{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.8, "eval_steps": 500, "global_step": 40000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.6401886940002441, "learning_rate": 3.6e-07, "loss": 1.3149, "step": 10 }, { "grad_norm": 1.8945657014846802, "learning_rate": 7.6e-07, "loss": 1.3163, "step": 20 }, { "grad_norm": 1.5859267711639404, "learning_rate": 1.16e-06, "loss": 1.2994, "step": 30 }, { "grad_norm": 1.2642822265625, "learning_rate": 1.56e-06, "loss": 1.297, "step": 40 }, { "grad_norm": 0.8344032168388367, "learning_rate": 1.96e-06, "loss": 1.2569, "step": 50 }, { "grad_norm": 0.49314084649086, "learning_rate": 2.36e-06, "loss": 1.2307, "step": 60 }, { "grad_norm": 0.32446572184562683, "learning_rate": 2.7600000000000003e-06, "loss": 1.203, "step": 70 }, { "grad_norm": 0.3321860134601593, "learning_rate": 3.1600000000000007e-06, "loss": 1.1971, "step": 80 }, { "grad_norm": 0.4447484612464905, "learning_rate": 3.5600000000000002e-06, "loss": 1.1866, "step": 90 }, { "grad_norm": 0.6310877799987793, "learning_rate": 3.96e-06, "loss": 1.1729, "step": 100 }, { "grad_norm": 0.47680121660232544, "learning_rate": 4.360000000000001e-06, "loss": 1.1282, "step": 110 }, { "grad_norm": 0.5493061542510986, "learning_rate": 4.76e-06, "loss": 1.0984, "step": 120 }, { "grad_norm": 0.5502547025680542, "learning_rate": 5.1600000000000006e-06, "loss": 1.0866, "step": 130 }, { "grad_norm": 0.3978302776813507, "learning_rate": 5.56e-06, "loss": 1.0817, "step": 140 }, { "grad_norm": 0.4087013304233551, "learning_rate": 5.9600000000000005e-06, "loss": 1.0658, "step": 150 }, { "grad_norm": 0.5459769368171692, "learning_rate": 6.360000000000001e-06, "loss": 1.0608, "step": 160 }, { "grad_norm": 0.519679069519043, "learning_rate": 6.76e-06, "loss": 1.0519, "step": 170 }, { "grad_norm": 0.5130072236061096, "learning_rate": 7.16e-06, "loss": 1.0519, "step": 180 }, { "grad_norm": 0.4664212167263031, "learning_rate": 7.5600000000000005e-06, "loss": 1.0414, "step": 190 }, { "grad_norm": 0.48712092638015747, "learning_rate": 7.96e-06, "loss": 1.0413, "step": 200 }, { "grad_norm": 0.47439441084861755, "learning_rate": 8.36e-06, "loss": 1.0382, "step": 210 }, { "grad_norm": 0.46334874629974365, "learning_rate": 8.76e-06, "loss": 1.0401, "step": 220 }, { "grad_norm": 0.4873521327972412, "learning_rate": 9.16e-06, "loss": 1.0294, "step": 230 }, { "grad_norm": 0.47923916578292847, "learning_rate": 9.560000000000002e-06, "loss": 1.0318, "step": 240 }, { "grad_norm": 0.42912131547927856, "learning_rate": 9.96e-06, "loss": 1.0239, "step": 250 }, { "grad_norm": 0.4370642602443695, "learning_rate": 1.036e-05, "loss": 1.028, "step": 260 }, { "grad_norm": 0.6046174168586731, "learning_rate": 1.076e-05, "loss": 1.0146, "step": 270 }, { "grad_norm": 0.8675687909126282, "learning_rate": 1.1160000000000002e-05, "loss": 0.9965, "step": 280 }, { "grad_norm": 1.0801470279693604, "learning_rate": 1.156e-05, "loss": 0.9631, "step": 290 }, { "grad_norm": 1.6936357021331787, "learning_rate": 1.196e-05, "loss": 0.9276, "step": 300 }, { "grad_norm": 1.015356421470642, "learning_rate": 1.236e-05, "loss": 0.9121, "step": 310 }, { "grad_norm": 1.3639075756072998, "learning_rate": 1.276e-05, "loss": 0.882, "step": 320 }, { "grad_norm": 1.4270645380020142, "learning_rate": 1.316e-05, "loss": 0.8411, "step": 330 }, { "grad_norm": 1.2292282581329346, "learning_rate": 1.356e-05, "loss": 0.8054, "step": 340 }, { "grad_norm": 1.5038810968399048, "learning_rate": 1.396e-05, "loss": 0.7666, "step": 350 }, { "grad_norm": 1.6285403966903687, "learning_rate": 1.4360000000000001e-05, "loss": 0.7208, "step": 360 }, { "grad_norm": 1.6673287153244019, "learning_rate": 1.4760000000000001e-05, "loss": 0.6717, "step": 370 }, { "grad_norm": 2.0559983253479004, "learning_rate": 1.5160000000000002e-05, "loss": 0.6369, "step": 380 }, { "grad_norm": 1.7887955904006958, "learning_rate": 1.556e-05, "loss": 0.6178, "step": 390 }, { "grad_norm": 1.8226420879364014, "learning_rate": 1.596e-05, "loss": 0.5882, "step": 400 }, { "grad_norm": 1.6779323816299438, "learning_rate": 1.636e-05, "loss": 0.5664, "step": 410 }, { "grad_norm": 1.654363751411438, "learning_rate": 1.6760000000000002e-05, "loss": 0.5242, "step": 420 }, { "grad_norm": 1.8008185625076294, "learning_rate": 1.7160000000000002e-05, "loss": 0.503, "step": 430 }, { "grad_norm": 2.010864496231079, "learning_rate": 1.756e-05, "loss": 0.4767, "step": 440 }, { "grad_norm": 1.4857263565063477, "learning_rate": 1.796e-05, "loss": 0.461, "step": 450 }, { "grad_norm": 1.8379981517791748, "learning_rate": 1.8360000000000004e-05, "loss": 0.4264, "step": 460 }, { "grad_norm": 2.315039873123169, "learning_rate": 1.876e-05, "loss": 0.4101, "step": 470 }, { "grad_norm": 2.186854362487793, "learning_rate": 1.916e-05, "loss": 0.3689, "step": 480 }, { "grad_norm": 2.3028440475463867, "learning_rate": 1.956e-05, "loss": 0.3323, "step": 490 }, { "grad_norm": 1.9360477924346924, "learning_rate": 1.9960000000000002e-05, "loss": 0.3115, "step": 500 }, { "grad_norm": 1.9707876443862915, "learning_rate": 2.036e-05, "loss": 0.2932, "step": 510 }, { "grad_norm": 1.7735759019851685, "learning_rate": 2.076e-05, "loss": 0.283, "step": 520 }, { "grad_norm": 1.597814917564392, "learning_rate": 2.116e-05, "loss": 0.273, "step": 530 }, { "grad_norm": 1.711505651473999, "learning_rate": 2.1560000000000004e-05, "loss": 0.2731, "step": 540 }, { "grad_norm": 1.8393617868423462, "learning_rate": 2.196e-05, "loss": 0.25, "step": 550 }, { "grad_norm": 1.7389594316482544, "learning_rate": 2.236e-05, "loss": 0.2398, "step": 560 }, { "grad_norm": 1.752402663230896, "learning_rate": 2.2760000000000002e-05, "loss": 0.2307, "step": 570 }, { "grad_norm": 2.040132522583008, "learning_rate": 2.3160000000000002e-05, "loss": 0.2114, "step": 580 }, { "grad_norm": 1.9711309671401978, "learning_rate": 2.356e-05, "loss": 0.1986, "step": 590 }, { "grad_norm": 2.056579113006592, "learning_rate": 2.396e-05, "loss": 0.1997, "step": 600 }, { "grad_norm": 1.6812186241149902, "learning_rate": 2.4360000000000004e-05, "loss": 0.1862, "step": 610 }, { "grad_norm": 1.5239371061325073, "learning_rate": 2.476e-05, "loss": 0.1885, "step": 620 }, { "grad_norm": 1.9391425848007202, "learning_rate": 2.516e-05, "loss": 0.176, "step": 630 }, { "grad_norm": 2.1032874584198, "learning_rate": 2.556e-05, "loss": 0.1607, "step": 640 }, { "grad_norm": 1.9331555366516113, "learning_rate": 2.5960000000000002e-05, "loss": 0.1512, "step": 650 }, { "grad_norm": 1.6696467399597168, "learning_rate": 2.6360000000000002e-05, "loss": 0.1471, "step": 660 }, { "grad_norm": 1.8185946941375732, "learning_rate": 2.676e-05, "loss": 0.1313, "step": 670 }, { "grad_norm": 1.7012594938278198, "learning_rate": 2.716e-05, "loss": 0.1212, "step": 680 }, { "grad_norm": 1.7694181203842163, "learning_rate": 2.7560000000000004e-05, "loss": 0.1189, "step": 690 }, { "grad_norm": 1.5570451021194458, "learning_rate": 2.7960000000000003e-05, "loss": 0.1159, "step": 700 }, { "grad_norm": 2.040879011154175, "learning_rate": 2.8360000000000003e-05, "loss": 0.1267, "step": 710 }, { "grad_norm": 2.0064566135406494, "learning_rate": 2.8760000000000002e-05, "loss": 0.1156, "step": 720 }, { "grad_norm": 1.5553596019744873, "learning_rate": 2.9160000000000005e-05, "loss": 0.0993, "step": 730 }, { "grad_norm": 1.4947905540466309, "learning_rate": 2.9559999999999998e-05, "loss": 0.1109, "step": 740 }, { "grad_norm": 2.195685386657715, "learning_rate": 2.9959999999999998e-05, "loss": 0.1163, "step": 750 }, { "grad_norm": 1.4413272142410278, "learning_rate": 3.036e-05, "loss": 0.1173, "step": 760 }, { "grad_norm": 1.55973482131958, "learning_rate": 3.076e-05, "loss": 0.1087, "step": 770 }, { "grad_norm": 1.8807109594345093, "learning_rate": 3.116e-05, "loss": 0.1133, "step": 780 }, { "grad_norm": 1.6292966604232788, "learning_rate": 3.156e-05, "loss": 0.1103, "step": 790 }, { "grad_norm": 1.3404572010040283, "learning_rate": 3.196e-05, "loss": 0.1092, "step": 800 }, { "grad_norm": 1.244386076927185, "learning_rate": 3.236e-05, "loss": 0.1037, "step": 810 }, { "grad_norm": 1.2594668865203857, "learning_rate": 3.2760000000000005e-05, "loss": 0.0993, "step": 820 }, { "grad_norm": 1.5505069494247437, "learning_rate": 3.316e-05, "loss": 0.1032, "step": 830 }, { "grad_norm": 1.2635992765426636, "learning_rate": 3.3560000000000004e-05, "loss": 0.0968, "step": 840 }, { "grad_norm": 1.6217236518859863, "learning_rate": 3.396e-05, "loss": 0.1031, "step": 850 }, { "grad_norm": 2.1571388244628906, "learning_rate": 3.436e-05, "loss": 0.1082, "step": 860 }, { "grad_norm": 1.3784488439559937, "learning_rate": 3.4760000000000006e-05, "loss": 0.0983, "step": 870 }, { "grad_norm": 1.2693129777908325, "learning_rate": 3.516e-05, "loss": 0.1013, "step": 880 }, { "grad_norm": 1.463412880897522, "learning_rate": 3.5560000000000005e-05, "loss": 0.0936, "step": 890 }, { "grad_norm": 1.6243302822113037, "learning_rate": 3.596e-05, "loss": 0.0878, "step": 900 }, { "grad_norm": 1.4497721195220947, "learning_rate": 3.636e-05, "loss": 0.0871, "step": 910 }, { "grad_norm": 0.9397141933441162, "learning_rate": 3.676e-05, "loss": 0.0863, "step": 920 }, { "grad_norm": 1.5587925910949707, "learning_rate": 3.716e-05, "loss": 0.0889, "step": 930 }, { "grad_norm": 1.3005576133728027, "learning_rate": 3.756e-05, "loss": 0.084, "step": 940 }, { "grad_norm": 1.2193269729614258, "learning_rate": 3.796e-05, "loss": 0.0909, "step": 950 }, { "grad_norm": 0.9646158814430237, "learning_rate": 3.836e-05, "loss": 0.0895, "step": 960 }, { "grad_norm": 1.4135878086090088, "learning_rate": 3.876e-05, "loss": 0.0896, "step": 970 }, { "grad_norm": 1.3432016372680664, "learning_rate": 3.9160000000000005e-05, "loss": 0.0922, "step": 980 }, { "grad_norm": 1.2035871744155884, "learning_rate": 3.956e-05, "loss": 0.0923, "step": 990 }, { "grad_norm": 1.2520123720169067, "learning_rate": 3.9960000000000004e-05, "loss": 0.0958, "step": 1000 }, { "grad_norm": 1.2242250442504883, "learning_rate": 4.0360000000000007e-05, "loss": 0.0867, "step": 1010 }, { "grad_norm": 1.252663016319275, "learning_rate": 4.076e-05, "loss": 0.0909, "step": 1020 }, { "grad_norm": 1.4396473169326782, "learning_rate": 4.1160000000000006e-05, "loss": 0.0929, "step": 1030 }, { "grad_norm": 1.2132742404937744, "learning_rate": 4.156e-05, "loss": 0.0862, "step": 1040 }, { "grad_norm": 1.1309123039245605, "learning_rate": 4.196e-05, "loss": 0.0796, "step": 1050 }, { "grad_norm": 0.9484611749649048, "learning_rate": 4.236e-05, "loss": 0.0781, "step": 1060 }, { "grad_norm": 0.9082660675048828, "learning_rate": 4.276e-05, "loss": 0.0806, "step": 1070 }, { "grad_norm": 1.2793047428131104, "learning_rate": 4.316e-05, "loss": 0.076, "step": 1080 }, { "grad_norm": 1.2996289730072021, "learning_rate": 4.356e-05, "loss": 0.0812, "step": 1090 }, { "grad_norm": 1.446408987045288, "learning_rate": 4.396e-05, "loss": 0.0863, "step": 1100 }, { "grad_norm": 1.1310473680496216, "learning_rate": 4.436e-05, "loss": 0.0732, "step": 1110 }, { "grad_norm": 1.1836631298065186, "learning_rate": 4.4760000000000005e-05, "loss": 0.0839, "step": 1120 }, { "grad_norm": 1.1174145936965942, "learning_rate": 4.516e-05, "loss": 0.0831, "step": 1130 }, { "grad_norm": 1.3114209175109863, "learning_rate": 4.5560000000000004e-05, "loss": 0.0809, "step": 1140 }, { "grad_norm": 1.1305618286132812, "learning_rate": 4.596e-05, "loss": 0.0873, "step": 1150 }, { "grad_norm": 1.0533171892166138, "learning_rate": 4.636e-05, "loss": 0.0781, "step": 1160 }, { "grad_norm": 1.1742881536483765, "learning_rate": 4.6760000000000006e-05, "loss": 0.0822, "step": 1170 }, { "grad_norm": 1.1944715976715088, "learning_rate": 4.716e-05, "loss": 0.0839, "step": 1180 }, { "grad_norm": 1.0838624238967896, "learning_rate": 4.7560000000000005e-05, "loss": 0.0768, "step": 1190 }, { "grad_norm": 1.0316380262374878, "learning_rate": 4.796e-05, "loss": 0.0823, "step": 1200 }, { "grad_norm": 0.9976275563240051, "learning_rate": 4.836e-05, "loss": 0.0751, "step": 1210 }, { "grad_norm": 0.7807214260101318, "learning_rate": 4.876e-05, "loss": 0.0668, "step": 1220 }, { "grad_norm": 1.175624132156372, "learning_rate": 4.9160000000000004e-05, "loss": 0.0707, "step": 1230 }, { "grad_norm": 1.2602707147598267, "learning_rate": 4.956e-05, "loss": 0.0791, "step": 1240 }, { "grad_norm": 1.0604500770568848, "learning_rate": 4.996e-05, "loss": 0.0773, "step": 1250 }, { "grad_norm": 0.9894817471504211, "learning_rate": 5.0360000000000006e-05, "loss": 0.0628, "step": 1260 }, { "grad_norm": 1.2252613306045532, "learning_rate": 5.076000000000001e-05, "loss": 0.0809, "step": 1270 }, { "grad_norm": 1.3312374353408813, "learning_rate": 5.1160000000000005e-05, "loss": 0.0725, "step": 1280 }, { "grad_norm": 1.1644082069396973, "learning_rate": 5.1559999999999994e-05, "loss": 0.0793, "step": 1290 }, { "grad_norm": 1.075032353401184, "learning_rate": 5.196e-05, "loss": 0.0771, "step": 1300 }, { "grad_norm": 0.9398416876792908, "learning_rate": 5.236e-05, "loss": 0.0659, "step": 1310 }, { "grad_norm": 0.9868608117103577, "learning_rate": 5.2759999999999996e-05, "loss": 0.082, "step": 1320 }, { "grad_norm": 0.8943616151809692, "learning_rate": 5.316e-05, "loss": 0.0716, "step": 1330 }, { "grad_norm": 0.9181155562400818, "learning_rate": 5.356e-05, "loss": 0.0651, "step": 1340 }, { "grad_norm": 0.9491348266601562, "learning_rate": 5.396e-05, "loss": 0.0745, "step": 1350 }, { "grad_norm": 1.0354893207550049, "learning_rate": 5.436e-05, "loss": 0.0764, "step": 1360 }, { "grad_norm": 0.8962678909301758, "learning_rate": 5.476e-05, "loss": 0.0692, "step": 1370 }, { "grad_norm": 1.0790551900863647, "learning_rate": 5.516e-05, "loss": 0.0664, "step": 1380 }, { "grad_norm": 0.828025758266449, "learning_rate": 5.556e-05, "loss": 0.0625, "step": 1390 }, { "grad_norm": 0.9519407749176025, "learning_rate": 5.596e-05, "loss": 0.0623, "step": 1400 }, { "grad_norm": 0.8859499096870422, "learning_rate": 5.636e-05, "loss": 0.0651, "step": 1410 }, { "grad_norm": 0.8652734160423279, "learning_rate": 5.6760000000000005e-05, "loss": 0.0629, "step": 1420 }, { "grad_norm": 1.0568864345550537, "learning_rate": 5.716e-05, "loss": 0.0775, "step": 1430 }, { "grad_norm": 0.8096494078636169, "learning_rate": 5.7560000000000005e-05, "loss": 0.0649, "step": 1440 }, { "grad_norm": 0.9572166800498962, "learning_rate": 5.796e-05, "loss": 0.0637, "step": 1450 }, { "grad_norm": 0.8617724180221558, "learning_rate": 5.8360000000000004e-05, "loss": 0.0683, "step": 1460 }, { "grad_norm": 0.9124924540519714, "learning_rate": 5.876000000000001e-05, "loss": 0.0627, "step": 1470 }, { "grad_norm": 0.869451642036438, "learning_rate": 5.916e-05, "loss": 0.0624, "step": 1480 }, { "grad_norm": 0.8673956394195557, "learning_rate": 5.9560000000000006e-05, "loss": 0.0671, "step": 1490 }, { "grad_norm": 0.6734641790390015, "learning_rate": 5.996e-05, "loss": 0.0631, "step": 1500 }, { "grad_norm": 0.7621887922286987, "learning_rate": 6.0360000000000005e-05, "loss": 0.0663, "step": 1510 }, { "grad_norm": 0.8125986456871033, "learning_rate": 6.076000000000001e-05, "loss": 0.0655, "step": 1520 }, { "grad_norm": 0.7236118316650391, "learning_rate": 6.116e-05, "loss": 0.0629, "step": 1530 }, { "grad_norm": 0.6856288313865662, "learning_rate": 6.156e-05, "loss": 0.0679, "step": 1540 }, { "grad_norm": 0.9942567348480225, "learning_rate": 6.196000000000001e-05, "loss": 0.0637, "step": 1550 }, { "grad_norm": 0.9096291661262512, "learning_rate": 6.236e-05, "loss": 0.064, "step": 1560 }, { "grad_norm": 0.9005893468856812, "learning_rate": 6.276e-05, "loss": 0.0636, "step": 1570 }, { "grad_norm": 0.9303935766220093, "learning_rate": 6.316000000000001e-05, "loss": 0.0614, "step": 1580 }, { "grad_norm": 0.6894645094871521, "learning_rate": 6.356000000000001e-05, "loss": 0.0621, "step": 1590 }, { "grad_norm": 0.6815313696861267, "learning_rate": 6.396e-05, "loss": 0.0671, "step": 1600 }, { "grad_norm": 0.6604280471801758, "learning_rate": 6.436e-05, "loss": 0.0585, "step": 1610 }, { "grad_norm": 0.7586976885795593, "learning_rate": 6.476e-05, "loss": 0.0557, "step": 1620 }, { "grad_norm": 0.9669461846351624, "learning_rate": 6.515999999999999e-05, "loss": 0.0622, "step": 1630 }, { "grad_norm": 0.8043057918548584, "learning_rate": 6.556e-05, "loss": 0.0646, "step": 1640 }, { "grad_norm": 0.7622464299201965, "learning_rate": 6.596e-05, "loss": 0.0591, "step": 1650 }, { "grad_norm": 0.752494215965271, "learning_rate": 6.636e-05, "loss": 0.0575, "step": 1660 }, { "grad_norm": 0.8137586712837219, "learning_rate": 6.676e-05, "loss": 0.0612, "step": 1670 }, { "grad_norm": 0.7831684947013855, "learning_rate": 6.716e-05, "loss": 0.0572, "step": 1680 }, { "grad_norm": 0.7652895450592041, "learning_rate": 6.756e-05, "loss": 0.0575, "step": 1690 }, { "grad_norm": 0.8497827053070068, "learning_rate": 6.796e-05, "loss": 0.0681, "step": 1700 }, { "grad_norm": 0.6377294659614563, "learning_rate": 6.836e-05, "loss": 0.0589, "step": 1710 }, { "grad_norm": 0.891244113445282, "learning_rate": 6.876e-05, "loss": 0.0639, "step": 1720 }, { "grad_norm": 0.7680009603500366, "learning_rate": 6.916000000000001e-05, "loss": 0.0616, "step": 1730 }, { "grad_norm": 0.7107235193252563, "learning_rate": 6.956e-05, "loss": 0.0546, "step": 1740 }, { "grad_norm": 0.6824011206626892, "learning_rate": 6.996e-05, "loss": 0.0577, "step": 1750 }, { "grad_norm": 0.9730836153030396, "learning_rate": 7.036e-05, "loss": 0.0534, "step": 1760 }, { "grad_norm": 0.9074521660804749, "learning_rate": 7.076000000000001e-05, "loss": 0.0671, "step": 1770 }, { "grad_norm": 0.644514799118042, "learning_rate": 7.116e-05, "loss": 0.0607, "step": 1780 }, { "grad_norm": 0.9121113419532776, "learning_rate": 7.156e-05, "loss": 0.0589, "step": 1790 }, { "grad_norm": 0.8029610514640808, "learning_rate": 7.196000000000001e-05, "loss": 0.0664, "step": 1800 }, { "grad_norm": 0.8757071495056152, "learning_rate": 7.236e-05, "loss": 0.0579, "step": 1810 }, { "grad_norm": 0.8301620483398438, "learning_rate": 7.276e-05, "loss": 0.0607, "step": 1820 }, { "grad_norm": 0.6021782755851746, "learning_rate": 7.316000000000001e-05, "loss": 0.0562, "step": 1830 }, { "grad_norm": 0.9442379474639893, "learning_rate": 7.356000000000001e-05, "loss": 0.0531, "step": 1840 }, { "grad_norm": 0.744971513748169, "learning_rate": 7.396e-05, "loss": 0.0584, "step": 1850 }, { "grad_norm": 0.7188789248466492, "learning_rate": 7.436000000000001e-05, "loss": 0.0561, "step": 1860 }, { "grad_norm": 0.6340389847755432, "learning_rate": 7.476000000000001e-05, "loss": 0.0566, "step": 1870 }, { "grad_norm": 0.7155025005340576, "learning_rate": 7.516e-05, "loss": 0.0563, "step": 1880 }, { "grad_norm": 0.662661612033844, "learning_rate": 7.556000000000002e-05, "loss": 0.0514, "step": 1890 }, { "grad_norm": 0.6533991694450378, "learning_rate": 7.596000000000001e-05, "loss": 0.0567, "step": 1900 }, { "grad_norm": 0.7804444432258606, "learning_rate": 7.636e-05, "loss": 0.0544, "step": 1910 }, { "grad_norm": 0.9438216090202332, "learning_rate": 7.676e-05, "loss": 0.0513, "step": 1920 }, { "grad_norm": 0.7270596623420715, "learning_rate": 7.716e-05, "loss": 0.0557, "step": 1930 }, { "grad_norm": 0.6200523972511292, "learning_rate": 7.756e-05, "loss": 0.0658, "step": 1940 }, { "grad_norm": 0.5124297738075256, "learning_rate": 7.796e-05, "loss": 0.0529, "step": 1950 }, { "grad_norm": 0.5669113993644714, "learning_rate": 7.836e-05, "loss": 0.0507, "step": 1960 }, { "grad_norm": 0.7467186450958252, "learning_rate": 7.876e-05, "loss": 0.054, "step": 1970 }, { "grad_norm": 0.6200578212738037, "learning_rate": 7.916e-05, "loss": 0.0599, "step": 1980 }, { "grad_norm": 0.691246509552002, "learning_rate": 7.956e-05, "loss": 0.0527, "step": 1990 }, { "grad_norm": 0.7206867337226868, "learning_rate": 7.996e-05, "loss": 0.0546, "step": 2000 }, { "grad_norm": 0.6038048267364502, "learning_rate": 8.036e-05, "loss": 0.0548, "step": 2010 }, { "grad_norm": 0.5746732950210571, "learning_rate": 8.076e-05, "loss": 0.0567, "step": 2020 }, { "grad_norm": 0.6559705138206482, "learning_rate": 8.116e-05, "loss": 0.0509, "step": 2030 }, { "grad_norm": 0.6057372689247131, "learning_rate": 8.156e-05, "loss": 0.0561, "step": 2040 }, { "grad_norm": 0.7387638092041016, "learning_rate": 8.196000000000001e-05, "loss": 0.0564, "step": 2050 }, { "grad_norm": 0.7291343808174133, "learning_rate": 8.236e-05, "loss": 0.0547, "step": 2060 }, { "grad_norm": 0.6363125443458557, "learning_rate": 8.276e-05, "loss": 0.0552, "step": 2070 }, { "grad_norm": 0.6508821845054626, "learning_rate": 8.316000000000001e-05, "loss": 0.0573, "step": 2080 }, { "grad_norm": 0.6537685990333557, "learning_rate": 8.356e-05, "loss": 0.0551, "step": 2090 }, { "grad_norm": 0.6614780426025391, "learning_rate": 8.396e-05, "loss": 0.0528, "step": 2100 }, { "grad_norm": 0.6560401320457458, "learning_rate": 8.436000000000001e-05, "loss": 0.0545, "step": 2110 }, { "grad_norm": 0.5808461904525757, "learning_rate": 8.476000000000001e-05, "loss": 0.0533, "step": 2120 }, { "grad_norm": 0.5264371037483215, "learning_rate": 8.516e-05, "loss": 0.0582, "step": 2130 }, { "grad_norm": 0.5954933762550354, "learning_rate": 8.556e-05, "loss": 0.0523, "step": 2140 }, { "grad_norm": 0.7801850438117981, "learning_rate": 8.596000000000001e-05, "loss": 0.0537, "step": 2150 }, { "grad_norm": 0.5850076079368591, "learning_rate": 8.636e-05, "loss": 0.0454, "step": 2160 }, { "grad_norm": 0.5922155380249023, "learning_rate": 8.676e-05, "loss": 0.0509, "step": 2170 }, { "grad_norm": 0.8190131187438965, "learning_rate": 8.716000000000001e-05, "loss": 0.0542, "step": 2180 }, { "grad_norm": 0.587446928024292, "learning_rate": 8.756000000000001e-05, "loss": 0.0553, "step": 2190 }, { "grad_norm": 0.6227629780769348, "learning_rate": 8.796e-05, "loss": 0.0487, "step": 2200 }, { "grad_norm": 0.6692546010017395, "learning_rate": 8.836000000000001e-05, "loss": 0.0476, "step": 2210 }, { "grad_norm": 0.7408078908920288, "learning_rate": 8.876e-05, "loss": 0.0489, "step": 2220 }, { "grad_norm": 0.586682140827179, "learning_rate": 8.916e-05, "loss": 0.0509, "step": 2230 }, { "grad_norm": 0.6249874234199524, "learning_rate": 8.956e-05, "loss": 0.0526, "step": 2240 }, { "grad_norm": 0.7161738872528076, "learning_rate": 8.996e-05, "loss": 0.0596, "step": 2250 }, { "grad_norm": 0.6018122434616089, "learning_rate": 9.036e-05, "loss": 0.0584, "step": 2260 }, { "grad_norm": 0.4742252826690674, "learning_rate": 9.076e-05, "loss": 0.0574, "step": 2270 }, { "grad_norm": 0.4201463758945465, "learning_rate": 9.116e-05, "loss": 0.0496, "step": 2280 }, { "grad_norm": 0.4542571008205414, "learning_rate": 9.156e-05, "loss": 0.0446, "step": 2290 }, { "grad_norm": 0.5460566282272339, "learning_rate": 9.196000000000001e-05, "loss": 0.0512, "step": 2300 }, { "grad_norm": 0.38791629672050476, "learning_rate": 9.236e-05, "loss": 0.052, "step": 2310 }, { "grad_norm": 0.5282217264175415, "learning_rate": 9.276e-05, "loss": 0.0476, "step": 2320 }, { "grad_norm": 0.5786757469177246, "learning_rate": 9.316000000000001e-05, "loss": 0.054, "step": 2330 }, { "grad_norm": 0.5746905207633972, "learning_rate": 9.356e-05, "loss": 0.0531, "step": 2340 }, { "grad_norm": 0.6251006722450256, "learning_rate": 9.396e-05, "loss": 0.0492, "step": 2350 }, { "grad_norm": 0.7246794700622559, "learning_rate": 9.436e-05, "loss": 0.0553, "step": 2360 }, { "grad_norm": 0.5014413595199585, "learning_rate": 9.476000000000001e-05, "loss": 0.0527, "step": 2370 }, { "grad_norm": 0.5953740477561951, "learning_rate": 9.516e-05, "loss": 0.0535, "step": 2380 }, { "grad_norm": 0.5419349074363708, "learning_rate": 9.556e-05, "loss": 0.0508, "step": 2390 }, { "grad_norm": 0.5652883052825928, "learning_rate": 9.596000000000001e-05, "loss": 0.0514, "step": 2400 }, { "grad_norm": 0.4460015594959259, "learning_rate": 9.636e-05, "loss": 0.0481, "step": 2410 }, { "grad_norm": 0.6789939999580383, "learning_rate": 9.676e-05, "loss": 0.0491, "step": 2420 }, { "grad_norm": 0.5071524381637573, "learning_rate": 9.716000000000001e-05, "loss": 0.0491, "step": 2430 }, { "grad_norm": 0.6390159726142883, "learning_rate": 9.756000000000001e-05, "loss": 0.0485, "step": 2440 }, { "grad_norm": 0.6449014544487, "learning_rate": 9.796e-05, "loss": 0.05, "step": 2450 }, { "grad_norm": 0.506130576133728, "learning_rate": 9.836000000000001e-05, "loss": 0.0529, "step": 2460 }, { "grad_norm": 0.5097265243530273, "learning_rate": 9.876000000000001e-05, "loss": 0.0463, "step": 2470 }, { "grad_norm": 0.4680483937263489, "learning_rate": 9.916e-05, "loss": 0.0554, "step": 2480 }, { "grad_norm": 0.7338577508926392, "learning_rate": 9.956e-05, "loss": 0.0483, "step": 2490 }, { "grad_norm": 0.46772661805152893, "learning_rate": 9.996000000000001e-05, "loss": 0.0513, "step": 2500 }, { "grad_norm": 0.5334926843643188, "learning_rate": 9.999999114196196e-05, "loss": 0.0502, "step": 2510 }, { "grad_norm": 0.5277791619300842, "learning_rate": 9.99999605215876e-05, "loss": 0.0487, "step": 2520 }, { "grad_norm": 0.683641254901886, "learning_rate": 9.999990802953179e-05, "loss": 0.0472, "step": 2530 }, { "grad_norm": 0.5659726858139038, "learning_rate": 9.99998336658175e-05, "loss": 0.0479, "step": 2540 }, { "grad_norm": 0.6926059126853943, "learning_rate": 9.999973743047727e-05, "loss": 0.0563, "step": 2550 }, { "grad_norm": 0.4505929946899414, "learning_rate": 9.999961932355319e-05, "loss": 0.0498, "step": 2560 }, { "grad_norm": 0.6377073526382446, "learning_rate": 9.999947934509693e-05, "loss": 0.0481, "step": 2570 }, { "grad_norm": 0.4898621737957001, "learning_rate": 9.999931749516971e-05, "loss": 0.052, "step": 2580 }, { "grad_norm": 0.5683681964874268, "learning_rate": 9.999913377384233e-05, "loss": 0.0467, "step": 2590 }, { "grad_norm": 0.5124620795249939, "learning_rate": 9.999892818119517e-05, "loss": 0.0499, "step": 2600 }, { "grad_norm": 0.645530641078949, "learning_rate": 9.999870071731814e-05, "loss": 0.0497, "step": 2610 }, { "grad_norm": 0.6242179870605469, "learning_rate": 9.999845138231076e-05, "loss": 0.0477, "step": 2620 }, { "grad_norm": 0.5192215442657471, "learning_rate": 9.999818017628208e-05, "loss": 0.0539, "step": 2630 }, { "grad_norm": 0.4066641628742218, "learning_rate": 9.999788709935078e-05, "loss": 0.049, "step": 2640 }, { "grad_norm": 0.4780401587486267, "learning_rate": 9.9997572151645e-05, "loss": 0.049, "step": 2650 }, { "grad_norm": 0.460101842880249, "learning_rate": 9.999723533330254e-05, "loss": 0.0483, "step": 2660 }, { "grad_norm": 0.5579531192779541, "learning_rate": 9.999687664447074e-05, "loss": 0.0505, "step": 2670 }, { "grad_norm": 0.45036232471466064, "learning_rate": 9.99964960853065e-05, "loss": 0.0462, "step": 2680 }, { "grad_norm": 0.6312862634658813, "learning_rate": 9.999609365597627e-05, "loss": 0.0491, "step": 2690 }, { "grad_norm": 0.6391765475273132, "learning_rate": 9.99956693566561e-05, "loss": 0.0495, "step": 2700 }, { "grad_norm": 0.5391504168510437, "learning_rate": 9.99952231875316e-05, "loss": 0.0532, "step": 2710 }, { "grad_norm": 0.39547762274742126, "learning_rate": 9.999475514879795e-05, "loss": 0.0457, "step": 2720 }, { "grad_norm": 0.5656508207321167, "learning_rate": 9.999426524065984e-05, "loss": 0.0486, "step": 2730 }, { "grad_norm": 0.45238953828811646, "learning_rate": 9.999375346333162e-05, "loss": 0.0493, "step": 2740 }, { "grad_norm": 0.586006760597229, "learning_rate": 9.999321981703715e-05, "loss": 0.0488, "step": 2750 }, { "grad_norm": 0.5166579484939575, "learning_rate": 9.999266430200985e-05, "loss": 0.0494, "step": 2760 }, { "grad_norm": 0.6533814072608948, "learning_rate": 9.999208691849271e-05, "loss": 0.0418, "step": 2770 }, { "grad_norm": 0.4532153904438019, "learning_rate": 9.999148766673832e-05, "loss": 0.0484, "step": 2780 }, { "grad_norm": 0.43949344754219055, "learning_rate": 9.999086654700881e-05, "loss": 0.0478, "step": 2790 }, { "grad_norm": 0.6996206045150757, "learning_rate": 9.999022355957588e-05, "loss": 0.0445, "step": 2800 }, { "grad_norm": 0.31124061346054077, "learning_rate": 9.998955870472079e-05, "loss": 0.0429, "step": 2810 }, { "grad_norm": 0.3255634605884552, "learning_rate": 9.998887198273437e-05, "loss": 0.0405, "step": 2820 }, { "grad_norm": 0.48686057329177856, "learning_rate": 9.998816339391701e-05, "loss": 0.0446, "step": 2830 }, { "grad_norm": 0.4439123272895813, "learning_rate": 9.998743293857868e-05, "loss": 0.0492, "step": 2840 }, { "grad_norm": 0.448848158121109, "learning_rate": 9.998668061703891e-05, "loss": 0.0417, "step": 2850 }, { "grad_norm": 0.5039438009262085, "learning_rate": 9.998590642962679e-05, "loss": 0.0414, "step": 2860 }, { "grad_norm": 0.46082308888435364, "learning_rate": 9.998511037668095e-05, "loss": 0.0431, "step": 2870 }, { "grad_norm": 0.38872405886650085, "learning_rate": 9.998429245854964e-05, "loss": 0.0422, "step": 2880 }, { "grad_norm": 0.4246666729450226, "learning_rate": 9.998345267559064e-05, "loss": 0.041, "step": 2890 }, { "grad_norm": 0.6673298478126526, "learning_rate": 9.998259102817129e-05, "loss": 0.0443, "step": 2900 }, { "grad_norm": 0.5126121044158936, "learning_rate": 9.99817075166685e-05, "loss": 0.0425, "step": 2910 }, { "grad_norm": 0.48576462268829346, "learning_rate": 9.998080214146878e-05, "loss": 0.0424, "step": 2920 }, { "grad_norm": 0.5307922959327698, "learning_rate": 9.997987490296813e-05, "loss": 0.0476, "step": 2930 }, { "grad_norm": 0.5131188631057739, "learning_rate": 9.99789258015722e-05, "loss": 0.0489, "step": 2940 }, { "grad_norm": 0.4211786687374115, "learning_rate": 9.997795483769611e-05, "loss": 0.0458, "step": 2950 }, { "grad_norm": 0.43335118889808655, "learning_rate": 9.997696201176462e-05, "loss": 0.0506, "step": 2960 }, { "grad_norm": 0.41095250844955444, "learning_rate": 9.997594732421203e-05, "loss": 0.0437, "step": 2970 }, { "grad_norm": 0.5414423942565918, "learning_rate": 9.997491077548217e-05, "loss": 0.0426, "step": 2980 }, { "grad_norm": 0.36229678988456726, "learning_rate": 9.997385236602851e-05, "loss": 0.0419, "step": 2990 }, { "grad_norm": 0.34936150908470154, "learning_rate": 9.997277209631399e-05, "loss": 0.0427, "step": 3000 }, { "grad_norm": 0.6067139506340027, "learning_rate": 9.997166996681118e-05, "loss": 0.0499, "step": 3010 }, { "grad_norm": 0.49866238236427307, "learning_rate": 9.997054597800218e-05, "loss": 0.0442, "step": 3020 }, { "grad_norm": 0.3345346748828888, "learning_rate": 9.996940013037866e-05, "loss": 0.0411, "step": 3030 }, { "grad_norm": 0.45981162786483765, "learning_rate": 9.996823242444186e-05, "loss": 0.0396, "step": 3040 }, { "grad_norm": 0.4103314280509949, "learning_rate": 9.996704286070258e-05, "loss": 0.0417, "step": 3050 }, { "grad_norm": 0.5575783848762512, "learning_rate": 9.996583143968115e-05, "loss": 0.0397, "step": 3060 }, { "grad_norm": 0.404458612203598, "learning_rate": 9.99645981619075e-05, "loss": 0.0414, "step": 3070 }, { "grad_norm": 0.36830517649650574, "learning_rate": 9.996334302792114e-05, "loss": 0.042, "step": 3080 }, { "grad_norm": 0.41416144371032715, "learning_rate": 9.996206603827105e-05, "loss": 0.0443, "step": 3090 }, { "grad_norm": 0.42167651653289795, "learning_rate": 9.996076719351587e-05, "loss": 0.0466, "step": 3100 }, { "grad_norm": 0.4996941387653351, "learning_rate": 9.995944649422374e-05, "loss": 0.0483, "step": 3110 }, { "grad_norm": 0.3907119035720825, "learning_rate": 9.995810394097239e-05, "loss": 0.0393, "step": 3120 }, { "grad_norm": 0.540632426738739, "learning_rate": 9.995673953434909e-05, "loss": 0.0495, "step": 3130 }, { "grad_norm": 0.4541913568973541, "learning_rate": 9.995535327495068e-05, "loss": 0.0401, "step": 3140 }, { "grad_norm": 0.5103195309638977, "learning_rate": 9.995394516338355e-05, "loss": 0.0454, "step": 3150 }, { "grad_norm": 0.4949314296245575, "learning_rate": 9.995251520026367e-05, "loss": 0.0367, "step": 3160 }, { "grad_norm": 0.4270646870136261, "learning_rate": 9.995106338621656e-05, "loss": 0.0413, "step": 3170 }, { "grad_norm": 0.3582088351249695, "learning_rate": 9.994958972187726e-05, "loss": 0.0422, "step": 3180 }, { "grad_norm": 0.5443354845046997, "learning_rate": 9.994809420789044e-05, "loss": 0.0464, "step": 3190 }, { "grad_norm": 0.4865322709083557, "learning_rate": 9.994657684491027e-05, "loss": 0.0416, "step": 3200 }, { "grad_norm": 0.44792330265045166, "learning_rate": 9.994503763360048e-05, "loss": 0.0384, "step": 3210 }, { "grad_norm": 0.5200015306472778, "learning_rate": 9.99434765746344e-05, "loss": 0.0448, "step": 3220 }, { "grad_norm": 0.3982260525226593, "learning_rate": 9.994189366869488e-05, "loss": 0.0484, "step": 3230 }, { "grad_norm": 0.35745036602020264, "learning_rate": 9.994028891647433e-05, "loss": 0.0415, "step": 3240 }, { "grad_norm": 0.38287830352783203, "learning_rate": 9.993866231867475e-05, "loss": 0.0425, "step": 3250 }, { "grad_norm": 0.4272738993167877, "learning_rate": 9.993701387600762e-05, "loss": 0.0436, "step": 3260 }, { "grad_norm": 0.560173749923706, "learning_rate": 9.993534358919408e-05, "loss": 0.0432, "step": 3270 }, { "grad_norm": 0.49206626415252686, "learning_rate": 9.993365145896473e-05, "loss": 0.0386, "step": 3280 }, { "grad_norm": 0.40479663014411926, "learning_rate": 9.993193748605977e-05, "loss": 0.0418, "step": 3290 }, { "grad_norm": 0.3908839821815491, "learning_rate": 9.993020167122898e-05, "loss": 0.0448, "step": 3300 }, { "grad_norm": 0.33099737763404846, "learning_rate": 9.992844401523164e-05, "loss": 0.0433, "step": 3310 }, { "grad_norm": 0.39015039801597595, "learning_rate": 9.992666451883661e-05, "loss": 0.0447, "step": 3320 }, { "grad_norm": 0.38739025592803955, "learning_rate": 9.99248631828223e-05, "loss": 0.0422, "step": 3330 }, { "grad_norm": 0.3889550566673279, "learning_rate": 9.99230400079767e-05, "loss": 0.0432, "step": 3340 }, { "grad_norm": 0.5984971523284912, "learning_rate": 9.992119499509728e-05, "loss": 0.0449, "step": 3350 }, { "grad_norm": 0.4535196125507355, "learning_rate": 9.991932814499114e-05, "loss": 0.0396, "step": 3360 }, { "grad_norm": 0.3221626281738281, "learning_rate": 9.991743945847493e-05, "loss": 0.0402, "step": 3370 }, { "grad_norm": 0.523133397102356, "learning_rate": 9.991552893637478e-05, "loss": 0.0394, "step": 3380 }, { "grad_norm": 0.5344473719596863, "learning_rate": 9.991359657952644e-05, "loss": 0.044, "step": 3390 }, { "grad_norm": 0.4005912244319916, "learning_rate": 9.991164238877519e-05, "loss": 0.0448, "step": 3400 }, { "grad_norm": 0.535717248916626, "learning_rate": 9.990966636497585e-05, "loss": 0.0462, "step": 3410 }, { "grad_norm": 0.4947056174278259, "learning_rate": 9.99076685089928e-05, "loss": 0.041, "step": 3420 }, { "grad_norm": 0.5003987550735474, "learning_rate": 9.990564882169998e-05, "loss": 0.0395, "step": 3430 }, { "grad_norm": 0.6668880581855774, "learning_rate": 9.990360730398088e-05, "loss": 0.0447, "step": 3440 }, { "grad_norm": 0.480677992105484, "learning_rate": 9.990154395672849e-05, "loss": 0.0368, "step": 3450 }, { "grad_norm": 0.36002954840660095, "learning_rate": 9.989945878084541e-05, "loss": 0.0424, "step": 3460 }, { "grad_norm": 0.48320573568344116, "learning_rate": 9.989735177724378e-05, "loss": 0.0459, "step": 3470 }, { "grad_norm": 0.3653718829154968, "learning_rate": 9.989522294684526e-05, "loss": 0.0411, "step": 3480 }, { "grad_norm": 0.3104921579360962, "learning_rate": 9.989307229058107e-05, "loss": 0.04, "step": 3490 }, { "grad_norm": 0.4608096182346344, "learning_rate": 9.989089980939202e-05, "loss": 0.0367, "step": 3500 }, { "grad_norm": 0.44114038348197937, "learning_rate": 9.988870550422835e-05, "loss": 0.045, "step": 3510 }, { "grad_norm": 0.3840514123439789, "learning_rate": 9.988648937604999e-05, "loss": 0.0401, "step": 3520 }, { "grad_norm": 0.3829437792301178, "learning_rate": 9.988425142582632e-05, "loss": 0.0412, "step": 3530 }, { "grad_norm": 0.5176998972892761, "learning_rate": 9.98819916545363e-05, "loss": 0.0405, "step": 3540 }, { "grad_norm": 0.38363850116729736, "learning_rate": 9.987971006316844e-05, "loss": 0.0398, "step": 3550 }, { "grad_norm": 0.5735336542129517, "learning_rate": 9.987740665272077e-05, "loss": 0.0412, "step": 3560 }, { "grad_norm": 0.41206949949264526, "learning_rate": 9.98750814242009e-05, "loss": 0.0407, "step": 3570 }, { "grad_norm": 0.45664188265800476, "learning_rate": 9.987273437862594e-05, "loss": 0.0414, "step": 3580 }, { "grad_norm": 0.45491260290145874, "learning_rate": 9.987036551702259e-05, "loss": 0.0394, "step": 3590 }, { "grad_norm": 0.3253578841686249, "learning_rate": 9.986797484042706e-05, "loss": 0.0388, "step": 3600 }, { "grad_norm": 0.42229610681533813, "learning_rate": 9.986556234988512e-05, "loss": 0.0403, "step": 3610 }, { "grad_norm": 0.332193523645401, "learning_rate": 9.986312804645205e-05, "loss": 0.039, "step": 3620 }, { "grad_norm": 0.3959410488605499, "learning_rate": 9.986067193119273e-05, "loss": 0.0423, "step": 3630 }, { "grad_norm": 0.3983510136604309, "learning_rate": 9.985819400518153e-05, "loss": 0.0375, "step": 3640 }, { "grad_norm": 0.4366046190261841, "learning_rate": 9.985569426950239e-05, "loss": 0.0368, "step": 3650 }, { "grad_norm": 0.4879259169101715, "learning_rate": 9.985317272524876e-05, "loss": 0.0358, "step": 3660 }, { "grad_norm": 0.40022969245910645, "learning_rate": 9.985062937352366e-05, "loss": 0.0365, "step": 3670 }, { "grad_norm": 0.33493074774742126, "learning_rate": 9.984806421543966e-05, "loss": 0.0374, "step": 3680 }, { "grad_norm": 0.3884679079055786, "learning_rate": 9.984547725211881e-05, "loss": 0.0442, "step": 3690 }, { "grad_norm": 0.4603472650051117, "learning_rate": 9.984286848469276e-05, "loss": 0.035, "step": 3700 }, { "grad_norm": 0.31187203526496887, "learning_rate": 9.984023791430266e-05, "loss": 0.0363, "step": 3710 }, { "grad_norm": 0.35833173990249634, "learning_rate": 9.983758554209924e-05, "loss": 0.0376, "step": 3720 }, { "grad_norm": 0.3557681441307068, "learning_rate": 9.983491136924268e-05, "loss": 0.0403, "step": 3730 }, { "grad_norm": 0.4058561325073242, "learning_rate": 9.983221539690282e-05, "loss": 0.0409, "step": 3740 }, { "grad_norm": 0.31761378049850464, "learning_rate": 9.982949762625892e-05, "loss": 0.0374, "step": 3750 }, { "grad_norm": 0.4844707250595093, "learning_rate": 9.982675805849986e-05, "loss": 0.0415, "step": 3760 }, { "grad_norm": 0.40428078174591064, "learning_rate": 9.982399669482399e-05, "loss": 0.0322, "step": 3770 }, { "grad_norm": 0.485557496547699, "learning_rate": 9.982121353643924e-05, "loss": 0.0394, "step": 3780 }, { "grad_norm": 0.4449644684791565, "learning_rate": 9.981840858456306e-05, "loss": 0.0372, "step": 3790 }, { "grad_norm": 0.40097731351852417, "learning_rate": 9.981558184042243e-05, "loss": 0.0413, "step": 3800 }, { "grad_norm": 0.42122742533683777, "learning_rate": 9.981273330525387e-05, "loss": 0.0423, "step": 3810 }, { "grad_norm": 0.3714742958545685, "learning_rate": 9.980986298030341e-05, "loss": 0.0464, "step": 3820 }, { "grad_norm": 0.29959240555763245, "learning_rate": 9.980697086682662e-05, "loss": 0.0357, "step": 3830 }, { "grad_norm": 0.3232460916042328, "learning_rate": 9.980405696608866e-05, "loss": 0.0359, "step": 3840 }, { "grad_norm": 0.3844173550605774, "learning_rate": 9.980112127936409e-05, "loss": 0.039, "step": 3850 }, { "grad_norm": 0.3285090923309326, "learning_rate": 9.979816380793717e-05, "loss": 0.0392, "step": 3860 }, { "grad_norm": 0.38894811272621155, "learning_rate": 9.979518455310151e-05, "loss": 0.0336, "step": 3870 }, { "grad_norm": 0.40085750818252563, "learning_rate": 9.97921835161604e-05, "loss": 0.0377, "step": 3880 }, { "grad_norm": 0.4129883348941803, "learning_rate": 9.978916069842656e-05, "loss": 0.0421, "step": 3890 }, { "grad_norm": 0.3532140851020813, "learning_rate": 9.97861161012223e-05, "loss": 0.0428, "step": 3900 }, { "grad_norm": 0.40982306003570557, "learning_rate": 9.978304972587942e-05, "loss": 0.0381, "step": 3910 }, { "grad_norm": 0.4418235421180725, "learning_rate": 9.977996157373925e-05, "loss": 0.0447, "step": 3920 }, { "grad_norm": 0.35203543305397034, "learning_rate": 9.977685164615265e-05, "loss": 0.0343, "step": 3930 }, { "grad_norm": 0.3843795359134674, "learning_rate": 9.977371994448002e-05, "loss": 0.0393, "step": 3940 }, { "grad_norm": 0.35057252645492554, "learning_rate": 9.977056647009127e-05, "loss": 0.0394, "step": 3950 }, { "grad_norm": 0.4007676839828491, "learning_rate": 9.976739122436582e-05, "loss": 0.0374, "step": 3960 }, { "grad_norm": 0.4060984253883362, "learning_rate": 9.976419420869265e-05, "loss": 0.0346, "step": 3970 }, { "grad_norm": 0.2825969457626343, "learning_rate": 9.976097542447025e-05, "loss": 0.0343, "step": 3980 }, { "grad_norm": 0.4013255834579468, "learning_rate": 9.97577348731066e-05, "loss": 0.0362, "step": 3990 }, { "grad_norm": 0.480281263589859, "learning_rate": 9.975447255601927e-05, "loss": 0.0397, "step": 4000 }, { "grad_norm": 0.4256203770637512, "learning_rate": 9.975118847463525e-05, "loss": 0.0354, "step": 4010 }, { "grad_norm": 0.36489900946617126, "learning_rate": 9.974788263039114e-05, "loss": 0.0347, "step": 4020 }, { "grad_norm": 0.3175728917121887, "learning_rate": 9.974455502473303e-05, "loss": 0.0378, "step": 4030 }, { "grad_norm": 0.32769936323165894, "learning_rate": 9.974120565911652e-05, "loss": 0.0345, "step": 4040 }, { "grad_norm": 0.38217347860336304, "learning_rate": 9.973783453500674e-05, "loss": 0.0366, "step": 4050 }, { "grad_norm": 0.4321262538433075, "learning_rate": 9.973444165387835e-05, "loss": 0.0372, "step": 4060 }, { "grad_norm": 0.4731610417366028, "learning_rate": 9.973102701721549e-05, "loss": 0.0313, "step": 4070 }, { "grad_norm": 0.31932616233825684, "learning_rate": 9.972759062651184e-05, "loss": 0.0367, "step": 4080 }, { "grad_norm": 0.3345484733581543, "learning_rate": 9.972413248327059e-05, "loss": 0.0346, "step": 4090 }, { "grad_norm": 0.3954044282436371, "learning_rate": 9.972065258900447e-05, "loss": 0.0327, "step": 4100 }, { "grad_norm": 0.4238516688346863, "learning_rate": 9.971715094523569e-05, "loss": 0.0374, "step": 4110 }, { "grad_norm": 0.371136337518692, "learning_rate": 9.971362755349598e-05, "loss": 0.0378, "step": 4120 }, { "grad_norm": 0.365295946598053, "learning_rate": 9.971008241532662e-05, "loss": 0.0394, "step": 4130 }, { "grad_norm": 0.41665971279144287, "learning_rate": 9.970651553227835e-05, "loss": 0.035, "step": 4140 }, { "grad_norm": 0.3277624845504761, "learning_rate": 9.970292690591143e-05, "loss": 0.0375, "step": 4150 }, { "grad_norm": 0.40566933155059814, "learning_rate": 9.969931653779569e-05, "loss": 0.0345, "step": 4160 }, { "grad_norm": 0.3218092620372772, "learning_rate": 9.969568442951038e-05, "loss": 0.0379, "step": 4170 }, { "grad_norm": 0.38037654757499695, "learning_rate": 9.969203058264436e-05, "loss": 0.0379, "step": 4180 }, { "grad_norm": 0.34302985668182373, "learning_rate": 9.96883549987959e-05, "loss": 0.0402, "step": 4190 }, { "grad_norm": 0.40920278429985046, "learning_rate": 9.968465767957287e-05, "loss": 0.0348, "step": 4200 }, { "grad_norm": 0.36454278230667114, "learning_rate": 9.968093862659256e-05, "loss": 0.0342, "step": 4210 }, { "grad_norm": 0.38657626509666443, "learning_rate": 9.967719784148182e-05, "loss": 0.034, "step": 4220 }, { "grad_norm": 0.29038262367248535, "learning_rate": 9.967343532587702e-05, "loss": 0.0346, "step": 4230 }, { "grad_norm": 0.4502968490123749, "learning_rate": 9.966965108142399e-05, "loss": 0.035, "step": 4240 }, { "grad_norm": 0.4211561381816864, "learning_rate": 9.96658451097781e-05, "loss": 0.036, "step": 4250 }, { "grad_norm": 0.370222806930542, "learning_rate": 9.966201741260419e-05, "loss": 0.0366, "step": 4260 }, { "grad_norm": 0.4462031424045563, "learning_rate": 9.965816799157665e-05, "loss": 0.0376, "step": 4270 }, { "grad_norm": 0.43808743357658386, "learning_rate": 9.965429684837935e-05, "loss": 0.0362, "step": 4280 }, { "grad_norm": 0.3189237415790558, "learning_rate": 9.965040398470562e-05, "loss": 0.0321, "step": 4290 }, { "grad_norm": 0.5428202152252197, "learning_rate": 9.964648940225838e-05, "loss": 0.0324, "step": 4300 }, { "grad_norm": 0.37513720989227295, "learning_rate": 9.964255310274997e-05, "loss": 0.0345, "step": 4310 }, { "grad_norm": 0.37615296244621277, "learning_rate": 9.963859508790228e-05, "loss": 0.0389, "step": 4320 }, { "grad_norm": 0.33625322580337524, "learning_rate": 9.963461535944664e-05, "loss": 0.034, "step": 4330 }, { "grad_norm": 0.3464892506599426, "learning_rate": 9.963061391912399e-05, "loss": 0.0375, "step": 4340 }, { "grad_norm": 0.3498099744319916, "learning_rate": 9.962659076868463e-05, "loss": 0.0401, "step": 4350 }, { "grad_norm": 0.3511611223220825, "learning_rate": 9.962254590988846e-05, "loss": 0.0337, "step": 4360 }, { "grad_norm": 0.37777942419052124, "learning_rate": 9.961847934450481e-05, "loss": 0.0365, "step": 4370 }, { "grad_norm": 0.37865445017814636, "learning_rate": 9.961439107431257e-05, "loss": 0.039, "step": 4380 }, { "grad_norm": 0.32945024967193604, "learning_rate": 9.961028110110006e-05, "loss": 0.0421, "step": 4390 }, { "grad_norm": 0.5437269806861877, "learning_rate": 9.960614942666513e-05, "loss": 0.0356, "step": 4400 }, { "grad_norm": 0.39982321858406067, "learning_rate": 9.960199605281511e-05, "loss": 0.0386, "step": 4410 }, { "grad_norm": 0.3944191634654999, "learning_rate": 9.959782098136683e-05, "loss": 0.0365, "step": 4420 }, { "grad_norm": 0.34251973032951355, "learning_rate": 9.959362421414662e-05, "loss": 0.0338, "step": 4430 }, { "grad_norm": 0.4052746593952179, "learning_rate": 9.958940575299027e-05, "loss": 0.0359, "step": 4440 }, { "grad_norm": 0.35605090856552124, "learning_rate": 9.95851655997431e-05, "loss": 0.0363, "step": 4450 }, { "grad_norm": 0.42771291732788086, "learning_rate": 9.958090375625986e-05, "loss": 0.035, "step": 4460 }, { "grad_norm": 0.4660288989543915, "learning_rate": 9.957662022440486e-05, "loss": 0.0333, "step": 4470 }, { "grad_norm": 0.3814123868942261, "learning_rate": 9.957231500605187e-05, "loss": 0.0353, "step": 4480 }, { "grad_norm": 0.4554194211959839, "learning_rate": 9.95679881030841e-05, "loss": 0.0406, "step": 4490 }, { "grad_norm": 0.3063037395477295, "learning_rate": 9.95636395173943e-05, "loss": 0.0347, "step": 4500 }, { "grad_norm": 0.30606895685195923, "learning_rate": 9.95592692508847e-05, "loss": 0.0318, "step": 4510 }, { "grad_norm": 0.45677411556243896, "learning_rate": 9.9554877305467e-05, "loss": 0.0335, "step": 4520 }, { "grad_norm": 0.38935282826423645, "learning_rate": 9.955046368306237e-05, "loss": 0.0365, "step": 4530 }, { "grad_norm": 0.3629598617553711, "learning_rate": 9.954602838560153e-05, "loss": 0.0355, "step": 4540 }, { "grad_norm": 0.3336105942726135, "learning_rate": 9.954157141502456e-05, "loss": 0.0314, "step": 4550 }, { "grad_norm": 0.35257044434547424, "learning_rate": 9.953709277328112e-05, "loss": 0.0334, "step": 4560 }, { "grad_norm": 0.3599191904067993, "learning_rate": 9.953259246233032e-05, "loss": 0.0329, "step": 4570 }, { "grad_norm": 0.3616459369659424, "learning_rate": 9.952807048414077e-05, "loss": 0.038, "step": 4580 }, { "grad_norm": 0.434116005897522, "learning_rate": 9.95235268406905e-05, "loss": 0.0387, "step": 4590 }, { "grad_norm": 0.38972005248069763, "learning_rate": 9.951896153396708e-05, "loss": 0.0331, "step": 4600 }, { "grad_norm": 0.5667542815208435, "learning_rate": 9.95143745659675e-05, "loss": 0.0381, "step": 4610 }, { "grad_norm": 0.27837735414505005, "learning_rate": 9.95097659386983e-05, "loss": 0.0336, "step": 4620 }, { "grad_norm": 0.4152267277240753, "learning_rate": 9.950513565417542e-05, "loss": 0.0353, "step": 4630 }, { "grad_norm": 0.3352091312408447, "learning_rate": 9.95004837144243e-05, "loss": 0.036, "step": 4640 }, { "grad_norm": 0.29839518666267395, "learning_rate": 9.949581012147988e-05, "loss": 0.0337, "step": 4650 }, { "grad_norm": 0.2854501008987427, "learning_rate": 9.949111487738653e-05, "loss": 0.0349, "step": 4660 }, { "grad_norm": 0.4064194858074188, "learning_rate": 9.948639798419813e-05, "loss": 0.0339, "step": 4670 }, { "grad_norm": 0.35676681995391846, "learning_rate": 9.948165944397799e-05, "loss": 0.0299, "step": 4680 }, { "grad_norm": 0.4107678532600403, "learning_rate": 9.94768992587989e-05, "loss": 0.0338, "step": 4690 }, { "grad_norm": 0.4234924912452698, "learning_rate": 9.947211743074313e-05, "loss": 0.0429, "step": 4700 }, { "grad_norm": 0.3785814344882965, "learning_rate": 9.946731396190246e-05, "loss": 0.0365, "step": 4710 }, { "grad_norm": 0.30872663855552673, "learning_rate": 9.946248885437803e-05, "loss": 0.0307, "step": 4720 }, { "grad_norm": 0.29361608624458313, "learning_rate": 9.945764211028053e-05, "loss": 0.0325, "step": 4730 }, { "grad_norm": 0.44983023405075073, "learning_rate": 9.94527737317301e-05, "loss": 0.0341, "step": 4740 }, { "grad_norm": 0.3165139853954315, "learning_rate": 9.944788372085631e-05, "loss": 0.0342, "step": 4750 }, { "grad_norm": 0.3707144856452942, "learning_rate": 9.944297207979825e-05, "loss": 0.0355, "step": 4760 }, { "grad_norm": 0.32467520236968994, "learning_rate": 9.943803881070441e-05, "loss": 0.0382, "step": 4770 }, { "grad_norm": 0.2798452079296112, "learning_rate": 9.943308391573278e-05, "loss": 0.033, "step": 4780 }, { "grad_norm": 0.4505842924118042, "learning_rate": 9.942810739705079e-05, "loss": 0.0366, "step": 4790 }, { "grad_norm": 0.30525127053260803, "learning_rate": 9.942310925683538e-05, "loss": 0.0355, "step": 4800 }, { "grad_norm": 0.39801350235939026, "learning_rate": 9.941808949727285e-05, "loss": 0.031, "step": 4810 }, { "grad_norm": 0.38724657893180847, "learning_rate": 9.941304812055903e-05, "loss": 0.0338, "step": 4820 }, { "grad_norm": 0.35385754704475403, "learning_rate": 9.940798512889921e-05, "loss": 0.0325, "step": 4830 }, { "grad_norm": 0.2838450074195862, "learning_rate": 9.94029005245081e-05, "loss": 0.0338, "step": 4840 }, { "grad_norm": 0.3314788341522217, "learning_rate": 9.939779430960988e-05, "loss": 0.0337, "step": 4850 }, { "grad_norm": 0.366000235080719, "learning_rate": 9.939266648643817e-05, "loss": 0.0315, "step": 4860 }, { "grad_norm": 0.3744954466819763, "learning_rate": 9.938751705723607e-05, "loss": 0.0388, "step": 4870 }, { "grad_norm": 0.355292946100235, "learning_rate": 9.938234602425613e-05, "loss": 0.033, "step": 4880 }, { "grad_norm": 0.3489528298377991, "learning_rate": 9.93771533897603e-05, "loss": 0.0334, "step": 4890 }, { "grad_norm": 0.3268028199672699, "learning_rate": 9.937193915602004e-05, "loss": 0.0335, "step": 4900 }, { "grad_norm": 0.21146169304847717, "learning_rate": 9.936670332531621e-05, "loss": 0.0306, "step": 4910 }, { "grad_norm": 0.3154968023300171, "learning_rate": 9.936144589993916e-05, "loss": 0.035, "step": 4920 }, { "grad_norm": 0.45724233984947205, "learning_rate": 9.935616688218867e-05, "loss": 0.0356, "step": 4930 }, { "grad_norm": 0.3169020414352417, "learning_rate": 9.935086627437395e-05, "loss": 0.0306, "step": 4940 }, { "grad_norm": 0.28987300395965576, "learning_rate": 9.934554407881366e-05, "loss": 0.0354, "step": 4950 }, { "grad_norm": 0.2962575852870941, "learning_rate": 9.934020029783593e-05, "loss": 0.0353, "step": 4960 }, { "grad_norm": 0.2583666145801544, "learning_rate": 9.933483493377829e-05, "loss": 0.0296, "step": 4970 }, { "grad_norm": 0.304384708404541, "learning_rate": 9.932944798898774e-05, "loss": 0.0366, "step": 4980 }, { "grad_norm": 0.3262456953525543, "learning_rate": 9.932403946582072e-05, "loss": 0.0338, "step": 4990 }, { "grad_norm": 0.29961907863616943, "learning_rate": 9.93186093666431e-05, "loss": 0.0315, "step": 5000 }, { "grad_norm": 0.27681347727775574, "learning_rate": 9.931315769383018e-05, "loss": 0.0325, "step": 5010 }, { "grad_norm": 0.35694533586502075, "learning_rate": 9.930768444976672e-05, "loss": 0.0361, "step": 5020 }, { "grad_norm": 0.31411150097846985, "learning_rate": 9.93021896368469e-05, "loss": 0.0312, "step": 5030 }, { "grad_norm": 0.3105534613132477, "learning_rate": 9.929667325747431e-05, "loss": 0.0299, "step": 5040 }, { "grad_norm": 0.22749543190002441, "learning_rate": 9.929113531406205e-05, "loss": 0.0344, "step": 5050 }, { "grad_norm": 0.3893921673297882, "learning_rate": 9.928557580903257e-05, "loss": 0.0348, "step": 5060 }, { "grad_norm": 0.3666151165962219, "learning_rate": 9.927999474481779e-05, "loss": 0.0337, "step": 5070 }, { "grad_norm": 0.26170024275779724, "learning_rate": 9.927439212385907e-05, "loss": 0.0305, "step": 5080 }, { "grad_norm": 0.324009507894516, "learning_rate": 9.926876794860718e-05, "loss": 0.0336, "step": 5090 }, { "grad_norm": 0.3852055072784424, "learning_rate": 9.926312222152235e-05, "loss": 0.0345, "step": 5100 }, { "grad_norm": 0.309367835521698, "learning_rate": 9.925745494507414e-05, "loss": 0.0329, "step": 5110 }, { "grad_norm": 0.3883343040943146, "learning_rate": 9.925176612174169e-05, "loss": 0.0291, "step": 5120 }, { "grad_norm": 0.3420883119106293, "learning_rate": 9.924605575401346e-05, "loss": 0.0311, "step": 5130 }, { "grad_norm": 0.388572096824646, "learning_rate": 9.924032384438733e-05, "loss": 0.0336, "step": 5140 }, { "grad_norm": 0.35020217299461365, "learning_rate": 9.923457039537066e-05, "loss": 0.0326, "step": 5150 }, { "grad_norm": 0.3022792637348175, "learning_rate": 9.92287954094802e-05, "loss": 0.0335, "step": 5160 }, { "grad_norm": 0.33203238248825073, "learning_rate": 9.922299888924212e-05, "loss": 0.0294, "step": 5170 }, { "grad_norm": 0.3135257661342621, "learning_rate": 9.921718083719203e-05, "loss": 0.0281, "step": 5180 }, { "grad_norm": 0.3928318917751312, "learning_rate": 9.921134125587491e-05, "loss": 0.0345, "step": 5190 }, { "grad_norm": 0.3113667070865631, "learning_rate": 9.920548014784523e-05, "loss": 0.0325, "step": 5200 }, { "grad_norm": 0.3952057659626007, "learning_rate": 9.919959751566681e-05, "loss": 0.0376, "step": 5210 }, { "grad_norm": 0.3353860676288605, "learning_rate": 9.919369336191291e-05, "loss": 0.0383, "step": 5220 }, { "grad_norm": 0.3621787428855896, "learning_rate": 9.918776768916625e-05, "loss": 0.0348, "step": 5230 }, { "grad_norm": 0.39558669924736023, "learning_rate": 9.918182050001888e-05, "loss": 0.0291, "step": 5240 }, { "grad_norm": 0.3138406276702881, "learning_rate": 9.91758517970723e-05, "loss": 0.0348, "step": 5250 }, { "grad_norm": 0.3236713409423828, "learning_rate": 9.916986158293744e-05, "loss": 0.035, "step": 5260 }, { "grad_norm": 0.2943304479122162, "learning_rate": 9.916384986023463e-05, "loss": 0.0331, "step": 5270 }, { "grad_norm": 0.38190874457359314, "learning_rate": 9.915781663159359e-05, "loss": 0.0349, "step": 5280 }, { "grad_norm": 0.3932427763938904, "learning_rate": 9.915176189965346e-05, "loss": 0.0333, "step": 5290 }, { "grad_norm": 0.366428405046463, "learning_rate": 9.914568566706279e-05, "loss": 0.0368, "step": 5300 }, { "grad_norm": 0.4889920651912689, "learning_rate": 9.913958793647953e-05, "loss": 0.0406, "step": 5310 }, { "grad_norm": 0.3270202875137329, "learning_rate": 9.913346871057102e-05, "loss": 0.0407, "step": 5320 }, { "grad_norm": 0.36705857515335083, "learning_rate": 9.912732799201403e-05, "loss": 0.0319, "step": 5330 }, { "grad_norm": 0.37916940450668335, "learning_rate": 9.912116578349474e-05, "loss": 0.0345, "step": 5340 }, { "grad_norm": 0.316608726978302, "learning_rate": 9.911498208770867e-05, "loss": 0.0345, "step": 5350 }, { "grad_norm": 0.34297117590904236, "learning_rate": 9.910877690736078e-05, "loss": 0.029, "step": 5360 }, { "grad_norm": 0.22223787009716034, "learning_rate": 9.910255024516546e-05, "loss": 0.0304, "step": 5370 }, { "grad_norm": 0.31005755066871643, "learning_rate": 9.909630210384644e-05, "loss": 0.0322, "step": 5380 }, { "grad_norm": 0.24371714890003204, "learning_rate": 9.909003248613688e-05, "loss": 0.0302, "step": 5390 }, { "grad_norm": 0.28578463196754456, "learning_rate": 9.908374139477934e-05, "loss": 0.0287, "step": 5400 }, { "grad_norm": 0.2964169979095459, "learning_rate": 9.907742883252571e-05, "loss": 0.0336, "step": 5410 }, { "grad_norm": 0.2196938693523407, "learning_rate": 9.907109480213736e-05, "loss": 0.0319, "step": 5420 }, { "grad_norm": 0.3562096357345581, "learning_rate": 9.906473930638498e-05, "loss": 0.0351, "step": 5430 }, { "grad_norm": 0.4180792272090912, "learning_rate": 9.90583623480487e-05, "loss": 0.0326, "step": 5440 }, { "grad_norm": 0.3615745007991791, "learning_rate": 9.905196392991802e-05, "loss": 0.0341, "step": 5450 }, { "grad_norm": 0.23729492723941803, "learning_rate": 9.904554405479183e-05, "loss": 0.0331, "step": 5460 }, { "grad_norm": 0.31525325775146484, "learning_rate": 9.903910272547838e-05, "loss": 0.0318, "step": 5470 }, { "grad_norm": 0.30022549629211426, "learning_rate": 9.903263994479533e-05, "loss": 0.0329, "step": 5480 }, { "grad_norm": 0.3107682764530182, "learning_rate": 9.902615571556972e-05, "loss": 0.033, "step": 5490 }, { "grad_norm": 0.3192632794380188, "learning_rate": 9.9019650040638e-05, "loss": 0.0293, "step": 5500 }, { "grad_norm": 0.3539043962955475, "learning_rate": 9.901312292284594e-05, "loss": 0.0335, "step": 5510 }, { "grad_norm": 0.2970781922340393, "learning_rate": 9.900657436504875e-05, "loss": 0.0341, "step": 5520 }, { "grad_norm": 0.38731932640075684, "learning_rate": 9.900000437011097e-05, "loss": 0.0383, "step": 5530 }, { "grad_norm": 0.43598106503486633, "learning_rate": 9.899341294090653e-05, "loss": 0.0395, "step": 5540 }, { "grad_norm": 0.3633699119091034, "learning_rate": 9.898680008031877e-05, "loss": 0.0386, "step": 5550 }, { "grad_norm": 0.346344530582428, "learning_rate": 9.898016579124037e-05, "loss": 0.0361, "step": 5560 }, { "grad_norm": 0.3388567566871643, "learning_rate": 9.89735100765734e-05, "loss": 0.0357, "step": 5570 }, { "grad_norm": 0.36766669154167175, "learning_rate": 9.89668329392293e-05, "loss": 0.0309, "step": 5580 }, { "grad_norm": 0.2749458849430084, "learning_rate": 9.896013438212885e-05, "loss": 0.0325, "step": 5590 }, { "grad_norm": 0.25573599338531494, "learning_rate": 9.895341440820225e-05, "loss": 0.0315, "step": 5600 }, { "grad_norm": 0.35110175609588623, "learning_rate": 9.894667302038902e-05, "loss": 0.0314, "step": 5610 }, { "grad_norm": 0.42693403363227844, "learning_rate": 9.893991022163812e-05, "loss": 0.0325, "step": 5620 }, { "grad_norm": 0.2904145419597626, "learning_rate": 9.893312601490779e-05, "loss": 0.0291, "step": 5630 }, { "grad_norm": 0.3274777829647064, "learning_rate": 9.892632040316568e-05, "loss": 0.0289, "step": 5640 }, { "grad_norm": 0.2478848248720169, "learning_rate": 9.891949338938877e-05, "loss": 0.0287, "step": 5650 }, { "grad_norm": 0.291238009929657, "learning_rate": 9.89126449765635e-05, "loss": 0.0373, "step": 5660 }, { "grad_norm": 0.43003249168395996, "learning_rate": 9.890577516768551e-05, "loss": 0.029, "step": 5670 }, { "grad_norm": 0.32455873489379883, "learning_rate": 9.889888396575995e-05, "loss": 0.032, "step": 5680 }, { "grad_norm": 0.2745262384414673, "learning_rate": 9.889197137380125e-05, "loss": 0.0281, "step": 5690 }, { "grad_norm": 0.3585900664329529, "learning_rate": 9.888503739483321e-05, "loss": 0.0317, "step": 5700 }, { "grad_norm": 0.4106243848800659, "learning_rate": 9.887808203188897e-05, "loss": 0.0297, "step": 5710 }, { "grad_norm": 0.3652435839176178, "learning_rate": 9.887110528801108e-05, "loss": 0.0335, "step": 5720 }, { "grad_norm": 0.28336411714553833, "learning_rate": 9.886410716625137e-05, "loss": 0.0337, "step": 5730 }, { "grad_norm": 0.33693593740463257, "learning_rate": 9.885708766967108e-05, "loss": 0.0354, "step": 5740 }, { "grad_norm": 0.32107090950012207, "learning_rate": 9.885004680134076e-05, "loss": 0.0333, "step": 5750 }, { "grad_norm": 0.26424849033355713, "learning_rate": 9.884298456434033e-05, "loss": 0.0291, "step": 5760 }, { "grad_norm": 0.2907412648200989, "learning_rate": 9.883590096175905e-05, "loss": 0.0321, "step": 5770 }, { "grad_norm": 0.31279417872428894, "learning_rate": 9.882879599669554e-05, "loss": 0.0316, "step": 5780 }, { "grad_norm": 0.33282291889190674, "learning_rate": 9.882166967225775e-05, "loss": 0.034, "step": 5790 }, { "grad_norm": 0.358163058757782, "learning_rate": 9.881452199156296e-05, "loss": 0.031, "step": 5800 }, { "grad_norm": 0.25389009714126587, "learning_rate": 9.880735295773782e-05, "loss": 0.0287, "step": 5810 }, { "grad_norm": 0.36743617057800293, "learning_rate": 9.88001625739183e-05, "loss": 0.0324, "step": 5820 }, { "grad_norm": 0.3328872323036194, "learning_rate": 9.879295084324976e-05, "loss": 0.0334, "step": 5830 }, { "grad_norm": 0.34347450733184814, "learning_rate": 9.87857177688868e-05, "loss": 0.0324, "step": 5840 }, { "grad_norm": 0.28508445620536804, "learning_rate": 9.877846335399343e-05, "loss": 0.0322, "step": 5850 }, { "grad_norm": 0.2935417890548706, "learning_rate": 9.8771187601743e-05, "loss": 0.0334, "step": 5860 }, { "grad_norm": 0.29842033982276917, "learning_rate": 9.876389051531813e-05, "loss": 0.0347, "step": 5870 }, { "grad_norm": 0.385525107383728, "learning_rate": 9.875657209791088e-05, "loss": 0.0326, "step": 5880 }, { "grad_norm": 0.3795137107372284, "learning_rate": 9.874923235272248e-05, "loss": 0.03, "step": 5890 }, { "grad_norm": 0.4169817864894867, "learning_rate": 9.874187128296369e-05, "loss": 0.0315, "step": 5900 }, { "grad_norm": 0.22202163934707642, "learning_rate": 9.873448889185439e-05, "loss": 0.0313, "step": 5910 }, { "grad_norm": 0.4209665358066559, "learning_rate": 9.872708518262397e-05, "loss": 0.0299, "step": 5920 }, { "grad_norm": 0.3770769238471985, "learning_rate": 9.871966015851101e-05, "loss": 0.0261, "step": 5930 }, { "grad_norm": 0.3174689710140228, "learning_rate": 9.87122138227635e-05, "loss": 0.0324, "step": 5940 }, { "grad_norm": 0.26743605732917786, "learning_rate": 9.87047461786387e-05, "loss": 0.0266, "step": 5950 }, { "grad_norm": 0.4305480420589447, "learning_rate": 9.869725722940323e-05, "loss": 0.0285, "step": 5960 }, { "grad_norm": 0.2827478051185608, "learning_rate": 9.868974697833299e-05, "loss": 0.0311, "step": 5970 }, { "grad_norm": 0.2934468686580658, "learning_rate": 9.868221542871324e-05, "loss": 0.0282, "step": 5980 }, { "grad_norm": 0.2977944016456604, "learning_rate": 9.867466258383853e-05, "loss": 0.0309, "step": 5990 }, { "grad_norm": 0.3600994050502777, "learning_rate": 9.866708844701272e-05, "loss": 0.033, "step": 6000 }, { "grad_norm": 0.3719973564147949, "learning_rate": 9.865949302154899e-05, "loss": 0.0276, "step": 6010 }, { "grad_norm": 0.28464677929878235, "learning_rate": 9.865187631076987e-05, "loss": 0.0276, "step": 6020 }, { "grad_norm": 0.3651869595050812, "learning_rate": 9.864423831800712e-05, "loss": 0.0267, "step": 6030 }, { "grad_norm": 0.3451078534126282, "learning_rate": 9.863657904660191e-05, "loss": 0.0314, "step": 6040 }, { "grad_norm": 0.2570732533931732, "learning_rate": 9.862889849990462e-05, "loss": 0.0284, "step": 6050 }, { "grad_norm": 0.3290814757347107, "learning_rate": 9.8621196681275e-05, "loss": 0.0264, "step": 6060 }, { "grad_norm": 0.39716318249702454, "learning_rate": 9.861347359408211e-05, "loss": 0.0312, "step": 6070 }, { "grad_norm": 0.30969712138175964, "learning_rate": 9.860572924170426e-05, "loss": 0.0335, "step": 6080 }, { "grad_norm": 0.3009736239910126, "learning_rate": 9.85979636275291e-05, "loss": 0.0258, "step": 6090 }, { "grad_norm": 0.3390500843524933, "learning_rate": 9.859017675495357e-05, "loss": 0.0339, "step": 6100 }, { "grad_norm": 0.31303486227989197, "learning_rate": 9.858236862738392e-05, "loss": 0.0288, "step": 6110 }, { "grad_norm": 0.29130563139915466, "learning_rate": 9.85745392482357e-05, "loss": 0.0343, "step": 6120 }, { "grad_norm": 0.3054701089859009, "learning_rate": 9.856668862093372e-05, "loss": 0.0281, "step": 6130 }, { "grad_norm": 0.32072675228118896, "learning_rate": 9.855881674891214e-05, "loss": 0.0276, "step": 6140 }, { "grad_norm": 0.3291424512863159, "learning_rate": 9.855092363561437e-05, "loss": 0.0327, "step": 6150 }, { "grad_norm": 0.3477396070957184, "learning_rate": 9.854300928449312e-05, "loss": 0.0286, "step": 6160 }, { "grad_norm": 0.2765977382659912, "learning_rate": 9.85350736990104e-05, "loss": 0.0306, "step": 6170 }, { "grad_norm": 0.3779289722442627, "learning_rate": 9.852711688263752e-05, "loss": 0.029, "step": 6180 }, { "grad_norm": 0.2800227105617523, "learning_rate": 9.851913883885503e-05, "loss": 0.027, "step": 6190 }, { "grad_norm": 0.31865155696868896, "learning_rate": 9.851113957115283e-05, "loss": 0.0288, "step": 6200 }, { "grad_norm": 0.22526144981384277, "learning_rate": 9.850311908303005e-05, "loss": 0.0294, "step": 6210 }, { "grad_norm": 0.32080957293510437, "learning_rate": 9.849507737799514e-05, "loss": 0.032, "step": 6220 }, { "grad_norm": 0.2765636146068573, "learning_rate": 9.84870144595658e-05, "loss": 0.0298, "step": 6230 }, { "grad_norm": 0.2436542510986328, "learning_rate": 9.847893033126903e-05, "loss": 0.0285, "step": 6240 }, { "grad_norm": 0.2776239514350891, "learning_rate": 9.847082499664111e-05, "loss": 0.0334, "step": 6250 }, { "grad_norm": 0.342976838350296, "learning_rate": 9.846269845922758e-05, "loss": 0.0333, "step": 6260 }, { "grad_norm": 0.3209812641143799, "learning_rate": 9.845455072258326e-05, "loss": 0.0287, "step": 6270 }, { "grad_norm": 0.4500429928302765, "learning_rate": 9.844638179027226e-05, "loss": 0.032, "step": 6280 }, { "grad_norm": 0.33174213767051697, "learning_rate": 9.843819166586795e-05, "loss": 0.028, "step": 6290 }, { "grad_norm": 0.3553401529788971, "learning_rate": 9.842998035295294e-05, "loss": 0.0353, "step": 6300 }, { "grad_norm": 0.3887661397457123, "learning_rate": 9.842174785511918e-05, "loss": 0.0315, "step": 6310 }, { "grad_norm": 0.33735039830207825, "learning_rate": 9.841349417596779e-05, "loss": 0.0307, "step": 6320 }, { "grad_norm": 0.24815963208675385, "learning_rate": 9.840521931910926e-05, "loss": 0.031, "step": 6330 }, { "grad_norm": 0.34863215684890747, "learning_rate": 9.839692328816327e-05, "loss": 0.0313, "step": 6340 }, { "grad_norm": 0.3684467375278473, "learning_rate": 9.838860608675879e-05, "loss": 0.0329, "step": 6350 }, { "grad_norm": 0.3091985583305359, "learning_rate": 9.838026771853406e-05, "loss": 0.0351, "step": 6360 }, { "grad_norm": 0.342666894197464, "learning_rate": 9.837190818713655e-05, "loss": 0.0319, "step": 6370 }, { "grad_norm": 0.32199227809906006, "learning_rate": 9.836352749622299e-05, "loss": 0.0324, "step": 6380 }, { "grad_norm": 0.2459018975496292, "learning_rate": 9.835512564945941e-05, "loss": 0.0313, "step": 6390 }, { "grad_norm": 0.30317413806915283, "learning_rate": 9.834670265052104e-05, "loss": 0.0275, "step": 6400 }, { "grad_norm": 0.2592995762825012, "learning_rate": 9.833825850309239e-05, "loss": 0.0331, "step": 6410 }, { "grad_norm": 0.33231961727142334, "learning_rate": 9.832979321086723e-05, "loss": 0.0334, "step": 6420 }, { "grad_norm": 0.3258611857891083, "learning_rate": 9.832130677754854e-05, "loss": 0.0296, "step": 6430 }, { "grad_norm": 0.3048709034919739, "learning_rate": 9.831279920684861e-05, "loss": 0.0305, "step": 6440 }, { "grad_norm": 0.31239446997642517, "learning_rate": 9.830427050248891e-05, "loss": 0.03, "step": 6450 }, { "grad_norm": 0.3263850510120392, "learning_rate": 9.829572066820021e-05, "loss": 0.029, "step": 6460 }, { "grad_norm": 0.2974740266799927, "learning_rate": 9.828714970772247e-05, "loss": 0.0295, "step": 6470 }, { "grad_norm": 0.3160155713558197, "learning_rate": 9.827855762480493e-05, "loss": 0.033, "step": 6480 }, { "grad_norm": 0.33003395795822144, "learning_rate": 9.826994442320608e-05, "loss": 0.0308, "step": 6490 }, { "grad_norm": 0.27782559394836426, "learning_rate": 9.82613101066936e-05, "loss": 0.0268, "step": 6500 }, { "grad_norm": 0.3297159671783447, "learning_rate": 9.825265467904445e-05, "loss": 0.0289, "step": 6510 }, { "grad_norm": 0.2415226399898529, "learning_rate": 9.82439781440448e-05, "loss": 0.0277, "step": 6520 }, { "grad_norm": 0.2889556288719177, "learning_rate": 9.823528050549006e-05, "loss": 0.0285, "step": 6530 }, { "grad_norm": 0.3066122829914093, "learning_rate": 9.822656176718487e-05, "loss": 0.028, "step": 6540 }, { "grad_norm": 0.3318462669849396, "learning_rate": 9.82178219329431e-05, "loss": 0.0292, "step": 6550 }, { "grad_norm": 0.4174264669418335, "learning_rate": 9.820906100658789e-05, "loss": 0.0313, "step": 6560 }, { "grad_norm": 0.3039760887622833, "learning_rate": 9.82002789919515e-05, "loss": 0.0348, "step": 6570 }, { "grad_norm": 0.2670954465866089, "learning_rate": 9.819147589287554e-05, "loss": 0.0292, "step": 6580 }, { "grad_norm": 0.32622256875038147, "learning_rate": 9.818265171321074e-05, "loss": 0.0299, "step": 6590 }, { "grad_norm": 0.2653975486755371, "learning_rate": 9.817380645681714e-05, "loss": 0.0328, "step": 6600 }, { "grad_norm": 0.33552253246307373, "learning_rate": 9.816494012756392e-05, "loss": 0.0325, "step": 6610 }, { "grad_norm": 0.3273544907569885, "learning_rate": 9.815605272932953e-05, "loss": 0.0253, "step": 6620 }, { "grad_norm": 0.29956215620040894, "learning_rate": 9.814714426600162e-05, "loss": 0.0336, "step": 6630 }, { "grad_norm": 0.29610592126846313, "learning_rate": 9.813821474147704e-05, "loss": 0.0314, "step": 6640 }, { "grad_norm": 0.2388060986995697, "learning_rate": 9.812926415966189e-05, "loss": 0.0274, "step": 6650 }, { "grad_norm": 0.25543105602264404, "learning_rate": 9.812029252447143e-05, "loss": 0.0321, "step": 6660 }, { "grad_norm": 0.24750353395938873, "learning_rate": 9.811129983983018e-05, "loss": 0.03, "step": 6670 }, { "grad_norm": 0.36393025517463684, "learning_rate": 9.810228610967185e-05, "loss": 0.0311, "step": 6680 }, { "grad_norm": 0.3045990765094757, "learning_rate": 9.809325133793934e-05, "loss": 0.03, "step": 6690 }, { "grad_norm": 0.2678467333316803, "learning_rate": 9.808419552858477e-05, "loss": 0.0309, "step": 6700 }, { "grad_norm": 0.30948904156684875, "learning_rate": 9.807511868556944e-05, "loss": 0.0325, "step": 6710 }, { "grad_norm": 0.2672676146030426, "learning_rate": 9.806602081286392e-05, "loss": 0.0297, "step": 6720 }, { "grad_norm": 0.2727344334125519, "learning_rate": 9.805690191444785e-05, "loss": 0.0268, "step": 6730 }, { "grad_norm": 0.2761107385158539, "learning_rate": 9.804776199431022e-05, "loss": 0.0299, "step": 6740 }, { "grad_norm": 0.2995108366012573, "learning_rate": 9.80386010564491e-05, "loss": 0.0275, "step": 6750 }, { "grad_norm": 0.32105541229248047, "learning_rate": 9.802941910487183e-05, "loss": 0.0278, "step": 6760 }, { "grad_norm": 0.23851805925369263, "learning_rate": 9.802021614359487e-05, "loss": 0.0272, "step": 6770 }, { "grad_norm": 0.27570438385009766, "learning_rate": 9.801099217664394e-05, "loss": 0.0278, "step": 6780 }, { "grad_norm": 0.2814277708530426, "learning_rate": 9.800174720805388e-05, "loss": 0.0267, "step": 6790 }, { "grad_norm": 0.24848848581314087, "learning_rate": 9.799248124186878e-05, "loss": 0.0277, "step": 6800 }, { "grad_norm": 0.29091838002204895, "learning_rate": 9.798319428214187e-05, "loss": 0.0298, "step": 6810 }, { "grad_norm": 0.3515194058418274, "learning_rate": 9.79738863329356e-05, "loss": 0.0313, "step": 6820 }, { "grad_norm": 0.2511124908924103, "learning_rate": 9.79645573983216e-05, "loss": 0.0272, "step": 6830 }, { "grad_norm": 0.2849966287612915, "learning_rate": 9.795520748238062e-05, "loss": 0.0285, "step": 6840 }, { "grad_norm": 0.27315571904182434, "learning_rate": 9.794583658920264e-05, "loss": 0.0308, "step": 6850 }, { "grad_norm": 0.29518911242485046, "learning_rate": 9.793644472288683e-05, "loss": 0.033, "step": 6860 }, { "grad_norm": 0.2919248044490814, "learning_rate": 9.79270318875415e-05, "loss": 0.0253, "step": 6870 }, { "grad_norm": 0.3057709336280823, "learning_rate": 9.791759808728416e-05, "loss": 0.0307, "step": 6880 }, { "grad_norm": 0.24380561709403992, "learning_rate": 9.790814332624143e-05, "loss": 0.025, "step": 6890 }, { "grad_norm": 0.24087966978549957, "learning_rate": 9.789866760854919e-05, "loss": 0.0254, "step": 6900 }, { "grad_norm": 0.29995715618133545, "learning_rate": 9.788917093835242e-05, "loss": 0.0245, "step": 6910 }, { "grad_norm": 0.4264205992221832, "learning_rate": 9.787965331980528e-05, "loss": 0.0288, "step": 6920 }, { "grad_norm": 0.29289960861206055, "learning_rate": 9.787011475707111e-05, "loss": 0.0294, "step": 6930 }, { "grad_norm": 0.30965739488601685, "learning_rate": 9.786055525432241e-05, "loss": 0.0273, "step": 6940 }, { "grad_norm": 0.3240816593170166, "learning_rate": 9.78509748157408e-05, "loss": 0.0257, "step": 6950 }, { "grad_norm": 0.29325491189956665, "learning_rate": 9.784137344551713e-05, "loss": 0.0292, "step": 6960 }, { "grad_norm": 0.2223806232213974, "learning_rate": 9.783175114785134e-05, "loss": 0.0252, "step": 6970 }, { "grad_norm": 0.283572793006897, "learning_rate": 9.782210792695254e-05, "loss": 0.0274, "step": 6980 }, { "grad_norm": 0.2902943789958954, "learning_rate": 9.781244378703904e-05, "loss": 0.0262, "step": 6990 }, { "grad_norm": 0.30948299169540405, "learning_rate": 9.780275873233824e-05, "loss": 0.0275, "step": 7000 }, { "grad_norm": 0.24978826940059662, "learning_rate": 9.77930527670867e-05, "loss": 0.0299, "step": 7010 }, { "grad_norm": 0.269963800907135, "learning_rate": 9.778332589553016e-05, "loss": 0.0299, "step": 7020 }, { "grad_norm": 0.264514297246933, "learning_rate": 9.777357812192349e-05, "loss": 0.0252, "step": 7030 }, { "grad_norm": 0.3384951949119568, "learning_rate": 9.776380945053066e-05, "loss": 0.0277, "step": 7040 }, { "grad_norm": 0.24846996366977692, "learning_rate": 9.775401988562486e-05, "loss": 0.0258, "step": 7050 }, { "grad_norm": 0.2733401656150818, "learning_rate": 9.774420943148835e-05, "loss": 0.0281, "step": 7060 }, { "grad_norm": 0.31899774074554443, "learning_rate": 9.773437809241257e-05, "loss": 0.0254, "step": 7070 }, { "grad_norm": 0.3077262043952942, "learning_rate": 9.772452587269808e-05, "loss": 0.0273, "step": 7080 }, { "grad_norm": 0.2500836253166199, "learning_rate": 9.771465277665457e-05, "loss": 0.0278, "step": 7090 }, { "grad_norm": 0.2937280535697937, "learning_rate": 9.770475880860089e-05, "loss": 0.0249, "step": 7100 }, { "grad_norm": 0.2563726007938385, "learning_rate": 9.769484397286495e-05, "loss": 0.0269, "step": 7110 }, { "grad_norm": 0.2707694470882416, "learning_rate": 9.768490827378388e-05, "loss": 0.0274, "step": 7120 }, { "grad_norm": 0.2777843475341797, "learning_rate": 9.767495171570387e-05, "loss": 0.0236, "step": 7130 }, { "grad_norm": 0.3606092035770416, "learning_rate": 9.766497430298027e-05, "loss": 0.0288, "step": 7140 }, { "grad_norm": 0.2568415701389313, "learning_rate": 9.765497603997754e-05, "loss": 0.0248, "step": 7150 }, { "grad_norm": 0.38418450951576233, "learning_rate": 9.764495693106926e-05, "loss": 0.0276, "step": 7160 }, { "grad_norm": 0.31452423334121704, "learning_rate": 9.76349169806381e-05, "loss": 0.0262, "step": 7170 }, { "grad_norm": 0.36276009678840637, "learning_rate": 9.762485619307591e-05, "loss": 0.0286, "step": 7180 }, { "grad_norm": 0.32332664728164673, "learning_rate": 9.76147745727836e-05, "loss": 0.0265, "step": 7190 }, { "grad_norm": 0.3577694892883301, "learning_rate": 9.760467212417124e-05, "loss": 0.028, "step": 7200 }, { "grad_norm": 0.3289055824279785, "learning_rate": 9.759454885165796e-05, "loss": 0.0293, "step": 7210 }, { "grad_norm": 0.3744608163833618, "learning_rate": 9.7584404759672e-05, "loss": 0.0282, "step": 7220 }, { "grad_norm": 0.2586668133735657, "learning_rate": 9.75742398526508e-05, "loss": 0.0272, "step": 7230 }, { "grad_norm": 0.29433903098106384, "learning_rate": 9.756405413504077e-05, "loss": 0.0245, "step": 7240 }, { "grad_norm": 0.3648439943790436, "learning_rate": 9.755384761129752e-05, "loss": 0.0246, "step": 7250 }, { "grad_norm": 0.320087730884552, "learning_rate": 9.754362028588572e-05, "loss": 0.0254, "step": 7260 }, { "grad_norm": 0.3326079249382019, "learning_rate": 9.753337216327917e-05, "loss": 0.0256, "step": 7270 }, { "grad_norm": 0.2626769542694092, "learning_rate": 9.752310324796073e-05, "loss": 0.0241, "step": 7280 }, { "grad_norm": 0.2519007921218872, "learning_rate": 9.751281354442235e-05, "loss": 0.0273, "step": 7290 }, { "grad_norm": 0.2648239731788635, "learning_rate": 9.750250305716513e-05, "loss": 0.0269, "step": 7300 }, { "grad_norm": 0.34812217950820923, "learning_rate": 9.749217179069923e-05, "loss": 0.0254, "step": 7310 }, { "grad_norm": 0.2994368374347687, "learning_rate": 9.748181974954386e-05, "loss": 0.0269, "step": 7320 }, { "grad_norm": 0.2660892903804779, "learning_rate": 9.74714469382274e-05, "loss": 0.0301, "step": 7330 }, { "grad_norm": 0.2892247438430786, "learning_rate": 9.746105336128724e-05, "loss": 0.0279, "step": 7340 }, { "grad_norm": 0.28384554386138916, "learning_rate": 9.745063902326988e-05, "loss": 0.0283, "step": 7350 }, { "grad_norm": 0.19578608870506287, "learning_rate": 9.744020392873093e-05, "loss": 0.0253, "step": 7360 }, { "grad_norm": 0.2858792841434479, "learning_rate": 9.742974808223504e-05, "loss": 0.0262, "step": 7370 }, { "grad_norm": 0.3058668076992035, "learning_rate": 9.741927148835596e-05, "loss": 0.0255, "step": 7380 }, { "grad_norm": 0.3022083342075348, "learning_rate": 9.740877415167648e-05, "loss": 0.0275, "step": 7390 }, { "grad_norm": 0.29682692885398865, "learning_rate": 9.739825607678854e-05, "loss": 0.0283, "step": 7400 }, { "grad_norm": 0.3016035854816437, "learning_rate": 9.738771726829308e-05, "loss": 0.0315, "step": 7410 }, { "grad_norm": 0.34295621514320374, "learning_rate": 9.737715773080011e-05, "loss": 0.0282, "step": 7420 }, { "grad_norm": 0.3245771825313568, "learning_rate": 9.736657746892877e-05, "loss": 0.0357, "step": 7430 }, { "grad_norm": 0.26664912700653076, "learning_rate": 9.735597648730718e-05, "loss": 0.0293, "step": 7440 }, { "grad_norm": 0.28579097986221313, "learning_rate": 9.734535479057262e-05, "loss": 0.0278, "step": 7450 }, { "grad_norm": 0.2451818287372589, "learning_rate": 9.733471238337136e-05, "loss": 0.028, "step": 7460 }, { "grad_norm": 0.33957910537719727, "learning_rate": 9.732404927035871e-05, "loss": 0.0325, "step": 7470 }, { "grad_norm": 0.3972800374031067, "learning_rate": 9.731336545619915e-05, "loss": 0.0262, "step": 7480 }, { "grad_norm": 0.329716295003891, "learning_rate": 9.73026609455661e-05, "loss": 0.029, "step": 7490 }, { "grad_norm": 0.2676087021827698, "learning_rate": 9.72919357431421e-05, "loss": 0.0289, "step": 7500 }, { "grad_norm": 0.35349008440971375, "learning_rate": 9.72811898536187e-05, "loss": 0.0251, "step": 7510 }, { "grad_norm": 0.2381967306137085, "learning_rate": 9.72704232816965e-05, "loss": 0.023, "step": 7520 }, { "grad_norm": 0.22329097986221313, "learning_rate": 9.725963603208519e-05, "loss": 0.0268, "step": 7530 }, { "grad_norm": 0.23530860245227814, "learning_rate": 9.724882810950348e-05, "loss": 0.0283, "step": 7540 }, { "grad_norm": 0.24802492558956146, "learning_rate": 9.723799951867913e-05, "loss": 0.0277, "step": 7550 }, { "grad_norm": 0.25947779417037964, "learning_rate": 9.722715026434889e-05, "loss": 0.0284, "step": 7560 }, { "grad_norm": 0.26385605335235596, "learning_rate": 9.721628035125866e-05, "loss": 0.0238, "step": 7570 }, { "grad_norm": 0.22414451837539673, "learning_rate": 9.720538978416325e-05, "loss": 0.027, "step": 7580 }, { "grad_norm": 0.29326093196868896, "learning_rate": 9.71944785678266e-05, "loss": 0.0222, "step": 7590 }, { "grad_norm": 0.27058354020118713, "learning_rate": 9.718354670702161e-05, "loss": 0.0224, "step": 7600 }, { "grad_norm": 0.31051915884017944, "learning_rate": 9.717259420653029e-05, "loss": 0.026, "step": 7610 }, { "grad_norm": 0.3452931046485901, "learning_rate": 9.716162107114361e-05, "loss": 0.0264, "step": 7620 }, { "grad_norm": 0.28981807827949524, "learning_rate": 9.71506273056616e-05, "loss": 0.0256, "step": 7630 }, { "grad_norm": 0.28817668557167053, "learning_rate": 9.713961291489331e-05, "loss": 0.0311, "step": 7640 }, { "grad_norm": 0.2584552764892578, "learning_rate": 9.71285779036568e-05, "loss": 0.028, "step": 7650 }, { "grad_norm": 0.3475687801837921, "learning_rate": 9.711752227677916e-05, "loss": 0.0217, "step": 7660 }, { "grad_norm": 0.4387996792793274, "learning_rate": 9.710644603909649e-05, "loss": 0.0261, "step": 7670 }, { "grad_norm": 0.3155224919319153, "learning_rate": 9.709534919545393e-05, "loss": 0.0264, "step": 7680 }, { "grad_norm": 0.24571619927883148, "learning_rate": 9.708423175070563e-05, "loss": 0.0278, "step": 7690 }, { "grad_norm": 0.3032851815223694, "learning_rate": 9.707309370971468e-05, "loss": 0.0257, "step": 7700 }, { "grad_norm": 0.2854657769203186, "learning_rate": 9.70619350773533e-05, "loss": 0.0269, "step": 7710 }, { "grad_norm": 0.3204635977745056, "learning_rate": 9.70507558585026e-05, "loss": 0.0225, "step": 7720 }, { "grad_norm": 0.1806948035955429, "learning_rate": 9.703955605805281e-05, "loss": 0.0254, "step": 7730 }, { "grad_norm": 0.28693827986717224, "learning_rate": 9.702833568090306e-05, "loss": 0.0248, "step": 7740 }, { "grad_norm": 0.29066547751426697, "learning_rate": 9.701709473196154e-05, "loss": 0.0263, "step": 7750 }, { "grad_norm": 0.26554927229881287, "learning_rate": 9.700583321614541e-05, "loss": 0.0263, "step": 7760 }, { "grad_norm": 0.3386361300945282, "learning_rate": 9.699455113838085e-05, "loss": 0.0238, "step": 7770 }, { "grad_norm": 0.31081080436706543, "learning_rate": 9.698324850360303e-05, "loss": 0.0256, "step": 7780 }, { "grad_norm": 0.27737879753112793, "learning_rate": 9.69719253167561e-05, "loss": 0.0224, "step": 7790 }, { "grad_norm": 0.2834523916244507, "learning_rate": 9.69605815827932e-05, "loss": 0.0293, "step": 7800 }, { "grad_norm": 0.2754724323749542, "learning_rate": 9.694921730667647e-05, "loss": 0.0282, "step": 7810 }, { "grad_norm": 0.2540271580219269, "learning_rate": 9.693783249337705e-05, "loss": 0.0252, "step": 7820 }, { "grad_norm": 0.21979637444019318, "learning_rate": 9.692642714787501e-05, "loss": 0.0238, "step": 7830 }, { "grad_norm": 0.3163398802280426, "learning_rate": 9.691500127515945e-05, "loss": 0.0246, "step": 7840 }, { "grad_norm": 0.27172935009002686, "learning_rate": 9.690355488022844e-05, "loss": 0.0213, "step": 7850 }, { "grad_norm": 0.2636539340019226, "learning_rate": 9.689208796808902e-05, "loss": 0.0231, "step": 7860 }, { "grad_norm": 0.25851958990097046, "learning_rate": 9.68806005437572e-05, "loss": 0.022, "step": 7870 }, { "grad_norm": 0.31418976187705994, "learning_rate": 9.686909261225796e-05, "loss": 0.0287, "step": 7880 }, { "grad_norm": 0.2511616051197052, "learning_rate": 9.685756417862531e-05, "loss": 0.0231, "step": 7890 }, { "grad_norm": 0.215915709733963, "learning_rate": 9.684601524790212e-05, "loss": 0.0241, "step": 7900 }, { "grad_norm": 0.24998539686203003, "learning_rate": 9.68344458251403e-05, "loss": 0.0288, "step": 7910 }, { "grad_norm": 0.2416553944349289, "learning_rate": 9.682285591540072e-05, "loss": 0.0253, "step": 7920 }, { "grad_norm": 0.30611878633499146, "learning_rate": 9.681124552375322e-05, "loss": 0.0268, "step": 7930 }, { "grad_norm": 0.23860104382038116, "learning_rate": 9.679961465527654e-05, "loss": 0.0269, "step": 7940 }, { "grad_norm": 0.3028927445411682, "learning_rate": 9.678796331505843e-05, "loss": 0.0217, "step": 7950 }, { "grad_norm": 0.2753969728946686, "learning_rate": 9.677629150819558e-05, "loss": 0.0253, "step": 7960 }, { "grad_norm": 0.23730877041816711, "learning_rate": 9.676459923979366e-05, "loss": 0.0248, "step": 7970 }, { "grad_norm": 0.2380107194185257, "learning_rate": 9.675288651496723e-05, "loss": 0.0257, "step": 7980 }, { "grad_norm": 0.31520572304725647, "learning_rate": 9.674115333883986e-05, "loss": 0.0223, "step": 7990 }, { "grad_norm": 0.28640514612197876, "learning_rate": 9.672939971654402e-05, "loss": 0.0221, "step": 8000 }, { "grad_norm": 0.23724617063999176, "learning_rate": 9.671762565322117e-05, "loss": 0.0237, "step": 8010 }, { "grad_norm": 0.2490135133266449, "learning_rate": 9.670583115402166e-05, "loss": 0.027, "step": 8020 }, { "grad_norm": 0.25580787658691406, "learning_rate": 9.669401622410482e-05, "loss": 0.0221, "step": 8030 }, { "grad_norm": 0.3217068910598755, "learning_rate": 9.668218086863887e-05, "loss": 0.0264, "step": 8040 }, { "grad_norm": 0.24185623228549957, "learning_rate": 9.667032509280103e-05, "loss": 0.0226, "step": 8050 }, { "grad_norm": 0.2768763303756714, "learning_rate": 9.665844890177743e-05, "loss": 0.0226, "step": 8060 }, { "grad_norm": 0.22353193163871765, "learning_rate": 9.664655230076309e-05, "loss": 0.0256, "step": 8070 }, { "grad_norm": 0.2809293568134308, "learning_rate": 9.663463529496199e-05, "loss": 0.0257, "step": 8080 }, { "grad_norm": 0.31753069162368774, "learning_rate": 9.662269788958705e-05, "loss": 0.0236, "step": 8090 }, { "grad_norm": 0.2565624415874481, "learning_rate": 9.661074008986008e-05, "loss": 0.0243, "step": 8100 }, { "grad_norm": 0.22407996654510498, "learning_rate": 9.659876190101184e-05, "loss": 0.0281, "step": 8110 }, { "grad_norm": 0.3168320655822754, "learning_rate": 9.6586763328282e-05, "loss": 0.0259, "step": 8120 }, { "grad_norm": 0.27397286891937256, "learning_rate": 9.657474437691913e-05, "loss": 0.0245, "step": 8130 }, { "grad_norm": 0.34911566972732544, "learning_rate": 9.656270505218073e-05, "loss": 0.0265, "step": 8140 }, { "grad_norm": 0.2502925992012024, "learning_rate": 9.655064535933321e-05, "loss": 0.0262, "step": 8150 }, { "grad_norm": 0.25083398818969727, "learning_rate": 9.653856530365189e-05, "loss": 0.0244, "step": 8160 }, { "grad_norm": 0.29719749093055725, "learning_rate": 9.6526464890421e-05, "loss": 0.0235, "step": 8170 }, { "grad_norm": 0.3182894289493561, "learning_rate": 9.651434412493367e-05, "loss": 0.0251, "step": 8180 }, { "grad_norm": 0.22344397008419037, "learning_rate": 9.650220301249195e-05, "loss": 0.0242, "step": 8190 }, { "grad_norm": 0.27554601430892944, "learning_rate": 9.649004155840675e-05, "loss": 0.0202, "step": 8200 }, { "grad_norm": 0.2753584682941437, "learning_rate": 9.647785976799792e-05, "loss": 0.0224, "step": 8210 }, { "grad_norm": 0.2775469720363617, "learning_rate": 9.646565764659417e-05, "loss": 0.0217, "step": 8220 }, { "grad_norm": 0.3759711682796478, "learning_rate": 9.645343519953314e-05, "loss": 0.0231, "step": 8230 }, { "grad_norm": 0.3116992115974426, "learning_rate": 9.644119243216135e-05, "loss": 0.0214, "step": 8240 }, { "grad_norm": 0.22607944905757904, "learning_rate": 9.642892934983417e-05, "loss": 0.0221, "step": 8250 }, { "grad_norm": 0.2502546012401581, "learning_rate": 9.641664595791593e-05, "loss": 0.0245, "step": 8260 }, { "grad_norm": 0.209568589925766, "learning_rate": 9.640434226177977e-05, "loss": 0.0259, "step": 8270 }, { "grad_norm": 0.27021053433418274, "learning_rate": 9.639201826680777e-05, "loss": 0.0244, "step": 8280 }, { "grad_norm": 0.2539534568786621, "learning_rate": 9.637967397839083e-05, "loss": 0.0241, "step": 8290 }, { "grad_norm": 0.297733336687088, "learning_rate": 9.63673094019288e-05, "loss": 0.0214, "step": 8300 }, { "grad_norm": 0.23181241750717163, "learning_rate": 9.635492454283035e-05, "loss": 0.0248, "step": 8310 }, { "grad_norm": 0.31577733159065247, "learning_rate": 9.634251940651302e-05, "loss": 0.0248, "step": 8320 }, { "grad_norm": 0.2532350420951843, "learning_rate": 9.633009399840327e-05, "loss": 0.0219, "step": 8330 }, { "grad_norm": 0.3019818067550659, "learning_rate": 9.631764832393639e-05, "loss": 0.0237, "step": 8340 }, { "grad_norm": 0.19652318954467773, "learning_rate": 9.630518238855651e-05, "loss": 0.0209, "step": 8350 }, { "grad_norm": 0.3000640273094177, "learning_rate": 9.629269619771668e-05, "loss": 0.0206, "step": 8360 }, { "grad_norm": 0.2535010874271393, "learning_rate": 9.628018975687878e-05, "loss": 0.0213, "step": 8370 }, { "grad_norm": 0.25915199518203735, "learning_rate": 9.626766307151355e-05, "loss": 0.0229, "step": 8380 }, { "grad_norm": 0.31963661313056946, "learning_rate": 9.62551161471006e-05, "loss": 0.0235, "step": 8390 }, { "grad_norm": 0.2695634663105011, "learning_rate": 9.624254898912837e-05, "loss": 0.0235, "step": 8400 }, { "grad_norm": 0.3161536753177643, "learning_rate": 9.622996160309414e-05, "loss": 0.022, "step": 8410 }, { "grad_norm": 0.2878867983818054, "learning_rate": 9.62173539945041e-05, "loss": 0.0222, "step": 8420 }, { "grad_norm": 0.2850261926651001, "learning_rate": 9.620472616887323e-05, "loss": 0.0203, "step": 8430 }, { "grad_norm": 0.2819199562072754, "learning_rate": 9.619207813172536e-05, "loss": 0.0239, "step": 8440 }, { "grad_norm": 0.2379349172115326, "learning_rate": 9.617940988859318e-05, "loss": 0.0239, "step": 8450 }, { "grad_norm": 0.31933924555778503, "learning_rate": 9.616672144501821e-05, "loss": 0.027, "step": 8460 }, { "grad_norm": 0.2723536789417267, "learning_rate": 9.615401280655081e-05, "loss": 0.0256, "step": 8470 }, { "grad_norm": 0.29725322127342224, "learning_rate": 9.614128397875017e-05, "loss": 0.0227, "step": 8480 }, { "grad_norm": 0.2414250522851944, "learning_rate": 9.612853496718429e-05, "loss": 0.0216, "step": 8490 }, { "grad_norm": 0.25188010931015015, "learning_rate": 9.611576577743004e-05, "loss": 0.0227, "step": 8500 }, { "grad_norm": 0.3515135943889618, "learning_rate": 9.610297641507313e-05, "loss": 0.0244, "step": 8510 }, { "grad_norm": 0.294953852891922, "learning_rate": 9.6090166885708e-05, "loss": 0.0263, "step": 8520 }, { "grad_norm": 0.24369871616363525, "learning_rate": 9.607733719493798e-05, "loss": 0.0278, "step": 8530 }, { "grad_norm": 0.28182312846183777, "learning_rate": 9.606448734837527e-05, "loss": 0.0259, "step": 8540 }, { "grad_norm": 0.29961687326431274, "learning_rate": 9.605161735164079e-05, "loss": 0.0242, "step": 8550 }, { "grad_norm": 0.30143076181411743, "learning_rate": 9.60387272103643e-05, "loss": 0.0254, "step": 8560 }, { "grad_norm": 0.24056176841259003, "learning_rate": 9.60258169301844e-05, "loss": 0.0206, "step": 8570 }, { "grad_norm": 0.224228173494339, "learning_rate": 9.601288651674851e-05, "loss": 0.023, "step": 8580 }, { "grad_norm": 0.24074408411979675, "learning_rate": 9.59999359757128e-05, "loss": 0.023, "step": 8590 }, { "grad_norm": 0.23999366164207458, "learning_rate": 9.598696531274227e-05, "loss": 0.0211, "step": 8600 }, { "grad_norm": 0.357965350151062, "learning_rate": 9.597397453351076e-05, "loss": 0.0225, "step": 8610 }, { "grad_norm": 0.2597416341304779, "learning_rate": 9.596096364370087e-05, "loss": 0.0233, "step": 8620 }, { "grad_norm": 0.29217636585235596, "learning_rate": 9.594793264900399e-05, "loss": 0.0206, "step": 8630 }, { "grad_norm": 0.2660157084465027, "learning_rate": 9.593488155512032e-05, "loss": 0.0231, "step": 8640 }, { "grad_norm": 0.18962566554546356, "learning_rate": 9.592181036775886e-05, "loss": 0.0234, "step": 8650 }, { "grad_norm": 0.2694273293018341, "learning_rate": 9.590871909263741e-05, "loss": 0.0246, "step": 8660 }, { "grad_norm": 0.32581937313079834, "learning_rate": 9.589560773548252e-05, "loss": 0.0253, "step": 8670 }, { "grad_norm": 0.2393707036972046, "learning_rate": 9.588247630202952e-05, "loss": 0.0252, "step": 8680 }, { "grad_norm": 0.2798861265182495, "learning_rate": 9.586932479802258e-05, "loss": 0.023, "step": 8690 }, { "grad_norm": 0.309002161026001, "learning_rate": 9.585615322921462e-05, "loss": 0.0218, "step": 8700 }, { "grad_norm": 0.19604802131652832, "learning_rate": 9.58429616013673e-05, "loss": 0.0215, "step": 8710 }, { "grad_norm": 0.3414742648601532, "learning_rate": 9.58297499202511e-05, "loss": 0.0193, "step": 8720 }, { "grad_norm": 0.33937686681747437, "learning_rate": 9.581651819164525e-05, "loss": 0.0246, "step": 8730 }, { "grad_norm": 0.26246705651283264, "learning_rate": 9.580326642133777e-05, "loss": 0.0237, "step": 8740 }, { "grad_norm": 0.2491609752178192, "learning_rate": 9.578999461512544e-05, "loss": 0.0273, "step": 8750 }, { "grad_norm": 0.24505281448364258, "learning_rate": 9.577670277881378e-05, "loss": 0.0221, "step": 8760 }, { "grad_norm": 0.25111615657806396, "learning_rate": 9.57633909182171e-05, "loss": 0.019, "step": 8770 }, { "grad_norm": 0.20893190801143646, "learning_rate": 9.575005903915847e-05, "loss": 0.02, "step": 8780 }, { "grad_norm": 0.24406443536281586, "learning_rate": 9.573670714746972e-05, "loss": 0.0198, "step": 8790 }, { "grad_norm": 0.23990103602409363, "learning_rate": 9.572333524899136e-05, "loss": 0.0246, "step": 8800 }, { "grad_norm": 0.22216957807540894, "learning_rate": 9.570994334957278e-05, "loss": 0.0229, "step": 8810 }, { "grad_norm": 0.23326730728149414, "learning_rate": 9.569653145507201e-05, "loss": 0.0236, "step": 8820 }, { "grad_norm": 0.2843077480792999, "learning_rate": 9.568309957135587e-05, "loss": 0.028, "step": 8830 }, { "grad_norm": 0.2572084963321686, "learning_rate": 9.566964770429995e-05, "loss": 0.0234, "step": 8840 }, { "grad_norm": 0.27594682574272156, "learning_rate": 9.565617585978853e-05, "loss": 0.0257, "step": 8850 }, { "grad_norm": 0.239472433924675, "learning_rate": 9.564268404371466e-05, "loss": 0.0244, "step": 8860 }, { "grad_norm": 0.22838908433914185, "learning_rate": 9.562917226198013e-05, "loss": 0.0251, "step": 8870 }, { "grad_norm": 0.31908294558525085, "learning_rate": 9.561564052049543e-05, "loss": 0.0223, "step": 8880 }, { "grad_norm": 0.24166066944599152, "learning_rate": 9.560208882517982e-05, "loss": 0.0257, "step": 8890 }, { "grad_norm": 0.304926335811615, "learning_rate": 9.558851718196127e-05, "loss": 0.0192, "step": 8900 }, { "grad_norm": 0.2896885871887207, "learning_rate": 9.557492559677646e-05, "loss": 0.0213, "step": 8910 }, { "grad_norm": 0.2170281857252121, "learning_rate": 9.556131407557082e-05, "loss": 0.0204, "step": 8920 }, { "grad_norm": 0.25211814045906067, "learning_rate": 9.554768262429853e-05, "loss": 0.0209, "step": 8930 }, { "grad_norm": 0.28438377380371094, "learning_rate": 9.553403124892239e-05, "loss": 0.02, "step": 8940 }, { "grad_norm": 0.38985973596572876, "learning_rate": 9.552035995541402e-05, "loss": 0.0232, "step": 8950 }, { "grad_norm": 0.3205428421497345, "learning_rate": 9.550666874975368e-05, "loss": 0.0259, "step": 8960 }, { "grad_norm": 0.25402259826660156, "learning_rate": 9.549295763793038e-05, "loss": 0.0226, "step": 8970 }, { "grad_norm": 0.16712293028831482, "learning_rate": 9.547922662594183e-05, "loss": 0.0242, "step": 8980 }, { "grad_norm": 0.3072149455547333, "learning_rate": 9.546547571979443e-05, "loss": 0.0222, "step": 8990 }, { "grad_norm": 0.206142395734787, "learning_rate": 9.545170492550331e-05, "loss": 0.0258, "step": 9000 }, { "grad_norm": 0.32440656423568726, "learning_rate": 9.543791424909226e-05, "loss": 0.0259, "step": 9010 }, { "grad_norm": 0.24773822724819183, "learning_rate": 9.542410369659382e-05, "loss": 0.0207, "step": 9020 }, { "grad_norm": 0.20210446417331696, "learning_rate": 9.541027327404916e-05, "loss": 0.024, "step": 9030 }, { "grad_norm": 0.28600460290908813, "learning_rate": 9.539642298750821e-05, "loss": 0.023, "step": 9040 }, { "grad_norm": 0.23598986864089966, "learning_rate": 9.538255284302954e-05, "loss": 0.0198, "step": 9050 }, { "grad_norm": 0.20501859486103058, "learning_rate": 9.536866284668043e-05, "loss": 0.0204, "step": 9060 }, { "grad_norm": 0.2832529544830322, "learning_rate": 9.53547530045368e-05, "loss": 0.0222, "step": 9070 }, { "grad_norm": 0.3008792996406555, "learning_rate": 9.534082332268335e-05, "loss": 0.0237, "step": 9080 }, { "grad_norm": 0.2605428695678711, "learning_rate": 9.532687380721334e-05, "loss": 0.0221, "step": 9090 }, { "grad_norm": 0.2775779962539673, "learning_rate": 9.53129044642288e-05, "loss": 0.0224, "step": 9100 }, { "grad_norm": 0.2296871393918991, "learning_rate": 9.529891529984039e-05, "loss": 0.0224, "step": 9110 }, { "grad_norm": 0.24996177852153778, "learning_rate": 9.528490632016743e-05, "loss": 0.0226, "step": 9120 }, { "grad_norm": 0.30968156456947327, "learning_rate": 9.52708775313379e-05, "loss": 0.0223, "step": 9130 }, { "grad_norm": 0.29486387968063354, "learning_rate": 9.525682893948853e-05, "loss": 0.0224, "step": 9140 }, { "grad_norm": 0.2608632445335388, "learning_rate": 9.524276055076463e-05, "loss": 0.0204, "step": 9150 }, { "grad_norm": 0.25088000297546387, "learning_rate": 9.522867237132016e-05, "loss": 0.0208, "step": 9160 }, { "grad_norm": 0.29771706461906433, "learning_rate": 9.52145644073178e-05, "loss": 0.0239, "step": 9170 }, { "grad_norm": 0.29451388120651245, "learning_rate": 9.520043666492884e-05, "loss": 0.0203, "step": 9180 }, { "grad_norm": 0.23304559290409088, "learning_rate": 9.518628915033323e-05, "loss": 0.0221, "step": 9190 }, { "grad_norm": 0.2494295835494995, "learning_rate": 9.517212186971957e-05, "loss": 0.022, "step": 9200 }, { "grad_norm": 0.23576436936855316, "learning_rate": 9.515793482928515e-05, "loss": 0.0232, "step": 9210 }, { "grad_norm": 0.2726961672306061, "learning_rate": 9.514372803523582e-05, "loss": 0.0246, "step": 9220 }, { "grad_norm": 0.2853924334049225, "learning_rate": 9.512950149378613e-05, "loss": 0.0205, "step": 9230 }, { "grad_norm": 0.3077450096607208, "learning_rate": 9.511525521115925e-05, "loss": 0.0235, "step": 9240 }, { "grad_norm": 0.23647278547286987, "learning_rate": 9.510098919358698e-05, "loss": 0.0204, "step": 9250 }, { "grad_norm": 0.2955886423587799, "learning_rate": 9.508670344730979e-05, "loss": 0.0204, "step": 9260 }, { "grad_norm": 0.21963225305080414, "learning_rate": 9.507239797857672e-05, "loss": 0.0216, "step": 9270 }, { "grad_norm": 0.2760219871997833, "learning_rate": 9.505807279364548e-05, "loss": 0.0219, "step": 9280 }, { "grad_norm": 0.26910877227783203, "learning_rate": 9.504372789878239e-05, "loss": 0.0219, "step": 9290 }, { "grad_norm": 0.2729431986808777, "learning_rate": 9.502936330026239e-05, "loss": 0.0224, "step": 9300 }, { "grad_norm": 0.2641892731189728, "learning_rate": 9.501497900436906e-05, "loss": 0.0198, "step": 9310 }, { "grad_norm": 0.27855247259140015, "learning_rate": 9.500057501739457e-05, "loss": 0.0195, "step": 9320 }, { "grad_norm": 0.2845800220966339, "learning_rate": 9.49861513456397e-05, "loss": 0.0208, "step": 9330 }, { "grad_norm": 0.2375309318304062, "learning_rate": 9.497170799541388e-05, "loss": 0.0192, "step": 9340 }, { "grad_norm": 0.30093836784362793, "learning_rate": 9.495724497303508e-05, "loss": 0.0224, "step": 9350 }, { "grad_norm": 0.24772000312805176, "learning_rate": 9.494276228482998e-05, "loss": 0.0208, "step": 9360 }, { "grad_norm": 0.232768252491951, "learning_rate": 9.492825993713374e-05, "loss": 0.0178, "step": 9370 }, { "grad_norm": 0.23551689088344574, "learning_rate": 9.491373793629023e-05, "loss": 0.021, "step": 9380 }, { "grad_norm": 0.2408761978149414, "learning_rate": 9.489919628865182e-05, "loss": 0.0181, "step": 9390 }, { "grad_norm": 0.28038832545280457, "learning_rate": 9.488463500057955e-05, "loss": 0.0213, "step": 9400 }, { "grad_norm": 0.3298349678516388, "learning_rate": 9.487005407844302e-05, "loss": 0.0189, "step": 9410 }, { "grad_norm": 0.2700117230415344, "learning_rate": 9.485545352862039e-05, "loss": 0.0222, "step": 9420 }, { "grad_norm": 0.1923956722021103, "learning_rate": 9.48408333574985e-05, "loss": 0.023, "step": 9430 }, { "grad_norm": 0.31371650099754333, "learning_rate": 9.482619357147264e-05, "loss": 0.0206, "step": 9440 }, { "grad_norm": 0.25857922434806824, "learning_rate": 9.481153417694679e-05, "loss": 0.0228, "step": 9450 }, { "grad_norm": 0.18859760463237762, "learning_rate": 9.479685518033347e-05, "loss": 0.019, "step": 9460 }, { "grad_norm": 0.2824561297893524, "learning_rate": 9.478215658805375e-05, "loss": 0.0192, "step": 9470 }, { "grad_norm": 0.34904444217681885, "learning_rate": 9.47674384065373e-05, "loss": 0.0228, "step": 9480 }, { "grad_norm": 0.2556479573249817, "learning_rate": 9.475270064222237e-05, "loss": 0.0216, "step": 9490 }, { "grad_norm": 0.23711754381656647, "learning_rate": 9.473794330155572e-05, "loss": 0.0219, "step": 9500 }, { "grad_norm": 0.22615624964237213, "learning_rate": 9.472316639099276e-05, "loss": 0.0217, "step": 9510 }, { "grad_norm": 0.22031806409358978, "learning_rate": 9.470836991699739e-05, "loss": 0.0188, "step": 9520 }, { "grad_norm": 0.19253870844841003, "learning_rate": 9.469355388604208e-05, "loss": 0.0161, "step": 9530 }, { "grad_norm": 0.18653284013271332, "learning_rate": 9.467871830460787e-05, "loss": 0.0192, "step": 9540 }, { "grad_norm": 0.25115537643432617, "learning_rate": 9.466386317918436e-05, "loss": 0.0249, "step": 9550 }, { "grad_norm": 0.37245863676071167, "learning_rate": 9.464898851626969e-05, "loss": 0.0224, "step": 9560 }, { "grad_norm": 0.27260255813598633, "learning_rate": 9.463409432237051e-05, "loss": 0.0208, "step": 9570 }, { "grad_norm": 0.2518012523651123, "learning_rate": 9.461918060400209e-05, "loss": 0.0232, "step": 9580 }, { "grad_norm": 0.31925898790359497, "learning_rate": 9.460424736768816e-05, "loss": 0.0247, "step": 9590 }, { "grad_norm": 0.2036992609500885, "learning_rate": 9.458929461996105e-05, "loss": 0.0208, "step": 9600 }, { "grad_norm": 0.2276897132396698, "learning_rate": 9.457432236736158e-05, "loss": 0.0237, "step": 9610 }, { "grad_norm": 0.23287692666053772, "learning_rate": 9.455933061643916e-05, "loss": 0.018, "step": 9620 }, { "grad_norm": 0.2919159531593323, "learning_rate": 9.454431937375164e-05, "loss": 0.0202, "step": 9630 }, { "grad_norm": 0.2592717111110687, "learning_rate": 9.452928864586547e-05, "loss": 0.0238, "step": 9640 }, { "grad_norm": 0.33278924226760864, "learning_rate": 9.451423843935563e-05, "loss": 0.0235, "step": 9650 }, { "grad_norm": 0.23154206573963165, "learning_rate": 9.449916876080553e-05, "loss": 0.0225, "step": 9660 }, { "grad_norm": 0.26563605666160583, "learning_rate": 9.44840796168072e-05, "loss": 0.0176, "step": 9670 }, { "grad_norm": 0.20860138535499573, "learning_rate": 9.446897101396115e-05, "loss": 0.0177, "step": 9680 }, { "grad_norm": 0.271651029586792, "learning_rate": 9.445384295887638e-05, "loss": 0.0225, "step": 9690 }, { "grad_norm": 0.26481950283050537, "learning_rate": 9.443869545817043e-05, "loss": 0.0222, "step": 9700 }, { "grad_norm": 0.24151715636253357, "learning_rate": 9.442352851846929e-05, "loss": 0.0225, "step": 9710 }, { "grad_norm": 0.2635901868343353, "learning_rate": 9.440834214640755e-05, "loss": 0.0212, "step": 9720 }, { "grad_norm": 0.24533984065055847, "learning_rate": 9.439313634862823e-05, "loss": 0.0215, "step": 9730 }, { "grad_norm": 0.17376433312892914, "learning_rate": 9.437791113178282e-05, "loss": 0.0218, "step": 9740 }, { "grad_norm": 0.2715037167072296, "learning_rate": 9.43626665025314e-05, "loss": 0.0234, "step": 9750 }, { "grad_norm": 0.30662694573402405, "learning_rate": 9.434740246754248e-05, "loss": 0.0205, "step": 9760 }, { "grad_norm": 0.27739012241363525, "learning_rate": 9.433211903349304e-05, "loss": 0.0186, "step": 9770 }, { "grad_norm": 0.23447513580322266, "learning_rate": 9.431681620706858e-05, "loss": 0.0179, "step": 9780 }, { "grad_norm": 0.30972781777381897, "learning_rate": 9.43014939949631e-05, "loss": 0.0198, "step": 9790 }, { "grad_norm": 0.23031923174858093, "learning_rate": 9.428615240387904e-05, "loss": 0.0194, "step": 9800 }, { "grad_norm": 0.29500728845596313, "learning_rate": 9.427079144052732e-05, "loss": 0.0272, "step": 9810 }, { "grad_norm": 0.3305860161781311, "learning_rate": 9.425541111162739e-05, "loss": 0.0229, "step": 9820 }, { "grad_norm": 0.2628893256187439, "learning_rate": 9.424001142390709e-05, "loss": 0.0248, "step": 9830 }, { "grad_norm": 0.1825234591960907, "learning_rate": 9.422459238410277e-05, "loss": 0.0226, "step": 9840 }, { "grad_norm": 0.27743107080459595, "learning_rate": 9.420915399895926e-05, "loss": 0.0233, "step": 9850 }, { "grad_norm": 0.3016844093799591, "learning_rate": 9.419369627522981e-05, "loss": 0.0208, "step": 9860 }, { "grad_norm": 0.23935569822788239, "learning_rate": 9.417821921967618e-05, "loss": 0.0179, "step": 9870 }, { "grad_norm": 0.294528067111969, "learning_rate": 9.416272283906855e-05, "loss": 0.0191, "step": 9880 }, { "grad_norm": 0.3194102346897125, "learning_rate": 9.414720714018554e-05, "loss": 0.0211, "step": 9890 }, { "grad_norm": 0.24373705685138702, "learning_rate": 9.413167212981427e-05, "loss": 0.0202, "step": 9900 }, { "grad_norm": 0.19963878393173218, "learning_rate": 9.41161178147503e-05, "loss": 0.0205, "step": 9910 }, { "grad_norm": 0.25904932618141174, "learning_rate": 9.410054420179755e-05, "loss": 0.0185, "step": 9920 }, { "grad_norm": 0.25355279445648193, "learning_rate": 9.408495129776851e-05, "loss": 0.0189, "step": 9930 }, { "grad_norm": 0.2866595983505249, "learning_rate": 9.406933910948403e-05, "loss": 0.0171, "step": 9940 }, { "grad_norm": 0.2677626609802246, "learning_rate": 9.40537076437734e-05, "loss": 0.0249, "step": 9950 }, { "grad_norm": 0.24794548749923706, "learning_rate": 9.403805690747436e-05, "loss": 0.0196, "step": 9960 }, { "grad_norm": 0.2662081718444824, "learning_rate": 9.402238690743308e-05, "loss": 0.0188, "step": 9970 }, { "grad_norm": 0.2326134592294693, "learning_rate": 9.400669765050413e-05, "loss": 0.0159, "step": 9980 }, { "grad_norm": 0.2483377754688263, "learning_rate": 9.399098914355055e-05, "loss": 0.0181, "step": 9990 }, { "grad_norm": 0.28269460797309875, "learning_rate": 9.397526139344378e-05, "loss": 0.0201, "step": 10000 }, { "grad_norm": 0.21608909964561462, "learning_rate": 9.395951440706362e-05, "loss": 0.0169, "step": 10010 }, { "grad_norm": 0.2812051475048065, "learning_rate": 9.394374819129839e-05, "loss": 0.0205, "step": 10020 }, { "grad_norm": 0.25992244482040405, "learning_rate": 9.392796275304474e-05, "loss": 0.0217, "step": 10030 }, { "grad_norm": 0.2358676940202713, "learning_rate": 9.391215809920778e-05, "loss": 0.0174, "step": 10040 }, { "grad_norm": 0.25701212882995605, "learning_rate": 9.389633423670099e-05, "loss": 0.019, "step": 10050 }, { "grad_norm": 0.16488133370876312, "learning_rate": 9.388049117244626e-05, "loss": 0.0177, "step": 10060 }, { "grad_norm": 0.2120034098625183, "learning_rate": 9.386462891337389e-05, "loss": 0.0175, "step": 10070 }, { "grad_norm": 0.22943192720413208, "learning_rate": 9.384874746642257e-05, "loss": 0.0258, "step": 10080 }, { "grad_norm": 0.3567817211151123, "learning_rate": 9.383284683853937e-05, "loss": 0.0184, "step": 10090 }, { "grad_norm": 0.19279062747955322, "learning_rate": 9.381692703667981e-05, "loss": 0.0177, "step": 10100 }, { "grad_norm": 0.21944560110569, "learning_rate": 9.380098806780771e-05, "loss": 0.0161, "step": 10110 }, { "grad_norm": 0.26716071367263794, "learning_rate": 9.378502993889533e-05, "loss": 0.019, "step": 10120 }, { "grad_norm": 0.23674671351909637, "learning_rate": 9.376905265692329e-05, "loss": 0.0217, "step": 10130 }, { "grad_norm": 0.21083150804042816, "learning_rate": 9.37530562288806e-05, "loss": 0.0182, "step": 10140 }, { "grad_norm": 0.28007814288139343, "learning_rate": 9.373704066176465e-05, "loss": 0.0179, "step": 10150 }, { "grad_norm": 0.25921449065208435, "learning_rate": 9.372100596258118e-05, "loss": 0.0201, "step": 10160 }, { "grad_norm": 0.1855979561805725, "learning_rate": 9.370495213834433e-05, "loss": 0.0204, "step": 10170 }, { "grad_norm": 0.2711697816848755, "learning_rate": 9.368887919607657e-05, "loss": 0.0207, "step": 10180 }, { "grad_norm": 0.2301127165555954, "learning_rate": 9.367278714280876e-05, "loss": 0.0195, "step": 10190 }, { "grad_norm": 0.17540240287780762, "learning_rate": 9.36566759855801e-05, "loss": 0.0192, "step": 10200 }, { "grad_norm": 0.25268253684043884, "learning_rate": 9.36405457314382e-05, "loss": 0.0215, "step": 10210 }, { "grad_norm": 0.24855005741119385, "learning_rate": 9.36243963874389e-05, "loss": 0.0234, "step": 10220 }, { "grad_norm": 0.2017088681459427, "learning_rate": 9.360822796064655e-05, "loss": 0.0193, "step": 10230 }, { "grad_norm": 0.2046915888786316, "learning_rate": 9.359204045813372e-05, "loss": 0.0179, "step": 10240 }, { "grad_norm": 0.21998535096645355, "learning_rate": 9.357583388698141e-05, "loss": 0.0166, "step": 10250 }, { "grad_norm": 0.2539973855018616, "learning_rate": 9.35596082542789e-05, "loss": 0.0219, "step": 10260 }, { "grad_norm": 0.18894517421722412, "learning_rate": 9.354336356712383e-05, "loss": 0.0212, "step": 10270 }, { "grad_norm": 0.25493210554122925, "learning_rate": 9.35270998326222e-05, "loss": 0.0206, "step": 10280 }, { "grad_norm": 0.24909581243991852, "learning_rate": 9.351081705788831e-05, "loss": 0.0169, "step": 10290 }, { "grad_norm": 0.22944337129592896, "learning_rate": 9.349451525004477e-05, "loss": 0.0191, "step": 10300 }, { "grad_norm": 0.20833955705165863, "learning_rate": 9.347819441622261e-05, "loss": 0.0185, "step": 10310 }, { "grad_norm": 0.3538637161254883, "learning_rate": 9.346185456356105e-05, "loss": 0.0172, "step": 10320 }, { "grad_norm": 0.26469504833221436, "learning_rate": 9.344549569920774e-05, "loss": 0.0206, "step": 10330 }, { "grad_norm": 0.2403581738471985, "learning_rate": 9.342911783031858e-05, "loss": 0.0185, "step": 10340 }, { "grad_norm": 0.2301088571548462, "learning_rate": 9.341272096405782e-05, "loss": 0.0238, "step": 10350 }, { "grad_norm": 0.2804258167743683, "learning_rate": 9.3396305107598e-05, "loss": 0.0212, "step": 10360 }, { "grad_norm": 0.21539585292339325, "learning_rate": 9.337987026811998e-05, "loss": 0.0188, "step": 10370 }, { "grad_norm": 0.22304101288318634, "learning_rate": 9.33634164528129e-05, "loss": 0.0175, "step": 10380 }, { "grad_norm": 0.28763920068740845, "learning_rate": 9.334694366887424e-05, "loss": 0.0198, "step": 10390 }, { "grad_norm": 0.2419157475233078, "learning_rate": 9.333045192350973e-05, "loss": 0.0204, "step": 10400 }, { "grad_norm": 0.1978996992111206, "learning_rate": 9.331394122393345e-05, "loss": 0.0184, "step": 10410 }, { "grad_norm": 0.24739816784858704, "learning_rate": 9.329741157736771e-05, "loss": 0.0152, "step": 10420 }, { "grad_norm": 0.23371540009975433, "learning_rate": 9.328086299104317e-05, "loss": 0.02, "step": 10430 }, { "grad_norm": 0.2428014874458313, "learning_rate": 9.326429547219872e-05, "loss": 0.0192, "step": 10440 }, { "grad_norm": 0.3076070249080658, "learning_rate": 9.324770902808155e-05, "loss": 0.0181, "step": 10450 }, { "grad_norm": 0.24483181536197662, "learning_rate": 9.323110366594717e-05, "loss": 0.0189, "step": 10460 }, { "grad_norm": 0.19933010637760162, "learning_rate": 9.32144793930593e-05, "loss": 0.0182, "step": 10470 }, { "grad_norm": 0.2416263222694397, "learning_rate": 9.319783621668996e-05, "loss": 0.0177, "step": 10480 }, { "grad_norm": 0.24554041028022766, "learning_rate": 9.318117414411947e-05, "loss": 0.0176, "step": 10490 }, { "grad_norm": 0.26497769355773926, "learning_rate": 9.316449318263635e-05, "loss": 0.0178, "step": 10500 }, { "grad_norm": 0.22803758084774017, "learning_rate": 9.314779333953744e-05, "loss": 0.0202, "step": 10510 }, { "grad_norm": 0.2899167239665985, "learning_rate": 9.313107462212781e-05, "loss": 0.021, "step": 10520 }, { "grad_norm": 0.2137484848499298, "learning_rate": 9.311433703772082e-05, "loss": 0.0203, "step": 10530 }, { "grad_norm": 0.23829394578933716, "learning_rate": 9.3097580593638e-05, "loss": 0.0191, "step": 10540 }, { "grad_norm": 0.22926506400108337, "learning_rate": 9.308080529720926e-05, "loss": 0.0209, "step": 10550 }, { "grad_norm": 0.2344760000705719, "learning_rate": 9.306401115577264e-05, "loss": 0.018, "step": 10560 }, { "grad_norm": 0.28586333990097046, "learning_rate": 9.304719817667447e-05, "loss": 0.0207, "step": 10570 }, { "grad_norm": 0.1966448724269867, "learning_rate": 9.303036636726934e-05, "loss": 0.0173, "step": 10580 }, { "grad_norm": 0.2780885696411133, "learning_rate": 9.301351573492003e-05, "loss": 0.023, "step": 10590 }, { "grad_norm": 0.20525626838207245, "learning_rate": 9.299664628699758e-05, "loss": 0.0209, "step": 10600 }, { "grad_norm": 0.27389705181121826, "learning_rate": 9.297975803088129e-05, "loss": 0.0229, "step": 10610 }, { "grad_norm": 0.23744262754917145, "learning_rate": 9.296285097395864e-05, "loss": 0.0241, "step": 10620 }, { "grad_norm": 0.3115321397781372, "learning_rate": 9.294592512362533e-05, "loss": 0.0179, "step": 10630 }, { "grad_norm": 0.2978227138519287, "learning_rate": 9.292898048728533e-05, "loss": 0.0235, "step": 10640 }, { "grad_norm": 0.17475973069667816, "learning_rate": 9.29120170723508e-05, "loss": 0.0194, "step": 10650 }, { "grad_norm": 0.28474026918411255, "learning_rate": 9.28950348862421e-05, "loss": 0.0203, "step": 10660 }, { "grad_norm": 0.25000232458114624, "learning_rate": 9.287803393638781e-05, "loss": 0.0181, "step": 10670 }, { "grad_norm": 0.23203131556510925, "learning_rate": 9.286101423022474e-05, "loss": 0.0161, "step": 10680 }, { "grad_norm": 0.2135542780160904, "learning_rate": 9.284397577519788e-05, "loss": 0.0173, "step": 10690 }, { "grad_norm": 0.14061014354228973, "learning_rate": 9.282691857876043e-05, "loss": 0.0189, "step": 10700 }, { "grad_norm": 0.24653032422065735, "learning_rate": 9.280984264837377e-05, "loss": 0.0171, "step": 10710 }, { "grad_norm": 0.32819023728370667, "learning_rate": 9.279274799150752e-05, "loss": 0.0209, "step": 10720 }, { "grad_norm": 0.21283037960529327, "learning_rate": 9.277563461563945e-05, "loss": 0.0205, "step": 10730 }, { "grad_norm": 0.2716512978076935, "learning_rate": 9.275850252825555e-05, "loss": 0.0184, "step": 10740 }, { "grad_norm": 0.2231690138578415, "learning_rate": 9.274135173684994e-05, "loss": 0.017, "step": 10750 }, { "grad_norm": 0.2893235385417938, "learning_rate": 9.272418224892498e-05, "loss": 0.0183, "step": 10760 }, { "grad_norm": 0.2521395981311798, "learning_rate": 9.27069940719912e-05, "loss": 0.0196, "step": 10770 }, { "grad_norm": 0.27969545125961304, "learning_rate": 9.268978721356727e-05, "loss": 0.0183, "step": 10780 }, { "grad_norm": 0.257140576839447, "learning_rate": 9.267256168118008e-05, "loss": 0.0223, "step": 10790 }, { "grad_norm": 0.24515564739704132, "learning_rate": 9.265531748236463e-05, "loss": 0.0202, "step": 10800 }, { "grad_norm": 0.22425413131713867, "learning_rate": 9.263805462466416e-05, "loss": 0.0153, "step": 10810 }, { "grad_norm": 0.25628969073295593, "learning_rate": 9.262077311562998e-05, "loss": 0.0205, "step": 10820 }, { "grad_norm": 0.2316131889820099, "learning_rate": 9.260347296282165e-05, "loss": 0.021, "step": 10830 }, { "grad_norm": 0.2972199618816376, "learning_rate": 9.258615417380683e-05, "loss": 0.0197, "step": 10840 }, { "grad_norm": 0.25416114926338196, "learning_rate": 9.256881675616133e-05, "loss": 0.0165, "step": 10850 }, { "grad_norm": 0.23889869451522827, "learning_rate": 9.255146071746917e-05, "loss": 0.0208, "step": 10860 }, { "grad_norm": 0.19452854990959167, "learning_rate": 9.253408606532241e-05, "loss": 0.019, "step": 10870 }, { "grad_norm": 0.19842718541622162, "learning_rate": 9.251669280732137e-05, "loss": 0.0179, "step": 10880 }, { "grad_norm": 0.5598425269126892, "learning_rate": 9.249928095107441e-05, "loss": 0.0193, "step": 10890 }, { "grad_norm": 0.2181270718574524, "learning_rate": 9.248185050419811e-05, "loss": 0.0162, "step": 10900 }, { "grad_norm": 0.20888836681842804, "learning_rate": 9.24644014743171e-05, "loss": 0.0194, "step": 10910 }, { "grad_norm": 0.16374655067920685, "learning_rate": 9.24469338690642e-05, "loss": 0.019, "step": 10920 }, { "grad_norm": 0.24447907507419586, "learning_rate": 9.242944769608033e-05, "loss": 0.0177, "step": 10930 }, { "grad_norm": 0.21491360664367676, "learning_rate": 9.241194296301454e-05, "loss": 0.0179, "step": 10940 }, { "grad_norm": 0.13870009779930115, "learning_rate": 9.239441967752397e-05, "loss": 0.0152, "step": 10950 }, { "grad_norm": 0.22660881280899048, "learning_rate": 9.237687784727393e-05, "loss": 0.0148, "step": 10960 }, { "grad_norm": 0.21305285394191742, "learning_rate": 9.235931747993781e-05, "loss": 0.0148, "step": 10970 }, { "grad_norm": 0.1817772090435028, "learning_rate": 9.234173858319707e-05, "loss": 0.0185, "step": 10980 }, { "grad_norm": 0.2243172824382782, "learning_rate": 9.23241411647414e-05, "loss": 0.0196, "step": 10990 }, { "grad_norm": 0.24859297275543213, "learning_rate": 9.230652523226841e-05, "loss": 0.0187, "step": 11000 }, { "grad_norm": 0.26410484313964844, "learning_rate": 9.2288890793484e-05, "loss": 0.0173, "step": 11010 }, { "grad_norm": 0.2054034322500229, "learning_rate": 9.227123785610199e-05, "loss": 0.0175, "step": 11020 }, { "grad_norm": 0.21280667185783386, "learning_rate": 9.225356642784443e-05, "loss": 0.0191, "step": 11030 }, { "grad_norm": 0.19988057017326355, "learning_rate": 9.223587651644139e-05, "loss": 0.0181, "step": 11040 }, { "grad_norm": 0.26031792163848877, "learning_rate": 9.221816812963104e-05, "loss": 0.0205, "step": 11050 }, { "grad_norm": 0.2584247887134552, "learning_rate": 9.22004412751596e-05, "loss": 0.0179, "step": 11060 }, { "grad_norm": 0.26613327860832214, "learning_rate": 9.218269596078146e-05, "loss": 0.0187, "step": 11070 }, { "grad_norm": 0.23883669078350067, "learning_rate": 9.216493219425895e-05, "loss": 0.02, "step": 11080 }, { "grad_norm": 0.3305536210536957, "learning_rate": 9.21471499833626e-05, "loss": 0.0173, "step": 11090 }, { "grad_norm": 0.23506855964660645, "learning_rate": 9.212934933587094e-05, "loss": 0.0185, "step": 11100 }, { "grad_norm": 0.25789836049079895, "learning_rate": 9.211153025957056e-05, "loss": 0.0179, "step": 11110 }, { "grad_norm": 0.2348192185163498, "learning_rate": 9.209369276225614e-05, "loss": 0.0172, "step": 11120 }, { "grad_norm": 0.20443031191825867, "learning_rate": 9.207583685173042e-05, "loss": 0.0186, "step": 11130 }, { "grad_norm": 0.23772911727428436, "learning_rate": 9.205796253580417e-05, "loss": 0.0197, "step": 11140 }, { "grad_norm": 0.18360383808612823, "learning_rate": 9.204006982229621e-05, "loss": 0.0192, "step": 11150 }, { "grad_norm": 0.23849916458129883, "learning_rate": 9.202215871903346e-05, "loss": 0.0147, "step": 11160 }, { "grad_norm": 0.16656629741191864, "learning_rate": 9.20042292338508e-05, "loss": 0.0166, "step": 11170 }, { "grad_norm": 0.1823406219482422, "learning_rate": 9.198628137459123e-05, "loss": 0.0194, "step": 11180 }, { "grad_norm": 0.2135552167892456, "learning_rate": 9.196831514910572e-05, "loss": 0.0167, "step": 11190 }, { "grad_norm": 0.18353556096553802, "learning_rate": 9.195033056525332e-05, "loss": 0.0176, "step": 11200 }, { "grad_norm": 0.2681991159915924, "learning_rate": 9.193232763090114e-05, "loss": 0.021, "step": 11210 }, { "grad_norm": 0.2665809392929077, "learning_rate": 9.191430635392422e-05, "loss": 0.0164, "step": 11220 }, { "grad_norm": 0.1974109262228012, "learning_rate": 9.18962667422057e-05, "loss": 0.0167, "step": 11230 }, { "grad_norm": 0.1896234005689621, "learning_rate": 9.187820880363671e-05, "loss": 0.0168, "step": 11240 }, { "grad_norm": 0.22469575703144073, "learning_rate": 9.186013254611643e-05, "loss": 0.0177, "step": 11250 }, { "grad_norm": 0.1996164321899414, "learning_rate": 9.1842037977552e-05, "loss": 0.019, "step": 11260 }, { "grad_norm": 0.2855274975299835, "learning_rate": 9.182392510585862e-05, "loss": 0.0234, "step": 11270 }, { "grad_norm": 0.1962725669145584, "learning_rate": 9.180579393895946e-05, "loss": 0.0167, "step": 11280 }, { "grad_norm": 0.1875135451555252, "learning_rate": 9.178764448478572e-05, "loss": 0.0158, "step": 11290 }, { "grad_norm": 0.2569526433944702, "learning_rate": 9.176947675127658e-05, "loss": 0.017, "step": 11300 }, { "grad_norm": 0.22981354594230652, "learning_rate": 9.175129074637924e-05, "loss": 0.0207, "step": 11310 }, { "grad_norm": 0.19379164278507233, "learning_rate": 9.173308647804884e-05, "loss": 0.0149, "step": 11320 }, { "grad_norm": 0.23140884935855865, "learning_rate": 9.171486395424859e-05, "loss": 0.0184, "step": 11330 }, { "grad_norm": 0.245038703083992, "learning_rate": 9.16966231829496e-05, "loss": 0.0198, "step": 11340 }, { "grad_norm": 0.24129366874694824, "learning_rate": 9.167836417213105e-05, "loss": 0.0195, "step": 11350 }, { "grad_norm": 0.1795470118522644, "learning_rate": 9.166008692978001e-05, "loss": 0.0158, "step": 11360 }, { "grad_norm": 0.28362950682640076, "learning_rate": 9.164179146389158e-05, "loss": 0.0179, "step": 11370 }, { "grad_norm": 0.25175175070762634, "learning_rate": 9.162347778246882e-05, "loss": 0.0204, "step": 11380 }, { "grad_norm": 0.27670517563819885, "learning_rate": 9.160514589352276e-05, "loss": 0.0185, "step": 11390 }, { "grad_norm": 0.23314358294010162, "learning_rate": 9.15867958050724e-05, "loss": 0.0181, "step": 11400 }, { "grad_norm": 0.20065483450889587, "learning_rate": 9.156842752514466e-05, "loss": 0.0198, "step": 11410 }, { "grad_norm": 0.21583005785942078, "learning_rate": 9.155004106177447e-05, "loss": 0.0215, "step": 11420 }, { "grad_norm": 0.26928848028182983, "learning_rate": 9.153163642300471e-05, "loss": 0.0205, "step": 11430 }, { "grad_norm": 0.26945260167121887, "learning_rate": 9.151321361688616e-05, "loss": 0.0189, "step": 11440 }, { "grad_norm": 0.25080248713493347, "learning_rate": 9.149477265147762e-05, "loss": 0.0163, "step": 11450 }, { "grad_norm": 0.19466321170330048, "learning_rate": 9.147631353484574e-05, "loss": 0.0182, "step": 11460 }, { "grad_norm": 0.23434223234653473, "learning_rate": 9.145783627506522e-05, "loss": 0.0192, "step": 11470 }, { "grad_norm": 0.21764545142650604, "learning_rate": 9.143934088021861e-05, "loss": 0.0178, "step": 11480 }, { "grad_norm": 0.16071482002735138, "learning_rate": 9.142082735839645e-05, "loss": 0.0159, "step": 11490 }, { "grad_norm": 0.134405255317688, "learning_rate": 9.140229571769715e-05, "loss": 0.0199, "step": 11500 }, { "grad_norm": 0.22431392967700958, "learning_rate": 9.138374596622709e-05, "loss": 0.0176, "step": 11510 }, { "grad_norm": 0.34809592366218567, "learning_rate": 9.136517811210059e-05, "loss": 0.0183, "step": 11520 }, { "grad_norm": 0.2679460048675537, "learning_rate": 9.134659216343984e-05, "loss": 0.0182, "step": 11530 }, { "grad_norm": 0.28422755002975464, "learning_rate": 9.132798812837494e-05, "loss": 0.02, "step": 11540 }, { "grad_norm": 0.2886165976524353, "learning_rate": 9.130936601504396e-05, "loss": 0.0157, "step": 11550 }, { "grad_norm": 0.22985675930976868, "learning_rate": 9.129072583159284e-05, "loss": 0.0205, "step": 11560 }, { "grad_norm": 0.2613084018230438, "learning_rate": 9.127206758617542e-05, "loss": 0.0192, "step": 11570 }, { "grad_norm": 0.21359777450561523, "learning_rate": 9.125339128695346e-05, "loss": 0.0164, "step": 11580 }, { "grad_norm": 0.18640129268169403, "learning_rate": 9.123469694209659e-05, "loss": 0.0154, "step": 11590 }, { "grad_norm": 0.255043625831604, "learning_rate": 9.121598455978239e-05, "loss": 0.0152, "step": 11600 }, { "grad_norm": 0.24625079333782196, "learning_rate": 9.119725414819624e-05, "loss": 0.0182, "step": 11610 }, { "grad_norm": 0.21849319338798523, "learning_rate": 9.117850571553149e-05, "loss": 0.02, "step": 11620 }, { "grad_norm": 0.24009265005588531, "learning_rate": 9.115973926998935e-05, "loss": 0.0153, "step": 11630 }, { "grad_norm": 0.20689012110233307, "learning_rate": 9.114095481977888e-05, "loss": 0.0164, "step": 11640 }, { "grad_norm": 0.23804868757724762, "learning_rate": 9.112215237311703e-05, "loss": 0.0168, "step": 11650 }, { "grad_norm": 0.253255695104599, "learning_rate": 9.110333193822867e-05, "loss": 0.0158, "step": 11660 }, { "grad_norm": 0.2433309108018875, "learning_rate": 9.108449352334645e-05, "loss": 0.0138, "step": 11670 }, { "grad_norm": 0.19066566228866577, "learning_rate": 9.106563713671094e-05, "loss": 0.0154, "step": 11680 }, { "grad_norm": 0.2008248269557953, "learning_rate": 9.104676278657061e-05, "loss": 0.0204, "step": 11690 }, { "grad_norm": 0.22830387949943542, "learning_rate": 9.102787048118169e-05, "loss": 0.0191, "step": 11700 }, { "grad_norm": 0.25851309299468994, "learning_rate": 9.100896022880834e-05, "loss": 0.0175, "step": 11710 }, { "grad_norm": 0.25062787532806396, "learning_rate": 9.099003203772254e-05, "loss": 0.0185, "step": 11720 }, { "grad_norm": 0.19001345336437225, "learning_rate": 9.097108591620413e-05, "loss": 0.0146, "step": 11730 }, { "grad_norm": 0.14623180031776428, "learning_rate": 9.095212187254078e-05, "loss": 0.0149, "step": 11740 }, { "grad_norm": 0.2551042139530182, "learning_rate": 9.093313991502801e-05, "loss": 0.0158, "step": 11750 }, { "grad_norm": 0.245066300034523, "learning_rate": 9.091414005196917e-05, "loss": 0.0175, "step": 11760 }, { "grad_norm": 0.3145027458667755, "learning_rate": 9.089512229167545e-05, "loss": 0.0162, "step": 11770 }, { "grad_norm": 0.2046465128660202, "learning_rate": 9.087608664246587e-05, "loss": 0.0209, "step": 11780 }, { "grad_norm": 0.1765555590391159, "learning_rate": 9.085703311266727e-05, "loss": 0.0151, "step": 11790 }, { "grad_norm": 0.2854391038417816, "learning_rate": 9.083796171061429e-05, "loss": 0.0216, "step": 11800 }, { "grad_norm": 0.3248160183429718, "learning_rate": 9.081887244464941e-05, "loss": 0.0185, "step": 11810 }, { "grad_norm": 0.24133481085300446, "learning_rate": 9.079976532312297e-05, "loss": 0.0217, "step": 11820 }, { "grad_norm": 0.20805026590824127, "learning_rate": 9.078064035439301e-05, "loss": 0.0178, "step": 11830 }, { "grad_norm": 0.25035861134529114, "learning_rate": 9.07614975468255e-05, "loss": 0.0174, "step": 11840 }, { "grad_norm": 0.24027961492538452, "learning_rate": 9.074233690879412e-05, "loss": 0.0185, "step": 11850 }, { "grad_norm": 0.22835983335971832, "learning_rate": 9.072315844868038e-05, "loss": 0.0189, "step": 11860 }, { "grad_norm": 0.21079373359680176, "learning_rate": 9.07039621748736e-05, "loss": 0.0198, "step": 11870 }, { "grad_norm": 0.24112087488174438, "learning_rate": 9.06847480957709e-05, "loss": 0.0173, "step": 11880 }, { "grad_norm": 0.17388151586055756, "learning_rate": 9.066551621977713e-05, "loss": 0.0177, "step": 11890 }, { "grad_norm": 0.3177785873413086, "learning_rate": 9.064626655530501e-05, "loss": 0.0266, "step": 11900 }, { "grad_norm": 0.2170921266078949, "learning_rate": 9.062699911077497e-05, "loss": 0.0173, "step": 11910 }, { "grad_norm": 0.314103364944458, "learning_rate": 9.060771389461524e-05, "loss": 0.0182, "step": 11920 }, { "grad_norm": 0.2900501787662506, "learning_rate": 9.058841091526187e-05, "loss": 0.0175, "step": 11930 }, { "grad_norm": 0.25115472078323364, "learning_rate": 9.056909018115858e-05, "loss": 0.0163, "step": 11940 }, { "grad_norm": 0.2428686022758484, "learning_rate": 9.054975170075697e-05, "loss": 0.0201, "step": 11950 }, { "grad_norm": 0.22682823240756989, "learning_rate": 9.053039548251631e-05, "loss": 0.0197, "step": 11960 }, { "grad_norm": 0.1916792243719101, "learning_rate": 9.051102153490368e-05, "loss": 0.016, "step": 11970 }, { "grad_norm": 0.1472385972738266, "learning_rate": 9.04916298663939e-05, "loss": 0.0205, "step": 11980 }, { "grad_norm": 0.28521600365638733, "learning_rate": 9.047222048546955e-05, "loss": 0.0173, "step": 11990 }, { "grad_norm": 0.21148289740085602, "learning_rate": 9.045279340062097e-05, "loss": 0.0202, "step": 12000 }, { "grad_norm": 0.2704308331012726, "learning_rate": 9.043334862034618e-05, "loss": 0.0214, "step": 12010 }, { "grad_norm": 0.17110873758792877, "learning_rate": 9.041388615315102e-05, "loss": 0.0185, "step": 12020 }, { "grad_norm": 0.2233845591545105, "learning_rate": 9.039440600754905e-05, "loss": 0.0189, "step": 12030 }, { "grad_norm": 0.306430459022522, "learning_rate": 9.037490819206151e-05, "loss": 0.0201, "step": 12040 }, { "grad_norm": 0.229330375790596, "learning_rate": 9.035539271521744e-05, "loss": 0.017, "step": 12050 }, { "grad_norm": 0.19436146318912506, "learning_rate": 9.033585958555356e-05, "loss": 0.016, "step": 12060 }, { "grad_norm": 0.19887708127498627, "learning_rate": 9.031630881161431e-05, "loss": 0.0174, "step": 12070 }, { "grad_norm": 0.17754781246185303, "learning_rate": 9.029674040195186e-05, "loss": 0.0172, "step": 12080 }, { "grad_norm": 0.2013232558965683, "learning_rate": 9.027715436512613e-05, "loss": 0.0162, "step": 12090 }, { "grad_norm": 0.2513226866722107, "learning_rate": 9.02575507097047e-05, "loss": 0.0177, "step": 12100 }, { "grad_norm": 0.20022538304328918, "learning_rate": 9.023792944426286e-05, "loss": 0.0165, "step": 12110 }, { "grad_norm": 0.22361764311790466, "learning_rate": 9.021829057738364e-05, "loss": 0.0192, "step": 12120 }, { "grad_norm": 0.22778363525867462, "learning_rate": 9.019863411765775e-05, "loss": 0.0164, "step": 12130 }, { "grad_norm": 0.21175678074359894, "learning_rate": 9.017896007368357e-05, "loss": 0.0143, "step": 12140 }, { "grad_norm": 0.27405425906181335, "learning_rate": 9.015926845406722e-05, "loss": 0.0175, "step": 12150 }, { "grad_norm": 0.24207757413387299, "learning_rate": 9.013955926742245e-05, "loss": 0.0189, "step": 12160 }, { "grad_norm": 0.2595878541469574, "learning_rate": 9.011983252237077e-05, "loss": 0.0192, "step": 12170 }, { "grad_norm": 0.1792474389076233, "learning_rate": 9.01000882275413e-05, "loss": 0.0158, "step": 12180 }, { "grad_norm": 0.21619385480880737, "learning_rate": 9.008032639157088e-05, "loss": 0.0163, "step": 12190 }, { "grad_norm": 0.255868136882782, "learning_rate": 9.006054702310401e-05, "loss": 0.0192, "step": 12200 }, { "grad_norm": 0.19579842686653137, "learning_rate": 9.004075013079283e-05, "loss": 0.0154, "step": 12210 }, { "grad_norm": 0.2118227481842041, "learning_rate": 9.00209357232972e-05, "loss": 0.017, "step": 12220 }, { "grad_norm": 0.2707882523536682, "learning_rate": 9.000110380928461e-05, "loss": 0.0148, "step": 12230 }, { "grad_norm": 0.2244562953710556, "learning_rate": 8.998125439743021e-05, "loss": 0.016, "step": 12240 }, { "grad_norm": 0.2366371750831604, "learning_rate": 8.996138749641682e-05, "loss": 0.0183, "step": 12250 }, { "grad_norm": 0.20890265703201294, "learning_rate": 8.994150311493488e-05, "loss": 0.0171, "step": 12260 }, { "grad_norm": 0.21935853362083435, "learning_rate": 8.992160126168247e-05, "loss": 0.0198, "step": 12270 }, { "grad_norm": 0.1801203191280365, "learning_rate": 8.99016819453654e-05, "loss": 0.0174, "step": 12280 }, { "grad_norm": 0.21536654233932495, "learning_rate": 8.988174517469702e-05, "loss": 0.0165, "step": 12290 }, { "grad_norm": 0.24891389906406403, "learning_rate": 8.986179095839835e-05, "loss": 0.0182, "step": 12300 }, { "grad_norm": 0.3325229585170746, "learning_rate": 8.984181930519804e-05, "loss": 0.0195, "step": 12310 }, { "grad_norm": 0.23842647671699524, "learning_rate": 8.982183022383237e-05, "loss": 0.017, "step": 12320 }, { "grad_norm": 0.1979597508907318, "learning_rate": 8.980182372304525e-05, "loss": 0.0181, "step": 12330 }, { "grad_norm": 0.2200346738100052, "learning_rate": 8.97817998115882e-05, "loss": 0.0206, "step": 12340 }, { "grad_norm": 0.19707956910133362, "learning_rate": 8.976175849822038e-05, "loss": 0.0173, "step": 12350 }, { "grad_norm": 0.23864668607711792, "learning_rate": 8.97416997917085e-05, "loss": 0.0235, "step": 12360 }, { "grad_norm": 0.21148669719696045, "learning_rate": 8.972162370082695e-05, "loss": 0.0179, "step": 12370 }, { "grad_norm": 0.2258552759885788, "learning_rate": 8.97015302343577e-05, "loss": 0.0181, "step": 12380 }, { "grad_norm": 0.23386920988559723, "learning_rate": 8.968141940109027e-05, "loss": 0.0167, "step": 12390 }, { "grad_norm": 0.19266775250434875, "learning_rate": 8.966129120982188e-05, "loss": 0.0193, "step": 12400 }, { "grad_norm": 0.23731395602226257, "learning_rate": 8.964114566935724e-05, "loss": 0.0168, "step": 12410 }, { "grad_norm": 0.14120477437973022, "learning_rate": 8.962098278850871e-05, "loss": 0.0156, "step": 12420 }, { "grad_norm": 0.18466182053089142, "learning_rate": 8.960080257609622e-05, "loss": 0.0131, "step": 12430 }, { "grad_norm": 0.24158520996570587, "learning_rate": 8.95806050409473e-05, "loss": 0.0156, "step": 12440 }, { "grad_norm": 0.23820075392723083, "learning_rate": 8.9560390191897e-05, "loss": 0.0232, "step": 12450 }, { "grad_norm": 0.25189337134361267, "learning_rate": 8.9540158037788e-05, "loss": 0.0164, "step": 12460 }, { "grad_norm": 0.2547222673892975, "learning_rate": 8.951990858747054e-05, "loss": 0.0205, "step": 12470 }, { "grad_norm": 0.2187165915966034, "learning_rate": 8.94996418498024e-05, "loss": 0.0155, "step": 12480 }, { "grad_norm": 0.23191112279891968, "learning_rate": 8.947935783364896e-05, "loss": 0.0164, "step": 12490 }, { "grad_norm": 0.2584964632987976, "learning_rate": 8.945905654788311e-05, "loss": 0.0159, "step": 12500 }, { "grad_norm": 0.17940166592597961, "learning_rate": 8.943873800138535e-05, "loss": 0.0144, "step": 12510 }, { "grad_norm": 0.23682662844657898, "learning_rate": 8.94184022030437e-05, "loss": 0.0166, "step": 12520 }, { "grad_norm": 0.20393379032611847, "learning_rate": 8.939804916175372e-05, "loss": 0.0182, "step": 12530 }, { "grad_norm": 0.275746613740921, "learning_rate": 8.93776788864185e-05, "loss": 0.017, "step": 12540 }, { "grad_norm": 0.25151345133781433, "learning_rate": 8.935729138594873e-05, "loss": 0.0162, "step": 12550 }, { "grad_norm": 0.2015291005373001, "learning_rate": 8.933688666926258e-05, "loss": 0.0148, "step": 12560 }, { "grad_norm": 0.20738251507282257, "learning_rate": 8.931646474528575e-05, "loss": 0.0151, "step": 12570 }, { "grad_norm": 0.23546241223812103, "learning_rate": 8.929602562295151e-05, "loss": 0.0149, "step": 12580 }, { "grad_norm": 0.15644299983978271, "learning_rate": 8.92755693112006e-05, "loss": 0.015, "step": 12590 }, { "grad_norm": 0.2271387279033661, "learning_rate": 8.925509581898136e-05, "loss": 0.016, "step": 12600 }, { "grad_norm": 0.19471533596515656, "learning_rate": 8.923460515524951e-05, "loss": 0.0155, "step": 12610 }, { "grad_norm": 0.21621279418468475, "learning_rate": 8.921409732896842e-05, "loss": 0.018, "step": 12620 }, { "grad_norm": 0.18788500130176544, "learning_rate": 8.919357234910887e-05, "loss": 0.0149, "step": 12630 }, { "grad_norm": 0.25758692622184753, "learning_rate": 8.917303022464923e-05, "loss": 0.0156, "step": 12640 }, { "grad_norm": 0.24748460948467255, "learning_rate": 8.915247096457531e-05, "loss": 0.016, "step": 12650 }, { "grad_norm": 0.18178975582122803, "learning_rate": 8.91318945778804e-05, "loss": 0.0127, "step": 12660 }, { "grad_norm": 0.21342265605926514, "learning_rate": 8.911130107356534e-05, "loss": 0.0123, "step": 12670 }, { "grad_norm": 0.26807481050491333, "learning_rate": 8.90906904606384e-05, "loss": 0.0163, "step": 12680 }, { "grad_norm": 0.2555883824825287, "learning_rate": 8.90700627481154e-05, "loss": 0.017, "step": 12690 }, { "grad_norm": 0.22883762419223785, "learning_rate": 8.904941794501957e-05, "loss": 0.0145, "step": 12700 }, { "grad_norm": 0.18842226266860962, "learning_rate": 8.902875606038166e-05, "loss": 0.0151, "step": 12710 }, { "grad_norm": 0.23377849161624908, "learning_rate": 8.900807710323989e-05, "loss": 0.0165, "step": 12720 }, { "grad_norm": 0.23490281403064728, "learning_rate": 8.898738108263993e-05, "loss": 0.0153, "step": 12730 }, { "grad_norm": 0.21357516944408417, "learning_rate": 8.896666800763491e-05, "loss": 0.015, "step": 12740 }, { "grad_norm": 0.1871136575937271, "learning_rate": 8.894593788728546e-05, "loss": 0.0156, "step": 12750 }, { "grad_norm": 0.22777219116687775, "learning_rate": 8.892519073065961e-05, "loss": 0.0167, "step": 12760 }, { "grad_norm": 0.19241909682750702, "learning_rate": 8.89044265468329e-05, "loss": 0.0175, "step": 12770 }, { "grad_norm": 0.25077682733535767, "learning_rate": 8.888364534488827e-05, "loss": 0.019, "step": 12780 }, { "grad_norm": 0.20531544089317322, "learning_rate": 8.886284713391613e-05, "loss": 0.0161, "step": 12790 }, { "grad_norm": 0.2811039388179779, "learning_rate": 8.884203192301431e-05, "loss": 0.017, "step": 12800 }, { "grad_norm": 0.2918158173561096, "learning_rate": 8.88211997212881e-05, "loss": 0.0149, "step": 12810 }, { "grad_norm": 0.28741735219955444, "learning_rate": 8.880035053785023e-05, "loss": 0.0174, "step": 12820 }, { "grad_norm": 0.22104701399803162, "learning_rate": 8.877948438182083e-05, "loss": 0.0158, "step": 12830 }, { "grad_norm": 0.22504712641239166, "learning_rate": 8.875860126232745e-05, "loss": 0.0158, "step": 12840 }, { "grad_norm": 0.25644367933273315, "learning_rate": 8.87377011885051e-05, "loss": 0.016, "step": 12850 }, { "grad_norm": 0.2318536937236786, "learning_rate": 8.871678416949617e-05, "loss": 0.018, "step": 12860 }, { "grad_norm": 0.278922438621521, "learning_rate": 8.869585021445046e-05, "loss": 0.0156, "step": 12870 }, { "grad_norm": 0.17995630204677582, "learning_rate": 8.867489933252521e-05, "loss": 0.0147, "step": 12880 }, { "grad_norm": 0.21339863538742065, "learning_rate": 8.865393153288504e-05, "loss": 0.018, "step": 12890 }, { "grad_norm": 0.1941353678703308, "learning_rate": 8.8632946824702e-05, "loss": 0.0164, "step": 12900 }, { "grad_norm": 0.21743018925189972, "learning_rate": 8.86119452171555e-05, "loss": 0.0145, "step": 12910 }, { "grad_norm": 0.1750767081975937, "learning_rate": 8.859092671943234e-05, "loss": 0.016, "step": 12920 }, { "grad_norm": 0.21684449911117554, "learning_rate": 8.856989134072676e-05, "loss": 0.0157, "step": 12930 }, { "grad_norm": 0.2288552224636078, "learning_rate": 8.85488390902403e-05, "loss": 0.0175, "step": 12940 }, { "grad_norm": 0.19532908499240875, "learning_rate": 8.852776997718199e-05, "loss": 0.0153, "step": 12950 }, { "grad_norm": 0.23534490168094635, "learning_rate": 8.850668401076812e-05, "loss": 0.0169, "step": 12960 }, { "grad_norm": 0.19093267619609833, "learning_rate": 8.848558120022246e-05, "loss": 0.0144, "step": 12970 }, { "grad_norm": 0.1819402575492859, "learning_rate": 8.846446155477603e-05, "loss": 0.0167, "step": 12980 }, { "grad_norm": 0.268886536359787, "learning_rate": 8.844332508366735e-05, "loss": 0.0225, "step": 12990 }, { "grad_norm": 0.2621830999851227, "learning_rate": 8.84221717961422e-05, "loss": 0.0152, "step": 13000 }, { "grad_norm": 0.17516319453716278, "learning_rate": 8.840100170145374e-05, "loss": 0.0158, "step": 13010 }, { "grad_norm": 0.18356585502624512, "learning_rate": 8.837981480886249e-05, "loss": 0.0158, "step": 13020 }, { "grad_norm": 0.19135905802249908, "learning_rate": 8.835861112763633e-05, "loss": 0.0158, "step": 13030 }, { "grad_norm": 0.1786131113767624, "learning_rate": 8.833739066705044e-05, "loss": 0.0205, "step": 13040 }, { "grad_norm": 0.23772183060646057, "learning_rate": 8.831615343638742e-05, "loss": 0.0181, "step": 13050 }, { "grad_norm": 0.25261181592941284, "learning_rate": 8.829489944493711e-05, "loss": 0.0158, "step": 13060 }, { "grad_norm": 0.2413497418165207, "learning_rate": 8.827362870199675e-05, "loss": 0.0145, "step": 13070 }, { "grad_norm": 0.23130391538143158, "learning_rate": 8.825234121687089e-05, "loss": 0.0139, "step": 13080 }, { "grad_norm": 0.2268231213092804, "learning_rate": 8.823103699887139e-05, "loss": 0.0148, "step": 13090 }, { "grad_norm": 0.22658002376556396, "learning_rate": 8.820971605731745e-05, "loss": 0.0127, "step": 13100 }, { "grad_norm": 0.2857726514339447, "learning_rate": 8.818837840153556e-05, "loss": 0.0172, "step": 13110 }, { "grad_norm": 0.2606637179851532, "learning_rate": 8.816702404085952e-05, "loss": 0.0153, "step": 13120 }, { "grad_norm": 0.21300466358661652, "learning_rate": 8.814565298463048e-05, "loss": 0.018, "step": 13130 }, { "grad_norm": 0.20583316683769226, "learning_rate": 8.812426524219688e-05, "loss": 0.0178, "step": 13140 }, { "grad_norm": 0.21922001242637634, "learning_rate": 8.81028608229144e-05, "loss": 0.0195, "step": 13150 }, { "grad_norm": 0.253351092338562, "learning_rate": 8.808143973614611e-05, "loss": 0.0163, "step": 13160 }, { "grad_norm": 0.2331327199935913, "learning_rate": 8.806000199126228e-05, "loss": 0.0186, "step": 13170 }, { "grad_norm": 0.2720649540424347, "learning_rate": 8.803854759764052e-05, "loss": 0.0147, "step": 13180 }, { "grad_norm": 0.2434234917163849, "learning_rate": 8.801707656466572e-05, "loss": 0.0171, "step": 13190 }, { "grad_norm": 0.24383078515529633, "learning_rate": 8.799558890173003e-05, "loss": 0.0178, "step": 13200 }, { "grad_norm": 0.2232900708913803, "learning_rate": 8.79740846182329e-05, "loss": 0.0161, "step": 13210 }, { "grad_norm": 0.23591405153274536, "learning_rate": 8.7952563723581e-05, "loss": 0.0174, "step": 13220 }, { "grad_norm": 0.22698666155338287, "learning_rate": 8.793102622718834e-05, "loss": 0.0188, "step": 13230 }, { "grad_norm": 0.22234176099300385, "learning_rate": 8.790947213847613e-05, "loss": 0.0188, "step": 13240 }, { "grad_norm": 0.1794581264257431, "learning_rate": 8.788790146687286e-05, "loss": 0.0146, "step": 13250 }, { "grad_norm": 0.28435543179512024, "learning_rate": 8.786631422181429e-05, "loss": 0.0164, "step": 13260 }, { "grad_norm": 0.1640920341014862, "learning_rate": 8.78447104127434e-05, "loss": 0.0139, "step": 13270 }, { "grad_norm": 0.1934587061405182, "learning_rate": 8.782309004911042e-05, "loss": 0.0155, "step": 13280 }, { "grad_norm": 0.18053412437438965, "learning_rate": 8.780145314037286e-05, "loss": 0.0132, "step": 13290 }, { "grad_norm": 0.2604879140853882, "learning_rate": 8.777979969599542e-05, "loss": 0.0164, "step": 13300 }, { "grad_norm": 0.1752013862133026, "learning_rate": 8.775812972545006e-05, "loss": 0.0152, "step": 13310 }, { "grad_norm": 0.2255670130252838, "learning_rate": 8.773644323821596e-05, "loss": 0.0149, "step": 13320 }, { "grad_norm": 0.19538767635822296, "learning_rate": 8.771474024377953e-05, "loss": 0.0155, "step": 13330 }, { "grad_norm": 0.2076835185289383, "learning_rate": 8.769302075163438e-05, "loss": 0.0141, "step": 13340 }, { "grad_norm": 0.20142938196659088, "learning_rate": 8.767128477128137e-05, "loss": 0.0137, "step": 13350 }, { "grad_norm": 0.24773965775966644, "learning_rate": 8.764953231222854e-05, "loss": 0.0151, "step": 13360 }, { "grad_norm": 0.22969695925712585, "learning_rate": 8.762776338399119e-05, "loss": 0.0157, "step": 13370 }, { "grad_norm": 0.2214985191822052, "learning_rate": 8.760597799609176e-05, "loss": 0.0148, "step": 13380 }, { "grad_norm": 0.22393836081027985, "learning_rate": 8.758417615805992e-05, "loss": 0.0134, "step": 13390 }, { "grad_norm": 0.1673772782087326, "learning_rate": 8.756235787943254e-05, "loss": 0.0155, "step": 13400 }, { "grad_norm": 0.16354836523532867, "learning_rate": 8.754052316975367e-05, "loss": 0.0138, "step": 13410 }, { "grad_norm": 0.18902483582496643, "learning_rate": 8.751867203857455e-05, "loss": 0.0148, "step": 13420 }, { "grad_norm": 0.2009773999452591, "learning_rate": 8.749680449545363e-05, "loss": 0.0169, "step": 13430 }, { "grad_norm": 0.2766724228858948, "learning_rate": 8.747492054995649e-05, "loss": 0.0183, "step": 13440 }, { "grad_norm": 0.18602275848388672, "learning_rate": 8.745302021165595e-05, "loss": 0.017, "step": 13450 }, { "grad_norm": 0.24693943560123444, "learning_rate": 8.743110349013192e-05, "loss": 0.0177, "step": 13460 }, { "grad_norm": 0.1863868534564972, "learning_rate": 8.740917039497153e-05, "loss": 0.0201, "step": 13470 }, { "grad_norm": 0.210929736495018, "learning_rate": 8.738722093576906e-05, "loss": 0.0165, "step": 13480 }, { "grad_norm": 0.3053784966468811, "learning_rate": 8.736525512212597e-05, "loss": 0.0144, "step": 13490 }, { "grad_norm": 0.18047882616519928, "learning_rate": 8.734327296365084e-05, "loss": 0.0138, "step": 13500 }, { "grad_norm": 0.1734144389629364, "learning_rate": 8.732127446995939e-05, "loss": 0.016, "step": 13510 }, { "grad_norm": 0.2090451866388321, "learning_rate": 8.729925965067454e-05, "loss": 0.0132, "step": 13520 }, { "grad_norm": 0.18955454230308533, "learning_rate": 8.72772285154263e-05, "loss": 0.0132, "step": 13530 }, { "grad_norm": 0.1496249884366989, "learning_rate": 8.725518107385187e-05, "loss": 0.0138, "step": 13540 }, { "grad_norm": 0.21354801952838898, "learning_rate": 8.72331173355955e-05, "loss": 0.0143, "step": 13550 }, { "grad_norm": 0.19274447858333588, "learning_rate": 8.721103731030867e-05, "loss": 0.0186, "step": 13560 }, { "grad_norm": 0.30528634786605835, "learning_rate": 8.718894100764989e-05, "loss": 0.0153, "step": 13570 }, { "grad_norm": 0.29635849595069885, "learning_rate": 8.716682843728485e-05, "loss": 0.0144, "step": 13580 }, { "grad_norm": 0.2386879324913025, "learning_rate": 8.714469960888634e-05, "loss": 0.0161, "step": 13590 }, { "grad_norm": 0.20125404000282288, "learning_rate": 8.712255453213427e-05, "loss": 0.0133, "step": 13600 }, { "grad_norm": 0.2016288936138153, "learning_rate": 8.710039321671563e-05, "loss": 0.0164, "step": 13610 }, { "grad_norm": 0.24475125968456268, "learning_rate": 8.707821567232456e-05, "loss": 0.0152, "step": 13620 }, { "grad_norm": 0.14486651122570038, "learning_rate": 8.705602190866225e-05, "loss": 0.016, "step": 13630 }, { "grad_norm": 0.15197031199932098, "learning_rate": 8.703381193543701e-05, "loss": 0.0155, "step": 13640 }, { "grad_norm": 0.23353761434555054, "learning_rate": 8.701158576236423e-05, "loss": 0.0152, "step": 13650 }, { "grad_norm": 0.200966939330101, "learning_rate": 8.69893433991664e-05, "loss": 0.0129, "step": 13660 }, { "grad_norm": 0.20917513966560364, "learning_rate": 8.69670848555731e-05, "loss": 0.0157, "step": 13670 }, { "grad_norm": 0.22075428068637848, "learning_rate": 8.694481014132096e-05, "loss": 0.0155, "step": 13680 }, { "grad_norm": 0.2797356843948364, "learning_rate": 8.69225192661537e-05, "loss": 0.0172, "step": 13690 }, { "grad_norm": 0.21685753762722015, "learning_rate": 8.690021223982208e-05, "loss": 0.0151, "step": 13700 }, { "grad_norm": 0.18821197748184204, "learning_rate": 8.687788907208398e-05, "loss": 0.0165, "step": 13710 }, { "grad_norm": 0.21577297151088715, "learning_rate": 8.685554977270431e-05, "loss": 0.0174, "step": 13720 }, { "grad_norm": 0.23349544405937195, "learning_rate": 8.683319435145503e-05, "loss": 0.0151, "step": 13730 }, { "grad_norm": 0.2419474869966507, "learning_rate": 8.681082281811517e-05, "loss": 0.0174, "step": 13740 }, { "grad_norm": 0.2709105908870697, "learning_rate": 8.67884351824708e-05, "loss": 0.0129, "step": 13750 }, { "grad_norm": 0.21233297884464264, "learning_rate": 8.676603145431501e-05, "loss": 0.0147, "step": 13760 }, { "grad_norm": 0.17243273556232452, "learning_rate": 8.674361164344799e-05, "loss": 0.0153, "step": 13770 }, { "grad_norm": 0.2119709700345993, "learning_rate": 8.672117575967688e-05, "loss": 0.018, "step": 13780 }, { "grad_norm": 0.2274838089942932, "learning_rate": 8.669872381281595e-05, "loss": 0.0155, "step": 13790 }, { "grad_norm": 0.23185117542743683, "learning_rate": 8.667625581268639e-05, "loss": 0.0147, "step": 13800 }, { "grad_norm": 0.1754036247730255, "learning_rate": 8.665377176911651e-05, "loss": 0.0138, "step": 13810 }, { "grad_norm": 0.1926077902317047, "learning_rate": 8.663127169194159e-05, "loss": 0.0153, "step": 13820 }, { "grad_norm": 0.18572399020195007, "learning_rate": 8.660875559100389e-05, "loss": 0.0132, "step": 13830 }, { "grad_norm": 0.20062310993671417, "learning_rate": 8.658622347615274e-05, "loss": 0.0141, "step": 13840 }, { "grad_norm": 0.20048755407333374, "learning_rate": 8.656367535724448e-05, "loss": 0.016, "step": 13850 }, { "grad_norm": 0.25398576259613037, "learning_rate": 8.65411112441424e-05, "loss": 0.0177, "step": 13860 }, { "grad_norm": 0.15257631242275238, "learning_rate": 8.651853114671679e-05, "loss": 0.0157, "step": 13870 }, { "grad_norm": 0.20968742668628693, "learning_rate": 8.649593507484499e-05, "loss": 0.0157, "step": 13880 }, { "grad_norm": 0.16515901684761047, "learning_rate": 8.647332303841126e-05, "loss": 0.0145, "step": 13890 }, { "grad_norm": 0.16739679872989655, "learning_rate": 8.645069504730689e-05, "loss": 0.016, "step": 13900 }, { "grad_norm": 0.21535839140415192, "learning_rate": 8.64280511114301e-05, "loss": 0.0125, "step": 13910 }, { "grad_norm": 0.28381016850471497, "learning_rate": 8.640539124068617e-05, "loss": 0.0145, "step": 13920 }, { "grad_norm": 0.18303322792053223, "learning_rate": 8.638271544498727e-05, "loss": 0.015, "step": 13930 }, { "grad_norm": 0.22304949164390564, "learning_rate": 8.636002373425257e-05, "loss": 0.0142, "step": 13940 }, { "grad_norm": 0.19821374118328094, "learning_rate": 8.633731611840817e-05, "loss": 0.0163, "step": 13950 }, { "grad_norm": 0.24977527558803558, "learning_rate": 8.631459260738717e-05, "loss": 0.0143, "step": 13960 }, { "grad_norm": 0.19390562176704407, "learning_rate": 8.62918532111296e-05, "loss": 0.0148, "step": 13970 }, { "grad_norm": 0.21048428118228912, "learning_rate": 8.626909793958248e-05, "loss": 0.0169, "step": 13980 }, { "grad_norm": 0.2365146279335022, "learning_rate": 8.624632680269969e-05, "loss": 0.018, "step": 13990 }, { "grad_norm": 0.1944269835948944, "learning_rate": 8.622353981044212e-05, "loss": 0.0203, "step": 14000 }, { "grad_norm": 0.1678372472524643, "learning_rate": 8.620073697277757e-05, "loss": 0.0127, "step": 14010 }, { "grad_norm": 0.18243327736854553, "learning_rate": 8.617791829968079e-05, "loss": 0.0137, "step": 14020 }, { "grad_norm": 0.25175967812538147, "learning_rate": 8.615508380113344e-05, "loss": 0.0125, "step": 14030 }, { "grad_norm": 0.22590848803520203, "learning_rate": 8.613223348712408e-05, "loss": 0.0153, "step": 14040 }, { "grad_norm": 0.2392529845237732, "learning_rate": 8.610936736764824e-05, "loss": 0.0171, "step": 14050 }, { "grad_norm": 0.20959992706775665, "learning_rate": 8.608648545270833e-05, "loss": 0.0158, "step": 14060 }, { "grad_norm": 0.18744762241840363, "learning_rate": 8.606358775231366e-05, "loss": 0.0162, "step": 14070 }, { "grad_norm": 0.24289417266845703, "learning_rate": 8.60406742764805e-05, "loss": 0.014, "step": 14080 }, { "grad_norm": 0.2292727530002594, "learning_rate": 8.601774503523195e-05, "loss": 0.0162, "step": 14090 }, { "grad_norm": 0.15884733200073242, "learning_rate": 8.599480003859805e-05, "loss": 0.0148, "step": 14100 }, { "grad_norm": 0.17099329829216003, "learning_rate": 8.597183929661573e-05, "loss": 0.014, "step": 14110 }, { "grad_norm": 0.13494643568992615, "learning_rate": 8.594886281932879e-05, "loss": 0.0138, "step": 14120 }, { "grad_norm": 0.23635946214199066, "learning_rate": 8.59258706167879e-05, "loss": 0.0145, "step": 14130 }, { "grad_norm": 0.15357069671154022, "learning_rate": 8.590286269905068e-05, "loss": 0.0122, "step": 14140 }, { "grad_norm": 0.2180400788784027, "learning_rate": 8.587983907618154e-05, "loss": 0.0161, "step": 14150 }, { "grad_norm": 0.21714597940444946, "learning_rate": 8.585679975825178e-05, "loss": 0.0165, "step": 14160 }, { "grad_norm": 0.3220981955528259, "learning_rate": 8.583374475533962e-05, "loss": 0.0155, "step": 14170 }, { "grad_norm": 0.1914060115814209, "learning_rate": 8.581067407753009e-05, "loss": 0.018, "step": 14180 }, { "grad_norm": 0.20937086641788483, "learning_rate": 8.578758773491507e-05, "loss": 0.0153, "step": 14190 }, { "grad_norm": 0.19443756341934204, "learning_rate": 8.576448573759332e-05, "loss": 0.0163, "step": 14200 }, { "grad_norm": 0.16645048558712006, "learning_rate": 8.574136809567044e-05, "loss": 0.0107, "step": 14210 }, { "grad_norm": 0.23980863392353058, "learning_rate": 8.57182348192589e-05, "loss": 0.0131, "step": 14220 }, { "grad_norm": 0.25855135917663574, "learning_rate": 8.569508591847792e-05, "loss": 0.0141, "step": 14230 }, { "grad_norm": 0.21802735328674316, "learning_rate": 8.567192140345367e-05, "loss": 0.0171, "step": 14240 }, { "grad_norm": 0.21974194049835205, "learning_rate": 8.564874128431906e-05, "loss": 0.0162, "step": 14250 }, { "grad_norm": 0.25683102011680603, "learning_rate": 8.562554557121389e-05, "loss": 0.014, "step": 14260 }, { "grad_norm": 0.2032209038734436, "learning_rate": 8.560233427428475e-05, "loss": 0.0154, "step": 14270 }, { "grad_norm": 0.1902657151222229, "learning_rate": 8.557910740368503e-05, "loss": 0.0149, "step": 14280 }, { "grad_norm": 0.21987296640872955, "learning_rate": 8.555586496957495e-05, "loss": 0.015, "step": 14290 }, { "grad_norm": 0.18675370514392853, "learning_rate": 8.553260698212155e-05, "loss": 0.0186, "step": 14300 }, { "grad_norm": 0.1780502051115036, "learning_rate": 8.550933345149868e-05, "loss": 0.0137, "step": 14310 }, { "grad_norm": 0.19038142263889313, "learning_rate": 8.548604438788696e-05, "loss": 0.0157, "step": 14320 }, { "grad_norm": 0.15622252225875854, "learning_rate": 8.546273980147383e-05, "loss": 0.0154, "step": 14330 }, { "grad_norm": 0.21982739865779877, "learning_rate": 8.543941970245348e-05, "loss": 0.0129, "step": 14340 }, { "grad_norm": 0.180136576294899, "learning_rate": 8.541608410102693e-05, "loss": 0.0148, "step": 14350 }, { "grad_norm": 0.20178085565567017, "learning_rate": 8.539273300740195e-05, "loss": 0.0145, "step": 14360 }, { "grad_norm": 0.17430616915225983, "learning_rate": 8.536936643179313e-05, "loss": 0.012, "step": 14370 }, { "grad_norm": 0.1889248788356781, "learning_rate": 8.534598438442179e-05, "loss": 0.0123, "step": 14380 }, { "grad_norm": 0.22322812676429749, "learning_rate": 8.532258687551603e-05, "loss": 0.0167, "step": 14390 }, { "grad_norm": 0.24100732803344727, "learning_rate": 8.529917391531071e-05, "loss": 0.0144, "step": 14400 }, { "grad_norm": 0.24971657991409302, "learning_rate": 8.527574551404747e-05, "loss": 0.016, "step": 14410 }, { "grad_norm": 0.19462698698043823, "learning_rate": 8.525230168197468e-05, "loss": 0.015, "step": 14420 }, { "grad_norm": 0.20197591185569763, "learning_rate": 8.522884242934745e-05, "loss": 0.0127, "step": 14430 }, { "grad_norm": 0.21552525460720062, "learning_rate": 8.520536776642768e-05, "loss": 0.0145, "step": 14440 }, { "grad_norm": 0.23887455463409424, "learning_rate": 8.5181877703484e-05, "loss": 0.015, "step": 14450 }, { "grad_norm": 0.20110422372817993, "learning_rate": 8.51583722507917e-05, "loss": 0.0144, "step": 14460 }, { "grad_norm": 0.21691909432411194, "learning_rate": 8.513485141863293e-05, "loss": 0.0171, "step": 14470 }, { "grad_norm": 0.17417795956134796, "learning_rate": 8.511131521729647e-05, "loss": 0.0134, "step": 14480 }, { "grad_norm": 0.19722072780132294, "learning_rate": 8.508776365707787e-05, "loss": 0.0148, "step": 14490 }, { "grad_norm": 0.2367250919342041, "learning_rate": 8.506419674827934e-05, "loss": 0.0142, "step": 14500 }, { "grad_norm": 0.38434094190597534, "learning_rate": 8.50406145012099e-05, "loss": 0.0166, "step": 14510 }, { "grad_norm": 0.20360474288463593, "learning_rate": 8.501701692618519e-05, "loss": 0.0142, "step": 14520 }, { "grad_norm": 0.2321837991476059, "learning_rate": 8.499340403352761e-05, "loss": 0.0148, "step": 14530 }, { "grad_norm": 0.22018307447433472, "learning_rate": 8.496977583356623e-05, "loss": 0.0179, "step": 14540 }, { "grad_norm": 0.1777219921350479, "learning_rate": 8.494613233663684e-05, "loss": 0.0136, "step": 14550 }, { "grad_norm": 0.1491110771894455, "learning_rate": 8.492247355308189e-05, "loss": 0.014, "step": 14560 }, { "grad_norm": 0.18252411484718323, "learning_rate": 8.489879949325056e-05, "loss": 0.0162, "step": 14570 }, { "grad_norm": 0.24680772423744202, "learning_rate": 8.487511016749868e-05, "loss": 0.0151, "step": 14580 }, { "grad_norm": 0.23217321932315826, "learning_rate": 8.485140558618874e-05, "loss": 0.0142, "step": 14590 }, { "grad_norm": 0.17018473148345947, "learning_rate": 8.482768575968995e-05, "loss": 0.013, "step": 14600 }, { "grad_norm": 0.15557672083377838, "learning_rate": 8.480395069837818e-05, "loss": 0.0144, "step": 14610 }, { "grad_norm": 0.2567611336708069, "learning_rate": 8.478020041263595e-05, "loss": 0.0163, "step": 14620 }, { "grad_norm": 0.22482895851135254, "learning_rate": 8.475643491285242e-05, "loss": 0.0164, "step": 14630 }, { "grad_norm": 0.2013397067785263, "learning_rate": 8.473265420942345e-05, "loss": 0.0151, "step": 14640 }, { "grad_norm": 0.22968493402004242, "learning_rate": 8.470885831275151e-05, "loss": 0.0127, "step": 14650 }, { "grad_norm": 0.168733611702919, "learning_rate": 8.468504723324574e-05, "loss": 0.0174, "step": 14660 }, { "grad_norm": 0.2146560549736023, "learning_rate": 8.466122098132193e-05, "loss": 0.0137, "step": 14670 }, { "grad_norm": 0.14688140153884888, "learning_rate": 8.463737956740245e-05, "loss": 0.015, "step": 14680 }, { "grad_norm": 0.20747263729572296, "learning_rate": 8.461352300191639e-05, "loss": 0.0145, "step": 14690 }, { "grad_norm": 0.1828698068857193, "learning_rate": 8.45896512952994e-05, "loss": 0.0148, "step": 14700 }, { "grad_norm": 0.1762228161096573, "learning_rate": 8.456576445799377e-05, "loss": 0.0112, "step": 14710 }, { "grad_norm": 0.20817849040031433, "learning_rate": 8.454186250044844e-05, "loss": 0.0145, "step": 14720 }, { "grad_norm": 0.22076599299907684, "learning_rate": 8.451794543311892e-05, "loss": 0.0145, "step": 14730 }, { "grad_norm": 0.19213898479938507, "learning_rate": 8.449401326646736e-05, "loss": 0.017, "step": 14740 }, { "grad_norm": 0.15518803894519806, "learning_rate": 8.447006601096248e-05, "loss": 0.0141, "step": 14750 }, { "grad_norm": 0.21783936023712158, "learning_rate": 8.444610367707964e-05, "loss": 0.0122, "step": 14760 }, { "grad_norm": 0.14661931991577148, "learning_rate": 8.442212627530078e-05, "loss": 0.0143, "step": 14770 }, { "grad_norm": 0.1741727739572525, "learning_rate": 8.439813381611441e-05, "loss": 0.0127, "step": 14780 }, { "grad_norm": 0.2288089245557785, "learning_rate": 8.437412631001567e-05, "loss": 0.0146, "step": 14790 }, { "grad_norm": 0.20780333876609802, "learning_rate": 8.435010376750626e-05, "loss": 0.0127, "step": 14800 }, { "grad_norm": 0.22321802377700806, "learning_rate": 8.432606619909442e-05, "loss": 0.0119, "step": 14810 }, { "grad_norm": 0.21799057722091675, "learning_rate": 8.430201361529506e-05, "loss": 0.0145, "step": 14820 }, { "grad_norm": 0.1665000021457672, "learning_rate": 8.427794602662954e-05, "loss": 0.0141, "step": 14830 }, { "grad_norm": 0.15132303535938263, "learning_rate": 8.425386344362586e-05, "loss": 0.0139, "step": 14840 }, { "grad_norm": 0.20641767978668213, "learning_rate": 8.422976587681859e-05, "loss": 0.0137, "step": 14850 }, { "grad_norm": 0.15357741713523865, "learning_rate": 8.42056533367488e-05, "loss": 0.0112, "step": 14860 }, { "grad_norm": 0.17750313878059387, "learning_rate": 8.41815258339641e-05, "loss": 0.0114, "step": 14870 }, { "grad_norm": 0.23370198905467987, "learning_rate": 8.415738337901874e-05, "loss": 0.0148, "step": 14880 }, { "grad_norm": 0.17446325719356537, "learning_rate": 8.413322598247342e-05, "loss": 0.0145, "step": 14890 }, { "grad_norm": 0.2266729176044464, "learning_rate": 8.41090536548954e-05, "loss": 0.0131, "step": 14900 }, { "grad_norm": 0.2071094959974289, "learning_rate": 8.408486640685849e-05, "loss": 0.0129, "step": 14910 }, { "grad_norm": 0.19125568866729736, "learning_rate": 8.4060664248943e-05, "loss": 0.0177, "step": 14920 }, { "grad_norm": 0.21832269430160522, "learning_rate": 8.40364471917358e-05, "loss": 0.0157, "step": 14930 }, { "grad_norm": 0.17632268369197845, "learning_rate": 8.401221524583024e-05, "loss": 0.0167, "step": 14940 }, { "grad_norm": 0.2568971514701843, "learning_rate": 8.398796842182619e-05, "loss": 0.0132, "step": 14950 }, { "grad_norm": 0.24805079400539398, "learning_rate": 8.396370673033006e-05, "loss": 0.0138, "step": 14960 }, { "grad_norm": 0.14267311990261078, "learning_rate": 8.39394301819547e-05, "loss": 0.0193, "step": 14970 }, { "grad_norm": 0.21560396254062653, "learning_rate": 8.391513878731949e-05, "loss": 0.0149, "step": 14980 }, { "grad_norm": 0.1859932839870453, "learning_rate": 8.389083255705037e-05, "loss": 0.0133, "step": 14990 }, { "grad_norm": 0.23598697781562805, "learning_rate": 8.386651150177968e-05, "loss": 0.0161, "step": 15000 }, { "grad_norm": 0.13201259076595306, "learning_rate": 8.384217563214627e-05, "loss": 0.0132, "step": 15010 }, { "grad_norm": 0.2015417516231537, "learning_rate": 8.381782495879546e-05, "loss": 0.0144, "step": 15020 }, { "grad_norm": 0.22176051139831543, "learning_rate": 8.37934594923791e-05, "loss": 0.0126, "step": 15030 }, { "grad_norm": 0.2201116681098938, "learning_rate": 8.376907924355546e-05, "loss": 0.0127, "step": 15040 }, { "grad_norm": 0.2633454203605652, "learning_rate": 8.374468422298926e-05, "loss": 0.0141, "step": 15050 }, { "grad_norm": 0.22914880514144897, "learning_rate": 8.372027444135176e-05, "loss": 0.012, "step": 15060 }, { "grad_norm": 0.2098378837108612, "learning_rate": 8.36958499093206e-05, "loss": 0.0136, "step": 15070 }, { "grad_norm": 0.22794660925865173, "learning_rate": 8.367141063757988e-05, "loss": 0.0139, "step": 15080 }, { "grad_norm": 0.2820660471916199, "learning_rate": 8.364695663682022e-05, "loss": 0.0139, "step": 15090 }, { "grad_norm": 0.24393253028392792, "learning_rate": 8.362248791773858e-05, "loss": 0.0138, "step": 15100 }, { "grad_norm": 0.20994621515274048, "learning_rate": 8.359800449103842e-05, "loss": 0.0148, "step": 15110 }, { "grad_norm": 0.18136835098266602, "learning_rate": 8.357350636742967e-05, "loss": 0.0135, "step": 15120 }, { "grad_norm": 0.20085476338863373, "learning_rate": 8.354899355762858e-05, "loss": 0.0139, "step": 15130 }, { "grad_norm": 0.21443253755569458, "learning_rate": 8.352446607235791e-05, "loss": 0.0183, "step": 15140 }, { "grad_norm": 0.19582930207252502, "learning_rate": 8.349992392234683e-05, "loss": 0.0168, "step": 15150 }, { "grad_norm": 0.21877464652061462, "learning_rate": 8.347536711833088e-05, "loss": 0.0171, "step": 15160 }, { "grad_norm": 0.2102109044790268, "learning_rate": 8.345079567105206e-05, "loss": 0.0148, "step": 15170 }, { "grad_norm": 0.24954672157764435, "learning_rate": 8.342620959125875e-05, "loss": 0.0147, "step": 15180 }, { "grad_norm": 0.18947342038154602, "learning_rate": 8.340160888970574e-05, "loss": 0.0166, "step": 15190 }, { "grad_norm": 0.22434794902801514, "learning_rate": 8.337699357715422e-05, "loss": 0.0135, "step": 15200 }, { "grad_norm": 0.18232351541519165, "learning_rate": 8.335236366437174e-05, "loss": 0.015, "step": 15210 }, { "grad_norm": 0.20749615132808685, "learning_rate": 8.332771916213227e-05, "loss": 0.0135, "step": 15220 }, { "grad_norm": 0.20341046154499054, "learning_rate": 8.330306008121618e-05, "loss": 0.0129, "step": 15230 }, { "grad_norm": 0.22228018939495087, "learning_rate": 8.327838643241014e-05, "loss": 0.0132, "step": 15240 }, { "grad_norm": 0.2750200033187866, "learning_rate": 8.32536982265073e-05, "loss": 0.0134, "step": 15250 }, { "grad_norm": 0.1898784041404724, "learning_rate": 8.32289954743071e-05, "loss": 0.0138, "step": 15260 }, { "grad_norm": 0.19248391687870026, "learning_rate": 8.320427818661532e-05, "loss": 0.0132, "step": 15270 }, { "grad_norm": 0.18432924151420593, "learning_rate": 8.317954637424419e-05, "loss": 0.0151, "step": 15280 }, { "grad_norm": 0.18091033399105072, "learning_rate": 8.315480004801221e-05, "loss": 0.0133, "step": 15290 }, { "grad_norm": 0.22201557457447052, "learning_rate": 8.31300392187443e-05, "loss": 0.0133, "step": 15300 }, { "grad_norm": 0.20445020496845245, "learning_rate": 8.31052638972717e-05, "loss": 0.0123, "step": 15310 }, { "grad_norm": 0.22878718376159668, "learning_rate": 8.30804740944319e-05, "loss": 0.0153, "step": 15320 }, { "grad_norm": 0.19531208276748657, "learning_rate": 8.305566982106887e-05, "loss": 0.0131, "step": 15330 }, { "grad_norm": 0.3049273192882538, "learning_rate": 8.303085108803282e-05, "loss": 0.0151, "step": 15340 }, { "grad_norm": 0.2322346568107605, "learning_rate": 8.300601790618032e-05, "loss": 0.0148, "step": 15350 }, { "grad_norm": 0.23208405077457428, "learning_rate": 8.298117028637423e-05, "loss": 0.0142, "step": 15360 }, { "grad_norm": 0.21119150519371033, "learning_rate": 8.295630823948375e-05, "loss": 0.017, "step": 15370 }, { "grad_norm": 0.2214541733264923, "learning_rate": 8.293143177638438e-05, "loss": 0.0173, "step": 15380 }, { "grad_norm": 0.18076449632644653, "learning_rate": 8.290654090795795e-05, "loss": 0.0141, "step": 15390 }, { "grad_norm": 0.19351746141910553, "learning_rate": 8.288163564509253e-05, "loss": 0.0163, "step": 15400 }, { "grad_norm": 0.2070235162973404, "learning_rate": 8.285671599868257e-05, "loss": 0.0157, "step": 15410 }, { "grad_norm": 0.18620090186595917, "learning_rate": 8.283178197962875e-05, "loss": 0.0156, "step": 15420 }, { "grad_norm": 0.15074557065963745, "learning_rate": 8.280683359883808e-05, "loss": 0.0156, "step": 15430 }, { "grad_norm": 0.18230272829532623, "learning_rate": 8.278187086722377e-05, "loss": 0.0163, "step": 15440 }, { "grad_norm": 0.1810087114572525, "learning_rate": 8.275689379570544e-05, "loss": 0.0138, "step": 15450 }, { "grad_norm": 0.16751958429813385, "learning_rate": 8.273190239520887e-05, "loss": 0.0147, "step": 15460 }, { "grad_norm": 0.21276356279850006, "learning_rate": 8.270689667666616e-05, "loss": 0.016, "step": 15470 }, { "grad_norm": 0.19457021355628967, "learning_rate": 8.268187665101563e-05, "loss": 0.0161, "step": 15480 }, { "grad_norm": 0.17939449846744537, "learning_rate": 8.265684232920194e-05, "loss": 0.0216, "step": 15490 }, { "grad_norm": 0.22985893487930298, "learning_rate": 8.263179372217592e-05, "loss": 0.0179, "step": 15500 }, { "grad_norm": 0.19748789072036743, "learning_rate": 8.260673084089468e-05, "loss": 0.0168, "step": 15510 }, { "grad_norm": 0.2245890200138092, "learning_rate": 8.25816536963216e-05, "loss": 0.0153, "step": 15520 }, { "grad_norm": 0.18914680182933807, "learning_rate": 8.255656229942626e-05, "loss": 0.0141, "step": 15530 }, { "grad_norm": 0.1940603405237198, "learning_rate": 8.253145666118447e-05, "loss": 0.0127, "step": 15540 }, { "grad_norm": 0.20917953550815582, "learning_rate": 8.250633679257834e-05, "loss": 0.0157, "step": 15550 }, { "grad_norm": 0.23818765580654144, "learning_rate": 8.24812027045961e-05, "loss": 0.014, "step": 15560 }, { "grad_norm": 0.2166333943605423, "learning_rate": 8.245605440823226e-05, "loss": 0.0138, "step": 15570 }, { "grad_norm": 0.1811836063861847, "learning_rate": 8.243089191448759e-05, "loss": 0.0125, "step": 15580 }, { "grad_norm": 0.19130009412765503, "learning_rate": 8.240571523436896e-05, "loss": 0.0155, "step": 15590 }, { "grad_norm": 0.18153251707553864, "learning_rate": 8.238052437888955e-05, "loss": 0.0148, "step": 15600 }, { "grad_norm": 0.19900715351104736, "learning_rate": 8.235531935906866e-05, "loss": 0.0143, "step": 15610 }, { "grad_norm": 0.15515442192554474, "learning_rate": 8.233010018593185e-05, "loss": 0.0123, "step": 15620 }, { "grad_norm": 0.15772539377212524, "learning_rate": 8.230486687051082e-05, "loss": 0.0134, "step": 15630 }, { "grad_norm": 0.1765284538269043, "learning_rate": 8.22796194238435e-05, "loss": 0.0139, "step": 15640 }, { "grad_norm": 0.18094630539417267, "learning_rate": 8.225435785697397e-05, "loss": 0.0137, "step": 15650 }, { "grad_norm": 0.22996364533901215, "learning_rate": 8.222908218095252e-05, "loss": 0.0149, "step": 15660 }, { "grad_norm": 0.19521331787109375, "learning_rate": 8.220379240683553e-05, "loss": 0.0142, "step": 15670 }, { "grad_norm": 0.24973252415657043, "learning_rate": 8.217848854568566e-05, "loss": 0.014, "step": 15680 }, { "grad_norm": 0.15194422006607056, "learning_rate": 8.215317060857167e-05, "loss": 0.0129, "step": 15690 }, { "grad_norm": 0.18245375156402588, "learning_rate": 8.212783860656847e-05, "loss": 0.0112, "step": 15700 }, { "grad_norm": 0.17329463362693787, "learning_rate": 8.210249255075717e-05, "loss": 0.0121, "step": 15710 }, { "grad_norm": 0.2138417512178421, "learning_rate": 8.207713245222496e-05, "loss": 0.0133, "step": 15720 }, { "grad_norm": 0.25326699018478394, "learning_rate": 8.205175832206522e-05, "loss": 0.0152, "step": 15730 }, { "grad_norm": 0.22895698249340057, "learning_rate": 8.202637017137747e-05, "loss": 0.015, "step": 15740 }, { "grad_norm": 0.2192966789007187, "learning_rate": 8.200096801126734e-05, "loss": 0.0137, "step": 15750 }, { "grad_norm": 0.25613531470298767, "learning_rate": 8.197555185284658e-05, "loss": 0.0155, "step": 15760 }, { "grad_norm": 0.17924650013446808, "learning_rate": 8.19501217072331e-05, "loss": 0.0147, "step": 15770 }, { "grad_norm": 0.1897151917219162, "learning_rate": 8.192467758555091e-05, "loss": 0.0141, "step": 15780 }, { "grad_norm": 0.18003207445144653, "learning_rate": 8.189921949893014e-05, "loss": 0.0142, "step": 15790 }, { "grad_norm": 0.18742339313030243, "learning_rate": 8.187374745850702e-05, "loss": 0.0137, "step": 15800 }, { "grad_norm": 0.19231750071048737, "learning_rate": 8.184826147542387e-05, "loss": 0.0128, "step": 15810 }, { "grad_norm": 0.21481953561306, "learning_rate": 8.18227615608291e-05, "loss": 0.0182, "step": 15820 }, { "grad_norm": 0.18483859300613403, "learning_rate": 8.179724772587732e-05, "loss": 0.0126, "step": 15830 }, { "grad_norm": 0.17660047113895416, "learning_rate": 8.177171998172908e-05, "loss": 0.0178, "step": 15840 }, { "grad_norm": 0.1728576272726059, "learning_rate": 8.174617833955112e-05, "loss": 0.0138, "step": 15850 }, { "grad_norm": 0.21393314003944397, "learning_rate": 8.172062281051618e-05, "loss": 0.0147, "step": 15860 }, { "grad_norm": 0.16815169155597687, "learning_rate": 8.169505340580316e-05, "loss": 0.0132, "step": 15870 }, { "grad_norm": 0.25227484107017517, "learning_rate": 8.166947013659696e-05, "loss": 0.0148, "step": 15880 }, { "grad_norm": 0.19321474432945251, "learning_rate": 8.164387301408858e-05, "loss": 0.0151, "step": 15890 }, { "grad_norm": 0.29287952184677124, "learning_rate": 8.161826204947507e-05, "loss": 0.0188, "step": 15900 }, { "grad_norm": 0.20355576276779175, "learning_rate": 8.159263725395954e-05, "loss": 0.0182, "step": 15910 }, { "grad_norm": 0.2491297423839569, "learning_rate": 8.156699863875115e-05, "loss": 0.02, "step": 15920 }, { "grad_norm": 0.21967370808124542, "learning_rate": 8.154134621506508e-05, "loss": 0.0152, "step": 15930 }, { "grad_norm": 0.3160310983657837, "learning_rate": 8.15156799941226e-05, "loss": 0.0144, "step": 15940 }, { "grad_norm": 0.18960490822792053, "learning_rate": 8.148999998715096e-05, "loss": 0.0173, "step": 15950 }, { "grad_norm": 0.1715354025363922, "learning_rate": 8.146430620538348e-05, "loss": 0.0141, "step": 15960 }, { "grad_norm": 0.16950874030590057, "learning_rate": 8.143859866005949e-05, "loss": 0.014, "step": 15970 }, { "grad_norm": 0.20162555575370789, "learning_rate": 8.141287736242438e-05, "loss": 0.0136, "step": 15980 }, { "grad_norm": 0.1490805596113205, "learning_rate": 8.138714232372945e-05, "loss": 0.014, "step": 15990 }, { "grad_norm": 0.1909923553466797, "learning_rate": 8.136139355523213e-05, "loss": 0.0164, "step": 16000 }, { "grad_norm": 0.21604950726032257, "learning_rate": 8.13356310681958e-05, "loss": 0.0142, "step": 16010 }, { "grad_norm": 0.18406419456005096, "learning_rate": 8.130985487388982e-05, "loss": 0.0156, "step": 16020 }, { "grad_norm": 0.17565736174583435, "learning_rate": 8.12840649835896e-05, "loss": 0.0134, "step": 16030 }, { "grad_norm": 0.15658347308635712, "learning_rate": 8.125826140857653e-05, "loss": 0.0143, "step": 16040 }, { "grad_norm": 0.20856918394565582, "learning_rate": 8.123244416013792e-05, "loss": 0.0142, "step": 16050 }, { "grad_norm": 0.332287460565567, "learning_rate": 8.120661324956716e-05, "loss": 0.0159, "step": 16060 }, { "grad_norm": 0.19089867174625397, "learning_rate": 8.118076868816353e-05, "loss": 0.0128, "step": 16070 }, { "grad_norm": 0.18979990482330322, "learning_rate": 8.115491048723234e-05, "loss": 0.0145, "step": 16080 }, { "grad_norm": 0.19986410439014435, "learning_rate": 8.112903865808486e-05, "loss": 0.012, "step": 16090 }, { "grad_norm": 0.18363726139068604, "learning_rate": 8.110315321203826e-05, "loss": 0.013, "step": 16100 }, { "grad_norm": 0.2015133500099182, "learning_rate": 8.107725416041575e-05, "loss": 0.0125, "step": 16110 }, { "grad_norm": 0.20112544298171997, "learning_rate": 8.105134151454645e-05, "loss": 0.0128, "step": 16120 }, { "grad_norm": 0.1677011400461197, "learning_rate": 8.102541528576544e-05, "loss": 0.0143, "step": 16130 }, { "grad_norm": 0.22358575463294983, "learning_rate": 8.099947548541371e-05, "loss": 0.0133, "step": 16140 }, { "grad_norm": 0.20383265614509583, "learning_rate": 8.097352212483824e-05, "loss": 0.0138, "step": 16150 }, { "grad_norm": 0.2501258850097656, "learning_rate": 8.094755521539187e-05, "loss": 0.0147, "step": 16160 }, { "grad_norm": 0.1807689517736435, "learning_rate": 8.092157476843346e-05, "loss": 0.0127, "step": 16170 }, { "grad_norm": 0.1771225482225418, "learning_rate": 8.089558079532769e-05, "loss": 0.0146, "step": 16180 }, { "grad_norm": 0.2019912153482437, "learning_rate": 8.086957330744522e-05, "loss": 0.0134, "step": 16190 }, { "grad_norm": 0.26713356375694275, "learning_rate": 8.084355231616265e-05, "loss": 0.0132, "step": 16200 }, { "grad_norm": 0.2008967250585556, "learning_rate": 8.08175178328624e-05, "loss": 0.013, "step": 16210 }, { "grad_norm": 0.1896194964647293, "learning_rate": 8.079146986893284e-05, "loss": 0.0144, "step": 16220 }, { "grad_norm": 0.12849321961402893, "learning_rate": 8.076540843576826e-05, "loss": 0.0137, "step": 16230 }, { "grad_norm": 0.28719332814216614, "learning_rate": 8.073933354476879e-05, "loss": 0.0152, "step": 16240 }, { "grad_norm": 0.2001834660768509, "learning_rate": 8.07132452073405e-05, "loss": 0.0124, "step": 16250 }, { "grad_norm": 0.20847351849079132, "learning_rate": 8.068714343489529e-05, "loss": 0.0169, "step": 16260 }, { "grad_norm": 0.21027639508247375, "learning_rate": 8.066102823885097e-05, "loss": 0.0148, "step": 16270 }, { "grad_norm": 0.20712073147296906, "learning_rate": 8.063489963063125e-05, "loss": 0.017, "step": 16280 }, { "grad_norm": 0.18911238014698029, "learning_rate": 8.06087576216656e-05, "loss": 0.0117, "step": 16290 }, { "grad_norm": 0.20391049981117249, "learning_rate": 8.058260222338946e-05, "loss": 0.0136, "step": 16300 }, { "grad_norm": 0.20889630913734436, "learning_rate": 8.05564334472441e-05, "loss": 0.0152, "step": 16310 }, { "grad_norm": 0.21257750689983368, "learning_rate": 8.053025130467664e-05, "loss": 0.0134, "step": 16320 }, { "grad_norm": 0.21835851669311523, "learning_rate": 8.050405580714e-05, "loss": 0.0116, "step": 16330 }, { "grad_norm": 0.17834635078907013, "learning_rate": 8.047784696609301e-05, "loss": 0.0113, "step": 16340 }, { "grad_norm": 0.16305570304393768, "learning_rate": 8.045162479300029e-05, "loss": 0.014, "step": 16350 }, { "grad_norm": 0.22485600411891937, "learning_rate": 8.042538929933232e-05, "loss": 0.0164, "step": 16360 }, { "grad_norm": 0.18647557497024536, "learning_rate": 8.039914049656539e-05, "loss": 0.0115, "step": 16370 }, { "grad_norm": 0.1617908924818039, "learning_rate": 8.037287839618163e-05, "loss": 0.0109, "step": 16380 }, { "grad_norm": 0.18964210152626038, "learning_rate": 8.034660300966898e-05, "loss": 0.0126, "step": 16390 }, { "grad_norm": 0.17816808819770813, "learning_rate": 8.032031434852115e-05, "loss": 0.0161, "step": 16400 }, { "grad_norm": 0.18561047315597534, "learning_rate": 8.029401242423773e-05, "loss": 0.0145, "step": 16410 }, { "grad_norm": 0.21064558625221252, "learning_rate": 8.026769724832406e-05, "loss": 0.0145, "step": 16420 }, { "grad_norm": 0.23553688824176788, "learning_rate": 8.02413688322913e-05, "loss": 0.0114, "step": 16430 }, { "grad_norm": 0.21091671288013458, "learning_rate": 8.021502718765639e-05, "loss": 0.0117, "step": 16440 }, { "grad_norm": 0.23021844029426575, "learning_rate": 8.018867232594204e-05, "loss": 0.015, "step": 16450 }, { "grad_norm": 0.24437488615512848, "learning_rate": 8.016230425867679e-05, "loss": 0.0171, "step": 16460 }, { "grad_norm": 0.16909712553024292, "learning_rate": 8.013592299739494e-05, "loss": 0.0126, "step": 16470 }, { "grad_norm": 0.14658832550048828, "learning_rate": 8.01095285536365e-05, "loss": 0.0137, "step": 16480 }, { "grad_norm": 0.14560791850090027, "learning_rate": 8.008312093894735e-05, "loss": 0.0119, "step": 16490 }, { "grad_norm": 0.16564151644706726, "learning_rate": 8.005670016487903e-05, "loss": 0.012, "step": 16500 }, { "grad_norm": 0.23175086081027985, "learning_rate": 8.003026624298892e-05, "loss": 0.0102, "step": 16510 }, { "grad_norm": 0.16667188704013824, "learning_rate": 8.00038191848401e-05, "loss": 0.0134, "step": 16520 }, { "grad_norm": 0.20373383164405823, "learning_rate": 7.997735900200142e-05, "loss": 0.0144, "step": 16530 }, { "grad_norm": 0.1875503957271576, "learning_rate": 7.995088570604745e-05, "loss": 0.013, "step": 16540 }, { "grad_norm": 0.14034579694271088, "learning_rate": 7.992439930855852e-05, "loss": 0.0109, "step": 16550 }, { "grad_norm": 0.22447502613067627, "learning_rate": 7.989789982112067e-05, "loss": 0.0143, "step": 16560 }, { "grad_norm": 0.1913711130619049, "learning_rate": 7.987138725532572e-05, "loss": 0.0147, "step": 16570 }, { "grad_norm": 0.17126096785068512, "learning_rate": 7.98448616227711e-05, "loss": 0.0123, "step": 16580 }, { "grad_norm": 0.15150243043899536, "learning_rate": 7.981832293506004e-05, "loss": 0.013, "step": 16590 }, { "grad_norm": 0.14691191911697388, "learning_rate": 7.97917712038015e-05, "loss": 0.0122, "step": 16600 }, { "grad_norm": 0.15040814876556396, "learning_rate": 7.976520644061007e-05, "loss": 0.0101, "step": 16610 }, { "grad_norm": 0.22900432348251343, "learning_rate": 7.973862865710611e-05, "loss": 0.0125, "step": 16620 }, { "grad_norm": 0.14741674065589905, "learning_rate": 7.971203786491563e-05, "loss": 0.0156, "step": 16630 }, { "grad_norm": 0.23117634654045105, "learning_rate": 7.968543407567034e-05, "loss": 0.0131, "step": 16640 }, { "grad_norm": 0.1545952558517456, "learning_rate": 7.965881730100765e-05, "loss": 0.0168, "step": 16650 }, { "grad_norm": 0.20839174091815948, "learning_rate": 7.963218755257064e-05, "loss": 0.012, "step": 16660 }, { "grad_norm": 0.20110207796096802, "learning_rate": 7.960554484200807e-05, "loss": 0.0152, "step": 16670 }, { "grad_norm": 0.14714448153972626, "learning_rate": 7.957888918097436e-05, "loss": 0.0119, "step": 16680 }, { "grad_norm": 0.19018179178237915, "learning_rate": 7.955222058112962e-05, "loss": 0.0108, "step": 16690 }, { "grad_norm": 0.18157163262367249, "learning_rate": 7.952553905413955e-05, "loss": 0.0113, "step": 16700 }, { "grad_norm": 0.16330525279045105, "learning_rate": 7.949884461167563e-05, "loss": 0.0137, "step": 16710 }, { "grad_norm": 0.13949407637119293, "learning_rate": 7.947213726541484e-05, "loss": 0.0113, "step": 16720 }, { "grad_norm": 0.26493871212005615, "learning_rate": 7.944541702703996e-05, "loss": 0.0126, "step": 16730 }, { "grad_norm": 0.18622152507305145, "learning_rate": 7.941868390823928e-05, "loss": 0.0163, "step": 16740 }, { "grad_norm": 0.20533102750778198, "learning_rate": 7.939193792070677e-05, "loss": 0.0134, "step": 16750 }, { "grad_norm": 0.18917490541934967, "learning_rate": 7.936517907614207e-05, "loss": 0.0136, "step": 16760 }, { "grad_norm": 0.2670118510723114, "learning_rate": 7.933840738625036e-05, "loss": 0.0144, "step": 16770 }, { "grad_norm": 0.23025615513324738, "learning_rate": 7.931162286274252e-05, "loss": 0.0144, "step": 16780 }, { "grad_norm": 0.1937585324048996, "learning_rate": 7.928482551733503e-05, "loss": 0.0126, "step": 16790 }, { "grad_norm": 0.17140038311481476, "learning_rate": 7.925801536174991e-05, "loss": 0.0126, "step": 16800 }, { "grad_norm": 0.15472757816314697, "learning_rate": 7.923119240771484e-05, "loss": 0.0095, "step": 16810 }, { "grad_norm": 0.2222132533788681, "learning_rate": 7.920435666696313e-05, "loss": 0.0124, "step": 16820 }, { "grad_norm": 0.21090425550937653, "learning_rate": 7.917750815123362e-05, "loss": 0.0128, "step": 16830 }, { "grad_norm": 0.16505002975463867, "learning_rate": 7.915064687227075e-05, "loss": 0.0119, "step": 16840 }, { "grad_norm": 0.16034339368343353, "learning_rate": 7.912377284182458e-05, "loss": 0.0143, "step": 16850 }, { "grad_norm": 0.2239542007446289, "learning_rate": 7.90968860716507e-05, "loss": 0.012, "step": 16860 }, { "grad_norm": 0.19641442596912384, "learning_rate": 7.906998657351033e-05, "loss": 0.0123, "step": 16870 }, { "grad_norm": 0.17314167320728302, "learning_rate": 7.90430743591702e-05, "loss": 0.0131, "step": 16880 }, { "grad_norm": 0.19985292851924896, "learning_rate": 7.901614944040263e-05, "loss": 0.0121, "step": 16890 }, { "grad_norm": 0.15974561870098114, "learning_rate": 7.89892118289855e-05, "loss": 0.0132, "step": 16900 }, { "grad_norm": 0.17452605068683624, "learning_rate": 7.896226153670223e-05, "loss": 0.0114, "step": 16910 }, { "grad_norm": 0.2433462291955948, "learning_rate": 7.89352985753418e-05, "loss": 0.0168, "step": 16920 }, { "grad_norm": 0.17155568301677704, "learning_rate": 7.890832295669871e-05, "loss": 0.0116, "step": 16930 }, { "grad_norm": 0.15582984685897827, "learning_rate": 7.888133469257303e-05, "loss": 0.0126, "step": 16940 }, { "grad_norm": 0.1882326751947403, "learning_rate": 7.885433379477034e-05, "loss": 0.0145, "step": 16950 }, { "grad_norm": 0.16054409742355347, "learning_rate": 7.882732027510174e-05, "loss": 0.0147, "step": 16960 }, { "grad_norm": 0.17706015706062317, "learning_rate": 7.880029414538386e-05, "loss": 0.0147, "step": 16970 }, { "grad_norm": 0.17251335084438324, "learning_rate": 7.877325541743888e-05, "loss": 0.0109, "step": 16980 }, { "grad_norm": 0.22736628353595734, "learning_rate": 7.874620410309443e-05, "loss": 0.0117, "step": 16990 }, { "grad_norm": 0.1712624579668045, "learning_rate": 7.871914021418368e-05, "loss": 0.0141, "step": 17000 }, { "grad_norm": 0.2757171094417572, "learning_rate": 7.869206376254528e-05, "loss": 0.0186, "step": 17010 }, { "grad_norm": 0.18450172245502472, "learning_rate": 7.86649747600234e-05, "loss": 0.0152, "step": 17020 }, { "grad_norm": 0.19964875280857086, "learning_rate": 7.86378732184677e-05, "loss": 0.0124, "step": 17030 }, { "grad_norm": 0.21817098557949066, "learning_rate": 7.86107591497333e-05, "loss": 0.0144, "step": 17040 }, { "grad_norm": 0.182329922914505, "learning_rate": 7.858363256568083e-05, "loss": 0.0139, "step": 17050 }, { "grad_norm": 0.1959627866744995, "learning_rate": 7.855649347817637e-05, "loss": 0.0147, "step": 17060 }, { "grad_norm": 0.1999945044517517, "learning_rate": 7.85293418990915e-05, "loss": 0.0138, "step": 17070 }, { "grad_norm": 0.16402937471866608, "learning_rate": 7.85021778403032e-05, "loss": 0.0111, "step": 17080 }, { "grad_norm": 0.19569021463394165, "learning_rate": 7.847500131369397e-05, "loss": 0.0142, "step": 17090 }, { "grad_norm": 0.155470609664917, "learning_rate": 7.844781233115176e-05, "loss": 0.0106, "step": 17100 }, { "grad_norm": 0.2566491961479187, "learning_rate": 7.842061090456996e-05, "loss": 0.0105, "step": 17110 }, { "grad_norm": 0.2471739500761032, "learning_rate": 7.839339704584736e-05, "loss": 0.0132, "step": 17120 }, { "grad_norm": 0.19571788609027863, "learning_rate": 7.836617076688828e-05, "loss": 0.0123, "step": 17130 }, { "grad_norm": 0.2076190710067749, "learning_rate": 7.833893207960236e-05, "loss": 0.0144, "step": 17140 }, { "grad_norm": 0.1908150315284729, "learning_rate": 7.831168099590478e-05, "loss": 0.0133, "step": 17150 }, { "grad_norm": 0.17001797258853912, "learning_rate": 7.828441752771607e-05, "loss": 0.0141, "step": 17160 }, { "grad_norm": 0.149556964635849, "learning_rate": 7.825714168696219e-05, "loss": 0.0109, "step": 17170 }, { "grad_norm": 0.1783851683139801, "learning_rate": 7.82298534855745e-05, "loss": 0.0124, "step": 17180 }, { "grad_norm": 0.20724186301231384, "learning_rate": 7.820255293548984e-05, "loss": 0.0164, "step": 17190 }, { "grad_norm": 0.19871698319911957, "learning_rate": 7.817524004865035e-05, "loss": 0.0113, "step": 17200 }, { "grad_norm": 0.13494695723056793, "learning_rate": 7.814791483700366e-05, "loss": 0.0098, "step": 17210 }, { "grad_norm": 0.20947898924350739, "learning_rate": 7.81205773125027e-05, "loss": 0.0129, "step": 17220 }, { "grad_norm": 0.1666402667760849, "learning_rate": 7.809322748710585e-05, "loss": 0.0144, "step": 17230 }, { "grad_norm": 0.1581578403711319, "learning_rate": 7.806586537277685e-05, "loss": 0.0146, "step": 17240 }, { "grad_norm": 0.14638042449951172, "learning_rate": 7.803849098148484e-05, "loss": 0.0119, "step": 17250 }, { "grad_norm": 0.18746882677078247, "learning_rate": 7.801110432520427e-05, "loss": 0.0131, "step": 17260 }, { "grad_norm": 0.15351933240890503, "learning_rate": 7.798370541591504e-05, "loss": 0.0136, "step": 17270 }, { "grad_norm": 0.17540200054645538, "learning_rate": 7.795629426560232e-05, "loss": 0.0124, "step": 17280 }, { "grad_norm": 0.1643502414226532, "learning_rate": 7.792887088625671e-05, "loss": 0.0108, "step": 17290 }, { "grad_norm": 0.20626762509346008, "learning_rate": 7.790143528987412e-05, "loss": 0.0144, "step": 17300 }, { "grad_norm": 0.18912595510482788, "learning_rate": 7.787398748845578e-05, "loss": 0.0109, "step": 17310 }, { "grad_norm": 0.1651771068572998, "learning_rate": 7.784652749400834e-05, "loss": 0.012, "step": 17320 }, { "grad_norm": 0.278839647769928, "learning_rate": 7.78190553185437e-05, "loss": 0.0151, "step": 17330 }, { "grad_norm": 0.2048417180776596, "learning_rate": 7.779157097407915e-05, "loss": 0.0129, "step": 17340 }, { "grad_norm": 0.21829654276371002, "learning_rate": 7.776407447263725e-05, "loss": 0.0152, "step": 17350 }, { "grad_norm": 0.168159618973732, "learning_rate": 7.773656582624593e-05, "loss": 0.0124, "step": 17360 }, { "grad_norm": 0.18229460716247559, "learning_rate": 7.770904504693837e-05, "loss": 0.0121, "step": 17370 }, { "grad_norm": 0.15205134451389313, "learning_rate": 7.768151214675314e-05, "loss": 0.013, "step": 17380 }, { "grad_norm": 0.18675142526626587, "learning_rate": 7.765396713773402e-05, "loss": 0.0129, "step": 17390 }, { "grad_norm": 0.2199348658323288, "learning_rate": 7.762641003193017e-05, "loss": 0.0156, "step": 17400 }, { "grad_norm": 0.19740426540374756, "learning_rate": 7.759884084139597e-05, "loss": 0.0128, "step": 17410 }, { "grad_norm": 0.23451554775238037, "learning_rate": 7.757125957819116e-05, "loss": 0.0118, "step": 17420 }, { "grad_norm": 0.2191203534603119, "learning_rate": 7.754366625438069e-05, "loss": 0.0118, "step": 17430 }, { "grad_norm": 0.2032255083322525, "learning_rate": 7.751606088203483e-05, "loss": 0.0131, "step": 17440 }, { "grad_norm": 0.2743101418018341, "learning_rate": 7.74884434732291e-05, "loss": 0.0138, "step": 17450 }, { "grad_norm": 0.1422770917415619, "learning_rate": 7.746081404004431e-05, "loss": 0.0114, "step": 17460 }, { "grad_norm": 0.15598182380199432, "learning_rate": 7.743317259456649e-05, "loss": 0.0132, "step": 17470 }, { "grad_norm": 0.24664343893527985, "learning_rate": 7.740551914888696e-05, "loss": 0.0133, "step": 17480 }, { "grad_norm": 0.13909658789634705, "learning_rate": 7.737785371510229e-05, "loss": 0.0124, "step": 17490 }, { "grad_norm": 0.1832265853881836, "learning_rate": 7.735017630531426e-05, "loss": 0.0132, "step": 17500 }, { "grad_norm": 0.24958227574825287, "learning_rate": 7.732248693162991e-05, "loss": 0.0141, "step": 17510 }, { "grad_norm": 0.18425600230693817, "learning_rate": 7.729478560616153e-05, "loss": 0.01, "step": 17520 }, { "grad_norm": 0.21296411752700806, "learning_rate": 7.726707234102659e-05, "loss": 0.0134, "step": 17530 }, { "grad_norm": 0.2786808907985687, "learning_rate": 7.723934714834785e-05, "loss": 0.0129, "step": 17540 }, { "grad_norm": 0.16840443015098572, "learning_rate": 7.721161004025323e-05, "loss": 0.0118, "step": 17550 }, { "grad_norm": 0.2019394338130951, "learning_rate": 7.718386102887588e-05, "loss": 0.013, "step": 17560 }, { "grad_norm": 0.1984574794769287, "learning_rate": 7.715610012635418e-05, "loss": 0.0144, "step": 17570 }, { "grad_norm": 0.19724710285663605, "learning_rate": 7.712832734483166e-05, "loss": 0.0134, "step": 17580 }, { "grad_norm": 0.17074300348758698, "learning_rate": 7.710054269645713e-05, "loss": 0.0133, "step": 17590 }, { "grad_norm": 0.19136108458042145, "learning_rate": 7.707274619338449e-05, "loss": 0.0111, "step": 17600 }, { "grad_norm": 0.22624105215072632, "learning_rate": 7.704493784777289e-05, "loss": 0.0123, "step": 17610 }, { "grad_norm": 0.14387543499469757, "learning_rate": 7.701711767178665e-05, "loss": 0.0115, "step": 17620 }, { "grad_norm": 0.20454372465610504, "learning_rate": 7.698928567759525e-05, "loss": 0.0137, "step": 17630 }, { "grad_norm": 0.182696133852005, "learning_rate": 7.696144187737335e-05, "loss": 0.0125, "step": 17640 }, { "grad_norm": 0.20369169116020203, "learning_rate": 7.69335862833008e-05, "loss": 0.0093, "step": 17650 }, { "grad_norm": 0.15515895187854767, "learning_rate": 7.690571890756256e-05, "loss": 0.0097, "step": 17660 }, { "grad_norm": 0.2032262086868286, "learning_rate": 7.687783976234877e-05, "loss": 0.012, "step": 17670 }, { "grad_norm": 0.18972934782505035, "learning_rate": 7.684994885985472e-05, "loss": 0.0155, "step": 17680 }, { "grad_norm": 0.1714819371700287, "learning_rate": 7.682204621228083e-05, "loss": 0.0119, "step": 17690 }, { "grad_norm": 0.18884821236133575, "learning_rate": 7.679413183183267e-05, "loss": 0.0139, "step": 17700 }, { "grad_norm": 0.14690712094306946, "learning_rate": 7.676620573072093e-05, "loss": 0.0117, "step": 17710 }, { "grad_norm": 0.15074053406715393, "learning_rate": 7.673826792116145e-05, "loss": 0.0144, "step": 17720 }, { "grad_norm": 0.15411476790905, "learning_rate": 7.671031841537519e-05, "loss": 0.0132, "step": 17730 }, { "grad_norm": 0.17561569809913635, "learning_rate": 7.668235722558819e-05, "loss": 0.0164, "step": 17740 }, { "grad_norm": 0.2420274317264557, "learning_rate": 7.665438436403161e-05, "loss": 0.0118, "step": 17750 }, { "grad_norm": 0.17214705049991608, "learning_rate": 7.662639984294178e-05, "loss": 0.018, "step": 17760 }, { "grad_norm": 0.16211435198783875, "learning_rate": 7.659840367456002e-05, "loss": 0.0129, "step": 17770 }, { "grad_norm": 0.12783071398735046, "learning_rate": 7.657039587113287e-05, "loss": 0.012, "step": 17780 }, { "grad_norm": 0.15509523451328278, "learning_rate": 7.654237644491182e-05, "loss": 0.0137, "step": 17790 }, { "grad_norm": 0.17439597845077515, "learning_rate": 7.651434540815358e-05, "loss": 0.0117, "step": 17800 }, { "grad_norm": 0.16544301807880402, "learning_rate": 7.648630277311986e-05, "loss": 0.0148, "step": 17810 }, { "grad_norm": 0.15848210453987122, "learning_rate": 7.645824855207744e-05, "loss": 0.0131, "step": 17820 }, { "grad_norm": 0.17910870909690857, "learning_rate": 7.643018275729821e-05, "loss": 0.0099, "step": 17830 }, { "grad_norm": 0.17907653748989105, "learning_rate": 7.640210540105911e-05, "loss": 0.0131, "step": 17840 }, { "grad_norm": 0.16509029269218445, "learning_rate": 7.637401649564213e-05, "loss": 0.0118, "step": 17850 }, { "grad_norm": 0.21781806647777557, "learning_rate": 7.63459160533343e-05, "loss": 0.0133, "step": 17860 }, { "grad_norm": 0.241594135761261, "learning_rate": 7.631780408642769e-05, "loss": 0.0153, "step": 17870 }, { "grad_norm": 0.2747977375984192, "learning_rate": 7.628968060721946e-05, "loss": 0.0118, "step": 17880 }, { "grad_norm": 0.27169549465179443, "learning_rate": 7.626154562801176e-05, "loss": 0.0153, "step": 17890 }, { "grad_norm": 0.17441566288471222, "learning_rate": 7.623339916111177e-05, "loss": 0.0137, "step": 17900 }, { "grad_norm": 0.19402961432933807, "learning_rate": 7.620524121883174e-05, "loss": 0.0152, "step": 17910 }, { "grad_norm": 0.1799282729625702, "learning_rate": 7.617707181348892e-05, "loss": 0.0155, "step": 17920 }, { "grad_norm": 0.18157483637332916, "learning_rate": 7.614889095740554e-05, "loss": 0.0116, "step": 17930 }, { "grad_norm": 0.18750829994678497, "learning_rate": 7.612069866290885e-05, "loss": 0.0138, "step": 17940 }, { "grad_norm": 0.20470339059829712, "learning_rate": 7.609249494233115e-05, "loss": 0.0103, "step": 17950 }, { "grad_norm": 0.2344280332326889, "learning_rate": 7.60642798080097e-05, "loss": 0.0152, "step": 17960 }, { "grad_norm": 0.19096671044826508, "learning_rate": 7.603605327228673e-05, "loss": 0.0133, "step": 17970 }, { "grad_norm": 0.2365921437740326, "learning_rate": 7.600781534750953e-05, "loss": 0.0124, "step": 17980 }, { "grad_norm": 0.12231196463108063, "learning_rate": 7.597956604603029e-05, "loss": 0.0138, "step": 17990 }, { "grad_norm": 0.2049366980791092, "learning_rate": 7.595130538020622e-05, "loss": 0.0115, "step": 18000 }, { "grad_norm": 0.19563253223896027, "learning_rate": 7.592303336239952e-05, "loss": 0.0133, "step": 18010 }, { "grad_norm": 0.1886538416147232, "learning_rate": 7.589475000497733e-05, "loss": 0.0112, "step": 18020 }, { "grad_norm": 0.1753348559141159, "learning_rate": 7.586645532031171e-05, "loss": 0.0109, "step": 18030 }, { "grad_norm": 0.1693079024553299, "learning_rate": 7.583814932077975e-05, "loss": 0.0112, "step": 18040 }, { "grad_norm": 0.14030446112155914, "learning_rate": 7.580983201876343e-05, "loss": 0.0124, "step": 18050 }, { "grad_norm": 0.20239360630512238, "learning_rate": 7.578150342664973e-05, "loss": 0.0139, "step": 18060 }, { "grad_norm": 0.24271516501903534, "learning_rate": 7.575316355683055e-05, "loss": 0.0145, "step": 18070 }, { "grad_norm": 0.20339468121528625, "learning_rate": 7.572481242170266e-05, "loss": 0.0125, "step": 18080 }, { "grad_norm": 0.18271459639072418, "learning_rate": 7.569645003366786e-05, "loss": 0.0137, "step": 18090 }, { "grad_norm": 0.19255180656909943, "learning_rate": 7.566807640513278e-05, "loss": 0.0143, "step": 18100 }, { "grad_norm": 0.17189113795757294, "learning_rate": 7.563969154850905e-05, "loss": 0.0112, "step": 18110 }, { "grad_norm": 0.22535055875778198, "learning_rate": 7.561129547621314e-05, "loss": 0.0176, "step": 18120 }, { "grad_norm": 0.18858055770397186, "learning_rate": 7.558288820066648e-05, "loss": 0.0141, "step": 18130 }, { "grad_norm": 0.16816578805446625, "learning_rate": 7.555446973429534e-05, "loss": 0.0111, "step": 18140 }, { "grad_norm": 0.15724124014377594, "learning_rate": 7.552604008953097e-05, "loss": 0.0096, "step": 18150 }, { "grad_norm": 0.21814119815826416, "learning_rate": 7.549759927880942e-05, "loss": 0.0123, "step": 18160 }, { "grad_norm": 0.1996580958366394, "learning_rate": 7.54691473145717e-05, "loss": 0.0122, "step": 18170 }, { "grad_norm": 0.2246442437171936, "learning_rate": 7.544068420926365e-05, "loss": 0.012, "step": 18180 }, { "grad_norm": 0.14493007957935333, "learning_rate": 7.5412209975336e-05, "loss": 0.0128, "step": 18190 }, { "grad_norm": 0.22834785282611847, "learning_rate": 7.538372462524435e-05, "loss": 0.0123, "step": 18200 }, { "grad_norm": 0.16393938660621643, "learning_rate": 7.535522817144915e-05, "loss": 0.0106, "step": 18210 }, { "grad_norm": 0.1759633868932724, "learning_rate": 7.532672062641575e-05, "loss": 0.0121, "step": 18220 }, { "grad_norm": 0.24654430150985718, "learning_rate": 7.529820200261427e-05, "loss": 0.0138, "step": 18230 }, { "grad_norm": 0.16910642385482788, "learning_rate": 7.526967231251977e-05, "loss": 0.013, "step": 18240 }, { "grad_norm": 0.15669231116771698, "learning_rate": 7.524113156861206e-05, "loss": 0.0124, "step": 18250 }, { "grad_norm": 0.15529100596904755, "learning_rate": 7.521257978337587e-05, "loss": 0.0116, "step": 18260 }, { "grad_norm": 0.16490167379379272, "learning_rate": 7.518401696930072e-05, "loss": 0.013, "step": 18270 }, { "grad_norm": 0.16230081021785736, "learning_rate": 7.515544313888092e-05, "loss": 0.0126, "step": 18280 }, { "grad_norm": 0.15758302807807922, "learning_rate": 7.512685830461568e-05, "loss": 0.0105, "step": 18290 }, { "grad_norm": 0.17590612173080444, "learning_rate": 7.509826247900892e-05, "loss": 0.0107, "step": 18300 }, { "grad_norm": 0.21104857325553894, "learning_rate": 7.50696556745695e-05, "loss": 0.0167, "step": 18310 }, { "grad_norm": 0.19481642544269562, "learning_rate": 7.504103790381095e-05, "loss": 0.0109, "step": 18320 }, { "grad_norm": 0.17435726523399353, "learning_rate": 7.501240917925167e-05, "loss": 0.0157, "step": 18330 }, { "grad_norm": 0.1838396042585373, "learning_rate": 7.498376951341483e-05, "loss": 0.0128, "step": 18340 }, { "grad_norm": 0.18683254718780518, "learning_rate": 7.495511891882844e-05, "loss": 0.0126, "step": 18350 }, { "grad_norm": 0.20800252258777618, "learning_rate": 7.492645740802517e-05, "loss": 0.0117, "step": 18360 }, { "grad_norm": 0.21343742311000824, "learning_rate": 7.489778499354259e-05, "loss": 0.0111, "step": 18370 }, { "grad_norm": 0.1595340371131897, "learning_rate": 7.486910168792298e-05, "loss": 0.0125, "step": 18380 }, { "grad_norm": 0.19294749200344086, "learning_rate": 7.484040750371337e-05, "loss": 0.01, "step": 18390 }, { "grad_norm": 0.17441585659980774, "learning_rate": 7.48117024534656e-05, "loss": 0.011, "step": 18400 }, { "grad_norm": 0.18727931380271912, "learning_rate": 7.478298654973621e-05, "loss": 0.0116, "step": 18410 }, { "grad_norm": 0.19633245468139648, "learning_rate": 7.475425980508654e-05, "loss": 0.0114, "step": 18420 }, { "grad_norm": 0.1921997219324112, "learning_rate": 7.47255222320826e-05, "loss": 0.0131, "step": 18430 }, { "grad_norm": 0.2119644582271576, "learning_rate": 7.469677384329522e-05, "loss": 0.0143, "step": 18440 }, { "grad_norm": 0.1979852169752121, "learning_rate": 7.46680146512999e-05, "loss": 0.0136, "step": 18450 }, { "grad_norm": 0.24530205130577087, "learning_rate": 7.463924466867692e-05, "loss": 0.011, "step": 18460 }, { "grad_norm": 0.1790321320295334, "learning_rate": 7.461046390801121e-05, "loss": 0.0114, "step": 18470 }, { "grad_norm": 0.18571199476718903, "learning_rate": 7.458167238189248e-05, "loss": 0.0108, "step": 18480 }, { "grad_norm": 0.2013978362083435, "learning_rate": 7.455287010291509e-05, "loss": 0.0149, "step": 18490 }, { "grad_norm": 0.19386424124240875, "learning_rate": 7.452405708367818e-05, "loss": 0.0121, "step": 18500 }, { "grad_norm": 0.21166321635246277, "learning_rate": 7.449523333678549e-05, "loss": 0.0123, "step": 18510 }, { "grad_norm": 0.14023329317569733, "learning_rate": 7.446639887484555e-05, "loss": 0.0125, "step": 18520 }, { "grad_norm": 0.15196268260478973, "learning_rate": 7.443755371047152e-05, "loss": 0.0111, "step": 18530 }, { "grad_norm": 0.2037297934293747, "learning_rate": 7.440869785628124e-05, "loss": 0.013, "step": 18540 }, { "grad_norm": 0.19182361662387848, "learning_rate": 7.437983132489727e-05, "loss": 0.0141, "step": 18550 }, { "grad_norm": 0.1650853306055069, "learning_rate": 7.435095412894678e-05, "loss": 0.0122, "step": 18560 }, { "grad_norm": 0.1313398778438568, "learning_rate": 7.432206628106165e-05, "loss": 0.0109, "step": 18570 }, { "grad_norm": 0.14341309666633606, "learning_rate": 7.429316779387842e-05, "loss": 0.0094, "step": 18580 }, { "grad_norm": 0.15422917902469635, "learning_rate": 7.426425868003825e-05, "loss": 0.0113, "step": 18590 }, { "grad_norm": 0.16208846867084503, "learning_rate": 7.423533895218698e-05, "loss": 0.0107, "step": 18600 }, { "grad_norm": 0.16285823285579681, "learning_rate": 7.420640862297507e-05, "loss": 0.0131, "step": 18610 }, { "grad_norm": 0.16573621332645416, "learning_rate": 7.417746770505766e-05, "loss": 0.0139, "step": 18620 }, { "grad_norm": 0.21183575689792633, "learning_rate": 7.414851621109444e-05, "loss": 0.0138, "step": 18630 }, { "grad_norm": 0.1331263929605484, "learning_rate": 7.411955415374982e-05, "loss": 0.011, "step": 18640 }, { "grad_norm": 0.19629089534282684, "learning_rate": 7.409058154569278e-05, "loss": 0.0114, "step": 18650 }, { "grad_norm": 0.15443626046180725, "learning_rate": 7.406159839959689e-05, "loss": 0.0112, "step": 18660 }, { "grad_norm": 0.19833770394325256, "learning_rate": 7.403260472814039e-05, "loss": 0.0104, "step": 18670 }, { "grad_norm": 0.2044007033109665, "learning_rate": 7.400360054400611e-05, "loss": 0.0137, "step": 18680 }, { "grad_norm": 0.1870245635509491, "learning_rate": 7.397458585988143e-05, "loss": 0.0126, "step": 18690 }, { "grad_norm": 0.19683417677879333, "learning_rate": 7.394556068845841e-05, "loss": 0.0113, "step": 18700 }, { "grad_norm": 0.21472954750061035, "learning_rate": 7.39165250424336e-05, "loss": 0.0113, "step": 18710 }, { "grad_norm": 0.1248503252863884, "learning_rate": 7.38874789345082e-05, "loss": 0.01, "step": 18720 }, { "grad_norm": 0.18112768232822418, "learning_rate": 7.385842237738794e-05, "loss": 0.0137, "step": 18730 }, { "grad_norm": 0.21326950192451477, "learning_rate": 7.382935538378318e-05, "loss": 0.0108, "step": 18740 }, { "grad_norm": 0.1938982456922531, "learning_rate": 7.38002779664088e-05, "loss": 0.0112, "step": 18750 }, { "grad_norm": 0.15143512189388275, "learning_rate": 7.377119013798422e-05, "loss": 0.0086, "step": 18760 }, { "grad_norm": 0.21318207681179047, "learning_rate": 7.374209191123349e-05, "loss": 0.011, "step": 18770 }, { "grad_norm": 0.20381887257099152, "learning_rate": 7.371298329888512e-05, "loss": 0.0114, "step": 18780 }, { "grad_norm": 0.14822277426719666, "learning_rate": 7.368386431367223e-05, "loss": 0.0102, "step": 18790 }, { "grad_norm": 0.2085358202457428, "learning_rate": 7.365473496833246e-05, "loss": 0.0119, "step": 18800 }, { "grad_norm": 0.12606675922870636, "learning_rate": 7.362559527560797e-05, "loss": 0.0106, "step": 18810 }, { "grad_norm": 0.18155249953269958, "learning_rate": 7.359644524824545e-05, "loss": 0.0142, "step": 18820 }, { "grad_norm": 0.21405895054340363, "learning_rate": 7.35672848989961e-05, "loss": 0.0115, "step": 18830 }, { "grad_norm": 0.26266056299209595, "learning_rate": 7.353811424061565e-05, "loss": 0.0122, "step": 18840 }, { "grad_norm": 0.17924107611179352, "learning_rate": 7.350893328586435e-05, "loss": 0.0107, "step": 18850 }, { "grad_norm": 0.208436518907547, "learning_rate": 7.347974204750696e-05, "loss": 0.0106, "step": 18860 }, { "grad_norm": 0.14565600454807281, "learning_rate": 7.345054053831266e-05, "loss": 0.0097, "step": 18870 }, { "grad_norm": 0.18950042128562927, "learning_rate": 7.342132877105523e-05, "loss": 0.0121, "step": 18880 }, { "grad_norm": 0.165813609957695, "learning_rate": 7.339210675851288e-05, "loss": 0.0132, "step": 18890 }, { "grad_norm": 0.2116619497537613, "learning_rate": 7.336287451346831e-05, "loss": 0.0103, "step": 18900 }, { "grad_norm": 0.17606577277183533, "learning_rate": 7.333363204870868e-05, "loss": 0.0121, "step": 18910 }, { "grad_norm": 0.16686955094337463, "learning_rate": 7.330437937702566e-05, "loss": 0.0131, "step": 18920 }, { "grad_norm": 0.21528185904026031, "learning_rate": 7.327511651121534e-05, "loss": 0.0129, "step": 18930 }, { "grad_norm": 0.17814429104328156, "learning_rate": 7.324584346407833e-05, "loss": 0.0112, "step": 18940 }, { "grad_norm": 0.1568351686000824, "learning_rate": 7.321656024841962e-05, "loss": 0.0125, "step": 18950 }, { "grad_norm": 0.1489446759223938, "learning_rate": 7.31872668770487e-05, "loss": 0.0127, "step": 18960 }, { "grad_norm": 0.17591091990470886, "learning_rate": 7.315796336277949e-05, "loss": 0.0108, "step": 18970 }, { "grad_norm": 0.17621080577373505, "learning_rate": 7.312864971843031e-05, "loss": 0.0114, "step": 18980 }, { "grad_norm": 0.14916348457336426, "learning_rate": 7.309932595682398e-05, "loss": 0.0117, "step": 18990 }, { "grad_norm": 0.19298994541168213, "learning_rate": 7.30699920907877e-05, "loss": 0.0132, "step": 19000 }, { "grad_norm": 0.16541431844234467, "learning_rate": 7.304064813315308e-05, "loss": 0.0115, "step": 19010 }, { "grad_norm": 0.1469961255788803, "learning_rate": 7.30112940967562e-05, "loss": 0.0109, "step": 19020 }, { "grad_norm": 0.14893992245197296, "learning_rate": 7.298192999443747e-05, "loss": 0.0131, "step": 19030 }, { "grad_norm": 0.1689481884241104, "learning_rate": 7.295255583904179e-05, "loss": 0.0104, "step": 19040 }, { "grad_norm": 0.19867761433124542, "learning_rate": 7.29231716434184e-05, "loss": 0.0116, "step": 19050 }, { "grad_norm": 0.16729967296123505, "learning_rate": 7.289377742042092e-05, "loss": 0.0107, "step": 19060 }, { "grad_norm": 0.12557503581047058, "learning_rate": 7.286437318290741e-05, "loss": 0.0136, "step": 19070 }, { "grad_norm": 0.18714050948619843, "learning_rate": 7.283495894374027e-05, "loss": 0.0154, "step": 19080 }, { "grad_norm": 0.13715846836566925, "learning_rate": 7.280553471578629e-05, "loss": 0.0114, "step": 19090 }, { "grad_norm": 0.19969865679740906, "learning_rate": 7.277610051191663e-05, "loss": 0.0117, "step": 19100 }, { "grad_norm": 0.1655474454164505, "learning_rate": 7.27466563450068e-05, "loss": 0.011, "step": 19110 }, { "grad_norm": 0.18666283786296844, "learning_rate": 7.271720222793668e-05, "loss": 0.0129, "step": 19120 }, { "grad_norm": 0.18705947697162628, "learning_rate": 7.268773817359054e-05, "loss": 0.0107, "step": 19130 }, { "grad_norm": 0.19990399479866028, "learning_rate": 7.265826419485689e-05, "loss": 0.0111, "step": 19140 }, { "grad_norm": 0.1833992600440979, "learning_rate": 7.26287803046287e-05, "loss": 0.0125, "step": 19150 }, { "grad_norm": 0.2192019522190094, "learning_rate": 7.259928651580322e-05, "loss": 0.0106, "step": 19160 }, { "grad_norm": 0.2461443841457367, "learning_rate": 7.256978284128202e-05, "loss": 0.0119, "step": 19170 }, { "grad_norm": 0.22517727315425873, "learning_rate": 7.254026929397101e-05, "loss": 0.0137, "step": 19180 }, { "grad_norm": 0.15226878225803375, "learning_rate": 7.251074588678044e-05, "loss": 0.0124, "step": 19190 }, { "grad_norm": 0.14343887567520142, "learning_rate": 7.248121263262482e-05, "loss": 0.011, "step": 19200 }, { "grad_norm": 0.12526266276836395, "learning_rate": 7.245166954442304e-05, "loss": 0.0093, "step": 19210 }, { "grad_norm": 0.13337427377700806, "learning_rate": 7.24221166350982e-05, "loss": 0.0092, "step": 19220 }, { "grad_norm": 0.1772834062576294, "learning_rate": 7.239255391757777e-05, "loss": 0.0111, "step": 19230 }, { "grad_norm": 0.19824360311031342, "learning_rate": 7.236298140479351e-05, "loss": 0.0103, "step": 19240 }, { "grad_norm": 0.17644356191158295, "learning_rate": 7.233339910968141e-05, "loss": 0.0139, "step": 19250 }, { "grad_norm": 0.18103212118148804, "learning_rate": 7.230380704518178e-05, "loss": 0.0102, "step": 19260 }, { "grad_norm": 0.21373401582241058, "learning_rate": 7.227420522423919e-05, "loss": 0.0117, "step": 19270 }, { "grad_norm": 0.21635258197784424, "learning_rate": 7.224459365980249e-05, "loss": 0.0133, "step": 19280 }, { "grad_norm": 0.16442619264125824, "learning_rate": 7.221497236482479e-05, "loss": 0.0126, "step": 19290 }, { "grad_norm": 0.14004424214363098, "learning_rate": 7.218534135226345e-05, "loss": 0.0097, "step": 19300 }, { "grad_norm": 0.14560888707637787, "learning_rate": 7.215570063508005e-05, "loss": 0.0099, "step": 19310 }, { "grad_norm": 0.15902572870254517, "learning_rate": 7.21260502262405e-05, "loss": 0.0122, "step": 19320 }, { "grad_norm": 0.13771773874759674, "learning_rate": 7.209639013871487e-05, "loss": 0.0133, "step": 19330 }, { "grad_norm": 0.15242278575897217, "learning_rate": 7.206672038547746e-05, "loss": 0.0119, "step": 19340 }, { "grad_norm": 0.15112042427062988, "learning_rate": 7.203704097950687e-05, "loss": 0.0102, "step": 19350 }, { "grad_norm": 0.18977764248847961, "learning_rate": 7.200735193378587e-05, "loss": 0.0112, "step": 19360 }, { "grad_norm": 0.1985086053609848, "learning_rate": 7.197765326130145e-05, "loss": 0.0103, "step": 19370 }, { "grad_norm": 0.17211055755615234, "learning_rate": 7.194794497504481e-05, "loss": 0.0108, "step": 19380 }, { "grad_norm": 0.16687899827957153, "learning_rate": 7.19182270880114e-05, "loss": 0.0126, "step": 19390 }, { "grad_norm": 0.18941353261470795, "learning_rate": 7.188849961320079e-05, "loss": 0.0116, "step": 19400 }, { "grad_norm": 0.18274323642253876, "learning_rate": 7.185876256361681e-05, "loss": 0.013, "step": 19410 }, { "grad_norm": 0.14908260107040405, "learning_rate": 7.182901595226745e-05, "loss": 0.0099, "step": 19420 }, { "grad_norm": 0.1887970268726349, "learning_rate": 7.179925979216491e-05, "loss": 0.0104, "step": 19430 }, { "grad_norm": 0.18476919829845428, "learning_rate": 7.17694940963255e-05, "loss": 0.0107, "step": 19440 }, { "grad_norm": 0.19148029386997223, "learning_rate": 7.17397188777698e-05, "loss": 0.012, "step": 19450 }, { "grad_norm": 0.23468837141990662, "learning_rate": 7.170993414952245e-05, "loss": 0.0102, "step": 19460 }, { "grad_norm": 0.22746720910072327, "learning_rate": 7.168013992461233e-05, "loss": 0.0131, "step": 19470 }, { "grad_norm": 0.21804428100585938, "learning_rate": 7.165033621607245e-05, "loss": 0.0138, "step": 19480 }, { "grad_norm": 0.20864491164684296, "learning_rate": 7.162052303693995e-05, "loss": 0.0107, "step": 19490 }, { "grad_norm": 0.1613885462284088, "learning_rate": 7.159070040025614e-05, "loss": 0.0098, "step": 19500 }, { "grad_norm": 0.21238669753074646, "learning_rate": 7.156086831906646e-05, "loss": 0.0137, "step": 19510 }, { "grad_norm": 0.19478407502174377, "learning_rate": 7.153102680642044e-05, "loss": 0.0103, "step": 19520 }, { "grad_norm": 0.17616073787212372, "learning_rate": 7.150117587537182e-05, "loss": 0.0117, "step": 19530 }, { "grad_norm": 0.13292518258094788, "learning_rate": 7.147131553897838e-05, "loss": 0.0105, "step": 19540 }, { "grad_norm": 0.26289835572242737, "learning_rate": 7.144144581030205e-05, "loss": 0.0129, "step": 19550 }, { "grad_norm": 0.2064499855041504, "learning_rate": 7.141156670240889e-05, "loss": 0.0115, "step": 19560 }, { "grad_norm": 0.17051909863948822, "learning_rate": 7.138167822836899e-05, "loss": 0.0133, "step": 19570 }, { "grad_norm": 0.20623430609703064, "learning_rate": 7.135178040125661e-05, "loss": 0.0095, "step": 19580 }, { "grad_norm": 0.21441036462783813, "learning_rate": 7.13218732341501e-05, "loss": 0.0123, "step": 19590 }, { "grad_norm": 0.22051948308944702, "learning_rate": 7.129195674013181e-05, "loss": 0.0136, "step": 19600 }, { "grad_norm": 0.1402568817138672, "learning_rate": 7.126203093228829e-05, "loss": 0.0103, "step": 19610 }, { "grad_norm": 0.17667120695114136, "learning_rate": 7.123209582371006e-05, "loss": 0.0118, "step": 19620 }, { "grad_norm": 0.1691339612007141, "learning_rate": 7.120215142749179e-05, "loss": 0.0098, "step": 19630 }, { "grad_norm": 0.16157539188861847, "learning_rate": 7.117219775673218e-05, "loss": 0.0127, "step": 19640 }, { "grad_norm": 0.17221374809741974, "learning_rate": 7.114223482453394e-05, "loss": 0.0104, "step": 19650 }, { "grad_norm": 0.2151339054107666, "learning_rate": 7.11122626440039e-05, "loss": 0.0107, "step": 19660 }, { "grad_norm": 0.20373307168483734, "learning_rate": 7.10822812282529e-05, "loss": 0.0109, "step": 19670 }, { "grad_norm": 0.18632051348686218, "learning_rate": 7.105229059039586e-05, "loss": 0.0121, "step": 19680 }, { "grad_norm": 0.1845545917749405, "learning_rate": 7.102229074355168e-05, "loss": 0.01, "step": 19690 }, { "grad_norm": 0.15541477501392365, "learning_rate": 7.099228170084332e-05, "loss": 0.0088, "step": 19700 }, { "grad_norm": 0.18808230757713318, "learning_rate": 7.096226347539771e-05, "loss": 0.0123, "step": 19710 }, { "grad_norm": 0.1887749433517456, "learning_rate": 7.093223608034592e-05, "loss": 0.0097, "step": 19720 }, { "grad_norm": 0.20691046118736267, "learning_rate": 7.090219952882291e-05, "loss": 0.0146, "step": 19730 }, { "grad_norm": 0.19972388446331024, "learning_rate": 7.087215383396768e-05, "loss": 0.0112, "step": 19740 }, { "grad_norm": 0.20749326050281525, "learning_rate": 7.084209900892325e-05, "loss": 0.0108, "step": 19750 }, { "grad_norm": 0.2301788628101349, "learning_rate": 7.081203506683663e-05, "loss": 0.0102, "step": 19760 }, { "grad_norm": 0.221941739320755, "learning_rate": 7.078196202085879e-05, "loss": 0.0139, "step": 19770 }, { "grad_norm": 0.19017480313777924, "learning_rate": 7.075187988414473e-05, "loss": 0.0106, "step": 19780 }, { "grad_norm": 0.17279358208179474, "learning_rate": 7.072178866985334e-05, "loss": 0.0112, "step": 19790 }, { "grad_norm": 0.15703575313091278, "learning_rate": 7.069168839114764e-05, "loss": 0.009, "step": 19800 }, { "grad_norm": 0.16718357801437378, "learning_rate": 7.066157906119441e-05, "loss": 0.011, "step": 19810 }, { "grad_norm": 0.18201936781406403, "learning_rate": 7.063146069316456e-05, "loss": 0.0121, "step": 19820 }, { "grad_norm": 0.2503196597099304, "learning_rate": 7.060133330023287e-05, "loss": 0.0133, "step": 19830 }, { "grad_norm": 0.18651476502418518, "learning_rate": 7.057119689557807e-05, "loss": 0.0136, "step": 19840 }, { "grad_norm": 0.1670781672000885, "learning_rate": 7.054105149238287e-05, "loss": 0.011, "step": 19850 }, { "grad_norm": 0.25149157643318176, "learning_rate": 7.051089710383387e-05, "loss": 0.0144, "step": 19860 }, { "grad_norm": 0.17136549949645996, "learning_rate": 7.048073374312166e-05, "loss": 0.0142, "step": 19870 }, { "grad_norm": 0.18975761532783508, "learning_rate": 7.045056142344067e-05, "loss": 0.0131, "step": 19880 }, { "grad_norm": 0.20178429782390594, "learning_rate": 7.042038015798934e-05, "loss": 0.0117, "step": 19890 }, { "grad_norm": 0.18146274983882904, "learning_rate": 7.039018995996995e-05, "loss": 0.0122, "step": 19900 }, { "grad_norm": 0.18776217103004456, "learning_rate": 7.035999084258876e-05, "loss": 0.0144, "step": 19910 }, { "grad_norm": 0.22703418135643005, "learning_rate": 7.032978281905583e-05, "loss": 0.0107, "step": 19920 }, { "grad_norm": 0.17703606188297272, "learning_rate": 7.029956590258522e-05, "loss": 0.0113, "step": 19930 }, { "grad_norm": 0.1451052576303482, "learning_rate": 7.026934010639483e-05, "loss": 0.0122, "step": 19940 }, { "grad_norm": 0.17386291921138763, "learning_rate": 7.023910544370645e-05, "loss": 0.0105, "step": 19950 }, { "grad_norm": 0.23453392088413239, "learning_rate": 7.020886192774573e-05, "loss": 0.0126, "step": 19960 }, { "grad_norm": 0.2455233484506607, "learning_rate": 7.017860957174222e-05, "loss": 0.014, "step": 19970 }, { "grad_norm": 0.15349385142326355, "learning_rate": 7.014834838892934e-05, "loss": 0.0094, "step": 19980 }, { "grad_norm": 0.13492955267429352, "learning_rate": 7.011807839254435e-05, "loss": 0.0097, "step": 19990 }, { "grad_norm": 0.2587132155895233, "learning_rate": 7.008779959582837e-05, "loss": 0.0127, "step": 20000 }, { "grad_norm": 0.2557176649570465, "learning_rate": 7.005751201202639e-05, "loss": 0.0149, "step": 20010 }, { "grad_norm": 0.16680429875850677, "learning_rate": 7.002721565438724e-05, "loss": 0.0113, "step": 20020 }, { "grad_norm": 0.1871728003025055, "learning_rate": 6.999691053616353e-05, "loss": 0.0115, "step": 20030 }, { "grad_norm": 0.21564066410064697, "learning_rate": 6.996659667061178e-05, "loss": 0.0106, "step": 20040 }, { "grad_norm": 0.1538303941488266, "learning_rate": 6.993627407099229e-05, "loss": 0.0098, "step": 20050 }, { "grad_norm": 0.12237869203090668, "learning_rate": 6.990594275056921e-05, "loss": 0.0113, "step": 20060 }, { "grad_norm": 0.16063931584358215, "learning_rate": 6.987560272261049e-05, "loss": 0.0123, "step": 20070 }, { "grad_norm": 0.19766253232955933, "learning_rate": 6.984525400038788e-05, "loss": 0.01, "step": 20080 }, { "grad_norm": 0.1990915685892105, "learning_rate": 6.981489659717693e-05, "loss": 0.0133, "step": 20090 }, { "grad_norm": 0.1967361867427826, "learning_rate": 6.978453052625703e-05, "loss": 0.0155, "step": 20100 }, { "grad_norm": 0.2124536633491516, "learning_rate": 6.975415580091131e-05, "loss": 0.0129, "step": 20110 }, { "grad_norm": 0.20776978135108948, "learning_rate": 6.972377243442672e-05, "loss": 0.0135, "step": 20120 }, { "grad_norm": 0.18071985244750977, "learning_rate": 6.969338044009396e-05, "loss": 0.0123, "step": 20130 }, { "grad_norm": 0.22488372027873993, "learning_rate": 6.966297983120753e-05, "loss": 0.0115, "step": 20140 }, { "grad_norm": 0.22876937687397003, "learning_rate": 6.963257062106571e-05, "loss": 0.0146, "step": 20150 }, { "grad_norm": 0.2054663598537445, "learning_rate": 6.960215282297048e-05, "loss": 0.0102, "step": 20160 }, { "grad_norm": 0.21780245006084442, "learning_rate": 6.957172645022765e-05, "loss": 0.0097, "step": 20170 }, { "grad_norm": 0.13549740612506866, "learning_rate": 6.954129151614673e-05, "loss": 0.0105, "step": 20180 }, { "grad_norm": 0.14841488003730774, "learning_rate": 6.9510848034041e-05, "loss": 0.0107, "step": 20190 }, { "grad_norm": 0.14922398328781128, "learning_rate": 6.94803960172275e-05, "loss": 0.0096, "step": 20200 }, { "grad_norm": 0.1702589988708496, "learning_rate": 6.944993547902692e-05, "loss": 0.0112, "step": 20210 }, { "grad_norm": 0.2228366583585739, "learning_rate": 6.941946643276379e-05, "loss": 0.012, "step": 20220 }, { "grad_norm": 0.2307967096567154, "learning_rate": 6.938898889176626e-05, "loss": 0.009, "step": 20230 }, { "grad_norm": 0.2519766390323639, "learning_rate": 6.935850286936627e-05, "loss": 0.0098, "step": 20240 }, { "grad_norm": 0.17736345529556274, "learning_rate": 6.932800837889943e-05, "loss": 0.0123, "step": 20250 }, { "grad_norm": 0.2867749333381653, "learning_rate": 6.92975054337051e-05, "loss": 0.0137, "step": 20260 }, { "grad_norm": 0.2639591097831726, "learning_rate": 6.926699404712627e-05, "loss": 0.01, "step": 20270 }, { "grad_norm": 0.23023684322834015, "learning_rate": 6.923647423250967e-05, "loss": 0.0111, "step": 20280 }, { "grad_norm": 0.17157422006130219, "learning_rate": 6.920594600320567e-05, "loss": 0.0109, "step": 20290 }, { "grad_norm": 0.15140603482723236, "learning_rate": 6.91754093725684e-05, "loss": 0.01, "step": 20300 }, { "grad_norm": 0.15259721875190735, "learning_rate": 6.914486435395561e-05, "loss": 0.0146, "step": 20310 }, { "grad_norm": 0.2087843418121338, "learning_rate": 6.911431096072871e-05, "loss": 0.0106, "step": 20320 }, { "grad_norm": 0.19032873213291168, "learning_rate": 6.90837492062528e-05, "loss": 0.0116, "step": 20330 }, { "grad_norm": 0.14664989709854126, "learning_rate": 6.905317910389664e-05, "loss": 0.0113, "step": 20340 }, { "grad_norm": 0.2548518180847168, "learning_rate": 6.902260066703262e-05, "loss": 0.0129, "step": 20350 }, { "grad_norm": 0.1683628410100937, "learning_rate": 6.89920139090368e-05, "loss": 0.011, "step": 20360 }, { "grad_norm": 0.2066972851753235, "learning_rate": 6.896141884328886e-05, "loss": 0.0116, "step": 20370 }, { "grad_norm": 0.19190669059753418, "learning_rate": 6.893081548317211e-05, "loss": 0.0125, "step": 20380 }, { "grad_norm": 0.15936867892742157, "learning_rate": 6.890020384207353e-05, "loss": 0.0128, "step": 20390 }, { "grad_norm": 0.10639575868844986, "learning_rate": 6.886958393338366e-05, "loss": 0.0122, "step": 20400 }, { "grad_norm": 0.1789289116859436, "learning_rate": 6.883895577049668e-05, "loss": 0.011, "step": 20410 }, { "grad_norm": 0.1584610790014267, "learning_rate": 6.880831936681045e-05, "loss": 0.0077, "step": 20420 }, { "grad_norm": 0.13735057413578033, "learning_rate": 6.87776747357263e-05, "loss": 0.0085, "step": 20430 }, { "grad_norm": 0.16049465537071228, "learning_rate": 6.874702189064927e-05, "loss": 0.0118, "step": 20440 }, { "grad_norm": 0.179994136095047, "learning_rate": 6.871636084498796e-05, "loss": 0.0108, "step": 20450 }, { "grad_norm": 0.1767672747373581, "learning_rate": 6.868569161215453e-05, "loss": 0.0118, "step": 20460 }, { "grad_norm": 0.22899216413497925, "learning_rate": 6.865501420556477e-05, "loss": 0.0109, "step": 20470 }, { "grad_norm": 0.21053467690944672, "learning_rate": 6.862432863863802e-05, "loss": 0.0122, "step": 20480 }, { "grad_norm": 0.20258045196533203, "learning_rate": 6.859363492479716e-05, "loss": 0.0138, "step": 20490 }, { "grad_norm": 0.1911376267671585, "learning_rate": 6.856293307746868e-05, "loss": 0.0114, "step": 20500 }, { "grad_norm": 0.20449240505695343, "learning_rate": 6.853222311008263e-05, "loss": 0.0109, "step": 20510 }, { "grad_norm": 0.1852523237466812, "learning_rate": 6.850150503607256e-05, "loss": 0.0159, "step": 20520 }, { "grad_norm": 0.21257971227169037, "learning_rate": 6.847077886887562e-05, "loss": 0.0132, "step": 20530 }, { "grad_norm": 0.14600904285907745, "learning_rate": 6.844004462193248e-05, "loss": 0.0087, "step": 20540 }, { "grad_norm": 0.20207323133945465, "learning_rate": 6.840930230868734e-05, "loss": 0.0121, "step": 20550 }, { "grad_norm": 0.15506303310394287, "learning_rate": 6.837855194258791e-05, "loss": 0.0136, "step": 20560 }, { "grad_norm": 0.20277737081050873, "learning_rate": 6.834779353708548e-05, "loss": 0.0118, "step": 20570 }, { "grad_norm": 0.17725859582424164, "learning_rate": 6.831702710563478e-05, "loss": 0.0126, "step": 20580 }, { "grad_norm": 0.11764238774776459, "learning_rate": 6.828625266169416e-05, "loss": 0.0106, "step": 20590 }, { "grad_norm": 0.14060883224010468, "learning_rate": 6.825547021872534e-05, "loss": 0.01, "step": 20600 }, { "grad_norm": 0.15700827538967133, "learning_rate": 6.822467979019365e-05, "loss": 0.0114, "step": 20610 }, { "grad_norm": 0.1654394119977951, "learning_rate": 6.819388138956783e-05, "loss": 0.0085, "step": 20620 }, { "grad_norm": 0.1699107587337494, "learning_rate": 6.816307503032019e-05, "loss": 0.0116, "step": 20630 }, { "grad_norm": 0.21599388122558594, "learning_rate": 6.813226072592645e-05, "loss": 0.0114, "step": 20640 }, { "grad_norm": 0.2059008777141571, "learning_rate": 6.810143848986583e-05, "loss": 0.013, "step": 20650 }, { "grad_norm": 0.1566775143146515, "learning_rate": 6.807060833562106e-05, "loss": 0.014, "step": 20660 }, { "grad_norm": 0.15280960500240326, "learning_rate": 6.803977027667825e-05, "loss": 0.0138, "step": 20670 }, { "grad_norm": 0.20486606657505035, "learning_rate": 6.800892432652704e-05, "loss": 0.0135, "step": 20680 }, { "grad_norm": 0.1487714946269989, "learning_rate": 6.797807049866049e-05, "loss": 0.012, "step": 20690 }, { "grad_norm": 0.11399101465940475, "learning_rate": 6.794720880657512e-05, "loss": 0.0128, "step": 20700 }, { "grad_norm": 0.15832512080669403, "learning_rate": 6.791633926377085e-05, "loss": 0.0122, "step": 20710 }, { "grad_norm": 0.15588177740573883, "learning_rate": 6.788546188375114e-05, "loss": 0.0108, "step": 20720 }, { "grad_norm": 0.1265900731086731, "learning_rate": 6.785457668002274e-05, "loss": 0.0129, "step": 20730 }, { "grad_norm": 0.22834330797195435, "learning_rate": 6.78236836660959e-05, "loss": 0.0139, "step": 20740 }, { "grad_norm": 0.18703217804431915, "learning_rate": 6.779278285548427e-05, "loss": 0.0116, "step": 20750 }, { "grad_norm": 0.23664888739585876, "learning_rate": 6.776187426170493e-05, "loss": 0.0124, "step": 20760 }, { "grad_norm": 0.24347366392612457, "learning_rate": 6.773095789827837e-05, "loss": 0.0144, "step": 20770 }, { "grad_norm": 0.1955549418926239, "learning_rate": 6.770003377872841e-05, "loss": 0.0094, "step": 20780 }, { "grad_norm": 0.17129245400428772, "learning_rate": 6.766910191658233e-05, "loss": 0.0111, "step": 20790 }, { "grad_norm": 0.19179879128932953, "learning_rate": 6.763816232537079e-05, "loss": 0.0103, "step": 20800 }, { "grad_norm": 0.19190002977848053, "learning_rate": 6.76072150186278e-05, "loss": 0.0096, "step": 20810 }, { "grad_norm": 0.2360352724790573, "learning_rate": 6.757626000989077e-05, "loss": 0.012, "step": 20820 }, { "grad_norm": 0.20795121788978577, "learning_rate": 6.754529731270048e-05, "loss": 0.0125, "step": 20830 }, { "grad_norm": 0.17308831214904785, "learning_rate": 6.751432694060105e-05, "loss": 0.0108, "step": 20840 }, { "grad_norm": 0.17128589749336243, "learning_rate": 6.748334890713999e-05, "loss": 0.0114, "step": 20850 }, { "grad_norm": 0.17323458194732666, "learning_rate": 6.745236322586813e-05, "loss": 0.0118, "step": 20860 }, { "grad_norm": 0.1393457055091858, "learning_rate": 6.742136991033967e-05, "loss": 0.0092, "step": 20870 }, { "grad_norm": 0.17811328172683716, "learning_rate": 6.739036897411215e-05, "loss": 0.0104, "step": 20880 }, { "grad_norm": 0.13496756553649902, "learning_rate": 6.73593604307464e-05, "loss": 0.0096, "step": 20890 }, { "grad_norm": 0.1577647626399994, "learning_rate": 6.732834429380663e-05, "loss": 0.0088, "step": 20900 }, { "grad_norm": 0.19083021581172943, "learning_rate": 6.729732057686036e-05, "loss": 0.0116, "step": 20910 }, { "grad_norm": 0.21718701720237732, "learning_rate": 6.726628929347839e-05, "loss": 0.0105, "step": 20920 }, { "grad_norm": 0.15996325016021729, "learning_rate": 6.723525045723488e-05, "loss": 0.0105, "step": 20930 }, { "grad_norm": 0.22548678517341614, "learning_rate": 6.720420408170728e-05, "loss": 0.0137, "step": 20940 }, { "grad_norm": 0.1759725660085678, "learning_rate": 6.71731501804763e-05, "loss": 0.0104, "step": 20950 }, { "grad_norm": 0.18935760855674744, "learning_rate": 6.714208876712602e-05, "loss": 0.0105, "step": 20960 }, { "grad_norm": 0.17410387098789215, "learning_rate": 6.71110198552437e-05, "loss": 0.0104, "step": 20970 }, { "grad_norm": 0.1575106829404831, "learning_rate": 6.707994345841996e-05, "loss": 0.0115, "step": 20980 }, { "grad_norm": 0.17498980462551117, "learning_rate": 6.70488595902487e-05, "loss": 0.0094, "step": 20990 }, { "grad_norm": 0.17812250554561615, "learning_rate": 6.7017768264327e-05, "loss": 0.0134, "step": 21000 }, { "grad_norm": 0.16137534379959106, "learning_rate": 6.698666949425532e-05, "loss": 0.0106, "step": 21010 }, { "grad_norm": 0.2201918363571167, "learning_rate": 6.69555632936373e-05, "loss": 0.0126, "step": 21020 }, { "grad_norm": 0.17056263983249664, "learning_rate": 6.692444967607983e-05, "loss": 0.0093, "step": 21030 }, { "grad_norm": 0.18417535722255707, "learning_rate": 6.689332865519309e-05, "loss": 0.0111, "step": 21040 }, { "grad_norm": 0.20942354202270508, "learning_rate": 6.686220024459045e-05, "loss": 0.0137, "step": 21050 }, { "grad_norm": 0.18547572195529938, "learning_rate": 6.683106445788856e-05, "loss": 0.0098, "step": 21060 }, { "grad_norm": 0.1989092379808426, "learning_rate": 6.679992130870723e-05, "loss": 0.0136, "step": 21070 }, { "grad_norm": 0.1380550116300583, "learning_rate": 6.676877081066957e-05, "loss": 0.0108, "step": 21080 }, { "grad_norm": 0.20721712708473206, "learning_rate": 6.673761297740186e-05, "loss": 0.0112, "step": 21090 }, { "grad_norm": 0.1794164478778839, "learning_rate": 6.670644782253358e-05, "loss": 0.0112, "step": 21100 }, { "grad_norm": 0.16613364219665527, "learning_rate": 6.667527535969744e-05, "loss": 0.0124, "step": 21110 }, { "grad_norm": 0.1685854196548462, "learning_rate": 6.664409560252933e-05, "loss": 0.0101, "step": 21120 }, { "grad_norm": 0.18442267179489136, "learning_rate": 6.661290856466833e-05, "loss": 0.0119, "step": 21130 }, { "grad_norm": 0.21810148656368256, "learning_rate": 6.658171425975672e-05, "loss": 0.0114, "step": 21140 }, { "grad_norm": 0.2250090092420578, "learning_rate": 6.655051270143994e-05, "loss": 0.0128, "step": 21150 }, { "grad_norm": 0.19217421114444733, "learning_rate": 6.651930390336662e-05, "loss": 0.0104, "step": 21160 }, { "grad_norm": 0.11608275026082993, "learning_rate": 6.648808787918854e-05, "loss": 0.0112, "step": 21170 }, { "grad_norm": 0.16948024928569794, "learning_rate": 6.645686464256065e-05, "loss": 0.0118, "step": 21180 }, { "grad_norm": 0.1503266841173172, "learning_rate": 6.642563420714106e-05, "loss": 0.0127, "step": 21190 }, { "grad_norm": 0.21751253306865692, "learning_rate": 6.639439658659102e-05, "loss": 0.0115, "step": 21200 }, { "grad_norm": 0.2414206564426422, "learning_rate": 6.636315179457492e-05, "loss": 0.0099, "step": 21210 }, { "grad_norm": 0.1922888606786728, "learning_rate": 6.633189984476031e-05, "loss": 0.0118, "step": 21220 }, { "grad_norm": 0.14598584175109863, "learning_rate": 6.630064075081785e-05, "loss": 0.0108, "step": 21230 }, { "grad_norm": 0.166251540184021, "learning_rate": 6.626937452642132e-05, "loss": 0.0101, "step": 21240 }, { "grad_norm": 0.19690532982349396, "learning_rate": 6.623810118524765e-05, "loss": 0.0098, "step": 21250 }, { "grad_norm": 0.18482495844364166, "learning_rate": 6.620682074097682e-05, "loss": 0.01, "step": 21260 }, { "grad_norm": 0.15995000302791595, "learning_rate": 6.6175533207292e-05, "loss": 0.0128, "step": 21270 }, { "grad_norm": 0.18622662127017975, "learning_rate": 6.61442385978794e-05, "loss": 0.011, "step": 21280 }, { "grad_norm": 0.17273657023906708, "learning_rate": 6.611293692642836e-05, "loss": 0.0122, "step": 21290 }, { "grad_norm": 0.2058594524860382, "learning_rate": 6.608162820663128e-05, "loss": 0.0105, "step": 21300 }, { "grad_norm": 0.2248818278312683, "learning_rate": 6.605031245218365e-05, "loss": 0.0131, "step": 21310 }, { "grad_norm": 0.16864942014217377, "learning_rate": 6.601898967678408e-05, "loss": 0.0115, "step": 21320 }, { "grad_norm": 0.14728300273418427, "learning_rate": 6.598765989413419e-05, "loss": 0.0076, "step": 21330 }, { "grad_norm": 0.16560344398021698, "learning_rate": 6.595632311793867e-05, "loss": 0.012, "step": 21340 }, { "grad_norm": 0.1930578052997589, "learning_rate": 6.592497936190534e-05, "loss": 0.0106, "step": 21350 }, { "grad_norm": 0.2529297173023224, "learning_rate": 6.5893628639745e-05, "loss": 0.0102, "step": 21360 }, { "grad_norm": 0.19682936370372772, "learning_rate": 6.586227096517151e-05, "loss": 0.0093, "step": 21370 }, { "grad_norm": 0.1323523074388504, "learning_rate": 6.58309063519018e-05, "loss": 0.008, "step": 21380 }, { "grad_norm": 0.13490447402000427, "learning_rate": 6.579953481365582e-05, "loss": 0.0108, "step": 21390 }, { "grad_norm": 0.2003723680973053, "learning_rate": 6.576815636415652e-05, "loss": 0.0104, "step": 21400 }, { "grad_norm": 0.14707720279693604, "learning_rate": 6.573677101712991e-05, "loss": 0.0108, "step": 21410 }, { "grad_norm": 0.1835016906261444, "learning_rate": 6.570537878630502e-05, "loss": 0.0112, "step": 21420 }, { "grad_norm": 0.17271868884563446, "learning_rate": 6.56739796854139e-05, "loss": 0.0108, "step": 21430 }, { "grad_norm": 0.19244974851608276, "learning_rate": 6.564257372819153e-05, "loss": 0.0115, "step": 21440 }, { "grad_norm": 0.1960185319185257, "learning_rate": 6.561116092837596e-05, "loss": 0.0145, "step": 21450 }, { "grad_norm": 0.13664588332176208, "learning_rate": 6.557974129970825e-05, "loss": 0.009, "step": 21460 }, { "grad_norm": 0.1453867405653, "learning_rate": 6.554831485593239e-05, "loss": 0.0082, "step": 21470 }, { "grad_norm": 0.14215624332427979, "learning_rate": 6.551688161079534e-05, "loss": 0.0092, "step": 21480 }, { "grad_norm": 0.1641642451286316, "learning_rate": 6.548544157804712e-05, "loss": 0.0098, "step": 21490 }, { "grad_norm": 0.12978896498680115, "learning_rate": 6.545399477144065e-05, "loss": 0.0119, "step": 21500 }, { "grad_norm": 0.2904324233531952, "learning_rate": 6.542254120473184e-05, "loss": 0.0089, "step": 21510 }, { "grad_norm": 0.15688800811767578, "learning_rate": 6.539108089167953e-05, "loss": 0.0106, "step": 21520 }, { "grad_norm": 0.13740496337413788, "learning_rate": 6.535961384604554e-05, "loss": 0.0102, "step": 21530 }, { "grad_norm": 0.15430253744125366, "learning_rate": 6.532814008159461e-05, "loss": 0.0117, "step": 21540 }, { "grad_norm": 0.18426825106143951, "learning_rate": 6.529665961209446e-05, "loss": 0.0104, "step": 21550 }, { "grad_norm": 0.153671994805336, "learning_rate": 6.52651724513157e-05, "loss": 0.0102, "step": 21560 }, { "grad_norm": 0.19146676361560822, "learning_rate": 6.523367861303189e-05, "loss": 0.0098, "step": 21570 }, { "grad_norm": 0.16493569314479828, "learning_rate": 6.520217811101951e-05, "loss": 0.0126, "step": 21580 }, { "grad_norm": 0.17784230411052704, "learning_rate": 6.517067095905793e-05, "loss": 0.0116, "step": 21590 }, { "grad_norm": 0.1521739810705185, "learning_rate": 6.513915717092948e-05, "loss": 0.0134, "step": 21600 }, { "grad_norm": 0.19599667191505432, "learning_rate": 6.510763676041935e-05, "loss": 0.011, "step": 21610 }, { "grad_norm": 0.16778433322906494, "learning_rate": 6.50761097413156e-05, "loss": 0.0114, "step": 21620 }, { "grad_norm": 0.17359516024589539, "learning_rate": 6.504457612740928e-05, "loss": 0.011, "step": 21630 }, { "grad_norm": 0.1705867201089859, "learning_rate": 6.501303593249423e-05, "loss": 0.0115, "step": 21640 }, { "grad_norm": 0.16556194424629211, "learning_rate": 6.49814891703672e-05, "loss": 0.0097, "step": 21650 }, { "grad_norm": 0.25361791253089905, "learning_rate": 6.494993585482783e-05, "loss": 0.0113, "step": 21660 }, { "grad_norm": 0.15869159996509552, "learning_rate": 6.49183759996786e-05, "loss": 0.0113, "step": 21670 }, { "grad_norm": 0.16147175431251526, "learning_rate": 6.488680961872486e-05, "loss": 0.0112, "step": 21680 }, { "grad_norm": 0.19576740264892578, "learning_rate": 6.485523672577485e-05, "loss": 0.0114, "step": 21690 }, { "grad_norm": 0.1917080134153366, "learning_rate": 6.482365733463959e-05, "loss": 0.0111, "step": 21700 }, { "grad_norm": 0.1329818218946457, "learning_rate": 6.4792071459133e-05, "loss": 0.0105, "step": 21710 }, { "grad_norm": 0.1917017251253128, "learning_rate": 6.476047911307179e-05, "loss": 0.0154, "step": 21720 }, { "grad_norm": 0.136601060628891, "learning_rate": 6.472888031027556e-05, "loss": 0.0124, "step": 21730 }, { "grad_norm": 0.1730380356311798, "learning_rate": 6.469727506456666e-05, "loss": 0.0097, "step": 21740 }, { "grad_norm": 0.13472667336463928, "learning_rate": 6.466566338977034e-05, "loss": 0.01, "step": 21750 }, { "grad_norm": 0.16887542605400085, "learning_rate": 6.46340452997146e-05, "loss": 0.0109, "step": 21760 }, { "grad_norm": 0.19112178683280945, "learning_rate": 6.460242080823025e-05, "loss": 0.013, "step": 21770 }, { "grad_norm": 0.2168537676334381, "learning_rate": 6.457078992915093e-05, "loss": 0.0103, "step": 21780 }, { "grad_norm": 0.16052329540252686, "learning_rate": 6.453915267631308e-05, "loss": 0.01, "step": 21790 }, { "grad_norm": 0.2119000405073166, "learning_rate": 6.450750906355589e-05, "loss": 0.0096, "step": 21800 }, { "grad_norm": 0.14916768670082092, "learning_rate": 6.447585910472134e-05, "loss": 0.0105, "step": 21810 }, { "grad_norm": 0.1469573676586151, "learning_rate": 6.444420281365424e-05, "loss": 0.0088, "step": 21820 }, { "grad_norm": 0.16011612117290497, "learning_rate": 6.441254020420208e-05, "loss": 0.0106, "step": 21830 }, { "grad_norm": 0.14239658415317535, "learning_rate": 6.438087129021522e-05, "loss": 0.009, "step": 21840 }, { "grad_norm": 0.22075901925563812, "learning_rate": 6.434919608554663e-05, "loss": 0.0121, "step": 21850 }, { "grad_norm": 0.14896665513515472, "learning_rate": 6.43175146040522e-05, "loss": 0.0101, "step": 21860 }, { "grad_norm": 0.1685275435447693, "learning_rate": 6.428582685959045e-05, "loss": 0.0112, "step": 21870 }, { "grad_norm": 0.18524758517742157, "learning_rate": 6.42541328660227e-05, "loss": 0.0093, "step": 21880 }, { "grad_norm": 0.1714809089899063, "learning_rate": 6.422243263721293e-05, "loss": 0.0103, "step": 21890 }, { "grad_norm": 0.1477171629667282, "learning_rate": 6.419072618702794e-05, "loss": 0.0106, "step": 21900 }, { "grad_norm": 0.1463361233472824, "learning_rate": 6.41590135293372e-05, "loss": 0.0102, "step": 21910 }, { "grad_norm": 0.16443908214569092, "learning_rate": 6.41272946780129e-05, "loss": 0.0096, "step": 21920 }, { "grad_norm": 0.1701793074607849, "learning_rate": 6.409556964692996e-05, "loss": 0.0103, "step": 21930 }, { "grad_norm": 0.14943654835224152, "learning_rate": 6.406383844996595e-05, "loss": 0.011, "step": 21940 }, { "grad_norm": 0.17078867554664612, "learning_rate": 6.403210110100121e-05, "loss": 0.0079, "step": 21950 }, { "grad_norm": 0.1703493297100067, "learning_rate": 6.400035761391871e-05, "loss": 0.0091, "step": 21960 }, { "grad_norm": 0.12790130078792572, "learning_rate": 6.396860800260416e-05, "loss": 0.012, "step": 21970 }, { "grad_norm": 0.16735929250717163, "learning_rate": 6.393685228094588e-05, "loss": 0.0116, "step": 21980 }, { "grad_norm": 0.15113922953605652, "learning_rate": 6.390509046283493e-05, "loss": 0.0102, "step": 21990 }, { "grad_norm": 0.19218267500400543, "learning_rate": 6.387332256216498e-05, "loss": 0.0084, "step": 22000 }, { "grad_norm": 0.09919473528862, "learning_rate": 6.384154859283243e-05, "loss": 0.0092, "step": 22010 }, { "grad_norm": 0.16832999885082245, "learning_rate": 6.380976856873627e-05, "loss": 0.0099, "step": 22020 }, { "grad_norm": 0.19938506186008453, "learning_rate": 6.377798250377815e-05, "loss": 0.0094, "step": 22030 }, { "grad_norm": 0.21254943311214447, "learning_rate": 6.374619041186238e-05, "loss": 0.0111, "step": 22040 }, { "grad_norm": 0.16009533405303955, "learning_rate": 6.371439230689592e-05, "loss": 0.0099, "step": 22050 }, { "grad_norm": 0.136356383562088, "learning_rate": 6.368258820278833e-05, "loss": 0.0085, "step": 22060 }, { "grad_norm": 0.1945640593767166, "learning_rate": 6.365077811345182e-05, "loss": 0.0102, "step": 22070 }, { "grad_norm": 0.15300767123699188, "learning_rate": 6.361896205280117e-05, "loss": 0.009, "step": 22080 }, { "grad_norm": 0.21733640134334564, "learning_rate": 6.358714003475384e-05, "loss": 0.0105, "step": 22090 }, { "grad_norm": 0.19893081486225128, "learning_rate": 6.355531207322983e-05, "loss": 0.0128, "step": 22100 }, { "grad_norm": 0.15497083961963654, "learning_rate": 6.35234781821518e-05, "loss": 0.0097, "step": 22110 }, { "grad_norm": 0.14221924543380737, "learning_rate": 6.349163837544497e-05, "loss": 0.0095, "step": 22120 }, { "grad_norm": 0.17029717564582825, "learning_rate": 6.345979266703714e-05, "loss": 0.0129, "step": 22130 }, { "grad_norm": 0.14612993597984314, "learning_rate": 6.342794107085872e-05, "loss": 0.0101, "step": 22140 }, { "grad_norm": 0.11504451930522919, "learning_rate": 6.339608360084265e-05, "loss": 0.0092, "step": 22150 }, { "grad_norm": 0.17563201487064362, "learning_rate": 6.336422027092452e-05, "loss": 0.0097, "step": 22160 }, { "grad_norm": 0.12266577035188675, "learning_rate": 6.333235109504241e-05, "loss": 0.0101, "step": 22170 }, { "grad_norm": 0.13025866448879242, "learning_rate": 6.330047608713695e-05, "loss": 0.0105, "step": 22180 }, { "grad_norm": 0.1562608927488327, "learning_rate": 6.32685952611514e-05, "loss": 0.0092, "step": 22190 }, { "grad_norm": 0.14903868734836578, "learning_rate": 6.323670863103149e-05, "loss": 0.0096, "step": 22200 }, { "grad_norm": 0.16525746881961823, "learning_rate": 6.32048162107255e-05, "loss": 0.0101, "step": 22210 }, { "grad_norm": 0.14443136751651764, "learning_rate": 6.317291801418431e-05, "loss": 0.0117, "step": 22220 }, { "grad_norm": 0.14665141701698303, "learning_rate": 6.31410140553612e-05, "loss": 0.0098, "step": 22230 }, { "grad_norm": 0.20483767986297607, "learning_rate": 6.310910434821211e-05, "loss": 0.011, "step": 22240 }, { "grad_norm": 0.1462428867816925, "learning_rate": 6.307718890669541e-05, "loss": 0.0142, "step": 22250 }, { "grad_norm": 0.20139579474925995, "learning_rate": 6.304526774477198e-05, "loss": 0.0119, "step": 22260 }, { "grad_norm": 0.15433475375175476, "learning_rate": 6.301334087640526e-05, "loss": 0.0093, "step": 22270 }, { "grad_norm": 0.19898121058940887, "learning_rate": 6.298140831556112e-05, "loss": 0.0099, "step": 22280 }, { "grad_norm": 0.16369757056236267, "learning_rate": 6.294947007620794e-05, "loss": 0.0103, "step": 22290 }, { "grad_norm": 0.2097531408071518, "learning_rate": 6.291752617231661e-05, "loss": 0.0084, "step": 22300 }, { "grad_norm": 0.17679183185100555, "learning_rate": 6.288557661786047e-05, "loss": 0.01, "step": 22310 }, { "grad_norm": 0.22262082993984222, "learning_rate": 6.285362142681534e-05, "loss": 0.0095, "step": 22320 }, { "grad_norm": 0.1916545182466507, "learning_rate": 6.28216606131595e-05, "loss": 0.0133, "step": 22330 }, { "grad_norm": 0.15567471086978912, "learning_rate": 6.27896941908737e-05, "loss": 0.0102, "step": 22340 }, { "grad_norm": 0.12696465849876404, "learning_rate": 6.275772217394115e-05, "loss": 0.009, "step": 22350 }, { "grad_norm": 0.157383993268013, "learning_rate": 6.27257445763475e-05, "loss": 0.0096, "step": 22360 }, { "grad_norm": 0.14419983327388763, "learning_rate": 6.269376141208082e-05, "loss": 0.0099, "step": 22370 }, { "grad_norm": 0.15996843576431274, "learning_rate": 6.266177269513162e-05, "loss": 0.0084, "step": 22380 }, { "grad_norm": 0.17143651843070984, "learning_rate": 6.262977843949289e-05, "loss": 0.0116, "step": 22390 }, { "grad_norm": 0.22546066343784332, "learning_rate": 6.259777865915997e-05, "loss": 0.012, "step": 22400 }, { "grad_norm": 0.20518095791339874, "learning_rate": 6.256577336813069e-05, "loss": 0.0096, "step": 22410 }, { "grad_norm": 0.12031462788581848, "learning_rate": 6.253376258040521e-05, "loss": 0.0116, "step": 22420 }, { "grad_norm": 0.22356678545475006, "learning_rate": 6.250174630998616e-05, "loss": 0.0125, "step": 22430 }, { "grad_norm": 0.16171026229858398, "learning_rate": 6.246972457087854e-05, "loss": 0.0126, "step": 22440 }, { "grad_norm": 0.16515053808689117, "learning_rate": 6.243769737708974e-05, "loss": 0.0093, "step": 22450 }, { "grad_norm": 0.21975581347942352, "learning_rate": 6.240566474262955e-05, "loss": 0.0092, "step": 22460 }, { "grad_norm": 0.1697973757982254, "learning_rate": 6.237362668151012e-05, "loss": 0.0097, "step": 22470 }, { "grad_norm": 0.18939056992530823, "learning_rate": 6.2341583207746e-05, "loss": 0.0095, "step": 22480 }, { "grad_norm": 0.14284341037273407, "learning_rate": 6.23095343353541e-05, "loss": 0.0098, "step": 22490 }, { "grad_norm": 0.11992690712213516, "learning_rate": 6.227748007835366e-05, "loss": 0.0103, "step": 22500 }, { "grad_norm": 0.12769977748394012, "learning_rate": 6.224542045076632e-05, "loss": 0.0097, "step": 22510 }, { "grad_norm": 0.10506825894117355, "learning_rate": 6.221335546661606e-05, "loss": 0.0095, "step": 22520 }, { "grad_norm": 0.09810345619916916, "learning_rate": 6.218128513992917e-05, "loss": 0.0096, "step": 22530 }, { "grad_norm": 0.1633436530828476, "learning_rate": 6.214920948473433e-05, "loss": 0.0108, "step": 22540 }, { "grad_norm": 0.1509043276309967, "learning_rate": 6.211712851506249e-05, "loss": 0.0091, "step": 22550 }, { "grad_norm": 0.2184428572654724, "learning_rate": 6.208504224494696e-05, "loss": 0.0105, "step": 22560 }, { "grad_norm": 0.18132556974887848, "learning_rate": 6.205295068842338e-05, "loss": 0.0099, "step": 22570 }, { "grad_norm": 0.11002297699451447, "learning_rate": 6.202085385952968e-05, "loss": 0.0086, "step": 22580 }, { "grad_norm": 0.15185409784317017, "learning_rate": 6.19887517723061e-05, "loss": 0.0091, "step": 22590 }, { "grad_norm": 0.16703252494335175, "learning_rate": 6.19566444407952e-05, "loss": 0.0102, "step": 22600 }, { "grad_norm": 0.23625801503658295, "learning_rate": 6.192453187904178e-05, "loss": 0.0106, "step": 22610 }, { "grad_norm": 0.1514369696378708, "learning_rate": 6.189241410109303e-05, "loss": 0.0084, "step": 22620 }, { "grad_norm": 0.19141139090061188, "learning_rate": 6.186029112099829e-05, "loss": 0.0099, "step": 22630 }, { "grad_norm": 0.1725449115037918, "learning_rate": 6.182816295280926e-05, "loss": 0.0094, "step": 22640 }, { "grad_norm": 0.17025558650493622, "learning_rate": 6.179602961057993e-05, "loss": 0.012, "step": 22650 }, { "grad_norm": 0.12669697403907776, "learning_rate": 6.176389110836647e-05, "loss": 0.0098, "step": 22660 }, { "grad_norm": 0.20962145924568176, "learning_rate": 6.173174746022738e-05, "loss": 0.016, "step": 22670 }, { "grad_norm": 0.22365134954452515, "learning_rate": 6.169959868022338e-05, "loss": 0.0117, "step": 22680 }, { "grad_norm": 0.19159626960754395, "learning_rate": 6.166744478241742e-05, "loss": 0.0116, "step": 22690 }, { "grad_norm": 0.2035933881998062, "learning_rate": 6.163528578087473e-05, "loss": 0.0107, "step": 22700 }, { "grad_norm": 0.16985948383808136, "learning_rate": 6.160312168966274e-05, "loss": 0.0116, "step": 22710 }, { "grad_norm": 0.13212890923023224, "learning_rate": 6.157095252285112e-05, "loss": 0.0092, "step": 22720 }, { "grad_norm": 0.13431879878044128, "learning_rate": 6.153877829451174e-05, "loss": 0.0092, "step": 22730 }, { "grad_norm": 0.135501891374588, "learning_rate": 6.150659901871872e-05, "loss": 0.0105, "step": 22740 }, { "grad_norm": 0.10139214247465134, "learning_rate": 6.147441470954836e-05, "loss": 0.0097, "step": 22750 }, { "grad_norm": 0.16580139100551605, "learning_rate": 6.144222538107918e-05, "loss": 0.0083, "step": 22760 }, { "grad_norm": 0.16427136957645416, "learning_rate": 6.141003104739188e-05, "loss": 0.0108, "step": 22770 }, { "grad_norm": 0.1716408133506775, "learning_rate": 6.137783172256935e-05, "loss": 0.0097, "step": 22780 }, { "grad_norm": 0.13547708094120026, "learning_rate": 6.134562742069669e-05, "loss": 0.01, "step": 22790 }, { "grad_norm": 0.17049188911914825, "learning_rate": 6.131341815586111e-05, "loss": 0.0099, "step": 22800 }, { "grad_norm": 0.2219429910182953, "learning_rate": 6.128120394215209e-05, "loss": 0.0127, "step": 22810 }, { "grad_norm": 0.1628614068031311, "learning_rate": 6.124898479366116e-05, "loss": 0.008, "step": 22820 }, { "grad_norm": 0.2187962681055069, "learning_rate": 6.121676072448214e-05, "loss": 0.0108, "step": 22830 }, { "grad_norm": 0.15445823967456818, "learning_rate": 6.118453174871089e-05, "loss": 0.009, "step": 22840 }, { "grad_norm": 0.20160919427871704, "learning_rate": 6.115229788044548e-05, "loss": 0.0113, "step": 22850 }, { "grad_norm": 0.20955802500247955, "learning_rate": 6.112005913378607e-05, "loss": 0.0105, "step": 22860 }, { "grad_norm": 0.2079155147075653, "learning_rate": 6.1087815522835e-05, "loss": 0.0105, "step": 22870 }, { "grad_norm": 0.13898858428001404, "learning_rate": 6.105556706169673e-05, "loss": 0.0104, "step": 22880 }, { "grad_norm": 0.1489875167608261, "learning_rate": 6.102331376447782e-05, "loss": 0.0094, "step": 22890 }, { "grad_norm": 0.1784362643957138, "learning_rate": 6.099105564528697e-05, "loss": 0.012, "step": 22900 }, { "grad_norm": 0.12367185205221176, "learning_rate": 6.095879271823495e-05, "loss": 0.0099, "step": 22910 }, { "grad_norm": 0.12994655966758728, "learning_rate": 6.09265249974347e-05, "loss": 0.0105, "step": 22920 }, { "grad_norm": 0.1879769265651703, "learning_rate": 6.089425249700118e-05, "loss": 0.0089, "step": 22930 }, { "grad_norm": 0.1394907385110855, "learning_rate": 6.0861975231051505e-05, "loss": 0.0093, "step": 22940 }, { "grad_norm": 0.1825711876153946, "learning_rate": 6.082969321370484e-05, "loss": 0.0107, "step": 22950 }, { "grad_norm": 0.22647172212600708, "learning_rate": 6.079740645908243e-05, "loss": 0.0103, "step": 22960 }, { "grad_norm": 0.19141080975532532, "learning_rate": 6.07651149813076e-05, "loss": 0.0101, "step": 22970 }, { "grad_norm": 0.10959036648273468, "learning_rate": 6.073281879450574e-05, "loss": 0.0081, "step": 22980 }, { "grad_norm": 0.1814756840467453, "learning_rate": 6.070051791280431e-05, "loss": 0.0075, "step": 22990 }, { "grad_norm": 0.18760697543621063, "learning_rate": 6.066821235033281e-05, "loss": 0.0107, "step": 23000 }, { "grad_norm": 0.15414190292358398, "learning_rate": 6.063590212122277e-05, "loss": 0.0097, "step": 23010 }, { "grad_norm": 0.15072192251682281, "learning_rate": 6.0603587239607794e-05, "loss": 0.0098, "step": 23020 }, { "grad_norm": 0.17472437024116516, "learning_rate": 6.057126771962353e-05, "loss": 0.008, "step": 23030 }, { "grad_norm": 0.1961478739976883, "learning_rate": 6.05389435754076e-05, "loss": 0.0106, "step": 23040 }, { "grad_norm": 0.18475131690502167, "learning_rate": 6.0506614821099705e-05, "loss": 0.0089, "step": 23050 }, { "grad_norm": 0.1881481260061264, "learning_rate": 6.047428147084153e-05, "loss": 0.0112, "step": 23060 }, { "grad_norm": 0.15381397306919098, "learning_rate": 6.044194353877679e-05, "loss": 0.0092, "step": 23070 }, { "grad_norm": 0.23715443909168243, "learning_rate": 6.04096010390512e-05, "loss": 0.0119, "step": 23080 }, { "grad_norm": 0.1962587535381317, "learning_rate": 6.0377253985812445e-05, "loss": 0.0103, "step": 23090 }, { "grad_norm": 0.17410367727279663, "learning_rate": 6.0344902393210235e-05, "loss": 0.0096, "step": 23100 }, { "grad_norm": 0.17661510407924652, "learning_rate": 6.031254627539625e-05, "loss": 0.0093, "step": 23110 }, { "grad_norm": 0.18359529972076416, "learning_rate": 6.028018564652418e-05, "loss": 0.0077, "step": 23120 }, { "grad_norm": 0.11725842207670212, "learning_rate": 6.024782052074964e-05, "loss": 0.0093, "step": 23130 }, { "grad_norm": 0.15117046236991882, "learning_rate": 6.021545091223026e-05, "loss": 0.0081, "step": 23140 }, { "grad_norm": 0.14373986423015594, "learning_rate": 6.0183076835125586e-05, "loss": 0.0102, "step": 23150 }, { "grad_norm": 0.1768474131822586, "learning_rate": 6.015069830359714e-05, "loss": 0.0106, "step": 23160 }, { "grad_norm": 0.1321270614862442, "learning_rate": 6.0118315331808384e-05, "loss": 0.009, "step": 23170 }, { "grad_norm": 0.1517127901315689, "learning_rate": 6.0085927933924745e-05, "loss": 0.0084, "step": 23180 }, { "grad_norm": 0.14846158027648926, "learning_rate": 6.0053536124113585e-05, "loss": 0.0117, "step": 23190 }, { "grad_norm": 0.11555562913417816, "learning_rate": 6.0021139916544144e-05, "loss": 0.0136, "step": 23200 }, { "grad_norm": 0.13129378855228424, "learning_rate": 5.998873932538764e-05, "loss": 0.0088, "step": 23210 }, { "grad_norm": 0.17859961092472076, "learning_rate": 5.995633436481721e-05, "loss": 0.0084, "step": 23220 }, { "grad_norm": 0.2773771584033966, "learning_rate": 5.9923925049007855e-05, "loss": 0.0098, "step": 23230 }, { "grad_norm": 0.19474296271800995, "learning_rate": 5.989151139213656e-05, "loss": 0.0098, "step": 23240 }, { "grad_norm": 0.1791786551475525, "learning_rate": 5.9859093408382094e-05, "loss": 0.0082, "step": 23250 }, { "grad_norm": 0.13379330933094025, "learning_rate": 5.982667111192523e-05, "loss": 0.0083, "step": 23260 }, { "grad_norm": 0.13014809787273407, "learning_rate": 5.979424451694857e-05, "loss": 0.0094, "step": 23270 }, { "grad_norm": 0.20605546236038208, "learning_rate": 5.97618136376366e-05, "loss": 0.0114, "step": 23280 }, { "grad_norm": 0.14216075837612152, "learning_rate": 5.97293784881757e-05, "loss": 0.0094, "step": 23290 }, { "grad_norm": 0.22280192375183105, "learning_rate": 5.969693908275411e-05, "loss": 0.0089, "step": 23300 }, { "grad_norm": 0.1867988258600235, "learning_rate": 5.9664495435561916e-05, "loss": 0.0097, "step": 23310 }, { "grad_norm": 0.1695389747619629, "learning_rate": 5.963204756079108e-05, "loss": 0.0115, "step": 23320 }, { "grad_norm": 0.15757451951503754, "learning_rate": 5.959959547263538e-05, "loss": 0.0089, "step": 23330 }, { "grad_norm": 0.1698025017976761, "learning_rate": 5.95671391852905e-05, "loss": 0.0077, "step": 23340 }, { "grad_norm": 0.16108590364456177, "learning_rate": 5.95346787129539e-05, "loss": 0.0093, "step": 23350 }, { "grad_norm": 0.1416700780391693, "learning_rate": 5.950221406982489e-05, "loss": 0.0103, "step": 23360 }, { "grad_norm": 0.11571353673934937, "learning_rate": 5.9469745270104625e-05, "loss": 0.0101, "step": 23370 }, { "grad_norm": 0.2070477455854416, "learning_rate": 5.9437272327996064e-05, "loss": 0.013, "step": 23380 }, { "grad_norm": 0.18133804202079773, "learning_rate": 5.940479525770395e-05, "loss": 0.0094, "step": 23390 }, { "grad_norm": 0.15582101047039032, "learning_rate": 5.9372314073434876e-05, "loss": 0.0099, "step": 23400 }, { "grad_norm": 0.11277022957801819, "learning_rate": 5.93398287893972e-05, "loss": 0.0101, "step": 23410 }, { "grad_norm": 0.19386158883571625, "learning_rate": 5.930733941980111e-05, "loss": 0.0132, "step": 23420 }, { "grad_norm": 0.17285168170928955, "learning_rate": 5.927484597885855e-05, "loss": 0.0129, "step": 23430 }, { "grad_norm": 0.14666034281253815, "learning_rate": 5.924234848078325e-05, "loss": 0.0097, "step": 23440 }, { "grad_norm": 0.15751126408576965, "learning_rate": 5.9209846939790715e-05, "loss": 0.0107, "step": 23450 }, { "grad_norm": 0.1552300602197647, "learning_rate": 5.917734137009824e-05, "loss": 0.011, "step": 23460 }, { "grad_norm": 0.18218748271465302, "learning_rate": 5.914483178592485e-05, "loss": 0.0107, "step": 23470 }, { "grad_norm": 0.16615097224712372, "learning_rate": 5.911231820149136e-05, "loss": 0.0087, "step": 23480 }, { "grad_norm": 0.18907147645950317, "learning_rate": 5.907980063102031e-05, "loss": 0.0074, "step": 23490 }, { "grad_norm": 0.177125483751297, "learning_rate": 5.904727908873597e-05, "loss": 0.0104, "step": 23500 }, { "grad_norm": 0.17649389803409576, "learning_rate": 5.90147535888644e-05, "loss": 0.0112, "step": 23510 }, { "grad_norm": 0.20461758971214294, "learning_rate": 5.898222414563333e-05, "loss": 0.0114, "step": 23520 }, { "grad_norm": 0.16256879270076752, "learning_rate": 5.8949690773272256e-05, "loss": 0.0113, "step": 23530 }, { "grad_norm": 0.1489347219467163, "learning_rate": 5.891715348601239e-05, "loss": 0.0115, "step": 23540 }, { "grad_norm": 0.16904598474502563, "learning_rate": 5.888461229808663e-05, "loss": 0.0091, "step": 23550 }, { "grad_norm": 0.17050881683826447, "learning_rate": 5.885206722372959e-05, "loss": 0.0083, "step": 23560 }, { "grad_norm": 0.15307481586933136, "learning_rate": 5.881951827717763e-05, "loss": 0.0078, "step": 23570 }, { "grad_norm": 0.1356300413608551, "learning_rate": 5.878696547266872e-05, "loss": 0.0124, "step": 23580 }, { "grad_norm": 0.11302834749221802, "learning_rate": 5.875440882444259e-05, "loss": 0.0092, "step": 23590 }, { "grad_norm": 0.17458340525627136, "learning_rate": 5.8721848346740617e-05, "loss": 0.0101, "step": 23600 }, { "grad_norm": 0.13192391395568848, "learning_rate": 5.868928405380585e-05, "loss": 0.0091, "step": 23610 }, { "grad_norm": 0.11425354331731796, "learning_rate": 5.865671595988304e-05, "loss": 0.0081, "step": 23620 }, { "grad_norm": 0.16480226814746857, "learning_rate": 5.862414407921857e-05, "loss": 0.0088, "step": 23630 }, { "grad_norm": 0.17932955920696259, "learning_rate": 5.8591568426060475e-05, "loss": 0.0094, "step": 23640 }, { "grad_norm": 0.14763322472572327, "learning_rate": 5.8558989014658464e-05, "loss": 0.0078, "step": 23650 }, { "grad_norm": 0.1499917209148407, "learning_rate": 5.8526405859263855e-05, "loss": 0.0084, "step": 23660 }, { "grad_norm": 0.2319207340478897, "learning_rate": 5.8493818974129667e-05, "loss": 0.0096, "step": 23670 }, { "grad_norm": 0.1915825456380844, "learning_rate": 5.846122837351048e-05, "loss": 0.0079, "step": 23680 }, { "grad_norm": 0.24994438886642456, "learning_rate": 5.842863407166255e-05, "loss": 0.009, "step": 23690 }, { "grad_norm": 0.13363829255104065, "learning_rate": 5.83960360828437e-05, "loss": 0.0085, "step": 23700 }, { "grad_norm": 0.1835077852010727, "learning_rate": 5.836343442131343e-05, "loss": 0.0086, "step": 23710 }, { "grad_norm": 0.15411287546157837, "learning_rate": 5.833082910133281e-05, "loss": 0.0105, "step": 23720 }, { "grad_norm": 0.18361972272396088, "learning_rate": 5.82982201371645e-05, "loss": 0.0089, "step": 23730 }, { "grad_norm": 0.14256909489631653, "learning_rate": 5.826560754307279e-05, "loss": 0.0108, "step": 23740 }, { "grad_norm": 0.1655866801738739, "learning_rate": 5.8232991333323516e-05, "loss": 0.0096, "step": 23750 }, { "grad_norm": 0.14625224471092224, "learning_rate": 5.8200371522184116e-05, "loss": 0.01, "step": 23760 }, { "grad_norm": 0.1902720332145691, "learning_rate": 5.816774812392362e-05, "loss": 0.0086, "step": 23770 }, { "grad_norm": 0.18527227640151978, "learning_rate": 5.8135121152812614e-05, "loss": 0.0105, "step": 23780 }, { "grad_norm": 0.21849049627780914, "learning_rate": 5.8102490623123204e-05, "loss": 0.0099, "step": 23790 }, { "grad_norm": 0.16464947164058685, "learning_rate": 5.806985654912914e-05, "loss": 0.0097, "step": 23800 }, { "grad_norm": 0.14453473687171936, "learning_rate": 5.8037218945105644e-05, "loss": 0.0086, "step": 23810 }, { "grad_norm": 0.18746471405029297, "learning_rate": 5.800457782532953e-05, "loss": 0.0111, "step": 23820 }, { "grad_norm": 0.12292444705963135, "learning_rate": 5.797193320407912e-05, "loss": 0.0079, "step": 23830 }, { "grad_norm": 0.10716307163238525, "learning_rate": 5.793928509563431e-05, "loss": 0.0085, "step": 23840 }, { "grad_norm": 0.14872783422470093, "learning_rate": 5.790663351427644e-05, "loss": 0.0096, "step": 23850 }, { "grad_norm": 0.17199794948101044, "learning_rate": 5.7873978474288484e-05, "loss": 0.0083, "step": 23860 }, { "grad_norm": 0.13342317938804626, "learning_rate": 5.78413199899548e-05, "loss": 0.0084, "step": 23870 }, { "grad_norm": 0.1458325982093811, "learning_rate": 5.780865807556136e-05, "loss": 0.0093, "step": 23880 }, { "grad_norm": 0.14551177620887756, "learning_rate": 5.777599274539559e-05, "loss": 0.0107, "step": 23890 }, { "grad_norm": 0.18537873029708862, "learning_rate": 5.77433240137464e-05, "loss": 0.01, "step": 23900 }, { "grad_norm": 0.15026608109474182, "learning_rate": 5.771065189490422e-05, "loss": 0.0094, "step": 23910 }, { "grad_norm": 0.1679491102695465, "learning_rate": 5.767797640316095e-05, "loss": 0.0082, "step": 23920 }, { "grad_norm": 0.1751236766576767, "learning_rate": 5.764529755280994e-05, "loss": 0.0077, "step": 23930 }, { "grad_norm": 0.167789489030838, "learning_rate": 5.761261535814603e-05, "loss": 0.0083, "step": 23940 }, { "grad_norm": 0.19181878864765167, "learning_rate": 5.757992983346554e-05, "loss": 0.0089, "step": 23950 }, { "grad_norm": 0.1933700442314148, "learning_rate": 5.754724099306621e-05, "loss": 0.0104, "step": 23960 }, { "grad_norm": 0.1239490956068039, "learning_rate": 5.751454885124727e-05, "loss": 0.0072, "step": 23970 }, { "grad_norm": 0.14007526636123657, "learning_rate": 5.748185342230935e-05, "loss": 0.0074, "step": 23980 }, { "grad_norm": 0.14329925179481506, "learning_rate": 5.744915472055456e-05, "loss": 0.009, "step": 23990 }, { "grad_norm": 0.1467640995979309, "learning_rate": 5.741645276028642e-05, "loss": 0.0093, "step": 24000 }, { "grad_norm": 0.14676660299301147, "learning_rate": 5.7383747555809873e-05, "loss": 0.0089, "step": 24010 }, { "grad_norm": 0.1495020091533661, "learning_rate": 5.7351039121431294e-05, "loss": 0.0104, "step": 24020 }, { "grad_norm": 0.13182789087295532, "learning_rate": 5.7318327471458454e-05, "loss": 0.0085, "step": 24030 }, { "grad_norm": 0.20185944437980652, "learning_rate": 5.728561262020054e-05, "loss": 0.0107, "step": 24040 }, { "grad_norm": 0.10718248039484024, "learning_rate": 5.725289458196814e-05, "loss": 0.0103, "step": 24050 }, { "grad_norm": 0.14702700078487396, "learning_rate": 5.722017337107324e-05, "loss": 0.0115, "step": 24060 }, { "grad_norm": 0.16052216291427612, "learning_rate": 5.7187449001829206e-05, "loss": 0.0098, "step": 24070 }, { "grad_norm": 0.15060575306415558, "learning_rate": 5.715472148855081e-05, "loss": 0.0084, "step": 24080 }, { "grad_norm": 0.18276582658290863, "learning_rate": 5.712199084555414e-05, "loss": 0.0114, "step": 24090 }, { "grad_norm": 0.1456592082977295, "learning_rate": 5.7089257087156734e-05, "loss": 0.0089, "step": 24100 }, { "grad_norm": 0.12430857867002487, "learning_rate": 5.70565202276774e-05, "loss": 0.0085, "step": 24110 }, { "grad_norm": 0.15847177803516388, "learning_rate": 5.70237802814364e-05, "loss": 0.0101, "step": 24120 }, { "grad_norm": 0.13500328361988068, "learning_rate": 5.6991037262755277e-05, "loss": 0.0085, "step": 24130 }, { "grad_norm": 0.13666090369224548, "learning_rate": 5.6958291185956924e-05, "loss": 0.0097, "step": 24140 }, { "grad_norm": 0.1551935076713562, "learning_rate": 5.692554206536561e-05, "loss": 0.0088, "step": 24150 }, { "grad_norm": 0.13200466334819794, "learning_rate": 5.689278991530692e-05, "loss": 0.01, "step": 24160 }, { "grad_norm": 0.1769729107618332, "learning_rate": 5.686003475010773e-05, "loss": 0.0077, "step": 24170 }, { "grad_norm": 0.1648060381412506, "learning_rate": 5.682727658409628e-05, "loss": 0.0121, "step": 24180 }, { "grad_norm": 0.16070470213890076, "learning_rate": 5.67945154316021e-05, "loss": 0.0116, "step": 24190 }, { "grad_norm": 0.18123023211956024, "learning_rate": 5.676175130695602e-05, "loss": 0.0091, "step": 24200 }, { "grad_norm": 0.2394556701183319, "learning_rate": 5.672898422449022e-05, "loss": 0.0087, "step": 24210 }, { "grad_norm": 0.18369920551776886, "learning_rate": 5.66962141985381e-05, "loss": 0.0124, "step": 24220 }, { "grad_norm": 0.15011410415172577, "learning_rate": 5.666344124343439e-05, "loss": 0.01, "step": 24230 }, { "grad_norm": 0.14626722037792206, "learning_rate": 5.66306653735151e-05, "loss": 0.0092, "step": 24240 }, { "grad_norm": 0.18686443567276, "learning_rate": 5.659788660311749e-05, "loss": 0.0096, "step": 24250 }, { "grad_norm": 0.18956445157527924, "learning_rate": 5.656510494658014e-05, "loss": 0.0088, "step": 24260 }, { "grad_norm": 0.16100074350833893, "learning_rate": 5.6532320418242835e-05, "loss": 0.0086, "step": 24270 }, { "grad_norm": 0.13245223462581635, "learning_rate": 5.649953303244665e-05, "loss": 0.0091, "step": 24280 }, { "grad_norm": 0.18928034603595734, "learning_rate": 5.646674280353389e-05, "loss": 0.0129, "step": 24290 }, { "grad_norm": 0.13012899458408356, "learning_rate": 5.6433949745848135e-05, "loss": 0.0107, "step": 24300 }, { "grad_norm": 0.12411461770534515, "learning_rate": 5.6401153873734166e-05, "loss": 0.0109, "step": 24310 }, { "grad_norm": 0.10325054824352264, "learning_rate": 5.6368355201538036e-05, "loss": 0.009, "step": 24320 }, { "grad_norm": 0.122968889772892, "learning_rate": 5.633555374360697e-05, "loss": 0.0084, "step": 24330 }, { "grad_norm": 0.1636364758014679, "learning_rate": 5.630274951428944e-05, "loss": 0.0096, "step": 24340 }, { "grad_norm": 0.19007551670074463, "learning_rate": 5.626994252793517e-05, "loss": 0.0087, "step": 24350 }, { "grad_norm": 0.21164660155773163, "learning_rate": 5.6237132798895006e-05, "loss": 0.0108, "step": 24360 }, { "grad_norm": 0.14414788782596588, "learning_rate": 5.620432034152106e-05, "loss": 0.0102, "step": 24370 }, { "grad_norm": 0.1905253380537033, "learning_rate": 5.6171505170166604e-05, "loss": 0.0094, "step": 24380 }, { "grad_norm": 0.17770907282829285, "learning_rate": 5.613868729918612e-05, "loss": 0.0096, "step": 24390 }, { "grad_norm": 0.09498396515846252, "learning_rate": 5.6105866742935256e-05, "loss": 0.0087, "step": 24400 }, { "grad_norm": 0.22553133964538574, "learning_rate": 5.607304351577083e-05, "loss": 0.0102, "step": 24410 }, { "grad_norm": 0.12401127815246582, "learning_rate": 5.604021763205085e-05, "loss": 0.0083, "step": 24420 }, { "grad_norm": 0.22198164463043213, "learning_rate": 5.600738910613445e-05, "loss": 0.0102, "step": 24430 }, { "grad_norm": 0.18472342193126678, "learning_rate": 5.597455795238198e-05, "loss": 0.011, "step": 24440 }, { "grad_norm": 0.16024212539196014, "learning_rate": 5.5941724185154876e-05, "loss": 0.0095, "step": 24450 }, { "grad_norm": 0.13403628766536713, "learning_rate": 5.5908887818815746e-05, "loss": 0.0079, "step": 24460 }, { "grad_norm": 0.1791236847639084, "learning_rate": 5.587604886772834e-05, "loss": 0.0079, "step": 24470 }, { "grad_norm": 0.1322169154882431, "learning_rate": 5.584320734625752e-05, "loss": 0.0089, "step": 24480 }, { "grad_norm": 0.2079758495092392, "learning_rate": 5.581036326876928e-05, "loss": 0.0111, "step": 24490 }, { "grad_norm": 0.19254417717456818, "learning_rate": 5.5777516649630756e-05, "loss": 0.0126, "step": 24500 }, { "grad_norm": 0.13894963264465332, "learning_rate": 5.574466750321017e-05, "loss": 0.0089, "step": 24510 }, { "grad_norm": 0.1603253334760666, "learning_rate": 5.571181584387683e-05, "loss": 0.0102, "step": 24520 }, { "grad_norm": 0.1136169582605362, "learning_rate": 5.5678961686001196e-05, "loss": 0.0098, "step": 24530 }, { "grad_norm": 0.16871067881584167, "learning_rate": 5.5646105043954785e-05, "loss": 0.0127, "step": 24540 }, { "grad_norm": 0.11935514211654663, "learning_rate": 5.5613245932110205e-05, "loss": 0.0094, "step": 24550 }, { "grad_norm": 0.20700982213020325, "learning_rate": 5.5580384364841157e-05, "loss": 0.01, "step": 24560 }, { "grad_norm": 0.1157289668917656, "learning_rate": 5.554752035652237e-05, "loss": 0.0112, "step": 24570 }, { "grad_norm": 0.1842637062072754, "learning_rate": 5.5514653921529736e-05, "loss": 0.0084, "step": 24580 }, { "grad_norm": 0.13435623049736023, "learning_rate": 5.5481785074240114e-05, "loss": 0.0085, "step": 24590 }, { "grad_norm": 0.16202187538146973, "learning_rate": 5.544891382903146e-05, "loss": 0.0114, "step": 24600 }, { "grad_norm": 0.18020673096179962, "learning_rate": 5.541604020028277e-05, "loss": 0.0099, "step": 24610 }, { "grad_norm": 0.19251811504364014, "learning_rate": 5.5383164202374105e-05, "loss": 0.0111, "step": 24620 }, { "grad_norm": 0.1787234991788864, "learning_rate": 5.5350285849686515e-05, "loss": 0.0096, "step": 24630 }, { "grad_norm": 0.2504052221775055, "learning_rate": 5.531740515660213e-05, "loss": 0.0081, "step": 24640 }, { "grad_norm": 0.10568095743656158, "learning_rate": 5.528452213750408e-05, "loss": 0.0087, "step": 24650 }, { "grad_norm": 0.1681310534477234, "learning_rate": 5.52516368067765e-05, "loss": 0.0093, "step": 24660 }, { "grad_norm": 0.14793036878108978, "learning_rate": 5.5218749178804563e-05, "loss": 0.0102, "step": 24670 }, { "grad_norm": 0.15492045879364014, "learning_rate": 5.518585926797443e-05, "loss": 0.0103, "step": 24680 }, { "grad_norm": 0.17658020555973053, "learning_rate": 5.5152967088673265e-05, "loss": 0.0109, "step": 24690 }, { "grad_norm": 0.13920898735523224, "learning_rate": 5.512007265528924e-05, "loss": 0.0096, "step": 24700 }, { "grad_norm": 0.17936009168624878, "learning_rate": 5.508717598221148e-05, "loss": 0.0078, "step": 24710 }, { "grad_norm": 0.1703759729862213, "learning_rate": 5.505427708383012e-05, "loss": 0.0119, "step": 24720 }, { "grad_norm": 0.18596866726875305, "learning_rate": 5.502137597453624e-05, "loss": 0.0086, "step": 24730 }, { "grad_norm": 0.18658918142318726, "learning_rate": 5.4988472668721914e-05, "loss": 0.0096, "step": 24740 }, { "grad_norm": 0.11790794134140015, "learning_rate": 5.4955567180780164e-05, "loss": 0.0092, "step": 24750 }, { "grad_norm": 0.1601596474647522, "learning_rate": 5.492265952510497e-05, "loss": 0.0122, "step": 24760 }, { "grad_norm": 0.15226708352565765, "learning_rate": 5.4889749716091254e-05, "loss": 0.0085, "step": 24770 }, { "grad_norm": 0.14226096868515015, "learning_rate": 5.4856837768134885e-05, "loss": 0.0091, "step": 24780 }, { "grad_norm": 0.19087372720241547, "learning_rate": 5.482392369563267e-05, "loss": 0.0103, "step": 24790 }, { "grad_norm": 0.19576215744018555, "learning_rate": 5.4791007512982354e-05, "loss": 0.0087, "step": 24800 }, { "grad_norm": 0.21374139189720154, "learning_rate": 5.4758089234582566e-05, "loss": 0.0108, "step": 24810 }, { "grad_norm": 0.17007111012935638, "learning_rate": 5.47251688748329e-05, "loss": 0.0085, "step": 24820 }, { "grad_norm": 0.18705835938453674, "learning_rate": 5.4692246448133843e-05, "loss": 0.0089, "step": 24830 }, { "grad_norm": 0.1699766218662262, "learning_rate": 5.465932196888676e-05, "loss": 0.0083, "step": 24840 }, { "grad_norm": 0.18482382595539093, "learning_rate": 5.462639545149395e-05, "loss": 0.0098, "step": 24850 }, { "grad_norm": 0.13515114784240723, "learning_rate": 5.459346691035861e-05, "loss": 0.0072, "step": 24860 }, { "grad_norm": 0.1896362155675888, "learning_rate": 5.4560536359884776e-05, "loss": 0.0078, "step": 24870 }, { "grad_norm": 0.15254785120487213, "learning_rate": 5.45276038144774e-05, "loss": 0.0083, "step": 24880 }, { "grad_norm": 0.17463655769824982, "learning_rate": 5.449466928854229e-05, "loss": 0.0064, "step": 24890 }, { "grad_norm": 0.14858397841453552, "learning_rate": 5.446173279648613e-05, "loss": 0.0084, "step": 24900 }, { "grad_norm": 0.1323464959859848, "learning_rate": 5.442879435271646e-05, "loss": 0.008, "step": 24910 }, { "grad_norm": 0.16471095383167267, "learning_rate": 5.4395853971641675e-05, "loss": 0.0112, "step": 24920 }, { "grad_norm": 0.24587994813919067, "learning_rate": 5.4362911667671015e-05, "loss": 0.0093, "step": 24930 }, { "grad_norm": 0.19343839585781097, "learning_rate": 5.432996745521458e-05, "loss": 0.0089, "step": 24940 }, { "grad_norm": 0.11594853550195694, "learning_rate": 5.4297021348683264e-05, "loss": 0.0083, "step": 24950 }, { "grad_norm": 0.1557769477367401, "learning_rate": 5.426407336248882e-05, "loss": 0.0094, "step": 24960 }, { "grad_norm": 0.1667032092809677, "learning_rate": 5.423112351104382e-05, "loss": 0.0076, "step": 24970 }, { "grad_norm": 0.20383715629577637, "learning_rate": 5.4198171808761654e-05, "loss": 0.0111, "step": 24980 }, { "grad_norm": 0.16118177771568298, "learning_rate": 5.41652182700565e-05, "loss": 0.011, "step": 24990 }, { "grad_norm": 0.14757801592350006, "learning_rate": 5.413226290934338e-05, "loss": 0.0096, "step": 25000 }, { "grad_norm": 0.1326269954442978, "learning_rate": 5.4099305741038056e-05, "loss": 0.0084, "step": 25010 }, { "grad_norm": 0.1567969024181366, "learning_rate": 5.406634677955713e-05, "loss": 0.0085, "step": 25020 }, { "grad_norm": 0.18868760764598846, "learning_rate": 5.403338603931798e-05, "loss": 0.0094, "step": 25030 }, { "grad_norm": 0.14019161462783813, "learning_rate": 5.4000423534738745e-05, "loss": 0.0078, "step": 25040 }, { "grad_norm": 0.1392132043838501, "learning_rate": 5.396745928023835e-05, "loss": 0.0102, "step": 25050 }, { "grad_norm": 0.17088185250759125, "learning_rate": 5.3934493290236474e-05, "loss": 0.0103, "step": 25060 }, { "grad_norm": 0.18075264990329742, "learning_rate": 5.390152557915357e-05, "loss": 0.0115, "step": 25070 }, { "grad_norm": 0.19884738326072693, "learning_rate": 5.3868556161410796e-05, "loss": 0.0093, "step": 25080 }, { "grad_norm": 0.16397683322429657, "learning_rate": 5.383558505143015e-05, "loss": 0.0081, "step": 25090 }, { "grad_norm": 0.11553274840116501, "learning_rate": 5.38026122636343e-05, "loss": 0.0095, "step": 25100 }, { "grad_norm": 0.13938915729522705, "learning_rate": 5.3769637812446636e-05, "loss": 0.0086, "step": 25110 }, { "grad_norm": 0.13693255186080933, "learning_rate": 5.373666171229133e-05, "loss": 0.0098, "step": 25120 }, { "grad_norm": 0.1360224038362503, "learning_rate": 5.370368397759324e-05, "loss": 0.0086, "step": 25130 }, { "grad_norm": 0.18198657035827637, "learning_rate": 5.3670704622777944e-05, "loss": 0.0096, "step": 25140 }, { "grad_norm": 0.16317658126354218, "learning_rate": 5.3637723662271744e-05, "loss": 0.0086, "step": 25150 }, { "grad_norm": 0.17122642695903778, "learning_rate": 5.360474111050162e-05, "loss": 0.0079, "step": 25160 }, { "grad_norm": 0.16225238144397736, "learning_rate": 5.357175698189527e-05, "loss": 0.0084, "step": 25170 }, { "grad_norm": 0.13298451900482178, "learning_rate": 5.353877129088107e-05, "loss": 0.0089, "step": 25180 }, { "grad_norm": 0.1363864541053772, "learning_rate": 5.350578405188809e-05, "loss": 0.0086, "step": 25190 }, { "grad_norm": 0.1381729394197464, "learning_rate": 5.347279527934603e-05, "loss": 0.0062, "step": 25200 }, { "grad_norm": 0.1830565184354782, "learning_rate": 5.343980498768536e-05, "loss": 0.0071, "step": 25210 }, { "grad_norm": 0.21361270546913147, "learning_rate": 5.340681319133711e-05, "loss": 0.0118, "step": 25220 }, { "grad_norm": 0.17887625098228455, "learning_rate": 5.337381990473305e-05, "loss": 0.0103, "step": 25230 }, { "grad_norm": 0.1882539987564087, "learning_rate": 5.3340825142305504e-05, "loss": 0.009, "step": 25240 }, { "grad_norm": 0.1689753532409668, "learning_rate": 5.330782891848756e-05, "loss": 0.0077, "step": 25250 }, { "grad_norm": 0.1130693256855011, "learning_rate": 5.327483124771285e-05, "loss": 0.0084, "step": 25260 }, { "grad_norm": 0.11064855754375458, "learning_rate": 5.3241832144415706e-05, "loss": 0.0097, "step": 25270 }, { "grad_norm": 0.18950292468070984, "learning_rate": 5.320883162303104e-05, "loss": 0.0085, "step": 25280 }, { "grad_norm": 0.18336588144302368, "learning_rate": 5.3175829697994416e-05, "loss": 0.0108, "step": 25290 }, { "grad_norm": 0.16212618350982666, "learning_rate": 5.3142826383741975e-05, "loss": 0.0096, "step": 25300 }, { "grad_norm": 0.15392407774925232, "learning_rate": 5.3109821694710506e-05, "loss": 0.01, "step": 25310 }, { "grad_norm": 0.14503835141658783, "learning_rate": 5.307681564533736e-05, "loss": 0.0102, "step": 25320 }, { "grad_norm": 0.11327841132879257, "learning_rate": 5.3043808250060535e-05, "loss": 0.0079, "step": 25330 }, { "grad_norm": 0.13404740393161774, "learning_rate": 5.3010799523318566e-05, "loss": 0.0081, "step": 25340 }, { "grad_norm": 0.126858651638031, "learning_rate": 5.29777894795506e-05, "loss": 0.0074, "step": 25350 }, { "grad_norm": 0.16100452840328217, "learning_rate": 5.294477813319634e-05, "loss": 0.0079, "step": 25360 }, { "grad_norm": 0.19441978633403778, "learning_rate": 5.291176549869608e-05, "loss": 0.0096, "step": 25370 }, { "grad_norm": 0.12515397369861603, "learning_rate": 5.287875159049065e-05, "loss": 0.0098, "step": 25380 }, { "grad_norm": 0.1254092901945114, "learning_rate": 5.284573642302149e-05, "loss": 0.0096, "step": 25390 }, { "grad_norm": 0.13799872994422913, "learning_rate": 5.2812720010730543e-05, "loss": 0.0101, "step": 25400 }, { "grad_norm": 0.12374107539653778, "learning_rate": 5.277970236806029e-05, "loss": 0.0121, "step": 25410 }, { "grad_norm": 0.13182932138442993, "learning_rate": 5.2746683509453795e-05, "loss": 0.0107, "step": 25420 }, { "grad_norm": 0.31519389152526855, "learning_rate": 5.271366344935461e-05, "loss": 0.0086, "step": 25430 }, { "grad_norm": 0.129698246717453, "learning_rate": 5.268064220220683e-05, "loss": 0.0111, "step": 25440 }, { "grad_norm": 0.11940047889947891, "learning_rate": 5.2647619782455094e-05, "loss": 0.0075, "step": 25450 }, { "grad_norm": 0.12632475793361664, "learning_rate": 5.261459620454451e-05, "loss": 0.0099, "step": 25460 }, { "grad_norm": 0.21382153034210205, "learning_rate": 5.258157148292071e-05, "loss": 0.0134, "step": 25470 }, { "grad_norm": 0.187351256608963, "learning_rate": 5.2548545632029846e-05, "loss": 0.0122, "step": 25480 }, { "grad_norm": 0.13233990967273712, "learning_rate": 5.2515518666318534e-05, "loss": 0.0106, "step": 25490 }, { "grad_norm": 0.18836741149425507, "learning_rate": 5.248249060023389e-05, "loss": 0.009, "step": 25500 }, { "grad_norm": 0.09369895607233047, "learning_rate": 5.2449461448223517e-05, "loss": 0.0089, "step": 25510 }, { "grad_norm": 0.184903085231781, "learning_rate": 5.2416431224735486e-05, "loss": 0.0112, "step": 25520 }, { "grad_norm": 0.1144842654466629, "learning_rate": 5.238339994421836e-05, "loss": 0.0102, "step": 25530 }, { "grad_norm": 0.14158429205417633, "learning_rate": 5.235036762112111e-05, "loss": 0.0098, "step": 25540 }, { "grad_norm": 0.14127998054027557, "learning_rate": 5.2317334269893206e-05, "loss": 0.0072, "step": 25550 }, { "grad_norm": 0.13598570227622986, "learning_rate": 5.228429990498458e-05, "loss": 0.0076, "step": 25560 }, { "grad_norm": 0.11485196650028229, "learning_rate": 5.225126454084556e-05, "loss": 0.0088, "step": 25570 }, { "grad_norm": 0.15121911466121674, "learning_rate": 5.221822819192694e-05, "loss": 0.0106, "step": 25580 }, { "grad_norm": 0.17684027552604675, "learning_rate": 5.218519087267995e-05, "loss": 0.011, "step": 25590 }, { "grad_norm": 0.1576521247625351, "learning_rate": 5.2152152597556225e-05, "loss": 0.0107, "step": 25600 }, { "grad_norm": 0.1453453004360199, "learning_rate": 5.211911338100784e-05, "loss": 0.008, "step": 25610 }, { "grad_norm": 0.156765878200531, "learning_rate": 5.2086073237487264e-05, "loss": 0.0086, "step": 25620 }, { "grad_norm": 0.2190188318490982, "learning_rate": 5.205303218144739e-05, "loss": 0.0097, "step": 25630 }, { "grad_norm": 0.1852329820394516, "learning_rate": 5.2019990227341494e-05, "loss": 0.0081, "step": 25640 }, { "grad_norm": 0.17052233219146729, "learning_rate": 5.198694738962324e-05, "loss": 0.009, "step": 25650 }, { "grad_norm": 0.1623186618089676, "learning_rate": 5.19539036827467e-05, "loss": 0.0086, "step": 25660 }, { "grad_norm": 0.1390838921070099, "learning_rate": 5.192085912116629e-05, "loss": 0.0123, "step": 25670 }, { "grad_norm": 0.13097183406352997, "learning_rate": 5.188781371933685e-05, "loss": 0.0095, "step": 25680 }, { "grad_norm": 0.10683685541152954, "learning_rate": 5.185476749171356e-05, "loss": 0.0067, "step": 25690 }, { "grad_norm": 0.12746262550354004, "learning_rate": 5.1821720452751945e-05, "loss": 0.0102, "step": 25700 }, { "grad_norm": 0.11987859755754471, "learning_rate": 5.178867261690791e-05, "loss": 0.0103, "step": 25710 }, { "grad_norm": 0.14263421297073364, "learning_rate": 5.1755623998637715e-05, "loss": 0.0071, "step": 25720 }, { "grad_norm": 0.13106735050678253, "learning_rate": 5.172257461239794e-05, "loss": 0.0096, "step": 25730 }, { "grad_norm": 0.17796096205711365, "learning_rate": 5.168952447264548e-05, "loss": 0.0111, "step": 25740 }, { "grad_norm": 0.19277963042259216, "learning_rate": 5.165647359383763e-05, "loss": 0.0098, "step": 25750 }, { "grad_norm": 0.26839950680732727, "learning_rate": 5.1623421990431954e-05, "loss": 0.0094, "step": 25760 }, { "grad_norm": 0.14351381361484528, "learning_rate": 5.159036967688635e-05, "loss": 0.0083, "step": 25770 }, { "grad_norm": 0.18373283743858337, "learning_rate": 5.1557316667659015e-05, "loss": 0.0094, "step": 25780 }, { "grad_norm": 0.1699041724205017, "learning_rate": 5.152426297720846e-05, "loss": 0.0092, "step": 25790 }, { "grad_norm": 0.10603588074445724, "learning_rate": 5.1491208619993505e-05, "loss": 0.007, "step": 25800 }, { "grad_norm": 0.12289376556873322, "learning_rate": 5.145815361047321e-05, "loss": 0.0087, "step": 25810 }, { "grad_norm": 0.1568579375743866, "learning_rate": 5.1425097963107003e-05, "loss": 0.0095, "step": 25820 }, { "grad_norm": 0.10259394347667694, "learning_rate": 5.1392041692354544e-05, "loss": 0.0103, "step": 25830 }, { "grad_norm": 0.13704387843608856, "learning_rate": 5.1358984812675736e-05, "loss": 0.0087, "step": 25840 }, { "grad_norm": 0.10265596956014633, "learning_rate": 5.13259273385308e-05, "loss": 0.009, "step": 25850 }, { "grad_norm": 0.2048063725233078, "learning_rate": 5.12928692843802e-05, "loss": 0.0118, "step": 25860 }, { "grad_norm": 0.11683448404073715, "learning_rate": 5.125981066468465e-05, "loss": 0.0074, "step": 25870 }, { "grad_norm": 0.15408557653427124, "learning_rate": 5.122675149390514e-05, "loss": 0.0119, "step": 25880 }, { "grad_norm": 0.20561572909355164, "learning_rate": 5.119369178650282e-05, "loss": 0.0088, "step": 25890 }, { "grad_norm": 0.16854849457740784, "learning_rate": 5.1160631556939156e-05, "loss": 0.0091, "step": 25900 }, { "grad_norm": 0.2281162291765213, "learning_rate": 5.112757081967584e-05, "loss": 0.0108, "step": 25910 }, { "grad_norm": 0.20315952599048615, "learning_rate": 5.109450958917472e-05, "loss": 0.0105, "step": 25920 }, { "grad_norm": 0.12390350550413132, "learning_rate": 5.106144787989794e-05, "loss": 0.0074, "step": 25930 }, { "grad_norm": 0.11572965979576111, "learning_rate": 5.102838570630778e-05, "loss": 0.0071, "step": 25940 }, { "grad_norm": 0.17498810589313507, "learning_rate": 5.099532308286678e-05, "loss": 0.0085, "step": 25950 }, { "grad_norm": 0.19194817543029785, "learning_rate": 5.0962260024037645e-05, "loss": 0.0077, "step": 25960 }, { "grad_norm": 0.1395537555217743, "learning_rate": 5.0929196544283276e-05, "loss": 0.0068, "step": 25970 }, { "grad_norm": 0.15814076364040375, "learning_rate": 5.089613265806678e-05, "loss": 0.0078, "step": 25980 }, { "grad_norm": 0.18196912109851837, "learning_rate": 5.086306837985141e-05, "loss": 0.0073, "step": 25990 }, { "grad_norm": 0.17208664119243622, "learning_rate": 5.0830003724100604e-05, "loss": 0.0069, "step": 26000 }, { "grad_norm": 0.18526044487953186, "learning_rate": 5.0796938705277984e-05, "loss": 0.0122, "step": 26010 }, { "grad_norm": 0.1513615995645523, "learning_rate": 5.076387333784729e-05, "loss": 0.0077, "step": 26020 }, { "grad_norm": 0.1420505791902542, "learning_rate": 5.0730807636272445e-05, "loss": 0.0105, "step": 26030 }, { "grad_norm": 0.18183238804340363, "learning_rate": 5.0697741615017524e-05, "loss": 0.0077, "step": 26040 }, { "grad_norm": 0.12029215693473816, "learning_rate": 5.06646752885467e-05, "loss": 0.0072, "step": 26050 }, { "grad_norm": 0.14844438433647156, "learning_rate": 5.0631608671324324e-05, "loss": 0.0081, "step": 26060 }, { "grad_norm": 0.09931409358978271, "learning_rate": 5.059854177781487e-05, "loss": 0.0089, "step": 26070 }, { "grad_norm": 0.09935979545116425, "learning_rate": 5.05654746224829e-05, "loss": 0.0088, "step": 26080 }, { "grad_norm": 0.1788986176252365, "learning_rate": 5.053240721979311e-05, "loss": 0.0103, "step": 26090 }, { "grad_norm": 0.14829978346824646, "learning_rate": 5.049933958421031e-05, "loss": 0.0088, "step": 26100 }, { "grad_norm": 0.11087295413017273, "learning_rate": 5.046627173019943e-05, "loss": 0.008, "step": 26110 }, { "grad_norm": 0.12214794009923935, "learning_rate": 5.043320367222545e-05, "loss": 0.0066, "step": 26120 }, { "grad_norm": 0.2034122347831726, "learning_rate": 5.040013542475346e-05, "loss": 0.008, "step": 26130 }, { "grad_norm": 0.21374477446079254, "learning_rate": 5.0367067002248646e-05, "loss": 0.0082, "step": 26140 }, { "grad_norm": 0.11249958723783493, "learning_rate": 5.0333998419176254e-05, "loss": 0.0102, "step": 26150 }, { "grad_norm": 0.15939494967460632, "learning_rate": 5.030092969000161e-05, "loss": 0.0077, "step": 26160 }, { "grad_norm": 0.22043563425540924, "learning_rate": 5.026786082919008e-05, "loss": 0.0105, "step": 26170 }, { "grad_norm": 0.13516555726528168, "learning_rate": 5.0234791851207144e-05, "loss": 0.0067, "step": 26180 }, { "grad_norm": 0.17433863878250122, "learning_rate": 5.0201722770518256e-05, "loss": 0.0078, "step": 26190 }, { "grad_norm": 0.14495232701301575, "learning_rate": 5.0168653601589e-05, "loss": 0.0095, "step": 26200 }, { "grad_norm": 0.16050417721271515, "learning_rate": 5.01355843588849e-05, "loss": 0.0115, "step": 26210 }, { "grad_norm": 0.10825788974761963, "learning_rate": 5.010251505687161e-05, "loss": 0.0079, "step": 26220 }, { "grad_norm": 0.19126494228839874, "learning_rate": 5.006944571001474e-05, "loss": 0.0114, "step": 26230 }, { "grad_norm": 0.1599346399307251, "learning_rate": 5.003637633277997e-05, "loss": 0.0078, "step": 26240 }, { "grad_norm": 0.12217220664024353, "learning_rate": 5.0003306939632946e-05, "loss": 0.0072, "step": 26250 }, { "grad_norm": 0.12278939038515091, "learning_rate": 4.997023754503937e-05, "loss": 0.0087, "step": 26260 }, { "grad_norm": 0.13307350873947144, "learning_rate": 4.9937168163464897e-05, "loss": 0.0077, "step": 26270 }, { "grad_norm": 0.11153905838727951, "learning_rate": 4.990409880937519e-05, "loss": 0.006, "step": 26280 }, { "grad_norm": 0.10756944119930267, "learning_rate": 4.9871029497235936e-05, "loss": 0.0072, "step": 26290 }, { "grad_norm": 0.12187396734952927, "learning_rate": 4.9837960241512766e-05, "loss": 0.0091, "step": 26300 }, { "grad_norm": 0.17239302396774292, "learning_rate": 4.9804891056671265e-05, "loss": 0.0071, "step": 26310 }, { "grad_norm": 0.1274128556251526, "learning_rate": 4.9771821957177074e-05, "loss": 0.0087, "step": 26320 }, { "grad_norm": 0.1556788682937622, "learning_rate": 4.9738752957495694e-05, "loss": 0.0097, "step": 26330 }, { "grad_norm": 0.14743395149707794, "learning_rate": 4.970568407209265e-05, "loss": 0.0095, "step": 26340 }, { "grad_norm": 0.16305646300315857, "learning_rate": 4.967261531543339e-05, "loss": 0.0119, "step": 26350 }, { "grad_norm": 0.1562110185623169, "learning_rate": 4.963954670198332e-05, "loss": 0.0083, "step": 26360 }, { "grad_norm": 0.12147056311368942, "learning_rate": 4.960647824620776e-05, "loss": 0.0073, "step": 26370 }, { "grad_norm": 0.16774678230285645, "learning_rate": 4.9573409962571995e-05, "loss": 0.0096, "step": 26380 }, { "grad_norm": 0.14533740282058716, "learning_rate": 4.954034186554121e-05, "loss": 0.0098, "step": 26390 }, { "grad_norm": 0.13561925292015076, "learning_rate": 4.950727396958051e-05, "loss": 0.0078, "step": 26400 }, { "grad_norm": 0.20065830647945404, "learning_rate": 4.947420628915489e-05, "loss": 0.0093, "step": 26410 }, { "grad_norm": 0.10317568480968475, "learning_rate": 4.944113883872932e-05, "loss": 0.009, "step": 26420 }, { "grad_norm": 0.10967319458723068, "learning_rate": 4.9408071632768624e-05, "loss": 0.0087, "step": 26430 }, { "grad_norm": 0.1985549032688141, "learning_rate": 4.937500468573747e-05, "loss": 0.017, "step": 26440 }, { "grad_norm": 0.10036461800336838, "learning_rate": 4.934193801210053e-05, "loss": 0.0118, "step": 26450 }, { "grad_norm": 0.13834361732006073, "learning_rate": 4.930887162632224e-05, "loss": 0.0106, "step": 26460 }, { "grad_norm": 0.16471286118030548, "learning_rate": 4.9275805542866994e-05, "loss": 0.01, "step": 26470 }, { "grad_norm": 0.17868325114250183, "learning_rate": 4.9242739776199006e-05, "loss": 0.0104, "step": 26480 }, { "grad_norm": 0.1927725374698639, "learning_rate": 4.920967434078237e-05, "loss": 0.0107, "step": 26490 }, { "grad_norm": 0.12332227826118469, "learning_rate": 4.917660925108101e-05, "loss": 0.0083, "step": 26500 }, { "grad_norm": 0.10641764104366302, "learning_rate": 4.914354452155876e-05, "loss": 0.0093, "step": 26510 }, { "grad_norm": 0.17946335673332214, "learning_rate": 4.911048016667923e-05, "loss": 0.0096, "step": 26520 }, { "grad_norm": 0.1438552290201187, "learning_rate": 4.907741620090588e-05, "loss": 0.0099, "step": 26530 }, { "grad_norm": 0.18658311665058136, "learning_rate": 4.9044352638702046e-05, "loss": 0.0123, "step": 26540 }, { "grad_norm": 0.1537322700023651, "learning_rate": 4.9011289494530834e-05, "loss": 0.0094, "step": 26550 }, { "grad_norm": 0.2963276505470276, "learning_rate": 4.897822678285516e-05, "loss": 0.01, "step": 26560 }, { "grad_norm": 0.11203602701425552, "learning_rate": 4.8945164518137835e-05, "loss": 0.0106, "step": 26570 }, { "grad_norm": 0.14809967577457428, "learning_rate": 4.891210271484138e-05, "loss": 0.0107, "step": 26580 }, { "grad_norm": 0.11856665462255478, "learning_rate": 4.8879041387428164e-05, "loss": 0.0073, "step": 26590 }, { "grad_norm": 0.10313080251216888, "learning_rate": 4.8845980550360306e-05, "loss": 0.008, "step": 26600 }, { "grad_norm": 0.11114813387393951, "learning_rate": 4.8812920218099786e-05, "loss": 0.0106, "step": 26610 }, { "grad_norm": 0.10836158692836761, "learning_rate": 4.8779860405108294e-05, "loss": 0.01, "step": 26620 }, { "grad_norm": 0.26990795135498047, "learning_rate": 4.8746801125847285e-05, "loss": 0.0081, "step": 26630 }, { "grad_norm": 0.12698408961296082, "learning_rate": 4.8713742394778064e-05, "loss": 0.0079, "step": 26640 }, { "grad_norm": 0.10755690187215805, "learning_rate": 4.868068422636162e-05, "loss": 0.0078, "step": 26650 }, { "grad_norm": 0.12467428296804428, "learning_rate": 4.86476266350587e-05, "loss": 0.0092, "step": 26660 }, { "grad_norm": 0.16408249735832214, "learning_rate": 4.861456963532985e-05, "loss": 0.0084, "step": 26670 }, { "grad_norm": 0.10753535479307175, "learning_rate": 4.8581513241635316e-05, "loss": 0.0074, "step": 26680 }, { "grad_norm": 0.12001270800828934, "learning_rate": 4.8548457468435065e-05, "loss": 0.0078, "step": 26690 }, { "grad_norm": 0.16571955382823944, "learning_rate": 4.851540233018884e-05, "loss": 0.0087, "step": 26700 }, { "grad_norm": 0.1621694564819336, "learning_rate": 4.848234784135608e-05, "loss": 0.0097, "step": 26710 }, { "grad_norm": 0.19314199686050415, "learning_rate": 4.8449294016395916e-05, "loss": 0.0086, "step": 26720 }, { "grad_norm": 0.12604372203350067, "learning_rate": 4.8416240869767246e-05, "loss": 0.0072, "step": 26730 }, { "grad_norm": 0.1437527984380722, "learning_rate": 4.838318841592864e-05, "loss": 0.0061, "step": 26740 }, { "grad_norm": 0.13979856669902802, "learning_rate": 4.835013666933834e-05, "loss": 0.008, "step": 26750 }, { "grad_norm": 0.09638532251119614, "learning_rate": 4.83170856444543e-05, "loss": 0.0099, "step": 26760 }, { "grad_norm": 0.16591474413871765, "learning_rate": 4.828403535573419e-05, "loss": 0.007, "step": 26770 }, { "grad_norm": 0.14603076875209808, "learning_rate": 4.825098581763532e-05, "loss": 0.0108, "step": 26780 }, { "grad_norm": 0.11811744421720505, "learning_rate": 4.821793704461465e-05, "loss": 0.0063, "step": 26790 }, { "grad_norm": 0.13900168240070343, "learning_rate": 4.8184889051128884e-05, "loss": 0.0091, "step": 26800 }, { "grad_norm": 0.12443320453166962, "learning_rate": 4.81518418516343e-05, "loss": 0.0083, "step": 26810 }, { "grad_norm": 0.15667197108268738, "learning_rate": 4.811879546058688e-05, "loss": 0.0089, "step": 26820 }, { "grad_norm": 0.1769925206899643, "learning_rate": 4.808574989244224e-05, "loss": 0.0101, "step": 26830 }, { "grad_norm": 0.13002750277519226, "learning_rate": 4.8052705161655644e-05, "loss": 0.0084, "step": 26840 }, { "grad_norm": 0.11544995754957199, "learning_rate": 4.8019661282681936e-05, "loss": 0.0079, "step": 26850 }, { "grad_norm": 0.225580096244812, "learning_rate": 4.79866182699757e-05, "loss": 0.0069, "step": 26860 }, { "grad_norm": 0.22588399052619934, "learning_rate": 4.795357613799103e-05, "loss": 0.0098, "step": 26870 }, { "grad_norm": 0.2503620982170105, "learning_rate": 4.792053490118166e-05, "loss": 0.009, "step": 26880 }, { "grad_norm": 0.1696302890777588, "learning_rate": 4.788749457400099e-05, "loss": 0.0103, "step": 26890 }, { "grad_norm": 0.12080393731594086, "learning_rate": 4.7854455170901966e-05, "loss": 0.0086, "step": 26900 }, { "grad_norm": 0.1539495885372162, "learning_rate": 4.7821416706337136e-05, "loss": 0.0074, "step": 26910 }, { "grad_norm": 0.09114440530538559, "learning_rate": 4.7788379194758645e-05, "loss": 0.0089, "step": 26920 }, { "grad_norm": 0.1703728884458542, "learning_rate": 4.7755342650618246e-05, "loss": 0.0081, "step": 26930 }, { "grad_norm": 0.1063183918595314, "learning_rate": 4.7722307088367235e-05, "loss": 0.0078, "step": 26940 }, { "grad_norm": 0.23018421232700348, "learning_rate": 4.768927252245646e-05, "loss": 0.0105, "step": 26950 }, { "grad_norm": 0.15511538088321686, "learning_rate": 4.765623896733642e-05, "loss": 0.0098, "step": 26960 }, { "grad_norm": 0.13948042690753937, "learning_rate": 4.762320643745708e-05, "loss": 0.0086, "step": 26970 }, { "grad_norm": 0.1702926903963089, "learning_rate": 4.7590174947267966e-05, "loss": 0.0079, "step": 26980 }, { "grad_norm": 0.18395258486270905, "learning_rate": 4.755714451121823e-05, "loss": 0.0094, "step": 26990 }, { "grad_norm": 0.20380418002605438, "learning_rate": 4.7524115143756476e-05, "loss": 0.0077, "step": 27000 }, { "grad_norm": 0.15172937512397766, "learning_rate": 4.7491086859330854e-05, "loss": 0.0091, "step": 27010 }, { "grad_norm": 0.15123170614242554, "learning_rate": 4.74580596723891e-05, "loss": 0.0085, "step": 27020 }, { "grad_norm": 0.1406567394733429, "learning_rate": 4.742503359737841e-05, "loss": 0.009, "step": 27030 }, { "grad_norm": 0.14463558793067932, "learning_rate": 4.739200864874547e-05, "loss": 0.0077, "step": 27040 }, { "grad_norm": 0.14355133473873138, "learning_rate": 4.735898484093658e-05, "loss": 0.0075, "step": 27050 }, { "grad_norm": 0.17385853826999664, "learning_rate": 4.7325962188397444e-05, "loss": 0.0099, "step": 27060 }, { "grad_norm": 0.15324537456035614, "learning_rate": 4.729294070557329e-05, "loss": 0.0091, "step": 27070 }, { "grad_norm": 0.19629324972629547, "learning_rate": 4.7259920406908833e-05, "loss": 0.0092, "step": 27080 }, { "grad_norm": 0.13937361538410187, "learning_rate": 4.7226901306848306e-05, "loss": 0.0066, "step": 27090 }, { "grad_norm": 0.18010427057743073, "learning_rate": 4.719388341983535e-05, "loss": 0.0072, "step": 27100 }, { "grad_norm": 0.15190362930297852, "learning_rate": 4.71608667603131e-05, "loss": 0.0069, "step": 27110 }, { "grad_norm": 0.09779579192399979, "learning_rate": 4.712785134272422e-05, "loss": 0.0077, "step": 27120 }, { "grad_norm": 0.1456274390220642, "learning_rate": 4.7094837181510734e-05, "loss": 0.0077, "step": 27130 }, { "grad_norm": 0.1464596539735794, "learning_rate": 4.706182429111414e-05, "loss": 0.0068, "step": 27140 }, { "grad_norm": 0.21485207974910736, "learning_rate": 4.702881268597545e-05, "loss": 0.0098, "step": 27150 }, { "grad_norm": 0.15098707377910614, "learning_rate": 4.699580238053503e-05, "loss": 0.0104, "step": 27160 }, { "grad_norm": 0.18974517285823822, "learning_rate": 4.69627933892327e-05, "loss": 0.0108, "step": 27170 }, { "grad_norm": 0.14983804523944855, "learning_rate": 4.692978572650773e-05, "loss": 0.0092, "step": 27180 }, { "grad_norm": 0.1043538749217987, "learning_rate": 4.689677940679881e-05, "loss": 0.0104, "step": 27190 }, { "grad_norm": 0.14432166516780853, "learning_rate": 4.6863774444543984e-05, "loss": 0.0075, "step": 27200 }, { "grad_norm": 0.1312498152256012, "learning_rate": 4.6830770854180784e-05, "loss": 0.008, "step": 27210 }, { "grad_norm": 0.1486956775188446, "learning_rate": 4.679776865014608e-05, "loss": 0.0061, "step": 27220 }, { "grad_norm": 0.2027381956577301, "learning_rate": 4.676476784687615e-05, "loss": 0.0075, "step": 27230 }, { "grad_norm": 0.1484508365392685, "learning_rate": 4.673176845880669e-05, "loss": 0.0079, "step": 27240 }, { "grad_norm": 0.13058261573314667, "learning_rate": 4.669877050037275e-05, "loss": 0.0084, "step": 27250 }, { "grad_norm": 0.13401292264461517, "learning_rate": 4.666577398600874e-05, "loss": 0.0087, "step": 27260 }, { "grad_norm": 0.1323685497045517, "learning_rate": 4.663277893014846e-05, "loss": 0.0079, "step": 27270 }, { "grad_norm": 0.09244042634963989, "learning_rate": 4.659978534722508e-05, "loss": 0.0074, "step": 27280 }, { "grad_norm": 0.14284023642539978, "learning_rate": 4.6566793251671104e-05, "loss": 0.0082, "step": 27290 }, { "grad_norm": 0.1405511498451233, "learning_rate": 4.65338026579184e-05, "loss": 0.011, "step": 27300 }, { "grad_norm": 0.14764194190502167, "learning_rate": 4.650081358039817e-05, "loss": 0.0087, "step": 27310 }, { "grad_norm": 0.18516004085540771, "learning_rate": 4.646782603354098e-05, "loss": 0.0077, "step": 27320 }, { "grad_norm": 0.15365156531333923, "learning_rate": 4.643484003177666e-05, "loss": 0.0066, "step": 27330 }, { "grad_norm": 0.12926176190376282, "learning_rate": 4.6401855589534454e-05, "loss": 0.0073, "step": 27340 }, { "grad_norm": 0.14886344969272614, "learning_rate": 4.636887272124286e-05, "loss": 0.0071, "step": 27350 }, { "grad_norm": 0.14515838027000427, "learning_rate": 4.6335891441329685e-05, "loss": 0.0069, "step": 27360 }, { "grad_norm": 0.163442924618721, "learning_rate": 4.630291176422209e-05, "loss": 0.009, "step": 27370 }, { "grad_norm": 0.15796096622943878, "learning_rate": 4.62699337043465e-05, "loss": 0.0089, "step": 27380 }, { "grad_norm": 0.13514824211597443, "learning_rate": 4.623695727612861e-05, "loss": 0.0081, "step": 27390 }, { "grad_norm": 0.1224818006157875, "learning_rate": 4.620398249399347e-05, "loss": 0.0072, "step": 27400 }, { "grad_norm": 0.16166888177394867, "learning_rate": 4.617100937236535e-05, "loss": 0.0082, "step": 27410 }, { "grad_norm": 0.11624795943498611, "learning_rate": 4.613803792566781e-05, "loss": 0.008, "step": 27420 }, { "grad_norm": 0.14643535017967224, "learning_rate": 4.61050681683237e-05, "loss": 0.0101, "step": 27430 }, { "grad_norm": 0.22627896070480347, "learning_rate": 4.6072100114755104e-05, "loss": 0.0089, "step": 27440 }, { "grad_norm": 0.1593572050333023, "learning_rate": 4.603913377938336e-05, "loss": 0.0101, "step": 27450 }, { "grad_norm": 0.12138446420431137, "learning_rate": 4.600616917662905e-05, "loss": 0.0068, "step": 27460 }, { "grad_norm": 0.08974451571702957, "learning_rate": 4.597320632091206e-05, "loss": 0.0092, "step": 27470 }, { "grad_norm": 0.13080674409866333, "learning_rate": 4.594024522665142e-05, "loss": 0.0084, "step": 27480 }, { "grad_norm": 0.12999208271503448, "learning_rate": 4.590728590826543e-05, "loss": 0.0073, "step": 27490 }, { "grad_norm": 0.11806473135948181, "learning_rate": 4.587432838017165e-05, "loss": 0.0087, "step": 27500 }, { "grad_norm": 0.0952676311135292, "learning_rate": 4.58413726567868e-05, "loss": 0.007, "step": 27510 }, { "grad_norm": 0.13310304284095764, "learning_rate": 4.5808418752526814e-05, "loss": 0.008, "step": 27520 }, { "grad_norm": 0.13726706802845, "learning_rate": 4.5775466681806895e-05, "loss": 0.0063, "step": 27530 }, { "grad_norm": 0.13554976880550385, "learning_rate": 4.574251645904136e-05, "loss": 0.0083, "step": 27540 }, { "grad_norm": 0.11969169974327087, "learning_rate": 4.5709568098643776e-05, "loss": 0.0079, "step": 27550 }, { "grad_norm": 0.13308684527873993, "learning_rate": 4.567662161502687e-05, "loss": 0.0064, "step": 27560 }, { "grad_norm": 0.1467115730047226, "learning_rate": 4.5643677022602554e-05, "loss": 0.0091, "step": 27570 }, { "grad_norm": 0.1646692156791687, "learning_rate": 4.56107343357819e-05, "loss": 0.0083, "step": 27580 }, { "grad_norm": 0.09667795151472092, "learning_rate": 4.557779356897518e-05, "loss": 0.0096, "step": 27590 }, { "grad_norm": 0.13412897288799286, "learning_rate": 4.55448547365918e-05, "loss": 0.0074, "step": 27600 }, { "grad_norm": 0.08667556941509247, "learning_rate": 4.5511917853040315e-05, "loss": 0.009, "step": 27610 }, { "grad_norm": 0.1793396770954132, "learning_rate": 4.547898293272842e-05, "loss": 0.0116, "step": 27620 }, { "grad_norm": 0.22709552943706512, "learning_rate": 4.544604999006301e-05, "loss": 0.0092, "step": 27630 }, { "grad_norm": 0.17709824442863464, "learning_rate": 4.541311903945004e-05, "loss": 0.0081, "step": 27640 }, { "grad_norm": 0.1311289668083191, "learning_rate": 4.538019009529463e-05, "loss": 0.0084, "step": 27650 }, { "grad_norm": 0.14899371564388275, "learning_rate": 4.5347263172001035e-05, "loss": 0.0073, "step": 27660 }, { "grad_norm": 0.12574315071105957, "learning_rate": 4.531433828397261e-05, "loss": 0.0068, "step": 27670 }, { "grad_norm": 0.17460143566131592, "learning_rate": 4.528141544561179e-05, "loss": 0.0066, "step": 27680 }, { "grad_norm": 0.16737258434295654, "learning_rate": 4.5248494671320186e-05, "loss": 0.0086, "step": 27690 }, { "grad_norm": 0.14004531502723694, "learning_rate": 4.521557597549844e-05, "loss": 0.0088, "step": 27700 }, { "grad_norm": 0.17165717482566833, "learning_rate": 4.5182659372546284e-05, "loss": 0.0073, "step": 27710 }, { "grad_norm": 0.16718690097332, "learning_rate": 4.514974487686261e-05, "loss": 0.0059, "step": 27720 }, { "grad_norm": 0.11457915604114532, "learning_rate": 4.5116832502845306e-05, "loss": 0.0062, "step": 27730 }, { "grad_norm": 0.10730194300413132, "learning_rate": 4.508392226489136e-05, "loss": 0.0075, "step": 27740 }, { "grad_norm": 0.14459261298179626, "learning_rate": 4.505101417739683e-05, "loss": 0.0088, "step": 27750 }, { "grad_norm": 0.15105773508548737, "learning_rate": 4.501810825475685e-05, "loss": 0.0074, "step": 27760 }, { "grad_norm": 0.13704988360404968, "learning_rate": 4.4985204511365556e-05, "loss": 0.0102, "step": 27770 }, { "grad_norm": 0.11607807874679565, "learning_rate": 4.49523029616162e-05, "loss": 0.0064, "step": 27780 }, { "grad_norm": 0.13317421078681946, "learning_rate": 4.491940361990101e-05, "loss": 0.0082, "step": 27790 }, { "grad_norm": 0.14538483321666718, "learning_rate": 4.48865065006113e-05, "loss": 0.0076, "step": 27800 }, { "grad_norm": 0.16247016191482544, "learning_rate": 4.485361161813735e-05, "loss": 0.0081, "step": 27810 }, { "grad_norm": 0.2160690426826477, "learning_rate": 4.482071898686854e-05, "loss": 0.0094, "step": 27820 }, { "grad_norm": 0.14305977523326874, "learning_rate": 4.478782862119321e-05, "loss": 0.0072, "step": 27830 }, { "grad_norm": 0.13947628438472748, "learning_rate": 4.475494053549869e-05, "loss": 0.0069, "step": 27840 }, { "grad_norm": 0.14882729947566986, "learning_rate": 4.4722054744171405e-05, "loss": 0.0071, "step": 27850 }, { "grad_norm": 0.11473682522773743, "learning_rate": 4.468917126159669e-05, "loss": 0.0079, "step": 27860 }, { "grad_norm": 0.1530591994524002, "learning_rate": 4.465629010215887e-05, "loss": 0.009, "step": 27870 }, { "grad_norm": 0.13073429465293884, "learning_rate": 4.462341128024133e-05, "loss": 0.0089, "step": 27880 }, { "grad_norm": 0.14474846422672272, "learning_rate": 4.4590534810226356e-05, "loss": 0.0096, "step": 27890 }, { "grad_norm": 0.1513185054063797, "learning_rate": 4.455766070649523e-05, "loss": 0.0076, "step": 27900 }, { "grad_norm": 0.15572009980678558, "learning_rate": 4.4524788983428215e-05, "loss": 0.0106, "step": 27910 }, { "grad_norm": 0.1659800112247467, "learning_rate": 4.4491919655404515e-05, "loss": 0.0068, "step": 27920 }, { "grad_norm": 0.1433635801076889, "learning_rate": 4.445905273680228e-05, "loss": 0.0092, "step": 27930 }, { "grad_norm": 0.1397186815738678, "learning_rate": 4.442618824199865e-05, "loss": 0.007, "step": 27940 }, { "grad_norm": 0.13269540667533875, "learning_rate": 4.4393326185369635e-05, "loss": 0.0098, "step": 27950 }, { "grad_norm": 0.22368568181991577, "learning_rate": 4.436046658129024e-05, "loss": 0.0072, "step": 27960 }, { "grad_norm": 0.18292361497879028, "learning_rate": 4.432760944413433e-05, "loss": 0.0091, "step": 27970 }, { "grad_norm": 0.18586917221546173, "learning_rate": 4.4294754788274796e-05, "loss": 0.0077, "step": 27980 }, { "grad_norm": 0.16469186544418335, "learning_rate": 4.426190262808334e-05, "loss": 0.0099, "step": 27990 }, { "grad_norm": 0.18346966803073883, "learning_rate": 4.42290529779306e-05, "loss": 0.0077, "step": 28000 }, { "grad_norm": 0.11697409301996231, "learning_rate": 4.4196205852186176e-05, "loss": 0.0091, "step": 28010 }, { "grad_norm": 0.18374602496623993, "learning_rate": 4.416336126521848e-05, "loss": 0.0072, "step": 28020 }, { "grad_norm": 0.16694128513336182, "learning_rate": 4.4130519231394865e-05, "loss": 0.0085, "step": 28030 }, { "grad_norm": 0.1302739679813385, "learning_rate": 4.409767976508154e-05, "loss": 0.0077, "step": 28040 }, { "grad_norm": 0.1494555026292801, "learning_rate": 4.406484288064363e-05, "loss": 0.0105, "step": 28050 }, { "grad_norm": 0.16693352162837982, "learning_rate": 4.4032008592445055e-05, "loss": 0.0067, "step": 28060 }, { "grad_norm": 0.1640246957540512, "learning_rate": 4.3999176914848714e-05, "loss": 0.0067, "step": 28070 }, { "grad_norm": 0.14687533676624298, "learning_rate": 4.396634786221625e-05, "loss": 0.007, "step": 28080 }, { "grad_norm": 0.20133449137210846, "learning_rate": 4.3933521448908203e-05, "loss": 0.0057, "step": 28090 }, { "grad_norm": 0.12545138597488403, "learning_rate": 4.3900697689284e-05, "loss": 0.0094, "step": 28100 }, { "grad_norm": 0.1831011176109314, "learning_rate": 4.386787659770185e-05, "loss": 0.0071, "step": 28110 }, { "grad_norm": 0.14206328988075256, "learning_rate": 4.383505818851878e-05, "loss": 0.0067, "step": 28120 }, { "grad_norm": 0.1871981918811798, "learning_rate": 4.380224247609073e-05, "loss": 0.007, "step": 28130 }, { "grad_norm": 0.17728067934513092, "learning_rate": 4.3769429474772376e-05, "loss": 0.0077, "step": 28140 }, { "grad_norm": 0.19706743955612183, "learning_rate": 4.373661919891726e-05, "loss": 0.0062, "step": 28150 }, { "grad_norm": 0.10145439952611923, "learning_rate": 4.3703811662877674e-05, "loss": 0.0081, "step": 28160 }, { "grad_norm": 0.15407690405845642, "learning_rate": 4.36710068810048e-05, "loss": 0.0088, "step": 28170 }, { "grad_norm": 0.12307840585708618, "learning_rate": 4.363820486764853e-05, "loss": 0.0064, "step": 28180 }, { "grad_norm": 0.14958856999874115, "learning_rate": 4.360540563715758e-05, "loss": 0.009, "step": 28190 }, { "grad_norm": 0.16404154896736145, "learning_rate": 4.357260920387946e-05, "loss": 0.0072, "step": 28200 }, { "grad_norm": 0.11903741210699081, "learning_rate": 4.353981558216045e-05, "loss": 0.0082, "step": 28210 }, { "grad_norm": 0.13816912472248077, "learning_rate": 4.3507024786345565e-05, "loss": 0.0084, "step": 28220 }, { "grad_norm": 0.1196524053812027, "learning_rate": 4.3474236830778646e-05, "loss": 0.0069, "step": 28230 }, { "grad_norm": 0.13578474521636963, "learning_rate": 4.344145172980223e-05, "loss": 0.0091, "step": 28240 }, { "grad_norm": 0.14793013036251068, "learning_rate": 4.3408669497757656e-05, "loss": 0.0082, "step": 28250 }, { "grad_norm": 0.11074626445770264, "learning_rate": 4.3375890148984976e-05, "loss": 0.0085, "step": 28260 }, { "grad_norm": 0.15880554914474487, "learning_rate": 4.334311369782301e-05, "loss": 0.0064, "step": 28270 }, { "grad_norm": 0.1308412402868271, "learning_rate": 4.3310340158609244e-05, "loss": 0.0076, "step": 28280 }, { "grad_norm": 0.15407396852970123, "learning_rate": 4.327756954568e-05, "loss": 0.0069, "step": 28290 }, { "grad_norm": 0.16738440096378326, "learning_rate": 4.324480187337021e-05, "loss": 0.0077, "step": 28300 }, { "grad_norm": 0.08279238641262054, "learning_rate": 4.3212037156013596e-05, "loss": 0.0087, "step": 28310 }, { "grad_norm": 0.15732716023921967, "learning_rate": 4.317927540794251e-05, "loss": 0.0095, "step": 28320 }, { "grad_norm": 0.07117342948913574, "learning_rate": 4.314651664348812e-05, "loss": 0.0062, "step": 28330 }, { "grad_norm": 0.13863606750965118, "learning_rate": 4.3113760876980195e-05, "loss": 0.0064, "step": 28340 }, { "grad_norm": 0.10994623601436615, "learning_rate": 4.308100812274719e-05, "loss": 0.0061, "step": 28350 }, { "grad_norm": 0.12458813190460205, "learning_rate": 4.304825839511632e-05, "loss": 0.0058, "step": 28360 }, { "grad_norm": 0.14087443053722382, "learning_rate": 4.3015511708413395e-05, "loss": 0.0066, "step": 28370 }, { "grad_norm": 0.13031570613384247, "learning_rate": 4.298276807696294e-05, "loss": 0.0072, "step": 28380 }, { "grad_norm": 0.19550205767154694, "learning_rate": 4.2950027515088136e-05, "loss": 0.0074, "step": 28390 }, { "grad_norm": 0.13662177324295044, "learning_rate": 4.291729003711083e-05, "loss": 0.0057, "step": 28400 }, { "grad_norm": 0.19085757434368134, "learning_rate": 4.2884555657351455e-05, "loss": 0.0105, "step": 28410 }, { "grad_norm": 0.2661955952644348, "learning_rate": 4.285182439012921e-05, "loss": 0.0073, "step": 28420 }, { "grad_norm": 0.14414453506469727, "learning_rate": 4.281909624976183e-05, "loss": 0.0086, "step": 28430 }, { "grad_norm": 0.08106953650712967, "learning_rate": 4.278637125056568e-05, "loss": 0.0074, "step": 28440 }, { "grad_norm": 0.1643989086151123, "learning_rate": 4.2753649406855855e-05, "loss": 0.0083, "step": 28450 }, { "grad_norm": 0.22294284403324127, "learning_rate": 4.2720930732945966e-05, "loss": 0.0082, "step": 28460 }, { "grad_norm": 0.15255144238471985, "learning_rate": 4.268821524314826e-05, "loss": 0.0091, "step": 28470 }, { "grad_norm": 0.15241102874279022, "learning_rate": 4.2655502951773623e-05, "loss": 0.0085, "step": 28480 }, { "grad_norm": 0.12037118524312973, "learning_rate": 4.262279387313152e-05, "loss": 0.0071, "step": 28490 }, { "grad_norm": 0.17415282130241394, "learning_rate": 4.259008802153001e-05, "loss": 0.0078, "step": 28500 }, { "grad_norm": 0.10983295738697052, "learning_rate": 4.2557385411275735e-05, "loss": 0.0057, "step": 28510 }, { "grad_norm": 0.14464308321475983, "learning_rate": 4.252468605667395e-05, "loss": 0.0063, "step": 28520 }, { "grad_norm": 0.12025471031665802, "learning_rate": 4.2491989972028456e-05, "loss": 0.0079, "step": 28530 }, { "grad_norm": 0.12627336382865906, "learning_rate": 4.245929717164161e-05, "loss": 0.008, "step": 28540 }, { "grad_norm": 0.15427157282829285, "learning_rate": 4.242660766981438e-05, "loss": 0.0081, "step": 28550 }, { "grad_norm": 0.1430012732744217, "learning_rate": 4.2393921480846264e-05, "loss": 0.0083, "step": 28560 }, { "grad_norm": 0.1580684334039688, "learning_rate": 4.236123861903528e-05, "loss": 0.0081, "step": 28570 }, { "grad_norm": 0.1401800513267517, "learning_rate": 4.232855909867806e-05, "loss": 0.0071, "step": 28580 }, { "grad_norm": 0.12529698014259338, "learning_rate": 4.229588293406972e-05, "loss": 0.0068, "step": 28590 }, { "grad_norm": 0.12613633275032043, "learning_rate": 4.226321013950392e-05, "loss": 0.0062, "step": 28600 }, { "grad_norm": 0.17575603723526, "learning_rate": 4.223054072927286e-05, "loss": 0.0072, "step": 28610 }, { "grad_norm": 0.10603834688663483, "learning_rate": 4.219787471766724e-05, "loss": 0.0069, "step": 28620 }, { "grad_norm": 0.1281326413154602, "learning_rate": 4.216521211897629e-05, "loss": 0.0093, "step": 28630 }, { "grad_norm": 0.14935778081417084, "learning_rate": 4.2132552947487734e-05, "loss": 0.0064, "step": 28640 }, { "grad_norm": 0.18319158256053925, "learning_rate": 4.209989721748783e-05, "loss": 0.0082, "step": 28650 }, { "grad_norm": 0.1261778026819229, "learning_rate": 4.206724494326127e-05, "loss": 0.0059, "step": 28660 }, { "grad_norm": 0.10789172351360321, "learning_rate": 4.203459613909126e-05, "loss": 0.0061, "step": 28670 }, { "grad_norm": 0.1825706958770752, "learning_rate": 4.200195081925955e-05, "loss": 0.008, "step": 28680 }, { "grad_norm": 0.15148289501667023, "learning_rate": 4.196930899804627e-05, "loss": 0.0076, "step": 28690 }, { "grad_norm": 0.19923996925354004, "learning_rate": 4.1936670689730044e-05, "loss": 0.0089, "step": 28700 }, { "grad_norm": 0.17884714901447296, "learning_rate": 4.190403590858804e-05, "loss": 0.0076, "step": 28710 }, { "grad_norm": 0.12569214403629303, "learning_rate": 4.187140466889575e-05, "loss": 0.0057, "step": 28720 }, { "grad_norm": 0.10998191684484482, "learning_rate": 4.183877698492724e-05, "loss": 0.0077, "step": 28730 }, { "grad_norm": 0.10987468808889389, "learning_rate": 4.1806152870954935e-05, "loss": 0.0071, "step": 28740 }, { "grad_norm": 0.13963964581489563, "learning_rate": 4.177353234124976e-05, "loss": 0.0083, "step": 28750 }, { "grad_norm": 0.11367154121398926, "learning_rate": 4.1740915410081e-05, "loss": 0.0065, "step": 28760 }, { "grad_norm": 0.14432018995285034, "learning_rate": 4.170830209171646e-05, "loss": 0.008, "step": 28770 }, { "grad_norm": 0.1561034470796585, "learning_rate": 4.1675692400422286e-05, "loss": 0.008, "step": 28780 }, { "grad_norm": 0.14139051735401154, "learning_rate": 4.164308635046304e-05, "loss": 0.0063, "step": 28790 }, { "grad_norm": 0.11680980026721954, "learning_rate": 4.161048395610177e-05, "loss": 0.0065, "step": 28800 }, { "grad_norm": 0.1564752161502838, "learning_rate": 4.157788523159985e-05, "loss": 0.007, "step": 28810 }, { "grad_norm": 0.11426597833633423, "learning_rate": 4.154529019121706e-05, "loss": 0.0092, "step": 28820 }, { "grad_norm": 0.10029590874910355, "learning_rate": 4.1512698849211564e-05, "loss": 0.0062, "step": 28830 }, { "grad_norm": 0.13963806629180908, "learning_rate": 4.1480111219839954e-05, "loss": 0.0073, "step": 28840 }, { "grad_norm": 0.1106749027967453, "learning_rate": 4.144752731735714e-05, "loss": 0.0065, "step": 28850 }, { "grad_norm": 0.10804393142461777, "learning_rate": 4.1414947156016446e-05, "loss": 0.0061, "step": 28860 }, { "grad_norm": 0.17653991281986237, "learning_rate": 4.138237075006953e-05, "loss": 0.0069, "step": 28870 }, { "grad_norm": 0.13941730558872223, "learning_rate": 4.134979811376645e-05, "loss": 0.0069, "step": 28880 }, { "grad_norm": 0.08912739902734756, "learning_rate": 4.131722926135552e-05, "loss": 0.0066, "step": 28890 }, { "grad_norm": 0.12177515774965286, "learning_rate": 4.1284664207083526e-05, "loss": 0.0079, "step": 28900 }, { "grad_norm": 0.12719491124153137, "learning_rate": 4.125210296519551e-05, "loss": 0.0067, "step": 28910 }, { "grad_norm": 0.15825721621513367, "learning_rate": 4.121954554993484e-05, "loss": 0.006, "step": 28920 }, { "grad_norm": 0.16651098430156708, "learning_rate": 4.118699197554327e-05, "loss": 0.0095, "step": 28930 }, { "grad_norm": 0.10170853137969971, "learning_rate": 4.1154442256260814e-05, "loss": 0.0058, "step": 28940 }, { "grad_norm": 0.15266484022140503, "learning_rate": 4.1121896406325824e-05, "loss": 0.0072, "step": 28950 }, { "grad_norm": 0.15821045637130737, "learning_rate": 4.108935443997499e-05, "loss": 0.0101, "step": 28960 }, { "grad_norm": 0.15456902980804443, "learning_rate": 4.105681637144325e-05, "loss": 0.0076, "step": 28970 }, { "grad_norm": 0.14699554443359375, "learning_rate": 4.102428221496387e-05, "loss": 0.0076, "step": 28980 }, { "grad_norm": 0.19079245626926422, "learning_rate": 4.099175198476838e-05, "loss": 0.0082, "step": 28990 }, { "grad_norm": 0.16466794908046722, "learning_rate": 4.0959225695086635e-05, "loss": 0.0078, "step": 29000 }, { "grad_norm": 0.1305655837059021, "learning_rate": 4.0926703360146724e-05, "loss": 0.0069, "step": 29010 }, { "grad_norm": 0.1401958018541336, "learning_rate": 4.0894184994175e-05, "loss": 0.008, "step": 29020 }, { "grad_norm": 0.15661393105983734, "learning_rate": 4.086167061139614e-05, "loss": 0.0067, "step": 29030 }, { "grad_norm": 0.14112231135368347, "learning_rate": 4.082916022603303e-05, "loss": 0.0098, "step": 29040 }, { "grad_norm": 0.1777878850698471, "learning_rate": 4.079665385230678e-05, "loss": 0.0052, "step": 29050 }, { "grad_norm": 0.08252432942390442, "learning_rate": 4.076415150443683e-05, "loss": 0.0068, "step": 29060 }, { "grad_norm": 0.19716520607471466, "learning_rate": 4.073165319664079e-05, "loss": 0.0076, "step": 29070 }, { "grad_norm": 0.13176649808883667, "learning_rate": 4.069915894313451e-05, "loss": 0.0084, "step": 29080 }, { "grad_norm": 0.13359929621219635, "learning_rate": 4.066666875813212e-05, "loss": 0.0071, "step": 29090 }, { "grad_norm": 0.16766643524169922, "learning_rate": 4.063418265584589e-05, "loss": 0.0074, "step": 29100 }, { "grad_norm": 0.13836444914340973, "learning_rate": 4.060170065048636e-05, "loss": 0.0065, "step": 29110 }, { "grad_norm": 0.15904459357261658, "learning_rate": 4.056922275626227e-05, "loss": 0.0074, "step": 29120 }, { "grad_norm": 0.09511557966470718, "learning_rate": 4.053674898738056e-05, "loss": 0.0073, "step": 29130 }, { "grad_norm": 0.1139683648943901, "learning_rate": 4.050427935804633e-05, "loss": 0.0093, "step": 29140 }, { "grad_norm": 0.13235802948474884, "learning_rate": 4.047181388246293e-05, "loss": 0.0081, "step": 29150 }, { "grad_norm": 0.15228821337223053, "learning_rate": 4.043935257483185e-05, "loss": 0.0079, "step": 29160 }, { "grad_norm": 0.11099065095186234, "learning_rate": 4.0406895449352775e-05, "loss": 0.008, "step": 29170 }, { "grad_norm": 0.1364157795906067, "learning_rate": 4.0374442520223525e-05, "loss": 0.0079, "step": 29180 }, { "grad_norm": 0.14120294153690338, "learning_rate": 4.0341993801640155e-05, "loss": 0.0069, "step": 29190 }, { "grad_norm": 0.10320484638214111, "learning_rate": 4.030954930779681e-05, "loss": 0.0058, "step": 29200 }, { "grad_norm": 0.14853410422801971, "learning_rate": 4.027710905288582e-05, "loss": 0.007, "step": 29210 }, { "grad_norm": 0.15458475053310394, "learning_rate": 4.0244673051097666e-05, "loss": 0.0079, "step": 29220 }, { "grad_norm": 0.18790356814861298, "learning_rate": 4.021224131662095e-05, "loss": 0.0077, "step": 29230 }, { "grad_norm": 0.28429755568504333, "learning_rate": 4.0179813863642396e-05, "loss": 0.0095, "step": 29240 }, { "grad_norm": 0.16404642164707184, "learning_rate": 4.014739070634691e-05, "loss": 0.0083, "step": 29250 }, { "grad_norm": 0.17955516278743744, "learning_rate": 4.011497185891746e-05, "loss": 0.0066, "step": 29260 }, { "grad_norm": 0.09027164429426193, "learning_rate": 4.008255733553513e-05, "loss": 0.0069, "step": 29270 }, { "grad_norm": 0.1345921754837036, "learning_rate": 4.0050147150379194e-05, "loss": 0.0068, "step": 29280 }, { "grad_norm": 0.12986038625240326, "learning_rate": 4.001774131762692e-05, "loss": 0.0078, "step": 29290 }, { "grad_norm": 0.15255150198936462, "learning_rate": 3.9985339851453705e-05, "loss": 0.009, "step": 29300 }, { "grad_norm": 0.1381414383649826, "learning_rate": 3.9952942766033114e-05, "loss": 0.0069, "step": 29310 }, { "grad_norm": 0.14784982800483704, "learning_rate": 3.9920550075536686e-05, "loss": 0.007, "step": 29320 }, { "grad_norm": 0.11838902533054352, "learning_rate": 3.988816179413408e-05, "loss": 0.0077, "step": 29330 }, { "grad_norm": 0.10508493334054947, "learning_rate": 3.985577793599307e-05, "loss": 0.0057, "step": 29340 }, { "grad_norm": 0.1376771330833435, "learning_rate": 3.982339851527942e-05, "loss": 0.0075, "step": 29350 }, { "grad_norm": 0.13473562896251678, "learning_rate": 3.979102354615701e-05, "loss": 0.0079, "step": 29360 }, { "grad_norm": 0.17410540580749512, "learning_rate": 3.975865304278772e-05, "loss": 0.0072, "step": 29370 }, { "grad_norm": 0.17104153335094452, "learning_rate": 3.9726287019331546e-05, "loss": 0.0062, "step": 29380 }, { "grad_norm": 0.16404308378696442, "learning_rate": 3.9693925489946474e-05, "loss": 0.0069, "step": 29390 }, { "grad_norm": 0.11914695799350739, "learning_rate": 3.966156846878852e-05, "loss": 0.0062, "step": 29400 }, { "grad_norm": 0.12534192204475403, "learning_rate": 3.962921597001177e-05, "loss": 0.0055, "step": 29410 }, { "grad_norm": 0.18727122247219086, "learning_rate": 3.9596868007768294e-05, "loss": 0.0055, "step": 29420 }, { "grad_norm": 0.1155972108244896, "learning_rate": 3.9564524596208165e-05, "loss": 0.0082, "step": 29430 }, { "grad_norm": 0.1405165195465088, "learning_rate": 3.953218574947954e-05, "loss": 0.0075, "step": 29440 }, { "grad_norm": 0.15174277126789093, "learning_rate": 3.949985148172849e-05, "loss": 0.0073, "step": 29450 }, { "grad_norm": 0.15734711289405823, "learning_rate": 3.946752180709914e-05, "loss": 0.0074, "step": 29460 }, { "grad_norm": 0.14004652202129364, "learning_rate": 3.943519673973357e-05, "loss": 0.0054, "step": 29470 }, { "grad_norm": 0.16453224420547485, "learning_rate": 3.940287629377189e-05, "loss": 0.0079, "step": 29480 }, { "grad_norm": 0.13899318873882294, "learning_rate": 3.937056048335214e-05, "loss": 0.0058, "step": 29490 }, { "grad_norm": 0.1495346575975418, "learning_rate": 3.933824932261037e-05, "loss": 0.0073, "step": 29500 }, { "grad_norm": 0.10618031769990921, "learning_rate": 3.9305942825680576e-05, "loss": 0.0079, "step": 29510 }, { "grad_norm": 0.2144598364830017, "learning_rate": 3.9273641006694706e-05, "loss": 0.0077, "step": 29520 }, { "grad_norm": 0.1317562609910965, "learning_rate": 3.924134387978266e-05, "loss": 0.0056, "step": 29530 }, { "grad_norm": 0.13418632745742798, "learning_rate": 3.920905145907234e-05, "loss": 0.0095, "step": 29540 }, { "grad_norm": 0.11092503368854523, "learning_rate": 3.9176763758689525e-05, "loss": 0.007, "step": 29550 }, { "grad_norm": 0.1955035775899887, "learning_rate": 3.9144480792757925e-05, "loss": 0.0083, "step": 29560 }, { "grad_norm": 0.1378433257341385, "learning_rate": 3.911220257539926e-05, "loss": 0.007, "step": 29570 }, { "grad_norm": 0.1296093612909317, "learning_rate": 3.907992912073308e-05, "loss": 0.0064, "step": 29580 }, { "grad_norm": 0.11574333906173706, "learning_rate": 3.90476604428769e-05, "loss": 0.0065, "step": 29590 }, { "grad_norm": 0.19386203587055206, "learning_rate": 3.901539655594614e-05, "loss": 0.0081, "step": 29600 }, { "grad_norm": 0.16601872444152832, "learning_rate": 3.898313747405413e-05, "loss": 0.009, "step": 29610 }, { "grad_norm": 0.16095827519893646, "learning_rate": 3.895088321131205e-05, "loss": 0.0083, "step": 29620 }, { "grad_norm": 0.15072864294052124, "learning_rate": 3.891863378182907e-05, "loss": 0.0061, "step": 29630 }, { "grad_norm": 0.14029937982559204, "learning_rate": 3.8886389199712145e-05, "loss": 0.0061, "step": 29640 }, { "grad_norm": 0.1031031385064125, "learning_rate": 3.885414947906614e-05, "loss": 0.007, "step": 29650 }, { "grad_norm": 0.145295187830925, "learning_rate": 3.882191463399385e-05, "loss": 0.0067, "step": 29660 }, { "grad_norm": 0.11597221344709396, "learning_rate": 3.878968467859585e-05, "loss": 0.0059, "step": 29670 }, { "grad_norm": 0.14223605394363403, "learning_rate": 3.875745962697063e-05, "loss": 0.0063, "step": 29680 }, { "grad_norm": 0.12042297422885895, "learning_rate": 3.872523949321454e-05, "loss": 0.0069, "step": 29690 }, { "grad_norm": 0.15988516807556152, "learning_rate": 3.869302429142173e-05, "loss": 0.0071, "step": 29700 }, { "grad_norm": 0.1563008725643158, "learning_rate": 3.866081403568425e-05, "loss": 0.0082, "step": 29710 }, { "grad_norm": 0.12902937829494476, "learning_rate": 3.8628608740091934e-05, "loss": 0.007, "step": 29720 }, { "grad_norm": 0.12577860057353973, "learning_rate": 3.85964084187325e-05, "loss": 0.0104, "step": 29730 }, { "grad_norm": 0.11105851083993912, "learning_rate": 3.856421308569146e-05, "loss": 0.0059, "step": 29740 }, { "grad_norm": 0.12992660701274872, "learning_rate": 3.853202275505212e-05, "loss": 0.008, "step": 29750 }, { "grad_norm": 0.11627932637929916, "learning_rate": 3.849983744089565e-05, "loss": 0.0064, "step": 29760 }, { "grad_norm": 0.11482642590999603, "learning_rate": 3.8467657157301e-05, "loss": 0.0095, "step": 29770 }, { "grad_norm": 0.10210022330284119, "learning_rate": 3.8435481918344884e-05, "loss": 0.0068, "step": 29780 }, { "grad_norm": 0.13483014702796936, "learning_rate": 3.84033117381019e-05, "loss": 0.0076, "step": 29790 }, { "grad_norm": 0.13984467089176178, "learning_rate": 3.8371146630644335e-05, "loss": 0.0105, "step": 29800 }, { "grad_norm": 0.15887245535850525, "learning_rate": 3.833898661004231e-05, "loss": 0.0054, "step": 29810 }, { "grad_norm": 0.10416021198034286, "learning_rate": 3.830683169036372e-05, "loss": 0.0061, "step": 29820 }, { "grad_norm": 0.14099016785621643, "learning_rate": 3.827468188567422e-05, "loss": 0.008, "step": 29830 }, { "grad_norm": 0.2409650683403015, "learning_rate": 3.82425372100372e-05, "loss": 0.0071, "step": 29840 }, { "grad_norm": 0.2008189707994461, "learning_rate": 3.821039767751387e-05, "loss": 0.0088, "step": 29850 }, { "grad_norm": 0.17827241122722626, "learning_rate": 3.817826330216315e-05, "loss": 0.0087, "step": 29860 }, { "grad_norm": 0.18164938688278198, "learning_rate": 3.814613409804169e-05, "loss": 0.0072, "step": 29870 }, { "grad_norm": 0.1388385146856308, "learning_rate": 3.811401007920388e-05, "loss": 0.0089, "step": 29880 }, { "grad_norm": 0.16243232786655426, "learning_rate": 3.8081891259701895e-05, "loss": 0.009, "step": 29890 }, { "grad_norm": 0.15936732292175293, "learning_rate": 3.8049777653585584e-05, "loss": 0.0071, "step": 29900 }, { "grad_norm": 0.1549166440963745, "learning_rate": 3.80176692749025e-05, "loss": 0.0089, "step": 29910 }, { "grad_norm": 0.1066739559173584, "learning_rate": 3.798556613769799e-05, "loss": 0.0072, "step": 29920 }, { "grad_norm": 0.14979586005210876, "learning_rate": 3.795346825601502e-05, "loss": 0.0091, "step": 29930 }, { "grad_norm": 0.12176202237606049, "learning_rate": 3.7921375643894305e-05, "loss": 0.0086, "step": 29940 }, { "grad_norm": 0.10880230367183685, "learning_rate": 3.788928831537424e-05, "loss": 0.0071, "step": 29950 }, { "grad_norm": 0.1063804104924202, "learning_rate": 3.7857206284490934e-05, "loss": 0.0057, "step": 29960 }, { "grad_norm": 0.1372358202934265, "learning_rate": 3.7825129565278114e-05, "loss": 0.0063, "step": 29970 }, { "grad_norm": 0.10446113348007202, "learning_rate": 3.779305817176728e-05, "loss": 0.0056, "step": 29980 }, { "grad_norm": 0.11523381620645523, "learning_rate": 3.776099211798751e-05, "loss": 0.0064, "step": 29990 }, { "grad_norm": 0.11405347287654877, "learning_rate": 3.7728931417965584e-05, "loss": 0.0065, "step": 30000 }, { "grad_norm": 0.11971959471702576, "learning_rate": 3.769687608572598e-05, "loss": 0.0062, "step": 30010 }, { "grad_norm": 0.12417837977409363, "learning_rate": 3.766482613529076e-05, "loss": 0.01, "step": 30020 }, { "grad_norm": 0.15027642250061035, "learning_rate": 3.763278158067964e-05, "loss": 0.0078, "step": 30030 }, { "grad_norm": 0.06392830610275269, "learning_rate": 3.7600742435910045e-05, "loss": 0.0053, "step": 30040 }, { "grad_norm": 0.1126771792769432, "learning_rate": 3.756870871499696e-05, "loss": 0.0089, "step": 30050 }, { "grad_norm": 0.1868889480829239, "learning_rate": 3.7536680431953025e-05, "loss": 0.0092, "step": 30060 }, { "grad_norm": 0.15845850110054016, "learning_rate": 3.7504657600788485e-05, "loss": 0.0063, "step": 30070 }, { "grad_norm": 0.11748763173818588, "learning_rate": 3.747264023551125e-05, "loss": 0.0064, "step": 30080 }, { "grad_norm": 0.11419007182121277, "learning_rate": 3.744062835012678e-05, "loss": 0.0081, "step": 30090 }, { "grad_norm": 0.17741677165031433, "learning_rate": 3.740862195863814e-05, "loss": 0.0074, "step": 30100 }, { "grad_norm": 0.12907731533050537, "learning_rate": 3.737662107504606e-05, "loss": 0.0073, "step": 30110 }, { "grad_norm": 0.10978299379348755, "learning_rate": 3.734462571334879e-05, "loss": 0.0074, "step": 30120 }, { "grad_norm": 0.135283961892128, "learning_rate": 3.731263588754217e-05, "loss": 0.0067, "step": 30130 }, { "grad_norm": 0.17736156284809113, "learning_rate": 3.7280651611619665e-05, "loss": 0.0115, "step": 30140 }, { "grad_norm": 0.1678922027349472, "learning_rate": 3.724867289957229e-05, "loss": 0.0082, "step": 30150 }, { "grad_norm": 0.14171040058135986, "learning_rate": 3.7216699765388576e-05, "loss": 0.007, "step": 30160 }, { "grad_norm": 0.10894141346216202, "learning_rate": 3.71847322230547e-05, "loss": 0.0086, "step": 30170 }, { "grad_norm": 0.13622160255908966, "learning_rate": 3.7152770286554335e-05, "loss": 0.007, "step": 30180 }, { "grad_norm": 0.12662141025066376, "learning_rate": 3.712081396986872e-05, "loss": 0.0068, "step": 30190 }, { "grad_norm": 0.11002703011035919, "learning_rate": 3.708886328697663e-05, "loss": 0.0087, "step": 30200 }, { "grad_norm": 0.12086127698421478, "learning_rate": 3.7056918251854395e-05, "loss": 0.0081, "step": 30210 }, { "grad_norm": 0.15071018040180206, "learning_rate": 3.7024978878475846e-05, "loss": 0.0091, "step": 30220 }, { "grad_norm": 0.11592593044042587, "learning_rate": 3.699304518081233e-05, "loss": 0.0057, "step": 30230 }, { "grad_norm": 0.09958864748477936, "learning_rate": 3.696111717283277e-05, "loss": 0.0068, "step": 30240 }, { "grad_norm": 0.15545548498630524, "learning_rate": 3.692919486850355e-05, "loss": 0.0107, "step": 30250 }, { "grad_norm": 0.2327188104391098, "learning_rate": 3.689727828178854e-05, "loss": 0.0064, "step": 30260 }, { "grad_norm": 0.13522988557815552, "learning_rate": 3.6865367426649185e-05, "loss": 0.0075, "step": 30270 }, { "grad_norm": 0.16489216685295105, "learning_rate": 3.6833462317044346e-05, "loss": 0.009, "step": 30280 }, { "grad_norm": 0.13806888461112976, "learning_rate": 3.6801562966930406e-05, "loss": 0.0076, "step": 30290 }, { "grad_norm": 0.14771491289138794, "learning_rate": 3.676966939026123e-05, "loss": 0.0069, "step": 30300 }, { "grad_norm": 0.09687495231628418, "learning_rate": 3.6737781600988166e-05, "loss": 0.0061, "step": 30310 }, { "grad_norm": 0.13081125915050507, "learning_rate": 3.670589961305998e-05, "loss": 0.0081, "step": 30320 }, { "grad_norm": 0.10187942534685135, "learning_rate": 3.667402344042298e-05, "loss": 0.007, "step": 30330 }, { "grad_norm": 0.17208686470985413, "learning_rate": 3.664215309702086e-05, "loss": 0.008, "step": 30340 }, { "grad_norm": 0.1699693500995636, "learning_rate": 3.661028859679478e-05, "loss": 0.0071, "step": 30350 }, { "grad_norm": 0.1262141615152359, "learning_rate": 3.657842995368339e-05, "loss": 0.0075, "step": 30360 }, { "grad_norm": 0.12539905309677124, "learning_rate": 3.654657718162273e-05, "loss": 0.0081, "step": 30370 }, { "grad_norm": 0.17736218869686127, "learning_rate": 3.6514730294546275e-05, "loss": 0.0067, "step": 30380 }, { "grad_norm": 0.2241005003452301, "learning_rate": 3.648288930638492e-05, "loss": 0.0079, "step": 30390 }, { "grad_norm": 0.11748973280191422, "learning_rate": 3.645105423106704e-05, "loss": 0.0058, "step": 30400 }, { "grad_norm": 0.11837531626224518, "learning_rate": 3.6419225082518336e-05, "loss": 0.0071, "step": 30410 }, { "grad_norm": 0.14361822605133057, "learning_rate": 3.6387401874661975e-05, "loss": 0.0069, "step": 30420 }, { "grad_norm": 0.12440507113933563, "learning_rate": 3.6355584621418516e-05, "loss": 0.0043, "step": 30430 }, { "grad_norm": 0.06485583633184433, "learning_rate": 3.632377333670591e-05, "loss": 0.0052, "step": 30440 }, { "grad_norm": 0.20142780244350433, "learning_rate": 3.629196803443946e-05, "loss": 0.0068, "step": 30450 }, { "grad_norm": 0.1129593625664711, "learning_rate": 3.626016872853193e-05, "loss": 0.0058, "step": 30460 }, { "grad_norm": 0.15260049700737, "learning_rate": 3.6228375432893405e-05, "loss": 0.0065, "step": 30470 }, { "grad_norm": 0.10342555493116379, "learning_rate": 3.619658816143131e-05, "loss": 0.0056, "step": 30480 }, { "grad_norm": 0.1554076224565506, "learning_rate": 3.6164806928050544e-05, "loss": 0.0068, "step": 30490 }, { "grad_norm": 0.13829346001148224, "learning_rate": 3.613303174665327e-05, "loss": 0.007, "step": 30500 }, { "grad_norm": 0.13928543031215668, "learning_rate": 3.6101262631138996e-05, "loss": 0.0073, "step": 30510 }, { "grad_norm": 0.13167282938957214, "learning_rate": 3.606949959540467e-05, "loss": 0.007, "step": 30520 }, { "grad_norm": 0.14291056990623474, "learning_rate": 3.603774265334449e-05, "loss": 0.008, "step": 30530 }, { "grad_norm": 0.12085936218500137, "learning_rate": 3.600599181885003e-05, "loss": 0.0063, "step": 30540 }, { "grad_norm": 0.12151236832141876, "learning_rate": 3.597424710581018e-05, "loss": 0.0061, "step": 30550 }, { "grad_norm": 0.15447546541690826, "learning_rate": 3.5942508528111164e-05, "loss": 0.0083, "step": 30560 }, { "grad_norm": 0.17401531338691711, "learning_rate": 3.591077609963651e-05, "loss": 0.0098, "step": 30570 }, { "grad_norm": 0.16741633415222168, "learning_rate": 3.587904983426704e-05, "loss": 0.0092, "step": 30580 }, { "grad_norm": 0.13310733437538147, "learning_rate": 3.584732974588094e-05, "loss": 0.006, "step": 30590 }, { "grad_norm": 0.12579043209552765, "learning_rate": 3.5815615848353645e-05, "loss": 0.0065, "step": 30600 }, { "grad_norm": 0.10124915093183517, "learning_rate": 3.578390815555784e-05, "loss": 0.007, "step": 30610 }, { "grad_norm": 0.12385682761669159, "learning_rate": 3.5752206681363626e-05, "loss": 0.0052, "step": 30620 }, { "grad_norm": 0.1327909529209137, "learning_rate": 3.572051143963826e-05, "loss": 0.0078, "step": 30630 }, { "grad_norm": 0.11395905166864395, "learning_rate": 3.5688822444246294e-05, "loss": 0.0073, "step": 30640 }, { "grad_norm": 0.12066082656383514, "learning_rate": 3.565713970904963e-05, "loss": 0.006, "step": 30650 }, { "grad_norm": 0.17035065591335297, "learning_rate": 3.562546324790733e-05, "loss": 0.0078, "step": 30660 }, { "grad_norm": 0.10199645161628723, "learning_rate": 3.559379307467577e-05, "loss": 0.0052, "step": 30670 }, { "grad_norm": 0.1525876373052597, "learning_rate": 3.556212920320856e-05, "loss": 0.0064, "step": 30680 }, { "grad_norm": 0.13744014501571655, "learning_rate": 3.553047164735655e-05, "loss": 0.0071, "step": 30690 }, { "grad_norm": 0.13586294651031494, "learning_rate": 3.549882042096783e-05, "loss": 0.0066, "step": 30700 }, { "grad_norm": 0.1096530333161354, "learning_rate": 3.546717553788772e-05, "loss": 0.0071, "step": 30710 }, { "grad_norm": 0.12354957312345505, "learning_rate": 3.543553701195879e-05, "loss": 0.0073, "step": 30720 }, { "grad_norm": 0.14587031304836273, "learning_rate": 3.540390485702077e-05, "loss": 0.0082, "step": 30730 }, { "grad_norm": 0.14244696497917175, "learning_rate": 3.5372279086910645e-05, "loss": 0.0072, "step": 30740 }, { "grad_norm": 0.08880648016929626, "learning_rate": 3.5340659715462635e-05, "loss": 0.0051, "step": 30750 }, { "grad_norm": 0.11416425555944443, "learning_rate": 3.530904675650808e-05, "loss": 0.0069, "step": 30760 }, { "grad_norm": 0.13313227891921997, "learning_rate": 3.527744022387559e-05, "loss": 0.0074, "step": 30770 }, { "grad_norm": 0.1288822889328003, "learning_rate": 3.524584013139094e-05, "loss": 0.0081, "step": 30780 }, { "grad_norm": 0.15827634930610657, "learning_rate": 3.521424649287708e-05, "loss": 0.0052, "step": 30790 }, { "grad_norm": 0.12836244702339172, "learning_rate": 3.518265932215412e-05, "loss": 0.0083, "step": 30800 }, { "grad_norm": 0.132590651512146, "learning_rate": 3.51510786330394e-05, "loss": 0.0078, "step": 30810 }, { "grad_norm": 0.145486980676651, "learning_rate": 3.511950443934736e-05, "loss": 0.0075, "step": 30820 }, { "grad_norm": 0.13492612540721893, "learning_rate": 3.508793675488961e-05, "loss": 0.0082, "step": 30830 }, { "grad_norm": 0.14500509202480316, "learning_rate": 3.505637559347498e-05, "loss": 0.0055, "step": 30840 }, { "grad_norm": 0.16961878538131714, "learning_rate": 3.5024820968909345e-05, "loss": 0.0067, "step": 30850 }, { "grad_norm": 0.16823184490203857, "learning_rate": 3.4993272894995774e-05, "loss": 0.0076, "step": 30860 }, { "grad_norm": 0.1617727279663086, "learning_rate": 3.49617313855345e-05, "loss": 0.01, "step": 30870 }, { "grad_norm": 0.1401500552892685, "learning_rate": 3.4930196454322834e-05, "loss": 0.008, "step": 30880 }, { "grad_norm": 0.18811975419521332, "learning_rate": 3.48986681151552e-05, "loss": 0.0059, "step": 30890 }, { "grad_norm": 0.12619557976722717, "learning_rate": 3.4867146381823204e-05, "loss": 0.0068, "step": 30900 }, { "grad_norm": 0.1930829882621765, "learning_rate": 3.48356312681155e-05, "loss": 0.0067, "step": 30910 }, { "grad_norm": 0.10778233408927917, "learning_rate": 3.480412278781788e-05, "loss": 0.0069, "step": 30920 }, { "grad_norm": 0.10559656471014023, "learning_rate": 3.47726209547132e-05, "loss": 0.0066, "step": 30930 }, { "grad_norm": 0.12656421959400177, "learning_rate": 3.474112578258147e-05, "loss": 0.0055, "step": 30940 }, { "grad_norm": 0.12639710307121277, "learning_rate": 3.470963728519974e-05, "loss": 0.0061, "step": 30950 }, { "grad_norm": 0.1183728501200676, "learning_rate": 3.467815547634211e-05, "loss": 0.0072, "step": 30960 }, { "grad_norm": 0.08263039588928223, "learning_rate": 3.4646680369779836e-05, "loss": 0.0058, "step": 30970 }, { "grad_norm": 0.0837506353855133, "learning_rate": 3.46152119792812e-05, "loss": 0.0064, "step": 30980 }, { "grad_norm": 0.11186615377664566, "learning_rate": 3.45837503186115e-05, "loss": 0.0082, "step": 30990 }, { "grad_norm": 0.15561474859714508, "learning_rate": 3.455229540153318e-05, "loss": 0.006, "step": 31000 }, { "grad_norm": 0.14175774157047272, "learning_rate": 3.4520847241805665e-05, "loss": 0.0066, "step": 31010 }, { "grad_norm": 0.17799082398414612, "learning_rate": 3.4489405853185465e-05, "loss": 0.0069, "step": 31020 }, { "grad_norm": 0.10059880465269089, "learning_rate": 3.44579712494261e-05, "loss": 0.0055, "step": 31030 }, { "grad_norm": 0.1584777981042862, "learning_rate": 3.442654344427814e-05, "loss": 0.009, "step": 31040 }, { "grad_norm": 0.10719703137874603, "learning_rate": 3.439512245148916e-05, "loss": 0.0062, "step": 31050 }, { "grad_norm": 0.12123803049325943, "learning_rate": 3.4363708284803796e-05, "loss": 0.0061, "step": 31060 }, { "grad_norm": 0.10521633923053741, "learning_rate": 3.433230095796367e-05, "loss": 0.0064, "step": 31070 }, { "grad_norm": 0.12627023458480835, "learning_rate": 3.430090048470739e-05, "loss": 0.0081, "step": 31080 }, { "grad_norm": 0.14873476326465607, "learning_rate": 3.4269506878770585e-05, "loss": 0.0059, "step": 31090 }, { "grad_norm": 0.16274100542068481, "learning_rate": 3.4238120153885925e-05, "loss": 0.0076, "step": 31100 }, { "grad_norm": 0.14787952601909637, "learning_rate": 3.4206740323782994e-05, "loss": 0.0063, "step": 31110 }, { "grad_norm": 0.12443414330482483, "learning_rate": 3.4175367402188395e-05, "loss": 0.007, "step": 31120 }, { "grad_norm": 0.11954522877931595, "learning_rate": 3.414400140282574e-05, "loss": 0.0055, "step": 31130 }, { "grad_norm": 0.11595787107944489, "learning_rate": 3.411264233941556e-05, "loss": 0.0067, "step": 31140 }, { "grad_norm": 0.16976982355117798, "learning_rate": 3.408129022567537e-05, "loss": 0.0065, "step": 31150 }, { "grad_norm": 0.16359928250312805, "learning_rate": 3.404994507531968e-05, "loss": 0.008, "step": 31160 }, { "grad_norm": 0.133492574095726, "learning_rate": 3.40186069020599e-05, "loss": 0.0068, "step": 31170 }, { "grad_norm": 0.14365646243095398, "learning_rate": 3.3987275719604386e-05, "loss": 0.0068, "step": 31180 }, { "grad_norm": 0.11774184554815292, "learning_rate": 3.3955951541658526e-05, "loss": 0.0056, "step": 31190 }, { "grad_norm": 0.18561990559101105, "learning_rate": 3.392463438192454e-05, "loss": 0.0062, "step": 31200 }, { "grad_norm": 0.09722460806369781, "learning_rate": 3.38933242541016e-05, "loss": 0.007, "step": 31210 }, { "grad_norm": 0.14564825594425201, "learning_rate": 3.386202117188586e-05, "loss": 0.0084, "step": 31220 }, { "grad_norm": 0.12994705140590668, "learning_rate": 3.383072514897033e-05, "loss": 0.0074, "step": 31230 }, { "grad_norm": 0.10265808552503586, "learning_rate": 3.379943619904494e-05, "loss": 0.0098, "step": 31240 }, { "grad_norm": 0.1115187406539917, "learning_rate": 3.3768154335796583e-05, "loss": 0.0071, "step": 31250 }, { "grad_norm": 0.08994050323963165, "learning_rate": 3.3736879572908963e-05, "loss": 0.0052, "step": 31260 }, { "grad_norm": 0.11245670914649963, "learning_rate": 3.370561192406276e-05, "loss": 0.0066, "step": 31270 }, { "grad_norm": 0.10497291386127472, "learning_rate": 3.367435140293547e-05, "loss": 0.0087, "step": 31280 }, { "grad_norm": 0.15829382836818695, "learning_rate": 3.3643098023201545e-05, "loss": 0.0083, "step": 31290 }, { "grad_norm": 0.14849233627319336, "learning_rate": 3.361185179853225e-05, "loss": 0.0089, "step": 31300 }, { "grad_norm": 0.07903946936130524, "learning_rate": 3.358061274259574e-05, "loss": 0.0056, "step": 31310 }, { "grad_norm": 0.09798122942447662, "learning_rate": 3.3549380869057064e-05, "loss": 0.0059, "step": 31320 }, { "grad_norm": 0.08702096343040466, "learning_rate": 3.351815619157809e-05, "loss": 0.0065, "step": 31330 }, { "grad_norm": 0.13587549328804016, "learning_rate": 3.348693872381754e-05, "loss": 0.0072, "step": 31340 }, { "grad_norm": 0.151299387216568, "learning_rate": 3.3455728479431025e-05, "loss": 0.0078, "step": 31350 }, { "grad_norm": 0.15531370043754578, "learning_rate": 3.342452547207094e-05, "loss": 0.0082, "step": 31360 }, { "grad_norm": 0.107806496322155, "learning_rate": 3.3393329715386555e-05, "loss": 0.0078, "step": 31370 }, { "grad_norm": 0.1411067396402359, "learning_rate": 3.336214122302395e-05, "loss": 0.0072, "step": 31380 }, { "grad_norm": 0.14234529435634613, "learning_rate": 3.333096000862604e-05, "loss": 0.0101, "step": 31390 }, { "grad_norm": 0.094650037586689, "learning_rate": 3.3299786085832516e-05, "loss": 0.0067, "step": 31400 }, { "grad_norm": 0.13091914355754852, "learning_rate": 3.3268619468279956e-05, "loss": 0.0082, "step": 31410 }, { "grad_norm": 0.1327870488166809, "learning_rate": 3.323746016960167e-05, "loss": 0.0092, "step": 31420 }, { "grad_norm": 0.11131004244089127, "learning_rate": 3.3206308203427806e-05, "loss": 0.0067, "step": 31430 }, { "grad_norm": 0.11723223328590393, "learning_rate": 3.3175163583385254e-05, "loss": 0.0072, "step": 31440 }, { "grad_norm": 0.15537326037883759, "learning_rate": 3.314402632309779e-05, "loss": 0.0062, "step": 31450 }, { "grad_norm": 0.14631186425685883, "learning_rate": 3.311289643618587e-05, "loss": 0.0082, "step": 31460 }, { "grad_norm": 0.11024956405162811, "learning_rate": 3.308177393626676e-05, "loss": 0.0061, "step": 31470 }, { "grad_norm": 0.10818497091531754, "learning_rate": 3.305065883695453e-05, "loss": 0.006, "step": 31480 }, { "grad_norm": 0.12888634204864502, "learning_rate": 3.301955115185996e-05, "loss": 0.007, "step": 31490 }, { "grad_norm": 0.15411272644996643, "learning_rate": 3.29884508945906e-05, "loss": 0.0071, "step": 31500 }, { "grad_norm": 0.0972270667552948, "learning_rate": 3.295735807875079e-05, "loss": 0.0064, "step": 31510 }, { "grad_norm": 0.09508929401636124, "learning_rate": 3.292627271794156e-05, "loss": 0.0063, "step": 31520 }, { "grad_norm": 0.16375724971294403, "learning_rate": 3.289519482576069e-05, "loss": 0.0057, "step": 31530 }, { "grad_norm": 0.1046200543642044, "learning_rate": 3.286412441580274e-05, "loss": 0.0091, "step": 31540 }, { "grad_norm": 0.10605283826589584, "learning_rate": 3.283306150165896e-05, "loss": 0.0096, "step": 31550 }, { "grad_norm": 0.11424367874860764, "learning_rate": 3.2802006096917285e-05, "loss": 0.0055, "step": 31560 }, { "grad_norm": 0.11554358154535294, "learning_rate": 3.277095821516245e-05, "loss": 0.005, "step": 31570 }, { "grad_norm": 0.16191533207893372, "learning_rate": 3.273991786997583e-05, "loss": 0.0049, "step": 31580 }, { "grad_norm": 0.15623855590820312, "learning_rate": 3.2708885074935514e-05, "loss": 0.0062, "step": 31590 }, { "grad_norm": 0.18150894343852997, "learning_rate": 3.267785984361633e-05, "loss": 0.0066, "step": 31600 }, { "grad_norm": 0.1167236715555191, "learning_rate": 3.2646842189589754e-05, "loss": 0.0061, "step": 31610 }, { "grad_norm": 0.16296522319316864, "learning_rate": 3.261583212642397e-05, "loss": 0.0092, "step": 31620 }, { "grad_norm": 0.12939636409282684, "learning_rate": 3.25848296676838e-05, "loss": 0.0043, "step": 31630 }, { "grad_norm": 0.11428613215684891, "learning_rate": 3.255383482693083e-05, "loss": 0.0084, "step": 31640 }, { "grad_norm": 0.14518706500530243, "learning_rate": 3.252284761772322e-05, "loss": 0.007, "step": 31650 }, { "grad_norm": 0.1425611525774002, "learning_rate": 3.249186805361581e-05, "loss": 0.0082, "step": 31660 }, { "grad_norm": 0.14274564385414124, "learning_rate": 3.246089614816016e-05, "loss": 0.0078, "step": 31670 }, { "grad_norm": 0.09234541654586792, "learning_rate": 3.242993191490441e-05, "loss": 0.0067, "step": 31680 }, { "grad_norm": 0.07793637365102768, "learning_rate": 3.239897536739335e-05, "loss": 0.0054, "step": 31690 }, { "grad_norm": 0.12408098578453064, "learning_rate": 3.236802651916845e-05, "loss": 0.0072, "step": 31700 }, { "grad_norm": 0.14928951859474182, "learning_rate": 3.23370853837678e-05, "loss": 0.0061, "step": 31710 }, { "grad_norm": 0.17733943462371826, "learning_rate": 3.2306151974726056e-05, "loss": 0.0059, "step": 31720 }, { "grad_norm": 0.15351338684558868, "learning_rate": 3.227522630557459e-05, "loss": 0.0067, "step": 31730 }, { "grad_norm": 0.11399411410093307, "learning_rate": 3.224430838984132e-05, "loss": 0.0065, "step": 31740 }, { "grad_norm": 0.14089038968086243, "learning_rate": 3.2213398241050784e-05, "loss": 0.0077, "step": 31750 }, { "grad_norm": 0.14425501227378845, "learning_rate": 3.218249587272415e-05, "loss": 0.0064, "step": 31760 }, { "grad_norm": 0.19323429465293884, "learning_rate": 3.215160129837916e-05, "loss": 0.0074, "step": 31770 }, { "grad_norm": 0.12929467856884003, "learning_rate": 3.2120714531530146e-05, "loss": 0.0063, "step": 31780 }, { "grad_norm": 0.1546734869480133, "learning_rate": 3.2089835585688014e-05, "loss": 0.0071, "step": 31790 }, { "grad_norm": 0.19255970418453217, "learning_rate": 3.20589644743603e-05, "loss": 0.0077, "step": 31800 }, { "grad_norm": 0.07743442058563232, "learning_rate": 3.2028101211051054e-05, "loss": 0.0069, "step": 31810 }, { "grad_norm": 0.11569955945014954, "learning_rate": 3.19972458092609e-05, "loss": 0.0066, "step": 31820 }, { "grad_norm": 0.08392636477947235, "learning_rate": 3.196639828248707e-05, "loss": 0.0072, "step": 31830 }, { "grad_norm": 0.12035921216011047, "learning_rate": 3.193555864422329e-05, "loss": 0.0063, "step": 31840 }, { "grad_norm": 0.14068906009197235, "learning_rate": 3.19047269079599e-05, "loss": 0.0065, "step": 31850 }, { "grad_norm": 0.15814028680324554, "learning_rate": 3.1873903087183734e-05, "loss": 0.0063, "step": 31860 }, { "grad_norm": 0.1402445286512375, "learning_rate": 3.1843087195378176e-05, "loss": 0.006, "step": 31870 }, { "grad_norm": 0.12465289235115051, "learning_rate": 3.181227924602312e-05, "loss": 0.0058, "step": 31880 }, { "grad_norm": 0.08814354985952377, "learning_rate": 3.178147925259506e-05, "loss": 0.005, "step": 31890 }, { "grad_norm": 0.1717561036348343, "learning_rate": 3.175068722856693e-05, "loss": 0.0056, "step": 31900 }, { "grad_norm": 0.10810481011867523, "learning_rate": 3.17199031874082e-05, "loss": 0.007, "step": 31910 }, { "grad_norm": 0.12861844897270203, "learning_rate": 3.168912714258487e-05, "loss": 0.0064, "step": 31920 }, { "grad_norm": 0.11037667840719223, "learning_rate": 3.165835910755944e-05, "loss": 0.0057, "step": 31930 }, { "grad_norm": 0.16607710719108582, "learning_rate": 3.1627599095790874e-05, "loss": 0.0091, "step": 31940 }, { "grad_norm": 0.2064642608165741, "learning_rate": 3.159684712073462e-05, "loss": 0.0066, "step": 31950 }, { "grad_norm": 0.16052064299583435, "learning_rate": 3.1566103195842695e-05, "loss": 0.0099, "step": 31960 }, { "grad_norm": 0.12918981909751892, "learning_rate": 3.153536733456349e-05, "loss": 0.0057, "step": 31970 }, { "grad_norm": 0.11138524115085602, "learning_rate": 3.150463955034194e-05, "loss": 0.0081, "step": 31980 }, { "grad_norm": 0.1307319849729538, "learning_rate": 3.147391985661942e-05, "loss": 0.0057, "step": 31990 }, { "grad_norm": 0.14433376491069794, "learning_rate": 3.1443208266833755e-05, "loss": 0.006, "step": 32000 }, { "grad_norm": 0.11909041553735733, "learning_rate": 3.141250479441924e-05, "loss": 0.0078, "step": 32010 }, { "grad_norm": 0.16900333762168884, "learning_rate": 3.138180945280664e-05, "loss": 0.006, "step": 32020 }, { "grad_norm": 0.1575348824262619, "learning_rate": 3.1351122255423116e-05, "loss": 0.0056, "step": 32030 }, { "grad_norm": 0.0931181088089943, "learning_rate": 3.132044321569228e-05, "loss": 0.0056, "step": 32040 }, { "grad_norm": 0.13226689398288727, "learning_rate": 3.1289772347034226e-05, "loss": 0.0054, "step": 32050 }, { "grad_norm": 0.12317599356174469, "learning_rate": 3.1259109662865404e-05, "loss": 0.0069, "step": 32060 }, { "grad_norm": 0.12255436182022095, "learning_rate": 3.122845517659871e-05, "loss": 0.0049, "step": 32070 }, { "grad_norm": 0.13480713963508606, "learning_rate": 3.119780890164348e-05, "loss": 0.0068, "step": 32080 }, { "grad_norm": 0.12956906855106354, "learning_rate": 3.116717085140542e-05, "loss": 0.0067, "step": 32090 }, { "grad_norm": 0.12346787005662918, "learning_rate": 3.1136541039286657e-05, "loss": 0.0048, "step": 32100 }, { "grad_norm": 0.08059419691562653, "learning_rate": 3.110591947868571e-05, "loss": 0.0054, "step": 32110 }, { "grad_norm": 0.15764877200126648, "learning_rate": 3.107530618299751e-05, "loss": 0.0064, "step": 32120 }, { "grad_norm": 0.15475280582904816, "learning_rate": 3.104470116561333e-05, "loss": 0.0063, "step": 32130 }, { "grad_norm": 0.07670533657073975, "learning_rate": 3.1014104439920835e-05, "loss": 0.0066, "step": 32140 }, { "grad_norm": 0.11150666326284409, "learning_rate": 3.098351601930411e-05, "loss": 0.0079, "step": 32150 }, { "grad_norm": 0.12214116007089615, "learning_rate": 3.0952935917143533e-05, "loss": 0.0055, "step": 32160 }, { "grad_norm": 0.20226360857486725, "learning_rate": 3.092236414681588e-05, "loss": 0.0062, "step": 32170 }, { "grad_norm": 0.16545143723487854, "learning_rate": 3.089180072169431e-05, "loss": 0.0065, "step": 32180 }, { "grad_norm": 0.10139782726764679, "learning_rate": 3.08612456551483e-05, "loss": 0.0055, "step": 32190 }, { "grad_norm": 0.08709494769573212, "learning_rate": 3.083069896054363e-05, "loss": 0.0051, "step": 32200 }, { "grad_norm": 0.12813082337379456, "learning_rate": 3.080016065124253e-05, "loss": 0.0101, "step": 32210 }, { "grad_norm": 0.12832175195217133, "learning_rate": 3.076963074060344e-05, "loss": 0.0062, "step": 32220 }, { "grad_norm": 0.15618284046649933, "learning_rate": 3.07391092419812e-05, "loss": 0.0068, "step": 32230 }, { "grad_norm": 0.1578444391489029, "learning_rate": 3.070859616872695e-05, "loss": 0.0069, "step": 32240 }, { "grad_norm": 0.13999870419502258, "learning_rate": 3.067809153418816e-05, "loss": 0.0051, "step": 32250 }, { "grad_norm": 0.11180334538221359, "learning_rate": 3.0647595351708545e-05, "loss": 0.0085, "step": 32260 }, { "grad_norm": 0.10190362483263016, "learning_rate": 3.061710763462824e-05, "loss": 0.0061, "step": 32270 }, { "grad_norm": 0.20683345198631287, "learning_rate": 3.0586628396283564e-05, "loss": 0.0071, "step": 32280 }, { "grad_norm": 0.17284297943115234, "learning_rate": 3.0556157650007174e-05, "loss": 0.0071, "step": 32290 }, { "grad_norm": 0.15055955946445465, "learning_rate": 3.052569540912801e-05, "loss": 0.0065, "step": 32300 }, { "grad_norm": 0.13830052316188812, "learning_rate": 3.049524168697131e-05, "loss": 0.0054, "step": 32310 }, { "grad_norm": 0.09233562648296356, "learning_rate": 3.0464796496858538e-05, "loss": 0.0044, "step": 32320 }, { "grad_norm": 0.13038749992847443, "learning_rate": 3.0434359852107473e-05, "loss": 0.0061, "step": 32330 }, { "grad_norm": 0.09640044718980789, "learning_rate": 3.0403931766032135e-05, "loss": 0.0058, "step": 32340 }, { "grad_norm": 0.10686345398426056, "learning_rate": 3.0373512251942814e-05, "loss": 0.006, "step": 32350 }, { "grad_norm": 0.12459529936313629, "learning_rate": 3.0343101323146007e-05, "loss": 0.0068, "step": 32360 }, { "grad_norm": 0.1090453714132309, "learning_rate": 3.0312698992944523e-05, "loss": 0.0059, "step": 32370 }, { "grad_norm": 0.10343921929597855, "learning_rate": 3.028230527463736e-05, "loss": 0.0053, "step": 32380 }, { "grad_norm": 0.16327764093875885, "learning_rate": 3.0251920181519746e-05, "loss": 0.0086, "step": 32390 }, { "grad_norm": 0.18935753405094147, "learning_rate": 3.0221543726883183e-05, "loss": 0.0071, "step": 32400 }, { "grad_norm": 0.1746787577867508, "learning_rate": 3.019117592401535e-05, "loss": 0.0098, "step": 32410 }, { "grad_norm": 0.191737100481987, "learning_rate": 3.0160816786200137e-05, "loss": 0.007, "step": 32420 }, { "grad_norm": 0.15824460983276367, "learning_rate": 3.01304663267177e-05, "loss": 0.0056, "step": 32430 }, { "grad_norm": 0.0971779003739357, "learning_rate": 3.0100124558844338e-05, "loss": 0.0055, "step": 32440 }, { "grad_norm": 0.09041763842105865, "learning_rate": 3.0069791495852562e-05, "loss": 0.0063, "step": 32450 }, { "grad_norm": 0.09926420450210571, "learning_rate": 3.0039467151011113e-05, "loss": 0.0065, "step": 32460 }, { "grad_norm": 0.1770155429840088, "learning_rate": 3.000915153758487e-05, "loss": 0.0072, "step": 32470 }, { "grad_norm": 0.1190819963812828, "learning_rate": 2.9978844668834917e-05, "loss": 0.007, "step": 32480 }, { "grad_norm": 0.13815467059612274, "learning_rate": 2.994854655801849e-05, "loss": 0.006, "step": 32490 }, { "grad_norm": 0.11126867681741714, "learning_rate": 2.991825721838905e-05, "loss": 0.0051, "step": 32500 }, { "grad_norm": 0.11018216609954834, "learning_rate": 2.9887976663196166e-05, "loss": 0.0058, "step": 32510 }, { "grad_norm": 0.1848047524690628, "learning_rate": 2.985770490568556e-05, "loss": 0.0072, "step": 32520 }, { "grad_norm": 0.13350097835063934, "learning_rate": 2.9827441959099157e-05, "loss": 0.006, "step": 32530 }, { "grad_norm": 0.17179839313030243, "learning_rate": 2.979718783667499e-05, "loss": 0.0108, "step": 32540 }, { "grad_norm": 0.17066046595573425, "learning_rate": 2.9766942551647215e-05, "loss": 0.0059, "step": 32550 }, { "grad_norm": 0.14682257175445557, "learning_rate": 2.9736706117246188e-05, "loss": 0.0054, "step": 32560 }, { "grad_norm": 0.1516590416431427, "learning_rate": 2.9706478546698324e-05, "loss": 0.0061, "step": 32570 }, { "grad_norm": 0.14844442903995514, "learning_rate": 2.9676259853226195e-05, "loss": 0.0078, "step": 32580 }, { "grad_norm": 0.10091531276702881, "learning_rate": 2.964605005004849e-05, "loss": 0.0085, "step": 32590 }, { "grad_norm": 0.09751972556114197, "learning_rate": 2.9615849150380016e-05, "loss": 0.0099, "step": 32600 }, { "grad_norm": 0.1268218755722046, "learning_rate": 2.9585657167431637e-05, "loss": 0.005, "step": 32610 }, { "grad_norm": 0.13251028954982758, "learning_rate": 2.9555474114410388e-05, "loss": 0.0046, "step": 32620 }, { "grad_norm": 0.1009114682674408, "learning_rate": 2.9525300004519348e-05, "loss": 0.0072, "step": 32630 }, { "grad_norm": 0.10464303195476532, "learning_rate": 2.9495134850957696e-05, "loss": 0.0052, "step": 32640 }, { "grad_norm": 0.18210339546203613, "learning_rate": 2.9464978666920674e-05, "loss": 0.0047, "step": 32650 }, { "grad_norm": 0.15561768412590027, "learning_rate": 2.9434831465599662e-05, "loss": 0.0071, "step": 32660 }, { "grad_norm": 0.09953135997056961, "learning_rate": 2.9404693260182047e-05, "loss": 0.0058, "step": 32670 }, { "grad_norm": 0.1312778741121292, "learning_rate": 2.9374564063851295e-05, "loss": 0.0086, "step": 32680 }, { "grad_norm": 0.11798211932182312, "learning_rate": 2.9344443889786967e-05, "loss": 0.0063, "step": 32690 }, { "grad_norm": 0.16421104967594147, "learning_rate": 2.931433275116463e-05, "loss": 0.0067, "step": 32700 }, { "grad_norm": 0.08577638119459152, "learning_rate": 2.928423066115592e-05, "loss": 0.0048, "step": 32710 }, { "grad_norm": 0.12097060680389404, "learning_rate": 2.9254137632928537e-05, "loss": 0.0053, "step": 32720 }, { "grad_norm": 0.1263439953327179, "learning_rate": 2.9224053679646167e-05, "loss": 0.0075, "step": 32730 }, { "grad_norm": 0.09657727926969528, "learning_rate": 2.9193978814468547e-05, "loss": 0.0041, "step": 32740 }, { "grad_norm": 0.12645573914051056, "learning_rate": 2.9163913050551485e-05, "loss": 0.0063, "step": 32750 }, { "grad_norm": 0.12060090154409409, "learning_rate": 2.9133856401046735e-05, "loss": 0.0054, "step": 32760 }, { "grad_norm": 0.11130496114492416, "learning_rate": 2.910380887910209e-05, "loss": 0.007, "step": 32770 }, { "grad_norm": 0.1454237699508667, "learning_rate": 2.9073770497861386e-05, "loss": 0.0061, "step": 32780 }, { "grad_norm": 0.17473816871643066, "learning_rate": 2.904374127046442e-05, "loss": 0.0052, "step": 32790 }, { "grad_norm": 0.13412576913833618, "learning_rate": 2.9013721210046973e-05, "loss": 0.0058, "step": 32800 }, { "grad_norm": 0.1486349105834961, "learning_rate": 2.8983710329740882e-05, "loss": 0.0051, "step": 32810 }, { "grad_norm": 0.09377927333116531, "learning_rate": 2.8953708642673898e-05, "loss": 0.0046, "step": 32820 }, { "grad_norm": 0.15354914963245392, "learning_rate": 2.8923716161969782e-05, "loss": 0.0088, "step": 32830 }, { "grad_norm": 0.12007521837949753, "learning_rate": 2.8893732900748248e-05, "loss": 0.0049, "step": 32840 }, { "grad_norm": 0.09192389249801636, "learning_rate": 2.8863758872125036e-05, "loss": 0.0058, "step": 32850 }, { "grad_norm": 0.16077649593353271, "learning_rate": 2.883379408921175e-05, "loss": 0.0052, "step": 32860 }, { "grad_norm": 0.07937920838594437, "learning_rate": 2.880383856511606e-05, "loss": 0.0059, "step": 32870 }, { "grad_norm": 0.10558643192052841, "learning_rate": 2.8773892312941476e-05, "loss": 0.0074, "step": 32880 }, { "grad_norm": 0.08743251860141754, "learning_rate": 2.8743955345787555e-05, "loss": 0.0053, "step": 32890 }, { "grad_norm": 0.13768801093101501, "learning_rate": 2.87140276767497e-05, "loss": 0.0067, "step": 32900 }, { "grad_norm": 0.07842719554901123, "learning_rate": 2.868410931891934e-05, "loss": 0.004, "step": 32910 }, { "grad_norm": 0.14422743022441864, "learning_rate": 2.8654200285383748e-05, "loss": 0.0066, "step": 32920 }, { "grad_norm": 0.15881814062595367, "learning_rate": 2.8624300589226126e-05, "loss": 0.006, "step": 32930 }, { "grad_norm": 0.13258270919322968, "learning_rate": 2.859441024352568e-05, "loss": 0.0088, "step": 32940 }, { "grad_norm": 0.14985430240631104, "learning_rate": 2.8564529261357427e-05, "loss": 0.0072, "step": 32950 }, { "grad_norm": 0.14045162498950958, "learning_rate": 2.8534657655792307e-05, "loss": 0.0054, "step": 32960 }, { "grad_norm": 0.15355348587036133, "learning_rate": 2.8504795439897208e-05, "loss": 0.0051, "step": 32970 }, { "grad_norm": 0.13396872580051422, "learning_rate": 2.847494262673487e-05, "loss": 0.0058, "step": 32980 }, { "grad_norm": 0.12079613655805588, "learning_rate": 2.8445099229363925e-05, "loss": 0.0072, "step": 32990 }, { "grad_norm": 0.12581384181976318, "learning_rate": 2.841526526083886e-05, "loss": 0.0062, "step": 33000 }, { "grad_norm": 0.14064167439937592, "learning_rate": 2.838544073421011e-05, "loss": 0.0058, "step": 33010 }, { "grad_norm": 0.14616908133029938, "learning_rate": 2.8355625662523928e-05, "loss": 0.0062, "step": 33020 }, { "grad_norm": 0.11977989226579666, "learning_rate": 2.8325820058822407e-05, "loss": 0.0081, "step": 33030 }, { "grad_norm": 0.09075536578893661, "learning_rate": 2.829602393614357e-05, "loss": 0.0064, "step": 33040 }, { "grad_norm": 0.12440919876098633, "learning_rate": 2.8266237307521236e-05, "loss": 0.0074, "step": 33050 }, { "grad_norm": 0.07856209576129913, "learning_rate": 2.8236460185985074e-05, "loss": 0.0065, "step": 33060 }, { "grad_norm": 0.127189502120018, "learning_rate": 2.8206692584560646e-05, "loss": 0.0063, "step": 33070 }, { "grad_norm": 0.11204193532466888, "learning_rate": 2.8176934516269283e-05, "loss": 0.0079, "step": 33080 }, { "grad_norm": 0.11640162765979767, "learning_rate": 2.8147185994128167e-05, "loss": 0.0085, "step": 33090 }, { "grad_norm": 0.11792486906051636, "learning_rate": 2.8117447031150346e-05, "loss": 0.007, "step": 33100 }, { "grad_norm": 0.08733652532100677, "learning_rate": 2.8087717640344618e-05, "loss": 0.0061, "step": 33110 }, { "grad_norm": 0.10270404815673828, "learning_rate": 2.805799783471566e-05, "loss": 0.0059, "step": 33120 }, { "grad_norm": 0.10553299635648727, "learning_rate": 2.8028287627263895e-05, "loss": 0.0057, "step": 33130 }, { "grad_norm": 0.168716162443161, "learning_rate": 2.7998587030985608e-05, "loss": 0.0093, "step": 33140 }, { "grad_norm": 0.11237820237874985, "learning_rate": 2.7968896058872806e-05, "loss": 0.0068, "step": 33150 }, { "grad_norm": 0.09409064054489136, "learning_rate": 2.7939214723913388e-05, "loss": 0.0065, "step": 33160 }, { "grad_norm": 0.12603703141212463, "learning_rate": 2.7909543039090935e-05, "loss": 0.0083, "step": 33170 }, { "grad_norm": 0.13568492233753204, "learning_rate": 2.787988101738487e-05, "loss": 0.0073, "step": 33180 }, { "grad_norm": 0.14932911098003387, "learning_rate": 2.7850228671770334e-05, "loss": 0.0051, "step": 33190 }, { "grad_norm": 0.12421727180480957, "learning_rate": 2.782058601521832e-05, "loss": 0.0054, "step": 33200 }, { "grad_norm": 0.14322306215763092, "learning_rate": 2.77909530606955e-05, "loss": 0.0079, "step": 33210 }, { "grad_norm": 0.1453327089548111, "learning_rate": 2.7761329821164335e-05, "loss": 0.0061, "step": 33220 }, { "grad_norm": 0.14150173962116241, "learning_rate": 2.7731716309583062e-05, "loss": 0.0064, "step": 33230 }, { "grad_norm": 0.12096278369426727, "learning_rate": 2.7702112538905624e-05, "loss": 0.006, "step": 33240 }, { "grad_norm": 0.10163488239049911, "learning_rate": 2.7672518522081682e-05, "loss": 0.0061, "step": 33250 }, { "grad_norm": 0.1230865865945816, "learning_rate": 2.7642934272056726e-05, "loss": 0.0094, "step": 33260 }, { "grad_norm": 0.11817451566457748, "learning_rate": 2.7613359801771877e-05, "loss": 0.0068, "step": 33270 }, { "grad_norm": 0.09930537641048431, "learning_rate": 2.7583795124163996e-05, "loss": 0.0068, "step": 33280 }, { "grad_norm": 0.0990481749176979, "learning_rate": 2.755424025216572e-05, "loss": 0.0064, "step": 33290 }, { "grad_norm": 0.10883862525224686, "learning_rate": 2.7524695198705334e-05, "loss": 0.0054, "step": 33300 }, { "grad_norm": 0.174100860953331, "learning_rate": 2.749515997670683e-05, "loss": 0.0062, "step": 33310 }, { "grad_norm": 0.12302768230438232, "learning_rate": 2.746563459908995e-05, "loss": 0.0048, "step": 33320 }, { "grad_norm": 0.1712752878665924, "learning_rate": 2.7436119078770085e-05, "loss": 0.0099, "step": 33330 }, { "grad_norm": 0.1065274253487587, "learning_rate": 2.74066134286583e-05, "loss": 0.0063, "step": 33340 }, { "grad_norm": 0.10038530826568604, "learning_rate": 2.7377117661661412e-05, "loss": 0.0049, "step": 33350 }, { "grad_norm": 0.13690318167209625, "learning_rate": 2.734763179068183e-05, "loss": 0.0064, "step": 33360 }, { "grad_norm": 0.10249067097902298, "learning_rate": 2.7318155828617718e-05, "loss": 0.006, "step": 33370 }, { "grad_norm": 0.18274009227752686, "learning_rate": 2.7288689788362832e-05, "loss": 0.0064, "step": 33380 }, { "grad_norm": 0.13996925950050354, "learning_rate": 2.725923368280664e-05, "loss": 0.0065, "step": 33390 }, { "grad_norm": 0.13517865538597107, "learning_rate": 2.7229787524834237e-05, "loss": 0.0059, "step": 33400 }, { "grad_norm": 0.15508541464805603, "learning_rate": 2.7200351327326345e-05, "loss": 0.0068, "step": 33410 }, { "grad_norm": 0.13792890310287476, "learning_rate": 2.7170925103159407e-05, "loss": 0.0052, "step": 33420 }, { "grad_norm": 0.13161316514015198, "learning_rate": 2.714150886520542e-05, "loss": 0.006, "step": 33430 }, { "grad_norm": 0.17713890969753265, "learning_rate": 2.7112102626332038e-05, "loss": 0.0092, "step": 33440 }, { "grad_norm": 0.1105283796787262, "learning_rate": 2.7082706399402574e-05, "loss": 0.0054, "step": 33450 }, { "grad_norm": 0.13322852551937103, "learning_rate": 2.7053320197275934e-05, "loss": 0.0043, "step": 33460 }, { "grad_norm": 0.11120297759771347, "learning_rate": 2.7023944032806614e-05, "loss": 0.008, "step": 33470 }, { "grad_norm": 0.14554423093795776, "learning_rate": 2.699457791884478e-05, "loss": 0.0084, "step": 33480 }, { "grad_norm": 0.10419278591871262, "learning_rate": 2.6965221868236155e-05, "loss": 0.0057, "step": 33490 }, { "grad_norm": 0.08272016048431396, "learning_rate": 2.6935875893822077e-05, "loss": 0.005, "step": 33500 }, { "grad_norm": 0.15921299159526825, "learning_rate": 2.690654000843944e-05, "loss": 0.0087, "step": 33510 }, { "grad_norm": 0.0859428197145462, "learning_rate": 2.687721422492081e-05, "loss": 0.0047, "step": 33520 }, { "grad_norm": 0.12340313196182251, "learning_rate": 2.6847898556094252e-05, "loss": 0.0077, "step": 33530 }, { "grad_norm": 0.1370115876197815, "learning_rate": 2.6818593014783417e-05, "loss": 0.0053, "step": 33540 }, { "grad_norm": 0.10750164091587067, "learning_rate": 2.6789297613807585e-05, "loss": 0.0069, "step": 33550 }, { "grad_norm": 0.1302320659160614, "learning_rate": 2.676001236598154e-05, "loss": 0.0056, "step": 33560 }, { "grad_norm": 0.20198902487754822, "learning_rate": 2.6730737284115626e-05, "loss": 0.0058, "step": 33570 }, { "grad_norm": 0.14689449965953827, "learning_rate": 2.670147238101579e-05, "loss": 0.0071, "step": 33580 }, { "grad_norm": 0.1519101858139038, "learning_rate": 2.6672217669483473e-05, "loss": 0.0063, "step": 33590 }, { "grad_norm": 0.10151166468858719, "learning_rate": 2.6642973162315706e-05, "loss": 0.005, "step": 33600 }, { "grad_norm": 0.1071283370256424, "learning_rate": 2.6613738872304988e-05, "loss": 0.0073, "step": 33610 }, { "grad_norm": 0.10667894035577774, "learning_rate": 2.6584514812239437e-05, "loss": 0.0059, "step": 33620 }, { "grad_norm": 0.13162502646446228, "learning_rate": 2.655530099490261e-05, "loss": 0.0068, "step": 33630 }, { "grad_norm": 0.10990775376558304, "learning_rate": 2.6526097433073653e-05, "loss": 0.0055, "step": 33640 }, { "grad_norm": 0.12730854749679565, "learning_rate": 2.649690413952719e-05, "loss": 0.0071, "step": 33650 }, { "grad_norm": 0.09803076088428497, "learning_rate": 2.646772112703334e-05, "loss": 0.0053, "step": 33660 }, { "grad_norm": 0.10195334255695343, "learning_rate": 2.6438548408357768e-05, "loss": 0.0057, "step": 33670 }, { "grad_norm": 0.10923667252063751, "learning_rate": 2.6409385996261603e-05, "loss": 0.0049, "step": 33680 }, { "grad_norm": 0.16106006503105164, "learning_rate": 2.638023390350148e-05, "loss": 0.0077, "step": 33690 }, { "grad_norm": 0.11764510720968246, "learning_rate": 2.635109214282948e-05, "loss": 0.0046, "step": 33700 }, { "grad_norm": 0.10556923598051071, "learning_rate": 2.6321960726993254e-05, "loss": 0.0063, "step": 33710 }, { "grad_norm": 0.10245364159345627, "learning_rate": 2.6292839668735846e-05, "loss": 0.0057, "step": 33720 }, { "grad_norm": 0.12383749336004257, "learning_rate": 2.6263728980795778e-05, "loss": 0.0052, "step": 33730 }, { "grad_norm": 0.1051909476518631, "learning_rate": 2.6234628675907092e-05, "loss": 0.0043, "step": 33740 }, { "grad_norm": 0.0902271419763565, "learning_rate": 2.620553876679923e-05, "loss": 0.0055, "step": 33750 }, { "grad_norm": 0.05752139911055565, "learning_rate": 2.617645926619709e-05, "loss": 0.0051, "step": 33760 }, { "grad_norm": 0.12071316689252853, "learning_rate": 2.6147390186821064e-05, "loss": 0.0063, "step": 33770 }, { "grad_norm": 0.128805473446846, "learning_rate": 2.6118331541386955e-05, "loss": 0.0061, "step": 33780 }, { "grad_norm": 0.13908489048480988, "learning_rate": 2.6089283342605953e-05, "loss": 0.0062, "step": 33790 }, { "grad_norm": 0.11008013039827347, "learning_rate": 2.6060245603184797e-05, "loss": 0.0071, "step": 33800 }, { "grad_norm": 0.11250424385070801, "learning_rate": 2.6031218335825536e-05, "loss": 0.0061, "step": 33810 }, { "grad_norm": 0.12296626716852188, "learning_rate": 2.600220155322568e-05, "loss": 0.0044, "step": 33820 }, { "grad_norm": 0.16482701897621155, "learning_rate": 2.5973195268078186e-05, "loss": 0.0071, "step": 33830 }, { "grad_norm": 0.14870727062225342, "learning_rate": 2.5944199493071352e-05, "loss": 0.0053, "step": 33840 }, { "grad_norm": 0.12439948320388794, "learning_rate": 2.591521424088895e-05, "loss": 0.0064, "step": 33850 }, { "grad_norm": 0.08973588049411774, "learning_rate": 2.5886239524210086e-05, "loss": 0.0043, "step": 33860 }, { "grad_norm": 0.09481757134199142, "learning_rate": 2.5857275355709317e-05, "loss": 0.0049, "step": 33870 }, { "grad_norm": 0.1200161874294281, "learning_rate": 2.5828321748056533e-05, "loss": 0.006, "step": 33880 }, { "grad_norm": 0.13057288527488708, "learning_rate": 2.5799378713917012e-05, "loss": 0.0062, "step": 33890 }, { "grad_norm": 0.14206378161907196, "learning_rate": 2.5770446265951455e-05, "loss": 0.0073, "step": 33900 }, { "grad_norm": 0.09284420311450958, "learning_rate": 2.5741524416815877e-05, "loss": 0.0046, "step": 33910 }, { "grad_norm": 0.09526870399713516, "learning_rate": 2.571261317916166e-05, "loss": 0.0067, "step": 33920 }, { "grad_norm": 0.14302410185337067, "learning_rate": 2.5683712565635597e-05, "loss": 0.008, "step": 33930 }, { "grad_norm": 0.08149585127830505, "learning_rate": 2.565482258887979e-05, "loss": 0.0057, "step": 33940 }, { "grad_norm": 0.11676711589097977, "learning_rate": 2.5625943261531664e-05, "loss": 0.0052, "step": 33950 }, { "grad_norm": 0.08971579372882843, "learning_rate": 2.5597074596224067e-05, "loss": 0.0069, "step": 33960 }, { "grad_norm": 0.16288378834724426, "learning_rate": 2.5568216605585105e-05, "loss": 0.0067, "step": 33970 }, { "grad_norm": 0.17934663593769073, "learning_rate": 2.5539369302238235e-05, "loss": 0.0071, "step": 33980 }, { "grad_norm": 0.1718117594718933, "learning_rate": 2.5510532698802276e-05, "loss": 0.0059, "step": 33990 }, { "grad_norm": 0.0929725170135498, "learning_rate": 2.5481706807891324e-05, "loss": 0.0071, "step": 34000 }, { "grad_norm": 0.0882415845990181, "learning_rate": 2.5452891642114783e-05, "loss": 0.0046, "step": 34010 }, { "grad_norm": 0.12625473737716675, "learning_rate": 2.542408721407742e-05, "loss": 0.0061, "step": 34020 }, { "grad_norm": 0.11077640950679779, "learning_rate": 2.539529353637926e-05, "loss": 0.0077, "step": 34030 }, { "grad_norm": 0.098496213555336, "learning_rate": 2.5366510621615626e-05, "loss": 0.0062, "step": 34040 }, { "grad_norm": 0.11500569432973862, "learning_rate": 2.5337738482377132e-05, "loss": 0.0054, "step": 34050 }, { "grad_norm": 0.10508265346288681, "learning_rate": 2.5308977131249722e-05, "loss": 0.0087, "step": 34060 }, { "grad_norm": 0.1389881819486618, "learning_rate": 2.5280226580814554e-05, "loss": 0.0068, "step": 34070 }, { "grad_norm": 0.11655905842781067, "learning_rate": 2.525148684364813e-05, "loss": 0.0063, "step": 34080 }, { "grad_norm": 0.10040366649627686, "learning_rate": 2.522275793232215e-05, "loss": 0.0067, "step": 34090 }, { "grad_norm": 0.10428639501333237, "learning_rate": 2.519403985940365e-05, "loss": 0.0053, "step": 34100 }, { "grad_norm": 0.1135585755109787, "learning_rate": 2.5165332637454863e-05, "loss": 0.008, "step": 34110 }, { "grad_norm": 0.11215099692344666, "learning_rate": 2.513663627903334e-05, "loss": 0.0068, "step": 34120 }, { "grad_norm": 0.15129806101322174, "learning_rate": 2.5107950796691815e-05, "loss": 0.0078, "step": 34130 }, { "grad_norm": 0.1122749075293541, "learning_rate": 2.5079276202978276e-05, "loss": 0.0048, "step": 34140 }, { "grad_norm": 0.07896638661623001, "learning_rate": 2.505061251043601e-05, "loss": 0.0059, "step": 34150 }, { "grad_norm": 0.16812258958816528, "learning_rate": 2.502195973160346e-05, "loss": 0.0052, "step": 34160 }, { "grad_norm": 0.09644857048988342, "learning_rate": 2.4993317879014316e-05, "loss": 0.0053, "step": 34170 }, { "grad_norm": 0.11964108049869537, "learning_rate": 2.4964686965197526e-05, "loss": 0.0064, "step": 34180 }, { "grad_norm": 0.11381368339061737, "learning_rate": 2.4936067002677217e-05, "loss": 0.0053, "step": 34190 }, { "grad_norm": 0.1601971536874771, "learning_rate": 2.4907458003972724e-05, "loss": 0.005, "step": 34200 }, { "grad_norm": 0.15147529542446136, "learning_rate": 2.4878859981598572e-05, "loss": 0.0042, "step": 34210 }, { "grad_norm": 0.13925118744373322, "learning_rate": 2.4850272948064555e-05, "loss": 0.0053, "step": 34220 }, { "grad_norm": 0.12642847001552582, "learning_rate": 2.4821696915875596e-05, "loss": 0.0055, "step": 34230 }, { "grad_norm": 0.11187870055437088, "learning_rate": 2.4793131897531798e-05, "loss": 0.0065, "step": 34240 }, { "grad_norm": 0.08190196752548218, "learning_rate": 2.47645779055285e-05, "loss": 0.006, "step": 34250 }, { "grad_norm": 0.1477123647928238, "learning_rate": 2.4736034952356186e-05, "loss": 0.0082, "step": 34260 }, { "grad_norm": 0.1640128791332245, "learning_rate": 2.470750305050048e-05, "loss": 0.0057, "step": 34270 }, { "grad_norm": 0.16676567494869232, "learning_rate": 2.4678982212442253e-05, "loss": 0.0072, "step": 34280 }, { "grad_norm": 0.15542592108249664, "learning_rate": 2.4650472450657463e-05, "loss": 0.0062, "step": 34290 }, { "grad_norm": 0.11458322405815125, "learning_rate": 2.4621973777617237e-05, "loss": 0.005, "step": 34300 }, { "grad_norm": 0.158619225025177, "learning_rate": 2.459348620578789e-05, "loss": 0.0045, "step": 34310 }, { "grad_norm": 0.1252453476190567, "learning_rate": 2.456500974763083e-05, "loss": 0.0067, "step": 34320 }, { "grad_norm": 0.07362968474626541, "learning_rate": 2.4536544415602653e-05, "loss": 0.0053, "step": 34330 }, { "grad_norm": 0.09680784493684769, "learning_rate": 2.4508090222155027e-05, "loss": 0.0048, "step": 34340 }, { "grad_norm": 0.17922547459602356, "learning_rate": 2.4479647179734822e-05, "loss": 0.0061, "step": 34350 }, { "grad_norm": 0.13934382796287537, "learning_rate": 2.4451215300783953e-05, "loss": 0.0059, "step": 34360 }, { "grad_norm": 0.16711854934692383, "learning_rate": 2.442279459773953e-05, "loss": 0.005, "step": 34370 }, { "grad_norm": 0.11192802339792252, "learning_rate": 2.4394385083033712e-05, "loss": 0.0062, "step": 34380 }, { "grad_norm": 0.21630483865737915, "learning_rate": 2.4365986769093792e-05, "loss": 0.0088, "step": 34390 }, { "grad_norm": 0.1467602401971817, "learning_rate": 2.4337599668342127e-05, "loss": 0.0047, "step": 34400 }, { "grad_norm": 0.10494907945394516, "learning_rate": 2.4309223793196245e-05, "loss": 0.0054, "step": 34410 }, { "grad_norm": 0.14621210098266602, "learning_rate": 2.4280859156068698e-05, "loss": 0.0049, "step": 34420 }, { "grad_norm": 0.13687361776828766, "learning_rate": 2.4252505769367123e-05, "loss": 0.0055, "step": 34430 }, { "grad_norm": 0.08893392235040665, "learning_rate": 2.422416364549429e-05, "loss": 0.0042, "step": 34440 }, { "grad_norm": 0.19173145294189453, "learning_rate": 2.4195832796847994e-05, "loss": 0.0071, "step": 34450 }, { "grad_norm": 0.15659818053245544, "learning_rate": 2.416751323582109e-05, "loss": 0.0058, "step": 34460 }, { "grad_norm": 0.0805775448679924, "learning_rate": 2.4139204974801544e-05, "loss": 0.0049, "step": 34470 }, { "grad_norm": 0.13989172875881195, "learning_rate": 2.4110908026172347e-05, "loss": 0.0064, "step": 34480 }, { "grad_norm": 0.10565286129713058, "learning_rate": 2.4082622402311518e-05, "loss": 0.0056, "step": 34490 }, { "grad_norm": 0.1065058633685112, "learning_rate": 2.4054348115592183e-05, "loss": 0.0059, "step": 34500 }, { "grad_norm": 0.10832026600837708, "learning_rate": 2.4026085178382475e-05, "loss": 0.0065, "step": 34510 }, { "grad_norm": 0.1567484438419342, "learning_rate": 2.3997833603045523e-05, "loss": 0.0069, "step": 34520 }, { "grad_norm": 0.097134068608284, "learning_rate": 2.396959340193958e-05, "loss": 0.0059, "step": 34530 }, { "grad_norm": 0.09936913102865219, "learning_rate": 2.3941364587417836e-05, "loss": 0.005, "step": 34540 }, { "grad_norm": 0.08329066634178162, "learning_rate": 2.3913147171828526e-05, "loss": 0.0052, "step": 34550 }, { "grad_norm": 0.10421241074800491, "learning_rate": 2.388494116751494e-05, "loss": 0.0051, "step": 34560 }, { "grad_norm": 0.10444299131631851, "learning_rate": 2.3856746586815302e-05, "loss": 0.0062, "step": 34570 }, { "grad_norm": 0.07507981359958649, "learning_rate": 2.3828563442062924e-05, "loss": 0.0081, "step": 34580 }, { "grad_norm": 0.11690877377986908, "learning_rate": 2.3800391745586026e-05, "loss": 0.005, "step": 34590 }, { "grad_norm": 0.13726769387722015, "learning_rate": 2.377223150970791e-05, "loss": 0.0041, "step": 34600 }, { "grad_norm": 0.09613223373889923, "learning_rate": 2.3744082746746793e-05, "loss": 0.0059, "step": 34610 }, { "grad_norm": 0.1402672380208969, "learning_rate": 2.3715945469015893e-05, "loss": 0.0061, "step": 34620 }, { "grad_norm": 0.12723679840564728, "learning_rate": 2.368781968882343e-05, "loss": 0.007, "step": 34630 }, { "grad_norm": 0.08378996700048447, "learning_rate": 2.365970541847258e-05, "loss": 0.0062, "step": 34640 }, { "grad_norm": 0.15564332902431488, "learning_rate": 2.3631602670261445e-05, "loss": 0.0085, "step": 34650 }, { "grad_norm": 0.10493789613246918, "learning_rate": 2.3603511456483164e-05, "loss": 0.0046, "step": 34660 }, { "grad_norm": 0.13957317173480988, "learning_rate": 2.357543178942577e-05, "loss": 0.0051, "step": 34670 }, { "grad_norm": 0.09477540105581284, "learning_rate": 2.3547363681372242e-05, "loss": 0.0068, "step": 34680 }, { "grad_norm": 0.13119792938232422, "learning_rate": 2.3519307144600565e-05, "loss": 0.0055, "step": 34690 }, { "grad_norm": 0.09738360345363617, "learning_rate": 2.3491262191383607e-05, "loss": 0.0065, "step": 34700 }, { "grad_norm": 0.1311950534582138, "learning_rate": 2.3463228833989154e-05, "loss": 0.0062, "step": 34710 }, { "grad_norm": 0.20895199477672577, "learning_rate": 2.343520708467999e-05, "loss": 0.0061, "step": 34720 }, { "grad_norm": 0.12188415974378586, "learning_rate": 2.3407196955713767e-05, "loss": 0.0047, "step": 34730 }, { "grad_norm": 0.12886710464954376, "learning_rate": 2.337919845934306e-05, "loss": 0.0071, "step": 34740 }, { "grad_norm": 0.12034628540277481, "learning_rate": 2.335121160781535e-05, "loss": 0.0065, "step": 34750 }, { "grad_norm": 0.09959843754768372, "learning_rate": 2.3323236413373063e-05, "loss": 0.0047, "step": 34760 }, { "grad_norm": 0.19704420864582062, "learning_rate": 2.3295272888253494e-05, "loss": 0.005, "step": 34770 }, { "grad_norm": 0.08427391946315765, "learning_rate": 2.3267321044688815e-05, "loss": 0.005, "step": 34780 }, { "grad_norm": 0.11827975511550903, "learning_rate": 2.323938089490615e-05, "loss": 0.0053, "step": 34790 }, { "grad_norm": 0.1329450160264969, "learning_rate": 2.321145245112743e-05, "loss": 0.0046, "step": 34800 }, { "grad_norm": 0.1407606154680252, "learning_rate": 2.318353572556953e-05, "loss": 0.0063, "step": 34810 }, { "grad_norm": 0.14674392342567444, "learning_rate": 2.3155630730444183e-05, "loss": 0.0045, "step": 34820 }, { "grad_norm": 0.12385185062885284, "learning_rate": 2.3127737477957968e-05, "loss": 0.0066, "step": 34830 }, { "grad_norm": 0.17432749271392822, "learning_rate": 2.3099855980312324e-05, "loss": 0.005, "step": 34840 }, { "grad_norm": 0.12907254695892334, "learning_rate": 2.3071986249703602e-05, "loss": 0.005, "step": 34850 }, { "grad_norm": 0.08644958585500717, "learning_rate": 2.3044128298322953e-05, "loss": 0.005, "step": 34860 }, { "grad_norm": 0.08556610345840454, "learning_rate": 2.301628213835637e-05, "loss": 0.008, "step": 34870 }, { "grad_norm": 0.07086249440908432, "learning_rate": 2.2988447781984744e-05, "loss": 0.0055, "step": 34880 }, { "grad_norm": 0.08755135536193848, "learning_rate": 2.2960625241383753e-05, "loss": 0.0066, "step": 34890 }, { "grad_norm": 0.08791303634643555, "learning_rate": 2.2932814528723922e-05, "loss": 0.0045, "step": 34900 }, { "grad_norm": 0.15957427024841309, "learning_rate": 2.2905015656170582e-05, "loss": 0.0049, "step": 34910 }, { "grad_norm": 0.12368353456258774, "learning_rate": 2.287722863588394e-05, "loss": 0.005, "step": 34920 }, { "grad_norm": 0.0950416550040245, "learning_rate": 2.2849453480018962e-05, "loss": 0.0054, "step": 34930 }, { "grad_norm": 0.11984529346227646, "learning_rate": 2.282169020072543e-05, "loss": 0.0064, "step": 34940 }, { "grad_norm": 0.1536760777235031, "learning_rate": 2.2793938810147986e-05, "loss": 0.0051, "step": 34950 }, { "grad_norm": 0.12738732993602753, "learning_rate": 2.2766199320426007e-05, "loss": 0.0051, "step": 34960 }, { "grad_norm": 0.11021461337804794, "learning_rate": 2.2738471743693674e-05, "loss": 0.0047, "step": 34970 }, { "grad_norm": 0.13272227346897125, "learning_rate": 2.2710756092080006e-05, "loss": 0.0083, "step": 34980 }, { "grad_norm": 0.11192184686660767, "learning_rate": 2.2683052377708747e-05, "loss": 0.0059, "step": 34990 }, { "grad_norm": 0.14295169711112976, "learning_rate": 2.265536061269843e-05, "loss": 0.0057, "step": 35000 }, { "grad_norm": 0.13739238679409027, "learning_rate": 2.262768080916241e-05, "loss": 0.0101, "step": 35010 }, { "grad_norm": 0.13588003814220428, "learning_rate": 2.260001297920874e-05, "loss": 0.0057, "step": 35020 }, { "grad_norm": 0.13672487437725067, "learning_rate": 2.2572357134940297e-05, "loss": 0.0066, "step": 35030 }, { "grad_norm": 0.1135120689868927, "learning_rate": 2.2544713288454655e-05, "loss": 0.0058, "step": 35040 }, { "grad_norm": 0.12273205071687698, "learning_rate": 2.2517081451844213e-05, "loss": 0.0052, "step": 35050 }, { "grad_norm": 0.09584689885377884, "learning_rate": 2.248946163719605e-05, "loss": 0.0047, "step": 35060 }, { "grad_norm": 0.1670379489660263, "learning_rate": 2.2461853856591996e-05, "loss": 0.0067, "step": 35070 }, { "grad_norm": 0.19908767938613892, "learning_rate": 2.2434258122108664e-05, "loss": 0.0097, "step": 35080 }, { "grad_norm": 0.1868046671152115, "learning_rate": 2.2406674445817356e-05, "loss": 0.0063, "step": 35090 }, { "grad_norm": 0.1673387587070465, "learning_rate": 2.2379102839784083e-05, "loss": 0.0061, "step": 35100 }, { "grad_norm": 0.11417462676763535, "learning_rate": 2.2351543316069645e-05, "loss": 0.0089, "step": 35110 }, { "grad_norm": 0.1417246162891388, "learning_rate": 2.2323995886729493e-05, "loss": 0.0062, "step": 35120 }, { "grad_norm": 0.11772923171520233, "learning_rate": 2.2296460563813786e-05, "loss": 0.0059, "step": 35130 }, { "grad_norm": 0.12023168057203293, "learning_rate": 2.2268937359367457e-05, "loss": 0.0051, "step": 35140 }, { "grad_norm": 0.11053407937288284, "learning_rate": 2.224142628543006e-05, "loss": 0.0054, "step": 35150 }, { "grad_norm": 0.17011408507823944, "learning_rate": 2.2213927354035863e-05, "loss": 0.005, "step": 35160 }, { "grad_norm": 0.123911052942276, "learning_rate": 2.218644057721387e-05, "loss": 0.0048, "step": 35170 }, { "grad_norm": 0.12632562220096588, "learning_rate": 2.2158965966987703e-05, "loss": 0.0079, "step": 35180 }, { "grad_norm": 0.15502895414829254, "learning_rate": 2.2131503535375687e-05, "loss": 0.0071, "step": 35190 }, { "grad_norm": 0.12610439956188202, "learning_rate": 2.2104053294390846e-05, "loss": 0.0086, "step": 35200 }, { "grad_norm": 0.08372524380683899, "learning_rate": 2.2076615256040834e-05, "loss": 0.0053, "step": 35210 }, { "grad_norm": 0.1393045037984848, "learning_rate": 2.2049189432327966e-05, "loss": 0.0053, "step": 35220 }, { "grad_norm": 0.13714583218097687, "learning_rate": 2.202177583524926e-05, "loss": 0.0055, "step": 35230 }, { "grad_norm": 0.13145369291305542, "learning_rate": 2.1994374476796343e-05, "loss": 0.0061, "step": 35240 }, { "grad_norm": 0.0792953297495842, "learning_rate": 2.1966985368955477e-05, "loss": 0.006, "step": 35250 }, { "grad_norm": 0.09358973056077957, "learning_rate": 2.193960852370762e-05, "loss": 0.0066, "step": 35260 }, { "grad_norm": 0.15490475296974182, "learning_rate": 2.1912243953028317e-05, "loss": 0.0053, "step": 35270 }, { "grad_norm": 0.06490424275398254, "learning_rate": 2.1884891668887776e-05, "loss": 0.004, "step": 35280 }, { "grad_norm": 0.07684825360774994, "learning_rate": 2.1857551683250787e-05, "loss": 0.0065, "step": 35290 }, { "grad_norm": 0.08833988755941391, "learning_rate": 2.1830224008076827e-05, "loss": 0.0065, "step": 35300 }, { "grad_norm": 0.13491325080394745, "learning_rate": 2.1802908655319924e-05, "loss": 0.0055, "step": 35310 }, { "grad_norm": 0.14672917127609253, "learning_rate": 2.1775605636928738e-05, "loss": 0.0061, "step": 35320 }, { "grad_norm": 0.16205453872680664, "learning_rate": 2.1748314964846555e-05, "loss": 0.0065, "step": 35330 }, { "grad_norm": 0.10215020179748535, "learning_rate": 2.172103665101124e-05, "loss": 0.0063, "step": 35340 }, { "grad_norm": 0.12591278553009033, "learning_rate": 2.1693770707355226e-05, "loss": 0.0061, "step": 35350 }, { "grad_norm": 0.12244172394275665, "learning_rate": 2.1666517145805605e-05, "loss": 0.005, "step": 35360 }, { "grad_norm": 0.10559637099504471, "learning_rate": 2.1639275978283985e-05, "loss": 0.009, "step": 35370 }, { "grad_norm": 0.11655815690755844, "learning_rate": 2.1612047216706567e-05, "loss": 0.0045, "step": 35380 }, { "grad_norm": 0.11130638420581818, "learning_rate": 2.158483087298417e-05, "loss": 0.0062, "step": 35390 }, { "grad_norm": 0.16615867614746094, "learning_rate": 2.155762695902213e-05, "loss": 0.0059, "step": 35400 }, { "grad_norm": 0.07955991476774216, "learning_rate": 2.1530435486720363e-05, "loss": 0.0049, "step": 35410 }, { "grad_norm": 0.13495247066020966, "learning_rate": 2.1503256467973327e-05, "loss": 0.0073, "step": 35420 }, { "grad_norm": 0.056357305496931076, "learning_rate": 2.147608991467008e-05, "loss": 0.0058, "step": 35430 }, { "grad_norm": 0.1344807893037796, "learning_rate": 2.144893583869418e-05, "loss": 0.0079, "step": 35440 }, { "grad_norm": 0.1547788381576538, "learning_rate": 2.142179425192372e-05, "loss": 0.0062, "step": 35450 }, { "grad_norm": 0.12705327570438385, "learning_rate": 2.1394665166231393e-05, "loss": 0.0068, "step": 35460 }, { "grad_norm": 0.16363584995269775, "learning_rate": 2.1367548593484364e-05, "loss": 0.0065, "step": 35470 }, { "grad_norm": 0.166412353515625, "learning_rate": 2.134044454554433e-05, "loss": 0.0056, "step": 35480 }, { "grad_norm": 0.13690611720085144, "learning_rate": 2.1313353034267548e-05, "loss": 0.0036, "step": 35490 }, { "grad_norm": 0.07692886143922806, "learning_rate": 2.1286274071504735e-05, "loss": 0.0068, "step": 35500 }, { "grad_norm": 0.13267941772937775, "learning_rate": 2.125920766910119e-05, "loss": 0.0082, "step": 35510 }, { "grad_norm": 0.08841291815042496, "learning_rate": 2.1232153838896628e-05, "loss": 0.0061, "step": 35520 }, { "grad_norm": 0.13959051668643951, "learning_rate": 2.1205112592725366e-05, "loss": 0.0061, "step": 35530 }, { "grad_norm": 0.1251850575208664, "learning_rate": 2.1178083942416114e-05, "loss": 0.0062, "step": 35540 }, { "grad_norm": 0.15141332149505615, "learning_rate": 2.1151067899792164e-05, "loss": 0.0051, "step": 35550 }, { "grad_norm": 0.09681747853755951, "learning_rate": 2.1124064476671235e-05, "loss": 0.0059, "step": 35560 }, { "grad_norm": 0.11684110015630722, "learning_rate": 2.1097073684865516e-05, "loss": 0.0062, "step": 35570 }, { "grad_norm": 0.12480166554450989, "learning_rate": 2.107009553618174e-05, "loss": 0.0045, "step": 35580 }, { "grad_norm": 0.16156095266342163, "learning_rate": 2.1043130042421044e-05, "loss": 0.0055, "step": 35590 }, { "grad_norm": 0.1315840482711792, "learning_rate": 2.1016177215379053e-05, "loss": 0.0056, "step": 35600 }, { "grad_norm": 0.14412938058376312, "learning_rate": 2.0989237066845824e-05, "loss": 0.0061, "step": 35610 }, { "grad_norm": 0.1289074867963791, "learning_rate": 2.096230960860594e-05, "loss": 0.0069, "step": 35620 }, { "grad_norm": 0.10797052085399628, "learning_rate": 2.0935394852438366e-05, "loss": 0.0077, "step": 35630 }, { "grad_norm": 0.21066510677337646, "learning_rate": 2.090849281011651e-05, "loss": 0.0071, "step": 35640 }, { "grad_norm": 0.14368481934070587, "learning_rate": 2.088160349340827e-05, "loss": 0.0052, "step": 35650 }, { "grad_norm": 0.11517173051834106, "learning_rate": 2.085472691407594e-05, "loss": 0.0054, "step": 35660 }, { "grad_norm": 0.08217577636241913, "learning_rate": 2.082786308387622e-05, "loss": 0.0051, "step": 35670 }, { "grad_norm": 0.14754971861839294, "learning_rate": 2.0801012014560306e-05, "loss": 0.0058, "step": 35680 }, { "grad_norm": 0.1762806475162506, "learning_rate": 2.0774173717873746e-05, "loss": 0.007, "step": 35690 }, { "grad_norm": 0.16082265973091125, "learning_rate": 2.074734820555651e-05, "loss": 0.0062, "step": 35700 }, { "grad_norm": 0.11156415939331055, "learning_rate": 2.0720535489343014e-05, "loss": 0.0053, "step": 35710 }, { "grad_norm": 0.12413349002599716, "learning_rate": 2.0693735580962044e-05, "loss": 0.0082, "step": 35720 }, { "grad_norm": 0.11695197969675064, "learning_rate": 2.0666948492136766e-05, "loss": 0.0066, "step": 35730 }, { "grad_norm": 0.16160140931606293, "learning_rate": 2.0640174234584797e-05, "loss": 0.0059, "step": 35740 }, { "grad_norm": 0.11324487626552582, "learning_rate": 2.061341282001808e-05, "loss": 0.0055, "step": 35750 }, { "grad_norm": 0.128094881772995, "learning_rate": 2.0586664260142997e-05, "loss": 0.0049, "step": 35760 }, { "grad_norm": 0.11766497045755386, "learning_rate": 2.0559928566660237e-05, "loss": 0.0071, "step": 35770 }, { "grad_norm": 0.0983763262629509, "learning_rate": 2.0533205751264945e-05, "loss": 0.004, "step": 35780 }, { "grad_norm": 0.09665339440107346, "learning_rate": 2.050649582564656e-05, "loss": 0.004, "step": 35790 }, { "grad_norm": 0.12870456278324127, "learning_rate": 2.04797988014889e-05, "loss": 0.0059, "step": 35800 }, { "grad_norm": 0.11489642411470413, "learning_rate": 2.0453114690470194e-05, "loss": 0.0072, "step": 35810 }, { "grad_norm": 0.16780346632003784, "learning_rate": 2.0426443504262944e-05, "loss": 0.0046, "step": 35820 }, { "grad_norm": 0.08529996126890182, "learning_rate": 2.0399785254534027e-05, "loss": 0.0055, "step": 35830 }, { "grad_norm": 0.12564165890216827, "learning_rate": 2.0373139952944697e-05, "loss": 0.0055, "step": 35840 }, { "grad_norm": 0.09580177068710327, "learning_rate": 2.03465076111505e-05, "loss": 0.005, "step": 35850 }, { "grad_norm": 0.11573543399572372, "learning_rate": 2.0319888240801305e-05, "loss": 0.0036, "step": 35860 }, { "grad_norm": 0.10913408547639847, "learning_rate": 2.0293281853541375e-05, "loss": 0.0074, "step": 35870 }, { "grad_norm": 0.1267067939043045, "learning_rate": 2.0266688461009216e-05, "loss": 0.004, "step": 35880 }, { "grad_norm": 0.12682023644447327, "learning_rate": 2.0240108074837666e-05, "loss": 0.005, "step": 35890 }, { "grad_norm": 0.1423082947731018, "learning_rate": 2.021354070665393e-05, "loss": 0.005, "step": 35900 }, { "grad_norm": 0.18228401243686676, "learning_rate": 2.0186986368079465e-05, "loss": 0.0072, "step": 35910 }, { "grad_norm": 0.10067007690668106, "learning_rate": 2.016044507073001e-05, "loss": 0.0039, "step": 35920 }, { "grad_norm": 0.09454036504030228, "learning_rate": 2.013391682621567e-05, "loss": 0.0053, "step": 35930 }, { "grad_norm": 0.08748086541891098, "learning_rate": 2.0107401646140787e-05, "loss": 0.0045, "step": 35940 }, { "grad_norm": 0.07041362673044205, "learning_rate": 2.0080899542104003e-05, "loss": 0.0052, "step": 35950 }, { "grad_norm": 0.0801415666937828, "learning_rate": 2.0054410525698215e-05, "loss": 0.0044, "step": 35960 }, { "grad_norm": 0.14840467274188995, "learning_rate": 2.0027934608510653e-05, "loss": 0.006, "step": 35970 }, { "grad_norm": 0.12161961197853088, "learning_rate": 2.000147180212275e-05, "loss": 0.0054, "step": 35980 }, { "grad_norm": 0.19222289323806763, "learning_rate": 1.9975022118110275e-05, "loss": 0.0063, "step": 35990 }, { "grad_norm": 0.09274847060441971, "learning_rate": 1.994858556804318e-05, "loss": 0.0062, "step": 36000 }, { "grad_norm": 0.08763248473405838, "learning_rate": 1.9922162163485748e-05, "loss": 0.0057, "step": 36010 }, { "grad_norm": 0.1486961990594864, "learning_rate": 1.989575191599643e-05, "loss": 0.0051, "step": 36020 }, { "grad_norm": 0.11918248981237411, "learning_rate": 1.9869354837128012e-05, "loss": 0.0079, "step": 36030 }, { "grad_norm": 0.1112474650144577, "learning_rate": 1.984297093842745e-05, "loss": 0.0064, "step": 36040 }, { "grad_norm": 0.10586199164390564, "learning_rate": 1.981660023143594e-05, "loss": 0.0051, "step": 36050 }, { "grad_norm": 0.08783840388059616, "learning_rate": 1.9790242727688963e-05, "loss": 0.0064, "step": 36060 }, { "grad_norm": 0.13335925340652466, "learning_rate": 1.9763898438716165e-05, "loss": 0.0051, "step": 36070 }, { "grad_norm": 0.12142641097307205, "learning_rate": 1.973756737604142e-05, "loss": 0.0056, "step": 36080 }, { "grad_norm": 0.13358452916145325, "learning_rate": 1.9711249551182858e-05, "loss": 0.0065, "step": 36090 }, { "grad_norm": 0.12394564598798752, "learning_rate": 1.968494497565278e-05, "loss": 0.0075, "step": 36100 }, { "grad_norm": 0.118794746696949, "learning_rate": 1.9658653660957698e-05, "loss": 0.0053, "step": 36110 }, { "grad_norm": 0.10569504648447037, "learning_rate": 1.9632375618598302e-05, "loss": 0.0056, "step": 36120 }, { "grad_norm": 0.1340939700603485, "learning_rate": 1.960611086006955e-05, "loss": 0.005, "step": 36130 }, { "grad_norm": 0.08958996087312698, "learning_rate": 1.9579859396860513e-05, "loss": 0.004, "step": 36140 }, { "grad_norm": 0.17295940220355988, "learning_rate": 1.9553621240454452e-05, "loss": 0.005, "step": 36150 }, { "grad_norm": 0.08610067516565323, "learning_rate": 1.9527396402328864e-05, "loss": 0.0045, "step": 36160 }, { "grad_norm": 0.12216044962406158, "learning_rate": 1.9501184893955376e-05, "loss": 0.0076, "step": 36170 }, { "grad_norm": 0.07512973994016647, "learning_rate": 1.9474986726799765e-05, "loss": 0.0048, "step": 36180 }, { "grad_norm": 0.1330847144126892, "learning_rate": 1.944880191232204e-05, "loss": 0.0049, "step": 36190 }, { "grad_norm": 0.09460041671991348, "learning_rate": 1.9422630461976304e-05, "loss": 0.004, "step": 36200 }, { "grad_norm": 0.072636179625988, "learning_rate": 1.9396472387210835e-05, "loss": 0.0038, "step": 36210 }, { "grad_norm": 0.08492467552423477, "learning_rate": 1.9370327699468082e-05, "loss": 0.004, "step": 36220 }, { "grad_norm": 0.10375456511974335, "learning_rate": 1.9344196410184594e-05, "loss": 0.0075, "step": 36230 }, { "grad_norm": 0.182500958442688, "learning_rate": 1.9318078530791123e-05, "loss": 0.0057, "step": 36240 }, { "grad_norm": 0.105039082467556, "learning_rate": 1.9291974072712473e-05, "loss": 0.0054, "step": 36250 }, { "grad_norm": 0.17608702182769775, "learning_rate": 1.9265883047367656e-05, "loss": 0.0057, "step": 36260 }, { "grad_norm": 0.07847876101732254, "learning_rate": 1.9239805466169748e-05, "loss": 0.0053, "step": 36270 }, { "grad_norm": 0.11857032030820847, "learning_rate": 1.9213741340525992e-05, "loss": 0.0058, "step": 36280 }, { "grad_norm": 0.12176256626844406, "learning_rate": 1.9187690681837713e-05, "loss": 0.0057, "step": 36290 }, { "grad_norm": 0.11118478327989578, "learning_rate": 1.916165350150035e-05, "loss": 0.0048, "step": 36300 }, { "grad_norm": 0.1348710060119629, "learning_rate": 1.9135629810903434e-05, "loss": 0.0048, "step": 36310 }, { "grad_norm": 0.08655591309070587, "learning_rate": 1.9109619621430648e-05, "loss": 0.0041, "step": 36320 }, { "grad_norm": 0.10937072336673737, "learning_rate": 1.9083622944459712e-05, "loss": 0.0061, "step": 36330 }, { "grad_norm": 0.14272703230381012, "learning_rate": 1.9057639791362437e-05, "loss": 0.0059, "step": 36340 }, { "grad_norm": 0.11045803874731064, "learning_rate": 1.903167017350478e-05, "loss": 0.0047, "step": 36350 }, { "grad_norm": 0.2855115234851837, "learning_rate": 1.9005714102246707e-05, "loss": 0.0039, "step": 36360 }, { "grad_norm": 0.11477696895599365, "learning_rate": 1.897977158894227e-05, "loss": 0.0052, "step": 36370 }, { "grad_norm": 0.17791031301021576, "learning_rate": 1.8953842644939644e-05, "loss": 0.0062, "step": 36380 }, { "grad_norm": 0.12567931413650513, "learning_rate": 1.8927927281581015e-05, "loss": 0.0073, "step": 36390 }, { "grad_norm": 0.11271165311336517, "learning_rate": 1.890202551020262e-05, "loss": 0.0047, "step": 36400 }, { "grad_norm": 0.12896092236042023, "learning_rate": 1.8876137342134813e-05, "loss": 0.0049, "step": 36410 }, { "grad_norm": 0.12159454822540283, "learning_rate": 1.885026278870194e-05, "loss": 0.0044, "step": 36420 }, { "grad_norm": 0.08800026029348373, "learning_rate": 1.8824401861222384e-05, "loss": 0.0057, "step": 36430 }, { "grad_norm": 0.08750487864017487, "learning_rate": 1.879855457100864e-05, "loss": 0.0047, "step": 36440 }, { "grad_norm": 0.17297109961509705, "learning_rate": 1.877272092936717e-05, "loss": 0.0047, "step": 36450 }, { "grad_norm": 0.100660540163517, "learning_rate": 1.8746900947598467e-05, "loss": 0.0042, "step": 36460 }, { "grad_norm": 0.15037983655929565, "learning_rate": 1.87210946369971e-05, "loss": 0.005, "step": 36470 }, { "grad_norm": 0.07337719202041626, "learning_rate": 1.8695302008851602e-05, "loss": 0.0048, "step": 36480 }, { "grad_norm": 0.13309665024280548, "learning_rate": 1.8669523074444568e-05, "loss": 0.0051, "step": 36490 }, { "grad_norm": 0.14609773457050323, "learning_rate": 1.8643757845052545e-05, "loss": 0.0043, "step": 36500 }, { "grad_norm": 0.1037478819489479, "learning_rate": 1.8618006331946164e-05, "loss": 0.0038, "step": 36510 }, { "grad_norm": 0.058943700045347214, "learning_rate": 1.859226854638999e-05, "loss": 0.0034, "step": 36520 }, { "grad_norm": 0.05333798751235008, "learning_rate": 1.8566544499642587e-05, "loss": 0.0034, "step": 36530 }, { "grad_norm": 0.11318719387054443, "learning_rate": 1.854083420295656e-05, "loss": 0.0056, "step": 36540 }, { "grad_norm": 0.11858116090297699, "learning_rate": 1.8515137667578453e-05, "loss": 0.0047, "step": 36550 }, { "grad_norm": 0.07330646365880966, "learning_rate": 1.8489454904748784e-05, "loss": 0.004, "step": 36560 }, { "grad_norm": 0.11236634850502014, "learning_rate": 1.8463785925702094e-05, "loss": 0.0039, "step": 36570 }, { "grad_norm": 0.07832508534193039, "learning_rate": 1.843813074166686e-05, "loss": 0.0071, "step": 36580 }, { "grad_norm": 0.12992481887340546, "learning_rate": 1.841248936386551e-05, "loss": 0.0065, "step": 36590 }, { "grad_norm": 0.10245084762573242, "learning_rate": 1.8386861803514483e-05, "loss": 0.0051, "step": 36600 }, { "grad_norm": 0.11716794967651367, "learning_rate": 1.836124807182414e-05, "loss": 0.0063, "step": 36610 }, { "grad_norm": 0.08011887967586517, "learning_rate": 1.8335648179998783e-05, "loss": 0.0053, "step": 36620 }, { "grad_norm": 0.1382792890071869, "learning_rate": 1.8310062139236667e-05, "loss": 0.0049, "step": 36630 }, { "grad_norm": 0.11368222534656525, "learning_rate": 1.828448996073002e-05, "loss": 0.0046, "step": 36640 }, { "grad_norm": 0.12169136852025986, "learning_rate": 1.825893165566498e-05, "loss": 0.0066, "step": 36650 }, { "grad_norm": 0.12440536171197891, "learning_rate": 1.823338723522159e-05, "loss": 0.0037, "step": 36660 }, { "grad_norm": 0.11927314847707748, "learning_rate": 1.8207856710573878e-05, "loss": 0.0048, "step": 36670 }, { "grad_norm": 0.07962151616811752, "learning_rate": 1.8182340092889756e-05, "loss": 0.005, "step": 36680 }, { "grad_norm": 0.11436884105205536, "learning_rate": 1.8156837393331038e-05, "loss": 0.006, "step": 36690 }, { "grad_norm": 0.11253032833337784, "learning_rate": 1.81313486230535e-05, "loss": 0.005, "step": 36700 }, { "grad_norm": 0.13234944641590118, "learning_rate": 1.8105873793206773e-05, "loss": 0.005, "step": 36710 }, { "grad_norm": 0.12308768182992935, "learning_rate": 1.8080412914934436e-05, "loss": 0.006, "step": 36720 }, { "grad_norm": 0.07562725245952606, "learning_rate": 1.805496599937392e-05, "loss": 0.0045, "step": 36730 }, { "grad_norm": 0.11648557335138321, "learning_rate": 1.802953305765659e-05, "loss": 0.0051, "step": 36740 }, { "grad_norm": 0.07452497631311417, "learning_rate": 1.8004114100907653e-05, "loss": 0.0072, "step": 36750 }, { "grad_norm": 0.08504360169172287, "learning_rate": 1.7978709140246263e-05, "loss": 0.0062, "step": 36760 }, { "grad_norm": 0.09415525943040848, "learning_rate": 1.795331818678539e-05, "loss": 0.0041, "step": 36770 }, { "grad_norm": 0.09672126919031143, "learning_rate": 1.792794125163188e-05, "loss": 0.0051, "step": 36780 }, { "grad_norm": 0.1042947769165039, "learning_rate": 1.7902578345886507e-05, "loss": 0.0049, "step": 36790 }, { "grad_norm": 0.12187811732292175, "learning_rate": 1.7877229480643843e-05, "loss": 0.0055, "step": 36800 }, { "grad_norm": 0.10132759809494019, "learning_rate": 1.785189466699235e-05, "loss": 0.0039, "step": 36810 }, { "grad_norm": 0.17427724599838257, "learning_rate": 1.7826573916014317e-05, "loss": 0.0063, "step": 36820 }, { "grad_norm": 0.14001385867595673, "learning_rate": 1.7801267238785935e-05, "loss": 0.0071, "step": 36830 }, { "grad_norm": 0.1739109307527542, "learning_rate": 1.7775974646377187e-05, "loss": 0.0046, "step": 36840 }, { "grad_norm": 0.12634208798408508, "learning_rate": 1.775069614985189e-05, "loss": 0.0056, "step": 36850 }, { "grad_norm": 0.15979541838169098, "learning_rate": 1.7725431760267757e-05, "loss": 0.0057, "step": 36860 }, { "grad_norm": 0.12102089822292328, "learning_rate": 1.7700181488676277e-05, "loss": 0.0053, "step": 36870 }, { "grad_norm": 0.06289911270141602, "learning_rate": 1.767494534612275e-05, "loss": 0.005, "step": 36880 }, { "grad_norm": 0.1426238715648651, "learning_rate": 1.7649723343646367e-05, "loss": 0.004, "step": 36890 }, { "grad_norm": 0.14479610323905945, "learning_rate": 1.7624515492280063e-05, "loss": 0.0045, "step": 36900 }, { "grad_norm": 0.14452674984931946, "learning_rate": 1.7599321803050596e-05, "loss": 0.0084, "step": 36910 }, { "grad_norm": 0.12017496675252914, "learning_rate": 1.7574142286978574e-05, "loss": 0.0046, "step": 36920 }, { "grad_norm": 0.09172370284795761, "learning_rate": 1.7548976955078354e-05, "loss": 0.0038, "step": 36930 }, { "grad_norm": 0.21515795588493347, "learning_rate": 1.7523825818358092e-05, "loss": 0.006, "step": 36940 }, { "grad_norm": 0.16154401004314423, "learning_rate": 1.7498688887819785e-05, "loss": 0.0083, "step": 36950 }, { "grad_norm": 0.1096271350979805, "learning_rate": 1.7473566174459143e-05, "loss": 0.0088, "step": 36960 }, { "grad_norm": 0.17204268276691437, "learning_rate": 1.7448457689265735e-05, "loss": 0.0056, "step": 36970 }, { "grad_norm": 0.0768832117319107, "learning_rate": 1.7423363443222823e-05, "loss": 0.0045, "step": 36980 }, { "grad_norm": 0.14947646856307983, "learning_rate": 1.7398283447307524e-05, "loss": 0.0043, "step": 36990 }, { "grad_norm": 0.10147382318973541, "learning_rate": 1.737321771249066e-05, "loss": 0.0072, "step": 37000 }, { "grad_norm": 0.1514175534248352, "learning_rate": 1.7348166249736814e-05, "loss": 0.0055, "step": 37010 }, { "grad_norm": 0.0932316929101944, "learning_rate": 1.732312907000439e-05, "loss": 0.0047, "step": 37020 }, { "grad_norm": 0.14357200264930725, "learning_rate": 1.7298106184245478e-05, "loss": 0.0047, "step": 37030 }, { "grad_norm": 0.10588014125823975, "learning_rate": 1.727309760340592e-05, "loss": 0.0075, "step": 37040 }, { "grad_norm": 0.13895949721336365, "learning_rate": 1.724810333842536e-05, "loss": 0.0058, "step": 37050 }, { "grad_norm": 0.13330301642417908, "learning_rate": 1.7223123400237112e-05, "loss": 0.0066, "step": 37060 }, { "grad_norm": 0.08333956450223923, "learning_rate": 1.7198157799768238e-05, "loss": 0.0044, "step": 37070 }, { "grad_norm": 0.11831201612949371, "learning_rate": 1.7173206547939576e-05, "loss": 0.0046, "step": 37080 }, { "grad_norm": 0.1462329924106598, "learning_rate": 1.7148269655665628e-05, "loss": 0.0069, "step": 37090 }, { "grad_norm": 0.11955253779888153, "learning_rate": 1.712334713385463e-05, "loss": 0.0058, "step": 37100 }, { "grad_norm": 0.20091700553894043, "learning_rate": 1.709843899340856e-05, "loss": 0.0045, "step": 37110 }, { "grad_norm": 0.15097281336784363, "learning_rate": 1.7073545245223084e-05, "loss": 0.0041, "step": 37120 }, { "grad_norm": 0.09638073295354843, "learning_rate": 1.704866590018755e-05, "loss": 0.0047, "step": 37130 }, { "grad_norm": 0.09944487363100052, "learning_rate": 1.7023800969185054e-05, "loss": 0.0038, "step": 37140 }, { "grad_norm": 0.11814148724079132, "learning_rate": 1.6998950463092354e-05, "loss": 0.0049, "step": 37150 }, { "grad_norm": 0.12873774766921997, "learning_rate": 1.6974114392779907e-05, "loss": 0.0044, "step": 37160 }, { "grad_norm": 0.10112104564905167, "learning_rate": 1.694929276911183e-05, "loss": 0.0063, "step": 37170 }, { "grad_norm": 0.10188088566064835, "learning_rate": 1.6924485602945983e-05, "loss": 0.006, "step": 37180 }, { "grad_norm": 0.13601824641227722, "learning_rate": 1.6899692905133825e-05, "loss": 0.0082, "step": 37190 }, { "grad_norm": 0.1049005389213562, "learning_rate": 1.6874914686520565e-05, "loss": 0.0057, "step": 37200 }, { "grad_norm": 0.0940389484167099, "learning_rate": 1.6850150957945e-05, "loss": 0.0078, "step": 37210 }, { "grad_norm": 0.13369496166706085, "learning_rate": 1.682540173023966e-05, "loss": 0.0054, "step": 37220 }, { "grad_norm": 0.10775502026081085, "learning_rate": 1.6800667014230665e-05, "loss": 0.0068, "step": 37230 }, { "grad_norm": 0.15625184774398804, "learning_rate": 1.677594682073785e-05, "loss": 0.0039, "step": 37240 }, { "grad_norm": 0.12734733521938324, "learning_rate": 1.6751241160574653e-05, "loss": 0.0049, "step": 37250 }, { "grad_norm": 0.08227463811635971, "learning_rate": 1.6726550044548156e-05, "loss": 0.0044, "step": 37260 }, { "grad_norm": 0.2050698846578598, "learning_rate": 1.6701873483459125e-05, "loss": 0.0046, "step": 37270 }, { "grad_norm": 0.13732822239398956, "learning_rate": 1.6677211488101906e-05, "loss": 0.0063, "step": 37280 }, { "grad_norm": 0.10574778914451599, "learning_rate": 1.6652564069264475e-05, "loss": 0.0033, "step": 37290 }, { "grad_norm": 0.13121406733989716, "learning_rate": 1.662793123772849e-05, "loss": 0.005, "step": 37300 }, { "grad_norm": 0.09369825571775436, "learning_rate": 1.6603313004269168e-05, "loss": 0.0044, "step": 37310 }, { "grad_norm": 0.09712454676628113, "learning_rate": 1.657870937965536e-05, "loss": 0.0075, "step": 37320 }, { "grad_norm": 0.15048161149024963, "learning_rate": 1.655412037464952e-05, "loss": 0.0064, "step": 37330 }, { "grad_norm": 0.1740511804819107, "learning_rate": 1.652954600000773e-05, "loss": 0.0061, "step": 37340 }, { "grad_norm": 0.0868886187672615, "learning_rate": 1.650498626647965e-05, "loss": 0.007, "step": 37350 }, { "grad_norm": 0.0899910032749176, "learning_rate": 1.6480441184808526e-05, "loss": 0.0059, "step": 37360 }, { "grad_norm": 0.0760703757405281, "learning_rate": 1.6455910765731242e-05, "loss": 0.0047, "step": 37370 }, { "grad_norm": 0.1315239816904068, "learning_rate": 1.6431395019978217e-05, "loss": 0.0049, "step": 37380 }, { "grad_norm": 0.07496558874845505, "learning_rate": 1.6406893958273467e-05, "loss": 0.0062, "step": 37390 }, { "grad_norm": 0.12457121908664703, "learning_rate": 1.6382407591334602e-05, "loss": 0.0056, "step": 37400 }, { "grad_norm": 0.0840587466955185, "learning_rate": 1.635793592987279e-05, "loss": 0.0067, "step": 37410 }, { "grad_norm": 0.09696448594331741, "learning_rate": 1.633347898459275e-05, "loss": 0.0046, "step": 37420 }, { "grad_norm": 0.112822525203228, "learning_rate": 1.6309036766192805e-05, "loss": 0.0046, "step": 37430 }, { "grad_norm": 0.11487685143947601, "learning_rate": 1.628460928536478e-05, "loss": 0.0071, "step": 37440 }, { "grad_norm": 0.13965776562690735, "learning_rate": 1.6260196552794128e-05, "loss": 0.0049, "step": 37450 }, { "grad_norm": 0.09627663344144821, "learning_rate": 1.6235798579159766e-05, "loss": 0.0065, "step": 37460 }, { "grad_norm": 0.13434773683547974, "learning_rate": 1.6211415375134233e-05, "loss": 0.0064, "step": 37470 }, { "grad_norm": 0.1385190635919571, "learning_rate": 1.618704695138353e-05, "loss": 0.0053, "step": 37480 }, { "grad_norm": 0.11772424727678299, "learning_rate": 1.6162693318567273e-05, "loss": 0.0058, "step": 37490 }, { "grad_norm": 0.1575268805027008, "learning_rate": 1.613835448733856e-05, "loss": 0.0058, "step": 37500 }, { "grad_norm": 0.11068164557218552, "learning_rate": 1.6114030468344006e-05, "loss": 0.005, "step": 37510 }, { "grad_norm": 0.11896143108606339, "learning_rate": 1.6089721272223755e-05, "loss": 0.0064, "step": 37520 }, { "grad_norm": 0.18062567710876465, "learning_rate": 1.606542690961151e-05, "loss": 0.006, "step": 37530 }, { "grad_norm": 0.0953536108136177, "learning_rate": 1.6041147391134438e-05, "loss": 0.0052, "step": 37540 }, { "grad_norm": 0.09572085738182068, "learning_rate": 1.6016882727413196e-05, "loss": 0.0048, "step": 37550 }, { "grad_norm": 0.06458625942468643, "learning_rate": 1.5992632929062014e-05, "loss": 0.0049, "step": 37560 }, { "grad_norm": 0.10262694209814072, "learning_rate": 1.5968398006688557e-05, "loss": 0.0068, "step": 37570 }, { "grad_norm": 0.12517887353897095, "learning_rate": 1.594417797089399e-05, "loss": 0.0043, "step": 37580 }, { "grad_norm": 0.18705721199512482, "learning_rate": 1.5919972832273015e-05, "loss": 0.0048, "step": 37590 }, { "grad_norm": 0.09056536108255386, "learning_rate": 1.5895782601413767e-05, "loss": 0.0032, "step": 37600 }, { "grad_norm": 0.13295939564704895, "learning_rate": 1.587160728889785e-05, "loss": 0.0048, "step": 37610 }, { "grad_norm": 0.12326784431934357, "learning_rate": 1.58474469053004e-05, "loss": 0.0037, "step": 37620 }, { "grad_norm": 0.11854195594787598, "learning_rate": 1.5823301461189976e-05, "loss": 0.0042, "step": 37630 }, { "grad_norm": 0.13238069415092468, "learning_rate": 1.5799170967128606e-05, "loss": 0.0041, "step": 37640 }, { "grad_norm": 0.13475950062274933, "learning_rate": 1.5775055433671814e-05, "loss": 0.0057, "step": 37650 }, { "grad_norm": 0.060789935290813446, "learning_rate": 1.5750954871368535e-05, "loss": 0.0057, "step": 37660 }, { "grad_norm": 0.1291799247264862, "learning_rate": 1.5726869290761158e-05, "loss": 0.0063, "step": 37670 }, { "grad_norm": 0.15222537517547607, "learning_rate": 1.5702798702385574e-05, "loss": 0.0077, "step": 37680 }, { "grad_norm": 0.09563001245260239, "learning_rate": 1.5678743116771038e-05, "loss": 0.0061, "step": 37690 }, { "grad_norm": 0.061910200864076614, "learning_rate": 1.5654702544440313e-05, "loss": 0.0039, "step": 37700 }, { "grad_norm": 0.10267256945371628, "learning_rate": 1.5630676995909532e-05, "loss": 0.0054, "step": 37710 }, { "grad_norm": 0.0979897528886795, "learning_rate": 1.560666648168832e-05, "loss": 0.0047, "step": 37720 }, { "grad_norm": 0.06762511283159256, "learning_rate": 1.5582671012279663e-05, "loss": 0.0039, "step": 37730 }, { "grad_norm": 0.15329450368881226, "learning_rate": 1.5558690598179997e-05, "loss": 0.0048, "step": 37740 }, { "grad_norm": 0.06764054298400879, "learning_rate": 1.553472524987919e-05, "loss": 0.0047, "step": 37750 }, { "grad_norm": 0.06760740280151367, "learning_rate": 1.5510774977860485e-05, "loss": 0.0044, "step": 37760 }, { "grad_norm": 0.13680681586265564, "learning_rate": 1.5486839792600532e-05, "loss": 0.0036, "step": 37770 }, { "grad_norm": 0.1150963306427002, "learning_rate": 1.546291970456942e-05, "loss": 0.007, "step": 37780 }, { "grad_norm": 0.09556008130311966, "learning_rate": 1.5439014724230606e-05, "loss": 0.0038, "step": 37790 }, { "grad_norm": 0.14384309947490692, "learning_rate": 1.5415124862040907e-05, "loss": 0.0044, "step": 37800 }, { "grad_norm": 0.1406709998846054, "learning_rate": 1.53912501284506e-05, "loss": 0.0058, "step": 37810 }, { "grad_norm": 0.16014495491981506, "learning_rate": 1.5367390533903292e-05, "loss": 0.0058, "step": 37820 }, { "grad_norm": 0.1819959282875061, "learning_rate": 1.5343546088835964e-05, "loss": 0.0038, "step": 37830 }, { "grad_norm": 0.15117864310741425, "learning_rate": 1.531971680367901e-05, "loss": 0.0054, "step": 37840 }, { "grad_norm": 0.08564791083335876, "learning_rate": 1.529590268885616e-05, "loss": 0.0062, "step": 37850 }, { "grad_norm": 0.0668342188000679, "learning_rate": 1.5272103754784517e-05, "loss": 0.0041, "step": 37860 }, { "grad_norm": 0.12761473655700684, "learning_rate": 1.5248320011874523e-05, "loss": 0.0039, "step": 37870 }, { "grad_norm": 0.11815094202756882, "learning_rate": 1.5224551470530035e-05, "loss": 0.0042, "step": 37880 }, { "grad_norm": 0.0735870972275734, "learning_rate": 1.5200798141148193e-05, "loss": 0.0045, "step": 37890 }, { "grad_norm": 0.11850055307149887, "learning_rate": 1.5177060034119505e-05, "loss": 0.0037, "step": 37900 }, { "grad_norm": 0.08430856466293335, "learning_rate": 1.5153337159827835e-05, "loss": 0.0041, "step": 37910 }, { "grad_norm": 0.11714740097522736, "learning_rate": 1.5129629528650397e-05, "loss": 0.0042, "step": 37920 }, { "grad_norm": 0.12044750154018402, "learning_rate": 1.5105937150957671e-05, "loss": 0.0044, "step": 37930 }, { "grad_norm": 0.09648570418357849, "learning_rate": 1.5082260037113548e-05, "loss": 0.0042, "step": 37940 }, { "grad_norm": 0.11812524497509003, "learning_rate": 1.5058598197475187e-05, "loss": 0.004, "step": 37950 }, { "grad_norm": 0.08531465381383896, "learning_rate": 1.5034951642393064e-05, "loss": 0.0046, "step": 37960 }, { "grad_norm": 0.09053712338209152, "learning_rate": 1.5011320382211013e-05, "loss": 0.0052, "step": 37970 }, { "grad_norm": 0.11080653220415115, "learning_rate": 1.4987704427266136e-05, "loss": 0.0046, "step": 37980 }, { "grad_norm": 0.062652587890625, "learning_rate": 1.4964103787888839e-05, "loss": 0.0034, "step": 37990 }, { "grad_norm": 0.18443161249160767, "learning_rate": 1.4940518474402882e-05, "loss": 0.0067, "step": 38000 }, { "grad_norm": 0.1648959219455719, "learning_rate": 1.4916948497125249e-05, "loss": 0.0049, "step": 38010 }, { "grad_norm": 0.11861612647771835, "learning_rate": 1.4893393866366267e-05, "loss": 0.0049, "step": 38020 }, { "grad_norm": 0.08231619745492935, "learning_rate": 1.4869854592429506e-05, "loss": 0.0035, "step": 38030 }, { "grad_norm": 0.06489869952201843, "learning_rate": 1.4846330685611875e-05, "loss": 0.0041, "step": 38040 }, { "grad_norm": 0.1459994614124298, "learning_rate": 1.482282215620352e-05, "loss": 0.0051, "step": 38050 }, { "grad_norm": 0.11112802475690842, "learning_rate": 1.4799329014487857e-05, "loss": 0.0044, "step": 38060 }, { "grad_norm": 0.07169297337532043, "learning_rate": 1.4775851270741603e-05, "loss": 0.0036, "step": 38070 }, { "grad_norm": 0.19694244861602783, "learning_rate": 1.4752388935234719e-05, "loss": 0.005, "step": 38080 }, { "grad_norm": 0.1057288646697998, "learning_rate": 1.4728942018230401e-05, "loss": 0.0061, "step": 38090 }, { "grad_norm": 0.1850939244031906, "learning_rate": 1.4705510529985167e-05, "loss": 0.0076, "step": 38100 }, { "grad_norm": 0.08252188563346863, "learning_rate": 1.4682094480748726e-05, "loss": 0.0049, "step": 38110 }, { "grad_norm": 0.19355912506580353, "learning_rate": 1.4658693880764035e-05, "loss": 0.0057, "step": 38120 }, { "grad_norm": 0.15778371691703796, "learning_rate": 1.463530874026735e-05, "loss": 0.0066, "step": 38130 }, { "grad_norm": 0.0869639590382576, "learning_rate": 1.461193906948809e-05, "loss": 0.0035, "step": 38140 }, { "grad_norm": 0.15060779452323914, "learning_rate": 1.4588584878648975e-05, "loss": 0.004, "step": 38150 }, { "grad_norm": 0.11430522799491882, "learning_rate": 1.4565246177965896e-05, "loss": 0.005, "step": 38160 }, { "grad_norm": 0.14043205976486206, "learning_rate": 1.4541922977648015e-05, "loss": 0.0065, "step": 38170 }, { "grad_norm": 0.09348661452531815, "learning_rate": 1.451861528789768e-05, "loss": 0.0035, "step": 38180 }, { "grad_norm": 0.12339182943105698, "learning_rate": 1.4495323118910454e-05, "loss": 0.0052, "step": 38190 }, { "grad_norm": 0.123970165848732, "learning_rate": 1.4472046480875145e-05, "loss": 0.0043, "step": 38200 }, { "grad_norm": 0.08225594460964203, "learning_rate": 1.4448785383973745e-05, "loss": 0.0048, "step": 38210 }, { "grad_norm": 0.14001905918121338, "learning_rate": 1.4425539838381418e-05, "loss": 0.0042, "step": 38220 }, { "grad_norm": 0.09166620671749115, "learning_rate": 1.4402309854266594e-05, "loss": 0.0062, "step": 38230 }, { "grad_norm": 0.08644409477710724, "learning_rate": 1.4379095441790846e-05, "loss": 0.0042, "step": 38240 }, { "grad_norm": 0.14785800874233246, "learning_rate": 1.4355896611108921e-05, "loss": 0.0054, "step": 38250 }, { "grad_norm": 0.08684661239385605, "learning_rate": 1.433271337236881e-05, "loss": 0.0052, "step": 38260 }, { "grad_norm": 0.09989015758037567, "learning_rate": 1.4309545735711638e-05, "loss": 0.004, "step": 38270 }, { "grad_norm": 0.16299943625926971, "learning_rate": 1.4286393711271696e-05, "loss": 0.0045, "step": 38280 }, { "grad_norm": 0.12453996390104294, "learning_rate": 1.4263257309176497e-05, "loss": 0.0055, "step": 38290 }, { "grad_norm": 0.07762715965509415, "learning_rate": 1.4240136539546678e-05, "loss": 0.0048, "step": 38300 }, { "grad_norm": 0.09308665245771408, "learning_rate": 1.4217031412496029e-05, "loss": 0.0054, "step": 38310 }, { "grad_norm": 0.09124010056257248, "learning_rate": 1.4193941938131554e-05, "loss": 0.0045, "step": 38320 }, { "grad_norm": 0.11105766147375107, "learning_rate": 1.4170868126553356e-05, "loss": 0.0062, "step": 38330 }, { "grad_norm": 0.09806174039840698, "learning_rate": 1.4147809987854682e-05, "loss": 0.0079, "step": 38340 }, { "grad_norm": 0.12990885972976685, "learning_rate": 1.4124767532121985e-05, "loss": 0.0051, "step": 38350 }, { "grad_norm": 0.10705257207155228, "learning_rate": 1.41017407694348e-05, "loss": 0.0043, "step": 38360 }, { "grad_norm": 0.19210286438465118, "learning_rate": 1.4078729709865801e-05, "loss": 0.0062, "step": 38370 }, { "grad_norm": 0.17896047234535217, "learning_rate": 1.4055734363480833e-05, "loss": 0.0054, "step": 38380 }, { "grad_norm": 0.18029293417930603, "learning_rate": 1.4032754740338815e-05, "loss": 0.0041, "step": 38390 }, { "grad_norm": 0.1362297236919403, "learning_rate": 1.4009790850491844e-05, "loss": 0.0042, "step": 38400 }, { "grad_norm": 0.17958331108093262, "learning_rate": 1.3986842703985076e-05, "loss": 0.0079, "step": 38410 }, { "grad_norm": 0.11109627783298492, "learning_rate": 1.3963910310856842e-05, "loss": 0.0058, "step": 38420 }, { "grad_norm": 0.11841098964214325, "learning_rate": 1.394099368113853e-05, "loss": 0.0054, "step": 38430 }, { "grad_norm": 0.14389263093471527, "learning_rate": 1.3918092824854644e-05, "loss": 0.0041, "step": 38440 }, { "grad_norm": 0.11343884468078613, "learning_rate": 1.3895207752022821e-05, "loss": 0.0046, "step": 38450 }, { "grad_norm": 0.08708581328392029, "learning_rate": 1.3872338472653756e-05, "loss": 0.0062, "step": 38460 }, { "grad_norm": 0.10411164909601212, "learning_rate": 1.3849484996751233e-05, "loss": 0.0064, "step": 38470 }, { "grad_norm": 0.08140737563371658, "learning_rate": 1.382664733431217e-05, "loss": 0.0052, "step": 38480 }, { "grad_norm": 0.09714365005493164, "learning_rate": 1.3803825495326522e-05, "loss": 0.0057, "step": 38490 }, { "grad_norm": 0.08678720891475677, "learning_rate": 1.3781019489777324e-05, "loss": 0.0065, "step": 38500 }, { "grad_norm": 0.10036497563123703, "learning_rate": 1.375822932764072e-05, "loss": 0.0064, "step": 38510 }, { "grad_norm": 0.09586604684591293, "learning_rate": 1.373545501888589e-05, "loss": 0.0042, "step": 38520 }, { "grad_norm": 0.09163336455821991, "learning_rate": 1.3712696573475092e-05, "loss": 0.0045, "step": 38530 }, { "grad_norm": 0.07474157214164734, "learning_rate": 1.368995400136363e-05, "loss": 0.0067, "step": 38540 }, { "grad_norm": 0.11704439669847488, "learning_rate": 1.3667227312499903e-05, "loss": 0.0066, "step": 38550 }, { "grad_norm": 0.08758202940225601, "learning_rate": 1.3644516516825323e-05, "loss": 0.0044, "step": 38560 }, { "grad_norm": 0.11179694533348083, "learning_rate": 1.3621821624274356e-05, "loss": 0.005, "step": 38570 }, { "grad_norm": 0.1074770838022232, "learning_rate": 1.3599142644774537e-05, "loss": 0.0045, "step": 38580 }, { "grad_norm": 0.1389814019203186, "learning_rate": 1.357647958824641e-05, "loss": 0.006, "step": 38590 }, { "grad_norm": 0.08833569288253784, "learning_rate": 1.3553832464603555e-05, "loss": 0.0051, "step": 38600 }, { "grad_norm": 0.11645843088626862, "learning_rate": 1.3531201283752625e-05, "loss": 0.0055, "step": 38610 }, { "grad_norm": 0.16623106598854065, "learning_rate": 1.350858605559323e-05, "loss": 0.0079, "step": 38620 }, { "grad_norm": 0.09016714245080948, "learning_rate": 1.348598679001808e-05, "loss": 0.0044, "step": 38630 }, { "grad_norm": 0.12094587832689285, "learning_rate": 1.3463403496912819e-05, "loss": 0.0049, "step": 38640 }, { "grad_norm": 0.11157714575529099, "learning_rate": 1.3440836186156187e-05, "loss": 0.0057, "step": 38650 }, { "grad_norm": 0.1422162801027298, "learning_rate": 1.3418284867619852e-05, "loss": 0.0055, "step": 38660 }, { "grad_norm": 0.12196248769760132, "learning_rate": 1.3395749551168567e-05, "loss": 0.004, "step": 38670 }, { "grad_norm": 0.093611940741539, "learning_rate": 1.3373230246660024e-05, "loss": 0.0054, "step": 38680 }, { "grad_norm": 0.10657846182584763, "learning_rate": 1.335072696394492e-05, "loss": 0.0045, "step": 38690 }, { "grad_norm": 0.07851336896419525, "learning_rate": 1.3328239712866986e-05, "loss": 0.0035, "step": 38700 }, { "grad_norm": 0.0711018368601799, "learning_rate": 1.3305768503262889e-05, "loss": 0.0047, "step": 38710 }, { "grad_norm": 0.07999423891305923, "learning_rate": 1.3283313344962305e-05, "loss": 0.005, "step": 38720 }, { "grad_norm": 0.1906633824110031, "learning_rate": 1.3260874247787863e-05, "loss": 0.0057, "step": 38730 }, { "grad_norm": 0.11454883962869644, "learning_rate": 1.3238451221555226e-05, "loss": 0.0051, "step": 38740 }, { "grad_norm": 0.18745696544647217, "learning_rate": 1.3216044276072965e-05, "loss": 0.0064, "step": 38750 }, { "grad_norm": 0.13018538057804108, "learning_rate": 1.3193653421142622e-05, "loss": 0.004, "step": 38760 }, { "grad_norm": 0.14640365540981293, "learning_rate": 1.3171278666558756e-05, "loss": 0.0063, "step": 38770 }, { "grad_norm": 0.12228034436702728, "learning_rate": 1.3148920022108819e-05, "loss": 0.0045, "step": 38780 }, { "grad_norm": 0.11348465830087662, "learning_rate": 1.312657749757324e-05, "loss": 0.0039, "step": 38790 }, { "grad_norm": 0.06340811401605606, "learning_rate": 1.3104251102725412e-05, "loss": 0.0053, "step": 38800 }, { "grad_norm": 0.11035912483930588, "learning_rate": 1.3081940847331659e-05, "loss": 0.0055, "step": 38810 }, { "grad_norm": 0.0763712078332901, "learning_rate": 1.3059646741151222e-05, "loss": 0.0055, "step": 38820 }, { "grad_norm": 0.16599734127521515, "learning_rate": 1.3037368793936334e-05, "loss": 0.0048, "step": 38830 }, { "grad_norm": 0.06403690576553345, "learning_rate": 1.3015107015432104e-05, "loss": 0.0054, "step": 38840 }, { "grad_norm": 0.09747210890054703, "learning_rate": 1.2992861415376584e-05, "loss": 0.0053, "step": 38850 }, { "grad_norm": 0.09489373117685318, "learning_rate": 1.2970632003500782e-05, "loss": 0.0074, "step": 38860 }, { "grad_norm": 0.19747419655323029, "learning_rate": 1.294841878952856e-05, "loss": 0.0051, "step": 38870 }, { "grad_norm": 0.1969018578529358, "learning_rate": 1.2926221783176779e-05, "loss": 0.0051, "step": 38880 }, { "grad_norm": 0.12793155014514923, "learning_rate": 1.290404099415512e-05, "loss": 0.0048, "step": 38890 }, { "grad_norm": 0.10087696462869644, "learning_rate": 1.2881876432166246e-05, "loss": 0.006, "step": 38900 }, { "grad_norm": 0.07866119593381882, "learning_rate": 1.2859728106905678e-05, "loss": 0.0048, "step": 38910 }, { "grad_norm": 0.11441414803266525, "learning_rate": 1.2837596028061832e-05, "loss": 0.0044, "step": 38920 }, { "grad_norm": 0.08069837093353271, "learning_rate": 1.2815480205316055e-05, "loss": 0.0036, "step": 38930 }, { "grad_norm": 0.08354388922452927, "learning_rate": 1.279338064834255e-05, "loss": 0.0048, "step": 38940 }, { "grad_norm": 0.16285890340805054, "learning_rate": 1.27712973668084e-05, "loss": 0.0062, "step": 38950 }, { "grad_norm": 0.11082272976636887, "learning_rate": 1.2749230370373605e-05, "loss": 0.0047, "step": 38960 }, { "grad_norm": 0.13486243784427643, "learning_rate": 1.2727179668691009e-05, "loss": 0.0046, "step": 38970 }, { "grad_norm": 0.1265772581100464, "learning_rate": 1.2705145271406327e-05, "loss": 0.0049, "step": 38980 }, { "grad_norm": 0.07502716779708862, "learning_rate": 1.2683127188158172e-05, "loss": 0.0055, "step": 38990 }, { "grad_norm": 0.13626955449581146, "learning_rate": 1.2661125428577997e-05, "loss": 0.0048, "step": 39000 }, { "grad_norm": 0.1763816922903061, "learning_rate": 1.2639140002290101e-05, "loss": 0.0057, "step": 39010 }, { "grad_norm": 0.08281766623258591, "learning_rate": 1.261717091891168e-05, "loss": 0.0045, "step": 39020 }, { "grad_norm": 0.053243253380060196, "learning_rate": 1.2595218188052755e-05, "loss": 0.0055, "step": 39030 }, { "grad_norm": 0.16343140602111816, "learning_rate": 1.2573281819316174e-05, "loss": 0.0063, "step": 39040 }, { "grad_norm": 0.10122862458229065, "learning_rate": 1.2551361822297674e-05, "loss": 0.0035, "step": 39050 }, { "grad_norm": 0.07717835158109665, "learning_rate": 1.2529458206585804e-05, "loss": 0.0061, "step": 39060 }, { "grad_norm": 0.15549367666244507, "learning_rate": 1.250757098176194e-05, "loss": 0.0062, "step": 39070 }, { "grad_norm": 0.09871308505535126, "learning_rate": 1.2485700157400294e-05, "loss": 0.0035, "step": 39080 }, { "grad_norm": 0.09467627853155136, "learning_rate": 1.246384574306792e-05, "loss": 0.0044, "step": 39090 }, { "grad_norm": 0.06395968049764633, "learning_rate": 1.2442007748324664e-05, "loss": 0.0058, "step": 39100 }, { "grad_norm": 0.14825424551963806, "learning_rate": 1.2420186182723237e-05, "loss": 0.0033, "step": 39110 }, { "grad_norm": 0.13562481105327606, "learning_rate": 1.23983810558091e-05, "loss": 0.0069, "step": 39120 }, { "grad_norm": 0.08121367543935776, "learning_rate": 1.2376592377120582e-05, "loss": 0.0064, "step": 39130 }, { "grad_norm": 0.1501750946044922, "learning_rate": 1.2354820156188768e-05, "loss": 0.0057, "step": 39140 }, { "grad_norm": 0.08120913803577423, "learning_rate": 1.2333064402537597e-05, "loss": 0.0035, "step": 39150 }, { "grad_norm": 0.12833070755004883, "learning_rate": 1.2311325125683754e-05, "loss": 0.0039, "step": 39160 }, { "grad_norm": 0.05129116028547287, "learning_rate": 1.228960233513673e-05, "loss": 0.0037, "step": 39170 }, { "grad_norm": 0.05639931187033653, "learning_rate": 1.2267896040398836e-05, "loss": 0.0033, "step": 39180 }, { "grad_norm": 0.09174815565347672, "learning_rate": 1.2246206250965125e-05, "loss": 0.0054, "step": 39190 }, { "grad_norm": 0.10790887475013733, "learning_rate": 1.2224532976323444e-05, "loss": 0.0041, "step": 39200 }, { "grad_norm": 0.11204113066196442, "learning_rate": 1.220287622595444e-05, "loss": 0.0049, "step": 39210 }, { "grad_norm": 0.08481264114379883, "learning_rate": 1.2181236009331497e-05, "loss": 0.0049, "step": 39220 }, { "grad_norm": 0.11370692402124405, "learning_rate": 1.2159612335920784e-05, "loss": 0.0042, "step": 39230 }, { "grad_norm": 0.19396774470806122, "learning_rate": 1.2138005215181219e-05, "loss": 0.0083, "step": 39240 }, { "grad_norm": 0.07509177178144455, "learning_rate": 1.2116414656564512e-05, "loss": 0.0032, "step": 39250 }, { "grad_norm": 0.08673141151666641, "learning_rate": 1.2094840669515095e-05, "loss": 0.0034, "step": 39260 }, { "grad_norm": 0.13223159313201904, "learning_rate": 1.2073283263470152e-05, "loss": 0.0038, "step": 39270 }, { "grad_norm": 0.12556444108486176, "learning_rate": 1.2051742447859653e-05, "loss": 0.0064, "step": 39280 }, { "grad_norm": 0.13173574209213257, "learning_rate": 1.203021823210626e-05, "loss": 0.007, "step": 39290 }, { "grad_norm": 0.1543571799993515, "learning_rate": 1.2008710625625397e-05, "loss": 0.0043, "step": 39300 }, { "grad_norm": 0.12127561122179031, "learning_rate": 1.1987219637825236e-05, "loss": 0.0058, "step": 39310 }, { "grad_norm": 0.13770323991775513, "learning_rate": 1.1965745278106656e-05, "loss": 0.0049, "step": 39320 }, { "grad_norm": 0.04483555629849434, "learning_rate": 1.1944287555863259e-05, "loss": 0.0063, "step": 39330 }, { "grad_norm": 0.06032385304570198, "learning_rate": 1.1922846480481409e-05, "loss": 0.0032, "step": 39340 }, { "grad_norm": 0.1464250236749649, "learning_rate": 1.1901422061340133e-05, "loss": 0.0056, "step": 39350 }, { "grad_norm": 0.07771863788366318, "learning_rate": 1.1880014307811227e-05, "loss": 0.0053, "step": 39360 }, { "grad_norm": 0.09136692434549332, "learning_rate": 1.1858623229259142e-05, "loss": 0.004, "step": 39370 }, { "grad_norm": 0.137529656291008, "learning_rate": 1.1837248835041092e-05, "loss": 0.0048, "step": 39380 }, { "grad_norm": 0.07865653187036514, "learning_rate": 1.1815891134506935e-05, "loss": 0.0042, "step": 39390 }, { "grad_norm": 0.10815715789794922, "learning_rate": 1.1794550136999277e-05, "loss": 0.0033, "step": 39400 }, { "grad_norm": 0.1912461817264557, "learning_rate": 1.1773225851853387e-05, "loss": 0.0041, "step": 39410 }, { "grad_norm": 0.09588003903627396, "learning_rate": 1.1751918288397235e-05, "loss": 0.0045, "step": 39420 }, { "grad_norm": 0.14704681932926178, "learning_rate": 1.1730627455951449e-05, "loss": 0.0054, "step": 39430 }, { "grad_norm": 0.08976522088050842, "learning_rate": 1.17093533638294e-05, "loss": 0.0058, "step": 39440 }, { "grad_norm": 0.0574084110558033, "learning_rate": 1.168809602133708e-05, "loss": 0.0048, "step": 39450 }, { "grad_norm": 0.1954418122768402, "learning_rate": 1.1666855437773162e-05, "loss": 0.0045, "step": 39460 }, { "grad_norm": 0.09084783494472504, "learning_rate": 1.1645631622429026e-05, "loss": 0.004, "step": 39470 }, { "grad_norm": 0.06856070458889008, "learning_rate": 1.1624424584588673e-05, "loss": 0.0041, "step": 39480 }, { "grad_norm": 0.10831570625305176, "learning_rate": 1.1603234333528768e-05, "loss": 0.005, "step": 39490 }, { "grad_norm": 0.09703007340431213, "learning_rate": 1.1582060878518674e-05, "loss": 0.0041, "step": 39500 }, { "grad_norm": 0.16222110390663147, "learning_rate": 1.1560904228820375e-05, "loss": 0.0087, "step": 39510 }, { "grad_norm": 0.12753687798976898, "learning_rate": 1.1539764393688478e-05, "loss": 0.0059, "step": 39520 }, { "grad_norm": 0.09119443595409393, "learning_rate": 1.151864138237031e-05, "loss": 0.0034, "step": 39530 }, { "grad_norm": 0.09505545347929001, "learning_rate": 1.1497535204105769e-05, "loss": 0.0043, "step": 39540 }, { "grad_norm": 0.09377448260784149, "learning_rate": 1.14764458681274e-05, "loss": 0.0041, "step": 39550 }, { "grad_norm": 0.05275110900402069, "learning_rate": 1.1455373383660434e-05, "loss": 0.0037, "step": 39560 }, { "grad_norm": 0.07693563401699066, "learning_rate": 1.1434317759922664e-05, "loss": 0.0036, "step": 39570 }, { "grad_norm": 0.11868257075548172, "learning_rate": 1.1413279006124528e-05, "loss": 0.0059, "step": 39580 }, { "grad_norm": 0.12015243619680405, "learning_rate": 1.1392257131469119e-05, "loss": 0.0051, "step": 39590 }, { "grad_norm": 0.14458253979682922, "learning_rate": 1.1371252145152095e-05, "loss": 0.0037, "step": 39600 }, { "grad_norm": 0.08844292908906937, "learning_rate": 1.1350264056361776e-05, "loss": 0.0052, "step": 39610 }, { "grad_norm": 0.15725857019424438, "learning_rate": 1.1329292874279029e-05, "loss": 0.0036, "step": 39620 }, { "grad_norm": 0.09098515659570694, "learning_rate": 1.1308338608077401e-05, "loss": 0.0038, "step": 39630 }, { "grad_norm": 0.07918034493923187, "learning_rate": 1.1287401266922981e-05, "loss": 0.0036, "step": 39640 }, { "grad_norm": 0.06544831395149231, "learning_rate": 1.126648085997446e-05, "loss": 0.0032, "step": 39650 }, { "grad_norm": 0.11387473344802856, "learning_rate": 1.124557739638316e-05, "loss": 0.0038, "step": 39660 }, { "grad_norm": 0.15392811596393585, "learning_rate": 1.1224690885292955e-05, "loss": 0.0041, "step": 39670 }, { "grad_norm": 0.11467603594064713, "learning_rate": 1.1203821335840303e-05, "loss": 0.0046, "step": 39680 }, { "grad_norm": 0.09804652631282806, "learning_rate": 1.1182968757154278e-05, "loss": 0.0043, "step": 39690 }, { "grad_norm": 0.08341359347105026, "learning_rate": 1.1162133158356492e-05, "loss": 0.0047, "step": 39700 }, { "grad_norm": 0.08583025634288788, "learning_rate": 1.1141314548561133e-05, "loss": 0.0056, "step": 39710 }, { "grad_norm": 0.09407664835453033, "learning_rate": 1.1120512936874994e-05, "loss": 0.004, "step": 39720 }, { "grad_norm": 0.19505669176578522, "learning_rate": 1.109972833239739e-05, "loss": 0.0048, "step": 39730 }, { "grad_norm": 0.07876951992511749, "learning_rate": 1.1078960744220218e-05, "loss": 0.0073, "step": 39740 }, { "grad_norm": 0.0844556987285614, "learning_rate": 1.1058210181427914e-05, "loss": 0.0036, "step": 39750 }, { "grad_norm": 0.132882222533226, "learning_rate": 1.10374766530975e-05, "loss": 0.0044, "step": 39760 }, { "grad_norm": 0.13872261345386505, "learning_rate": 1.1016760168298518e-05, "loss": 0.0043, "step": 39770 }, { "grad_norm": 0.06947003304958344, "learning_rate": 1.0996060736093044e-05, "loss": 0.0046, "step": 39780 }, { "grad_norm": 0.11032518744468689, "learning_rate": 1.0975378365535749e-05, "loss": 0.0061, "step": 39790 }, { "grad_norm": 0.08696574717760086, "learning_rate": 1.0954713065673777e-05, "loss": 0.0052, "step": 39800 }, { "grad_norm": 0.1165047287940979, "learning_rate": 1.0934064845546832e-05, "loss": 0.0045, "step": 39810 }, { "grad_norm": 0.09750648587942123, "learning_rate": 1.0913433714187166e-05, "loss": 0.0057, "step": 39820 }, { "grad_norm": 0.1144886314868927, "learning_rate": 1.089281968061951e-05, "loss": 0.0044, "step": 39830 }, { "grad_norm": 0.15474142134189606, "learning_rate": 1.0872222753861172e-05, "loss": 0.0048, "step": 39840 }, { "grad_norm": 0.11221309006214142, "learning_rate": 1.0851642942921925e-05, "loss": 0.0061, "step": 39850 }, { "grad_norm": 0.08756054192781448, "learning_rate": 1.0831080256804093e-05, "loss": 0.0043, "step": 39860 }, { "grad_norm": 0.09257573634386063, "learning_rate": 1.0810534704502478e-05, "loss": 0.004, "step": 39870 }, { "grad_norm": 0.08835425227880478, "learning_rate": 1.0790006295004423e-05, "loss": 0.0041, "step": 39880 }, { "grad_norm": 0.1479598581790924, "learning_rate": 1.076949503728974e-05, "loss": 0.0066, "step": 39890 }, { "grad_norm": 0.0981711819767952, "learning_rate": 1.074900094033074e-05, "loss": 0.0057, "step": 39900 }, { "grad_norm": 0.15275788307189941, "learning_rate": 1.0728524013092256e-05, "loss": 0.0051, "step": 39910 }, { "grad_norm": 0.06904864311218262, "learning_rate": 1.0708064264531581e-05, "loss": 0.0052, "step": 39920 }, { "grad_norm": 0.09652280807495117, "learning_rate": 1.068762170359851e-05, "loss": 0.0064, "step": 39930 }, { "grad_norm": 0.09056758135557175, "learning_rate": 1.0667196339235297e-05, "loss": 0.0037, "step": 39940 }, { "grad_norm": 0.04544631391763687, "learning_rate": 1.0646788180376716e-05, "loss": 0.003, "step": 39950 }, { "grad_norm": 0.15917742252349854, "learning_rate": 1.0626397235949976e-05, "loss": 0.0052, "step": 39960 }, { "grad_norm": 0.10801336169242859, "learning_rate": 1.0606023514874751e-05, "loss": 0.0032, "step": 39970 }, { "grad_norm": 0.07901004701852798, "learning_rate": 1.0585667026063233e-05, "loss": 0.0043, "step": 39980 }, { "grad_norm": 0.10070598870515823, "learning_rate": 1.0565327778420025e-05, "loss": 0.0034, "step": 39990 }, { "grad_norm": 0.11555691808462143, "learning_rate": 1.0545005780842193e-05, "loss": 0.0035, "step": 40000 } ], "logging_steps": 10, "max_steps": 50000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 48, "trial_name": null, "trial_params": null }