208, "learning_rate": 3.4818300782186e-05, "loss": 0.006, "step": 61830 }, { "grad_norm": 0.14104270935058594, "learning_rate": 3.4802547538177236e-05, "loss": 0.006, "step": 61840 }, { "grad_norm": 0.10690160095691681, "learning_rate": 3.478679595613905e-05, "loss": 0.0057, "step": 61850 }, { "grad_norm": 0.09359824657440186, "learning_rate": 3.4771046037793975e-05, "loss": 0.0052, "step": 61860 }, { "grad_norm": 0.11990922689437866, "learning_rate": 3.4755297784864416e-05, "loss": 0.0074, "step": 61870 }, { "grad_norm": 0.14474499225616455, "learning_rate": 3.473955119907257e-05, "loss": 0.0053, "step": 61880 }, { "grad_norm": 0.09992917627096176, "learning_rate": 3.472380628214046e-05, "loss": 0.0047, "step": 61890 }, { "grad_norm": 0.11262785643339157, "learning_rate": 3.4708063035789906e-05, "loss": 0.008, "step": 61900 }, { "grad_norm": 0.1538023054599762, "learning_rate": 3.4692321461742614e-05, "loss": 0.0065, "step": 61910 }, { "grad_norm": 0.07006890326738358, "learning_rate": 3.4676581561720023e-05, "loss": 0.0059, "step": 61920 }, { "grad_norm": 0.12206080555915833, "learning_rate": 3.466084333744344e-05, "loss": 0.0052, "step": 61930 }, { "grad_norm": 0.10841158777475357, "learning_rate": 3.4645106790633964e-05, "loss": 0.0068, "step": 61940 }, { "grad_norm": 0.10896730422973633, "learning_rate": 3.4629371923012526e-05, "loss": 0.007, "step": 61950 }, { "grad_norm": 0.14699667692184448, "learning_rate": 3.4613638736299856e-05, "loss": 0.0086, "step": 61960 }, { "grad_norm": 0.12074822187423706, "learning_rate": 3.459790723221652e-05, "loss": 0.0071, "step": 61970 }, { "grad_norm": 0.15199816226959229, "learning_rate": 3.45821774124829e-05, "loss": 0.0072, "step": 61980 }, { "grad_norm": 0.09079240262508392, "learning_rate": 3.456644927881918e-05, "loss": 0.0047, "step": 61990 }, { "grad_norm": 0.1554608941078186, "learning_rate": 3.4550722832945354e-05, "loss": 0.005, "step": 62000 }, { "grad_norm": 0.11564938724040985, "learning_rate": 3.453499807658126e-05, "loss": 0.0048, "step": 62010 }, { "grad_norm": 0.06821953505277634, "learning_rate": 3.451927501144653e-05, "loss": 0.0052, "step": 62020 }, { "grad_norm": 0.08371082693338394, "learning_rate": 3.450355363926059e-05, "loss": 0.0046, "step": 62030 }, { "grad_norm": 0.13496491312980652, "learning_rate": 3.448783396174273e-05, "loss": 0.0057, "step": 62040 }, { "grad_norm": 0.1379287987947464, "learning_rate": 3.4472115980612034e-05, "loss": 0.009, "step": 62050 }, { "grad_norm": 0.15293601155281067, "learning_rate": 3.4456399697587396e-05, "loss": 0.0053, "step": 62060 }, { "grad_norm": 0.14069300889968872, "learning_rate": 3.4440685114387506e-05, "loss": 0.0067, "step": 62070 }, { "grad_norm": 0.12874962389469147, "learning_rate": 3.442497223273091e-05, "loss": 0.0069, "step": 62080 }, { "grad_norm": 0.09996237605810165, "learning_rate": 3.4409261054335926e-05, "loss": 0.0053, "step": 62090 }, { "grad_norm": 0.11105115711688995, "learning_rate": 3.439355158092069e-05, "loss": 0.0063, "step": 62100 }, { "grad_norm": 0.11936410516500473, "learning_rate": 3.437784381420322e-05, "loss": 0.0048, "step": 62110 }, { "grad_norm": 0.13623163104057312, "learning_rate": 3.436213775590126e-05, "loss": 0.0061, "step": 62120 }, { "grad_norm": 0.11499369144439697, "learning_rate": 3.4346433407732405e-05, "loss": 0.0066, "step": 62130 }, { "grad_norm": 0.1947389543056488, "learning_rate": 3.433073077141405e-05, "loss": 0.0054, "step": 62140 }, { "grad_norm": 0.1527576744556427, "learning_rate": 3.431502984866343e-05, "loss": 0.0065, "step": 62150 }, { "grad_norm": 0.12380548566579819, "learning_rate": 3.429933064119753e-05, "loss": 0.006, "step": 62160 }, { "grad_norm": 0.12661920487880707, "learning_rate": 3.4283633150733255e-05, "loss": 0.0051, "step": 62170 }, { "grad_norm": 0.15984083712100983, "learning_rate": 3.4267937378987225e-05, "loss": 0.0073, "step": 62180 }, { "grad_norm": 0.11523260176181793, "learning_rate": 3.425224332767592e-05, "loss": 0.0046, "step": 62190 }, { "grad_norm": 0.16169045865535736, "learning_rate": 3.4236550998515595e-05, "loss": 0.0066, "step": 62200 }, { "grad_norm": 0.11961939185857773, "learning_rate": 3.422086039322235e-05, "loss": 0.0072, "step": 62210 }, { "grad_norm": 0.15694688260555267, "learning_rate": 3.420517151351209e-05, "loss": 0.0052, "step": 62220 }, { "grad_norm": 0.1505674421787262, "learning_rate": 3.418948436110053e-05, "loss": 0.0063, "step": 62230 }, { "grad_norm": 0.1338135004043579, "learning_rate": 3.417379893770319e-05, "loss": 0.0062, "step": 62240 }, { "grad_norm": 0.08527067303657532, "learning_rate": 3.41581152450354e-05, "loss": 0.0047, "step": 62250 }, { "grad_norm": 0.1215362697839737, "learning_rate": 3.414243328481231e-05, "loss": 0.0048, "step": 62260 }, { "grad_norm": 0.10567105561494827, "learning_rate": 3.412675305874888e-05, "loss": 0.0051, "step": 62270 }, { "grad_norm": 0.12054005265235901, "learning_rate": 3.4111074568559876e-05, "loss": 0.0048, "step": 62280 }, { "grad_norm": 0.1261935830116272, "learning_rate": 3.4095397815959846e-05, "loss": 0.0072, "step": 62290 }, { "grad_norm": 0.23138539493083954, "learning_rate": 3.407972280266322e-05, "loss": 0.0081, "step": 62300 }, { "grad_norm": 0.16347327828407288, "learning_rate": 3.4064049530384166e-05, "loss": 0.0056, "step": 62310 }, { "grad_norm": 0.10300679504871368, "learning_rate": 3.40483780008367e-05, "loss": 0.0042, "step": 62320 }, { "grad_norm": 0.11811584234237671, "learning_rate": 3.4032708215734635e-05, "loss": 0.0082, "step": 62330 }, { "grad_norm": 0.07533368468284607, "learning_rate": 3.4017040176791584e-05, "loss": 0.0055, "step": 62340 }, { "grad_norm": 0.0947766900062561, "learning_rate": 3.4001373885720976e-05, "loss": 0.0058, "step": 62350 }, { "grad_norm": 0.09569492936134338, "learning_rate": 3.398570934423608e-05, "loss": 0.0068, "step": 62360 }, { "grad_norm": 0.12371012568473816, "learning_rate": 3.397004655404994e-05, "loss": 0.008, "step": 62370 }, { "grad_norm": 0.13117308914661407, "learning_rate": 3.395438551687541e-05, "loss": 0.0078, "step": 62380 }, { "grad_norm": 0.142241433262825, "learning_rate": 3.393872623442515e-05, "loss": 0.0059, "step": 62390 }, { "grad_norm": 0.10850682109594345, "learning_rate": 3.392306870841164e-05, "loss": 0.005, "step": 62400 }, { "grad_norm": 0.11341812461614609, "learning_rate": 3.390741294054715e-05, "loss": 0.0081, "step": 62410 }, { "grad_norm": 0.11283061653375626, "learning_rate": 3.38917589325438e-05, "loss": 0.0066, "step": 62420 }, { "grad_norm": 0.15585358440876007, "learning_rate": 3.387610668611349e-05, "loss": 0.0072, "step": 62430 }, { "grad_norm": 0.11767840385437012, "learning_rate": 3.38604562029679e-05, "loss": 0.0045, "step": 62440 }, { "grad_norm": 0.11668401956558228, "learning_rate": 3.384480748481855e-05, "loss": 0.0061, "step": 62450 }, { "grad_norm": 0.0876309871673584, "learning_rate": 3.382916053337676e-05, "loss": 0.0066, "step": 62460 }, { "grad_norm": 0.07326582819223404, "learning_rate": 3.381351535035366e-05, "loss": 0.0049, "step": 62470 }, { "grad_norm": 0.11861557513475418, "learning_rate": 3.379787193746019e-05, "loss": 0.0056, "step": 62480 }, { "grad_norm": 0.085109643638134, "learning_rate": 3.378223029640709e-05, "loss": 0.0052, "step": 62490 }, { "grad_norm": 0.13431468605995178, "learning_rate": 3.376659042890489e-05, "loss": 0.0058, "step": 62500 }, { "grad_norm": 0.12439657002687454, "learning_rate": 3.375095233666397e-05, "loss": 0.0075, "step": 62510 }, { "grad_norm": 0.08188112825155258, "learning_rate": 3.373531602139447e-05, "loss": 0.0046, "step": 62520 }, { "grad_norm": 0.1308506280183792, "learning_rate": 3.371968148480635e-05, "loss": 0.0091, "step": 62530 }, { "grad_norm": 0.16855557262897491, "learning_rate": 3.370404872860937e-05, "loss": 0.0065, "step": 62540 }, { "grad_norm": 0.1471112072467804, "learning_rate": 3.3688417754513145e-05, "loss": 0.0053, "step": 62550 }, { "grad_norm": 0.155540332198143, "learning_rate": 3.367278856422703e-05, "loss": 0.0084, "step": 62560 }, { "grad_norm": 0.09917524456977844, "learning_rate": 3.365716115946021e-05, "loss": 0.0062, "step": 62570 }, { "grad_norm": 0.09294203668832779, "learning_rate": 3.3641535541921666e-05, "loss": 0.0076, "step": 62580 }, { "grad_norm": 0.06444181501865387, "learning_rate": 3.36259117133202e-05, "loss": 0.0043, "step": 62590 }, { "grad_norm": 0.11453602463006973, "learning_rate": 3.36102896753644e-05, "loss": 0.0061, "step": 62600 }, { "grad_norm": 0.13998809456825256, "learning_rate": 3.35946694297627e-05, "loss": 0.0047, "step": 62610 }, { "grad_norm": 0.09839006513357162, "learning_rate": 3.357905097822328e-05, "loss": 0.0051, "step": 62620 }, { "grad_norm": 0.13683271408081055, "learning_rate": 3.356343432245416e-05, "loss": 0.0055, "step": 62630 }, { "grad_norm": 0.07829683274030685, "learning_rate": 3.354781946416315e-05, "loss": 0.0048, "step": 62640 }, { "grad_norm": 0.2127513885498047, "learning_rate": 3.3532206405057864e-05, "loss": 0.0064, "step": 62650 }, { "grad_norm": 0.12281148880720139, "learning_rate": 3.351659514684573e-05, "loss": 0.0072, "step": 62660 }, { "grad_norm": 0.14518262445926666, "learning_rate": 3.350098569123395e-05, "loss": 0.0057, "step": 62670 }, { "grad_norm": 0.1281837821006775, "learning_rate": 3.348537803992958e-05, "loss": 0.005, "step": 62680 }, { "grad_norm": 0.09233179688453674, "learning_rate": 3.346977219463945e-05, "loss": 0.0061, "step": 62690 }, { "grad_norm": 0.09929517656564713, "learning_rate": 3.345416815707017e-05, "loss": 0.005, "step": 62700 }, { "grad_norm": 0.10663005709648132, "learning_rate": 3.3438565928928185e-05, "loss": 0.0053, "step": 62710 }, { "grad_norm": 0.12084472179412842, "learning_rate": 3.3422965511919736e-05, "loss": 0.0047, "step": 62720 }, { "grad_norm": 0.11922460794448853, "learning_rate": 3.3407366907750846e-05, "loss": 0.0044, "step": 62730 }, { "grad_norm": 0.11467578262090683, "learning_rate": 3.339177011812737e-05, "loss": 0.0053, "step": 62740 }, { "grad_norm": 0.09892000257968903, "learning_rate": 3.3376175144754935e-05, "loss": 0.006, "step": 62750 }, { "grad_norm": 0.11513732373714447, "learning_rate": 3.3360581989339004e-05, "loss": 0.0054, "step": 62760 }, { "grad_norm": 0.13273368775844574, "learning_rate": 3.3344990653584804e-05, "loss": 0.0069, "step": 62770 }, { "grad_norm": 0.10896346718072891, "learning_rate": 3.332940113919739e-05, "loss": 0.0062, "step": 62780 }, { "grad_norm": 0.1356351673603058, "learning_rate": 3.331381344788161e-05, "loss": 0.0047, "step": 62790 }, { "grad_norm": 0.11354304105043411, "learning_rate": 3.3298227581342076e-05, "loss": 0.0047, "step": 62800 }, { "grad_norm": 0.07966813445091248, "learning_rate": 3.328264354128329e-05, "loss": 0.0054, "step": 62810 }, { "grad_norm": 0.17046110332012177, "learning_rate": 3.3267061329409465e-05, "loss": 0.0061, "step": 62820 }, { "grad_norm": 0.16025224328041077, "learning_rate": 3.325148094742466e-05, "loss": 0.0054, "step": 62830 }, { "grad_norm": 0.14416253566741943, "learning_rate": 3.323590239703271e-05, "loss": 0.0056, "step": 62840 }, { "grad_norm": 0.2551876902580261, "learning_rate": 3.322032567993728e-05, "loss": 0.0073, "step": 62850 }, { "grad_norm": 0.12147192656993866, "learning_rate": 3.3204750797841774e-05, "loss": 0.0043, "step": 62860 }, { "grad_norm": 0.12936435639858246, "learning_rate": 3.31891777524495e-05, "loss": 0.0061, "step": 62870 }, { "grad_norm": 0.10409268736839294, "learning_rate": 3.317360654546347e-05, "loss": 0.0054, "step": 62880 }, { "grad_norm": 0.12434700131416321, "learning_rate": 3.3158037178586534e-05, "loss": 0.0083, "step": 62890 }, { "grad_norm": 0.10428010672330856, "learning_rate": 3.3142469653521316e-05, "loss": 0.0057, "step": 62900 }, { "grad_norm": 0.10349854081869125, "learning_rate": 3.3126903971970284e-05, "loss": 0.0059, "step": 62910 }, { "grad_norm": 0.09415256232023239, "learning_rate": 3.311134013563564e-05, "loss": 0.0061, "step": 62920 }, { "grad_norm": 0.1582220494747162, "learning_rate": 3.309577814621947e-05, "loss": 0.0055, "step": 62930 }, { "grad_norm": 0.16070283949375153, "learning_rate": 3.308021800542359e-05, "loss": 0.0065, "step": 62940 }, { "grad_norm": 0.08676904439926147, "learning_rate": 3.306465971494963e-05, "loss": 0.0055, "step": 62950 }, { "grad_norm": 0.12491322308778763, "learning_rate": 3.304910327649901e-05, "loss": 0.0047, "step": 62960 }, { "grad_norm": 0.11041376739740372, "learning_rate": 3.303354869177297e-05, "loss": 0.0051, "step": 62970 }, { "grad_norm": 0.18454530835151672, "learning_rate": 3.301799596247255e-05, "loss": 0.0071, "step": 62980 }, { "grad_norm": 0.12683628499507904, "learning_rate": 3.300244509029854e-05, "loss": 0.0045, "step": 62990 }, { "grad_norm": 0.14105048775672913, "learning_rate": 3.2986896076951585e-05, "loss": 0.0065, "step": 63000 }, { "grad_norm": 0.12153096497058868, "learning_rate": 3.297134892413211e-05, "loss": 0.0041, "step": 63010 }, { "grad_norm": 0.12982077896595, "learning_rate": 3.295580363354031e-05, "loss": 0.0072, "step": 63020 }, { "grad_norm": 0.19077032804489136, "learning_rate": 3.2940260206876194e-05, "loss": 0.0071, "step": 63030 }, { "grad_norm": 0.12191440910100937, "learning_rate": 3.292471864583958e-05, "loss": 0.005, "step": 63040 }, { "grad_norm": 0.13556385040283203, "learning_rate": 3.290917895213004e-05, "loss": 0.0053, "step": 63050 }, { "grad_norm": 0.10332562774419785, "learning_rate": 3.289364112744701e-05, "loss": 0.0053, "step": 63060 }, { "grad_norm": 0.05959196761250496, "learning_rate": 3.287810517348967e-05, "loss": 0.004, "step": 63070 }, { "grad_norm": 0.12212206423282623, "learning_rate": 3.2862571091957014e-05, "loss": 0.0046, "step": 63080 }, { "grad_norm": 0.09818733483552933, "learning_rate": 3.284703888454781e-05, "loss": 0.0048, "step": 63090 }, { "grad_norm": 0.1242058277130127, "learning_rate": 3.2831508552960646e-05, "loss": 0.007, "step": 63100 }, { "grad_norm": 0.13166005909442902, "learning_rate": 3.281598009889387e-05, "loss": 0.0056, "step": 63110 }, { "grad_norm": 0.11261007934808731, "learning_rate": 3.28004535240457e-05, "loss": 0.0054, "step": 63120 }, { "grad_norm": 0.1286424696445465, "learning_rate": 3.278492883011407e-05, "loss": 0.0048, "step": 63130 }, { "grad_norm": 0.11114831268787384, "learning_rate": 3.276940601879675e-05, "loss": 0.0051, "step": 63140 }, { "grad_norm": 0.0931008830666542, "learning_rate": 3.275388509179127e-05, "loss": 0.0056, "step": 63150 }, { "grad_norm": 0.12783896923065186, "learning_rate": 3.2738366050794985e-05, "loss": 0.0054, "step": 63160 }, { "grad_norm": 0.16236361861228943, "learning_rate": 3.2722848897505034e-05, "loss": 0.0063, "step": 63170 }, { "grad_norm": 0.16125114262104034, "learning_rate": 3.2707333633618356e-05, "loss": 0.0066, "step": 63180 }, { "grad_norm": 0.1499990075826645, "learning_rate": 3.269182026083167e-05, "loss": 0.0068, "step": 63190 }, { "grad_norm": 0.14535106718540192, "learning_rate": 3.2676308780841516e-05, "loss": 0.0056, "step": 63200 }, { "grad_norm": 0.11504475027322769, "learning_rate": 3.266079919534416e-05, "loss": 0.0053, "step": 63210 }, { "grad_norm": 0.09468992054462433, "learning_rate": 3.264529150603575e-05, "loss": 0.0051, "step": 63220 }, { "grad_norm": 0.1155816987156868, "learning_rate": 3.262978571461217e-05, "loss": 0.0044, "step": 63230 }, { "grad_norm": 0.10258252918720245, "learning_rate": 3.26142818227691e-05, "loss": 0.0084, "step": 63240 }, { "grad_norm": 0.13319581747055054, "learning_rate": 3.259877983220203e-05, "loss": 0.0073, "step": 63250 }, { "grad_norm": 0.12374390661716461, "learning_rate": 3.2583279744606246e-05, "loss": 0.0051, "step": 63260 }, { "grad_norm": 0.13024526834487915, "learning_rate": 3.25677815616768e-05, "loss": 0.0049, "step": 63270 }, { "grad_norm": 0.14734245836734772, "learning_rate": 3.255228528510856e-05, "loss": 0.0044, "step": 63280 }, { "grad_norm": 0.16809888184070587, "learning_rate": 3.253679091659617e-05, "loss": 0.007, "step": 63290 }, { "grad_norm": 0.2535930871963501, "learning_rate": 3.252129845783404e-05, "loss": 0.0052, "step": 63300 }, { "grad_norm": 0.10382410138845444, "learning_rate": 3.2505807910516456e-05, "loss": 0.0054, "step": 63310 }, { "grad_norm": 0.13693654537200928, "learning_rate": 3.249031927633742e-05, "loss": 0.0085, "step": 63320 }, { "grad_norm": 0.153512105345726, "learning_rate": 3.2474832556990737e-05, "loss": 0.0059, "step": 63330 }, { "grad_norm": 0.09816155582666397, "learning_rate": 3.245934775417002e-05, "loss": 0.0059, "step": 63340 }, { "grad_norm": 0.1782534420490265, "learning_rate": 3.244386486956866e-05, "loss": 0.0081, "step": 63350 }, { "grad_norm": 0.1257011443376541, "learning_rate": 3.2428383904879846e-05, "loss": 0.0048, "step": 63360 }, { "grad_norm": 0.14113005995750427, "learning_rate": 3.241290486179652e-05, "loss": 0.0057, "step": 63370 }, { "grad_norm": 0.08219698071479797, "learning_rate": 3.239742774201151e-05, "loss": 0.0061, "step": 63380 }, { "grad_norm": 0.0946110263466835, "learning_rate": 3.238195254721733e-05, "loss": 0.0057, "step": 63390 }, { "grad_norm": 0.14239902794361115, "learning_rate": 3.236647927910633e-05, "loss": 0.007, "step": 63400 }, { "grad_norm": 0.10564707964658737, "learning_rate": 3.2351007939370646e-05, "loss": 0.0056, "step": 63410 }, { "grad_norm": 0.19857439398765564, "learning_rate": 3.2335538529702184e-05, "loss": 0.0052, "step": 63420 }, { "grad_norm": 0.11283423751592636, "learning_rate": 3.232007105179268e-05, "loss": 0.0042, "step": 63430 }, { "grad_norm": 0.14481519162654877, "learning_rate": 3.230460550733363e-05, "loss": 0.0087, "step": 63440 }, { "grad_norm": 0.2535015642642975, "learning_rate": 3.2289141898016315e-05, "loss": 0.0073, "step": 63450 }, { "grad_norm": 0.11982689797878265, "learning_rate": 3.227368022553181e-05, "loss": 0.0072, "step": 63460 }, { "grad_norm": 0.09105724096298218, "learning_rate": 3.225822049157099e-05, "loss": 0.0057, "step": 63470 }, { "grad_norm": 0.12520937621593475, "learning_rate": 3.2242762697824494e-05, "loss": 0.0066, "step": 63480 }, { "grad_norm": 0.08794232457876205, "learning_rate": 3.222730684598279e-05, "loss": 0.0063, "step": 63490 }, { "grad_norm": 0.11816427111625671, "learning_rate": 3.221185293773605e-05, "loss": 0.0074, "step": 63500 }, { "grad_norm": 0.1871040016412735, "learning_rate": 3.2196400974774354e-05, "loss": 0.0064, "step": 63510 }, { "grad_norm": 0.1682208776473999, "learning_rate": 3.2180950958787485e-05, "loss": 0.0052, "step": 63520 }, { "grad_norm": 0.14653615653514862, "learning_rate": 3.2165502891465024e-05, "loss": 0.0059, "step": 63530 }, { "grad_norm": 0.10321485251188278, "learning_rate": 3.215005677449636e-05, "loss": 0.0052, "step": 63540 }, { "grad_norm": 0.10834953933954239, "learning_rate": 3.2134612609570644e-05, "loss": 0.0051, "step": 63550 }, { "grad_norm": 0.09424460679292679, "learning_rate": 3.211917039837682e-05, "loss": 0.0063, "step": 63560 }, { "grad_norm": 0.08005436509847641, "learning_rate": 3.210373014260365e-05, "loss": 0.0044, "step": 63570 }, { "grad_norm": 0.11791496723890305, "learning_rate": 3.208829184393965e-05, "loss": 0.0074, "step": 63580 }, { "grad_norm": 0.14307552576065063, "learning_rate": 3.207285550407312e-05, "loss": 0.006, "step": 63590 }, { "grad_norm": 0.1127048209309578, "learning_rate": 3.205742112469217e-05, "loss": 0.005, "step": 63600 }, { "grad_norm": 0.15324105322360992, "learning_rate": 3.204198870748465e-05, "loss": 0.0045, "step": 63610 }, { "grad_norm": 0.2033691555261612, "learning_rate": 3.202655825413824e-05, "loss": 0.0067, "step": 63620 }, { "grad_norm": 0.13142643868923187, "learning_rate": 3.2011129766340416e-05, "loss": 0.0049, "step": 63630 }, { "grad_norm": 0.09651050716638565, "learning_rate": 3.1995703245778384e-05, "loss": 0.0056, "step": 63640 }, { "grad_norm": 0.16812065243721008, "learning_rate": 3.1980278694139186e-05, "loss": 0.0057, "step": 63650 }, { "grad_norm": 0.1517782360315323, "learning_rate": 3.196485611310962e-05, "loss": 0.0052, "step": 63660 }, { "grad_norm": 0.11384948343038559, "learning_rate": 3.1949435504376257e-05, "loss": 0.0049, "step": 63670 }, { "grad_norm": 0.125797301530838, "learning_rate": 3.1934016869625505e-05, "loss": 0.0051, "step": 63680 }, { "grad_norm": 0.0935787782073021, "learning_rate": 3.191860021054349e-05, "loss": 0.0061, "step": 63690 }, { "grad_norm": 0.144463911652565, "learning_rate": 3.190318552881618e-05, "loss": 0.0053, "step": 63700 }, { "grad_norm": 0.12584979832172394, "learning_rate": 3.1887772826129285e-05, "loss": 0.0056, "step": 63710 }, { "grad_norm": 0.12607744336128235, "learning_rate": 3.1872362104168335e-05, "loss": 0.0053, "step": 63720 }, { "grad_norm": 0.1269170343875885, "learning_rate": 3.185695336461861e-05, "loss": 0.0053, "step": 63730 }, { "grad_norm": 0.14092940092086792, "learning_rate": 3.184154660916519e-05, "loss": 0.0059, "step": 63740 }, { "grad_norm": 0.1078699380159378, "learning_rate": 3.1826141839492906e-05, "loss": 0.0063, "step": 63750 }, { "grad_norm": 0.1858637034893036, "learning_rate": 3.181073905728644e-05, "loss": 0.005, "step": 63760 }, { "grad_norm": 0.12058551609516144, "learning_rate": 3.1795338264230214e-05, "loss": 0.005, "step": 63770 }, { "grad_norm": 0.1331520676612854, "learning_rate": 3.1779939462008426e-05, "loss": 0.0061, "step": 63780 }, { "grad_norm": 0.08387591689825058, "learning_rate": 3.1764542652305066e-05, "loss": 0.0045, "step": 63790 }, { "grad_norm": 0.12032485753297806, "learning_rate": 3.17491478368039e-05, "loss": 0.0046, "step": 63800 }, { "grad_norm": 0.13179680705070496, "learning_rate": 3.173375501718847e-05, "loss": 0.0054, "step": 63810 }, { "grad_norm": 0.12217686325311661, "learning_rate": 3.171836419514215e-05, "loss": 0.0065, "step": 63820 }, { "grad_norm": 0.12381867319345474, "learning_rate": 3.170297537234803e-05, "loss": 0.0059, "step": 63830 }, { "grad_norm": 0.11063820123672485, "learning_rate": 3.168758855048901e-05, "loss": 0.005, "step": 63840 }, { "grad_norm": 0.1232314258813858, "learning_rate": 3.1672203731247785e-05, "loss": 0.0069, "step": 63850 }, { "grad_norm": 0.11354940384626389, "learning_rate": 3.16568209163068e-05, "loss": 0.0071, "step": 63860 }, { "grad_norm": 0.13725712895393372, "learning_rate": 3.1641440107348295e-05, "loss": 0.0059, "step": 63870 }, { "grad_norm": 0.1211550310254097, "learning_rate": 3.162606130605429e-05, "loss": 0.0058, "step": 63880 }, { "grad_norm": 0.076677605509758, "learning_rate": 3.16106845141066e-05, "loss": 0.0056, "step": 63890 }, { "grad_norm": 0.11829233914613724, "learning_rate": 3.159530973318682e-05, "loss": 0.0055, "step": 63900 }, { "grad_norm": 0.1562664806842804, "learning_rate": 3.1579936964976286e-05, "loss": 0.007, "step": 63910 }, { "grad_norm": 0.14071007072925568, "learning_rate": 3.156456621115615e-05, "loss": 0.0056, "step": 63920 }, { "grad_norm": 0.1359720081090927, "learning_rate": 3.154919747340734e-05, "loss": 0.0062, "step": 63930 }, { "grad_norm": 0.12355729937553406, "learning_rate": 3.153383075341054e-05, "loss": 0.006, "step": 63940 }, { "grad_norm": 0.15391771495342255, "learning_rate": 3.1518466052846264e-05, "loss": 0.0059, "step": 63950 }, { "grad_norm": 0.13792456686496735, "learning_rate": 3.150310337339474e-05, "loss": 0.0057, "step": 63960 }, { "grad_norm": 0.12540414929389954, "learning_rate": 3.148774271673603e-05, "loss": 0.0065, "step": 63970 }, { "grad_norm": 0.10414727032184601, "learning_rate": 3.147238408454995e-05, "loss": 0.0055, "step": 63980 }, { "grad_norm": 0.06287407875061035, "learning_rate": 3.1457027478516086e-05, "loss": 0.0041, "step": 63990 }, { "grad_norm": 0.1437513679265976, "learning_rate": 3.144167290031382e-05, "loss": 0.0061, "step": 64000 }, { "grad_norm": 0.12890298664569855, "learning_rate": 3.142632035162227e-05, "loss": 0.0054, "step": 64010 }, { "grad_norm": 0.16787868738174438, "learning_rate": 3.141096983412044e-05, "loss": 0.0057, "step": 64020 }, { "grad_norm": 0.0948387160897255, "learning_rate": 3.139562134948699e-05, "loss": 0.0049, "step": 64030 }, { "grad_norm": 0.10481506586074829, "learning_rate": 3.1380274899400424e-05, "loss": 0.0045, "step": 64040 }, { "grad_norm": 0.10218816995620728, "learning_rate": 3.1364930485539e-05, "loss": 0.0065, "step": 64050 }, { "grad_norm": 0.13175129890441895, "learning_rate": 3.134958810958076e-05, "loss": 0.0039, "step": 64060 }, { "grad_norm": 0.13530193269252777, "learning_rate": 3.13342477732035e-05, "loss": 0.0045, "step": 64070 }, { "grad_norm": 0.08950992673635483, "learning_rate": 3.131890947808487e-05, "loss": 0.007, "step": 64080 }, { "grad_norm": 0.07594949007034302, "learning_rate": 3.13035732259022e-05, "loss": 0.005, "step": 64090 }, { "grad_norm": 0.12585759162902832, "learning_rate": 3.1288239018332656e-05, "loss": 0.0043, "step": 64100 }, { "grad_norm": 0.12776066362857819, "learning_rate": 3.1272906857053164e-05, "loss": 0.0058, "step": 64110 }, { "grad_norm": 0.09285092353820801, "learning_rate": 3.125757674374042e-05, "loss": 0.0047, "step": 64120 }, { "grad_norm": 0.13493217527866364, "learning_rate": 3.124224868007088e-05, "loss": 0.0066, "step": 64130 }, { "grad_norm": 0.08667929470539093, "learning_rate": 3.1226922667720854e-05, "loss": 0.0052, "step": 64140 }, { "grad_norm": 0.10166805237531662, "learning_rate": 3.121159870836633e-05, "loss": 0.0064, "step": 64150 }, { "grad_norm": 0.0900803953409195, "learning_rate": 3.119627680368312e-05, "loss": 0.0063, "step": 64160 }, { "grad_norm": 0.07689744234085083, "learning_rate": 3.1180956955346816e-05, "loss": 0.0032, "step": 64170 }, { "grad_norm": 0.1347583830356598, "learning_rate": 3.116563916503276e-05, "loss": 0.0057, "step": 64180 }, { "grad_norm": 0.09389356523752213, "learning_rate": 3.1150323434416095e-05, "loss": 0.0061, "step": 64190 }, { "grad_norm": 0.12304509431123734, "learning_rate": 3.1135009765171696e-05, "loss": 0.0078, "step": 64200 }, { "grad_norm": 0.09661424905061722, "learning_rate": 3.111969815897429e-05, "loss": 0.0052, "step": 64210 }, { "grad_norm": 0.15485292673110962, "learning_rate": 3.1104388617498295e-05, "loss": 0.0048, "step": 64220 }, { "grad_norm": 0.17468489706516266, "learning_rate": 3.1089081142417966e-05, "loss": 0.0069, "step": 64230 }, { "grad_norm": 0.13990308344364166, "learning_rate": 3.1073775735407285e-05, "loss": 0.0062, "step": 64240 }, { "grad_norm": 0.2261473387479782, "learning_rate": 3.1058472398140035e-05, "loss": 0.0055, "step": 64250 }, { "grad_norm": 0.13787788152694702, "learning_rate": 3.1043171132289736e-05, "loss": 0.0052, "step": 64260 }, { "grad_norm": 0.12574470043182373, "learning_rate": 3.1027871939529774e-05, "loss": 0.0069, "step": 64270 }, { "grad_norm": 0.10100802779197693, "learning_rate": 3.1012574821533214e-05, "loss": 0.0048, "step": 64280 }, { "grad_norm": 0.13123568892478943, "learning_rate": 3.099727977997291e-05, "loss": 0.005, "step": 64290 }, { "grad_norm": 0.12376101315021515, "learning_rate": 3.098198681652154e-05, "loss": 0.0066, "step": 64300 }, { "grad_norm": 0.0802508294582367, "learning_rate": 3.0966695932851475e-05, "loss": 0.0053, "step": 64310 }, { "grad_norm": 0.13354133069515228, "learning_rate": 3.0951407130634926e-05, "loss": 0.0054, "step": 64320 }, { "grad_norm": 0.15302298963069916, "learning_rate": 3.0936120411543864e-05, "loss": 0.005, "step": 64330 }, { "grad_norm": 0.1500989943742752, "learning_rate": 3.0920835777250015e-05, "loss": 0.0065, "step": 64340 }, { "grad_norm": 0.2596133053302765, "learning_rate": 3.090555322942488e-05, "loss": 0.0053, "step": 64350 }, { "grad_norm": 0.11023771017789841, "learning_rate": 3.089027276973973e-05, "loss": 0.0056, "step": 64360 }, { "grad_norm": 0.12116356939077377, "learning_rate": 3.0874994399865615e-05, "loss": 0.0048, "step": 64370 }, { "grad_norm": 0.14461106061935425, "learning_rate": 3.085971812147337e-05, "loss": 0.0058, "step": 64380 }, { "grad_norm": 0.1748594045639038, "learning_rate": 3.0844443936233557e-05, "loss": 0.0092, "step": 64390 }, { "grad_norm": 0.1264413446187973, "learning_rate": 3.082917184581657e-05, "loss": 0.006, "step": 64400 }, { "grad_norm": 0.09411966800689697, "learning_rate": 3.0813901851892513e-05, "loss": 0.0061, "step": 64410 }, { "grad_norm": 0.0722113847732544, "learning_rate": 3.0798633956131316e-05, "loss": 0.0054, "step": 64420 }, { "grad_norm": 0.091673843562603, "learning_rate": 3.078336816020264e-05, "loss": 0.0066, "step": 64430 }, { "grad_norm": 0.10915575176477432, "learning_rate": 3.0768104465775935e-05, "loss": 0.0048, "step": 64440 }, { "grad_norm": 0.08716952055692673, "learning_rate": 3.0752842874520396e-05, "loss": 0.0039, "step": 64450 }, { "grad_norm": 0.07732153683900833, "learning_rate": 3.073758338810503e-05, "loss": 0.005, "step": 64460 }, { "grad_norm": 0.09833717346191406, "learning_rate": 3.072232600819859e-05, "loss": 0.0053, "step": 64470 }, { "grad_norm": 0.10256505012512207, "learning_rate": 3.070707073646959e-05, "loss": 0.0051, "step": 64480 }, { "grad_norm": 0.0988328605890274, "learning_rate": 3.069181757458633e-05, "loss": 0.0053, "step": 64490 }, { "grad_norm": 0.13507907092571259, "learning_rate": 3.067656652421687e-05, "loss": 0.0051, "step": 64500 }, { "grad_norm": 0.09506387263536453, "learning_rate": 3.0661317587029025e-05, "loss": 0.0054, "step": 64510 }, { "grad_norm": 0.2265092432498932, "learning_rate": 3.064607076469043e-05, "loss": 0.0047, "step": 64520 }, { "grad_norm": 0.1367141753435135, "learning_rate": 3.063082605886845e-05, "loss": 0.0046, "step": 64530 }, { "grad_norm": 0.15370096266269684, "learning_rate": 3.061558347123019e-05, "loss": 0.0054, "step": 64540 }, { "grad_norm": 0.13200539350509644, "learning_rate": 3.060034300344259e-05, "loss": 0.0058, "step": 64550 }, { "grad_norm": 0.15325212478637695, "learning_rate": 3.058510465717231e-05, "loss": 0.0047, "step": 64560 }, { "grad_norm": 0.11885388940572739, "learning_rate": 3.0569868434085804e-05, "loss": 0.0058, "step": 64570 }, { "grad_norm": 0.13823562860488892, "learning_rate": 3.055463433584924e-05, "loss": 0.0065, "step": 64580 }, { "grad_norm": 0.21662820875644684, "learning_rate": 3.0539402364128656e-05, "loss": 0.0051, "step": 64590 }, { "grad_norm": 0.1295471340417862, "learning_rate": 3.052417252058977e-05, "loss": 0.0083, "step": 64600 }, { "grad_norm": 0.0765300914645195, "learning_rate": 3.0508944806898092e-05, "loss": 0.0032, "step": 64610 }, { "grad_norm": 0.11114230006933212, "learning_rate": 3.04937192247189e-05, "loss": 0.006, "step": 64620 }, { "grad_norm": 0.12585005164146423, "learning_rate": 3.047849577571725e-05, "loss": 0.0058, "step": 64630 }, { "grad_norm": 0.11657685786485672, "learning_rate": 3.046327446155795e-05, "loss": 0.0061, "step": 64640 }, { "grad_norm": 0.11574740707874298, "learning_rate": 3.0448055283905587e-05, "loss": 0.0051, "step": 64650 }, { "grad_norm": 0.12373869121074677, "learning_rate": 3.0432838244424494e-05, "loss": 0.0066, "step": 64660 }, { "grad_norm": 0.1736866682767868, "learning_rate": 3.04176233447788e-05, "loss": 0.0056, "step": 64670 }, { "grad_norm": 0.139027401804924, "learning_rate": 3.040241058663238e-05, "loss": 0.0066, "step": 64680 }, { "grad_norm": 0.1424001306295395, "learning_rate": 3.038719997164887e-05, "loss": 0.006, "step": 64690 }, { "grad_norm": 0.15069235861301422, "learning_rate": 3.0371991501491683e-05, "loss": 0.0053, "step": 64700 }, { "grad_norm": 0.11853821575641632, "learning_rate": 3.0356785177823975e-05, "loss": 0.0048, "step": 64710 }, { "grad_norm": 0.1469612419605255, "learning_rate": 3.0341581002308726e-05, "loss": 0.0039, "step": 64720 }, { "grad_norm": 0.09822509437799454, "learning_rate": 3.0326378976608627e-05, "loss": 0.008, "step": 64730 }, { "grad_norm": 0.13491137325763702, "learning_rate": 3.0311179102386146e-05, "loss": 0.0067, "step": 64740 }, { "grad_norm": 0.11126764863729477, "learning_rate": 3.0295981381303513e-05, "loss": 0.0069, "step": 64750 }, { "grad_norm": 0.12409946322441101, "learning_rate": 3.0280785815022727e-05, "loss": 0.0082, "step": 64760 }, { "grad_norm": 0.07686852663755417, "learning_rate": 3.0265592405205544e-05, "loss": 0.0054, "step": 64770 }, { "grad_norm": 0.07992562651634216, "learning_rate": 3.0250401153513517e-05, "loss": 0.0055, "step": 64780 }, { "grad_norm": 0.13412456214427948, "learning_rate": 3.0235212061607936e-05, "loss": 0.0053, "step": 64790 }, { "grad_norm": 0.12129499018192291, "learning_rate": 3.0220025131149842e-05, "loss": 0.0077, "step": 64800 }, { "grad_norm": 0.15272369980812073, "learning_rate": 3.020484036380007e-05, "loss": 0.0053, "step": 64810 }, { "grad_norm": 0.09155953675508499, "learning_rate": 3.0189657761219188e-05, "loss": 0.0062, "step": 64820 }, { "grad_norm": 0.11651226878166199, "learning_rate": 3.017447732506753e-05, "loss": 0.0088, "step": 64830 }, { "grad_norm": 0.13188447058200836, "learning_rate": 3.015929905700524e-05, "loss": 0.0053, "step": 64840 }, { "grad_norm": 0.09874758124351501, "learning_rate": 3.014412295869218e-05, "loss": 0.0066, "step": 64850 }, { "grad_norm": 0.16533523797988892, "learning_rate": 3.0128949031787977e-05, "loss": 0.0076, "step": 64860 }, { "grad_norm": 0.08295554667711258, "learning_rate": 3.0113777277952022e-05, "loss": 0.005, "step": 64870 }, { "grad_norm": 0.09560801833868027, "learning_rate": 3.0098607698843496e-05, "loss": 0.0054, "step": 64880 }, { "grad_norm": 0.1411675214767456, "learning_rate": 3.0083440296121312e-05, "loss": 0.0046, "step": 64890 }, { "grad_norm": 0.08345580101013184, "learning_rate": 3.0068275071444142e-05, "loss": 0.0051, "step": 64900 }, { "grad_norm": 0.09937959164381027, "learning_rate": 3.0053112026470442e-05, "loss": 0.0049, "step": 64910 }, { "grad_norm": 0.44887489080429077, "learning_rate": 3.0037951162858434e-05, "loss": 0.0054, "step": 64920 }, { "grad_norm": 0.18159587681293488, "learning_rate": 3.002279248226607e-05, "loss": 0.0068, "step": 64930 }, { "grad_norm": 0.1464875191450119, "learning_rate": 3.000763598635109e-05, "loss": 0.0047, "step": 64940 }, { "grad_norm": 0.07630900293588638, "learning_rate": 2.999248167677098e-05, "loss": 0.0044, "step": 64950 }, { "grad_norm": 0.11556798219680786, "learning_rate": 2.9977329555182976e-05, "loss": 0.0061, "step": 64960 }, { "grad_norm": 0.11479504406452179, "learning_rate": 2.9962179623244124e-05, "loss": 0.007, "step": 64970 }, { "grad_norm": 0.10125931352376938, "learning_rate": 2.9947031882611175e-05, "loss": 0.004, "step": 64980 }, { "grad_norm": 0.10272903740406036, "learning_rate": 2.9931886334940685e-05, "loss": 0.0056, "step": 64990 }, { "grad_norm": 0.11247284710407257, "learning_rate": 2.9916742981888923e-05, "loss": 0.006, "step": 65000 }, { "grad_norm": 0.0793221965432167, "learning_rate": 2.990160182511196e-05, "loss": 0.0052, "step": 65010 }, { "grad_norm": 0.17131707072257996, "learning_rate": 2.9886462866265574e-05, "loss": 0.0066, "step": 65020 }, { "grad_norm": 0.1362854689359665, "learning_rate": 2.9871326107005392e-05, "loss": 0.0067, "step": 65030 }, { "grad_norm": 0.11960215866565704, "learning_rate": 2.9856191548986735e-05, "loss": 0.0073, "step": 65040 }, { "grad_norm": 0.14768685400485992, "learning_rate": 2.9841059193864672e-05, "loss": 0.0068, "step": 65050 }, { "grad_norm": 0.15684917569160461, "learning_rate": 2.982592904329407e-05, "loss": 0.0076, "step": 65060 }, { "grad_norm": 0.11290057748556137, "learning_rate": 2.981080109892954e-05, "loss": 0.0086, "step": 65070 }, { "grad_norm": 0.10085092484951019, "learning_rate": 2.9795675362425434e-05, "loss": 0.0055, "step": 65080 }, { "grad_norm": 0.11403893679380417, "learning_rate": 2.97805518354359e-05, "loss": 0.0048, "step": 65090 }, { "grad_norm": 0.10636726766824722, "learning_rate": 2.9765430519614825e-05, "loss": 0.0065, "step": 65100 }, { "grad_norm": 0.10890036076307297, "learning_rate": 2.975031141661584e-05, "loss": 0.0058, "step": 65110 }, { "grad_norm": 0.08931279182434082, "learning_rate": 2.9735194528092347e-05, "loss": 0.0042, "step": 65120 }, { "grad_norm": 0.10203313082456589, "learning_rate": 2.9720079855697525e-05, "loss": 0.0055, "step": 65130 }, { "grad_norm": 0.15486954152584076, "learning_rate": 2.970496740108427e-05, "loss": 0.0052, "step": 65140 }, { "grad_norm": 0.08213362097740173, "learning_rate": 2.968985716590527e-05, "loss": 0.0052, "step": 65150 }, { "grad_norm": 0.07195544987916946, "learning_rate": 2.9674749151812953e-05, "loss": 0.0039, "step": 65160 }, { "grad_norm": 0.10510121285915375, "learning_rate": 2.965964336045951e-05, "loss": 0.0047, "step": 65170 }, { "grad_norm": 0.09853394329547882, "learning_rate": 2.9644539793496894e-05, "loss": 0.007, "step": 65180 }, { "grad_norm": 0.12321292608976364, "learning_rate": 2.9629438452576807e-05, "loss": 0.0043, "step": 65190 }, { "grad_norm": 0.12904179096221924, "learning_rate": 2.9614339339350698e-05, "loss": 0.0055, "step": 65200 }, { "grad_norm": 0.07500123977661133, "learning_rate": 2.959924245546979e-05, "loss": 0.0036, "step": 65210 }, { "grad_norm": 0.08640386164188385, "learning_rate": 2.9584147802585038e-05, "loss": 0.0043, "step": 65220 }, { "grad_norm": 0.15692876279354095, "learning_rate": 2.956905538234721e-05, "loss": 0.0067, "step": 65230 }, { "grad_norm": 0.08763938397169113, "learning_rate": 2.9553965196406764e-05, "loss": 0.0051, "step": 65240 }, { "grad_norm": 0.12857845425605774, "learning_rate": 2.9538877246413943e-05, "loss": 0.0045, "step": 65250 }, { "grad_norm": 0.1463548094034195, "learning_rate": 2.952379153401874e-05, "loss": 0.0049, "step": 65260 }, { "grad_norm": 0.14216671884059906, "learning_rate": 2.9508708060870914e-05, "loss": 0.0051, "step": 65270 }, { "grad_norm": 0.11035072803497314, "learning_rate": 2.9493626828619948e-05, "loss": 0.0068, "step": 65280 }, { "grad_norm": 0.12251604348421097, "learning_rate": 2.9478547838915137e-05, "loss": 0.0043, "step": 65290 }, { "grad_norm": 0.1591336876153946, "learning_rate": 2.946347109340548e-05, "loss": 0.0068, "step": 65300 }, { "grad_norm": 0.11905315518379211, "learning_rate": 2.944839659373976e-05, "loss": 0.0055, "step": 65310 }, { "grad_norm": 0.12487204372882843, "learning_rate": 2.943332434156647e-05, "loss": 0.0057, "step": 65320 }, { "grad_norm": 0.07938162237405777, "learning_rate": 2.9418254338533912e-05, "loss": 0.0045, "step": 65330 }, { "grad_norm": 0.09006299078464508, "learning_rate": 2.940318658629011e-05, "loss": 0.0053, "step": 65340 }, { "grad_norm": 0.12495651841163635, "learning_rate": 2.938812108648287e-05, "loss": 0.0071, "step": 65350 }, { "grad_norm": 0.07218986004590988, "learning_rate": 2.937305784075972e-05, "loss": 0.0114, "step": 65360 }, { "grad_norm": 0.09174472093582153, "learning_rate": 2.9357996850767937e-05, "loss": 0.0045, "step": 65370 }, { "grad_norm": 0.09287679195404053, "learning_rate": 2.9342938118154595e-05, "loss": 0.0062, "step": 65380 }, { "grad_norm": 0.09387326240539551, "learning_rate": 2.9327881644566485e-05, "loss": 0.004, "step": 65390 }, { "grad_norm": 0.09393352270126343, "learning_rate": 2.9312827431650157e-05, "loss": 0.0065, "step": 65400 }, { "grad_norm": 0.0961269736289978, "learning_rate": 2.92977754810519e-05, "loss": 0.0062, "step": 65410 }, { "grad_norm": 0.15028804540634155, "learning_rate": 2.928272579441781e-05, "loss": 0.0055, "step": 65420 }, { "grad_norm": 0.11757022887468338, "learning_rate": 2.9267678373393676e-05, "loss": 0.0057, "step": 65430 }, { "grad_norm": 0.11344081908464432, "learning_rate": 2.925263321962507e-05, "loss": 0.0053, "step": 65440 }, { "grad_norm": 0.12130019813776016, "learning_rate": 2.9237590334757292e-05, "loss": 0.0051, "step": 65450 }, { "grad_norm": 0.19715075194835663, "learning_rate": 2.922254972043543e-05, "loss": 0.0056, "step": 65460 }, { "grad_norm": 0.1638394445180893, "learning_rate": 2.9207511378304265e-05, "loss": 0.0072, "step": 65470 }, { "grad_norm": 0.13109090924263, "learning_rate": 2.9192475310008428e-05, "loss": 0.0051, "step": 65480 }, { "grad_norm": 0.11485620588064194, "learning_rate": 2.917744151719219e-05, "loss": 0.0046, "step": 65490 }, { "grad_norm": 0.11256985366344452, "learning_rate": 2.916241000149964e-05, "loss": 0.0052, "step": 65500 }, { "grad_norm": 0.0958135724067688, "learning_rate": 2.9147380764574622e-05, "loss": 0.006, "step": 65510 }, { "grad_norm": 0.11316895484924316, "learning_rate": 2.9132353808060674e-05, "loss": 0.0047, "step": 65520 }, { "grad_norm": 0.08837148547172546, "learning_rate": 2.9117329133601134e-05, "loss": 0.0056, "step": 65530 }, { "grad_norm": 0.16744175553321838, "learning_rate": 2.9102306742839093e-05, "loss": 0.0059, "step": 65540 }, { "grad_norm": 0.19008247554302216, "learning_rate": 2.9087286637417375e-05, "loss": 0.0053, "step": 65550 }, { "grad_norm": 0.12548886239528656, "learning_rate": 2.9072268818978537e-05, "loss": 0.0064, "step": 65560 }, { "grad_norm": 0.1266688108444214, "learning_rate": 2.905725328916493e-05, "loss": 0.0054, "step": 65570 }, { "grad_norm": 0.16507229208946228, "learning_rate": 2.9042240049618608e-05, "loss": 0.0057, "step": 65580 }, { "grad_norm": 0.15902647376060486, "learning_rate": 2.9027229101981413e-05, "loss": 0.0053, "step": 65590 }, { "grad_norm": 0.1054774597287178, "learning_rate": 2.901222044789488e-05, "loss": 0.0061, "step": 65600 }, { "grad_norm": 0.0874776542186737, "learning_rate": 2.89972140890004e-05, "loss": 0.0068, "step": 65610 }, { "grad_norm": 0.11090093851089478, "learning_rate": 2.8982210026938988e-05, "loss": 0.0039, "step": 65620 }, { "grad_norm": 0.12308239191770554, "learning_rate": 2.896720826335151e-05, "loss": 0.0043, "step": 65630 }, { "grad_norm": 0.07184872031211853, "learning_rate": 2.8952208799878493e-05, "loss": 0.0051, "step": 65640 }, { "grad_norm": 0.10123254358768463, "learning_rate": 2.89372116381603e-05, "loss": 0.0053, "step": 65650 }, { "grad_norm": 0.2081994116306305, "learning_rate": 2.8922216779836932e-05, "loss": 0.005, "step": 65660 }, { "grad_norm": 0.08571726083755493, "learning_rate": 2.8907224226548285e-05, "loss": 0.0055, "step": 65670 }, { "grad_norm": 0.14296337962150574, "learning_rate": 2.8892233979933868e-05, "loss": 0.0046, "step": 65680 }, { "grad_norm": 0.13532066345214844, "learning_rate": 2.8877246041633026e-05, "loss": 0.0073, "step": 65690 }, { "grad_norm": 0.13702034950256348, "learning_rate": 2.886226041328478e-05, "loss": 0.0059, "step": 65700 }, { "grad_norm": 0.07255818694829941, "learning_rate": 2.8847277096527957e-05, "loss": 0.0052, "step": 65710 }, { "grad_norm": 0.08018283545970917, "learning_rate": 2.8832296093001105e-05, "loss": 0.0074, "step": 65720 }, { "grad_norm": 0.10437286645174026, "learning_rate": 2.8817317404342547e-05, "loss": 0.0057, "step": 65730 }, { "grad_norm": 0.07633588463068008, "learning_rate": 2.880234103219029e-05, "loss": 0.0052, "step": 65740 }, { "grad_norm": 0.07566521316766739, "learning_rate": 2.878736697818215e-05, "loss": 0.004, "step": 65750 }, { "grad_norm": 0.0729522705078125, "learning_rate": 2.877239524395568e-05, "loss": 0.0047, "step": 65760 }, { "grad_norm": 0.09079056978225708, "learning_rate": 2.8757425831148134e-05, "loss": 0.0045, "step": 65770 }, { "grad_norm": 0.10807789117097855, "learning_rate": 2.874245874139658e-05, "loss": 0.0079, "step": 65780 }, { "grad_norm": 0.15598922967910767, "learning_rate": 2.8727493976337737e-05, "loss": 0.0049, "step": 65790 }, { "grad_norm": 0.10979945212602615, "learning_rate": 2.8712531537608218e-05, "loss": 0.0042, "step": 65800 }, { "grad_norm": 0.09117446094751358, "learning_rate": 2.8697571426844217e-05, "loss": 0.0065, "step": 65810 }, { "grad_norm": 0.17001435160636902, "learning_rate": 2.86826136456818e-05, "loss": 0.0052, "step": 65820 }, { "grad_norm": 0.15942272543907166, "learning_rate": 2.8667658195756685e-05, "loss": 0.0051, "step": 65830 }, { "grad_norm": 0.15342888236045837, "learning_rate": 2.8652705078704412e-05, "loss": 0.0047, "step": 65840 }, { "grad_norm": 0.06643209606409073, "learning_rate": 2.863775429616018e-05, "loss": 0.0042, "step": 65850 }, { "grad_norm": 0.06987457722425461, "learning_rate": 2.862280584975906e-05, "loss": 0.004, "step": 65860 }, { "grad_norm": 0.11469888687133789, "learning_rate": 2.8607859741135735e-05, "loss": 0.0056, "step": 65870 }, { "grad_norm": 0.11185837537050247, "learning_rate": 2.8592915971924723e-05, "loss": 0.0064, "step": 65880 }, { "grad_norm": 0.15812912583351135, "learning_rate": 2.8577974543760212e-05, "loss": 0.005, "step": 65890 }, { "grad_norm": 0.14046867191791534, "learning_rate": 2.8563035458276222e-05, "loss": 0.005, "step": 65900 }, { "grad_norm": 0.1339891105890274, "learning_rate": 2.8548098717106424e-05, "loss": 0.0062, "step": 65910 }, { "grad_norm": 0.10141867399215698, "learning_rate": 2.85331643218843e-05, "loss": 0.0057, "step": 65920 }, { "grad_norm": 0.14630241692066193, "learning_rate": 2.8518232274243046e-05, "loss": 0.0051, "step": 65930 }, { "grad_norm": 0.1127098947763443, "learning_rate": 2.8503302575815638e-05, "loss": 0.0048, "step": 65940 }, { "grad_norm": 0.09513290971517563, "learning_rate": 2.8488375228234716e-05, "loss": 0.0039, "step": 65950 }, { "grad_norm": 0.12078773230314255, "learning_rate": 2.8473450233132738e-05, "loss": 0.0056, "step": 65960 }, { "grad_norm": 0.10042659193277359, "learning_rate": 2.8458527592141898e-05, "loss": 0.0058, "step": 65970 }, { "grad_norm": 0.12347353994846344, "learning_rate": 2.8443607306894073e-05, "loss": 0.0047, "step": 65980 }, { "grad_norm": 0.1185273677110672, "learning_rate": 2.8428689379020946e-05, "loss": 0.0061, "step": 65990 }, { "grad_norm": 0.1024065613746643, "learning_rate": 2.841377381015391e-05, "loss": 0.0049, "step": 66000 }, { "grad_norm": 0.12216198444366455, "learning_rate": 2.8398860601924144e-05, "loss": 0.0044, "step": 66010 }, { "grad_norm": 0.13270355761051178, "learning_rate": 2.838394975596248e-05, "loss": 0.0051, "step": 66020 }, { "grad_norm": 0.12636183202266693, "learning_rate": 2.83690412738996e-05, "loss": 0.0039, "step": 66030 }, { "grad_norm": 0.12162748724222183, "learning_rate": 2.835413515736579e-05, "loss": 0.0051, "step": 66040 }, { "grad_norm": 0.09477367252111435, "learning_rate": 2.8339231407991267e-05, "loss": 0.0048, "step": 66050 }, { "grad_norm": 0.08280177414417267, "learning_rate": 2.832433002740581e-05, "loss": 0.0047, "step": 66060 }, { "grad_norm": 0.1337149739265442, "learning_rate": 2.8309431017239047e-05, "loss": 0.0077, "step": 66070 }, { "grad_norm": 0.12831974029541016, "learning_rate": 2.829453437912028e-05, "loss": 0.0051, "step": 66080 }, { "grad_norm": 0.1080055758357048, "learning_rate": 2.8279640114678617e-05, "loss": 0.006, "step": 66090 }, { "grad_norm": 0.10630493611097336, "learning_rate": 2.8264748225542836e-05, "loss": 0.0049, "step": 66100 }, { "grad_norm": 0.07355258613824844, "learning_rate": 2.8249858713341516e-05, "loss": 0.0064, "step": 66110 }, { "grad_norm": 0.1363835483789444, "learning_rate": 2.823497157970294e-05, "loss": 0.0073, "step": 66120 }, { "grad_norm": 0.08929687738418579, "learning_rate": 2.8220086826255164e-05, "loss": 0.0046, "step": 66130 }, { "grad_norm": 0.11674396693706512, "learning_rate": 2.8205204454625934e-05, "loss": 0.0058, "step": 66140 }, { "grad_norm": 0.10154606401920319, "learning_rate": 2.8190324466442798e-05, "loss": 0.0044, "step": 66150 }, { "grad_norm": 0.07312164455652237, "learning_rate": 2.8175446863332965e-05, "loss": 0.0055, "step": 66160 }, { "grad_norm": 0.12106971442699432, "learning_rate": 2.816057164692345e-05, "loss": 0.0047, "step": 66170 }, { "grad_norm": 0.12333239614963531, "learning_rate": 2.8145698818840992e-05, "loss": 0.0068, "step": 66180 }, { "grad_norm": 0.09462462365627289, "learning_rate": 2.8130828380712065e-05, "loss": 0.0071, "step": 66190 }, { "grad_norm": 0.12578493356704712, "learning_rate": 2.811596033416285e-05, "loss": 0.0054, "step": 66200 }, { "grad_norm": 0.09629254043102264, "learning_rate": 2.8101094680819307e-05, "loss": 0.0052, "step": 66210 }, { "grad_norm": 0.059701647609472275, "learning_rate": 2.808623142230715e-05, "loss": 0.0051, "step": 66220 }, { "grad_norm": 0.1301959604024887, "learning_rate": 2.8071370560251757e-05, "loss": 0.0045, "step": 66230 }, { "grad_norm": 0.11987816542387009, "learning_rate": 2.805651209627831e-05, "loss": 0.0043, "step": 66240 }, { "grad_norm": 0.15114089846611023, "learning_rate": 2.804165603201171e-05, "loss": 0.0049, "step": 66250 }, { "grad_norm": 0.15427595376968384, "learning_rate": 2.802680236907661e-05, "loss": 0.007, "step": 66260 }, { "grad_norm": 0.11095479875802994, "learning_rate": 2.8011951109097344e-05, "loss": 0.0041, "step": 66270 }, { "grad_norm": 0.08499818295240402, "learning_rate": 2.7997102253698072e-05, "loss": 0.0049, "step": 66280 }, { "grad_norm": 0.08644920587539673, "learning_rate": 2.7982255804502588e-05, "loss": 0.0052, "step": 66290 }, { "grad_norm": 0.1068243607878685, "learning_rate": 2.7967411763134516e-05, "loss": 0.0054, "step": 66300 }, { "grad_norm": 0.15232184529304504, "learning_rate": 2.7952570131217166e-05, "loss": 0.0074, "step": 66310 }, { "grad_norm": 0.11995317041873932, "learning_rate": 2.7937730910373618e-05, "loss": 0.0043, "step": 66320 }, { "grad_norm": 0.10209439694881439, "learning_rate": 2.7922894102226628e-05, "loss": 0.0067, "step": 66330 }, { "grad_norm": 0.08951754122972488, "learning_rate": 2.790805970839876e-05, "loss": 0.006, "step": 66340 }, { "grad_norm": 0.0751689225435257, "learning_rate": 2.789322773051225e-05, "loss": 0.004, "step": 66350 }, { "grad_norm": 0.15278470516204834, "learning_rate": 2.7878398170189124e-05, "loss": 0.0042, "step": 66360 }, { "grad_norm": 0.1021941751241684, "learning_rate": 2.7863571029051112e-05, "loss": 0.007, "step": 66370 }, { "grad_norm": 0.13224662840366364, "learning_rate": 2.784874630871971e-05, "loss": 0.0041, "step": 66380 }, { "grad_norm": 0.142435222864151, "learning_rate": 2.7833924010816086e-05, "loss": 0.0048, "step": 66390 }, { "grad_norm": 0.1175817921757698, "learning_rate": 2.7819104136961226e-05, "loss": 0.0058, "step": 66400 }, { "grad_norm": 0.1297113448381424, "learning_rate": 2.780428668877577e-05, "loss": 0.0049, "step": 66410 }, { "grad_norm": 0.13082532584667206, "learning_rate": 2.7789471667880152e-05, "loss": 0.0079, "step": 66420 }, { "grad_norm": 0.09651421010494232, "learning_rate": 2.7774659075894523e-05, "loss": 0.0072, "step": 66430 }, { "grad_norm": 0.10672298073768616, "learning_rate": 2.7759848914438778e-05, "loss": 0.0057, "step": 66440 }, { "grad_norm": 0.1383606344461441, "learning_rate": 2.7745041185132496e-05, "loss": 0.0044, "step": 66450 }, { "grad_norm": 0.1503715068101883, "learning_rate": 2.773023588959506e-05, "loss": 0.0088, "step": 66460 }, { "grad_norm": 0.139196515083313, "learning_rate": 2.7715433029445555e-05, "loss": 0.006, "step": 66470 }, { "grad_norm": 0.11174939572811127, "learning_rate": 2.7700632606302766e-05, "loss": 0.0064, "step": 66480 }, { "grad_norm": 0.10174679756164551, "learning_rate": 2.768583462178528e-05, "loss": 0.0057, "step": 66490 }, { "grad_norm": 0.10373058915138245, "learning_rate": 2.767103907751136e-05, "loss": 0.0036, "step": 66500 }, { "grad_norm": 0.16300661861896515, "learning_rate": 2.7656245975099065e-05, "loss": 0.0052, "step": 66510 }, { "grad_norm": 0.14759013056755066, "learning_rate": 2.7641455316166093e-05, "loss": 0.005, "step": 66520 }, { "grad_norm": 0.12290707230567932, "learning_rate": 2.7626667102329973e-05, "loss": 0.0044, "step": 66530 }, { "grad_norm": 0.1328609436750412, "learning_rate": 2.761188133520788e-05, "loss": 0.0053, "step": 66540 }, { "grad_norm": 0.12809239327907562, "learning_rate": 2.7597098016416784e-05, "loss": 0.005, "step": 66550 }, { "grad_norm": 0.11899831891059875, "learning_rate": 2.7582317147573368e-05, "loss": 0.006, "step": 66560 }, { "grad_norm": 0.11547955125570297, "learning_rate": 2.7567538730294064e-05, "loss": 0.0036, "step": 66570 }, { "grad_norm": 0.11093103140592575, "learning_rate": 2.7552762766194973e-05, "loss": 0.0053, "step": 66580 }, { "grad_norm": 0.09158015251159668, "learning_rate": 2.753798925689202e-05, "loss": 0.0079, "step": 66590 }, { "grad_norm": 0.16287806630134583, "learning_rate": 2.7523218204000774e-05, "loss": 0.0071, "step": 66600 }, { "grad_norm": 0.10895472764968872, "learning_rate": 2.7508449609136612e-05, "loss": 0.0039, "step": 66610 }, { "grad_norm": 0.09224008768796921, "learning_rate": 2.749368347391454e-05, "loss": 0.0043, "step": 66620 }, { "grad_norm": 0.11620213836431503, "learning_rate": 2.7478919799949443e-05, "loss": 0.0045, "step": 66630 }, { "grad_norm": 0.15095853805541992, "learning_rate": 2.7464158588855803e-05, "loss": 0.0054, "step": 66640 }, { "grad_norm": 0.1561315357685089, "learning_rate": 2.7449399842247918e-05, "loss": 0.0057, "step": 66650 }, { "grad_norm": 0.08937028050422668, "learning_rate": 2.7434643561739736e-05, "loss": 0.0044, "step": 66660 }, { "grad_norm": 0.113884836435318, "learning_rate": 2.7419889748945016e-05, "loss": 0.0051, "step": 66670 }, { "grad_norm": 0.11346250772476196, "learning_rate": 2.7405138405477203e-05, "loss": 0.0056, "step": 66680 }, { "grad_norm": 0.15217138826847076, "learning_rate": 2.73903895329495e-05, "loss": 0.0056, "step": 66690 }, { "grad_norm": 0.11209623515605927, "learning_rate": 2.737564313297478e-05, "loss": 0.0046, "step": 66700 }, { "grad_norm": 0.09390133619308472, "learning_rate": 2.736089920716572e-05, "loss": 0.0043, "step": 66710 }, { "grad_norm": 0.17752698063850403, "learning_rate": 2.7346157757134706e-05, "loss": 0.0068, "step": 66720 }, { "grad_norm": 0.082334503531456, "learning_rate": 2.733141878449379e-05, "loss": 0.0054, "step": 66730 }, { "grad_norm": 0.0893101692199707, "learning_rate": 2.7316682290854844e-05, "loss": 0.004, "step": 66740 }, { "grad_norm": 0.11146488040685654, "learning_rate": 2.7301948277829414e-05, "loss": 0.0081, "step": 66750 }, { "grad_norm": 0.10974671691656113, "learning_rate": 2.7287216747028814e-05, "loss": 0.0077, "step": 66760 }, { "grad_norm": 0.17356249690055847, "learning_rate": 2.7272487700064025e-05, "loss": 0.0065, "step": 66770 }, { "grad_norm": 0.15107719600200653, "learning_rate": 2.725776113854583e-05, "loss": 0.0046, "step": 66780 }, { "grad_norm": 0.11091279983520508, "learning_rate": 2.7243037064084664e-05, "loss": 0.0058, "step": 66790 }, { "grad_norm": 0.11789222061634064, "learning_rate": 2.7228315478290766e-05, "loss": 0.005, "step": 66800 }, { "grad_norm": 0.093501016497612, "learning_rate": 2.721359638277402e-05, "loss": 0.0059, "step": 66810 }, { "grad_norm": 0.13182149827480316, "learning_rate": 2.7198879779144148e-05, "loss": 0.0048, "step": 66820 }, { "grad_norm": 0.11410156637430191, "learning_rate": 2.7184165669010485e-05, "loss": 0.0051, "step": 66830 }, { "grad_norm": 0.13241761922836304, "learning_rate": 2.716945405398218e-05, "loss": 0.0062, "step": 66840 }, { "grad_norm": 0.1723746806383133, "learning_rate": 2.715474493566803e-05, "loss": 0.0047, "step": 66850 }, { "grad_norm": 0.11273668706417084, "learning_rate": 2.7140038315676652e-05, "loss": 0.0072, "step": 66860 }, { "grad_norm": 0.12902633845806122, "learning_rate": 2.7125334195616275e-05, "loss": 0.0053, "step": 66870 }, { "grad_norm": 0.11357533931732178, "learning_rate": 2.7110632577094997e-05, "loss": 0.0045, "step": 66880 }, { "grad_norm": 0.07340236008167267, "learning_rate": 2.709593346172051e-05, "loss": 0.0069, "step": 66890 }, { "grad_norm": 0.1125086098909378, "learning_rate": 2.708123685110032e-05, "loss": 0.0059, "step": 66900 }, { "grad_norm": 0.11900102347135544, "learning_rate": 2.7066542746841588e-05, "loss": 0.0045, "step": 66910 }, { "grad_norm": 0.10696262866258621, "learning_rate": 2.7051851150551267e-05, "loss": 0.0066, "step": 66920 }, { "grad_norm": 0.07666905969381332, "learning_rate": 2.7037162063835996e-05, "loss": 0.0076, "step": 66930 }, { "grad_norm": 0.08137094229459763, "learning_rate": 2.702247548830218e-05, "loss": 0.0052, "step": 66940 }, { "grad_norm": 0.1708681732416153, "learning_rate": 2.7007791425555883e-05, "loss": 0.0035, "step": 66950 }, { "grad_norm": 0.10567349195480347, "learning_rate": 2.6993109877202945e-05, "loss": 0.0056, "step": 66960 }, { "grad_norm": 0.12759649753570557, "learning_rate": 2.6978430844848935e-05, "loss": 0.0051, "step": 66970 }, { "grad_norm": 0.13715428113937378, "learning_rate": 2.69637543300991e-05, "loss": 0.0045, "step": 66980 }, { "grad_norm": 0.09061299264431, "learning_rate": 2.694908033455848e-05, "loss": 0.0044, "step": 66990 }, { "grad_norm": 0.17642498016357422, "learning_rate": 2.6934408859831733e-05, "loss": 0.0037, "step": 67000 }, { "grad_norm": 0.14711174368858337, "learning_rate": 2.69197399075234e-05, "loss": 0.0052, "step": 67010 }, { "grad_norm": 0.08656801283359528, "learning_rate": 2.6905073479237584e-05, "loss": 0.0054, "step": 67020 }, { "grad_norm": 0.12810015678405762, "learning_rate": 2.6890409576578235e-05, "loss": 0.0054, "step": 67030 }, { "grad_norm": 0.10579416155815125, "learning_rate": 2.687574820114892e-05, "loss": 0.0044, "step": 67040 }, { "grad_norm": 0.16077172756195068, "learning_rate": 2.686108935455305e-05, "loss": 0.0042, "step": 67050 }, { "grad_norm": 0.15889130532741547, "learning_rate": 2.6846433038393603e-05, "loss": 0.006, "step": 67060 }, { "grad_norm": 0.14497628808021545, "learning_rate": 2.6831779254273482e-05, "loss": 0.0048, "step": 67070 }, { "grad_norm": 0.12368078529834747, "learning_rate": 2.6817128003795122e-05, "loss": 0.0042, "step": 67080 }, { "grad_norm": 0.13047409057617188, "learning_rate": 2.680247928856081e-05, "loss": 0.0057, "step": 67090 }, { "grad_norm": 0.17372511327266693, "learning_rate": 2.6787833110172466e-05, "loss": 0.0051, "step": 67100 }, { "grad_norm": 0.1095457375049591, "learning_rate": 2.6773189470231807e-05, "loss": 0.0058, "step": 67110 }, { "grad_norm": 0.10135465115308762, "learning_rate": 2.675854837034022e-05, "loss": 0.0057, "step": 67120 }, { "grad_norm": 0.11237550526857376, "learning_rate": 2.6743909812098833e-05, "loss": 0.0041, "step": 67130 }, { "grad_norm": 0.11088212579488754, "learning_rate": 2.67292737971085e-05, "loss": 0.0041, "step": 67140 }, { "grad_norm": 0.06679287552833557, "learning_rate": 2.671464032696982e-05, "loss": 0.0053, "step": 67150 }, { "grad_norm": 0.13539369404315948, "learning_rate": 2.6700009403283055e-05, "loss": 0.0073, "step": 67160 }, { "grad_norm": 0.15466654300689697, "learning_rate": 2.6685381027648214e-05, "loss": 0.0048, "step": 67170 }, { "grad_norm": 0.08789556473493576, "learning_rate": 2.6670755201665086e-05, "loss": 0.0044, "step": 67180 }, { "grad_norm": 0.07607682049274445, "learning_rate": 2.6656131926933066e-05, "loss": 0.0046, "step": 67190 }, { "grad_norm": 0.08306961506605148, "learning_rate": 2.664151120505136e-05, "loss": 0.0053, "step": 67200 }, { "grad_norm": 0.15127506852149963, "learning_rate": 2.6626893037618873e-05, "loss": 0.0064, "step": 67210 }, { "grad_norm": 0.13822245597839355, "learning_rate": 2.6612277426234243e-05, "loss": 0.0051, "step": 67220 }, { "grad_norm": 0.14920653402805328, "learning_rate": 2.659766437249577e-05, "loss": 0.0049, "step": 67230 }, { "grad_norm": 0.10409195721149445, "learning_rate": 2.6583053878001557e-05, "loss": 0.0047, "step": 67240 }, { "grad_norm": 0.12659282982349396, "learning_rate": 2.6568445944349323e-05, "loss": 0.0045, "step": 67250 }, { "grad_norm": 0.1853860467672348, "learning_rate": 2.6553840573136657e-05, "loss": 0.0052, "step": 67260 }, { "grad_norm": 0.07644278556108475, "learning_rate": 2.6539237765960706e-05, "loss": 0.0054, "step": 67270 }, { "grad_norm": 0.1920984387397766, "learning_rate": 2.6524637524418465e-05, "loss": 0.0075, "step": 67280 }, { "grad_norm": 0.11215865612030029, "learning_rate": 2.651003985010655e-05, "loss": 0.0086, "step": 67290 }, { "grad_norm": 0.08399554342031479, "learning_rate": 2.649544474462138e-05, "loss": 0.0055, "step": 67300 }, { "grad_norm": 0.10025707632303238, "learning_rate": 2.6480852209559015e-05, "loss": 0.0051, "step": 67310 }, { "grad_norm": 0.10084450244903564, "learning_rate": 2.6466262246515296e-05, "loss": 0.0044, "step": 67320 }, { "grad_norm": 0.15172109007835388, "learning_rate": 2.6451674857085746e-05, "loss": 0.004, "step": 67330 }, { "grad_norm": 0.08399905264377594, "learning_rate": 2.6437090042865655e-05, "loss": 0.0054, "step": 67340 }, { "grad_norm": 0.1330430954694748, "learning_rate": 2.6422507805449947e-05, "loss": 0.0052, "step": 67350 }, { "grad_norm": 0.1014239564538002, "learning_rate": 2.6407928146433357e-05, "loss": 0.0048, "step": 67360 }, { "grad_norm": 0.1141059473156929, "learning_rate": 2.639335106741026e-05, "loss": 0.0045, "step": 67370 }, { "grad_norm": 0.06135743856430054, "learning_rate": 2.6378776569974796e-05, "loss": 0.0059, "step": 67380 }, { "grad_norm": 0.08778037130832672, "learning_rate": 2.6364204655720814e-05, "loss": 0.0056, "step": 67390 }, { "grad_norm": 0.07262758910655975, "learning_rate": 2.6349635326241894e-05, "loss": 0.004, "step": 67400 }, { "grad_norm": 0.12583987414836884, "learning_rate": 2.633506858313127e-05, "loss": 0.0064, "step": 67410 }, { "grad_norm": 0.13563133776187897, "learning_rate": 2.6320504427981975e-05, "loss": 0.0042, "step": 67420 }, { "grad_norm": 0.1262681931257248, "learning_rate": 2.6305942862386734e-05, "loss": 0.0078, "step": 67430 }, { "grad_norm": 0.12068738788366318, "learning_rate": 2.629138388793793e-05, "loss": 0.0048, "step": 67440 }, { "grad_norm": 0.13533686101436615, "learning_rate": 2.6276827506227743e-05, "loss": 0.0058, "step": 67450 }, { "grad_norm": 0.10716971755027771, "learning_rate": 2.6262273718848037e-05, "loss": 0.0049, "step": 67460 }, { "grad_norm": 0.13635966181755066, "learning_rate": 2.6247722527390407e-05, "loss": 0.0048, "step": 67470 }, { "grad_norm": 0.10889003425836563, "learning_rate": 2.6233173933446104e-05, "loss": 0.0077, "step": 67480 }, { "grad_norm": 0.11304029077291489, "learning_rate": 2.6218627938606188e-05, "loss": 0.0069, "step": 67490 }, { "grad_norm": 0.09816870093345642, "learning_rate": 2.620408454446135e-05, "loss": 0.0054, "step": 67500 }, { "grad_norm": 0.14257557690143585, "learning_rate": 2.6189543752602046e-05, "loss": 0.005, "step": 67510 }, { "grad_norm": 0.12148000299930573, "learning_rate": 2.6175005564618444e-05, "loss": 0.0062, "step": 67520 }, { "grad_norm": 0.11757359653711319, "learning_rate": 2.6160469982100428e-05, "loss": 0.0071, "step": 67530 }, { "grad_norm": 0.0919453352689743, "learning_rate": 2.6145937006637555e-05, "loss": 0.0036, "step": 67540 }, { "grad_norm": 0.1448967009782791, "learning_rate": 2.6131406639819166e-05, "loss": 0.0037, "step": 67550 }, { "grad_norm": 0.09517555683851242, "learning_rate": 2.611687888323424e-05, "loss": 0.0047, "step": 67560 }, { "grad_norm": 0.13261306285858154, "learning_rate": 2.610235373847153e-05, "loss": 0.0044, "step": 67570 }, { "grad_norm": 0.11492098867893219, "learning_rate": 2.6087831207119494e-05, "loss": 0.0047, "step": 67580 }, { "grad_norm": 0.16554071009159088, "learning_rate": 2.60733112907663e-05, "loss": 0.0044, "step": 67590 }, { "grad_norm": 0.0756220817565918, "learning_rate": 2.6058793990999786e-05, "loss": 0.0045, "step": 67600 }, { "grad_norm": 0.13681116700172424, "learning_rate": 2.6044279309407593e-05, "loss": 0.0053, "step": 67610 }, { "grad_norm": 0.11215928196907043, "learning_rate": 2.6029767247576976e-05, "loss": 0.0057, "step": 67620 }, { "grad_norm": 0.12956173717975616, "learning_rate": 2.601525780709497e-05, "loss": 0.0056, "step": 67630 }, { "grad_norm": 0.10784135013818741, "learning_rate": 2.6000750989548318e-05, "loss": 0.0047, "step": 67640 }, { "grad_norm": 0.11775650084018707, "learning_rate": 2.5986246796523462e-05, "loss": 0.0043, "step": 67650 }, { "grad_norm": 0.12519796192646027, "learning_rate": 2.5971745229606546e-05, "loss": 0.0045, "step": 67660 }, { "grad_norm": 0.0834108516573906, "learning_rate": 2.5957246290383442e-05, "loss": 0.0059, "step": 67670 }, { "grad_norm": 0.1579824835062027, "learning_rate": 2.594274998043975e-05, "loss": 0.0065, "step": 67680 }, { "grad_norm": 0.12634195387363434, "learning_rate": 2.5928256301360733e-05, "loss": 0.0049, "step": 67690 }, { "grad_norm": 0.06969202309846878, "learning_rate": 2.591376525473142e-05, "loss": 0.0069, "step": 67700 }, { "grad_norm": 0.11628543585538864, "learning_rate": 2.5899276842136523e-05, "loss": 0.0062, "step": 67710 }, { "grad_norm": 0.1372608095407486, "learning_rate": 2.5884791065160495e-05, "loss": 0.0062, "step": 67720 }, { "grad_norm": 0.12858714163303375, "learning_rate": 2.5870307925387436e-05, "loss": 0.0064, "step": 67730 }, { "grad_norm": 0.15289439260959625, "learning_rate": 2.585582742440125e-05, "loss": 0.0052, "step": 67740 }, { "grad_norm": 0.11724267154932022, "learning_rate": 2.5841349563785455e-05, "loss": 0.0047, "step": 67750 }, { "grad_norm": 0.1291476935148239, "learning_rate": 2.5826874345123355e-05, "loss": 0.0074, "step": 67760 }, { "grad_norm": 0.13074658811092377, "learning_rate": 2.581240176999792e-05, "loss": 0.0049, "step": 67770 }, { "grad_norm": 0.11578287929296494, "learning_rate": 2.579793183999189e-05, "loss": 0.0046, "step": 67780 }, { "grad_norm": 0.16203707456588745, "learning_rate": 2.5783464556687627e-05, "loss": 0.0051, "step": 67790 }, { "grad_norm": 0.08251847326755524, "learning_rate": 2.576899992166729e-05, "loss": 0.0042, "step": 67800 }, { "grad_norm": 0.10882317274808884, "learning_rate": 2.5754537936512668e-05, "loss": 0.007, "step": 67810 }, { "grad_norm": 0.07327092438936234, "learning_rate": 2.574007860280535e-05, "loss": 0.0037, "step": 67820 }, { "grad_norm": 0.14677263796329498, "learning_rate": 2.5725621922126518e-05, "loss": 0.0054, "step": 67830 }, { "grad_norm": 0.1779220551252365, "learning_rate": 2.571116789605722e-05, "loss": 0.0049, "step": 67840 }, { "grad_norm": 0.0823485478758812, "learning_rate": 2.5696716526178054e-05, "loss": 0.0051, "step": 67850 }, { "grad_norm": 0.1373339593410492, "learning_rate": 2.568226781406946e-05, "loss": 0.0033, "step": 67860 }, { "grad_norm": 0.13961003720760345, "learning_rate": 2.5667821761311474e-05, "loss": 0.0068, "step": 67870 }, { "grad_norm": 0.10061261057853699, "learning_rate": 2.5653378369483917e-05, "loss": 0.0047, "step": 67880 }, { "grad_norm": 0.12539170682430267, "learning_rate": 2.5638937640166292e-05, "loss": 0.0056, "step": 67890 }, { "grad_norm": 0.1209067851305008, "learning_rate": 2.5624499574937842e-05, "loss": 0.0049, "step": 67900 }, { "grad_norm": 0.09092459082603455, "learning_rate": 2.5610064175377456e-05, "loss": 0.0078, "step": 67910 }, { "grad_norm": 0.12097529321908951, "learning_rate": 2.5595631443063782e-05, "loss": 0.0044, "step": 67920 }, { "grad_norm": 0.08569561690092087, "learning_rate": 2.5581201379575177e-05, "loss": 0.004, "step": 67930 }, { "grad_norm": 0.13688263297080994, "learning_rate": 2.5566773986489655e-05, "loss": 0.0057, "step": 67940 }, { "grad_norm": 0.10547429323196411, "learning_rate": 2.5552349265385e-05, "loss": 0.0044, "step": 67950 }, { "grad_norm": 0.15092648565769196, "learning_rate": 2.5537927217838675e-05, "loss": 0.0042, "step": 67960 }, { "grad_norm": 0.08654393255710602, "learning_rate": 2.5523507845427864e-05, "loss": 0.0069, "step": 67970 }, { "grad_norm": 0.09963885694742203, "learning_rate": 2.5509091149729413e-05, "loss": 0.0041, "step": 67980 }, { "grad_norm": 0.1201714426279068, "learning_rate": 2.5494677132319962e-05, "loss": 0.0054, "step": 67990 }, { "grad_norm": 0.09690962731838226, "learning_rate": 2.548026579477575e-05, "loss": 0.0066, "step": 68000 }, { "grad_norm": 0.11889202147722244, "learning_rate": 2.5465857138672834e-05, "loss": 0.0046, "step": 68010 }, { "grad_norm": 0.07040698081254959, "learning_rate": 2.5451451165586848e-05, "loss": 0.0031, "step": 68020 }, { "grad_norm": 0.09284055978059769, "learning_rate": 2.543704787709329e-05, "loss": 0.0042, "step": 68030 }, { "grad_norm": 0.09045374393463135, "learning_rate": 2.542264727476723e-05, "loss": 0.0054, "step": 68040 }, { "grad_norm": 0.09729855507612228, "learning_rate": 2.5408249360183535e-05, "loss": 0.0058, "step": 68050 }, { "grad_norm": 0.12524856626987457, "learning_rate": 2.5393854134916694e-05, "loss": 0.006, "step": 68060 }, { "grad_norm": 0.13201457262039185, "learning_rate": 2.5379461600540993e-05, "loss": 0.0048, "step": 68070 }, { "grad_norm": 0.13184970617294312, "learning_rate": 2.5365071758630317e-05, "loss": 0.0066, "step": 68080 }, { "grad_norm": 0.10343848913908005, "learning_rate": 2.535068461075839e-05, "loss": 0.0055, "step": 68090 }, { "grad_norm": 0.19117730855941772, "learning_rate": 2.5336300158498516e-05, "loss": 0.005, "step": 68100 }, { "grad_norm": 0.10927309840917587, "learning_rate": 2.5321918403423793e-05, "loss": 0.0044, "step": 68110 }, { "grad_norm": 0.10027018189430237, "learning_rate": 2.5307539347106957e-05, "loss": 0.0046, "step": 68120 }, { "grad_norm": 0.09433623403310776, "learning_rate": 2.529316299112049e-05, "loss": 0.0052, "step": 68130 }, { "grad_norm": 0.1577429324388504, "learning_rate": 2.5278789337036567e-05, "loss": 0.0057, "step": 68140 }, { "grad_norm": 0.14245173335075378, "learning_rate": 2.52644183864271e-05, "loss": 0.0049, "step": 68150 }, { "grad_norm": 0.1450122743844986, "learning_rate": 2.525005014086363e-05, "loss": 0.0047, "step": 68160 }, { "grad_norm": 0.10890938341617584, "learning_rate": 2.5235684601917465e-05, "loss": 0.0046, "step": 68170 }, { "grad_norm": 0.17025253176689148, "learning_rate": 2.5221321771159618e-05, "loss": 0.0047, "step": 68180 }, { "grad_norm": 0.07949446141719818, "learning_rate": 2.5206961650160743e-05, "loss": 0.0043, "step": 68190 }, { "grad_norm": 0.10983049124479294, "learning_rate": 2.519260424049129e-05, "loss": 0.006, "step": 68200 }, { "grad_norm": 0.1312108337879181, "learning_rate": 2.5178249543721304e-05, "loss": 0.006, "step": 68210 }, { "grad_norm": 0.23094990849494934, "learning_rate": 2.5163897561420668e-05, "loss": 0.0072, "step": 68220 }, { "grad_norm": 0.10751669853925705, "learning_rate": 2.514954829515883e-05, "loss": 0.0063, "step": 68230 }, { "grad_norm": 0.15182438492774963, "learning_rate": 2.513520174650505e-05, "loss": 0.0059, "step": 68240 }, { "grad_norm": 0.13654343783855438, "learning_rate": 2.5120857917028207e-05, "loss": 0.0055, "step": 68250 }, { "grad_norm": 0.15214911103248596, "learning_rate": 2.5106516808296953e-05, "loss": 0.0043, "step": 68260 }, { "grad_norm": 0.12099780142307281, "learning_rate": 2.5092178421879552e-05, "loss": 0.0043, "step": 68270 }, { "grad_norm": 0.10000063478946686, "learning_rate": 2.5077842759344116e-05, "loss": 0.0043, "step": 68280 }, { "grad_norm": 0.10641973465681076, "learning_rate": 2.5063509822258304e-05, "loss": 0.0073, "step": 68290 }, { "grad_norm": 0.12321434915065765, "learning_rate": 2.5049179612189576e-05, "loss": 0.0048, "step": 68300 }, { "grad_norm": 0.1417333036661148, "learning_rate": 2.5034852130705043e-05, "loss": 0.0055, "step": 68310 }, { "grad_norm": 0.08912932872772217, "learning_rate": 2.5020527379371572e-05, "loss": 0.0048, "step": 68320 }, { "grad_norm": 0.13845615088939667, "learning_rate": 2.500620535975564e-05, "loss": 0.0049, "step": 68330 }, { "grad_norm": 0.07382988184690475, "learning_rate": 2.4991886073423515e-05, "loss": 0.005, "step": 68340 }, { "grad_norm": 0.10821476578712463, "learning_rate": 2.4977569521941134e-05, "loss": 0.0042, "step": 68350 }, { "grad_norm": 0.08630312234163284, "learning_rate": 2.4963255706874143e-05, "loss": 0.0039, "step": 68360 }, { "grad_norm": 0.14510944485664368, "learning_rate": 2.4948944629787853e-05, "loss": 0.005, "step": 68370 }, { "grad_norm": 0.12660139799118042, "learning_rate": 2.493463629224731e-05, "loss": 0.004, "step": 68380 }, { "grad_norm": 0.1031523272395134, "learning_rate": 2.492033069581728e-05, "loss": 0.0034, "step": 68390 }, { "grad_norm": 0.08861250430345535, "learning_rate": 2.4906027842062137e-05, "loss": 0.0056, "step": 68400 }, { "grad_norm": 0.10056524723768234, "learning_rate": 2.48917277325461e-05, "loss": 0.0039, "step": 68410 }, { "grad_norm": 0.08879955112934113, "learning_rate": 2.487743036883295e-05, "loss": 0.0056, "step": 68420 }, { "grad_norm": 0.17061187326908112, "learning_rate": 2.486313575248626e-05, "loss": 0.0042, "step": 68430 }, { "grad_norm": 0.14827971160411835, "learning_rate": 2.4848843885069235e-05, "loss": 0.0038, "step": 68440 }, { "grad_norm": 0.29508817195892334, "learning_rate": 2.483455476814484e-05, "loss": 0.0049, "step": 68450 }, { "grad_norm": 0.14797937870025635, "learning_rate": 2.4820268403275672e-05, "loss": 0.005, "step": 68460 }, { "grad_norm": 0.06427179276943207, "learning_rate": 2.480598479202413e-05, "loss": 0.0034, "step": 68470 }, { "grad_norm": 0.09501741081476212, "learning_rate": 2.4791703935952193e-05, "loss": 0.0049, "step": 68480 }, { "grad_norm": 0.15827037394046783, "learning_rate": 2.477742583662163e-05, "loss": 0.0046, "step": 68490 }, { "grad_norm": 0.0998869389295578, "learning_rate": 2.476315049559384e-05, "loss": 0.0057, "step": 68500 }, { "grad_norm": 0.11492506414651871, "learning_rate": 2.4748877914429987e-05, "loss": 0.0083, "step": 68510 }, { "grad_norm": 0.1042397990822792, "learning_rate": 2.4734608094690857e-05, "loss": 0.0044, "step": 68520 }, { "grad_norm": 0.11524323374032974, "learning_rate": 2.4720341037937006e-05, "loss": 0.0037, "step": 68530 }, { "grad_norm": 0.12223101407289505, "learning_rate": 2.4706076745728645e-05, "loss": 0.0038, "step": 68540 }, { "grad_norm": 0.12319614738225937, "learning_rate": 2.469181521962573e-05, "loss": 0.0045, "step": 68550 }, { "grad_norm": 0.09896278381347656, "learning_rate": 2.467755646118783e-05, "loss": 0.0049, "step": 68560 }, { "grad_norm": 0.18747535347938538, "learning_rate": 2.46633004719743e-05, "loss": 0.0057, "step": 68570 }, { "grad_norm": 0.12904861569404602, "learning_rate": 2.4649047253544123e-05, "loss": 0.0048, "step": 68580 }, { "grad_norm": 0.08771666139364243, "learning_rate": 2.4634796807456016e-05, "loss": 0.0042, "step": 68590 }, { "grad_norm": 0.1109798476099968, "learning_rate": 2.46205491352684e-05, "loss": 0.0059, "step": 68600 }, { "grad_norm": 0.10053396224975586, "learning_rate": 2.4606304238539375e-05, "loss": 0.0044, "step": 68610 }, { "grad_norm": 0.06760876625776291, "learning_rate": 2.4592062118826753e-05, "loss": 0.006, "step": 68620 }, { "grad_norm": 0.08792532980442047, "learning_rate": 2.4577822777688004e-05, "loss": 0.0039, "step": 68630 }, { "grad_norm": 0.21594426035881042, "learning_rate": 2.4563586216680347e-05, "loss": 0.0063, "step": 68640 }, { "grad_norm": 0.11977411061525345, "learning_rate": 2.4549352437360622e-05, "loss": 0.0059, "step": 68650 }, { "grad_norm": 0.1285611391067505, "learning_rate": 2.4535121441285493e-05, "loss": 0.0041, "step": 68660 }, { "grad_norm": 0.1206798329949379, "learning_rate": 2.4520893230011172e-05, "loss": 0.0051, "step": 68670 }, { "grad_norm": 0.09938936680555344, "learning_rate": 2.450666780509368e-05, "loss": 0.0043, "step": 68680 }, { "grad_norm": 0.09014317393302917, "learning_rate": 2.4492445168088656e-05, "loss": 0.0039, "step": 68690 }, { "grad_norm": 0.09397868067026138, "learning_rate": 2.447822532055149e-05, "loss": 0.0039, "step": 68700 }, { "grad_norm": 0.11690425872802734, "learning_rate": 2.4464008264037212e-05, "loss": 0.007, "step": 68710 }, { "grad_norm": 0.12027287483215332, "learning_rate": 2.4449794000100605e-05, "loss": 0.0074, "step": 68720 }, { "grad_norm": 0.14156819880008698, "learning_rate": 2.4435582530296108e-05, "loss": 0.0044, "step": 68730 }, { "grad_norm": 0.08390264958143234, "learning_rate": 2.4421373856177887e-05, "loss": 0.0057, "step": 68740 }, { "grad_norm": 0.1434953212738037, "learning_rate": 2.440716797929975e-05, "loss": 0.005, "step": 68750 }, { "grad_norm": 0.16587108373641968, "learning_rate": 2.439296490121526e-05, "loss": 0.0048, "step": 68760 }, { "grad_norm": 0.14006146788597107, "learning_rate": 2.4378764623477608e-05, "loss": 0.0052, "step": 68770 }, { "grad_norm": 0.07622548937797546, "learning_rate": 2.436456714763974e-05, "loss": 0.0041, "step": 68780 }, { "grad_norm": 0.1131749227643013, "learning_rate": 2.4350372475254262e-05, "loss": 0.0049, "step": 68790 }, { "grad_norm": 0.07588174194097519, "learning_rate": 2.433618060787351e-05, "loss": 0.0053, "step": 68800 }, { "grad_norm": 0.10511911660432816, "learning_rate": 2.432199154704944e-05, "loss": 0.0057, "step": 68810 }, { "grad_norm": 0.08492796123027802, "learning_rate": 2.430780529433377e-05, "loss": 0.007, "step": 68820 }, { "grad_norm": 0.08520055562257767, "learning_rate": 2.42936218512779e-05, "loss": 0.0088, "step": 68830 }, { "grad_norm": 0.14043521881103516, "learning_rate": 2.427944121943288e-05, "loss": 0.0051, "step": 68840 }, { "grad_norm": 0.10290079563856125, "learning_rate": 2.4265263400349497e-05, "loss": 0.0055, "step": 68850 }, { "grad_norm": 0.17477433383464813, "learning_rate": 2.4251088395578213e-05, "loss": 0.0064, "step": 68860 }, { "grad_norm": 0.16328713297843933, "learning_rate": 2.4236916206669214e-05, "loss": 0.0062, "step": 68870 }, { "grad_norm": 0.0932508036494255, "learning_rate": 2.4222746835172304e-05, "loss": 0.004, "step": 68880 }, { "grad_norm": 0.15182553231716156, "learning_rate": 2.4208580282637065e-05, "loss": 0.0055, "step": 68890 }, { "grad_norm": 0.14790405333042145, "learning_rate": 2.4194416550612687e-05, "loss": 0.0053, "step": 68900 }, { "grad_norm": 0.108685702085495, "learning_rate": 2.418025564064812e-05, "loss": 0.0046, "step": 68910 }, { "grad_norm": 0.09817440807819366, "learning_rate": 2.4166097554291978e-05, "loss": 0.0048, "step": 68920 }, { "grad_norm": 0.11365779489278793, "learning_rate": 2.4151942293092583e-05, "loss": 0.0039, "step": 68930 }, { "grad_norm": 0.12333571910858154, "learning_rate": 2.41377898585979e-05, "loss": 0.0055, "step": 68940 }, { "grad_norm": 0.20560520887374878, "learning_rate": 2.4123640252355652e-05, "loss": 0.0068, "step": 68950 }, { "grad_norm": 0.11207633465528488, "learning_rate": 2.410949347591318e-05, "loss": 0.0044, "step": 68960 }, { "grad_norm": 0.11402539163827896, "learning_rate": 2.4095349530817578e-05, "loss": 0.0059, "step": 68970 }, { "grad_norm": 0.10099085420370102, "learning_rate": 2.4081208418615603e-05, "loss": 0.0046, "step": 68980 }, { "grad_norm": 0.10063830018043518, "learning_rate": 2.4067070140853725e-05, "loss": 0.0046, "step": 68990 }, { "grad_norm": 0.1791531890630722, "learning_rate": 2.405293469907805e-05, "loss": 0.0047, "step": 69000 }, { "grad_norm": 0.15637001395225525, "learning_rate": 2.4038802094834446e-05, "loss": 0.0044, "step": 69010 }, { "grad_norm": 0.21447619795799255, "learning_rate": 2.4024672329668395e-05, "loss": 0.0051, "step": 69020 }, { "grad_norm": 0.11882328242063522, "learning_rate": 2.4010545405125125e-05, "loss": 0.0046, "step": 69030 }, { "grad_norm": 0.15675611793994904, "learning_rate": 2.3996421322749528e-05, "loss": 0.0035, "step": 69040 }, { "grad_norm": 0.13561539351940155, "learning_rate": 2.3982300084086224e-05, "loss": 0.0072, "step": 69050 }, { "grad_norm": 0.14836500585079193, "learning_rate": 2.396818169067944e-05, "loss": 0.0043, "step": 69060 }, { "grad_norm": 0.07437289506196976, "learning_rate": 2.3954066144073178e-05, "loss": 0.0033, "step": 69070 }, { "grad_norm": 0.10995552688837051, "learning_rate": 2.39399534458111e-05, "loss": 0.0042, "step": 69080 }, { "grad_norm": 0.07509522140026093, "learning_rate": 2.392584359743651e-05, "loss": 0.0039, "step": 69090 }, { "grad_norm": 0.09385998547077179, "learning_rate": 2.3911736600492463e-05, "loss": 0.0051, "step": 69100 }, { "grad_norm": 0.14667938649654388, "learning_rate": 2.3897632456521672e-05, "loss": 0.0052, "step": 69110 }, { "grad_norm": 0.1307367980480194, "learning_rate": 2.388353116706658e-05, "loss": 0.0045, "step": 69120 }, { "grad_norm": 0.09786703437566757, "learning_rate": 2.3869432733669232e-05, "loss": 0.0051, "step": 69130 }, { "grad_norm": 0.16783110797405243, "learning_rate": 2.385533715787145e-05, "loss": 0.0069, "step": 69140 }, { "grad_norm": 0.16245169937610626, "learning_rate": 2.3841244441214673e-05, "loss": 0.0038, "step": 69150 }, { "grad_norm": 0.12702515721321106, "learning_rate": 2.3827154585240073e-05, "loss": 0.0068, "step": 69160 }, { "grad_norm": 0.1447792947292328, "learning_rate": 2.3813067591488498e-05, "loss": 0.005, "step": 69170 }, { "grad_norm": 0.09350819140672684, "learning_rate": 2.3798983461500494e-05, "loss": 0.0041, "step": 69180 }, { "grad_norm": 0.09051215648651123, "learning_rate": 2.3784902196816256e-05, "loss": 0.0048, "step": 69190 }, { "grad_norm": 0.08662062138319016, "learning_rate": 2.3770823798975716e-05, "loss": 0.0038, "step": 69200 }, { "grad_norm": 0.11440127342939377, "learning_rate": 2.3756748269518424e-05, "loss": 0.0045, "step": 69210 }, { "grad_norm": 0.11936602741479874, "learning_rate": 2.374267560998371e-05, "loss": 0.0036, "step": 69220 }, { "grad_norm": 0.08836521953344345, "learning_rate": 2.3728605821910477e-05, "loss": 0.0045, "step": 69230 }, { "grad_norm": 0.10408083349466324, "learning_rate": 2.3714538906837453e-05, "loss": 0.0052, "step": 69240 }, { "grad_norm": 0.09360120445489883, "learning_rate": 2.3700474866302913e-05, "loss": 0.0047, "step": 69250 }, { "grad_norm": 0.10338149964809418, "learning_rate": 2.3686413701844917e-05, "loss": 0.0042, "step": 69260 }, { "grad_norm": 0.19283559918403625, "learning_rate": 2.367235541500114e-05, "loss": 0.0052, "step": 69270 }, { "grad_norm": 0.07400038838386536, "learning_rate": 2.3658300007308993e-05, "loss": 0.0057, "step": 69280 }, { "grad_norm": 0.1440303474664688, "learning_rate": 2.364424748030555e-05, "loss": 0.0053, "step": 69290 }, { "grad_norm": 0.12672185897827148, "learning_rate": 2.3630197835527595e-05, "loss": 0.0058, "step": 69300 }, { "grad_norm": 0.17452408373355865, "learning_rate": 2.3616151074511546e-05, "loss": 0.0048, "step": 69310 }, { "grad_norm": 0.07753998786211014, "learning_rate": 2.3602107198793545e-05, "loss": 0.0045, "step": 69320 }, { "grad_norm": 0.08218876272439957, "learning_rate": 2.3588066209909427e-05, "loss": 0.0069, "step": 69330 }, { "grad_norm": 0.12479298561811447, "learning_rate": 2.3574028109394662e-05, "loss": 0.0042, "step": 69340 }, { "grad_norm": 0.10194674879312515, "learning_rate": 2.3559992898784454e-05, "loss": 0.0036, "step": 69350 }, { "grad_norm": 0.14066797494888306, "learning_rate": 2.354596057961367e-05, "loss": 0.0042, "step": 69360 }, { "grad_norm": 0.12304932624101639, "learning_rate": 2.353193115341688e-05, "loss": 0.0038, "step": 69370 }, { "grad_norm": 0.08620977401733398, "learning_rate": 2.3517904621728294e-05, "loss": 0.006, "step": 69380 }, { "grad_norm": 0.11066890507936478, "learning_rate": 2.3503880986081855e-05, "loss": 0.0061, "step": 69390 }, { "grad_norm": 0.13054679334163666, "learning_rate": 2.3489860248011136e-05, "loss": 0.0043, "step": 69400 }, { "grad_norm": 0.11753305047750473, "learning_rate": 2.3475842409049464e-05, "loss": 0.0049, "step": 69410 }, { "grad_norm": 0.0990568995475769, "learning_rate": 2.346182747072975e-05, "loss": 0.005, "step": 69420 }, { "grad_norm": 0.11076505482196808, "learning_rate": 2.3447815434584718e-05, "loss": 0.0038, "step": 69430 }, { "grad_norm": 0.21419385075569153, "learning_rate": 2.3433806302146656e-05, "loss": 0.0031, "step": 69440 }, { "grad_norm": 0.11260959506034851, "learning_rate": 2.3419800074947612e-05, "loss": 0.0068, "step": 69450 }, { "grad_norm": 0.10266973823308945, "learning_rate": 2.3405796754519248e-05, "loss": 0.0036, "step": 69460 }, { "grad_norm": 0.11524347215890884, "learning_rate": 2.3391796342392992e-05, "loss": 0.0046, "step": 69470 }, { "grad_norm": 0.12722745537757874, "learning_rate": 2.337779884009984e-05, "loss": 0.0042, "step": 69480 }, { "grad_norm": 0.17959344387054443, "learning_rate": 2.3363804249170624e-05, "loss": 0.0048, "step": 69490 }, { "grad_norm": 0.16255426406860352, "learning_rate": 2.3349812571135703e-05, "loss": 0.0066, "step": 69500 }, { "grad_norm": 0.10932575911283493, "learning_rate": 2.333582380752523e-05, "loss": 0.0046, "step": 69510 }, { "grad_norm": 0.08259030431509018, "learning_rate": 2.3321837959868963e-05, "loss": 0.0041, "step": 69520 }, { "grad_norm": 0.14057406783103943, "learning_rate": 2.3307855029696383e-05, "loss": 0.004, "step": 69530 }, { "grad_norm": 0.13399484753608704, "learning_rate": 2.3293875018536648e-05, "loss": 0.0047, "step": 69540 }, { "grad_norm": 0.0867665708065033, "learning_rate": 2.327989792791861e-05, "loss": 0.0035, "step": 69550 }, { "grad_norm": 0.09419603645801544, "learning_rate": 2.326592375937074e-05, "loss": 0.0052, "step": 69560 }, { "grad_norm": 0.14000892639160156, "learning_rate": 2.3251952514421248e-05, "loss": 0.0049, "step": 69570 }, { "grad_norm": 0.07198715209960938, "learning_rate": 2.323798419459804e-05, "loss": 0.0034, "step": 69580 }, { "grad_norm": 0.07682482898235321, "learning_rate": 2.3224018801428625e-05, "loss": 0.0033, "step": 69590 }, { "grad_norm": 0.09594769775867462, "learning_rate": 2.3210056336440273e-05, "loss": 0.0064, "step": 69600 }, { "grad_norm": 0.17068131268024445, "learning_rate": 2.3196096801159844e-05, "loss": 0.0059, "step": 69610 }, { "grad_norm": 0.1590355634689331, "learning_rate": 2.3182140197114012e-05, "loss": 0.0065, "step": 69620 }, { "grad_norm": 0.08274412155151367, "learning_rate": 2.3168186525828985e-05, "loss": 0.0033, "step": 69630 }, { "grad_norm": 0.08334967494010925, "learning_rate": 2.3154235788830757e-05, "loss": 0.0078, "step": 69640 }, { "grad_norm": 0.13036081194877625, "learning_rate": 2.3140287987644927e-05, "loss": 0.004, "step": 69650 }, { "grad_norm": 0.08807156980037689, "learning_rate": 2.3126343123796833e-05, "loss": 0.0041, "step": 69660 }, { "grad_norm": 0.11676820367574692, "learning_rate": 2.3112401198811418e-05, "loss": 0.004, "step": 69670 }, { "grad_norm": 0.1273014098405838, "learning_rate": 2.3098462214213418e-05, "loss": 0.0049, "step": 69680 }, { "grad_norm": 0.06576233357191086, "learning_rate": 2.308452617152712e-05, "loss": 0.004, "step": 69690 }, { "grad_norm": 0.09626993536949158, "learning_rate": 2.3070593072276603e-05, "loss": 0.0042, "step": 69700 }, { "grad_norm": 0.08781086653470993, "learning_rate": 2.3056662917985518e-05, "loss": 0.0041, "step": 69710 }, { "grad_norm": 0.1471552848815918, "learning_rate": 2.3042735710177283e-05, "loss": 0.006, "step": 69720 }, { "grad_norm": 0.10221575200557709, "learning_rate": 2.302881145037492e-05, "loss": 0.0051, "step": 69730 }, { "grad_norm": 0.13403983414173126, "learning_rate": 2.3014890140101192e-05, "loss": 0.0047, "step": 69740 }, { "grad_norm": 0.12277276813983917, "learning_rate": 2.30009717808785e-05, "loss": 0.0057, "step": 69750 }, { "grad_norm": 0.07907502353191376, "learning_rate": 2.2987056374228967e-05, "loss": 0.0041, "step": 69760 }, { "grad_norm": 0.08158616721630096, "learning_rate": 2.297314392167431e-05, "loss": 0.0057, "step": 69770 }, { "grad_norm": 0.08417416363954544, "learning_rate": 2.295923442473601e-05, "loss": 0.005, "step": 69780 }, { "grad_norm": 0.09648572653532028, "learning_rate": 2.294532788493519e-05, "loss": 0.0035, "step": 69790 }, { "grad_norm": 0.129542738199234, "learning_rate": 2.293142430379262e-05, "loss": 0.0044, "step": 69800 }, { "grad_norm": 0.11759237200021744, "learning_rate": 2.291752368282879e-05, "loss": 0.0047, "step": 69810 }, { "grad_norm": 0.09824778139591217, "learning_rate": 2.290362602356385e-05, "loss": 0.0043, "step": 69820 }, { "grad_norm": 0.10944972932338715, "learning_rate": 2.288973132751765e-05, "loss": 0.006, "step": 69830 }, { "grad_norm": 0.08969604969024658, "learning_rate": 2.287583959620965e-05, "loss": 0.0047, "step": 69840 }, { "grad_norm": 0.1506504863500595, "learning_rate": 2.2861950831159074e-05, "loss": 0.005, "step": 69850 }, { "grad_norm": 0.11400124430656433, "learning_rate": 2.284806503388472e-05, "loss": 0.0051, "step": 69860 }, { "grad_norm": 0.08610143512487411, "learning_rate": 2.283418220590518e-05, "loss": 0.0062, "step": 69870 }, { "grad_norm": 0.26371005177497864, "learning_rate": 2.282030234873862e-05, "loss": 0.0051, "step": 69880 }, { "grad_norm": 0.09826876223087311, "learning_rate": 2.2806425463902948e-05, "loss": 0.0073, "step": 69890 }, { "grad_norm": 0.10051871091127396, "learning_rate": 2.279255155291568e-05, "loss": 0.0041, "step": 69900 }, { "grad_norm": 0.13760092854499817, "learning_rate": 2.277868061729409e-05, "loss": 0.0054, "step": 69910 }, { "grad_norm": 0.1333603709936142, "learning_rate": 2.276481265855504e-05, "loss": 0.0037, "step": 69920 }, { "grad_norm": 0.14888308942317963, "learning_rate": 2.2750947678215128e-05, "loss": 0.0067, "step": 69930 }, { "grad_norm": 0.11603779345750809, "learning_rate": 2.2737085677790616e-05, "loss": 0.0045, "step": 69940 }, { "grad_norm": 0.1161251813173294, "learning_rate": 2.2723226658797443e-05, "loss": 0.0045, "step": 69950 }, { "grad_norm": 0.16338187456130981, "learning_rate": 2.270937062275117e-05, "loss": 0.0046, "step": 69960 }, { "grad_norm": 0.12057612836360931, "learning_rate": 2.2695517571167113e-05, "loss": 0.004, "step": 69970 }, { "grad_norm": 0.14866909384727478, "learning_rate": 2.268166750556019e-05, "loss": 0.0041, "step": 69980 }, { "grad_norm": 0.12653633952140808, "learning_rate": 2.266782042744503e-05, "loss": 0.0056, "step": 69990 }, { "grad_norm": 0.14464950561523438, "learning_rate": 2.2653976338335935e-05, "loss": 0.0057, "step": 70000 }, { "grad_norm": 0.10014479607343674, "learning_rate": 2.2640135239746895e-05, "loss": 0.0041, "step": 70010 }, { "grad_norm": 0.09918352961540222, "learning_rate": 2.2626297133191515e-05, "loss": 0.0039, "step": 70020 }, { "grad_norm": 0.09926871210336685, "learning_rate": 2.261246202018312e-05, "loss": 0.0066, "step": 70030 }, { "grad_norm": 0.08681771904230118, "learning_rate": 2.2598629902234718e-05, "loss": 0.0049, "step": 70040 }, { "grad_norm": 0.0856526717543602, "learning_rate": 2.258480078085894e-05, "loss": 0.0052, "step": 70050 }, { "grad_norm": 0.12876728177070618, "learning_rate": 2.2570974657568126e-05, "loss": 0.0036, "step": 70060 }, { "grad_norm": 0.12649698555469513, "learning_rate": 2.2557151533874286e-05, "loss": 0.0047, "step": 70070 }, { "grad_norm": 0.1228742003440857, "learning_rate": 2.2543331411289116e-05, "loss": 0.0044, "step": 70080 }, { "grad_norm": 0.12300373613834381, "learning_rate": 2.252951429132392e-05, "loss": 0.0056, "step": 70090 }, { "grad_norm": 0.23937322199344635, "learning_rate": 2.2515700175489762e-05, "loss": 0.0049, "step": 70100 }, { "grad_norm": 0.12692470848560333, "learning_rate": 2.2501889065297288e-05, "loss": 0.0053, "step": 70110 }, { "grad_norm": 0.16497254371643066, "learning_rate": 2.2488080962256887e-05, "loss": 0.0054, "step": 70120 }, { "grad_norm": 0.09388063848018646, "learning_rate": 2.2474275867878585e-05, "loss": 0.005, "step": 70130 }, { "grad_norm": 0.10028354823589325, "learning_rate": 2.246047378367211e-05, "loss": 0.0041, "step": 70140 }, { "grad_norm": 0.10582606494426727, "learning_rate": 2.2446674711146798e-05, "loss": 0.0039, "step": 70150 }, { "grad_norm": 0.11163730919361115, "learning_rate": 2.2432878651811734e-05, "loss": 0.0061, "step": 70160 }, { "grad_norm": 0.08237574994564056, "learning_rate": 2.2419085607175594e-05, "loss": 0.0035, "step": 70170 }, { "grad_norm": 0.0907307043671608, "learning_rate": 2.240529557874678e-05, "loss": 0.006, "step": 70180 }, { "grad_norm": 0.1413305401802063, "learning_rate": 2.2391508568033358e-05, "loss": 0.0065, "step": 70190 }, { "grad_norm": 0.12397753447294235, "learning_rate": 2.237772457654307e-05, "loss": 0.0044, "step": 70200 }, { "grad_norm": 0.12141814827919006, "learning_rate": 2.236394360578327e-05, "loss": 0.0061, "step": 70210 }, { "grad_norm": 0.12682150304317474, "learning_rate": 2.2350165657261073e-05, "loss": 0.0062, "step": 70220 }, { "grad_norm": 0.14048388600349426, "learning_rate": 2.2336390732483164e-05, "loss": 0.0066, "step": 70230 }, { "grad_norm": 0.14524777233600616, "learning_rate": 2.2322618832955976e-05, "loss": 0.0067, "step": 70240 }, { "grad_norm": 0.11020954698324203, "learning_rate": 2.2308849960185578e-05, "loss": 0.0053, "step": 70250 }, { "grad_norm": 0.11433427035808563, "learning_rate": 2.229508411567774e-05, "loss": 0.0052, "step": 70260 }, { "grad_norm": 0.07974661886692047, "learning_rate": 2.2281321300937823e-05, "loss": 0.0039, "step": 70270 }, { "grad_norm": 0.13461492955684662, "learning_rate": 2.2267561517470937e-05, "loss": 0.0053, "step": 70280 }, { "grad_norm": 0.10217311233282089, "learning_rate": 2.2253804766781845e-05, "loss": 0.0041, "step": 70290 }, { "grad_norm": 0.10897638648748398, "learning_rate": 2.2240051050374933e-05, "loss": 0.0053, "step": 70300 }, { "grad_norm": 0.08357168734073639, "learning_rate": 2.22263003697543e-05, "loss": 0.0043, "step": 70310 }, { "grad_norm": 0.07874852418899536, "learning_rate": 2.2212552726423702e-05, "loss": 0.005, "step": 70320 }, { "grad_norm": 0.09996170550584793, "learning_rate": 2.2198808121886578e-05, "loss": 0.005, "step": 70330 }, { "grad_norm": 0.10335786640644073, "learning_rate": 2.218506655764598e-05, "loss": 0.0049, "step": 70340 }, { "grad_norm": 0.12526333332061768, "learning_rate": 2.2171328035204704e-05, "loss": 0.0044, "step": 70350 }, { "grad_norm": 0.0822681412100792, "learning_rate": 2.2157592556065133e-05, "loss": 0.0045, "step": 70360 }, { "grad_norm": 0.07759920507669449, "learning_rate": 2.2143860121729382e-05, "loss": 0.005, "step": 70370 }, { "grad_norm": 0.1432054042816162, "learning_rate": 2.2130130733699206e-05, "loss": 0.0052, "step": 70380 }, { "grad_norm": 0.0991683304309845, "learning_rate": 2.2116404393476052e-05, "loss": 0.0043, "step": 70390 }, { "grad_norm": 0.13635951280593872, "learning_rate": 2.2102681102560967e-05, "loss": 0.0046, "step": 70400 }, { "grad_norm": 0.11683199554681778, "learning_rate": 2.208896086245476e-05, "loss": 0.005, "step": 70410 }, { "grad_norm": 0.11995876580476761, "learning_rate": 2.2075243674657808e-05, "loss": 0.0053, "step": 70420 }, { "grad_norm": 0.22079716622829437, "learning_rate": 2.206152954067024e-05, "loss": 0.0062, "step": 70430 }, { "grad_norm": 0.08617820590734482, "learning_rate": 2.2047818461991758e-05, "loss": 0.0035, "step": 70440 }, { "grad_norm": 0.07737227529287338, "learning_rate": 2.2034110440121864e-05, "loss": 0.0064, "step": 70450 }, { "grad_norm": 0.08193743973970413, "learning_rate": 2.2020405476559585e-05, "loss": 0.004, "step": 70460 }, { "grad_norm": 0.08879584819078445, "learning_rate": 2.2006703572803722e-05, "loss": 0.0043, "step": 70470 }, { "grad_norm": 0.12311224639415741, "learning_rate": 2.1993004730352646e-05, "loss": 0.0045, "step": 70480 }, { "grad_norm": 0.1708538681268692, "learning_rate": 2.1979308950704464e-05, "loss": 0.0042, "step": 70490 }, { "grad_norm": 0.19615483283996582, "learning_rate": 2.1965616235356923e-05, "loss": 0.0039, "step": 70500 }, { "grad_norm": 0.12435417622327805, "learning_rate": 2.1951926585807465e-05, "loss": 0.0041, "step": 70510 }, { "grad_norm": 0.1532076895236969, "learning_rate": 2.193824000355313e-05, "loss": 0.0037, "step": 70520 }, { "grad_norm": 0.1035144105553627, "learning_rate": 2.192455649009067e-05, "loss": 0.0068, "step": 70530 }, { "grad_norm": 0.08094708621501923, "learning_rate": 2.1910876046916523e-05, "loss": 0.004, "step": 70540 }, { "grad_norm": 0.14299796521663666, "learning_rate": 2.1897198675526725e-05, "loss": 0.0037, "step": 70550 }, { "grad_norm": 0.14437010884284973, "learning_rate": 2.1883524377417024e-05, "loss": 0.0033, "step": 70560 }, { "grad_norm": 0.08480873703956604, "learning_rate": 2.1869853154082825e-05, "loss": 0.0038, "step": 70570 }, { "grad_norm": 0.08813030272722244, "learning_rate": 2.1856185007019215e-05, "loss": 0.0058, "step": 70580 }, { "grad_norm": 0.12253367900848389, "learning_rate": 2.1842519937720874e-05, "loss": 0.0065, "step": 70590 }, { "grad_norm": 0.1471424549818039, "learning_rate": 2.182885794768224e-05, "loss": 0.0041, "step": 70600 }, { "grad_norm": 0.15024620294570923, "learning_rate": 2.1815199038397326e-05, "loss": 0.0044, "step": 70610 }, { "grad_norm": 0.12650395929813385, "learning_rate": 2.1801543211359893e-05, "loss": 0.0048, "step": 70620 }, { "grad_norm": 0.08892273902893066, "learning_rate": 2.178789046806326e-05, "loss": 0.005, "step": 70630 }, { "grad_norm": 0.14011250436306, "learning_rate": 2.1774240810000547e-05, "loss": 0.004, "step": 70640 }, { "grad_norm": 0.1079520434141159, "learning_rate": 2.17605942386644e-05, "loss": 0.0041, "step": 70650 }, { "grad_norm": 0.10426183044910431, "learning_rate": 2.174695075554723e-05, "loss": 0.0046, "step": 70660 }, { "grad_norm": 0.08969978243112564, "learning_rate": 2.173331036214103e-05, "loss": 0.0034, "step": 70670 }, { "grad_norm": 0.10847999155521393, "learning_rate": 2.171967305993753e-05, "loss": 0.0048, "step": 70680 }, { "grad_norm": 0.09538548439741135, "learning_rate": 2.1706038850428024e-05, "loss": 0.0043, "step": 70690 }, { "grad_norm": 0.10792329907417297, "learning_rate": 2.169240773510361e-05, "loss": 0.0053, "step": 70700 }, { "grad_norm": 0.10557807981967926, "learning_rate": 2.1678779715454906e-05, "loss": 0.0063, "step": 70710 }, { "grad_norm": 0.13146735727787018, "learning_rate": 2.166515479297229e-05, "loss": 0.0071, "step": 70720 }, { "grad_norm": 0.16646170616149902, "learning_rate": 2.1651532969145732e-05, "loss": 0.0043, "step": 70730 }, { "grad_norm": 0.2131461203098297, "learning_rate": 2.1637914245464903e-05, "loss": 0.0044, "step": 70740 }, { "grad_norm": 0.11770832538604736, "learning_rate": 2.1624298623419133e-05, "loss": 0.0054, "step": 70750 }, { "grad_norm": 0.07460350543260574, "learning_rate": 2.1610686104497413e-05, "loss": 0.0048, "step": 70760 }, { "grad_norm": 0.11424284428358078, "learning_rate": 2.1597076690188366e-05, "loss": 0.0051, "step": 70770 }, { "grad_norm": 0.14679867029190063, "learning_rate": 2.1583470381980315e-05, "loss": 0.0046, "step": 70780 }, { "grad_norm": 0.12360113114118576, "learning_rate": 2.156986718136123e-05, "loss": 0.0036, "step": 70790 }, { "grad_norm": 0.1444239318370819, "learning_rate": 2.155626708981871e-05, "loss": 0.0057, "step": 70800 }, { "grad_norm": 0.1720474362373352, "learning_rate": 2.1542670108840075e-05, "loss": 0.0037, "step": 70810 }, { "grad_norm": 0.13993056118488312, "learning_rate": 2.152907623991221e-05, "loss": 0.0061, "step": 70820 }, { "grad_norm": 0.09361322969198227, "learning_rate": 2.1515485484521802e-05, "loss": 0.0071, "step": 70830 }, { "grad_norm": 0.08892687410116196, "learning_rate": 2.150189784415506e-05, "loss": 0.0056, "step": 70840 }, { "grad_norm": 0.13789516687393188, "learning_rate": 2.148831332029794e-05, "loss": 0.0038, "step": 70850 }, { "grad_norm": 0.0923498272895813, "learning_rate": 2.147473191443599e-05, "loss": 0.0037, "step": 70860 }, { "grad_norm": 0.07298527657985687, "learning_rate": 2.1461153628054492e-05, "loss": 0.0033, "step": 70870 }, { "grad_norm": 0.12725065648555756, "learning_rate": 2.1447578462638284e-05, "loss": 0.004, "step": 70880 }, { "grad_norm": 0.07627851516008377, "learning_rate": 2.143400641967201e-05, "loss": 0.0046, "step": 70890 }, { "grad_norm": 0.2393353283405304, "learning_rate": 2.1420437500639823e-05, "loss": 0.0057, "step": 70900 }, { "grad_norm": 0.14865699410438538, "learning_rate": 2.1406871707025646e-05, "loss": 0.0047, "step": 70910 }, { "grad_norm": 0.17918920516967773, "learning_rate": 2.1393309040312963e-05, "loss": 0.0036, "step": 70920 }, { "grad_norm": 0.11245867609977722, "learning_rate": 2.137974950198502e-05, "loss": 0.0065, "step": 70930 }, { "grad_norm": 0.13896220922470093, "learning_rate": 2.1366193093524617e-05, "loss": 0.0062, "step": 70940 }, { "grad_norm": 0.15609636902809143, "learning_rate": 2.1352639816414288e-05, "loss": 0.0042, "step": 70950 }, { "grad_norm": 0.07869519293308258, "learning_rate": 2.1339089672136202e-05, "loss": 0.0037, "step": 70960 }, { "grad_norm": 0.15541419386863708, "learning_rate": 2.1325542662172193e-05, "loss": 0.0044, "step": 70970 }, { "grad_norm": 0.10305290669202805, "learning_rate": 2.1311998788003712e-05, "loss": 0.0057, "step": 70980 }, { "grad_norm": 0.16208024322986603, "learning_rate": 2.1298458051111907e-05, "loss": 0.0059, "step": 70990 }, { "grad_norm": 0.11767373234033585, "learning_rate": 2.1284920452977602e-05, "loss": 0.0052, "step": 71000 }, { "grad_norm": 0.14710727334022522, "learning_rate": 2.12713859950812e-05, "loss": 0.004, "step": 71010 }, { "grad_norm": 0.12437896430492401, "learning_rate": 2.1257854678902832e-05, "loss": 0.0058, "step": 71020 }, { "grad_norm": 0.1403089314699173, "learning_rate": 2.124432650592227e-05, "loss": 0.0053, "step": 71030 }, { "grad_norm": 0.09898845106363297, "learning_rate": 2.1230801477618937e-05, "loss": 0.0041, "step": 71040 }, { "grad_norm": 0.11550861597061157, "learning_rate": 2.121727959547189e-05, "loss": 0.0049, "step": 71050 }, { "grad_norm": 0.08640869706869125, "learning_rate": 2.1203760860959886e-05, "loss": 0.0045, "step": 71060 }, { "grad_norm": 0.09288482367992401, "learning_rate": 2.119024527556127e-05, "loss": 0.0048, "step": 71070 }, { "grad_norm": 0.09621001034975052, "learning_rate": 2.117673284075415e-05, "loss": 0.006, "step": 71080 }, { "grad_norm": 0.13293232023715973, "learning_rate": 2.1163223558016165e-05, "loss": 0.0092, "step": 71090 }, { "grad_norm": 0.1271759569644928, "learning_rate": 2.1149717428824722e-05, "loss": 0.005, "step": 71100 }, { "grad_norm": 0.12451375275850296, "learning_rate": 2.1136214454656784e-05, "loss": 0.0052, "step": 71110 }, { "grad_norm": 0.08389939367771149, "learning_rate": 2.1122714636989056e-05, "loss": 0.0046, "step": 71120 }, { "grad_norm": 0.08729615062475204, "learning_rate": 2.1109217977297823e-05, "loss": 0.0047, "step": 71130 }, { "grad_norm": 0.10931576043367386, "learning_rate": 2.1095724477059077e-05, "loss": 0.0047, "step": 71140 }, { "grad_norm": 0.09142220765352249, "learning_rate": 2.1082234137748434e-05, "loss": 0.0057, "step": 71150 }, { "grad_norm": 0.1046425849199295, "learning_rate": 2.106874696084122e-05, "loss": 0.0051, "step": 71160 }, { "grad_norm": 0.09507410228252411, "learning_rate": 2.1055262947812315e-05, "loss": 0.0039, "step": 71170 }, { "grad_norm": 0.12814980745315552, "learning_rate": 2.104178210013636e-05, "loss": 0.0039, "step": 71180 }, { "grad_norm": 0.11173806339502335, "learning_rate": 2.1028304419287554e-05, "loss": 0.0056, "step": 71190 }, { "grad_norm": 0.12525556981563568, "learning_rate": 2.101482990673982e-05, "loss": 0.0048, "step": 71200 }, { "grad_norm": 0.1121157854795456, "learning_rate": 2.1001358563966707e-05, "loss": 0.0041, "step": 71210 }, { "grad_norm": 0.11458763480186462, "learning_rate": 2.098789039244144e-05, "loss": 0.0037, "step": 71220 }, { "grad_norm": 0.18923155963420868, "learning_rate": 2.097442539363685e-05, "loss": 0.0054, "step": 71230 }, { "grad_norm": 0.1294352412223816, "learning_rate": 2.0960963569025448e-05, "loss": 0.004, "step": 71240 }, { "grad_norm": 0.10135049372911453, "learning_rate": 2.0947504920079435e-05, "loss": 0.005, "step": 71250 }, { "grad_norm": 0.2243369221687317, "learning_rate": 2.093404944827059e-05, "loss": 0.0052, "step": 71260 }, { "grad_norm": 0.13158372044563293, "learning_rate": 2.0920597155070394e-05, "loss": 0.004, "step": 71270 }, { "grad_norm": 0.34886038303375244, "learning_rate": 2.090714804194997e-05, "loss": 0.0039, "step": 71280 }, { "grad_norm": 0.13449087738990784, "learning_rate": 2.0893702110380114e-05, "loss": 0.0046, "step": 71290 }, { "grad_norm": 0.1333703249692917, "learning_rate": 2.0880259361831216e-05, "loss": 0.0044, "step": 71300 }, { "grad_norm": 0.10427599400281906, "learning_rate": 2.0866819797773397e-05, "loss": 0.0034, "step": 71310 }, { "grad_norm": 0.10042718797922134, "learning_rate": 2.085338341967634e-05, "loss": 0.0039, "step": 71320 }, { "grad_norm": 0.11541467905044556, "learning_rate": 2.083995022900946e-05, "loss": 0.0054, "step": 71330 }, { "grad_norm": 0.12922509014606476, "learning_rate": 2.0826520227241768e-05, "loss": 0.0038, "step": 71340 }, { "grad_norm": 0.1502513438463211, "learning_rate": 2.081309341584199e-05, "loss": 0.0055, "step": 71350 }, { "grad_norm": 0.11551980674266815, "learning_rate": 2.0799669796278413e-05, "loss": 0.0062, "step": 71360 }, { "grad_norm": 0.05899566039443016, "learning_rate": 2.078624937001906e-05, "loss": 0.0035, "step": 71370 }, { "grad_norm": 0.13605765998363495, "learning_rate": 2.0772832138531544e-05, "loss": 0.0047, "step": 71380 }, { "grad_norm": 0.11976085603237152, "learning_rate": 2.0759418103283158e-05, "loss": 0.0046, "step": 71390 }, { "grad_norm": 0.10767682641744614, "learning_rate": 2.0746007265740846e-05, "loss": 0.0053, "step": 71400 }, { "grad_norm": 0.11493312567472458, "learning_rate": 2.0732599627371215e-05, "loss": 0.0055, "step": 71410 }, { "grad_norm": 0.13176263868808746, "learning_rate": 2.0719195189640468e-05, "loss": 0.0044, "step": 71420 }, { "grad_norm": 0.11458224803209305, "learning_rate": 2.0705793954014528e-05, "loss": 0.0043, "step": 71430 }, { "grad_norm": 0.15535837411880493, "learning_rate": 2.06923959219589e-05, "loss": 0.0048, "step": 71440 }, { "grad_norm": 0.10090438276529312, "learning_rate": 2.0679001094938793e-05, "loss": 0.0044, "step": 71450 }, { "grad_norm": 0.07222702354192734, "learning_rate": 2.066560947441904e-05, "loss": 0.0042, "step": 71460 }, { "grad_norm": 0.07945044338703156, "learning_rate": 2.0652221061864152e-05, "loss": 0.0043, "step": 71470 }, { "grad_norm": 0.11372816562652588, "learning_rate": 2.0638835858738226e-05, "loss": 0.0052, "step": 71480 }, { "grad_norm": 0.12784713506698608, "learning_rate": 2.0625453866505068e-05, "loss": 0.0051, "step": 71490 }, { "grad_norm": 0.12073580175638199, "learning_rate": 2.0612075086628123e-05, "loss": 0.0043, "step": 71500 }, { "grad_norm": 0.09698071330785751, "learning_rate": 2.059869952057045e-05, "loss": 0.0049, "step": 71510 }, { "grad_norm": 0.11612558364868164, "learning_rate": 2.0585327169794793e-05, "loss": 0.0046, "step": 71520 }, { "grad_norm": 0.14235416054725647, "learning_rate": 2.0571958035763534e-05, "loss": 0.0036, "step": 71530 }, { "grad_norm": 0.2611130177974701, "learning_rate": 2.0558592119938715e-05, "loss": 0.0059, "step": 71540 }, { "grad_norm": 0.1499037891626358, "learning_rate": 2.054522942378198e-05, "loss": 0.004, "step": 71550 }, { "grad_norm": 0.1232827678322792, "learning_rate": 2.053186994875469e-05, "loss": 0.0037, "step": 71560 }, { "grad_norm": 0.18283523619174957, "learning_rate": 2.0518513696317775e-05, "loss": 0.0081, "step": 71570 }, { "grad_norm": 0.13643550872802734, "learning_rate": 2.050516066793188e-05, "loss": 0.0039, "step": 71580 }, { "grad_norm": 0.1987657994031906, "learning_rate": 2.0491810865057276e-05, "loss": 0.0051, "step": 71590 }, { "grad_norm": 0.10880408436059952, "learning_rate": 2.0478464289153882e-05, "loss": 0.0042, "step": 71600 }, { "grad_norm": 0.10246706008911133, "learning_rate": 2.0465120941681232e-05, "loss": 0.0048, "step": 71610 }, { "grad_norm": 0.17797507345676422, "learning_rate": 2.0451780824098566e-05, "loss": 0.0051, "step": 71620 }, { "grad_norm": 0.1829768419265747, "learning_rate": 2.043844393786471e-05, "loss": 0.0042, "step": 71630 }, { "grad_norm": 0.10572196543216705, "learning_rate": 2.0425110284438193e-05, "loss": 0.0044, "step": 71640 }, { "grad_norm": 0.15199345350265503, "learning_rate": 2.041177986527712e-05, "loss": 0.0041, "step": 71650 }, { "grad_norm": 0.06863581389188766, "learning_rate": 2.0398452681839347e-05, "loss": 0.0058, "step": 71660 }, { "grad_norm": 0.1427212655544281, "learning_rate": 2.0385128735582272e-05, "loss": 0.0036, "step": 71670 }, { "grad_norm": 0.11088032275438309, "learning_rate": 2.0371808027963003e-05, "loss": 0.0048, "step": 71680 }, { "grad_norm": 0.15821535885334015, "learning_rate": 2.035849056043825e-05, "loss": 0.0051, "step": 71690 }, { "grad_norm": 0.12440570443868637, "learning_rate": 2.03451763344644e-05, "loss": 0.0056, "step": 71700 }, { "grad_norm": 0.11024143546819687, "learning_rate": 2.0331865351497482e-05, "loss": 0.0033, "step": 71710 }, { "grad_norm": 0.09207265824079514, "learning_rate": 2.0318557612993185e-05, "loss": 0.0043, "step": 71720 }, { "grad_norm": 0.1511860191822052, "learning_rate": 2.0305253120406782e-05, "loss": 0.0046, "step": 71730 }, { "grad_norm": 0.1774395853281021, "learning_rate": 2.029195187519325e-05, "loss": 0.0055, "step": 71740 }, { "grad_norm": 0.12327598035335541, "learning_rate": 2.0278653878807217e-05, "loss": 0.0043, "step": 71750 }, { "grad_norm": 0.09748660773038864, "learning_rate": 2.0265359132702892e-05, "loss": 0.004, "step": 71760 }, { "grad_norm": 0.17610254883766174, "learning_rate": 2.0252067638334182e-05, "loss": 0.0052, "step": 71770 }, { "grad_norm": 0.08231455087661743, "learning_rate": 2.023877939715464e-05, "loss": 0.004, "step": 71780 }, { "grad_norm": 0.13272352516651154, "learning_rate": 2.0225494410617448e-05, "loss": 0.0048, "step": 71790 }, { "grad_norm": 0.06731178611516953, "learning_rate": 2.02122126801754e-05, "loss": 0.0049, "step": 71800 }, { "grad_norm": 0.10826251655817032, "learning_rate": 2.0198934207281018e-05, "loss": 0.0036, "step": 71810 }, { "grad_norm": 0.1508534550666809, "learning_rate": 2.0185658993386357e-05, "loss": 0.005, "step": 71820 }, { "grad_norm": 0.10182453691959381, "learning_rate": 2.0172387039943226e-05, "loss": 0.0061, "step": 71830 }, { "grad_norm": 0.11931747943162918, "learning_rate": 2.0159118348402968e-05, "loss": 0.0052, "step": 71840 }, { "grad_norm": 0.15204615890979767, "learning_rate": 2.0145852920216697e-05, "loss": 0.0055, "step": 71850 }, { "grad_norm": 0.09423400461673737, "learning_rate": 2.0132590756835045e-05, "loss": 0.0045, "step": 71860 }, { "grad_norm": 0.11212503910064697, "learning_rate": 2.0119331859708386e-05, "loss": 0.0055, "step": 71870 }, { "grad_norm": 0.2251509428024292, "learning_rate": 2.0106076230286654e-05, "loss": 0.0055, "step": 71880 }, { "grad_norm": 0.12732428312301636, "learning_rate": 2.0092823870019488e-05, "loss": 0.0061, "step": 71890 }, { "grad_norm": 0.11321311444044113, "learning_rate": 2.0079574780356115e-05, "loss": 0.0047, "step": 71900 }, { "grad_norm": 0.10733946412801743, "learning_rate": 2.0066328962745486e-05, "loss": 0.0045, "step": 71910 }, { "grad_norm": 0.13477280735969543, "learning_rate": 2.00530864186361e-05, "loss": 0.0051, "step": 71920 }, { "grad_norm": 0.2803754508495331, "learning_rate": 2.003984714947618e-05, "loss": 0.0041, "step": 71930 }, { "grad_norm": 0.08973104506731033, "learning_rate": 2.0026611156713515e-05, "loss": 0.0037, "step": 71940 }, { "grad_norm": 0.15738213062286377, "learning_rate": 2.0013378441795584e-05, "loss": 0.0053, "step": 71950 }, { "grad_norm": 0.29731038212776184, "learning_rate": 2.0000149006169506e-05, "loss": 0.0084, "step": 71960 }, { "grad_norm": 0.2272898554801941, "learning_rate": 1.9986922851282048e-05, "loss": 0.0056, "step": 71970 }, { "grad_norm": 0.12519893050193787, "learning_rate": 1.997369997857956e-05, "loss": 0.0056, "step": 71980 }, { "grad_norm": 0.11014207452535629, "learning_rate": 1.9960480389508103e-05, "loss": 0.0042, "step": 71990 }, { "grad_norm": 0.09913090616464615, "learning_rate": 1.9947264085513363e-05, "loss": 0.0043, "step": 72000 }, { "grad_norm": 0.0823809951543808, "learning_rate": 1.993405106804062e-05, "loss": 0.0042, "step": 72010 }, { "grad_norm": 0.0831952914595604, "learning_rate": 1.9920841338534867e-05, "loss": 0.0064, "step": 72020 }, { "grad_norm": 0.11702752858400345, "learning_rate": 1.9907634898440646e-05, "loss": 0.0039, "step": 72030 }, { "grad_norm": 0.07337554544210434, "learning_rate": 1.989443174920227e-05, "loss": 0.0062, "step": 72040 }, { "grad_norm": 0.11733593791723251, "learning_rate": 1.988123189226355e-05, "loss": 0.0057, "step": 72050 }, { "grad_norm": 0.11227550357580185, "learning_rate": 1.986803532906805e-05, "loss": 0.0035, "step": 72060 }, { "grad_norm": 0.11824999004602432, "learning_rate": 1.985484206105889e-05, "loss": 0.0047, "step": 72070 }, { "grad_norm": 0.12350758165121078, "learning_rate": 1.9841652089678887e-05, "loss": 0.0044, "step": 72080 }, { "grad_norm": 0.26364922523498535, "learning_rate": 1.9828465416370435e-05, "loss": 0.0052, "step": 72090 }, { "grad_norm": 0.06666755676269531, "learning_rate": 1.9815282042575685e-05, "loss": 0.0043, "step": 72100 }, { "grad_norm": 0.1434168517589569, "learning_rate": 1.9802101969736285e-05, "loss": 0.0064, "step": 72110 }, { "grad_norm": 0.13419942557811737, "learning_rate": 1.9788925199293633e-05, "loss": 0.0038, "step": 72120 }, { "grad_norm": 0.11464712023735046, "learning_rate": 1.977575173268868e-05, "loss": 0.004, "step": 72130 }, { "grad_norm": 0.17320148646831512, "learning_rate": 1.9762581571362094e-05, "loss": 0.006, "step": 72140 }, { "grad_norm": 0.3110577464103699, "learning_rate": 1.9749414716754112e-05, "loss": 0.0046, "step": 72150 }, { "grad_norm": 0.07774689793586731, "learning_rate": 1.973625117030465e-05, "loss": 0.0043, "step": 72160 }, { "grad_norm": 0.1445581614971161, "learning_rate": 1.972309093345326e-05, "loss": 0.006, "step": 72170 }, { "grad_norm": 0.07476057857275009, "learning_rate": 1.970993400763915e-05, "loss": 0.0044, "step": 72180 }, { "grad_norm": 0.14021900296211243, "learning_rate": 1.969678039430109e-05, "loss": 0.0057, "step": 72190 }, { "grad_norm": 0.0813734382390976, "learning_rate": 1.9683630094877564e-05, "loss": 0.0036, "step": 72200 }, { "grad_norm": 0.10790058970451355, "learning_rate": 1.9670483110806696e-05, "loss": 0.0041, "step": 72210 }, { "grad_norm": 0.0970708429813385, "learning_rate": 1.9657339443526172e-05, "loss": 0.0045, "step": 72220 }, { "grad_norm": 0.06275716423988342, "learning_rate": 1.964419909447339e-05, "loss": 0.0046, "step": 72230 }, { "grad_norm": 0.10240230709314346, "learning_rate": 1.963106206508536e-05, "loss": 0.0042, "step": 72240 }, { "grad_norm": 0.09319502860307693, "learning_rate": 1.9617928356798738e-05, "loss": 0.0035, "step": 72250 }, { "grad_norm": 0.08667860925197601, "learning_rate": 1.960479797104977e-05, "loss": 0.0057, "step": 72260 }, { "grad_norm": 0.08245627582073212, "learning_rate": 1.9591670909274418e-05, "loss": 0.0042, "step": 72270 }, { "grad_norm": 0.15047915279865265, "learning_rate": 1.9578547172908184e-05, "loss": 0.0041, "step": 72280 }, { "grad_norm": 0.2610003352165222, "learning_rate": 1.9565426763386325e-05, "loss": 0.0051, "step": 72290 }, { "grad_norm": 0.13844847679138184, "learning_rate": 1.9552309682143622e-05, "loss": 0.0031, "step": 72300 }, { "grad_norm": 0.07093483209609985, "learning_rate": 1.9539195930614574e-05, "loss": 0.0039, "step": 72310 }, { "grad_norm": 0.1374605894088745, "learning_rate": 1.952608551023324e-05, "loss": 0.0072, "step": 72320 }, { "grad_norm": 0.1061137318611145, "learning_rate": 1.9512978422433397e-05, "loss": 0.0058, "step": 72330 }, { "grad_norm": 0.1283036321401596, "learning_rate": 1.9499874668648376e-05, "loss": 0.0047, "step": 72340 }, { "grad_norm": 0.08396779745817184, "learning_rate": 1.9486774250311213e-05, "loss": 0.0032, "step": 72350 }, { "grad_norm": 0.1701124757528305, "learning_rate": 1.9473677168854537e-05, "loss": 0.0045, "step": 72360 }, { "grad_norm": 0.12212231010198593, "learning_rate": 1.946058342571065e-05, "loss": 0.0043, "step": 72370 }, { "grad_norm": 0.07652111351490021, "learning_rate": 1.9447493022311424e-05, "loss": 0.0048, "step": 72380 }, { "grad_norm": 0.08574585616588593, "learning_rate": 1.9434405960088442e-05, "loss": 0.003, "step": 72390 }, { "grad_norm": 0.09364054352045059, "learning_rate": 1.9421322240472857e-05, "loss": 0.0063, "step": 72400 }, { "grad_norm": 0.10536283999681473, "learning_rate": 1.9408241864895494e-05, "loss": 0.003, "step": 72410 }, { "grad_norm": 0.12389175593852997, "learning_rate": 1.939516483478681e-05, "loss": 0.0047, "step": 72420 }, { "grad_norm": 0.08938847482204437, "learning_rate": 1.9382091151576902e-05, "loss": 0.0055, "step": 72430 }, { "grad_norm": 0.16114689409732819, "learning_rate": 1.9369020816695454e-05, "loss": 0.005, "step": 72440 }, { "grad_norm": 0.10240277647972107, "learning_rate": 1.935595383157184e-05, "loss": 0.0051, "step": 72450 }, { "grad_norm": 0.1478051245212555, "learning_rate": 1.9342890197635057e-05, "loss": 0.0058, "step": 72460 }, { "grad_norm": 0.12152387201786041, "learning_rate": 1.9329829916313685e-05, "loss": 0.007, "step": 72470 }, { "grad_norm": 0.12496862560510635, "learning_rate": 1.9316772989036013e-05, "loss": 0.0039, "step": 72480 }, { "grad_norm": 0.10034544765949249, "learning_rate": 1.930371941722991e-05, "loss": 0.0045, "step": 72490 }, { "grad_norm": 0.06387636810541153, "learning_rate": 1.929066920232292e-05, "loss": 0.0046, "step": 72500 }, { "grad_norm": 0.049968089908361435, "learning_rate": 1.9277622345742152e-05, "loss": 0.0031, "step": 72510 }, { "grad_norm": 0.08872903138399124, "learning_rate": 1.926457884891444e-05, "loss": 0.0046, "step": 72520 }, { "grad_norm": 0.11330396682024002, "learning_rate": 1.9251538713266147e-05, "loss": 0.0048, "step": 72530 }, { "grad_norm": 0.1133635863661766, "learning_rate": 1.9238501940223358e-05, "loss": 0.0043, "step": 72540 }, { "grad_norm": 0.10396547615528107, "learning_rate": 1.922546853121175e-05, "loss": 0.0042, "step": 72550 }, { "grad_norm": 0.09004805237054825, "learning_rate": 1.9212438487656647e-05, "loss": 0.0054, "step": 72560 }, { "grad_norm": 0.12216287851333618, "learning_rate": 1.9199411810982965e-05, "loss": 0.0043, "step": 72570 }, { "grad_norm": 0.18261092901229858, "learning_rate": 1.918638850261532e-05, "loss": 0.0053, "step": 72580 }, { "grad_norm": 0.1570814847946167, "learning_rate": 1.9173368563977885e-05, "loss": 0.0043, "step": 72590 }, { "grad_norm": 0.09890995174646378, "learning_rate": 1.9160351996494513e-05, "loss": 0.0053, "step": 72600 }, { "grad_norm": 0.07307405769824982, "learning_rate": 1.9147338801588677e-05, "loss": 0.0049, "step": 72610 }, { "grad_norm": 0.16484495997428894, "learning_rate": 1.913432898068351e-05, "loss": 0.004, "step": 72620 }, { "grad_norm": 0.08523988723754883, "learning_rate": 1.912132253520169e-05, "loss": 0.0064, "step": 72630 }, { "grad_norm": 0.13300438225269318, "learning_rate": 1.910831946656564e-05, "loss": 0.0045, "step": 72640 }, { "grad_norm": 0.12085719406604767, "learning_rate": 1.909531977619731e-05, "loss": 0.0047, "step": 72650 }, { "grad_norm": 0.10540743172168732, "learning_rate": 1.908232346551834e-05, "loss": 0.0041, "step": 72660 }, { "grad_norm": 0.11622938513755798, "learning_rate": 1.9069330535949993e-05, "loss": 0.0034, "step": 72670 }, { "grad_norm": 0.08848654478788376, "learning_rate": 1.9056340988913167e-05, "loss": 0.005, "step": 72680 }, { "grad_norm": 0.08231404423713684, "learning_rate": 1.9043354825828357e-05, "loss": 0.005, "step": 72690 }, { "grad_norm": 0.10497330129146576, "learning_rate": 1.9030372048115724e-05, "loss": 0.0033, "step": 72700 }, { "grad_norm": 0.18979187309741974, "learning_rate": 1.9017392657195056e-05, "loss": 0.0036, "step": 72710 }, { "grad_norm": 0.12009826302528381, "learning_rate": 1.9004416654485724e-05, "loss": 0.0037, "step": 72720 }, { "grad_norm": 0.1624445766210556, "learning_rate": 1.8991444041406785e-05, "loss": 0.0039, "step": 72730 }, { "grad_norm": 0.10237869620323181, "learning_rate": 1.8978474819376912e-05, "loss": 0.0044, "step": 72740 }, { "grad_norm": 0.11861313134431839, "learning_rate": 1.8965508989814406e-05, "loss": 0.0057, "step": 72750 }, { "grad_norm": 0.08817575871944427, "learning_rate": 1.8952546554137158e-05, "loss": 0.0042, "step": 72760 }, { "grad_norm": 0.126016303896904, "learning_rate": 1.8939587513762757e-05, "loss": 0.0036, "step": 72770 }, { "grad_norm": 0.10849139094352722, "learning_rate": 1.892663187010835e-05, "loss": 0.0042, "step": 72780 }, { "grad_norm": 0.17178337275981903, "learning_rate": 1.891367962459077e-05, "loss": 0.0044, "step": 72790 }, { "grad_norm": 0.20953312516212463, "learning_rate": 1.890073077862644e-05, "loss": 0.0054, "step": 72800 }, { "grad_norm": 0.10013555735349655, "learning_rate": 1.888778533363146e-05, "loss": 0.0039, "step": 72810 }, { "grad_norm": 0.10471262037754059, "learning_rate": 1.8874843291021478e-05, "loss": 0.0048, "step": 72820 }, { "grad_norm": 0.06303361803293228, "learning_rate": 1.8861904652211853e-05, "loss": 0.0055, "step": 72830 }, { "grad_norm": 0.10338710248470306, "learning_rate": 1.8848969418617507e-05, "loss": 0.0036, "step": 72840 }, { "grad_norm": 0.05573286861181259, "learning_rate": 1.8836037591653044e-05, "loss": 0.0061, "step": 72850 }, { "grad_norm": 0.07477005571126938, "learning_rate": 1.882310917273262e-05, "loss": 0.0044, "step": 72860 }, { "grad_norm": 0.17684388160705566, "learning_rate": 1.881018416327014e-05, "loss": 0.0054, "step": 72870 }, { "grad_norm": 0.10657444596290588, "learning_rate": 1.8797262564679014e-05, "loss": 0.0043, "step": 72880 }, { "grad_norm": 0.25772330164909363, "learning_rate": 1.8784344378372355e-05, "loss": 0.0036, "step": 72890 }, { "grad_norm": 0.0690503716468811, "learning_rate": 1.8771429605762843e-05, "loss": 0.0042, "step": 72900 }, { "grad_norm": 0.12733975052833557, "learning_rate": 1.8758518248262835e-05, "loss": 0.0055, "step": 72910 }, { "grad_norm": 0.11974797397851944, "learning_rate": 1.8745610307284305e-05, "loss": 0.0035, "step": 72920 }, { "grad_norm": 0.10510794073343277, "learning_rate": 1.8732705784238862e-05, "loss": 0.0035, "step": 72930 }, { "grad_norm": 0.09618252515792847, "learning_rate": 1.8719804680537682e-05, "loss": 0.0045, "step": 72940 }, { "grad_norm": 0.12494613975286484, "learning_rate": 1.870690699759164e-05, "loss": 0.0048, "step": 72950 }, { "grad_norm": 0.08755359053611755, "learning_rate": 1.8694012736811213e-05, "loss": 0.0028, "step": 72960 }, { "grad_norm": 0.08632951229810715, "learning_rate": 1.8681121899606463e-05, "loss": 0.0071, "step": 72970 }, { "grad_norm": 0.11359383910894394, "learning_rate": 1.8668234487387144e-05, "loss": 0.0044, "step": 72980 }, { "grad_norm": 0.0630013570189476, "learning_rate": 1.8655350501562592e-05, "loss": 0.0065, "step": 72990 }, { "grad_norm": 0.10767028480768204, "learning_rate": 1.8642469943541796e-05, "loss": 0.0036, "step": 73000 }, { "grad_norm": 0.1876743584871292, "learning_rate": 1.8629592814733328e-05, "loss": 0.0047, "step": 73010 }, { "grad_norm": 0.12320443987846375, "learning_rate": 1.8616719116545438e-05, "loss": 0.0058, "step": 73020 }, { "grad_norm": 0.09422978013753891, "learning_rate": 1.8603848850385952e-05, "loss": 0.0049, "step": 73030 }, { "grad_norm": 0.10180195420980453, "learning_rate": 1.859098201766236e-05, "loss": 0.0046, "step": 73040 }, { "grad_norm": 0.1031602993607521, "learning_rate": 1.8578118619781726e-05, "loss": 0.0049, "step": 73050 }, { "grad_norm": 0.08662787079811096, "learning_rate": 1.856525865815083e-05, "loss": 0.0031, "step": 73060 }, { "grad_norm": 0.08421162515878677, "learning_rate": 1.8552402134175966e-05, "loss": 0.0051, "step": 73070 }, { "grad_norm": 0.1114954724907875, "learning_rate": 1.8539549049263143e-05, "loss": 0.005, "step": 73080 }, { "grad_norm": 0.07725262641906738, "learning_rate": 1.8526699404817917e-05, "loss": 0.0042, "step": 73090 }, { "grad_norm": 0.1323823183774948, "learning_rate": 1.8513853202245546e-05, "loss": 0.0041, "step": 73100 }, { "grad_norm": 0.08825461566448212, "learning_rate": 1.850101044295081e-05, "loss": 0.0056, "step": 73110 }, { "grad_norm": 0.13788624107837677, "learning_rate": 1.848817112833825e-05, "loss": 0.0052, "step": 73120 }, { "grad_norm": 0.08748745918273926, "learning_rate": 1.8475335259811905e-05, "loss": 0.0036, "step": 73130 }, { "grad_norm": 0.135191410779953, "learning_rate": 1.8462502838775508e-05, "loss": 0.0035, "step": 73140 }, { "grad_norm": 0.12078099697828293, "learning_rate": 1.8449673866632378e-05, "loss": 0.0049, "step": 73150 }, { "grad_norm": 0.15410377085208893, "learning_rate": 1.8436848344785473e-05, "loss": 0.0042, "step": 73160 }, { "grad_norm": 0.07086362689733505, "learning_rate": 1.8424026274637378e-05, "loss": 0.0043, "step": 73170 }, { "grad_norm": 0.1514081060886383, "learning_rate": 1.8411207657590312e-05, "loss": 0.0064, "step": 73180 }, { "grad_norm": 0.14181160926818848, "learning_rate": 1.839839249504607e-05, "loss": 0.0076, "step": 73190 }, { "grad_norm": 0.10253295302391052, "learning_rate": 1.8385580788406108e-05, "loss": 0.0036, "step": 73200 }, { "grad_norm": 0.16828399896621704, "learning_rate": 1.837277253907152e-05, "loss": 0.004, "step": 73210 }, { "grad_norm": 0.1089962050318718, "learning_rate": 1.8359967748442957e-05, "loss": 0.0032, "step": 73220 }, { "grad_norm": 0.07238627970218658, "learning_rate": 1.8347166417920774e-05, "loss": 0.006, "step": 73230 }, { "grad_norm": 0.11095543950796127, "learning_rate": 1.833436854890484e-05, "loss": 0.0042, "step": 73240 }, { "grad_norm": 0.13075560331344604, "learning_rate": 1.8321574142794796e-05, "loss": 0.0043, "step": 73250 }, { "grad_norm": 0.13523997366428375, "learning_rate": 1.8308783200989748e-05, "loss": 0.0047, "step": 73260 }, { "grad_norm": 0.12491282820701599, "learning_rate": 1.829599572488855e-05, "loss": 0.0067, "step": 73270 }, { "grad_norm": 0.09594250470399857, "learning_rate": 1.828321171588957e-05, "loss": 0.0038, "step": 73280 }, { "grad_norm": 0.12320961058139801, "learning_rate": 1.827043117539089e-05, "loss": 0.0036, "step": 73290 }, { "grad_norm": 0.12167779356241226, "learning_rate": 1.8257654104790116e-05, "loss": 0.0054, "step": 73300 }, { "grad_norm": 0.06217437982559204, "learning_rate": 1.82448805054846e-05, "loss": 0.003, "step": 73310 }, { "grad_norm": 0.07941659539937973, "learning_rate": 1.823211037887119e-05, "loss": 0.0034, "step": 73320 }, { "grad_norm": 0.10704908519983292, "learning_rate": 1.821934372634645e-05, "loss": 0.005, "step": 73330 }, { "grad_norm": 0.13502237200737, "learning_rate": 1.8206580549306472e-05, "loss": 0.0052, "step": 73340 }, { "grad_norm": 0.07892511785030365, "learning_rate": 1.819382084914707e-05, "loss": 0.0035, "step": 73350 }, { "grad_norm": 0.13735482096672058, "learning_rate": 1.8181064627263577e-05, "loss": 0.0063, "step": 73360 }, { "grad_norm": 0.1292763650417328, "learning_rate": 1.8168311885051022e-05, "loss": 0.0049, "step": 73370 }, { "grad_norm": 0.11051172763109207, "learning_rate": 1.815556262390402e-05, "loss": 0.0044, "step": 73380 }, { "grad_norm": 0.12544599175453186, "learning_rate": 1.8142816845216826e-05, "loss": 0.0053, "step": 73390 }, { "grad_norm": 0.08996856212615967, "learning_rate": 1.8130074550383276e-05, "loss": 0.0033, "step": 73400 }, { "grad_norm": 0.1313905417919159, "learning_rate": 1.8117335740796848e-05, "loss": 0.0042, "step": 73410 }, { "grad_norm": 0.12538400292396545, "learning_rate": 1.810460041785067e-05, "loss": 0.0039, "step": 73420 }, { "grad_norm": 0.05691730976104736, "learning_rate": 1.809186858293742e-05, "loss": 0.005, "step": 73430 }, { "grad_norm": 0.06893326342105865, "learning_rate": 1.8079140237449448e-05, "loss": 0.0044, "step": 73440 }, { "grad_norm": 0.09025907516479492, "learning_rate": 1.8066415382778707e-05, "loss": 0.0054, "step": 73450 }, { "grad_norm": 0.050729431211948395, "learning_rate": 1.8053694020316786e-05, "loss": 0.0041, "step": 73460 }, { "grad_norm": 0.0888541042804718, "learning_rate": 1.8040976151454837e-05, "loss": 0.0041, "step": 73470 }, { "grad_norm": 0.12940388917922974, "learning_rate": 1.8028261777583704e-05, "loss": 0.0041, "step": 73480 }, { "grad_norm": 0.13330629467964172, "learning_rate": 1.8015550900093758e-05, "loss": 0.0079, "step": 73490 }, { "grad_norm": 0.13494540750980377, "learning_rate": 1.800284352037512e-05, "loss": 0.0049, "step": 73500 }, { "grad_norm": 0.08414231985807419, "learning_rate": 1.799013963981738e-05, "loss": 0.0031, "step": 73510 }, { "grad_norm": 0.14144068956375122, "learning_rate": 1.797743925980987e-05, "loss": 0.0044, "step": 73520 }, { "grad_norm": 0.26320019364356995, "learning_rate": 1.7964742381741435e-05, "loss": 0.0045, "step": 73530 }, { "grad_norm": 0.08900491148233414, "learning_rate": 1.7952049007000627e-05, "loss": 0.0055, "step": 73540 }, { "grad_norm": 0.11326610296964645, "learning_rate": 1.7939359136975535e-05, "loss": 0.0046, "step": 73550 }, { "grad_norm": 0.12740597128868103, "learning_rate": 1.7926672773053932e-05, "loss": 0.0055, "step": 73560 }, { "grad_norm": 0.17511694133281708, "learning_rate": 1.791398991662317e-05, "loss": 0.0041, "step": 73570 }, { "grad_norm": 0.16986365616321564, "learning_rate": 1.7901310569070244e-05, "loss": 0.005, "step": 73580 }, { "grad_norm": 0.16586194932460785, "learning_rate": 1.7888634731781724e-05, "loss": 0.0066, "step": 73590 }, { "grad_norm": 0.16413672268390656, "learning_rate": 1.7875962406143844e-05, "loss": 0.0049, "step": 73600 }, { "grad_norm": 0.09877388179302216, "learning_rate": 1.7863293593542403e-05, "loss": 0.003, "step": 73610 }, { "grad_norm": 0.17875054478645325, "learning_rate": 1.7850628295362853e-05, "loss": 0.004, "step": 73620 }, { "grad_norm": 0.16403360664844513, "learning_rate": 1.7837966512990263e-05, "loss": 0.0065, "step": 73630 }, { "grad_norm": 0.1074395626783371, "learning_rate": 1.782530824780931e-05, "loss": 0.0037, "step": 73640 }, { "grad_norm": 0.13874664902687073, "learning_rate": 1.7812653501204263e-05, "loss": 0.0036, "step": 73650 }, { "grad_norm": 0.09616593271493912, "learning_rate": 1.780000227455903e-05, "loss": 0.0057, "step": 73660 }, { "grad_norm": 0.1059647649526596, "learning_rate": 1.778735456925716e-05, "loss": 0.0042, "step": 73670 }, { "grad_norm": 0.11392177641391754, "learning_rate": 1.7774710386681737e-05, "loss": 0.0044, "step": 73680 }, { "grad_norm": 0.1403152197599411, "learning_rate": 1.7762069728215545e-05, "loss": 0.0032, "step": 73690 }, { "grad_norm": 0.09713644534349442, "learning_rate": 1.7749432595240932e-05, "loss": 0.0044, "step": 73700 }, { "grad_norm": 0.1256985366344452, "learning_rate": 1.773679898913991e-05, "loss": 0.0063, "step": 73710 }, { "grad_norm": 0.12408836930990219, "learning_rate": 1.7724168911294015e-05, "loss": 0.0046, "step": 73720 }, { "grad_norm": 0.1407010853290558, "learning_rate": 1.7711542363084504e-05, "loss": 0.0052, "step": 73730 }, { "grad_norm": 0.1456044465303421, "learning_rate": 1.7698919345892158e-05, "loss": 0.0057, "step": 73740 }, { "grad_norm": 0.15084464848041534, "learning_rate": 1.768629986109743e-05, "loss": 0.0039, "step": 73750 }, { "grad_norm": 0.06901401281356812, "learning_rate": 1.7673683910080357e-05, "loss": 0.0041, "step": 73760 }, { "grad_norm": 0.08454424887895584, "learning_rate": 1.7661071494220638e-05, "loss": 0.0048, "step": 73770 }, { "grad_norm": 0.131178081035614, "learning_rate": 1.7648462614897498e-05, "loss": 0.004, "step": 73780 }, { "grad_norm": 0.181399866938591, "learning_rate": 1.7635857273489863e-05, "loss": 0.0039, "step": 73790 }, { "grad_norm": 0.11213760077953339, "learning_rate": 1.7623255471376198e-05, "loss": 0.0069, "step": 73800 }, { "grad_norm": 0.09829484671354294, "learning_rate": 1.7610657209934646e-05, "loss": 0.0036, "step": 73810 }, { "grad_norm": 0.13285118341445923, "learning_rate": 1.7598062490542915e-05, "loss": 0.0049, "step": 73820 }, { "grad_norm": 0.18151932954788208, "learning_rate": 1.7585471314578377e-05, "loss": 0.0044, "step": 73830 }, { "grad_norm": 0.12262971699237823, "learning_rate": 1.7572883683417945e-05, "loss": 0.0042, "step": 73840 }, { "grad_norm": 0.10154440253973007, "learning_rate": 1.7560299598438212e-05, "loss": 0.0066, "step": 73850 }, { "grad_norm": 0.16244255006313324, "learning_rate": 1.7547719061015334e-05, "loss": 0.0045, "step": 73860 }, { "grad_norm": 0.1268542855978012, "learning_rate": 1.75351420725251e-05, "loss": 0.0055, "step": 73870 }, { "grad_norm": 0.1226014494895935, "learning_rate": 1.7522568634342923e-05, "loss": 0.0072, "step": 73880 }, { "grad_norm": 0.07917377352714539, "learning_rate": 1.7509998747843825e-05, "loss": 0.0049, "step": 73890 }, { "grad_norm": 0.1147032305598259, "learning_rate": 1.7497432414402403e-05, "loss": 0.0038, "step": 73900 }, { "grad_norm": 0.07119573652744293, "learning_rate": 1.7484869635392903e-05, "loss": 0.0034, "step": 73910 }, { "grad_norm": 0.10844562202692032, "learning_rate": 1.747231041218919e-05, "loss": 0.0027, "step": 73920 }, { "grad_norm": 0.07177639752626419, "learning_rate": 1.7459754746164692e-05, "loss": 0.0039, "step": 73930 }, { "grad_norm": 0.1085168719291687, "learning_rate": 1.744720263869249e-05, "loss": 0.0039, "step": 73940 }, { "grad_norm": 0.08248411864042282, "learning_rate": 1.7434654091145263e-05, "loss": 0.0037, "step": 73950 }, { "grad_norm": 0.10095607489347458, "learning_rate": 1.742210910489532e-05, "loss": 0.0056, "step": 73960 }, { "grad_norm": 0.14023388922214508, "learning_rate": 1.7409567681314532e-05, "loss": 0.004, "step": 73970 }, { "grad_norm": 0.17088398337364197, "learning_rate": 1.7397029821774434e-05, "loss": 0.0044, "step": 73980 }, { "grad_norm": 0.09181055426597595, "learning_rate": 1.7384495527646126e-05, "loss": 0.0067, "step": 73990 }, { "grad_norm": 0.13442163169384003, "learning_rate": 1.737196480030035e-05, "loss": 0.0028, "step": 74000 }, { "grad_norm": 0.10787847638130188, "learning_rate": 1.7359437641107447e-05, "loss": 0.0035, "step": 74010 }, { "grad_norm": 0.17590385675430298, "learning_rate": 1.7346914051437387e-05, "loss": 0.005, "step": 74020 }, { "grad_norm": 0.10650581866502762, "learning_rate": 1.733439403265969e-05, "loss": 0.0041, "step": 74030 }, { "grad_norm": 0.10330922156572342, "learning_rate": 1.732187758614357e-05, "loss": 0.0038, "step": 74040 }, { "grad_norm": 0.07309837639331818, "learning_rate": 1.730936471325777e-05, "loss": 0.0044, "step": 74050 }, { "grad_norm": 0.1073337271809578, "learning_rate": 1.7296855415370718e-05, "loss": 0.0041, "step": 74060 }, { "grad_norm": 0.07507620751857758, "learning_rate": 1.728434969385035e-05, "loss": 0.0039, "step": 74070 }, { "grad_norm": 0.08894797414541245, "learning_rate": 1.7271847550064346e-05, "loss": 0.0038, "step": 74080 }, { "grad_norm": 0.07363785058259964, "learning_rate": 1.725934898537987e-05, "loss": 0.0052, "step": 74090 }, { "grad_norm": 0.18389320373535156, "learning_rate": 1.7246854001163783e-05, "loss": 0.0066, "step": 74100 }, { "grad_norm": 0.14443953335285187, "learning_rate": 1.7234362598782476e-05, "loss": 0.0049, "step": 74110 }, { "grad_norm": 0.08389116823673248, "learning_rate": 1.722187477960202e-05, "loss": 0.0049, "step": 74120 }, { "grad_norm": 0.13599760830402374, "learning_rate": 1.7209390544988057e-05, "loss": 0.0042, "step": 74130 }, { "grad_norm": 0.09404083341360092, "learning_rate": 1.7196909896305856e-05, "loss": 0.0053, "step": 74140 }, { "grad_norm": 0.16277582943439484, "learning_rate": 1.718443283492026e-05, "loss": 0.0072, "step": 74150 }, { "grad_norm": 0.1773546189069748, "learning_rate": 1.7171959362195746e-05, "loss": 0.0043, "step": 74160 }, { "grad_norm": 0.1779634952545166, "learning_rate": 1.715948947949642e-05, "loss": 0.0057, "step": 74170 }, { "grad_norm": 0.07000698894262314, "learning_rate": 1.714702318818593e-05, "loss": 0.0046, "step": 74180 }, { "grad_norm": 0.18956226110458374, "learning_rate": 1.7134560489627592e-05, "loss": 0.0041, "step": 74190 }, { "grad_norm": 0.11099651455879211, "learning_rate": 1.712210138518431e-05, "loss": 0.005, "step": 74200 }, { "grad_norm": 0.1306428462266922, "learning_rate": 1.71096458762186e-05, "loss": 0.0042, "step": 74210 }, { "grad_norm": 0.11696288734674454, "learning_rate": 1.7097193964092552e-05, "loss": 0.0063, "step": 74220 }, { "grad_norm": 0.12662164866924286, "learning_rate": 1.708474565016792e-05, "loss": 0.0042, "step": 74230 }, { "grad_norm": 0.16509343683719635, "learning_rate": 1.7072300935806e-05, "loss": 0.0051, "step": 74240 }, { "grad_norm": 0.20594444870948792, "learning_rate": 1.705985982236776e-05, "loss": 0.0044, "step": 74250 }, { "grad_norm": 0.07647334039211273, "learning_rate": 1.7047422311213697e-05, "loss": 0.0057, "step": 74260 }, { "grad_norm": 0.09795340895652771, "learning_rate": 1.703498840370401e-05, "loss": 0.0033, "step": 74270 }, { "grad_norm": 0.12765508890151978, "learning_rate": 1.702255810119842e-05, "loss": 0.004, "step": 74280 }, { "grad_norm": 0.10017505288124084, "learning_rate": 1.7010131405056305e-05, "loss": 0.0035, "step": 74290 }, { "grad_norm": 0.06997241824865341, "learning_rate": 1.6997708316636613e-05, "loss": 0.0057, "step": 74300 }, { "grad_norm": 0.08890984952449799, "learning_rate": 1.6985288837297932e-05, "loss": 0.0054, "step": 74310 }, { "grad_norm": 0.08861149847507477, "learning_rate": 1.6972872968398394e-05, "loss": 0.0072, "step": 74320 }, { "grad_norm": 0.08522123843431473, "learning_rate": 1.6960460711295845e-05, "loss": 0.0051, "step": 74330 }, { "grad_norm": 0.12643641233444214, "learning_rate": 1.6948052067347623e-05, "loss": 0.0035, "step": 74340 }, { "grad_norm": 0.09757325053215027, "learning_rate": 1.693564703791075e-05, "loss": 0.006, "step": 74350 }, { "grad_norm": 0.0795222595334053, "learning_rate": 1.6923245624341793e-05, "loss": 0.0054, "step": 74360 }, { "grad_norm": 0.12806500494480133, "learning_rate": 1.6910847827996962e-05, "loss": 0.003, "step": 74370 }, { "grad_norm": 0.04534194990992546, "learning_rate": 1.689845365023206e-05, "loss": 0.004, "step": 74380 }, { "grad_norm": 0.12206366658210754, "learning_rate": 1.6886063092402505e-05, "loss": 0.005, "step": 74390 }, { "grad_norm": 0.09331014752388, "learning_rate": 1.6873676155863317e-05, "loss": 0.0046, "step": 74400 }, { "grad_norm": 0.12189368903636932, "learning_rate": 1.6861292841969085e-05, "loss": 0.0035, "step": 74410 }, { "grad_norm": 0.1253080517053604, "learning_rate": 1.684891315207406e-05, "loss": 0.0043, "step": 74420 }, { "grad_norm": 0.16984252631664276, "learning_rate": 1.6836537087532034e-05, "loss": 0.0042, "step": 74430 }, { "grad_norm": 0.1285495012998581, "learning_rate": 1.6824164649696465e-05, "loss": 0.005, "step": 74440 }, { "grad_norm": 0.1479395180940628, "learning_rate": 1.681179583992034e-05, "loss": 0.0036, "step": 74450 }, { "grad_norm": 0.06448052078485489, "learning_rate": 1.679943065955636e-05, "loss": 0.0041, "step": 74460 }, { "grad_norm": 0.09715543687343597, "learning_rate": 1.678706910995671e-05, "loss": 0.0051, "step": 74470 }, { "grad_norm": 0.14777818322181702, "learning_rate": 1.6774711192473262e-05, "loss": 0.004, "step": 74480 }, { "grad_norm": 0.09057815372943878, "learning_rate": 1.676235690845743e-05, "loss": 0.0048, "step": 74490 }, { "grad_norm": 0.09961722791194916, "learning_rate": 1.6750006259260285e-05, "loss": 0.0042, "step": 74500 }, { "grad_norm": 0.07012491673231125, "learning_rate": 1.673765924623244e-05, "loss": 0.0036, "step": 74510 }, { "grad_norm": 0.09006272256374359, "learning_rate": 1.67253158707242e-05, "loss": 0.0034, "step": 74520 }, { "grad_norm": 0.10391465574502945, "learning_rate": 1.671297613408537e-05, "loss": 0.0037, "step": 74530 }, { "grad_norm": 0.24673405289649963, "learning_rate": 1.6700640037665444e-05, "loss": 0.0028, "step": 74540 }, { "grad_norm": 0.1610715687274933, "learning_rate": 1.668830758281344e-05, "loss": 0.0056, "step": 74550 }, { "grad_norm": 0.1388847976922989, "learning_rate": 1.6675978770878052e-05, "loss": 0.0052, "step": 74560 }, { "grad_norm": 0.07407216727733612, "learning_rate": 1.6663653603207506e-05, "loss": 0.0045, "step": 74570 }, { "grad_norm": 0.06653151661157608, "learning_rate": 1.6651332081149685e-05, "loss": 0.0035, "step": 74580 }, { "grad_norm": 0.0767473503947258, "learning_rate": 1.6639014206052046e-05, "loss": 0.0038, "step": 74590 }, { "grad_norm": 0.22421886026859283, "learning_rate": 1.6626699979261657e-05, "loss": 0.0057, "step": 74600 }, { "grad_norm": 0.1000160351395607, "learning_rate": 1.6614389402125197e-05, "loss": 0.0049, "step": 74610 }, { "grad_norm": 0.1442827433347702, "learning_rate": 1.66020824759889e-05, "loss": 0.0036, "step": 74620 }, { "grad_norm": 0.09195852279663086, "learning_rate": 1.6589779202198675e-05, "loss": 0.0027, "step": 74630 }, { "grad_norm": 0.22604109346866608, "learning_rate": 1.6577479582099924e-05, "loss": 0.0049, "step": 74640 }, { "grad_norm": 0.0674547553062439, "learning_rate": 1.6565183617037788e-05, "loss": 0.0036, "step": 74650 }, { "grad_norm": 0.09646089375019073, "learning_rate": 1.6552891308356894e-05, "loss": 0.0041, "step": 74660 }, { "grad_norm": 0.11639365553855896, "learning_rate": 1.6540602657401526e-05, "loss": 0.0055, "step": 74670 }, { "grad_norm": 0.14845414459705353, "learning_rate": 1.6528317665515537e-05, "loss": 0.0042, "step": 74680 }, { "grad_norm": 0.15181580185890198, "learning_rate": 1.6516036334042416e-05, "loss": 0.0045, "step": 74690 }, { "grad_norm": 0.1042054072022438, "learning_rate": 1.650375866432519e-05, "loss": 0.0042, "step": 74700 }, { "grad_norm": 0.06972004473209381, "learning_rate": 1.649148465770659e-05, "loss": 0.0037, "step": 74710 }, { "grad_norm": 0.1361074149608612, "learning_rate": 1.647921431552883e-05, "loss": 0.0073, "step": 74720 }, { "grad_norm": 0.13202866911888123, "learning_rate": 1.646694763913381e-05, "loss": 0.0057, "step": 74730 }, { "grad_norm": 0.12155993282794952, "learning_rate": 1.6454684629862966e-05, "loss": 0.0055, "step": 74740 }, { "grad_norm": 0.10331284999847412, "learning_rate": 1.6442425289057388e-05, "loss": 0.0042, "step": 74750 }, { "grad_norm": 0.12020819634199142, "learning_rate": 1.6430169618057718e-05, "loss": 0.0055, "step": 74760 }, { "grad_norm": 0.15422362089157104, "learning_rate": 1.641791761820422e-05, "loss": 0.0032, "step": 74770 }, { "grad_norm": 0.15200816094875336, "learning_rate": 1.640566929083676e-05, "loss": 0.0043, "step": 74780 }, { "grad_norm": 0.11624590307474136, "learning_rate": 1.6393424637294823e-05, "loss": 0.008, "step": 74790 }, { "grad_norm": 0.10903260856866837, "learning_rate": 1.638118365891742e-05, "loss": 0.0045, "step": 74800 }, { "grad_norm": 0.16809509694576263, "learning_rate": 1.636894635704323e-05, "loss": 0.0048, "step": 74810 }, { "grad_norm": 0.10092202574014664, "learning_rate": 1.6356712733010517e-05, "loss": 0.0032, "step": 74820 }, { "grad_norm": 0.09725219756364822, "learning_rate": 1.6344482788157096e-05, "loss": 0.0053, "step": 74830 }, { "grad_norm": 0.12819203734397888, "learning_rate": 1.6332256523820444e-05, "loss": 0.0064, "step": 74840 }, { "grad_norm": 0.166395902633667, "learning_rate": 1.6320033941337597e-05, "loss": 0.0053, "step": 74850 }, { "grad_norm": 0.11678451299667358, "learning_rate": 1.6307815042045216e-05, "loss": 0.0039, "step": 74860 }, { "grad_norm": 0.13258525729179382, "learning_rate": 1.6295599827279513e-05, "loss": 0.0043, "step": 74870 }, { "grad_norm": 0.17057351768016815, "learning_rate": 1.6283388298376356e-05, "loss": 0.0033, "step": 74880 }, { "grad_norm": 0.14738772809505463, "learning_rate": 1.6271180456671125e-05, "loss": 0.0036, "step": 74890 }, { "grad_norm": 0.06400374323129654, "learning_rate": 1.625897630349893e-05, "loss": 0.0035, "step": 74900 }, { "grad_norm": 0.1484963595867157, "learning_rate": 1.6246775840194335e-05, "loss": 0.0065, "step": 74910 }, { "grad_norm": 0.11179530620574951, "learning_rate": 1.6234579068091604e-05, "loss": 0.0032, "step": 74920 }, { "grad_norm": 0.06881362199783325, "learning_rate": 1.6222385988524524e-05, "loss": 0.004, "step": 74930 }, { "grad_norm": 0.12325289100408554, "learning_rate": 1.6210196602826544e-05, "loss": 0.0039, "step": 74940 }, { "grad_norm": 0.11759991943836212, "learning_rate": 1.6198010912330646e-05, "loss": 0.0061, "step": 74950 }, { "grad_norm": 0.10568298399448395, "learning_rate": 1.6185828918369456e-05, "loss": 0.0068, "step": 74960 }, { "grad_norm": 0.10791289061307907, "learning_rate": 1.617365062227517e-05, "loss": 0.0041, "step": 74970 }, { "grad_norm": 0.0470602810382843, "learning_rate": 1.616147602537962e-05, "loss": 0.0037, "step": 74980 }, { "grad_norm": 0.1658175140619278, "learning_rate": 1.614930512901415e-05, "loss": 0.0036, "step": 74990 }, { "grad_norm": 0.10122112184762955, "learning_rate": 1.6137137934509804e-05, "loss": 0.004, "step": 75000 }, { "grad_norm": 0.12668144702911377, "learning_rate": 1.6124974443197123e-05, "loss": 0.0064, "step": 75010 }, { "grad_norm": 0.12030289322137833, "learning_rate": 1.6112814656406306e-05, "loss": 0.0045, "step": 75020 }, { "grad_norm": 0.1339724212884903, "learning_rate": 1.6100658575467125e-05, "loss": 0.006, "step": 75030 }, { "grad_norm": 0.09356119483709335, "learning_rate": 1.6088506201708974e-05, "loss": 0.0033, "step": 75040 }, { "grad_norm": 0.12062900513410568, "learning_rate": 1.6076357536460785e-05, "loss": 0.0043, "step": 75050 }, { "grad_norm": 0.14741773903369904, "learning_rate": 1.606421258105113e-05, "loss": 0.0059, "step": 75060 }, { "grad_norm": 0.06210764870047569, "learning_rate": 1.6052071336808183e-05, "loss": 0.0046, "step": 75070 }, { "grad_norm": 0.15509328246116638, "learning_rate": 1.6039933805059658e-05, "loss": 0.0068, "step": 75080 }, { "grad_norm": 0.14467623829841614, "learning_rate": 1.6027799987132915e-05, "loss": 0.0071, "step": 75090 }, { "grad_norm": 0.05938009172677994, "learning_rate": 1.6015669884354888e-05, "loss": 0.0047, "step": 75100 }, { "grad_norm": 0.06396040320396423, "learning_rate": 1.6003543498052125e-05, "loss": 0.0062, "step": 75110 }, { "grad_norm": 0.15519586205482483, "learning_rate": 1.5991420829550715e-05, "loss": 0.003, "step": 75120 }, { "grad_norm": 0.11827424168586731, "learning_rate": 1.5979301880176405e-05, "loss": 0.0053, "step": 75130 }, { "grad_norm": 0.08610949665307999, "learning_rate": 1.5967186651254485e-05, "loss": 0.0044, "step": 75140 }, { "grad_norm": 0.08774873614311218, "learning_rate": 1.5955075144109853e-05, "loss": 0.0067, "step": 75150 }, { "grad_norm": 0.09380020201206207, "learning_rate": 1.594296736006703e-05, "loss": 0.0039, "step": 75160 }, { "grad_norm": 0.10257484763860703, "learning_rate": 1.59308633004501e-05, "loss": 0.0046, "step": 75170 }, { "grad_norm": 0.1680845469236374, "learning_rate": 1.5918762966582724e-05, "loss": 0.0052, "step": 75180 }, { "grad_norm": 0.10647910833358765, "learning_rate": 1.5906666359788203e-05, "loss": 0.004, "step": 75190 }, { "grad_norm": 0.11731456220149994, "learning_rate": 1.589457348138938e-05, "loss": 0.0085, "step": 75200 }, { "grad_norm": 0.11926549673080444, "learning_rate": 1.588248433270872e-05, "loss": 0.0036, "step": 75210 }, { "grad_norm": 0.08322262763977051, "learning_rate": 1.587039891506828e-05, "loss": 0.0043, "step": 75220 }, { "grad_norm": 0.10044930130243301, "learning_rate": 1.5858317229789715e-05, "loss": 0.0068, "step": 75230 }, { "grad_norm": 0.08387904614210129, "learning_rate": 1.584623927819423e-05, "loss": 0.0052, "step": 75240 }, { "grad_norm": 0.11316324025392532, "learning_rate": 1.583416506160269e-05, "loss": 0.0058, "step": 75250 }, { "grad_norm": 0.10846099257469177, "learning_rate": 1.5822094581335472e-05, "loss": 0.0051, "step": 75260 }, { "grad_norm": 0.11494994163513184, "learning_rate": 1.5810027838712604e-05, "loss": 0.004, "step": 75270 }, { "grad_norm": 0.16459274291992188, "learning_rate": 1.5797964835053687e-05, "loss": 0.005, "step": 75280 }, { "grad_norm": 0.10508731007575989, "learning_rate": 1.5785905571677933e-05, "loss": 0.0053, "step": 75290 }, { "grad_norm": 0.10733170807361603, "learning_rate": 1.5773850049904086e-05, "loss": 0.0032, "step": 75300 }, { "grad_norm": 0.10591060668230057, "learning_rate": 1.5761798271050543e-05, "loss": 0.0048, "step": 75310 }, { "grad_norm": 0.28860124945640564, "learning_rate": 1.5749750236435277e-05, "loss": 0.0031, "step": 75320 }, { "grad_norm": 0.12627768516540527, "learning_rate": 1.5737705947375815e-05, "loss": 0.0042, "step": 75330 }, { "grad_norm": 0.10104615986347198, "learning_rate": 1.5725665405189322e-05, "loss": 0.0029, "step": 75340 }, { "grad_norm": 0.1710216999053955, "learning_rate": 1.571362861119253e-05, "loss": 0.0039, "step": 75350 }, { "grad_norm": 0.11196841299533844, "learning_rate": 1.570159556670177e-05, "loss": 0.0049, "step": 75360 }, { "grad_norm": 0.13443811237812042, "learning_rate": 1.5689566273032946e-05, "loss": 0.0035, "step": 75370 }, { "grad_norm": 0.08777894079685211, "learning_rate": 1.5677540731501583e-05, "loss": 0.0068, "step": 75380 }, { "grad_norm": 0.12253684550523758, "learning_rate": 1.5665518943422746e-05, "loss": 0.0044, "step": 75390 }, { "grad_norm": 0.12726864218711853, "learning_rate": 1.5653500910111134e-05, "loss": 0.0047, "step": 75400 }, { "grad_norm": 0.11552369594573975, "learning_rate": 1.5641486632881023e-05, "loss": 0.0033, "step": 75410 }, { "grad_norm": 0.1339944750070572, "learning_rate": 1.56294761130463e-05, "loss": 0.0044, "step": 75420 }, { "grad_norm": 0.18632473051548004, "learning_rate": 1.5617469351920373e-05, "loss": 0.004, "step": 75430 }, { "grad_norm": 0.13883592188358307, "learning_rate": 1.5605466350816315e-05, "loss": 0.0036, "step": 75440 }, { "grad_norm": 0.13142159581184387, "learning_rate": 1.559346711104674e-05, "loss": 0.0049, "step": 75450 }, { "grad_norm": 0.06749240309000015, "learning_rate": 1.5581471633923883e-05, "loss": 0.0034, "step": 75460 }, { "grad_norm": 0.12284647673368454, "learning_rate": 1.5569479920759505e-05, "loss": 0.0029, "step": 75470 }, { "grad_norm": 0.09051982313394547, "learning_rate": 1.5557491972865072e-05, "loss": 0.0034, "step": 75480 }, { "grad_norm": 0.11462472379207611, "learning_rate": 1.554550779155152e-05, "loss": 0.0062, "step": 75490 }, { "grad_norm": 0.08477507531642914, "learning_rate": 1.5533527378129453e-05, "loss": 0.0038, "step": 75500 }, { "grad_norm": 0.08274912089109421, "learning_rate": 1.552155073390899e-05, "loss": 0.0054, "step": 75510 }, { "grad_norm": 0.14004650712013245, "learning_rate": 1.5509577860199903e-05, "loss": 0.0042, "step": 75520 }, { "grad_norm": 0.21687592566013336, "learning_rate": 1.5497608758311525e-05, "loss": 0.0058, "step": 75530 }, { "grad_norm": 0.12710504233837128, "learning_rate": 1.5485643429552797e-05, "loss": 0.0035, "step": 75540 }, { "grad_norm": 0.15678496658802032, "learning_rate": 1.5473681875232197e-05, "loss": 0.0068, "step": 75550 }, { "grad_norm": 0.10948637872934341, "learning_rate": 1.5461724096657837e-05, "loss": 0.0051, "step": 75560 }, { "grad_norm": 0.11339117586612701, "learning_rate": 1.5449770095137416e-05, "loss": 0.004, "step": 75570 }, { "grad_norm": 0.11469101905822754, "learning_rate": 1.5437819871978178e-05, "loss": 0.0029, "step": 75580 }, { "grad_norm": 0.1573711782693863, "learning_rate": 1.5425873428486992e-05, "loss": 0.0031, "step": 75590 }, { "grad_norm": 0.1281205266714096, "learning_rate": 1.5413930765970312e-05, "loss": 0.0029, "step": 75600 }, { "grad_norm": 0.10864143073558807, "learning_rate": 1.5401991885734175e-05, "loss": 0.0039, "step": 75610 }, { "grad_norm": 0.09185242652893066, "learning_rate": 1.5390056789084174e-05, "loss": 0.0052, "step": 75620 }, { "grad_norm": 0.21794390678405762, "learning_rate": 1.5378125477325545e-05, "loss": 0.004, "step": 75630 }, { "grad_norm": 0.04732928425073624, "learning_rate": 1.5366197951763046e-05, "loss": 0.0042, "step": 75640 }, { "grad_norm": 0.1350087672472, "learning_rate": 1.5354274213701075e-05, "loss": 0.0036, "step": 75650 }, { "grad_norm": 0.05726164951920509, "learning_rate": 1.5342354264443566e-05, "loss": 0.0049, "step": 75660 }, { "grad_norm": 0.12877407670021057, "learning_rate": 1.533043810529411e-05, "loss": 0.0041, "step": 75670 }, { "grad_norm": 0.06822559237480164, "learning_rate": 1.5318525737555807e-05, "loss": 0.0045, "step": 75680 }, { "grad_norm": 0.17370329797267914, "learning_rate": 1.53066171625314e-05, "loss": 0.0053, "step": 75690 }, { "grad_norm": 0.1497793346643448, "learning_rate": 1.529471238152317e-05, "loss": 0.0031, "step": 75700 }, { "grad_norm": 0.07998921722173691, "learning_rate": 1.5282811395833023e-05, "loss": 0.0032, "step": 75710 }, { "grad_norm": 0.12637069821357727, "learning_rate": 1.5270914206762397e-05, "loss": 0.0051, "step": 75720 }, { "grad_norm": 0.1407594084739685, "learning_rate": 1.525902081561241e-05, "loss": 0.0046, "step": 75730 }, { "grad_norm": 0.11167572438716888, "learning_rate": 1.5247131223683664e-05, "loss": 0.0043, "step": 75740 }, { "grad_norm": 0.13291803002357483, "learning_rate": 1.5235245432276408e-05, "loss": 0.0033, "step": 75750 }, { "grad_norm": 0.07948756217956543, "learning_rate": 1.5223363442690424e-05, "loss": 0.0045, "step": 75760 }, { "grad_norm": 0.055643003433942795, "learning_rate": 1.5211485256225133e-05, "loss": 0.0034, "step": 75770 }, { "grad_norm": 0.12415704876184464, "learning_rate": 1.519961087417951e-05, "loss": 0.006, "step": 75780 }, { "grad_norm": 0.15487603843212128, "learning_rate": 1.5187740297852137e-05, "loss": 0.0039, "step": 75790 }, { "grad_norm": 0.09588801860809326, "learning_rate": 1.517587352854113e-05, "loss": 0.0042, "step": 75800 }, { "grad_norm": 0.10067823529243469, "learning_rate": 1.516401056754423e-05, "loss": 0.0034, "step": 75810 }, { "grad_norm": 0.08618202060461044, "learning_rate": 1.5152151416158783e-05, "loss": 0.0033, "step": 75820 }, { "grad_norm": 0.11122225970029831, "learning_rate": 1.5140296075681642e-05, "loss": 0.007, "step": 75830 }, { "grad_norm": 0.12230801582336426, "learning_rate": 1.5128444547409331e-05, "loss": 0.0033, "step": 75840 }, { "grad_norm": 0.07157284021377563, "learning_rate": 1.5116596832637864e-05, "loss": 0.0043, "step": 75850 }, { "grad_norm": 0.05792519450187683, "learning_rate": 1.5104752932662952e-05, "loss": 0.0056, "step": 75860 }, { "grad_norm": 0.09589669853448868, "learning_rate": 1.5092912848779778e-05, "loss": 0.0039, "step": 75870 }, { "grad_norm": 0.10163421183824539, "learning_rate": 1.5081076582283194e-05, "loss": 0.0037, "step": 75880 }, { "grad_norm": 0.06214575096964836, "learning_rate": 1.5069244134467553e-05, "loss": 0.0044, "step": 75890 }, { "grad_norm": 0.09965787827968597, "learning_rate": 1.5057415506626876e-05, "loss": 0.004, "step": 75900 }, { "grad_norm": 0.07583426684141159, "learning_rate": 1.5045590700054668e-05, "loss": 0.0038, "step": 75910 }, { "grad_norm": 0.11339717358350754, "learning_rate": 1.5033769716044144e-05, "loss": 0.0054, "step": 75920 }, { "grad_norm": 0.15879689157009125, "learning_rate": 1.5021952555887975e-05, "loss": 0.0045, "step": 75930 }, { "grad_norm": 0.09749045968055725, "learning_rate": 1.50101392208785e-05, "loss": 0.0044, "step": 75940 }, { "grad_norm": 0.15378256142139435, "learning_rate": 1.4998329712307579e-05, "loss": 0.0045, "step": 75950 }, { "grad_norm": 0.12292494624853134, "learning_rate": 1.4986524031466708e-05, "loss": 0.0039, "step": 75960 }, { "grad_norm": 0.17121966183185577, "learning_rate": 1.4974722179646916e-05, "loss": 0.005, "step": 75970 }, { "grad_norm": 0.1309392750263214, "learning_rate": 1.4962924158138847e-05, "loss": 0.0041, "step": 75980 }, { "grad_norm": 0.10683467984199524, "learning_rate": 1.4951129968232714e-05, "loss": 0.0032, "step": 75990 }, { "grad_norm": 0.06756845861673355, "learning_rate": 1.4939339611218329e-05, "loss": 0.0051, "step": 76000 }, { "grad_norm": 0.11514542996883392, "learning_rate": 1.4927553088385032e-05, "loss": 0.004, "step": 76010 }, { "grad_norm": 0.0883399099111557, "learning_rate": 1.4915770401021801e-05, "loss": 0.0034, "step": 76020 }, { "grad_norm": 0.20289571583271027, "learning_rate": 1.4903991550417185e-05, "loss": 0.0047, "step": 76030 }, { "grad_norm": 0.2170751839876175, "learning_rate": 1.4892216537859277e-05, "loss": 0.003, "step": 76040 }, { "grad_norm": 0.15828755497932434, "learning_rate": 1.4880445364635775e-05, "loss": 0.0062, "step": 76050 }, { "grad_norm": 0.148670956492424, "learning_rate": 1.4868678032033973e-05, "loss": 0.0039, "step": 76060 }, { "grad_norm": 0.133160799741745, "learning_rate": 1.4856914541340739e-05, "loss": 0.0043, "step": 76070 }, { "grad_norm": 0.13933569192886353, "learning_rate": 1.4845154893842472e-05, "loss": 0.0032, "step": 76080 }, { "grad_norm": 0.11314121633768082, "learning_rate": 1.4833399090825234e-05, "loss": 0.0053, "step": 76090 }, { "grad_norm": 0.11539257317781448, "learning_rate": 1.4821647133574562e-05, "loss": 0.0036, "step": 76100 }, { "grad_norm": 0.1257971078157425, "learning_rate": 1.4809899023375706e-05, "loss": 0.0049, "step": 76110 }, { "grad_norm": 0.17378494143486023, "learning_rate": 1.4798154761513366e-05, "loss": 0.0057, "step": 76120 }, { "grad_norm": 0.1093987226486206, "learning_rate": 1.4786414349271916e-05, "loss": 0.0044, "step": 76130 }, { "grad_norm": 0.12780851125717163, "learning_rate": 1.4774677787935226e-05, "loss": 0.0041, "step": 76140 }, { "grad_norm": 0.2493186891078949, "learning_rate": 1.4762945078786838e-05, "loss": 0.0042, "step": 76150 }, { "grad_norm": 0.09983917325735092, "learning_rate": 1.4751216223109777e-05, "loss": 0.0047, "step": 76160 }, { "grad_norm": 0.09576746821403503, "learning_rate": 1.4739491222186714e-05, "loss": 0.0056, "step": 76170 }, { "grad_norm": 0.07880068570375443, "learning_rate": 1.4727770077299879e-05, "loss": 0.0031, "step": 76180 }, { "grad_norm": 0.17043709754943848, "learning_rate": 1.4716052789731094e-05, "loss": 0.0036, "step": 76190 }, { "grad_norm": 0.21948781609535217, "learning_rate": 1.470433936076171e-05, "loss": 0.0037, "step": 76200 }, { "grad_norm": 0.1731037050485611, "learning_rate": 1.4692629791672719e-05, "loss": 0.0046, "step": 76210 }, { "grad_norm": 0.1207890510559082, "learning_rate": 1.468092408374464e-05, "loss": 0.0062, "step": 76220 }, { "grad_norm": 0.13331498205661774, "learning_rate": 1.4669222238257602e-05, "loss": 0.0052, "step": 76230 }, { "grad_norm": 0.10224760323762894, "learning_rate": 1.4657524256491301e-05, "loss": 0.0036, "step": 76240 }, { "grad_norm": 0.11263717710971832, "learning_rate": 1.464583013972502e-05, "loss": 0.0039, "step": 76250 }, { "grad_norm": 0.07100008428096771, "learning_rate": 1.4634139889237591e-05, "loss": 0.0041, "step": 76260 }, { "grad_norm": 0.19719167053699493, "learning_rate": 1.4622453506307448e-05, "loss": 0.005, "step": 76270 }, { "grad_norm": 0.11159035563468933, "learning_rate": 1.461077099221262e-05, "loss": 0.0041, "step": 76280 }, { "grad_norm": 0.12379682809114456, "learning_rate": 1.4599092348230652e-05, "loss": 0.004, "step": 76290 }, { "grad_norm": 0.15039673447608948, "learning_rate": 1.4587417575638724e-05, "loss": 0.0063, "step": 76300 }, { "grad_norm": 0.07884006202220917, "learning_rate": 1.4575746675713563e-05, "loss": 0.0056, "step": 76310 }, { "grad_norm": 0.1741846352815628, "learning_rate": 1.4564079649731504e-05, "loss": 0.0036, "step": 76320 }, { "grad_norm": 0.2510077655315399, "learning_rate": 1.4552416498968402e-05, "loss": 0.0043, "step": 76330 }, { "grad_norm": 0.13709968328475952, "learning_rate": 1.4540757224699747e-05, "loss": 0.0033, "step": 76340 }, { "grad_norm": 0.07528514415025711, "learning_rate": 1.452910182820056e-05, "loss": 0.0052, "step": 76350 }, { "grad_norm": 0.07247256487607956, "learning_rate": 1.4517450310745468e-05, "loss": 0.0033, "step": 76360 }, { "grad_norm": 0.10857312381267548, "learning_rate": 1.450580267360866e-05, "loss": 0.0043, "step": 76370 }, { "grad_norm": 0.1235072910785675, "learning_rate": 1.4494158918063922e-05, "loss": 0.0051, "step": 76380 }, { "grad_norm": 0.07865875214338303, "learning_rate": 1.4482519045384568e-05, "loss": 0.0036, "step": 76390 }, { "grad_norm": 0.1073717549443245, "learning_rate": 1.4470883056843548e-05, "loss": 0.0036, "step": 76400 }, { "grad_norm": 0.09109239280223846, "learning_rate": 1.4459250953713322e-05, "loss": 0.005, "step": 76410 }, { "grad_norm": 0.15286099910736084, "learning_rate": 1.444762273726598e-05, "loss": 0.0039, "step": 76420 }, { "grad_norm": 0.1705797016620636, "learning_rate": 1.443599840877316e-05, "loss": 0.004, "step": 76430 }, { "grad_norm": 0.12487123906612396, "learning_rate": 1.4424377969506103e-05, "loss": 0.004, "step": 76440 }, { "grad_norm": 0.10804013162851334, "learning_rate": 1.4412761420735565e-05, "loss": 0.0057, "step": 76450 }, { "grad_norm": 0.12659424543380737, "learning_rate": 1.4401148763731953e-05, "loss": 0.004, "step": 76460 }, { "grad_norm": 0.13163340091705322, "learning_rate": 1.4389539999765173e-05, "loss": 0.0036, "step": 76470 }, { "grad_norm": 0.13769367337226868, "learning_rate": 1.4377935130104758e-05, "loss": 0.0064, "step": 76480 }, { "grad_norm": 0.12922152876853943, "learning_rate": 1.4366334156019801e-05, "loss": 0.0053, "step": 76490 }, { "grad_norm": 0.1354692578315735, "learning_rate": 1.435473707877898e-05, "loss": 0.006, "step": 76500 }, { "grad_norm": 0.0991237685084343, "learning_rate": 1.4343143899650508e-05, "loss": 0.0043, "step": 76510 }, { "grad_norm": 0.06767316162586212, "learning_rate": 1.4331554619902204e-05, "loss": 0.0037, "step": 76520 }, { "grad_norm": 0.12048842757940292, "learning_rate": 1.4319969240801484e-05, "loss": 0.0053, "step": 76530 }, { "grad_norm": 0.1262214630842209, "learning_rate": 1.430838776361526e-05, "loss": 0.0041, "step": 76540 }, { "grad_norm": 0.1132708489894867, "learning_rate": 1.4296810189610093e-05, "loss": 0.0053, "step": 76550 }, { "grad_norm": 0.11574958264827728, "learning_rate": 1.4285236520052087e-05, "loss": 0.0047, "step": 76560 }, { "grad_norm": 0.12222365289926529, "learning_rate": 1.4273666756206933e-05, "loss": 0.0039, "step": 76570 }, { "grad_norm": 0.0475970096886158, "learning_rate": 1.4262100899339848e-05, "loss": 0.0048, "step": 76580 }, { "grad_norm": 0.12933743000030518, "learning_rate": 1.4250538950715696e-05, "loss": 0.0058, "step": 76590 }, { "grad_norm": 0.25056764483451843, "learning_rate": 1.4238980911598842e-05, "loss": 0.0041, "step": 76600 }, { "grad_norm": 0.14902134239673615, "learning_rate": 1.4227426783253267e-05, "loss": 0.0034, "step": 76610 }, { "grad_norm": 0.1664838343858719, "learning_rate": 1.4215876566942521e-05, "loss": 0.0038, "step": 76620 }, { "grad_norm": 0.08541479706764221, "learning_rate": 1.4204330263929721e-05, "loss": 0.0044, "step": 76630 }, { "grad_norm": 0.06696383655071259, "learning_rate": 1.4192787875477537e-05, "loss": 0.004, "step": 76640 }, { "grad_norm": 0.11879736185073853, "learning_rate": 1.4181249402848246e-05, "loss": 0.004, "step": 76650 }, { "grad_norm": 0.14324863255023956, "learning_rate": 1.4169714847303656e-05, "loss": 0.0032, "step": 76660 }, { "grad_norm": 0.0975709781050682, "learning_rate": 1.4158184210105197e-05, "loss": 0.0032, "step": 76670 }, { "grad_norm": 0.1241474598646164, "learning_rate": 1.4146657492513787e-05, "loss": 0.0036, "step": 76680 }, { "grad_norm": 0.112086221575737, "learning_rate": 1.4135134695790047e-05, "loss": 0.0036, "step": 76690 }, { "grad_norm": 0.1399272233247757, "learning_rate": 1.4123615821194037e-05, "loss": 0.004, "step": 76700 }, { "grad_norm": 0.08696075528860092, "learning_rate": 1.4112100869985473e-05, "loss": 0.0039, "step": 76710 }, { "grad_norm": 0.11515114456415176, "learning_rate": 1.4100589843423589e-05, "loss": 0.0041, "step": 76720 }, { "grad_norm": 0.09897264838218689, "learning_rate": 1.408908274276723e-05, "loss": 0.0042, "step": 76730 }, { "grad_norm": 0.12458847463130951, "learning_rate": 1.4077579569274786e-05, "loss": 0.0041, "step": 76740 }, { "grad_norm": 0.10522831231355667, "learning_rate": 1.4066080324204245e-05, "loss": 0.005, "step": 76750 }, { "grad_norm": 0.08996798098087311, "learning_rate": 1.405458500881312e-05, "loss": 0.0031, "step": 76760 }, { "grad_norm": 0.1306101232767105, "learning_rate": 1.4043093624358538e-05, "loss": 0.0035, "step": 76770 }, { "grad_norm": 0.1605299562215805, "learning_rate": 1.4031606172097189e-05, "loss": 0.0046, "step": 76780 }, { "grad_norm": 0.14263613522052765, "learning_rate": 1.40201226532853e-05, "loss": 0.003, "step": 76790 }, { "grad_norm": 0.1920090615749359, "learning_rate": 1.4008643069178701e-05, "loss": 0.0036, "step": 76800 }, { "grad_norm": 0.1286948323249817, "learning_rate": 1.399716742103278e-05, "loss": 0.0038, "step": 76810 }, { "grad_norm": 0.08505254238843918, "learning_rate": 1.398569571010252e-05, "loss": 0.0035, "step": 76820 }, { "grad_norm": 0.12682388722896576, "learning_rate": 1.3974227937642414e-05, "loss": 0.004, "step": 76830 }, { "grad_norm": 0.12528933584690094, "learning_rate": 1.3962764104906595e-05, "loss": 0.0037, "step": 76840 }, { "grad_norm": 0.06997460126876831, "learning_rate": 1.3951304213148696e-05, "loss": 0.004, "step": 76850 }, { "grad_norm": 0.06636141240596771, "learning_rate": 1.3939848263621979e-05, "loss": 0.0038, "step": 76860 }, { "grad_norm": 0.0985696092247963, "learning_rate": 1.3928396257579219e-05, "loss": 0.0037, "step": 76870 }, { "grad_norm": 0.18779361248016357, "learning_rate": 1.3916948196272833e-05, "loss": 0.0034, "step": 76880 }, { "grad_norm": 0.23667262494564056, "learning_rate": 1.3905504080954734e-05, "loss": 0.0037, "step": 76890 }, { "grad_norm": 0.0964885726571083, "learning_rate": 1.3894063912876454e-05, "loss": 0.0038, "step": 76900 }, { "grad_norm": 0.13475589454174042, "learning_rate": 1.3882627693289046e-05, "loss": 0.0051, "step": 76910 }, { "grad_norm": 0.0996757373213768, "learning_rate": 1.3871195423443183e-05, "loss": 0.0037, "step": 76920 }, { "grad_norm": 0.1270604133605957, "learning_rate": 1.3859767104589044e-05, "loss": 0.0052, "step": 76930 }, { "grad_norm": 0.0699038952589035, "learning_rate": 1.3848342737976467e-05, "loss": 0.004, "step": 76940 }, { "grad_norm": 0.1773063689470291, "learning_rate": 1.3836922324854757e-05, "loss": 0.0046, "step": 76950 }, { "grad_norm": 0.11025097221136093, "learning_rate": 1.382550586647287e-05, "loss": 0.003, "step": 76960 }, { "grad_norm": 0.12169064581394196, "learning_rate": 1.3814093364079257e-05, "loss": 0.0049, "step": 76970 }, { "grad_norm": 0.25019553303718567, "learning_rate": 1.3802684818921991e-05, "loss": 0.0034, "step": 76980 }, { "grad_norm": 0.0676213949918747, "learning_rate": 1.3791280232248687e-05, "loss": 0.004, "step": 76990 }, { "grad_norm": 0.11359814554452896, "learning_rate": 1.3779879605306555e-05, "loss": 0.0044, "step": 77000 }, { "grad_norm": 0.09183169901371002, "learning_rate": 1.376848293934232e-05, "loss": 0.0033, "step": 77010 }, { "grad_norm": 0.09572699666023254, "learning_rate": 1.3757090235602316e-05, "loss": 0.0044, "step": 77020 }, { "grad_norm": 0.15963733196258545, "learning_rate": 1.3745701495332447e-05, "loss": 0.0038, "step": 77030 }, { "grad_norm": 0.1302589476108551, "learning_rate": 1.3734316719778145e-05, "loss": 0.0029, "step": 77040 }, { "grad_norm": 0.10400904715061188, "learning_rate": 1.3722935910184454e-05, "loss": 0.004, "step": 77050 }, { "grad_norm": 0.12400984019041061, "learning_rate": 1.3711559067795921e-05, "loss": 0.0043, "step": 77060 }, { "grad_norm": 0.08171380311250687, "learning_rate": 1.3700186193856756e-05, "loss": 0.0044, "step": 77070 }, { "grad_norm": 0.09762326627969742, "learning_rate": 1.368881728961064e-05, "loss": 0.0033, "step": 77080 }, { "grad_norm": 0.08986189216375351, "learning_rate": 1.3677452356300885e-05, "loss": 0.0034, "step": 77090 }, { "grad_norm": 0.11552688479423523, "learning_rate": 1.3666091395170321e-05, "loss": 0.0038, "step": 77100 }, { "grad_norm": 0.09067854285240173, "learning_rate": 1.3654734407461384e-05, "loss": 0.0032, "step": 77110 }, { "grad_norm": 0.09679592400789261, "learning_rate": 1.3643381394416022e-05, "loss": 0.0037, "step": 77120 }, { "grad_norm": 0.1103653535246849, "learning_rate": 1.3632032357275838e-05, "loss": 0.0054, "step": 77130 }, { "grad_norm": 0.10747970640659332, "learning_rate": 1.3620687297281898e-05, "loss": 0.0039, "step": 77140 }, { "grad_norm": 0.11657579243183136, "learning_rate": 1.3609346215674917e-05, "loss": 0.0034, "step": 77150 }, { "grad_norm": 0.12401965260505676, "learning_rate": 1.3598009113695109e-05, "loss": 0.0028, "step": 77160 }, { "grad_norm": 0.09426939487457275, "learning_rate": 1.358667599258231e-05, "loss": 0.0029, "step": 77170 }, { "grad_norm": 0.1672511249780655, "learning_rate": 1.357534685357586e-05, "loss": 0.0036, "step": 77180 }, { "grad_norm": 0.09463915973901749, "learning_rate": 1.3564021697914715e-05, "loss": 0.0044, "step": 77190 }, { "grad_norm": 0.10730574280023575, "learning_rate": 1.3552700526837381e-05, "loss": 0.0048, "step": 77200 }, { "grad_norm": 0.1133648082613945, "learning_rate": 1.3541383341581925e-05, "loss": 0.0031, "step": 77210 }, { "grad_norm": 0.1893383413553238, "learning_rate": 1.3530070143385965e-05, "loss": 0.0044, "step": 77220 }, { "grad_norm": 0.1442035585641861, "learning_rate": 1.3518760933486702e-05, "loss": 0.0036, "step": 77230 }, { "grad_norm": 0.11833397299051285, "learning_rate": 1.3507455713120908e-05, "loss": 0.0036, "step": 77240 }, { "grad_norm": 0.09682110697031021, "learning_rate": 1.3496154483524876e-05, "loss": 0.0038, "step": 77250 }, { "grad_norm": 0.14529559016227722, "learning_rate": 1.3484857245934506e-05, "loss": 0.0036, "step": 77260 }, { "grad_norm": 0.17070665955543518, "learning_rate": 1.3473564001585248e-05, "loss": 0.0042, "step": 77270 }, { "grad_norm": 0.14434772729873657, "learning_rate": 1.3462274751712129e-05, "loss": 0.0031, "step": 77280 }, { "grad_norm": 0.08234110474586487, "learning_rate": 1.3450989497549698e-05, "loss": 0.004, "step": 77290 }, { "grad_norm": 0.16390061378479004, "learning_rate": 1.3439708240332115e-05, "loss": 0.005, "step": 77300 }, { "grad_norm": 0.11646008491516113, "learning_rate": 1.3428430981293043e-05, "loss": 0.0041, "step": 77310 }, { "grad_norm": 0.16187265515327454, "learning_rate": 1.3417157721665808e-05, "loss": 0.0035, "step": 77320 }, { "grad_norm": 0.15231075882911682, "learning_rate": 1.3405888462683186e-05, "loss": 0.0035, "step": 77330 }, { "grad_norm": 0.10324560105800629, "learning_rate": 1.3394623205577595e-05, "loss": 0.0031, "step": 77340 }, { "grad_norm": 0.11416297405958176, "learning_rate": 1.3383361951580963e-05, "loss": 0.0041, "step": 77350 }, { "grad_norm": 0.1287793666124344, "learning_rate": 1.337210470192483e-05, "loss": 0.0036, "step": 77360 }, { "grad_norm": 0.12267296016216278, "learning_rate": 1.3360851457840245e-05, "loss": 0.0033, "step": 77370 }, { "grad_norm": 0.12449076026678085, "learning_rate": 1.3349602220557856e-05, "loss": 0.0035, "step": 77380 }, { "grad_norm": 0.08131015300750732, "learning_rate": 1.3338356991307865e-05, "loss": 0.0053, "step": 77390 }, { "grad_norm": 0.08123630285263062, "learning_rate": 1.3327115771320048e-05, "loss": 0.0052, "step": 77400 }, { "grad_norm": 0.06655960530042648, "learning_rate": 1.3315878561823697e-05, "loss": 0.0046, "step": 77410 }, { "grad_norm": 0.1349460482597351, "learning_rate": 1.330464536404773e-05, "loss": 0.0041, "step": 77420 }, { "grad_norm": 0.08891482651233673, "learning_rate": 1.3293416179220563e-05, "loss": 0.0049, "step": 77430 }, { "grad_norm": 0.12583769857883453, "learning_rate": 1.328219100857021e-05, "loss": 0.0038, "step": 77440 }, { "grad_norm": 0.10296371579170227, "learning_rate": 1.327096985332425e-05, "loss": 0.0031, "step": 77450 }, { "grad_norm": 0.05979745462536812, "learning_rate": 1.3259752714709816e-05, "loss": 0.0029, "step": 77460 }, { "grad_norm": 0.15709388256072998, "learning_rate": 1.3248539593953569e-05, "loss": 0.0077, "step": 77470 }, { "grad_norm": 0.11931859701871872, "learning_rate": 1.323733049228178e-05, "loss": 0.0032, "step": 77480 }, { "grad_norm": 0.10736209154129028, "learning_rate": 1.3226125410920276e-05, "loss": 0.0042, "step": 77490 }, { "grad_norm": 0.1402117758989334, "learning_rate": 1.3214924351094388e-05, "loss": 0.0052, "step": 77500 }, { "grad_norm": 0.13326574862003326, "learning_rate": 1.3203727314029069e-05, "loss": 0.0033, "step": 77510 }, { "grad_norm": 0.07085131853818893, "learning_rate": 1.3192534300948811e-05, "loss": 0.0031, "step": 77520 }, { "grad_norm": 0.10556729137897491, "learning_rate": 1.3181345313077675e-05, "loss": 0.0032, "step": 77530 }, { "grad_norm": 0.08053689450025558, "learning_rate": 1.3170160351639243e-05, "loss": 0.0052, "step": 77540 }, { "grad_norm": 0.08700897544622421, "learning_rate": 1.3158979417856721e-05, "loss": 0.0042, "step": 77550 }, { "grad_norm": 0.1630600243806839, "learning_rate": 1.31478025129528e-05, "loss": 0.0047, "step": 77560 }, { "grad_norm": 0.1382155567407608, "learning_rate": 1.3136629638149794e-05, "loss": 0.0036, "step": 77570 }, { "grad_norm": 0.0861920639872551, "learning_rate": 1.3125460794669547e-05, "loss": 0.0073, "step": 77580 }, { "grad_norm": 0.10711530596017838, "learning_rate": 1.3114295983733483e-05, "loss": 0.0037, "step": 77590 }, { "grad_norm": 0.09529254585504532, "learning_rate": 1.3103135206562534e-05, "loss": 0.0069, "step": 77600 }, { "grad_norm": 0.06453339010477066, "learning_rate": 1.3091978464377264e-05, "loss": 0.0062, "step": 77610 }, { "grad_norm": 0.17420291900634766, "learning_rate": 1.3080825758397724e-05, "loss": 0.006, "step": 77620 }, { "grad_norm": 0.11934878677129745, "learning_rate": 1.3069677089843568e-05, "loss": 0.0031, "step": 77630 }, { "grad_norm": 0.10808655619621277, "learning_rate": 1.305853245993401e-05, "loss": 0.0042, "step": 77640 }, { "grad_norm": 0.11208553612232208, "learning_rate": 1.3047391869887809e-05, "loss": 0.0033, "step": 77650 }, { "grad_norm": 0.1879415214061737, "learning_rate": 1.3036255320923263e-05, "loss": 0.0077, "step": 77660 }, { "grad_norm": 0.12252660095691681, "learning_rate": 1.3025122814258273e-05, "loss": 0.0027, "step": 77670 }, { "grad_norm": 0.10830583423376083, "learning_rate": 1.3013994351110249e-05, "loss": 0.0033, "step": 77680 }, { "grad_norm": 0.0889056921005249, "learning_rate": 1.3002869932696194e-05, "loss": 0.0032, "step": 77690 }, { "grad_norm": 0.10832775384187698, "learning_rate": 1.2991749560232664e-05, "loss": 0.0025, "step": 77700 }, { "grad_norm": 0.15704505145549774, "learning_rate": 1.2980633234935768e-05, "loss": 0.0028, "step": 77710 }, { "grad_norm": 0.13813675940036774, "learning_rate": 1.2969520958021153e-05, "loss": 0.0044, "step": 77720 }, { "grad_norm": 0.12050702422857285, "learning_rate": 1.295841273070405e-05, "loss": 0.0058, "step": 77730 }, { "grad_norm": 0.09786980599164963, "learning_rate": 1.2947308554199255e-05, "loss": 0.0043, "step": 77740 }, { "grad_norm": 0.15327651798725128, "learning_rate": 1.2936208429721075e-05, "loss": 0.0047, "step": 77750 }, { "grad_norm": 0.10553131997585297, "learning_rate": 1.2925112358483416e-05, "loss": 0.0049, "step": 77760 }, { "grad_norm": 0.1388179361820221, "learning_rate": 1.2914020341699728e-05, "loss": 0.0042, "step": 77770 }, { "grad_norm": 0.0891619622707367, "learning_rate": 1.2902932380583032e-05, "loss": 0.0034, "step": 77780 }, { "grad_norm": 0.11922384053468704, "learning_rate": 1.2891848476345863e-05, "loss": 0.0053, "step": 77790 }, { "grad_norm": 0.14505045115947723, "learning_rate": 1.2880768630200368e-05, "loss": 0.0033, "step": 77800 }, { "grad_norm": 0.07440382987260818, "learning_rate": 1.286969284335819e-05, "loss": 0.0046, "step": 77810 }, { "grad_norm": 0.0890616923570633, "learning_rate": 1.2858621117030578e-05, "loss": 0.0034, "step": 77820 }, { "grad_norm": 0.07011111080646515, "learning_rate": 1.2847553452428324e-05, "loss": 0.0034, "step": 77830 }, { "grad_norm": 0.12210285663604736, "learning_rate": 1.283648985076178e-05, "loss": 0.0044, "step": 77840 }, { "grad_norm": 0.10460149496793747, "learning_rate": 1.2825430313240817e-05, "loss": 0.0034, "step": 77850 }, { "grad_norm": 0.0812988132238388, "learning_rate": 1.281437484107491e-05, "loss": 0.0034, "step": 77860 }, { "grad_norm": 0.13629373908042908, "learning_rate": 1.2803323435473058e-05, "loss": 0.0053, "step": 77870 }, { "grad_norm": 0.10643793642520905, "learning_rate": 1.2792276097643841e-05, "loss": 0.005, "step": 77880 }, { "grad_norm": 0.10326030105352402, "learning_rate": 1.2781232828795342e-05, "loss": 0.0029, "step": 77890 }, { "grad_norm": 0.11355608701705933, "learning_rate": 1.2770193630135292e-05, "loss": 0.0046, "step": 77900 }, { "grad_norm": 0.1909584105014801, "learning_rate": 1.2759158502870877e-05, "loss": 0.0049, "step": 77910 }, { "grad_norm": 0.0788412019610405, "learning_rate": 1.2748127448208913e-05, "loss": 0.0034, "step": 77920 }, { "grad_norm": 0.10846599191427231, "learning_rate": 1.2737100467355706e-05, "loss": 0.0033, "step": 77930 }, { "grad_norm": 0.18665407598018646, "learning_rate": 1.2726077561517169e-05, "loss": 0.0039, "step": 77940 }, { "grad_norm": 0.09583204984664917, "learning_rate": 1.2715058731898755e-05, "loss": 0.0041, "step": 77950 }, { "grad_norm": 0.11029483377933502, "learning_rate": 1.2704043979705471e-05, "loss": 0.0032, "step": 77960 }, { "grad_norm": 0.2315102517604828, "learning_rate": 1.269303330614185e-05, "loss": 0.0058, "step": 77970 }, { "grad_norm": 0.17136463522911072, "learning_rate": 1.2682026712412015e-05, "loss": 0.0035, "step": 77980 }, { "grad_norm": 0.10233257710933685, "learning_rate": 1.2671024199719644e-05, "loss": 0.003, "step": 77990 }, { "grad_norm": 0.055314965546131134, "learning_rate": 1.2660025769267936e-05, "loss": 0.0033, "step": 78000 }, { "grad_norm": 0.12730856239795685, "learning_rate": 1.2649031422259667e-05, "loss": 0.0036, "step": 78010 }, { "grad_norm": 0.12585265934467316, "learning_rate": 1.2638041159897163e-05, "loss": 0.0046, "step": 78020 }, { "grad_norm": 0.13819095492362976, "learning_rate": 1.262705498338232e-05, "loss": 0.004, "step": 78030 }, { "grad_norm": 0.14544466137886047, "learning_rate": 1.2616072893916542e-05, "loss": 0.0043, "step": 78040 }, { "grad_norm": 0.13284897804260254, "learning_rate": 1.2605094892700836e-05, "loss": 0.0039, "step": 78050 }, { "grad_norm": 0.11957251280546188, "learning_rate": 1.259412098093572e-05, "loss": 0.0089, "step": 78060 }, { "grad_norm": 0.11944567412137985, "learning_rate": 1.258315115982131e-05, "loss": 0.004, "step": 78070 }, { "grad_norm": 0.19280724227428436, "learning_rate": 1.2572185430557203e-05, "loss": 0.0034, "step": 78080 }, { "grad_norm": 0.05251491442322731, "learning_rate": 1.2561223794342659e-05, "loss": 0.0055, "step": 78090 }, { "grad_norm": 0.09061411768198013, "learning_rate": 1.255026625237637e-05, "loss": 0.0037, "step": 78100 }, { "grad_norm": 0.08030953258275986, "learning_rate": 1.2539312805856678e-05, "loss": 0.0045, "step": 78110 }, { "grad_norm": 0.11380399018526077, "learning_rate": 1.25283634559814e-05, "loss": 0.0028, "step": 78120 }, { "grad_norm": 0.07885225862264633, "learning_rate": 1.2517418203947972e-05, "loss": 0.0032, "step": 78130 }, { "grad_norm": 0.08529531210660934, "learning_rate": 1.2506477050953308e-05, "loss": 0.0043, "step": 78140 }, { "grad_norm": 0.12499728053808212, "learning_rate": 1.2495539998193972e-05, "loss": 0.0032, "step": 78150 }, { "grad_norm": 0.1098850816488266, "learning_rate": 1.2484607046865975e-05, "loss": 0.0046, "step": 78160 }, { "grad_norm": 0.19257095456123352, "learning_rate": 1.2473678198164968e-05, "loss": 0.004, "step": 78170 }, { "grad_norm": 0.09207341820001602, "learning_rate": 1.2462753453286075e-05, "loss": 0.0043, "step": 78180 }, { "grad_norm": 0.058253563940525055, "learning_rate": 1.2451832813424031e-05, "loss": 0.0027, "step": 78190 }, { "grad_norm": 0.12154845893383026, "learning_rate": 1.2440916279773095e-05, "loss": 0.0027, "step": 78200 }, { "grad_norm": 0.12432615458965302, "learning_rate": 1.2430003853527101e-05, "loss": 0.0042, "step": 78210 }, { "grad_norm": 0.1462649255990982, "learning_rate": 1.2419095535879383e-05, "loss": 0.005, "step": 78220 }, { "grad_norm": 0.10997068136930466, "learning_rate": 1.2408191328022878e-05, "loss": 0.0053, "step": 78230 }, { "grad_norm": 0.10829761624336243, "learning_rate": 1.2397291231150066e-05, "loss": 0.0044, "step": 78240 }, { "grad_norm": 0.0844283476471901, "learning_rate": 1.2386395246452937e-05, "loss": 0.0041, "step": 78250 }, { "grad_norm": 0.12332886457443237, "learning_rate": 1.2375503375123083e-05, "loss": 0.0047, "step": 78260 }, { "grad_norm": 0.07864763587713242, "learning_rate": 1.2364615618351583e-05, "loss": 0.0026, "step": 78270 }, { "grad_norm": 0.08188987523317337, "learning_rate": 1.2353731977329169e-05, "loss": 0.0038, "step": 78280 }, { "grad_norm": 0.12737655639648438, "learning_rate": 1.2342852453246007e-05, "loss": 0.004, "step": 78290 }, { "grad_norm": 0.24362137913703918, "learning_rate": 1.2331977047291898e-05, "loss": 0.0042, "step": 78300 }, { "grad_norm": 0.088801309466362, "learning_rate": 1.2321105760656137e-05, "loss": 0.0031, "step": 78310 }, { "grad_norm": 0.09167657792568207, "learning_rate": 1.2310238594527613e-05, "loss": 0.0046, "step": 78320 }, { "grad_norm": 0.10741263628005981, "learning_rate": 1.2299375550094693e-05, "loss": 0.0033, "step": 78330 }, { "grad_norm": 0.11613868921995163, "learning_rate": 1.2288516628545421e-05, "loss": 0.0049, "step": 78340 }, { "grad_norm": 0.11276353150606155, "learning_rate": 1.2277661831067255e-05, "loss": 0.0034, "step": 78350 }, { "grad_norm": 0.12024834752082825, "learning_rate": 1.2266811158847286e-05, "loss": 0.005, "step": 78360 }, { "grad_norm": 0.1843898892402649, "learning_rate": 1.2255964613072107e-05, "loss": 0.0042, "step": 78370 }, { "grad_norm": 0.09516575932502747, "learning_rate": 1.2245122194927905e-05, "loss": 0.005, "step": 78380 }, { "grad_norm": 0.14101648330688477, "learning_rate": 1.2234283905600364e-05, "loss": 0.0041, "step": 78390 }, { "grad_norm": 0.12622138857841492, "learning_rate": 1.222344974627475e-05, "loss": 0.0068, "step": 78400 }, { "grad_norm": 0.13290999829769135, "learning_rate": 1.2212619718135875e-05, "loss": 0.0054, "step": 78410 }, { "grad_norm": 0.12344422191381454, "learning_rate": 1.2201793822368118e-05, "loss": 0.0035, "step": 78420 }, { "grad_norm": 0.12105172127485275, "learning_rate": 1.2190972060155337e-05, "loss": 0.0036, "step": 78430 }, { "grad_norm": 0.07556873559951782, "learning_rate": 1.2180154432681013e-05, "loss": 0.0033, "step": 78440 }, { "grad_norm": 0.09299574047327042, "learning_rate": 1.2169340941128148e-05, "loss": 0.0043, "step": 78450 }, { "grad_norm": 0.22994868457317352, "learning_rate": 1.2158531586679267e-05, "loss": 0.005, "step": 78460 }, { "grad_norm": 0.06867256760597229, "learning_rate": 1.2147726370516477e-05, "loss": 0.0027, "step": 78470 }, { "grad_norm": 0.15505509078502655, "learning_rate": 1.2136925293821422e-05, "loss": 0.0029, "step": 78480 }, { "grad_norm": 0.05353221669793129, "learning_rate": 1.2126128357775306e-05, "loss": 0.0027, "step": 78490 }, { "grad_norm": 0.07289476692676544, "learning_rate": 1.2115335563558838e-05, "loss": 0.0043, "step": 78500 }, { "grad_norm": 0.13012826442718506, "learning_rate": 1.2104546912352327e-05, "loss": 0.0041, "step": 78510 }, { "grad_norm": 0.15486761927604675, "learning_rate": 1.2093762405335557e-05, "loss": 0.0054, "step": 78520 }, { "grad_norm": 0.15494905412197113, "learning_rate": 1.2082982043687974e-05, "loss": 0.0035, "step": 78530 }, { "grad_norm": 0.1404842883348465, "learning_rate": 1.207220582858845e-05, "loss": 0.0035, "step": 78540 }, { "grad_norm": 0.09097137302160263, "learning_rate": 1.206143376121549e-05, "loss": 0.0046, "step": 78550 }, { "grad_norm": 0.10472392290830612, "learning_rate": 1.2050665842747078e-05, "loss": 0.0041, "step": 78560 }, { "grad_norm": 0.20288026332855225, "learning_rate": 1.2039902074360804e-05, "loss": 0.0066, "step": 78570 }, { "grad_norm": 0.08074217289686203, "learning_rate": 1.2029142457233755e-05, "loss": 0.0054, "step": 78580 }, { "grad_norm": 0.12057527154684067, "learning_rate": 1.2018386992542601e-05, "loss": 0.0043, "step": 78590 }, { "grad_norm": 0.06414493173360825, "learning_rate": 1.2007635681463542e-05, "loss": 0.0041, "step": 78600 }, { "grad_norm": 0.09253256022930145, "learning_rate": 1.1996888525172334e-05, "loss": 0.0032, "step": 78610 }, { "grad_norm": 0.13782799243927002, "learning_rate": 1.198614552484425e-05, "loss": 0.0028, "step": 78620 }, { "grad_norm": 0.11270835995674133, "learning_rate": 1.197540668165415e-05, "loss": 0.0032, "step": 78630 }, { "grad_norm": 0.1418779045343399, "learning_rate": 1.1964671996776395e-05, "loss": 0.0026, "step": 78640 }, { "grad_norm": 0.13412684202194214, "learning_rate": 1.1953941471384922e-05, "loss": 0.0034, "step": 78650 }, { "grad_norm": 0.07553649693727493, "learning_rate": 1.1943215106653205e-05, "loss": 0.0031, "step": 78660 }, { "grad_norm": 0.04289562255144119, "learning_rate": 1.193249290375429e-05, "loss": 0.0037, "step": 78670 }, { "grad_norm": 0.12017403542995453, "learning_rate": 1.1921774863860696e-05, "loss": 0.0038, "step": 78680 }, { "grad_norm": 0.12384487688541412, "learning_rate": 1.1911060988144556e-05, "loss": 0.0034, "step": 78690 }, { "grad_norm": 0.08497379720211029, "learning_rate": 1.1900351277777533e-05, "loss": 0.0036, "step": 78700 }, { "grad_norm": 0.07165880501270294, "learning_rate": 1.1889645733930798e-05, "loss": 0.0027, "step": 78710 }, { "grad_norm": 0.08105740696191788, "learning_rate": 1.1878944357775107e-05, "loss": 0.0061, "step": 78720 }, { "grad_norm": 0.09885044395923615, "learning_rate": 1.1868247150480749e-05, "loss": 0.0036, "step": 78730 }, { "grad_norm": 0.11852908879518509, "learning_rate": 1.1857554113217567e-05, "loss": 0.0049, "step": 78740 }, { "grad_norm": 0.21694065630435944, "learning_rate": 1.1846865247154903e-05, "loss": 0.0046, "step": 78750 }, { "grad_norm": 0.0886978879570961, "learning_rate": 1.1836180553461711e-05, "loss": 0.0023, "step": 78760 }, { "grad_norm": 0.2866148054599762, "learning_rate": 1.1825500033306419e-05, "loss": 0.0059, "step": 78770 }, { "grad_norm": 0.1865861415863037, "learning_rate": 1.1814823687857052e-05, "loss": 0.005, "step": 78780 }, { "grad_norm": 0.12053247541189194, "learning_rate": 1.1804151518281153e-05, "loss": 0.0024, "step": 78790 }, { "grad_norm": 0.1030297502875328, "learning_rate": 1.179348352574583e-05, "loss": 0.0028, "step": 78800 }, { "grad_norm": 0.08341579139232635, "learning_rate": 1.1782819711417697e-05, "loss": 0.0045, "step": 78810 }, { "grad_norm": 0.1009315475821495, "learning_rate": 1.1772160076462951e-05, "loss": 0.0025, "step": 78820 }, { "grad_norm": 0.08306997269392014, "learning_rate": 1.1761504622047293e-05, "loss": 0.0029, "step": 78830 }, { "grad_norm": 0.11098434776067734, "learning_rate": 1.1750853349335999e-05, "loss": 0.0043, "step": 78840 }, { "grad_norm": 0.15791882574558258, "learning_rate": 1.1740206259493874e-05, "loss": 0.0027, "step": 78850 }, { "grad_norm": 0.11509793251752853, "learning_rate": 1.1729563353685286e-05, "loss": 0.0057, "step": 78860 }, { "grad_norm": 0.06851843744516373, "learning_rate": 1.1718924633074091e-05, "loss": 0.0028, "step": 78870 }, { "grad_norm": 0.08771979063749313, "learning_rate": 1.1708290098823765e-05, "loss": 0.0034, "step": 78880 }, { "grad_norm": 0.10390956699848175, "learning_rate": 1.1697659752097245e-05, "loss": 0.0046, "step": 78890 }, { "grad_norm": 0.15464743971824646, "learning_rate": 1.1687033594057068e-05, "loss": 0.004, "step": 78900 }, { "grad_norm": 0.08332914859056473, "learning_rate": 1.1676411625865297e-05, "loss": 0.0031, "step": 78910 }, { "grad_norm": 0.12170197814702988, "learning_rate": 1.1665793848683542e-05, "loss": 0.004, "step": 78920 }, { "grad_norm": 0.10358671098947525, "learning_rate": 1.1655180263672927e-05, "loss": 0.0078, "step": 78930 }, { "grad_norm": 0.10839106142520905, "learning_rate": 1.1644570871994142e-05, "loss": 0.0065, "step": 78940 }, { "grad_norm": 0.10686599463224411, "learning_rate": 1.163396567480744e-05, "loss": 0.0042, "step": 78950 }, { "grad_norm": 0.08160672336816788, "learning_rate": 1.1623364673272552e-05, "loss": 0.003, "step": 78960 }, { "grad_norm": 0.08006078749895096, "learning_rate": 1.1612767868548807e-05, "loss": 0.0035, "step": 78970 }, { "grad_norm": 0.09637409448623657, "learning_rate": 1.1602175261795057e-05, "loss": 0.0027, "step": 78980 }, { "grad_norm": 0.1254757046699524, "learning_rate": 1.1591586854169696e-05, "loss": 0.002, "step": 78990 }, { "grad_norm": 0.090090312063694, "learning_rate": 1.158100264683064e-05, "loss": 0.0047, "step": 79000 }, { "grad_norm": 0.08734862506389618, "learning_rate": 1.1570422640935386e-05, "loss": 0.0023, "step": 79010 }, { "grad_norm": 0.06328096985816956, "learning_rate": 1.155984683764092e-05, "loss": 0.0048, "step": 79020 }, { "grad_norm": 0.09918750822544098, "learning_rate": 1.1549275238103812e-05, "loss": 0.004, "step": 79030 }, { "grad_norm": 0.089544877409935, "learning_rate": 1.1538707843480152e-05, "loss": 0.0044, "step": 79040 }, { "grad_norm": 0.15641361474990845, "learning_rate": 1.1528144654925593e-05, "loss": 0.0055, "step": 79050 }, { "grad_norm": 0.10784242302179337, "learning_rate": 1.1517585673595277e-05, "loss": 0.0042, "step": 79060 }, { "grad_norm": 0.0623735636472702, "learning_rate": 1.150703090064395e-05, "loss": 0.0029, "step": 79070 }, { "grad_norm": 0.15380604565143585, "learning_rate": 1.1496480337225835e-05, "loss": 0.0038, "step": 79080 }, { "grad_norm": 0.08177772164344788, "learning_rate": 1.1485933984494751e-05, "loss": 0.0035, "step": 79090 }, { "grad_norm": 0.07046306878328323, "learning_rate": 1.1475391843604e-05, "loss": 0.0034, "step": 79100 }, { "grad_norm": 0.08191192895174026, "learning_rate": 1.1464853915706503e-05, "loss": 0.0065, "step": 79110 }, { "grad_norm": 0.1587974727153778, "learning_rate": 1.1454320201954626e-05, "loss": 0.0027, "step": 79120 }, { "grad_norm": 0.0926685631275177, "learning_rate": 1.1443790703500356e-05, "loss": 0.0042, "step": 79130 }, { "grad_norm": 0.1613403558731079, "learning_rate": 1.1433265421495154e-05, "loss": 0.0055, "step": 79140 }, { "grad_norm": 0.08279422670602798, "learning_rate": 1.1422744357090065e-05, "loss": 0.0042, "step": 79150 }, { "grad_norm": 0.3086225986480713, "learning_rate": 1.1412227511435647e-05, "loss": 0.0043, "step": 79160 }, { "grad_norm": 0.10147164016962051, "learning_rate": 1.1401714885682025e-05, "loss": 0.0027, "step": 79170 }, { "grad_norm": 0.1337011456489563, "learning_rate": 1.1391206480978823e-05, "loss": 0.0032, "step": 79180 }, { "grad_norm": 0.125499427318573, "learning_rate": 1.1380702298475227e-05, "loss": 0.004, "step": 79190 }, { "grad_norm": 0.14413483440876007, "learning_rate": 1.1370202339319985e-05, "loss": 0.0039, "step": 79200 }, { "grad_norm": 0.09011676907539368, "learning_rate": 1.1359706604661319e-05, "loss": 0.005, "step": 79210 }, { "grad_norm": 0.07741501182317734, "learning_rate": 1.1349215095647042e-05, "loss": 0.0047, "step": 79220 }, { "grad_norm": 0.1511210799217224, "learning_rate": 1.1338727813424493e-05, "loss": 0.0036, "step": 79230 }, { "grad_norm": 0.049683813005685806, "learning_rate": 1.132824475914056e-05, "loss": 0.003, "step": 79240 }, { "grad_norm": 0.08789700269699097, "learning_rate": 1.131776593394162e-05, "loss": 0.0039, "step": 79250 }, { "grad_norm": 0.12922507524490356, "learning_rate": 1.1307291338973657e-05, "loss": 0.0045, "step": 79260 }, { "grad_norm": 0.12468075752258301, "learning_rate": 1.1296820975382121e-05, "loss": 0.0035, "step": 79270 }, { "grad_norm": 0.06560339033603668, "learning_rate": 1.1286354844312074e-05, "loss": 0.0031, "step": 79280 }, { "grad_norm": 0.07944752275943756, "learning_rate": 1.1275892946908023e-05, "loss": 0.0034, "step": 79290 }, { "grad_norm": 0.26217904686927795, "learning_rate": 1.126543528431413e-05, "loss": 0.0056, "step": 79300 }, { "grad_norm": 0.08328374475240707, "learning_rate": 1.125498185767398e-05, "loss": 0.0057, "step": 79310 }, { "grad_norm": 0.11716878414154053, "learning_rate": 1.1244532668130781e-05, "loss": 0.005, "step": 79320 }, { "grad_norm": 0.13609515130519867, "learning_rate": 1.1234087716827203e-05, "loss": 0.004, "step": 79330 }, { "grad_norm": 0.0938011109828949, "learning_rate": 1.1223647004905524e-05, "loss": 0.0033, "step": 79340 }, { "grad_norm": 0.09651768207550049, "learning_rate": 1.1213210533507485e-05, "loss": 0.005, "step": 79350 }, { "grad_norm": 0.2222820520401001, "learning_rate": 1.1202778303774447e-05, "loss": 0.0032, "step": 79360 }, { "grad_norm": 0.12799884378910065, "learning_rate": 1.1192350316847234e-05, "loss": 0.0041, "step": 79370 }, { "grad_norm": 0.20152249932289124, "learning_rate": 1.1181926573866258e-05, "loss": 0.0051, "step": 79380 }, { "grad_norm": 0.1165679469704628, "learning_rate": 1.1171507075971416e-05, "loss": 0.0027, "step": 79390 }, { "grad_norm": 0.14887428283691406, "learning_rate": 1.1161091824302177e-05, "loss": 0.0051, "step": 79400 }, { "grad_norm": 0.1796470731496811, "learning_rate": 1.115068081999755e-05, "loss": 0.0048, "step": 79410 }, { "grad_norm": 0.19467377662658691, "learning_rate": 1.1140274064196071e-05, "loss": 0.0035, "step": 79420 }, { "grad_norm": 0.10814373195171356, "learning_rate": 1.112987155803578e-05, "loss": 0.006, "step": 79430 }, { "grad_norm": 0.10996866971254349, "learning_rate": 1.1119473302654298e-05, "loss": 0.0028, "step": 79440 }, { "grad_norm": 0.1921953409910202, "learning_rate": 1.1109079299188769e-05, "loss": 0.0036, "step": 79450 }, { "grad_norm": 0.08791518211364746, "learning_rate": 1.1098689548775847e-05, "loss": 0.0034, "step": 79460 }, { "grad_norm": 0.18902643024921417, "learning_rate": 1.1088304052551762e-05, "loss": 0.0047, "step": 79470 }, { "grad_norm": 0.17343267798423767, "learning_rate": 1.1077922811652215e-05, "loss": 0.0036, "step": 79480 }, { "grad_norm": 0.11542883515357971, "learning_rate": 1.1067545827212534e-05, "loss": 0.0042, "step": 79490 }, { "grad_norm": 0.0830584466457367, "learning_rate": 1.1057173100367496e-05, "loss": 0.0039, "step": 79500 }, { "grad_norm": 0.3092349171638489, "learning_rate": 1.1046804632251474e-05, "loss": 0.0041, "step": 79510 }, { "grad_norm": 0.08010753244161606, "learning_rate": 1.1036440423998318e-05, "loss": 0.0031, "step": 79520 }, { "grad_norm": 0.12938274443149567, "learning_rate": 1.1026080476741469e-05, "loss": 0.0045, "step": 79530 }, { "grad_norm": 0.13049820065498352, "learning_rate": 1.1015724791613836e-05, "loss": 0.0051, "step": 79540 }, { "grad_norm": 0.053832996636629105, "learning_rate": 1.1005373369747951e-05, "loss": 0.0029, "step": 79550 }, { "grad_norm": 0.07693225890398026, "learning_rate": 1.09950262122758e-05, "loss": 0.0049, "step": 79560 }, { "grad_norm": 0.0954306498169899, "learning_rate": 1.0984683320328948e-05, "loss": 0.0036, "step": 79570 }, { "grad_norm": 0.09939320385456085, "learning_rate": 1.0974344695038458e-05, "loss": 0.0025, "step": 79580 }, { "grad_norm": 0.23276394605636597, "learning_rate": 1.0964010337534968e-05, "loss": 0.0036, "step": 79590 }, { "grad_norm": 0.11317809671163559, "learning_rate": 1.0953680248948611e-05, "loss": 0.0037, "step": 79600 }, { "grad_norm": 0.08817215263843536, "learning_rate": 1.0943354430409069e-05, "loss": 0.003, "step": 79610 }, { "grad_norm": 0.13406217098236084, "learning_rate": 1.0933032883045574e-05, "loss": 0.0049, "step": 79620 }, { "grad_norm": 0.12273978441953659, "learning_rate": 1.0922715607986872e-05, "loss": 0.0036, "step": 79630 }, { "grad_norm": 0.0860995426774025, "learning_rate": 1.0912402606361234e-05, "loss": 0.0027, "step": 79640 }, { "grad_norm": 0.14117074012756348, "learning_rate": 1.090209387929647e-05, "loss": 0.0042, "step": 79650 }, { "grad_norm": 0.1230582445859909, "learning_rate": 1.0891789427919957e-05, "loss": 0.0041, "step": 79660 }, { "grad_norm": 0.079294353723526, "learning_rate": 1.088148925335854e-05, "loss": 0.0026, "step": 79670 }, { "grad_norm": 0.09893196076154709, "learning_rate": 1.087119335673864e-05, "loss": 0.0041, "step": 79680 }, { "grad_norm": 0.1232108622789383, "learning_rate": 1.0860901739186208e-05, "loss": 0.0048, "step": 79690 }, { "grad_norm": 0.12176123261451721, "learning_rate": 1.085061440182673e-05, "loss": 0.004, "step": 79700 }, { "grad_norm": 0.1110648512840271, "learning_rate": 1.0840331345785193e-05, "loss": 0.0037, "step": 79710 }, { "grad_norm": 0.09230968356132507, "learning_rate": 1.083005257218615e-05, "loss": 0.004, "step": 79720 }, { "grad_norm": 0.04161534830927849, "learning_rate": 1.0819778082153642e-05, "loss": 0.0033, "step": 79730 }, { "grad_norm": 0.09203929454088211, "learning_rate": 1.0809507876811325e-05, "loss": 0.005, "step": 79740 }, { "grad_norm": 0.16691777110099792, "learning_rate": 1.0799241957282292e-05, "loss": 0.0041, "step": 79750 }, { "grad_norm": 0.1665607988834381, "learning_rate": 1.078898032468923e-05, "loss": 0.004, "step": 79760 }, { "grad_norm": 0.08610966801643372, "learning_rate": 1.0778722980154315e-05, "loss": 0.0053, "step": 79770 }, { "grad_norm": 0.12312226742506027, "learning_rate": 1.0768469924799302e-05, "loss": 0.0044, "step": 79780 }, { "grad_norm": 0.11269199103116989, "learning_rate": 1.0758221159745419e-05, "loss": 0.0032, "step": 79790 }, { "grad_norm": 0.10593114793300629, "learning_rate": 1.074797668611347e-05, "loss": 0.003, "step": 79800 }, { "grad_norm": 0.10608579218387604, "learning_rate": 1.0737736505023777e-05, "loss": 0.0044, "step": 79810 }, { "grad_norm": 0.11738818138837814, "learning_rate": 1.0727500617596203e-05, "loss": 0.0044, "step": 79820 }, { "grad_norm": 0.09065189212560654, "learning_rate": 1.0717269024950104e-05, "loss": 0.0047, "step": 79830 }, { "grad_norm": 0.12332657724618912, "learning_rate": 1.070704172820442e-05, "loss": 0.0033, "step": 79840 }, { "grad_norm": 0.08711962401866913, "learning_rate": 1.0696818728477559e-05, "loss": 0.0029, "step": 79850 }, { "grad_norm": 0.08452321588993073, "learning_rate": 1.0686600026887511e-05, "loss": 0.005, "step": 79860 }, { "grad_norm": 0.1530192643404007, "learning_rate": 1.0676385624551777e-05, "loss": 0.0042, "step": 79870 }, { "grad_norm": 0.11341656744480133, "learning_rate": 1.0666175522587402e-05, "loss": 0.0048, "step": 79880 }, { "grad_norm": 0.09511221945285797, "learning_rate": 1.0655969722110925e-05, "loss": 0.0034, "step": 79890 }, { "grad_norm": 0.15990708768367767, "learning_rate": 1.0645768224238439e-05, "loss": 0.0042, "step": 79900 }, { "grad_norm": 0.084876149892807, "learning_rate": 1.063557103008559e-05, "loss": 0.003, "step": 79910 }, { "grad_norm": 0.11784721910953522, "learning_rate": 1.0625378140767494e-05, "loss": 0.0038, "step": 79920 }, { "grad_norm": 0.1286875158548355, "learning_rate": 1.0615189557398846e-05, "loss": 0.0047, "step": 79930 }, { "grad_norm": 0.12088554352521896, "learning_rate": 1.060500528109386e-05, "loss": 0.0035, "step": 79940 }, { "grad_norm": 0.13762107491493225, "learning_rate": 1.0594825312966272e-05, "loss": 0.0042, "step": 79950 }, { "grad_norm": 0.09547898173332214, "learning_rate": 1.0584649654129336e-05, "loss": 0.0047, "step": 79960 }, { "grad_norm": 0.08627932518720627, "learning_rate": 1.0574478305695863e-05, "loss": 0.0039, "step": 79970 }, { "grad_norm": 0.12106577306985855, "learning_rate": 1.0564311268778154e-05, "loss": 0.0035, "step": 79980 }, { "grad_norm": 0.12638206779956818, "learning_rate": 1.0554148544488073e-05, "loss": 0.0043, "step": 79990 }, { "grad_norm": 0.15515606105327606, "learning_rate": 1.0543990133937004e-05, "loss": 0.004, "step": 80000 }, { "grad_norm": 0.12147919088602066, "learning_rate": 1.0533836038235868e-05, "loss": 0.0051, "step": 80010 }, { "grad_norm": 0.07590822875499725, "learning_rate": 1.052368625849507e-05, "loss": 0.0036, "step": 80020 }, { "grad_norm": 0.0937921404838562, "learning_rate": 1.05135407958246e-05, "loss": 0.004, "step": 80030 }, { "grad_norm": 0.08866564929485321, "learning_rate": 1.0503399651333934e-05, "loss": 0.0045, "step": 80040 }, { "grad_norm": 0.15269455313682556, "learning_rate": 1.04932628261321e-05, "loss": 0.0038, "step": 80050 }, { "grad_norm": 0.10522153973579407, "learning_rate": 1.0483130321327644e-05, "loss": 0.0033, "step": 80060 }, { "grad_norm": 0.11988537013530731, "learning_rate": 1.0473002138028654e-05, "loss": 0.0027, "step": 80070 }, { "grad_norm": 0.11747787147760391, "learning_rate": 1.0462878277342713e-05, "loss": 0.0031, "step": 80080 }, { "grad_norm": 0.13336944580078125, "learning_rate": 1.0452758740376972e-05, "loss": 0.0057, "step": 80090 }, { "grad_norm": 0.0740288719534874, "learning_rate": 1.0442643528238067e-05, "loss": 0.0032, "step": 80100 }, { "grad_norm": 0.09399433434009552, "learning_rate": 1.0432532642032188e-05, "loss": 0.0034, "step": 80110 }, { "grad_norm": 0.11132555454969406, "learning_rate": 1.042242608286505e-05, "loss": 0.0027, "step": 80120 }, { "grad_norm": 0.13082890212535858, "learning_rate": 1.041232385184191e-05, "loss": 0.004, "step": 80130 }, { "grad_norm": 0.09501571208238602, "learning_rate": 1.0402225950067495e-05, "loss": 0.0045, "step": 80140 }, { "grad_norm": 0.13346317410469055, "learning_rate": 1.0392132378646119e-05, "loss": 0.0045, "step": 80150 }, { "grad_norm": 0.11833155155181885, "learning_rate": 1.0382043138681607e-05, "loss": 0.0024, "step": 80160 }, { "grad_norm": 0.14105485379695892, "learning_rate": 1.0371958231277284e-05, "loss": 0.0038, "step": 80170 }, { "grad_norm": 0.0675247460603714, "learning_rate": 1.0361877657536024e-05, "loss": 0.0031, "step": 80180 }, { "grad_norm": 0.09687498211860657, "learning_rate": 1.035180141856023e-05, "loss": 0.005, "step": 80190 }, { "grad_norm": 0.09036502987146378, "learning_rate": 1.0341729515451836e-05, "loss": 0.0036, "step": 80200 }, { "grad_norm": 0.10374952107667923, "learning_rate": 1.0331661949312265e-05, "loss": 0.0027, "step": 80210 }, { "grad_norm": 0.14713434875011444, "learning_rate": 1.0321598721242514e-05, "loss": 0.0038, "step": 80220 }, { "grad_norm": 0.07725553214550018, "learning_rate": 1.0311539832343064e-05, "loss": 0.0031, "step": 80230 }, { "grad_norm": 0.14493277668952942, "learning_rate": 1.0301485283713942e-05, "loss": 0.0041, "step": 80240 }, { "grad_norm": 0.11521246284246445, "learning_rate": 1.0291435076454703e-05, "loss": 0.0027, "step": 80250 }, { "grad_norm": 0.08825612813234329, "learning_rate": 1.028138921166444e-05, "loss": 0.0027, "step": 80260 }, { "grad_norm": 0.13526701927185059, "learning_rate": 1.0271347690441719e-05, "loss": 0.0057, "step": 80270 }, { "grad_norm": 0.07656583189964294, "learning_rate": 1.0261310513884698e-05, "loss": 0.0039, "step": 80280 }, { "grad_norm": 0.07586728781461716, "learning_rate": 1.0251277683090998e-05, "loss": 0.0052, "step": 80290 }, { "grad_norm": 0.1055072769522667, "learning_rate": 1.0241249199157826e-05, "loss": 0.0032, "step": 80300 }, { "grad_norm": 0.09375981241464615, "learning_rate": 1.0231225063181837e-05, "loss": 0.0032, "step": 80310 }, { "grad_norm": 0.09434811025857925, "learning_rate": 1.0221205276259316e-05, "loss": 0.0049, "step": 80320 }, { "grad_norm": 0.1820243000984192, "learning_rate": 1.0211189839485958e-05, "loss": 0.003, "step": 80330 }, { "grad_norm": 0.15429939329624176, "learning_rate": 1.0201178753957075e-05, "loss": 0.0058, "step": 80340 }, { "grad_norm": 0.10003774613142014, "learning_rate": 1.0191172020767432e-05, "loss": 0.0038, "step": 80350 }, { "grad_norm": 0.06910580396652222, "learning_rate": 1.0181169641011362e-05, "loss": 0.0038, "step": 80360 }, { "grad_norm": 0.10195465385913849, "learning_rate": 1.0171171615782721e-05, "loss": 0.0029, "step": 80370 }, { "grad_norm": 0.1986168622970581, "learning_rate": 1.0161177946174866e-05, "loss": 0.003, "step": 80380 }, { "grad_norm": 0.08823033422231674, "learning_rate": 1.0151188633280706e-05, "loss": 0.0031, "step": 80390 }, { "grad_norm": 0.09810147434473038, "learning_rate": 1.0141203678192635e-05, "loss": 0.0035, "step": 80400 }, { "grad_norm": 0.13593608140945435, "learning_rate": 1.0131223082002612e-05, "loss": 0.0041, "step": 80410 }, { "grad_norm": 0.13184568285942078, "learning_rate": 1.012124684580208e-05, "loss": 0.0033, "step": 80420 }, { "grad_norm": 0.09019463509321213, "learning_rate": 1.0111274970682039e-05, "loss": 0.0036, "step": 80430 }, { "grad_norm": 0.04458237066864967, "learning_rate": 1.010130745773299e-05, "loss": 0.0034, "step": 80440 }, { "grad_norm": 0.1342093050479889, "learning_rate": 1.0091344308044986e-05, "loss": 0.0033, "step": 80450 }, { "grad_norm": 0.1423124074935913, "learning_rate": 1.0081385522707554e-05, "loss": 0.0031, "step": 80460 }, { "grad_norm": 0.1642480194568634, "learning_rate": 1.0071431102809803e-05, "loss": 0.0038, "step": 80470 }, { "grad_norm": 0.10758435726165771, "learning_rate": 1.0061481049440297e-05, "loss": 0.0049, "step": 80480 }, { "grad_norm": 0.12544962763786316, "learning_rate": 1.0051535363687187e-05, "loss": 0.0068, "step": 80490 }, { "grad_norm": 0.06257496029138565, "learning_rate": 1.0041594046638087e-05, "loss": 0.0049, "step": 80500 }, { "grad_norm": 0.1770443320274353, "learning_rate": 1.0031657099380204e-05, "loss": 0.0034, "step": 80510 }, { "grad_norm": 0.09695564210414886, "learning_rate": 1.0021724523000204e-05, "loss": 0.0056, "step": 80520 }, { "grad_norm": 0.10214580595493317, "learning_rate": 1.001179631858431e-05, "loss": 0.0032, "step": 80530 }, { "grad_norm": 0.1366969496011734, "learning_rate": 1.0001872487218245e-05, "loss": 0.0031, "step": 80540 }, { "grad_norm": 0.22239938378334045, "learning_rate": 9.991953029987272e-06, "loss": 0.0037, "step": 80550 }, { "grad_norm": 0.10403753817081451, "learning_rate": 9.982037947976147e-06, "loss": 0.0034, "step": 80560 }, { "grad_norm": 0.21079549193382263, "learning_rate": 9.972127242269208e-06, "loss": 0.0062, "step": 80570 }, { "grad_norm": 0.2608744502067566, "learning_rate": 9.962220913950238e-06, "loss": 0.0066, "step": 80580 }, { "grad_norm": 0.10657057166099548, "learning_rate": 9.952318964102591e-06, "loss": 0.0036, "step": 80590 }, { "grad_norm": 0.09687908738851547, "learning_rate": 9.942421393809148e-06, "loss": 0.0033, "step": 80600 }, { "grad_norm": 0.28507867455482483, "learning_rate": 9.932528204152264e-06, "loss": 0.0031, "step": 80610 }, { "grad_norm": 0.11901699006557465, "learning_rate": 9.922639396213856e-06, "loss": 0.0045, "step": 80620 }, { "grad_norm": 0.05117171257734299, "learning_rate": 9.912754971075344e-06, "loss": 0.0039, "step": 80630 }, { "grad_norm": 0.09728718549013138, "learning_rate": 9.902874929817696e-06, "loss": 0.0042, "step": 80640 }, { "grad_norm": 0.10061751306056976, "learning_rate": 9.892999273521342e-06, "loss": 0.0033, "step": 80650 }, { "grad_norm": 0.08954387158155441, "learning_rate": 9.883128003266302e-06, "loss": 0.0037, "step": 80660 }, { "grad_norm": 0.10040194541215897, "learning_rate": 9.87326112013206e-06, "loss": 0.0039, "step": 80670 }, { "grad_norm": 0.13352587819099426, "learning_rate": 9.863398625197662e-06, "loss": 0.0046, "step": 80680 }, { "grad_norm": 0.08940327167510986, "learning_rate": 9.853540519541615e-06, "loss": 0.003, "step": 80690 }, { "grad_norm": 0.1040831059217453, "learning_rate": 9.843686804242053e-06, "loss": 0.0039, "step": 80700 }, { "grad_norm": 0.15288761258125305, "learning_rate": 9.833837480376506e-06, "loss": 0.0048, "step": 80710 }, { "grad_norm": 0.13628201186656952, "learning_rate": 9.82399254902212e-06, "loss": 0.0035, "step": 80720 }, { "grad_norm": 0.06801103055477142, "learning_rate": 9.814152011255494e-06, "loss": 0.0026, "step": 80730 }, { "grad_norm": 0.16006171703338623, "learning_rate": 9.804315868152797e-06, "loss": 0.0035, "step": 80740 }, { "grad_norm": 0.17624178528785706, "learning_rate": 9.79448412078966e-06, "loss": 0.0035, "step": 80750 }, { "grad_norm": 0.10150022804737091, "learning_rate": 9.784656770241312e-06, "loss": 0.0061, "step": 80760 }, { "grad_norm": 0.08494458347558975, "learning_rate": 9.77483381758243e-06, "loss": 0.0034, "step": 80770 }, { "grad_norm": 0.11493906378746033, "learning_rate": 9.76501526388725e-06, "loss": 0.0038, "step": 80780 }, { "grad_norm": 0.10881315171718597, "learning_rate": 9.7552011102295e-06, "loss": 0.003, "step": 80790 }, { "grad_norm": 0.1508631408214569, "learning_rate": 9.745391357682449e-06, "loss": 0.0029, "step": 80800 }, { "grad_norm": 0.06849001348018646, "learning_rate": 9.735586007318875e-06, "loss": 0.0032, "step": 80810 }, { "grad_norm": 0.20228159427642822, "learning_rate": 9.725785060211096e-06, "loss": 0.0041, "step": 80820 }, { "grad_norm": 0.08163496851921082, "learning_rate": 9.715988517430896e-06, "loss": 0.0049, "step": 80830 }, { "grad_norm": 0.12903718650341034, "learning_rate": 9.706196380049625e-06, "loss": 0.0064, "step": 80840 }, { "grad_norm": 0.09430189430713654, "learning_rate": 9.69640864913815e-06, "loss": 0.0041, "step": 80850 }, { "grad_norm": 0.12508659064769745, "learning_rate": 9.686625325766818e-06, "loss": 0.0041, "step": 80860 }, { "grad_norm": 0.12067711353302002, "learning_rate": 9.67684641100554e-06, "loss": 0.0053, "step": 80870 }, { "grad_norm": 0.11933194100856781, "learning_rate": 9.667071905923692e-06, "loss": 0.0023, "step": 80880 }, { "grad_norm": 0.13190041482448578, "learning_rate": 9.657301811590242e-06, "loss": 0.0031, "step": 80890 }, { "grad_norm": 0.11081229150295258, "learning_rate": 9.647536129073603e-06, "loss": 0.0041, "step": 80900 }, { "grad_norm": 0.11197306960821152, "learning_rate": 9.637774859441751e-06, "loss": 0.0028, "step": 80910 }, { "grad_norm": 0.08571261167526245, "learning_rate": 9.628018003762152e-06, "loss": 0.004, "step": 80920 }, { "grad_norm": 0.0905691608786583, "learning_rate": 9.618265563101813e-06, "loss": 0.0026, "step": 80930 }, { "grad_norm": 0.12058455497026443, "learning_rate": 9.608517538527211e-06, "loss": 0.0044, "step": 80940 }, { "grad_norm": 0.06381074339151382, "learning_rate": 9.598773931104431e-06, "loss": 0.0034, "step": 80950 }, { "grad_norm": 0.05260483920574188, "learning_rate": 9.589034741898983e-06, "loss": 0.0047, "step": 80960 }, { "grad_norm": 0.36656612157821655, "learning_rate": 9.579299971975947e-06, "loss": 0.004, "step": 80970 }, { "grad_norm": 0.15911634266376495, "learning_rate": 9.569569622399882e-06, "loss": 0.0031, "step": 80980 }, { "grad_norm": 0.0971560925245285, "learning_rate": 9.559843694234916e-06, "loss": 0.0042, "step": 80990 }, { "grad_norm": 0.12784378230571747, "learning_rate": 9.550122188544625e-06, "loss": 0.0055, "step": 81000 }, { "grad_norm": 0.11028789728879929, "learning_rate": 9.540405106392152e-06, "loss": 0.0046, "step": 81010 }, { "grad_norm": 0.0529247522354126, "learning_rate": 9.53069244884015e-06, "loss": 0.0029, "step": 81020 }, { "grad_norm": 0.07433217018842697, "learning_rate": 9.52098421695079e-06, "loss": 0.0029, "step": 81030 }, { "grad_norm": 0.14141814410686493, "learning_rate": 9.511280411785728e-06, "loss": 0.0048, "step": 81040 }, { "grad_norm": 0.12887534499168396, "learning_rate": 9.50158103440616e-06, "loss": 0.0049, "step": 81050 }, { "grad_norm": 0.06588256359100342, "learning_rate": 9.491886085872815e-06, "loss": 0.0038, "step": 81060 }, { "grad_norm": 0.13371676206588745, "learning_rate": 9.482195567245895e-06, "loss": 0.0052, "step": 81070 }, { "grad_norm": 0.11854440718889236, "learning_rate": 9.472509479585151e-06, "loss": 0.0031, "step": 81080 }, { "grad_norm": 0.12449335306882858, "learning_rate": 9.462827823949833e-06, "loss": 0.0042, "step": 81090 }, { "grad_norm": 0.13764262199401855, "learning_rate": 9.453150601398735e-06, "loss": 0.0044, "step": 81100 }, { "grad_norm": 0.119191974401474, "learning_rate": 9.44347781299011e-06, "loss": 0.0046, "step": 81110 }, { "grad_norm": 0.12080944329500198, "learning_rate": 9.43380945978179e-06, "loss": 0.0029, "step": 81120 }, { "grad_norm": 0.07251476496458054, "learning_rate": 9.424145542831047e-06, "loss": 0.0039, "step": 81130 }, { "grad_norm": 0.16269588470458984, "learning_rate": 9.414486063194778e-06, "loss": 0.0045, "step": 81140 }, { "grad_norm": 0.11192392557859421, "learning_rate": 9.404831021929273e-06, "loss": 0.0042, "step": 81150 }, { "grad_norm": 0.14323873817920685, "learning_rate": 9.395180420090427e-06, "loss": 0.0027, "step": 81160 }, { "grad_norm": 0.13255691528320312, "learning_rate": 9.385534258733591e-06, "loss": 0.0035, "step": 81170 }, { "grad_norm": 0.15317364037036896, "learning_rate": 9.375892538913677e-06, "loss": 0.0044, "step": 81180 }, { "grad_norm": 0.2204854041337967, "learning_rate": 9.36625526168507e-06, "loss": 0.0051, "step": 81190 }, { "grad_norm": 0.2085660845041275, "learning_rate": 9.356622428101686e-06, "loss": 0.0043, "step": 81200 }, { "grad_norm": 0.09646205604076385, "learning_rate": 9.346994039216972e-06, "loss": 0.0031, "step": 81210 }, { "grad_norm": 0.10483411699533463, "learning_rate": 9.337370096083886e-06, "loss": 0.0035, "step": 81220 }, { "grad_norm": 0.15192440152168274, "learning_rate": 9.327750599754854e-06, "loss": 0.0034, "step": 81230 }, { "grad_norm": 0.13443103432655334, "learning_rate": 9.318135551281875e-06, "loss": 0.0029, "step": 81240 }, { "grad_norm": 0.11022670567035675, "learning_rate": 9.308524951716424e-06, "loss": 0.0031, "step": 81250 }, { "grad_norm": 0.12616662681102753, "learning_rate": 9.298918802109497e-06, "loss": 0.0052, "step": 81260 }, { "grad_norm": 0.11732427030801773, "learning_rate": 9.289317103511625e-06, "loss": 0.005, "step": 81270 }, { "grad_norm": 0.09504887461662292, "learning_rate": 9.279719856972835e-06, "loss": 0.0041, "step": 81280 }, { "grad_norm": 0.07841915637254715, "learning_rate": 9.270127063542645e-06, "loss": 0.0042, "step": 81290 }, { "grad_norm": 0.1559361219406128, "learning_rate": 9.260538724270124e-06, "loss": 0.004, "step": 81300 }, { "grad_norm": 0.19920332729816437, "learning_rate": 9.250954840203852e-06, "loss": 0.0025, "step": 81310 }, { "grad_norm": 0.08554182201623917, "learning_rate": 9.241375412391878e-06, "loss": 0.0032, "step": 81320 }, { "grad_norm": 0.08152375370264053, "learning_rate": 9.23180044188181e-06, "loss": 0.0042, "step": 81330 }, { "grad_norm": 0.1428239792585373, "learning_rate": 9.222229929720755e-06, "loss": 0.0048, "step": 81340 }, { "grad_norm": 0.15789289772510529, "learning_rate": 9.212663876955336e-06, "loss": 0.0048, "step": 81350 }, { "grad_norm": 0.12096759676933289, "learning_rate": 9.20310228463166e-06, "loss": 0.0033, "step": 81360 }, { "grad_norm": 0.1213068887591362, "learning_rate": 9.19354515379539e-06, "loss": 0.0057, "step": 81370 }, { "grad_norm": 0.08252610266208649, "learning_rate": 9.183992485491666e-06, "loss": 0.0038, "step": 81380 }, { "grad_norm": 0.0853504091501236, "learning_rate": 9.174444280765154e-06, "loss": 0.0044, "step": 81390 }, { "grad_norm": 0.08121205121278763, "learning_rate": 9.164900540660032e-06, "loss": 0.0049, "step": 81400 }, { "grad_norm": 0.07931079715490341, "learning_rate": 9.155361266220008e-06, "loss": 0.0027, "step": 81410 }, { "grad_norm": 0.09351528435945511, "learning_rate": 9.145826458488254e-06, "loss": 0.0036, "step": 81420 }, { "grad_norm": 0.1142294779419899, "learning_rate": 9.136296118507509e-06, "loss": 0.0042, "step": 81430 }, { "grad_norm": 0.11896838992834091, "learning_rate": 9.126770247319966e-06, "loss": 0.0037, "step": 81440 }, { "grad_norm": 0.17263571918010712, "learning_rate": 9.11724884596738e-06, "loss": 0.0033, "step": 81450 }, { "grad_norm": 0.16436992585659027, "learning_rate": 9.107731915490996e-06, "loss": 0.0028, "step": 81460 }, { "grad_norm": 0.2308327704668045, "learning_rate": 9.098219456931579e-06, "loss": 0.0046, "step": 81470 }, { "grad_norm": 0.13100895285606384, "learning_rate": 9.088711471329376e-06, "loss": 0.0028, "step": 81480 }, { "grad_norm": 0.11909476667642593, "learning_rate": 9.079207959724184e-06, "loss": 0.0047, "step": 81490 }, { "grad_norm": 0.10535233467817307, "learning_rate": 9.069708923155279e-06, "loss": 0.0034, "step": 81500 }, { "grad_norm": 0.12582562863826752, "learning_rate": 9.060214362661468e-06, "loss": 0.0037, "step": 81510 }, { "grad_norm": 0.12450666725635529, "learning_rate": 9.050724279281064e-06, "loss": 0.0038, "step": 81520 }, { "grad_norm": 0.10954143106937408, "learning_rate": 9.041238674051894e-06, "loss": 0.0041, "step": 81530 }, { "grad_norm": 0.16127881407737732, "learning_rate": 9.031757548011277e-06, "loss": 0.0049, "step": 81540 }, { "grad_norm": 0.16073794662952423, "learning_rate": 9.022280902196051e-06, "loss": 0.0037, "step": 81550 }, { "grad_norm": 0.22449719905853271, "learning_rate": 9.012808737642592e-06, "loss": 0.0028, "step": 81560 }, { "grad_norm": 0.13126447796821594, "learning_rate": 9.003341055386739e-06, "loss": 0.0057, "step": 81570 }, { "grad_norm": 0.11634434759616852, "learning_rate": 8.993877856463862e-06, "loss": 0.0071, "step": 81580 }, { "grad_norm": 0.10134931653738022, "learning_rate": 8.984419141908856e-06, "loss": 0.0029, "step": 81590 }, { "grad_norm": 0.07972719520330429, "learning_rate": 8.974964912756117e-06, "loss": 0.0031, "step": 81600 }, { "grad_norm": 0.11023054271936417, "learning_rate": 8.965515170039523e-06, "loss": 0.0053, "step": 81610 }, { "grad_norm": 0.1271434724330902, "learning_rate": 8.956069914792514e-06, "loss": 0.0047, "step": 81620 }, { "grad_norm": 0.15625931322574615, "learning_rate": 8.946629148047969e-06, "loss": 0.005, "step": 81630 }, { "grad_norm": 0.11554814875125885, "learning_rate": 8.93719287083834e-06, "loss": 0.003, "step": 81640 }, { "grad_norm": 0.11772696673870087, "learning_rate": 8.927761084195569e-06, "loss": 0.0033, "step": 81650 }, { "grad_norm": 0.13250641524791718, "learning_rate": 8.918333789151096e-06, "loss": 0.0033, "step": 81660 }, { "grad_norm": 0.1559293270111084, "learning_rate": 8.90891098673587e-06, "loss": 0.0035, "step": 81670 }, { "grad_norm": 0.11380369961261749, "learning_rate": 8.89949267798037e-06, "loss": 0.0062, "step": 81680 }, { "grad_norm": 0.08513014018535614, "learning_rate": 8.890078863914542e-06, "loss": 0.0059, "step": 81690 }, { "grad_norm": 0.12565311789512634, "learning_rate": 8.880669545567894e-06, "loss": 0.0028, "step": 81700 }, { "grad_norm": 0.20167843997478485, "learning_rate": 8.871264723969381e-06, "loss": 0.0034, "step": 81710 }, { "grad_norm": 0.11383001506328583, "learning_rate": 8.86186440014754e-06, "loss": 0.0038, "step": 81720 }, { "grad_norm": 0.07304908335208893, "learning_rate": 8.852468575130351e-06, "loss": 0.0063, "step": 81730 }, { "grad_norm": 0.11023984849452972, "learning_rate": 8.843077249945347e-06, "loss": 0.0032, "step": 81740 }, { "grad_norm": 0.0736594945192337, "learning_rate": 8.833690425619523e-06, "loss": 0.0028, "step": 81750 }, { "grad_norm": 0.0985010638833046, "learning_rate": 8.82430810317942e-06, "loss": 0.0035, "step": 81760 }, { "grad_norm": 0.16763731837272644, "learning_rate": 8.814930283651074e-06, "loss": 0.0028, "step": 81770 }, { "grad_norm": 0.15060175955295563, "learning_rate": 8.805556968060048e-06, "loss": 0.0035, "step": 81780 }, { "grad_norm": 0.08201233297586441, "learning_rate": 8.796188157431362e-06, "loss": 0.0045, "step": 81790 }, { "grad_norm": 0.07623586803674698, "learning_rate": 8.786823852789595e-06, "loss": 0.0049, "step": 81800 }, { "grad_norm": 0.11155882477760315, "learning_rate": 8.777464055158813e-06, "loss": 0.0043, "step": 81810 }, { "grad_norm": 0.10267210006713867, "learning_rate": 8.768108765562578e-06, "loss": 0.0031, "step": 81820 }, { "grad_norm": 0.045910149812698364, "learning_rate": 8.758757985023975e-06, "loss": 0.0027, "step": 81830 }, { "grad_norm": 0.07937157154083252, "learning_rate": 8.749411714565598e-06, "loss": 0.0028, "step": 81840 }, { "grad_norm": 0.07959090918302536, "learning_rate": 8.74006995520954e-06, "loss": 0.0028, "step": 81850 }, { "grad_norm": 0.1781461089849472, "learning_rate": 8.730732707977397e-06, "loss": 0.0043, "step": 81860 }, { "grad_norm": 0.11358129978179932, "learning_rate": 8.721399973890281e-06, "loss": 0.0049, "step": 81870 }, { "grad_norm": 0.07725227624177933, "learning_rate": 8.712071753968792e-06, "loss": 0.0033, "step": 81880 }, { "grad_norm": 0.11363226920366287, "learning_rate": 8.702748049233072e-06, "loss": 0.0031, "step": 81890 }, { "grad_norm": 0.14845629036426544, "learning_rate": 8.693428860702718e-06, "loss": 0.0042, "step": 81900 }, { "grad_norm": 0.12572094798088074, "learning_rate": 8.684114189396898e-06, "loss": 0.0037, "step": 81910 }, { "grad_norm": 0.09292511641979218, "learning_rate": 8.674804036334222e-06, "loss": 0.0031, "step": 81920 }, { "grad_norm": 0.11165757477283478, "learning_rate": 8.665498402532857e-06, "loss": 0.0036, "step": 81930 }, { "grad_norm": 0.09272956103086472, "learning_rate": 8.656197289010432e-06, "loss": 0.0051, "step": 81940 }, { "grad_norm": 0.12207894772291183, "learning_rate": 8.646900696784122e-06, "loss": 0.0058, "step": 81950 }, { "grad_norm": 0.08286938071250916, "learning_rate": 8.637608626870553e-06, "loss": 0.0027, "step": 81960 }, { "grad_norm": 0.0696222111582756, "learning_rate": 8.628321080285944e-06, "loss": 0.0036, "step": 81970 }, { "grad_norm": 0.12744946777820587, "learning_rate": 8.619038058045925e-06, "loss": 0.0049, "step": 81980 }, { "grad_norm": 0.0660964772105217, "learning_rate": 8.609759561165704e-06, "loss": 0.0053, "step": 81990 }, { "grad_norm": 0.09966526925563812, "learning_rate": 8.600485590659934e-06, "loss": 0.0034, "step": 82000 }, { "grad_norm": 0.08917135000228882, "learning_rate": 8.591216147542813e-06, "loss": 0.0041, "step": 82010 }, { "grad_norm": 0.07499386370182037, "learning_rate": 8.581951232828045e-06, "loss": 0.0031, "step": 82020 }, { "grad_norm": 0.21987082064151764, "learning_rate": 8.572690847528825e-06, "loss": 0.0035, "step": 82030 }, { "grad_norm": 0.17539772391319275, "learning_rate": 8.563434992657838e-06, "loss": 0.0043, "step": 82040 }, { "grad_norm": 0.11108379811048508, "learning_rate": 8.554183669227307e-06, "loss": 0.004, "step": 82050 }, { "grad_norm": 0.16633211076259613, "learning_rate": 8.544936878248937e-06, "loss": 0.004, "step": 82060 }, { "grad_norm": 0.09329453110694885, "learning_rate": 8.535694620733942e-06, "loss": 0.0036, "step": 82070 }, { "grad_norm": 0.10853869467973709, "learning_rate": 8.526456897693053e-06, "loss": 0.0027, "step": 82080 }, { "grad_norm": 0.12008479237556458, "learning_rate": 8.517223710136456e-06, "loss": 0.003, "step": 82090 }, { "grad_norm": 0.11418452113866806, "learning_rate": 8.507995059073931e-06, "loss": 0.0038, "step": 82100 }, { "grad_norm": 0.15266115963459015, "learning_rate": 8.498770945514673e-06, "loss": 0.0043, "step": 82110 }, { "grad_norm": 0.1646433174610138, "learning_rate": 8.489551370467436e-06, "loss": 0.0033, "step": 82120 }, { "grad_norm": 0.1160622388124466, "learning_rate": 8.480336334940442e-06, "loss": 0.0032, "step": 82130 }, { "grad_norm": 0.09823844581842422, "learning_rate": 8.471125839941458e-06, "loss": 0.003, "step": 82140 }, { "grad_norm": 0.11272481083869934, "learning_rate": 8.461919886477682e-06, "loss": 0.0025, "step": 82150 }, { "grad_norm": 0.1510685384273529, "learning_rate": 8.452718475555927e-06, "loss": 0.0035, "step": 82160 }, { "grad_norm": 0.05407440662384033, "learning_rate": 8.443521608182398e-06, "loss": 0.0029, "step": 82170 }, { "grad_norm": 0.16786210238933563, "learning_rate": 8.43432928536288e-06, "loss": 0.0031, "step": 82180 }, { "grad_norm": 0.09900354593992233, "learning_rate": 8.425141508102607e-06, "loss": 0.0028, "step": 82190 }, { "grad_norm": 0.0962815210223198, "learning_rate": 8.415958277406365e-06, "loss": 0.003, "step": 82200 }, { "grad_norm": 0.15872906148433685, "learning_rate": 8.406779594278392e-06, "loss": 0.0035, "step": 82210 }, { "grad_norm": 0.11459430307149887, "learning_rate": 8.397605459722468e-06, "loss": 0.0031, "step": 82220 }, { "grad_norm": 0.11760460585355759, "learning_rate": 8.388435874741863e-06, "loss": 0.0035, "step": 82230 }, { "grad_norm": 0.1410849541425705, "learning_rate": 8.37927084033936e-06, "loss": 0.0078, "step": 82240 }, { "grad_norm": 0.07804874330759048, "learning_rate": 8.370110357517208e-06, "loss": 0.0029, "step": 82250 }, { "grad_norm": 0.09655944257974625, "learning_rate": 8.360954427277201e-06, "loss": 0.0031, "step": 82260 }, { "grad_norm": 0.1061936765909195, "learning_rate": 8.351803050620627e-06, "loss": 0.0031, "step": 82270 }, { "grad_norm": 0.049568839371204376, "learning_rate": 8.34265622854824e-06, "loss": 0.0043, "step": 82280 }, { "grad_norm": 0.21295641362667084, "learning_rate": 8.333513962060335e-06, "loss": 0.0042, "step": 82290 }, { "grad_norm": 0.5003190040588379, "learning_rate": 8.324376252156701e-06, "loss": 0.0054, "step": 82300 }, { "grad_norm": 0.048602145165205, "learning_rate": 8.315243099836633e-06, "loss": 0.0034, "step": 82310 }, { "grad_norm": 0.1919163167476654, "learning_rate": 8.3061145060989e-06, "loss": 0.0038, "step": 82320 }, { "grad_norm": 0.1586100310087204, "learning_rate": 8.296990471941806e-06, "loss": 0.0037, "step": 82330 }, { "grad_norm": 0.15336564183235168, "learning_rate": 8.287870998363112e-06, "loss": 0.0068, "step": 82340 }, { "grad_norm": 0.11063138395547867, "learning_rate": 8.278756086360156e-06, "loss": 0.0033, "step": 82350 }, { "grad_norm": 0.12187618017196655, "learning_rate": 8.269645736929693e-06, "loss": 0.0033, "step": 82360 }, { "grad_norm": 0.1292835772037506, "learning_rate": 8.260539951068048e-06, "loss": 0.0031, "step": 82370 }, { "grad_norm": 0.13043294847011566, "learning_rate": 8.251438729770983e-06, "loss": 0.003, "step": 82380 }, { "grad_norm": 0.13802874088287354, "learning_rate": 8.242342074033826e-06, "loss": 0.0039, "step": 82390 }, { "grad_norm": 0.18004484474658966, "learning_rate": 8.233249984851344e-06, "loss": 0.0051, "step": 82400 }, { "grad_norm": 0.15441563725471497, "learning_rate": 8.224162463217849e-06, "loss": 0.0048, "step": 82410 }, { "grad_norm": 0.10423332452774048, "learning_rate": 8.215079510127139e-06, "loss": 0.0041, "step": 82420 }, { "grad_norm": 0.08604420721530914, "learning_rate": 8.206001126572521e-06, "loss": 0.0068, "step": 82430 }, { "grad_norm": 0.09913425892591476, "learning_rate": 8.196927313546776e-06, "loss": 0.0069, "step": 82440 }, { "grad_norm": 0.10228677093982697, "learning_rate": 8.18785807204222e-06, "loss": 0.003, "step": 82450 }, { "grad_norm": 0.07514099031686783, "learning_rate": 8.178793403050627e-06, "loss": 0.0033, "step": 82460 }, { "grad_norm": 0.13412827253341675, "learning_rate": 8.16973330756331e-06, "loss": 0.0039, "step": 82470 }, { "grad_norm": 0.13841423392295837, "learning_rate": 8.16067778657107e-06, "loss": 0.0029, "step": 82480 }, { "grad_norm": 0.10476895421743393, "learning_rate": 8.151626841064214e-06, "loss": 0.0038, "step": 82490 }, { "grad_norm": 0.08856423944234848, "learning_rate": 8.142580472032518e-06, "loss": 0.0053, "step": 82500 }, { "grad_norm": 0.08117131143808365, "learning_rate": 8.133538680465285e-06, "loss": 0.0033, "step": 82510 }, { "grad_norm": 0.10698432475328445, "learning_rate": 8.124501467351336e-06, "loss": 0.0039, "step": 82520 }, { "grad_norm": 0.1226981058716774, "learning_rate": 8.11546883367893e-06, "loss": 0.0041, "step": 82530 }, { "grad_norm": 0.0923372432589531, "learning_rate": 8.106440780435882e-06, "loss": 0.0034, "step": 82540 }, { "grad_norm": 0.11320623755455017, "learning_rate": 8.097417308609485e-06, "loss": 0.0029, "step": 82550 }, { "grad_norm": 0.07951397448778152, "learning_rate": 8.088398419186544e-06, "loss": 0.0022, "step": 82560 }, { "grad_norm": 0.08110750466585159, "learning_rate": 8.079384113153321e-06, "loss": 0.0034, "step": 82570 }, { "grad_norm": 0.10729916393756866, "learning_rate": 8.070374391495644e-06, "loss": 0.0034, "step": 82580 }, { "grad_norm": 0.1045849621295929, "learning_rate": 8.061369255198764e-06, "loss": 0.0052, "step": 82590 }, { "grad_norm": 0.18382951617240906, "learning_rate": 8.052368705247493e-06, "loss": 0.0026, "step": 82600 }, { "grad_norm": 0.10071160644292831, "learning_rate": 8.043372742626114e-06, "loss": 0.0036, "step": 82610 }, { "grad_norm": 0.15029984712600708, "learning_rate": 8.034381368318422e-06, "loss": 0.0023, "step": 82620 }, { "grad_norm": 0.13199493288993835, "learning_rate": 8.025394583307677e-06, "loss": 0.0034, "step": 82630 }, { "grad_norm": 0.12402576208114624, "learning_rate": 8.016412388576683e-06, "loss": 0.0029, "step": 82640 }, { "grad_norm": 0.18965472280979156, "learning_rate": 8.007434785107699e-06, "loss": 0.004, "step": 82650 }, { "grad_norm": 0.16004788875579834, "learning_rate": 7.998461773882509e-06, "loss": 0.0025, "step": 82660 }, { "grad_norm": 0.1743587702512741, "learning_rate": 7.989493355882393e-06, "loss": 0.0033, "step": 82670 }, { "grad_norm": 0.13333894312381744, "learning_rate": 7.980529532088137e-06, "loss": 0.0028, "step": 82680 }, { "grad_norm": 0.08019594103097916, "learning_rate": 7.97157030347998e-06, "loss": 0.0029, "step": 82690 }, { "grad_norm": 0.09011273086071014, "learning_rate": 7.96261567103772e-06, "loss": 0.0043, "step": 82700 }, { "grad_norm": 0.24786598980426788, "learning_rate": 7.953665635740593e-06, "loss": 0.0034, "step": 82710 }, { "grad_norm": 0.09296990931034088, "learning_rate": 7.94472019856738e-06, "loss": 0.0031, "step": 82720 }, { "grad_norm": 0.1357198804616928, "learning_rate": 7.935779360496337e-06, "loss": 0.0056, "step": 82730 }, { "grad_norm": 0.11837385594844818, "learning_rate": 7.92684312250523e-06, "loss": 0.0032, "step": 82740 }, { "grad_norm": 0.0906326174736023, "learning_rate": 7.917911485571289e-06, "loss": 0.0026, "step": 82750 }, { "grad_norm": 0.11947588622570038, "learning_rate": 7.90898445067128e-06, "loss": 0.0057, "step": 82760 }, { "grad_norm": 0.22726184129714966, "learning_rate": 7.900062018781462e-06, "loss": 0.0039, "step": 82770 }, { "grad_norm": 0.09637151658535004, "learning_rate": 7.891144190877553e-06, "loss": 0.0034, "step": 82780 }, { "grad_norm": 0.11575561761856079, "learning_rate": 7.882230967934806e-06, "loss": 0.0041, "step": 82790 }, { "grad_norm": 0.15591830015182495, "learning_rate": 7.873322350927959e-06, "loss": 0.0028, "step": 82800 }, { "grad_norm": 0.2017267495393753, "learning_rate": 7.864418340831253e-06, "loss": 0.0044, "step": 82810 }, { "grad_norm": 0.07484236359596252, "learning_rate": 7.855518938618395e-06, "loss": 0.0022, "step": 82820 }, { "grad_norm": 0.11761019378900528, "learning_rate": 7.846624145262633e-06, "loss": 0.0033, "step": 82830 }, { "grad_norm": 0.09330493956804276, "learning_rate": 7.837733961736672e-06, "loss": 0.0032, "step": 82840 }, { "grad_norm": 0.06734901666641235, "learning_rate": 7.828848389012727e-06, "loss": 0.0038, "step": 82850 }, { "grad_norm": 0.09747618436813354, "learning_rate": 7.819967428062524e-06, "loss": 0.0031, "step": 82860 }, { "grad_norm": 0.1428147256374359, "learning_rate": 7.811091079857274e-06, "loss": 0.0034, "step": 82870 }, { "grad_norm": 0.08772511780261993, "learning_rate": 7.802219345367667e-06, "loss": 0.0038, "step": 82880 }, { "grad_norm": 0.07879100739955902, "learning_rate": 7.793352225563916e-06, "loss": 0.0026, "step": 82890 }, { "grad_norm": 0.10510575771331787, "learning_rate": 7.7844897214157e-06, "loss": 0.0032, "step": 82900 }, { "grad_norm": 0.04552914947271347, "learning_rate": 7.775631833892228e-06, "loss": 0.0032, "step": 82910 }, { "grad_norm": 0.10674386471509933, "learning_rate": 7.76677856396215e-06, "loss": 0.0046, "step": 82920 }, { "grad_norm": 0.09620343893766403, "learning_rate": 7.757929912593697e-06, "loss": 0.0037, "step": 82930 }, { "grad_norm": 0.07935260236263275, "learning_rate": 7.749085880754509e-06, "loss": 0.0039, "step": 82940 }, { "grad_norm": 0.08324584364891052, "learning_rate": 7.740246469411777e-06, "loss": 0.0022, "step": 82950 }, { "grad_norm": 0.0771995559334755, "learning_rate": 7.731411679532142e-06, "loss": 0.0035, "step": 82960 }, { "grad_norm": 0.10448640584945679, "learning_rate": 7.722581512081779e-06, "loss": 0.005, "step": 82970 }, { "grad_norm": 0.1015014499425888, "learning_rate": 7.713755968026331e-06, "loss": 0.0043, "step": 82980 }, { "grad_norm": 0.13095085322856903, "learning_rate": 7.704935048330969e-06, "loss": 0.003, "step": 82990 }, { "grad_norm": 0.09425433725118637, "learning_rate": 7.696118753960313e-06, "loss": 0.0035, "step": 83000 }, { "grad_norm": 0.13173280656337738, "learning_rate": 7.687307085878503e-06, "loss": 0.0023, "step": 83010 }, { "grad_norm": 0.1278132200241089, "learning_rate": 7.678500045049186e-06, "loss": 0.0032, "step": 83020 }, { "grad_norm": 0.07425855100154877, "learning_rate": 7.669697632435463e-06, "loss": 0.0034, "step": 83030 }, { "grad_norm": 0.11161134392023087, "learning_rate": 7.660899848999963e-06, "loss": 0.0047, "step": 83040 }, { "grad_norm": 0.150432288646698, "learning_rate": 7.652106695704803e-06, "loss": 0.004, "step": 83050 }, { "grad_norm": 0.1158176064491272, "learning_rate": 7.643318173511598e-06, "loss": 0.0048, "step": 83060 }, { "grad_norm": 0.14213618636131287, "learning_rate": 7.634534283381423e-06, "loss": 0.0031, "step": 83070 }, { "grad_norm": 0.10260511934757233, "learning_rate": 7.625755026274889e-06, "loss": 0.003, "step": 83080 }, { "grad_norm": 0.12057013809680939, "learning_rate": 7.616980403152069e-06, "loss": 0.0032, "step": 83090 }, { "grad_norm": 0.05661454424262047, "learning_rate": 7.608210414972561e-06, "loss": 0.0055, "step": 83100 }, { "grad_norm": 0.1833786815404892, "learning_rate": 7.599445062695404e-06, "loss": 0.0042, "step": 83110 }, { "grad_norm": 0.2057371586561203, "learning_rate": 7.590684347279214e-06, "loss": 0.0047, "step": 83120 }, { "grad_norm": 0.06297498196363449, "learning_rate": 7.581928269682004e-06, "loss": 0.0039, "step": 83130 }, { "grad_norm": 0.14162027835845947, "learning_rate": 7.573176830861361e-06, "loss": 0.0031, "step": 83140 }, { "grad_norm": 0.11148854345083237, "learning_rate": 7.5644300317743e-06, "loss": 0.0031, "step": 83150 }, { "grad_norm": 0.17742189764976501, "learning_rate": 7.555687873377376e-06, "loss": 0.0039, "step": 83160 }, { "grad_norm": 0.1296369731426239, "learning_rate": 7.546950356626592e-06, "loss": 0.0046, "step": 83170 }, { "grad_norm": 0.09603054076433182, "learning_rate": 7.538217482477516e-06, "loss": 0.0031, "step": 83180 }, { "grad_norm": 0.16827458143234253, "learning_rate": 7.529489251885119e-06, "loss": 0.0057, "step": 83190 }, { "grad_norm": 0.15759152173995972, "learning_rate": 7.520765665803941e-06, "loss": 0.0029, "step": 83200 }, { "grad_norm": 0.08378294110298157, "learning_rate": 7.512046725187949e-06, "loss": 0.0028, "step": 83210 }, { "grad_norm": 0.14088550209999084, "learning_rate": 7.503332430990645e-06, "loss": 0.0033, "step": 83220 }, { "grad_norm": 0.13624307513237, "learning_rate": 7.494622784165017e-06, "loss": 0.0034, "step": 83230 }, { "grad_norm": 0.1296480894088745, "learning_rate": 7.485917785663543e-06, "loss": 0.0065, "step": 83240 }, { "grad_norm": 0.08558741956949234, "learning_rate": 7.477217436438172e-06, "loss": 0.0042, "step": 83250 }, { "grad_norm": 0.13230927288532257, "learning_rate": 7.46852173744037e-06, "loss": 0.0057, "step": 83260 }, { "grad_norm": 0.09083113819360733, "learning_rate": 7.4598306896211004e-06, "loss": 0.0056, "step": 83270 }, { "grad_norm": 0.06153198704123497, "learning_rate": 7.451144293930773e-06, "loss": 0.0035, "step": 83280 }, { "grad_norm": 0.07934565842151642, "learning_rate": 7.4424625513193455e-06, "loss": 0.0031, "step": 83290 }, { "grad_norm": 0.12588192522525787, "learning_rate": 7.433785462736209e-06, "loss": 0.0024, "step": 83300 }, { "grad_norm": 0.05928992107510567, "learning_rate": 7.425113029130315e-06, "loss": 0.0038, "step": 83310 }, { "grad_norm": 0.08145612478256226, "learning_rate": 7.416445251450044e-06, "loss": 0.0052, "step": 83320 }, { "grad_norm": 0.09548228234052658, "learning_rate": 7.4077821306433035e-06, "loss": 0.0036, "step": 83330 }, { "grad_norm": 0.108424611389637, "learning_rate": 7.399123667657465e-06, "loss": 0.0041, "step": 83340 }, { "grad_norm": 0.07403428852558136, "learning_rate": 7.390469863439426e-06, "loss": 0.0051, "step": 83350 }, { "grad_norm": 0.054752107709646225, "learning_rate": 7.381820718935512e-06, "loss": 0.0022, "step": 83360 }, { "grad_norm": 0.1351126730442047, "learning_rate": 7.373176235091633e-06, "loss": 0.003, "step": 83370 }, { "grad_norm": 0.15584596991539001, "learning_rate": 7.364536412853101e-06, "loss": 0.0045, "step": 83380 }, { "grad_norm": 0.11695010960102081, "learning_rate": 7.355901253164776e-06, "loss": 0.0029, "step": 83390 }, { "grad_norm": 0.10561239719390869, "learning_rate": 7.34727075697097e-06, "loss": 0.0042, "step": 83400 }, { "grad_norm": 0.11806154996156693, "learning_rate": 7.3386449252155095e-06, "loss": 0.004, "step": 83410 }, { "grad_norm": 0.09467723965644836, "learning_rate": 7.330023758841692e-06, "loss": 0.0031, "step": 83420 }, { "grad_norm": 0.0931183397769928, "learning_rate": 7.321407258792323e-06, "loss": 0.0034, "step": 83430 }, { "grad_norm": 0.07674963027238846, "learning_rate": 7.312795426009694e-06, "loss": 0.0041, "step": 83440 }, { "grad_norm": 0.16633448004722595, "learning_rate": 7.304188261435585e-06, "loss": 0.0042, "step": 83450 }, { "grad_norm": 0.10420531779527664, "learning_rate": 7.295585766011242e-06, "loss": 0.0038, "step": 83460 }, { "grad_norm": 0.12903518974781036, "learning_rate": 7.286987940677436e-06, "loss": 0.0037, "step": 83470 }, { "grad_norm": 0.14161406457424164, "learning_rate": 7.278394786374426e-06, "loss": 0.0041, "step": 83480 }, { "grad_norm": 0.0806138813495636, "learning_rate": 7.269806304041915e-06, "loss": 0.0046, "step": 83490 }, { "grad_norm": 0.10715804249048233, "learning_rate": 7.261222494619147e-06, "loss": 0.0044, "step": 83500 }, { "grad_norm": 0.10937659442424774, "learning_rate": 7.252643359044825e-06, "loss": 0.0046, "step": 83510 }, { "grad_norm": 0.10347946733236313, "learning_rate": 7.244068898257172e-06, "loss": 0.0037, "step": 83520 }, { "grad_norm": 0.051866043359041214, "learning_rate": 7.2354991131938495e-06, "loss": 0.0035, "step": 83530 }, { "grad_norm": 0.06849698722362518, "learning_rate": 7.2269340047920574e-06, "loss": 0.0029, "step": 83540 }, { "grad_norm": 0.06168193370103836, "learning_rate": 7.218373573988435e-06, "loss": 0.0033, "step": 83550 }, { "grad_norm": 0.12328789383172989, "learning_rate": 7.209817821719178e-06, "loss": 0.006, "step": 83560 }, { "grad_norm": 0.0777924507856369, "learning_rate": 7.201266748919894e-06, "loss": 0.0035, "step": 83570 }, { "grad_norm": 0.055805351585149765, "learning_rate": 7.19272035652575e-06, "loss": 0.0028, "step": 83580 }, { "grad_norm": 0.10109960287809372, "learning_rate": 7.184178645471334e-06, "loss": 0.0039, "step": 83590 }, { "grad_norm": 0.06182949244976044, "learning_rate": 7.175641616690775e-06, "loss": 0.0027, "step": 83600 }, { "grad_norm": 0.10851379483938217, "learning_rate": 7.167109271117661e-06, "loss": 0.0065, "step": 83610 }, { "grad_norm": 0.07899559289216995, "learning_rate": 7.158581609685072e-06, "loss": 0.0061, "step": 83620 }, { "grad_norm": 0.10898265242576599, "learning_rate": 7.150058633325591e-06, "loss": 0.0028, "step": 83630 }, { "grad_norm": 0.0634598433971405, "learning_rate": 7.141540342971287e-06, "loss": 0.0038, "step": 83640 }, { "grad_norm": 0.07381905615329742, "learning_rate": 7.133026739553677e-06, "loss": 0.003, "step": 83650 }, { "grad_norm": 0.19288122653961182, "learning_rate": 7.124517824003829e-06, "loss": 0.0091, "step": 83660 }, { "grad_norm": 0.09542778134346008, "learning_rate": 7.116013597252236e-06, "loss": 0.004, "step": 83670 }, { "grad_norm": 0.07509828358888626, "learning_rate": 7.107514060228921e-06, "loss": 0.0044, "step": 83680 }, { "grad_norm": 0.14773111045360565, "learning_rate": 7.099019213863378e-06, "loss": 0.0046, "step": 83690 }, { "grad_norm": 0.1174469143152237, "learning_rate": 7.090529059084611e-06, "loss": 0.0033, "step": 83700 }, { "grad_norm": 0.1237708255648613, "learning_rate": 7.082043596821058e-06, "loss": 0.0029, "step": 83710 }, { "grad_norm": 0.07948441058397293, "learning_rate": 7.07356282800069e-06, "loss": 0.0041, "step": 83720 }, { "grad_norm": 0.13728415966033936, "learning_rate": 7.0650867535509645e-06, "loss": 0.0048, "step": 83730 }, { "grad_norm": 0.07387731969356537, "learning_rate": 7.0566153743987895e-06, "loss": 0.0031, "step": 83740 }, { "grad_norm": 0.19091419875621796, "learning_rate": 7.04814869147059e-06, "loss": 0.0042, "step": 83750 }, { "grad_norm": 0.09502150863409042, "learning_rate": 7.0396867056922725e-06, "loss": 0.0033, "step": 83760 }, { "grad_norm": 0.07331975549459457, "learning_rate": 7.031229417989244e-06, "loss": 0.0028, "step": 83770 }, { "grad_norm": 0.10588335990905762, "learning_rate": 7.022776829286348e-06, "loss": 0.0033, "step": 83780 }, { "grad_norm": 0.12893889844417572, "learning_rate": 7.014328940507975e-06, "loss": 0.0044, "step": 83790 }, { "grad_norm": 0.12825781106948853, "learning_rate": 7.005885752577945e-06, "loss": 0.0031, "step": 83800 }, { "grad_norm": 0.11431801319122314, "learning_rate": 6.997447266419615e-06, "loss": 0.0028, "step": 83810 }, { "grad_norm": 0.06393536925315857, "learning_rate": 6.989013482955797e-06, "loss": 0.0027, "step": 83820 }, { "grad_norm": 0.0712638571858406, "learning_rate": 6.980584403108808e-06, "loss": 0.0059, "step": 83830 }, { "grad_norm": 0.13172493875026703, "learning_rate": 6.972160027800417e-06, "loss": 0.0038, "step": 83840 }, { "grad_norm": 0.06897953152656555, "learning_rate": 6.963740357951931e-06, "loss": 0.0025, "step": 83850 }, { "grad_norm": 0.09953277558088303, "learning_rate": 6.95532539448408e-06, "loss": 0.0025, "step": 83860 }, { "grad_norm": 0.23267556726932526, "learning_rate": 6.946915138317128e-06, "loss": 0.0038, "step": 83870 }, { "grad_norm": 0.16485977172851562, "learning_rate": 6.9385095903708065e-06, "loss": 0.0037, "step": 83880 }, { "grad_norm": 0.07242505252361298, "learning_rate": 6.930108751564346e-06, "loss": 0.0055, "step": 83890 }, { "grad_norm": 0.09336481243371964, "learning_rate": 6.921712622816428e-06, "loss": 0.0038, "step": 83900 }, { "grad_norm": 0.17631638050079346, "learning_rate": 6.913321205045259e-06, "loss": 0.0029, "step": 83910 }, { "grad_norm": 0.12721019983291626, "learning_rate": 6.904934499168491e-06, "loss": 0.003, "step": 83920 }, { "grad_norm": 0.1005140095949173, "learning_rate": 6.896552506103299e-06, "loss": 0.0031, "step": 83930 }, { "grad_norm": 0.05805299058556557, "learning_rate": 6.888175226766313e-06, "loss": 0.0039, "step": 83940 }, { "grad_norm": 0.39109814167022705, "learning_rate": 6.879802662073686e-06, "loss": 0.0038, "step": 83950 }, { "grad_norm": 0.12202560156583786, "learning_rate": 6.871434812940991e-06, "loss": 0.0034, "step": 83960 }, { "grad_norm": 0.10220654308795929, "learning_rate": 6.863071680283345e-06, "loss": 0.004, "step": 83970 }, { "grad_norm": 0.0967525839805603, "learning_rate": 6.854713265015339e-06, "loss": 0.0027, "step": 83980 }, { "grad_norm": 0.1890559047460556, "learning_rate": 6.8463595680510025e-06, "loss": 0.0044, "step": 83990 }, { "grad_norm": 0.09539195150136948, "learning_rate": 6.838010590303911e-06, "loss": 0.007, "step": 84000 }, { "grad_norm": 0.08442394435405731, "learning_rate": 6.829666332687085e-06, "loss": 0.0025, "step": 84010 }, { "grad_norm": 0.10252492129802704, "learning_rate": 6.821326796113059e-06, "loss": 0.0034, "step": 84020 }, { "grad_norm": 0.10038813203573227, "learning_rate": 6.812991981493799e-06, "loss": 0.0074, "step": 84030 }, { "grad_norm": 0.09155953675508499, "learning_rate": 6.804661889740821e-06, "loss": 0.0049, "step": 84040 }, { "grad_norm": 0.13749246299266815, "learning_rate": 6.7963365217650625e-06, "loss": 0.0046, "step": 84050 }, { "grad_norm": 0.14952675998210907, "learning_rate": 6.788015878476983e-06, "loss": 0.0041, "step": 84060 }, { "grad_norm": 0.059884652495384216, "learning_rate": 6.779699960786523e-06, "loss": 0.0031, "step": 84070 }, { "grad_norm": 0.12887398898601532, "learning_rate": 6.771388769603104e-06, "loss": 0.0026, "step": 84080 }, { "grad_norm": 0.1465480625629425, "learning_rate": 6.763082305835605e-06, "loss": 0.0032, "step": 84090 }, { "grad_norm": 0.0825447365641594, "learning_rate": 6.754780570392433e-06, "loss": 0.0029, "step": 84100 }, { "grad_norm": 0.06982220709323883, "learning_rate": 6.7464835641814215e-06, "loss": 0.0028, "step": 84110 }, { "grad_norm": 0.10295608639717102, "learning_rate": 6.7381912881099515e-06, "loss": 0.0042, "step": 84120 }, { "grad_norm": 0.18483741581439972, "learning_rate": 6.72990374308482e-06, "loss": 0.0036, "step": 84130 }, { "grad_norm": 0.0868806391954422, "learning_rate": 6.721620930012379e-06, "loss": 0.0037, "step": 84140 }, { "grad_norm": 0.06715232878923416, "learning_rate": 6.713342849798393e-06, "loss": 0.0033, "step": 84150 }, { "grad_norm": 0.11778873205184937, "learning_rate": 6.705069503348166e-06, "loss": 0.0026, "step": 84160 }, { "grad_norm": 0.12615545094013214, "learning_rate": 6.696800891566435e-06, "loss": 0.0038, "step": 84170 }, { "grad_norm": 0.08089244365692139, "learning_rate": 6.688537015357449e-06, "loss": 0.0057, "step": 84180 }, { "grad_norm": 0.1956840455532074, "learning_rate": 6.680277875624941e-06, "loss": 0.0041, "step": 84190 }, { "grad_norm": 0.13221915066242218, "learning_rate": 6.672023473272121e-06, "loss": 0.005, "step": 84200 }, { "grad_norm": 0.11827091127634048, "learning_rate": 6.663773809201667e-06, "loss": 0.0042, "step": 84210 }, { "grad_norm": 0.13696710765361786, "learning_rate": 6.655528884315748e-06, "loss": 0.0027, "step": 84220 }, { "grad_norm": 0.1871328204870224, "learning_rate": 6.647288699516036e-06, "loss": 0.0043, "step": 84230 }, { "grad_norm": 0.14717571437358856, "learning_rate": 6.639053255703642e-06, "loss": 0.0051, "step": 84240 }, { "grad_norm": 0.12334898859262466, "learning_rate": 6.6308225537791925e-06, "loss": 0.0036, "step": 84250 }, { "grad_norm": 0.06419141590595245, "learning_rate": 6.622596594642777e-06, "loss": 0.0025, "step": 84260 }, { "grad_norm": 0.10002105683088303, "learning_rate": 6.614375379193999e-06, "loss": 0.0037, "step": 84270 }, { "grad_norm": 0.11261548846960068, "learning_rate": 6.606158908331889e-06, "loss": 0.0029, "step": 84280 }, { "grad_norm": 0.12270487844944, "learning_rate": 6.597947182955005e-06, "loss": 0.0038, "step": 84290 }, { "grad_norm": 0.10216362029314041, "learning_rate": 6.589740203961359e-06, "loss": 0.0029, "step": 84300 }, { "grad_norm": 0.10420535504817963, "learning_rate": 6.581537972248464e-06, "loss": 0.0041, "step": 84310 }, { "grad_norm": 0.2183918058872223, "learning_rate": 6.573340488713275e-06, "loss": 0.0037, "step": 84320 }, { "grad_norm": 0.11303756386041641, "learning_rate": 6.565147754252304e-06, "loss": 0.0031, "step": 84330 }, { "grad_norm": 0.0811690017580986, "learning_rate": 6.556959769761462e-06, "loss": 0.0032, "step": 84340 }, { "grad_norm": 0.059989020228385925, "learning_rate": 6.548776536136192e-06, "loss": 0.0053, "step": 84350 }, { "grad_norm": 0.0864698514342308, "learning_rate": 6.54059805427138e-06, "loss": 0.0033, "step": 84360 }, { "grad_norm": 0.19774045050144196, "learning_rate": 6.532424325061437e-06, "loss": 0.0034, "step": 84370 }, { "grad_norm": 0.17242787778377533, "learning_rate": 6.5242553494001935e-06, "loss": 0.0046, "step": 84380 }, { "grad_norm": 0.1036582961678505, "learning_rate": 6.516091128181046e-06, "loss": 0.0025, "step": 84390 }, { "grad_norm": 0.11736901849508286, "learning_rate": 6.507931662296785e-06, "loss": 0.0031, "step": 84400 }, { "grad_norm": 0.14660859107971191, "learning_rate": 6.4997769526397354e-06, "loss": 0.0035, "step": 84410 }, { "grad_norm": 0.055792491883039474, "learning_rate": 6.491627000101663e-06, "loss": 0.0028, "step": 84420 }, { "grad_norm": 0.14416709542274475, "learning_rate": 6.4834818055738545e-06, "loss": 0.0025, "step": 84430 }, { "grad_norm": 0.06464514881372452, "learning_rate": 6.4753413699470465e-06, "loss": 0.0035, "step": 84440 }, { "grad_norm": 0.09935448318719864, "learning_rate": 6.467205694111478e-06, "loss": 0.0029, "step": 84450 }, { "grad_norm": 0.12695899605751038, "learning_rate": 6.459074778956836e-06, "loss": 0.0036, "step": 84460 }, { "grad_norm": 0.1202310100197792, "learning_rate": 6.450948625372311e-06, "loss": 0.0064, "step": 84470 }, { "grad_norm": 0.09641741961240768, "learning_rate": 6.442827234246585e-06, "loss": 0.0031, "step": 84480 }, { "grad_norm": 0.12936389446258545, "learning_rate": 6.43471060646777e-06, "loss": 0.0045, "step": 84490 }, { "grad_norm": 0.1425914764404297, "learning_rate": 6.426598742923512e-06, "loss": 0.0032, "step": 84500 }, { "grad_norm": 0.10990263521671295, "learning_rate": 6.418491644500885e-06, "loss": 0.0028, "step": 84510 }, { "grad_norm": 0.06125270947813988, "learning_rate": 6.410389312086512e-06, "loss": 0.006, "step": 84520 }, { "grad_norm": 0.0662326067686081, "learning_rate": 6.4022917465664135e-06, "loss": 0.0038, "step": 84530 }, { "grad_norm": 0.14001308381557465, "learning_rate": 6.3941989488261575e-06, "loss": 0.0035, "step": 84540 }, { "grad_norm": 0.1162634864449501, "learning_rate": 6.386110919750726e-06, "loss": 0.0024, "step": 84550 }, { "grad_norm": 0.0799131691455841, "learning_rate": 6.378027660224651e-06, "loss": 0.0037, "step": 84560 }, { "grad_norm": 0.11785053461790085, "learning_rate": 6.369949171131856e-06, "loss": 0.0037, "step": 84570 }, { "grad_norm": 0.12806981801986694, "learning_rate": 6.361875453355848e-06, "loss": 0.0024, "step": 84580 }, { "grad_norm": 0.08616615831851959, "learning_rate": 6.353806507779525e-06, "loss": 0.0046, "step": 84590 }, { "grad_norm": 0.07194452732801437, "learning_rate": 6.3457423352853044e-06, "loss": 0.0038, "step": 84600 }, { "grad_norm": 0.1313903033733368, "learning_rate": 6.337682936755062e-06, "loss": 0.004, "step": 84610 }, { "grad_norm": 0.16475872695446014, "learning_rate": 6.329628313070179e-06, "loss": 0.0031, "step": 84620 }, { "grad_norm": 0.27319175004959106, "learning_rate": 6.321578465111477e-06, "loss": 0.0039, "step": 84630 }, { "grad_norm": 0.13119573891162872, "learning_rate": 6.31353339375928e-06, "loss": 0.0034, "step": 84640 }, { "grad_norm": 0.09104766696691513, "learning_rate": 6.305493099893395e-06, "loss": 0.003, "step": 84650 }, { "grad_norm": 0.1187794879078865, "learning_rate": 6.297457584393096e-06, "loss": 0.0032, "step": 84660 }, { "grad_norm": 0.11652545630931854, "learning_rate": 6.289426848137126e-06, "loss": 0.0032, "step": 84670 }, { "grad_norm": 0.1825007051229477, "learning_rate": 6.281400892003713e-06, "loss": 0.003, "step": 84680 }, { "grad_norm": 0.1199454516172409, "learning_rate": 6.273379716870581e-06, "loss": 0.0025, "step": 84690 }, { "grad_norm": 0.12464325875043869, "learning_rate": 6.265363323614887e-06, "loss": 0.0042, "step": 84700 }, { "grad_norm": 0.07363849878311157, "learning_rate": 6.257351713113308e-06, "loss": 0.0029, "step": 84710 }, { "grad_norm": 0.10960374772548676, "learning_rate": 6.249344886241981e-06, "loss": 0.0041, "step": 84720 }, { "grad_norm": 0.11774203926324844, "learning_rate": 6.2413428438765295e-06, "loss": 0.0037, "step": 84730 }, { "grad_norm": 0.08039553463459015, "learning_rate": 6.233345586892026e-06, "loss": 0.0031, "step": 84740 }, { "grad_norm": 0.0803665891289711, "learning_rate": 6.225353116163052e-06, "loss": 0.0034, "step": 84750 }, { "grad_norm": 0.10180163383483887, "learning_rate": 6.217365432563627e-06, "loss": 0.0036, "step": 84760 }, { "grad_norm": 0.11533136665821075, "learning_rate": 6.209382536967318e-06, "loss": 0.0027, "step": 84770 }, { "grad_norm": 0.15052203834056854, "learning_rate": 6.201404430247082e-06, "loss": 0.0056, "step": 84780 }, { "grad_norm": 0.09776905924081802, "learning_rate": 6.1934311132754165e-06, "loss": 0.0028, "step": 84790 }, { "grad_norm": 0.09030544757843018, "learning_rate": 6.185462586924257e-06, "loss": 0.0039, "step": 84800 }, { "grad_norm": 0.20352423191070557, "learning_rate": 6.177498852065039e-06, "loss": 0.0035, "step": 84810 }, { "grad_norm": 0.15771159529685974, "learning_rate": 6.169539909568655e-06, "loss": 0.0034, "step": 84820 }, { "grad_norm": 0.06593481451272964, "learning_rate": 6.161585760305488e-06, "loss": 0.0026, "step": 84830 }, { "grad_norm": 0.08260320872068405, "learning_rate": 6.153636405145391e-06, "loss": 0.0038, "step": 84840 }, { "grad_norm": 0.15521687269210815, "learning_rate": 6.145691844957702e-06, "loss": 0.0025, "step": 84850 }, { "grad_norm": 0.11987220495939255, "learning_rate": 6.1377520806112145e-06, "loss": 0.0038, "step": 84860 }, { "grad_norm": 0.08188367635011673, "learning_rate": 6.129817112974218e-06, "loss": 0.0045, "step": 84870 }, { "grad_norm": 0.16655832529067993, "learning_rate": 6.121886942914457e-06, "loss": 0.0054, "step": 84880 }, { "grad_norm": 0.10853315144777298, "learning_rate": 6.11396157129917e-06, "loss": 0.0029, "step": 84890 }, { "grad_norm": 0.1715324968099594, "learning_rate": 6.106040998995066e-06, "loss": 0.0037, "step": 84900 }, { "grad_norm": 0.11971507221460342, "learning_rate": 6.098125226868329e-06, "loss": 0.0045, "step": 84910 }, { "grad_norm": 0.10602565854787827, "learning_rate": 6.090214255784604e-06, "loss": 0.0053, "step": 84920 }, { "grad_norm": 0.05687568336725235, "learning_rate": 6.0823080866090346e-06, "loss": 0.0038, "step": 84930 }, { "grad_norm": 0.10042914748191833, "learning_rate": 6.074406720206233e-06, "loss": 0.005, "step": 84940 }, { "grad_norm": 0.16292153298854828, "learning_rate": 6.066510157440259e-06, "loss": 0.0029, "step": 84950 }, { "grad_norm": 0.11834800988435745, "learning_rate": 6.058618399174676e-06, "loss": 0.003, "step": 84960 }, { "grad_norm": 0.14140458405017853, "learning_rate": 6.050731446272528e-06, "loss": 0.004, "step": 84970 }, { "grad_norm": 0.06800831109285355, "learning_rate": 6.042849299596321e-06, "loss": 0.0034, "step": 84980 }, { "grad_norm": 0.09512348473072052, "learning_rate": 6.034971960008013e-06, "loss": 0.0057, "step": 84990 }, { "grad_norm": 0.1460428237915039, "learning_rate": 6.027099428369082e-06, "loss": 0.0037, "step": 85000 }, { "grad_norm": 0.1483001410961151, "learning_rate": 6.019231705540435e-06, "loss": 0.0038, "step": 85010 }, { "grad_norm": 0.06817420572042465, "learning_rate": 6.011368792382488e-06, "loss": 0.0037, "step": 85020 }, { "grad_norm": 0.06279002130031586, "learning_rate": 6.00351068975511e-06, "loss": 0.0036, "step": 85030 }, { "grad_norm": 0.16598919034004211, "learning_rate": 5.9956573985176674e-06, "loss": 0.0028, "step": 85040 }, { "grad_norm": 0.11687695980072021, "learning_rate": 5.987808919528959e-06, "loss": 0.0045, "step": 85050 }, { "grad_norm": 0.12249191105365753, "learning_rate": 5.979965253647307e-06, "loss": 0.0031, "step": 85060 }, { "grad_norm": 0.09968535602092743, "learning_rate": 5.9721264017304655e-06, "loss": 0.0034, "step": 85070 }, { "grad_norm": 0.1490401178598404, "learning_rate": 5.964292364635682e-06, "loss": 0.0041, "step": 85080 }, { "grad_norm": 0.05407125502824783, "learning_rate": 5.956463143219682e-06, "loss": 0.0054, "step": 85090 }, { "grad_norm": 0.16702422499656677, "learning_rate": 5.948638738338663e-06, "loss": 0.0039, "step": 85100 }, { "grad_norm": 0.087676040828228, "learning_rate": 5.940819150848271e-06, "loss": 0.0029, "step": 85110 }, { "grad_norm": 0.25716152787208557, "learning_rate": 5.933004381603663e-06, "loss": 0.0027, "step": 85120 }, { "grad_norm": 0.05445568263530731, "learning_rate": 5.92519443145943e-06, "loss": 0.0048, "step": 85130 }, { "grad_norm": 0.1521713137626648, "learning_rate": 5.917389301269671e-06, "loss": 0.0032, "step": 85140 }, { "grad_norm": 0.05144519731402397, "learning_rate": 5.9095889918879435e-06, "loss": 0.0028, "step": 85150 }, { "grad_norm": 0.05358356237411499, "learning_rate": 5.901793504167286e-06, "loss": 0.003, "step": 85160 }, { "grad_norm": 0.08364232629537582, "learning_rate": 5.894002838960172e-06, "loss": 0.0038, "step": 85170 }, { "grad_norm": 0.09404610842466354, "learning_rate": 5.886216997118604e-06, "loss": 0.0029, "step": 85180 }, { "grad_norm": 0.12981882691383362, "learning_rate": 5.878435979494023e-06, "loss": 0.0039, "step": 85190 }, { "grad_norm": 0.09269268810749054, "learning_rate": 5.870659786937344e-06, "loss": 0.0036, "step": 85200 }, { "grad_norm": 0.1314881294965744, "learning_rate": 5.862888420298962e-06, "loss": 0.0029, "step": 85210 }, { "grad_norm": 0.05044576898217201, "learning_rate": 5.855121880428743e-06, "loss": 0.0052, "step": 85220 }, { "grad_norm": 0.1397450715303421, "learning_rate": 5.84736016817603e-06, "loss": 0.0052, "step": 85230 }, { "grad_norm": 0.07512795180082321, "learning_rate": 5.839603284389622e-06, "loss": 0.0034, "step": 85240 }, { "grad_norm": 0.09486548602581024, "learning_rate": 5.831851229917812e-06, "loss": 0.0026, "step": 85250 }, { "grad_norm": 0.16799089312553406, "learning_rate": 5.824104005608338e-06, "loss": 0.0024, "step": 85260 }, { "grad_norm": 0.15076811611652374, "learning_rate": 5.81636161230843e-06, "loss": 0.0028, "step": 85270 }, { "grad_norm": 0.28051677346229553, "learning_rate": 5.808624050864791e-06, "loss": 0.0028, "step": 85280 }, { "grad_norm": 0.10395002365112305, "learning_rate": 5.800891322123592e-06, "loss": 0.0045, "step": 85290 }, { "grad_norm": 0.07597383111715317, "learning_rate": 5.793163426930454e-06, "loss": 0.0043, "step": 85300 }, { "grad_norm": 0.1749325692653656, "learning_rate": 5.785440366130512e-06, "loss": 0.0033, "step": 85310 }, { "grad_norm": 0.1569644808769226, "learning_rate": 5.777722140568326e-06, "loss": 0.0033, "step": 85320 }, { "grad_norm": 0.09858129918575287, "learning_rate": 5.7700087510879685e-06, "loss": 0.0039, "step": 85330 }, { "grad_norm": 0.1068454384803772, "learning_rate": 5.762300198532933e-06, "loss": 0.0052, "step": 85340 }, { "grad_norm": 0.15805454552173615, "learning_rate": 5.7545964837462606e-06, "loss": 0.0027, "step": 85350 }, { "grad_norm": 0.09685593098402023, "learning_rate": 5.746897607570384e-06, "loss": 0.0034, "step": 85360 }, { "grad_norm": 0.0934644490480423, "learning_rate": 5.7392035708472626e-06, "loss": 0.0031, "step": 85370 }, { "grad_norm": 0.13068966567516327, "learning_rate": 5.731514374418278e-06, "loss": 0.0054, "step": 85380 }, { "grad_norm": 0.09062746912240982, "learning_rate": 5.7238300191243325e-06, "loss": 0.003, "step": 85390 }, { "grad_norm": 0.11807575821876526, "learning_rate": 5.716150505805762e-06, "loss": 0.0044, "step": 85400 }, { "grad_norm": 0.06764450669288635, "learning_rate": 5.708475835302407e-06, "loss": 0.0028, "step": 85410 }, { "grad_norm": 0.0776691734790802, "learning_rate": 5.70080600845353e-06, "loss": 0.0038, "step": 85420 }, { "grad_norm": 0.10171095281839371, "learning_rate": 5.693141026097909e-06, "loss": 0.0036, "step": 85430 }, { "grad_norm": 0.12495449930429459, "learning_rate": 5.6854808890737745e-06, "loss": 0.0052, "step": 85440 }, { "grad_norm": 0.09150774776935577, "learning_rate": 5.6778255982188225e-06, "loss": 0.0048, "step": 85450 }, { "grad_norm": 0.09140560030937195, "learning_rate": 5.670175154370217e-06, "loss": 0.0024, "step": 85460 }, { "grad_norm": 0.14845550060272217, "learning_rate": 5.662529558364616e-06, "loss": 0.003, "step": 85470 }, { "grad_norm": 0.09917158633470535, "learning_rate": 5.654888811038128e-06, "loss": 0.0045, "step": 85480 }, { "grad_norm": 0.10729033499956131, "learning_rate": 5.647252913226314e-06, "loss": 0.0032, "step": 85490 }, { "grad_norm": 0.2283126264810562, "learning_rate": 5.639621865764255e-06, "loss": 0.0028, "step": 85500 }, { "grad_norm": 0.07244555652141571, "learning_rate": 5.63199566948644e-06, "loss": 0.0041, "step": 85510 }, { "grad_norm": 0.09653643518686295, "learning_rate": 5.6243743252268835e-06, "loss": 0.0031, "step": 85520 }, { "grad_norm": 0.07901393622159958, "learning_rate": 5.616757833819009e-06, "loss": 0.0034, "step": 85530 }, { "grad_norm": 0.1274781972169876, "learning_rate": 5.609146196095794e-06, "loss": 0.006, "step": 85540 }, { "grad_norm": 0.08817362785339355, "learning_rate": 5.6015394128896005e-06, "loss": 0.0065, "step": 85550 }, { "grad_norm": 0.08456137031316757, "learning_rate": 5.593937485032308e-06, "loss": 0.003, "step": 85560 }, { "grad_norm": 0.12632569670677185, "learning_rate": 5.586340413355246e-06, "loss": 0.0044, "step": 85570 }, { "grad_norm": 0.23541302978992462, "learning_rate": 5.578748198689227e-06, "loss": 0.0025, "step": 85580 }, { "grad_norm": 0.12042930722236633, "learning_rate": 5.571160841864504e-06, "loss": 0.0037, "step": 85590 }, { "grad_norm": 0.08623544126749039, "learning_rate": 5.563578343710846e-06, "loss": 0.0032, "step": 85600 }, { "grad_norm": 0.07584749907255173, "learning_rate": 5.556000705057446e-06, "loss": 0.0037, "step": 85610 }, { "grad_norm": 0.09552605450153351, "learning_rate": 5.548427926732996e-06, "loss": 0.0031, "step": 85620 }, { "grad_norm": 0.0449296198785305, "learning_rate": 5.540860009565629e-06, "loss": 0.0026, "step": 85630 }, { "grad_norm": 0.08621620386838913, "learning_rate": 5.533296954382966e-06, "loss": 0.0034, "step": 85640 }, { "grad_norm": 0.0700192078948021, "learning_rate": 5.525738762012095e-06, "loss": 0.0044, "step": 85650 }, { "grad_norm": 0.07514241337776184, "learning_rate": 5.518185433279577e-06, "loss": 0.0029, "step": 85660 }, { "grad_norm": 0.16046348214149475, "learning_rate": 5.510636969011407e-06, "loss": 0.003, "step": 85670 }, { "grad_norm": 0.15345102548599243, "learning_rate": 5.503093370033096e-06, "loss": 0.0038, "step": 85680 }, { "grad_norm": 0.1555137038230896, "learning_rate": 5.495554637169603e-06, "loss": 0.0028, "step": 85690 }, { "grad_norm": 0.09036116302013397, "learning_rate": 5.488020771245328e-06, "loss": 0.0032, "step": 85700 }, { "grad_norm": 0.07178892940282822, "learning_rate": 5.480491773084195e-06, "loss": 0.0031, "step": 85710 }, { "grad_norm": 0.09165825694799423, "learning_rate": 5.472967643509525e-06, "loss": 0.004, "step": 85720 }, { "grad_norm": 0.1400989294052124, "learning_rate": 5.46544838334419e-06, "loss": 0.0027, "step": 85730 }, { "grad_norm": 0.0855206549167633, "learning_rate": 5.4579339934104525e-06, "loss": 0.0037, "step": 85740 }, { "grad_norm": 0.14518281817436218, "learning_rate": 5.450424474530091e-06, "loss": 0.0031, "step": 85750 }, { "grad_norm": 0.13324832916259766, "learning_rate": 5.442919827524323e-06, "loss": 0.0025, "step": 85760 }, { "grad_norm": 0.16827575862407684, "learning_rate": 5.435420053213863e-06, "loss": 0.0033, "step": 85770 }, { "grad_norm": 0.12929996848106384, "learning_rate": 5.427925152418845e-06, "loss": 0.0024, "step": 85780 }, { "grad_norm": 0.11346284300088882, "learning_rate": 5.420435125958934e-06, "loss": 0.0033, "step": 85790 }, { "grad_norm": 0.1539023518562317, "learning_rate": 5.412949974653209e-06, "loss": 0.004, "step": 85800 }, { "grad_norm": 0.1179019957780838, "learning_rate": 5.405469699320248e-06, "loss": 0.0037, "step": 85810 }, { "grad_norm": 0.08850623667240143, "learning_rate": 5.397994300778064e-06, "loss": 0.0059, "step": 85820 }, { "grad_norm": 0.10237102955579758, "learning_rate": 5.39052377984417e-06, "loss": 0.0047, "step": 85830 }, { "grad_norm": 0.1539357602596283, "learning_rate": 5.383058137335523e-06, "loss": 0.0031, "step": 85840 }, { "grad_norm": 0.15328937768936157, "learning_rate": 5.375597374068553e-06, "loss": 0.0028, "step": 85850 }, { "grad_norm": 0.09038375318050385, "learning_rate": 5.368141490859163e-06, "loss": 0.0039, "step": 85860 }, { "grad_norm": 0.10431952029466629, "learning_rate": 5.360690488522729e-06, "loss": 0.0033, "step": 85870 }, { "grad_norm": 0.1169818863272667, "learning_rate": 5.353244367874055e-06, "loss": 0.0031, "step": 85880 }, { "grad_norm": 0.22179466485977173, "learning_rate": 5.345803129727456e-06, "loss": 0.0032, "step": 85890 }, { "grad_norm": 0.11772232502698898, "learning_rate": 5.338366774896697e-06, "loss": 0.0048, "step": 85900 }, { "grad_norm": 0.07372826337814331, "learning_rate": 5.330935304194995e-06, "loss": 0.0038, "step": 85910 }, { "grad_norm": 0.12233791500329971, "learning_rate": 5.323508718435044e-06, "loss": 0.0024, "step": 85920 }, { "grad_norm": 0.07261829823255539, "learning_rate": 5.316087018429011e-06, "loss": 0.0036, "step": 85930 }, { "grad_norm": 0.08882499486207962, "learning_rate": 5.30867020498853e-06, "loss": 0.003, "step": 85940 }, { "grad_norm": 0.10106945037841797, "learning_rate": 5.30125827892467e-06, "loss": 0.0037, "step": 85950 }, { "grad_norm": 0.09999462217092514, "learning_rate": 5.293851241048015e-06, "loss": 0.0052, "step": 85960 }, { "grad_norm": 0.17155322432518005, "learning_rate": 5.286449092168555e-06, "loss": 0.0032, "step": 85970 }, { "grad_norm": 0.14401055872440338, "learning_rate": 5.2790518330958105e-06, "loss": 0.0027, "step": 85980 }, { "grad_norm": 0.12021270394325256, "learning_rate": 5.271659464638717e-06, "loss": 0.0033, "step": 85990 }, { "grad_norm": 0.18178990483283997, "learning_rate": 5.264271987605701e-06, "loss": 0.0035, "step": 86000 }, { "grad_norm": 0.034223802387714386, "learning_rate": 5.256889402804632e-06, "loss": 0.0022, "step": 86010 }, { "grad_norm": 0.07299596816301346, "learning_rate": 5.249511711042881e-06, "loss": 0.003, "step": 86020 }, { "grad_norm": 0.10211484134197235, "learning_rate": 5.242138913127237e-06, "loss": 0.0042, "step": 86030 }, { "grad_norm": 0.10888327658176422, "learning_rate": 5.234771009863987e-06, "loss": 0.0029, "step": 86040 }, { "grad_norm": 0.08366430550813675, "learning_rate": 5.227408002058881e-06, "loss": 0.0035, "step": 86050 }, { "grad_norm": 0.13158883154392242, "learning_rate": 5.220049890517126e-06, "loss": 0.004, "step": 86060 }, { "grad_norm": 0.08085855841636658, "learning_rate": 5.212696676043377e-06, "loss": 0.0039, "step": 86070 }, { "grad_norm": 0.0972011536359787, "learning_rate": 5.205348359441798e-06, "loss": 0.0024, "step": 86080 }, { "grad_norm": 0.19520801305770874, "learning_rate": 5.198004941515966e-06, "loss": 0.0028, "step": 86090 }, { "grad_norm": 0.0962362065911293, "learning_rate": 5.190666423068952e-06, "loss": 0.0028, "step": 86100 }, { "grad_norm": 0.14541979134082794, "learning_rate": 5.183332804903284e-06, "loss": 0.0033, "step": 86110 }, { "grad_norm": 0.07065048068761826, "learning_rate": 5.176004087820979e-06, "loss": 0.0039, "step": 86120 }, { "grad_norm": 0.15662139654159546, "learning_rate": 5.168680272623455e-06, "loss": 0.0043, "step": 86130 }, { "grad_norm": 0.1953447014093399, "learning_rate": 5.161361360111661e-06, "loss": 0.0029, "step": 86140 }, { "grad_norm": 0.05743347480893135, "learning_rate": 5.154047351085984e-06, "loss": 0.0031, "step": 86150 }, { "grad_norm": 0.09433639049530029, "learning_rate": 5.1467382463462385e-06, "loss": 0.0027, "step": 86160 }, { "grad_norm": 0.036170605570077896, "learning_rate": 5.139434046691788e-06, "loss": 0.0022, "step": 86170 }, { "grad_norm": 0.28163185715675354, "learning_rate": 5.13213475292137e-06, "loss": 0.0024, "step": 86180 }, { "grad_norm": 0.09957470744848251, "learning_rate": 5.12484036583325e-06, "loss": 0.0038, "step": 86190 }, { "grad_norm": 0.06073421984910965, "learning_rate": 5.117550886225109e-06, "loss": 0.0035, "step": 86200 }, { "grad_norm": 0.13731320202350616, "learning_rate": 5.11026631489413e-06, "loss": 0.0021, "step": 86210 }, { "grad_norm": 0.16470220685005188, "learning_rate": 5.102986652636926e-06, "loss": 0.0039, "step": 86220 }, { "grad_norm": 0.12610137462615967, "learning_rate": 5.095711900249606e-06, "loss": 0.0051, "step": 86230 }, { "grad_norm": 0.0939207449555397, "learning_rate": 5.088442058527715e-06, "loss": 0.003, "step": 86240 }, { "grad_norm": 0.10141682624816895, "learning_rate": 5.08117712826629e-06, "loss": 0.003, "step": 86250 }, { "grad_norm": 0.08894967287778854, "learning_rate": 5.073917110259785e-06, "loss": 0.0039, "step": 86260 }, { "grad_norm": 0.11065414547920227, "learning_rate": 5.066662005302175e-06, "loss": 0.0033, "step": 86270 }, { "grad_norm": 0.04512261599302292, "learning_rate": 5.059411814186838e-06, "loss": 0.0047, "step": 86280 }, { "grad_norm": 0.08781083673238754, "learning_rate": 5.05216653770666e-06, "loss": 0.0018, "step": 86290 }, { "grad_norm": 0.047770727425813675, "learning_rate": 5.044926176653969e-06, "loss": 0.0029, "step": 86300 }, { "grad_norm": 0.09571905434131622, "learning_rate": 5.037690731820571e-06, "loss": 0.0034, "step": 86310 }, { "grad_norm": 0.08009449392557144, "learning_rate": 5.030460203997711e-06, "loss": 0.0027, "step": 86320 }, { "grad_norm": 0.15131504833698273, "learning_rate": 5.0232345939761174e-06, "loss": 0.0036, "step": 86330 }, { "grad_norm": 0.17252686619758606, "learning_rate": 5.016013902545957e-06, "loss": 0.0029, "step": 86340 }, { "grad_norm": 0.10570110380649567, "learning_rate": 5.008798130496889e-06, "loss": 0.0029, "step": 86350 }, { "grad_norm": 0.20123925805091858, "learning_rate": 5.001587278618008e-06, "loss": 0.0033, "step": 86360 }, { "grad_norm": 0.04409029334783554, "learning_rate": 4.994381347697891e-06, "loss": 0.0024, "step": 86370 }, { "grad_norm": 0.0941522866487503, "learning_rate": 4.987180338524572e-06, "loss": 0.0031, "step": 86380 }, { "grad_norm": 0.05975306034088135, "learning_rate": 4.9799842518855275e-06, "loss": 0.0039, "step": 86390 }, { "grad_norm": 0.09617183357477188, "learning_rate": 4.9727930885677285e-06, "loss": 0.0048, "step": 86400 }, { "grad_norm": 0.0819408968091011, "learning_rate": 4.965606849357573e-06, "loss": 0.0027, "step": 86410 }, { "grad_norm": 0.17028789222240448, "learning_rate": 4.9584255350409445e-06, "loss": 0.0037, "step": 86420 }, { "grad_norm": 0.2756916880607605, "learning_rate": 4.951249146403175e-06, "loss": 0.0053, "step": 86430 }, { "grad_norm": 0.17592483758926392, "learning_rate": 4.9440776842290815e-06, "loss": 0.003, "step": 86440 }, { "grad_norm": 0.13487063348293304, "learning_rate": 4.936911149302903e-06, "loss": 0.004, "step": 86450 }, { "grad_norm": 0.15001215040683746, "learning_rate": 4.929749542408379e-06, "loss": 0.0039, "step": 86460 }, { "grad_norm": 0.057160161435604095, "learning_rate": 4.922592864328673e-06, "loss": 0.0035, "step": 86470 }, { "grad_norm": 0.12227009236812592, "learning_rate": 4.915441115846436e-06, "loss": 0.0036, "step": 86480 }, { "grad_norm": 0.08031502366065979, "learning_rate": 4.908294297743782e-06, "loss": 0.0024, "step": 86490 }, { "grad_norm": 0.03881870582699776, "learning_rate": 4.901152410802273e-06, "loss": 0.0056, "step": 86500 }, { "grad_norm": 0.11306646466255188, "learning_rate": 4.894015455802919e-06, "loss": 0.0025, "step": 86510 }, { "grad_norm": 0.049130629748106, "learning_rate": 4.8868834335262346e-06, "loss": 0.0025, "step": 86520 }, { "grad_norm": 0.08896009624004364, "learning_rate": 4.87975634475214e-06, "loss": 0.0034, "step": 86530 }, { "grad_norm": 0.12082202732563019, "learning_rate": 4.8726341902600614e-06, "loss": 0.0019, "step": 86540 }, { "grad_norm": 0.09946572780609131, "learning_rate": 4.865516970828843e-06, "loss": 0.0026, "step": 86550 }, { "grad_norm": 0.2700616419315338, "learning_rate": 4.85840468723685e-06, "loss": 0.0035, "step": 86560 }, { "grad_norm": 0.08779799193143845, "learning_rate": 4.85129734026184e-06, "loss": 0.0028, "step": 86570 }, { "grad_norm": 0.06485693156719208, "learning_rate": 4.844194930681073e-06, "loss": 0.0026, "step": 86580 }, { "grad_norm": 0.07324442267417908, "learning_rate": 4.837097459271267e-06, "loss": 0.003, "step": 86590 }, { "grad_norm": 0.08154937624931335, "learning_rate": 4.830004926808573e-06, "loss": 0.0027, "step": 86600 }, { "grad_norm": 0.09134859591722488, "learning_rate": 4.822917334068627e-06, "loss": 0.0031, "step": 86610 }, { "grad_norm": 0.1496216058731079, "learning_rate": 4.81583468182652e-06, "loss": 0.0024, "step": 86620 }, { "grad_norm": 0.18317019939422607, "learning_rate": 4.808756970856809e-06, "loss": 0.0026, "step": 86630 }, { "grad_norm": 0.1338932365179062, "learning_rate": 4.801684201933476e-06, "loss": 0.0068, "step": 86640 }, { "grad_norm": 0.1304151862859726, "learning_rate": 4.794616375830019e-06, "loss": 0.007, "step": 86650 }, { "grad_norm": 0.09220434725284576, "learning_rate": 4.787553493319341e-06, "loss": 0.003, "step": 86660 }, { "grad_norm": 0.2934677302837372, "learning_rate": 4.780495555173836e-06, "loss": 0.0035, "step": 86670 }, { "grad_norm": 0.14627031981945038, "learning_rate": 4.7734425621653475e-06, "loss": 0.0034, "step": 86680 }, { "grad_norm": 0.11052171140909195, "learning_rate": 4.766394515065198e-06, "loss": 0.0051, "step": 86690 }, { "grad_norm": 0.10883722454309464, "learning_rate": 4.75935141464412e-06, "loss": 0.0031, "step": 86700 }, { "grad_norm": 0.08899638056755066, "learning_rate": 4.75231326167237e-06, "loss": 0.0041, "step": 86710 }, { "grad_norm": 0.07041613012552261, "learning_rate": 4.745280056919599e-06, "loss": 0.0035, "step": 86720 }, { "grad_norm": 0.11824311316013336, "learning_rate": 4.738251801154975e-06, "loss": 0.003, "step": 86730 }, { "grad_norm": 0.04334807023406029, "learning_rate": 4.731228495147061e-06, "loss": 0.0022, "step": 86740 }, { "grad_norm": 0.07175502926111221, "learning_rate": 4.7242101396639595e-06, "loss": 0.0047, "step": 86750 }, { "grad_norm": 0.0675983652472496, "learning_rate": 4.717196735473151e-06, "loss": 0.003, "step": 86760 }, { "grad_norm": 0.08774147927761078, "learning_rate": 4.710188283341643e-06, "loss": 0.0029, "step": 86770 }, { "grad_norm": 0.31634223461151123, "learning_rate": 4.703184784035835e-06, "loss": 0.004, "step": 86780 }, { "grad_norm": 0.22479775547981262, "learning_rate": 4.696186238321642e-06, "loss": 0.0036, "step": 86790 }, { "grad_norm": 0.08415810763835907, "learning_rate": 4.689192646964408e-06, "loss": 0.0029, "step": 86800 }, { "grad_norm": 0.06332094222307205, "learning_rate": 4.682204010728952e-06, "loss": 0.0026, "step": 86810 }, { "grad_norm": 0.10761581361293793, "learning_rate": 4.67522033037952e-06, "loss": 0.0028, "step": 86820 }, { "grad_norm": 0.07860587537288666, "learning_rate": 4.668241606679846e-06, "loss": 0.0024, "step": 86830 }, { "grad_norm": 0.09806251525878906, "learning_rate": 4.66126784039313e-06, "loss": 0.0034, "step": 86840 }, { "grad_norm": 0.10899841040372849, "learning_rate": 4.6542990322819836e-06, "loss": 0.003, "step": 86850 }, { "grad_norm": 0.04468033090233803, "learning_rate": 4.64733518310852e-06, "loss": 0.0044, "step": 86860 }, { "grad_norm": 0.056178461760282516, "learning_rate": 4.640376293634291e-06, "loss": 0.0021, "step": 86870 }, { "grad_norm": 0.05158097669482231, "learning_rate": 4.633422364620327e-06, "loss": 0.0034, "step": 86880 }, { "grad_norm": 0.10118051618337631, "learning_rate": 4.6264733968270745e-06, "loss": 0.0029, "step": 86890 }, { "grad_norm": 0.07746763527393341, "learning_rate": 4.619529391014482e-06, "loss": 0.0034, "step": 86900 }, { "grad_norm": 0.12642353773117065, "learning_rate": 4.61259034794192e-06, "loss": 0.0045, "step": 86910 }, { "grad_norm": 0.09933974593877792, "learning_rate": 4.605656268368247e-06, "loss": 0.0026, "step": 86920 }, { "grad_norm": 0.1310509443283081, "learning_rate": 4.598727153051735e-06, "loss": 0.0046, "step": 86930 }, { "grad_norm": 0.10734664648771286, "learning_rate": 4.591803002750178e-06, "loss": 0.0035, "step": 86940 }, { "grad_norm": 0.0618901401758194, "learning_rate": 4.58488381822077e-06, "loss": 0.0039, "step": 86950 }, { "grad_norm": 0.12599174678325653, "learning_rate": 4.5779696002201946e-06, "loss": 0.003, "step": 86960 }, { "grad_norm": 0.09747610986232758, "learning_rate": 4.571060349504563e-06, "loss": 0.0047, "step": 86970 }, { "grad_norm": 0.11291864514350891, "learning_rate": 4.564156066829478e-06, "loss": 0.0022, "step": 86980 }, { "grad_norm": 0.19877289235591888, "learning_rate": 4.557256752949957e-06, "loss": 0.0035, "step": 86990 }, { "grad_norm": 0.12551048398017883, "learning_rate": 4.550362408620534e-06, "loss": 0.0039, "step": 87000 }, { "grad_norm": 0.16159170866012573, "learning_rate": 4.543473034595136e-06, "loss": 0.006, "step": 87010 }, { "grad_norm": 0.0921575054526329, "learning_rate": 4.536588631627192e-06, "loss": 0.0026, "step": 87020 }, { "grad_norm": 0.10170509666204453, "learning_rate": 4.5297092004695565e-06, "loss": 0.0033, "step": 87030 }, { "grad_norm": 0.08945589512586594, "learning_rate": 4.522834741874555e-06, "loss": 0.0027, "step": 87040 }, { "grad_norm": 0.0921831950545311, "learning_rate": 4.515965256593974e-06, "loss": 0.0035, "step": 87050 }, { "grad_norm": 0.2892703115940094, "learning_rate": 4.509100745379058e-06, "loss": 0.006, "step": 87060 }, { "grad_norm": 0.14819715917110443, "learning_rate": 4.502241208980479e-06, "loss": 0.004, "step": 87070 }, { "grad_norm": 0.0977223739027977, "learning_rate": 4.495386648148398e-06, "loss": 0.0029, "step": 87080 }, { "grad_norm": 0.08601552248001099, "learning_rate": 4.488537063632431e-06, "loss": 0.0055, "step": 87090 }, { "grad_norm": 0.12171522527933121, "learning_rate": 4.481692456181608e-06, "loss": 0.004, "step": 87100 }, { "grad_norm": 0.13339795172214508, "learning_rate": 4.474852826544479e-06, "loss": 0.0043, "step": 87110 }, { "grad_norm": 0.07350623607635498, "learning_rate": 4.468018175468974e-06, "loss": 0.0026, "step": 87120 }, { "grad_norm": 0.14452745020389557, "learning_rate": 4.461188503702568e-06, "loss": 0.0045, "step": 87130 }, { "grad_norm": 0.11008471250534058, "learning_rate": 4.454363811992107e-06, "loss": 0.0029, "step": 87140 }, { "grad_norm": 0.08062984049320221, "learning_rate": 4.447544101083956e-06, "loss": 0.0017, "step": 87150 }, { "grad_norm": 0.11512500047683716, "learning_rate": 4.440729371723884e-06, "loss": 0.0041, "step": 87160 }, { "grad_norm": 0.09523488581180573, "learning_rate": 4.433919624657162e-06, "loss": 0.0031, "step": 87170 }, { "grad_norm": 0.06785088777542114, "learning_rate": 4.427114860628462e-06, "loss": 0.0032, "step": 87180 }, { "grad_norm": 0.15919294953346252, "learning_rate": 4.420315080381982e-06, "loss": 0.0036, "step": 87190 }, { "grad_norm": 0.15769341588020325, "learning_rate": 4.4135202846613045e-06, "loss": 0.0028, "step": 87200 }, { "grad_norm": 0.1640254110097885, "learning_rate": 4.406730474209525e-06, "loss": 0.0043, "step": 87210 }, { "grad_norm": 0.17210344970226288, "learning_rate": 4.399945649769138e-06, "loss": 0.0051, "step": 87220 }, { "grad_norm": 0.10425747185945511, "learning_rate": 4.393165812082151e-06, "loss": 0.0041, "step": 87230 }, { "grad_norm": 0.14228947460651398, "learning_rate": 4.386390961889969e-06, "loss": 0.0033, "step": 87240 }, { "grad_norm": 0.11663234978914261, "learning_rate": 4.3796210999334895e-06, "loss": 0.0028, "step": 87250 }, { "grad_norm": 0.09392917156219482, "learning_rate": 4.372856226953065e-06, "loss": 0.0041, "step": 87260 }, { "grad_norm": 0.17219345271587372, "learning_rate": 4.366096343688486e-06, "loss": 0.0031, "step": 87270 }, { "grad_norm": 0.1547825038433075, "learning_rate": 4.359341450878995e-06, "loss": 0.0039, "step": 87280 }, { "grad_norm": 0.06873888522386551, "learning_rate": 4.352591549263302e-06, "loss": 0.0046, "step": 87290 }, { "grad_norm": 0.0674171969294548, "learning_rate": 4.345846639579582e-06, "loss": 0.0043, "step": 87300 }, { "grad_norm": 0.1130194216966629, "learning_rate": 4.339106722565417e-06, "loss": 0.0029, "step": 87310 }, { "grad_norm": 0.06688360869884491, "learning_rate": 4.332371798957896e-06, "loss": 0.003, "step": 87320 }, { "grad_norm": 0.09893203526735306, "learning_rate": 4.3256418694935295e-06, "loss": 0.0033, "step": 87330 }, { "grad_norm": 0.11801386624574661, "learning_rate": 4.3189169349083125e-06, "loss": 0.0026, "step": 87340 }, { "grad_norm": 0.11036112904548645, "learning_rate": 4.312196995937645e-06, "loss": 0.0022, "step": 87350 }, { "grad_norm": 0.2684260606765747, "learning_rate": 4.305482053316434e-06, "loss": 0.0026, "step": 87360 }, { "grad_norm": 0.16379913687705994, "learning_rate": 4.2987721077789925e-06, "loss": 0.0032, "step": 87370 }, { "grad_norm": 0.07817158102989197, "learning_rate": 4.2920671600591325e-06, "loss": 0.0027, "step": 87380 }, { "grad_norm": 0.06416840106248856, "learning_rate": 4.285367210890084e-06, "loss": 0.0024, "step": 87390 }, { "grad_norm": 0.06953904777765274, "learning_rate": 4.278672261004551e-06, "loss": 0.0021, "step": 87400 }, { "grad_norm": 0.10244684666395187, "learning_rate": 4.2719823111346704e-06, "loss": 0.0032, "step": 87410 }, { "grad_norm": 0.07243230938911438, "learning_rate": 4.265297362012061e-06, "loss": 0.0026, "step": 87420 }, { "grad_norm": 0.08597361296415329, "learning_rate": 4.258617414367766e-06, "loss": 0.0026, "step": 87430 }, { "grad_norm": 0.1046937108039856, "learning_rate": 4.251942468932296e-06, "loss": 0.0028, "step": 87440 }, { "grad_norm": 0.18374964594841003, "learning_rate": 4.2452725264356175e-06, "loss": 0.0025, "step": 87450 }, { "grad_norm": 0.19723495841026306, "learning_rate": 4.2386075876071575e-06, "loss": 0.0025, "step": 87460 }, { "grad_norm": 0.10088082402944565, "learning_rate": 4.231947653175755e-06, "loss": 0.0034, "step": 87470 }, { "grad_norm": 0.12368778884410858, "learning_rate": 4.225292723869761e-06, "loss": 0.0035, "step": 87480 }, { "grad_norm": 0.1287606656551361, "learning_rate": 4.218642800416916e-06, "loss": 0.0026, "step": 87490 }, { "grad_norm": 0.1846238225698471, "learning_rate": 4.211997883544472e-06, "loss": 0.0062, "step": 87500 }, { "grad_norm": 0.06751491874456406, "learning_rate": 4.205357973979096e-06, "loss": 0.0036, "step": 87510 }, { "grad_norm": 0.11911787837743759, "learning_rate": 4.198723072446925e-06, "loss": 0.0042, "step": 87520 }, { "grad_norm": 0.05711512267589569, "learning_rate": 4.192093179673534e-06, "loss": 0.0025, "step": 87530 }, { "grad_norm": 0.06628524512052536, "learning_rate": 4.185468296383959e-06, "loss": 0.0032, "step": 87540 }, { "grad_norm": 0.1649901568889618, "learning_rate": 4.1788484233027005e-06, "loss": 0.0039, "step": 87550 }, { "grad_norm": 0.1236700713634491, "learning_rate": 4.172233561153677e-06, "loss": 0.0033, "step": 87560 }, { "grad_norm": 0.12706027925014496, "learning_rate": 4.165623710660288e-06, "loss": 0.0025, "step": 87570 }, { "grad_norm": 0.20259349048137665, "learning_rate": 4.1590188725453835e-06, "loss": 0.0043, "step": 87580 }, { "grad_norm": 0.0894789844751358, "learning_rate": 4.152419047531259e-06, "loss": 0.0026, "step": 87590 }, { "grad_norm": 0.09960606694221497, "learning_rate": 4.145824236339651e-06, "loss": 0.0056, "step": 87600 }, { "grad_norm": 0.12948006391525269, "learning_rate": 4.1392344396917695e-06, "loss": 0.0033, "step": 87610 }, { "grad_norm": 0.11541000753641129, "learning_rate": 4.132649658308252e-06, "loss": 0.0048, "step": 87620 }, { "grad_norm": 0.1282479614019394, "learning_rate": 4.126069892909213e-06, "loss": 0.0032, "step": 87630 }, { "grad_norm": 0.11746028810739517, "learning_rate": 4.119495144214191e-06, "loss": 0.0036, "step": 87640 }, { "grad_norm": 0.1644294112920761, "learning_rate": 4.112925412942214e-06, "loss": 0.0022, "step": 87650 }, { "grad_norm": 0.1202564463019371, "learning_rate": 4.106360699811712e-06, "loss": 0.0035, "step": 87660 }, { "grad_norm": 0.13969208300113678, "learning_rate": 4.099801005540616e-06, "loss": 0.0036, "step": 87670 }, { "grad_norm": 0.09440166503190994, "learning_rate": 4.093246330846256e-06, "loss": 0.0041, "step": 87680 }, { "grad_norm": 0.12687890231609344, "learning_rate": 4.086696676445467e-06, "loss": 0.004, "step": 87690 }, { "grad_norm": 0.2358497977256775, "learning_rate": 4.0801520430544914e-06, "loss": 0.0058, "step": 87700 }, { "grad_norm": 0.15092797577381134, "learning_rate": 4.073612431389062e-06, "loss": 0.0039, "step": 87710 }, { "grad_norm": 0.13433058559894562, "learning_rate": 4.067077842164319e-06, "loss": 0.0034, "step": 87720 }, { "grad_norm": 0.1241644099354744, "learning_rate": 4.0605482760948945e-06, "loss": 0.0027, "step": 87730 }, { "grad_norm": 0.08205144852399826, "learning_rate": 4.054023733894829e-06, "loss": 0.0032, "step": 87740 }, { "grad_norm": 0.1439240574836731, "learning_rate": 4.047504216277653e-06, "loss": 0.0023, "step": 87750 }, { "grad_norm": 0.1297004669904709, "learning_rate": 4.040989723956323e-06, "loss": 0.0035, "step": 87760 }, { "grad_norm": 0.06484046578407288, "learning_rate": 4.03448025764327e-06, "loss": 0.0028, "step": 87770 }, { "grad_norm": 0.10188659280538559, "learning_rate": 4.027975818050339e-06, "loss": 0.0028, "step": 87780 }, { "grad_norm": 0.10215181857347488, "learning_rate": 4.021476405888858e-06, "loss": 0.0038, "step": 87790 }, { "grad_norm": 0.11965293437242508, "learning_rate": 4.01498202186959e-06, "loss": 0.0021, "step": 87800 }, { "grad_norm": 0.12678955495357513, "learning_rate": 4.008492666702751e-06, "loss": 0.0025, "step": 87810 }, { "grad_norm": 0.06369120627641678, "learning_rate": 4.002008341098001e-06, "loss": 0.0028, "step": 87820 }, { "grad_norm": 0.12838730216026306, "learning_rate": 3.995529045764462e-06, "loss": 0.0032, "step": 87830 }, { "grad_norm": 0.05575452372431755, "learning_rate": 3.989054781410712e-06, "loss": 0.0044, "step": 87840 }, { "grad_norm": 0.08390261977910995, "learning_rate": 3.982585548744744e-06, "loss": 0.005, "step": 87850 }, { "grad_norm": 0.08294016122817993, "learning_rate": 3.9761213484740435e-06, "loss": 0.0043, "step": 87860 }, { "grad_norm": 0.08954343944787979, "learning_rate": 3.9696621813055055e-06, "loss": 0.0027, "step": 87870 }, { "grad_norm": 0.1441616266965866, "learning_rate": 3.9632080479455044e-06, "loss": 0.0032, "step": 87880 }, { "grad_norm": 0.40170320868492126, "learning_rate": 3.956758949099859e-06, "loss": 0.0041, "step": 87890 }, { "grad_norm": 0.350488543510437, "learning_rate": 3.9503148854738385e-06, "loss": 0.0042, "step": 87900 }, { "grad_norm": 0.09234870970249176, "learning_rate": 3.943875857772139e-06, "loss": 0.0026, "step": 87910 }, { "grad_norm": 0.09204641729593277, "learning_rate": 3.9374418666989375e-06, "loss": 0.0026, "step": 87920 }, { "grad_norm": 0.1716911494731903, "learning_rate": 3.931012912957838e-06, "loss": 0.0035, "step": 87930 }, { "grad_norm": 0.11437768489122391, "learning_rate": 3.9245889972519045e-06, "loss": 0.0019, "step": 87940 }, { "grad_norm": 0.14317898452281952, "learning_rate": 3.9181701202836265e-06, "loss": 0.0029, "step": 87950 }, { "grad_norm": 0.11556316912174225, "learning_rate": 3.911756282755003e-06, "loss": 0.0042, "step": 87960 }, { "grad_norm": 0.11572090536355972, "learning_rate": 3.905347485367411e-06, "loss": 0.0044, "step": 87970 }, { "grad_norm": 0.12531645596027374, "learning_rate": 3.898943728821725e-06, "loss": 0.0022, "step": 87980 }, { "grad_norm": 0.11395423859357834, "learning_rate": 3.892545013818227e-06, "loss": 0.0043, "step": 87990 }, { "grad_norm": 0.09464828670024872, "learning_rate": 3.886151341056687e-06, "loss": 0.005, "step": 88000 }, { "grad_norm": 0.07558848708868027, "learning_rate": 3.8797627112363066e-06, "loss": 0.0026, "step": 88010 }, { "grad_norm": 0.09880896657705307, "learning_rate": 3.873379125055743e-06, "loss": 0.0034, "step": 88020 }, { "grad_norm": 0.32596859335899353, "learning_rate": 3.867000583213082e-06, "loss": 0.0024, "step": 88030 }, { "grad_norm": 0.0809214636683464, "learning_rate": 3.860627086405871e-06, "loss": 0.0027, "step": 88040 }, { "grad_norm": 0.08945733308792114, "learning_rate": 3.854258635331126e-06, "loss": 0.0055, "step": 88050 }, { "grad_norm": 0.15211419761180878, "learning_rate": 3.8478952306852665e-06, "loss": 0.0035, "step": 88060 }, { "grad_norm": 0.11663034558296204, "learning_rate": 3.841536873164198e-06, "loss": 0.0027, "step": 88070 }, { "grad_norm": 0.1335754096508026, "learning_rate": 3.8351835634632626e-06, "loss": 0.0037, "step": 88080 }, { "grad_norm": 0.10181716829538345, "learning_rate": 3.828835302277245e-06, "loss": 0.0023, "step": 88090 }, { "grad_norm": 0.09246259927749634, "learning_rate": 3.822492090300378e-06, "loss": 0.0034, "step": 88100 }, { "grad_norm": 0.18094272911548615, "learning_rate": 3.8161539282263566e-06, "loss": 0.0042, "step": 88110 }, { "grad_norm": 0.13120996952056885, "learning_rate": 3.8098208167482995e-06, "loss": 0.0027, "step": 88120 }, { "grad_norm": 0.06289990991353989, "learning_rate": 3.803492756558796e-06, "loss": 0.0029, "step": 88130 }, { "grad_norm": 0.11864057183265686, "learning_rate": 3.797169748349855e-06, "loss": 0.0046, "step": 88140 }, { "grad_norm": 0.06543558835983276, "learning_rate": 3.790851792812977e-06, "loss": 0.0038, "step": 88150 }, { "grad_norm": 0.07410234957933426, "learning_rate": 3.784538890639072e-06, "loss": 0.0036, "step": 88160 }, { "grad_norm": 0.08062516897916794, "learning_rate": 3.7782310425185153e-06, "loss": 0.0033, "step": 88170 }, { "grad_norm": 0.04726797342300415, "learning_rate": 3.77192824914111e-06, "loss": 0.0019, "step": 88180 }, { "grad_norm": 0.11549672484397888, "learning_rate": 3.7656305111961333e-06, "loss": 0.0043, "step": 88190 }, { "grad_norm": 0.10994330793619156, "learning_rate": 3.7593378293722782e-06, "loss": 0.0035, "step": 88200 }, { "grad_norm": 0.12274885922670364, "learning_rate": 3.7530502043577333e-06, "loss": 0.0032, "step": 88210 }, { "grad_norm": 0.08475007116794586, "learning_rate": 3.746767636840076e-06, "loss": 0.0024, "step": 88220 }, { "grad_norm": 0.15933243930339813, "learning_rate": 3.74049012750638e-06, "loss": 0.0036, "step": 88230 }, { "grad_norm": 0.20305149257183075, "learning_rate": 3.734217677043128e-06, "loss": 0.0029, "step": 88240 }, { "grad_norm": 0.03944534808397293, "learning_rate": 3.727950286136267e-06, "loss": 0.0025, "step": 88250 }, { "grad_norm": 0.135105699300766, "learning_rate": 3.7216879554711925e-06, "loss": 0.0048, "step": 88260 }, { "grad_norm": 0.0822555273771286, "learning_rate": 3.715430685732757e-06, "loss": 0.0021, "step": 88270 }, { "grad_norm": 0.1660318225622177, "learning_rate": 3.7091784776052243e-06, "loss": 0.0037, "step": 88280 }, { "grad_norm": 0.11991376429796219, "learning_rate": 3.7029313317723356e-06, "loss": 0.0028, "step": 88290 }, { "grad_norm": 0.06732548028230667, "learning_rate": 3.6966892489172787e-06, "loss": 0.0032, "step": 88300 }, { "grad_norm": 0.11046479642391205, "learning_rate": 3.6904522297226575e-06, "loss": 0.0035, "step": 88310 }, { "grad_norm": 0.0771593451499939, "learning_rate": 3.6842202748705647e-06, "loss": 0.0026, "step": 88320 }, { "grad_norm": 0.06449420005083084, "learning_rate": 3.677993385042494e-06, "loss": 0.0045, "step": 88330 }, { "grad_norm": 0.04127972573041916, "learning_rate": 3.67177156091944e-06, "loss": 0.0029, "step": 88340 }, { "grad_norm": 0.07216352224349976, "learning_rate": 3.6655548031817866e-06, "loss": 0.0034, "step": 88350 }, { "grad_norm": 0.09219256788492203, "learning_rate": 3.6593431125094e-06, "loss": 0.0043, "step": 88360 }, { "grad_norm": 0.08626027405261993, "learning_rate": 3.6531364895815766e-06, "loss": 0.0025, "step": 88370 }, { "grad_norm": 0.09803038835525513, "learning_rate": 3.6469349350770677e-06, "loss": 0.0031, "step": 88380 }, { "grad_norm": 0.06950859725475311, "learning_rate": 3.6407384496740525e-06, "loss": 0.003, "step": 88390 }, { "grad_norm": 0.13681446015834808, "learning_rate": 3.6345470340501954e-06, "loss": 0.0044, "step": 88400 }, { "grad_norm": 0.0956176221370697, "learning_rate": 3.6283606888825537e-06, "loss": 0.0033, "step": 88410 }, { "grad_norm": 0.13920748233795166, "learning_rate": 3.622179414847676e-06, "loss": 0.005, "step": 88420 }, { "grad_norm": 0.07849764823913574, "learning_rate": 3.616003212621527e-06, "loss": 0.0029, "step": 88430 }, { "grad_norm": 0.13089720904827118, "learning_rate": 3.609832082879533e-06, "loss": 0.0033, "step": 88440 }, { "grad_norm": 0.09375660866498947, "learning_rate": 3.6036660262965485e-06, "loss": 0.0036, "step": 88450 }, { "grad_norm": 0.0957164466381073, "learning_rate": 3.597505043546895e-06, "loss": 0.0028, "step": 88460 }, { "grad_norm": 0.1726132482290268, "learning_rate": 3.5913491353043227e-06, "loss": 0.0036, "step": 88470 }, { "grad_norm": 0.1415078490972519, "learning_rate": 3.5851983022420375e-06, "loss": 0.0033, "step": 88480 }, { "grad_norm": 0.04526561498641968, "learning_rate": 3.5790525450326797e-06, "loss": 0.003, "step": 88490 }, { "grad_norm": 0.0958034098148346, "learning_rate": 3.572911864348344e-06, "loss": 0.0022, "step": 88500 }, { "grad_norm": 0.1673373281955719, "learning_rate": 3.5667762608605717e-06, "loss": 0.0032, "step": 88510 }, { "grad_norm": 0.15414966642856598, "learning_rate": 3.5606457352403255e-06, "loss": 0.0029, "step": 88520 }, { "grad_norm": 0.12508846819400787, "learning_rate": 3.554520288158042e-06, "loss": 0.0059, "step": 88530 }, { "grad_norm": 0.08593127131462097, "learning_rate": 3.5483999202835903e-06, "loss": 0.0025, "step": 88540 }, { "grad_norm": 0.07359173893928528, "learning_rate": 3.5422846322862913e-06, "loss": 0.0034, "step": 88550 }, { "grad_norm": 0.1001458391547203, "learning_rate": 3.5361744248348925e-06, "loss": 0.0029, "step": 88560 }, { "grad_norm": 0.07103387266397476, "learning_rate": 3.5300692985976047e-06, "loss": 0.0027, "step": 88570 }, { "grad_norm": 0.06898268312215805, "learning_rate": 3.52396925424206e-06, "loss": 0.0038, "step": 88580 }, { "grad_norm": 0.06819163262844086, "learning_rate": 3.5178742924353747e-06, "loss": 0.0024, "step": 88590 }, { "grad_norm": 0.13343338668346405, "learning_rate": 3.511784413844066e-06, "loss": 0.0025, "step": 88600 }, { "grad_norm": 0.10082706809043884, "learning_rate": 3.5056996191341287e-06, "loss": 0.0034, "step": 88610 }, { "grad_norm": 0.11468556523323059, "learning_rate": 3.4996199089709692e-06, "loss": 0.0028, "step": 88620 }, { "grad_norm": 0.12934105098247528, "learning_rate": 3.4935452840194725e-06, "loss": 0.0038, "step": 88630 }, { "grad_norm": 0.16499945521354675, "learning_rate": 3.48747574494393e-06, "loss": 0.0032, "step": 88640 }, { "grad_norm": 0.05961531400680542, "learning_rate": 3.4814112924081154e-06, "loss": 0.0027, "step": 88650 }, { "grad_norm": 0.04923977330327034, "learning_rate": 3.475351927075221e-06, "loss": 0.0024, "step": 88660 }, { "grad_norm": 0.12450390309095383, "learning_rate": 3.469297649607894e-06, "loss": 0.0028, "step": 88670 }, { "grad_norm": 0.07218541949987411, "learning_rate": 3.463248460668217e-06, "loss": 0.004, "step": 88680 }, { "grad_norm": 0.12322960793972015, "learning_rate": 3.457204360917726e-06, "loss": 0.0026, "step": 88690 }, { "grad_norm": 0.171687051653862, "learning_rate": 3.4511653510173827e-06, "loss": 0.0039, "step": 88700 }, { "grad_norm": 0.055776793509721756, "learning_rate": 3.4451314316276074e-06, "loss": 0.0029, "step": 88710 }, { "grad_norm": 0.13533064723014832, "learning_rate": 3.4391026034082674e-06, "loss": 0.0033, "step": 88720 }, { "grad_norm": 0.10931286960840225, "learning_rate": 3.433078867018674e-06, "loss": 0.0032, "step": 88730 }, { "grad_norm": 0.1171535924077034, "learning_rate": 3.427060223117551e-06, "loss": 0.005, "step": 88740 }, { "grad_norm": 0.10108904540538788, "learning_rate": 3.4210466723631042e-06, "loss": 0.0023, "step": 88750 }, { "grad_norm": 0.14852523803710938, "learning_rate": 3.415038215412969e-06, "loss": 0.0042, "step": 88760 }, { "grad_norm": 0.07840505242347717, "learning_rate": 3.4090348529242088e-06, "loss": 0.0037, "step": 88770 }, { "grad_norm": 0.12572526931762695, "learning_rate": 3.4030365855533487e-06, "loss": 0.003, "step": 88780 }, { "grad_norm": 0.1187434196472168, "learning_rate": 3.397043413956347e-06, "loss": 0.0047, "step": 88790 }, { "grad_norm": 0.07110773026943207, "learning_rate": 3.3910553387886237e-06, "loss": 0.0019, "step": 88800 }, { "grad_norm": 0.056099820882081985, "learning_rate": 3.3850723607049996e-06, "loss": 0.0037, "step": 88810 }, { "grad_norm": 0.1312280148267746, "learning_rate": 3.3790944803597902e-06, "loss": 0.0027, "step": 88820 }, { "grad_norm": 0.09173731505870819, "learning_rate": 3.373121698406706e-06, "loss": 0.0033, "step": 88830 }, { "grad_norm": 0.14837992191314697, "learning_rate": 3.3671540154989302e-06, "loss": 0.0047, "step": 88840 }, { "grad_norm": 0.10832253098487854, "learning_rate": 3.361191432289079e-06, "loss": 0.004, "step": 88850 }, { "grad_norm": 0.1642812043428421, "learning_rate": 3.3552339494292207e-06, "loss": 0.0038, "step": 88860 }, { "grad_norm": 0.07953982800245285, "learning_rate": 3.349281567570839e-06, "loss": 0.0029, "step": 88870 }, { "grad_norm": 0.11517000198364258, "learning_rate": 3.3433342873648965e-06, "loss": 0.004, "step": 88880 }, { "grad_norm": 0.14587034285068512, "learning_rate": 3.337392109461762e-06, "loss": 0.0037, "step": 88890 }, { "grad_norm": 0.06580626964569092, "learning_rate": 3.331455034511266e-06, "loss": 0.0025, "step": 88900 }, { "grad_norm": 0.14044393599033356, "learning_rate": 3.325523063162689e-06, "loss": 0.0037, "step": 88910 }, { "grad_norm": 0.07541549950838089, "learning_rate": 3.3195961960647393e-06, "loss": 0.0026, "step": 88920 }, { "grad_norm": 0.1807510107755661, "learning_rate": 3.31367443386556e-06, "loss": 0.0035, "step": 88930 }, { "grad_norm": 0.14172029495239258, "learning_rate": 3.3077577772127556e-06, "loss": 0.0024, "step": 88940 }, { "grad_norm": 0.09874605387449265, "learning_rate": 3.301846226753358e-06, "loss": 0.0032, "step": 88950 }, { "grad_norm": 0.051232025027275085, "learning_rate": 3.2959397831338444e-06, "loss": 0.0039, "step": 88960 }, { "grad_norm": 0.10111095011234283, "learning_rate": 3.2900384470001432e-06, "loss": 0.0034, "step": 88970 }, { "grad_norm": 0.1718132495880127, "learning_rate": 3.28414221899761e-06, "loss": 0.0029, "step": 88980 }, { "grad_norm": 0.16540028154850006, "learning_rate": 3.278251099771046e-06, "loss": 0.0038, "step": 88990 }, { "grad_norm": 0.04662125185132027, "learning_rate": 3.2723650899646906e-06, "loss": 0.0023, "step": 89000 }, { "grad_norm": 0.11815433949232101, "learning_rate": 3.2664841902222466e-06, "loss": 0.0029, "step": 89010 }, { "grad_norm": 0.11774808913469315, "learning_rate": 3.2606084011868208e-06, "loss": 0.003, "step": 89020 }, { "grad_norm": 0.033705517649650574, "learning_rate": 3.254737723500989e-06, "loss": 0.0023, "step": 89030 }, { "grad_norm": 0.17720234394073486, "learning_rate": 3.2488721578067595e-06, "loss": 0.0041, "step": 89040 }, { "grad_norm": 0.07831855863332748, "learning_rate": 3.243011704745591e-06, "loss": 0.0032, "step": 89050 }, { "grad_norm": 0.10215355455875397, "learning_rate": 3.237156364958355e-06, "loss": 0.0023, "step": 89060 }, { "grad_norm": 0.11623119562864304, "learning_rate": 3.231306139085394e-06, "loss": 0.0042, "step": 89070 }, { "grad_norm": 0.19056101143360138, "learning_rate": 3.2254610277664797e-06, "loss": 0.0039, "step": 89080 }, { "grad_norm": 0.06085546314716339, "learning_rate": 3.2196210316408183e-06, "loss": 0.0033, "step": 89090 }, { "grad_norm": 0.10781678557395935, "learning_rate": 3.2137861513470644e-06, "loss": 0.003, "step": 89100 }, { "grad_norm": 0.1351899355649948, "learning_rate": 3.2079563875233252e-06, "loss": 0.003, "step": 89110 }, { "grad_norm": 0.08641154319047928, "learning_rate": 3.2021317408071184e-06, "loss": 0.0045, "step": 89120 }, { "grad_norm": 0.05986398085951805, "learning_rate": 3.196312211835434e-06, "loss": 0.0023, "step": 89130 }, { "grad_norm": 0.1638065129518509, "learning_rate": 3.1904978012446686e-06, "loss": 0.0033, "step": 89140 }, { "grad_norm": 0.08201110363006592, "learning_rate": 3.184688509670691e-06, "loss": 0.0027, "step": 89150 }, { "grad_norm": 0.09040310978889465, "learning_rate": 3.178884337748783e-06, "loss": 0.0029, "step": 89160 }, { "grad_norm": 0.21940819919109344, "learning_rate": 3.173085286113697e-06, "loss": 0.0036, "step": 89170 }, { "grad_norm": 0.13651272654533386, "learning_rate": 3.1672913553995986e-06, "loss": 0.0033, "step": 89180 }, { "grad_norm": 0.07298759371042252, "learning_rate": 3.1615025462401138e-06, "loss": 0.0035, "step": 89190 }, { "grad_norm": 0.0720658153295517, "learning_rate": 3.1557188592682808e-06, "loss": 0.0031, "step": 89200 }, { "grad_norm": 0.0769738256931305, "learning_rate": 3.1499402951166045e-06, "loss": 0.0034, "step": 89210 }, { "grad_norm": 0.10838306695222855, "learning_rate": 3.144166854417019e-06, "loss": 0.0038, "step": 89220 }, { "grad_norm": 0.14920072257518768, "learning_rate": 3.1383985378009073e-06, "loss": 0.004, "step": 89230 }, { "grad_norm": 0.0910579189658165, "learning_rate": 3.1326353458990654e-06, "loss": 0.0042, "step": 89240 }, { "grad_norm": 0.07545853406190872, "learning_rate": 3.126877279341761e-06, "loss": 0.0027, "step": 89250 }, { "grad_norm": 0.12189323455095291, "learning_rate": 3.1211243387586964e-06, "loss": 0.0036, "step": 89260 }, { "grad_norm": 0.045178212225437164, "learning_rate": 3.115376524778979e-06, "loss": 0.0026, "step": 89270 }, { "grad_norm": 0.06314920634031296, "learning_rate": 3.109633838031195e-06, "loss": 0.0029, "step": 89280 }, { "grad_norm": 0.09758679568767548, "learning_rate": 3.103896279143359e-06, "loss": 0.0029, "step": 89290 }, { "grad_norm": 0.20157073438167572, "learning_rate": 3.098163848742924e-06, "loss": 0.0032, "step": 89300 }, { "grad_norm": 0.09676389396190643, "learning_rate": 3.0924365474567675e-06, "loss": 0.0044, "step": 89310 }, { "grad_norm": 0.11264258623123169, "learning_rate": 3.086714375911237e-06, "loss": 0.0028, "step": 89320 }, { "grad_norm": 0.0892946720123291, "learning_rate": 3.0809973347320775e-06, "loss": 0.0039, "step": 89330 }, { "grad_norm": 0.09653974324464798, "learning_rate": 3.075285424544516e-06, "loss": 0.0038, "step": 89340 }, { "grad_norm": 0.15911872684955597, "learning_rate": 3.069578645973176e-06, "loss": 0.0022, "step": 89350 }, { "grad_norm": 0.10834889858961105, "learning_rate": 3.0638769996421735e-06, "loss": 0.0026, "step": 89360 }, { "grad_norm": 0.06318704038858414, "learning_rate": 3.058180486175005e-06, "loss": 0.0018, "step": 89370 }, { "grad_norm": 0.07521826028823853, "learning_rate": 3.052489106194645e-06, "loss": 0.0027, "step": 89380 }, { "grad_norm": 0.05393045023083687, "learning_rate": 3.0468028603234946e-06, "loss": 0.0024, "step": 89390 }, { "grad_norm": 0.08087045699357986, "learning_rate": 3.04112174918339e-06, "loss": 0.0028, "step": 89400 }, { "grad_norm": 0.10663386434316635, "learning_rate": 3.0354457733955954e-06, "loss": 0.0022, "step": 89410 }, { "grad_norm": 0.10757981240749359, "learning_rate": 3.029774933580859e-06, "loss": 0.0048, "step": 89420 }, { "grad_norm": 0.18328304588794708, "learning_rate": 3.0241092303593065e-06, "loss": 0.003, "step": 89430 }, { "grad_norm": 0.052604567259550095, "learning_rate": 3.018448664350554e-06, "loss": 0.0024, "step": 89440 }, { "grad_norm": 0.14395762979984283, "learning_rate": 3.0127932361736055e-06, "loss": 0.0036, "step": 89450 }, { "grad_norm": 0.09234602749347687, "learning_rate": 3.00714294644695e-06, "loss": 0.0028, "step": 89460 }, { "grad_norm": 0.04689933732151985, "learning_rate": 3.0014977957884828e-06, "loss": 0.0021, "step": 89470 }, { "grad_norm": 0.1040067970752716, "learning_rate": 2.99585778481557e-06, "loss": 0.0051, "step": 89480 }, { "grad_norm": 0.15426206588745117, "learning_rate": 2.990222914144969e-06, "loss": 0.0025, "step": 89490 }, { "grad_norm": 0.07380201667547226, "learning_rate": 2.984593184392914e-06, "loss": 0.0033, "step": 89500 }, { "grad_norm": 0.14247487485408783, "learning_rate": 2.978968596175069e-06, "loss": 0.0029, "step": 89510 }, { "grad_norm": 0.15582266449928284, "learning_rate": 2.9733491501065137e-06, "loss": 0.0037, "step": 89520 }, { "grad_norm": 0.18795008957386017, "learning_rate": 2.9677348468018005e-06, "loss": 0.0036, "step": 89530 }, { "grad_norm": 0.08861282467842102, "learning_rate": 2.9621256868748726e-06, "loss": 0.0039, "step": 89540 }, { "grad_norm": 0.26476696133613586, "learning_rate": 2.956521670939183e-06, "loss": 0.0029, "step": 89550 }, { "grad_norm": 0.08971808105707169, "learning_rate": 2.950922799607542e-06, "loss": 0.0071, "step": 89560 }, { "grad_norm": 0.05270574241876602, "learning_rate": 2.9453290734922535e-06, "loss": 0.0029, "step": 89570 }, { "grad_norm": 0.07051508128643036, "learning_rate": 2.939740493205029e-06, "loss": 0.0033, "step": 89580 }, { "grad_norm": 0.04153670743107796, "learning_rate": 2.9341570593570345e-06, "loss": 0.004, "step": 89590 }, { "grad_norm": 0.11026795208454132, "learning_rate": 2.928578772558843e-06, "loss": 0.0029, "step": 89600 }, { "grad_norm": 0.10105002671480179, "learning_rate": 2.923005633420528e-06, "loss": 0.0036, "step": 89610 }, { "grad_norm": 0.12699532508850098, "learning_rate": 2.9174376425515237e-06, "loss": 0.0034, "step": 89620 }, { "grad_norm": 0.19006460905075073, "learning_rate": 2.91187480056076e-06, "loss": 0.0033, "step": 89630 }, { "grad_norm": 0.1108209639787674, "learning_rate": 2.9063171080565664e-06, "loss": 0.0021, "step": 89640 }, { "grad_norm": 0.133755624294281, "learning_rate": 2.900764565646741e-06, "loss": 0.0041, "step": 89650 }, { "grad_norm": 0.06278710067272186, "learning_rate": 2.895217173938475e-06, "loss": 0.0027, "step": 89660 }, { "grad_norm": 0.06721924990415573, "learning_rate": 2.8896749335384445e-06, "loss": 0.0025, "step": 89670 }, { "grad_norm": 0.07396034896373749, "learning_rate": 2.884137845052731e-06, "loss": 0.0024, "step": 89680 }, { "grad_norm": 0.10156256705522537, "learning_rate": 2.8786059090868734e-06, "loss": 0.0037, "step": 89690 }, { "grad_norm": 0.15770883858203888, "learning_rate": 2.87307912624582e-06, "loss": 0.0043, "step": 89700 }, { "grad_norm": 0.1239241436123848, "learning_rate": 2.867557497133977e-06, "loss": 0.0027, "step": 89710 }, { "grad_norm": 0.15032362937927246, "learning_rate": 2.862041022355194e-06, "loss": 0.0029, "step": 89720 }, { "grad_norm": 0.367891401052475, "learning_rate": 2.8565297025127226e-06, "loss": 0.0045, "step": 89730 }, { "grad_norm": 0.10374535620212555, "learning_rate": 2.8510235382092853e-06, "loss": 0.0045, "step": 89740 }, { "grad_norm": 0.051163800060749054, "learning_rate": 2.845522530047029e-06, "loss": 0.0021, "step": 89750 }, { "grad_norm": 0.17564888298511505, "learning_rate": 2.8400266786275387e-06, "loss": 0.005, "step": 89760 }, { "grad_norm": 0.08791756629943848, "learning_rate": 2.834535984551817e-06, "loss": 0.0038, "step": 89770 }, { "grad_norm": 0.11515305191278458, "learning_rate": 2.8290504484203397e-06, "loss": 0.0039, "step": 89780 }, { "grad_norm": 0.03906302899122238, "learning_rate": 2.8235700708329705e-06, "loss": 0.002, "step": 89790 }, { "grad_norm": 0.06073249876499176, "learning_rate": 2.818094852389058e-06, "loss": 0.0046, "step": 89800 }, { "grad_norm": 0.09506627917289734, "learning_rate": 2.8126247936873516e-06, "loss": 0.0024, "step": 89810 }, { "grad_norm": 0.24764809012413025, "learning_rate": 2.8071598953260614e-06, "loss": 0.0018, "step": 89820 }, { "grad_norm": 0.2221674770116806, "learning_rate": 2.8017001579028035e-06, "loss": 0.0027, "step": 89830 }, { "grad_norm": 0.0768623948097229, "learning_rate": 2.7962455820146617e-06, "loss": 0.0044, "step": 89840 }, { "grad_norm": 0.07753052562475204, "learning_rate": 2.7907961682581253e-06, "loss": 0.0028, "step": 89850 }, { "grad_norm": 0.08199840784072876, "learning_rate": 2.7853519172291453e-06, "loss": 0.0035, "step": 89860 }, { "grad_norm": 0.20715287327766418, "learning_rate": 2.7799128295230904e-06, "loss": 0.0022, "step": 89870 }, { "grad_norm": 0.09080992639064789, "learning_rate": 2.774478905734784e-06, "loss": 0.0024, "step": 89880 }, { "grad_norm": 0.1194123849272728, "learning_rate": 2.7690501464584516e-06, "loss": 0.0044, "step": 89890 }, { "grad_norm": 0.07398968189954758, "learning_rate": 2.763626552287796e-06, "loss": 0.0032, "step": 89900 }, { "grad_norm": 0.1514037549495697, "learning_rate": 2.7582081238159083e-06, "loss": 0.0051, "step": 89910 }, { "grad_norm": 0.22967980802059174, "learning_rate": 2.7527948616353548e-06, "loss": 0.0034, "step": 89920 }, { "grad_norm": 0.18927016854286194, "learning_rate": 2.7473867663381226e-06, "loss": 0.0032, "step": 89930 }, { "grad_norm": 0.10583575069904327, "learning_rate": 2.7419838385156383e-06, "loss": 0.0033, "step": 89940 }, { "grad_norm": 0.16361679136753082, "learning_rate": 2.7365860787587407e-06, "loss": 0.0034, "step": 89950 }, { "grad_norm": 0.08024729788303375, "learning_rate": 2.7311934876577295e-06, "loss": 0.0033, "step": 89960 }, { "grad_norm": 0.08787974715232849, "learning_rate": 2.725806065802339e-06, "loss": 0.0027, "step": 89970 }, { "grad_norm": 0.14802615344524384, "learning_rate": 2.720423813781714e-06, "loss": 0.0026, "step": 89980 }, { "grad_norm": 0.06318432092666626, "learning_rate": 2.715046732184462e-06, "loss": 0.0027, "step": 89990 }, { "grad_norm": 0.1319194883108139, "learning_rate": 2.7096748215986013e-06, "loss": 0.004, "step": 90000 }, { "grad_norm": 0.11483099311590195, "learning_rate": 2.704308082611612e-06, "loss": 0.0043, "step": 90010 }, { "grad_norm": 0.0955561101436615, "learning_rate": 2.698946515810374e-06, "loss": 0.0043, "step": 90020 }, { "grad_norm": 0.09366268664598465, "learning_rate": 2.6935901217812367e-06, "loss": 0.0026, "step": 90030 }, { "grad_norm": 0.12947092950344086, "learning_rate": 2.688238901109952e-06, "loss": 0.0032, "step": 90040 }, { "grad_norm": 0.14589573442935944, "learning_rate": 2.68289285438173e-06, "loss": 0.0029, "step": 90050 }, { "grad_norm": 0.12804090976715088, "learning_rate": 2.6775519821812033e-06, "loss": 0.0037, "step": 90060 }, { "grad_norm": 0.08781939744949341, "learning_rate": 2.672216285092449e-06, "loss": 0.0027, "step": 90070 }, { "grad_norm": 0.14490382373332977, "learning_rate": 2.666885763698962e-06, "loss": 0.0037, "step": 90080 }, { "grad_norm": 0.17034126818180084, "learning_rate": 2.661560418583692e-06, "loss": 0.0026, "step": 90090 }, { "grad_norm": 0.11189134418964386, "learning_rate": 2.656240250328995e-06, "loss": 0.0044, "step": 90100 }, { "grad_norm": 0.10101239383220673, "learning_rate": 2.650925259516679e-06, "loss": 0.003, "step": 90110 }, { "grad_norm": 0.4727228283882141, "learning_rate": 2.645615446727995e-06, "loss": 0.0022, "step": 90120 }, { "grad_norm": 0.10645093768835068, "learning_rate": 2.6403108125436115e-06, "loss": 0.0025, "step": 90130 }, { "grad_norm": 0.13086996972560883, "learning_rate": 2.6350113575436264e-06, "loss": 0.0041, "step": 90140 }, { "grad_norm": 0.11461703479290009, "learning_rate": 2.629717082307598e-06, "loss": 0.0032, "step": 90150 }, { "grad_norm": 0.18910019099712372, "learning_rate": 2.6244279874144793e-06, "loss": 0.0049, "step": 90160 }, { "grad_norm": 0.20887866616249084, "learning_rate": 2.6191440734426862e-06, "loss": 0.0022, "step": 90170 }, { "grad_norm": 0.07227269560098648, "learning_rate": 2.613865340970062e-06, "loss": 0.0028, "step": 90180 }, { "grad_norm": 0.09079885482788086, "learning_rate": 2.608591790573889e-06, "loss": 0.0023, "step": 90190 }, { "grad_norm": 0.13580666482448578, "learning_rate": 2.603323422830861e-06, "loss": 0.0032, "step": 90200 }, { "grad_norm": 0.04917800799012184, "learning_rate": 2.5980602383171183e-06, "loss": 0.0033, "step": 90210 }, { "grad_norm": 0.186879962682724, "learning_rate": 2.5928022376082496e-06, "loss": 0.0027, "step": 90220 }, { "grad_norm": 0.08173477649688721, "learning_rate": 2.5875494212792505e-06, "loss": 0.0032, "step": 90230 }, { "grad_norm": 0.13538184762001038, "learning_rate": 2.5823017899045565e-06, "loss": 0.0045, "step": 90240 }, { "grad_norm": 0.11328143626451492, "learning_rate": 2.577059344058047e-06, "loss": 0.0024, "step": 90250 }, { "grad_norm": 0.09573114663362503, "learning_rate": 2.5718220843130415e-06, "loss": 0.0028, "step": 90260 }, { "grad_norm": 0.13394726812839508, "learning_rate": 2.5665900112422535e-06, "loss": 0.0045, "step": 90270 }, { "grad_norm": 0.04827103018760681, "learning_rate": 2.561363125417876e-06, "loss": 0.0021, "step": 90280 }, { "grad_norm": 0.06901531666517258, "learning_rate": 2.556141427411496e-06, "loss": 0.0026, "step": 90290 }, { "grad_norm": 0.0862589180469513, "learning_rate": 2.550924917794156e-06, "loss": 0.0024, "step": 90300 }, { "grad_norm": 0.13419663906097412, "learning_rate": 2.545713597136329e-06, "loss": 0.0032, "step": 90310 }, { "grad_norm": 0.16429206728935242, "learning_rate": 2.5405074660079298e-06, "loss": 0.0043, "step": 90320 }, { "grad_norm": 0.12273652106523514, "learning_rate": 2.5353065249782647e-06, "loss": 0.0029, "step": 90330 }, { "grad_norm": 0.13462881743907928, "learning_rate": 2.530110774616129e-06, "loss": 0.0041, "step": 90340 }, { "grad_norm": 0.1355067491531372, "learning_rate": 2.524920215489701e-06, "loss": 0.0034, "step": 90350 }, { "grad_norm": 0.10108359903097153, "learning_rate": 2.519734848166627e-06, "loss": 0.0033, "step": 90360 }, { "grad_norm": 0.04939502850174904, "learning_rate": 2.514554673213948e-06, "loss": 0.0026, "step": 90370 }, { "grad_norm": 0.1002960130572319, "learning_rate": 2.5093796911981936e-06, "loss": 0.0038, "step": 90380 }, { "grad_norm": 0.06722813099622726, "learning_rate": 2.5042099026852672e-06, "loss": 0.002, "step": 90390 }, { "grad_norm": 0.1771184206008911, "learning_rate": 2.499045308240544e-06, "loss": 0.0044, "step": 90400 }, { "grad_norm": 0.09738084673881531, "learning_rate": 2.4938859084288003e-06, "loss": 0.0025, "step": 90410 }, { "grad_norm": 0.21347369253635406, "learning_rate": 2.4887317038142734e-06, "loss": 0.0043, "step": 90420 }, { "grad_norm": 0.06489714235067368, "learning_rate": 2.4835826949606124e-06, "loss": 0.0031, "step": 90430 }, { "grad_norm": 0.1770186424255371, "learning_rate": 2.4784388824309167e-06, "loss": 0.0029, "step": 90440 }, { "grad_norm": 0.09109929949045181, "learning_rate": 2.473300266787687e-06, "loss": 0.0046, "step": 90450 }, { "grad_norm": 0.13690003752708435, "learning_rate": 2.46816684859289e-06, "loss": 0.0031, "step": 90460 }, { "grad_norm": 0.11628375202417374, "learning_rate": 2.4630386284079043e-06, "loss": 0.0023, "step": 90470 }, { "grad_norm": 0.1422605961561203, "learning_rate": 2.4579156067935427e-06, "loss": 0.004, "step": 90480 }, { "grad_norm": 0.07920090854167938, "learning_rate": 2.452797784310046e-06, "loss": 0.0025, "step": 90490 }, { "grad_norm": 0.12631024420261383, "learning_rate": 2.447685161517105e-06, "loss": 0.0021, "step": 90500 }, { "grad_norm": 0.09911639988422394, "learning_rate": 2.4425777389738224e-06, "loss": 0.0027, "step": 90510 }, { "grad_norm": 0.15317267179489136, "learning_rate": 2.43747551723873e-06, "loss": 0.0046, "step": 90520 }, { "grad_norm": 0.20060031116008759, "learning_rate": 2.4323784968698136e-06, "loss": 0.0048, "step": 90530 }, { "grad_norm": 0.062433794140815735, "learning_rate": 2.4272866784244565e-06, "loss": 0.0021, "step": 90540 }, { "grad_norm": 0.11053086072206497, "learning_rate": 2.422200062459518e-06, "loss": 0.0028, "step": 90550 }, { "grad_norm": 0.2826220393180847, "learning_rate": 2.41711864953123e-06, "loss": 0.0047, "step": 90560 }, { "grad_norm": 0.13457168638706207, "learning_rate": 2.412042440195322e-06, "loss": 0.0057, "step": 90570 }, { "grad_norm": 0.08348053693771362, "learning_rate": 2.4069714350068993e-06, "loss": 0.0061, "step": 90580 }, { "grad_norm": 0.14524035155773163, "learning_rate": 2.4019056345205294e-06, "loss": 0.0035, "step": 90590 }, { "grad_norm": 0.07650905102491379, "learning_rate": 2.396845039290191e-06, "loss": 0.0024, "step": 90600 }, { "grad_norm": 0.08859895169734955, "learning_rate": 2.3917896498693194e-06, "loss": 0.0036, "step": 90610 }, { "grad_norm": 0.08538147807121277, "learning_rate": 2.3867394668107335e-06, "loss": 0.0039, "step": 90620 }, { "grad_norm": 0.1545225977897644, "learning_rate": 2.3816944906667527e-06, "loss": 0.0024, "step": 90630 }, { "grad_norm": 0.07264062762260437, "learning_rate": 2.3766547219890634e-06, "loss": 0.0037, "step": 90640 }, { "grad_norm": 0.09015937894582748, "learning_rate": 2.371620161328819e-06, "loss": 0.0034, "step": 90650 }, { "grad_norm": 0.1228237971663475, "learning_rate": 2.366590809236574e-06, "loss": 0.0038, "step": 90660 }, { "grad_norm": 0.15730343759059906, "learning_rate": 2.3615666662623494e-06, "loss": 0.0032, "step": 90670 }, { "grad_norm": 0.08829399943351746, "learning_rate": 2.3565477329555664e-06, "loss": 0.0029, "step": 90680 }, { "grad_norm": 0.04063351824879646, "learning_rate": 2.3515340098650975e-06, "loss": 0.003, "step": 90690 }, { "grad_norm": 0.08691377937793732, "learning_rate": 2.3465254975392258e-06, "loss": 0.0027, "step": 90700 }, { "grad_norm": 0.0781366303563118, "learning_rate": 2.3415221965256805e-06, "loss": 0.0024, "step": 90710 }, { "grad_norm": 0.06054304167628288, "learning_rate": 2.3365241073716237e-06, "loss": 0.0034, "step": 90720 }, { "grad_norm": 0.08387605845928192, "learning_rate": 2.3315312306236183e-06, "loss": 0.0033, "step": 90730 }, { "grad_norm": 0.11225122958421707, "learning_rate": 2.326543566827699e-06, "loss": 0.0034, "step": 90740 }, { "grad_norm": 0.12351324409246445, "learning_rate": 2.3215611165292862e-06, "loss": 0.0047, "step": 90750 }, { "grad_norm": 0.0807250589132309, "learning_rate": 2.3165838802732765e-06, "loss": 0.004, "step": 90760 }, { "grad_norm": 0.0784771516919136, "learning_rate": 2.311611858603957e-06, "loss": 0.0031, "step": 90770 }, { "grad_norm": 0.13245868682861328, "learning_rate": 2.3066450520650763e-06, "loss": 0.0029, "step": 90780 }, { "grad_norm": 0.03919927030801773, "learning_rate": 2.3016834611997773e-06, "loss": 0.0052, "step": 90790 }, { "grad_norm": 0.08408334851264954, "learning_rate": 2.296727086550665e-06, "loss": 0.0044, "step": 90800 }, { "grad_norm": 0.1669754534959793, "learning_rate": 2.2917759286597507e-06, "loss": 0.0044, "step": 90810 }, { "grad_norm": 0.1023552268743515, "learning_rate": 2.2868299880685006e-06, "loss": 0.0027, "step": 90820 }, { "grad_norm": 0.1911320835351944, "learning_rate": 2.2818892653177827e-06, "loss": 0.0037, "step": 90830 }, { "grad_norm": 0.11151022464036942, "learning_rate": 2.2769537609479196e-06, "loss": 0.0033, "step": 90840 }, { "grad_norm": 0.12385554611682892, "learning_rate": 2.2720234754986357e-06, "loss": 0.0023, "step": 90850 }, { "grad_norm": 0.0809771791100502, "learning_rate": 2.2670984095091108e-06, "loss": 0.0023, "step": 90860 }, { "grad_norm": 0.14233237504959106, "learning_rate": 2.262178563517936e-06, "loss": 0.0033, "step": 90870 }, { "grad_norm": 0.10927853733301163, "learning_rate": 2.257263938063137e-06, "loss": 0.0042, "step": 90880 }, { "grad_norm": 0.1348179578781128, "learning_rate": 2.2523545336821715e-06, "loss": 0.0047, "step": 90890 }, { "grad_norm": 0.26916927099227905, "learning_rate": 2.247450350911939e-06, "loss": 0.0038, "step": 90900 }, { "grad_norm": 0.11402680724859238, "learning_rate": 2.2425513902887328e-06, "loss": 0.0068, "step": 90910 }, { "grad_norm": 0.13119840621948242, "learning_rate": 2.2376576523483015e-06, "loss": 0.0034, "step": 90920 }, { "grad_norm": 0.06490004807710648, "learning_rate": 2.232769137625829e-06, "loss": 0.0038, "step": 90930 }, { "grad_norm": 0.07891571521759033, "learning_rate": 2.227885846655903e-06, "loss": 0.0038, "step": 90940 }, { "grad_norm": 0.08451838046312332, "learning_rate": 2.2230077799725523e-06, "loss": 0.0037, "step": 90950 }, { "grad_norm": 0.06880062073469162, "learning_rate": 2.2181349381092396e-06, "loss": 0.0025, "step": 90960 }, { "grad_norm": 0.09452743083238602, "learning_rate": 2.2132673215988552e-06, "loss": 0.0022, "step": 90970 }, { "grad_norm": 0.09408309310674667, "learning_rate": 2.2084049309737065e-06, "loss": 0.004, "step": 90980 }, { "grad_norm": 0.14042994379997253, "learning_rate": 2.2035477667655513e-06, "loss": 0.0028, "step": 90990 }, { "grad_norm": 0.06473281234502792, "learning_rate": 2.1986958295055315e-06, "loss": 0.0023, "step": 91000 }, { "grad_norm": 0.14105355739593506, "learning_rate": 2.193849119724284e-06, "loss": 0.0025, "step": 91010 }, { "grad_norm": 0.11726590991020203, "learning_rate": 2.1890076379518175e-06, "loss": 0.0036, "step": 91020 }, { "grad_norm": 0.12188500910997391, "learning_rate": 2.1841713847175925e-06, "loss": 0.0023, "step": 91030 }, { "grad_norm": 0.05842864140868187, "learning_rate": 2.1793403605504968e-06, "loss": 0.0032, "step": 91040 }, { "grad_norm": 0.13167724013328552, "learning_rate": 2.1745145659788413e-06, "loss": 0.0033, "step": 91050 }, { "grad_norm": 0.23371991515159607, "learning_rate": 2.1696940015303645e-06, "loss": 0.0028, "step": 91060 }, { "grad_norm": 0.09018134325742722, "learning_rate": 2.1648786677322397e-06, "loss": 0.0045, "step": 91070 }, { "grad_norm": 0.1577141433954239, "learning_rate": 2.1600685651110673e-06, "loss": 0.0034, "step": 91080 }, { "grad_norm": 0.07564052194356918, "learning_rate": 2.1552636941928717e-06, "loss": 0.0034, "step": 91090 }, { "grad_norm": 0.07935521006584167, "learning_rate": 2.1504640555031043e-06, "loss": 0.0034, "step": 91100 }, { "grad_norm": 0.10680373758077621, "learning_rate": 2.1456696495666506e-06, "loss": 0.0039, "step": 91110 }, { "grad_norm": 0.04985867068171501, "learning_rate": 2.1408804769078084e-06, "loss": 0.0022, "step": 91120 }, { "grad_norm": 0.04349647834897041, "learning_rate": 2.136096538050325e-06, "loss": 0.0045, "step": 91130 }, { "grad_norm": 0.08950909972190857, "learning_rate": 2.13131783351736e-06, "loss": 0.0029, "step": 91140 }, { "grad_norm": 0.09144969284534454, "learning_rate": 2.126544363831512e-06, "loss": 0.0034, "step": 91150 }, { "grad_norm": 0.18015503883361816, "learning_rate": 2.121776129514791e-06, "loss": 0.0025, "step": 91160 }, { "grad_norm": 0.17426054179668427, "learning_rate": 2.1170131310886523e-06, "loss": 0.0038, "step": 91170 }, { "grad_norm": 0.21430648863315582, "learning_rate": 2.112255369073968e-06, "loss": 0.0033, "step": 91180 }, { "grad_norm": 0.37044817209243774, "learning_rate": 2.1075028439910328e-06, "loss": 0.0031, "step": 91190 }, { "grad_norm": 0.12636317312717438, "learning_rate": 2.1027555563595813e-06, "loss": 0.0028, "step": 91200 }, { "grad_norm": 0.12333405762910843, "learning_rate": 2.0980135066987706e-06, "loss": 0.0028, "step": 91210 }, { "grad_norm": 0.1454010158777237, "learning_rate": 2.093276695527191e-06, "loss": 0.0038, "step": 91220 }, { "grad_norm": 0.16773459315299988, "learning_rate": 2.0885451233628396e-06, "loss": 0.0045, "step": 91230 }, { "grad_norm": 0.08380237221717834, "learning_rate": 2.0838187907231633e-06, "loss": 0.0045, "step": 91240 }, { "grad_norm": 0.123326875269413, "learning_rate": 2.0790976981250153e-06, "loss": 0.0045, "step": 91250 }, { "grad_norm": 0.2261296957731247, "learning_rate": 2.0743818460847e-06, "loss": 0.0033, "step": 91260 }, { "grad_norm": 0.25141507387161255, "learning_rate": 2.069671235117926e-06, "loss": 0.0034, "step": 91270 }, { "grad_norm": 0.08094882220029831, "learning_rate": 2.0649658657398484e-06, "loss": 0.0028, "step": 91280 }, { "grad_norm": 0.30417197942733765, "learning_rate": 2.060265738465034e-06, "loss": 0.0024, "step": 91290 }, { "grad_norm": 0.09615931659936905, "learning_rate": 2.055570853807487e-06, "loss": 0.0025, "step": 91300 }, { "grad_norm": 0.1358153522014618, "learning_rate": 2.0508812122806196e-06, "loss": 0.0037, "step": 91310 }, { "grad_norm": 0.09837467223405838, "learning_rate": 2.046196814397294e-06, "loss": 0.0022, "step": 91320 }, { "grad_norm": 0.07370778918266296, "learning_rate": 2.0415176606697896e-06, "loss": 0.0026, "step": 91330 }, { "grad_norm": 0.11327166855335236, "learning_rate": 2.0368437516098126e-06, "loss": 0.0036, "step": 91340 }, { "grad_norm": 0.13306714594364166, "learning_rate": 2.032175087728483e-06, "loss": 0.0035, "step": 91350 }, { "grad_norm": 0.1234428882598877, "learning_rate": 2.027511669536375e-06, "loss": 0.0024, "step": 91360 }, { "grad_norm": 0.06786354631185532, "learning_rate": 2.0228534975434643e-06, "loss": 0.0053, "step": 91370 }, { "grad_norm": 0.10245595872402191, "learning_rate": 2.0182005722591548e-06, "loss": 0.003, "step": 91380 }, { "grad_norm": 0.10767069458961487, "learning_rate": 2.0135528941922944e-06, "loss": 0.0031, "step": 91390 }, { "grad_norm": 0.12927059829235077, "learning_rate": 2.008910463851149e-06, "loss": 0.0024, "step": 91400 }, { "grad_norm": 0.1614651083946228, "learning_rate": 2.0042732817433963e-06, "loss": 0.0022, "step": 91410 }, { "grad_norm": 0.15033866465091705, "learning_rate": 1.999641348376158e-06, "loss": 0.0033, "step": 91420 }, { "grad_norm": 0.12008870393037796, "learning_rate": 1.9950146642559787e-06, "loss": 0.0026, "step": 91430 }, { "grad_norm": 0.03681626543402672, "learning_rate": 1.990393229888815e-06, "loss": 0.0022, "step": 91440 }, { "grad_norm": 0.125443235039711, "learning_rate": 1.9857770457800684e-06, "loss": 0.0048, "step": 91450 }, { "grad_norm": 0.1394132673740387, "learning_rate": 1.9811661124345517e-06, "loss": 0.0029, "step": 91460 }, { "grad_norm": 0.10277043282985687, "learning_rate": 1.976560430356522e-06, "loss": 0.0053, "step": 91470 }, { "grad_norm": 0.12323231250047684, "learning_rate": 1.9719600000496383e-06, "loss": 0.0029, "step": 91480 }, { "grad_norm": 0.1408642828464508, "learning_rate": 1.967364822017004e-06, "loss": 0.0037, "step": 91490 }, { "grad_norm": 0.15541422367095947, "learning_rate": 1.962774896761127e-06, "loss": 0.0026, "step": 91500 }, { "grad_norm": 0.056264858692884445, "learning_rate": 1.958190224783968e-06, "loss": 0.0024, "step": 91510 }, { "grad_norm": 0.12372898310422897, "learning_rate": 1.9536108065868975e-06, "loss": 0.0028, "step": 91520 }, { "grad_norm": 0.10022952407598495, "learning_rate": 1.949036642670715e-06, "loss": 0.0055, "step": 91530 }, { "grad_norm": 0.17118524014949799, "learning_rate": 1.9444677335356377e-06, "loss": 0.0023, "step": 91540 }, { "grad_norm": 0.097080759704113, "learning_rate": 1.939904079681326e-06, "loss": 0.0032, "step": 91550 }, { "grad_norm": 0.26108455657958984, "learning_rate": 1.935345681606837e-06, "loss": 0.0036, "step": 91560 }, { "grad_norm": 0.14978361129760742, "learning_rate": 1.930792539810694e-06, "loss": 0.0022, "step": 91570 }, { "grad_norm": 0.1497868150472641, "learning_rate": 1.9262446547907877e-06, "loss": 0.0031, "step": 91580 }, { "grad_norm": 0.1664258986711502, "learning_rate": 1.921702027044503e-06, "loss": 0.0031, "step": 91590 }, { "grad_norm": 0.15042422711849213, "learning_rate": 1.917164657068593e-06, "loss": 0.0025, "step": 91600 }, { "grad_norm": 0.38407889008522034, "learning_rate": 1.912632545359272e-06, "loss": 0.0028, "step": 91610 }, { "grad_norm": 0.10394388437271118, "learning_rate": 1.9081056924121488e-06, "loss": 0.0039, "step": 91620 }, { "grad_norm": 0.12269585579633713, "learning_rate": 1.9035840987222887e-06, "loss": 0.0033, "step": 91630 }, { "grad_norm": 0.08037354797124863, "learning_rate": 1.8990677647841515e-06, "loss": 0.0038, "step": 91640 }, { "grad_norm": 0.05856267362833023, "learning_rate": 1.8945566910916534e-06, "loss": 0.0021, "step": 91650 }, { "grad_norm": 0.13028469681739807, "learning_rate": 1.8900508781381054e-06, "loss": 0.0033, "step": 91660 }, { "grad_norm": 0.0863872542977333, "learning_rate": 1.8855503264162633e-06, "loss": 0.0035, "step": 91670 }, { "grad_norm": 0.10678023099899292, "learning_rate": 1.8810550364182998e-06, "loss": 0.0039, "step": 91680 }, { "grad_norm": 0.10523762553930283, "learning_rate": 1.8765650086358112e-06, "loss": 0.0027, "step": 91690 }, { "grad_norm": 0.09018941223621368, "learning_rate": 1.8720802435598151e-06, "loss": 0.0036, "step": 91700 }, { "grad_norm": 0.1498570293188095, "learning_rate": 1.8676007416807694e-06, "loss": 0.0022, "step": 91710 }, { "grad_norm": 0.03959173709154129, "learning_rate": 1.8631265034885437e-06, "loss": 0.0047, "step": 91720 }, { "grad_norm": 0.06755200028419495, "learning_rate": 1.858657529472424e-06, "loss": 0.0047, "step": 91730 }, { "grad_norm": 0.08644239604473114, "learning_rate": 1.8541938201211473e-06, "loss": 0.0031, "step": 91740 }, { "grad_norm": 0.05772208794951439, "learning_rate": 1.8497353759228398e-06, "loss": 0.0035, "step": 91750 }, { "grad_norm": 0.1271832138299942, "learning_rate": 1.8452821973650835e-06, "loss": 0.0032, "step": 91760 }, { "grad_norm": 0.150387704372406, "learning_rate": 1.840834284934856e-06, "loss": 0.0043, "step": 91770 }, { "grad_norm": 0.03807058557868004, "learning_rate": 1.8363916391185954e-06, "loss": 0.0017, "step": 91780 }, { "grad_norm": 0.16895461082458496, "learning_rate": 1.8319542604021245e-06, "loss": 0.0051, "step": 91790 }, { "grad_norm": 0.07791095972061157, "learning_rate": 1.8275221492707273e-06, "loss": 0.0047, "step": 91800 }, { "grad_norm": 0.09926360845565796, "learning_rate": 1.8230953062090717e-06, "loss": 0.0034, "step": 91810 }, { "grad_norm": 0.10744553059339523, "learning_rate": 1.8186737317012926e-06, "loss": 0.0029, "step": 91820 }, { "grad_norm": 0.27315300703048706, "learning_rate": 1.8142574262308977e-06, "loss": 0.0035, "step": 91830 }, { "grad_norm": 0.14515039324760437, "learning_rate": 1.8098463902808727e-06, "loss": 0.0051, "step": 91840 }, { "grad_norm": 0.09254801273345947, "learning_rate": 1.805440624333593e-06, "loss": 0.0024, "step": 91850 }, { "grad_norm": 0.06808379292488098, "learning_rate": 1.8010401288708734e-06, "loss": 0.0023, "step": 91860 }, { "grad_norm": 0.10203877091407776, "learning_rate": 1.7966449043739343e-06, "loss": 0.0032, "step": 91870 }, { "grad_norm": 0.22005727887153625, "learning_rate": 1.79225495132343e-06, "loss": 0.003, "step": 91880 }, { "grad_norm": 0.062327392399311066, "learning_rate": 1.7878702701994487e-06, "loss": 0.0025, "step": 91890 }, { "grad_norm": 0.16778354346752167, "learning_rate": 1.783490861481496e-06, "loss": 0.0024, "step": 91900 }, { "grad_norm": 0.14856179058551788, "learning_rate": 1.7791167256484832e-06, "loss": 0.0035, "step": 91910 }, { "grad_norm": 0.17838971316814423, "learning_rate": 1.774747863178766e-06, "loss": 0.0045, "step": 91920 }, { "grad_norm": 0.11129548400640488, "learning_rate": 1.7703842745501186e-06, "loss": 0.0029, "step": 91930 }, { "grad_norm": 0.22081856429576874, "learning_rate": 1.7660259602397366e-06, "loss": 0.0031, "step": 91940 }, { "grad_norm": 0.141891747713089, "learning_rate": 1.7616729207242388e-06, "loss": 0.0047, "step": 91950 }, { "grad_norm": 0.09125497192144394, "learning_rate": 1.7573251564796556e-06, "loss": 0.0026, "step": 91960 }, { "grad_norm": 0.13184021413326263, "learning_rate": 1.7529826679814676e-06, "loss": 0.0026, "step": 91970 }, { "grad_norm": 0.14854736626148224, "learning_rate": 1.7486454557045562e-06, "loss": 0.0055, "step": 91980 }, { "grad_norm": 0.11285325139760971, "learning_rate": 1.7443135201232364e-06, "loss": 0.003, "step": 91990 }, { "grad_norm": 0.08595361560583115, "learning_rate": 1.7399868617112347e-06, "loss": 0.0038, "step": 92000 }, { "grad_norm": 0.1483195275068283, "learning_rate": 1.7356654809417171e-06, "loss": 0.0027, "step": 92010 }, { "grad_norm": 0.05636192113161087, "learning_rate": 1.7313493782872448e-06, "loss": 0.0053, "step": 92020 }, { "grad_norm": 0.030721621587872505, "learning_rate": 1.7270385542198453e-06, "loss": 0.0026, "step": 92030 }, { "grad_norm": 0.07418230921030045, "learning_rate": 1.7227330092109307e-06, "loss": 0.003, "step": 92040 }, { "grad_norm": 0.10580306500196457, "learning_rate": 1.718432743731352e-06, "loss": 0.0053, "step": 92050 }, { "grad_norm": 0.11359909176826477, "learning_rate": 1.7141377582513718e-06, "loss": 0.0033, "step": 92060 }, { "grad_norm": 0.0822015032172203, "learning_rate": 1.709848053240698e-06, "loss": 0.0025, "step": 92070 }, { "grad_norm": 0.09574254602193832, "learning_rate": 1.7055636291684328e-06, "loss": 0.0082, "step": 92080 }, { "grad_norm": 0.1081201508641243, "learning_rate": 1.7012844865031185e-06, "loss": 0.0029, "step": 92090 }, { "grad_norm": 0.06912694871425629, "learning_rate": 1.6970106257127138e-06, "loss": 0.002, "step": 92100 }, { "grad_norm": 0.5404561161994934, "learning_rate": 1.6927420472646117e-06, "loss": 0.0033, "step": 92110 }, { "grad_norm": 0.17566229403018951, "learning_rate": 1.6884787516256106e-06, "loss": 0.0021, "step": 92120 }, { "grad_norm": 0.2040206640958786, "learning_rate": 1.6842207392619325e-06, "loss": 0.0032, "step": 92130 }, { "grad_norm": 0.1668044477701187, "learning_rate": 1.679968010639238e-06, "loss": 0.0039, "step": 92140 }, { "grad_norm": 0.08530048280954361, "learning_rate": 1.6757205662225829e-06, "loss": 0.0035, "step": 92150 }, { "grad_norm": 0.11350247263908386, "learning_rate": 1.6714784064764844e-06, "loss": 0.0027, "step": 92160 }, { "grad_norm": 0.14905677735805511, "learning_rate": 1.6672415318648383e-06, "loss": 0.0059, "step": 92170 }, { "grad_norm": 0.08411303907632828, "learning_rate": 1.6630099428510016e-06, "loss": 0.003, "step": 92180 }, { "grad_norm": 0.1003541573882103, "learning_rate": 1.6587836398977153e-06, "loss": 0.0031, "step": 92190 }, { "grad_norm": 0.1235307827591896, "learning_rate": 1.6545626234671762e-06, "loss": 0.0042, "step": 92200 }, { "grad_norm": 0.10057804733514786, "learning_rate": 1.650346894020971e-06, "loss": 0.0045, "step": 92210 }, { "grad_norm": 0.06624999642372131, "learning_rate": 1.6461364520201472e-06, "loss": 0.0036, "step": 92220 }, { "grad_norm": 0.06937529891729355, "learning_rate": 1.6419312979251366e-06, "loss": 0.0047, "step": 92230 }, { "grad_norm": 0.11927247792482376, "learning_rate": 1.6377314321958214e-06, "loss": 0.0036, "step": 92240 }, { "grad_norm": 0.20451293885707855, "learning_rate": 1.6335368552914787e-06, "loss": 0.0028, "step": 92250 }, { "grad_norm": 0.09823481738567352, "learning_rate": 1.6293475676708358e-06, "loss": 0.0031, "step": 92260 }, { "grad_norm": 0.20942457020282745, "learning_rate": 1.6251635697920097e-06, "loss": 0.0024, "step": 92270 }, { "grad_norm": 0.06842217594385147, "learning_rate": 1.6209848621125624e-06, "loss": 0.003, "step": 92280 }, { "grad_norm": 0.15236712992191315, "learning_rate": 1.6168114450894778e-06, "loss": 0.0049, "step": 92290 }, { "grad_norm": 0.07253444194793701, "learning_rate": 1.612643319179158e-06, "loss": 0.0027, "step": 92300 }, { "grad_norm": 0.12913653254508972, "learning_rate": 1.6084804848374046e-06, "loss": 0.0038, "step": 92310 }, { "grad_norm": 0.06218607723712921, "learning_rate": 1.6043229425194817e-06, "loss": 0.0034, "step": 92320 }, { "grad_norm": 0.09171473234891891, "learning_rate": 1.6001706926800308e-06, "loss": 0.0039, "step": 92330 }, { "grad_norm": 0.0861026868224144, "learning_rate": 1.5960237357731444e-06, "loss": 0.0022, "step": 92340 }, { "grad_norm": 0.09020489454269409, "learning_rate": 1.5918820722523264e-06, "loss": 0.0035, "step": 92350 }, { "grad_norm": 0.1278231292963028, "learning_rate": 1.587745702570509e-06, "loss": 0.0032, "step": 92360 }, { "grad_norm": 0.076954685151577, "learning_rate": 1.5836146271800356e-06, "loss": 0.0023, "step": 92370 }, { "grad_norm": 0.10679067671298981, "learning_rate": 1.5794888465326728e-06, "loss": 0.0051, "step": 92380 }, { "grad_norm": 0.08948031067848206, "learning_rate": 1.5753683610796155e-06, "loss": 0.0025, "step": 92390 }, { "grad_norm": 0.14910553395748138, "learning_rate": 1.5712531712714528e-06, "loss": 0.0029, "step": 92400 }, { "grad_norm": 0.10535509884357452, "learning_rate": 1.5671432775582474e-06, "loss": 0.0021, "step": 92410 }, { "grad_norm": 0.11279062926769257, "learning_rate": 1.563038680389428e-06, "loss": 0.0026, "step": 92420 }, { "grad_norm": 0.07059676945209503, "learning_rate": 1.5589393802138808e-06, "loss": 0.0035, "step": 92430 }, { "grad_norm": 0.12252197414636612, "learning_rate": 1.554845377479891e-06, "loss": 0.0026, "step": 92440 }, { "grad_norm": 0.11104022711515427, "learning_rate": 1.550756672635173e-06, "loss": 0.0069, "step": 92450 }, { "grad_norm": 0.0929887443780899, "learning_rate": 1.5466732661268635e-06, "loss": 0.0039, "step": 92460 }, { "grad_norm": 0.08183830231428146, "learning_rate": 1.5425951584015163e-06, "loss": 0.0022, "step": 92470 }, { "grad_norm": 0.09989620745182037, "learning_rate": 1.5385223499051026e-06, "loss": 0.0043, "step": 92480 }, { "grad_norm": 0.14985717833042145, "learning_rate": 1.5344548410830328e-06, "loss": 0.0039, "step": 92490 }, { "grad_norm": 0.10542551428079605, "learning_rate": 1.5303926323801121e-06, "loss": 0.0034, "step": 92500 }, { "grad_norm": 0.12908314168453217, "learning_rate": 1.5263357242405795e-06, "loss": 0.003, "step": 92510 }, { "grad_norm": 0.0869763121008873, "learning_rate": 1.5222841171080915e-06, "loss": 0.0035, "step": 92520 }, { "grad_norm": 0.16638176143169403, "learning_rate": 1.5182378114257268e-06, "loss": 0.0043, "step": 92530 }, { "grad_norm": 0.10847163200378418, "learning_rate": 1.5141968076359869e-06, "loss": 0.0021, "step": 92540 }, { "grad_norm": 0.0676828995347023, "learning_rate": 1.5101611061807908e-06, "loss": 0.003, "step": 92550 }, { "grad_norm": 0.14356745779514313, "learning_rate": 1.506130707501463e-06, "loss": 0.0043, "step": 92560 }, { "grad_norm": 0.17629998922348022, "learning_rate": 1.5021056120387734e-06, "loss": 0.0038, "step": 92570 }, { "grad_norm": 0.12546417117118835, "learning_rate": 1.4980858202329084e-06, "loss": 0.0032, "step": 92580 }, { "grad_norm": 0.13250799477100372, "learning_rate": 1.4940713325234445e-06, "loss": 0.003, "step": 92590 }, { "grad_norm": 0.1881714016199112, "learning_rate": 1.4900621493494137e-06, "loss": 0.0045, "step": 92600 }, { "grad_norm": 0.12162227928638458, "learning_rate": 1.4860582711492544e-06, "loss": 0.0039, "step": 92610 }, { "grad_norm": 0.1765725463628769, "learning_rate": 1.4820596983608325e-06, "loss": 0.0035, "step": 92620 }, { "grad_norm": 0.14239850640296936, "learning_rate": 1.4780664314214044e-06, "loss": 0.0031, "step": 92630 }, { "grad_norm": 0.07991405576467514, "learning_rate": 1.474078470767687e-06, "loss": 0.0036, "step": 92640 }, { "grad_norm": 0.07818373292684555, "learning_rate": 1.470095816835787e-06, "loss": 0.0039, "step": 92650 }, { "grad_norm": 0.135764479637146, "learning_rate": 1.4661184700612395e-06, "loss": 0.0029, "step": 92660 }, { "grad_norm": 0.08415409922599792, "learning_rate": 1.4621464308790133e-06, "loss": 0.0023, "step": 92670 }, { "grad_norm": 0.04614971950650215, "learning_rate": 1.4581796997234775e-06, "loss": 0.0046, "step": 92680 }, { "grad_norm": 0.06341792643070221, "learning_rate": 1.4542182770284297e-06, "loss": 0.0034, "step": 92690 }, { "grad_norm": 0.1487775593996048, "learning_rate": 1.450262163227084e-06, "loss": 0.0026, "step": 92700 }, { "grad_norm": 0.16769060492515564, "learning_rate": 1.4463113587520671e-06, "loss": 0.0042, "step": 92710 }, { "grad_norm": 0.14419062435626984, "learning_rate": 1.4423658640354443e-06, "loss": 0.0028, "step": 92720 }, { "grad_norm": 0.10372154414653778, "learning_rate": 1.438425679508687e-06, "loss": 0.004, "step": 92730 }, { "grad_norm": 0.1697673201560974, "learning_rate": 1.4344908056026895e-06, "loss": 0.0027, "step": 92740 }, { "grad_norm": 0.088951475918293, "learning_rate": 1.4305612427477521e-06, "loss": 0.0036, "step": 92750 }, { "grad_norm": 0.16423264145851135, "learning_rate": 1.42663699137362e-06, "loss": 0.0032, "step": 92760 }, { "grad_norm": 0.09278523176908493, "learning_rate": 1.4227180519094386e-06, "loss": 0.0032, "step": 92770 }, { "grad_norm": 0.13922245800495148, "learning_rate": 1.4188044247837706e-06, "loss": 0.0029, "step": 92780 }, { "grad_norm": 0.0370146743953228, "learning_rate": 1.4148961104246128e-06, "loss": 0.0039, "step": 92790 }, { "grad_norm": 0.14287789165973663, "learning_rate": 1.4109931092593731e-06, "loss": 0.0027, "step": 92800 }, { "grad_norm": 0.08308630436658859, "learning_rate": 1.4070954217148658e-06, "loss": 0.0035, "step": 92810 }, { "grad_norm": 0.21690499782562256, "learning_rate": 1.4032030482173497e-06, "loss": 0.0055, "step": 92820 }, { "grad_norm": 0.07229091227054596, "learning_rate": 1.3993159891924901e-06, "loss": 0.0031, "step": 92830 }, { "grad_norm": 0.0792807787656784, "learning_rate": 1.3954342450653524e-06, "loss": 0.0022, "step": 92840 }, { "grad_norm": 0.15955257415771484, "learning_rate": 1.3915578162604582e-06, "loss": 0.0035, "step": 92850 }, { "grad_norm": 0.10957278311252594, "learning_rate": 1.3876867032017126e-06, "loss": 0.0026, "step": 92860 }, { "grad_norm": 0.1581050604581833, "learning_rate": 1.383820906312472e-06, "loss": 0.0047, "step": 92870 }, { "grad_norm": 0.06303814053535461, "learning_rate": 1.3799604260154752e-06, "loss": 0.0021, "step": 92880 }, { "grad_norm": 0.07551950216293335, "learning_rate": 1.376105262732913e-06, "loss": 0.0018, "step": 92890 }, { "grad_norm": 0.12568818032741547, "learning_rate": 1.37225541688637e-06, "loss": 0.0031, "step": 92900 }, { "grad_norm": 0.13555239140987396, "learning_rate": 1.3684108888968594e-06, "loss": 0.0029, "step": 92910 }, { "grad_norm": 0.14631803333759308, "learning_rate": 1.364571679184823e-06, "loss": 0.0024, "step": 92920 }, { "grad_norm": 0.10989569872617722, "learning_rate": 1.3607377881701077e-06, "loss": 0.0035, "step": 92930 }, { "grad_norm": 0.11586987227201462, "learning_rate": 1.3569092162719787e-06, "loss": 0.0038, "step": 92940 }, { "grad_norm": 0.12363184988498688, "learning_rate": 1.353085963909123e-06, "loss": 0.0036, "step": 92950 }, { "grad_norm": 0.08634407818317413, "learning_rate": 1.3492680314996454e-06, "loss": 0.0048, "step": 92960 }, { "grad_norm": 0.12820853292942047, "learning_rate": 1.345455419461078e-06, "loss": 0.0031, "step": 92970 }, { "grad_norm": 0.11430187523365021, "learning_rate": 1.3416481282103431e-06, "loss": 0.0027, "step": 92980 }, { "grad_norm": 0.1581081598997116, "learning_rate": 1.337846158163819e-06, "loss": 0.0036, "step": 92990 }, { "grad_norm": 0.20664933323860168, "learning_rate": 1.334049509737273e-06, "loss": 0.004, "step": 93000 }, { "grad_norm": 0.1659306138753891, "learning_rate": 1.3302581833459116e-06, "loss": 0.0025, "step": 93010 }, { "grad_norm": 0.1609792858362198, "learning_rate": 1.3264721794043366e-06, "loss": 0.0026, "step": 93020 }, { "grad_norm": 0.1068691685795784, "learning_rate": 1.3226914983265836e-06, "loss": 0.0023, "step": 93030 }, { "grad_norm": 0.07220451533794403, "learning_rate": 1.3189161405260997e-06, "loss": 0.0022, "step": 93040 }, { "grad_norm": 0.09247501939535141, "learning_rate": 1.3151461064157656e-06, "loss": 0.0029, "step": 93050 }, { "grad_norm": 0.09504722058773041, "learning_rate": 1.3113813964078459e-06, "loss": 0.0038, "step": 93060 }, { "grad_norm": 0.09638846665620804, "learning_rate": 1.3076220109140614e-06, "loss": 0.0023, "step": 93070 }, { "grad_norm": 0.23356744647026062, "learning_rate": 1.3038679503455275e-06, "loss": 0.0024, "step": 93080 }, { "grad_norm": 0.09316865354776382, "learning_rate": 1.3001192151127773e-06, "loss": 0.0039, "step": 93090 }, { "grad_norm": 0.08894342184066772, "learning_rate": 1.2963758056257657e-06, "loss": 0.0038, "step": 93100 }, { "grad_norm": 0.1667255163192749, "learning_rate": 1.2926377222938768e-06, "loss": 0.0035, "step": 93110 }, { "grad_norm": 0.16500285267829895, "learning_rate": 1.2889049655259e-06, "loss": 0.0029, "step": 93120 }, { "grad_norm": 0.14304614067077637, "learning_rate": 1.2851775357300312e-06, "loss": 0.0032, "step": 93130 }, { "grad_norm": 0.13031809031963348, "learning_rate": 1.281455433313916e-06, "loss": 0.0034, "step": 93140 }, { "grad_norm": 0.06152145937085152, "learning_rate": 1.2777386586845796e-06, "loss": 0.0045, "step": 93150 }, { "grad_norm": 0.18364317715168, "learning_rate": 1.2740272122484964e-06, "loss": 0.0046, "step": 93160 }, { "grad_norm": 0.13595160841941833, "learning_rate": 1.2703210944115308e-06, "loss": 0.0034, "step": 93170 }, { "grad_norm": 0.09888707846403122, "learning_rate": 1.2666203055789916e-06, "loss": 0.0033, "step": 93180 }, { "grad_norm": 0.07613600045442581, "learning_rate": 1.2629248461555832e-06, "loss": 0.0025, "step": 93190 }, { "grad_norm": 0.10113557428121567, "learning_rate": 1.2592347165454431e-06, "loss": 0.003, "step": 93200 }, { "grad_norm": 0.2518876791000366, "learning_rate": 1.2555499171521102e-06, "loss": 0.0033, "step": 93210 }, { "grad_norm": 0.10210328549146652, "learning_rate": 1.251870448378556e-06, "loss": 0.0045, "step": 93220 }, { "grad_norm": 0.10525929182767868, "learning_rate": 1.2481963106271478e-06, "loss": 0.0034, "step": 93230 }, { "grad_norm": 0.10874660313129425, "learning_rate": 1.2445275042997084e-06, "loss": 0.0036, "step": 93240 }, { "grad_norm": 0.09934725612401962, "learning_rate": 1.2408640297974283e-06, "loss": 0.0027, "step": 93250 }, { "grad_norm": 0.15218086540699005, "learning_rate": 1.2372058875209535e-06, "loss": 0.0032, "step": 93260 }, { "grad_norm": 0.16315454244613647, "learning_rate": 1.2335530778703252e-06, "loss": 0.0028, "step": 93270 }, { "grad_norm": 0.18237115442752838, "learning_rate": 1.2299056012450182e-06, "loss": 0.0031, "step": 93280 }, { "grad_norm": 0.06999314576387405, "learning_rate": 1.2262634580439026e-06, "loss": 0.0024, "step": 93290 }, { "grad_norm": 0.24174296855926514, "learning_rate": 1.2226266486652926e-06, "loss": 0.002, "step": 93300 }, { "grad_norm": 0.18464510142803192, "learning_rate": 1.2189951735068927e-06, "loss": 0.0056, "step": 93310 }, { "grad_norm": 0.13229376077651978, "learning_rate": 1.2153690329658407e-06, "loss": 0.003, "step": 93320 }, { "grad_norm": 0.21224737167358398, "learning_rate": 1.2117482274386859e-06, "loss": 0.0035, "step": 93330 }, { "grad_norm": 0.14579369127750397, "learning_rate": 1.2081327573213953e-06, "loss": 0.0024, "step": 93340 }, { "grad_norm": 0.03402925655245781, "learning_rate": 1.2045226230093465e-06, "loss": 0.0024, "step": 93350 }, { "grad_norm": 0.1404917687177658, "learning_rate": 1.2009178248973296e-06, "loss": 0.0023, "step": 93360 }, { "grad_norm": 0.13746009767055511, "learning_rate": 1.1973183633795849e-06, "loss": 0.0045, "step": 93370 }, { "grad_norm": 0.0924902930855751, "learning_rate": 1.193724238849725e-06, "loss": 0.0034, "step": 93380 }, { "grad_norm": 0.1499052792787552, "learning_rate": 1.190135451700808e-06, "loss": 0.0024, "step": 93390 }, { "grad_norm": 0.07948025315999985, "learning_rate": 1.186552002325292e-06, "loss": 0.0021, "step": 93400 }, { "grad_norm": 0.0838591530919075, "learning_rate": 1.1829738911150578e-06, "loss": 0.0025, "step": 93410 }, { "grad_norm": 0.07789753377437592, "learning_rate": 1.1794011184613985e-06, "loss": 0.0044, "step": 93420 }, { "grad_norm": 0.08966407924890518, "learning_rate": 1.17583368475504e-06, "loss": 0.0031, "step": 93430 }, { "grad_norm": 0.1938808113336563, "learning_rate": 1.1722715903860982e-06, "loss": 0.0027, "step": 93440 }, { "grad_norm": 0.10967370867729187, "learning_rate": 1.168714835744128e-06, "loss": 0.0025, "step": 93450 }, { "grad_norm": 0.17763729393482208, "learning_rate": 1.165163421218085e-06, "loss": 0.0032, "step": 93460 }, { "grad_norm": 0.06870397925376892, "learning_rate": 1.1616173471963477e-06, "loss": 0.0032, "step": 93470 }, { "grad_norm": 0.18353043496608734, "learning_rate": 1.1580766140667109e-06, "loss": 0.0021, "step": 93480 }, { "grad_norm": 0.20334622263908386, "learning_rate": 1.1545412222163764e-06, "loss": 0.0033, "step": 93490 }, { "grad_norm": 0.0694529190659523, "learning_rate": 1.1510111720319795e-06, "loss": 0.0021, "step": 93500 }, { "grad_norm": 0.06506811827421188, "learning_rate": 1.1474864638995663e-06, "loss": 0.0026, "step": 93510 }, { "grad_norm": 0.08305585384368896, "learning_rate": 1.1439670982045735e-06, "loss": 0.0032, "step": 93520 }, { "grad_norm": 0.12008076161146164, "learning_rate": 1.1404530753318877e-06, "loss": 0.0047, "step": 93530 }, { "grad_norm": 0.1259225457906723, "learning_rate": 1.136944395665801e-06, "loss": 0.0034, "step": 93540 }, { "grad_norm": 0.132046639919281, "learning_rate": 1.1334410595900014e-06, "loss": 0.0042, "step": 93550 }, { "grad_norm": 0.09862737357616425, "learning_rate": 1.129943067487621e-06, "loss": 0.0042, "step": 93560 }, { "grad_norm": 0.34864476323127747, "learning_rate": 1.126450419741193e-06, "loss": 0.0028, "step": 93570 }, { "grad_norm": 0.13976289331912994, "learning_rate": 1.1229631167326671e-06, "loss": 0.0032, "step": 93580 }, { "grad_norm": 0.03230443224310875, "learning_rate": 1.1194811588434052e-06, "loss": 0.0026, "step": 93590 }, { "grad_norm": 0.07670683413743973, "learning_rate": 1.1160045464542023e-06, "loss": 0.0025, "step": 93600 }, { "grad_norm": 0.2355380803346634, "learning_rate": 1.1125332799452325e-06, "loss": 0.0049, "step": 93610 }, { "grad_norm": 0.12543538212776184, "learning_rate": 1.109067359696131e-06, "loss": 0.0033, "step": 93620 }, { "grad_norm": 0.051756851375103, "learning_rate": 1.1056067860859166e-06, "loss": 0.0034, "step": 93630 }, { "grad_norm": 0.1453370898962021, "learning_rate": 1.102151559493031e-06, "loss": 0.0064, "step": 93640 }, { "grad_norm": 0.1153695285320282, "learning_rate": 1.0987016802953332e-06, "loss": 0.0047, "step": 93650 }, { "grad_norm": 0.0756043940782547, "learning_rate": 1.0952571488701048e-06, "loss": 0.0029, "step": 93660 }, { "grad_norm": 0.08367900550365448, "learning_rate": 1.0918179655940163e-06, "loss": 0.0023, "step": 93670 }, { "grad_norm": 0.1289311647415161, "learning_rate": 1.088384130843184e-06, "loss": 0.003, "step": 93680 }, { "grad_norm": 0.1427566111087799, "learning_rate": 1.0849556449931296e-06, "loss": 0.0031, "step": 93690 }, { "grad_norm": 0.15961098670959473, "learning_rate": 1.0815325084187866e-06, "loss": 0.0025, "step": 93700 }, { "grad_norm": 0.11764712631702423, "learning_rate": 1.0781147214944943e-06, "loss": 0.0041, "step": 93710 }, { "grad_norm": 0.08332698047161102, "learning_rate": 1.0747022845940313e-06, "loss": 0.0032, "step": 93720 }, { "grad_norm": 0.08257374912500381, "learning_rate": 1.0712951980905605e-06, "loss": 0.0024, "step": 93730 }, { "grad_norm": 0.1016821339726448, "learning_rate": 1.0678934623566839e-06, "loss": 0.0054, "step": 93740 }, { "grad_norm": 0.1928769052028656, "learning_rate": 1.0644970777644091e-06, "loss": 0.0026, "step": 93750 }, { "grad_norm": 0.17332762479782104, "learning_rate": 1.0611060446851673e-06, "loss": 0.0019, "step": 93760 }, { "grad_norm": 0.10959156602621078, "learning_rate": 1.0577203634897836e-06, "loss": 0.0046, "step": 93770 }, { "grad_norm": 0.12421264499425888, "learning_rate": 1.0543400345485178e-06, "loss": 0.0017, "step": 93780 }, { "grad_norm": 0.09756293892860413, "learning_rate": 1.0509650582310404e-06, "loss": 0.0035, "step": 93790 }, { "grad_norm": 0.1045152097940445, "learning_rate": 1.0475954349064287e-06, "loss": 0.0026, "step": 93800 }, { "grad_norm": 0.14106257259845734, "learning_rate": 1.0442311649431825e-06, "loss": 0.0028, "step": 93810 }, { "grad_norm": 0.13704344630241394, "learning_rate": 1.0408722487092071e-06, "loss": 0.0033, "step": 93820 }, { "grad_norm": 0.057807642966508865, "learning_rate": 1.0375186865718424e-06, "loss": 0.0037, "step": 93830 }, { "grad_norm": 0.05156861990690231, "learning_rate": 1.0341704788978113e-06, "loss": 0.0026, "step": 93840 }, { "grad_norm": 0.14366193115711212, "learning_rate": 1.0308276260532879e-06, "loss": 0.0031, "step": 93850 }, { "grad_norm": 0.09264501929283142, "learning_rate": 1.0274901284038241e-06, "loss": 0.004, "step": 93860 }, { "grad_norm": 0.09117680788040161, "learning_rate": 1.0241579863144114e-06, "loss": 0.0033, "step": 93870 }, { "grad_norm": 0.07243797928094864, "learning_rate": 1.0208312001494413e-06, "loss": 0.0026, "step": 93880 }, { "grad_norm": 0.23205049335956573, "learning_rate": 1.0175097702727453e-06, "loss": 0.0031, "step": 93890 }, { "grad_norm": 0.13221971690654755, "learning_rate": 1.0141936970475275e-06, "loss": 0.0032, "step": 93900 }, { "grad_norm": 0.09647756814956665, "learning_rate": 1.0108829808364417e-06, "loss": 0.0036, "step": 93910 }, { "grad_norm": 0.09296903759241104, "learning_rate": 1.0075776220015376e-06, "loss": 0.0053, "step": 93920 }, { "grad_norm": 0.0654500424861908, "learning_rate": 1.0042776209042816e-06, "loss": 0.0024, "step": 93930 }, { "grad_norm": 0.125204399228096, "learning_rate": 1.000982977905568e-06, "loss": 0.0034, "step": 93940 }, { "grad_norm": 0.09960131347179413, "learning_rate": 9.976936933656867e-07, "loss": 0.0037, "step": 93950 }, { "grad_norm": 0.0733104795217514, "learning_rate": 9.944097676443442e-07, "loss": 0.0066, "step": 93960 }, { "grad_norm": 0.1318519115447998, "learning_rate": 9.911312011006757e-07, "loss": 0.0029, "step": 93970 }, { "grad_norm": 0.08793556690216064, "learning_rate": 9.878579940932109e-07, "loss": 0.0033, "step": 93980 }, { "grad_norm": 0.1399121731519699, "learning_rate": 9.84590146979908e-07, "loss": 0.0025, "step": 93990 }, { "grad_norm": 0.1557869017124176, "learning_rate": 9.813276601181365e-07, "loss": 0.0027, "step": 94000 }, { "grad_norm": 0.15069957077503204, "learning_rate": 9.780705338646724e-07, "loss": 0.0024, "step": 94010 }, { "grad_norm": 0.13396687805652618, "learning_rate": 9.748187685757083e-07, "loss": 0.0025, "step": 94020 }, { "grad_norm": 0.16302397847175598, "learning_rate": 9.715723646068543e-07, "loss": 0.0026, "step": 94030 }, { "grad_norm": 0.056142598390579224, "learning_rate": 9.68331322313143e-07, "loss": 0.0046, "step": 94040 }, { "grad_norm": 0.054015353322029114, "learning_rate": 9.650956420489909e-07, "loss": 0.0037, "step": 94050 }, { "grad_norm": 0.11234013736248016, "learning_rate": 9.618653241682595e-07, "loss": 0.0032, "step": 94060 }, { "grad_norm": 0.12053877860307693, "learning_rate": 9.586403690242053e-07, "loss": 0.0036, "step": 94070 }, { "grad_norm": 0.13155077397823334, "learning_rate": 9.554207769695122e-07, "loss": 0.0034, "step": 94080 }, { "grad_norm": 0.11910031735897064, "learning_rate": 9.522065483562659e-07, "loss": 0.0028, "step": 94090 }, { "grad_norm": 0.1162976548075676, "learning_rate": 9.489976835359681e-07, "loss": 0.0029, "step": 94100 }, { "grad_norm": 0.08784779161214828, "learning_rate": 9.457941828595329e-07, "loss": 0.0023, "step": 94110 }, { "grad_norm": 0.11103342473506927, "learning_rate": 9.425960466772965e-07, "loss": 0.0054, "step": 94120 }, { "grad_norm": 0.07723983377218246, "learning_rate": 9.394032753390014e-07, "loss": 0.0031, "step": 94130 }, { "grad_norm": 0.10303441435098648, "learning_rate": 9.362158691938017e-07, "loss": 0.0024, "step": 94140 }, { "grad_norm": 0.1051163449883461, "learning_rate": 9.330338285902685e-07, "loss": 0.0046, "step": 94150 }, { "grad_norm": 0.12827549874782562, "learning_rate": 9.298571538763845e-07, "loss": 0.0033, "step": 94160 }, { "grad_norm": 0.19149164855480194, "learning_rate": 9.266858453995441e-07, "loss": 0.0035, "step": 94170 }, { "grad_norm": 0.14189957082271576, "learning_rate": 9.235199035065645e-07, "loss": 0.0024, "step": 94180 }, { "grad_norm": 0.060825150460004807, "learning_rate": 9.203593285436574e-07, "loss": 0.0032, "step": 94190 }, { "grad_norm": 0.15511946380138397, "learning_rate": 9.172041208564686e-07, "loss": 0.0025, "step": 94200 }, { "grad_norm": 0.22864945232868195, "learning_rate": 9.140542807900332e-07, "loss": 0.0038, "step": 94210 }, { "grad_norm": 0.08124639838933945, "learning_rate": 9.109098086888313e-07, "loss": 0.0023, "step": 94220 }, { "grad_norm": 0.09474582970142365, "learning_rate": 9.077707048967265e-07, "loss": 0.0044, "step": 94230 }, { "grad_norm": 0.1478898823261261, "learning_rate": 9.046369697570057e-07, "loss": 0.0032, "step": 94240 }, { "grad_norm": 0.07447382062673569, "learning_rate": 9.015086036123721e-07, "loss": 0.0031, "step": 94250 }, { "grad_norm": 0.1333392858505249, "learning_rate": 8.983856068049413e-07, "loss": 0.0027, "step": 94260 }, { "grad_norm": 0.11344830691814423, "learning_rate": 8.952679796762398e-07, "loss": 0.0021, "step": 94270 }, { "grad_norm": 0.09137359261512756, "learning_rate": 8.921557225671951e-07, "loss": 0.0036, "step": 94280 }, { "grad_norm": 0.09183676540851593, "learning_rate": 8.890488358181792e-07, "loss": 0.0052, "step": 94290 }, { "grad_norm": 0.16547739505767822, "learning_rate": 8.859473197689372e-07, "loss": 0.003, "step": 94300 }, { "grad_norm": 0.06283023208379745, "learning_rate": 8.828511747586588e-07, "loss": 0.0026, "step": 94310 }, { "grad_norm": 0.11744806170463562, "learning_rate": 8.797604011259287e-07, "loss": 0.0035, "step": 94320 }, { "grad_norm": 0.09273470193147659, "learning_rate": 8.766749992087487e-07, "loss": 0.0025, "step": 94330 }, { "grad_norm": 0.11535141617059708, "learning_rate": 8.735949693445378e-07, "loss": 0.0032, "step": 94340 }, { "grad_norm": 0.17748653888702393, "learning_rate": 8.705203118701211e-07, "loss": 0.003, "step": 94350 }, { "grad_norm": 0.09794247895479202, "learning_rate": 8.674510271217295e-07, "loss": 0.0021, "step": 94360 }, { "grad_norm": 0.14961646497249603, "learning_rate": 8.643871154350336e-07, "loss": 0.0041, "step": 94370 }, { "grad_norm": 0.05681855231523514, "learning_rate": 8.613285771450818e-07, "loss": 0.0019, "step": 94380 }, { "grad_norm": 0.06624075770378113, "learning_rate": 8.582754125863623e-07, "loss": 0.004, "step": 94390 }, { "grad_norm": 0.1412057876586914, "learning_rate": 8.55227622092758e-07, "loss": 0.0041, "step": 94400 }, { "grad_norm": 0.09182711690664291, "learning_rate": 8.521852059975743e-07, "loss": 0.0031, "step": 94410 }, { "grad_norm": 0.10376296192407608, "learning_rate": 8.491481646335231e-07, "loss": 0.0022, "step": 94420 }, { "grad_norm": 0.0912332758307457, "learning_rate": 8.46116498332733e-07, "loss": 0.0022, "step": 94430 }, { "grad_norm": 0.08432524651288986, "learning_rate": 8.430902074267333e-07, "loss": 0.0025, "step": 94440 }, { "grad_norm": 0.13813680410385132, "learning_rate": 8.400692922464925e-07, "loss": 0.0036, "step": 94450 }, { "grad_norm": 0.06132470443844795, "learning_rate": 8.370537531223577e-07, "loss": 0.0025, "step": 94460 }, { "grad_norm": 0.1308145970106125, "learning_rate": 8.340435903841092e-07, "loss": 0.0022, "step": 94470 }, { "grad_norm": 0.10360884666442871, "learning_rate": 8.310388043609341e-07, "loss": 0.0031, "step": 94480 }, { "grad_norm": 0.04598936066031456, "learning_rate": 8.280393953814303e-07, "loss": 0.0043, "step": 94490 }, { "grad_norm": 0.12394615262746811, "learning_rate": 8.250453637736077e-07, "loss": 0.0034, "step": 94500 }, { "grad_norm": 0.14906686544418335, "learning_rate": 8.22056709864899e-07, "loss": 0.0029, "step": 94510 }, { "grad_norm": 0.11062861233949661, "learning_rate": 8.190734339821204e-07, "loss": 0.0022, "step": 94520 }, { "grad_norm": 0.07996483147144318, "learning_rate": 8.160955364515333e-07, "loss": 0.0022, "step": 94530 }, { "grad_norm": 0.08387111872434616, "learning_rate": 8.131230175987936e-07, "loss": 0.0025, "step": 94540 }, { "grad_norm": 0.12001194059848785, "learning_rate": 8.101558777489693e-07, "loss": 0.004, "step": 94550 }, { "grad_norm": 0.25027766823768616, "learning_rate": 8.071941172265396e-07, "loss": 0.0036, "step": 94560 }, { "grad_norm": 0.1770133227109909, "learning_rate": 8.042377363553955e-07, "loss": 0.0031, "step": 94570 }, { "grad_norm": 0.07149757444858551, "learning_rate": 8.01286735458856e-07, "loss": 0.0026, "step": 94580 }, { "grad_norm": 0.14124543964862823, "learning_rate": 7.983411148596298e-07, "loss": 0.0029, "step": 94590 }, { "grad_norm": 0.13592560589313507, "learning_rate": 7.954008748798425e-07, "loss": 0.0024, "step": 94600 }, { "grad_norm": 0.09382785856723785, "learning_rate": 7.924660158410369e-07, "loss": 0.0041, "step": 94610 }, { "grad_norm": 0.164045512676239, "learning_rate": 7.895365380641728e-07, "loss": 0.0036, "step": 94620 }, { "grad_norm": 0.13081231713294983, "learning_rate": 7.866124418695941e-07, "loss": 0.0028, "step": 94630 }, { "grad_norm": 0.16053275763988495, "learning_rate": 7.83693727577095e-07, "loss": 0.0033, "step": 94640 }, { "grad_norm": 0.11653471738100052, "learning_rate": 7.807803955058535e-07, "loss": 0.0027, "step": 94650 }, { "grad_norm": 0.12104268372058868, "learning_rate": 7.778724459744702e-07, "loss": 0.0023, "step": 94660 }, { "grad_norm": 0.15898150205612183, "learning_rate": 7.749698793009464e-07, "loss": 0.0035, "step": 94670 }, { "grad_norm": 0.08567960560321808, "learning_rate": 7.720726958027169e-07, "loss": 0.0029, "step": 94680 }, { "grad_norm": 0.10959099233150482, "learning_rate": 7.691808957965951e-07, "loss": 0.0024, "step": 94690 }, { "grad_norm": 0.22819991409778595, "learning_rate": 7.662944795988336e-07, "loss": 0.0032, "step": 94700 }, { "grad_norm": 0.12027819454669952, "learning_rate": 7.634134475250854e-07, "loss": 0.0029, "step": 94710 }, { "grad_norm": 0.04898136481642723, "learning_rate": 7.605377998904262e-07, "loss": 0.0023, "step": 94720 }, { "grad_norm": 0.06299544125795364, "learning_rate": 7.5766753700931e-07, "loss": 0.0024, "step": 94730 }, { "grad_norm": 0.05580824613571167, "learning_rate": 7.54802659195647e-07, "loss": 0.0054, "step": 94740 }, { "grad_norm": 0.10283726453781128, "learning_rate": 7.519431667627253e-07, "loss": 0.0027, "step": 94750 }, { "grad_norm": 0.2556968331336975, "learning_rate": 7.490890600232503e-07, "loss": 0.0042, "step": 94760 }, { "grad_norm": 0.15025892853736877, "learning_rate": 7.462403392893502e-07, "loss": 0.0029, "step": 94770 }, { "grad_norm": 0.09513290226459503, "learning_rate": 7.433970048725536e-07, "loss": 0.005, "step": 94780 }, { "grad_norm": 0.05767650902271271, "learning_rate": 7.405590570838062e-07, "loss": 0.0044, "step": 94790 }, { "grad_norm": 0.05058445408940315, "learning_rate": 7.377264962334596e-07, "loss": 0.0027, "step": 94800 }, { "grad_norm": 0.1305893510580063, "learning_rate": 7.348993226312773e-07, "loss": 0.0042, "step": 94810 }, { "grad_norm": 0.11384066939353943, "learning_rate": 7.32077536586434e-07, "loss": 0.0022, "step": 94820 }, { "grad_norm": 0.11196094006299973, "learning_rate": 7.292611384075221e-07, "loss": 0.0027, "step": 94830 }, { "grad_norm": 0.21554546058177948, "learning_rate": 7.264501284025338e-07, "loss": 0.004, "step": 94840 }, { "grad_norm": 0.13638609647750854, "learning_rate": 7.236445068788789e-07, "loss": 0.0034, "step": 94850 }, { "grad_norm": 0.08656656742095947, "learning_rate": 7.20844274143373e-07, "loss": 0.0025, "step": 94860 }, { "grad_norm": 0.108225978910923, "learning_rate": 7.180494305022489e-07, "loss": 0.0032, "step": 94870 }, { "grad_norm": 0.08001166582107544, "learning_rate": 7.152599762611456e-07, "loss": 0.0018, "step": 94880 }, { "grad_norm": 0.41240394115448, "learning_rate": 7.124759117251079e-07, "loss": 0.0041, "step": 94890 }, { "grad_norm": 0.19394215941429138, "learning_rate": 7.096972371986033e-07, "loss": 0.0036, "step": 94900 }, { "grad_norm": 0.0961318090558052, "learning_rate": 7.069239529855054e-07, "loss": 0.0043, "step": 94910 }, { "grad_norm": 0.14031486213207245, "learning_rate": 7.041560593890939e-07, "loss": 0.0032, "step": 94920 }, { "grad_norm": 0.09730614721775055, "learning_rate": 7.013935567120599e-07, "loss": 0.003, "step": 94930 }, { "grad_norm": 0.0979338139295578, "learning_rate": 6.986364452565009e-07, "loss": 0.003, "step": 94940 }, { "grad_norm": 0.14647112786769867, "learning_rate": 6.95884725323942e-07, "loss": 0.0025, "step": 94950 }, { "grad_norm": 0.09650632739067078, "learning_rate": 6.931383972152982e-07, "loss": 0.0034, "step": 94960 }, { "grad_norm": 0.1395847201347351, "learning_rate": 6.903974612309127e-07, "loss": 0.0035, "step": 94970 }, { "grad_norm": 0.15495501458644867, "learning_rate": 6.876619176705179e-07, "loss": 0.0025, "step": 94980 }, { "grad_norm": 0.22534029185771942, "learning_rate": 6.8493176683328e-07, "loss": 0.0031, "step": 94990 }, { "grad_norm": 0.13777446746826172, "learning_rate": 6.822070090177657e-07, "loss": 0.0032, "step": 95000 }, { "grad_norm": 0.17400863766670227, "learning_rate": 6.794876445219367e-07, "loss": 0.0032, "step": 95010 }, { "grad_norm": 0.08680423349142075, "learning_rate": 6.767736736431885e-07, "loss": 0.0029, "step": 95020 }, { "grad_norm": 0.05745792016386986, "learning_rate": 6.740650966783113e-07, "loss": 0.0025, "step": 95030 }, { "grad_norm": 0.08246426284313202, "learning_rate": 6.713619139235239e-07, "loss": 0.0024, "step": 95040 }, { "grad_norm": 0.08824706077575684, "learning_rate": 6.686641256744286e-07, "loss": 0.003, "step": 95050 }, { "grad_norm": 0.08746457099914551, "learning_rate": 6.659717322260562e-07, "loss": 0.004, "step": 95060 }, { "grad_norm": 0.0883544385433197, "learning_rate": 6.632847338728377e-07, "loss": 0.003, "step": 95070 }, { "grad_norm": 0.06714078038930893, "learning_rate": 6.60603130908627e-07, "loss": 0.003, "step": 95080 }, { "grad_norm": 0.0846787840127945, "learning_rate": 6.579269236266783e-07, "loss": 0.0031, "step": 95090 }, { "grad_norm": 0.06555343419313431, "learning_rate": 6.552561123196632e-07, "loss": 0.0026, "step": 95100 }, { "grad_norm": 0.0846661925315857, "learning_rate": 6.525906972796425e-07, "loss": 0.0024, "step": 95110 }, { "grad_norm": 0.10109704732894897, "learning_rate": 6.499306787981108e-07, "loss": 0.0028, "step": 95120 }, { "grad_norm": 0.15543623268604279, "learning_rate": 6.472760571659631e-07, "loss": 0.0036, "step": 95130 }, { "grad_norm": 0.11987543106079102, "learning_rate": 6.446268326735061e-07, "loss": 0.0018, "step": 95140 }, { "grad_norm": 0.3199917674064636, "learning_rate": 6.41983005610458e-07, "loss": 0.0042, "step": 95150 }, { "grad_norm": 0.1598476618528366, "learning_rate": 6.393445762659378e-07, "loss": 0.0038, "step": 95160 }, { "grad_norm": 0.09325215965509415, "learning_rate": 6.36711544928481e-07, "loss": 0.005, "step": 95170 }, { "grad_norm": 0.13377922773361206, "learning_rate": 6.340839118860298e-07, "loss": 0.0024, "step": 95180 }, { "grad_norm": 0.1678907722234726, "learning_rate": 6.314616774259485e-07, "loss": 0.0033, "step": 95190 }, { "grad_norm": 0.10543013364076614, "learning_rate": 6.288448418349858e-07, "loss": 0.0033, "step": 95200 }, { "grad_norm": 0.06570970267057419, "learning_rate": 6.262334053993235e-07, "loss": 0.0031, "step": 95210 }, { "grad_norm": 0.11350898444652557, "learning_rate": 6.236273684045501e-07, "loss": 0.0042, "step": 95220 }, { "grad_norm": 0.13202758133411407, "learning_rate": 6.210267311356488e-07, "loss": 0.0024, "step": 95230 }, { "grad_norm": 0.10977665334939957, "learning_rate": 6.184314938770197e-07, "loss": 0.004, "step": 95240 }, { "grad_norm": 0.048084162175655365, "learning_rate": 6.158416569124858e-07, "loss": 0.0023, "step": 95250 }, { "grad_norm": 0.18293552100658417, "learning_rate": 6.132572205252541e-07, "loss": 0.0036, "step": 95260 }, { "grad_norm": 0.1890249252319336, "learning_rate": 6.106781849979648e-07, "loss": 0.003, "step": 95270 }, { "grad_norm": 0.12134765088558197, "learning_rate": 6.081045506126482e-07, "loss": 0.0036, "step": 95280 }, { "grad_norm": 0.1779281497001648, "learning_rate": 6.055363176507678e-07, "loss": 0.0025, "step": 95290 }, { "grad_norm": 0.1153242215514183, "learning_rate": 6.029734863931602e-07, "loss": 0.0033, "step": 95300 }, { "grad_norm": 0.04900037497282028, "learning_rate": 6.004160571201179e-07, "loss": 0.0022, "step": 95310 }, { "grad_norm": 0.06585390120744705, "learning_rate": 5.978640301112947e-07, "loss": 0.0049, "step": 95320 }, { "grad_norm": 0.1281442642211914, "learning_rate": 5.9531740564579e-07, "loss": 0.0022, "step": 95330 }, { "grad_norm": 0.12513014674186707, "learning_rate": 5.927761840020918e-07, "loss": 0.0024, "step": 95340 }, { "grad_norm": 0.12931302189826965, "learning_rate": 5.902403654581113e-07, "loss": 0.0053, "step": 95350 }, { "grad_norm": 0.06832953542470932, "learning_rate": 5.877099502911542e-07, "loss": 0.0029, "step": 95360 }, { "grad_norm": 0.13418929278850555, "learning_rate": 5.851849387779496e-07, "loss": 0.0029, "step": 95370 }, { "grad_norm": 0.07000388205051422, "learning_rate": 5.826653311946262e-07, "loss": 0.0036, "step": 95380 }, { "grad_norm": 0.07954390347003937, "learning_rate": 5.801511278167249e-07, "loss": 0.0024, "step": 95390 }, { "grad_norm": 0.14226728677749634, "learning_rate": 5.776423289191868e-07, "loss": 0.0031, "step": 95400 }, { "grad_norm": 0.05927286669611931, "learning_rate": 5.751389347763869e-07, "loss": 0.0025, "step": 95410 }, { "grad_norm": 0.0892196074128151, "learning_rate": 5.726409456620841e-07, "loss": 0.0028, "step": 95420 }, { "grad_norm": 0.06792337447404861, "learning_rate": 5.701483618494541e-07, "loss": 0.0034, "step": 95430 }, { "grad_norm": 0.11938890069723129, "learning_rate": 5.67661183611079e-07, "loss": 0.0039, "step": 95440 }, { "grad_norm": 0.1277032345533371, "learning_rate": 5.651794112189579e-07, "loss": 0.0037, "step": 95450 }, { "grad_norm": 0.1432633250951767, "learning_rate": 5.627030449444903e-07, "loss": 0.0022, "step": 95460 }, { "grad_norm": 0.07576493918895721, "learning_rate": 5.60232085058493e-07, "loss": 0.0023, "step": 95470 }, { "grad_norm": 0.125864177942276, "learning_rate": 5.577665318311776e-07, "loss": 0.0021, "step": 95480 }, { "grad_norm": 0.039515215903520584, "learning_rate": 5.553063855321838e-07, "loss": 0.0026, "step": 95490 }, { "grad_norm": 0.16630245745182037, "learning_rate": 5.528516464305467e-07, "loss": 0.0023, "step": 95500 }, { "grad_norm": 0.15845425426959991, "learning_rate": 5.504023147947068e-07, "loss": 0.004, "step": 95510 }, { "grad_norm": 0.0992572009563446, "learning_rate": 5.479583908925279e-07, "loss": 0.003, "step": 95520 }, { "grad_norm": 0.04958257079124451, "learning_rate": 5.455198749912626e-07, "loss": 0.0026, "step": 95530 }, { "grad_norm": 0.09994708746671677, "learning_rate": 5.430867673575979e-07, "loss": 0.0025, "step": 95540 }, { "grad_norm": 0.09449252486228943, "learning_rate": 5.406590682575985e-07, "loss": 0.0047, "step": 95550 }, { "grad_norm": 0.0626937747001648, "learning_rate": 5.382367779567687e-07, "loss": 0.0025, "step": 95560 }, { "grad_norm": 0.0837935283780098, "learning_rate": 5.358198967200023e-07, "loss": 0.0029, "step": 95570 }, { "grad_norm": 0.05464575067162514, "learning_rate": 5.334084248115989e-07, "loss": 0.0022, "step": 95580 }, { "grad_norm": 0.08971698582172394, "learning_rate": 5.310023624952754e-07, "loss": 0.0026, "step": 95590 }, { "grad_norm": 0.07222498953342438, "learning_rate": 5.286017100341656e-07, "loss": 0.0025, "step": 95600 }, { "grad_norm": 0.07081642001867294, "learning_rate": 5.262064676907874e-07, "loss": 0.0042, "step": 95610 }, { "grad_norm": 0.17559200525283813, "learning_rate": 5.238166357270924e-07, "loss": 0.0031, "step": 95620 }, { "grad_norm": 0.19867706298828125, "learning_rate": 5.214322144044215e-07, "loss": 0.005, "step": 95630 }, { "grad_norm": 0.10535337775945663, "learning_rate": 5.190532039835327e-07, "loss": 0.0025, "step": 95640 }, { "grad_norm": 0.07817097753286362, "learning_rate": 5.166796047245903e-07, "loss": 0.0027, "step": 95650 }, { "grad_norm": 0.06470517069101334, "learning_rate": 5.143114168871699e-07, "loss": 0.0022, "step": 95660 }, { "grad_norm": 0.30326157808303833, "learning_rate": 5.119486407302532e-07, "loss": 0.0022, "step": 95670 }, { "grad_norm": 0.11322575807571411, "learning_rate": 5.095912765122335e-07, "loss": 0.0029, "step": 95680 }, { "grad_norm": 0.09132619947195053, "learning_rate": 5.07239324490899e-07, "loss": 0.0024, "step": 95690 }, { "grad_norm": 0.13081949949264526, "learning_rate": 5.048927849234608e-07, "loss": 0.0035, "step": 95700 }, { "grad_norm": 0.10048296302556992, "learning_rate": 5.025516580665358e-07, "loss": 0.0058, "step": 95710 }, { "grad_norm": 0.06977514177560806, "learning_rate": 5.002159441761412e-07, "loss": 0.0023, "step": 95720 }, { "grad_norm": 0.0766293928027153, "learning_rate": 4.978856435077062e-07, "loss": 0.0028, "step": 95730 }, { "grad_norm": 0.1481875479221344, "learning_rate": 4.955607563160769e-07, "loss": 0.0025, "step": 95740 }, { "grad_norm": 0.049263227730989456, "learning_rate": 4.932412828554944e-07, "loss": 0.0018, "step": 95750 }, { "grad_norm": 0.06873604655265808, "learning_rate": 4.909272233796114e-07, "loss": 0.0038, "step": 95760 }, { "grad_norm": 0.1520785391330719, "learning_rate": 4.88618578141492e-07, "loss": 0.0028, "step": 95770 }, { "grad_norm": 0.09707970917224884, "learning_rate": 4.863153473936011e-07, "loss": 0.0042, "step": 95780 }, { "grad_norm": 0.0674888864159584, "learning_rate": 4.840175313878259e-07, "loss": 0.0037, "step": 95790 }, { "grad_norm": 0.037604428827762604, "learning_rate": 4.817251303754433e-07, "loss": 0.0038, "step": 95800 }, { "grad_norm": 0.14654052257537842, "learning_rate": 4.794381446071583e-07, "loss": 0.0034, "step": 95810 }, { "grad_norm": 0.13473738729953766, "learning_rate": 4.771565743330597e-07, "loss": 0.0025, "step": 95820 }, { "grad_norm": 0.1454174667596817, "learning_rate": 4.7488041980266463e-07, "loss": 0.0028, "step": 95830 }, { "grad_norm": 0.046176571398973465, "learning_rate": 4.726096812648795e-07, "loss": 0.0026, "step": 95840 }, { "grad_norm": 0.10329321771860123, "learning_rate": 4.703443589680445e-07, "loss": 0.0028, "step": 95850 }, { "grad_norm": 0.06280430406332016, "learning_rate": 4.6808445315987826e-07, "loss": 0.0029, "step": 95860 }, { "grad_norm": 0.10782533138990402, "learning_rate": 4.658299640875274e-07, "loss": 0.0028, "step": 95870 }, { "grad_norm": 0.14332027733325958, "learning_rate": 4.635808919975393e-07, "loss": 0.0027, "step": 95880 }, { "grad_norm": 0.09860710054636002, "learning_rate": 4.6133723713586706e-07, "loss": 0.0022, "step": 95890 }, { "grad_norm": 0.1278102695941925, "learning_rate": 4.590989997478756e-07, "loss": 0.0035, "step": 95900 }, { "grad_norm": 0.07726452499628067, "learning_rate": 4.5686618007833027e-07, "loss": 0.0025, "step": 95910 }, { "grad_norm": 0.074996717274189, "learning_rate": 4.54638778371419e-07, "loss": 0.002, "step": 95920 }, { "grad_norm": 0.09350799769163132, "learning_rate": 4.524167948707192e-07, "loss": 0.0038, "step": 95930 }, { "grad_norm": 0.1267843246459961, "learning_rate": 4.5020022981921986e-07, "loss": 0.0029, "step": 95940 }, { "grad_norm": 0.1940329670906067, "learning_rate": 4.47989083459327e-07, "loss": 0.0054, "step": 95950 }, { "grad_norm": 0.11318768560886383, "learning_rate": 4.457833560328528e-07, "loss": 0.0052, "step": 95960 }, { "grad_norm": 0.049568839371204376, "learning_rate": 4.4358304778100434e-07, "loss": 0.0038, "step": 95970 }, { "grad_norm": 0.16492560505867004, "learning_rate": 4.413881589444058e-07, "loss": 0.0032, "step": 95980 }, { "grad_norm": 0.0589275099337101, "learning_rate": 4.391986897630873e-07, "loss": 0.004, "step": 95990 }, { "grad_norm": 0.130768284201622, "learning_rate": 4.370146404764908e-07, "loss": 0.0033, "step": 96000 }, { "grad_norm": 0.06330393254756927, "learning_rate": 4.3483601132345285e-07, "loss": 0.004, "step": 96010 }, { "grad_norm": 0.12901075184345245, "learning_rate": 4.32662802542233e-07, "loss": 0.0049, "step": 96020 }, { "grad_norm": 0.1871391385793686, "learning_rate": 4.3049501437047444e-07, "loss": 0.0025, "step": 96030 }, { "grad_norm": 0.07724542170763016, "learning_rate": 4.283326470452653e-07, "loss": 0.0032, "step": 96040 }, { "grad_norm": 0.12089687585830688, "learning_rate": 4.2617570080306647e-07, "loss": 0.0023, "step": 96050 }, { "grad_norm": 0.2002982795238495, "learning_rate": 4.2402417587975585e-07, "loss": 0.0024, "step": 96060 }, { "grad_norm": 0.1600438952445984, "learning_rate": 4.218780725106286e-07, "loss": 0.0035, "step": 96070 }, { "grad_norm": 0.08528461307287216, "learning_rate": 4.197373909303748e-07, "loss": 0.0042, "step": 96080 }, { "grad_norm": 0.15910165011882782, "learning_rate": 4.17602131373096e-07, "loss": 0.0033, "step": 96090 }, { "grad_norm": 0.08176888525485992, "learning_rate": 4.154722940723055e-07, "loss": 0.0036, "step": 96100 }, { "grad_norm": 0.02847439795732498, "learning_rate": 4.1334787926091136e-07, "loss": 0.0022, "step": 96110 }, { "grad_norm": 0.06857344508171082, "learning_rate": 4.1122888717124444e-07, "loss": 0.0036, "step": 96120 }, { "grad_norm": 0.18366624414920807, "learning_rate": 4.091153180350249e-07, "loss": 0.0037, "step": 96130 }, { "grad_norm": 0.15170733630657196, "learning_rate": 4.0700717208340125e-07, "loss": 0.0036, "step": 96140 }, { "grad_norm": 0.1158607006072998, "learning_rate": 4.049044495469112e-07, "loss": 0.0024, "step": 96150 }, { "grad_norm": 0.10345764458179474, "learning_rate": 4.0280715065549866e-07, "loss": 0.0043, "step": 96160 }, { "grad_norm": 0.07515393197536469, "learning_rate": 4.0071527563853575e-07, "loss": 0.0029, "step": 96170 }, { "grad_norm": 0.06205298751592636, "learning_rate": 3.9862882472477823e-07, "loss": 0.0022, "step": 96180 }, { "grad_norm": 0.04853576049208641, "learning_rate": 3.9654779814239375e-07, "loss": 0.0027, "step": 96190 }, { "grad_norm": 0.11991991102695465, "learning_rate": 3.944721961189668e-07, "loss": 0.0028, "step": 96200 }, { "grad_norm": 0.1659150868654251, "learning_rate": 3.9240201888147697e-07, "loss": 0.0027, "step": 96210 }, { "grad_norm": 0.1353267878293991, "learning_rate": 3.9033726665632096e-07, "loss": 0.0028, "step": 96220 }, { "grad_norm": 0.13053855299949646, "learning_rate": 3.882779396692959e-07, "loss": 0.003, "step": 96230 }, { "grad_norm": 0.10030391812324524, "learning_rate": 3.8622403814559947e-07, "loss": 0.0023, "step": 96240 }, { "grad_norm": 0.11767589300870895, "learning_rate": 3.841755623098575e-07, "loss": 0.002, "step": 96250 }, { "grad_norm": 0.0811874121427536, "learning_rate": 3.821325123860742e-07, "loss": 0.0032, "step": 96260 }, { "grad_norm": 0.08310721814632416, "learning_rate": 3.800948885976818e-07, "loss": 0.0024, "step": 96270 }, { "grad_norm": 0.1263190656900406, "learning_rate": 3.780626911675078e-07, "loss": 0.0041, "step": 96280 }, { "grad_norm": 0.10278084129095078, "learning_rate": 3.76035920317791e-07, "loss": 0.0029, "step": 96290 }, { "grad_norm": 0.12099821865558624, "learning_rate": 3.740145762701819e-07, "loss": 0.0022, "step": 96300 }, { "grad_norm": 0.10220253467559814, "learning_rate": 3.7199865924572585e-07, "loss": 0.0037, "step": 96310 }, { "grad_norm": 0.13397952914237976, "learning_rate": 3.699881694648799e-07, "loss": 0.0032, "step": 96320 }, { "grad_norm": 0.13681086897850037, "learning_rate": 3.679831071475126e-07, "loss": 0.0051, "step": 96330 }, { "grad_norm": 0.038410283625125885, "learning_rate": 3.65983472512893e-07, "loss": 0.004, "step": 96340 }, { "grad_norm": 0.09042636305093765, "learning_rate": 3.6398926577969616e-07, "loss": 0.0022, "step": 96350 }, { "grad_norm": 0.06613211333751678, "learning_rate": 3.620004871660032e-07, "loss": 0.0037, "step": 96360 }, { "grad_norm": 0.18390502035617828, "learning_rate": 3.600171368893124e-07, "loss": 0.0036, "step": 96370 }, { "grad_norm": 0.06901424378156662, "learning_rate": 3.580392151665113e-07, "loss": 0.0023, "step": 96380 }, { "grad_norm": 0.09238816797733307, "learning_rate": 3.5606672221391024e-07, "loss": 0.003, "step": 96390 }, { "grad_norm": 0.04502619802951813, "learning_rate": 3.540996582472089e-07, "loss": 0.0024, "step": 96400 }, { "grad_norm": 0.1922593116760254, "learning_rate": 3.521380234815297e-07, "loss": 0.0043, "step": 96410 }, { "grad_norm": 0.10478883236646652, "learning_rate": 3.5018181813138985e-07, "loss": 0.003, "step": 96420 }, { "grad_norm": 0.06827367842197418, "learning_rate": 3.482310424107238e-07, "loss": 0.0035, "step": 96430 }, { "grad_norm": 0.13102836906909943, "learning_rate": 3.4628569653285537e-07, "loss": 0.0051, "step": 96440 }, { "grad_norm": 0.04091911017894745, "learning_rate": 3.44345780710531e-07, "loss": 0.0023, "step": 96450 }, { "grad_norm": 0.07223982363939285, "learning_rate": 3.424112951558978e-07, "loss": 0.0035, "step": 96460 }, { "grad_norm": 0.04186321794986725, "learning_rate": 3.40482240080503e-07, "loss": 0.0021, "step": 96470 }, { "grad_norm": 0.23207148909568787, "learning_rate": 3.3855861569531133e-07, "loss": 0.0023, "step": 96480 }, { "grad_norm": 0.11822246015071869, "learning_rate": 3.3664042221068226e-07, "loss": 0.0037, "step": 96490 }, { "grad_norm": 0.0678933635354042, "learning_rate": 3.347276598363924e-07, "loss": 0.0026, "step": 96500 }, { "grad_norm": 0.06731155514717102, "learning_rate": 3.328203287816134e-07, "loss": 0.0039, "step": 96510 }, { "grad_norm": 0.10386389493942261, "learning_rate": 3.3091842925493386e-07, "loss": 0.0047, "step": 96520 }, { "grad_norm": 0.18050259351730347, "learning_rate": 3.2902196146433196e-07, "loss": 0.0026, "step": 96530 }, { "grad_norm": 0.08317311108112335, "learning_rate": 3.2713092561720836e-07, "loss": 0.0024, "step": 96540 }, { "grad_norm": 0.05300386622548103, "learning_rate": 3.2524532192036996e-07, "loss": 0.0037, "step": 96550 }, { "grad_norm": 0.12292318791151047, "learning_rate": 3.2336515058001836e-07, "loss": 0.0043, "step": 96560 }, { "grad_norm": 0.036505915224552155, "learning_rate": 3.2149041180176143e-07, "loss": 0.0032, "step": 96570 }, { "grad_norm": 0.17695361375808716, "learning_rate": 3.1962110579062954e-07, "loss": 0.0042, "step": 96580 }, { "grad_norm": 0.136360764503479, "learning_rate": 3.177572327510314e-07, "loss": 0.0026, "step": 96590 }, { "grad_norm": 0.09722766280174255, "learning_rate": 3.158987928868151e-07, "loss": 0.0031, "step": 96600 }, { "grad_norm": 0.08700592815876007, "learning_rate": 3.1404578640120144e-07, "loss": 0.0039, "step": 96610 }, { "grad_norm": 0.12055033445358276, "learning_rate": 3.1219821349683933e-07, "loss": 0.0034, "step": 96620 }, { "grad_norm": 0.16502754390239716, "learning_rate": 3.103560743757783e-07, "loss": 0.0028, "step": 96630 }, { "grad_norm": 0.13191375136375427, "learning_rate": 3.0851936923946834e-07, "loss": 0.0025, "step": 96640 }, { "grad_norm": 0.12655292451381683, "learning_rate": 3.06688098288771e-07, "loss": 0.0024, "step": 96650 }, { "grad_norm": 0.16739313304424286, "learning_rate": 3.0486226172394825e-07, "loss": 0.0033, "step": 96660 }, { "grad_norm": 0.04887184500694275, "learning_rate": 3.030418597446738e-07, "loss": 0.0025, "step": 96670 }, { "grad_norm": 0.1815863400697708, "learning_rate": 3.012268925500272e-07, "loss": 0.0018, "step": 96680 }, { "grad_norm": 0.061731237918138504, "learning_rate": 2.994173603384831e-07, "loss": 0.0037, "step": 96690 }, { "grad_norm": 0.14505700767040253, "learning_rate": 2.9761326330793317e-07, "loss": 0.0037, "step": 96700 }, { "grad_norm": 0.1288665235042572, "learning_rate": 2.958146016556751e-07, "loss": 0.0024, "step": 96710 }, { "grad_norm": 0.061425745487213135, "learning_rate": 2.94021375578396e-07, "loss": 0.0031, "step": 96720 }, { "grad_norm": 0.08022382110357285, "learning_rate": 2.9223358527221135e-07, "loss": 0.0027, "step": 96730 }, { "grad_norm": 0.09244974702596664, "learning_rate": 2.904512309326257e-07, "loss": 0.0034, "step": 96740 }, { "grad_norm": 0.11245540529489517, "learning_rate": 2.8867431275455545e-07, "loss": 0.0031, "step": 96750 }, { "grad_norm": 0.07248827815055847, "learning_rate": 2.86902830932323e-07, "loss": 0.0028, "step": 96760 }, { "grad_norm": 0.09718122333288193, "learning_rate": 2.851367856596565e-07, "loss": 0.0034, "step": 96770 }, { "grad_norm": 0.11652539670467377, "learning_rate": 2.83376177129685e-07, "loss": 0.0027, "step": 96780 }, { "grad_norm": 0.11024495214223862, "learning_rate": 2.8162100553494884e-07, "loss": 0.0034, "step": 96790 }, { "grad_norm": 0.13248930871486664, "learning_rate": 2.798712710673834e-07, "loss": 0.0051, "step": 96800 }, { "grad_norm": 0.10876121371984482, "learning_rate": 2.7812697391834676e-07, "loss": 0.0026, "step": 96810 }, { "grad_norm": 0.0739123672246933, "learning_rate": 2.7638811427858627e-07, "loss": 0.0023, "step": 96820 }, { "grad_norm": 0.08076518028974533, "learning_rate": 2.746546923382665e-07, "loss": 0.0022, "step": 96830 }, { "grad_norm": 0.10560530424118042, "learning_rate": 2.729267082869469e-07, "loss": 0.0036, "step": 96840 }, { "grad_norm": 0.0721503347158432, "learning_rate": 2.712041623135986e-07, "loss": 0.0033, "step": 96850 }, { "grad_norm": 0.13390037417411804, "learning_rate": 2.694870546065931e-07, "loss": 0.0036, "step": 96860 }, { "grad_norm": 0.24702274799346924, "learning_rate": 2.677753853537246e-07, "loss": 0.0032, "step": 96870 }, { "grad_norm": 0.07930004596710205, "learning_rate": 2.660691547421601e-07, "loss": 0.0044, "step": 96880 }, { "grad_norm": 0.035971056669950485, "learning_rate": 2.6436836295850033e-07, "loss": 0.0047, "step": 96890 }, { "grad_norm": 0.14539721608161926, "learning_rate": 2.6267301018874093e-07, "loss": 0.0023, "step": 96900 }, { "grad_norm": 0.07183855026960373, "learning_rate": 2.6098309661828355e-07, "loss": 0.0027, "step": 96910 }, { "grad_norm": 0.13955548405647278, "learning_rate": 2.592986224319305e-07, "loss": 0.0036, "step": 96920 }, { "grad_norm": 0.1375655233860016, "learning_rate": 2.576195878139009e-07, "loss": 0.003, "step": 96930 }, { "grad_norm": 0.039704252034425735, "learning_rate": 2.5594599294780364e-07, "loss": 0.0019, "step": 96940 }, { "grad_norm": 0.0704093798995018, "learning_rate": 2.542778380166644e-07, "loss": 0.0023, "step": 96950 }, { "grad_norm": 0.08957300335168839, "learning_rate": 2.526151232029095e-07, "loss": 0.005, "step": 96960 }, { "grad_norm": 0.08081039041280746, "learning_rate": 2.5095784868836567e-07, "loss": 0.0031, "step": 96970 }, { "grad_norm": 0.09528979659080505, "learning_rate": 2.493060146542825e-07, "loss": 0.005, "step": 96980 }, { "grad_norm": 0.11095095425844193, "learning_rate": 2.476596212812876e-07, "loss": 0.0036, "step": 96990 }, { "grad_norm": 0.045400816947221756, "learning_rate": 2.4601866874943703e-07, "loss": 0.003, "step": 97000 }, { "grad_norm": 0.0520249605178833, "learning_rate": 2.4438315723818163e-07, "loss": 0.0026, "step": 97010 }, { "grad_norm": 0.1559118926525116, "learning_rate": 2.4275308692637833e-07, "loss": 0.0025, "step": 97020 }, { "grad_norm": 0.11297350376844406, "learning_rate": 2.411284579922901e-07, "loss": 0.0035, "step": 97030 }, { "grad_norm": 0.03881179541349411, "learning_rate": 2.395092706135804e-07, "loss": 0.0037, "step": 97040 }, { "grad_norm": 0.18620485067367554, "learning_rate": 2.3789552496731872e-07, "loss": 0.0047, "step": 97050 }, { "grad_norm": 0.0772264376282692, "learning_rate": 2.3628722122999158e-07, "loss": 0.0021, "step": 97060 }, { "grad_norm": 0.07957375794649124, "learning_rate": 2.3468435957747503e-07, "loss": 0.0046, "step": 97070 }, { "grad_norm": 0.12329074740409851, "learning_rate": 2.330869401850566e-07, "loss": 0.003, "step": 97080 }, { "grad_norm": 0.17427699267864227, "learning_rate": 2.3149496322742437e-07, "loss": 0.0024, "step": 97090 }, { "grad_norm": 0.05544830486178398, "learning_rate": 2.2990842887868346e-07, "loss": 0.0028, "step": 97100 }, { "grad_norm": 0.06044328957796097, "learning_rate": 2.2832733731232292e-07, "loss": 0.0026, "step": 97110 }, { "grad_norm": 0.10961803048849106, "learning_rate": 2.2675168870124886e-07, "loss": 0.0031, "step": 97120 }, { "grad_norm": 0.05355238541960716, "learning_rate": 2.2518148321778455e-07, "loss": 0.0025, "step": 97130 }, { "grad_norm": 0.12169649451971054, "learning_rate": 2.2361672103363706e-07, "loss": 0.0026, "step": 97140 }, { "grad_norm": 0.09667903929948807, "learning_rate": 2.220574023199251e-07, "loss": 0.0034, "step": 97150 }, { "grad_norm": 0.1093805804848671, "learning_rate": 2.2050352724717894e-07, "loss": 0.003, "step": 97160 }, { "grad_norm": 0.08355095237493515, "learning_rate": 2.1895509598532372e-07, "loss": 0.0027, "step": 97170 }, { "grad_norm": 0.18209488689899445, "learning_rate": 2.1741210870369066e-07, "loss": 0.0031, "step": 97180 }, { "grad_norm": 0.12894181907176971, "learning_rate": 2.1587456557102814e-07, "loss": 0.0033, "step": 97190 }, { "grad_norm": 0.059409257024526596, "learning_rate": 2.143424667554683e-07, "loss": 0.0026, "step": 97200 }, { "grad_norm": 0.1062798723578453, "learning_rate": 2.128158124245716e-07, "loss": 0.0032, "step": 97210 }, { "grad_norm": 0.14179417490959167, "learning_rate": 2.1129460274527668e-07, "loss": 0.0035, "step": 97220 }, { "grad_norm": 0.11460397392511368, "learning_rate": 2.0977883788395046e-07, "loss": 0.0033, "step": 97230 }, { "grad_norm": 0.08533518761396408, "learning_rate": 2.0826851800634928e-07, "loss": 0.0027, "step": 97240 }, { "grad_norm": 0.03966989740729332, "learning_rate": 2.0676364327764653e-07, "loss": 0.0026, "step": 97250 }, { "grad_norm": 0.07914827018976212, "learning_rate": 2.0526421386240502e-07, "loss": 0.0018, "step": 97260 }, { "grad_norm": 0.1150270476937294, "learning_rate": 2.0377022992461025e-07, "loss": 0.0029, "step": 97270 }, { "grad_norm": 0.10667160153388977, "learning_rate": 2.0228169162763156e-07, "loss": 0.006, "step": 97280 }, { "grad_norm": 0.15003113448619843, "learning_rate": 2.0079859913425536e-07, "loss": 0.003, "step": 97290 }, { "grad_norm": 0.07480377703905106, "learning_rate": 1.9932095260667417e-07, "loss": 0.0026, "step": 97300 }, { "grad_norm": 0.13252142071723938, "learning_rate": 1.9784875220648092e-07, "loss": 0.0045, "step": 97310 }, { "grad_norm": 0.23936690390110016, "learning_rate": 1.9638199809466907e-07, "loss": 0.0024, "step": 97320 }, { "grad_norm": 0.2028360217809677, "learning_rate": 1.9492069043164364e-07, "loss": 0.0034, "step": 97330 }, { "grad_norm": 0.146353617310524, "learning_rate": 1.9346482937721567e-07, "loss": 0.0031, "step": 97340 }, { "grad_norm": 0.09586599469184875, "learning_rate": 1.920144150905856e-07, "loss": 0.0024, "step": 97350 }, { "grad_norm": 0.16846998035907745, "learning_rate": 1.9056944773037655e-07, "loss": 0.0029, "step": 97360 }, { "grad_norm": 0.10937477648258209, "learning_rate": 1.8912992745460102e-07, "loss": 0.0019, "step": 97370 }, { "grad_norm": 0.05675110965967178, "learning_rate": 1.876958544206886e-07, "loss": 0.0042, "step": 97380 }, { "grad_norm": 0.10108522325754166, "learning_rate": 1.8626722878546942e-07, "loss": 0.0049, "step": 97390 }, { "grad_norm": 0.11170496046543121, "learning_rate": 1.8484405070516851e-07, "loss": 0.0027, "step": 97400 }, { "grad_norm": 0.12166718393564224, "learning_rate": 1.8342632033542805e-07, "loss": 0.0033, "step": 97410 }, { "grad_norm": 0.16177386045455933, "learning_rate": 1.8201403783129066e-07, "loss": 0.0035, "step": 97420 }, { "grad_norm": 0.06930186599493027, "learning_rate": 1.806072033471884e-07, "loss": 0.0053, "step": 97430 }, { "grad_norm": 0.17915478348731995, "learning_rate": 1.79205817036987e-07, "loss": 0.0028, "step": 97440 }, { "grad_norm": 0.13222423195838928, "learning_rate": 1.7780987905393065e-07, "loss": 0.0033, "step": 97450 }, { "grad_norm": 0.039893489331007004, "learning_rate": 1.764193895506805e-07, "loss": 0.0027, "step": 97460 }, { "grad_norm": 0.07482753694057465, "learning_rate": 1.7503434867929824e-07, "loss": 0.0026, "step": 97470 }, { "grad_norm": 0.1351305991411209, "learning_rate": 1.7365475659125165e-07, "loss": 0.0042, "step": 97480 }, { "grad_norm": 0.16277346014976501, "learning_rate": 1.7228061343740332e-07, "loss": 0.0098, "step": 97490 }, { "grad_norm": 0.11842642724514008, "learning_rate": 1.7091191936803308e-07, "loss": 0.0026, "step": 97500 }, { "grad_norm": 0.11528899520635605, "learning_rate": 1.6954867453281563e-07, "loss": 0.0029, "step": 97510 }, { "grad_norm": 0.1683022677898407, "learning_rate": 1.6819087908084285e-07, "loss": 0.0046, "step": 97520 }, { "grad_norm": 0.09895655512809753, "learning_rate": 1.6683853316058485e-07, "loss": 0.0019, "step": 97530 }, { "grad_norm": 0.06764096021652222, "learning_rate": 1.6549163691995106e-07, "loss": 0.0029, "step": 97540 }, { "grad_norm": 0.1865319162607193, "learning_rate": 1.641501905062237e-07, "loss": 0.003, "step": 97550 }, { "grad_norm": 0.12981756031513214, "learning_rate": 1.628141940661021e-07, "loss": 0.005, "step": 97560 }, { "grad_norm": 0.11822565644979477, "learning_rate": 1.6148364774569158e-07, "loss": 0.0026, "step": 97570 }, { "grad_norm": 0.07823007553815842, "learning_rate": 1.6015855169050355e-07, "loss": 0.0037, "step": 97580 }, { "grad_norm": 0.12575389444828033, "learning_rate": 1.5883890604543873e-07, "loss": 0.0026, "step": 97590 }, { "grad_norm": 0.05081149563193321, "learning_rate": 1.5752471095482057e-07, "loss": 0.0043, "step": 97600 }, { "grad_norm": 0.10470612347126007, "learning_rate": 1.5621596656235638e-07, "loss": 0.003, "step": 97610 }, { "grad_norm": 0.09776286035776138, "learning_rate": 1.5491267301117608e-07, "loss": 0.0054, "step": 97620 }, { "grad_norm": 0.06512879580259323, "learning_rate": 1.536148304438101e-07, "loss": 0.0037, "step": 97630 }, { "grad_norm": 0.12826858460903168, "learning_rate": 1.5232243900217823e-07, "loss": 0.005, "step": 97640 }, { "grad_norm": 0.07114163786172867, "learning_rate": 1.51035498827623e-07, "loss": 0.0023, "step": 97650 }, { "grad_norm": 0.09286166727542877, "learning_rate": 1.497540100608763e-07, "loss": 0.002, "step": 97660 }, { "grad_norm": 0.04338790848851204, "learning_rate": 1.4847797284208153e-07, "loss": 0.0025, "step": 97670 }, { "grad_norm": 0.19105537235736847, "learning_rate": 1.472073873107882e-07, "loss": 0.0024, "step": 97680 }, { "grad_norm": 0.17400327324867249, "learning_rate": 1.459422536059407e-07, "loss": 0.004, "step": 97690 }, { "grad_norm": 0.11201420426368713, "learning_rate": 1.4468257186589508e-07, "loss": 0.0038, "step": 97700 }, { "grad_norm": 0.10910457372665405, "learning_rate": 1.4342834222840217e-07, "loss": 0.0032, "step": 97710 }, { "grad_norm": 0.034947529435157776, "learning_rate": 1.4217956483063566e-07, "loss": 0.0018, "step": 97720 }, { "grad_norm": 0.06913989037275314, "learning_rate": 1.409362398091474e-07, "loss": 0.0028, "step": 97730 }, { "grad_norm": 0.0740179643034935, "learning_rate": 1.3969836729990638e-07, "loss": 0.0027, "step": 97740 }, { "grad_norm": 0.08562498539686203, "learning_rate": 1.384659474382932e-07, "loss": 0.0024, "step": 97750 }, { "grad_norm": 0.08583921194076538, "learning_rate": 1.372389803590779e-07, "loss": 0.0034, "step": 97760 }, { "grad_norm": 0.1439894288778305, "learning_rate": 1.360174661964364e-07, "loss": 0.003, "step": 97770 }, { "grad_norm": 0.0702928900718689, "learning_rate": 1.3480140508396188e-07, "loss": 0.0032, "step": 97780 }, { "grad_norm": 0.08594033122062683, "learning_rate": 1.335907971546313e-07, "loss": 0.0031, "step": 97790 }, { "grad_norm": 0.09220763295888901, "learning_rate": 1.3238564254083875e-07, "loss": 0.0029, "step": 97800 }, { "grad_norm": 0.07323311269283295, "learning_rate": 1.3118594137437323e-07, "loss": 0.0033, "step": 97810 }, { "grad_norm": 0.07452400773763657, "learning_rate": 1.2999169378644094e-07, "loss": 0.0027, "step": 97820 }, { "grad_norm": 0.20315471291542053, "learning_rate": 1.2880289990763738e-07, "loss": 0.0038, "step": 97830 }, { "grad_norm": 0.047468919306993484, "learning_rate": 1.2761955986796968e-07, "loss": 0.0027, "step": 97840 }, { "grad_norm": 0.16627807915210724, "learning_rate": 1.264416737968399e-07, "loss": 0.0032, "step": 97850 }, { "grad_norm": 0.10655507445335388, "learning_rate": 1.2526924182307275e-07, "loss": 0.005, "step": 97860 }, { "grad_norm": 0.1909611076116562, "learning_rate": 1.2410226407487124e-07, "loss": 0.003, "step": 97870 }, { "grad_norm": 0.05424521490931511, "learning_rate": 1.229407406798555e-07, "loss": 0.0023, "step": 97880 }, { "grad_norm": 0.044930875301361084, "learning_rate": 1.2178467176505726e-07, "loss": 0.0023, "step": 97890 }, { "grad_norm": 0.04887603968381882, "learning_rate": 1.2063405745689205e-07, "loss": 0.003, "step": 97900 }, { "grad_norm": 0.09113325923681259, "learning_rate": 1.1948889788119812e-07, "loss": 0.0042, "step": 97910 }, { "grad_norm": 0.05851174518465996, "learning_rate": 1.1834919316320303e-07, "loss": 0.0033, "step": 97920 }, { "grad_norm": 0.08350042998790741, "learning_rate": 1.1721494342754048e-07, "loss": 0.0031, "step": 97930 }, { "grad_norm": 0.05666443333029747, "learning_rate": 1.1608614879825563e-07, "loss": 0.0028, "step": 97940 }, { "grad_norm": 0.1798010617494583, "learning_rate": 1.1496280939879422e-07, "loss": 0.0045, "step": 97950 }, { "grad_norm": 0.13306553661823273, "learning_rate": 1.1384492535199687e-07, "loss": 0.0023, "step": 97960 }, { "grad_norm": 0.07070175558328629, "learning_rate": 1.1273249678011578e-07, "loss": 0.0021, "step": 97970 }, { "grad_norm": 0.046159300953149796, "learning_rate": 1.1162552380480362e-07, "loss": 0.0048, "step": 97980 }, { "grad_norm": 0.08870332688093185, "learning_rate": 1.1052400654711359e-07, "loss": 0.0035, "step": 97990 }, { "grad_norm": 0.1114366203546524, "learning_rate": 1.0942794512751597e-07, "loss": 0.0038, "step": 98000 }, { "grad_norm": 0.19311441481113434, "learning_rate": 1.0833733966587045e-07, "loss": 0.0048, "step": 98010 }, { "grad_norm": 0.14932076632976532, "learning_rate": 1.0725219028143718e-07, "loss": 0.0029, "step": 98020 }, { "grad_norm": 0.06865562498569489, "learning_rate": 1.0617249709289345e-07, "loss": 0.0047, "step": 98030 }, { "grad_norm": 0.09192389249801636, "learning_rate": 1.0509826021831703e-07, "loss": 0.0029, "step": 98040 }, { "grad_norm": 0.1460018903017044, "learning_rate": 1.0402947977517508e-07, "loss": 0.0026, "step": 98050 }, { "grad_norm": 0.0482543408870697, "learning_rate": 1.0296615588035185e-07, "loss": 0.0022, "step": 98060 }, { "grad_norm": 0.12601062655448914, "learning_rate": 1.0190828865012659e-07, "loss": 0.0027, "step": 98070 }, { "grad_norm": 0.14623083174228668, "learning_rate": 1.008558782001956e-07, "loss": 0.0027, "step": 98080 }, { "grad_norm": 0.09874515980482101, "learning_rate": 9.980892464563906e-08, "loss": 0.0028, "step": 98090 }, { "grad_norm": 0.09379791468381882, "learning_rate": 9.876742810095985e-08, "loss": 0.0045, "step": 98100 }, { "grad_norm": 0.1158272922039032, "learning_rate": 9.77313886800446e-08, "loss": 0.0028, "step": 98110 }, { "grad_norm": 0.126408189535141, "learning_rate": 9.670080649619717e-08, "loss": 0.0028, "step": 98120 }, { "grad_norm": 0.21956712007522583, "learning_rate": 9.567568166212737e-08, "loss": 0.0054, "step": 98130 }, { "grad_norm": 0.10506752133369446, "learning_rate": 9.465601428992888e-08, "loss": 0.0053, "step": 98140 }, { "grad_norm": 0.042966946959495544, "learning_rate": 9.364180449111803e-08, "loss": 0.003, "step": 98150 }, { "grad_norm": 0.1551857441663742, "learning_rate": 9.263305237661169e-08, "loss": 0.0035, "step": 98160 }, { "grad_norm": 0.15382863581180573, "learning_rate": 9.162975805671603e-08, "loss": 0.005, "step": 98170 }, { "grad_norm": 0.17326632142066956, "learning_rate": 9.063192164115442e-08, "loss": 0.004, "step": 98180 }, { "grad_norm": 0.04861123859882355, "learning_rate": 8.963954323904511e-08, "loss": 0.0027, "step": 98190 }, { "grad_norm": 0.12329528480768204, "learning_rate": 8.865262295891796e-08, "loss": 0.0037, "step": 98200 }, { "grad_norm": 0.09181713312864304, "learning_rate": 8.767116090870331e-08, "loss": 0.0036, "step": 98210 }, { "grad_norm": 0.04682282358407974, "learning_rate": 8.669515719572086e-08, "loss": 0.003, "step": 98220 }, { "grad_norm": 0.08346761018037796, "learning_rate": 8.572461192671855e-08, "loss": 0.0039, "step": 98230 }, { "grad_norm": 0.08991818130016327, "learning_rate": 8.475952520782815e-08, "loss": 0.0044, "step": 98240 }, { "grad_norm": 0.10239174962043762, "learning_rate": 8.379989714458747e-08, "loss": 0.0022, "step": 98250 }, { "grad_norm": 0.11663763225078583, "learning_rate": 8.284572784194034e-08, "loss": 0.0031, "step": 98260 }, { "grad_norm": 0.05023202300071716, "learning_rate": 8.189701740424216e-08, "loss": 0.0031, "step": 98270 }, { "grad_norm": 0.10572415590286255, "learning_rate": 8.095376593522663e-08, "loss": 0.0027, "step": 98280 }, { "grad_norm": 0.03970269113779068, "learning_rate": 8.001597353806123e-08, "loss": 0.0027, "step": 98290 }, { "grad_norm": 0.11526350677013397, "learning_rate": 7.90836403152917e-08, "loss": 0.0027, "step": 98300 }, { "grad_norm": 0.1404576450586319, "learning_rate": 7.815676636888092e-08, "loss": 0.0045, "step": 98310 }, { "grad_norm": 0.168972909450531, "learning_rate": 7.723535180019226e-08, "loss": 0.003, "step": 98320 }, { "grad_norm": 0.21917064487934113, "learning_rate": 7.6319396709984e-08, "loss": 0.0038, "step": 98330 }, { "grad_norm": 0.0922110304236412, "learning_rate": 7.540890119843158e-08, "loss": 0.002, "step": 98340 }, { "grad_norm": 0.24423161149024963, "learning_rate": 7.450386536509979e-08, "loss": 0.0018, "step": 98350 }, { "grad_norm": 0.08411676436662674, "learning_rate": 7.360428930895947e-08, "loss": 0.0031, "step": 98360 }, { "grad_norm": 0.05120910704135895, "learning_rate": 7.271017312839302e-08, "loss": 0.0042, "step": 98370 }, { "grad_norm": 0.1157933846116066, "learning_rate": 7.182151692118333e-08, "loss": 0.0045, "step": 98380 }, { "grad_norm": 0.1878184676170349, "learning_rate": 7.093832078449713e-08, "loss": 0.0024, "step": 98390 }, { "grad_norm": 0.12332573533058167, "learning_rate": 7.006058481493493e-08, "loss": 0.0038, "step": 98400 }, { "grad_norm": 0.11059872806072235, "learning_rate": 6.918830910847552e-08, "loss": 0.0023, "step": 98410 }, { "grad_norm": 0.14314085245132446, "learning_rate": 6.832149376052033e-08, "loss": 0.0024, "step": 98420 }, { "grad_norm": 0.12047569453716278, "learning_rate": 6.746013886584912e-08, "loss": 0.0037, "step": 98430 }, { "grad_norm": 0.21560560166835785, "learning_rate": 6.660424451866431e-08, "loss": 0.0031, "step": 98440 }, { "grad_norm": 0.06090700626373291, "learning_rate": 6.575381081256327e-08, "loss": 0.0025, "step": 98450 }, { "grad_norm": 0.21260280907154083, "learning_rate": 6.49088378405549e-08, "loss": 0.0024, "step": 98460 }, { "grad_norm": 0.17744305729866028, "learning_rate": 6.406932569503755e-08, "loss": 0.0019, "step": 98470 }, { "grad_norm": 0.10194661468267441, "learning_rate": 6.323527446782662e-08, "loss": 0.003, "step": 98480 }, { "grad_norm": 0.10904590785503387, "learning_rate": 6.240668425012142e-08, "loss": 0.0021, "step": 98490 }, { "grad_norm": 0.16125261783599854, "learning_rate": 6.158355513254388e-08, "loss": 0.0025, "step": 98500 }, { "grad_norm": 0.08679782599210739, "learning_rate": 6.076588720510534e-08, "loss": 0.0027, "step": 98510 }, { "grad_norm": 0.11418114602565765, "learning_rate": 5.995368055722316e-08, "loss": 0.0042, "step": 98520 }, { "grad_norm": 0.16091087460517883, "learning_rate": 5.914693527773185e-08, "loss": 0.0028, "step": 98530 }, { "grad_norm": 0.13661183416843414, "learning_rate": 5.834565145483861e-08, "loss": 0.0037, "step": 98540 }, { "grad_norm": 0.16354383528232574, "learning_rate": 5.754982917618446e-08, "loss": 0.0026, "step": 98550 }, { "grad_norm": 0.12915386259555817, "learning_rate": 5.675946852878866e-08, "loss": 0.0031, "step": 98560 }, { "grad_norm": 0.10182313621044159, "learning_rate": 5.59745695990932e-08, "loss": 0.0023, "step": 98570 }, { "grad_norm": 0.15123379230499268, "learning_rate": 5.51951324729294e-08, "loss": 0.002, "step": 98580 }, { "grad_norm": 0.10681262612342834, "learning_rate": 5.4421157235529093e-08, "loss": 0.0023, "step": 98590 }, { "grad_norm": 0.09565936774015427, "learning_rate": 5.365264397154679e-08, "loss": 0.0029, "step": 98600 }, { "grad_norm": 0.1938035637140274, "learning_rate": 5.288959276501526e-08, "loss": 0.0043, "step": 98610 }, { "grad_norm": 0.040781281888484955, "learning_rate": 5.2132003699378876e-08, "loss": 0.0031, "step": 98620 }, { "grad_norm": 0.11228805780410767, "learning_rate": 5.137987685749912e-08, "loss": 0.0022, "step": 98630 }, { "grad_norm": 0.17649903893470764, "learning_rate": 5.0633212321610227e-08, "loss": 0.0049, "step": 98640 }, { "grad_norm": 0.04848748818039894, "learning_rate": 4.989201017338019e-08, "loss": 0.003, "step": 98650 }, { "grad_norm": 0.06649629771709442, "learning_rate": 4.9156270493860846e-08, "loss": 0.004, "step": 98660 }, { "grad_norm": 0.04721957817673683, "learning_rate": 4.842599336351561e-08, "loss": 0.0025, "step": 98670 }, { "grad_norm": 0.08156514167785645, "learning_rate": 4.7701178862197274e-08, "loss": 0.0028, "step": 98680 }, { "grad_norm": 0.08433841168880463, "learning_rate": 4.698182706918131e-08, "loss": 0.0026, "step": 98690 }, { "grad_norm": 0.09602056443691254, "learning_rate": 4.6267938063121466e-08, "loss": 0.0046, "step": 98700 }, { "grad_norm": 0.054765909910202026, "learning_rate": 4.555951192209973e-08, "loss": 0.0029, "step": 98710 }, { "grad_norm": 0.06661861389875412, "learning_rate": 4.485654872358747e-08, "loss": 0.0027, "step": 98720 }, { "grad_norm": 0.0928291529417038, "learning_rate": 4.415904854445097e-08, "loss": 0.0056, "step": 98730 }, { "grad_norm": 0.11751092970371246, "learning_rate": 4.34670114609792e-08, "loss": 0.0033, "step": 98740 }, { "grad_norm": 0.12302130460739136, "learning_rate": 4.278043754884498e-08, "loss": 0.0036, "step": 98750 }, { "grad_norm": 0.13241888582706451, "learning_rate": 4.209932688312712e-08, "loss": 0.0032, "step": 98760 }, { "grad_norm": 0.08677788078784943, "learning_rate": 4.14236795383216e-08, "loss": 0.0028, "step": 98770 }, { "grad_norm": 0.07830854505300522, "learning_rate": 4.075349558830821e-08, "loss": 0.0022, "step": 98780 }, { "grad_norm": 0.09107000380754471, "learning_rate": 4.008877510638387e-08, "loss": 0.0023, "step": 98790 }, { "grad_norm": 0.15608185529708862, "learning_rate": 3.942951816523488e-08, "loss": 0.0035, "step": 98800 }, { "grad_norm": 0.17838536202907562, "learning_rate": 3.8775724836959125e-08, "loss": 0.003, "step": 98810 }, { "grad_norm": 0.0646522119641304, "learning_rate": 3.812739519305497e-08, "loss": 0.0039, "step": 98820 }, { "grad_norm": 0.0996401458978653, "learning_rate": 3.748452930442126e-08, "loss": 0.0028, "step": 98830 }, { "grad_norm": 0.05371544882655144, "learning_rate": 3.6847127241362855e-08, "loss": 0.002, "step": 98840 }, { "grad_norm": 0.07681956142187119, "learning_rate": 3.621518907358512e-08, "loss": 0.0025, "step": 98850 }, { "grad_norm": 0.16249258816242218, "learning_rate": 3.558871487019388e-08, "loss": 0.005, "step": 98860 }, { "grad_norm": 0.12505139410495758, "learning_rate": 3.4967704699701007e-08, "loss": 0.004, "step": 98870 }, { "grad_norm": 0.03277362510561943, "learning_rate": 3.435215863001884e-08, "loss": 0.0032, "step": 98880 }, { "grad_norm": 0.09139536321163177, "learning_rate": 3.374207672846019e-08, "loss": 0.003, "step": 98890 }, { "grad_norm": 0.05046888068318367, "learning_rate": 3.313745906174392e-08, "loss": 0.003, "step": 98900 }, { "grad_norm": 0.13189518451690674, "learning_rate": 3.253830569599492e-08, "loss": 0.0028, "step": 98910 }, { "grad_norm": 0.13226135075092316, "learning_rate": 3.1944616696727436e-08, "loss": 0.0033, "step": 98920 }, { "grad_norm": 0.14053551852703094, "learning_rate": 3.135639212887287e-08, "loss": 0.0028, "step": 98930 }, { "grad_norm": 0.10635548830032349, "learning_rate": 3.077363205675754e-08, "loss": 0.0026, "step": 98940 }, { "grad_norm": 0.059805385768413544, "learning_rate": 3.0196336544113804e-08, "loss": 0.0036, "step": 98950 }, { "grad_norm": 0.07876714318990707, "learning_rate": 2.9624505654063383e-08, "loss": 0.0023, "step": 98960 }, { "grad_norm": 0.11378622055053711, "learning_rate": 2.90581394491507e-08, "loss": 0.0059, "step": 98970 }, { "grad_norm": 0.11520421504974365, "learning_rate": 2.8497237991309545e-08, "loss": 0.0025, "step": 98980 }, { "grad_norm": 0.07258860766887665, "learning_rate": 2.7941801341879735e-08, "loss": 0.0025, "step": 98990 }, { "grad_norm": 0.10524914413690567, "learning_rate": 2.7391829561601578e-08, "loss": 0.0023, "step": 99000 }, { "grad_norm": 0.1365942806005478, "learning_rate": 2.6847322710621403e-08, "loss": 0.0018, "step": 99010 }, { "grad_norm": 0.1458994746208191, "learning_rate": 2.6308280848486022e-08, "loss": 0.0029, "step": 99020 }, { "grad_norm": 0.12961414456367493, "learning_rate": 2.5774704034137177e-08, "loss": 0.0035, "step": 99030 }, { "grad_norm": 0.06602068990468979, "learning_rate": 2.524659232593929e-08, "loss": 0.0021, "step": 99040 }, { "grad_norm": 0.23742827773094177, "learning_rate": 2.472394578163506e-08, "loss": 0.003, "step": 99050 }, { "grad_norm": 0.0599246546626091, "learning_rate": 2.4206764458378772e-08, "loss": 0.0036, "step": 99060 }, { "grad_norm": 0.04903441295027733, "learning_rate": 2.3695048412736285e-08, "loss": 0.0021, "step": 99070 }, { "grad_norm": 0.11795556545257568, "learning_rate": 2.3188797700668395e-08, "loss": 0.0025, "step": 99080 }, { "grad_norm": 0.1068144217133522, "learning_rate": 2.268801237753082e-08, "loss": 0.0023, "step": 99090 }, { "grad_norm": 0.14957167208194733, "learning_rate": 2.2192692498090862e-08, "loss": 0.0031, "step": 99100 }, { "grad_norm": 0.10089729726314545, "learning_rate": 2.170283811652185e-08, "loss": 0.003, "step": 99110 }, { "grad_norm": 0.07376433163881302, "learning_rate": 2.12184492863865e-08, "loss": 0.0031, "step": 99120 }, { "grad_norm": 0.07446866482496262, "learning_rate": 2.073952606066465e-08, "loss": 0.0033, "step": 99130 }, { "grad_norm": 0.15488353371620178, "learning_rate": 2.026606849171997e-08, "loss": 0.0036, "step": 99140 }, { "grad_norm": 0.17895717918872833, "learning_rate": 1.9798076631333262e-08, "loss": 0.0026, "step": 99150 }, { "grad_norm": 0.09212162345647812, "learning_rate": 1.933555053069136e-08, "loss": 0.0021, "step": 99160 }, { "grad_norm": 0.04556843265891075, "learning_rate": 1.887849024036492e-08, "loss": 0.0025, "step": 99170 }, { "grad_norm": 0.15259774029254913, "learning_rate": 1.8426895810341736e-08, "loss": 0.0052, "step": 99180 }, { "grad_norm": 0.10482978820800781, "learning_rate": 1.798076729000453e-08, "loss": 0.0024, "step": 99190 }, { "grad_norm": 0.11458390206098557, "learning_rate": 1.75401047281476e-08, "loss": 0.0019, "step": 99200 }, { "grad_norm": 0.11137744784355164, "learning_rate": 1.7104908172954626e-08, "loss": 0.0043, "step": 99210 }, { "grad_norm": 0.09961707144975662, "learning_rate": 1.6675177672020868e-08, "loss": 0.0041, "step": 99220 }, { "grad_norm": 0.1366555392742157, "learning_rate": 1.6250913272342073e-08, "loss": 0.0022, "step": 99230 }, { "grad_norm": 0.1503373086452484, "learning_rate": 1.5832115020314453e-08, "loss": 0.0029, "step": 99240 }, { "grad_norm": 0.0827169269323349, "learning_rate": 1.5418782961734712e-08, "loss": 0.0027, "step": 99250 }, { "grad_norm": 0.12708023190498352, "learning_rate": 1.501091714181113e-08, "loss": 0.003, "step": 99260 }, { "grad_norm": 0.08083398640155792, "learning_rate": 1.460851760513582e-08, "loss": 0.0027, "step": 99270 }, { "grad_norm": 0.09448296576738358, "learning_rate": 1.4211584395723566e-08, "loss": 0.0022, "step": 99280 }, { "grad_norm": 0.07897444814443588, "learning_rate": 1.382011755697854e-08, "loss": 0.0031, "step": 99290 }, { "grad_norm": 0.11175759136676788, "learning_rate": 1.3434117131716494e-08, "loss": 0.0018, "step": 99300 }, { "grad_norm": 0.22659608721733093, "learning_rate": 1.3053583162142557e-08, "loss": 0.0028, "step": 99310 }, { "grad_norm": 0.13676542043685913, "learning_rate": 1.2678515689873438e-08, "loss": 0.0035, "step": 99320 }, { "grad_norm": 0.12560534477233887, "learning_rate": 1.2308914755931877e-08, "loss": 0.0031, "step": 99330 }, { "grad_norm": 0.08297558128833771, "learning_rate": 1.1944780400729993e-08, "loss": 0.0025, "step": 99340 }, { "grad_norm": 0.15777146816253662, "learning_rate": 1.1586112664085935e-08, "loss": 0.0055, "step": 99350 }, { "grad_norm": 0.18475289642810822, "learning_rate": 1.1232911585234984e-08, "loss": 0.0026, "step": 99360 }, { "grad_norm": 0.11157451570034027, "learning_rate": 1.0885177202790697e-08, "loss": 0.0048, "step": 99370 }, { "grad_norm": 0.14133521914482117, "learning_rate": 1.0542909554789315e-08, "loss": 0.0028, "step": 99380 }, { "grad_norm": 0.11229553073644638, "learning_rate": 1.0206108678656456e-08, "loss": 0.0036, "step": 99390 }, { "grad_norm": 0.1692482978105545, "learning_rate": 9.874774611223769e-09, "loss": 0.0027, "step": 99400 }, { "grad_norm": 0.12117423117160797, "learning_rate": 9.548907388728933e-09, "loss": 0.0034, "step": 99410 }, { "grad_norm": 0.0461953729391098, "learning_rate": 9.228507046804557e-09, "loss": 0.0036, "step": 99420 }, { "grad_norm": 0.11964564025402069, "learning_rate": 8.91357362048928e-09, "loss": 0.0026, "step": 99430 }, { "grad_norm": 0.14431226253509521, "learning_rate": 8.604107144227769e-09, "loss": 0.0026, "step": 99440 }, { "grad_norm": 0.06552307307720184, "learning_rate": 8.300107651859623e-09, "loss": 0.0033, "step": 99450 }, { "grad_norm": 0.1581479161977768, "learning_rate": 8.001575176630472e-09, "loss": 0.004, "step": 99460 }, { "grad_norm": 0.06302393227815628, "learning_rate": 7.708509751186422e-09, "loss": 0.0031, "step": 99470 }, { "grad_norm": 0.15429799258708954, "learning_rate": 7.420911407579611e-09, "loss": 0.0069, "step": 99480 }, { "grad_norm": 0.12091081589460373, "learning_rate": 7.13878017725711e-09, "loss": 0.0022, "step": 99490 }, { "grad_norm": 0.08707492798566818, "learning_rate": 6.8621160910720125e-09, "loss": 0.002, "step": 99500 }, { "grad_norm": 0.10716226696968079, "learning_rate": 6.59091917928345e-09, "loss": 0.0028, "step": 99510 }, { "grad_norm": 0.07265043258666992, "learning_rate": 6.32518947155103e-09, "loss": 0.0018, "step": 99520 }, { "grad_norm": 0.1897108107805252, "learning_rate": 6.064926996929288e-09, "loss": 0.0049, "step": 99530 }, { "grad_norm": 0.08403678983449936, "learning_rate": 5.810131783884343e-09, "loss": 0.004, "step": 99540 }, { "grad_norm": 0.10495772957801819, "learning_rate": 5.5608038602772415e-09, "loss": 0.0035, "step": 99550 }, { "grad_norm": 0.09705901890993118, "learning_rate": 5.316943253375062e-09, "loss": 0.0027, "step": 99560 }, { "grad_norm": 0.0795309841632843, "learning_rate": 5.07854998984536e-09, "loss": 0.004, "step": 99570 }, { "grad_norm": 0.1313929408788681, "learning_rate": 4.845624095756174e-09, "loss": 0.0044, "step": 99580 }, { "grad_norm": 0.08917403221130371, "learning_rate": 4.618165596587121e-09, "loss": 0.0026, "step": 99590 }, { "grad_norm": 0.1293553113937378, "learning_rate": 4.396174517207197e-09, "loss": 0.0041, "step": 99600 }, { "grad_norm": 0.07875996828079224, "learning_rate": 4.179650881896979e-09, "loss": 0.0022, "step": 99610 }, { "grad_norm": 0.0867341160774231, "learning_rate": 3.968594714331975e-09, "loss": 0.0025, "step": 99620 }, { "grad_norm": 0.11657476425170898, "learning_rate": 3.763006037593719e-09, "loss": 0.0048, "step": 99630 }, { "grad_norm": 0.10908109694719315, "learning_rate": 3.5628848741586786e-09, "loss": 0.0043, "step": 99640 }, { "grad_norm": 0.09792764484882355, "learning_rate": 3.368231245926001e-09, "loss": 0.0036, "step": 99650 }, { "grad_norm": 0.06857029348611832, "learning_rate": 3.179045174167561e-09, "loss": 0.0025, "step": 99660 }, { "grad_norm": 0.16978518664836884, "learning_rate": 2.9953266795834657e-09, "loss": 0.0025, "step": 99670 }, { "grad_norm": 0.08516296744346619, "learning_rate": 2.817075782263201e-09, "loss": 0.0033, "step": 99680 }, { "grad_norm": 0.1313149631023407, "learning_rate": 2.6442925016967325e-09, "loss": 0.0029, "step": 99690 }, { "grad_norm": 0.14324021339416504, "learning_rate": 2.4769768567800555e-09, "loss": 0.005, "step": 99700 }, { "grad_norm": 0.1223907545208931, "learning_rate": 2.315128865809646e-09, "loss": 0.0022, "step": 99710 }, { "grad_norm": 0.10990612208843231, "learning_rate": 2.1587485464880097e-09, "loss": 0.0025, "step": 99720 }, { "grad_norm": 0.04133474454283714, "learning_rate": 2.0078359159125813e-09, "loss": 0.0034, "step": 99730 }, { "grad_norm": 0.0978722795844078, "learning_rate": 1.8623909905923776e-09, "loss": 0.002, "step": 99740 }, { "grad_norm": 0.08313076943159103, "learning_rate": 1.7224137864257916e-09, "loss": 0.0022, "step": 99750 }, { "grad_norm": 0.09785076230764389, "learning_rate": 1.5879043187283505e-09, "loss": 0.0027, "step": 99760 }, { "grad_norm": 0.09090428799390793, "learning_rate": 1.458862602204958e-09, "loss": 0.0042, "step": 99770 }, { "grad_norm": 0.08860364556312561, "learning_rate": 1.3352886509720996e-09, "loss": 0.0026, "step": 99780 }, { "grad_norm": 0.08172298967838287, "learning_rate": 1.2171824785356389e-09, "loss": 0.0031, "step": 99790 }, { "grad_norm": 0.1431373953819275, "learning_rate": 1.1045440978185718e-09, "loss": 0.0028, "step": 99800 }, { "grad_norm": 0.2302117794752121, "learning_rate": 9.973735211332714e-10, "loss": 0.0023, "step": 99810 }, { "grad_norm": 0.03434794396162033, "learning_rate": 8.956707602036929e-10, "loss": 0.0031, "step": 99820 }, { "grad_norm": 0.1786935031414032, "learning_rate": 7.994358261542712e-10, "loss": 0.0022, "step": 99830 }, { "grad_norm": 0.10923613607883453, "learning_rate": 7.086687295043692e-10, "loss": 0.0022, "step": 99840 }, { "grad_norm": 0.09203235059976578, "learning_rate": 6.23369480179381e-10, "loss": 0.0029, "step": 99850 }, { "grad_norm": 0.18970797955989838, "learning_rate": 5.43538087510731e-10, "loss": 0.0043, "step": 99860 }, { "grad_norm": 0.07536488771438599, "learning_rate": 4.691745602303233e-10, "loss": 0.0044, "step": 99870 }, { "grad_norm": 0.1362202763557434, "learning_rate": 4.0027890646499034e-10, "loss": 0.0026, "step": 99880 }, { "grad_norm": 0.05484582856297493, "learning_rate": 3.368511337531466e-10, "loss": 0.0035, "step": 99890 }, { "grad_norm": 0.20611049234867096, "learning_rate": 2.788912490281348e-10, "loss": 0.0025, "step": 99900 }, { "grad_norm": 0.06766333431005478, "learning_rate": 2.2639925863487954e-10, "loss": 0.0037, "step": 99910 }, { "grad_norm": 0.06827490031719208, "learning_rate": 1.7937516830213163e-10, "loss": 0.003, "step": 99920 }, { "grad_norm": 0.088434137403965, "learning_rate": 1.3781898318687702e-10, "loss": 0.0021, "step": 99930 }, { "grad_norm": 0.08590427786111832, "learning_rate": 1.0173070781882566e-10, "loss": 0.0028, "step": 99940 }, { "grad_norm": 0.1147371381521225, "learning_rate": 7.111034615592261e-11, "loss": 0.0031, "step": 99950 }, { "grad_norm": 0.05324018746614456, "learning_rate": 4.595790153993918e-11, "loss": 0.0019, "step": 99960 }, { "grad_norm": 0.11983740329742432, "learning_rate": 2.6273376729779585e-11, "loss": 0.0035, "step": 99970 }, { "grad_norm": 0.14406542479991913, "learning_rate": 1.2056773868174276e-11, "loss": 0.0033, "step": 99980 }, { "grad_norm": 0.07417076081037521, "learning_rate": 3.3080945149865926e-12, "loss": 0.0021, "step": 99990 }, { "grad_norm": 0.050665825605392456, "learning_rate": 2.7339625008337977e-14, "loss": 0.0037, "step": 100000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 5000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 48, "trial_name": null, "trial_params": null }