{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.95, "eval_steps": 500, "global_step": 47500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 3.546887159347534, "learning_rate": 3.6e-07, "loss": 0.4304, "step": 10 }, { "grad_norm": 3.7994141578674316, "learning_rate": 7.6e-07, "loss": 0.4576, "step": 20 }, { "grad_norm": 3.498992443084717, "learning_rate": 1.16e-06, "loss": 0.4086, "step": 30 }, { "grad_norm": 3.643303155899048, "learning_rate": 1.56e-06, "loss": 0.3905, "step": 40 }, { "grad_norm": 3.060072898864746, "learning_rate": 1.96e-06, "loss": 0.3135, "step": 50 }, { "grad_norm": 1.780480146408081, "learning_rate": 2.36e-06, "loss": 0.2512, "step": 60 }, { "grad_norm": 0.9976758360862732, "learning_rate": 2.7600000000000003e-06, "loss": 0.1636, "step": 70 }, { "grad_norm": 0.5895625352859497, "learning_rate": 3.1600000000000007e-06, "loss": 0.1143, "step": 80 }, { "grad_norm": 0.5999687910079956, "learning_rate": 3.5600000000000002e-06, "loss": 0.1053, "step": 90 }, { "grad_norm": 0.5172633528709412, "learning_rate": 3.96e-06, "loss": 0.093, "step": 100 }, { "grad_norm": 0.5282970666885376, "learning_rate": 4.360000000000001e-06, "loss": 0.0794, "step": 110 }, { "grad_norm": 0.4997275769710541, "learning_rate": 4.76e-06, "loss": 0.0757, "step": 120 }, { "grad_norm": 0.34583181142807007, "learning_rate": 5.1600000000000006e-06, "loss": 0.0628, "step": 130 }, { "grad_norm": 0.4450480043888092, "learning_rate": 5.56e-06, "loss": 0.0679, "step": 140 }, { "grad_norm": 0.4181276857852936, "learning_rate": 5.9600000000000005e-06, "loss": 0.0569, "step": 150 }, { "grad_norm": 0.3868604600429535, "learning_rate": 6.360000000000001e-06, "loss": 0.0537, "step": 160 }, { "grad_norm": 0.3434831202030182, "learning_rate": 6.76e-06, "loss": 0.0487, "step": 170 }, { "grad_norm": 0.2950899600982666, "learning_rate": 7.16e-06, "loss": 0.0569, "step": 180 }, { "grad_norm": 0.572747528553009, "learning_rate": 7.5600000000000005e-06, "loss": 0.049, "step": 190 }, { "grad_norm": 0.7449966669082642, "learning_rate": 7.96e-06, "loss": 0.0482, "step": 200 }, { "grad_norm": 0.4956739544868469, "learning_rate": 8.36e-06, "loss": 0.043, "step": 210 }, { "grad_norm": 0.23666192591190338, "learning_rate": 8.76e-06, "loss": 0.046, "step": 220 }, { "grad_norm": 0.3957290053367615, "learning_rate": 9.16e-06, "loss": 0.0456, "step": 230 }, { "grad_norm": 0.3977077305316925, "learning_rate": 9.560000000000002e-06, "loss": 0.0559, "step": 240 }, { "grad_norm": 0.3658420741558075, "learning_rate": 9.96e-06, "loss": 0.0436, "step": 250 }, { "grad_norm": 0.49545809626579285, "learning_rate": 1.036e-05, "loss": 0.042, "step": 260 }, { "grad_norm": 0.3103584349155426, "learning_rate": 1.076e-05, "loss": 0.0408, "step": 270 }, { "grad_norm": 0.2993907630443573, "learning_rate": 1.1160000000000002e-05, "loss": 0.041, "step": 280 }, { "grad_norm": 0.3652917444705963, "learning_rate": 1.156e-05, "loss": 0.0449, "step": 290 }, { "grad_norm": 0.329866498708725, "learning_rate": 1.196e-05, "loss": 0.0439, "step": 300 }, { "grad_norm": 0.2595917284488678, "learning_rate": 1.236e-05, "loss": 0.0451, "step": 310 }, { "grad_norm": 0.36765289306640625, "learning_rate": 1.276e-05, "loss": 0.0403, "step": 320 }, { "grad_norm": 0.48923471570014954, "learning_rate": 1.316e-05, "loss": 0.039, "step": 330 }, { "grad_norm": 0.2769504189491272, "learning_rate": 1.356e-05, "loss": 0.0381, "step": 340 }, { "grad_norm": 0.3513057827949524, "learning_rate": 1.396e-05, "loss": 0.0354, "step": 350 }, { "grad_norm": 0.2992415726184845, "learning_rate": 1.4360000000000001e-05, "loss": 0.0386, "step": 360 }, { "grad_norm": 0.36486807465553284, "learning_rate": 1.4760000000000001e-05, "loss": 0.0358, "step": 370 }, { "grad_norm": 0.30597713589668274, "learning_rate": 1.5160000000000002e-05, "loss": 0.042, "step": 380 }, { "grad_norm": 0.435866117477417, "learning_rate": 1.556e-05, "loss": 0.0402, "step": 390 }, { "grad_norm": 0.35244303941726685, "learning_rate": 1.596e-05, "loss": 0.0405, "step": 400 }, { "grad_norm": 0.23862473666667938, "learning_rate": 1.636e-05, "loss": 0.0367, "step": 410 }, { "grad_norm": 0.3165937662124634, "learning_rate": 1.6760000000000002e-05, "loss": 0.0319, "step": 420 }, { "grad_norm": 0.6266329884529114, "learning_rate": 1.7160000000000002e-05, "loss": 0.041, "step": 430 }, { "grad_norm": 0.3811008036136627, "learning_rate": 1.756e-05, "loss": 0.0386, "step": 440 }, { "grad_norm": 0.28754332661628723, "learning_rate": 1.796e-05, "loss": 0.0357, "step": 450 }, { "grad_norm": 0.3561721444129944, "learning_rate": 1.8360000000000004e-05, "loss": 0.0382, "step": 460 }, { "grad_norm": 0.2620997428894043, "learning_rate": 1.876e-05, "loss": 0.0409, "step": 470 }, { "grad_norm": 0.22939203679561615, "learning_rate": 1.916e-05, "loss": 0.0397, "step": 480 }, { "grad_norm": 0.4310348331928253, "learning_rate": 1.956e-05, "loss": 0.0498, "step": 490 }, { "grad_norm": 0.47144097089767456, "learning_rate": 1.9960000000000002e-05, "loss": 0.0467, "step": 500 }, { "grad_norm": 0.35166212916374207, "learning_rate": 2.036e-05, "loss": 0.0469, "step": 510 }, { "grad_norm": 0.471002459526062, "learning_rate": 2.076e-05, "loss": 0.0466, "step": 520 }, { "grad_norm": 0.39209192991256714, "learning_rate": 2.116e-05, "loss": 0.0494, "step": 530 }, { "grad_norm": 0.35149291157722473, "learning_rate": 2.1560000000000004e-05, "loss": 0.0435, "step": 540 }, { "grad_norm": 0.3114643394947052, "learning_rate": 2.196e-05, "loss": 0.0466, "step": 550 }, { "grad_norm": 0.45628663897514343, "learning_rate": 2.236e-05, "loss": 0.0436, "step": 560 }, { "grad_norm": 0.3895990550518036, "learning_rate": 2.2760000000000002e-05, "loss": 0.0405, "step": 570 }, { "grad_norm": 0.22049616277217865, "learning_rate": 2.3160000000000002e-05, "loss": 0.0372, "step": 580 }, { "grad_norm": 0.19829969108104706, "learning_rate": 2.356e-05, "loss": 0.0386, "step": 590 }, { "grad_norm": 0.2223646342754364, "learning_rate": 2.396e-05, "loss": 0.0445, "step": 600 }, { "grad_norm": 0.24074546992778778, "learning_rate": 2.4360000000000004e-05, "loss": 0.0397, "step": 610 }, { "grad_norm": 0.3903791308403015, "learning_rate": 2.476e-05, "loss": 0.0368, "step": 620 }, { "grad_norm": 0.3028348386287689, "learning_rate": 2.516e-05, "loss": 0.0396, "step": 630 }, { "grad_norm": 0.34927648305892944, "learning_rate": 2.556e-05, "loss": 0.0336, "step": 640 }, { "grad_norm": 0.27082231640815735, "learning_rate": 2.5960000000000002e-05, "loss": 0.0378, "step": 650 }, { "grad_norm": 0.3758169114589691, "learning_rate": 2.6360000000000002e-05, "loss": 0.0474, "step": 660 }, { "grad_norm": 0.257578045129776, "learning_rate": 2.676e-05, "loss": 0.0335, "step": 670 }, { "grad_norm": 0.2775658965110779, "learning_rate": 2.716e-05, "loss": 0.0405, "step": 680 }, { "grad_norm": 0.364395409822464, "learning_rate": 2.7560000000000004e-05, "loss": 0.0369, "step": 690 }, { "grad_norm": 0.18235783278942108, "learning_rate": 2.7960000000000003e-05, "loss": 0.0353, "step": 700 }, { "grad_norm": 0.32438424229621887, "learning_rate": 2.8360000000000003e-05, "loss": 0.0381, "step": 710 }, { "grad_norm": 0.38805416226387024, "learning_rate": 2.8760000000000002e-05, "loss": 0.0385, "step": 720 }, { "grad_norm": 0.36651045083999634, "learning_rate": 2.9160000000000005e-05, "loss": 0.0387, "step": 730 }, { "grad_norm": 0.30540627241134644, "learning_rate": 2.9559999999999998e-05, "loss": 0.031, "step": 740 }, { "grad_norm": 0.31893932819366455, "learning_rate": 2.9959999999999998e-05, "loss": 0.0402, "step": 750 }, { "grad_norm": 0.22543233633041382, "learning_rate": 3.036e-05, "loss": 0.0376, "step": 760 }, { "grad_norm": 0.40267693996429443, "learning_rate": 3.076e-05, "loss": 0.0367, "step": 770 }, { "grad_norm": 0.5144368410110474, "learning_rate": 3.116e-05, "loss": 0.0435, "step": 780 }, { "grad_norm": 0.3178098499774933, "learning_rate": 3.156e-05, "loss": 0.0326, "step": 790 }, { "grad_norm": 0.23506543040275574, "learning_rate": 3.196e-05, "loss": 0.0308, "step": 800 }, { "grad_norm": 0.20743806660175323, "learning_rate": 3.236e-05, "loss": 0.0348, "step": 810 }, { "grad_norm": 0.355976402759552, "learning_rate": 3.2760000000000005e-05, "loss": 0.033, "step": 820 }, { "grad_norm": 0.2427765130996704, "learning_rate": 3.316e-05, "loss": 0.0437, "step": 830 }, { "grad_norm": 0.20342418551445007, "learning_rate": 3.3560000000000004e-05, "loss": 0.0377, "step": 840 }, { "grad_norm": 0.4396277368068695, "learning_rate": 3.396e-05, "loss": 0.0375, "step": 850 }, { "grad_norm": 0.25973987579345703, "learning_rate": 3.436e-05, "loss": 0.0331, "step": 860 }, { "grad_norm": 0.3297395408153534, "learning_rate": 3.4760000000000006e-05, "loss": 0.042, "step": 870 }, { "grad_norm": 0.2635456919670105, "learning_rate": 3.516e-05, "loss": 0.0305, "step": 880 }, { "grad_norm": 0.263938307762146, "learning_rate": 3.5560000000000005e-05, "loss": 0.0344, "step": 890 }, { "grad_norm": 0.22394773364067078, "learning_rate": 3.596e-05, "loss": 0.0328, "step": 900 }, { "grad_norm": 0.30775463581085205, "learning_rate": 3.636e-05, "loss": 0.0302, "step": 910 }, { "grad_norm": 0.1720799207687378, "learning_rate": 3.676e-05, "loss": 0.0277, "step": 920 }, { "grad_norm": 0.29309728741645813, "learning_rate": 3.716e-05, "loss": 0.0384, "step": 930 }, { "grad_norm": 0.28079354763031006, "learning_rate": 3.756e-05, "loss": 0.0368, "step": 940 }, { "grad_norm": 0.2308836728334427, "learning_rate": 3.796e-05, "loss": 0.0295, "step": 950 }, { "grad_norm": 0.3147638142108917, "learning_rate": 3.836e-05, "loss": 0.0344, "step": 960 }, { "grad_norm": 0.2953590154647827, "learning_rate": 3.876e-05, "loss": 0.0377, "step": 970 }, { "grad_norm": 0.43050825595855713, "learning_rate": 3.9160000000000005e-05, "loss": 0.0326, "step": 980 }, { "grad_norm": 0.41555044054985046, "learning_rate": 3.956e-05, "loss": 0.0348, "step": 990 }, { "grad_norm": 0.23179574310779572, "learning_rate": 3.9960000000000004e-05, "loss": 0.0367, "step": 1000 }, { "grad_norm": 0.2136465162038803, "learning_rate": 4.0360000000000007e-05, "loss": 0.0358, "step": 1010 }, { "grad_norm": 0.37341630458831787, "learning_rate": 4.076e-05, "loss": 0.0414, "step": 1020 }, { "grad_norm": 0.3190579414367676, "learning_rate": 4.1160000000000006e-05, "loss": 0.0344, "step": 1030 }, { "grad_norm": 0.20750872790813446, "learning_rate": 4.156e-05, "loss": 0.0356, "step": 1040 }, { "grad_norm": 0.47596025466918945, "learning_rate": 4.196e-05, "loss": 0.0462, "step": 1050 }, { "grad_norm": 0.25302228331565857, "learning_rate": 4.236e-05, "loss": 0.0375, "step": 1060 }, { "grad_norm": 0.4475952684879303, "learning_rate": 4.276e-05, "loss": 0.047, "step": 1070 }, { "grad_norm": 0.5108179450035095, "learning_rate": 4.316e-05, "loss": 0.0474, "step": 1080 }, { "grad_norm": 0.28000229597091675, "learning_rate": 4.356e-05, "loss": 0.039, "step": 1090 }, { "grad_norm": 0.4322735667228699, "learning_rate": 4.396e-05, "loss": 0.0357, "step": 1100 }, { "grad_norm": 0.3499084711074829, "learning_rate": 4.436e-05, "loss": 0.0427, "step": 1110 }, { "grad_norm": 0.2859272062778473, "learning_rate": 4.4760000000000005e-05, "loss": 0.0348, "step": 1120 }, { "grad_norm": 0.210103839635849, "learning_rate": 4.516e-05, "loss": 0.0394, "step": 1130 }, { "grad_norm": 0.3881371319293976, "learning_rate": 4.5560000000000004e-05, "loss": 0.0457, "step": 1140 }, { "grad_norm": 0.3659462630748749, "learning_rate": 4.596e-05, "loss": 0.0404, "step": 1150 }, { "grad_norm": 0.23163630068302155, "learning_rate": 4.636e-05, "loss": 0.0334, "step": 1160 }, { "grad_norm": 0.3244779706001282, "learning_rate": 4.6760000000000006e-05, "loss": 0.0392, "step": 1170 }, { "grad_norm": 0.21283097565174103, "learning_rate": 4.716e-05, "loss": 0.0312, "step": 1180 }, { "grad_norm": 0.2749260663986206, "learning_rate": 4.7560000000000005e-05, "loss": 0.0382, "step": 1190 }, { "grad_norm": 0.35258209705352783, "learning_rate": 4.796e-05, "loss": 0.0358, "step": 1200 }, { "grad_norm": 0.437526136636734, "learning_rate": 4.836e-05, "loss": 0.0366, "step": 1210 }, { "grad_norm": 0.40374132990837097, "learning_rate": 4.876e-05, "loss": 0.0354, "step": 1220 }, { "grad_norm": 0.24649934470653534, "learning_rate": 4.9160000000000004e-05, "loss": 0.0363, "step": 1230 }, { "grad_norm": 0.24509717524051666, "learning_rate": 4.956e-05, "loss": 0.0422, "step": 1240 }, { "grad_norm": 0.35285699367523193, "learning_rate": 4.996e-05, "loss": 0.0386, "step": 1250 }, { "grad_norm": 0.30946505069732666, "learning_rate": 5.0360000000000006e-05, "loss": 0.0364, "step": 1260 }, { "grad_norm": 0.2514175474643707, "learning_rate": 5.076000000000001e-05, "loss": 0.0356, "step": 1270 }, { "grad_norm": 0.3689185082912445, "learning_rate": 5.1160000000000005e-05, "loss": 0.0338, "step": 1280 }, { "grad_norm": 0.3503323495388031, "learning_rate": 5.1559999999999994e-05, "loss": 0.033, "step": 1290 }, { "grad_norm": 0.28486448526382446, "learning_rate": 5.196e-05, "loss": 0.0315, "step": 1300 }, { "grad_norm": 0.22150872647762299, "learning_rate": 5.236e-05, "loss": 0.0329, "step": 1310 }, { "grad_norm": 0.33501961827278137, "learning_rate": 5.2759999999999996e-05, "loss": 0.0355, "step": 1320 }, { "grad_norm": 0.20952826738357544, "learning_rate": 5.316e-05, "loss": 0.0299, "step": 1330 }, { "grad_norm": 0.22369424998760223, "learning_rate": 5.356e-05, "loss": 0.0308, "step": 1340 }, { "grad_norm": 0.25281545519828796, "learning_rate": 5.396e-05, "loss": 0.0396, "step": 1350 }, { "grad_norm": 0.38692280650138855, "learning_rate": 5.436e-05, "loss": 0.039, "step": 1360 }, { "grad_norm": 0.26798105239868164, "learning_rate": 5.476e-05, "loss": 0.0323, "step": 1370 }, { "grad_norm": 0.17496569454669952, "learning_rate": 5.516e-05, "loss": 0.0306, "step": 1380 }, { "grad_norm": 0.42651912569999695, "learning_rate": 5.556e-05, "loss": 0.0319, "step": 1390 }, { "grad_norm": 0.3260864019393921, "learning_rate": 5.596e-05, "loss": 0.0346, "step": 1400 }, { "grad_norm": 0.3656407594680786, "learning_rate": 5.636e-05, "loss": 0.0428, "step": 1410 }, { "grad_norm": 0.22460691630840302, "learning_rate": 5.6760000000000005e-05, "loss": 0.0444, "step": 1420 }, { "grad_norm": 0.24131491780281067, "learning_rate": 5.716e-05, "loss": 0.0352, "step": 1430 }, { "grad_norm": 0.345666766166687, "learning_rate": 5.7560000000000005e-05, "loss": 0.0411, "step": 1440 }, { "grad_norm": 0.36719000339508057, "learning_rate": 5.796e-05, "loss": 0.0428, "step": 1450 }, { "grad_norm": 0.3920629322528839, "learning_rate": 5.8360000000000004e-05, "loss": 0.0476, "step": 1460 }, { "grad_norm": 0.3591307997703552, "learning_rate": 5.876000000000001e-05, "loss": 0.0476, "step": 1470 }, { "grad_norm": 0.23891742527484894, "learning_rate": 5.916e-05, "loss": 0.042, "step": 1480 }, { "grad_norm": 0.3538036346435547, "learning_rate": 5.9560000000000006e-05, "loss": 0.042, "step": 1490 }, { "grad_norm": 0.21692420542240143, "learning_rate": 5.996e-05, "loss": 0.0505, "step": 1500 }, { "grad_norm": 0.20069529116153717, "learning_rate": 6.0360000000000005e-05, "loss": 0.0379, "step": 1510 }, { "grad_norm": 0.2941264808177948, "learning_rate": 6.076000000000001e-05, "loss": 0.045, "step": 1520 }, { "grad_norm": 0.21788308024406433, "learning_rate": 6.116e-05, "loss": 0.047, "step": 1530 }, { "grad_norm": 0.25625187158584595, "learning_rate": 6.156e-05, "loss": 0.0431, "step": 1540 }, { "grad_norm": 0.31483879685401917, "learning_rate": 6.196000000000001e-05, "loss": 0.0417, "step": 1550 }, { "grad_norm": 0.2393040806055069, "learning_rate": 6.236e-05, "loss": 0.0427, "step": 1560 }, { "grad_norm": 0.22680608928203583, "learning_rate": 6.276e-05, "loss": 0.0371, "step": 1570 }, { "grad_norm": 0.24070757627487183, "learning_rate": 6.316000000000001e-05, "loss": 0.0365, "step": 1580 }, { "grad_norm": 0.30589523911476135, "learning_rate": 6.356000000000001e-05, "loss": 0.0407, "step": 1590 }, { "grad_norm": 0.22735585272312164, "learning_rate": 6.396e-05, "loss": 0.0365, "step": 1600 }, { "grad_norm": 0.25094056129455566, "learning_rate": 6.436e-05, "loss": 0.0358, "step": 1610 }, { "grad_norm": 0.18573275208473206, "learning_rate": 6.476e-05, "loss": 0.0372, "step": 1620 }, { "grad_norm": 0.41015535593032837, "learning_rate": 6.515999999999999e-05, "loss": 0.0682, "step": 1630 }, { "grad_norm": 0.45345717668533325, "learning_rate": 6.556e-05, "loss": 0.0503, "step": 1640 }, { "grad_norm": 0.4436343312263489, "learning_rate": 6.596e-05, "loss": 0.0436, "step": 1650 }, { "grad_norm": 0.1902451068162918, "learning_rate": 6.636e-05, "loss": 0.04, "step": 1660 }, { "grad_norm": 0.17136135697364807, "learning_rate": 6.676e-05, "loss": 0.0373, "step": 1670 }, { "grad_norm": 0.19156162440776825, "learning_rate": 6.716e-05, "loss": 0.0362, "step": 1680 }, { "grad_norm": 0.3431842029094696, "learning_rate": 6.756e-05, "loss": 0.0364, "step": 1690 }, { "grad_norm": 0.1849825084209442, "learning_rate": 6.796e-05, "loss": 0.0405, "step": 1700 }, { "grad_norm": 0.4013414680957794, "learning_rate": 6.836e-05, "loss": 0.0439, "step": 1710 }, { "grad_norm": 0.2063896358013153, "learning_rate": 6.876e-05, "loss": 0.0383, "step": 1720 }, { "grad_norm": 0.2561852037906647, "learning_rate": 6.916000000000001e-05, "loss": 0.0362, "step": 1730 }, { "grad_norm": 0.337148517370224, "learning_rate": 6.956e-05, "loss": 0.0392, "step": 1740 }, { "grad_norm": 0.23878096044063568, "learning_rate": 6.996e-05, "loss": 0.0443, "step": 1750 }, { "grad_norm": 0.4548408091068268, "learning_rate": 7.036e-05, "loss": 0.0374, "step": 1760 }, { "grad_norm": 0.259794145822525, "learning_rate": 7.076000000000001e-05, "loss": 0.0383, "step": 1770 }, { "grad_norm": 0.22533714771270752, "learning_rate": 7.116e-05, "loss": 0.0402, "step": 1780 }, { "grad_norm": 0.2734662592411041, "learning_rate": 7.156e-05, "loss": 0.0364, "step": 1790 }, { "grad_norm": 0.28625717759132385, "learning_rate": 7.196000000000001e-05, "loss": 0.034, "step": 1800 }, { "grad_norm": 0.2431894689798355, "learning_rate": 7.236e-05, "loss": 0.0353, "step": 1810 }, { "grad_norm": 0.49855026602745056, "learning_rate": 7.276e-05, "loss": 0.038, "step": 1820 }, { "grad_norm": 0.41488510370254517, "learning_rate": 7.316000000000001e-05, "loss": 0.0486, "step": 1830 }, { "grad_norm": 0.25711095333099365, "learning_rate": 7.356000000000001e-05, "loss": 0.0432, "step": 1840 }, { "grad_norm": 0.22361280024051666, "learning_rate": 7.396e-05, "loss": 0.0476, "step": 1850 }, { "grad_norm": 0.36687353253364563, "learning_rate": 7.436000000000001e-05, "loss": 0.0509, "step": 1860 }, { "grad_norm": 0.22232168912887573, "learning_rate": 7.476000000000001e-05, "loss": 0.0397, "step": 1870 }, { "grad_norm": 0.26704034209251404, "learning_rate": 7.516e-05, "loss": 0.0339, "step": 1880 }, { "grad_norm": 0.36167219281196594, "learning_rate": 7.556000000000002e-05, "loss": 0.0394, "step": 1890 }, { "grad_norm": 0.30486559867858887, "learning_rate": 7.596000000000001e-05, "loss": 0.0415, "step": 1900 }, { "grad_norm": 0.3025766909122467, "learning_rate": 7.636e-05, "loss": 0.0499, "step": 1910 }, { "grad_norm": 0.31434720754623413, "learning_rate": 7.676e-05, "loss": 0.0378, "step": 1920 }, { "grad_norm": 0.1845294088125229, "learning_rate": 7.716e-05, "loss": 0.0384, "step": 1930 }, { "grad_norm": 0.22529694437980652, "learning_rate": 7.756e-05, "loss": 0.0346, "step": 1940 }, { "grad_norm": 0.29431846737861633, "learning_rate": 7.796e-05, "loss": 0.038, "step": 1950 }, { "grad_norm": 0.3329464793205261, "learning_rate": 7.836e-05, "loss": 0.0372, "step": 1960 }, { "grad_norm": 0.3270244896411896, "learning_rate": 7.876e-05, "loss": 0.0351, "step": 1970 }, { "grad_norm": 0.4252513349056244, "learning_rate": 7.916e-05, "loss": 0.045, "step": 1980 }, { "grad_norm": 0.23762854933738708, "learning_rate": 7.956e-05, "loss": 0.0388, "step": 1990 }, { "grad_norm": 0.2894483804702759, "learning_rate": 7.996e-05, "loss": 0.038, "step": 2000 }, { "grad_norm": 0.28248271346092224, "learning_rate": 8.036e-05, "loss": 0.0447, "step": 2010 }, { "grad_norm": 0.45229101181030273, "learning_rate": 8.076e-05, "loss": 0.0386, "step": 2020 }, { "grad_norm": 0.41498902440071106, "learning_rate": 8.116e-05, "loss": 0.0343, "step": 2030 }, { "grad_norm": 0.18084469437599182, "learning_rate": 8.156e-05, "loss": 0.0402, "step": 2040 }, { "grad_norm": 0.2664511203765869, "learning_rate": 8.196000000000001e-05, "loss": 0.0409, "step": 2050 }, { "grad_norm": 0.3644930124282837, "learning_rate": 8.236e-05, "loss": 0.0335, "step": 2060 }, { "grad_norm": 0.3970622420310974, "learning_rate": 8.276e-05, "loss": 0.0327, "step": 2070 }, { "grad_norm": 0.277560830116272, "learning_rate": 8.316000000000001e-05, "loss": 0.0339, "step": 2080 }, { "grad_norm": 0.2795353829860687, "learning_rate": 8.356e-05, "loss": 0.036, "step": 2090 }, { "grad_norm": 0.2472517490386963, "learning_rate": 8.396e-05, "loss": 0.0428, "step": 2100 }, { "grad_norm": 0.29143717885017395, "learning_rate": 8.436000000000001e-05, "loss": 0.0434, "step": 2110 }, { "grad_norm": 0.19028492271900177, "learning_rate": 8.476000000000001e-05, "loss": 0.0354, "step": 2120 }, { "grad_norm": 0.15771502256393433, "learning_rate": 8.516e-05, "loss": 0.0373, "step": 2130 }, { "grad_norm": 0.20394255220890045, "learning_rate": 8.556e-05, "loss": 0.0314, "step": 2140 }, { "grad_norm": 0.428763747215271, "learning_rate": 8.596000000000001e-05, "loss": 0.0411, "step": 2150 }, { "grad_norm": 0.20813021063804626, "learning_rate": 8.636e-05, "loss": 0.0324, "step": 2160 }, { "grad_norm": 0.16151902079582214, "learning_rate": 8.676e-05, "loss": 0.0356, "step": 2170 }, { "grad_norm": 0.18731603026390076, "learning_rate": 8.716000000000001e-05, "loss": 0.033, "step": 2180 }, { "grad_norm": 0.3230607807636261, "learning_rate": 8.756000000000001e-05, "loss": 0.0333, "step": 2190 }, { "grad_norm": 0.243938609957695, "learning_rate": 8.796e-05, "loss": 0.0297, "step": 2200 }, { "grad_norm": 0.14579905569553375, "learning_rate": 8.836000000000001e-05, "loss": 0.0356, "step": 2210 }, { "grad_norm": 0.3296683132648468, "learning_rate": 8.876e-05, "loss": 0.0309, "step": 2220 }, { "grad_norm": 0.24329544603824615, "learning_rate": 8.916e-05, "loss": 0.0348, "step": 2230 }, { "grad_norm": 0.2949528396129608, "learning_rate": 8.956e-05, "loss": 0.0312, "step": 2240 }, { "grad_norm": 0.3855993151664734, "learning_rate": 8.996e-05, "loss": 0.0337, "step": 2250 }, { "grad_norm": 0.19163888692855835, "learning_rate": 9.036e-05, "loss": 0.0323, "step": 2260 }, { "grad_norm": 0.23634874820709229, "learning_rate": 9.076e-05, "loss": 0.0304, "step": 2270 }, { "grad_norm": 0.2060225009918213, "learning_rate": 9.116e-05, "loss": 0.0299, "step": 2280 }, { "grad_norm": 0.4028884768486023, "learning_rate": 9.156e-05, "loss": 0.0369, "step": 2290 }, { "grad_norm": 0.26010340452194214, "learning_rate": 9.196000000000001e-05, "loss": 0.0298, "step": 2300 }, { "grad_norm": 0.16423682868480682, "learning_rate": 9.236e-05, "loss": 0.0331, "step": 2310 }, { "grad_norm": 0.24090035259723663, "learning_rate": 9.276e-05, "loss": 0.0303, "step": 2320 }, { "grad_norm": 0.30139198899269104, "learning_rate": 9.316000000000001e-05, "loss": 0.0356, "step": 2330 }, { "grad_norm": 0.30917999148368835, "learning_rate": 9.356e-05, "loss": 0.0343, "step": 2340 }, { "grad_norm": 0.27409639954566956, "learning_rate": 9.396e-05, "loss": 0.0344, "step": 2350 }, { "grad_norm": 0.4897935092449188, "learning_rate": 9.436e-05, "loss": 0.0349, "step": 2360 }, { "grad_norm": 0.21246276795864105, "learning_rate": 9.476000000000001e-05, "loss": 0.0404, "step": 2370 }, { "grad_norm": 0.38638386130332947, "learning_rate": 9.516e-05, "loss": 0.0376, "step": 2380 }, { "grad_norm": 0.2519444525241852, "learning_rate": 9.556e-05, "loss": 0.0372, "step": 2390 }, { "grad_norm": 0.17996981739997864, "learning_rate": 9.596000000000001e-05, "loss": 0.0363, "step": 2400 }, { "grad_norm": 0.2842499613761902, "learning_rate": 9.636e-05, "loss": 0.0346, "step": 2410 }, { "grad_norm": 0.40104976296424866, "learning_rate": 9.676e-05, "loss": 0.0446, "step": 2420 }, { "grad_norm": 0.31160107254981995, "learning_rate": 9.716000000000001e-05, "loss": 0.0416, "step": 2430 }, { "grad_norm": 0.3118152916431427, "learning_rate": 9.756000000000001e-05, "loss": 0.0356, "step": 2440 }, { "grad_norm": 0.2705709636211395, "learning_rate": 9.796e-05, "loss": 0.0313, "step": 2450 }, { "grad_norm": 0.20069251954555511, "learning_rate": 9.836000000000001e-05, "loss": 0.0307, "step": 2460 }, { "grad_norm": 0.2991482615470886, "learning_rate": 9.876000000000001e-05, "loss": 0.0296, "step": 2470 }, { "grad_norm": 0.3668324053287506, "learning_rate": 9.916e-05, "loss": 0.0404, "step": 2480 }, { "grad_norm": 0.2138020098209381, "learning_rate": 9.956e-05, "loss": 0.0318, "step": 2490 }, { "grad_norm": 0.2162335216999054, "learning_rate": 9.996000000000001e-05, "loss": 0.0306, "step": 2500 }, { "grad_norm": 0.19371595978736877, "learning_rate": 9.999999114196196e-05, "loss": 0.0306, "step": 2510 }, { "grad_norm": 0.17482741177082062, "learning_rate": 9.99999605215876e-05, "loss": 0.0359, "step": 2520 }, { "grad_norm": 0.26738423109054565, "learning_rate": 9.999990802953179e-05, "loss": 0.029, "step": 2530 }, { "grad_norm": 0.18339166045188904, "learning_rate": 9.99998336658175e-05, "loss": 0.0284, "step": 2540 }, { "grad_norm": 0.18262645602226257, "learning_rate": 9.999973743047727e-05, "loss": 0.0318, "step": 2550 }, { "grad_norm": 0.27301132678985596, "learning_rate": 9.999961932355319e-05, "loss": 0.0314, "step": 2560 }, { "grad_norm": 0.3916976749897003, "learning_rate": 9.999947934509693e-05, "loss": 0.0307, "step": 2570 }, { "grad_norm": 0.28756073117256165, "learning_rate": 9.999931749516971e-05, "loss": 0.0316, "step": 2580 }, { "grad_norm": 0.17592065036296844, "learning_rate": 9.999913377384233e-05, "loss": 0.0313, "step": 2590 }, { "grad_norm": 0.18975912034511566, "learning_rate": 9.999892818119517e-05, "loss": 0.0298, "step": 2600 }, { "grad_norm": 0.22800566256046295, "learning_rate": 9.999870071731814e-05, "loss": 0.0293, "step": 2610 }, { "grad_norm": 0.2720426619052887, "learning_rate": 9.999845138231076e-05, "loss": 0.0296, "step": 2620 }, { "grad_norm": 0.29745715856552124, "learning_rate": 9.999818017628208e-05, "loss": 0.0329, "step": 2630 }, { "grad_norm": 0.259875625371933, "learning_rate": 9.999788709935078e-05, "loss": 0.0325, "step": 2640 }, { "grad_norm": 0.14151740074157715, "learning_rate": 9.9997572151645e-05, "loss": 0.0359, "step": 2650 }, { "grad_norm": 0.2415122389793396, "learning_rate": 9.999723533330254e-05, "loss": 0.0375, "step": 2660 }, { "grad_norm": 0.3514879643917084, "learning_rate": 9.999687664447074e-05, "loss": 0.037, "step": 2670 }, { "grad_norm": 0.3118593692779541, "learning_rate": 9.99964960853065e-05, "loss": 0.0376, "step": 2680 }, { "grad_norm": 0.22724445164203644, "learning_rate": 9.999609365597627e-05, "loss": 0.0331, "step": 2690 }, { "grad_norm": 0.31846368312835693, "learning_rate": 9.99956693566561e-05, "loss": 0.0262, "step": 2700 }, { "grad_norm": 0.3444663882255554, "learning_rate": 9.99952231875316e-05, "loss": 0.0302, "step": 2710 }, { "grad_norm": 0.43765294551849365, "learning_rate": 9.999475514879795e-05, "loss": 0.0408, "step": 2720 }, { "grad_norm": 0.20872987806797028, "learning_rate": 9.999426524065984e-05, "loss": 0.0347, "step": 2730 }, { "grad_norm": 0.20433704555034637, "learning_rate": 9.999375346333162e-05, "loss": 0.0278, "step": 2740 }, { "grad_norm": 0.24376459419727325, "learning_rate": 9.999321981703715e-05, "loss": 0.033, "step": 2750 }, { "grad_norm": 0.21670112013816833, "learning_rate": 9.999266430200985e-05, "loss": 0.0303, "step": 2760 }, { "grad_norm": 0.29209405183792114, "learning_rate": 9.999208691849271e-05, "loss": 0.0337, "step": 2770 }, { "grad_norm": 0.25079649686813354, "learning_rate": 9.999148766673832e-05, "loss": 0.0351, "step": 2780 }, { "grad_norm": 0.21562227606773376, "learning_rate": 9.999086654700881e-05, "loss": 0.0311, "step": 2790 }, { "grad_norm": 0.33722633123397827, "learning_rate": 9.999022355957588e-05, "loss": 0.0297, "step": 2800 }, { "grad_norm": 0.3871658742427826, "learning_rate": 9.998955870472079e-05, "loss": 0.0472, "step": 2810 }, { "grad_norm": 0.3341105580329895, "learning_rate": 9.998887198273437e-05, "loss": 0.0376, "step": 2820 }, { "grad_norm": 0.1737353801727295, "learning_rate": 9.998816339391701e-05, "loss": 0.0345, "step": 2830 }, { "grad_norm": 0.29043614864349365, "learning_rate": 9.998743293857868e-05, "loss": 0.0381, "step": 2840 }, { "grad_norm": 0.2747464179992676, "learning_rate": 9.998668061703891e-05, "loss": 0.0356, "step": 2850 }, { "grad_norm": 0.21665728092193604, "learning_rate": 9.998590642962679e-05, "loss": 0.0285, "step": 2860 }, { "grad_norm": 0.28115952014923096, "learning_rate": 9.998511037668095e-05, "loss": 0.0284, "step": 2870 }, { "grad_norm": 0.19157476723194122, "learning_rate": 9.998429245854964e-05, "loss": 0.0246, "step": 2880 }, { "grad_norm": 0.2383480966091156, "learning_rate": 9.998345267559064e-05, "loss": 0.0332, "step": 2890 }, { "grad_norm": 0.37341463565826416, "learning_rate": 9.998259102817129e-05, "loss": 0.0446, "step": 2900 }, { "grad_norm": 0.23299415409564972, "learning_rate": 9.99817075166685e-05, "loss": 0.0337, "step": 2910 }, { "grad_norm": 0.255562424659729, "learning_rate": 9.998080214146878e-05, "loss": 0.0338, "step": 2920 }, { "grad_norm": 0.14506320655345917, "learning_rate": 9.997987490296813e-05, "loss": 0.0322, "step": 2930 }, { "grad_norm": 0.1869095265865326, "learning_rate": 9.99789258015722e-05, "loss": 0.0319, "step": 2940 }, { "grad_norm": 0.3102479875087738, "learning_rate": 9.997795483769611e-05, "loss": 0.0339, "step": 2950 }, { "grad_norm": 0.40752512216567993, "learning_rate": 9.997696201176462e-05, "loss": 0.0399, "step": 2960 }, { "grad_norm": 0.15188740193843842, "learning_rate": 9.997594732421203e-05, "loss": 0.0348, "step": 2970 }, { "grad_norm": 0.269216388463974, "learning_rate": 9.997491077548217e-05, "loss": 0.0371, "step": 2980 }, { "grad_norm": 0.23330360651016235, "learning_rate": 9.997385236602851e-05, "loss": 0.0386, "step": 2990 }, { "grad_norm": 0.3250981569290161, "learning_rate": 9.997277209631399e-05, "loss": 0.0384, "step": 3000 }, { "grad_norm": 0.13273845613002777, "learning_rate": 9.997166996681118e-05, "loss": 0.0324, "step": 3010 }, { "grad_norm": 0.22593702375888824, "learning_rate": 9.997054597800218e-05, "loss": 0.0351, "step": 3020 }, { "grad_norm": 0.36467450857162476, "learning_rate": 9.996940013037866e-05, "loss": 0.0335, "step": 3030 }, { "grad_norm": 0.30431199073791504, "learning_rate": 9.996823242444186e-05, "loss": 0.036, "step": 3040 }, { "grad_norm": 0.21686606109142303, "learning_rate": 9.996704286070258e-05, "loss": 0.0339, "step": 3050 }, { "grad_norm": 0.3570975363254547, "learning_rate": 9.996583143968115e-05, "loss": 0.0429, "step": 3060 }, { "grad_norm": 0.3031894266605377, "learning_rate": 9.99645981619075e-05, "loss": 0.0328, "step": 3070 }, { "grad_norm": 0.2704267203807831, "learning_rate": 9.996334302792114e-05, "loss": 0.0335, "step": 3080 }, { "grad_norm": 0.48929500579833984, "learning_rate": 9.996206603827105e-05, "loss": 0.0446, "step": 3090 }, { "grad_norm": 0.2787121534347534, "learning_rate": 9.996076719351587e-05, "loss": 0.0362, "step": 3100 }, { "grad_norm": 0.20141683518886566, "learning_rate": 9.995944649422374e-05, "loss": 0.0317, "step": 3110 }, { "grad_norm": 0.21804708242416382, "learning_rate": 9.995810394097239e-05, "loss": 0.0302, "step": 3120 }, { "grad_norm": 0.19731181859970093, "learning_rate": 9.995673953434909e-05, "loss": 0.0298, "step": 3130 }, { "grad_norm": 0.3206290006637573, "learning_rate": 9.995535327495068e-05, "loss": 0.0345, "step": 3140 }, { "grad_norm": 0.12807269394397736, "learning_rate": 9.995394516338355e-05, "loss": 0.0305, "step": 3150 }, { "grad_norm": 0.21773996949195862, "learning_rate": 9.995251520026367e-05, "loss": 0.0387, "step": 3160 }, { "grad_norm": 0.23196645081043243, "learning_rate": 9.995106338621656e-05, "loss": 0.034, "step": 3170 }, { "grad_norm": 0.2112317532300949, "learning_rate": 9.994958972187726e-05, "loss": 0.0404, "step": 3180 }, { "grad_norm": 0.3830588161945343, "learning_rate": 9.994809420789044e-05, "loss": 0.0443, "step": 3190 }, { "grad_norm": 0.2603774070739746, "learning_rate": 9.994657684491027e-05, "loss": 0.0439, "step": 3200 }, { "grad_norm": 0.2760973870754242, "learning_rate": 9.994503763360048e-05, "loss": 0.0346, "step": 3210 }, { "grad_norm": 0.28766441345214844, "learning_rate": 9.99434765746344e-05, "loss": 0.0404, "step": 3220 }, { "grad_norm": 0.3500135838985443, "learning_rate": 9.994189366869488e-05, "loss": 0.041, "step": 3230 }, { "grad_norm": 0.24025273323059082, "learning_rate": 9.994028891647433e-05, "loss": 0.033, "step": 3240 }, { "grad_norm": 0.21475470066070557, "learning_rate": 9.993866231867475e-05, "loss": 0.0447, "step": 3250 }, { "grad_norm": 0.27520301938056946, "learning_rate": 9.993701387600762e-05, "loss": 0.0333, "step": 3260 }, { "grad_norm": 0.47402632236480713, "learning_rate": 9.993534358919408e-05, "loss": 0.0332, "step": 3270 }, { "grad_norm": 0.15420302748680115, "learning_rate": 9.993365145896473e-05, "loss": 0.034, "step": 3280 }, { "grad_norm": 0.21054524183273315, "learning_rate": 9.993193748605977e-05, "loss": 0.0304, "step": 3290 }, { "grad_norm": 0.15617547929286957, "learning_rate": 9.993020167122898e-05, "loss": 0.0306, "step": 3300 }, { "grad_norm": 0.27951154112815857, "learning_rate": 9.992844401523164e-05, "loss": 0.037, "step": 3310 }, { "grad_norm": 0.21944785118103027, "learning_rate": 9.992666451883661e-05, "loss": 0.036, "step": 3320 }, { "grad_norm": 0.20105627179145813, "learning_rate": 9.99248631828223e-05, "loss": 0.0325, "step": 3330 }, { "grad_norm": 0.20850282907485962, "learning_rate": 9.99230400079767e-05, "loss": 0.0354, "step": 3340 }, { "grad_norm": 0.16059856116771698, "learning_rate": 9.992119499509728e-05, "loss": 0.0318, "step": 3350 }, { "grad_norm": 0.29408615827560425, "learning_rate": 9.991932814499114e-05, "loss": 0.0354, "step": 3360 }, { "grad_norm": 0.1958850473165512, "learning_rate": 9.991743945847493e-05, "loss": 0.0364, "step": 3370 }, { "grad_norm": 0.23799031972885132, "learning_rate": 9.991552893637478e-05, "loss": 0.0536, "step": 3380 }, { "grad_norm": 0.1916004866361618, "learning_rate": 9.991359657952644e-05, "loss": 0.0356, "step": 3390 }, { "grad_norm": 0.2162080556154251, "learning_rate": 9.991164238877519e-05, "loss": 0.0345, "step": 3400 }, { "grad_norm": 0.23842526972293854, "learning_rate": 9.990966636497585e-05, "loss": 0.0358, "step": 3410 }, { "grad_norm": 0.3486901521682739, "learning_rate": 9.99076685089928e-05, "loss": 0.0348, "step": 3420 }, { "grad_norm": 0.23161612451076508, "learning_rate": 9.990564882169998e-05, "loss": 0.0339, "step": 3430 }, { "grad_norm": 0.17545653879642487, "learning_rate": 9.990360730398088e-05, "loss": 0.0379, "step": 3440 }, { "grad_norm": 0.2715107202529907, "learning_rate": 9.990154395672849e-05, "loss": 0.0333, "step": 3450 }, { "grad_norm": 0.24865931272506714, "learning_rate": 9.989945878084541e-05, "loss": 0.034, "step": 3460 }, { "grad_norm": 0.47707056999206543, "learning_rate": 9.989735177724378e-05, "loss": 0.041, "step": 3470 }, { "grad_norm": 0.2523977756500244, "learning_rate": 9.989522294684526e-05, "loss": 0.0444, "step": 3480 }, { "grad_norm": 0.3211061656475067, "learning_rate": 9.989307229058107e-05, "loss": 0.0393, "step": 3490 }, { "grad_norm": 0.14338338375091553, "learning_rate": 9.989089980939202e-05, "loss": 0.0322, "step": 3500 }, { "grad_norm": 0.18760667741298676, "learning_rate": 9.988870550422835e-05, "loss": 0.0323, "step": 3510 }, { "grad_norm": 0.18979564309120178, "learning_rate": 9.988648937604999e-05, "loss": 0.0358, "step": 3520 }, { "grad_norm": 0.27958962321281433, "learning_rate": 9.988425142582632e-05, "loss": 0.0362, "step": 3530 }, { "grad_norm": 0.1969999372959137, "learning_rate": 9.98819916545363e-05, "loss": 0.0307, "step": 3540 }, { "grad_norm": 0.2835831940174103, "learning_rate": 9.987971006316844e-05, "loss": 0.0353, "step": 3550 }, { "grad_norm": 0.20378942787647247, "learning_rate": 9.987740665272077e-05, "loss": 0.0291, "step": 3560 }, { "grad_norm": 0.25527918338775635, "learning_rate": 9.98750814242009e-05, "loss": 0.0364, "step": 3570 }, { "grad_norm": 0.1931353658437729, "learning_rate": 9.987273437862594e-05, "loss": 0.0386, "step": 3580 }, { "grad_norm": 0.2937192916870117, "learning_rate": 9.987036551702259e-05, "loss": 0.0451, "step": 3590 }, { "grad_norm": 0.2482585608959198, "learning_rate": 9.986797484042706e-05, "loss": 0.038, "step": 3600 }, { "grad_norm": 0.19169314205646515, "learning_rate": 9.986556234988512e-05, "loss": 0.0407, "step": 3610 }, { "grad_norm": 0.2570359706878662, "learning_rate": 9.986312804645205e-05, "loss": 0.0347, "step": 3620 }, { "grad_norm": 0.2957741618156433, "learning_rate": 9.986067193119273e-05, "loss": 0.0306, "step": 3630 }, { "grad_norm": 0.30272936820983887, "learning_rate": 9.985819400518153e-05, "loss": 0.0376, "step": 3640 }, { "grad_norm": 0.24428710341453552, "learning_rate": 9.985569426950239e-05, "loss": 0.0349, "step": 3650 }, { "grad_norm": 0.31840094923973083, "learning_rate": 9.985317272524876e-05, "loss": 0.0367, "step": 3660 }, { "grad_norm": 0.28941047191619873, "learning_rate": 9.985062937352366e-05, "loss": 0.0351, "step": 3670 }, { "grad_norm": 0.22137905657291412, "learning_rate": 9.984806421543966e-05, "loss": 0.0329, "step": 3680 }, { "grad_norm": 0.36693549156188965, "learning_rate": 9.984547725211881e-05, "loss": 0.0354, "step": 3690 }, { "grad_norm": 0.10560542345046997, "learning_rate": 9.984286848469276e-05, "loss": 0.0334, "step": 3700 }, { "grad_norm": 0.17493203282356262, "learning_rate": 9.984023791430266e-05, "loss": 0.0318, "step": 3710 }, { "grad_norm": 0.19725535809993744, "learning_rate": 9.983758554209924e-05, "loss": 0.0361, "step": 3720 }, { "grad_norm": 0.20076237618923187, "learning_rate": 9.983491136924268e-05, "loss": 0.0311, "step": 3730 }, { "grad_norm": 0.28355684876441956, "learning_rate": 9.983221539690282e-05, "loss": 0.0309, "step": 3740 }, { "grad_norm": 0.20672239363193512, "learning_rate": 9.982949762625892e-05, "loss": 0.039, "step": 3750 }, { "grad_norm": 0.15799486637115479, "learning_rate": 9.982675805849986e-05, "loss": 0.034, "step": 3760 }, { "grad_norm": 0.31827205419540405, "learning_rate": 9.982399669482399e-05, "loss": 0.0379, "step": 3770 }, { "grad_norm": 0.301687091588974, "learning_rate": 9.982121353643924e-05, "loss": 0.0337, "step": 3780 }, { "grad_norm": 0.1850714236497879, "learning_rate": 9.981840858456306e-05, "loss": 0.0326, "step": 3790 }, { "grad_norm": 0.32770493626594543, "learning_rate": 9.981558184042243e-05, "loss": 0.0485, "step": 3800 }, { "grad_norm": 0.12370344996452332, "learning_rate": 9.981273330525387e-05, "loss": 0.0325, "step": 3810 }, { "grad_norm": 0.3020925223827362, "learning_rate": 9.980986298030341e-05, "loss": 0.0344, "step": 3820 }, { "grad_norm": 0.17183339595794678, "learning_rate": 9.980697086682662e-05, "loss": 0.0292, "step": 3830 }, { "grad_norm": 0.14575307071208954, "learning_rate": 9.980405696608866e-05, "loss": 0.0262, "step": 3840 }, { "grad_norm": 0.37982285022735596, "learning_rate": 9.980112127936409e-05, "loss": 0.0389, "step": 3850 }, { "grad_norm": 0.20314639806747437, "learning_rate": 9.979816380793717e-05, "loss": 0.0308, "step": 3860 }, { "grad_norm": 0.15913701057434082, "learning_rate": 9.979518455310151e-05, "loss": 0.0309, "step": 3870 }, { "grad_norm": 0.2194487452507019, "learning_rate": 9.97921835161604e-05, "loss": 0.0293, "step": 3880 }, { "grad_norm": 0.20236681401729584, "learning_rate": 9.978916069842656e-05, "loss": 0.0346, "step": 3890 }, { "grad_norm": 0.31532812118530273, "learning_rate": 9.97861161012223e-05, "loss": 0.0332, "step": 3900 }, { "grad_norm": 0.3902089595794678, "learning_rate": 9.978304972587942e-05, "loss": 0.0288, "step": 3910 }, { "grad_norm": 0.30774828791618347, "learning_rate": 9.977996157373925e-05, "loss": 0.0337, "step": 3920 }, { "grad_norm": 0.22786326706409454, "learning_rate": 9.977685164615265e-05, "loss": 0.0324, "step": 3930 }, { "grad_norm": 0.4762822687625885, "learning_rate": 9.977371994448002e-05, "loss": 0.0419, "step": 3940 }, { "grad_norm": 0.21602769196033478, "learning_rate": 9.977056647009127e-05, "loss": 0.0387, "step": 3950 }, { "grad_norm": 0.2650911808013916, "learning_rate": 9.976739122436582e-05, "loss": 0.0452, "step": 3960 }, { "grad_norm": 0.2481309175491333, "learning_rate": 9.976419420869265e-05, "loss": 0.0339, "step": 3970 }, { "grad_norm": 0.19427870213985443, "learning_rate": 9.976097542447025e-05, "loss": 0.0367, "step": 3980 }, { "grad_norm": 0.36334729194641113, "learning_rate": 9.97577348731066e-05, "loss": 0.0356, "step": 3990 }, { "grad_norm": 0.35066524147987366, "learning_rate": 9.975447255601927e-05, "loss": 0.0361, "step": 4000 }, { "grad_norm": 0.27320626378059387, "learning_rate": 9.975118847463525e-05, "loss": 0.0335, "step": 4010 }, { "grad_norm": 0.22617504000663757, "learning_rate": 9.974788263039114e-05, "loss": 0.0394, "step": 4020 }, { "grad_norm": 0.2856038212776184, "learning_rate": 9.974455502473303e-05, "loss": 0.0387, "step": 4030 }, { "grad_norm": 0.19029085338115692, "learning_rate": 9.974120565911652e-05, "loss": 0.029, "step": 4040 }, { "grad_norm": 0.15553484857082367, "learning_rate": 9.973783453500674e-05, "loss": 0.0329, "step": 4050 }, { "grad_norm": 0.28226858377456665, "learning_rate": 9.973444165387835e-05, "loss": 0.0317, "step": 4060 }, { "grad_norm": 0.19541215896606445, "learning_rate": 9.973102701721549e-05, "loss": 0.0336, "step": 4070 }, { "grad_norm": 0.15213556587696075, "learning_rate": 9.972759062651184e-05, "loss": 0.0363, "step": 4080 }, { "grad_norm": 0.2711688280105591, "learning_rate": 9.972413248327059e-05, "loss": 0.0359, "step": 4090 }, { "grad_norm": 0.16989386081695557, "learning_rate": 9.972065258900447e-05, "loss": 0.0313, "step": 4100 }, { "grad_norm": 0.18336175382137299, "learning_rate": 9.971715094523569e-05, "loss": 0.0301, "step": 4110 }, { "grad_norm": 0.3852640688419342, "learning_rate": 9.971362755349598e-05, "loss": 0.0341, "step": 4120 }, { "grad_norm": 0.18842417001724243, "learning_rate": 9.971008241532662e-05, "loss": 0.0329, "step": 4130 }, { "grad_norm": 0.24652604758739471, "learning_rate": 9.970651553227835e-05, "loss": 0.033, "step": 4140 }, { "grad_norm": 0.2415955513715744, "learning_rate": 9.970292690591143e-05, "loss": 0.0352, "step": 4150 }, { "grad_norm": 0.23447279632091522, "learning_rate": 9.969931653779569e-05, "loss": 0.0353, "step": 4160 }, { "grad_norm": 0.19060108065605164, "learning_rate": 9.969568442951038e-05, "loss": 0.0332, "step": 4170 }, { "grad_norm": 0.1465197205543518, "learning_rate": 9.969203058264436e-05, "loss": 0.0322, "step": 4180 }, { "grad_norm": 0.13435815274715424, "learning_rate": 9.96883549987959e-05, "loss": 0.0307, "step": 4190 }, { "grad_norm": 0.1653425693511963, "learning_rate": 9.968465767957287e-05, "loss": 0.0293, "step": 4200 }, { "grad_norm": 0.15403082966804504, "learning_rate": 9.968093862659256e-05, "loss": 0.0301, "step": 4210 }, { "grad_norm": 0.2657291889190674, "learning_rate": 9.967719784148182e-05, "loss": 0.0332, "step": 4220 }, { "grad_norm": 0.255513072013855, "learning_rate": 9.967343532587702e-05, "loss": 0.0444, "step": 4230 }, { "grad_norm": 0.39845913648605347, "learning_rate": 9.966965108142399e-05, "loss": 0.0385, "step": 4240 }, { "grad_norm": 0.23027995228767395, "learning_rate": 9.96658451097781e-05, "loss": 0.0383, "step": 4250 }, { "grad_norm": 0.21409447491168976, "learning_rate": 9.966201741260419e-05, "loss": 0.0325, "step": 4260 }, { "grad_norm": 0.18850845098495483, "learning_rate": 9.965816799157665e-05, "loss": 0.0342, "step": 4270 }, { "grad_norm": 0.16579729318618774, "learning_rate": 9.965429684837935e-05, "loss": 0.0341, "step": 4280 }, { "grad_norm": 0.3190186023712158, "learning_rate": 9.965040398470562e-05, "loss": 0.0324, "step": 4290 }, { "grad_norm": 0.23656348884105682, "learning_rate": 9.964648940225838e-05, "loss": 0.0282, "step": 4300 }, { "grad_norm": 0.2679075300693512, "learning_rate": 9.964255310274997e-05, "loss": 0.0299, "step": 4310 }, { "grad_norm": 0.24237652122974396, "learning_rate": 9.963859508790228e-05, "loss": 0.0475, "step": 4320 }, { "grad_norm": 0.21356217563152313, "learning_rate": 9.963461535944664e-05, "loss": 0.0287, "step": 4330 }, { "grad_norm": 0.15440313518047333, "learning_rate": 9.963061391912399e-05, "loss": 0.0303, "step": 4340 }, { "grad_norm": 0.2094208151102066, "learning_rate": 9.962659076868463e-05, "loss": 0.0269, "step": 4350 }, { "grad_norm": 0.16545724868774414, "learning_rate": 9.962254590988846e-05, "loss": 0.0231, "step": 4360 }, { "grad_norm": 0.18463630974292755, "learning_rate": 9.961847934450481e-05, "loss": 0.0292, "step": 4370 }, { "grad_norm": 0.3282734751701355, "learning_rate": 9.961439107431257e-05, "loss": 0.0309, "step": 4380 }, { "grad_norm": 0.1885005235671997, "learning_rate": 9.961028110110006e-05, "loss": 0.0256, "step": 4390 }, { "grad_norm": 0.1549755483865738, "learning_rate": 9.960614942666513e-05, "loss": 0.0334, "step": 4400 }, { "grad_norm": 0.21130788326263428, "learning_rate": 9.960199605281511e-05, "loss": 0.032, "step": 4410 }, { "grad_norm": 0.22729507088661194, "learning_rate": 9.959782098136683e-05, "loss": 0.0452, "step": 4420 }, { "grad_norm": 0.2556120753288269, "learning_rate": 9.959362421414662e-05, "loss": 0.0382, "step": 4430 }, { "grad_norm": 0.2137247920036316, "learning_rate": 9.958940575299027e-05, "loss": 0.0389, "step": 4440 }, { "grad_norm": 0.27707722783088684, "learning_rate": 9.95851655997431e-05, "loss": 0.0392, "step": 4450 }, { "grad_norm": 0.27385061979293823, "learning_rate": 9.958090375625986e-05, "loss": 0.0361, "step": 4460 }, { "grad_norm": 0.1825522780418396, "learning_rate": 9.957662022440486e-05, "loss": 0.0311, "step": 4470 }, { "grad_norm": 0.2570458948612213, "learning_rate": 9.957231500605187e-05, "loss": 0.0365, "step": 4480 }, { "grad_norm": 0.19899700582027435, "learning_rate": 9.95679881030841e-05, "loss": 0.0356, "step": 4490 }, { "grad_norm": 0.2581806480884552, "learning_rate": 9.95636395173943e-05, "loss": 0.0343, "step": 4500 }, { "grad_norm": 0.1983656883239746, "learning_rate": 9.95592692508847e-05, "loss": 0.0479, "step": 4510 }, { "grad_norm": 0.3941843807697296, "learning_rate": 9.9554877305467e-05, "loss": 0.0332, "step": 4520 }, { "grad_norm": 0.15437911450862885, "learning_rate": 9.955046368306237e-05, "loss": 0.0349, "step": 4530 }, { "grad_norm": 0.3618836998939514, "learning_rate": 9.954602838560153e-05, "loss": 0.0432, "step": 4540 }, { "grad_norm": 0.3203836679458618, "learning_rate": 9.954157141502456e-05, "loss": 0.0418, "step": 4550 }, { "grad_norm": 0.3236807584762573, "learning_rate": 9.953709277328112e-05, "loss": 0.0372, "step": 4560 }, { "grad_norm": 0.21433743834495544, "learning_rate": 9.953259246233032e-05, "loss": 0.0407, "step": 4570 }, { "grad_norm": 0.33789488673210144, "learning_rate": 9.952807048414077e-05, "loss": 0.0368, "step": 4580 }, { "grad_norm": 0.2252611666917801, "learning_rate": 9.95235268406905e-05, "loss": 0.0385, "step": 4590 }, { "grad_norm": 0.3154430091381073, "learning_rate": 9.951896153396708e-05, "loss": 0.0395, "step": 4600 }, { "grad_norm": 0.21723824739456177, "learning_rate": 9.95143745659675e-05, "loss": 0.0329, "step": 4610 }, { "grad_norm": 0.23007597029209137, "learning_rate": 9.95097659386983e-05, "loss": 0.0359, "step": 4620 }, { "grad_norm": 0.4641176164150238, "learning_rate": 9.950513565417542e-05, "loss": 0.0372, "step": 4630 }, { "grad_norm": 0.21306543052196503, "learning_rate": 9.95004837144243e-05, "loss": 0.042, "step": 4640 }, { "grad_norm": 0.3119349181652069, "learning_rate": 9.949581012147988e-05, "loss": 0.036, "step": 4650 }, { "grad_norm": 0.1636851727962494, "learning_rate": 9.949111487738653e-05, "loss": 0.0368, "step": 4660 }, { "grad_norm": 0.2606201469898224, "learning_rate": 9.948639798419813e-05, "loss": 0.0309, "step": 4670 }, { "grad_norm": 0.18877443671226501, "learning_rate": 9.948165944397799e-05, "loss": 0.0366, "step": 4680 }, { "grad_norm": 0.1638733297586441, "learning_rate": 9.94768992587989e-05, "loss": 0.0429, "step": 4690 }, { "grad_norm": 0.20404328405857086, "learning_rate": 9.947211743074313e-05, "loss": 0.0391, "step": 4700 }, { "grad_norm": 0.2528764009475708, "learning_rate": 9.946731396190246e-05, "loss": 0.0356, "step": 4710 }, { "grad_norm": 0.32524198293685913, "learning_rate": 9.946248885437803e-05, "loss": 0.0378, "step": 4720 }, { "grad_norm": 0.25637343525886536, "learning_rate": 9.945764211028053e-05, "loss": 0.0373, "step": 4730 }, { "grad_norm": 0.24615006148815155, "learning_rate": 9.94527737317301e-05, "loss": 0.0377, "step": 4740 }, { "grad_norm": 0.2802336513996124, "learning_rate": 9.944788372085631e-05, "loss": 0.0376, "step": 4750 }, { "grad_norm": 0.1714850515127182, "learning_rate": 9.944297207979825e-05, "loss": 0.0363, "step": 4760 }, { "grad_norm": 0.43740227818489075, "learning_rate": 9.943803881070441e-05, "loss": 0.0388, "step": 4770 }, { "grad_norm": 0.34240666031837463, "learning_rate": 9.943308391573278e-05, "loss": 0.0437, "step": 4780 }, { "grad_norm": 0.22508257627487183, "learning_rate": 9.942810739705079e-05, "loss": 0.0348, "step": 4790 }, { "grad_norm": 0.21646109223365784, "learning_rate": 9.942310925683538e-05, "loss": 0.0377, "step": 4800 }, { "grad_norm": 0.1803000569343567, "learning_rate": 9.941808949727285e-05, "loss": 0.0346, "step": 4810 }, { "grad_norm": 0.22460804879665375, "learning_rate": 9.941304812055903e-05, "loss": 0.0354, "step": 4820 }, { "grad_norm": 0.14627967774868011, "learning_rate": 9.940798512889921e-05, "loss": 0.0329, "step": 4830 }, { "grad_norm": 0.262909859418869, "learning_rate": 9.94029005245081e-05, "loss": 0.037, "step": 4840 }, { "grad_norm": 0.28126463294029236, "learning_rate": 9.939779430960988e-05, "loss": 0.0361, "step": 4850 }, { "grad_norm": 0.3700762987136841, "learning_rate": 9.939266648643817e-05, "loss": 0.0382, "step": 4860 }, { "grad_norm": 0.2545032501220703, "learning_rate": 9.938751705723607e-05, "loss": 0.0347, "step": 4870 }, { "grad_norm": 0.3824107348918915, "learning_rate": 9.938234602425613e-05, "loss": 0.0398, "step": 4880 }, { "grad_norm": 0.35656481981277466, "learning_rate": 9.93771533897603e-05, "loss": 0.0444, "step": 4890 }, { "grad_norm": 0.22767634689807892, "learning_rate": 9.937193915602004e-05, "loss": 0.0348, "step": 4900 }, { "grad_norm": 0.3630760908126831, "learning_rate": 9.936670332531621e-05, "loss": 0.035, "step": 4910 }, { "grad_norm": 0.2548968493938446, "learning_rate": 9.936144589993916e-05, "loss": 0.0381, "step": 4920 }, { "grad_norm": 0.2245481312274933, "learning_rate": 9.935616688218867e-05, "loss": 0.0309, "step": 4930 }, { "grad_norm": 0.18964073061943054, "learning_rate": 9.935086627437395e-05, "loss": 0.0333, "step": 4940 }, { "grad_norm": 0.19784630835056305, "learning_rate": 9.934554407881366e-05, "loss": 0.0329, "step": 4950 }, { "grad_norm": 0.20977304875850677, "learning_rate": 9.934020029783593e-05, "loss": 0.0306, "step": 4960 }, { "grad_norm": 0.18179257214069366, "learning_rate": 9.933483493377829e-05, "loss": 0.0346, "step": 4970 }, { "grad_norm": 0.1779179573059082, "learning_rate": 9.932944798898774e-05, "loss": 0.0329, "step": 4980 }, { "grad_norm": 0.1523599624633789, "learning_rate": 9.932403946582072e-05, "loss": 0.0326, "step": 4990 }, { "grad_norm": 0.3131561875343323, "learning_rate": 9.93186093666431e-05, "loss": 0.037, "step": 5000 }, { "grad_norm": 0.21304652094841003, "learning_rate": 9.931315769383018e-05, "loss": 0.0346, "step": 5010 }, { "grad_norm": 0.3146451711654663, "learning_rate": 9.930768444976672e-05, "loss": 0.0332, "step": 5020 }, { "grad_norm": 0.23027372360229492, "learning_rate": 9.93021896368469e-05, "loss": 0.031, "step": 5030 }, { "grad_norm": 0.2079070806503296, "learning_rate": 9.929667325747431e-05, "loss": 0.0328, "step": 5040 }, { "grad_norm": 0.23538382351398468, "learning_rate": 9.929113531406205e-05, "loss": 0.0354, "step": 5050 }, { "grad_norm": 0.14390914142131805, "learning_rate": 9.928557580903257e-05, "loss": 0.026, "step": 5060 }, { "grad_norm": 0.302728533744812, "learning_rate": 9.927999474481779e-05, "loss": 0.034, "step": 5070 }, { "grad_norm": 0.24539592862129211, "learning_rate": 9.927439212385907e-05, "loss": 0.0306, "step": 5080 }, { "grad_norm": 0.23590905964374542, "learning_rate": 9.926876794860718e-05, "loss": 0.0314, "step": 5090 }, { "grad_norm": 0.25182044506073, "learning_rate": 9.926312222152235e-05, "loss": 0.028, "step": 5100 }, { "grad_norm": 0.21525833010673523, "learning_rate": 9.925745494507414e-05, "loss": 0.0343, "step": 5110 }, { "grad_norm": 0.3479357659816742, "learning_rate": 9.925176612174169e-05, "loss": 0.0358, "step": 5120 }, { "grad_norm": 0.25972801446914673, "learning_rate": 9.924605575401346e-05, "loss": 0.0357, "step": 5130 }, { "grad_norm": 0.32694458961486816, "learning_rate": 9.924032384438733e-05, "loss": 0.0319, "step": 5140 }, { "grad_norm": 0.13578437268733978, "learning_rate": 9.923457039537066e-05, "loss": 0.0281, "step": 5150 }, { "grad_norm": 0.1742972731590271, "learning_rate": 9.92287954094802e-05, "loss": 0.0424, "step": 5160 }, { "grad_norm": 0.34418854117393494, "learning_rate": 9.922299888924212e-05, "loss": 0.0349, "step": 5170 }, { "grad_norm": 0.15793560445308685, "learning_rate": 9.921718083719203e-05, "loss": 0.028, "step": 5180 }, { "grad_norm": 0.3268117904663086, "learning_rate": 9.921134125587491e-05, "loss": 0.0377, "step": 5190 }, { "grad_norm": 0.15807700157165527, "learning_rate": 9.920548014784523e-05, "loss": 0.0282, "step": 5200 }, { "grad_norm": 0.18670429289340973, "learning_rate": 9.919959751566681e-05, "loss": 0.0295, "step": 5210 }, { "grad_norm": 0.20216429233551025, "learning_rate": 9.919369336191291e-05, "loss": 0.0289, "step": 5220 }, { "grad_norm": 0.13919037580490112, "learning_rate": 9.918776768916625e-05, "loss": 0.0259, "step": 5230 }, { "grad_norm": 0.14145107567310333, "learning_rate": 9.918182050001888e-05, "loss": 0.0272, "step": 5240 }, { "grad_norm": 0.28772035241127014, "learning_rate": 9.91758517970723e-05, "loss": 0.0348, "step": 5250 }, { "grad_norm": 0.2046770304441452, "learning_rate": 9.916986158293744e-05, "loss": 0.0298, "step": 5260 }, { "grad_norm": 0.20647412538528442, "learning_rate": 9.916384986023463e-05, "loss": 0.0314, "step": 5270 }, { "grad_norm": 0.4430522620677948, "learning_rate": 9.915781663159359e-05, "loss": 0.0364, "step": 5280 }, { "grad_norm": 0.2890566885471344, "learning_rate": 9.915176189965346e-05, "loss": 0.0352, "step": 5290 }, { "grad_norm": 0.1602289229631424, "learning_rate": 9.914568566706279e-05, "loss": 0.0304, "step": 5300 }, { "grad_norm": 0.19222886860370636, "learning_rate": 9.913958793647953e-05, "loss": 0.031, "step": 5310 }, { "grad_norm": 0.3227500021457672, "learning_rate": 9.913346871057102e-05, "loss": 0.0302, "step": 5320 }, { "grad_norm": 0.2612934112548828, "learning_rate": 9.912732799201403e-05, "loss": 0.0354, "step": 5330 }, { "grad_norm": 0.21372072398662567, "learning_rate": 9.912116578349474e-05, "loss": 0.0325, "step": 5340 }, { "grad_norm": 0.2392779439687729, "learning_rate": 9.911498208770867e-05, "loss": 0.0337, "step": 5350 }, { "grad_norm": 0.24878734350204468, "learning_rate": 9.910877690736078e-05, "loss": 0.0337, "step": 5360 }, { "grad_norm": 0.15085849165916443, "learning_rate": 9.910255024516546e-05, "loss": 0.0301, "step": 5370 }, { "grad_norm": 0.2632475197315216, "learning_rate": 9.909630210384644e-05, "loss": 0.0313, "step": 5380 }, { "grad_norm": 0.20074355602264404, "learning_rate": 9.909003248613688e-05, "loss": 0.0264, "step": 5390 }, { "grad_norm": 0.22696655988693237, "learning_rate": 9.908374139477934e-05, "loss": 0.0325, "step": 5400 }, { "grad_norm": 0.1294875591993332, "learning_rate": 9.907742883252571e-05, "loss": 0.0305, "step": 5410 }, { "grad_norm": 0.22502388060092926, "learning_rate": 9.907109480213736e-05, "loss": 0.031, "step": 5420 }, { "grad_norm": 0.30472806096076965, "learning_rate": 9.906473930638498e-05, "loss": 0.0267, "step": 5430 }, { "grad_norm": 0.17444226145744324, "learning_rate": 9.90583623480487e-05, "loss": 0.0294, "step": 5440 }, { "grad_norm": 0.25106313824653625, "learning_rate": 9.905196392991802e-05, "loss": 0.0269, "step": 5450 }, { "grad_norm": 0.16473530232906342, "learning_rate": 9.904554405479183e-05, "loss": 0.0278, "step": 5460 }, { "grad_norm": 0.222516730427742, "learning_rate": 9.903910272547838e-05, "loss": 0.0296, "step": 5470 }, { "grad_norm": 0.15642692148685455, "learning_rate": 9.903263994479533e-05, "loss": 0.0293, "step": 5480 }, { "grad_norm": 0.16150133311748505, "learning_rate": 9.902615571556972e-05, "loss": 0.0287, "step": 5490 }, { "grad_norm": 0.22681839764118195, "learning_rate": 9.9019650040638e-05, "loss": 0.0318, "step": 5500 }, { "grad_norm": 0.17593121528625488, "learning_rate": 9.901312292284594e-05, "loss": 0.0343, "step": 5510 }, { "grad_norm": 0.20200780034065247, "learning_rate": 9.900657436504875e-05, "loss": 0.0306, "step": 5520 }, { "grad_norm": 0.21468722820281982, "learning_rate": 9.900000437011097e-05, "loss": 0.0273, "step": 5530 }, { "grad_norm": 0.17968478798866272, "learning_rate": 9.899341294090653e-05, "loss": 0.0274, "step": 5540 }, { "grad_norm": 0.1535012125968933, "learning_rate": 9.898680008031877e-05, "loss": 0.0278, "step": 5550 }, { "grad_norm": 0.20988669991493225, "learning_rate": 9.898016579124037e-05, "loss": 0.0295, "step": 5560 }, { "grad_norm": 0.3133999705314636, "learning_rate": 9.89735100765734e-05, "loss": 0.031, "step": 5570 }, { "grad_norm": 0.2789193391799927, "learning_rate": 9.89668329392293e-05, "loss": 0.032, "step": 5580 }, { "grad_norm": 0.1425546407699585, "learning_rate": 9.896013438212885e-05, "loss": 0.0281, "step": 5590 }, { "grad_norm": 0.3163287043571472, "learning_rate": 9.895341440820225e-05, "loss": 0.0335, "step": 5600 }, { "grad_norm": 0.14179451763629913, "learning_rate": 9.894667302038902e-05, "loss": 0.0286, "step": 5610 }, { "grad_norm": 0.1697067767381668, "learning_rate": 9.893991022163812e-05, "loss": 0.0287, "step": 5620 }, { "grad_norm": 0.16119073331356049, "learning_rate": 9.893312601490779e-05, "loss": 0.0267, "step": 5630 }, { "grad_norm": 0.2651679813861847, "learning_rate": 9.892632040316568e-05, "loss": 0.0265, "step": 5640 }, { "grad_norm": 0.2817445993423462, "learning_rate": 9.891949338938877e-05, "loss": 0.0382, "step": 5650 }, { "grad_norm": 0.25626832246780396, "learning_rate": 9.89126449765635e-05, "loss": 0.0335, "step": 5660 }, { "grad_norm": 0.40750598907470703, "learning_rate": 9.890577516768551e-05, "loss": 0.0351, "step": 5670 }, { "grad_norm": 0.21621447801589966, "learning_rate": 9.889888396575995e-05, "loss": 0.0352, "step": 5680 }, { "grad_norm": 0.2993229329586029, "learning_rate": 9.889197137380125e-05, "loss": 0.0378, "step": 5690 }, { "grad_norm": 0.19217102229595184, "learning_rate": 9.888503739483321e-05, "loss": 0.0314, "step": 5700 }, { "grad_norm": 0.17199227213859558, "learning_rate": 9.887808203188897e-05, "loss": 0.0332, "step": 5710 }, { "grad_norm": 0.41894465684890747, "learning_rate": 9.887110528801108e-05, "loss": 0.0326, "step": 5720 }, { "grad_norm": 0.1307428926229477, "learning_rate": 9.886410716625137e-05, "loss": 0.0309, "step": 5730 }, { "grad_norm": 0.2829993665218353, "learning_rate": 9.885708766967108e-05, "loss": 0.0345, "step": 5740 }, { "grad_norm": 0.16609127819538116, "learning_rate": 9.885004680134076e-05, "loss": 0.0275, "step": 5750 }, { "grad_norm": 0.22742082178592682, "learning_rate": 9.884298456434033e-05, "loss": 0.0345, "step": 5760 }, { "grad_norm": 0.2634093761444092, "learning_rate": 9.883590096175905e-05, "loss": 0.0303, "step": 5770 }, { "grad_norm": 0.24008606374263763, "learning_rate": 9.882879599669554e-05, "loss": 0.0365, "step": 5780 }, { "grad_norm": 0.19459116458892822, "learning_rate": 9.882166967225775e-05, "loss": 0.029, "step": 5790 }, { "grad_norm": 0.2329580932855606, "learning_rate": 9.881452199156296e-05, "loss": 0.034, "step": 5800 }, { "grad_norm": 0.1736326366662979, "learning_rate": 9.880735295773782e-05, "loss": 0.0286, "step": 5810 }, { "grad_norm": 0.3220580816268921, "learning_rate": 9.88001625739183e-05, "loss": 0.0329, "step": 5820 }, { "grad_norm": 0.2502436935901642, "learning_rate": 9.879295084324976e-05, "loss": 0.0324, "step": 5830 }, { "grad_norm": 0.17087696492671967, "learning_rate": 9.87857177688868e-05, "loss": 0.0329, "step": 5840 }, { "grad_norm": 0.1899125576019287, "learning_rate": 9.877846335399343e-05, "loss": 0.0277, "step": 5850 }, { "grad_norm": 0.17140303552150726, "learning_rate": 9.8771187601743e-05, "loss": 0.0317, "step": 5860 }, { "grad_norm": 0.20988869667053223, "learning_rate": 9.876389051531813e-05, "loss": 0.0356, "step": 5870 }, { "grad_norm": 0.37597113847732544, "learning_rate": 9.875657209791088e-05, "loss": 0.0354, "step": 5880 }, { "grad_norm": 0.19212889671325684, "learning_rate": 9.874923235272248e-05, "loss": 0.0327, "step": 5890 }, { "grad_norm": 0.2571043372154236, "learning_rate": 9.874187128296369e-05, "loss": 0.0415, "step": 5900 }, { "grad_norm": 0.2088250070810318, "learning_rate": 9.873448889185439e-05, "loss": 0.0344, "step": 5910 }, { "grad_norm": 0.29061388969421387, "learning_rate": 9.872708518262397e-05, "loss": 0.0284, "step": 5920 }, { "grad_norm": 0.19758892059326172, "learning_rate": 9.871966015851101e-05, "loss": 0.0261, "step": 5930 }, { "grad_norm": 0.15616217255592346, "learning_rate": 9.87122138227635e-05, "loss": 0.0311, "step": 5940 }, { "grad_norm": 0.18786180019378662, "learning_rate": 9.87047461786387e-05, "loss": 0.0309, "step": 5950 }, { "grad_norm": 0.14114540815353394, "learning_rate": 9.869725722940323e-05, "loss": 0.0277, "step": 5960 }, { "grad_norm": 0.3614485561847687, "learning_rate": 9.868974697833299e-05, "loss": 0.0294, "step": 5970 }, { "grad_norm": 0.16046743094921112, "learning_rate": 9.868221542871324e-05, "loss": 0.0309, "step": 5980 }, { "grad_norm": 0.30641698837280273, "learning_rate": 9.867466258383853e-05, "loss": 0.0378, "step": 5990 }, { "grad_norm": 0.27425140142440796, "learning_rate": 9.866708844701272e-05, "loss": 0.043, "step": 6000 }, { "grad_norm": 0.2457849383354187, "learning_rate": 9.865949302154899e-05, "loss": 0.0367, "step": 6010 }, { "grad_norm": 0.26137685775756836, "learning_rate": 9.865187631076987e-05, "loss": 0.032, "step": 6020 }, { "grad_norm": 0.18645578622817993, "learning_rate": 9.864423831800712e-05, "loss": 0.0288, "step": 6030 }, { "grad_norm": 0.23391863703727722, "learning_rate": 9.863657904660191e-05, "loss": 0.0313, "step": 6040 }, { "grad_norm": 0.2586894631385803, "learning_rate": 9.862889849990462e-05, "loss": 0.0363, "step": 6050 }, { "grad_norm": 0.19233191013336182, "learning_rate": 9.8621196681275e-05, "loss": 0.0297, "step": 6060 }, { "grad_norm": 0.19794979691505432, "learning_rate": 9.861347359408211e-05, "loss": 0.0315, "step": 6070 }, { "grad_norm": 0.27164047956466675, "learning_rate": 9.860572924170426e-05, "loss": 0.0336, "step": 6080 }, { "grad_norm": 0.149839386343956, "learning_rate": 9.85979636275291e-05, "loss": 0.0326, "step": 6090 }, { "grad_norm": 0.23455004394054413, "learning_rate": 9.859017675495357e-05, "loss": 0.0309, "step": 6100 }, { "grad_norm": 0.26491403579711914, "learning_rate": 9.858236862738392e-05, "loss": 0.0319, "step": 6110 }, { "grad_norm": 0.18559223413467407, "learning_rate": 9.85745392482357e-05, "loss": 0.0335, "step": 6120 }, { "grad_norm": 0.15594851970672607, "learning_rate": 9.856668862093372e-05, "loss": 0.027, "step": 6130 }, { "grad_norm": 0.232744038105011, "learning_rate": 9.855881674891214e-05, "loss": 0.0286, "step": 6140 }, { "grad_norm": 0.33513158559799194, "learning_rate": 9.855092363561437e-05, "loss": 0.041, "step": 6150 }, { "grad_norm": 0.2549797594547272, "learning_rate": 9.854300928449312e-05, "loss": 0.0438, "step": 6160 }, { "grad_norm": 0.2565065622329712, "learning_rate": 9.85350736990104e-05, "loss": 0.0331, "step": 6170 }, { "grad_norm": 0.16878558695316315, "learning_rate": 9.852711688263752e-05, "loss": 0.0302, "step": 6180 }, { "grad_norm": 0.1590256690979004, "learning_rate": 9.851913883885503e-05, "loss": 0.0323, "step": 6190 }, { "grad_norm": 0.6190244555473328, "learning_rate": 9.851113957115283e-05, "loss": 0.0287, "step": 6200 }, { "grad_norm": 0.19470371305942535, "learning_rate": 9.850311908303005e-05, "loss": 0.0368, "step": 6210 }, { "grad_norm": 0.2859501838684082, "learning_rate": 9.849507737799514e-05, "loss": 0.0305, "step": 6220 }, { "grad_norm": 0.15032227337360382, "learning_rate": 9.84870144595658e-05, "loss": 0.0317, "step": 6230 }, { "grad_norm": 0.178712397813797, "learning_rate": 9.847893033126903e-05, "loss": 0.0308, "step": 6240 }, { "grad_norm": 0.18251362442970276, "learning_rate": 9.847082499664111e-05, "loss": 0.0313, "step": 6250 }, { "grad_norm": 0.12848588824272156, "learning_rate": 9.846269845922758e-05, "loss": 0.025, "step": 6260 }, { "grad_norm": 0.25801461935043335, "learning_rate": 9.845455072258326e-05, "loss": 0.0261, "step": 6270 }, { "grad_norm": 0.14103084802627563, "learning_rate": 9.844638179027226e-05, "loss": 0.0302, "step": 6280 }, { "grad_norm": 0.21695253252983093, "learning_rate": 9.843819166586795e-05, "loss": 0.028, "step": 6290 }, { "grad_norm": 0.147353857755661, "learning_rate": 9.842998035295294e-05, "loss": 0.0313, "step": 6300 }, { "grad_norm": 0.18848997354507446, "learning_rate": 9.842174785511918e-05, "loss": 0.0257, "step": 6310 }, { "grad_norm": 0.40367844700813293, "learning_rate": 9.841349417596779e-05, "loss": 0.0282, "step": 6320 }, { "grad_norm": 0.17865662276744843, "learning_rate": 9.840521931910926e-05, "loss": 0.0277, "step": 6330 }, { "grad_norm": 0.2877201735973358, "learning_rate": 9.839692328816327e-05, "loss": 0.0389, "step": 6340 }, { "grad_norm": 0.16584505140781403, "learning_rate": 9.838860608675879e-05, "loss": 0.0305, "step": 6350 }, { "grad_norm": 0.16781306266784668, "learning_rate": 9.838026771853406e-05, "loss": 0.0299, "step": 6360 }, { "grad_norm": 0.20933431386947632, "learning_rate": 9.837190818713655e-05, "loss": 0.0312, "step": 6370 }, { "grad_norm": 0.5052786469459534, "learning_rate": 9.836352749622299e-05, "loss": 0.0301, "step": 6380 }, { "grad_norm": 0.24118857085704803, "learning_rate": 9.835512564945941e-05, "loss": 0.029, "step": 6390 }, { "grad_norm": 0.26470041275024414, "learning_rate": 9.834670265052104e-05, "loss": 0.0323, "step": 6400 }, { "grad_norm": 0.17050804197788239, "learning_rate": 9.833825850309239e-05, "loss": 0.0253, "step": 6410 }, { "grad_norm": 0.21122606098651886, "learning_rate": 9.832979321086723e-05, "loss": 0.0299, "step": 6420 }, { "grad_norm": 0.20102821290493011, "learning_rate": 9.832130677754854e-05, "loss": 0.0285, "step": 6430 }, { "grad_norm": 0.17979910969734192, "learning_rate": 9.831279920684861e-05, "loss": 0.028, "step": 6440 }, { "grad_norm": 0.17756310105323792, "learning_rate": 9.830427050248891e-05, "loss": 0.0248, "step": 6450 }, { "grad_norm": 0.20044486224651337, "learning_rate": 9.829572066820021e-05, "loss": 0.027, "step": 6460 }, { "grad_norm": 0.156093567609787, "learning_rate": 9.828714970772247e-05, "loss": 0.0253, "step": 6470 }, { "grad_norm": 0.30599749088287354, "learning_rate": 9.827855762480493e-05, "loss": 0.0252, "step": 6480 }, { "grad_norm": 0.2826230227947235, "learning_rate": 9.826994442320608e-05, "loss": 0.0286, "step": 6490 }, { "grad_norm": 0.11617647856473923, "learning_rate": 9.82613101066936e-05, "loss": 0.0224, "step": 6500 }, { "grad_norm": 0.1599779576063156, "learning_rate": 9.825265467904445e-05, "loss": 0.0239, "step": 6510 }, { "grad_norm": 0.2639065086841583, "learning_rate": 9.82439781440448e-05, "loss": 0.028, "step": 6520 }, { "grad_norm": 0.16641873121261597, "learning_rate": 9.823528050549006e-05, "loss": 0.0222, "step": 6530 }, { "grad_norm": 0.20306938886642456, "learning_rate": 9.822656176718487e-05, "loss": 0.0301, "step": 6540 }, { "grad_norm": 0.3540564477443695, "learning_rate": 9.82178219329431e-05, "loss": 0.026, "step": 6550 }, { "grad_norm": 0.185908704996109, "learning_rate": 9.820906100658789e-05, "loss": 0.0281, "step": 6560 }, { "grad_norm": 0.23113799095153809, "learning_rate": 9.82002789919515e-05, "loss": 0.0278, "step": 6570 }, { "grad_norm": 0.18169401586055756, "learning_rate": 9.819147589287554e-05, "loss": 0.022, "step": 6580 }, { "grad_norm": 0.3472957909107208, "learning_rate": 9.818265171321074e-05, "loss": 0.0334, "step": 6590 }, { "grad_norm": 0.29523760080337524, "learning_rate": 9.817380645681714e-05, "loss": 0.0282, "step": 6600 }, { "grad_norm": 0.21449755132198334, "learning_rate": 9.816494012756392e-05, "loss": 0.0238, "step": 6610 }, { "grad_norm": 0.12915940582752228, "learning_rate": 9.815605272932953e-05, "loss": 0.0253, "step": 6620 }, { "grad_norm": 0.17374096810817719, "learning_rate": 9.814714426600162e-05, "loss": 0.025, "step": 6630 }, { "grad_norm": 0.10276525467634201, "learning_rate": 9.813821474147704e-05, "loss": 0.0244, "step": 6640 }, { "grad_norm": 0.17661243677139282, "learning_rate": 9.812926415966189e-05, "loss": 0.0283, "step": 6650 }, { "grad_norm": 0.26773005723953247, "learning_rate": 9.812029252447143e-05, "loss": 0.0232, "step": 6660 }, { "grad_norm": 0.14911393821239471, "learning_rate": 9.811129983983018e-05, "loss": 0.0203, "step": 6670 }, { "grad_norm": 0.214217409491539, "learning_rate": 9.810228610967185e-05, "loss": 0.0242, "step": 6680 }, { "grad_norm": 0.20019859075546265, "learning_rate": 9.809325133793934e-05, "loss": 0.0256, "step": 6690 }, { "grad_norm": 0.2735741436481476, "learning_rate": 9.808419552858477e-05, "loss": 0.024, "step": 6700 }, { "grad_norm": 0.26965072751045227, "learning_rate": 9.807511868556944e-05, "loss": 0.0239, "step": 6710 }, { "grad_norm": 0.28713011741638184, "learning_rate": 9.806602081286392e-05, "loss": 0.0265, "step": 6720 }, { "grad_norm": 0.2899388372898102, "learning_rate": 9.805690191444785e-05, "loss": 0.0269, "step": 6730 }, { "grad_norm": 0.1663220226764679, "learning_rate": 9.804776199431022e-05, "loss": 0.035, "step": 6740 }, { "grad_norm": 0.16675089299678802, "learning_rate": 9.80386010564491e-05, "loss": 0.0271, "step": 6750 }, { "grad_norm": 0.27202853560447693, "learning_rate": 9.802941910487183e-05, "loss": 0.0276, "step": 6760 }, { "grad_norm": 0.13992314040660858, "learning_rate": 9.802021614359487e-05, "loss": 0.0264, "step": 6770 }, { "grad_norm": 0.19707123935222626, "learning_rate": 9.801099217664394e-05, "loss": 0.0262, "step": 6780 }, { "grad_norm": 0.18878750503063202, "learning_rate": 9.800174720805388e-05, "loss": 0.0233, "step": 6790 }, { "grad_norm": 0.36596420407295227, "learning_rate": 9.799248124186878e-05, "loss": 0.0272, "step": 6800 }, { "grad_norm": 0.14866149425506592, "learning_rate": 9.798319428214187e-05, "loss": 0.0203, "step": 6810 }, { "grad_norm": 0.27526065707206726, "learning_rate": 9.79738863329356e-05, "loss": 0.027, "step": 6820 }, { "grad_norm": 0.179055318236351, "learning_rate": 9.79645573983216e-05, "loss": 0.0233, "step": 6830 }, { "grad_norm": 0.23409713804721832, "learning_rate": 9.795520748238062e-05, "loss": 0.0255, "step": 6840 }, { "grad_norm": 0.21187832951545715, "learning_rate": 9.794583658920264e-05, "loss": 0.0254, "step": 6850 }, { "grad_norm": 0.2223644256591797, "learning_rate": 9.793644472288683e-05, "loss": 0.0234, "step": 6860 }, { "grad_norm": 0.32143792510032654, "learning_rate": 9.79270318875415e-05, "loss": 0.031, "step": 6870 }, { "grad_norm": 0.19008681178092957, "learning_rate": 9.791759808728416e-05, "loss": 0.03, "step": 6880 }, { "grad_norm": 0.21130914986133575, "learning_rate": 9.790814332624143e-05, "loss": 0.0268, "step": 6890 }, { "grad_norm": 0.2718564569950104, "learning_rate": 9.789866760854919e-05, "loss": 0.0304, "step": 6900 }, { "grad_norm": 0.19052070379257202, "learning_rate": 9.788917093835242e-05, "loss": 0.0261, "step": 6910 }, { "grad_norm": 0.1565798968076706, "learning_rate": 9.787965331980528e-05, "loss": 0.0281, "step": 6920 }, { "grad_norm": 0.1880509853363037, "learning_rate": 9.787011475707111e-05, "loss": 0.029, "step": 6930 }, { "grad_norm": 0.19279849529266357, "learning_rate": 9.786055525432241e-05, "loss": 0.0236, "step": 6940 }, { "grad_norm": 0.19657890498638153, "learning_rate": 9.78509748157408e-05, "loss": 0.026, "step": 6950 }, { "grad_norm": 0.39481261372566223, "learning_rate": 9.784137344551713e-05, "loss": 0.0273, "step": 6960 }, { "grad_norm": 0.1548413634300232, "learning_rate": 9.783175114785134e-05, "loss": 0.0294, "step": 6970 }, { "grad_norm": 0.1807706356048584, "learning_rate": 9.782210792695254e-05, "loss": 0.033, "step": 6980 }, { "grad_norm": 0.20189127326011658, "learning_rate": 9.781244378703904e-05, "loss": 0.0267, "step": 6990 }, { "grad_norm": 0.22851674258708954, "learning_rate": 9.780275873233824e-05, "loss": 0.0303, "step": 7000 }, { "grad_norm": 0.2267526388168335, "learning_rate": 9.77930527670867e-05, "loss": 0.0345, "step": 7010 }, { "grad_norm": 0.25363412499427795, "learning_rate": 9.778332589553016e-05, "loss": 0.0274, "step": 7020 }, { "grad_norm": 0.2676391899585724, "learning_rate": 9.777357812192349e-05, "loss": 0.0394, "step": 7030 }, { "grad_norm": 0.21532562375068665, "learning_rate": 9.776380945053066e-05, "loss": 0.0276, "step": 7040 }, { "grad_norm": 0.1833852380514145, "learning_rate": 9.775401988562486e-05, "loss": 0.029, "step": 7050 }, { "grad_norm": 0.23947596549987793, "learning_rate": 9.774420943148835e-05, "loss": 0.0291, "step": 7060 }, { "grad_norm": 0.20723333954811096, "learning_rate": 9.773437809241257e-05, "loss": 0.0288, "step": 7070 }, { "grad_norm": 0.16840918362140656, "learning_rate": 9.772452587269808e-05, "loss": 0.0387, "step": 7080 }, { "grad_norm": 0.3655005395412445, "learning_rate": 9.771465277665457e-05, "loss": 0.0336, "step": 7090 }, { "grad_norm": 0.31590718030929565, "learning_rate": 9.770475880860089e-05, "loss": 0.0313, "step": 7100 }, { "grad_norm": 0.34025660157203674, "learning_rate": 9.769484397286495e-05, "loss": 0.0327, "step": 7110 }, { "grad_norm": 0.26090914011001587, "learning_rate": 9.768490827378388e-05, "loss": 0.0313, "step": 7120 }, { "grad_norm": 0.2383711040019989, "learning_rate": 9.767495171570387e-05, "loss": 0.0304, "step": 7130 }, { "grad_norm": 0.2416400909423828, "learning_rate": 9.766497430298027e-05, "loss": 0.0476, "step": 7140 }, { "grad_norm": 0.204457625746727, "learning_rate": 9.765497603997754e-05, "loss": 0.039, "step": 7150 }, { "grad_norm": 0.373238742351532, "learning_rate": 9.764495693106926e-05, "loss": 0.0361, "step": 7160 }, { "grad_norm": 0.24625295400619507, "learning_rate": 9.76349169806381e-05, "loss": 0.0318, "step": 7170 }, { "grad_norm": 0.3686094880104065, "learning_rate": 9.762485619307591e-05, "loss": 0.0339, "step": 7180 }, { "grad_norm": 0.2004641741514206, "learning_rate": 9.76147745727836e-05, "loss": 0.036, "step": 7190 }, { "grad_norm": 0.345085084438324, "learning_rate": 9.760467212417124e-05, "loss": 0.0328, "step": 7200 }, { "grad_norm": 0.23240892589092255, "learning_rate": 9.759454885165796e-05, "loss": 0.0347, "step": 7210 }, { "grad_norm": 0.1739155799150467, "learning_rate": 9.7584404759672e-05, "loss": 0.0355, "step": 7220 }, { "grad_norm": 0.2311200648546219, "learning_rate": 9.75742398526508e-05, "loss": 0.0319, "step": 7230 }, { "grad_norm": 0.23508308827877045, "learning_rate": 9.756405413504077e-05, "loss": 0.0316, "step": 7240 }, { "grad_norm": 0.14462445676326752, "learning_rate": 9.755384761129752e-05, "loss": 0.0309, "step": 7250 }, { "grad_norm": 0.2957500219345093, "learning_rate": 9.754362028588572e-05, "loss": 0.0281, "step": 7260 }, { "grad_norm": 0.22799858450889587, "learning_rate": 9.753337216327917e-05, "loss": 0.0311, "step": 7270 }, { "grad_norm": 0.2241506725549698, "learning_rate": 9.752310324796073e-05, "loss": 0.0292, "step": 7280 }, { "grad_norm": 0.3766356110572815, "learning_rate": 9.751281354442235e-05, "loss": 0.0335, "step": 7290 }, { "grad_norm": 0.2708972990512848, "learning_rate": 9.750250305716513e-05, "loss": 0.034, "step": 7300 }, { "grad_norm": 0.16895490884780884, "learning_rate": 9.749217179069923e-05, "loss": 0.0296, "step": 7310 }, { "grad_norm": 0.1945081204175949, "learning_rate": 9.748181974954386e-05, "loss": 0.0303, "step": 7320 }, { "grad_norm": 0.15132448077201843, "learning_rate": 9.74714469382274e-05, "loss": 0.0286, "step": 7330 }, { "grad_norm": 0.19937288761138916, "learning_rate": 9.746105336128724e-05, "loss": 0.0353, "step": 7340 }, { "grad_norm": 0.31199100613594055, "learning_rate": 9.745063902326988e-05, "loss": 0.0328, "step": 7350 }, { "grad_norm": 0.2256213277578354, "learning_rate": 9.744020392873093e-05, "loss": 0.0331, "step": 7360 }, { "grad_norm": 0.23606042563915253, "learning_rate": 9.742974808223504e-05, "loss": 0.0326, "step": 7370 }, { "grad_norm": 0.15146282315254211, "learning_rate": 9.741927148835596e-05, "loss": 0.0295, "step": 7380 }, { "grad_norm": 0.16701896488666534, "learning_rate": 9.740877415167648e-05, "loss": 0.0304, "step": 7390 }, { "grad_norm": 0.1321386843919754, "learning_rate": 9.739825607678854e-05, "loss": 0.0317, "step": 7400 }, { "grad_norm": 0.20294056832790375, "learning_rate": 9.738771726829308e-05, "loss": 0.0336, "step": 7410 }, { "grad_norm": 0.2281942218542099, "learning_rate": 9.737715773080011e-05, "loss": 0.0355, "step": 7420 }, { "grad_norm": 0.18607783317565918, "learning_rate": 9.736657746892877e-05, "loss": 0.0337, "step": 7430 }, { "grad_norm": 0.20770321786403656, "learning_rate": 9.735597648730718e-05, "loss": 0.036, "step": 7440 }, { "grad_norm": 0.16197611391544342, "learning_rate": 9.734535479057262e-05, "loss": 0.0304, "step": 7450 }, { "grad_norm": 0.21719564497470856, "learning_rate": 9.733471238337136e-05, "loss": 0.0327, "step": 7460 }, { "grad_norm": 0.43547365069389343, "learning_rate": 9.732404927035871e-05, "loss": 0.0373, "step": 7470 }, { "grad_norm": 0.33916744589805603, "learning_rate": 9.731336545619915e-05, "loss": 0.0389, "step": 7480 }, { "grad_norm": 0.26533442735671997, "learning_rate": 9.73026609455661e-05, "loss": 0.0368, "step": 7490 }, { "grad_norm": 0.2545904219150543, "learning_rate": 9.72919357431421e-05, "loss": 0.0324, "step": 7500 }, { "grad_norm": 0.4475284218788147, "learning_rate": 9.72811898536187e-05, "loss": 0.0312, "step": 7510 }, { "grad_norm": 0.21808183193206787, "learning_rate": 9.72704232816965e-05, "loss": 0.032, "step": 7520 }, { "grad_norm": 0.15662431716918945, "learning_rate": 9.725963603208519e-05, "loss": 0.032, "step": 7530 }, { "grad_norm": 0.19488798081874847, "learning_rate": 9.724882810950348e-05, "loss": 0.0351, "step": 7540 }, { "grad_norm": 0.19852109253406525, "learning_rate": 9.723799951867913e-05, "loss": 0.0316, "step": 7550 }, { "grad_norm": 0.19987745583057404, "learning_rate": 9.722715026434889e-05, "loss": 0.0343, "step": 7560 }, { "grad_norm": 0.2385755479335785, "learning_rate": 9.721628035125866e-05, "loss": 0.0411, "step": 7570 }, { "grad_norm": 0.32667335867881775, "learning_rate": 9.720538978416325e-05, "loss": 0.0375, "step": 7580 }, { "grad_norm": 0.20085924863815308, "learning_rate": 9.71944785678266e-05, "loss": 0.0372, "step": 7590 }, { "grad_norm": 0.31946179270744324, "learning_rate": 9.718354670702161e-05, "loss": 0.0441, "step": 7600 }, { "grad_norm": 0.1726575642824173, "learning_rate": 9.717259420653029e-05, "loss": 0.0381, "step": 7610 }, { "grad_norm": 0.17952850461006165, "learning_rate": 9.716162107114361e-05, "loss": 0.0344, "step": 7620 }, { "grad_norm": 0.17608606815338135, "learning_rate": 9.71506273056616e-05, "loss": 0.0357, "step": 7630 }, { "grad_norm": 0.19620245695114136, "learning_rate": 9.713961291489331e-05, "loss": 0.033, "step": 7640 }, { "grad_norm": 0.206907719373703, "learning_rate": 9.71285779036568e-05, "loss": 0.0334, "step": 7650 }, { "grad_norm": 0.16825468838214874, "learning_rate": 9.711752227677916e-05, "loss": 0.0316, "step": 7660 }, { "grad_norm": 0.3277825713157654, "learning_rate": 9.710644603909649e-05, "loss": 0.035, "step": 7670 }, { "grad_norm": 0.22960041463375092, "learning_rate": 9.709534919545393e-05, "loss": 0.0342, "step": 7680 }, { "grad_norm": 0.21421344578266144, "learning_rate": 9.708423175070563e-05, "loss": 0.0334, "step": 7690 }, { "grad_norm": 0.17156197130680084, "learning_rate": 9.707309370971468e-05, "loss": 0.0262, "step": 7700 }, { "grad_norm": 0.355979323387146, "learning_rate": 9.70619350773533e-05, "loss": 0.0343, "step": 7710 }, { "grad_norm": 0.21902745962142944, "learning_rate": 9.70507558585026e-05, "loss": 0.0363, "step": 7720 }, { "grad_norm": 0.24764594435691833, "learning_rate": 9.703955605805281e-05, "loss": 0.0327, "step": 7730 }, { "grad_norm": 0.20965486764907837, "learning_rate": 9.702833568090306e-05, "loss": 0.0347, "step": 7740 }, { "grad_norm": 0.20050080120563507, "learning_rate": 9.701709473196154e-05, "loss": 0.03, "step": 7750 }, { "grad_norm": 0.1833415925502777, "learning_rate": 9.700583321614541e-05, "loss": 0.0351, "step": 7760 }, { "grad_norm": 0.16146747767925262, "learning_rate": 9.699455113838085e-05, "loss": 0.0326, "step": 7770 }, { "grad_norm": 0.2562040388584137, "learning_rate": 9.698324850360303e-05, "loss": 0.0315, "step": 7780 }, { "grad_norm": 0.2509492337703705, "learning_rate": 9.69719253167561e-05, "loss": 0.0331, "step": 7790 }, { "grad_norm": 0.29203492403030396, "learning_rate": 9.69605815827932e-05, "loss": 0.0313, "step": 7800 }, { "grad_norm": 0.1657320261001587, "learning_rate": 9.694921730667647e-05, "loss": 0.0341, "step": 7810 }, { "grad_norm": 0.16629011929035187, "learning_rate": 9.693783249337705e-05, "loss": 0.0304, "step": 7820 }, { "grad_norm": 0.22137151658535004, "learning_rate": 9.692642714787501e-05, "loss": 0.0327, "step": 7830 }, { "grad_norm": 0.17869582772254944, "learning_rate": 9.691500127515945e-05, "loss": 0.0321, "step": 7840 }, { "grad_norm": 0.2143394649028778, "learning_rate": 9.690355488022844e-05, "loss": 0.0268, "step": 7850 }, { "grad_norm": 0.1830715835094452, "learning_rate": 9.689208796808902e-05, "loss": 0.0314, "step": 7860 }, { "grad_norm": 0.2362588793039322, "learning_rate": 9.68806005437572e-05, "loss": 0.0338, "step": 7870 }, { "grad_norm": 0.2782280445098877, "learning_rate": 9.686909261225796e-05, "loss": 0.0309, "step": 7880 }, { "grad_norm": 0.1941898763179779, "learning_rate": 9.685756417862531e-05, "loss": 0.0287, "step": 7890 }, { "grad_norm": 0.14631469547748566, "learning_rate": 9.684601524790212e-05, "loss": 0.0376, "step": 7900 }, { "grad_norm": 0.25684571266174316, "learning_rate": 9.68344458251403e-05, "loss": 0.0368, "step": 7910 }, { "grad_norm": 0.2534716725349426, "learning_rate": 9.682285591540072e-05, "loss": 0.039, "step": 7920 }, { "grad_norm": 0.1781902015209198, "learning_rate": 9.681124552375322e-05, "loss": 0.0295, "step": 7930 }, { "grad_norm": 0.2834855020046234, "learning_rate": 9.679961465527654e-05, "loss": 0.0305, "step": 7940 }, { "grad_norm": 0.24427521228790283, "learning_rate": 9.678796331505843e-05, "loss": 0.0278, "step": 7950 }, { "grad_norm": 0.10230432450771332, "learning_rate": 9.677629150819558e-05, "loss": 0.0292, "step": 7960 }, { "grad_norm": 0.23986569046974182, "learning_rate": 9.676459923979366e-05, "loss": 0.0254, "step": 7970 }, { "grad_norm": 0.18821099400520325, "learning_rate": 9.675288651496723e-05, "loss": 0.0277, "step": 7980 }, { "grad_norm": 0.12477374076843262, "learning_rate": 9.674115333883986e-05, "loss": 0.0289, "step": 7990 }, { "grad_norm": 0.23939037322998047, "learning_rate": 9.672939971654402e-05, "loss": 0.027, "step": 8000 }, { "grad_norm": 0.19288049638271332, "learning_rate": 9.671762565322117e-05, "loss": 0.0271, "step": 8010 }, { "grad_norm": 0.21086320281028748, "learning_rate": 9.670583115402166e-05, "loss": 0.0278, "step": 8020 }, { "grad_norm": 0.11230416595935822, "learning_rate": 9.669401622410482e-05, "loss": 0.0244, "step": 8030 }, { "grad_norm": 0.261640340089798, "learning_rate": 9.668218086863887e-05, "loss": 0.0279, "step": 8040 }, { "grad_norm": 0.20032133162021637, "learning_rate": 9.667032509280103e-05, "loss": 0.0318, "step": 8050 }, { "grad_norm": 0.23750869929790497, "learning_rate": 9.665844890177743e-05, "loss": 0.0269, "step": 8060 }, { "grad_norm": 0.1459992378950119, "learning_rate": 9.664655230076309e-05, "loss": 0.0315, "step": 8070 }, { "grad_norm": 0.3633784055709839, "learning_rate": 9.663463529496199e-05, "loss": 0.0375, "step": 8080 }, { "grad_norm": 0.25279682874679565, "learning_rate": 9.662269788958705e-05, "loss": 0.0286, "step": 8090 }, { "grad_norm": 0.20803655683994293, "learning_rate": 9.661074008986008e-05, "loss": 0.0296, "step": 8100 }, { "grad_norm": 0.21200431883335114, "learning_rate": 9.659876190101184e-05, "loss": 0.0366, "step": 8110 }, { "grad_norm": 0.12428006529808044, "learning_rate": 9.6586763328282e-05, "loss": 0.0258, "step": 8120 }, { "grad_norm": 0.394804447889328, "learning_rate": 9.657474437691913e-05, "loss": 0.0391, "step": 8130 }, { "grad_norm": 0.2254556566476822, "learning_rate": 9.656270505218073e-05, "loss": 0.0311, "step": 8140 }, { "grad_norm": 0.20987841486930847, "learning_rate": 9.655064535933321e-05, "loss": 0.0277, "step": 8150 }, { "grad_norm": 0.18746742606163025, "learning_rate": 9.653856530365189e-05, "loss": 0.0285, "step": 8160 }, { "grad_norm": 0.33110281825065613, "learning_rate": 9.6526464890421e-05, "loss": 0.0327, "step": 8170 }, { "grad_norm": 0.16701211035251617, "learning_rate": 9.651434412493367e-05, "loss": 0.0273, "step": 8180 }, { "grad_norm": 0.16969642043113708, "learning_rate": 9.650220301249195e-05, "loss": 0.0304, "step": 8190 }, { "grad_norm": 0.20343460142612457, "learning_rate": 9.649004155840675e-05, "loss": 0.0258, "step": 8200 }, { "grad_norm": 0.34077075123786926, "learning_rate": 9.647785976799792e-05, "loss": 0.0274, "step": 8210 }, { "grad_norm": 0.24407491087913513, "learning_rate": 9.646565764659417e-05, "loss": 0.0279, "step": 8220 }, { "grad_norm": 0.22233925759792328, "learning_rate": 9.645343519953314e-05, "loss": 0.0302, "step": 8230 }, { "grad_norm": 0.26643991470336914, "learning_rate": 9.644119243216135e-05, "loss": 0.0273, "step": 8240 }, { "grad_norm": 0.11854870617389679, "learning_rate": 9.642892934983417e-05, "loss": 0.0268, "step": 8250 }, { "grad_norm": 0.17468130588531494, "learning_rate": 9.641664595791593e-05, "loss": 0.024, "step": 8260 }, { "grad_norm": 0.1359395980834961, "learning_rate": 9.640434226177977e-05, "loss": 0.0252, "step": 8270 }, { "grad_norm": 0.2752685546875, "learning_rate": 9.639201826680777e-05, "loss": 0.0331, "step": 8280 }, { "grad_norm": 0.15378624200820923, "learning_rate": 9.637967397839083e-05, "loss": 0.0263, "step": 8290 }, { "grad_norm": 0.22038336098194122, "learning_rate": 9.63673094019288e-05, "loss": 0.0299, "step": 8300 }, { "grad_norm": 0.2861631512641907, "learning_rate": 9.635492454283035e-05, "loss": 0.0228, "step": 8310 }, { "grad_norm": 0.11572781205177307, "learning_rate": 9.634251940651302e-05, "loss": 0.0242, "step": 8320 }, { "grad_norm": 0.1797153651714325, "learning_rate": 9.633009399840327e-05, "loss": 0.0244, "step": 8330 }, { "grad_norm": 0.15558268129825592, "learning_rate": 9.631764832393639e-05, "loss": 0.0242, "step": 8340 }, { "grad_norm": 0.26697802543640137, "learning_rate": 9.630518238855651e-05, "loss": 0.0312, "step": 8350 }, { "grad_norm": 0.232689768075943, "learning_rate": 9.629269619771668e-05, "loss": 0.0275, "step": 8360 }, { "grad_norm": 0.1805265098810196, "learning_rate": 9.628018975687878e-05, "loss": 0.0269, "step": 8370 }, { "grad_norm": 0.1682014763355255, "learning_rate": 9.626766307151355e-05, "loss": 0.0256, "step": 8380 }, { "grad_norm": 0.1641201674938202, "learning_rate": 9.62551161471006e-05, "loss": 0.0241, "step": 8390 }, { "grad_norm": 0.14837408065795898, "learning_rate": 9.624254898912837e-05, "loss": 0.022, "step": 8400 }, { "grad_norm": 0.15358154475688934, "learning_rate": 9.622996160309414e-05, "loss": 0.0297, "step": 8410 }, { "grad_norm": 0.19763629138469696, "learning_rate": 9.62173539945041e-05, "loss": 0.0259, "step": 8420 }, { "grad_norm": 0.33756837248802185, "learning_rate": 9.620472616887323e-05, "loss": 0.0302, "step": 8430 }, { "grad_norm": 0.19611786305904388, "learning_rate": 9.619207813172536e-05, "loss": 0.0264, "step": 8440 }, { "grad_norm": 0.18838128447532654, "learning_rate": 9.617940988859318e-05, "loss": 0.0318, "step": 8450 }, { "grad_norm": 0.22106847167015076, "learning_rate": 9.616672144501821e-05, "loss": 0.0273, "step": 8460 }, { "grad_norm": 0.1808529943227768, "learning_rate": 9.615401280655081e-05, "loss": 0.0249, "step": 8470 }, { "grad_norm": 0.17592599987983704, "learning_rate": 9.614128397875017e-05, "loss": 0.0258, "step": 8480 }, { "grad_norm": 0.3061319589614868, "learning_rate": 9.612853496718429e-05, "loss": 0.0279, "step": 8490 }, { "grad_norm": 0.46806827187538147, "learning_rate": 9.611576577743004e-05, "loss": 0.0271, "step": 8500 }, { "grad_norm": 0.23881563544273376, "learning_rate": 9.610297641507313e-05, "loss": 0.0268, "step": 8510 }, { "grad_norm": 0.15583860874176025, "learning_rate": 9.6090166885708e-05, "loss": 0.0237, "step": 8520 }, { "grad_norm": 0.15887945890426636, "learning_rate": 9.607733719493798e-05, "loss": 0.0281, "step": 8530 }, { "grad_norm": 0.19332051277160645, "learning_rate": 9.606448734837527e-05, "loss": 0.0267, "step": 8540 }, { "grad_norm": 0.2027137279510498, "learning_rate": 9.605161735164079e-05, "loss": 0.0301, "step": 8550 }, { "grad_norm": 0.18600745499134064, "learning_rate": 9.60387272103643e-05, "loss": 0.0264, "step": 8560 }, { "grad_norm": 0.28709813952445984, "learning_rate": 9.60258169301844e-05, "loss": 0.0453, "step": 8570 }, { "grad_norm": 0.46798235177993774, "learning_rate": 9.601288651674851e-05, "loss": 0.0384, "step": 8580 }, { "grad_norm": 0.21497027575969696, "learning_rate": 9.59999359757128e-05, "loss": 0.0325, "step": 8590 }, { "grad_norm": 0.2217976152896881, "learning_rate": 9.598696531274227e-05, "loss": 0.0298, "step": 8600 }, { "grad_norm": 0.2594149112701416, "learning_rate": 9.597397453351076e-05, "loss": 0.0312, "step": 8610 }, { "grad_norm": 0.32169273495674133, "learning_rate": 9.596096364370087e-05, "loss": 0.0359, "step": 8620 }, { "grad_norm": 0.2941077947616577, "learning_rate": 9.594793264900399e-05, "loss": 0.0322, "step": 8630 }, { "grad_norm": 0.198715940117836, "learning_rate": 9.593488155512032e-05, "loss": 0.0317, "step": 8640 }, { "grad_norm": 0.23065036535263062, "learning_rate": 9.592181036775886e-05, "loss": 0.035, "step": 8650 }, { "grad_norm": 0.2459569126367569, "learning_rate": 9.590871909263741e-05, "loss": 0.0294, "step": 8660 }, { "grad_norm": 0.27135980129241943, "learning_rate": 9.589560773548252e-05, "loss": 0.0295, "step": 8670 }, { "grad_norm": 0.21448233723640442, "learning_rate": 9.588247630202952e-05, "loss": 0.023, "step": 8680 }, { "grad_norm": 0.19892236590385437, "learning_rate": 9.586932479802258e-05, "loss": 0.0275, "step": 8690 }, { "grad_norm": 0.1945362538099289, "learning_rate": 9.585615322921462e-05, "loss": 0.028, "step": 8700 }, { "grad_norm": 0.10635978728532791, "learning_rate": 9.58429616013673e-05, "loss": 0.0301, "step": 8710 }, { "grad_norm": 0.3648577630519867, "learning_rate": 9.58297499202511e-05, "loss": 0.0331, "step": 8720 }, { "grad_norm": 0.18359939754009247, "learning_rate": 9.581651819164525e-05, "loss": 0.0307, "step": 8730 }, { "grad_norm": 0.16747413575649261, "learning_rate": 9.580326642133777e-05, "loss": 0.0275, "step": 8740 }, { "grad_norm": 0.17895664274692535, "learning_rate": 9.578999461512544e-05, "loss": 0.0337, "step": 8750 }, { "grad_norm": 0.18911032378673553, "learning_rate": 9.577670277881378e-05, "loss": 0.0292, "step": 8760 }, { "grad_norm": 0.2490876317024231, "learning_rate": 9.57633909182171e-05, "loss": 0.033, "step": 8770 }, { "grad_norm": 0.12209452688694, "learning_rate": 9.575005903915847e-05, "loss": 0.0313, "step": 8780 }, { "grad_norm": 0.1720394641160965, "learning_rate": 9.573670714746972e-05, "loss": 0.0285, "step": 8790 }, { "grad_norm": 0.12659679353237152, "learning_rate": 9.572333524899136e-05, "loss": 0.0295, "step": 8800 }, { "grad_norm": 0.1482001394033432, "learning_rate": 9.570994334957278e-05, "loss": 0.0259, "step": 8810 }, { "grad_norm": 0.186662495136261, "learning_rate": 9.569653145507201e-05, "loss": 0.0252, "step": 8820 }, { "grad_norm": 0.2562216520309448, "learning_rate": 9.568309957135587e-05, "loss": 0.0267, "step": 8830 }, { "grad_norm": 0.23675741255283356, "learning_rate": 9.566964770429995e-05, "loss": 0.0248, "step": 8840 }, { "grad_norm": 0.14415250718593597, "learning_rate": 9.565617585978853e-05, "loss": 0.0245, "step": 8850 }, { "grad_norm": 0.1938789188861847, "learning_rate": 9.564268404371466e-05, "loss": 0.0255, "step": 8860 }, { "grad_norm": 0.3551104664802551, "learning_rate": 9.562917226198013e-05, "loss": 0.0293, "step": 8870 }, { "grad_norm": 0.1842319518327713, "learning_rate": 9.561564052049543e-05, "loss": 0.0304, "step": 8880 }, { "grad_norm": 0.22825148701667786, "learning_rate": 9.560208882517982e-05, "loss": 0.032, "step": 8890 }, { "grad_norm": 0.21792633831501007, "learning_rate": 9.558851718196127e-05, "loss": 0.0323, "step": 8900 }, { "grad_norm": 0.20210599899291992, "learning_rate": 9.557492559677646e-05, "loss": 0.031, "step": 8910 }, { "grad_norm": 0.26681315898895264, "learning_rate": 9.556131407557082e-05, "loss": 0.0337, "step": 8920 }, { "grad_norm": 0.2747860550880432, "learning_rate": 9.554768262429853e-05, "loss": 0.0258, "step": 8930 }, { "grad_norm": 0.24156400561332703, "learning_rate": 9.553403124892239e-05, "loss": 0.0288, "step": 8940 }, { "grad_norm": 0.2318858802318573, "learning_rate": 9.552035995541402e-05, "loss": 0.0307, "step": 8950 }, { "grad_norm": 0.1383676826953888, "learning_rate": 9.550666874975368e-05, "loss": 0.0255, "step": 8960 }, { "grad_norm": 0.2579522132873535, "learning_rate": 9.549295763793038e-05, "loss": 0.0278, "step": 8970 }, { "grad_norm": 0.14902204275131226, "learning_rate": 9.547922662594183e-05, "loss": 0.0304, "step": 8980 }, { "grad_norm": 0.2142041027545929, "learning_rate": 9.546547571979443e-05, "loss": 0.0388, "step": 8990 }, { "grad_norm": 0.12199276685714722, "learning_rate": 9.545170492550331e-05, "loss": 0.0319, "step": 9000 }, { "grad_norm": 0.22957295179367065, "learning_rate": 9.543791424909226e-05, "loss": 0.0314, "step": 9010 }, { "grad_norm": 0.24963994324207306, "learning_rate": 9.542410369659382e-05, "loss": 0.0293, "step": 9020 }, { "grad_norm": 0.23506206274032593, "learning_rate": 9.541027327404916e-05, "loss": 0.034, "step": 9030 }, { "grad_norm": 0.397797554731369, "learning_rate": 9.539642298750821e-05, "loss": 0.0345, "step": 9040 }, { "grad_norm": 0.14651277661323547, "learning_rate": 9.538255284302954e-05, "loss": 0.0289, "step": 9050 }, { "grad_norm": 0.22680282592773438, "learning_rate": 9.536866284668043e-05, "loss": 0.0351, "step": 9060 }, { "grad_norm": 0.17493416368961334, "learning_rate": 9.53547530045368e-05, "loss": 0.0261, "step": 9070 }, { "grad_norm": 0.09485752880573273, "learning_rate": 9.534082332268335e-05, "loss": 0.0263, "step": 9080 }, { "grad_norm": 0.1981823742389679, "learning_rate": 9.532687380721334e-05, "loss": 0.0335, "step": 9090 }, { "grad_norm": 0.2526615560054779, "learning_rate": 9.53129044642288e-05, "loss": 0.0349, "step": 9100 }, { "grad_norm": 0.3105442225933075, "learning_rate": 9.529891529984039e-05, "loss": 0.0286, "step": 9110 }, { "grad_norm": 0.24266771972179413, "learning_rate": 9.528490632016743e-05, "loss": 0.0274, "step": 9120 }, { "grad_norm": 0.16535833477973938, "learning_rate": 9.52708775313379e-05, "loss": 0.0353, "step": 9130 }, { "grad_norm": 0.14487645030021667, "learning_rate": 9.525682893948853e-05, "loss": 0.0238, "step": 9140 }, { "grad_norm": 0.19166326522827148, "learning_rate": 9.524276055076463e-05, "loss": 0.0314, "step": 9150 }, { "grad_norm": 0.11872252821922302, "learning_rate": 9.522867237132016e-05, "loss": 0.0249, "step": 9160 }, { "grad_norm": 0.11931612342596054, "learning_rate": 9.52145644073178e-05, "loss": 0.0276, "step": 9170 }, { "grad_norm": 0.22768862545490265, "learning_rate": 9.520043666492884e-05, "loss": 0.0244, "step": 9180 }, { "grad_norm": 0.11656907945871353, "learning_rate": 9.518628915033323e-05, "loss": 0.0236, "step": 9190 }, { "grad_norm": 0.28544414043426514, "learning_rate": 9.517212186971957e-05, "loss": 0.0268, "step": 9200 }, { "grad_norm": 0.1744229644536972, "learning_rate": 9.515793482928515e-05, "loss": 0.0303, "step": 9210 }, { "grad_norm": 0.2983028292655945, "learning_rate": 9.514372803523582e-05, "loss": 0.0291, "step": 9220 }, { "grad_norm": 0.2030583769083023, "learning_rate": 9.512950149378613e-05, "loss": 0.0249, "step": 9230 }, { "grad_norm": 0.25356048345565796, "learning_rate": 9.511525521115925e-05, "loss": 0.0292, "step": 9240 }, { "grad_norm": 0.1919245421886444, "learning_rate": 9.510098919358698e-05, "loss": 0.0244, "step": 9250 }, { "grad_norm": 0.11149285733699799, "learning_rate": 9.508670344730979e-05, "loss": 0.0246, "step": 9260 }, { "grad_norm": 0.17343157529830933, "learning_rate": 9.507239797857672e-05, "loss": 0.0263, "step": 9270 }, { "grad_norm": 0.19590269029140472, "learning_rate": 9.505807279364548e-05, "loss": 0.0235, "step": 9280 }, { "grad_norm": 0.33931785821914673, "learning_rate": 9.504372789878239e-05, "loss": 0.0323, "step": 9290 }, { "grad_norm": 0.22265200316905975, "learning_rate": 9.502936330026239e-05, "loss": 0.0301, "step": 9300 }, { "grad_norm": 0.14442463219165802, "learning_rate": 9.501497900436906e-05, "loss": 0.0274, "step": 9310 }, { "grad_norm": 0.18601106107234955, "learning_rate": 9.500057501739457e-05, "loss": 0.0354, "step": 9320 }, { "grad_norm": 0.2454259693622589, "learning_rate": 9.49861513456397e-05, "loss": 0.0309, "step": 9330 }, { "grad_norm": 0.158721461892128, "learning_rate": 9.497170799541388e-05, "loss": 0.0293, "step": 9340 }, { "grad_norm": 0.21941454708576202, "learning_rate": 9.495724497303508e-05, "loss": 0.0238, "step": 9350 }, { "grad_norm": 0.13507384061813354, "learning_rate": 9.494276228482998e-05, "loss": 0.0233, "step": 9360 }, { "grad_norm": 0.175194650888443, "learning_rate": 9.492825993713374e-05, "loss": 0.0316, "step": 9370 }, { "grad_norm": 0.17412765324115753, "learning_rate": 9.491373793629023e-05, "loss": 0.0243, "step": 9380 }, { "grad_norm": 0.23291094601154327, "learning_rate": 9.489919628865182e-05, "loss": 0.0253, "step": 9390 }, { "grad_norm": 0.18446090817451477, "learning_rate": 9.488463500057955e-05, "loss": 0.0285, "step": 9400 }, { "grad_norm": 0.20466332137584686, "learning_rate": 9.487005407844302e-05, "loss": 0.0318, "step": 9410 }, { "grad_norm": 0.3191821873188019, "learning_rate": 9.485545352862039e-05, "loss": 0.0273, "step": 9420 }, { "grad_norm": 0.18797391653060913, "learning_rate": 9.48408333574985e-05, "loss": 0.0236, "step": 9430 }, { "grad_norm": 0.2725661098957062, "learning_rate": 9.482619357147264e-05, "loss": 0.0264, "step": 9440 }, { "grad_norm": 0.23266351222991943, "learning_rate": 9.481153417694679e-05, "loss": 0.0249, "step": 9450 }, { "grad_norm": 0.15745942294597626, "learning_rate": 9.479685518033347e-05, "loss": 0.0242, "step": 9460 }, { "grad_norm": 0.2934640645980835, "learning_rate": 9.478215658805375e-05, "loss": 0.0349, "step": 9470 }, { "grad_norm": 0.1925497204065323, "learning_rate": 9.47674384065373e-05, "loss": 0.0242, "step": 9480 }, { "grad_norm": 0.21698984503746033, "learning_rate": 9.475270064222237e-05, "loss": 0.0277, "step": 9490 }, { "grad_norm": 0.1941937804222107, "learning_rate": 9.473794330155572e-05, "loss": 0.0245, "step": 9500 }, { "grad_norm": 0.13564187288284302, "learning_rate": 9.472316639099276e-05, "loss": 0.0245, "step": 9510 }, { "grad_norm": 0.2408691644668579, "learning_rate": 9.470836991699739e-05, "loss": 0.0261, "step": 9520 }, { "grad_norm": 0.20509204268455505, "learning_rate": 9.469355388604208e-05, "loss": 0.0246, "step": 9530 }, { "grad_norm": 0.5675477981567383, "learning_rate": 9.467871830460787e-05, "loss": 0.0303, "step": 9540 }, { "grad_norm": 0.19734472036361694, "learning_rate": 9.466386317918436e-05, "loss": 0.0264, "step": 9550 }, { "grad_norm": 0.19212095439434052, "learning_rate": 9.464898851626969e-05, "loss": 0.0288, "step": 9560 }, { "grad_norm": 0.1828514188528061, "learning_rate": 9.463409432237051e-05, "loss": 0.0343, "step": 9570 }, { "grad_norm": 0.19556401669979095, "learning_rate": 9.461918060400209e-05, "loss": 0.0276, "step": 9580 }, { "grad_norm": 0.15438193082809448, "learning_rate": 9.460424736768816e-05, "loss": 0.0247, "step": 9590 }, { "grad_norm": 0.18921053409576416, "learning_rate": 9.458929461996105e-05, "loss": 0.022, "step": 9600 }, { "grad_norm": 0.16722707450389862, "learning_rate": 9.457432236736158e-05, "loss": 0.0205, "step": 9610 }, { "grad_norm": 0.14956772327423096, "learning_rate": 9.455933061643916e-05, "loss": 0.0244, "step": 9620 }, { "grad_norm": 0.10803282260894775, "learning_rate": 9.454431937375164e-05, "loss": 0.0238, "step": 9630 }, { "grad_norm": 0.20751719176769257, "learning_rate": 9.452928864586547e-05, "loss": 0.0231, "step": 9640 }, { "grad_norm": 0.1028289794921875, "learning_rate": 9.451423843935563e-05, "loss": 0.0249, "step": 9650 }, { "grad_norm": 0.16202932596206665, "learning_rate": 9.449916876080553e-05, "loss": 0.023, "step": 9660 }, { "grad_norm": 0.2520557641983032, "learning_rate": 9.44840796168072e-05, "loss": 0.0226, "step": 9670 }, { "grad_norm": 0.11422251909971237, "learning_rate": 9.446897101396115e-05, "loss": 0.0208, "step": 9680 }, { "grad_norm": 0.1259026676416397, "learning_rate": 9.445384295887638e-05, "loss": 0.0234, "step": 9690 }, { "grad_norm": 0.2066774070262909, "learning_rate": 9.443869545817043e-05, "loss": 0.0242, "step": 9700 }, { "grad_norm": 0.1951393485069275, "learning_rate": 9.442352851846929e-05, "loss": 0.0261, "step": 9710 }, { "grad_norm": 0.2697119414806366, "learning_rate": 9.440834214640755e-05, "loss": 0.0249, "step": 9720 }, { "grad_norm": 0.19335058331489563, "learning_rate": 9.439313634862823e-05, "loss": 0.0316, "step": 9730 }, { "grad_norm": 0.2528826594352722, "learning_rate": 9.437791113178282e-05, "loss": 0.0255, "step": 9740 }, { "grad_norm": 0.17375832796096802, "learning_rate": 9.43626665025314e-05, "loss": 0.0269, "step": 9750 }, { "grad_norm": 0.1841871738433838, "learning_rate": 9.434740246754248e-05, "loss": 0.0252, "step": 9760 }, { "grad_norm": 0.15324634313583374, "learning_rate": 9.433211903349304e-05, "loss": 0.0268, "step": 9770 }, { "grad_norm": 0.14595438539981842, "learning_rate": 9.431681620706858e-05, "loss": 0.0223, "step": 9780 }, { "grad_norm": 0.13418641686439514, "learning_rate": 9.43014939949631e-05, "loss": 0.0236, "step": 9790 }, { "grad_norm": 0.17360521852970123, "learning_rate": 9.428615240387904e-05, "loss": 0.0209, "step": 9800 }, { "grad_norm": 0.1763627827167511, "learning_rate": 9.427079144052732e-05, "loss": 0.0266, "step": 9810 }, { "grad_norm": 0.18888193368911743, "learning_rate": 9.425541111162739e-05, "loss": 0.0261, "step": 9820 }, { "grad_norm": 0.24116818606853485, "learning_rate": 9.424001142390709e-05, "loss": 0.027, "step": 9830 }, { "grad_norm": 0.11465919017791748, "learning_rate": 9.422459238410277e-05, "loss": 0.0284, "step": 9840 }, { "grad_norm": 0.2211756706237793, "learning_rate": 9.420915399895926e-05, "loss": 0.0276, "step": 9850 }, { "grad_norm": 0.10762306302785873, "learning_rate": 9.419369627522981e-05, "loss": 0.0234, "step": 9860 }, { "grad_norm": 0.19748252630233765, "learning_rate": 9.417821921967618e-05, "loss": 0.0232, "step": 9870 }, { "grad_norm": 0.14223244786262512, "learning_rate": 9.416272283906855e-05, "loss": 0.0264, "step": 9880 }, { "grad_norm": 0.11364094167947769, "learning_rate": 9.414720714018554e-05, "loss": 0.0221, "step": 9890 }, { "grad_norm": 0.10003038495779037, "learning_rate": 9.413167212981427e-05, "loss": 0.0237, "step": 9900 }, { "grad_norm": 0.20074430108070374, "learning_rate": 9.41161178147503e-05, "loss": 0.0277, "step": 9910 }, { "grad_norm": 0.156886488199234, "learning_rate": 9.410054420179755e-05, "loss": 0.0312, "step": 9920 }, { "grad_norm": 0.14896804094314575, "learning_rate": 9.408495129776851e-05, "loss": 0.0267, "step": 9930 }, { "grad_norm": 0.19259119033813477, "learning_rate": 9.406933910948403e-05, "loss": 0.0243, "step": 9940 }, { "grad_norm": 0.16491249203681946, "learning_rate": 9.40537076437734e-05, "loss": 0.027, "step": 9950 }, { "grad_norm": 0.3012794554233551, "learning_rate": 9.403805690747436e-05, "loss": 0.0248, "step": 9960 }, { "grad_norm": 0.20177023112773895, "learning_rate": 9.402238690743308e-05, "loss": 0.0262, "step": 9970 }, { "grad_norm": 0.13988752663135529, "learning_rate": 9.400669765050413e-05, "loss": 0.0295, "step": 9980 }, { "grad_norm": 0.22218534350395203, "learning_rate": 9.399098914355055e-05, "loss": 0.0303, "step": 9990 }, { "grad_norm": 0.20827670395374298, "learning_rate": 9.397526139344378e-05, "loss": 0.03, "step": 10000 }, { "grad_norm": 0.14256614446640015, "learning_rate": 9.395951440706362e-05, "loss": 0.0305, "step": 10010 }, { "grad_norm": 0.1894553005695343, "learning_rate": 9.394374819129839e-05, "loss": 0.0317, "step": 10020 }, { "grad_norm": 0.22861234843730927, "learning_rate": 9.392796275304474e-05, "loss": 0.0279, "step": 10030 }, { "grad_norm": 0.37326815724372864, "learning_rate": 9.391215809920778e-05, "loss": 0.0328, "step": 10040 }, { "grad_norm": 0.3018113970756531, "learning_rate": 9.389633423670099e-05, "loss": 0.0313, "step": 10050 }, { "grad_norm": 0.19832219183444977, "learning_rate": 9.388049117244626e-05, "loss": 0.0333, "step": 10060 }, { "grad_norm": 0.15533499419689178, "learning_rate": 9.386462891337389e-05, "loss": 0.024, "step": 10070 }, { "grad_norm": 0.13324320316314697, "learning_rate": 9.384874746642257e-05, "loss": 0.029, "step": 10080 }, { "grad_norm": 0.1469622105360031, "learning_rate": 9.383284683853937e-05, "loss": 0.0262, "step": 10090 }, { "grad_norm": 0.10446592420339584, "learning_rate": 9.381692703667981e-05, "loss": 0.0268, "step": 10100 }, { "grad_norm": 0.107459656894207, "learning_rate": 9.380098806780771e-05, "loss": 0.0272, "step": 10110 }, { "grad_norm": 0.3089190125465393, "learning_rate": 9.378502993889533e-05, "loss": 0.0279, "step": 10120 }, { "grad_norm": 0.24992476403713226, "learning_rate": 9.376905265692329e-05, "loss": 0.0305, "step": 10130 }, { "grad_norm": 0.20231027901172638, "learning_rate": 9.37530562288806e-05, "loss": 0.0246, "step": 10140 }, { "grad_norm": 0.11830668896436691, "learning_rate": 9.373704066176465e-05, "loss": 0.0244, "step": 10150 }, { "grad_norm": 0.2521790862083435, "learning_rate": 9.372100596258118e-05, "loss": 0.028, "step": 10160 }, { "grad_norm": 0.1622917652130127, "learning_rate": 9.370495213834433e-05, "loss": 0.0287, "step": 10170 }, { "grad_norm": 0.14759665727615356, "learning_rate": 9.368887919607657e-05, "loss": 0.0306, "step": 10180 }, { "grad_norm": 0.28902843594551086, "learning_rate": 9.367278714280876e-05, "loss": 0.0298, "step": 10190 }, { "grad_norm": 0.16423001885414124, "learning_rate": 9.36566759855801e-05, "loss": 0.0269, "step": 10200 }, { "grad_norm": 0.17179207503795624, "learning_rate": 9.36405457314382e-05, "loss": 0.0265, "step": 10210 }, { "grad_norm": 0.14147871732711792, "learning_rate": 9.36243963874389e-05, "loss": 0.0224, "step": 10220 }, { "grad_norm": 0.22500640153884888, "learning_rate": 9.360822796064655e-05, "loss": 0.0278, "step": 10230 }, { "grad_norm": 0.12022972851991653, "learning_rate": 9.359204045813372e-05, "loss": 0.0232, "step": 10240 }, { "grad_norm": 0.18424491584300995, "learning_rate": 9.357583388698141e-05, "loss": 0.0233, "step": 10250 }, { "grad_norm": 0.0795210748910904, "learning_rate": 9.35596082542789e-05, "loss": 0.0206, "step": 10260 }, { "grad_norm": 0.19963368773460388, "learning_rate": 9.354336356712383e-05, "loss": 0.0233, "step": 10270 }, { "grad_norm": 0.27325281500816345, "learning_rate": 9.35270998326222e-05, "loss": 0.0286, "step": 10280 }, { "grad_norm": 0.1301131248474121, "learning_rate": 9.351081705788831e-05, "loss": 0.0304, "step": 10290 }, { "grad_norm": 0.15214189887046814, "learning_rate": 9.349451525004477e-05, "loss": 0.0258, "step": 10300 }, { "grad_norm": 0.22416934370994568, "learning_rate": 9.347819441622261e-05, "loss": 0.025, "step": 10310 }, { "grad_norm": 0.17889364063739777, "learning_rate": 9.346185456356105e-05, "loss": 0.0252, "step": 10320 }, { "grad_norm": 0.1656249761581421, "learning_rate": 9.344549569920774e-05, "loss": 0.0218, "step": 10330 }, { "grad_norm": 0.1737261265516281, "learning_rate": 9.342911783031858e-05, "loss": 0.0268, "step": 10340 }, { "grad_norm": 0.15242594480514526, "learning_rate": 9.341272096405782e-05, "loss": 0.0286, "step": 10350 }, { "grad_norm": 0.21342678368091583, "learning_rate": 9.3396305107598e-05, "loss": 0.0243, "step": 10360 }, { "grad_norm": 0.10044273734092712, "learning_rate": 9.337987026811998e-05, "loss": 0.0214, "step": 10370 }, { "grad_norm": 0.1325065791606903, "learning_rate": 9.33634164528129e-05, "loss": 0.0244, "step": 10380 }, { "grad_norm": 0.18671075999736786, "learning_rate": 9.334694366887424e-05, "loss": 0.0251, "step": 10390 }, { "grad_norm": 0.1273147165775299, "learning_rate": 9.333045192350973e-05, "loss": 0.0234, "step": 10400 }, { "grad_norm": 0.17724689841270447, "learning_rate": 9.331394122393345e-05, "loss": 0.0216, "step": 10410 }, { "grad_norm": 0.10811752825975418, "learning_rate": 9.329741157736771e-05, "loss": 0.0223, "step": 10420 }, { "grad_norm": 0.1716107577085495, "learning_rate": 9.328086299104317e-05, "loss": 0.021, "step": 10430 }, { "grad_norm": 0.15581032633781433, "learning_rate": 9.326429547219872e-05, "loss": 0.0202, "step": 10440 }, { "grad_norm": 0.14490415155887604, "learning_rate": 9.324770902808155e-05, "loss": 0.0282, "step": 10450 }, { "grad_norm": 0.19357691705226898, "learning_rate": 9.323110366594717e-05, "loss": 0.0237, "step": 10460 }, { "grad_norm": 0.09082087874412537, "learning_rate": 9.32144793930593e-05, "loss": 0.0239, "step": 10470 }, { "grad_norm": 0.13346919417381287, "learning_rate": 9.319783621668996e-05, "loss": 0.0218, "step": 10480 }, { "grad_norm": 0.2565051019191742, "learning_rate": 9.318117414411947e-05, "loss": 0.0223, "step": 10490 }, { "grad_norm": 0.1448710411787033, "learning_rate": 9.316449318263635e-05, "loss": 0.0212, "step": 10500 }, { "grad_norm": 0.3247460126876831, "learning_rate": 9.314779333953744e-05, "loss": 0.03, "step": 10510 }, { "grad_norm": 0.16570185124874115, "learning_rate": 9.313107462212781e-05, "loss": 0.0252, "step": 10520 }, { "grad_norm": 0.09775561839342117, "learning_rate": 9.311433703772082e-05, "loss": 0.0243, "step": 10530 }, { "grad_norm": 0.16968899965286255, "learning_rate": 9.3097580593638e-05, "loss": 0.0251, "step": 10540 }, { "grad_norm": 0.15288765728473663, "learning_rate": 9.308080529720926e-05, "loss": 0.0206, "step": 10550 }, { "grad_norm": 0.17861849069595337, "learning_rate": 9.306401115577264e-05, "loss": 0.023, "step": 10560 }, { "grad_norm": 0.1292240470647812, "learning_rate": 9.304719817667447e-05, "loss": 0.023, "step": 10570 }, { "grad_norm": 0.19977819919586182, "learning_rate": 9.303036636726934e-05, "loss": 0.0237, "step": 10580 }, { "grad_norm": 0.1568721979856491, "learning_rate": 9.301351573492003e-05, "loss": 0.0253, "step": 10590 }, { "grad_norm": 0.12127631902694702, "learning_rate": 9.299664628699758e-05, "loss": 0.031, "step": 10600 }, { "grad_norm": 0.27846670150756836, "learning_rate": 9.297975803088129e-05, "loss": 0.0363, "step": 10610 }, { "grad_norm": 0.2618520259857178, "learning_rate": 9.296285097395864e-05, "loss": 0.036, "step": 10620 }, { "grad_norm": 0.1976354420185089, "learning_rate": 9.294592512362533e-05, "loss": 0.0254, "step": 10630 }, { "grad_norm": 0.16750720143318176, "learning_rate": 9.292898048728533e-05, "loss": 0.0294, "step": 10640 }, { "grad_norm": 0.16544842720031738, "learning_rate": 9.29120170723508e-05, "loss": 0.0306, "step": 10650 }, { "grad_norm": 0.16380594670772552, "learning_rate": 9.28950348862421e-05, "loss": 0.0278, "step": 10660 }, { "grad_norm": 0.24492202699184418, "learning_rate": 9.287803393638781e-05, "loss": 0.0274, "step": 10670 }, { "grad_norm": 0.24125748872756958, "learning_rate": 9.286101423022474e-05, "loss": 0.0228, "step": 10680 }, { "grad_norm": 0.13368958234786987, "learning_rate": 9.284397577519788e-05, "loss": 0.0258, "step": 10690 }, { "grad_norm": 0.35185500979423523, "learning_rate": 9.282691857876043e-05, "loss": 0.0346, "step": 10700 }, { "grad_norm": 0.18140169978141785, "learning_rate": 9.280984264837377e-05, "loss": 0.0279, "step": 10710 }, { "grad_norm": 0.1706695258617401, "learning_rate": 9.279274799150752e-05, "loss": 0.0247, "step": 10720 }, { "grad_norm": 0.15451665222644806, "learning_rate": 9.277563461563945e-05, "loss": 0.0223, "step": 10730 }, { "grad_norm": 0.2883654534816742, "learning_rate": 9.275850252825555e-05, "loss": 0.024, "step": 10740 }, { "grad_norm": 0.12247186154127121, "learning_rate": 9.274135173684994e-05, "loss": 0.0271, "step": 10750 }, { "grad_norm": 0.18518204987049103, "learning_rate": 9.272418224892498e-05, "loss": 0.0288, "step": 10760 }, { "grad_norm": 0.4371016025543213, "learning_rate": 9.27069940719912e-05, "loss": 0.0319, "step": 10770 }, { "grad_norm": 0.1401866227388382, "learning_rate": 9.268978721356727e-05, "loss": 0.0278, "step": 10780 }, { "grad_norm": 0.14271169900894165, "learning_rate": 9.267256168118008e-05, "loss": 0.0276, "step": 10790 }, { "grad_norm": 0.2443968653678894, "learning_rate": 9.265531748236463e-05, "loss": 0.0309, "step": 10800 }, { "grad_norm": 0.12920017540454865, "learning_rate": 9.263805462466416e-05, "loss": 0.0241, "step": 10810 }, { "grad_norm": 0.20362626016139984, "learning_rate": 9.262077311562998e-05, "loss": 0.0259, "step": 10820 }, { "grad_norm": 0.14300614595413208, "learning_rate": 9.260347296282165e-05, "loss": 0.0298, "step": 10830 }, { "grad_norm": 0.14910003542900085, "learning_rate": 9.258615417380683e-05, "loss": 0.0296, "step": 10840 }, { "grad_norm": 0.2715282738208771, "learning_rate": 9.256881675616133e-05, "loss": 0.0299, "step": 10850 }, { "grad_norm": 0.1515108197927475, "learning_rate": 9.255146071746917e-05, "loss": 0.0263, "step": 10860 }, { "grad_norm": 0.3001886308193207, "learning_rate": 9.253408606532241e-05, "loss": 0.0291, "step": 10870 }, { "grad_norm": 0.20886549353599548, "learning_rate": 9.251669280732137e-05, "loss": 0.0279, "step": 10880 }, { "grad_norm": 0.17907705903053284, "learning_rate": 9.249928095107441e-05, "loss": 0.0264, "step": 10890 }, { "grad_norm": 0.3081241846084595, "learning_rate": 9.248185050419811e-05, "loss": 0.027, "step": 10900 }, { "grad_norm": 0.12968753278255463, "learning_rate": 9.24644014743171e-05, "loss": 0.0267, "step": 10910 }, { "grad_norm": 0.23722772300243378, "learning_rate": 9.24469338690642e-05, "loss": 0.0265, "step": 10920 }, { "grad_norm": 0.19146083295345306, "learning_rate": 9.242944769608033e-05, "loss": 0.0291, "step": 10930 }, { "grad_norm": 0.21786966919898987, "learning_rate": 9.241194296301454e-05, "loss": 0.0296, "step": 10940 }, { "grad_norm": 0.1289423406124115, "learning_rate": 9.239441967752397e-05, "loss": 0.0252, "step": 10950 }, { "grad_norm": 0.11162551492452621, "learning_rate": 9.237687784727393e-05, "loss": 0.0225, "step": 10960 }, { "grad_norm": 0.15841056406497955, "learning_rate": 9.235931747993781e-05, "loss": 0.0216, "step": 10970 }, { "grad_norm": 0.1155828982591629, "learning_rate": 9.234173858319707e-05, "loss": 0.0247, "step": 10980 }, { "grad_norm": 0.18403734266757965, "learning_rate": 9.23241411647414e-05, "loss": 0.027, "step": 10990 }, { "grad_norm": 0.16119953989982605, "learning_rate": 9.230652523226841e-05, "loss": 0.0225, "step": 11000 }, { "grad_norm": 0.1839390993118286, "learning_rate": 9.2288890793484e-05, "loss": 0.0246, "step": 11010 }, { "grad_norm": 0.19916290044784546, "learning_rate": 9.227123785610199e-05, "loss": 0.0275, "step": 11020 }, { "grad_norm": 0.23609048128128052, "learning_rate": 9.225356642784443e-05, "loss": 0.0274, "step": 11030 }, { "grad_norm": 0.11946256458759308, "learning_rate": 9.223587651644139e-05, "loss": 0.0234, "step": 11040 }, { "grad_norm": 0.13928553462028503, "learning_rate": 9.221816812963104e-05, "loss": 0.024, "step": 11050 }, { "grad_norm": 0.2345762699842453, "learning_rate": 9.22004412751596e-05, "loss": 0.0274, "step": 11060 }, { "grad_norm": 0.18900668621063232, "learning_rate": 9.218269596078146e-05, "loss": 0.0237, "step": 11070 }, { "grad_norm": 0.18135787546634674, "learning_rate": 9.216493219425895e-05, "loss": 0.025, "step": 11080 }, { "grad_norm": 0.2835157811641693, "learning_rate": 9.21471499833626e-05, "loss": 0.0254, "step": 11090 }, { "grad_norm": 0.16777755320072174, "learning_rate": 9.212934933587094e-05, "loss": 0.0267, "step": 11100 }, { "grad_norm": 0.20649214088916779, "learning_rate": 9.211153025957056e-05, "loss": 0.0271, "step": 11110 }, { "grad_norm": 0.124750055372715, "learning_rate": 9.209369276225614e-05, "loss": 0.0259, "step": 11120 }, { "grad_norm": 0.1328166425228119, "learning_rate": 9.207583685173042e-05, "loss": 0.0303, "step": 11130 }, { "grad_norm": 0.1476275771856308, "learning_rate": 9.205796253580417e-05, "loss": 0.0244, "step": 11140 }, { "grad_norm": 0.17589181661605835, "learning_rate": 9.204006982229621e-05, "loss": 0.0206, "step": 11150 }, { "grad_norm": 0.1737859696149826, "learning_rate": 9.202215871903346e-05, "loss": 0.023, "step": 11160 }, { "grad_norm": 0.15725189447402954, "learning_rate": 9.20042292338508e-05, "loss": 0.0258, "step": 11170 }, { "grad_norm": 0.2742244601249695, "learning_rate": 9.198628137459123e-05, "loss": 0.0283, "step": 11180 }, { "grad_norm": 0.14861378073692322, "learning_rate": 9.196831514910572e-05, "loss": 0.0302, "step": 11190 }, { "grad_norm": 0.1360122561454773, "learning_rate": 9.195033056525332e-05, "loss": 0.0275, "step": 11200 }, { "grad_norm": 0.16328176856040955, "learning_rate": 9.193232763090114e-05, "loss": 0.0244, "step": 11210 }, { "grad_norm": 0.23138655722141266, "learning_rate": 9.191430635392422e-05, "loss": 0.0256, "step": 11220 }, { "grad_norm": 0.25190597772598267, "learning_rate": 9.18962667422057e-05, "loss": 0.0324, "step": 11230 }, { "grad_norm": 0.17456254363059998, "learning_rate": 9.187820880363671e-05, "loss": 0.0269, "step": 11240 }, { "grad_norm": 0.15429823100566864, "learning_rate": 9.186013254611643e-05, "loss": 0.024, "step": 11250 }, { "grad_norm": 0.10476001352071762, "learning_rate": 9.1842037977552e-05, "loss": 0.0237, "step": 11260 }, { "grad_norm": 0.15772661566734314, "learning_rate": 9.182392510585862e-05, "loss": 0.0242, "step": 11270 }, { "grad_norm": 0.3030480146408081, "learning_rate": 9.180579393895946e-05, "loss": 0.0289, "step": 11280 }, { "grad_norm": 0.2906154692173004, "learning_rate": 9.178764448478572e-05, "loss": 0.032, "step": 11290 }, { "grad_norm": 0.1398652195930481, "learning_rate": 9.176947675127658e-05, "loss": 0.0264, "step": 11300 }, { "grad_norm": 0.21249565482139587, "learning_rate": 9.175129074637924e-05, "loss": 0.0244, "step": 11310 }, { "grad_norm": 0.13397446274757385, "learning_rate": 9.173308647804884e-05, "loss": 0.0249, "step": 11320 }, { "grad_norm": 0.09672483801841736, "learning_rate": 9.171486395424859e-05, "loss": 0.0214, "step": 11330 }, { "grad_norm": 0.13977332413196564, "learning_rate": 9.16966231829496e-05, "loss": 0.02, "step": 11340 }, { "grad_norm": 0.35313770174980164, "learning_rate": 9.167836417213105e-05, "loss": 0.0358, "step": 11350 }, { "grad_norm": 0.2588861584663391, "learning_rate": 9.166008692978001e-05, "loss": 0.025, "step": 11360 }, { "grad_norm": 0.16267822682857513, "learning_rate": 9.164179146389158e-05, "loss": 0.0227, "step": 11370 }, { "grad_norm": 0.12888717651367188, "learning_rate": 9.162347778246882e-05, "loss": 0.0262, "step": 11380 }, { "grad_norm": 0.140599325299263, "learning_rate": 9.160514589352276e-05, "loss": 0.027, "step": 11390 }, { "grad_norm": 0.09762305021286011, "learning_rate": 9.15867958050724e-05, "loss": 0.0286, "step": 11400 }, { "grad_norm": 0.3656361997127533, "learning_rate": 9.156842752514466e-05, "loss": 0.0277, "step": 11410 }, { "grad_norm": 0.3457493484020233, "learning_rate": 9.155004106177447e-05, "loss": 0.0345, "step": 11420 }, { "grad_norm": 0.2387632578611374, "learning_rate": 9.153163642300471e-05, "loss": 0.0283, "step": 11430 }, { "grad_norm": 0.16031228005886078, "learning_rate": 9.151321361688616e-05, "loss": 0.0244, "step": 11440 }, { "grad_norm": 0.19651342928409576, "learning_rate": 9.149477265147762e-05, "loss": 0.0249, "step": 11450 }, { "grad_norm": 0.22957594692707062, "learning_rate": 9.147631353484574e-05, "loss": 0.0276, "step": 11460 }, { "grad_norm": 0.2448572814464569, "learning_rate": 9.145783627506522e-05, "loss": 0.0234, "step": 11470 }, { "grad_norm": 0.183720663189888, "learning_rate": 9.143934088021861e-05, "loss": 0.0233, "step": 11480 }, { "grad_norm": 0.23072363436222076, "learning_rate": 9.142082735839645e-05, "loss": 0.0257, "step": 11490 }, { "grad_norm": 0.22348326444625854, "learning_rate": 9.140229571769715e-05, "loss": 0.0265, "step": 11500 }, { "grad_norm": 0.1618289053440094, "learning_rate": 9.138374596622709e-05, "loss": 0.0233, "step": 11510 }, { "grad_norm": 0.11589346081018448, "learning_rate": 9.136517811210059e-05, "loss": 0.022, "step": 11520 }, { "grad_norm": 0.1504592001438141, "learning_rate": 9.134659216343984e-05, "loss": 0.0228, "step": 11530 }, { "grad_norm": 0.11205630749464035, "learning_rate": 9.132798812837494e-05, "loss": 0.0233, "step": 11540 }, { "grad_norm": 0.1745501309633255, "learning_rate": 9.130936601504396e-05, "loss": 0.0276, "step": 11550 }, { "grad_norm": 0.15102963149547577, "learning_rate": 9.129072583159284e-05, "loss": 0.0242, "step": 11560 }, { "grad_norm": 0.3424801528453827, "learning_rate": 9.127206758617542e-05, "loss": 0.0267, "step": 11570 }, { "grad_norm": 0.24818848073482513, "learning_rate": 9.125339128695346e-05, "loss": 0.0261, "step": 11580 }, { "grad_norm": 0.1927766650915146, "learning_rate": 9.123469694209659e-05, "loss": 0.0285, "step": 11590 }, { "grad_norm": 0.3037368059158325, "learning_rate": 9.121598455978239e-05, "loss": 0.0235, "step": 11600 }, { "grad_norm": 0.22853721678256989, "learning_rate": 9.119725414819624e-05, "loss": 0.0279, "step": 11610 }, { "grad_norm": 0.3341428339481354, "learning_rate": 9.117850571553149e-05, "loss": 0.0297, "step": 11620 }, { "grad_norm": 0.22105348110198975, "learning_rate": 9.115973926998935e-05, "loss": 0.0333, "step": 11630 }, { "grad_norm": 0.26903393864631653, "learning_rate": 9.114095481977888e-05, "loss": 0.0261, "step": 11640 }, { "grad_norm": 0.26358288526535034, "learning_rate": 9.112215237311703e-05, "loss": 0.0247, "step": 11650 }, { "grad_norm": 0.22725167870521545, "learning_rate": 9.110333193822867e-05, "loss": 0.0259, "step": 11660 }, { "grad_norm": 0.1354473978281021, "learning_rate": 9.108449352334645e-05, "loss": 0.0218, "step": 11670 }, { "grad_norm": 0.15586887300014496, "learning_rate": 9.106563713671094e-05, "loss": 0.0246, "step": 11680 }, { "grad_norm": 0.2514376640319824, "learning_rate": 9.104676278657061e-05, "loss": 0.0252, "step": 11690 }, { "grad_norm": 0.12024156004190445, "learning_rate": 9.102787048118169e-05, "loss": 0.0227, "step": 11700 }, { "grad_norm": 0.1335715502500534, "learning_rate": 9.100896022880834e-05, "loss": 0.0246, "step": 11710 }, { "grad_norm": 0.18523810803890228, "learning_rate": 9.099003203772254e-05, "loss": 0.0257, "step": 11720 }, { "grad_norm": 0.14470262825489044, "learning_rate": 9.097108591620413e-05, "loss": 0.0209, "step": 11730 }, { "grad_norm": 0.1370673030614853, "learning_rate": 9.095212187254078e-05, "loss": 0.021, "step": 11740 }, { "grad_norm": 0.1558285802602768, "learning_rate": 9.093313991502801e-05, "loss": 0.0276, "step": 11750 }, { "grad_norm": 0.1336604356765747, "learning_rate": 9.091414005196917e-05, "loss": 0.0215, "step": 11760 }, { "grad_norm": 0.23674245178699493, "learning_rate": 9.089512229167545e-05, "loss": 0.0259, "step": 11770 }, { "grad_norm": 0.1861560195684433, "learning_rate": 9.087608664246587e-05, "loss": 0.0232, "step": 11780 }, { "grad_norm": 0.24700342118740082, "learning_rate": 9.085703311266727e-05, "loss": 0.0263, "step": 11790 }, { "grad_norm": 0.1834622025489807, "learning_rate": 9.083796171061429e-05, "loss": 0.0234, "step": 11800 }, { "grad_norm": 0.14310893416404724, "learning_rate": 9.081887244464941e-05, "loss": 0.0274, "step": 11810 }, { "grad_norm": 0.28935006260871887, "learning_rate": 9.079976532312297e-05, "loss": 0.0246, "step": 11820 }, { "grad_norm": 0.3506048619747162, "learning_rate": 9.078064035439301e-05, "loss": 0.0324, "step": 11830 }, { "grad_norm": 0.17652106285095215, "learning_rate": 9.07614975468255e-05, "loss": 0.0255, "step": 11840 }, { "grad_norm": 0.20761016011238098, "learning_rate": 9.074233690879412e-05, "loss": 0.0289, "step": 11850 }, { "grad_norm": 0.19520506262779236, "learning_rate": 9.072315844868038e-05, "loss": 0.0305, "step": 11860 }, { "grad_norm": 0.18565016984939575, "learning_rate": 9.07039621748736e-05, "loss": 0.0258, "step": 11870 }, { "grad_norm": 0.26602184772491455, "learning_rate": 9.06847480957709e-05, "loss": 0.0382, "step": 11880 }, { "grad_norm": 0.17707689106464386, "learning_rate": 9.066551621977713e-05, "loss": 0.0337, "step": 11890 }, { "grad_norm": 0.16856016218662262, "learning_rate": 9.064626655530501e-05, "loss": 0.0336, "step": 11900 }, { "grad_norm": 0.26813915371894836, "learning_rate": 9.062699911077497e-05, "loss": 0.0351, "step": 11910 }, { "grad_norm": 0.14353857934474945, "learning_rate": 9.060771389461524e-05, "loss": 0.0303, "step": 11920 }, { "grad_norm": 0.12744522094726562, "learning_rate": 9.058841091526187e-05, "loss": 0.033, "step": 11930 }, { "grad_norm": 0.1708512157201767, "learning_rate": 9.056909018115858e-05, "loss": 0.0302, "step": 11940 }, { "grad_norm": 0.16533511877059937, "learning_rate": 9.054975170075697e-05, "loss": 0.0293, "step": 11950 }, { "grad_norm": 0.14049147069454193, "learning_rate": 9.053039548251631e-05, "loss": 0.0273, "step": 11960 }, { "grad_norm": 0.3542535603046417, "learning_rate": 9.051102153490368e-05, "loss": 0.0309, "step": 11970 }, { "grad_norm": 0.2610064148902893, "learning_rate": 9.04916298663939e-05, "loss": 0.0346, "step": 11980 }, { "grad_norm": 0.17077268660068512, "learning_rate": 9.047222048546955e-05, "loss": 0.0344, "step": 11990 }, { "grad_norm": 0.18327781558036804, "learning_rate": 9.045279340062097e-05, "loss": 0.0386, "step": 12000 }, { "grad_norm": 0.21738217771053314, "learning_rate": 9.043334862034618e-05, "loss": 0.0352, "step": 12010 }, { "grad_norm": 0.15498077869415283, "learning_rate": 9.041388615315102e-05, "loss": 0.0331, "step": 12020 }, { "grad_norm": 0.18832312524318695, "learning_rate": 9.039440600754905e-05, "loss": 0.0326, "step": 12030 }, { "grad_norm": 0.48808372020721436, "learning_rate": 9.037490819206151e-05, "loss": 0.0363, "step": 12040 }, { "grad_norm": 0.2886384427547455, "learning_rate": 9.035539271521744e-05, "loss": 0.0358, "step": 12050 }, { "grad_norm": 0.18595030903816223, "learning_rate": 9.033585958555356e-05, "loss": 0.0313, "step": 12060 }, { "grad_norm": 0.18493163585662842, "learning_rate": 9.031630881161431e-05, "loss": 0.0407, "step": 12070 }, { "grad_norm": 0.4531078040599823, "learning_rate": 9.029674040195186e-05, "loss": 0.0361, "step": 12080 }, { "grad_norm": 0.24609829485416412, "learning_rate": 9.027715436512613e-05, "loss": 0.0345, "step": 12090 }, { "grad_norm": 0.26544055342674255, "learning_rate": 9.02575507097047e-05, "loss": 0.0355, "step": 12100 }, { "grad_norm": 0.21323391795158386, "learning_rate": 9.023792944426286e-05, "loss": 0.0316, "step": 12110 }, { "grad_norm": 0.25255087018013, "learning_rate": 9.021829057738364e-05, "loss": 0.0298, "step": 12120 }, { "grad_norm": 0.1319267451763153, "learning_rate": 9.019863411765775e-05, "loss": 0.0318, "step": 12130 }, { "grad_norm": 0.41611841320991516, "learning_rate": 9.017896007368357e-05, "loss": 0.0331, "step": 12140 }, { "grad_norm": 0.13953761756420135, "learning_rate": 9.015926845406722e-05, "loss": 0.029, "step": 12150 }, { "grad_norm": 0.20199425518512726, "learning_rate": 9.013955926742245e-05, "loss": 0.0283, "step": 12160 }, { "grad_norm": 0.19459670782089233, "learning_rate": 9.011983252237077e-05, "loss": 0.0335, "step": 12170 }, { "grad_norm": 0.25395467877388, "learning_rate": 9.01000882275413e-05, "loss": 0.0352, "step": 12180 }, { "grad_norm": 0.1681405007839203, "learning_rate": 9.008032639157088e-05, "loss": 0.0293, "step": 12190 }, { "grad_norm": 0.19608472287654877, "learning_rate": 9.006054702310401e-05, "loss": 0.0348, "step": 12200 }, { "grad_norm": 0.3112753927707672, "learning_rate": 9.004075013079283e-05, "loss": 0.0435, "step": 12210 }, { "grad_norm": 0.21253244578838348, "learning_rate": 9.00209357232972e-05, "loss": 0.0282, "step": 12220 }, { "grad_norm": 0.14917145669460297, "learning_rate": 9.000110380928461e-05, "loss": 0.0315, "step": 12230 }, { "grad_norm": 0.26894471049308777, "learning_rate": 8.998125439743021e-05, "loss": 0.0285, "step": 12240 }, { "grad_norm": 0.2329196333885193, "learning_rate": 8.996138749641682e-05, "loss": 0.0372, "step": 12250 }, { "grad_norm": 0.24931462109088898, "learning_rate": 8.994150311493488e-05, "loss": 0.0361, "step": 12260 }, { "grad_norm": 0.1417238861322403, "learning_rate": 8.992160126168247e-05, "loss": 0.0328, "step": 12270 }, { "grad_norm": 0.22672100365161896, "learning_rate": 8.99016819453654e-05, "loss": 0.0302, "step": 12280 }, { "grad_norm": 0.26941338181495667, "learning_rate": 8.988174517469702e-05, "loss": 0.0337, "step": 12290 }, { "grad_norm": 0.2360243946313858, "learning_rate": 8.986179095839835e-05, "loss": 0.0282, "step": 12300 }, { "grad_norm": 0.17020995914936066, "learning_rate": 8.984181930519804e-05, "loss": 0.0283, "step": 12310 }, { "grad_norm": 0.19942134618759155, "learning_rate": 8.982183022383237e-05, "loss": 0.0371, "step": 12320 }, { "grad_norm": 0.23441345989704132, "learning_rate": 8.980182372304525e-05, "loss": 0.031, "step": 12330 }, { "grad_norm": 0.13388529419898987, "learning_rate": 8.97817998115882e-05, "loss": 0.0292, "step": 12340 }, { "grad_norm": 0.21052882075309753, "learning_rate": 8.976175849822038e-05, "loss": 0.0328, "step": 12350 }, { "grad_norm": 0.17240451276302338, "learning_rate": 8.97416997917085e-05, "loss": 0.0361, "step": 12360 }, { "grad_norm": 0.1798332780599594, "learning_rate": 8.972162370082695e-05, "loss": 0.0292, "step": 12370 }, { "grad_norm": 0.19350981712341309, "learning_rate": 8.97015302343577e-05, "loss": 0.0306, "step": 12380 }, { "grad_norm": 0.24617336690425873, "learning_rate": 8.968141940109027e-05, "loss": 0.0319, "step": 12390 }, { "grad_norm": 0.22788472473621368, "learning_rate": 8.966129120982188e-05, "loss": 0.0304, "step": 12400 }, { "grad_norm": 0.25289151072502136, "learning_rate": 8.964114566935724e-05, "loss": 0.0359, "step": 12410 }, { "grad_norm": 0.14540259540081024, "learning_rate": 8.962098278850871e-05, "loss": 0.0261, "step": 12420 }, { "grad_norm": 0.26354914903640747, "learning_rate": 8.960080257609622e-05, "loss": 0.0345, "step": 12430 }, { "grad_norm": 0.23131446540355682, "learning_rate": 8.95806050409473e-05, "loss": 0.0326, "step": 12440 }, { "grad_norm": 0.22772054374217987, "learning_rate": 8.9560390191897e-05, "loss": 0.0301, "step": 12450 }, { "grad_norm": 0.1778029203414917, "learning_rate": 8.9540158037788e-05, "loss": 0.0316, "step": 12460 }, { "grad_norm": 0.14835110306739807, "learning_rate": 8.951990858747054e-05, "loss": 0.0316, "step": 12470 }, { "grad_norm": 0.12899358570575714, "learning_rate": 8.94996418498024e-05, "loss": 0.0283, "step": 12480 }, { "grad_norm": 0.20283843576908112, "learning_rate": 8.947935783364896e-05, "loss": 0.0275, "step": 12490 }, { "grad_norm": 0.2698703110218048, "learning_rate": 8.945905654788311e-05, "loss": 0.0318, "step": 12500 }, { "grad_norm": 0.23668335378170013, "learning_rate": 8.943873800138535e-05, "loss": 0.0324, "step": 12510 }, { "grad_norm": 0.19840703904628754, "learning_rate": 8.94184022030437e-05, "loss": 0.0271, "step": 12520 }, { "grad_norm": 0.12493161112070084, "learning_rate": 8.939804916175372e-05, "loss": 0.0274, "step": 12530 }, { "grad_norm": 0.20083282887935638, "learning_rate": 8.93776788864185e-05, "loss": 0.0275, "step": 12540 }, { "grad_norm": 0.34195706248283386, "learning_rate": 8.935729138594873e-05, "loss": 0.0271, "step": 12550 }, { "grad_norm": 0.20940043032169342, "learning_rate": 8.933688666926258e-05, "loss": 0.0279, "step": 12560 }, { "grad_norm": 0.13522934913635254, "learning_rate": 8.931646474528575e-05, "loss": 0.0254, "step": 12570 }, { "grad_norm": 0.23213256895542145, "learning_rate": 8.929602562295151e-05, "loss": 0.0249, "step": 12580 }, { "grad_norm": 0.11456428468227386, "learning_rate": 8.92755693112006e-05, "loss": 0.0281, "step": 12590 }, { "grad_norm": 0.1268170177936554, "learning_rate": 8.925509581898136e-05, "loss": 0.0258, "step": 12600 }, { "grad_norm": 0.10621478408575058, "learning_rate": 8.923460515524951e-05, "loss": 0.022, "step": 12610 }, { "grad_norm": 0.1789689064025879, "learning_rate": 8.921409732896842e-05, "loss": 0.0244, "step": 12620 }, { "grad_norm": 0.12427803874015808, "learning_rate": 8.919357234910887e-05, "loss": 0.0232, "step": 12630 }, { "grad_norm": 0.2554054260253906, "learning_rate": 8.917303022464923e-05, "loss": 0.0312, "step": 12640 }, { "grad_norm": 0.10979140549898148, "learning_rate": 8.915247096457531e-05, "loss": 0.0235, "step": 12650 }, { "grad_norm": 0.22511038184165955, "learning_rate": 8.91318945778804e-05, "loss": 0.029, "step": 12660 }, { "grad_norm": 0.2394827902317047, "learning_rate": 8.911130107356534e-05, "loss": 0.0276, "step": 12670 }, { "grad_norm": 0.17942224442958832, "learning_rate": 8.90906904606384e-05, "loss": 0.0262, "step": 12680 }, { "grad_norm": 0.18728674948215485, "learning_rate": 8.90700627481154e-05, "loss": 0.0261, "step": 12690 }, { "grad_norm": 0.17803725600242615, "learning_rate": 8.904941794501957e-05, "loss": 0.0261, "step": 12700 }, { "grad_norm": 0.15869572758674622, "learning_rate": 8.902875606038166e-05, "loss": 0.0233, "step": 12710 }, { "grad_norm": 0.2261994332075119, "learning_rate": 8.900807710323989e-05, "loss": 0.0228, "step": 12720 }, { "grad_norm": 0.29880020022392273, "learning_rate": 8.898738108263993e-05, "loss": 0.0265, "step": 12730 }, { "grad_norm": 0.13413943350315094, "learning_rate": 8.896666800763491e-05, "loss": 0.0245, "step": 12740 }, { "grad_norm": 0.13502272963523865, "learning_rate": 8.894593788728546e-05, "loss": 0.0236, "step": 12750 }, { "grad_norm": 0.30857524275779724, "learning_rate": 8.892519073065961e-05, "loss": 0.0296, "step": 12760 }, { "grad_norm": 0.2542788088321686, "learning_rate": 8.89044265468329e-05, "loss": 0.031, "step": 12770 }, { "grad_norm": 0.294565349817276, "learning_rate": 8.888364534488827e-05, "loss": 0.0273, "step": 12780 }, { "grad_norm": 0.12762665748596191, "learning_rate": 8.886284713391613e-05, "loss": 0.0277, "step": 12790 }, { "grad_norm": 0.17158719897270203, "learning_rate": 8.884203192301431e-05, "loss": 0.0262, "step": 12800 }, { "grad_norm": 0.1293683797121048, "learning_rate": 8.88211997212881e-05, "loss": 0.0287, "step": 12810 }, { "grad_norm": 0.10607478767633438, "learning_rate": 8.880035053785023e-05, "loss": 0.0265, "step": 12820 }, { "grad_norm": 0.3105185329914093, "learning_rate": 8.877948438182083e-05, "loss": 0.0286, "step": 12830 }, { "grad_norm": 0.20478717982769012, "learning_rate": 8.875860126232745e-05, "loss": 0.0257, "step": 12840 }, { "grad_norm": 0.1499645709991455, "learning_rate": 8.87377011885051e-05, "loss": 0.0249, "step": 12850 }, { "grad_norm": 0.493356853723526, "learning_rate": 8.871678416949617e-05, "loss": 0.0306, "step": 12860 }, { "grad_norm": 0.2577926516532898, "learning_rate": 8.869585021445046e-05, "loss": 0.0251, "step": 12870 }, { "grad_norm": 0.3269892930984497, "learning_rate": 8.867489933252521e-05, "loss": 0.0245, "step": 12880 }, { "grad_norm": 0.16286231577396393, "learning_rate": 8.865393153288504e-05, "loss": 0.026, "step": 12890 }, { "grad_norm": 0.19377857446670532, "learning_rate": 8.8632946824702e-05, "loss": 0.0289, "step": 12900 }, { "grad_norm": 0.2406398504972458, "learning_rate": 8.86119452171555e-05, "loss": 0.0222, "step": 12910 }, { "grad_norm": 0.1499948650598526, "learning_rate": 8.859092671943234e-05, "loss": 0.0177, "step": 12920 }, { "grad_norm": 0.15853965282440186, "learning_rate": 8.856989134072676e-05, "loss": 0.0212, "step": 12930 }, { "grad_norm": 0.14564277231693268, "learning_rate": 8.85488390902403e-05, "loss": 0.0237, "step": 12940 }, { "grad_norm": 0.12619641423225403, "learning_rate": 8.852776997718199e-05, "loss": 0.0242, "step": 12950 }, { "grad_norm": 0.12413664907217026, "learning_rate": 8.850668401076812e-05, "loss": 0.0224, "step": 12960 }, { "grad_norm": 0.1705208271741867, "learning_rate": 8.848558120022246e-05, "loss": 0.0219, "step": 12970 }, { "grad_norm": 0.2111823409795761, "learning_rate": 8.846446155477603e-05, "loss": 0.0223, "step": 12980 }, { "grad_norm": 0.10970038920640945, "learning_rate": 8.844332508366735e-05, "loss": 0.0261, "step": 12990 }, { "grad_norm": 0.1461486667394638, "learning_rate": 8.84221717961422e-05, "loss": 0.026, "step": 13000 }, { "grad_norm": 0.18444067239761353, "learning_rate": 8.840100170145374e-05, "loss": 0.031, "step": 13010 }, { "grad_norm": 0.24409931898117065, "learning_rate": 8.837981480886249e-05, "loss": 0.0282, "step": 13020 }, { "grad_norm": 0.20660032331943512, "learning_rate": 8.835861112763633e-05, "loss": 0.0296, "step": 13030 }, { "grad_norm": 0.10484491288661957, "learning_rate": 8.833739066705044e-05, "loss": 0.0223, "step": 13040 }, { "grad_norm": 0.15518608689308167, "learning_rate": 8.831615343638742e-05, "loss": 0.0199, "step": 13050 }, { "grad_norm": 0.08768881857395172, "learning_rate": 8.829489944493711e-05, "loss": 0.0232, "step": 13060 }, { "grad_norm": 0.21857772767543793, "learning_rate": 8.827362870199675e-05, "loss": 0.0292, "step": 13070 }, { "grad_norm": 0.2238813191652298, "learning_rate": 8.825234121687089e-05, "loss": 0.0266, "step": 13080 }, { "grad_norm": 0.2027789056301117, "learning_rate": 8.823103699887139e-05, "loss": 0.0261, "step": 13090 }, { "grad_norm": 0.14643041789531708, "learning_rate": 8.820971605731745e-05, "loss": 0.0282, "step": 13100 }, { "grad_norm": 0.18460208177566528, "learning_rate": 8.818837840153556e-05, "loss": 0.0244, "step": 13110 }, { "grad_norm": 0.09474723786115646, "learning_rate": 8.816702404085952e-05, "loss": 0.0222, "step": 13120 }, { "grad_norm": 0.22745859622955322, "learning_rate": 8.814565298463048e-05, "loss": 0.0298, "step": 13130 }, { "grad_norm": 0.2074170708656311, "learning_rate": 8.812426524219688e-05, "loss": 0.0266, "step": 13140 }, { "grad_norm": 0.2885584831237793, "learning_rate": 8.81028608229144e-05, "loss": 0.026, "step": 13150 }, { "grad_norm": 0.1990453153848648, "learning_rate": 8.808143973614611e-05, "loss": 0.0218, "step": 13160 }, { "grad_norm": 0.20547853410243988, "learning_rate": 8.806000199126228e-05, "loss": 0.0226, "step": 13170 }, { "grad_norm": 0.11701852828264236, "learning_rate": 8.803854759764052e-05, "loss": 0.0249, "step": 13180 }, { "grad_norm": 0.10917941480875015, "learning_rate": 8.801707656466572e-05, "loss": 0.0222, "step": 13190 }, { "grad_norm": 0.2012253701686859, "learning_rate": 8.799558890173003e-05, "loss": 0.0237, "step": 13200 }, { "grad_norm": 0.11587274074554443, "learning_rate": 8.79740846182329e-05, "loss": 0.0193, "step": 13210 }, { "grad_norm": 0.094405896961689, "learning_rate": 8.7952563723581e-05, "loss": 0.0211, "step": 13220 }, { "grad_norm": 0.21044966578483582, "learning_rate": 8.793102622718834e-05, "loss": 0.0234, "step": 13230 }, { "grad_norm": 0.2559979557991028, "learning_rate": 8.790947213847613e-05, "loss": 0.0275, "step": 13240 }, { "grad_norm": 0.24414798617362976, "learning_rate": 8.788790146687286e-05, "loss": 0.027, "step": 13250 }, { "grad_norm": 0.14309091866016388, "learning_rate": 8.786631422181429e-05, "loss": 0.0268, "step": 13260 }, { "grad_norm": 0.15944774448871613, "learning_rate": 8.78447104127434e-05, "loss": 0.0259, "step": 13270 }, { "grad_norm": 0.14574779570102692, "learning_rate": 8.782309004911042e-05, "loss": 0.0257, "step": 13280 }, { "grad_norm": 0.1657639592885971, "learning_rate": 8.780145314037286e-05, "loss": 0.0253, "step": 13290 }, { "grad_norm": 0.18873851001262665, "learning_rate": 8.777979969599542e-05, "loss": 0.0216, "step": 13300 }, { "grad_norm": 0.16527268290519714, "learning_rate": 8.775812972545006e-05, "loss": 0.0257, "step": 13310 }, { "grad_norm": 0.20751304924488068, "learning_rate": 8.773644323821596e-05, "loss": 0.0278, "step": 13320 }, { "grad_norm": 0.23123952746391296, "learning_rate": 8.771474024377953e-05, "loss": 0.026, "step": 13330 }, { "grad_norm": 0.22234408557415009, "learning_rate": 8.769302075163438e-05, "loss": 0.025, "step": 13340 }, { "grad_norm": 0.18309885263442993, "learning_rate": 8.767128477128137e-05, "loss": 0.0233, "step": 13350 }, { "grad_norm": 0.13593953847885132, "learning_rate": 8.764953231222854e-05, "loss": 0.0227, "step": 13360 }, { "grad_norm": 0.19761766493320465, "learning_rate": 8.762776338399119e-05, "loss": 0.0273, "step": 13370 }, { "grad_norm": 0.22679683566093445, "learning_rate": 8.760597799609176e-05, "loss": 0.0247, "step": 13380 }, { "grad_norm": 0.13534146547317505, "learning_rate": 8.758417615805992e-05, "loss": 0.0248, "step": 13390 }, { "grad_norm": 0.18132942914962769, "learning_rate": 8.756235787943254e-05, "loss": 0.0279, "step": 13400 }, { "grad_norm": 0.1691446453332901, "learning_rate": 8.754052316975367e-05, "loss": 0.0241, "step": 13410 }, { "grad_norm": 0.08949816972017288, "learning_rate": 8.751867203857455e-05, "loss": 0.0226, "step": 13420 }, { "grad_norm": 0.13856717944145203, "learning_rate": 8.749680449545363e-05, "loss": 0.0239, "step": 13430 }, { "grad_norm": 0.15877442061901093, "learning_rate": 8.747492054995649e-05, "loss": 0.0243, "step": 13440 }, { "grad_norm": 0.2790378928184509, "learning_rate": 8.745302021165595e-05, "loss": 0.0277, "step": 13450 }, { "grad_norm": 0.12271425873041153, "learning_rate": 8.743110349013192e-05, "loss": 0.0273, "step": 13460 }, { "grad_norm": 0.13791388273239136, "learning_rate": 8.740917039497153e-05, "loss": 0.0323, "step": 13470 }, { "grad_norm": 0.17134079337120056, "learning_rate": 8.738722093576906e-05, "loss": 0.0285, "step": 13480 }, { "grad_norm": 0.23411187529563904, "learning_rate": 8.736525512212597e-05, "loss": 0.0284, "step": 13490 }, { "grad_norm": 0.1594974845647812, "learning_rate": 8.734327296365084e-05, "loss": 0.0228, "step": 13500 }, { "grad_norm": 0.27501997351646423, "learning_rate": 8.732127446995939e-05, "loss": 0.0291, "step": 13510 }, { "grad_norm": 0.2828959822654724, "learning_rate": 8.729925965067454e-05, "loss": 0.0285, "step": 13520 }, { "grad_norm": 0.17789416015148163, "learning_rate": 8.72772285154263e-05, "loss": 0.0297, "step": 13530 }, { "grad_norm": 0.14135226607322693, "learning_rate": 8.725518107385187e-05, "loss": 0.0263, "step": 13540 }, { "grad_norm": 0.22237519919872284, "learning_rate": 8.72331173355955e-05, "loss": 0.021, "step": 13550 }, { "grad_norm": 0.2023538500070572, "learning_rate": 8.721103731030867e-05, "loss": 0.0277, "step": 13560 }, { "grad_norm": 0.18108880519866943, "learning_rate": 8.718894100764989e-05, "loss": 0.0245, "step": 13570 }, { "grad_norm": 0.16642694175243378, "learning_rate": 8.716682843728485e-05, "loss": 0.0227, "step": 13580 }, { "grad_norm": 0.11417503654956818, "learning_rate": 8.714469960888634e-05, "loss": 0.0278, "step": 13590 }, { "grad_norm": 0.17281696200370789, "learning_rate": 8.712255453213427e-05, "loss": 0.0298, "step": 13600 }, { "grad_norm": 0.12351976335048676, "learning_rate": 8.710039321671563e-05, "loss": 0.0212, "step": 13610 }, { "grad_norm": 0.1986982673406601, "learning_rate": 8.707821567232456e-05, "loss": 0.028, "step": 13620 }, { "grad_norm": 0.2808137536048889, "learning_rate": 8.705602190866225e-05, "loss": 0.0224, "step": 13630 }, { "grad_norm": 0.22478850185871124, "learning_rate": 8.703381193543701e-05, "loss": 0.0215, "step": 13640 }, { "grad_norm": 0.3245067596435547, "learning_rate": 8.701158576236423e-05, "loss": 0.0282, "step": 13650 }, { "grad_norm": 0.31963613629341125, "learning_rate": 8.69893433991664e-05, "loss": 0.0247, "step": 13660 }, { "grad_norm": 0.14251863956451416, "learning_rate": 8.69670848555731e-05, "loss": 0.024, "step": 13670 }, { "grad_norm": 0.11417273432016373, "learning_rate": 8.694481014132096e-05, "loss": 0.0237, "step": 13680 }, { "grad_norm": 0.291910320520401, "learning_rate": 8.69225192661537e-05, "loss": 0.0276, "step": 13690 }, { "grad_norm": 0.1578601598739624, "learning_rate": 8.690021223982208e-05, "loss": 0.0229, "step": 13700 }, { "grad_norm": 0.14454133808612823, "learning_rate": 8.687788907208398e-05, "loss": 0.0214, "step": 13710 }, { "grad_norm": 0.142221137881279, "learning_rate": 8.685554977270431e-05, "loss": 0.0203, "step": 13720 }, { "grad_norm": 0.21294209361076355, "learning_rate": 8.683319435145503e-05, "loss": 0.023, "step": 13730 }, { "grad_norm": 0.15506942570209503, "learning_rate": 8.681082281811517e-05, "loss": 0.0203, "step": 13740 }, { "grad_norm": 0.07853997498750687, "learning_rate": 8.67884351824708e-05, "loss": 0.0224, "step": 13750 }, { "grad_norm": 0.1950102150440216, "learning_rate": 8.676603145431501e-05, "loss": 0.036, "step": 13760 }, { "grad_norm": 0.23546645045280457, "learning_rate": 8.674361164344799e-05, "loss": 0.0276, "step": 13770 }, { "grad_norm": 0.2064974009990692, "learning_rate": 8.672117575967688e-05, "loss": 0.0231, "step": 13780 }, { "grad_norm": 0.1726672649383545, "learning_rate": 8.669872381281595e-05, "loss": 0.0247, "step": 13790 }, { "grad_norm": 0.1792387068271637, "learning_rate": 8.667625581268639e-05, "loss": 0.0222, "step": 13800 }, { "grad_norm": 0.11981657892465591, "learning_rate": 8.665377176911651e-05, "loss": 0.0291, "step": 13810 }, { "grad_norm": 0.1995706409215927, "learning_rate": 8.663127169194159e-05, "loss": 0.0294, "step": 13820 }, { "grad_norm": 0.20342764258384705, "learning_rate": 8.660875559100389e-05, "loss": 0.0303, "step": 13830 }, { "grad_norm": 0.15755179524421692, "learning_rate": 8.658622347615274e-05, "loss": 0.0287, "step": 13840 }, { "grad_norm": 0.1616755723953247, "learning_rate": 8.656367535724448e-05, "loss": 0.0211, "step": 13850 }, { "grad_norm": 0.1624089628458023, "learning_rate": 8.65411112441424e-05, "loss": 0.0262, "step": 13860 }, { "grad_norm": 0.12262984365224838, "learning_rate": 8.651853114671679e-05, "loss": 0.0266, "step": 13870 }, { "grad_norm": 0.1511944979429245, "learning_rate": 8.649593507484499e-05, "loss": 0.0252, "step": 13880 }, { "grad_norm": 0.27257564663887024, "learning_rate": 8.647332303841126e-05, "loss": 0.0281, "step": 13890 }, { "grad_norm": 0.128867506980896, "learning_rate": 8.645069504730689e-05, "loss": 0.0234, "step": 13900 }, { "grad_norm": 0.2236945480108261, "learning_rate": 8.64280511114301e-05, "loss": 0.0268, "step": 13910 }, { "grad_norm": 0.19506539404392242, "learning_rate": 8.640539124068617e-05, "loss": 0.0284, "step": 13920 }, { "grad_norm": 0.17988009750843048, "learning_rate": 8.638271544498727e-05, "loss": 0.028, "step": 13930 }, { "grad_norm": 0.16376721858978271, "learning_rate": 8.636002373425257e-05, "loss": 0.0289, "step": 13940 }, { "grad_norm": 0.21842984855175018, "learning_rate": 8.633731611840817e-05, "loss": 0.0301, "step": 13950 }, { "grad_norm": 0.23731964826583862, "learning_rate": 8.631459260738717e-05, "loss": 0.0292, "step": 13960 }, { "grad_norm": 0.2533927857875824, "learning_rate": 8.62918532111296e-05, "loss": 0.041, "step": 13970 }, { "grad_norm": 0.12117312848567963, "learning_rate": 8.626909793958248e-05, "loss": 0.0362, "step": 13980 }, { "grad_norm": 0.2476797252893448, "learning_rate": 8.624632680269969e-05, "loss": 0.0349, "step": 13990 }, { "grad_norm": 0.15771779417991638, "learning_rate": 8.622353981044212e-05, "loss": 0.0323, "step": 14000 }, { "grad_norm": 0.19183768332004547, "learning_rate": 8.620073697277757e-05, "loss": 0.0275, "step": 14010 }, { "grad_norm": 0.2084181010723114, "learning_rate": 8.617791829968079e-05, "loss": 0.0329, "step": 14020 }, { "grad_norm": 0.23016518354415894, "learning_rate": 8.615508380113344e-05, "loss": 0.0301, "step": 14030 }, { "grad_norm": 0.2970336377620697, "learning_rate": 8.613223348712408e-05, "loss": 0.025, "step": 14040 }, { "grad_norm": 0.2614697515964508, "learning_rate": 8.610936736764824e-05, "loss": 0.0256, "step": 14050 }, { "grad_norm": 0.08489828556776047, "learning_rate": 8.608648545270833e-05, "loss": 0.0252, "step": 14060 }, { "grad_norm": 0.25782886147499084, "learning_rate": 8.606358775231366e-05, "loss": 0.0269, "step": 14070 }, { "grad_norm": 0.1358519047498703, "learning_rate": 8.60406742764805e-05, "loss": 0.026, "step": 14080 }, { "grad_norm": 0.24629026651382446, "learning_rate": 8.601774503523195e-05, "loss": 0.0278, "step": 14090 }, { "grad_norm": 0.10684795677661896, "learning_rate": 8.599480003859805e-05, "loss": 0.0278, "step": 14100 }, { "grad_norm": 0.11192042380571365, "learning_rate": 8.597183929661573e-05, "loss": 0.0265, "step": 14110 }, { "grad_norm": 0.0918087437748909, "learning_rate": 8.594886281932879e-05, "loss": 0.0242, "step": 14120 }, { "grad_norm": 0.29323622584342957, "learning_rate": 8.59258706167879e-05, "loss": 0.0343, "step": 14130 }, { "grad_norm": 0.21045097708702087, "learning_rate": 8.590286269905068e-05, "loss": 0.029, "step": 14140 }, { "grad_norm": 0.18044006824493408, "learning_rate": 8.587983907618154e-05, "loss": 0.0271, "step": 14150 }, { "grad_norm": 0.3148023188114166, "learning_rate": 8.585679975825178e-05, "loss": 0.0318, "step": 14160 }, { "grad_norm": 0.27816861867904663, "learning_rate": 8.583374475533962e-05, "loss": 0.0276, "step": 14170 }, { "grad_norm": 0.14777015149593353, "learning_rate": 8.581067407753009e-05, "loss": 0.0262, "step": 14180 }, { "grad_norm": 0.2677735686302185, "learning_rate": 8.578758773491507e-05, "loss": 0.0283, "step": 14190 }, { "grad_norm": 0.18756107985973358, "learning_rate": 8.576448573759332e-05, "loss": 0.0266, "step": 14200 }, { "grad_norm": 0.16123470664024353, "learning_rate": 8.574136809567044e-05, "loss": 0.0248, "step": 14210 }, { "grad_norm": 0.11205974221229553, "learning_rate": 8.57182348192589e-05, "loss": 0.0263, "step": 14220 }, { "grad_norm": 0.1082172766327858, "learning_rate": 8.569508591847792e-05, "loss": 0.0255, "step": 14230 }, { "grad_norm": 0.17424112558364868, "learning_rate": 8.567192140345367e-05, "loss": 0.0239, "step": 14240 }, { "grad_norm": 0.2005939930677414, "learning_rate": 8.564874128431906e-05, "loss": 0.0273, "step": 14250 }, { "grad_norm": 0.1417471170425415, "learning_rate": 8.562554557121389e-05, "loss": 0.0223, "step": 14260 }, { "grad_norm": 0.23350897431373596, "learning_rate": 8.560233427428475e-05, "loss": 0.0277, "step": 14270 }, { "grad_norm": 0.17361761629581451, "learning_rate": 8.557910740368503e-05, "loss": 0.0255, "step": 14280 }, { "grad_norm": 0.15574586391448975, "learning_rate": 8.555586496957495e-05, "loss": 0.0201, "step": 14290 }, { "grad_norm": 0.15925124287605286, "learning_rate": 8.553260698212155e-05, "loss": 0.0248, "step": 14300 }, { "grad_norm": 0.1716056913137436, "learning_rate": 8.550933345149868e-05, "loss": 0.0223, "step": 14310 }, { "grad_norm": 0.10551566630601883, "learning_rate": 8.548604438788696e-05, "loss": 0.0364, "step": 14320 }, { "grad_norm": 0.28443384170532227, "learning_rate": 8.546273980147383e-05, "loss": 0.0243, "step": 14330 }, { "grad_norm": 0.12783917784690857, "learning_rate": 8.543941970245348e-05, "loss": 0.0259, "step": 14340 }, { "grad_norm": 0.3350085914134979, "learning_rate": 8.541608410102693e-05, "loss": 0.0256, "step": 14350 }, { "grad_norm": 0.19016240537166595, "learning_rate": 8.539273300740195e-05, "loss": 0.0213, "step": 14360 }, { "grad_norm": 0.12859532237052917, "learning_rate": 8.536936643179313e-05, "loss": 0.0247, "step": 14370 }, { "grad_norm": 0.10515488684177399, "learning_rate": 8.534598438442179e-05, "loss": 0.0228, "step": 14380 }, { "grad_norm": 0.333721786737442, "learning_rate": 8.532258687551603e-05, "loss": 0.0307, "step": 14390 }, { "grad_norm": 0.16950500011444092, "learning_rate": 8.529917391531071e-05, "loss": 0.0231, "step": 14400 }, { "grad_norm": 0.2004522979259491, "learning_rate": 8.527574551404747e-05, "loss": 0.023, "step": 14410 }, { "grad_norm": 0.1349787712097168, "learning_rate": 8.525230168197468e-05, "loss": 0.0241, "step": 14420 }, { "grad_norm": 0.12191593647003174, "learning_rate": 8.522884242934745e-05, "loss": 0.0208, "step": 14430 }, { "grad_norm": 0.1600448191165924, "learning_rate": 8.520536776642768e-05, "loss": 0.0189, "step": 14440 }, { "grad_norm": 0.13497143983840942, "learning_rate": 8.5181877703484e-05, "loss": 0.0238, "step": 14450 }, { "grad_norm": 0.1167539581656456, "learning_rate": 8.51583722507917e-05, "loss": 0.0222, "step": 14460 }, { "grad_norm": 0.14289745688438416, "learning_rate": 8.513485141863293e-05, "loss": 0.0221, "step": 14470 }, { "grad_norm": 0.14250092208385468, "learning_rate": 8.511131521729647e-05, "loss": 0.02, "step": 14480 }, { "grad_norm": 0.182058185338974, "learning_rate": 8.508776365707787e-05, "loss": 0.0217, "step": 14490 }, { "grad_norm": 0.19491468369960785, "learning_rate": 8.506419674827934e-05, "loss": 0.027, "step": 14500 }, { "grad_norm": 0.15335336327552795, "learning_rate": 8.50406145012099e-05, "loss": 0.0234, "step": 14510 }, { "grad_norm": 0.15077511966228485, "learning_rate": 8.501701692618519e-05, "loss": 0.0204, "step": 14520 }, { "grad_norm": 0.18675583600997925, "learning_rate": 8.499340403352761e-05, "loss": 0.027, "step": 14530 }, { "grad_norm": 0.17432399094104767, "learning_rate": 8.496977583356623e-05, "loss": 0.0216, "step": 14540 }, { "grad_norm": 0.29915153980255127, "learning_rate": 8.494613233663684e-05, "loss": 0.021, "step": 14550 }, { "grad_norm": 0.1310018002986908, "learning_rate": 8.492247355308189e-05, "loss": 0.0198, "step": 14560 }, { "grad_norm": 0.1165018230676651, "learning_rate": 8.489879949325056e-05, "loss": 0.0191, "step": 14570 }, { "grad_norm": 0.2030206173658371, "learning_rate": 8.487511016749868e-05, "loss": 0.0213, "step": 14580 }, { "grad_norm": 0.19784769415855408, "learning_rate": 8.485140558618874e-05, "loss": 0.0245, "step": 14590 }, { "grad_norm": 0.2011437565088272, "learning_rate": 8.482768575968995e-05, "loss": 0.024, "step": 14600 }, { "grad_norm": 0.14564409852027893, "learning_rate": 8.480395069837818e-05, "loss": 0.02, "step": 14610 }, { "grad_norm": 0.2588787376880646, "learning_rate": 8.478020041263595e-05, "loss": 0.0207, "step": 14620 }, { "grad_norm": 0.1602122187614441, "learning_rate": 8.475643491285242e-05, "loss": 0.0223, "step": 14630 }, { "grad_norm": 0.14000558853149414, "learning_rate": 8.473265420942345e-05, "loss": 0.0205, "step": 14640 }, { "grad_norm": 0.25795283913612366, "learning_rate": 8.470885831275151e-05, "loss": 0.0237, "step": 14650 }, { "grad_norm": 0.10268189013004303, "learning_rate": 8.468504723324574e-05, "loss": 0.0229, "step": 14660 }, { "grad_norm": 0.10830003768205643, "learning_rate": 8.466122098132193e-05, "loss": 0.023, "step": 14670 }, { "grad_norm": 0.13836075365543365, "learning_rate": 8.463737956740245e-05, "loss": 0.0206, "step": 14680 }, { "grad_norm": 0.22415222227573395, "learning_rate": 8.461352300191639e-05, "loss": 0.0213, "step": 14690 }, { "grad_norm": 0.168512761592865, "learning_rate": 8.45896512952994e-05, "loss": 0.0224, "step": 14700 }, { "grad_norm": 0.25563710927963257, "learning_rate": 8.456576445799377e-05, "loss": 0.024, "step": 14710 }, { "grad_norm": 0.14428181946277618, "learning_rate": 8.454186250044844e-05, "loss": 0.0229, "step": 14720 }, { "grad_norm": 0.13446752727031708, "learning_rate": 8.451794543311892e-05, "loss": 0.0242, "step": 14730 }, { "grad_norm": 0.1406097561120987, "learning_rate": 8.449401326646736e-05, "loss": 0.0272, "step": 14740 }, { "grad_norm": 0.20316682755947113, "learning_rate": 8.447006601096248e-05, "loss": 0.0224, "step": 14750 }, { "grad_norm": 0.14668036997318268, "learning_rate": 8.444610367707964e-05, "loss": 0.0237, "step": 14760 }, { "grad_norm": 0.20204393565654755, "learning_rate": 8.442212627530078e-05, "loss": 0.022, "step": 14770 }, { "grad_norm": 0.2514674663543701, "learning_rate": 8.439813381611441e-05, "loss": 0.0225, "step": 14780 }, { "grad_norm": 0.20360106229782104, "learning_rate": 8.437412631001567e-05, "loss": 0.0212, "step": 14790 }, { "grad_norm": 0.25110694766044617, "learning_rate": 8.435010376750626e-05, "loss": 0.0251, "step": 14800 }, { "grad_norm": 0.16208726167678833, "learning_rate": 8.432606619909442e-05, "loss": 0.0229, "step": 14810 }, { "grad_norm": 0.20707300305366516, "learning_rate": 8.430201361529506e-05, "loss": 0.0259, "step": 14820 }, { "grad_norm": 0.12442222982645035, "learning_rate": 8.427794602662954e-05, "loss": 0.0261, "step": 14830 }, { "grad_norm": 0.3182677924633026, "learning_rate": 8.425386344362586e-05, "loss": 0.0292, "step": 14840 }, { "grad_norm": 0.16128374636173248, "learning_rate": 8.422976587681859e-05, "loss": 0.0339, "step": 14850 }, { "grad_norm": 0.19701898097991943, "learning_rate": 8.42056533367488e-05, "loss": 0.0219, "step": 14860 }, { "grad_norm": 0.15341833233833313, "learning_rate": 8.41815258339641e-05, "loss": 0.023, "step": 14870 }, { "grad_norm": 0.24340444803237915, "learning_rate": 8.415738337901874e-05, "loss": 0.0253, "step": 14880 }, { "grad_norm": 0.2128688097000122, "learning_rate": 8.413322598247342e-05, "loss": 0.0225, "step": 14890 }, { "grad_norm": 0.27050313353538513, "learning_rate": 8.41090536548954e-05, "loss": 0.0242, "step": 14900 }, { "grad_norm": 0.23896484076976776, "learning_rate": 8.408486640685849e-05, "loss": 0.0298, "step": 14910 }, { "grad_norm": 0.18118129670619965, "learning_rate": 8.4060664248943e-05, "loss": 0.023, "step": 14920 }, { "grad_norm": 0.23430445790290833, "learning_rate": 8.40364471917358e-05, "loss": 0.0229, "step": 14930 }, { "grad_norm": 0.14800837635993958, "learning_rate": 8.401221524583024e-05, "loss": 0.0275, "step": 14940 }, { "grad_norm": 0.3556208312511444, "learning_rate": 8.398796842182619e-05, "loss": 0.0303, "step": 14950 }, { "grad_norm": 0.2210242599248886, "learning_rate": 8.396370673033006e-05, "loss": 0.0261, "step": 14960 }, { "grad_norm": 0.18728074431419373, "learning_rate": 8.39394301819547e-05, "loss": 0.0261, "step": 14970 }, { "grad_norm": 0.17648845911026, "learning_rate": 8.391513878731949e-05, "loss": 0.026, "step": 14980 }, { "grad_norm": 0.1396275758743286, "learning_rate": 8.389083255705037e-05, "loss": 0.0274, "step": 14990 }, { "grad_norm": 0.33052030205726624, "learning_rate": 8.386651150177968e-05, "loss": 0.0298, "step": 15000 }, { "grad_norm": 0.14298813045024872, "learning_rate": 8.384217563214627e-05, "loss": 0.0234, "step": 15010 }, { "grad_norm": 0.18515165150165558, "learning_rate": 8.381782495879546e-05, "loss": 0.0248, "step": 15020 }, { "grad_norm": 0.255903035402298, "learning_rate": 8.37934594923791e-05, "loss": 0.0238, "step": 15030 }, { "grad_norm": 0.22934821248054504, "learning_rate": 8.376907924355546e-05, "loss": 0.03, "step": 15040 }, { "grad_norm": 0.2082718014717102, "learning_rate": 8.374468422298926e-05, "loss": 0.0346, "step": 15050 }, { "grad_norm": 0.21001972258090973, "learning_rate": 8.372027444135176e-05, "loss": 0.0327, "step": 15060 }, { "grad_norm": 0.13658297061920166, "learning_rate": 8.36958499093206e-05, "loss": 0.0269, "step": 15070 }, { "grad_norm": 0.1951613426208496, "learning_rate": 8.367141063757988e-05, "loss": 0.026, "step": 15080 }, { "grad_norm": 0.1377798467874527, "learning_rate": 8.364695663682022e-05, "loss": 0.0291, "step": 15090 }, { "grad_norm": 0.14343157410621643, "learning_rate": 8.362248791773858e-05, "loss": 0.0259, "step": 15100 }, { "grad_norm": 0.09735367447137833, "learning_rate": 8.359800449103842e-05, "loss": 0.0251, "step": 15110 }, { "grad_norm": 0.10362179577350616, "learning_rate": 8.357350636742967e-05, "loss": 0.0239, "step": 15120 }, { "grad_norm": 0.2162770926952362, "learning_rate": 8.354899355762858e-05, "loss": 0.0238, "step": 15130 }, { "grad_norm": 0.14809058606624603, "learning_rate": 8.352446607235791e-05, "loss": 0.0204, "step": 15140 }, { "grad_norm": 0.17303027212619781, "learning_rate": 8.349992392234683e-05, "loss": 0.0271, "step": 15150 }, { "grad_norm": 0.14254681766033173, "learning_rate": 8.347536711833088e-05, "loss": 0.0239, "step": 15160 }, { "grad_norm": 0.13662225008010864, "learning_rate": 8.345079567105206e-05, "loss": 0.0286, "step": 15170 }, { "grad_norm": 0.1813633143901825, "learning_rate": 8.342620959125875e-05, "loss": 0.0239, "step": 15180 }, { "grad_norm": 0.15609247982501984, "learning_rate": 8.340160888970574e-05, "loss": 0.0247, "step": 15190 }, { "grad_norm": 0.16026583313941956, "learning_rate": 8.337699357715422e-05, "loss": 0.0288, "step": 15200 }, { "grad_norm": 0.14674435555934906, "learning_rate": 8.335236366437174e-05, "loss": 0.0259, "step": 15210 }, { "grad_norm": 0.2604127824306488, "learning_rate": 8.332771916213227e-05, "loss": 0.0241, "step": 15220 }, { "grad_norm": 0.12884654104709625, "learning_rate": 8.330306008121618e-05, "loss": 0.0224, "step": 15230 }, { "grad_norm": 0.14738677442073822, "learning_rate": 8.327838643241014e-05, "loss": 0.0229, "step": 15240 }, { "grad_norm": 0.23179928958415985, "learning_rate": 8.32536982265073e-05, "loss": 0.041, "step": 15250 }, { "grad_norm": 0.15356892347335815, "learning_rate": 8.32289954743071e-05, "loss": 0.0275, "step": 15260 }, { "grad_norm": 0.1543225795030594, "learning_rate": 8.320427818661532e-05, "loss": 0.0266, "step": 15270 }, { "grad_norm": 0.11480923742055893, "learning_rate": 8.317954637424419e-05, "loss": 0.0248, "step": 15280 }, { "grad_norm": 0.20845839381217957, "learning_rate": 8.315480004801221e-05, "loss": 0.0273, "step": 15290 }, { "grad_norm": 0.17743591964244843, "learning_rate": 8.31300392187443e-05, "loss": 0.0238, "step": 15300 }, { "grad_norm": 0.1123829260468483, "learning_rate": 8.31052638972717e-05, "loss": 0.0294, "step": 15310 }, { "grad_norm": 0.20873740315437317, "learning_rate": 8.30804740944319e-05, "loss": 0.0234, "step": 15320 }, { "grad_norm": 0.19704008102416992, "learning_rate": 8.305566982106887e-05, "loss": 0.0252, "step": 15330 }, { "grad_norm": 0.2385089248418808, "learning_rate": 8.303085108803282e-05, "loss": 0.0266, "step": 15340 }, { "grad_norm": 0.1622633934020996, "learning_rate": 8.300601790618032e-05, "loss": 0.029, "step": 15350 }, { "grad_norm": 0.14742417633533478, "learning_rate": 8.298117028637423e-05, "loss": 0.0242, "step": 15360 }, { "grad_norm": 0.2569963037967682, "learning_rate": 8.295630823948375e-05, "loss": 0.0269, "step": 15370 }, { "grad_norm": 0.18300127983093262, "learning_rate": 8.293143177638438e-05, "loss": 0.0227, "step": 15380 }, { "grad_norm": 0.27549365162849426, "learning_rate": 8.290654090795795e-05, "loss": 0.0227, "step": 15390 }, { "grad_norm": 0.1610165685415268, "learning_rate": 8.288163564509253e-05, "loss": 0.0246, "step": 15400 }, { "grad_norm": 0.19065764546394348, "learning_rate": 8.285671599868257e-05, "loss": 0.0207, "step": 15410 }, { "grad_norm": 0.18817760050296783, "learning_rate": 8.283178197962875e-05, "loss": 0.0235, "step": 15420 }, { "grad_norm": 0.16477525234222412, "learning_rate": 8.280683359883808e-05, "loss": 0.0323, "step": 15430 }, { "grad_norm": 0.14546342194080353, "learning_rate": 8.278187086722377e-05, "loss": 0.0251, "step": 15440 }, { "grad_norm": 0.19196739792823792, "learning_rate": 8.275689379570544e-05, "loss": 0.028, "step": 15450 }, { "grad_norm": 0.30997928977012634, "learning_rate": 8.273190239520887e-05, "loss": 0.0251, "step": 15460 }, { "grad_norm": 0.1956174522638321, "learning_rate": 8.270689667666616e-05, "loss": 0.0213, "step": 15470 }, { "grad_norm": 0.29317983984947205, "learning_rate": 8.268187665101563e-05, "loss": 0.0262, "step": 15480 }, { "grad_norm": 0.1946917623281479, "learning_rate": 8.265684232920194e-05, "loss": 0.0257, "step": 15490 }, { "grad_norm": 0.1777428686618805, "learning_rate": 8.263179372217592e-05, "loss": 0.0251, "step": 15500 }, { "grad_norm": 0.19356271624565125, "learning_rate": 8.260673084089468e-05, "loss": 0.0265, "step": 15510 }, { "grad_norm": 0.18561837077140808, "learning_rate": 8.25816536963216e-05, "loss": 0.0214, "step": 15520 }, { "grad_norm": 0.17156819999217987, "learning_rate": 8.255656229942626e-05, "loss": 0.023, "step": 15530 }, { "grad_norm": 0.10137134045362473, "learning_rate": 8.253145666118447e-05, "loss": 0.0206, "step": 15540 }, { "grad_norm": 0.13892777264118195, "learning_rate": 8.250633679257834e-05, "loss": 0.0236, "step": 15550 }, { "grad_norm": 0.11903750151395798, "learning_rate": 8.24812027045961e-05, "loss": 0.0243, "step": 15560 }, { "grad_norm": 0.14451394975185394, "learning_rate": 8.245605440823226e-05, "loss": 0.0265, "step": 15570 }, { "grad_norm": 0.21398355066776276, "learning_rate": 8.243089191448759e-05, "loss": 0.0297, "step": 15580 }, { "grad_norm": 0.1681848168373108, "learning_rate": 8.240571523436896e-05, "loss": 0.0244, "step": 15590 }, { "grad_norm": 0.15860651433467865, "learning_rate": 8.238052437888955e-05, "loss": 0.0203, "step": 15600 }, { "grad_norm": 0.186679869890213, "learning_rate": 8.235531935906866e-05, "loss": 0.0259, "step": 15610 }, { "grad_norm": 0.1795722395181656, "learning_rate": 8.233010018593185e-05, "loss": 0.0216, "step": 15620 }, { "grad_norm": 0.17581893503665924, "learning_rate": 8.230486687051082e-05, "loss": 0.0229, "step": 15630 }, { "grad_norm": 0.1254018396139145, "learning_rate": 8.22796194238435e-05, "loss": 0.0257, "step": 15640 }, { "grad_norm": 0.1289043128490448, "learning_rate": 8.225435785697397e-05, "loss": 0.0223, "step": 15650 }, { "grad_norm": 0.16260433197021484, "learning_rate": 8.222908218095252e-05, "loss": 0.0256, "step": 15660 }, { "grad_norm": 0.2003840208053589, "learning_rate": 8.220379240683553e-05, "loss": 0.0216, "step": 15670 }, { "grad_norm": 0.1960928589105606, "learning_rate": 8.217848854568566e-05, "loss": 0.0232, "step": 15680 }, { "grad_norm": 0.17682155966758728, "learning_rate": 8.215317060857167e-05, "loss": 0.0256, "step": 15690 }, { "grad_norm": 0.15924878418445587, "learning_rate": 8.212783860656847e-05, "loss": 0.0236, "step": 15700 }, { "grad_norm": 0.20281349122524261, "learning_rate": 8.210249255075717e-05, "loss": 0.0255, "step": 15710 }, { "grad_norm": 0.11973901838064194, "learning_rate": 8.207713245222496e-05, "loss": 0.0212, "step": 15720 }, { "grad_norm": 0.2756851613521576, "learning_rate": 8.205175832206522e-05, "loss": 0.0264, "step": 15730 }, { "grad_norm": 0.1597694605588913, "learning_rate": 8.202637017137747e-05, "loss": 0.0211, "step": 15740 }, { "grad_norm": 0.19632987678050995, "learning_rate": 8.200096801126734e-05, "loss": 0.0274, "step": 15750 }, { "grad_norm": 0.12295802682638168, "learning_rate": 8.197555185284658e-05, "loss": 0.0248, "step": 15760 }, { "grad_norm": 0.1891784369945526, "learning_rate": 8.19501217072331e-05, "loss": 0.0233, "step": 15770 }, { "grad_norm": 0.1807405799627304, "learning_rate": 8.192467758555091e-05, "loss": 0.0235, "step": 15780 }, { "grad_norm": 0.15460771322250366, "learning_rate": 8.189921949893014e-05, "loss": 0.0259, "step": 15790 }, { "grad_norm": 0.21737658977508545, "learning_rate": 8.187374745850702e-05, "loss": 0.0253, "step": 15800 }, { "grad_norm": 0.17103755474090576, "learning_rate": 8.184826147542387e-05, "loss": 0.0247, "step": 15810 }, { "grad_norm": 0.138347789645195, "learning_rate": 8.18227615608291e-05, "loss": 0.0254, "step": 15820 }, { "grad_norm": 0.12279053777456284, "learning_rate": 8.179724772587732e-05, "loss": 0.0266, "step": 15830 }, { "grad_norm": 0.30938902497291565, "learning_rate": 8.177171998172908e-05, "loss": 0.0255, "step": 15840 }, { "grad_norm": 0.2410738468170166, "learning_rate": 8.174617833955112e-05, "loss": 0.0242, "step": 15850 }, { "grad_norm": 0.17531630396842957, "learning_rate": 8.172062281051618e-05, "loss": 0.0219, "step": 15860 }, { "grad_norm": 0.12139636278152466, "learning_rate": 8.169505340580316e-05, "loss": 0.0271, "step": 15870 }, { "grad_norm": 0.16688264906406403, "learning_rate": 8.166947013659696e-05, "loss": 0.0266, "step": 15880 }, { "grad_norm": 0.4184979498386383, "learning_rate": 8.164387301408858e-05, "loss": 0.0255, "step": 15890 }, { "grad_norm": 0.1794683188199997, "learning_rate": 8.161826204947507e-05, "loss": 0.022, "step": 15900 }, { "grad_norm": 0.27082037925720215, "learning_rate": 8.159263725395954e-05, "loss": 0.0284, "step": 15910 }, { "grad_norm": 0.1775353103876114, "learning_rate": 8.156699863875115e-05, "loss": 0.026, "step": 15920 }, { "grad_norm": 0.22942304611206055, "learning_rate": 8.154134621506508e-05, "loss": 0.0287, "step": 15930 }, { "grad_norm": 0.15214106440544128, "learning_rate": 8.15156799941226e-05, "loss": 0.0242, "step": 15940 }, { "grad_norm": 0.13037334382534027, "learning_rate": 8.148999998715096e-05, "loss": 0.0245, "step": 15950 }, { "grad_norm": 0.19044993817806244, "learning_rate": 8.146430620538348e-05, "loss": 0.0218, "step": 15960 }, { "grad_norm": 0.19453240931034088, "learning_rate": 8.143859866005949e-05, "loss": 0.0277, "step": 15970 }, { "grad_norm": 0.14189013838768005, "learning_rate": 8.141287736242438e-05, "loss": 0.0221, "step": 15980 }, { "grad_norm": 0.1980692446231842, "learning_rate": 8.138714232372945e-05, "loss": 0.0255, "step": 15990 }, { "grad_norm": 0.2746286988258362, "learning_rate": 8.136139355523213e-05, "loss": 0.0246, "step": 16000 }, { "grad_norm": 0.1597728431224823, "learning_rate": 8.13356310681958e-05, "loss": 0.0243, "step": 16010 }, { "grad_norm": 0.15229928493499756, "learning_rate": 8.130985487388982e-05, "loss": 0.0216, "step": 16020 }, { "grad_norm": 0.23269055783748627, "learning_rate": 8.12840649835896e-05, "loss": 0.0276, "step": 16030 }, { "grad_norm": 0.27160289883613586, "learning_rate": 8.125826140857653e-05, "loss": 0.0266, "step": 16040 }, { "grad_norm": 0.13231338560581207, "learning_rate": 8.123244416013792e-05, "loss": 0.0244, "step": 16050 }, { "grad_norm": 0.17549753189086914, "learning_rate": 8.120661324956716e-05, "loss": 0.0232, "step": 16060 }, { "grad_norm": 0.16320841014385223, "learning_rate": 8.118076868816353e-05, "loss": 0.0228, "step": 16070 }, { "grad_norm": 0.20702895522117615, "learning_rate": 8.115491048723234e-05, "loss": 0.0238, "step": 16080 }, { "grad_norm": 0.09377454966306686, "learning_rate": 8.112903865808486e-05, "loss": 0.0222, "step": 16090 }, { "grad_norm": 0.2166322022676468, "learning_rate": 8.110315321203826e-05, "loss": 0.0247, "step": 16100 }, { "grad_norm": 0.11542031168937683, "learning_rate": 8.107725416041575e-05, "loss": 0.0252, "step": 16110 }, { "grad_norm": 0.19221794605255127, "learning_rate": 8.105134151454645e-05, "loss": 0.0266, "step": 16120 }, { "grad_norm": 0.21267648041248322, "learning_rate": 8.102541528576544e-05, "loss": 0.027, "step": 16130 }, { "grad_norm": 0.14468573033809662, "learning_rate": 8.099947548541371e-05, "loss": 0.021, "step": 16140 }, { "grad_norm": 0.15879003703594208, "learning_rate": 8.097352212483824e-05, "loss": 0.0245, "step": 16150 }, { "grad_norm": 0.16174913942813873, "learning_rate": 8.094755521539187e-05, "loss": 0.0216, "step": 16160 }, { "grad_norm": 0.11671948432922363, "learning_rate": 8.092157476843346e-05, "loss": 0.024, "step": 16170 }, { "grad_norm": 0.1213458850979805, "learning_rate": 8.089558079532769e-05, "loss": 0.0209, "step": 16180 }, { "grad_norm": 0.20576351881027222, "learning_rate": 8.086957330744522e-05, "loss": 0.0231, "step": 16190 }, { "grad_norm": 0.11858894675970078, "learning_rate": 8.084355231616265e-05, "loss": 0.0202, "step": 16200 }, { "grad_norm": 0.19460827112197876, "learning_rate": 8.08175178328624e-05, "loss": 0.025, "step": 16210 }, { "grad_norm": 0.29583850502967834, "learning_rate": 8.079146986893284e-05, "loss": 0.0191, "step": 16220 }, { "grad_norm": 0.13035158812999725, "learning_rate": 8.076540843576826e-05, "loss": 0.0274, "step": 16230 }, { "grad_norm": 0.2494681179523468, "learning_rate": 8.073933354476879e-05, "loss": 0.0246, "step": 16240 }, { "grad_norm": 0.20748019218444824, "learning_rate": 8.07132452073405e-05, "loss": 0.0269, "step": 16250 }, { "grad_norm": 0.17545288801193237, "learning_rate": 8.068714343489529e-05, "loss": 0.0252, "step": 16260 }, { "grad_norm": 0.15458698570728302, "learning_rate": 8.066102823885097e-05, "loss": 0.0236, "step": 16270 }, { "grad_norm": 0.20118902623653412, "learning_rate": 8.063489963063125e-05, "loss": 0.0238, "step": 16280 }, { "grad_norm": 0.15505757927894592, "learning_rate": 8.06087576216656e-05, "loss": 0.0237, "step": 16290 }, { "grad_norm": 0.13696008920669556, "learning_rate": 8.058260222338946e-05, "loss": 0.0228, "step": 16300 }, { "grad_norm": 0.28453612327575684, "learning_rate": 8.05564334472441e-05, "loss": 0.0263, "step": 16310 }, { "grad_norm": 0.10435628890991211, "learning_rate": 8.053025130467664e-05, "loss": 0.0205, "step": 16320 }, { "grad_norm": 0.1721748411655426, "learning_rate": 8.050405580714e-05, "loss": 0.0262, "step": 16330 }, { "grad_norm": 0.11688947677612305, "learning_rate": 8.047784696609301e-05, "loss": 0.0219, "step": 16340 }, { "grad_norm": 0.12492858618497849, "learning_rate": 8.045162479300029e-05, "loss": 0.0205, "step": 16350 }, { "grad_norm": 0.11167500168085098, "learning_rate": 8.042538929933232e-05, "loss": 0.0214, "step": 16360 }, { "grad_norm": 0.10890022665262222, "learning_rate": 8.039914049656539e-05, "loss": 0.0276, "step": 16370 }, { "grad_norm": 0.17493639886379242, "learning_rate": 8.037287839618163e-05, "loss": 0.0214, "step": 16380 }, { "grad_norm": 0.16130830347537994, "learning_rate": 8.034660300966898e-05, "loss": 0.0282, "step": 16390 }, { "grad_norm": 0.16807138919830322, "learning_rate": 8.032031434852115e-05, "loss": 0.0223, "step": 16400 }, { "grad_norm": 0.11891430616378784, "learning_rate": 8.029401242423773e-05, "loss": 0.0245, "step": 16410 }, { "grad_norm": 0.1967582255601883, "learning_rate": 8.026769724832406e-05, "loss": 0.0225, "step": 16420 }, { "grad_norm": 0.11507226526737213, "learning_rate": 8.02413688322913e-05, "loss": 0.0251, "step": 16430 }, { "grad_norm": 0.1429753601551056, "learning_rate": 8.021502718765639e-05, "loss": 0.0213, "step": 16440 }, { "grad_norm": 0.12615175545215607, "learning_rate": 8.018867232594204e-05, "loss": 0.0219, "step": 16450 }, { "grad_norm": 0.20520026981830597, "learning_rate": 8.016230425867679e-05, "loss": 0.0246, "step": 16460 }, { "grad_norm": 0.15678098797798157, "learning_rate": 8.013592299739494e-05, "loss": 0.0223, "step": 16470 }, { "grad_norm": 0.23705065250396729, "learning_rate": 8.01095285536365e-05, "loss": 0.0283, "step": 16480 }, { "grad_norm": 0.15505141019821167, "learning_rate": 8.008312093894735e-05, "loss": 0.0245, "step": 16490 }, { "grad_norm": 0.28769364953041077, "learning_rate": 8.005670016487903e-05, "loss": 0.021, "step": 16500 }, { "grad_norm": 0.17428459227085114, "learning_rate": 8.003026624298892e-05, "loss": 0.0215, "step": 16510 }, { "grad_norm": 0.221037819981575, "learning_rate": 8.00038191848401e-05, "loss": 0.0236, "step": 16520 }, { "grad_norm": 0.15114785730838776, "learning_rate": 7.997735900200142e-05, "loss": 0.0253, "step": 16530 }, { "grad_norm": 0.14206506311893463, "learning_rate": 7.995088570604745e-05, "loss": 0.0237, "step": 16540 }, { "grad_norm": 0.13496527075767517, "learning_rate": 7.992439930855852e-05, "loss": 0.0228, "step": 16550 }, { "grad_norm": 0.30143505334854126, "learning_rate": 7.989789982112067e-05, "loss": 0.0258, "step": 16560 }, { "grad_norm": 0.17858761548995972, "learning_rate": 7.987138725532572e-05, "loss": 0.0259, "step": 16570 }, { "grad_norm": 0.16605859994888306, "learning_rate": 7.98448616227711e-05, "loss": 0.0275, "step": 16580 }, { "grad_norm": 0.14019808173179626, "learning_rate": 7.981832293506004e-05, "loss": 0.0203, "step": 16590 }, { "grad_norm": 0.119663305580616, "learning_rate": 7.97917712038015e-05, "loss": 0.0171, "step": 16600 }, { "grad_norm": 0.10277310758829117, "learning_rate": 7.976520644061007e-05, "loss": 0.0198, "step": 16610 }, { "grad_norm": 0.23090237379074097, "learning_rate": 7.973862865710611e-05, "loss": 0.0206, "step": 16620 }, { "grad_norm": 0.12018050253391266, "learning_rate": 7.971203786491563e-05, "loss": 0.0193, "step": 16630 }, { "grad_norm": 0.2991129159927368, "learning_rate": 7.968543407567034e-05, "loss": 0.024, "step": 16640 }, { "grad_norm": 0.16128286719322205, "learning_rate": 7.965881730100765e-05, "loss": 0.025, "step": 16650 }, { "grad_norm": 0.2368653118610382, "learning_rate": 7.963218755257064e-05, "loss": 0.0209, "step": 16660 }, { "grad_norm": 0.1276330053806305, "learning_rate": 7.960554484200807e-05, "loss": 0.0279, "step": 16670 }, { "grad_norm": 0.22741807997226715, "learning_rate": 7.957888918097436e-05, "loss": 0.0237, "step": 16680 }, { "grad_norm": 0.1133093535900116, "learning_rate": 7.955222058112962e-05, "loss": 0.0228, "step": 16690 }, { "grad_norm": 0.1707005351781845, "learning_rate": 7.952553905413955e-05, "loss": 0.0211, "step": 16700 }, { "grad_norm": 0.17020264267921448, "learning_rate": 7.949884461167563e-05, "loss": 0.0267, "step": 16710 }, { "grad_norm": 0.26252827048301697, "learning_rate": 7.947213726541484e-05, "loss": 0.0313, "step": 16720 }, { "grad_norm": 0.20318211615085602, "learning_rate": 7.944541702703996e-05, "loss": 0.0291, "step": 16730 }, { "grad_norm": 0.2048724740743637, "learning_rate": 7.941868390823928e-05, "loss": 0.0247, "step": 16740 }, { "grad_norm": 0.21575702726840973, "learning_rate": 7.939193792070677e-05, "loss": 0.0281, "step": 16750 }, { "grad_norm": 0.15151962637901306, "learning_rate": 7.936517907614207e-05, "loss": 0.0264, "step": 16760 }, { "grad_norm": 0.18361331522464752, "learning_rate": 7.933840738625036e-05, "loss": 0.0233, "step": 16770 }, { "grad_norm": 0.2723053991794586, "learning_rate": 7.931162286274252e-05, "loss": 0.0274, "step": 16780 }, { "grad_norm": 0.12038698047399521, "learning_rate": 7.928482551733503e-05, "loss": 0.0237, "step": 16790 }, { "grad_norm": 0.22718192636966705, "learning_rate": 7.925801536174991e-05, "loss": 0.0258, "step": 16800 }, { "grad_norm": 0.26419663429260254, "learning_rate": 7.923119240771484e-05, "loss": 0.0248, "step": 16810 }, { "grad_norm": 0.15009157359600067, "learning_rate": 7.920435666696313e-05, "loss": 0.0233, "step": 16820 }, { "grad_norm": 0.10382391512393951, "learning_rate": 7.917750815123362e-05, "loss": 0.0227, "step": 16830 }, { "grad_norm": 0.19447314739227295, "learning_rate": 7.915064687227075e-05, "loss": 0.0255, "step": 16840 }, { "grad_norm": 0.09877549856901169, "learning_rate": 7.912377284182458e-05, "loss": 0.0232, "step": 16850 }, { "grad_norm": 0.16257713735103607, "learning_rate": 7.90968860716507e-05, "loss": 0.0261, "step": 16860 }, { "grad_norm": 0.3268103003501892, "learning_rate": 7.906998657351033e-05, "loss": 0.0212, "step": 16870 }, { "grad_norm": 0.14974159002304077, "learning_rate": 7.90430743591702e-05, "loss": 0.0231, "step": 16880 }, { "grad_norm": 0.2291640341281891, "learning_rate": 7.901614944040263e-05, "loss": 0.0229, "step": 16890 }, { "grad_norm": 0.24672865867614746, "learning_rate": 7.89892118289855e-05, "loss": 0.0233, "step": 16900 }, { "grad_norm": 0.11947348713874817, "learning_rate": 7.896226153670223e-05, "loss": 0.0202, "step": 16910 }, { "grad_norm": 0.09590490162372589, "learning_rate": 7.89352985753418e-05, "loss": 0.0228, "step": 16920 }, { "grad_norm": 0.2839618921279907, "learning_rate": 7.890832295669871e-05, "loss": 0.0218, "step": 16930 }, { "grad_norm": 0.17555569112300873, "learning_rate": 7.888133469257303e-05, "loss": 0.0232, "step": 16940 }, { "grad_norm": 0.11672692000865936, "learning_rate": 7.885433379477034e-05, "loss": 0.0262, "step": 16950 }, { "grad_norm": 0.1490037888288498, "learning_rate": 7.882732027510174e-05, "loss": 0.0248, "step": 16960 }, { "grad_norm": 0.18415844440460205, "learning_rate": 7.880029414538386e-05, "loss": 0.0241, "step": 16970 }, { "grad_norm": 0.16993752121925354, "learning_rate": 7.877325541743888e-05, "loss": 0.0233, "step": 16980 }, { "grad_norm": 0.1393348127603531, "learning_rate": 7.874620410309443e-05, "loss": 0.026, "step": 16990 }, { "grad_norm": 0.41215842962265015, "learning_rate": 7.871914021418368e-05, "loss": 0.0254, "step": 17000 }, { "grad_norm": 0.19645445048809052, "learning_rate": 7.869206376254528e-05, "loss": 0.0343, "step": 17010 }, { "grad_norm": 0.23986053466796875, "learning_rate": 7.86649747600234e-05, "loss": 0.0262, "step": 17020 }, { "grad_norm": 0.1107553243637085, "learning_rate": 7.86378732184677e-05, "loss": 0.0241, "step": 17030 }, { "grad_norm": 0.3125312328338623, "learning_rate": 7.86107591497333e-05, "loss": 0.0235, "step": 17040 }, { "grad_norm": 0.14076215028762817, "learning_rate": 7.858363256568083e-05, "loss": 0.022, "step": 17050 }, { "grad_norm": 0.35658401250839233, "learning_rate": 7.855649347817637e-05, "loss": 0.03, "step": 17060 }, { "grad_norm": 0.22077420353889465, "learning_rate": 7.85293418990915e-05, "loss": 0.0202, "step": 17070 }, { "grad_norm": 0.1788552701473236, "learning_rate": 7.85021778403032e-05, "loss": 0.0235, "step": 17080 }, { "grad_norm": 0.13915926218032837, "learning_rate": 7.847500131369397e-05, "loss": 0.0204, "step": 17090 }, { "grad_norm": 0.1514752358198166, "learning_rate": 7.844781233115176e-05, "loss": 0.0211, "step": 17100 }, { "grad_norm": 0.18184047937393188, "learning_rate": 7.842061090456996e-05, "loss": 0.0234, "step": 17110 }, { "grad_norm": 0.20057259500026703, "learning_rate": 7.839339704584736e-05, "loss": 0.0229, "step": 17120 }, { "grad_norm": 0.1821182817220688, "learning_rate": 7.836617076688828e-05, "loss": 0.0232, "step": 17130 }, { "grad_norm": 0.1375734508037567, "learning_rate": 7.833893207960236e-05, "loss": 0.0202, "step": 17140 }, { "grad_norm": 0.14398056268692017, "learning_rate": 7.831168099590478e-05, "loss": 0.0246, "step": 17150 }, { "grad_norm": 0.10219105333089828, "learning_rate": 7.828441752771607e-05, "loss": 0.0215, "step": 17160 }, { "grad_norm": 0.18676277995109558, "learning_rate": 7.825714168696219e-05, "loss": 0.0247, "step": 17170 }, { "grad_norm": 0.19722314178943634, "learning_rate": 7.82298534855745e-05, "loss": 0.0213, "step": 17180 }, { "grad_norm": 0.20480626821517944, "learning_rate": 7.820255293548984e-05, "loss": 0.0182, "step": 17190 }, { "grad_norm": 0.1071401834487915, "learning_rate": 7.817524004865035e-05, "loss": 0.019, "step": 17200 }, { "grad_norm": 0.18247689306735992, "learning_rate": 7.814791483700366e-05, "loss": 0.017, "step": 17210 }, { "grad_norm": 0.1843661069869995, "learning_rate": 7.81205773125027e-05, "loss": 0.0196, "step": 17220 }, { "grad_norm": 0.17794500291347504, "learning_rate": 7.809322748710585e-05, "loss": 0.0173, "step": 17230 }, { "grad_norm": 0.1578107625246048, "learning_rate": 7.806586537277685e-05, "loss": 0.0189, "step": 17240 }, { "grad_norm": 0.11679024994373322, "learning_rate": 7.803849098148484e-05, "loss": 0.0237, "step": 17250 }, { "grad_norm": 0.23203414678573608, "learning_rate": 7.801110432520427e-05, "loss": 0.0199, "step": 17260 }, { "grad_norm": 0.14914384484291077, "learning_rate": 7.798370541591504e-05, "loss": 0.0205, "step": 17270 }, { "grad_norm": 0.09532899409532547, "learning_rate": 7.795629426560232e-05, "loss": 0.0205, "step": 17280 }, { "grad_norm": 0.10361646860837936, "learning_rate": 7.792887088625671e-05, "loss": 0.0215, "step": 17290 }, { "grad_norm": 0.23419632017612457, "learning_rate": 7.790143528987412e-05, "loss": 0.0183, "step": 17300 }, { "grad_norm": 0.12516137957572937, "learning_rate": 7.787398748845578e-05, "loss": 0.0192, "step": 17310 }, { "grad_norm": 0.060483191162347794, "learning_rate": 7.784652749400834e-05, "loss": 0.0171, "step": 17320 }, { "grad_norm": 0.1350524127483368, "learning_rate": 7.78190553185437e-05, "loss": 0.019, "step": 17330 }, { "grad_norm": 0.1296003758907318, "learning_rate": 7.779157097407915e-05, "loss": 0.0199, "step": 17340 }, { "grad_norm": 0.13308016955852509, "learning_rate": 7.776407447263725e-05, "loss": 0.0185, "step": 17350 }, { "grad_norm": 0.1865130364894867, "learning_rate": 7.773656582624593e-05, "loss": 0.0234, "step": 17360 }, { "grad_norm": 0.16397128999233246, "learning_rate": 7.770904504693837e-05, "loss": 0.0211, "step": 17370 }, { "grad_norm": 0.11006537824869156, "learning_rate": 7.768151214675314e-05, "loss": 0.0184, "step": 17380 }, { "grad_norm": 0.10319047421216965, "learning_rate": 7.765396713773402e-05, "loss": 0.0175, "step": 17390 }, { "grad_norm": 0.12832072377204895, "learning_rate": 7.762641003193017e-05, "loss": 0.0214, "step": 17400 }, { "grad_norm": 0.10078369081020355, "learning_rate": 7.759884084139597e-05, "loss": 0.0188, "step": 17410 }, { "grad_norm": 0.14952224493026733, "learning_rate": 7.757125957819116e-05, "loss": 0.0179, "step": 17420 }, { "grad_norm": 0.09266462922096252, "learning_rate": 7.754366625438069e-05, "loss": 0.0176, "step": 17430 }, { "grad_norm": 0.14560377597808838, "learning_rate": 7.751606088203483e-05, "loss": 0.0202, "step": 17440 }, { "grad_norm": 0.09854321926832199, "learning_rate": 7.74884434732291e-05, "loss": 0.0214, "step": 17450 }, { "grad_norm": 0.1630842238664627, "learning_rate": 7.746081404004431e-05, "loss": 0.0194, "step": 17460 }, { "grad_norm": 0.10661561787128448, "learning_rate": 7.743317259456649e-05, "loss": 0.0182, "step": 17470 }, { "grad_norm": 0.21082107722759247, "learning_rate": 7.740551914888696e-05, "loss": 0.0229, "step": 17480 }, { "grad_norm": 0.09071831405162811, "learning_rate": 7.737785371510229e-05, "loss": 0.0194, "step": 17490 }, { "grad_norm": 0.07333268970251083, "learning_rate": 7.735017630531426e-05, "loss": 0.0172, "step": 17500 }, { "grad_norm": 0.16343384981155396, "learning_rate": 7.732248693162991e-05, "loss": 0.0267, "step": 17510 }, { "grad_norm": 0.19926156103610992, "learning_rate": 7.729478560616153e-05, "loss": 0.0218, "step": 17520 }, { "grad_norm": 0.13478857278823853, "learning_rate": 7.726707234102659e-05, "loss": 0.0238, "step": 17530 }, { "grad_norm": 0.15853066742420197, "learning_rate": 7.723934714834785e-05, "loss": 0.0214, "step": 17540 }, { "grad_norm": 0.10333265364170074, "learning_rate": 7.721161004025323e-05, "loss": 0.0207, "step": 17550 }, { "grad_norm": 0.1435878425836563, "learning_rate": 7.718386102887588e-05, "loss": 0.0267, "step": 17560 }, { "grad_norm": 0.19932906329631805, "learning_rate": 7.715610012635418e-05, "loss": 0.024, "step": 17570 }, { "grad_norm": 0.19467371702194214, "learning_rate": 7.712832734483166e-05, "loss": 0.0247, "step": 17580 }, { "grad_norm": 0.156673401594162, "learning_rate": 7.710054269645713e-05, "loss": 0.0249, "step": 17590 }, { "grad_norm": 0.23290520906448364, "learning_rate": 7.707274619338449e-05, "loss": 0.0273, "step": 17600 }, { "grad_norm": 0.264850914478302, "learning_rate": 7.704493784777289e-05, "loss": 0.0279, "step": 17610 }, { "grad_norm": 0.13690559566020966, "learning_rate": 7.701711767178665e-05, "loss": 0.0219, "step": 17620 }, { "grad_norm": 0.15131904184818268, "learning_rate": 7.698928567759525e-05, "loss": 0.0203, "step": 17630 }, { "grad_norm": 0.14699803292751312, "learning_rate": 7.696144187737335e-05, "loss": 0.0209, "step": 17640 }, { "grad_norm": 0.18376532196998596, "learning_rate": 7.69335862833008e-05, "loss": 0.0232, "step": 17650 }, { "grad_norm": 0.35508790612220764, "learning_rate": 7.690571890756256e-05, "loss": 0.0293, "step": 17660 }, { "grad_norm": 0.295881062746048, "learning_rate": 7.687783976234877e-05, "loss": 0.0261, "step": 17670 }, { "grad_norm": 0.19836308062076569, "learning_rate": 7.684994885985472e-05, "loss": 0.0299, "step": 17680 }, { "grad_norm": 0.14700235426425934, "learning_rate": 7.682204621228083e-05, "loss": 0.0249, "step": 17690 }, { "grad_norm": 0.15064272284507751, "learning_rate": 7.679413183183267e-05, "loss": 0.0298, "step": 17700 }, { "grad_norm": 0.27549028396606445, "learning_rate": 7.676620573072093e-05, "loss": 0.0266, "step": 17710 }, { "grad_norm": 0.1761283427476883, "learning_rate": 7.673826792116145e-05, "loss": 0.0236, "step": 17720 }, { "grad_norm": 0.118454709649086, "learning_rate": 7.671031841537519e-05, "loss": 0.0248, "step": 17730 }, { "grad_norm": 0.3039356470108032, "learning_rate": 7.668235722558819e-05, "loss": 0.0334, "step": 17740 }, { "grad_norm": 0.1445164531469345, "learning_rate": 7.665438436403161e-05, "loss": 0.0329, "step": 17750 }, { "grad_norm": 0.1711435765028, "learning_rate": 7.662639984294178e-05, "loss": 0.0285, "step": 17760 }, { "grad_norm": 0.10974177718162537, "learning_rate": 7.659840367456002e-05, "loss": 0.0248, "step": 17770 }, { "grad_norm": 0.26655930280685425, "learning_rate": 7.657039587113287e-05, "loss": 0.0295, "step": 17780 }, { "grad_norm": 0.15826477110385895, "learning_rate": 7.654237644491182e-05, "loss": 0.0232, "step": 17790 }, { "grad_norm": 0.14192484319210052, "learning_rate": 7.651434540815358e-05, "loss": 0.0271, "step": 17800 }, { "grad_norm": 0.16068659722805023, "learning_rate": 7.648630277311986e-05, "loss": 0.0307, "step": 17810 }, { "grad_norm": 0.2368188053369522, "learning_rate": 7.645824855207744e-05, "loss": 0.0293, "step": 17820 }, { "grad_norm": 0.15246719121932983, "learning_rate": 7.643018275729821e-05, "loss": 0.0279, "step": 17830 }, { "grad_norm": 0.1347208321094513, "learning_rate": 7.640210540105911e-05, "loss": 0.0259, "step": 17840 }, { "grad_norm": 0.10072208940982819, "learning_rate": 7.637401649564213e-05, "loss": 0.0259, "step": 17850 }, { "grad_norm": 0.13779497146606445, "learning_rate": 7.63459160533343e-05, "loss": 0.0249, "step": 17860 }, { "grad_norm": 0.17426112294197083, "learning_rate": 7.631780408642769e-05, "loss": 0.0283, "step": 17870 }, { "grad_norm": 0.25488370656967163, "learning_rate": 7.628968060721946e-05, "loss": 0.0238, "step": 17880 }, { "grad_norm": 0.19185343384742737, "learning_rate": 7.626154562801176e-05, "loss": 0.0236, "step": 17890 }, { "grad_norm": 0.22942133247852325, "learning_rate": 7.623339916111177e-05, "loss": 0.0249, "step": 17900 }, { "grad_norm": 0.31379249691963196, "learning_rate": 7.620524121883174e-05, "loss": 0.025, "step": 17910 }, { "grad_norm": 0.19422808289527893, "learning_rate": 7.617707181348892e-05, "loss": 0.0251, "step": 17920 }, { "grad_norm": 0.152276873588562, "learning_rate": 7.614889095740554e-05, "loss": 0.0244, "step": 17930 }, { "grad_norm": 0.12929724156856537, "learning_rate": 7.612069866290885e-05, "loss": 0.0242, "step": 17940 }, { "grad_norm": 0.09090364724397659, "learning_rate": 7.609249494233115e-05, "loss": 0.0273, "step": 17950 }, { "grad_norm": 0.26183071732521057, "learning_rate": 7.60642798080097e-05, "loss": 0.0267, "step": 17960 }, { "grad_norm": 0.16109226644039154, "learning_rate": 7.603605327228673e-05, "loss": 0.0213, "step": 17970 }, { "grad_norm": 0.14907123148441315, "learning_rate": 7.600781534750953e-05, "loss": 0.0297, "step": 17980 }, { "grad_norm": 0.11058637499809265, "learning_rate": 7.597956604603029e-05, "loss": 0.0255, "step": 17990 }, { "grad_norm": 0.14439459145069122, "learning_rate": 7.595130538020622e-05, "loss": 0.0253, "step": 18000 }, { "grad_norm": 0.1558893471956253, "learning_rate": 7.592303336239952e-05, "loss": 0.033, "step": 18010 }, { "grad_norm": 0.2112971991300583, "learning_rate": 7.589475000497733e-05, "loss": 0.0275, "step": 18020 }, { "grad_norm": 0.12226345390081406, "learning_rate": 7.586645532031171e-05, "loss": 0.0266, "step": 18030 }, { "grad_norm": 0.21455135941505432, "learning_rate": 7.583814932077975e-05, "loss": 0.0259, "step": 18040 }, { "grad_norm": 0.17794297635555267, "learning_rate": 7.580983201876343e-05, "loss": 0.0271, "step": 18050 }, { "grad_norm": 0.4291195571422577, "learning_rate": 7.578150342664973e-05, "loss": 0.024, "step": 18060 }, { "grad_norm": 0.16342157125473022, "learning_rate": 7.575316355683055e-05, "loss": 0.0268, "step": 18070 }, { "grad_norm": 0.16921553015708923, "learning_rate": 7.572481242170266e-05, "loss": 0.0261, "step": 18080 }, { "grad_norm": 0.13644464313983917, "learning_rate": 7.569645003366786e-05, "loss": 0.0221, "step": 18090 }, { "grad_norm": 0.1099173054099083, "learning_rate": 7.566807640513278e-05, "loss": 0.0286, "step": 18100 }, { "grad_norm": 0.2143206149339676, "learning_rate": 7.563969154850905e-05, "loss": 0.0246, "step": 18110 }, { "grad_norm": 0.1640574336051941, "learning_rate": 7.561129547621314e-05, "loss": 0.0257, "step": 18120 }, { "grad_norm": 0.1983177214860916, "learning_rate": 7.558288820066648e-05, "loss": 0.0256, "step": 18130 }, { "grad_norm": 0.16515256464481354, "learning_rate": 7.555446973429534e-05, "loss": 0.0237, "step": 18140 }, { "grad_norm": 0.14967605471611023, "learning_rate": 7.552604008953097e-05, "loss": 0.0251, "step": 18150 }, { "grad_norm": 0.13660411536693573, "learning_rate": 7.549759927880942e-05, "loss": 0.0253, "step": 18160 }, { "grad_norm": 0.15892140567302704, "learning_rate": 7.54691473145717e-05, "loss": 0.024, "step": 18170 }, { "grad_norm": 0.1060541495680809, "learning_rate": 7.544068420926365e-05, "loss": 0.0239, "step": 18180 }, { "grad_norm": 0.07779771834611893, "learning_rate": 7.5412209975336e-05, "loss": 0.0245, "step": 18190 }, { "grad_norm": 0.2576160728931427, "learning_rate": 7.538372462524435e-05, "loss": 0.0229, "step": 18200 }, { "grad_norm": 0.10141865909099579, "learning_rate": 7.535522817144915e-05, "loss": 0.0225, "step": 18210 }, { "grad_norm": 0.23376202583312988, "learning_rate": 7.532672062641575e-05, "loss": 0.0226, "step": 18220 }, { "grad_norm": 0.13870766758918762, "learning_rate": 7.529820200261427e-05, "loss": 0.0205, "step": 18230 }, { "grad_norm": 0.17489242553710938, "learning_rate": 7.526967231251977e-05, "loss": 0.0226, "step": 18240 }, { "grad_norm": 0.1776604950428009, "learning_rate": 7.524113156861206e-05, "loss": 0.026, "step": 18250 }, { "grad_norm": 0.1973070204257965, "learning_rate": 7.521257978337587e-05, "loss": 0.0229, "step": 18260 }, { "grad_norm": 0.2215711921453476, "learning_rate": 7.518401696930072e-05, "loss": 0.0231, "step": 18270 }, { "grad_norm": 0.18442007899284363, "learning_rate": 7.515544313888092e-05, "loss": 0.02, "step": 18280 }, { "grad_norm": 0.1293327361345291, "learning_rate": 7.512685830461568e-05, "loss": 0.0199, "step": 18290 }, { "grad_norm": 0.29949522018432617, "learning_rate": 7.509826247900892e-05, "loss": 0.0197, "step": 18300 }, { "grad_norm": 0.12854838371276855, "learning_rate": 7.50696556745695e-05, "loss": 0.0221, "step": 18310 }, { "grad_norm": 0.1747557818889618, "learning_rate": 7.504103790381095e-05, "loss": 0.0206, "step": 18320 }, { "grad_norm": 0.1648508459329605, "learning_rate": 7.501240917925167e-05, "loss": 0.019, "step": 18330 }, { "grad_norm": 0.15439848601818085, "learning_rate": 7.498376951341483e-05, "loss": 0.0215, "step": 18340 }, { "grad_norm": 0.15037524700164795, "learning_rate": 7.495511891882844e-05, "loss": 0.0187, "step": 18350 }, { "grad_norm": 0.1596088409423828, "learning_rate": 7.492645740802517e-05, "loss": 0.0232, "step": 18360 }, { "grad_norm": 0.13529236614704132, "learning_rate": 7.489778499354259e-05, "loss": 0.0206, "step": 18370 }, { "grad_norm": 0.21919625997543335, "learning_rate": 7.486910168792298e-05, "loss": 0.022, "step": 18380 }, { "grad_norm": 0.0990520566701889, "learning_rate": 7.484040750371337e-05, "loss": 0.02, "step": 18390 }, { "grad_norm": 0.22740072011947632, "learning_rate": 7.48117024534656e-05, "loss": 0.0227, "step": 18400 }, { "grad_norm": 0.2349415421485901, "learning_rate": 7.478298654973621e-05, "loss": 0.022, "step": 18410 }, { "grad_norm": 0.13808032870292664, "learning_rate": 7.475425980508654e-05, "loss": 0.0254, "step": 18420 }, { "grad_norm": 0.09155923873186111, "learning_rate": 7.47255222320826e-05, "loss": 0.0203, "step": 18430 }, { "grad_norm": 0.2210679054260254, "learning_rate": 7.469677384329522e-05, "loss": 0.0225, "step": 18440 }, { "grad_norm": 0.10925287008285522, "learning_rate": 7.46680146512999e-05, "loss": 0.0237, "step": 18450 }, { "grad_norm": 0.12954969704151154, "learning_rate": 7.463924466867692e-05, "loss": 0.0213, "step": 18460 }, { "grad_norm": 0.2873817980289459, "learning_rate": 7.461046390801121e-05, "loss": 0.0246, "step": 18470 }, { "grad_norm": 0.12612652778625488, "learning_rate": 7.458167238189248e-05, "loss": 0.0211, "step": 18480 }, { "grad_norm": 0.1561654508113861, "learning_rate": 7.455287010291509e-05, "loss": 0.021, "step": 18490 }, { "grad_norm": 0.13164976239204407, "learning_rate": 7.452405708367818e-05, "loss": 0.0236, "step": 18500 }, { "grad_norm": 0.1379188448190689, "learning_rate": 7.449523333678549e-05, "loss": 0.02, "step": 18510 }, { "grad_norm": 0.1725103110074997, "learning_rate": 7.446639887484555e-05, "loss": 0.0206, "step": 18520 }, { "grad_norm": 0.18867331743240356, "learning_rate": 7.443755371047152e-05, "loss": 0.0203, "step": 18530 }, { "grad_norm": 0.11301585286855698, "learning_rate": 7.440869785628124e-05, "loss": 0.0194, "step": 18540 }, { "grad_norm": 0.0893312394618988, "learning_rate": 7.437983132489727e-05, "loss": 0.0211, "step": 18550 }, { "grad_norm": 0.11590966582298279, "learning_rate": 7.435095412894678e-05, "loss": 0.0254, "step": 18560 }, { "grad_norm": 0.24453361332416534, "learning_rate": 7.432206628106165e-05, "loss": 0.0186, "step": 18570 }, { "grad_norm": 0.19959725439548492, "learning_rate": 7.429316779387842e-05, "loss": 0.0203, "step": 18580 }, { "grad_norm": 0.19017216563224792, "learning_rate": 7.426425868003825e-05, "loss": 0.0202, "step": 18590 }, { "grad_norm": 0.21124348044395447, "learning_rate": 7.423533895218698e-05, "loss": 0.0235, "step": 18600 }, { "grad_norm": 0.17565004527568817, "learning_rate": 7.420640862297507e-05, "loss": 0.0207, "step": 18610 }, { "grad_norm": 0.1205807626247406, "learning_rate": 7.417746770505766e-05, "loss": 0.0236, "step": 18620 }, { "grad_norm": 0.29606008529663086, "learning_rate": 7.414851621109444e-05, "loss": 0.0391, "step": 18630 }, { "grad_norm": 0.21004149317741394, "learning_rate": 7.411955415374982e-05, "loss": 0.0188, "step": 18640 }, { "grad_norm": 0.15400291979312897, "learning_rate": 7.409058154569278e-05, "loss": 0.0174, "step": 18650 }, { "grad_norm": 0.13973748683929443, "learning_rate": 7.406159839959689e-05, "loss": 0.0218, "step": 18660 }, { "grad_norm": 0.15598422288894653, "learning_rate": 7.403260472814039e-05, "loss": 0.0218, "step": 18670 }, { "grad_norm": 0.19498351216316223, "learning_rate": 7.400360054400611e-05, "loss": 0.023, "step": 18680 }, { "grad_norm": 0.09299484640359879, "learning_rate": 7.397458585988143e-05, "loss": 0.0199, "step": 18690 }, { "grad_norm": 0.0758303701877594, "learning_rate": 7.394556068845841e-05, "loss": 0.0193, "step": 18700 }, { "grad_norm": 0.08000867813825607, "learning_rate": 7.39165250424336e-05, "loss": 0.0197, "step": 18710 }, { "grad_norm": 0.09825547784566879, "learning_rate": 7.38874789345082e-05, "loss": 0.0195, "step": 18720 }, { "grad_norm": 0.21979963779449463, "learning_rate": 7.385842237738794e-05, "loss": 0.0315, "step": 18730 }, { "grad_norm": 0.1596507728099823, "learning_rate": 7.382935538378318e-05, "loss": 0.0209, "step": 18740 }, { "grad_norm": 0.09030274301767349, "learning_rate": 7.38002779664088e-05, "loss": 0.0183, "step": 18750 }, { "grad_norm": 0.2293124943971634, "learning_rate": 7.377119013798422e-05, "loss": 0.0239, "step": 18760 }, { "grad_norm": 0.12906889617443085, "learning_rate": 7.374209191123349e-05, "loss": 0.0269, "step": 18770 }, { "grad_norm": 0.2525954246520996, "learning_rate": 7.371298329888512e-05, "loss": 0.028, "step": 18780 }, { "grad_norm": 0.15157979726791382, "learning_rate": 7.368386431367223e-05, "loss": 0.0194, "step": 18790 }, { "grad_norm": 0.14895053207874298, "learning_rate": 7.365473496833246e-05, "loss": 0.0209, "step": 18800 }, { "grad_norm": 0.13078810274600983, "learning_rate": 7.362559527560797e-05, "loss": 0.0242, "step": 18810 }, { "grad_norm": 0.12644436955451965, "learning_rate": 7.359644524824545e-05, "loss": 0.0233, "step": 18820 }, { "grad_norm": 0.10153555870056152, "learning_rate": 7.35672848989961e-05, "loss": 0.0218, "step": 18830 }, { "grad_norm": 0.1298433393239975, "learning_rate": 7.353811424061565e-05, "loss": 0.0216, "step": 18840 }, { "grad_norm": 0.13640207052230835, "learning_rate": 7.350893328586435e-05, "loss": 0.0229, "step": 18850 }, { "grad_norm": 0.1726335883140564, "learning_rate": 7.347974204750696e-05, "loss": 0.0247, "step": 18860 }, { "grad_norm": 0.2481941133737564, "learning_rate": 7.345054053831266e-05, "loss": 0.0228, "step": 18870 }, { "grad_norm": 0.16100434958934784, "learning_rate": 7.342132877105523e-05, "loss": 0.0204, "step": 18880 }, { "grad_norm": 0.22357209026813507, "learning_rate": 7.339210675851288e-05, "loss": 0.021, "step": 18890 }, { "grad_norm": 0.10320866107940674, "learning_rate": 7.336287451346831e-05, "loss": 0.0284, "step": 18900 }, { "grad_norm": 0.19500026106834412, "learning_rate": 7.333363204870868e-05, "loss": 0.0212, "step": 18910 }, { "grad_norm": 0.10353472083806992, "learning_rate": 7.330437937702566e-05, "loss": 0.0212, "step": 18920 }, { "grad_norm": 0.2677724063396454, "learning_rate": 7.327511651121534e-05, "loss": 0.0245, "step": 18930 }, { "grad_norm": 0.12865523993968964, "learning_rate": 7.324584346407833e-05, "loss": 0.0216, "step": 18940 }, { "grad_norm": 0.1253816932439804, "learning_rate": 7.321656024841962e-05, "loss": 0.0197, "step": 18950 }, { "grad_norm": 0.23254697024822235, "learning_rate": 7.31872668770487e-05, "loss": 0.021, "step": 18960 }, { "grad_norm": 0.13927045464515686, "learning_rate": 7.315796336277949e-05, "loss": 0.0227, "step": 18970 }, { "grad_norm": 0.1076284795999527, "learning_rate": 7.312864971843031e-05, "loss": 0.0223, "step": 18980 }, { "grad_norm": 0.16564922034740448, "learning_rate": 7.309932595682398e-05, "loss": 0.0222, "step": 18990 }, { "grad_norm": 0.19841599464416504, "learning_rate": 7.30699920907877e-05, "loss": 0.0219, "step": 19000 }, { "grad_norm": 0.15057730674743652, "learning_rate": 7.304064813315308e-05, "loss": 0.0227, "step": 19010 }, { "grad_norm": 0.18821962177753448, "learning_rate": 7.30112940967562e-05, "loss": 0.0239, "step": 19020 }, { "grad_norm": 0.10442099720239639, "learning_rate": 7.298192999443747e-05, "loss": 0.0197, "step": 19030 }, { "grad_norm": 0.1071915552020073, "learning_rate": 7.295255583904179e-05, "loss": 0.0232, "step": 19040 }, { "grad_norm": 0.3097374737262726, "learning_rate": 7.29231716434184e-05, "loss": 0.0233, "step": 19050 }, { "grad_norm": 0.1663050800561905, "learning_rate": 7.289377742042092e-05, "loss": 0.0251, "step": 19060 }, { "grad_norm": 0.12828229367733002, "learning_rate": 7.286437318290741e-05, "loss": 0.0252, "step": 19070 }, { "grad_norm": 0.11713403463363647, "learning_rate": 7.283495894374027e-05, "loss": 0.0187, "step": 19080 }, { "grad_norm": 0.11808176338672638, "learning_rate": 7.280553471578629e-05, "loss": 0.021, "step": 19090 }, { "grad_norm": 0.199058398604393, "learning_rate": 7.277610051191663e-05, "loss": 0.0244, "step": 19100 }, { "grad_norm": 0.0972195714712143, "learning_rate": 7.27466563450068e-05, "loss": 0.0198, "step": 19110 }, { "grad_norm": 0.10236507654190063, "learning_rate": 7.271720222793668e-05, "loss": 0.0192, "step": 19120 }, { "grad_norm": 0.171878382563591, "learning_rate": 7.268773817359054e-05, "loss": 0.0273, "step": 19130 }, { "grad_norm": 0.11459974199533463, "learning_rate": 7.265826419485689e-05, "loss": 0.0242, "step": 19140 }, { "grad_norm": 0.17571379244327545, "learning_rate": 7.26287803046287e-05, "loss": 0.0216, "step": 19150 }, { "grad_norm": 0.1902099847793579, "learning_rate": 7.259928651580322e-05, "loss": 0.0226, "step": 19160 }, { "grad_norm": 0.13568131625652313, "learning_rate": 7.256978284128202e-05, "loss": 0.0208, "step": 19170 }, { "grad_norm": 0.11029903590679169, "learning_rate": 7.254026929397101e-05, "loss": 0.0215, "step": 19180 }, { "grad_norm": 0.19833141565322876, "learning_rate": 7.251074588678044e-05, "loss": 0.0367, "step": 19190 }, { "grad_norm": 0.15361285209655762, "learning_rate": 7.248121263262482e-05, "loss": 0.0195, "step": 19200 }, { "grad_norm": 0.07784315198659897, "learning_rate": 7.245166954442304e-05, "loss": 0.0215, "step": 19210 }, { "grad_norm": 0.18983027338981628, "learning_rate": 7.24221166350982e-05, "loss": 0.0204, "step": 19220 }, { "grad_norm": 0.19625234603881836, "learning_rate": 7.239255391757777e-05, "loss": 0.0208, "step": 19230 }, { "grad_norm": 0.11198417842388153, "learning_rate": 7.236298140479351e-05, "loss": 0.0215, "step": 19240 }, { "grad_norm": 0.10988398641347885, "learning_rate": 7.233339910968141e-05, "loss": 0.0202, "step": 19250 }, { "grad_norm": 0.14402520656585693, "learning_rate": 7.230380704518178e-05, "loss": 0.0207, "step": 19260 }, { "grad_norm": 0.1399247944355011, "learning_rate": 7.227420522423919e-05, "loss": 0.0181, "step": 19270 }, { "grad_norm": 0.16108128428459167, "learning_rate": 7.224459365980249e-05, "loss": 0.0207, "step": 19280 }, { "grad_norm": 0.22256852686405182, "learning_rate": 7.221497236482479e-05, "loss": 0.0264, "step": 19290 }, { "grad_norm": 0.14047247171401978, "learning_rate": 7.218534135226345e-05, "loss": 0.0218, "step": 19300 }, { "grad_norm": 0.19714735448360443, "learning_rate": 7.215570063508005e-05, "loss": 0.0193, "step": 19310 }, { "grad_norm": 0.16430068016052246, "learning_rate": 7.21260502262405e-05, "loss": 0.0266, "step": 19320 }, { "grad_norm": 0.17570507526397705, "learning_rate": 7.209639013871487e-05, "loss": 0.0332, "step": 19330 }, { "grad_norm": 0.2811044752597809, "learning_rate": 7.206672038547746e-05, "loss": 0.0358, "step": 19340 }, { "grad_norm": 0.1714031994342804, "learning_rate": 7.203704097950687e-05, "loss": 0.0237, "step": 19350 }, { "grad_norm": 0.14055369794368744, "learning_rate": 7.200735193378587e-05, "loss": 0.0189, "step": 19360 }, { "grad_norm": 0.13407772779464722, "learning_rate": 7.197765326130145e-05, "loss": 0.0186, "step": 19370 }, { "grad_norm": 0.16271239519119263, "learning_rate": 7.194794497504481e-05, "loss": 0.03, "step": 19380 }, { "grad_norm": 0.12761230766773224, "learning_rate": 7.19182270880114e-05, "loss": 0.0197, "step": 19390 }, { "grad_norm": 0.1998453438282013, "learning_rate": 7.188849961320079e-05, "loss": 0.0177, "step": 19400 }, { "grad_norm": 0.11661679297685623, "learning_rate": 7.185876256361681e-05, "loss": 0.019, "step": 19410 }, { "grad_norm": 0.17510806024074554, "learning_rate": 7.182901595226745e-05, "loss": 0.0185, "step": 19420 }, { "grad_norm": 0.32175493240356445, "learning_rate": 7.179925979216491e-05, "loss": 0.0258, "step": 19430 }, { "grad_norm": 0.15820470452308655, "learning_rate": 7.17694940963255e-05, "loss": 0.0181, "step": 19440 }, { "grad_norm": 0.181793212890625, "learning_rate": 7.17397188777698e-05, "loss": 0.0176, "step": 19450 }, { "grad_norm": 0.1559191793203354, "learning_rate": 7.170993414952245e-05, "loss": 0.0188, "step": 19460 }, { "grad_norm": 0.12947183847427368, "learning_rate": 7.168013992461233e-05, "loss": 0.0194, "step": 19470 }, { "grad_norm": 0.1583518534898758, "learning_rate": 7.165033621607245e-05, "loss": 0.018, "step": 19480 }, { "grad_norm": 0.2494782656431198, "learning_rate": 7.162052303693995e-05, "loss": 0.0267, "step": 19490 }, { "grad_norm": 0.18591833114624023, "learning_rate": 7.159070040025614e-05, "loss": 0.0209, "step": 19500 }, { "grad_norm": 0.16178050637245178, "learning_rate": 7.156086831906646e-05, "loss": 0.0246, "step": 19510 }, { "grad_norm": 0.1573534607887268, "learning_rate": 7.153102680642044e-05, "loss": 0.0176, "step": 19520 }, { "grad_norm": 0.16237477958202362, "learning_rate": 7.150117587537182e-05, "loss": 0.0177, "step": 19530 }, { "grad_norm": 0.144809752702713, "learning_rate": 7.147131553897838e-05, "loss": 0.0211, "step": 19540 }, { "grad_norm": 0.10452719032764435, "learning_rate": 7.144144581030205e-05, "loss": 0.0212, "step": 19550 }, { "grad_norm": 0.13580957055091858, "learning_rate": 7.141156670240889e-05, "loss": 0.0253, "step": 19560 }, { "grad_norm": 0.1450706273317337, "learning_rate": 7.138167822836899e-05, "loss": 0.0206, "step": 19570 }, { "grad_norm": 0.19058719277381897, "learning_rate": 7.135178040125661e-05, "loss": 0.0248, "step": 19580 }, { "grad_norm": 0.1932228058576584, "learning_rate": 7.13218732341501e-05, "loss": 0.0209, "step": 19590 }, { "grad_norm": 0.18677377700805664, "learning_rate": 7.129195674013181e-05, "loss": 0.0217, "step": 19600 }, { "grad_norm": 0.09881462156772614, "learning_rate": 7.126203093228829e-05, "loss": 0.0181, "step": 19610 }, { "grad_norm": 0.19635409116744995, "learning_rate": 7.123209582371006e-05, "loss": 0.0194, "step": 19620 }, { "grad_norm": 0.15277163684368134, "learning_rate": 7.120215142749179e-05, "loss": 0.0195, "step": 19630 }, { "grad_norm": 0.1439335197210312, "learning_rate": 7.117219775673218e-05, "loss": 0.0206, "step": 19640 }, { "grad_norm": 0.11591483652591705, "learning_rate": 7.114223482453394e-05, "loss": 0.0199, "step": 19650 }, { "grad_norm": 0.10234595835208893, "learning_rate": 7.11122626440039e-05, "loss": 0.0165, "step": 19660 }, { "grad_norm": 0.19243676960468292, "learning_rate": 7.10822812282529e-05, "loss": 0.0187, "step": 19670 }, { "grad_norm": 0.10637567937374115, "learning_rate": 7.105229059039586e-05, "loss": 0.022, "step": 19680 }, { "grad_norm": 0.2241574227809906, "learning_rate": 7.102229074355168e-05, "loss": 0.0242, "step": 19690 }, { "grad_norm": 0.14793656766414642, "learning_rate": 7.099228170084332e-05, "loss": 0.0226, "step": 19700 }, { "grad_norm": 0.17919600009918213, "learning_rate": 7.096226347539771e-05, "loss": 0.0237, "step": 19710 }, { "grad_norm": 0.2657794952392578, "learning_rate": 7.093223608034592e-05, "loss": 0.0258, "step": 19720 }, { "grad_norm": 0.23297525942325592, "learning_rate": 7.090219952882291e-05, "loss": 0.0202, "step": 19730 }, { "grad_norm": 0.12677112221717834, "learning_rate": 7.087215383396768e-05, "loss": 0.0227, "step": 19740 }, { "grad_norm": 0.11658245325088501, "learning_rate": 7.084209900892325e-05, "loss": 0.0186, "step": 19750 }, { "grad_norm": 0.18068240582942963, "learning_rate": 7.081203506683663e-05, "loss": 0.0255, "step": 19760 }, { "grad_norm": 0.16192898154258728, "learning_rate": 7.078196202085879e-05, "loss": 0.0211, "step": 19770 }, { "grad_norm": 0.13157080113887787, "learning_rate": 7.075187988414473e-05, "loss": 0.0235, "step": 19780 }, { "grad_norm": 0.1060536578297615, "learning_rate": 7.072178866985334e-05, "loss": 0.0223, "step": 19790 }, { "grad_norm": 0.13272570073604584, "learning_rate": 7.069168839114764e-05, "loss": 0.0216, "step": 19800 }, { "grad_norm": 0.15788386762142181, "learning_rate": 7.066157906119441e-05, "loss": 0.0216, "step": 19810 }, { "grad_norm": 0.12342218309640884, "learning_rate": 7.063146069316456e-05, "loss": 0.021, "step": 19820 }, { "grad_norm": 0.09969336539506912, "learning_rate": 7.060133330023287e-05, "loss": 0.0166, "step": 19830 }, { "grad_norm": 0.13102354109287262, "learning_rate": 7.057119689557807e-05, "loss": 0.0182, "step": 19840 }, { "grad_norm": 0.300716370344162, "learning_rate": 7.054105149238287e-05, "loss": 0.0265, "step": 19850 }, { "grad_norm": 0.17580799758434296, "learning_rate": 7.051089710383387e-05, "loss": 0.0242, "step": 19860 }, { "grad_norm": 0.16927897930145264, "learning_rate": 7.048073374312166e-05, "loss": 0.0204, "step": 19870 }, { "grad_norm": 0.10703068971633911, "learning_rate": 7.045056142344067e-05, "loss": 0.0244, "step": 19880 }, { "grad_norm": 0.22413307428359985, "learning_rate": 7.042038015798934e-05, "loss": 0.0239, "step": 19890 }, { "grad_norm": 0.34219804406166077, "learning_rate": 7.039018995996995e-05, "loss": 0.0323, "step": 19900 }, { "grad_norm": 0.18883943557739258, "learning_rate": 7.035999084258876e-05, "loss": 0.0221, "step": 19910 }, { "grad_norm": 0.19667290151119232, "learning_rate": 7.032978281905583e-05, "loss": 0.0221, "step": 19920 }, { "grad_norm": 0.21860942244529724, "learning_rate": 7.029956590258522e-05, "loss": 0.0236, "step": 19930 }, { "grad_norm": 0.1954776495695114, "learning_rate": 7.026934010639483e-05, "loss": 0.0265, "step": 19940 }, { "grad_norm": 0.16587869822978973, "learning_rate": 7.023910544370645e-05, "loss": 0.0191, "step": 19950 }, { "grad_norm": 0.16061946749687195, "learning_rate": 7.020886192774573e-05, "loss": 0.0251, "step": 19960 }, { "grad_norm": 0.15011416375637054, "learning_rate": 7.017860957174222e-05, "loss": 0.0247, "step": 19970 }, { "grad_norm": 0.27882856130599976, "learning_rate": 7.014834838892934e-05, "loss": 0.0226, "step": 19980 }, { "grad_norm": 0.15778112411499023, "learning_rate": 7.011807839254435e-05, "loss": 0.0202, "step": 19990 }, { "grad_norm": 0.18694469332695007, "learning_rate": 7.008779959582837e-05, "loss": 0.0238, "step": 20000 }, { "grad_norm": 0.1324436217546463, "learning_rate": 7.005751201202639e-05, "loss": 0.0214, "step": 20010 }, { "grad_norm": 0.10234185308218002, "learning_rate": 7.002721565438724e-05, "loss": 0.0211, "step": 20020 }, { "grad_norm": 0.19518201053142548, "learning_rate": 6.999691053616353e-05, "loss": 0.0206, "step": 20030 }, { "grad_norm": 0.1181177943944931, "learning_rate": 6.996659667061178e-05, "loss": 0.0361, "step": 20040 }, { "grad_norm": 0.17756035923957825, "learning_rate": 6.993627407099229e-05, "loss": 0.0254, "step": 20050 }, { "grad_norm": 0.18346203863620758, "learning_rate": 6.990594275056921e-05, "loss": 0.0198, "step": 20060 }, { "grad_norm": 0.21852871775627136, "learning_rate": 6.987560272261049e-05, "loss": 0.0251, "step": 20070 }, { "grad_norm": 0.1950143426656723, "learning_rate": 6.984525400038788e-05, "loss": 0.0237, "step": 20080 }, { "grad_norm": 0.17866334319114685, "learning_rate": 6.981489659717693e-05, "loss": 0.0247, "step": 20090 }, { "grad_norm": 0.1255737990140915, "learning_rate": 6.978453052625703e-05, "loss": 0.0218, "step": 20100 }, { "grad_norm": 0.11266205459833145, "learning_rate": 6.975415580091131e-05, "loss": 0.021, "step": 20110 }, { "grad_norm": 0.22826866805553436, "learning_rate": 6.972377243442672e-05, "loss": 0.0231, "step": 20120 }, { "grad_norm": 0.1635763943195343, "learning_rate": 6.969338044009396e-05, "loss": 0.0245, "step": 20130 }, { "grad_norm": 0.12229903042316437, "learning_rate": 6.966297983120753e-05, "loss": 0.0185, "step": 20140 }, { "grad_norm": 0.13904863595962524, "learning_rate": 6.963257062106571e-05, "loss": 0.0216, "step": 20150 }, { "grad_norm": 0.10664800554513931, "learning_rate": 6.960215282297048e-05, "loss": 0.0241, "step": 20160 }, { "grad_norm": 0.1772765815258026, "learning_rate": 6.957172645022765e-05, "loss": 0.0255, "step": 20170 }, { "grad_norm": 0.11257144808769226, "learning_rate": 6.954129151614673e-05, "loss": 0.0202, "step": 20180 }, { "grad_norm": 0.13260045647621155, "learning_rate": 6.9510848034041e-05, "loss": 0.0203, "step": 20190 }, { "grad_norm": 0.18073448538780212, "learning_rate": 6.94803960172275e-05, "loss": 0.0188, "step": 20200 }, { "grad_norm": 0.08625152707099915, "learning_rate": 6.944993547902692e-05, "loss": 0.0199, "step": 20210 }, { "grad_norm": 0.2956448793411255, "learning_rate": 6.941946643276379e-05, "loss": 0.019, "step": 20220 }, { "grad_norm": 0.2433054894208908, "learning_rate": 6.938898889176626e-05, "loss": 0.0232, "step": 20230 }, { "grad_norm": 0.29278799891471863, "learning_rate": 6.935850286936627e-05, "loss": 0.0204, "step": 20240 }, { "grad_norm": 0.16264909505844116, "learning_rate": 6.932800837889943e-05, "loss": 0.0219, "step": 20250 }, { "grad_norm": 0.11976873874664307, "learning_rate": 6.92975054337051e-05, "loss": 0.021, "step": 20260 }, { "grad_norm": 0.19113968312740326, "learning_rate": 6.926699404712627e-05, "loss": 0.018, "step": 20270 }, { "grad_norm": 0.12695389986038208, "learning_rate": 6.923647423250967e-05, "loss": 0.0202, "step": 20280 }, { "grad_norm": 0.15297257900238037, "learning_rate": 6.920594600320567e-05, "loss": 0.0263, "step": 20290 }, { "grad_norm": 0.1751624196767807, "learning_rate": 6.91754093725684e-05, "loss": 0.0184, "step": 20300 }, { "grad_norm": 0.18830394744873047, "learning_rate": 6.914486435395561e-05, "loss": 0.0237, "step": 20310 }, { "grad_norm": 0.12381619215011597, "learning_rate": 6.911431096072871e-05, "loss": 0.0199, "step": 20320 }, { "grad_norm": 0.13518542051315308, "learning_rate": 6.90837492062528e-05, "loss": 0.0246, "step": 20330 }, { "grad_norm": 0.2487478405237198, "learning_rate": 6.905317910389664e-05, "loss": 0.0242, "step": 20340 }, { "grad_norm": 0.15043795108795166, "learning_rate": 6.902260066703262e-05, "loss": 0.0245, "step": 20350 }, { "grad_norm": 0.08478659391403198, "learning_rate": 6.89920139090368e-05, "loss": 0.0206, "step": 20360 }, { "grad_norm": 0.14696666598320007, "learning_rate": 6.896141884328886e-05, "loss": 0.0234, "step": 20370 }, { "grad_norm": 0.19742754101753235, "learning_rate": 6.893081548317211e-05, "loss": 0.0349, "step": 20380 }, { "grad_norm": 0.1713370382785797, "learning_rate": 6.890020384207353e-05, "loss": 0.0221, "step": 20390 }, { "grad_norm": 0.11973272264003754, "learning_rate": 6.886958393338366e-05, "loss": 0.0208, "step": 20400 }, { "grad_norm": 0.2730211615562439, "learning_rate": 6.883895577049668e-05, "loss": 0.0218, "step": 20410 }, { "grad_norm": 0.1543956995010376, "learning_rate": 6.880831936681045e-05, "loss": 0.0223, "step": 20420 }, { "grad_norm": 0.16876022517681122, "learning_rate": 6.87776747357263e-05, "loss": 0.019, "step": 20430 }, { "grad_norm": 0.1947314739227295, "learning_rate": 6.874702189064927e-05, "loss": 0.0254, "step": 20440 }, { "grad_norm": 0.22109751403331757, "learning_rate": 6.871636084498796e-05, "loss": 0.0238, "step": 20450 }, { "grad_norm": 0.16074758768081665, "learning_rate": 6.868569161215453e-05, "loss": 0.0292, "step": 20460 }, { "grad_norm": 0.25203385949134827, "learning_rate": 6.865501420556477e-05, "loss": 0.0278, "step": 20470 }, { "grad_norm": 0.18252384662628174, "learning_rate": 6.862432863863802e-05, "loss": 0.0218, "step": 20480 }, { "grad_norm": 0.23880593478679657, "learning_rate": 6.859363492479716e-05, "loss": 0.0188, "step": 20490 }, { "grad_norm": 0.13990682363510132, "learning_rate": 6.856293307746868e-05, "loss": 0.0211, "step": 20500 }, { "grad_norm": 0.2027769684791565, "learning_rate": 6.853222311008263e-05, "loss": 0.0217, "step": 20510 }, { "grad_norm": 0.11480981856584549, "learning_rate": 6.850150503607256e-05, "loss": 0.0207, "step": 20520 }, { "grad_norm": 0.12700650095939636, "learning_rate": 6.847077886887562e-05, "loss": 0.0224, "step": 20530 }, { "grad_norm": 0.161909282207489, "learning_rate": 6.844004462193248e-05, "loss": 0.0199, "step": 20540 }, { "grad_norm": 0.1446620672941208, "learning_rate": 6.840930230868734e-05, "loss": 0.0225, "step": 20550 }, { "grad_norm": 0.13039126992225647, "learning_rate": 6.837855194258791e-05, "loss": 0.0213, "step": 20560 }, { "grad_norm": 0.10406218469142914, "learning_rate": 6.834779353708548e-05, "loss": 0.0199, "step": 20570 }, { "grad_norm": 0.11280099302530289, "learning_rate": 6.831702710563478e-05, "loss": 0.0205, "step": 20580 }, { "grad_norm": 0.10152328759431839, "learning_rate": 6.828625266169416e-05, "loss": 0.0242, "step": 20590 }, { "grad_norm": 0.16459554433822632, "learning_rate": 6.825547021872534e-05, "loss": 0.0243, "step": 20600 }, { "grad_norm": 0.15631091594696045, "learning_rate": 6.822467979019365e-05, "loss": 0.0229, "step": 20610 }, { "grad_norm": 0.19005495309829712, "learning_rate": 6.819388138956783e-05, "loss": 0.0224, "step": 20620 }, { "grad_norm": 0.1647394299507141, "learning_rate": 6.816307503032019e-05, "loss": 0.0219, "step": 20630 }, { "grad_norm": 0.17501676082611084, "learning_rate": 6.813226072592645e-05, "loss": 0.0234, "step": 20640 }, { "grad_norm": 0.10195299237966537, "learning_rate": 6.810143848986583e-05, "loss": 0.0173, "step": 20650 }, { "grad_norm": 0.1456180214881897, "learning_rate": 6.807060833562106e-05, "loss": 0.0193, "step": 20660 }, { "grad_norm": 0.15189601480960846, "learning_rate": 6.803977027667825e-05, "loss": 0.0177, "step": 20670 }, { "grad_norm": 0.1602291315793991, "learning_rate": 6.800892432652704e-05, "loss": 0.0219, "step": 20680 }, { "grad_norm": 0.14183950424194336, "learning_rate": 6.797807049866049e-05, "loss": 0.0197, "step": 20690 }, { "grad_norm": 0.13047075271606445, "learning_rate": 6.794720880657512e-05, "loss": 0.0232, "step": 20700 }, { "grad_norm": 0.12062834203243256, "learning_rate": 6.791633926377085e-05, "loss": 0.0235, "step": 20710 }, { "grad_norm": 0.15976816415786743, "learning_rate": 6.788546188375114e-05, "loss": 0.0247, "step": 20720 }, { "grad_norm": 0.17860227823257446, "learning_rate": 6.785457668002274e-05, "loss": 0.0339, "step": 20730 }, { "grad_norm": 0.13747118413448334, "learning_rate": 6.78236836660959e-05, "loss": 0.0188, "step": 20740 }, { "grad_norm": 0.1588360220193863, "learning_rate": 6.779278285548427e-05, "loss": 0.0265, "step": 20750 }, { "grad_norm": 0.12716518342494965, "learning_rate": 6.776187426170493e-05, "loss": 0.0222, "step": 20760 }, { "grad_norm": 0.14319385588169098, "learning_rate": 6.773095789827837e-05, "loss": 0.0226, "step": 20770 }, { "grad_norm": 0.2169671207666397, "learning_rate": 6.770003377872841e-05, "loss": 0.0248, "step": 20780 }, { "grad_norm": 0.1665796935558319, "learning_rate": 6.766910191658233e-05, "loss": 0.0203, "step": 20790 }, { "grad_norm": 0.2583707571029663, "learning_rate": 6.763816232537079e-05, "loss": 0.0283, "step": 20800 }, { "grad_norm": 0.13904346525669098, "learning_rate": 6.76072150186278e-05, "loss": 0.0209, "step": 20810 }, { "grad_norm": 0.1442999541759491, "learning_rate": 6.757626000989077e-05, "loss": 0.0266, "step": 20820 }, { "grad_norm": 0.15821100771427155, "learning_rate": 6.754529731270048e-05, "loss": 0.0213, "step": 20830 }, { "grad_norm": 0.1091063916683197, "learning_rate": 6.751432694060105e-05, "loss": 0.0225, "step": 20840 }, { "grad_norm": 0.14324553310871124, "learning_rate": 6.748334890713999e-05, "loss": 0.0232, "step": 20850 }, { "grad_norm": 0.12680791318416595, "learning_rate": 6.745236322586813e-05, "loss": 0.023, "step": 20860 }, { "grad_norm": 0.19800785183906555, "learning_rate": 6.742136991033967e-05, "loss": 0.0214, "step": 20870 }, { "grad_norm": 0.11702656745910645, "learning_rate": 6.739036897411215e-05, "loss": 0.0227, "step": 20880 }, { "grad_norm": 0.11580661684274673, "learning_rate": 6.73593604307464e-05, "loss": 0.024, "step": 20890 }, { "grad_norm": 0.20876221358776093, "learning_rate": 6.732834429380663e-05, "loss": 0.0222, "step": 20900 }, { "grad_norm": 0.22345447540283203, "learning_rate": 6.729732057686036e-05, "loss": 0.023, "step": 20910 }, { "grad_norm": 0.1313803493976593, "learning_rate": 6.726628929347839e-05, "loss": 0.0238, "step": 20920 }, { "grad_norm": 0.12039083987474442, "learning_rate": 6.723525045723488e-05, "loss": 0.0215, "step": 20930 }, { "grad_norm": 0.19077350199222565, "learning_rate": 6.720420408170728e-05, "loss": 0.0237, "step": 20940 }, { "grad_norm": 0.19846698641777039, "learning_rate": 6.71731501804763e-05, "loss": 0.0238, "step": 20950 }, { "grad_norm": 0.20457272231578827, "learning_rate": 6.714208876712602e-05, "loss": 0.0256, "step": 20960 }, { "grad_norm": 0.11036306619644165, "learning_rate": 6.71110198552437e-05, "loss": 0.0225, "step": 20970 }, { "grad_norm": 0.18789231777191162, "learning_rate": 6.707994345841996e-05, "loss": 0.0206, "step": 20980 }, { "grad_norm": 0.10098574310541153, "learning_rate": 6.70488595902487e-05, "loss": 0.0212, "step": 20990 }, { "grad_norm": 0.13513748347759247, "learning_rate": 6.7017768264327e-05, "loss": 0.0229, "step": 21000 }, { "grad_norm": 0.18276666104793549, "learning_rate": 6.698666949425532e-05, "loss": 0.0251, "step": 21010 }, { "grad_norm": 0.12766267359256744, "learning_rate": 6.69555632936373e-05, "loss": 0.0196, "step": 21020 }, { "grad_norm": 0.21303537487983704, "learning_rate": 6.692444967607983e-05, "loss": 0.0251, "step": 21030 }, { "grad_norm": 0.20740623772144318, "learning_rate": 6.689332865519309e-05, "loss": 0.0242, "step": 21040 }, { "grad_norm": 0.18969349563121796, "learning_rate": 6.686220024459045e-05, "loss": 0.0242, "step": 21050 }, { "grad_norm": 0.1244618222117424, "learning_rate": 6.683106445788856e-05, "loss": 0.0281, "step": 21060 }, { "grad_norm": 0.179499551653862, "learning_rate": 6.679992130870723e-05, "loss": 0.02, "step": 21070 }, { "grad_norm": 0.07998904585838318, "learning_rate": 6.676877081066957e-05, "loss": 0.0212, "step": 21080 }, { "grad_norm": 0.11966857314109802, "learning_rate": 6.673761297740186e-05, "loss": 0.024, "step": 21090 }, { "grad_norm": 0.19039694964885712, "learning_rate": 6.670644782253358e-05, "loss": 0.0232, "step": 21100 }, { "grad_norm": 0.1636977642774582, "learning_rate": 6.667527535969744e-05, "loss": 0.0233, "step": 21110 }, { "grad_norm": 0.15182612836360931, "learning_rate": 6.664409560252933e-05, "loss": 0.022, "step": 21120 }, { "grad_norm": 0.2697623372077942, "learning_rate": 6.661290856466833e-05, "loss": 0.0234, "step": 21130 }, { "grad_norm": 0.16953468322753906, "learning_rate": 6.658171425975672e-05, "loss": 0.0234, "step": 21140 }, { "grad_norm": 0.189499169588089, "learning_rate": 6.655051270143994e-05, "loss": 0.023, "step": 21150 }, { "grad_norm": 0.11593009531497955, "learning_rate": 6.651930390336662e-05, "loss": 0.0208, "step": 21160 }, { "grad_norm": 0.1449161171913147, "learning_rate": 6.648808787918854e-05, "loss": 0.0215, "step": 21170 }, { "grad_norm": 0.14532960951328278, "learning_rate": 6.645686464256065e-05, "loss": 0.0214, "step": 21180 }, { "grad_norm": 0.2567045986652374, "learning_rate": 6.642563420714106e-05, "loss": 0.0222, "step": 21190 }, { "grad_norm": 0.2644861042499542, "learning_rate": 6.639439658659102e-05, "loss": 0.024, "step": 21200 }, { "grad_norm": 0.18126939237117767, "learning_rate": 6.636315179457492e-05, "loss": 0.0206, "step": 21210 }, { "grad_norm": 0.10539720207452774, "learning_rate": 6.633189984476031e-05, "loss": 0.0215, "step": 21220 }, { "grad_norm": 0.16691669821739197, "learning_rate": 6.630064075081785e-05, "loss": 0.0227, "step": 21230 }, { "grad_norm": 0.1171063557267189, "learning_rate": 6.626937452642132e-05, "loss": 0.0205, "step": 21240 }, { "grad_norm": 0.11777897924184799, "learning_rate": 6.623810118524765e-05, "loss": 0.0198, "step": 21250 }, { "grad_norm": 0.16648812592029572, "learning_rate": 6.620682074097682e-05, "loss": 0.0247, "step": 21260 }, { "grad_norm": 0.1260433942079544, "learning_rate": 6.6175533207292e-05, "loss": 0.0235, "step": 21270 }, { "grad_norm": 0.1763058453798294, "learning_rate": 6.61442385978794e-05, "loss": 0.0232, "step": 21280 }, { "grad_norm": 0.16447776556015015, "learning_rate": 6.611293692642836e-05, "loss": 0.0232, "step": 21290 }, { "grad_norm": 0.08527208864688873, "learning_rate": 6.608162820663128e-05, "loss": 0.0181, "step": 21300 }, { "grad_norm": 0.19213199615478516, "learning_rate": 6.605031245218365e-05, "loss": 0.02, "step": 21310 }, { "grad_norm": 0.12062589079141617, "learning_rate": 6.601898967678408e-05, "loss": 0.0195, "step": 21320 }, { "grad_norm": 0.22167496383190155, "learning_rate": 6.598765989413419e-05, "loss": 0.0258, "step": 21330 }, { "grad_norm": 0.15082642436027527, "learning_rate": 6.595632311793867e-05, "loss": 0.0222, "step": 21340 }, { "grad_norm": 0.12876255810260773, "learning_rate": 6.592497936190534e-05, "loss": 0.0244, "step": 21350 }, { "grad_norm": 0.33417198061943054, "learning_rate": 6.5893628639745e-05, "loss": 0.0226, "step": 21360 }, { "grad_norm": 0.22610540688037872, "learning_rate": 6.586227096517151e-05, "loss": 0.0269, "step": 21370 }, { "grad_norm": 0.09005338698625565, "learning_rate": 6.58309063519018e-05, "loss": 0.023, "step": 21380 }, { "grad_norm": 0.1482081264257431, "learning_rate": 6.579953481365582e-05, "loss": 0.021, "step": 21390 }, { "grad_norm": 0.11192421615123749, "learning_rate": 6.576815636415652e-05, "loss": 0.0226, "step": 21400 }, { "grad_norm": 0.09996557235717773, "learning_rate": 6.573677101712991e-05, "loss": 0.0264, "step": 21410 }, { "grad_norm": 0.14708754420280457, "learning_rate": 6.570537878630502e-05, "loss": 0.0226, "step": 21420 }, { "grad_norm": 0.15672646462917328, "learning_rate": 6.56739796854139e-05, "loss": 0.0205, "step": 21430 }, { "grad_norm": 0.12119711935520172, "learning_rate": 6.564257372819153e-05, "loss": 0.0216, "step": 21440 }, { "grad_norm": 0.16402482986450195, "learning_rate": 6.561116092837596e-05, "loss": 0.0281, "step": 21450 }, { "grad_norm": 0.09522690623998642, "learning_rate": 6.557974129970825e-05, "loss": 0.0199, "step": 21460 }, { "grad_norm": 0.21436557173728943, "learning_rate": 6.554831485593239e-05, "loss": 0.0311, "step": 21470 }, { "grad_norm": 0.2567731440067291, "learning_rate": 6.551688161079534e-05, "loss": 0.0191, "step": 21480 }, { "grad_norm": 0.14158323407173157, "learning_rate": 6.548544157804712e-05, "loss": 0.0219, "step": 21490 }, { "grad_norm": 0.10362379252910614, "learning_rate": 6.545399477144065e-05, "loss": 0.0204, "step": 21500 }, { "grad_norm": 0.10467023402452469, "learning_rate": 6.542254120473184e-05, "loss": 0.0236, "step": 21510 }, { "grad_norm": 0.168877512216568, "learning_rate": 6.539108089167953e-05, "loss": 0.0243, "step": 21520 }, { "grad_norm": 0.17109183967113495, "learning_rate": 6.535961384604554e-05, "loss": 0.0196, "step": 21530 }, { "grad_norm": 0.15668727457523346, "learning_rate": 6.532814008159461e-05, "loss": 0.0206, "step": 21540 }, { "grad_norm": 0.1491089016199112, "learning_rate": 6.529665961209446e-05, "loss": 0.022, "step": 21550 }, { "grad_norm": 0.10925859957933426, "learning_rate": 6.52651724513157e-05, "loss": 0.0222, "step": 21560 }, { "grad_norm": 0.10044518113136292, "learning_rate": 6.523367861303189e-05, "loss": 0.0182, "step": 21570 }, { "grad_norm": 0.11945559829473495, "learning_rate": 6.520217811101951e-05, "loss": 0.021, "step": 21580 }, { "grad_norm": 0.1212381049990654, "learning_rate": 6.517067095905793e-05, "loss": 0.0233, "step": 21590 }, { "grad_norm": 0.14504580199718475, "learning_rate": 6.513915717092948e-05, "loss": 0.0203, "step": 21600 }, { "grad_norm": 0.10077040642499924, "learning_rate": 6.510763676041935e-05, "loss": 0.0234, "step": 21610 }, { "grad_norm": 0.11155501753091812, "learning_rate": 6.50761097413156e-05, "loss": 0.0179, "step": 21620 }, { "grad_norm": 0.14710856974124908, "learning_rate": 6.504457612740928e-05, "loss": 0.0236, "step": 21630 }, { "grad_norm": 0.1982964724302292, "learning_rate": 6.501303593249423e-05, "loss": 0.0193, "step": 21640 }, { "grad_norm": 0.14451365172863007, "learning_rate": 6.49814891703672e-05, "loss": 0.021, "step": 21650 }, { "grad_norm": 0.11699148267507553, "learning_rate": 6.494993585482783e-05, "loss": 0.022, "step": 21660 }, { "grad_norm": 0.1399221122264862, "learning_rate": 6.49183759996786e-05, "loss": 0.0194, "step": 21670 }, { "grad_norm": 0.07727060467004776, "learning_rate": 6.488680961872486e-05, "loss": 0.0188, "step": 21680 }, { "grad_norm": 0.12514692544937134, "learning_rate": 6.485523672577485e-05, "loss": 0.0202, "step": 21690 }, { "grad_norm": 0.12801101803779602, "learning_rate": 6.482365733463959e-05, "loss": 0.0211, "step": 21700 }, { "grad_norm": 0.11126521974802017, "learning_rate": 6.4792071459133e-05, "loss": 0.0208, "step": 21710 }, { "grad_norm": 0.20121507346630096, "learning_rate": 6.476047911307179e-05, "loss": 0.0236, "step": 21720 }, { "grad_norm": 0.1333550065755844, "learning_rate": 6.472888031027556e-05, "loss": 0.0191, "step": 21730 }, { "grad_norm": 0.08443306386470795, "learning_rate": 6.469727506456666e-05, "loss": 0.0203, "step": 21740 }, { "grad_norm": 0.12773814797401428, "learning_rate": 6.466566338977034e-05, "loss": 0.0185, "step": 21750 }, { "grad_norm": 0.1775703877210617, "learning_rate": 6.46340452997146e-05, "loss": 0.0237, "step": 21760 }, { "grad_norm": 0.2553989291191101, "learning_rate": 6.460242080823025e-05, "loss": 0.0216, "step": 21770 }, { "grad_norm": 0.13196568191051483, "learning_rate": 6.457078992915093e-05, "loss": 0.0255, "step": 21780 }, { "grad_norm": 0.1632147878408432, "learning_rate": 6.453915267631308e-05, "loss": 0.02, "step": 21790 }, { "grad_norm": 0.2072785496711731, "learning_rate": 6.450750906355589e-05, "loss": 0.0196, "step": 21800 }, { "grad_norm": 0.15197576582431793, "learning_rate": 6.447585910472134e-05, "loss": 0.0156, "step": 21810 }, { "grad_norm": 0.0933629721403122, "learning_rate": 6.444420281365424e-05, "loss": 0.0188, "step": 21820 }, { "grad_norm": 0.20844224095344543, "learning_rate": 6.441254020420208e-05, "loss": 0.0192, "step": 21830 }, { "grad_norm": 0.08144612610340118, "learning_rate": 6.438087129021522e-05, "loss": 0.0185, "step": 21840 }, { "grad_norm": 0.2003304660320282, "learning_rate": 6.434919608554663e-05, "loss": 0.0194, "step": 21850 }, { "grad_norm": 0.11649315059185028, "learning_rate": 6.43175146040522e-05, "loss": 0.0222, "step": 21860 }, { "grad_norm": 0.10940143465995789, "learning_rate": 6.428582685959045e-05, "loss": 0.0194, "step": 21870 }, { "grad_norm": 0.18323254585266113, "learning_rate": 6.42541328660227e-05, "loss": 0.0194, "step": 21880 }, { "grad_norm": 0.1999579519033432, "learning_rate": 6.422243263721293e-05, "loss": 0.0221, "step": 21890 }, { "grad_norm": 0.16443060338497162, "learning_rate": 6.419072618702794e-05, "loss": 0.0191, "step": 21900 }, { "grad_norm": 0.27817708253860474, "learning_rate": 6.41590135293372e-05, "loss": 0.0199, "step": 21910 }, { "grad_norm": 0.1334243267774582, "learning_rate": 6.41272946780129e-05, "loss": 0.0236, "step": 21920 }, { "grad_norm": 0.11424700915813446, "learning_rate": 6.409556964692996e-05, "loss": 0.0182, "step": 21930 }, { "grad_norm": 0.14793984591960907, "learning_rate": 6.406383844996595e-05, "loss": 0.0189, "step": 21940 }, { "grad_norm": 0.1452939510345459, "learning_rate": 6.403210110100121e-05, "loss": 0.0163, "step": 21950 }, { "grad_norm": 0.19714829325675964, "learning_rate": 6.400035761391871e-05, "loss": 0.0204, "step": 21960 }, { "grad_norm": 0.08246000111103058, "learning_rate": 6.396860800260416e-05, "loss": 0.0198, "step": 21970 }, { "grad_norm": 0.16223300993442535, "learning_rate": 6.393685228094588e-05, "loss": 0.0167, "step": 21980 }, { "grad_norm": 0.1726863533258438, "learning_rate": 6.390509046283493e-05, "loss": 0.0176, "step": 21990 }, { "grad_norm": 0.1551951766014099, "learning_rate": 6.387332256216498e-05, "loss": 0.0193, "step": 22000 }, { "grad_norm": 0.16653546690940857, "learning_rate": 6.384154859283243e-05, "loss": 0.0175, "step": 22010 }, { "grad_norm": 0.11502152681350708, "learning_rate": 6.380976856873627e-05, "loss": 0.0232, "step": 22020 }, { "grad_norm": 0.19051697850227356, "learning_rate": 6.377798250377815e-05, "loss": 0.0226, "step": 22030 }, { "grad_norm": 0.11483708024024963, "learning_rate": 6.374619041186238e-05, "loss": 0.0187, "step": 22040 }, { "grad_norm": 0.14772158861160278, "learning_rate": 6.371439230689592e-05, "loss": 0.0193, "step": 22050 }, { "grad_norm": 0.11089693009853363, "learning_rate": 6.368258820278833e-05, "loss": 0.0189, "step": 22060 }, { "grad_norm": 0.18116526305675507, "learning_rate": 6.365077811345182e-05, "loss": 0.018, "step": 22070 }, { "grad_norm": 0.14094626903533936, "learning_rate": 6.361896205280117e-05, "loss": 0.0176, "step": 22080 }, { "grad_norm": 0.10722627490758896, "learning_rate": 6.358714003475384e-05, "loss": 0.0226, "step": 22090 }, { "grad_norm": 0.10304684191942215, "learning_rate": 6.355531207322983e-05, "loss": 0.0191, "step": 22100 }, { "grad_norm": 0.13950349390506744, "learning_rate": 6.35234781821518e-05, "loss": 0.0293, "step": 22110 }, { "grad_norm": 0.16160981357097626, "learning_rate": 6.349163837544497e-05, "loss": 0.0235, "step": 22120 }, { "grad_norm": 0.13412386178970337, "learning_rate": 6.345979266703714e-05, "loss": 0.023, "step": 22130 }, { "grad_norm": 0.13222233951091766, "learning_rate": 6.342794107085872e-05, "loss": 0.0194, "step": 22140 }, { "grad_norm": 0.11136044561862946, "learning_rate": 6.339608360084265e-05, "loss": 0.0198, "step": 22150 }, { "grad_norm": 0.07997207343578339, "learning_rate": 6.336422027092452e-05, "loss": 0.0217, "step": 22160 }, { "grad_norm": 0.11016438156366348, "learning_rate": 6.333235109504241e-05, "loss": 0.0192, "step": 22170 }, { "grad_norm": 0.10973599553108215, "learning_rate": 6.330047608713695e-05, "loss": 0.0181, "step": 22180 }, { "grad_norm": 0.10290781408548355, "learning_rate": 6.32685952611514e-05, "loss": 0.0188, "step": 22190 }, { "grad_norm": 0.09816804528236389, "learning_rate": 6.323670863103149e-05, "loss": 0.021, "step": 22200 }, { "grad_norm": 0.2241678684949875, "learning_rate": 6.32048162107255e-05, "loss": 0.0246, "step": 22210 }, { "grad_norm": 0.08232192695140839, "learning_rate": 6.317291801418431e-05, "loss": 0.017, "step": 22220 }, { "grad_norm": 0.1499621868133545, "learning_rate": 6.31410140553612e-05, "loss": 0.0206, "step": 22230 }, { "grad_norm": 0.18691802024841309, "learning_rate": 6.310910434821211e-05, "loss": 0.0211, "step": 22240 }, { "grad_norm": 0.1430569589138031, "learning_rate": 6.307718890669541e-05, "loss": 0.0169, "step": 22250 }, { "grad_norm": 0.26246166229248047, "learning_rate": 6.304526774477198e-05, "loss": 0.0206, "step": 22260 }, { "grad_norm": 0.11854145675897598, "learning_rate": 6.301334087640526e-05, "loss": 0.0197, "step": 22270 }, { "grad_norm": 0.13761788606643677, "learning_rate": 6.298140831556112e-05, "loss": 0.0212, "step": 22280 }, { "grad_norm": 0.11462510377168655, "learning_rate": 6.294947007620794e-05, "loss": 0.0206, "step": 22290 }, { "grad_norm": 0.10117506235837936, "learning_rate": 6.291752617231661e-05, "loss": 0.0165, "step": 22300 }, { "grad_norm": 0.1030701994895935, "learning_rate": 6.288557661786047e-05, "loss": 0.0196, "step": 22310 }, { "grad_norm": 0.178611159324646, "learning_rate": 6.285362142681534e-05, "loss": 0.0248, "step": 22320 }, { "grad_norm": 0.14664903283119202, "learning_rate": 6.28216606131595e-05, "loss": 0.0184, "step": 22330 }, { "grad_norm": 0.1262696534395218, "learning_rate": 6.27896941908737e-05, "loss": 0.0214, "step": 22340 }, { "grad_norm": 0.19021201133728027, "learning_rate": 6.275772217394115e-05, "loss": 0.0216, "step": 22350 }, { "grad_norm": 0.07844226062297821, "learning_rate": 6.27257445763475e-05, "loss": 0.0189, "step": 22360 }, { "grad_norm": 0.15054461359977722, "learning_rate": 6.269376141208082e-05, "loss": 0.0219, "step": 22370 }, { "grad_norm": 0.09937486797571182, "learning_rate": 6.266177269513162e-05, "loss": 0.0191, "step": 22380 }, { "grad_norm": 0.115594282746315, "learning_rate": 6.262977843949289e-05, "loss": 0.0174, "step": 22390 }, { "grad_norm": 0.2068924903869629, "learning_rate": 6.259777865915997e-05, "loss": 0.0228, "step": 22400 }, { "grad_norm": 0.13294276595115662, "learning_rate": 6.256577336813069e-05, "loss": 0.0255, "step": 22410 }, { "grad_norm": 0.2178899645805359, "learning_rate": 6.253376258040521e-05, "loss": 0.0226, "step": 22420 }, { "grad_norm": 0.14829188585281372, "learning_rate": 6.250174630998616e-05, "loss": 0.0191, "step": 22430 }, { "grad_norm": 0.15978378057479858, "learning_rate": 6.246972457087854e-05, "loss": 0.0195, "step": 22440 }, { "grad_norm": 0.3161694407463074, "learning_rate": 6.243769737708974e-05, "loss": 0.0222, "step": 22450 }, { "grad_norm": 0.21655821800231934, "learning_rate": 6.240566474262955e-05, "loss": 0.0234, "step": 22460 }, { "grad_norm": 0.1630381941795349, "learning_rate": 6.237362668151012e-05, "loss": 0.0226, "step": 22470 }, { "grad_norm": 0.1752423495054245, "learning_rate": 6.2341583207746e-05, "loss": 0.0203, "step": 22480 }, { "grad_norm": 0.13929809629917145, "learning_rate": 6.23095343353541e-05, "loss": 0.0225, "step": 22490 }, { "grad_norm": 0.1404161900281906, "learning_rate": 6.227748007835366e-05, "loss": 0.0193, "step": 22500 }, { "grad_norm": 0.15204483270645142, "learning_rate": 6.224542045076632e-05, "loss": 0.0205, "step": 22510 }, { "grad_norm": 0.09104004502296448, "learning_rate": 6.221335546661606e-05, "loss": 0.0194, "step": 22520 }, { "grad_norm": 0.12158304452896118, "learning_rate": 6.218128513992917e-05, "loss": 0.0232, "step": 22530 }, { "grad_norm": 0.17206129431724548, "learning_rate": 6.214920948473433e-05, "loss": 0.0172, "step": 22540 }, { "grad_norm": 0.15667732059955597, "learning_rate": 6.211712851506249e-05, "loss": 0.0257, "step": 22550 }, { "grad_norm": 0.3459275960922241, "learning_rate": 6.208504224494696e-05, "loss": 0.0247, "step": 22560 }, { "grad_norm": 0.11727980524301529, "learning_rate": 6.205295068842338e-05, "loss": 0.0203, "step": 22570 }, { "grad_norm": 0.09232889860868454, "learning_rate": 6.202085385952968e-05, "loss": 0.0178, "step": 22580 }, { "grad_norm": 0.11615793406963348, "learning_rate": 6.19887517723061e-05, "loss": 0.0186, "step": 22590 }, { "grad_norm": 0.15512026846408844, "learning_rate": 6.19566444407952e-05, "loss": 0.0195, "step": 22600 }, { "grad_norm": 0.13204248249530792, "learning_rate": 6.192453187904178e-05, "loss": 0.0181, "step": 22610 }, { "grad_norm": 0.11089323461055756, "learning_rate": 6.189241410109303e-05, "loss": 0.0192, "step": 22620 }, { "grad_norm": 0.12498313933610916, "learning_rate": 6.186029112099829e-05, "loss": 0.0189, "step": 22630 }, { "grad_norm": 0.13219670951366425, "learning_rate": 6.182816295280926e-05, "loss": 0.0225, "step": 22640 }, { "grad_norm": 0.11207453906536102, "learning_rate": 6.179602961057993e-05, "loss": 0.0224, "step": 22650 }, { "grad_norm": 0.1230640560388565, "learning_rate": 6.176389110836647e-05, "loss": 0.0188, "step": 22660 }, { "grad_norm": 0.19143201410770416, "learning_rate": 6.173174746022738e-05, "loss": 0.0176, "step": 22670 }, { "grad_norm": 0.09030932933092117, "learning_rate": 6.169959868022338e-05, "loss": 0.0208, "step": 22680 }, { "grad_norm": 0.14578023552894592, "learning_rate": 6.166744478241742e-05, "loss": 0.022, "step": 22690 }, { "grad_norm": 0.22801591455936432, "learning_rate": 6.163528578087473e-05, "loss": 0.0205, "step": 22700 }, { "grad_norm": 0.10711626708507538, "learning_rate": 6.160312168966274e-05, "loss": 0.0236, "step": 22710 }, { "grad_norm": 0.13641317188739777, "learning_rate": 6.157095252285112e-05, "loss": 0.0193, "step": 22720 }, { "grad_norm": 0.23139715194702148, "learning_rate": 6.153877829451174e-05, "loss": 0.0206, "step": 22730 }, { "grad_norm": 0.11543009430170059, "learning_rate": 6.150659901871872e-05, "loss": 0.0195, "step": 22740 }, { "grad_norm": 0.17177586257457733, "learning_rate": 6.147441470954836e-05, "loss": 0.027, "step": 22750 }, { "grad_norm": 0.16033399105072021, "learning_rate": 6.144222538107918e-05, "loss": 0.0251, "step": 22760 }, { "grad_norm": 0.11732650548219681, "learning_rate": 6.141003104739188e-05, "loss": 0.0196, "step": 22770 }, { "grad_norm": 0.09834591299295425, "learning_rate": 6.137783172256935e-05, "loss": 0.0195, "step": 22780 }, { "grad_norm": 0.1343960165977478, "learning_rate": 6.134562742069669e-05, "loss": 0.0185, "step": 22790 }, { "grad_norm": 0.17192496359348297, "learning_rate": 6.131341815586111e-05, "loss": 0.0185, "step": 22800 }, { "grad_norm": 0.21187812089920044, "learning_rate": 6.128120394215209e-05, "loss": 0.0238, "step": 22810 }, { "grad_norm": 0.14980563521385193, "learning_rate": 6.124898479366116e-05, "loss": 0.0242, "step": 22820 }, { "grad_norm": 0.27396902441978455, "learning_rate": 6.121676072448214e-05, "loss": 0.0225, "step": 22830 }, { "grad_norm": 0.11874070763587952, "learning_rate": 6.118453174871089e-05, "loss": 0.0244, "step": 22840 }, { "grad_norm": 0.1944214105606079, "learning_rate": 6.115229788044548e-05, "loss": 0.0243, "step": 22850 }, { "grad_norm": 0.1463037133216858, "learning_rate": 6.112005913378607e-05, "loss": 0.0193, "step": 22860 }, { "grad_norm": 0.15291613340377808, "learning_rate": 6.1087815522835e-05, "loss": 0.021, "step": 22870 }, { "grad_norm": 0.16502845287322998, "learning_rate": 6.105556706169673e-05, "loss": 0.0254, "step": 22880 }, { "grad_norm": 0.13938288390636444, "learning_rate": 6.102331376447782e-05, "loss": 0.021, "step": 22890 }, { "grad_norm": 0.19227702915668488, "learning_rate": 6.099105564528697e-05, "loss": 0.0276, "step": 22900 }, { "grad_norm": 0.2324313074350357, "learning_rate": 6.095879271823495e-05, "loss": 0.0249, "step": 22910 }, { "grad_norm": 0.1999591588973999, "learning_rate": 6.09265249974347e-05, "loss": 0.0219, "step": 22920 }, { "grad_norm": 0.11714869737625122, "learning_rate": 6.089425249700118e-05, "loss": 0.0203, "step": 22930 }, { "grad_norm": 0.2629542350769043, "learning_rate": 6.0861975231051505e-05, "loss": 0.0184, "step": 22940 }, { "grad_norm": 0.28163647651672363, "learning_rate": 6.082969321370484e-05, "loss": 0.0246, "step": 22950 }, { "grad_norm": 0.14753638207912445, "learning_rate": 6.079740645908243e-05, "loss": 0.0196, "step": 22960 }, { "grad_norm": 0.11290643364191055, "learning_rate": 6.07651149813076e-05, "loss": 0.0196, "step": 22970 }, { "grad_norm": 0.11927596479654312, "learning_rate": 6.073281879450574e-05, "loss": 0.0212, "step": 22980 }, { "grad_norm": 0.2250404804944992, "learning_rate": 6.070051791280431e-05, "loss": 0.0232, "step": 22990 }, { "grad_norm": 0.11358234286308289, "learning_rate": 6.066821235033281e-05, "loss": 0.0202, "step": 23000 }, { "grad_norm": 0.12300065159797668, "learning_rate": 6.063590212122277e-05, "loss": 0.0205, "step": 23010 }, { "grad_norm": 0.14004109799861908, "learning_rate": 6.0603587239607794e-05, "loss": 0.0204, "step": 23020 }, { "grad_norm": 0.20232081413269043, "learning_rate": 6.057126771962353e-05, "loss": 0.0229, "step": 23030 }, { "grad_norm": 0.11722752451896667, "learning_rate": 6.05389435754076e-05, "loss": 0.0213, "step": 23040 }, { "grad_norm": 0.2813187539577484, "learning_rate": 6.0506614821099705e-05, "loss": 0.0273, "step": 23050 }, { "grad_norm": 0.13558019697666168, "learning_rate": 6.047428147084153e-05, "loss": 0.0226, "step": 23060 }, { "grad_norm": 0.19483564794063568, "learning_rate": 6.044194353877679e-05, "loss": 0.0195, "step": 23070 }, { "grad_norm": 0.14527082443237305, "learning_rate": 6.04096010390512e-05, "loss": 0.0218, "step": 23080 }, { "grad_norm": 0.19079972803592682, "learning_rate": 6.0377253985812445e-05, "loss": 0.0219, "step": 23090 }, { "grad_norm": 0.19036701321601868, "learning_rate": 6.0344902393210235e-05, "loss": 0.0246, "step": 23100 }, { "grad_norm": 0.11025292426347733, "learning_rate": 6.031254627539625e-05, "loss": 0.0241, "step": 23110 }, { "grad_norm": 0.1291188895702362, "learning_rate": 6.028018564652418e-05, "loss": 0.0248, "step": 23120 }, { "grad_norm": 0.15890341997146606, "learning_rate": 6.024782052074964e-05, "loss": 0.0192, "step": 23130 }, { "grad_norm": 0.21369940042495728, "learning_rate": 6.021545091223026e-05, "loss": 0.024, "step": 23140 }, { "grad_norm": 0.1384638547897339, "learning_rate": 6.0183076835125586e-05, "loss": 0.0275, "step": 23150 }, { "grad_norm": 0.17025332152843475, "learning_rate": 6.015069830359714e-05, "loss": 0.0204, "step": 23160 }, { "grad_norm": 0.21725787222385406, "learning_rate": 6.0118315331808384e-05, "loss": 0.0235, "step": 23170 }, { "grad_norm": 0.11118188500404358, "learning_rate": 6.0085927933924745e-05, "loss": 0.0205, "step": 23180 }, { "grad_norm": 0.16010572016239166, "learning_rate": 6.0053536124113585e-05, "loss": 0.0208, "step": 23190 }, { "grad_norm": 0.2508329451084137, "learning_rate": 6.0021139916544144e-05, "loss": 0.0222, "step": 23200 }, { "grad_norm": 0.1832638531923294, "learning_rate": 5.998873932538764e-05, "loss": 0.0213, "step": 23210 }, { "grad_norm": 0.12625901401042938, "learning_rate": 5.995633436481721e-05, "loss": 0.02, "step": 23220 }, { "grad_norm": 0.2811602056026459, "learning_rate": 5.9923925049007855e-05, "loss": 0.0207, "step": 23230 }, { "grad_norm": 0.13383036851882935, "learning_rate": 5.989151139213656e-05, "loss": 0.0187, "step": 23240 }, { "grad_norm": 0.09947169572114944, "learning_rate": 5.9859093408382094e-05, "loss": 0.0193, "step": 23250 }, { "grad_norm": 0.1160784438252449, "learning_rate": 5.982667111192523e-05, "loss": 0.0187, "step": 23260 }, { "grad_norm": 0.23421478271484375, "learning_rate": 5.979424451694857e-05, "loss": 0.0196, "step": 23270 }, { "grad_norm": 0.13008727133274078, "learning_rate": 5.97618136376366e-05, "loss": 0.0204, "step": 23280 }, { "grad_norm": 0.1250128448009491, "learning_rate": 5.97293784881757e-05, "loss": 0.0251, "step": 23290 }, { "grad_norm": 0.18699093163013458, "learning_rate": 5.969693908275411e-05, "loss": 0.021, "step": 23300 }, { "grad_norm": 0.15248438715934753, "learning_rate": 5.9664495435561916e-05, "loss": 0.0252, "step": 23310 }, { "grad_norm": 0.15986862778663635, "learning_rate": 5.963204756079108e-05, "loss": 0.0209, "step": 23320 }, { "grad_norm": 0.15880191326141357, "learning_rate": 5.959959547263538e-05, "loss": 0.0218, "step": 23330 }, { "grad_norm": 0.1096254512667656, "learning_rate": 5.95671391852905e-05, "loss": 0.0236, "step": 23340 }, { "grad_norm": 0.30492085218429565, "learning_rate": 5.95346787129539e-05, "loss": 0.0234, "step": 23350 }, { "grad_norm": 0.17465704679489136, "learning_rate": 5.950221406982489e-05, "loss": 0.0221, "step": 23360 }, { "grad_norm": 0.1484193354845047, "learning_rate": 5.9469745270104625e-05, "loss": 0.0225, "step": 23370 }, { "grad_norm": 0.1595427244901657, "learning_rate": 5.9437272327996064e-05, "loss": 0.0257, "step": 23380 }, { "grad_norm": 0.13194550573825836, "learning_rate": 5.940479525770395e-05, "loss": 0.0258, "step": 23390 }, { "grad_norm": 0.10010907799005508, "learning_rate": 5.9372314073434876e-05, "loss": 0.0232, "step": 23400 }, { "grad_norm": 0.2115546464920044, "learning_rate": 5.93398287893972e-05, "loss": 0.0251, "step": 23410 }, { "grad_norm": 0.20568594336509705, "learning_rate": 5.930733941980111e-05, "loss": 0.0195, "step": 23420 }, { "grad_norm": 0.11085113883018494, "learning_rate": 5.927484597885855e-05, "loss": 0.0213, "step": 23430 }, { "grad_norm": 0.23258736729621887, "learning_rate": 5.924234848078325e-05, "loss": 0.0206, "step": 23440 }, { "grad_norm": 0.22349779307842255, "learning_rate": 5.9209846939790715e-05, "loss": 0.0224, "step": 23450 }, { "grad_norm": 0.12842941284179688, "learning_rate": 5.917734137009824e-05, "loss": 0.0215, "step": 23460 }, { "grad_norm": 0.14574183523654938, "learning_rate": 5.914483178592485e-05, "loss": 0.0221, "step": 23470 }, { "grad_norm": 0.13483376801013947, "learning_rate": 5.911231820149136e-05, "loss": 0.0205, "step": 23480 }, { "grad_norm": 0.1837264746427536, "learning_rate": 5.907980063102031e-05, "loss": 0.0212, "step": 23490 }, { "grad_norm": 0.09197451919317245, "learning_rate": 5.904727908873597e-05, "loss": 0.0233, "step": 23500 }, { "grad_norm": 0.1423761546611786, "learning_rate": 5.90147535888644e-05, "loss": 0.0233, "step": 23510 }, { "grad_norm": 0.22563670575618744, "learning_rate": 5.898222414563333e-05, "loss": 0.0313, "step": 23520 }, { "grad_norm": 0.1381337195634842, "learning_rate": 5.8949690773272256e-05, "loss": 0.0221, "step": 23530 }, { "grad_norm": 0.18274126946926117, "learning_rate": 5.891715348601239e-05, "loss": 0.0262, "step": 23540 }, { "grad_norm": 0.11829206347465515, "learning_rate": 5.888461229808663e-05, "loss": 0.0217, "step": 23550 }, { "grad_norm": 0.11216428875923157, "learning_rate": 5.885206722372959e-05, "loss": 0.0208, "step": 23560 }, { "grad_norm": 0.10415170341730118, "learning_rate": 5.881951827717763e-05, "loss": 0.019, "step": 23570 }, { "grad_norm": 0.15479250252246857, "learning_rate": 5.878696547266872e-05, "loss": 0.0216, "step": 23580 }, { "grad_norm": 0.1280139982700348, "learning_rate": 5.875440882444259e-05, "loss": 0.0234, "step": 23590 }, { "grad_norm": 0.11368955671787262, "learning_rate": 5.8721848346740617e-05, "loss": 0.0186, "step": 23600 }, { "grad_norm": 0.144217386841774, "learning_rate": 5.868928405380585e-05, "loss": 0.0227, "step": 23610 }, { "grad_norm": 0.09783807396888733, "learning_rate": 5.865671595988304e-05, "loss": 0.0192, "step": 23620 }, { "grad_norm": 0.13858728110790253, "learning_rate": 5.862414407921857e-05, "loss": 0.0174, "step": 23630 }, { "grad_norm": 0.10957144945859909, "learning_rate": 5.8591568426060475e-05, "loss": 0.0223, "step": 23640 }, { "grad_norm": 0.17293034493923187, "learning_rate": 5.8558989014658464e-05, "loss": 0.0205, "step": 23650 }, { "grad_norm": 0.11470290273427963, "learning_rate": 5.8526405859263855e-05, "loss": 0.0197, "step": 23660 }, { "grad_norm": 0.0845152884721756, "learning_rate": 5.8493818974129667e-05, "loss": 0.0213, "step": 23670 }, { "grad_norm": 0.15347525477409363, "learning_rate": 5.846122837351048e-05, "loss": 0.0232, "step": 23680 }, { "grad_norm": 0.10071103274822235, "learning_rate": 5.842863407166255e-05, "loss": 0.0187, "step": 23690 }, { "grad_norm": 0.1316511482000351, "learning_rate": 5.83960360828437e-05, "loss": 0.0209, "step": 23700 }, { "grad_norm": 0.22864897549152374, "learning_rate": 5.836343442131343e-05, "loss": 0.0202, "step": 23710 }, { "grad_norm": 0.0758475810289383, "learning_rate": 5.833082910133281e-05, "loss": 0.0208, "step": 23720 }, { "grad_norm": 0.20465707778930664, "learning_rate": 5.82982201371645e-05, "loss": 0.0225, "step": 23730 }, { "grad_norm": 0.10585597902536392, "learning_rate": 5.826560754307279e-05, "loss": 0.019, "step": 23740 }, { "grad_norm": 0.14659827947616577, "learning_rate": 5.8232991333323516e-05, "loss": 0.0233, "step": 23750 }, { "grad_norm": 0.10795941948890686, "learning_rate": 5.8200371522184116e-05, "loss": 0.0209, "step": 23760 }, { "grad_norm": 0.07674072682857513, "learning_rate": 5.816774812392362e-05, "loss": 0.0206, "step": 23770 }, { "grad_norm": 0.10953204333782196, "learning_rate": 5.8135121152812614e-05, "loss": 0.0188, "step": 23780 }, { "grad_norm": 0.15253186225891113, "learning_rate": 5.8102490623123204e-05, "loss": 0.0194, "step": 23790 }, { "grad_norm": 0.10648788511753082, "learning_rate": 5.806985654912914e-05, "loss": 0.0185, "step": 23800 }, { "grad_norm": 0.26990801095962524, "learning_rate": 5.8037218945105644e-05, "loss": 0.0213, "step": 23810 }, { "grad_norm": 0.16410231590270996, "learning_rate": 5.800457782532953e-05, "loss": 0.0209, "step": 23820 }, { "grad_norm": 0.1362190991640091, "learning_rate": 5.797193320407912e-05, "loss": 0.0209, "step": 23830 }, { "grad_norm": 0.12427579611539841, "learning_rate": 5.793928509563431e-05, "loss": 0.022, "step": 23840 }, { "grad_norm": 0.13383392989635468, "learning_rate": 5.790663351427644e-05, "loss": 0.019, "step": 23850 }, { "grad_norm": 0.13574717938899994, "learning_rate": 5.7873978474288484e-05, "loss": 0.0201, "step": 23860 }, { "grad_norm": 0.1951388120651245, "learning_rate": 5.78413199899548e-05, "loss": 0.0227, "step": 23870 }, { "grad_norm": 0.061444059014320374, "learning_rate": 5.780865807556136e-05, "loss": 0.0185, "step": 23880 }, { "grad_norm": 0.19468937814235687, "learning_rate": 5.777599274539559e-05, "loss": 0.024, "step": 23890 }, { "grad_norm": 0.18282869458198547, "learning_rate": 5.77433240137464e-05, "loss": 0.02, "step": 23900 }, { "grad_norm": 0.21931082010269165, "learning_rate": 5.771065189490422e-05, "loss": 0.02, "step": 23910 }, { "grad_norm": 0.1696486473083496, "learning_rate": 5.767797640316095e-05, "loss": 0.0236, "step": 23920 }, { "grad_norm": 0.13328838348388672, "learning_rate": 5.764529755280994e-05, "loss": 0.02, "step": 23930 }, { "grad_norm": 0.12072398513555527, "learning_rate": 5.761261535814603e-05, "loss": 0.0168, "step": 23940 }, { "grad_norm": 0.08491555601358414, "learning_rate": 5.757992983346554e-05, "loss": 0.0172, "step": 23950 }, { "grad_norm": 0.2558961510658264, "learning_rate": 5.754724099306621e-05, "loss": 0.021, "step": 23960 }, { "grad_norm": 0.15857355296611786, "learning_rate": 5.751454885124727e-05, "loss": 0.0191, "step": 23970 }, { "grad_norm": 0.08240366727113724, "learning_rate": 5.748185342230935e-05, "loss": 0.0187, "step": 23980 }, { "grad_norm": 0.10284663736820221, "learning_rate": 5.744915472055456e-05, "loss": 0.0195, "step": 23990 }, { "grad_norm": 0.11856869608163834, "learning_rate": 5.741645276028642e-05, "loss": 0.0184, "step": 24000 }, { "grad_norm": 0.07984700053930283, "learning_rate": 5.7383747555809873e-05, "loss": 0.0168, "step": 24010 }, { "grad_norm": 0.13382591307163239, "learning_rate": 5.7351039121431294e-05, "loss": 0.0177, "step": 24020 }, { "grad_norm": 0.12416905164718628, "learning_rate": 5.7318327471458454e-05, "loss": 0.0218, "step": 24030 }, { "grad_norm": 0.1085289865732193, "learning_rate": 5.728561262020054e-05, "loss": 0.0196, "step": 24040 }, { "grad_norm": 0.17662374675273895, "learning_rate": 5.725289458196814e-05, "loss": 0.0216, "step": 24050 }, { "grad_norm": 0.09278015047311783, "learning_rate": 5.722017337107324e-05, "loss": 0.0212, "step": 24060 }, { "grad_norm": 0.11781104654073715, "learning_rate": 5.7187449001829206e-05, "loss": 0.0208, "step": 24070 }, { "grad_norm": 0.15079650282859802, "learning_rate": 5.715472148855081e-05, "loss": 0.0217, "step": 24080 }, { "grad_norm": 0.14518125355243683, "learning_rate": 5.712199084555414e-05, "loss": 0.0202, "step": 24090 }, { "grad_norm": 0.13356667757034302, "learning_rate": 5.7089257087156734e-05, "loss": 0.0246, "step": 24100 }, { "grad_norm": 0.15025806427001953, "learning_rate": 5.70565202276774e-05, "loss": 0.0187, "step": 24110 }, { "grad_norm": 0.09271620213985443, "learning_rate": 5.70237802814364e-05, "loss": 0.0211, "step": 24120 }, { "grad_norm": 0.12416230887174606, "learning_rate": 5.6991037262755277e-05, "loss": 0.0241, "step": 24130 }, { "grad_norm": 0.11875692009925842, "learning_rate": 5.6958291185956924e-05, "loss": 0.0223, "step": 24140 }, { "grad_norm": 0.12641146779060364, "learning_rate": 5.692554206536561e-05, "loss": 0.0177, "step": 24150 }, { "grad_norm": 0.11993242055177689, "learning_rate": 5.689278991530692e-05, "loss": 0.0207, "step": 24160 }, { "grad_norm": 0.11681359261274338, "learning_rate": 5.686003475010773e-05, "loss": 0.021, "step": 24170 }, { "grad_norm": 0.0873931422829628, "learning_rate": 5.682727658409628e-05, "loss": 0.0205, "step": 24180 }, { "grad_norm": 0.11899099498987198, "learning_rate": 5.67945154316021e-05, "loss": 0.0218, "step": 24190 }, { "grad_norm": 0.08579571545124054, "learning_rate": 5.676175130695602e-05, "loss": 0.0163, "step": 24200 }, { "grad_norm": 0.16540570557117462, "learning_rate": 5.672898422449022e-05, "loss": 0.0213, "step": 24210 }, { "grad_norm": 0.08555443584918976, "learning_rate": 5.66962141985381e-05, "loss": 0.0212, "step": 24220 }, { "grad_norm": 0.22644267976284027, "learning_rate": 5.666344124343439e-05, "loss": 0.0223, "step": 24230 }, { "grad_norm": 0.14623667299747467, "learning_rate": 5.66306653735151e-05, "loss": 0.0209, "step": 24240 }, { "grad_norm": 0.21199095249176025, "learning_rate": 5.659788660311749e-05, "loss": 0.0212, "step": 24250 }, { "grad_norm": 0.14466936886310577, "learning_rate": 5.656510494658014e-05, "loss": 0.0209, "step": 24260 }, { "grad_norm": 0.13327455520629883, "learning_rate": 5.6532320418242835e-05, "loss": 0.0206, "step": 24270 }, { "grad_norm": 0.12979334592819214, "learning_rate": 5.649953303244665e-05, "loss": 0.0185, "step": 24280 }, { "grad_norm": 0.24679690599441528, "learning_rate": 5.646674280353389e-05, "loss": 0.022, "step": 24290 }, { "grad_norm": 0.11144895106554031, "learning_rate": 5.6433949745848135e-05, "loss": 0.021, "step": 24300 }, { "grad_norm": 0.2912446856498718, "learning_rate": 5.6401153873734166e-05, "loss": 0.0219, "step": 24310 }, { "grad_norm": 0.1628635674715042, "learning_rate": 5.6368355201538036e-05, "loss": 0.0188, "step": 24320 }, { "grad_norm": 0.11480241268873215, "learning_rate": 5.633555374360697e-05, "loss": 0.0189, "step": 24330 }, { "grad_norm": 0.10571160167455673, "learning_rate": 5.630274951428944e-05, "loss": 0.02, "step": 24340 }, { "grad_norm": 0.1205609068274498, "learning_rate": 5.626994252793517e-05, "loss": 0.0229, "step": 24350 }, { "grad_norm": 0.11923234909772873, "learning_rate": 5.6237132798895006e-05, "loss": 0.0183, "step": 24360 }, { "grad_norm": 0.09586698561906815, "learning_rate": 5.620432034152106e-05, "loss": 0.0198, "step": 24370 }, { "grad_norm": 0.14136561751365662, "learning_rate": 5.6171505170166604e-05, "loss": 0.0184, "step": 24380 }, { "grad_norm": 0.07761842012405396, "learning_rate": 5.613868729918612e-05, "loss": 0.0158, "step": 24390 }, { "grad_norm": 0.1302981972694397, "learning_rate": 5.6105866742935256e-05, "loss": 0.0207, "step": 24400 }, { "grad_norm": 0.158559188246727, "learning_rate": 5.607304351577083e-05, "loss": 0.0213, "step": 24410 }, { "grad_norm": 0.07937338948249817, "learning_rate": 5.604021763205085e-05, "loss": 0.0205, "step": 24420 }, { "grad_norm": 0.14190024137496948, "learning_rate": 5.600738910613445e-05, "loss": 0.0216, "step": 24430 }, { "grad_norm": 0.10169994831085205, "learning_rate": 5.597455795238198e-05, "loss": 0.018, "step": 24440 }, { "grad_norm": 0.1636887788772583, "learning_rate": 5.5941724185154876e-05, "loss": 0.022, "step": 24450 }, { "grad_norm": 0.18032851815223694, "learning_rate": 5.5908887818815746e-05, "loss": 0.0219, "step": 24460 }, { "grad_norm": 0.17794863879680634, "learning_rate": 5.587604886772834e-05, "loss": 0.0208, "step": 24470 }, { "grad_norm": 0.1740294098854065, "learning_rate": 5.584320734625752e-05, "loss": 0.0206, "step": 24480 }, { "grad_norm": 0.11531418561935425, "learning_rate": 5.581036326876928e-05, "loss": 0.0191, "step": 24490 }, { "grad_norm": 0.16490326821804047, "learning_rate": 5.5777516649630756e-05, "loss": 0.0208, "step": 24500 }, { "grad_norm": 0.08763464540243149, "learning_rate": 5.574466750321017e-05, "loss": 0.0172, "step": 24510 }, { "grad_norm": 0.08968948572874069, "learning_rate": 5.571181584387683e-05, "loss": 0.0223, "step": 24520 }, { "grad_norm": 0.14001530408859253, "learning_rate": 5.5678961686001196e-05, "loss": 0.0184, "step": 24530 }, { "grad_norm": 0.13025102019309998, "learning_rate": 5.5646105043954785e-05, "loss": 0.0199, "step": 24540 }, { "grad_norm": 0.13107240200042725, "learning_rate": 5.5613245932110205e-05, "loss": 0.0194, "step": 24550 }, { "grad_norm": 0.07924743741750717, "learning_rate": 5.5580384364841157e-05, "loss": 0.0158, "step": 24560 }, { "grad_norm": 0.09501826018095016, "learning_rate": 5.554752035652237e-05, "loss": 0.0189, "step": 24570 }, { "grad_norm": 0.22992704808712006, "learning_rate": 5.5514653921529736e-05, "loss": 0.0181, "step": 24580 }, { "grad_norm": 0.14434237778186798, "learning_rate": 5.5481785074240114e-05, "loss": 0.0176, "step": 24590 }, { "grad_norm": 0.11151311546564102, "learning_rate": 5.544891382903146e-05, "loss": 0.0163, "step": 24600 }, { "grad_norm": 0.09913978725671768, "learning_rate": 5.541604020028277e-05, "loss": 0.0208, "step": 24610 }, { "grad_norm": 0.09604664146900177, "learning_rate": 5.5383164202374105e-05, "loss": 0.0162, "step": 24620 }, { "grad_norm": 0.1503785252571106, "learning_rate": 5.5350285849686515e-05, "loss": 0.0217, "step": 24630 }, { "grad_norm": 0.11098060756921768, "learning_rate": 5.531740515660213e-05, "loss": 0.0202, "step": 24640 }, { "grad_norm": 0.20954075455665588, "learning_rate": 5.528452213750408e-05, "loss": 0.0172, "step": 24650 }, { "grad_norm": 0.11847098916769028, "learning_rate": 5.52516368067765e-05, "loss": 0.021, "step": 24660 }, { "grad_norm": 0.13761067390441895, "learning_rate": 5.5218749178804563e-05, "loss": 0.0191, "step": 24670 }, { "grad_norm": 0.09543492645025253, "learning_rate": 5.518585926797443e-05, "loss": 0.0182, "step": 24680 }, { "grad_norm": 0.19036228954792023, "learning_rate": 5.5152967088673265e-05, "loss": 0.0181, "step": 24690 }, { "grad_norm": 0.22370974719524384, "learning_rate": 5.512007265528924e-05, "loss": 0.0186, "step": 24700 }, { "grad_norm": 0.14357028901576996, "learning_rate": 5.508717598221148e-05, "loss": 0.0198, "step": 24710 }, { "grad_norm": 0.13547323644161224, "learning_rate": 5.505427708383012e-05, "loss": 0.0176, "step": 24720 }, { "grad_norm": 0.1323752999305725, "learning_rate": 5.502137597453624e-05, "loss": 0.0201, "step": 24730 }, { "grad_norm": 0.143809512257576, "learning_rate": 5.4988472668721914e-05, "loss": 0.018, "step": 24740 }, { "grad_norm": 0.1270478218793869, "learning_rate": 5.4955567180780164e-05, "loss": 0.0197, "step": 24750 }, { "grad_norm": 0.17010723054409027, "learning_rate": 5.492265952510497e-05, "loss": 0.0233, "step": 24760 }, { "grad_norm": 0.1623261719942093, "learning_rate": 5.4889749716091254e-05, "loss": 0.0197, "step": 24770 }, { "grad_norm": 0.1566523313522339, "learning_rate": 5.4856837768134885e-05, "loss": 0.0195, "step": 24780 }, { "grad_norm": 0.16993916034698486, "learning_rate": 5.482392369563267e-05, "loss": 0.0202, "step": 24790 }, { "grad_norm": 0.11583800613880157, "learning_rate": 5.4791007512982354e-05, "loss": 0.0143, "step": 24800 }, { "grad_norm": 0.1058168038725853, "learning_rate": 5.4758089234582566e-05, "loss": 0.0186, "step": 24810 }, { "grad_norm": 0.21744045615196228, "learning_rate": 5.47251688748329e-05, "loss": 0.0192, "step": 24820 }, { "grad_norm": 0.2701205015182495, "learning_rate": 5.4692246448133843e-05, "loss": 0.0197, "step": 24830 }, { "grad_norm": 0.22121627628803253, "learning_rate": 5.465932196888676e-05, "loss": 0.0204, "step": 24840 }, { "grad_norm": 0.16861796379089355, "learning_rate": 5.462639545149395e-05, "loss": 0.0175, "step": 24850 }, { "grad_norm": 0.1395934522151947, "learning_rate": 5.459346691035861e-05, "loss": 0.0172, "step": 24860 }, { "grad_norm": 0.22276657819747925, "learning_rate": 5.4560536359884776e-05, "loss": 0.0206, "step": 24870 }, { "grad_norm": 0.09193438291549683, "learning_rate": 5.45276038144774e-05, "loss": 0.0158, "step": 24880 }, { "grad_norm": 0.13790106773376465, "learning_rate": 5.449466928854229e-05, "loss": 0.0193, "step": 24890 }, { "grad_norm": 0.12272381037473679, "learning_rate": 5.446173279648613e-05, "loss": 0.0198, "step": 24900 }, { "grad_norm": 0.10911954194307327, "learning_rate": 5.442879435271646e-05, "loss": 0.0181, "step": 24910 }, { "grad_norm": 0.13483071327209473, "learning_rate": 5.4395853971641675e-05, "loss": 0.0171, "step": 24920 }, { "grad_norm": 0.18806663155555725, "learning_rate": 5.4362911667671015e-05, "loss": 0.0181, "step": 24930 }, { "grad_norm": 0.13344749808311462, "learning_rate": 5.432996745521458e-05, "loss": 0.0172, "step": 24940 }, { "grad_norm": 0.23875457048416138, "learning_rate": 5.4297021348683264e-05, "loss": 0.0193, "step": 24950 }, { "grad_norm": 0.129985511302948, "learning_rate": 5.426407336248882e-05, "loss": 0.0175, "step": 24960 }, { "grad_norm": 0.10984055697917938, "learning_rate": 5.423112351104382e-05, "loss": 0.0179, "step": 24970 }, { "grad_norm": 0.17391051352024078, "learning_rate": 5.4198171808761654e-05, "loss": 0.0178, "step": 24980 }, { "grad_norm": 0.12792062759399414, "learning_rate": 5.41652182700565e-05, "loss": 0.0315, "step": 24990 }, { "grad_norm": 0.18537582457065582, "learning_rate": 5.413226290934338e-05, "loss": 0.0183, "step": 25000 }, { "grad_norm": 0.12797260284423828, "learning_rate": 5.4099305741038056e-05, "loss": 0.0189, "step": 25010 }, { "grad_norm": 0.19243748486042023, "learning_rate": 5.406634677955713e-05, "loss": 0.0199, "step": 25020 }, { "grad_norm": 0.15864787995815277, "learning_rate": 5.403338603931798e-05, "loss": 0.0267, "step": 25030 }, { "grad_norm": 0.1870935708284378, "learning_rate": 5.4000423534738745e-05, "loss": 0.0161, "step": 25040 }, { "grad_norm": 0.10862977057695389, "learning_rate": 5.396745928023835e-05, "loss": 0.0183, "step": 25050 }, { "grad_norm": 0.08676145225763321, "learning_rate": 5.3934493290236474e-05, "loss": 0.0166, "step": 25060 }, { "grad_norm": 0.16480650007724762, "learning_rate": 5.390152557915357e-05, "loss": 0.016, "step": 25070 }, { "grad_norm": 0.10029874742031097, "learning_rate": 5.3868556161410796e-05, "loss": 0.018, "step": 25080 }, { "grad_norm": 0.14859020709991455, "learning_rate": 5.383558505143015e-05, "loss": 0.019, "step": 25090 }, { "grad_norm": 0.1465212106704712, "learning_rate": 5.38026122636343e-05, "loss": 0.019, "step": 25100 }, { "grad_norm": 0.09826698899269104, "learning_rate": 5.3769637812446636e-05, "loss": 0.02, "step": 25110 }, { "grad_norm": 0.10209561139345169, "learning_rate": 5.373666171229133e-05, "loss": 0.016, "step": 25120 }, { "grad_norm": 0.1380016952753067, "learning_rate": 5.370368397759324e-05, "loss": 0.0325, "step": 25130 }, { "grad_norm": 0.09873092919588089, "learning_rate": 5.3670704622777944e-05, "loss": 0.0183, "step": 25140 }, { "grad_norm": 0.09751900285482407, "learning_rate": 5.3637723662271744e-05, "loss": 0.0173, "step": 25150 }, { "grad_norm": 0.12305276840925217, "learning_rate": 5.360474111050162e-05, "loss": 0.0161, "step": 25160 }, { "grad_norm": 0.15559633076190948, "learning_rate": 5.357175698189527e-05, "loss": 0.0159, "step": 25170 }, { "grad_norm": 0.13185887038707733, "learning_rate": 5.353877129088107e-05, "loss": 0.0163, "step": 25180 }, { "grad_norm": 0.09605438262224197, "learning_rate": 5.350578405188809e-05, "loss": 0.0182, "step": 25190 }, { "grad_norm": 0.1326892226934433, "learning_rate": 5.347279527934603e-05, "loss": 0.0127, "step": 25200 }, { "grad_norm": 0.09018957614898682, "learning_rate": 5.343980498768536e-05, "loss": 0.0189, "step": 25210 }, { "grad_norm": 0.11119314283132553, "learning_rate": 5.340681319133711e-05, "loss": 0.0157, "step": 25220 }, { "grad_norm": 0.1636216640472412, "learning_rate": 5.337381990473305e-05, "loss": 0.0196, "step": 25230 }, { "grad_norm": 0.09867366403341293, "learning_rate": 5.3340825142305504e-05, "loss": 0.0151, "step": 25240 }, { "grad_norm": 0.060477420687675476, "learning_rate": 5.330782891848756e-05, "loss": 0.0137, "step": 25250 }, { "grad_norm": 0.09956171363592148, "learning_rate": 5.327483124771285e-05, "loss": 0.0128, "step": 25260 }, { "grad_norm": 0.09019958227872849, "learning_rate": 5.3241832144415706e-05, "loss": 0.0161, "step": 25270 }, { "grad_norm": 0.08612820506095886, "learning_rate": 5.320883162303104e-05, "loss": 0.018, "step": 25280 }, { "grad_norm": 0.08250681310892105, "learning_rate": 5.3175829697994416e-05, "loss": 0.0164, "step": 25290 }, { "grad_norm": 0.19092310965061188, "learning_rate": 5.3142826383741975e-05, "loss": 0.0164, "step": 25300 }, { "grad_norm": 0.12894824147224426, "learning_rate": 5.3109821694710506e-05, "loss": 0.0164, "step": 25310 }, { "grad_norm": 0.08645032346248627, "learning_rate": 5.307681564533736e-05, "loss": 0.0142, "step": 25320 }, { "grad_norm": 0.16074679791927338, "learning_rate": 5.3043808250060535e-05, "loss": 0.0196, "step": 25330 }, { "grad_norm": 0.11445294320583344, "learning_rate": 5.3010799523318566e-05, "loss": 0.0154, "step": 25340 }, { "grad_norm": 0.16600598394870758, "learning_rate": 5.29777894795506e-05, "loss": 0.0155, "step": 25350 }, { "grad_norm": 0.10733803361654282, "learning_rate": 5.294477813319634e-05, "loss": 0.0173, "step": 25360 }, { "grad_norm": 0.16075535118579865, "learning_rate": 5.291176549869608e-05, "loss": 0.0183, "step": 25370 }, { "grad_norm": 0.11033749580383301, "learning_rate": 5.287875159049065e-05, "loss": 0.0147, "step": 25380 }, { "grad_norm": 0.07651697099208832, "learning_rate": 5.284573642302149e-05, "loss": 0.0148, "step": 25390 }, { "grad_norm": 0.1408582180738449, "learning_rate": 5.2812720010730543e-05, "loss": 0.0143, "step": 25400 }, { "grad_norm": 0.09967602044343948, "learning_rate": 5.277970236806029e-05, "loss": 0.0157, "step": 25410 }, { "grad_norm": 0.10386479645967484, "learning_rate": 5.2746683509453795e-05, "loss": 0.0154, "step": 25420 }, { "grad_norm": 0.15249982476234436, "learning_rate": 5.271366344935461e-05, "loss": 0.0205, "step": 25430 }, { "grad_norm": 0.13885003328323364, "learning_rate": 5.268064220220683e-05, "loss": 0.0166, "step": 25440 }, { "grad_norm": 0.07207741588354111, "learning_rate": 5.2647619782455094e-05, "loss": 0.0178, "step": 25450 }, { "grad_norm": 0.09264855831861496, "learning_rate": 5.261459620454451e-05, "loss": 0.0189, "step": 25460 }, { "grad_norm": 0.1539379060268402, "learning_rate": 5.258157148292071e-05, "loss": 0.0194, "step": 25470 }, { "grad_norm": 0.1274363398551941, "learning_rate": 5.2548545632029846e-05, "loss": 0.0177, "step": 25480 }, { "grad_norm": 0.14051900804042816, "learning_rate": 5.2515518666318534e-05, "loss": 0.0186, "step": 25490 }, { "grad_norm": 0.1573651283979416, "learning_rate": 5.248249060023389e-05, "loss": 0.019, "step": 25500 }, { "grad_norm": 0.13427536189556122, "learning_rate": 5.2449461448223517e-05, "loss": 0.0188, "step": 25510 }, { "grad_norm": 0.1880405843257904, "learning_rate": 5.2416431224735486e-05, "loss": 0.0215, "step": 25520 }, { "grad_norm": 0.0976545512676239, "learning_rate": 5.238339994421836e-05, "loss": 0.0219, "step": 25530 }, { "grad_norm": 0.24114465713500977, "learning_rate": 5.235036762112111e-05, "loss": 0.0208, "step": 25540 }, { "grad_norm": 0.1647106111049652, "learning_rate": 5.2317334269893206e-05, "loss": 0.0212, "step": 25550 }, { "grad_norm": 0.21255625784397125, "learning_rate": 5.228429990498458e-05, "loss": 0.0204, "step": 25560 }, { "grad_norm": 0.11623160541057587, "learning_rate": 5.225126454084556e-05, "loss": 0.0202, "step": 25570 }, { "grad_norm": 0.08812768012285233, "learning_rate": 5.221822819192694e-05, "loss": 0.0193, "step": 25580 }, { "grad_norm": 0.1399112492799759, "learning_rate": 5.218519087267995e-05, "loss": 0.0202, "step": 25590 }, { "grad_norm": 0.1371431201696396, "learning_rate": 5.2152152597556225e-05, "loss": 0.0183, "step": 25600 }, { "grad_norm": 0.07517122477293015, "learning_rate": 5.211911338100784e-05, "loss": 0.0197, "step": 25610 }, { "grad_norm": 0.17191968858242035, "learning_rate": 5.2086073237487264e-05, "loss": 0.0225, "step": 25620 }, { "grad_norm": 0.1740117073059082, "learning_rate": 5.205303218144739e-05, "loss": 0.0207, "step": 25630 }, { "grad_norm": 0.10459133982658386, "learning_rate": 5.2019990227341494e-05, "loss": 0.0201, "step": 25640 }, { "grad_norm": 0.10865145176649094, "learning_rate": 5.198694738962324e-05, "loss": 0.0235, "step": 25650 }, { "grad_norm": 0.1515170782804489, "learning_rate": 5.19539036827467e-05, "loss": 0.0174, "step": 25660 }, { "grad_norm": 0.20165230333805084, "learning_rate": 5.192085912116629e-05, "loss": 0.0275, "step": 25670 }, { "grad_norm": 0.08894245326519012, "learning_rate": 5.188781371933685e-05, "loss": 0.0213, "step": 25680 }, { "grad_norm": 0.12243223190307617, "learning_rate": 5.185476749171356e-05, "loss": 0.0231, "step": 25690 }, { "grad_norm": 0.12609820067882538, "learning_rate": 5.1821720452751945e-05, "loss": 0.0232, "step": 25700 }, { "grad_norm": 0.20857878029346466, "learning_rate": 5.178867261690791e-05, "loss": 0.0205, "step": 25710 }, { "grad_norm": 0.10319285094738007, "learning_rate": 5.1755623998637715e-05, "loss": 0.0215, "step": 25720 }, { "grad_norm": 0.14995017647743225, "learning_rate": 5.172257461239794e-05, "loss": 0.0234, "step": 25730 }, { "grad_norm": 0.112650565803051, "learning_rate": 5.168952447264548e-05, "loss": 0.0199, "step": 25740 }, { "grad_norm": 0.1329740285873413, "learning_rate": 5.165647359383763e-05, "loss": 0.0195, "step": 25750 }, { "grad_norm": 0.10040687024593353, "learning_rate": 5.1623421990431954e-05, "loss": 0.0199, "step": 25760 }, { "grad_norm": 0.13551706075668335, "learning_rate": 5.159036967688635e-05, "loss": 0.0194, "step": 25770 }, { "grad_norm": 0.07786986976861954, "learning_rate": 5.1557316667659015e-05, "loss": 0.0174, "step": 25780 }, { "grad_norm": 0.1735474169254303, "learning_rate": 5.152426297720846e-05, "loss": 0.0207, "step": 25790 }, { "grad_norm": 0.14483658969402313, "learning_rate": 5.1491208619993505e-05, "loss": 0.019, "step": 25800 }, { "grad_norm": 0.15433892607688904, "learning_rate": 5.145815361047321e-05, "loss": 0.0186, "step": 25810 }, { "grad_norm": 0.19623243808746338, "learning_rate": 5.1425097963107003e-05, "loss": 0.0178, "step": 25820 }, { "grad_norm": 0.15814310312271118, "learning_rate": 5.1392041692354544e-05, "loss": 0.0191, "step": 25830 }, { "grad_norm": 0.1507692188024521, "learning_rate": 5.1358984812675736e-05, "loss": 0.0203, "step": 25840 }, { "grad_norm": 0.10276682674884796, "learning_rate": 5.13259273385308e-05, "loss": 0.0197, "step": 25850 }, { "grad_norm": 0.1483384519815445, "learning_rate": 5.12928692843802e-05, "loss": 0.0159, "step": 25860 }, { "grad_norm": 0.10287940502166748, "learning_rate": 5.125981066468465e-05, "loss": 0.0193, "step": 25870 }, { "grad_norm": 0.11778636276721954, "learning_rate": 5.122675149390514e-05, "loss": 0.0228, "step": 25880 }, { "grad_norm": 0.13406050205230713, "learning_rate": 5.119369178650282e-05, "loss": 0.0194, "step": 25890 }, { "grad_norm": 0.11332123726606369, "learning_rate": 5.1160631556939156e-05, "loss": 0.0169, "step": 25900 }, { "grad_norm": 0.11487305164337158, "learning_rate": 5.112757081967584e-05, "loss": 0.0192, "step": 25910 }, { "grad_norm": 0.24424797296524048, "learning_rate": 5.109450958917472e-05, "loss": 0.0202, "step": 25920 }, { "grad_norm": 0.21943232417106628, "learning_rate": 5.106144787989794e-05, "loss": 0.0192, "step": 25930 }, { "grad_norm": 0.10108345001935959, "learning_rate": 5.102838570630778e-05, "loss": 0.0165, "step": 25940 }, { "grad_norm": 0.11081822961568832, "learning_rate": 5.099532308286678e-05, "loss": 0.0192, "step": 25950 }, { "grad_norm": 0.09137255698442459, "learning_rate": 5.0962260024037645e-05, "loss": 0.0184, "step": 25960 }, { "grad_norm": 0.11011040955781937, "learning_rate": 5.0929196544283276e-05, "loss": 0.0171, "step": 25970 }, { "grad_norm": 0.2006826251745224, "learning_rate": 5.089613265806678e-05, "loss": 0.0171, "step": 25980 }, { "grad_norm": 0.14403210580348969, "learning_rate": 5.086306837985141e-05, "loss": 0.0189, "step": 25990 }, { "grad_norm": 0.1256532520055771, "learning_rate": 5.0830003724100604e-05, "loss": 0.019, "step": 26000 }, { "grad_norm": 0.19228395819664001, "learning_rate": 5.0796938705277984e-05, "loss": 0.0214, "step": 26010 }, { "grad_norm": 0.13092152774333954, "learning_rate": 5.076387333784729e-05, "loss": 0.0197, "step": 26020 }, { "grad_norm": 0.2096746861934662, "learning_rate": 5.0730807636272445e-05, "loss": 0.0195, "step": 26030 }, { "grad_norm": 0.11529078334569931, "learning_rate": 5.0697741615017524e-05, "loss": 0.0202, "step": 26040 }, { "grad_norm": 0.13136759400367737, "learning_rate": 5.06646752885467e-05, "loss": 0.021, "step": 26050 }, { "grad_norm": 0.13823741674423218, "learning_rate": 5.0631608671324324e-05, "loss": 0.0181, "step": 26060 }, { "grad_norm": 0.13540634512901306, "learning_rate": 5.059854177781487e-05, "loss": 0.0213, "step": 26070 }, { "grad_norm": 0.22409923374652863, "learning_rate": 5.05654746224829e-05, "loss": 0.0227, "step": 26080 }, { "grad_norm": 0.1428065150976181, "learning_rate": 5.053240721979311e-05, "loss": 0.0228, "step": 26090 }, { "grad_norm": 0.1973944902420044, "learning_rate": 5.049933958421031e-05, "loss": 0.024, "step": 26100 }, { "grad_norm": 0.14620168507099152, "learning_rate": 5.046627173019943e-05, "loss": 0.0195, "step": 26110 }, { "grad_norm": 0.13027647137641907, "learning_rate": 5.043320367222545e-05, "loss": 0.0204, "step": 26120 }, { "grad_norm": 0.09962894022464752, "learning_rate": 5.040013542475346e-05, "loss": 0.0199, "step": 26130 }, { "grad_norm": 0.17661280930042267, "learning_rate": 5.0367067002248646e-05, "loss": 0.0194, "step": 26140 }, { "grad_norm": 0.13169653713703156, "learning_rate": 5.0333998419176254e-05, "loss": 0.0213, "step": 26150 }, { "grad_norm": 0.1086677685379982, "learning_rate": 5.030092969000161e-05, "loss": 0.0173, "step": 26160 }, { "grad_norm": 0.14144711196422577, "learning_rate": 5.026786082919008e-05, "loss": 0.0178, "step": 26170 }, { "grad_norm": 0.13494686782360077, "learning_rate": 5.0234791851207144e-05, "loss": 0.0177, "step": 26180 }, { "grad_norm": 0.2538880407810211, "learning_rate": 5.0201722770518256e-05, "loss": 0.0188, "step": 26190 }, { "grad_norm": 0.12381644546985626, "learning_rate": 5.0168653601589e-05, "loss": 0.0193, "step": 26200 }, { "grad_norm": 0.13811957836151123, "learning_rate": 5.01355843588849e-05, "loss": 0.0199, "step": 26210 }, { "grad_norm": 0.1790999472141266, "learning_rate": 5.010251505687161e-05, "loss": 0.0207, "step": 26220 }, { "grad_norm": 0.10878618061542511, "learning_rate": 5.006944571001474e-05, "loss": 0.0189, "step": 26230 }, { "grad_norm": 0.09973053634166718, "learning_rate": 5.003637633277997e-05, "loss": 0.0184, "step": 26240 }, { "grad_norm": 0.10597692430019379, "learning_rate": 5.0003306939632946e-05, "loss": 0.0189, "step": 26250 }, { "grad_norm": 0.18174836039543152, "learning_rate": 4.997023754503937e-05, "loss": 0.0237, "step": 26260 }, { "grad_norm": 0.1607823222875595, "learning_rate": 4.9937168163464897e-05, "loss": 0.0178, "step": 26270 }, { "grad_norm": 0.0844133198261261, "learning_rate": 4.990409880937519e-05, "loss": 0.0206, "step": 26280 }, { "grad_norm": 0.11796817928552628, "learning_rate": 4.9871029497235936e-05, "loss": 0.0227, "step": 26290 }, { "grad_norm": 0.16243799030780792, "learning_rate": 4.9837960241512766e-05, "loss": 0.0215, "step": 26300 }, { "grad_norm": 0.16505204141139984, "learning_rate": 4.9804891056671265e-05, "loss": 0.0243, "step": 26310 }, { "grad_norm": 0.14863210916519165, "learning_rate": 4.9771821957177074e-05, "loss": 0.0172, "step": 26320 }, { "grad_norm": 0.10342040657997131, "learning_rate": 4.9738752957495694e-05, "loss": 0.0197, "step": 26330 }, { "grad_norm": 0.21484734117984772, "learning_rate": 4.970568407209265e-05, "loss": 0.0178, "step": 26340 }, { "grad_norm": 0.1425163298845291, "learning_rate": 4.967261531543339e-05, "loss": 0.0181, "step": 26350 }, { "grad_norm": 0.10818415880203247, "learning_rate": 4.963954670198332e-05, "loss": 0.0215, "step": 26360 }, { "grad_norm": 0.1837388277053833, "learning_rate": 4.960647824620776e-05, "loss": 0.0174, "step": 26370 }, { "grad_norm": 0.1388355791568756, "learning_rate": 4.9573409962571995e-05, "loss": 0.0182, "step": 26380 }, { "grad_norm": 0.08592391014099121, "learning_rate": 4.954034186554121e-05, "loss": 0.0208, "step": 26390 }, { "grad_norm": 0.14612723886966705, "learning_rate": 4.950727396958051e-05, "loss": 0.0187, "step": 26400 }, { "grad_norm": 0.07593711465597153, "learning_rate": 4.947420628915489e-05, "loss": 0.0188, "step": 26410 }, { "grad_norm": 0.09109128266572952, "learning_rate": 4.944113883872932e-05, "loss": 0.0166, "step": 26420 }, { "grad_norm": 0.10305337607860565, "learning_rate": 4.9408071632768624e-05, "loss": 0.0199, "step": 26430 }, { "grad_norm": 0.11144543439149857, "learning_rate": 4.937500468573747e-05, "loss": 0.0209, "step": 26440 }, { "grad_norm": 0.09772215783596039, "learning_rate": 4.934193801210053e-05, "loss": 0.0194, "step": 26450 }, { "grad_norm": 0.14301171898841858, "learning_rate": 4.930887162632224e-05, "loss": 0.0199, "step": 26460 }, { "grad_norm": 0.10870107263326645, "learning_rate": 4.9275805542866994e-05, "loss": 0.0174, "step": 26470 }, { "grad_norm": 0.09711555391550064, "learning_rate": 4.9242739776199006e-05, "loss": 0.0162, "step": 26480 }, { "grad_norm": 0.16245613992214203, "learning_rate": 4.920967434078237e-05, "loss": 0.0198, "step": 26490 }, { "grad_norm": 0.1414850801229477, "learning_rate": 4.917660925108101e-05, "loss": 0.0207, "step": 26500 }, { "grad_norm": 0.12310337275266647, "learning_rate": 4.914354452155876e-05, "loss": 0.0216, "step": 26510 }, { "grad_norm": 0.17822641134262085, "learning_rate": 4.911048016667923e-05, "loss": 0.0155, "step": 26520 }, { "grad_norm": 0.13800691068172455, "learning_rate": 4.907741620090588e-05, "loss": 0.0184, "step": 26530 }, { "grad_norm": 0.16765223443508148, "learning_rate": 4.9044352638702046e-05, "loss": 0.0189, "step": 26540 }, { "grad_norm": 0.16433027386665344, "learning_rate": 4.9011289494530834e-05, "loss": 0.0173, "step": 26550 }, { "grad_norm": 0.12878847122192383, "learning_rate": 4.897822678285516e-05, "loss": 0.0192, "step": 26560 }, { "grad_norm": 0.08952873945236206, "learning_rate": 4.8945164518137835e-05, "loss": 0.0182, "step": 26570 }, { "grad_norm": 0.10211741924285889, "learning_rate": 4.891210271484138e-05, "loss": 0.0174, "step": 26580 }, { "grad_norm": 0.16652044653892517, "learning_rate": 4.8879041387428164e-05, "loss": 0.018, "step": 26590 }, { "grad_norm": 0.15215663611888885, "learning_rate": 4.8845980550360306e-05, "loss": 0.0193, "step": 26600 }, { "grad_norm": 0.16613085567951202, "learning_rate": 4.8812920218099786e-05, "loss": 0.0192, "step": 26610 }, { "grad_norm": 0.11897451430559158, "learning_rate": 4.8779860405108294e-05, "loss": 0.0192, "step": 26620 }, { "grad_norm": 0.15935321152210236, "learning_rate": 4.8746801125847285e-05, "loss": 0.0195, "step": 26630 }, { "grad_norm": 0.1448453962802887, "learning_rate": 4.8713742394778064e-05, "loss": 0.0185, "step": 26640 }, { "grad_norm": 0.08308782428503036, "learning_rate": 4.868068422636162e-05, "loss": 0.016, "step": 26650 }, { "grad_norm": 0.10476414859294891, "learning_rate": 4.86476266350587e-05, "loss": 0.0214, "step": 26660 }, { "grad_norm": 0.13787998259067535, "learning_rate": 4.861456963532985e-05, "loss": 0.0182, "step": 26670 }, { "grad_norm": 0.1410088837146759, "learning_rate": 4.8581513241635316e-05, "loss": 0.0176, "step": 26680 }, { "grad_norm": 0.10322398692369461, "learning_rate": 4.8548457468435065e-05, "loss": 0.0181, "step": 26690 }, { "grad_norm": 0.18767961859703064, "learning_rate": 4.851540233018884e-05, "loss": 0.0241, "step": 26700 }, { "grad_norm": 0.20954491198062897, "learning_rate": 4.848234784135608e-05, "loss": 0.0193, "step": 26710 }, { "grad_norm": 0.0700482651591301, "learning_rate": 4.8449294016395916e-05, "loss": 0.0181, "step": 26720 }, { "grad_norm": 0.13897287845611572, "learning_rate": 4.8416240869767246e-05, "loss": 0.0175, "step": 26730 }, { "grad_norm": 0.09216828644275665, "learning_rate": 4.838318841592864e-05, "loss": 0.018, "step": 26740 }, { "grad_norm": 0.16107209026813507, "learning_rate": 4.835013666933834e-05, "loss": 0.0232, "step": 26750 }, { "grad_norm": 0.1776239275932312, "learning_rate": 4.83170856444543e-05, "loss": 0.0193, "step": 26760 }, { "grad_norm": 0.15035825967788696, "learning_rate": 4.828403535573419e-05, "loss": 0.0177, "step": 26770 }, { "grad_norm": 0.12364953011274338, "learning_rate": 4.825098581763532e-05, "loss": 0.0171, "step": 26780 }, { "grad_norm": 0.11100003123283386, "learning_rate": 4.821793704461465e-05, "loss": 0.0212, "step": 26790 }, { "grad_norm": 0.10817791521549225, "learning_rate": 4.8184889051128884e-05, "loss": 0.019, "step": 26800 }, { "grad_norm": 0.16679084300994873, "learning_rate": 4.81518418516343e-05, "loss": 0.0188, "step": 26810 }, { "grad_norm": 0.1596277356147766, "learning_rate": 4.811879546058688e-05, "loss": 0.019, "step": 26820 }, { "grad_norm": 0.10004975646734238, "learning_rate": 4.808574989244224e-05, "loss": 0.0176, "step": 26830 }, { "grad_norm": 0.17470324039459229, "learning_rate": 4.8052705161655644e-05, "loss": 0.0207, "step": 26840 }, { "grad_norm": 0.18241150677204132, "learning_rate": 4.8019661282681936e-05, "loss": 0.0181, "step": 26850 }, { "grad_norm": 0.1102629005908966, "learning_rate": 4.79866182699757e-05, "loss": 0.0161, "step": 26860 }, { "grad_norm": 0.09830936044454575, "learning_rate": 4.795357613799103e-05, "loss": 0.0164, "step": 26870 }, { "grad_norm": 0.10913507640361786, "learning_rate": 4.792053490118166e-05, "loss": 0.0199, "step": 26880 }, { "grad_norm": 0.10473732650279999, "learning_rate": 4.788749457400099e-05, "loss": 0.016, "step": 26890 }, { "grad_norm": 0.08517294377088547, "learning_rate": 4.7854455170901966e-05, "loss": 0.0183, "step": 26900 }, { "grad_norm": 0.0871393084526062, "learning_rate": 4.7821416706337136e-05, "loss": 0.0193, "step": 26910 }, { "grad_norm": 0.23308482766151428, "learning_rate": 4.7788379194758645e-05, "loss": 0.0211, "step": 26920 }, { "grad_norm": 0.1539209485054016, "learning_rate": 4.7755342650618246e-05, "loss": 0.0162, "step": 26930 }, { "grad_norm": 0.21642053127288818, "learning_rate": 4.7722307088367235e-05, "loss": 0.0237, "step": 26940 }, { "grad_norm": 0.13216006755828857, "learning_rate": 4.768927252245646e-05, "loss": 0.0205, "step": 26950 }, { "grad_norm": 0.12247534841299057, "learning_rate": 4.765623896733642e-05, "loss": 0.0157, "step": 26960 }, { "grad_norm": 0.1369602531194687, "learning_rate": 4.762320643745708e-05, "loss": 0.0195, "step": 26970 }, { "grad_norm": 0.10881401598453522, "learning_rate": 4.7590174947267966e-05, "loss": 0.0159, "step": 26980 }, { "grad_norm": 0.141176238656044, "learning_rate": 4.755714451121823e-05, "loss": 0.0168, "step": 26990 }, { "grad_norm": 0.17925308644771576, "learning_rate": 4.7524115143756476e-05, "loss": 0.0154, "step": 27000 }, { "grad_norm": 0.11022309213876724, "learning_rate": 4.7491086859330854e-05, "loss": 0.0149, "step": 27010 }, { "grad_norm": 0.18371307849884033, "learning_rate": 4.74580596723891e-05, "loss": 0.0174, "step": 27020 }, { "grad_norm": 0.09356648474931717, "learning_rate": 4.742503359737841e-05, "loss": 0.0155, "step": 27030 }, { "grad_norm": 0.06679507344961166, "learning_rate": 4.739200864874547e-05, "loss": 0.0194, "step": 27040 }, { "grad_norm": 0.11878959834575653, "learning_rate": 4.735898484093658e-05, "loss": 0.0126, "step": 27050 }, { "grad_norm": 0.07777684181928635, "learning_rate": 4.7325962188397444e-05, "loss": 0.016, "step": 27060 }, { "grad_norm": 0.09014764428138733, "learning_rate": 4.729294070557329e-05, "loss": 0.016, "step": 27070 }, { "grad_norm": 0.08932525664567947, "learning_rate": 4.7259920406908833e-05, "loss": 0.0172, "step": 27080 }, { "grad_norm": 0.08869722485542297, "learning_rate": 4.7226901306848306e-05, "loss": 0.0139, "step": 27090 }, { "grad_norm": 0.06909776479005814, "learning_rate": 4.719388341983535e-05, "loss": 0.0151, "step": 27100 }, { "grad_norm": 0.16876688599586487, "learning_rate": 4.71608667603131e-05, "loss": 0.0152, "step": 27110 }, { "grad_norm": 0.1278250366449356, "learning_rate": 4.712785134272422e-05, "loss": 0.0175, "step": 27120 }, { "grad_norm": 0.09927810728549957, "learning_rate": 4.7094837181510734e-05, "loss": 0.0138, "step": 27130 }, { "grad_norm": 0.1355901062488556, "learning_rate": 4.706182429111414e-05, "loss": 0.0194, "step": 27140 }, { "grad_norm": 0.15220969915390015, "learning_rate": 4.702881268597545e-05, "loss": 0.0145, "step": 27150 }, { "grad_norm": 0.1939060539007187, "learning_rate": 4.699580238053503e-05, "loss": 0.0166, "step": 27160 }, { "grad_norm": 0.09744827449321747, "learning_rate": 4.69627933892327e-05, "loss": 0.0191, "step": 27170 }, { "grad_norm": 0.3422001302242279, "learning_rate": 4.692978572650773e-05, "loss": 0.019, "step": 27180 }, { "grad_norm": 0.15328210592269897, "learning_rate": 4.689677940679881e-05, "loss": 0.0172, "step": 27190 }, { "grad_norm": 0.13382789492607117, "learning_rate": 4.6863774444543984e-05, "loss": 0.0214, "step": 27200 }, { "grad_norm": 0.13293305039405823, "learning_rate": 4.6830770854180784e-05, "loss": 0.0156, "step": 27210 }, { "grad_norm": 0.11228286474943161, "learning_rate": 4.679776865014608e-05, "loss": 0.0156, "step": 27220 }, { "grad_norm": 0.16229510307312012, "learning_rate": 4.676476784687615e-05, "loss": 0.0145, "step": 27230 }, { "grad_norm": 0.1295761913061142, "learning_rate": 4.673176845880669e-05, "loss": 0.0147, "step": 27240 }, { "grad_norm": 0.21615873277187347, "learning_rate": 4.669877050037275e-05, "loss": 0.0161, "step": 27250 }, { "grad_norm": 0.08753505349159241, "learning_rate": 4.666577398600874e-05, "loss": 0.0152, "step": 27260 }, { "grad_norm": 0.09058281034231186, "learning_rate": 4.663277893014846e-05, "loss": 0.0132, "step": 27270 }, { "grad_norm": 0.10760592669248581, "learning_rate": 4.659978534722508e-05, "loss": 0.0167, "step": 27280 }, { "grad_norm": 0.10977942496538162, "learning_rate": 4.6566793251671104e-05, "loss": 0.0141, "step": 27290 }, { "grad_norm": 0.1220499724149704, "learning_rate": 4.65338026579184e-05, "loss": 0.0135, "step": 27300 }, { "grad_norm": 0.16670426726341248, "learning_rate": 4.650081358039817e-05, "loss": 0.0145, "step": 27310 }, { "grad_norm": 0.11875534057617188, "learning_rate": 4.646782603354098e-05, "loss": 0.0154, "step": 27320 }, { "grad_norm": 0.07876678556203842, "learning_rate": 4.643484003177666e-05, "loss": 0.0125, "step": 27330 }, { "grad_norm": 0.13179709017276764, "learning_rate": 4.6401855589534454e-05, "loss": 0.0144, "step": 27340 }, { "grad_norm": 0.183803528547287, "learning_rate": 4.636887272124286e-05, "loss": 0.0186, "step": 27350 }, { "grad_norm": 0.2558690011501312, "learning_rate": 4.6335891441329685e-05, "loss": 0.0175, "step": 27360 }, { "grad_norm": 0.11514898389577866, "learning_rate": 4.630291176422209e-05, "loss": 0.0141, "step": 27370 }, { "grad_norm": 0.1254771500825882, "learning_rate": 4.62699337043465e-05, "loss": 0.0154, "step": 27380 }, { "grad_norm": 0.12181437760591507, "learning_rate": 4.623695727612861e-05, "loss": 0.0143, "step": 27390 }, { "grad_norm": 0.09469976276159286, "learning_rate": 4.620398249399347e-05, "loss": 0.012, "step": 27400 }, { "grad_norm": 0.07761237025260925, "learning_rate": 4.617100937236535e-05, "loss": 0.0178, "step": 27410 }, { "grad_norm": 0.1353186070919037, "learning_rate": 4.613803792566781e-05, "loss": 0.0137, "step": 27420 }, { "grad_norm": 0.16341538727283478, "learning_rate": 4.61050681683237e-05, "loss": 0.0152, "step": 27430 }, { "grad_norm": 0.2227846086025238, "learning_rate": 4.6072100114755104e-05, "loss": 0.0171, "step": 27440 }, { "grad_norm": 0.1269064098596573, "learning_rate": 4.603913377938336e-05, "loss": 0.0161, "step": 27450 }, { "grad_norm": 0.1573556661605835, "learning_rate": 4.600616917662905e-05, "loss": 0.0156, "step": 27460 }, { "grad_norm": 0.16158980131149292, "learning_rate": 4.597320632091206e-05, "loss": 0.0181, "step": 27470 }, { "grad_norm": 0.20515380799770355, "learning_rate": 4.594024522665142e-05, "loss": 0.0168, "step": 27480 }, { "grad_norm": 0.1455192118883133, "learning_rate": 4.590728590826543e-05, "loss": 0.0146, "step": 27490 }, { "grad_norm": 0.1837523728609085, "learning_rate": 4.587432838017165e-05, "loss": 0.0192, "step": 27500 }, { "grad_norm": 0.25204119086265564, "learning_rate": 4.58413726567868e-05, "loss": 0.0185, "step": 27510 }, { "grad_norm": 0.14135578274726868, "learning_rate": 4.5808418752526814e-05, "loss": 0.0176, "step": 27520 }, { "grad_norm": 0.1017589196562767, "learning_rate": 4.5775466681806895e-05, "loss": 0.0171, "step": 27530 }, { "grad_norm": 0.08415919542312622, "learning_rate": 4.574251645904136e-05, "loss": 0.0164, "step": 27540 }, { "grad_norm": 0.14195770025253296, "learning_rate": 4.5709568098643776e-05, "loss": 0.0177, "step": 27550 }, { "grad_norm": 0.09311961382627487, "learning_rate": 4.567662161502687e-05, "loss": 0.0142, "step": 27560 }, { "grad_norm": 0.1615968495607376, "learning_rate": 4.5643677022602554e-05, "loss": 0.0169, "step": 27570 }, { "grad_norm": 0.10200619697570801, "learning_rate": 4.56107343357819e-05, "loss": 0.0157, "step": 27580 }, { "grad_norm": 0.11413545161485672, "learning_rate": 4.557779356897518e-05, "loss": 0.0144, "step": 27590 }, { "grad_norm": 0.22890160977840424, "learning_rate": 4.55448547365918e-05, "loss": 0.0164, "step": 27600 }, { "grad_norm": 0.23528869450092316, "learning_rate": 4.5511917853040315e-05, "loss": 0.0192, "step": 27610 }, { "grad_norm": 0.1316303163766861, "learning_rate": 4.547898293272842e-05, "loss": 0.0165, "step": 27620 }, { "grad_norm": 0.10665438324213028, "learning_rate": 4.544604999006301e-05, "loss": 0.0141, "step": 27630 }, { "grad_norm": 0.10205581784248352, "learning_rate": 4.541311903945004e-05, "loss": 0.0162, "step": 27640 }, { "grad_norm": 0.13362255692481995, "learning_rate": 4.538019009529463e-05, "loss": 0.0187, "step": 27650 }, { "grad_norm": 0.1636674404144287, "learning_rate": 4.5347263172001035e-05, "loss": 0.018, "step": 27660 }, { "grad_norm": 0.13753513991832733, "learning_rate": 4.531433828397261e-05, "loss": 0.0195, "step": 27670 }, { "grad_norm": 0.17160463333129883, "learning_rate": 4.528141544561179e-05, "loss": 0.0184, "step": 27680 }, { "grad_norm": 0.13083188235759735, "learning_rate": 4.5248494671320186e-05, "loss": 0.0134, "step": 27690 }, { "grad_norm": 0.12698732316493988, "learning_rate": 4.521557597549844e-05, "loss": 0.019, "step": 27700 }, { "grad_norm": 0.14578311145305634, "learning_rate": 4.5182659372546284e-05, "loss": 0.0128, "step": 27710 }, { "grad_norm": 0.05948005989193916, "learning_rate": 4.514974487686261e-05, "loss": 0.014, "step": 27720 }, { "grad_norm": 0.10094869881868362, "learning_rate": 4.5116832502845306e-05, "loss": 0.0141, "step": 27730 }, { "grad_norm": 0.08500009775161743, "learning_rate": 4.508392226489136e-05, "loss": 0.0141, "step": 27740 }, { "grad_norm": 0.15060745179653168, "learning_rate": 4.505101417739683e-05, "loss": 0.0147, "step": 27750 }, { "grad_norm": 0.10905767232179642, "learning_rate": 4.501810825475685e-05, "loss": 0.0155, "step": 27760 }, { "grad_norm": 0.1069747656583786, "learning_rate": 4.4985204511365556e-05, "loss": 0.0189, "step": 27770 }, { "grad_norm": 0.08511964976787567, "learning_rate": 4.49523029616162e-05, "loss": 0.0148, "step": 27780 }, { "grad_norm": 0.10559134185314178, "learning_rate": 4.491940361990101e-05, "loss": 0.0175, "step": 27790 }, { "grad_norm": 0.172398179769516, "learning_rate": 4.48865065006113e-05, "loss": 0.0151, "step": 27800 }, { "grad_norm": 0.1732216328382492, "learning_rate": 4.485361161813735e-05, "loss": 0.0165, "step": 27810 }, { "grad_norm": 0.1490965634584427, "learning_rate": 4.482071898686854e-05, "loss": 0.0148, "step": 27820 }, { "grad_norm": 0.13595439493656158, "learning_rate": 4.478782862119321e-05, "loss": 0.0162, "step": 27830 }, { "grad_norm": 0.10674428939819336, "learning_rate": 4.475494053549869e-05, "loss": 0.0177, "step": 27840 }, { "grad_norm": 0.10849172621965408, "learning_rate": 4.4722054744171405e-05, "loss": 0.014, "step": 27850 }, { "grad_norm": 0.08953342586755753, "learning_rate": 4.468917126159669e-05, "loss": 0.0149, "step": 27860 }, { "grad_norm": 0.08185054361820221, "learning_rate": 4.465629010215887e-05, "loss": 0.0126, "step": 27870 }, { "grad_norm": 0.20273356139659882, "learning_rate": 4.462341128024133e-05, "loss": 0.0329, "step": 27880 }, { "grad_norm": 0.09823652356863022, "learning_rate": 4.4590534810226356e-05, "loss": 0.0161, "step": 27890 }, { "grad_norm": 0.32951241731643677, "learning_rate": 4.455766070649523e-05, "loss": 0.0164, "step": 27900 }, { "grad_norm": 0.1671943962574005, "learning_rate": 4.4524788983428215e-05, "loss": 0.0162, "step": 27910 }, { "grad_norm": 0.12842509150505066, "learning_rate": 4.4491919655404515e-05, "loss": 0.0185, "step": 27920 }, { "grad_norm": 0.19343827664852142, "learning_rate": 4.445905273680228e-05, "loss": 0.0177, "step": 27930 }, { "grad_norm": 0.15876595675945282, "learning_rate": 4.442618824199865e-05, "loss": 0.019, "step": 27940 }, { "grad_norm": 0.18667052686214447, "learning_rate": 4.4393326185369635e-05, "loss": 0.0188, "step": 27950 }, { "grad_norm": 0.14317305386066437, "learning_rate": 4.436046658129024e-05, "loss": 0.0201, "step": 27960 }, { "grad_norm": 0.09363661706447601, "learning_rate": 4.432760944413433e-05, "loss": 0.0164, "step": 27970 }, { "grad_norm": 0.11733818054199219, "learning_rate": 4.4294754788274796e-05, "loss": 0.015, "step": 27980 }, { "grad_norm": 0.09818368405103683, "learning_rate": 4.426190262808334e-05, "loss": 0.0175, "step": 27990 }, { "grad_norm": 0.1483861654996872, "learning_rate": 4.42290529779306e-05, "loss": 0.021, "step": 28000 }, { "grad_norm": 0.10625140368938446, "learning_rate": 4.4196205852186176e-05, "loss": 0.0152, "step": 28010 }, { "grad_norm": 0.2157064527273178, "learning_rate": 4.416336126521848e-05, "loss": 0.0196, "step": 28020 }, { "grad_norm": 0.1008097305893898, "learning_rate": 4.4130519231394865e-05, "loss": 0.0168, "step": 28030 }, { "grad_norm": 0.0858251228928566, "learning_rate": 4.409767976508154e-05, "loss": 0.0161, "step": 28040 }, { "grad_norm": 0.12254098802804947, "learning_rate": 4.406484288064363e-05, "loss": 0.0174, "step": 28050 }, { "grad_norm": 0.17378363013267517, "learning_rate": 4.4032008592445055e-05, "loss": 0.0195, "step": 28060 }, { "grad_norm": 0.07331738620996475, "learning_rate": 4.3999176914848714e-05, "loss": 0.0176, "step": 28070 }, { "grad_norm": 0.09684018790721893, "learning_rate": 4.396634786221625e-05, "loss": 0.0152, "step": 28080 }, { "grad_norm": 0.10203785449266434, "learning_rate": 4.3933521448908203e-05, "loss": 0.0162, "step": 28090 }, { "grad_norm": 0.09676770865917206, "learning_rate": 4.3900697689284e-05, "loss": 0.0164, "step": 28100 }, { "grad_norm": 0.09036992490291595, "learning_rate": 4.386787659770185e-05, "loss": 0.0168, "step": 28110 }, { "grad_norm": 0.15003344416618347, "learning_rate": 4.383505818851878e-05, "loss": 0.0175, "step": 28120 }, { "grad_norm": 0.10756135731935501, "learning_rate": 4.380224247609073e-05, "loss": 0.0164, "step": 28130 }, { "grad_norm": 0.0960574671626091, "learning_rate": 4.3769429474772376e-05, "loss": 0.0158, "step": 28140 }, { "grad_norm": 0.22427961230278015, "learning_rate": 4.373661919891726e-05, "loss": 0.0204, "step": 28150 }, { "grad_norm": 0.1345623880624771, "learning_rate": 4.3703811662877674e-05, "loss": 0.0184, "step": 28160 }, { "grad_norm": 0.1620190441608429, "learning_rate": 4.36710068810048e-05, "loss": 0.0153, "step": 28170 }, { "grad_norm": 0.12581820785999298, "learning_rate": 4.363820486764853e-05, "loss": 0.0216, "step": 28180 }, { "grad_norm": 0.13581012189388275, "learning_rate": 4.360540563715758e-05, "loss": 0.02, "step": 28190 }, { "grad_norm": 0.16339106857776642, "learning_rate": 4.357260920387946e-05, "loss": 0.021, "step": 28200 }, { "grad_norm": 0.14750441908836365, "learning_rate": 4.353981558216045e-05, "loss": 0.0175, "step": 28210 }, { "grad_norm": 0.18277713656425476, "learning_rate": 4.3507024786345565e-05, "loss": 0.0159, "step": 28220 }, { "grad_norm": 0.1337823122739792, "learning_rate": 4.3474236830778646e-05, "loss": 0.0163, "step": 28230 }, { "grad_norm": 0.11990299820899963, "learning_rate": 4.344145172980223e-05, "loss": 0.0158, "step": 28240 }, { "grad_norm": 0.17158536612987518, "learning_rate": 4.3408669497757656e-05, "loss": 0.0205, "step": 28250 }, { "grad_norm": 0.14245294034481049, "learning_rate": 4.3375890148984976e-05, "loss": 0.0198, "step": 28260 }, { "grad_norm": 0.13005539774894714, "learning_rate": 4.334311369782301e-05, "loss": 0.015, "step": 28270 }, { "grad_norm": 0.09734495729207993, "learning_rate": 4.3310340158609244e-05, "loss": 0.0179, "step": 28280 }, { "grad_norm": 0.08828350156545639, "learning_rate": 4.327756954568e-05, "loss": 0.0139, "step": 28290 }, { "grad_norm": 0.07992541790008545, "learning_rate": 4.324480187337021e-05, "loss": 0.016, "step": 28300 }, { "grad_norm": 0.0793006494641304, "learning_rate": 4.3212037156013596e-05, "loss": 0.0172, "step": 28310 }, { "grad_norm": 0.08820235729217529, "learning_rate": 4.317927540794251e-05, "loss": 0.0143, "step": 28320 }, { "grad_norm": 0.10191942006349564, "learning_rate": 4.314651664348812e-05, "loss": 0.0151, "step": 28330 }, { "grad_norm": 0.05860336869955063, "learning_rate": 4.3113760876980195e-05, "loss": 0.0151, "step": 28340 }, { "grad_norm": 0.09814650565385818, "learning_rate": 4.308100812274719e-05, "loss": 0.0134, "step": 28350 }, { "grad_norm": 0.06594283133745193, "learning_rate": 4.304825839511632e-05, "loss": 0.0144, "step": 28360 }, { "grad_norm": 0.0785687118768692, "learning_rate": 4.3015511708413395e-05, "loss": 0.0153, "step": 28370 }, { "grad_norm": 0.14706523716449738, "learning_rate": 4.298276807696294e-05, "loss": 0.0157, "step": 28380 }, { "grad_norm": 0.13978008925914764, "learning_rate": 4.2950027515088136e-05, "loss": 0.0167, "step": 28390 }, { "grad_norm": 0.17818781733512878, "learning_rate": 4.291729003711083e-05, "loss": 0.0138, "step": 28400 }, { "grad_norm": 0.18675261735916138, "learning_rate": 4.2884555657351455e-05, "loss": 0.0153, "step": 28410 }, { "grad_norm": 0.13967323303222656, "learning_rate": 4.285182439012921e-05, "loss": 0.0169, "step": 28420 }, { "grad_norm": 0.20402269065380096, "learning_rate": 4.281909624976183e-05, "loss": 0.0151, "step": 28430 }, { "grad_norm": 0.1323353797197342, "learning_rate": 4.278637125056568e-05, "loss": 0.015, "step": 28440 }, { "grad_norm": 0.1380321979522705, "learning_rate": 4.2753649406855855e-05, "loss": 0.0149, "step": 28450 }, { "grad_norm": 0.19380758702754974, "learning_rate": 4.2720930732945966e-05, "loss": 0.018, "step": 28460 }, { "grad_norm": 0.09920495003461838, "learning_rate": 4.268821524314826e-05, "loss": 0.0156, "step": 28470 }, { "grad_norm": 0.18293380737304688, "learning_rate": 4.2655502951773623e-05, "loss": 0.0161, "step": 28480 }, { "grad_norm": 0.08246969431638718, "learning_rate": 4.262279387313152e-05, "loss": 0.0138, "step": 28490 }, { "grad_norm": 0.10635332018136978, "learning_rate": 4.259008802153001e-05, "loss": 0.0154, "step": 28500 }, { "grad_norm": 0.12403637170791626, "learning_rate": 4.2557385411275735e-05, "loss": 0.0156, "step": 28510 }, { "grad_norm": 0.09564485400915146, "learning_rate": 4.252468605667395e-05, "loss": 0.0163, "step": 28520 }, { "grad_norm": 0.1406458020210266, "learning_rate": 4.2491989972028456e-05, "loss": 0.0171, "step": 28530 }, { "grad_norm": 0.19423706829547882, "learning_rate": 4.245929717164161e-05, "loss": 0.0138, "step": 28540 }, { "grad_norm": 0.09453561156988144, "learning_rate": 4.242660766981438e-05, "loss": 0.0168, "step": 28550 }, { "grad_norm": 0.09043316543102264, "learning_rate": 4.2393921480846264e-05, "loss": 0.0183, "step": 28560 }, { "grad_norm": 0.10353461652994156, "learning_rate": 4.236123861903528e-05, "loss": 0.0184, "step": 28570 }, { "grad_norm": 0.07609079778194427, "learning_rate": 4.232855909867806e-05, "loss": 0.0168, "step": 28580 }, { "grad_norm": 0.10980664193630219, "learning_rate": 4.229588293406972e-05, "loss": 0.0169, "step": 28590 }, { "grad_norm": 0.12429125607013702, "learning_rate": 4.226321013950392e-05, "loss": 0.0152, "step": 28600 }, { "grad_norm": 0.09127883613109589, "learning_rate": 4.223054072927286e-05, "loss": 0.0218, "step": 28610 }, { "grad_norm": 0.11414151638746262, "learning_rate": 4.219787471766724e-05, "loss": 0.0178, "step": 28620 }, { "grad_norm": 0.14721769094467163, "learning_rate": 4.216521211897629e-05, "loss": 0.0166, "step": 28630 }, { "grad_norm": 0.10063809156417847, "learning_rate": 4.2132552947487734e-05, "loss": 0.0177, "step": 28640 }, { "grad_norm": 0.07522519677877426, "learning_rate": 4.209989721748783e-05, "loss": 0.0185, "step": 28650 }, { "grad_norm": 0.149352565407753, "learning_rate": 4.206724494326127e-05, "loss": 0.018, "step": 28660 }, { "grad_norm": 0.14373566210269928, "learning_rate": 4.203459613909126e-05, "loss": 0.0146, "step": 28670 }, { "grad_norm": 0.15822570025920868, "learning_rate": 4.200195081925955e-05, "loss": 0.0167, "step": 28680 }, { "grad_norm": 0.10240768641233444, "learning_rate": 4.196930899804627e-05, "loss": 0.0175, "step": 28690 }, { "grad_norm": 0.16040590405464172, "learning_rate": 4.1936670689730044e-05, "loss": 0.0173, "step": 28700 }, { "grad_norm": 0.0791720524430275, "learning_rate": 4.190403590858804e-05, "loss": 0.0157, "step": 28710 }, { "grad_norm": 0.12665113806724548, "learning_rate": 4.187140466889575e-05, "loss": 0.0172, "step": 28720 }, { "grad_norm": 0.12839318811893463, "learning_rate": 4.183877698492724e-05, "loss": 0.0225, "step": 28730 }, { "grad_norm": 0.12817995250225067, "learning_rate": 4.1806152870954935e-05, "loss": 0.0152, "step": 28740 }, { "grad_norm": 0.2131527066230774, "learning_rate": 4.177353234124976e-05, "loss": 0.0184, "step": 28750 }, { "grad_norm": 0.0911240428686142, "learning_rate": 4.1740915410081e-05, "loss": 0.0206, "step": 28760 }, { "grad_norm": 0.13821116089820862, "learning_rate": 4.170830209171646e-05, "loss": 0.0163, "step": 28770 }, { "grad_norm": 0.09605841338634491, "learning_rate": 4.1675692400422286e-05, "loss": 0.0168, "step": 28780 }, { "grad_norm": 0.1022643893957138, "learning_rate": 4.164308635046304e-05, "loss": 0.0211, "step": 28790 }, { "grad_norm": 0.09055076539516449, "learning_rate": 4.161048395610177e-05, "loss": 0.0157, "step": 28800 }, { "grad_norm": 0.07750216871500015, "learning_rate": 4.157788523159985e-05, "loss": 0.0142, "step": 28810 }, { "grad_norm": 0.09591007977724075, "learning_rate": 4.154529019121706e-05, "loss": 0.0178, "step": 28820 }, { "grad_norm": 0.08225135505199432, "learning_rate": 4.1512698849211564e-05, "loss": 0.0183, "step": 28830 }, { "grad_norm": 0.12533102929592133, "learning_rate": 4.1480111219839954e-05, "loss": 0.016, "step": 28840 }, { "grad_norm": 0.31087374687194824, "learning_rate": 4.144752731735714e-05, "loss": 0.0205, "step": 28850 }, { "grad_norm": 0.0966780036687851, "learning_rate": 4.1414947156016446e-05, "loss": 0.0185, "step": 28860 }, { "grad_norm": 0.10382996499538422, "learning_rate": 4.138237075006953e-05, "loss": 0.0184, "step": 28870 }, { "grad_norm": 0.18394234776496887, "learning_rate": 4.134979811376645e-05, "loss": 0.0214, "step": 28880 }, { "grad_norm": 0.12377340346574783, "learning_rate": 4.131722926135552e-05, "loss": 0.0176, "step": 28890 }, { "grad_norm": 0.09386216849088669, "learning_rate": 4.1284664207083526e-05, "loss": 0.0177, "step": 28900 }, { "grad_norm": 0.09689618647098541, "learning_rate": 4.125210296519551e-05, "loss": 0.0191, "step": 28910 }, { "grad_norm": 0.16184236109256744, "learning_rate": 4.121954554993484e-05, "loss": 0.0163, "step": 28920 }, { "grad_norm": 0.09769776463508606, "learning_rate": 4.118699197554327e-05, "loss": 0.017, "step": 28930 }, { "grad_norm": 0.19742369651794434, "learning_rate": 4.1154442256260814e-05, "loss": 0.018, "step": 28940 }, { "grad_norm": 0.1254173219203949, "learning_rate": 4.1121896406325824e-05, "loss": 0.0189, "step": 28950 }, { "grad_norm": 0.09636629372835159, "learning_rate": 4.108935443997499e-05, "loss": 0.0155, "step": 28960 }, { "grad_norm": 0.07726873457431793, "learning_rate": 4.105681637144325e-05, "loss": 0.0194, "step": 28970 }, { "grad_norm": 0.10546715557575226, "learning_rate": 4.102428221496387e-05, "loss": 0.0152, "step": 28980 }, { "grad_norm": 0.148087739944458, "learning_rate": 4.099175198476838e-05, "loss": 0.0177, "step": 28990 }, { "grad_norm": 0.1272495985031128, "learning_rate": 4.0959225695086635e-05, "loss": 0.0172, "step": 29000 }, { "grad_norm": 0.08903747797012329, "learning_rate": 4.0926703360146724e-05, "loss": 0.0162, "step": 29010 }, { "grad_norm": 0.13990357518196106, "learning_rate": 4.0894184994175e-05, "loss": 0.0172, "step": 29020 }, { "grad_norm": 0.11273425072431564, "learning_rate": 4.086167061139614e-05, "loss": 0.0147, "step": 29030 }, { "grad_norm": 0.1752830445766449, "learning_rate": 4.082916022603303e-05, "loss": 0.0162, "step": 29040 }, { "grad_norm": 0.13079899549484253, "learning_rate": 4.079665385230678e-05, "loss": 0.0167, "step": 29050 }, { "grad_norm": 0.13746868073940277, "learning_rate": 4.076415150443683e-05, "loss": 0.0159, "step": 29060 }, { "grad_norm": 0.0995606854557991, "learning_rate": 4.073165319664079e-05, "loss": 0.0164, "step": 29070 }, { "grad_norm": 0.07630464434623718, "learning_rate": 4.069915894313451e-05, "loss": 0.0157, "step": 29080 }, { "grad_norm": 0.15767614543437958, "learning_rate": 4.066666875813212e-05, "loss": 0.0145, "step": 29090 }, { "grad_norm": 0.11206874251365662, "learning_rate": 4.063418265584589e-05, "loss": 0.0174, "step": 29100 }, { "grad_norm": 0.14148865640163422, "learning_rate": 4.060170065048636e-05, "loss": 0.014, "step": 29110 }, { "grad_norm": 0.20816059410572052, "learning_rate": 4.056922275626227e-05, "loss": 0.0158, "step": 29120 }, { "grad_norm": 0.10252676159143448, "learning_rate": 4.053674898738056e-05, "loss": 0.014, "step": 29130 }, { "grad_norm": 0.08892812579870224, "learning_rate": 4.050427935804633e-05, "loss": 0.0166, "step": 29140 }, { "grad_norm": 0.09629368782043457, "learning_rate": 4.047181388246293e-05, "loss": 0.0156, "step": 29150 }, { "grad_norm": 0.11926700174808502, "learning_rate": 4.043935257483185e-05, "loss": 0.0189, "step": 29160 }, { "grad_norm": 0.17162184417247772, "learning_rate": 4.0406895449352775e-05, "loss": 0.0215, "step": 29170 }, { "grad_norm": 0.12343911081552505, "learning_rate": 4.0374442520223525e-05, "loss": 0.0163, "step": 29180 }, { "grad_norm": 0.1053028479218483, "learning_rate": 4.0341993801640155e-05, "loss": 0.0166, "step": 29190 }, { "grad_norm": 0.10060581564903259, "learning_rate": 4.030954930779681e-05, "loss": 0.0155, "step": 29200 }, { "grad_norm": 0.12881913781166077, "learning_rate": 4.027710905288582e-05, "loss": 0.0176, "step": 29210 }, { "grad_norm": 0.09160027652978897, "learning_rate": 4.0244673051097666e-05, "loss": 0.0171, "step": 29220 }, { "grad_norm": 0.152007058262825, "learning_rate": 4.021224131662095e-05, "loss": 0.0154, "step": 29230 }, { "grad_norm": 0.09946981072425842, "learning_rate": 4.0179813863642396e-05, "loss": 0.0185, "step": 29240 }, { "grad_norm": 0.12653522193431854, "learning_rate": 4.014739070634691e-05, "loss": 0.0161, "step": 29250 }, { "grad_norm": 0.11825774610042572, "learning_rate": 4.011497185891746e-05, "loss": 0.0165, "step": 29260 }, { "grad_norm": 0.07814688980579376, "learning_rate": 4.008255733553513e-05, "loss": 0.0153, "step": 29270 }, { "grad_norm": 0.14186938107013702, "learning_rate": 4.0050147150379194e-05, "loss": 0.0153, "step": 29280 }, { "grad_norm": 0.10690118372440338, "learning_rate": 4.001774131762692e-05, "loss": 0.0169, "step": 29290 }, { "grad_norm": 0.08291304856538773, "learning_rate": 3.9985339851453705e-05, "loss": 0.0157, "step": 29300 }, { "grad_norm": 0.08089806884527206, "learning_rate": 3.9952942766033114e-05, "loss": 0.0182, "step": 29310 }, { "grad_norm": 0.168727844953537, "learning_rate": 3.9920550075536686e-05, "loss": 0.0157, "step": 29320 }, { "grad_norm": 0.10597556829452515, "learning_rate": 3.988816179413408e-05, "loss": 0.0159, "step": 29330 }, { "grad_norm": 0.09812936931848526, "learning_rate": 3.985577793599307e-05, "loss": 0.0159, "step": 29340 }, { "grad_norm": 0.16281719505786896, "learning_rate": 3.982339851527942e-05, "loss": 0.0185, "step": 29350 }, { "grad_norm": 0.09427895396947861, "learning_rate": 3.979102354615701e-05, "loss": 0.015, "step": 29360 }, { "grad_norm": 0.12406350672245026, "learning_rate": 3.975865304278772e-05, "loss": 0.0182, "step": 29370 }, { "grad_norm": 0.07439212501049042, "learning_rate": 3.9726287019331546e-05, "loss": 0.0161, "step": 29380 }, { "grad_norm": 0.08185813575983047, "learning_rate": 3.9693925489946474e-05, "loss": 0.0116, "step": 29390 }, { "grad_norm": 0.07963480055332184, "learning_rate": 3.966156846878852e-05, "loss": 0.0171, "step": 29400 }, { "grad_norm": 0.21099311113357544, "learning_rate": 3.962921597001177e-05, "loss": 0.0162, "step": 29410 }, { "grad_norm": 0.09267430007457733, "learning_rate": 3.9596868007768294e-05, "loss": 0.0142, "step": 29420 }, { "grad_norm": 0.15554043650627136, "learning_rate": 3.9564524596208165e-05, "loss": 0.0179, "step": 29430 }, { "grad_norm": 0.09150592982769012, "learning_rate": 3.953218574947954e-05, "loss": 0.0155, "step": 29440 }, { "grad_norm": 0.09085676819086075, "learning_rate": 3.949985148172849e-05, "loss": 0.0165, "step": 29450 }, { "grad_norm": 0.11420532315969467, "learning_rate": 3.946752180709914e-05, "loss": 0.013, "step": 29460 }, { "grad_norm": 0.11606288701295853, "learning_rate": 3.943519673973357e-05, "loss": 0.0161, "step": 29470 }, { "grad_norm": 0.10056951642036438, "learning_rate": 3.940287629377189e-05, "loss": 0.0208, "step": 29480 }, { "grad_norm": 0.15955393016338348, "learning_rate": 3.937056048335214e-05, "loss": 0.0177, "step": 29490 }, { "grad_norm": 0.12290836125612259, "learning_rate": 3.933824932261037e-05, "loss": 0.0192, "step": 29500 }, { "grad_norm": 0.1302550882101059, "learning_rate": 3.9305942825680576e-05, "loss": 0.0191, "step": 29510 }, { "grad_norm": 0.11366842687129974, "learning_rate": 3.9273641006694706e-05, "loss": 0.0154, "step": 29520 }, { "grad_norm": 0.1926860362291336, "learning_rate": 3.924134387978266e-05, "loss": 0.0186, "step": 29530 }, { "grad_norm": 0.10610854625701904, "learning_rate": 3.920905145907234e-05, "loss": 0.0284, "step": 29540 }, { "grad_norm": 0.12294457852840424, "learning_rate": 3.9176763758689525e-05, "loss": 0.0146, "step": 29550 }, { "grad_norm": 0.06268675625324249, "learning_rate": 3.9144480792757925e-05, "loss": 0.015, "step": 29560 }, { "grad_norm": 0.07553467154502869, "learning_rate": 3.911220257539926e-05, "loss": 0.0189, "step": 29570 }, { "grad_norm": 0.17294403910636902, "learning_rate": 3.907992912073308e-05, "loss": 0.0186, "step": 29580 }, { "grad_norm": 0.167682483792305, "learning_rate": 3.90476604428769e-05, "loss": 0.0167, "step": 29590 }, { "grad_norm": 0.06763854622840881, "learning_rate": 3.901539655594614e-05, "loss": 0.0146, "step": 29600 }, { "grad_norm": 0.08417104184627533, "learning_rate": 3.898313747405413e-05, "loss": 0.0136, "step": 29610 }, { "grad_norm": 0.16775727272033691, "learning_rate": 3.895088321131205e-05, "loss": 0.0149, "step": 29620 }, { "grad_norm": 0.1347135454416275, "learning_rate": 3.891863378182907e-05, "loss": 0.0198, "step": 29630 }, { "grad_norm": 0.06423123180866241, "learning_rate": 3.8886389199712145e-05, "loss": 0.0149, "step": 29640 }, { "grad_norm": 0.09153858572244644, "learning_rate": 3.885414947906614e-05, "loss": 0.0191, "step": 29650 }, { "grad_norm": 0.15198618173599243, "learning_rate": 3.882191463399385e-05, "loss": 0.0187, "step": 29660 }, { "grad_norm": 0.12220696359872818, "learning_rate": 3.878968467859585e-05, "loss": 0.0181, "step": 29670 }, { "grad_norm": 0.22792915999889374, "learning_rate": 3.875745962697063e-05, "loss": 0.0204, "step": 29680 }, { "grad_norm": 0.1865498423576355, "learning_rate": 3.872523949321454e-05, "loss": 0.0177, "step": 29690 }, { "grad_norm": 0.1317993700504303, "learning_rate": 3.869302429142173e-05, "loss": 0.0159, "step": 29700 }, { "grad_norm": 0.15015891194343567, "learning_rate": 3.866081403568425e-05, "loss": 0.0169, "step": 29710 }, { "grad_norm": 0.07792578637599945, "learning_rate": 3.8628608740091934e-05, "loss": 0.0173, "step": 29720 }, { "grad_norm": 0.10128399729728699, "learning_rate": 3.85964084187325e-05, "loss": 0.0221, "step": 29730 }, { "grad_norm": 0.09661168605089188, "learning_rate": 3.856421308569146e-05, "loss": 0.0141, "step": 29740 }, { "grad_norm": 0.13371756672859192, "learning_rate": 3.853202275505212e-05, "loss": 0.0181, "step": 29750 }, { "grad_norm": 0.08753904700279236, "learning_rate": 3.849983744089565e-05, "loss": 0.0157, "step": 29760 }, { "grad_norm": 0.12786194682121277, "learning_rate": 3.8467657157301e-05, "loss": 0.0187, "step": 29770 }, { "grad_norm": 0.09486547857522964, "learning_rate": 3.8435481918344884e-05, "loss": 0.0161, "step": 29780 }, { "grad_norm": 0.12901930510997772, "learning_rate": 3.84033117381019e-05, "loss": 0.0162, "step": 29790 }, { "grad_norm": 0.10551179200410843, "learning_rate": 3.8371146630644335e-05, "loss": 0.0149, "step": 29800 }, { "grad_norm": 0.12360920757055283, "learning_rate": 3.833898661004231e-05, "loss": 0.0149, "step": 29810 }, { "grad_norm": 0.06931200623512268, "learning_rate": 3.830683169036372e-05, "loss": 0.0151, "step": 29820 }, { "grad_norm": 0.3111320734024048, "learning_rate": 3.827468188567422e-05, "loss": 0.0228, "step": 29830 }, { "grad_norm": 0.16833533346652985, "learning_rate": 3.82425372100372e-05, "loss": 0.0211, "step": 29840 }, { "grad_norm": 0.12329953908920288, "learning_rate": 3.821039767751387e-05, "loss": 0.0166, "step": 29850 }, { "grad_norm": 0.15812882781028748, "learning_rate": 3.817826330216315e-05, "loss": 0.0188, "step": 29860 }, { "grad_norm": 0.08431217074394226, "learning_rate": 3.814613409804169e-05, "loss": 0.0161, "step": 29870 }, { "grad_norm": 0.22672350704669952, "learning_rate": 3.811401007920388e-05, "loss": 0.0175, "step": 29880 }, { "grad_norm": 0.09623256325721741, "learning_rate": 3.8081891259701895e-05, "loss": 0.0161, "step": 29890 }, { "grad_norm": 0.16544169187545776, "learning_rate": 3.8049777653585584e-05, "loss": 0.0198, "step": 29900 }, { "grad_norm": 0.15724390745162964, "learning_rate": 3.80176692749025e-05, "loss": 0.0167, "step": 29910 }, { "grad_norm": 0.08946197479963303, "learning_rate": 3.798556613769799e-05, "loss": 0.0164, "step": 29920 }, { "grad_norm": 0.14684243500232697, "learning_rate": 3.795346825601502e-05, "loss": 0.0221, "step": 29930 }, { "grad_norm": 0.09321089088916779, "learning_rate": 3.7921375643894305e-05, "loss": 0.017, "step": 29940 }, { "grad_norm": 0.10987932980060577, "learning_rate": 3.788928831537424e-05, "loss": 0.016, "step": 29950 }, { "grad_norm": 0.11495782434940338, "learning_rate": 3.7857206284490934e-05, "loss": 0.021, "step": 29960 }, { "grad_norm": 0.16385021805763245, "learning_rate": 3.7825129565278114e-05, "loss": 0.019, "step": 29970 }, { "grad_norm": 0.1286260038614273, "learning_rate": 3.779305817176728e-05, "loss": 0.0165, "step": 29980 }, { "grad_norm": 0.11420753598213196, "learning_rate": 3.776099211798751e-05, "loss": 0.0179, "step": 29990 }, { "grad_norm": 0.11573547124862671, "learning_rate": 3.7728931417965584e-05, "loss": 0.0166, "step": 30000 }, { "grad_norm": 0.10831356048583984, "learning_rate": 3.769687608572598e-05, "loss": 0.0172, "step": 30010 }, { "grad_norm": 0.06655006110668182, "learning_rate": 3.766482613529076e-05, "loss": 0.0149, "step": 30020 }, { "grad_norm": 0.09493523836135864, "learning_rate": 3.763278158067964e-05, "loss": 0.0152, "step": 30030 }, { "grad_norm": 0.09262809157371521, "learning_rate": 3.7600742435910045e-05, "loss": 0.0156, "step": 30040 }, { "grad_norm": 0.16331705451011658, "learning_rate": 3.756870871499696e-05, "loss": 0.0195, "step": 30050 }, { "grad_norm": 0.12735716998577118, "learning_rate": 3.7536680431953025e-05, "loss": 0.0181, "step": 30060 }, { "grad_norm": 0.19103029370307922, "learning_rate": 3.7504657600788485e-05, "loss": 0.017, "step": 30070 }, { "grad_norm": 0.10436858981847763, "learning_rate": 3.747264023551125e-05, "loss": 0.0183, "step": 30080 }, { "grad_norm": 0.13365450501441956, "learning_rate": 3.744062835012678e-05, "loss": 0.0161, "step": 30090 }, { "grad_norm": 0.10276436060667038, "learning_rate": 3.740862195863814e-05, "loss": 0.0159, "step": 30100 }, { "grad_norm": 0.17011567950248718, "learning_rate": 3.737662107504606e-05, "loss": 0.0212, "step": 30110 }, { "grad_norm": 0.1250508725643158, "learning_rate": 3.734462571334879e-05, "loss": 0.0196, "step": 30120 }, { "grad_norm": 0.07100346684455872, "learning_rate": 3.731263588754217e-05, "loss": 0.0152, "step": 30130 }, { "grad_norm": 0.10656425356864929, "learning_rate": 3.7280651611619665e-05, "loss": 0.017, "step": 30140 }, { "grad_norm": 0.08024546504020691, "learning_rate": 3.724867289957229e-05, "loss": 0.0181, "step": 30150 }, { "grad_norm": 0.13578146696090698, "learning_rate": 3.7216699765388576e-05, "loss": 0.0125, "step": 30160 }, { "grad_norm": 0.11668247729539871, "learning_rate": 3.71847322230547e-05, "loss": 0.0205, "step": 30170 }, { "grad_norm": 0.10902078449726105, "learning_rate": 3.7152770286554335e-05, "loss": 0.0151, "step": 30180 }, { "grad_norm": 0.12188459187746048, "learning_rate": 3.712081396986872e-05, "loss": 0.0158, "step": 30190 }, { "grad_norm": 0.1061999723315239, "learning_rate": 3.708886328697663e-05, "loss": 0.0172, "step": 30200 }, { "grad_norm": 0.11156194657087326, "learning_rate": 3.7056918251854395e-05, "loss": 0.0162, "step": 30210 }, { "grad_norm": 0.12901969254016876, "learning_rate": 3.7024978878475846e-05, "loss": 0.015, "step": 30220 }, { "grad_norm": 0.12478885799646378, "learning_rate": 3.699304518081233e-05, "loss": 0.015, "step": 30230 }, { "grad_norm": 0.10870136320590973, "learning_rate": 3.696111717283277e-05, "loss": 0.0183, "step": 30240 }, { "grad_norm": 0.1264009028673172, "learning_rate": 3.692919486850355e-05, "loss": 0.0171, "step": 30250 }, { "grad_norm": 0.07006553560495377, "learning_rate": 3.689727828178854e-05, "loss": 0.0154, "step": 30260 }, { "grad_norm": 0.20249031484127045, "learning_rate": 3.6865367426649185e-05, "loss": 0.0193, "step": 30270 }, { "grad_norm": 0.10979447513818741, "learning_rate": 3.6833462317044346e-05, "loss": 0.0234, "step": 30280 }, { "grad_norm": 0.12342998385429382, "learning_rate": 3.6801562966930406e-05, "loss": 0.0187, "step": 30290 }, { "grad_norm": 0.08057952672243118, "learning_rate": 3.676966939026123e-05, "loss": 0.0169, "step": 30300 }, { "grad_norm": 0.1257845163345337, "learning_rate": 3.6737781600988166e-05, "loss": 0.0188, "step": 30310 }, { "grad_norm": 0.12154624611139297, "learning_rate": 3.670589961305998e-05, "loss": 0.0209, "step": 30320 }, { "grad_norm": 0.12529732286930084, "learning_rate": 3.667402344042298e-05, "loss": 0.0195, "step": 30330 }, { "grad_norm": 0.14939582347869873, "learning_rate": 3.664215309702086e-05, "loss": 0.017, "step": 30340 }, { "grad_norm": 0.14721624553203583, "learning_rate": 3.661028859679478e-05, "loss": 0.0181, "step": 30350 }, { "grad_norm": 0.13081203401088715, "learning_rate": 3.657842995368339e-05, "loss": 0.019, "step": 30360 }, { "grad_norm": 0.16177894175052643, "learning_rate": 3.654657718162273e-05, "loss": 0.0186, "step": 30370 }, { "grad_norm": 0.11641830950975418, "learning_rate": 3.6514730294546275e-05, "loss": 0.0183, "step": 30380 }, { "grad_norm": 0.11716515570878983, "learning_rate": 3.648288930638492e-05, "loss": 0.0188, "step": 30390 }, { "grad_norm": 0.10024742782115936, "learning_rate": 3.645105423106704e-05, "loss": 0.0172, "step": 30400 }, { "grad_norm": 0.08471894264221191, "learning_rate": 3.6419225082518336e-05, "loss": 0.0147, "step": 30410 }, { "grad_norm": 0.11306969076395035, "learning_rate": 3.6387401874661975e-05, "loss": 0.017, "step": 30420 }, { "grad_norm": 0.12010481208562851, "learning_rate": 3.6355584621418516e-05, "loss": 0.0169, "step": 30430 }, { "grad_norm": 0.1311640739440918, "learning_rate": 3.632377333670591e-05, "loss": 0.0167, "step": 30440 }, { "grad_norm": 0.09899280965328217, "learning_rate": 3.629196803443946e-05, "loss": 0.0173, "step": 30450 }, { "grad_norm": 0.09604940563440323, "learning_rate": 3.626016872853193e-05, "loss": 0.0165, "step": 30460 }, { "grad_norm": 0.10797473043203354, "learning_rate": 3.6228375432893405e-05, "loss": 0.0187, "step": 30470 }, { "grad_norm": 0.2832624614238739, "learning_rate": 3.619658816143131e-05, "loss": 0.0146, "step": 30480 }, { "grad_norm": 0.13156162202358246, "learning_rate": 3.6164806928050544e-05, "loss": 0.0191, "step": 30490 }, { "grad_norm": 0.09337524324655533, "learning_rate": 3.613303174665327e-05, "loss": 0.018, "step": 30500 }, { "grad_norm": 0.23236943781375885, "learning_rate": 3.6101262631138996e-05, "loss": 0.0174, "step": 30510 }, { "grad_norm": 0.09031796455383301, "learning_rate": 3.606949959540467e-05, "loss": 0.0155, "step": 30520 }, { "grad_norm": 0.10030386596918106, "learning_rate": 3.603774265334449e-05, "loss": 0.0176, "step": 30530 }, { "grad_norm": 0.1695040613412857, "learning_rate": 3.600599181885003e-05, "loss": 0.0191, "step": 30540 }, { "grad_norm": 0.08208048343658447, "learning_rate": 3.597424710581018e-05, "loss": 0.0154, "step": 30550 }, { "grad_norm": 0.0885346308350563, "learning_rate": 3.5942508528111164e-05, "loss": 0.0152, "step": 30560 }, { "grad_norm": 0.14507079124450684, "learning_rate": 3.591077609963651e-05, "loss": 0.0183, "step": 30570 }, { "grad_norm": 0.12053379416465759, "learning_rate": 3.587904983426704e-05, "loss": 0.0167, "step": 30580 }, { "grad_norm": 0.09738411009311676, "learning_rate": 3.584732974588094e-05, "loss": 0.0148, "step": 30590 }, { "grad_norm": 0.08518819510936737, "learning_rate": 3.5815615848353645e-05, "loss": 0.0151, "step": 30600 }, { "grad_norm": 0.08489366620779037, "learning_rate": 3.578390815555784e-05, "loss": 0.016, "step": 30610 }, { "grad_norm": 0.09590772539377213, "learning_rate": 3.5752206681363626e-05, "loss": 0.0172, "step": 30620 }, { "grad_norm": 0.11152923107147217, "learning_rate": 3.572051143963826e-05, "loss": 0.0155, "step": 30630 }, { "grad_norm": 0.08406022936105728, "learning_rate": 3.5688822444246294e-05, "loss": 0.0145, "step": 30640 }, { "grad_norm": 0.23871983587741852, "learning_rate": 3.565713970904963e-05, "loss": 0.0189, "step": 30650 }, { "grad_norm": 0.16241855919361115, "learning_rate": 3.562546324790733e-05, "loss": 0.0148, "step": 30660 }, { "grad_norm": 0.12486189603805542, "learning_rate": 3.559379307467577e-05, "loss": 0.017, "step": 30670 }, { "grad_norm": 0.13628806173801422, "learning_rate": 3.556212920320856e-05, "loss": 0.0148, "step": 30680 }, { "grad_norm": 0.10090593248605728, "learning_rate": 3.553047164735655e-05, "loss": 0.0145, "step": 30690 }, { "grad_norm": 0.17695067822933197, "learning_rate": 3.549882042096783e-05, "loss": 0.0156, "step": 30700 }, { "grad_norm": 0.10399454832077026, "learning_rate": 3.546717553788772e-05, "loss": 0.0145, "step": 30710 }, { "grad_norm": 0.09962844103574753, "learning_rate": 3.543553701195879e-05, "loss": 0.0149, "step": 30720 }, { "grad_norm": 0.18918897211551666, "learning_rate": 3.540390485702077e-05, "loss": 0.018, "step": 30730 }, { "grad_norm": 0.10591316223144531, "learning_rate": 3.5372279086910645e-05, "loss": 0.013, "step": 30740 }, { "grad_norm": 0.07743318378925323, "learning_rate": 3.5340659715462635e-05, "loss": 0.0138, "step": 30750 }, { "grad_norm": 0.06365974992513657, "learning_rate": 3.530904675650808e-05, "loss": 0.012, "step": 30760 }, { "grad_norm": 0.12026292085647583, "learning_rate": 3.527744022387559e-05, "loss": 0.014, "step": 30770 }, { "grad_norm": 0.12550930678844452, "learning_rate": 3.524584013139094e-05, "loss": 0.0136, "step": 30780 }, { "grad_norm": 0.08654236048460007, "learning_rate": 3.521424649287708e-05, "loss": 0.0136, "step": 30790 }, { "grad_norm": 0.1009642705321312, "learning_rate": 3.518265932215412e-05, "loss": 0.013, "step": 30800 }, { "grad_norm": 0.09941362589597702, "learning_rate": 3.51510786330394e-05, "loss": 0.0132, "step": 30810 }, { "grad_norm": 0.12424501776695251, "learning_rate": 3.511950443934736e-05, "loss": 0.0128, "step": 30820 }, { "grad_norm": 0.08927394449710846, "learning_rate": 3.508793675488961e-05, "loss": 0.0148, "step": 30830 }, { "grad_norm": 0.14450661838054657, "learning_rate": 3.505637559347498e-05, "loss": 0.0153, "step": 30840 }, { "grad_norm": 0.08942519128322601, "learning_rate": 3.5024820968909345e-05, "loss": 0.0112, "step": 30850 }, { "grad_norm": 0.11430623382329941, "learning_rate": 3.4993272894995774e-05, "loss": 0.0154, "step": 30860 }, { "grad_norm": 0.09538059681653976, "learning_rate": 3.49617313855345e-05, "loss": 0.0117, "step": 30870 }, { "grad_norm": 0.09211049228906631, "learning_rate": 3.4930196454322834e-05, "loss": 0.0138, "step": 30880 }, { "grad_norm": 0.0756223276257515, "learning_rate": 3.48986681151552e-05, "loss": 0.013, "step": 30890 }, { "grad_norm": 0.1292264759540558, "learning_rate": 3.4867146381823204e-05, "loss": 0.0156, "step": 30900 }, { "grad_norm": 0.15766341984272003, "learning_rate": 3.48356312681155e-05, "loss": 0.0167, "step": 30910 }, { "grad_norm": 0.10599891096353531, "learning_rate": 3.480412278781788e-05, "loss": 0.0144, "step": 30920 }, { "grad_norm": 0.14551255106925964, "learning_rate": 3.47726209547132e-05, "loss": 0.014, "step": 30930 }, { "grad_norm": 0.1266455054283142, "learning_rate": 3.474112578258147e-05, "loss": 0.0114, "step": 30940 }, { "grad_norm": 0.08223643153905869, "learning_rate": 3.470963728519974e-05, "loss": 0.0142, "step": 30950 }, { "grad_norm": 0.11211986839771271, "learning_rate": 3.467815547634211e-05, "loss": 0.0121, "step": 30960 }, { "grad_norm": 0.09551750123500824, "learning_rate": 3.4646680369779836e-05, "loss": 0.0142, "step": 30970 }, { "grad_norm": 0.11312823742628098, "learning_rate": 3.46152119792812e-05, "loss": 0.0122, "step": 30980 }, { "grad_norm": 0.08495239913463593, "learning_rate": 3.45837503186115e-05, "loss": 0.0136, "step": 30990 }, { "grad_norm": 0.24030593037605286, "learning_rate": 3.455229540153318e-05, "loss": 0.0135, "step": 31000 }, { "grad_norm": 0.10622945427894592, "learning_rate": 3.4520847241805665e-05, "loss": 0.0132, "step": 31010 }, { "grad_norm": 0.12401214987039566, "learning_rate": 3.4489405853185465e-05, "loss": 0.0122, "step": 31020 }, { "grad_norm": 0.13093475997447968, "learning_rate": 3.44579712494261e-05, "loss": 0.0123, "step": 31030 }, { "grad_norm": 0.1334972381591797, "learning_rate": 3.442654344427814e-05, "loss": 0.0155, "step": 31040 }, { "grad_norm": 0.11037059873342514, "learning_rate": 3.439512245148916e-05, "loss": 0.0118, "step": 31050 }, { "grad_norm": 0.07130108028650284, "learning_rate": 3.4363708284803796e-05, "loss": 0.01, "step": 31060 }, { "grad_norm": 0.11484188586473465, "learning_rate": 3.433230095796367e-05, "loss": 0.0139, "step": 31070 }, { "grad_norm": 0.17359618842601776, "learning_rate": 3.430090048470739e-05, "loss": 0.0113, "step": 31080 }, { "grad_norm": 0.11685415357351303, "learning_rate": 3.4269506878770585e-05, "loss": 0.0141, "step": 31090 }, { "grad_norm": 0.11646030098199844, "learning_rate": 3.4238120153885925e-05, "loss": 0.0124, "step": 31100 }, { "grad_norm": 0.11752079427242279, "learning_rate": 3.4206740323782994e-05, "loss": 0.015, "step": 31110 }, { "grad_norm": 0.10213872790336609, "learning_rate": 3.4175367402188395e-05, "loss": 0.0139, "step": 31120 }, { "grad_norm": 0.14382362365722656, "learning_rate": 3.414400140282574e-05, "loss": 0.0144, "step": 31130 }, { "grad_norm": 0.11855415254831314, "learning_rate": 3.411264233941556e-05, "loss": 0.014, "step": 31140 }, { "grad_norm": 0.08144217729568481, "learning_rate": 3.408129022567537e-05, "loss": 0.0142, "step": 31150 }, { "grad_norm": 0.1689598262310028, "learning_rate": 3.404994507531968e-05, "loss": 0.0151, "step": 31160 }, { "grad_norm": 0.11097928136587143, "learning_rate": 3.40186069020599e-05, "loss": 0.0143, "step": 31170 }, { "grad_norm": 0.0885418951511383, "learning_rate": 3.3987275719604386e-05, "loss": 0.0169, "step": 31180 }, { "grad_norm": 0.09361272305250168, "learning_rate": 3.3955951541658526e-05, "loss": 0.0132, "step": 31190 }, { "grad_norm": 0.07914308458566666, "learning_rate": 3.392463438192454e-05, "loss": 0.0125, "step": 31200 }, { "grad_norm": 0.09800460934638977, "learning_rate": 3.38933242541016e-05, "loss": 0.0152, "step": 31210 }, { "grad_norm": 0.07222364097833633, "learning_rate": 3.386202117188586e-05, "loss": 0.0145, "step": 31220 }, { "grad_norm": 0.10245613008737564, "learning_rate": 3.383072514897033e-05, "loss": 0.0148, "step": 31230 }, { "grad_norm": 0.08284533023834229, "learning_rate": 3.379943619904494e-05, "loss": 0.0154, "step": 31240 }, { "grad_norm": 0.07233686745166779, "learning_rate": 3.3768154335796583e-05, "loss": 0.0158, "step": 31250 }, { "grad_norm": 0.23261648416519165, "learning_rate": 3.3736879572908963e-05, "loss": 0.0237, "step": 31260 }, { "grad_norm": 0.148414745926857, "learning_rate": 3.370561192406276e-05, "loss": 0.0134, "step": 31270 }, { "grad_norm": 0.1334942728281021, "learning_rate": 3.367435140293547e-05, "loss": 0.0183, "step": 31280 }, { "grad_norm": 0.17216747999191284, "learning_rate": 3.3643098023201545e-05, "loss": 0.0176, "step": 31290 }, { "grad_norm": 0.26416513323783875, "learning_rate": 3.361185179853225e-05, "loss": 0.0186, "step": 31300 }, { "grad_norm": 0.09891235083341599, "learning_rate": 3.358061274259574e-05, "loss": 0.0215, "step": 31310 }, { "grad_norm": 0.13759854435920715, "learning_rate": 3.3549380869057064e-05, "loss": 0.0167, "step": 31320 }, { "grad_norm": 0.11958230286836624, "learning_rate": 3.351815619157809e-05, "loss": 0.0152, "step": 31330 }, { "grad_norm": 0.13253869116306305, "learning_rate": 3.348693872381754e-05, "loss": 0.0165, "step": 31340 }, { "grad_norm": 0.11425324529409409, "learning_rate": 3.3455728479431025e-05, "loss": 0.0159, "step": 31350 }, { "grad_norm": 0.17121241986751556, "learning_rate": 3.342452547207094e-05, "loss": 0.0179, "step": 31360 }, { "grad_norm": 0.25311049818992615, "learning_rate": 3.3393329715386555e-05, "loss": 0.0191, "step": 31370 }, { "grad_norm": 0.20991171896457672, "learning_rate": 3.336214122302395e-05, "loss": 0.0138, "step": 31380 }, { "grad_norm": 0.13011403381824493, "learning_rate": 3.333096000862604e-05, "loss": 0.0148, "step": 31390 }, { "grad_norm": 0.0823407992720604, "learning_rate": 3.3299786085832516e-05, "loss": 0.0133, "step": 31400 }, { "grad_norm": 0.12908224761486053, "learning_rate": 3.3268619468279956e-05, "loss": 0.0145, "step": 31410 }, { "grad_norm": 0.0460507906973362, "learning_rate": 3.323746016960167e-05, "loss": 0.0119, "step": 31420 }, { "grad_norm": 0.130389004945755, "learning_rate": 3.3206308203427806e-05, "loss": 0.0152, "step": 31430 }, { "grad_norm": 0.062096890062093735, "learning_rate": 3.3175163583385254e-05, "loss": 0.0148, "step": 31440 }, { "grad_norm": 0.09389208257198334, "learning_rate": 3.314402632309779e-05, "loss": 0.0165, "step": 31450 }, { "grad_norm": 0.13712245225906372, "learning_rate": 3.311289643618587e-05, "loss": 0.0151, "step": 31460 }, { "grad_norm": 0.08739010989665985, "learning_rate": 3.308177393626676e-05, "loss": 0.0136, "step": 31470 }, { "grad_norm": 0.13469651341438293, "learning_rate": 3.305065883695453e-05, "loss": 0.0174, "step": 31480 }, { "grad_norm": 0.1469908207654953, "learning_rate": 3.301955115185996e-05, "loss": 0.0173, "step": 31490 }, { "grad_norm": 0.10685190558433533, "learning_rate": 3.29884508945906e-05, "loss": 0.016, "step": 31500 }, { "grad_norm": 0.14740332961082458, "learning_rate": 3.295735807875079e-05, "loss": 0.015, "step": 31510 }, { "grad_norm": 0.0695749819278717, "learning_rate": 3.292627271794156e-05, "loss": 0.0155, "step": 31520 }, { "grad_norm": 0.17759492993354797, "learning_rate": 3.289519482576069e-05, "loss": 0.0205, "step": 31530 }, { "grad_norm": 0.07367175072431564, "learning_rate": 3.286412441580274e-05, "loss": 0.013, "step": 31540 }, { "grad_norm": 0.10648266971111298, "learning_rate": 3.283306150165896e-05, "loss": 0.0192, "step": 31550 }, { "grad_norm": 0.0465749055147171, "learning_rate": 3.2802006096917285e-05, "loss": 0.0172, "step": 31560 }, { "grad_norm": 0.12036252021789551, "learning_rate": 3.277095821516245e-05, "loss": 0.0183, "step": 31570 }, { "grad_norm": 0.08868297934532166, "learning_rate": 3.273991786997583e-05, "loss": 0.0152, "step": 31580 }, { "grad_norm": 0.15568195283412933, "learning_rate": 3.2708885074935514e-05, "loss": 0.0167, "step": 31590 }, { "grad_norm": 0.08161798864603043, "learning_rate": 3.267785984361633e-05, "loss": 0.0131, "step": 31600 }, { "grad_norm": 0.12133027613162994, "learning_rate": 3.2646842189589754e-05, "loss": 0.0183, "step": 31610 }, { "grad_norm": 0.18234436213970184, "learning_rate": 3.261583212642397e-05, "loss": 0.0163, "step": 31620 }, { "grad_norm": 0.08128337562084198, "learning_rate": 3.25848296676838e-05, "loss": 0.0148, "step": 31630 }, { "grad_norm": 0.11864530295133591, "learning_rate": 3.255383482693083e-05, "loss": 0.0135, "step": 31640 }, { "grad_norm": 0.10777943581342697, "learning_rate": 3.252284761772322e-05, "loss": 0.0168, "step": 31650 }, { "grad_norm": 0.06973695755004883, "learning_rate": 3.249186805361581e-05, "loss": 0.0149, "step": 31660 }, { "grad_norm": 0.17688700556755066, "learning_rate": 3.246089614816016e-05, "loss": 0.0141, "step": 31670 }, { "grad_norm": 0.07573945820331573, "learning_rate": 3.242993191490441e-05, "loss": 0.0143, "step": 31680 }, { "grad_norm": 0.13772642612457275, "learning_rate": 3.239897536739335e-05, "loss": 0.0147, "step": 31690 }, { "grad_norm": 0.10329511761665344, "learning_rate": 3.236802651916845e-05, "loss": 0.0155, "step": 31700 }, { "grad_norm": 0.0698712170124054, "learning_rate": 3.23370853837678e-05, "loss": 0.0152, "step": 31710 }, { "grad_norm": 0.12683488428592682, "learning_rate": 3.2306151974726056e-05, "loss": 0.0196, "step": 31720 }, { "grad_norm": 0.11599691957235336, "learning_rate": 3.227522630557459e-05, "loss": 0.0144, "step": 31730 }, { "grad_norm": 0.17264561355113983, "learning_rate": 3.224430838984132e-05, "loss": 0.0181, "step": 31740 }, { "grad_norm": 0.11665680259466171, "learning_rate": 3.2213398241050784e-05, "loss": 0.0146, "step": 31750 }, { "grad_norm": 0.10886938124895096, "learning_rate": 3.218249587272415e-05, "loss": 0.0183, "step": 31760 }, { "grad_norm": 0.115376316010952, "learning_rate": 3.215160129837916e-05, "loss": 0.0182, "step": 31770 }, { "grad_norm": 0.12014852464199066, "learning_rate": 3.2120714531530146e-05, "loss": 0.0207, "step": 31780 }, { "grad_norm": 0.08663515001535416, "learning_rate": 3.2089835585688014e-05, "loss": 0.0297, "step": 31790 }, { "grad_norm": 0.13826489448547363, "learning_rate": 3.20589644743603e-05, "loss": 0.0181, "step": 31800 }, { "grad_norm": 0.0750800147652626, "learning_rate": 3.2028101211051054e-05, "loss": 0.0188, "step": 31810 }, { "grad_norm": 0.10185640305280685, "learning_rate": 3.19972458092609e-05, "loss": 0.0165, "step": 31820 }, { "grad_norm": 0.08526410162448883, "learning_rate": 3.196639828248707e-05, "loss": 0.0196, "step": 31830 }, { "grad_norm": 0.12099305540323257, "learning_rate": 3.193555864422329e-05, "loss": 0.0208, "step": 31840 }, { "grad_norm": 0.16153036057949066, "learning_rate": 3.19047269079599e-05, "loss": 0.0176, "step": 31850 }, { "grad_norm": 0.08366875350475311, "learning_rate": 3.1873903087183734e-05, "loss": 0.0163, "step": 31860 }, { "grad_norm": 0.1316094994544983, "learning_rate": 3.1843087195378176e-05, "loss": 0.017, "step": 31870 }, { "grad_norm": 0.10698328912258148, "learning_rate": 3.181227924602312e-05, "loss": 0.0167, "step": 31880 }, { "grad_norm": 0.11395274102687836, "learning_rate": 3.178147925259506e-05, "loss": 0.0297, "step": 31890 }, { "grad_norm": 0.09745959937572479, "learning_rate": 3.175068722856693e-05, "loss": 0.015, "step": 31900 }, { "grad_norm": 0.10935021936893463, "learning_rate": 3.17199031874082e-05, "loss": 0.0189, "step": 31910 }, { "grad_norm": 0.11628981679677963, "learning_rate": 3.168912714258487e-05, "loss": 0.0213, "step": 31920 }, { "grad_norm": 0.10059048235416412, "learning_rate": 3.165835910755944e-05, "loss": 0.0143, "step": 31930 }, { "grad_norm": 0.09560398757457733, "learning_rate": 3.1627599095790874e-05, "loss": 0.0164, "step": 31940 }, { "grad_norm": 0.1360759288072586, "learning_rate": 3.159684712073462e-05, "loss": 0.0175, "step": 31950 }, { "grad_norm": 0.12907063961029053, "learning_rate": 3.1566103195842695e-05, "loss": 0.0172, "step": 31960 }, { "grad_norm": 0.16881725192070007, "learning_rate": 3.153536733456349e-05, "loss": 0.0148, "step": 31970 }, { "grad_norm": 0.09475138783454895, "learning_rate": 3.150463955034194e-05, "loss": 0.0129, "step": 31980 }, { "grad_norm": 0.13505849242210388, "learning_rate": 3.147391985661942e-05, "loss": 0.0194, "step": 31990 }, { "grad_norm": 0.0856519564986229, "learning_rate": 3.1443208266833755e-05, "loss": 0.0138, "step": 32000 }, { "grad_norm": 0.14035892486572266, "learning_rate": 3.141250479441924e-05, "loss": 0.0154, "step": 32010 }, { "grad_norm": 0.10118673741817474, "learning_rate": 3.138180945280664e-05, "loss": 0.0171, "step": 32020 }, { "grad_norm": 0.14266736805438995, "learning_rate": 3.1351122255423116e-05, "loss": 0.017, "step": 32030 }, { "grad_norm": 0.10239565372467041, "learning_rate": 3.132044321569228e-05, "loss": 0.0164, "step": 32040 }, { "grad_norm": 0.081191286444664, "learning_rate": 3.1289772347034226e-05, "loss": 0.0172, "step": 32050 }, { "grad_norm": 0.13465675711631775, "learning_rate": 3.1259109662865404e-05, "loss": 0.0195, "step": 32060 }, { "grad_norm": 0.14203085005283356, "learning_rate": 3.122845517659871e-05, "loss": 0.017, "step": 32070 }, { "grad_norm": 0.09144572168588638, "learning_rate": 3.119780890164348e-05, "loss": 0.0129, "step": 32080 }, { "grad_norm": 0.06736057996749878, "learning_rate": 3.116717085140542e-05, "loss": 0.0179, "step": 32090 }, { "grad_norm": 0.1647365689277649, "learning_rate": 3.1136541039286657e-05, "loss": 0.0166, "step": 32100 }, { "grad_norm": 0.0891546681523323, "learning_rate": 3.110591947868571e-05, "loss": 0.0168, "step": 32110 }, { "grad_norm": 0.1302339881658554, "learning_rate": 3.107530618299751e-05, "loss": 0.017, "step": 32120 }, { "grad_norm": 0.11974871903657913, "learning_rate": 3.104470116561333e-05, "loss": 0.016, "step": 32130 }, { "grad_norm": 0.0943177342414856, "learning_rate": 3.1014104439920835e-05, "loss": 0.0155, "step": 32140 }, { "grad_norm": 0.1418013721704483, "learning_rate": 3.098351601930411e-05, "loss": 0.0153, "step": 32150 }, { "grad_norm": 0.15603716671466827, "learning_rate": 3.0952935917143533e-05, "loss": 0.0175, "step": 32160 }, { "grad_norm": 0.09563218057155609, "learning_rate": 3.092236414681588e-05, "loss": 0.017, "step": 32170 }, { "grad_norm": 0.10898581892251968, "learning_rate": 3.089180072169431e-05, "loss": 0.0151, "step": 32180 }, { "grad_norm": 0.11443260312080383, "learning_rate": 3.08612456551483e-05, "loss": 0.0142, "step": 32190 }, { "grad_norm": 0.1532214879989624, "learning_rate": 3.083069896054363e-05, "loss": 0.0163, "step": 32200 }, { "grad_norm": 0.12182220071554184, "learning_rate": 3.080016065124253e-05, "loss": 0.0173, "step": 32210 }, { "grad_norm": 0.0819581151008606, "learning_rate": 3.076963074060344e-05, "loss": 0.0159, "step": 32220 }, { "grad_norm": 0.070229172706604, "learning_rate": 3.07391092419812e-05, "loss": 0.0144, "step": 32230 }, { "grad_norm": 0.12899130582809448, "learning_rate": 3.070859616872695e-05, "loss": 0.0189, "step": 32240 }, { "grad_norm": 0.18360593914985657, "learning_rate": 3.067809153418816e-05, "loss": 0.0173, "step": 32250 }, { "grad_norm": 0.08398856967687607, "learning_rate": 3.0647595351708545e-05, "loss": 0.018, "step": 32260 }, { "grad_norm": 0.16952916979789734, "learning_rate": 3.061710763462824e-05, "loss": 0.0154, "step": 32270 }, { "grad_norm": 0.10077431797981262, "learning_rate": 3.0586628396283564e-05, "loss": 0.0133, "step": 32280 }, { "grad_norm": 0.07653965055942535, "learning_rate": 3.0556157650007174e-05, "loss": 0.0166, "step": 32290 }, { "grad_norm": 0.10548686236143112, "learning_rate": 3.052569540912801e-05, "loss": 0.0161, "step": 32300 }, { "grad_norm": 0.11644591391086578, "learning_rate": 3.049524168697131e-05, "loss": 0.0175, "step": 32310 }, { "grad_norm": 0.12929858267307281, "learning_rate": 3.0464796496858538e-05, "loss": 0.0138, "step": 32320 }, { "grad_norm": 0.10356678068637848, "learning_rate": 3.0434359852107473e-05, "loss": 0.0174, "step": 32330 }, { "grad_norm": 0.1381758451461792, "learning_rate": 3.0403931766032135e-05, "loss": 0.0171, "step": 32340 }, { "grad_norm": 0.10000765323638916, "learning_rate": 3.0373512251942814e-05, "loss": 0.0167, "step": 32350 }, { "grad_norm": 0.1171807199716568, "learning_rate": 3.0343101323146007e-05, "loss": 0.014, "step": 32360 }, { "grad_norm": 0.13681140542030334, "learning_rate": 3.0312698992944523e-05, "loss": 0.016, "step": 32370 }, { "grad_norm": 0.0708189532160759, "learning_rate": 3.028230527463736e-05, "loss": 0.0179, "step": 32380 }, { "grad_norm": 0.13704654574394226, "learning_rate": 3.0251920181519746e-05, "loss": 0.0137, "step": 32390 }, { "grad_norm": 0.14370155334472656, "learning_rate": 3.0221543726883183e-05, "loss": 0.016, "step": 32400 }, { "grad_norm": 0.18216031789779663, "learning_rate": 3.019117592401535e-05, "loss": 0.0172, "step": 32410 }, { "grad_norm": 0.1789289265871048, "learning_rate": 3.0160816786200137e-05, "loss": 0.0164, "step": 32420 }, { "grad_norm": 0.1981777846813202, "learning_rate": 3.01304663267177e-05, "loss": 0.0199, "step": 32430 }, { "grad_norm": 0.08656928688287735, "learning_rate": 3.0100124558844338e-05, "loss": 0.0156, "step": 32440 }, { "grad_norm": 0.08987158536911011, "learning_rate": 3.0069791495852562e-05, "loss": 0.0147, "step": 32450 }, { "grad_norm": 0.07071316242218018, "learning_rate": 3.0039467151011113e-05, "loss": 0.0144, "step": 32460 }, { "grad_norm": 0.16426147520542145, "learning_rate": 3.000915153758487e-05, "loss": 0.0174, "step": 32470 }, { "grad_norm": 0.09662598371505737, "learning_rate": 2.9978844668834917e-05, "loss": 0.0162, "step": 32480 }, { "grad_norm": 0.15526869893074036, "learning_rate": 2.994854655801849e-05, "loss": 0.0158, "step": 32490 }, { "grad_norm": 0.1664523035287857, "learning_rate": 2.991825721838905e-05, "loss": 0.018, "step": 32500 }, { "grad_norm": 0.10511656850576401, "learning_rate": 2.9887976663196166e-05, "loss": 0.0161, "step": 32510 }, { "grad_norm": 0.12429153919219971, "learning_rate": 2.985770490568556e-05, "loss": 0.0146, "step": 32520 }, { "grad_norm": 0.15604425966739655, "learning_rate": 2.9827441959099157e-05, "loss": 0.013, "step": 32530 }, { "grad_norm": 0.23270706832408905, "learning_rate": 2.979718783667499e-05, "loss": 0.0168, "step": 32540 }, { "grad_norm": 0.11975052207708359, "learning_rate": 2.9766942551647215e-05, "loss": 0.0187, "step": 32550 }, { "grad_norm": 0.1336950957775116, "learning_rate": 2.9736706117246188e-05, "loss": 0.0174, "step": 32560 }, { "grad_norm": 0.11709416657686234, "learning_rate": 2.9706478546698324e-05, "loss": 0.0189, "step": 32570 }, { "grad_norm": 0.09103292971849442, "learning_rate": 2.9676259853226195e-05, "loss": 0.0167, "step": 32580 }, { "grad_norm": 0.09247385710477829, "learning_rate": 2.964605005004849e-05, "loss": 0.0152, "step": 32590 }, { "grad_norm": 0.09664727747440338, "learning_rate": 2.9615849150380016e-05, "loss": 0.016, "step": 32600 }, { "grad_norm": 0.0851827934384346, "learning_rate": 2.9585657167431637e-05, "loss": 0.0157, "step": 32610 }, { "grad_norm": 0.10089622437953949, "learning_rate": 2.9555474114410388e-05, "loss": 0.0174, "step": 32620 }, { "grad_norm": 0.05415542796254158, "learning_rate": 2.9525300004519348e-05, "loss": 0.0145, "step": 32630 }, { "grad_norm": 0.04965578392148018, "learning_rate": 2.9495134850957696e-05, "loss": 0.0137, "step": 32640 }, { "grad_norm": 0.12407049536705017, "learning_rate": 2.9464978666920674e-05, "loss": 0.0155, "step": 32650 }, { "grad_norm": 0.10865703225135803, "learning_rate": 2.9434831465599662e-05, "loss": 0.015, "step": 32660 }, { "grad_norm": 0.08495884388685226, "learning_rate": 2.9404693260182047e-05, "loss": 0.0154, "step": 32670 }, { "grad_norm": 0.15247787535190582, "learning_rate": 2.9374564063851295e-05, "loss": 0.0149, "step": 32680 }, { "grad_norm": 0.14399787783622742, "learning_rate": 2.9344443889786967e-05, "loss": 0.018, "step": 32690 }, { "grad_norm": 0.17344407737255096, "learning_rate": 2.931433275116463e-05, "loss": 0.0155, "step": 32700 }, { "grad_norm": 0.23016802966594696, "learning_rate": 2.928423066115592e-05, "loss": 0.0166, "step": 32710 }, { "grad_norm": 0.08369724452495575, "learning_rate": 2.9254137632928537e-05, "loss": 0.0133, "step": 32720 }, { "grad_norm": 0.10077754408121109, "learning_rate": 2.9224053679646167e-05, "loss": 0.0159, "step": 32730 }, { "grad_norm": 0.09104610234498978, "learning_rate": 2.9193978814468547e-05, "loss": 0.018, "step": 32740 }, { "grad_norm": 0.09493423998355865, "learning_rate": 2.9163913050551485e-05, "loss": 0.0145, "step": 32750 }, { "grad_norm": 0.09382162988185883, "learning_rate": 2.9133856401046735e-05, "loss": 0.0147, "step": 32760 }, { "grad_norm": 0.057239875197410583, "learning_rate": 2.910380887910209e-05, "loss": 0.0132, "step": 32770 }, { "grad_norm": 0.1053890511393547, "learning_rate": 2.9073770497861386e-05, "loss": 0.0162, "step": 32780 }, { "grad_norm": 0.09548116475343704, "learning_rate": 2.904374127046442e-05, "loss": 0.0136, "step": 32790 }, { "grad_norm": 0.156846284866333, "learning_rate": 2.9013721210046973e-05, "loss": 0.0142, "step": 32800 }, { "grad_norm": 0.08435763418674469, "learning_rate": 2.8983710329740882e-05, "loss": 0.012, "step": 32810 }, { "grad_norm": 0.15386436879634857, "learning_rate": 2.8953708642673898e-05, "loss": 0.0156, "step": 32820 }, { "grad_norm": 0.10514939576387405, "learning_rate": 2.8923716161969782e-05, "loss": 0.0167, "step": 32830 }, { "grad_norm": 0.10330986976623535, "learning_rate": 2.8893732900748248e-05, "loss": 0.0136, "step": 32840 }, { "grad_norm": 0.13616357743740082, "learning_rate": 2.8863758872125036e-05, "loss": 0.0156, "step": 32850 }, { "grad_norm": 0.08465185761451721, "learning_rate": 2.883379408921175e-05, "loss": 0.0164, "step": 32860 }, { "grad_norm": 0.09379305690526962, "learning_rate": 2.880383856511606e-05, "loss": 0.0143, "step": 32870 }, { "grad_norm": 0.11136734485626221, "learning_rate": 2.8773892312941476e-05, "loss": 0.0156, "step": 32880 }, { "grad_norm": 0.0799354538321495, "learning_rate": 2.8743955345787555e-05, "loss": 0.0138, "step": 32890 }, { "grad_norm": 0.07154149562120438, "learning_rate": 2.87140276767497e-05, "loss": 0.0172, "step": 32900 }, { "grad_norm": 0.1210436075925827, "learning_rate": 2.868410931891934e-05, "loss": 0.0157, "step": 32910 }, { "grad_norm": 0.1255929172039032, "learning_rate": 2.8654200285383748e-05, "loss": 0.0192, "step": 32920 }, { "grad_norm": 0.14734281599521637, "learning_rate": 2.8624300589226126e-05, "loss": 0.014, "step": 32930 }, { "grad_norm": 0.07386550307273865, "learning_rate": 2.859441024352568e-05, "loss": 0.0157, "step": 32940 }, { "grad_norm": 0.08223168551921844, "learning_rate": 2.8564529261357427e-05, "loss": 0.0149, "step": 32950 }, { "grad_norm": 0.06903425604104996, "learning_rate": 2.8534657655792307e-05, "loss": 0.0132, "step": 32960 }, { "grad_norm": 0.11313524842262268, "learning_rate": 2.8504795439897208e-05, "loss": 0.0146, "step": 32970 }, { "grad_norm": 0.09971831738948822, "learning_rate": 2.847494262673487e-05, "loss": 0.016, "step": 32980 }, { "grad_norm": 0.10465334355831146, "learning_rate": 2.8445099229363925e-05, "loss": 0.0147, "step": 32990 }, { "grad_norm": 0.06768713891506195, "learning_rate": 2.841526526083886e-05, "loss": 0.0142, "step": 33000 }, { "grad_norm": 0.09578166157007217, "learning_rate": 2.838544073421011e-05, "loss": 0.0165, "step": 33010 }, { "grad_norm": 0.09058388322591782, "learning_rate": 2.8355625662523928e-05, "loss": 0.0127, "step": 33020 }, { "grad_norm": 0.06327053904533386, "learning_rate": 2.8325820058822407e-05, "loss": 0.0178, "step": 33030 }, { "grad_norm": 0.08713361620903015, "learning_rate": 2.829602393614357e-05, "loss": 0.0142, "step": 33040 }, { "grad_norm": 0.12886159121990204, "learning_rate": 2.8266237307521236e-05, "loss": 0.0171, "step": 33050 }, { "grad_norm": 0.08262001723051071, "learning_rate": 2.8236460185985074e-05, "loss": 0.0149, "step": 33060 }, { "grad_norm": 0.10442963987588882, "learning_rate": 2.8206692584560646e-05, "loss": 0.0136, "step": 33070 }, { "grad_norm": 0.2011510133743286, "learning_rate": 2.8176934516269283e-05, "loss": 0.0196, "step": 33080 }, { "grad_norm": 0.09763792157173157, "learning_rate": 2.8147185994128167e-05, "loss": 0.0146, "step": 33090 }, { "grad_norm": 0.12986955046653748, "learning_rate": 2.8117447031150346e-05, "loss": 0.015, "step": 33100 }, { "grad_norm": 0.08516945689916611, "learning_rate": 2.8087717640344618e-05, "loss": 0.0134, "step": 33110 }, { "grad_norm": 0.09212279319763184, "learning_rate": 2.805799783471566e-05, "loss": 0.0139, "step": 33120 }, { "grad_norm": 0.0799071192741394, "learning_rate": 2.8028287627263895e-05, "loss": 0.0142, "step": 33130 }, { "grad_norm": 0.15357103943824768, "learning_rate": 2.7998587030985608e-05, "loss": 0.0189, "step": 33140 }, { "grad_norm": 0.11850754916667938, "learning_rate": 2.7968896058872806e-05, "loss": 0.0106, "step": 33150 }, { "grad_norm": 0.08785535395145416, "learning_rate": 2.7939214723913388e-05, "loss": 0.0135, "step": 33160 }, { "grad_norm": 0.09897345304489136, "learning_rate": 2.7909543039090935e-05, "loss": 0.0141, "step": 33170 }, { "grad_norm": 0.1505672186613083, "learning_rate": 2.787988101738487e-05, "loss": 0.0148, "step": 33180 }, { "grad_norm": 0.15139234066009521, "learning_rate": 2.7850228671770334e-05, "loss": 0.0126, "step": 33190 }, { "grad_norm": 0.08514704555273056, "learning_rate": 2.782058601521832e-05, "loss": 0.0134, "step": 33200 }, { "grad_norm": 0.10565556585788727, "learning_rate": 2.77909530606955e-05, "loss": 0.0169, "step": 33210 }, { "grad_norm": 0.09115009754896164, "learning_rate": 2.7761329821164335e-05, "loss": 0.0153, "step": 33220 }, { "grad_norm": 0.09325534850358963, "learning_rate": 2.7731716309583062e-05, "loss": 0.0284, "step": 33230 }, { "grad_norm": 0.17638243734836578, "learning_rate": 2.7702112538905624e-05, "loss": 0.0156, "step": 33240 }, { "grad_norm": 0.11377481371164322, "learning_rate": 2.7672518522081682e-05, "loss": 0.0164, "step": 33250 }, { "grad_norm": 0.09443870186805725, "learning_rate": 2.7642934272056726e-05, "loss": 0.0148, "step": 33260 }, { "grad_norm": 0.08369755744934082, "learning_rate": 2.7613359801771877e-05, "loss": 0.0169, "step": 33270 }, { "grad_norm": 0.09411031752824783, "learning_rate": 2.7583795124163996e-05, "loss": 0.0166, "step": 33280 }, { "grad_norm": 0.10715952515602112, "learning_rate": 2.755424025216572e-05, "loss": 0.0133, "step": 33290 }, { "grad_norm": 0.08291196078062057, "learning_rate": 2.7524695198705334e-05, "loss": 0.0157, "step": 33300 }, { "grad_norm": 0.095109723508358, "learning_rate": 2.749515997670683e-05, "loss": 0.0184, "step": 33310 }, { "grad_norm": 0.09985119849443436, "learning_rate": 2.746563459908995e-05, "loss": 0.0141, "step": 33320 }, { "grad_norm": 0.14229556918144226, "learning_rate": 2.7436119078770085e-05, "loss": 0.0136, "step": 33330 }, { "grad_norm": 0.07329757511615753, "learning_rate": 2.74066134286583e-05, "loss": 0.0148, "step": 33340 }, { "grad_norm": 0.16690559685230255, "learning_rate": 2.7377117661661412e-05, "loss": 0.0144, "step": 33350 }, { "grad_norm": 0.18770140409469604, "learning_rate": 2.734763179068183e-05, "loss": 0.0145, "step": 33360 }, { "grad_norm": 0.11830861866474152, "learning_rate": 2.7318155828617718e-05, "loss": 0.0181, "step": 33370 }, { "grad_norm": 0.13336293399333954, "learning_rate": 2.7288689788362832e-05, "loss": 0.0167, "step": 33380 }, { "grad_norm": 0.09586317837238312, "learning_rate": 2.725923368280664e-05, "loss": 0.0162, "step": 33390 }, { "grad_norm": 0.1074061319231987, "learning_rate": 2.7229787524834237e-05, "loss": 0.0124, "step": 33400 }, { "grad_norm": 0.1265571415424347, "learning_rate": 2.7200351327326345e-05, "loss": 0.0153, "step": 33410 }, { "grad_norm": 0.07367333769798279, "learning_rate": 2.7170925103159407e-05, "loss": 0.0139, "step": 33420 }, { "grad_norm": 0.08213625103235245, "learning_rate": 2.714150886520542e-05, "loss": 0.0164, "step": 33430 }, { "grad_norm": 0.09124640375375748, "learning_rate": 2.7112102626332038e-05, "loss": 0.0156, "step": 33440 }, { "grad_norm": 0.09997396171092987, "learning_rate": 2.7082706399402574e-05, "loss": 0.014, "step": 33450 }, { "grad_norm": 0.1534949392080307, "learning_rate": 2.7053320197275934e-05, "loss": 0.0142, "step": 33460 }, { "grad_norm": 0.09074445813894272, "learning_rate": 2.7023944032806614e-05, "loss": 0.016, "step": 33470 }, { "grad_norm": 0.06501207500696182, "learning_rate": 2.699457791884478e-05, "loss": 0.0135, "step": 33480 }, { "grad_norm": 0.12175188213586807, "learning_rate": 2.6965221868236155e-05, "loss": 0.014, "step": 33490 }, { "grad_norm": 0.10452725738286972, "learning_rate": 2.6935875893822077e-05, "loss": 0.0155, "step": 33500 }, { "grad_norm": 0.13386203348636627, "learning_rate": 2.690654000843944e-05, "loss": 0.0162, "step": 33510 }, { "grad_norm": 0.0994560718536377, "learning_rate": 2.687721422492081e-05, "loss": 0.0152, "step": 33520 }, { "grad_norm": 0.1321347951889038, "learning_rate": 2.6847898556094252e-05, "loss": 0.0159, "step": 33530 }, { "grad_norm": 0.08620515465736389, "learning_rate": 2.6818593014783417e-05, "loss": 0.0154, "step": 33540 }, { "grad_norm": 0.1161651536822319, "learning_rate": 2.6789297613807585e-05, "loss": 0.0164, "step": 33550 }, { "grad_norm": 0.137111097574234, "learning_rate": 2.676001236598154e-05, "loss": 0.014, "step": 33560 }, { "grad_norm": 0.09419606626033783, "learning_rate": 2.6730737284115626e-05, "loss": 0.0178, "step": 33570 }, { "grad_norm": 0.051841042935848236, "learning_rate": 2.670147238101579e-05, "loss": 0.0117, "step": 33580 }, { "grad_norm": 0.1003652885556221, "learning_rate": 2.6672217669483473e-05, "loss": 0.0127, "step": 33590 }, { "grad_norm": 0.06377436965703964, "learning_rate": 2.6642973162315706e-05, "loss": 0.0132, "step": 33600 }, { "grad_norm": 0.13560466468334198, "learning_rate": 2.6613738872304988e-05, "loss": 0.0172, "step": 33610 }, { "grad_norm": 0.13213852047920227, "learning_rate": 2.6584514812239437e-05, "loss": 0.0156, "step": 33620 }, { "grad_norm": 0.1304301619529724, "learning_rate": 2.655530099490261e-05, "loss": 0.0172, "step": 33630 }, { "grad_norm": 0.07750517874956131, "learning_rate": 2.6526097433073653e-05, "loss": 0.0171, "step": 33640 }, { "grad_norm": 0.06616625189781189, "learning_rate": 2.649690413952719e-05, "loss": 0.0149, "step": 33650 }, { "grad_norm": 0.16430149972438812, "learning_rate": 2.646772112703334e-05, "loss": 0.0139, "step": 33660 }, { "grad_norm": 0.17087088525295258, "learning_rate": 2.6438548408357768e-05, "loss": 0.0167, "step": 33670 }, { "grad_norm": 0.06854428350925446, "learning_rate": 2.6409385996261603e-05, "loss": 0.012, "step": 33680 }, { "grad_norm": 0.098394975066185, "learning_rate": 2.638023390350148e-05, "loss": 0.0177, "step": 33690 }, { "grad_norm": 0.09524883329868317, "learning_rate": 2.635109214282948e-05, "loss": 0.0146, "step": 33700 }, { "grad_norm": 0.08004920929670334, "learning_rate": 2.6321960726993254e-05, "loss": 0.0167, "step": 33710 }, { "grad_norm": 0.10895748436450958, "learning_rate": 2.6292839668735846e-05, "loss": 0.0154, "step": 33720 }, { "grad_norm": 0.1730445772409439, "learning_rate": 2.6263728980795778e-05, "loss": 0.0173, "step": 33730 }, { "grad_norm": 0.1046738401055336, "learning_rate": 2.6234628675907092e-05, "loss": 0.0105, "step": 33740 }, { "grad_norm": 0.10407592356204987, "learning_rate": 2.620553876679923e-05, "loss": 0.0129, "step": 33750 }, { "grad_norm": 0.10178419202566147, "learning_rate": 2.617645926619709e-05, "loss": 0.0173, "step": 33760 }, { "grad_norm": 0.13371239602565765, "learning_rate": 2.6147390186821064e-05, "loss": 0.0184, "step": 33770 }, { "grad_norm": 0.11351824551820755, "learning_rate": 2.6118331541386955e-05, "loss": 0.0144, "step": 33780 }, { "grad_norm": 0.15236645936965942, "learning_rate": 2.6089283342605953e-05, "loss": 0.0182, "step": 33790 }, { "grad_norm": 0.08492331951856613, "learning_rate": 2.6060245603184797e-05, "loss": 0.0138, "step": 33800 }, { "grad_norm": 0.12208270281553268, "learning_rate": 2.6031218335825536e-05, "loss": 0.014, "step": 33810 }, { "grad_norm": 0.060473840683698654, "learning_rate": 2.600220155322568e-05, "loss": 0.0137, "step": 33820 }, { "grad_norm": 0.07845289260149002, "learning_rate": 2.5973195268078186e-05, "loss": 0.0145, "step": 33830 }, { "grad_norm": 0.059548091143369675, "learning_rate": 2.5944199493071352e-05, "loss": 0.0143, "step": 33840 }, { "grad_norm": 0.08212003111839294, "learning_rate": 2.591521424088895e-05, "loss": 0.0181, "step": 33850 }, { "grad_norm": 0.08948994427919388, "learning_rate": 2.5886239524210086e-05, "loss": 0.0177, "step": 33860 }, { "grad_norm": 0.10129168629646301, "learning_rate": 2.5857275355709317e-05, "loss": 0.0146, "step": 33870 }, { "grad_norm": 0.07860589772462845, "learning_rate": 2.5828321748056533e-05, "loss": 0.0154, "step": 33880 }, { "grad_norm": 0.12721240520477295, "learning_rate": 2.5799378713917012e-05, "loss": 0.0161, "step": 33890 }, { "grad_norm": 0.10540831834077835, "learning_rate": 2.5770446265951455e-05, "loss": 0.0174, "step": 33900 }, { "grad_norm": 0.17109759151935577, "learning_rate": 2.5741524416815877e-05, "loss": 0.0171, "step": 33910 }, { "grad_norm": 0.19456441700458527, "learning_rate": 2.571261317916166e-05, "loss": 0.0319, "step": 33920 }, { "grad_norm": 0.11267895996570587, "learning_rate": 2.5683712565635597e-05, "loss": 0.0157, "step": 33930 }, { "grad_norm": 0.06341388076543808, "learning_rate": 2.565482258887979e-05, "loss": 0.0145, "step": 33940 }, { "grad_norm": 0.09213100373744965, "learning_rate": 2.5625943261531664e-05, "loss": 0.0215, "step": 33950 }, { "grad_norm": 0.1031760647892952, "learning_rate": 2.5597074596224067e-05, "loss": 0.0155, "step": 33960 }, { "grad_norm": 0.1451670378446579, "learning_rate": 2.5568216605585105e-05, "loss": 0.0143, "step": 33970 }, { "grad_norm": 0.06449520587921143, "learning_rate": 2.5539369302238235e-05, "loss": 0.015, "step": 33980 }, { "grad_norm": 0.1257781833410263, "learning_rate": 2.5510532698802276e-05, "loss": 0.0139, "step": 33990 }, { "grad_norm": 0.1721838265657425, "learning_rate": 2.5481706807891324e-05, "loss": 0.0185, "step": 34000 }, { "grad_norm": 0.14764827489852905, "learning_rate": 2.5452891642114783e-05, "loss": 0.018, "step": 34010 }, { "grad_norm": 0.13832758367061615, "learning_rate": 2.542408721407742e-05, "loss": 0.0179, "step": 34020 }, { "grad_norm": 0.08603151887655258, "learning_rate": 2.539529353637926e-05, "loss": 0.0165, "step": 34030 }, { "grad_norm": 0.09201102703809738, "learning_rate": 2.5366510621615626e-05, "loss": 0.018, "step": 34040 }, { "grad_norm": 0.0862312987446785, "learning_rate": 2.5337738482377132e-05, "loss": 0.0186, "step": 34050 }, { "grad_norm": 0.11595135182142258, "learning_rate": 2.5308977131249722e-05, "loss": 0.0159, "step": 34060 }, { "grad_norm": 0.0853378102183342, "learning_rate": 2.5280226580814554e-05, "loss": 0.0172, "step": 34070 }, { "grad_norm": 0.1015087142586708, "learning_rate": 2.525148684364813e-05, "loss": 0.0159, "step": 34080 }, { "grad_norm": 0.11621119827032089, "learning_rate": 2.522275793232215e-05, "loss": 0.0159, "step": 34090 }, { "grad_norm": 0.07081045210361481, "learning_rate": 2.519403985940365e-05, "loss": 0.0182, "step": 34100 }, { "grad_norm": 0.10338740795850754, "learning_rate": 2.5165332637454863e-05, "loss": 0.0156, "step": 34110 }, { "grad_norm": 0.13369743525981903, "learning_rate": 2.513663627903334e-05, "loss": 0.0166, "step": 34120 }, { "grad_norm": 0.10439546406269073, "learning_rate": 2.5107950796691815e-05, "loss": 0.0176, "step": 34130 }, { "grad_norm": 0.07091385871171951, "learning_rate": 2.5079276202978276e-05, "loss": 0.0123, "step": 34140 }, { "grad_norm": 0.09185173362493515, "learning_rate": 2.505061251043601e-05, "loss": 0.0167, "step": 34150 }, { "grad_norm": 0.08968105912208557, "learning_rate": 2.502195973160346e-05, "loss": 0.0151, "step": 34160 }, { "grad_norm": 0.13583847880363464, "learning_rate": 2.4993317879014316e-05, "loss": 0.0195, "step": 34170 }, { "grad_norm": 0.08304591476917267, "learning_rate": 2.4964686965197526e-05, "loss": 0.0165, "step": 34180 }, { "grad_norm": 0.1503482609987259, "learning_rate": 2.4936067002677217e-05, "loss": 0.0151, "step": 34190 }, { "grad_norm": 0.11033234745264053, "learning_rate": 2.4907458003972724e-05, "loss": 0.0183, "step": 34200 }, { "grad_norm": 0.11125573515892029, "learning_rate": 2.4878859981598572e-05, "loss": 0.0159, "step": 34210 }, { "grad_norm": 0.14217448234558105, "learning_rate": 2.4850272948064555e-05, "loss": 0.0155, "step": 34220 }, { "grad_norm": 0.05878425016999245, "learning_rate": 2.4821696915875596e-05, "loss": 0.0163, "step": 34230 }, { "grad_norm": 0.16804784536361694, "learning_rate": 2.4793131897531798e-05, "loss": 0.0132, "step": 34240 }, { "grad_norm": 0.14066040515899658, "learning_rate": 2.47645779055285e-05, "loss": 0.0148, "step": 34250 }, { "grad_norm": 0.1010214313864708, "learning_rate": 2.4736034952356186e-05, "loss": 0.0152, "step": 34260 }, { "grad_norm": 0.17920957505702972, "learning_rate": 2.470750305050048e-05, "loss": 0.0174, "step": 34270 }, { "grad_norm": 0.11033879220485687, "learning_rate": 2.4678982212442253e-05, "loss": 0.0159, "step": 34280 }, { "grad_norm": 0.15294095873832703, "learning_rate": 2.4650472450657463e-05, "loss": 0.015, "step": 34290 }, { "grad_norm": 0.10125763714313507, "learning_rate": 2.4621973777617237e-05, "loss": 0.013, "step": 34300 }, { "grad_norm": 0.18824443221092224, "learning_rate": 2.459348620578789e-05, "loss": 0.016, "step": 34310 }, { "grad_norm": 0.10057664662599564, "learning_rate": 2.456500974763083e-05, "loss": 0.0113, "step": 34320 }, { "grad_norm": 0.11675308644771576, "learning_rate": 2.4536544415602653e-05, "loss": 0.0137, "step": 34330 }, { "grad_norm": 0.09372752159833908, "learning_rate": 2.4508090222155027e-05, "loss": 0.0121, "step": 34340 }, { "grad_norm": 0.08533963561058044, "learning_rate": 2.4479647179734822e-05, "loss": 0.0173, "step": 34350 }, { "grad_norm": 0.1262001246213913, "learning_rate": 2.4451215300783953e-05, "loss": 0.0147, "step": 34360 }, { "grad_norm": 0.13095135986804962, "learning_rate": 2.442279459773953e-05, "loss": 0.0178, "step": 34370 }, { "grad_norm": 0.10117392987012863, "learning_rate": 2.4394385083033712e-05, "loss": 0.0157, "step": 34380 }, { "grad_norm": 0.08410656452178955, "learning_rate": 2.4365986769093792e-05, "loss": 0.0163, "step": 34390 }, { "grad_norm": 0.10728124529123306, "learning_rate": 2.4337599668342127e-05, "loss": 0.017, "step": 34400 }, { "grad_norm": 0.0895375907421112, "learning_rate": 2.4309223793196245e-05, "loss": 0.0148, "step": 34410 }, { "grad_norm": 0.07572023570537567, "learning_rate": 2.4280859156068698e-05, "loss": 0.016, "step": 34420 }, { "grad_norm": 0.07893357425928116, "learning_rate": 2.4252505769367123e-05, "loss": 0.0159, "step": 34430 }, { "grad_norm": 0.13179747760295868, "learning_rate": 2.422416364549429e-05, "loss": 0.0163, "step": 34440 }, { "grad_norm": 0.12453902512788773, "learning_rate": 2.4195832796847994e-05, "loss": 0.0153, "step": 34450 }, { "grad_norm": 0.09100793302059174, "learning_rate": 2.416751323582109e-05, "loss": 0.0132, "step": 34460 }, { "grad_norm": 0.13051246106624603, "learning_rate": 2.4139204974801544e-05, "loss": 0.019, "step": 34470 }, { "grad_norm": 0.07503299415111542, "learning_rate": 2.4110908026172347e-05, "loss": 0.0129, "step": 34480 }, { "grad_norm": 0.11433155834674835, "learning_rate": 2.4082622402311518e-05, "loss": 0.0162, "step": 34490 }, { "grad_norm": 0.09419433772563934, "learning_rate": 2.4054348115592183e-05, "loss": 0.013, "step": 34500 }, { "grad_norm": 0.12765252590179443, "learning_rate": 2.4026085178382475e-05, "loss": 0.0135, "step": 34510 }, { "grad_norm": 0.15180641412734985, "learning_rate": 2.3997833603045523e-05, "loss": 0.0135, "step": 34520 }, { "grad_norm": 0.13087978959083557, "learning_rate": 2.396959340193958e-05, "loss": 0.0176, "step": 34530 }, { "grad_norm": 0.048994049429893494, "learning_rate": 2.3941364587417836e-05, "loss": 0.0123, "step": 34540 }, { "grad_norm": 0.10687027126550674, "learning_rate": 2.3913147171828526e-05, "loss": 0.0133, "step": 34550 }, { "grad_norm": 0.09463488310575485, "learning_rate": 2.388494116751494e-05, "loss": 0.0129, "step": 34560 }, { "grad_norm": 0.09908181428909302, "learning_rate": 2.3856746586815302e-05, "loss": 0.0129, "step": 34570 }, { "grad_norm": 0.1057734414935112, "learning_rate": 2.3828563442062924e-05, "loss": 0.0163, "step": 34580 }, { "grad_norm": 0.09578301757574081, "learning_rate": 2.3800391745586026e-05, "loss": 0.0158, "step": 34590 }, { "grad_norm": 0.15219490230083466, "learning_rate": 2.377223150970791e-05, "loss": 0.017, "step": 34600 }, { "grad_norm": 0.1256960928440094, "learning_rate": 2.3744082746746793e-05, "loss": 0.0137, "step": 34610 }, { "grad_norm": 0.07909926027059555, "learning_rate": 2.3715945469015893e-05, "loss": 0.0146, "step": 34620 }, { "grad_norm": 0.1422232985496521, "learning_rate": 2.368781968882343e-05, "loss": 0.0134, "step": 34630 }, { "grad_norm": 0.18260043859481812, "learning_rate": 2.365970541847258e-05, "loss": 0.0169, "step": 34640 }, { "grad_norm": 0.09709475934505463, "learning_rate": 2.3631602670261445e-05, "loss": 0.0166, "step": 34650 }, { "grad_norm": 0.11419486254453659, "learning_rate": 2.3603511456483164e-05, "loss": 0.0124, "step": 34660 }, { "grad_norm": 0.08587288111448288, "learning_rate": 2.357543178942577e-05, "loss": 0.013, "step": 34670 }, { "grad_norm": 0.10416440665721893, "learning_rate": 2.3547363681372242e-05, "loss": 0.0146, "step": 34680 }, { "grad_norm": 0.07238782197237015, "learning_rate": 2.3519307144600565e-05, "loss": 0.0136, "step": 34690 }, { "grad_norm": 0.13180376589298248, "learning_rate": 2.3491262191383607e-05, "loss": 0.0156, "step": 34700 }, { "grad_norm": 0.06866597384214401, "learning_rate": 2.3463228833989154e-05, "loss": 0.0136, "step": 34710 }, { "grad_norm": 0.057211704552173615, "learning_rate": 2.343520708467999e-05, "loss": 0.0154, "step": 34720 }, { "grad_norm": 0.1052488312125206, "learning_rate": 2.3407196955713767e-05, "loss": 0.0133, "step": 34730 }, { "grad_norm": 0.1094396784901619, "learning_rate": 2.337919845934306e-05, "loss": 0.0133, "step": 34740 }, { "grad_norm": 0.04639771953225136, "learning_rate": 2.335121160781535e-05, "loss": 0.0111, "step": 34750 }, { "grad_norm": 0.15263094007968903, "learning_rate": 2.3323236413373063e-05, "loss": 0.0151, "step": 34760 }, { "grad_norm": 0.08036786317825317, "learning_rate": 2.3295272888253494e-05, "loss": 0.0098, "step": 34770 }, { "grad_norm": 0.05442390218377113, "learning_rate": 2.3267321044688815e-05, "loss": 0.0166, "step": 34780 }, { "grad_norm": 0.08241847902536392, "learning_rate": 2.323938089490615e-05, "loss": 0.0123, "step": 34790 }, { "grad_norm": 0.0758858323097229, "learning_rate": 2.321145245112743e-05, "loss": 0.012, "step": 34800 }, { "grad_norm": 0.07725507766008377, "learning_rate": 2.318353572556953e-05, "loss": 0.0111, "step": 34810 }, { "grad_norm": 0.15183426439762115, "learning_rate": 2.3155630730444183e-05, "loss": 0.0146, "step": 34820 }, { "grad_norm": 0.07021185010671616, "learning_rate": 2.3127737477957968e-05, "loss": 0.0107, "step": 34830 }, { "grad_norm": 0.05453687533736229, "learning_rate": 2.3099855980312324e-05, "loss": 0.0115, "step": 34840 }, { "grad_norm": 0.09715476632118225, "learning_rate": 2.3071986249703602e-05, "loss": 0.0145, "step": 34850 }, { "grad_norm": 0.1359994113445282, "learning_rate": 2.3044128298322953e-05, "loss": 0.0138, "step": 34860 }, { "grad_norm": 0.07627668231725693, "learning_rate": 2.301628213835637e-05, "loss": 0.0119, "step": 34870 }, { "grad_norm": 0.10772544890642166, "learning_rate": 2.2988447781984744e-05, "loss": 0.014, "step": 34880 }, { "grad_norm": 0.08190639317035675, "learning_rate": 2.2960625241383753e-05, "loss": 0.0141, "step": 34890 }, { "grad_norm": 0.0705597996711731, "learning_rate": 2.2932814528723922e-05, "loss": 0.0118, "step": 34900 }, { "grad_norm": 0.08993680030107498, "learning_rate": 2.2905015656170582e-05, "loss": 0.0123, "step": 34910 }, { "grad_norm": 0.0975923016667366, "learning_rate": 2.287722863588394e-05, "loss": 0.0126, "step": 34920 }, { "grad_norm": 0.06418600678443909, "learning_rate": 2.2849453480018962e-05, "loss": 0.0116, "step": 34930 }, { "grad_norm": 0.09771456569433212, "learning_rate": 2.282169020072543e-05, "loss": 0.0131, "step": 34940 }, { "grad_norm": 0.146568164229393, "learning_rate": 2.2793938810147986e-05, "loss": 0.0158, "step": 34950 }, { "grad_norm": 0.13979047536849976, "learning_rate": 2.2766199320426007e-05, "loss": 0.0157, "step": 34960 }, { "grad_norm": 0.14180417358875275, "learning_rate": 2.2738471743693674e-05, "loss": 0.0156, "step": 34970 }, { "grad_norm": 0.10710052400827408, "learning_rate": 2.2710756092080006e-05, "loss": 0.0172, "step": 34980 }, { "grad_norm": 0.07289191335439682, "learning_rate": 2.2683052377708747e-05, "loss": 0.0136, "step": 34990 }, { "grad_norm": 0.09918668121099472, "learning_rate": 2.265536061269843e-05, "loss": 0.0144, "step": 35000 }, { "grad_norm": 0.08655069768428802, "learning_rate": 2.262768080916241e-05, "loss": 0.0145, "step": 35010 }, { "grad_norm": 0.24430523812770844, "learning_rate": 2.260001297920874e-05, "loss": 0.0138, "step": 35020 }, { "grad_norm": 0.11520084738731384, "learning_rate": 2.2572357134940297e-05, "loss": 0.0169, "step": 35030 }, { "grad_norm": 0.0702512115240097, "learning_rate": 2.2544713288454655e-05, "loss": 0.0164, "step": 35040 }, { "grad_norm": 0.06676720827817917, "learning_rate": 2.2517081451844213e-05, "loss": 0.0134, "step": 35050 }, { "grad_norm": 0.0790090560913086, "learning_rate": 2.248946163719605e-05, "loss": 0.0138, "step": 35060 }, { "grad_norm": 0.09877689927816391, "learning_rate": 2.2461853856591996e-05, "loss": 0.0135, "step": 35070 }, { "grad_norm": 0.15283812582492828, "learning_rate": 2.2434258122108664e-05, "loss": 0.0164, "step": 35080 }, { "grad_norm": 0.09810866415500641, "learning_rate": 2.2406674445817356e-05, "loss": 0.014, "step": 35090 }, { "grad_norm": 0.10920583456754684, "learning_rate": 2.2379102839784083e-05, "loss": 0.0168, "step": 35100 }, { "grad_norm": 0.10070734471082687, "learning_rate": 2.2351543316069645e-05, "loss": 0.0142, "step": 35110 }, { "grad_norm": 0.07988037168979645, "learning_rate": 2.2323995886729493e-05, "loss": 0.0138, "step": 35120 }, { "grad_norm": 0.11074801534414291, "learning_rate": 2.2296460563813786e-05, "loss": 0.0167, "step": 35130 }, { "grad_norm": 0.07768822461366653, "learning_rate": 2.2268937359367457e-05, "loss": 0.0192, "step": 35140 }, { "grad_norm": 0.19123785197734833, "learning_rate": 2.224142628543006e-05, "loss": 0.0119, "step": 35150 }, { "grad_norm": 0.1575905680656433, "learning_rate": 2.2213927354035863e-05, "loss": 0.014, "step": 35160 }, { "grad_norm": 0.08604349941015244, "learning_rate": 2.218644057721387e-05, "loss": 0.0154, "step": 35170 }, { "grad_norm": 0.09084376692771912, "learning_rate": 2.2158965966987703e-05, "loss": 0.0131, "step": 35180 }, { "grad_norm": 0.11467776447534561, "learning_rate": 2.2131503535375687e-05, "loss": 0.0163, "step": 35190 }, { "grad_norm": 0.07502997666597366, "learning_rate": 2.2104053294390846e-05, "loss": 0.0147, "step": 35200 }, { "grad_norm": 0.08110924065113068, "learning_rate": 2.2076615256040834e-05, "loss": 0.0211, "step": 35210 }, { "grad_norm": 0.09231007099151611, "learning_rate": 2.2049189432327966e-05, "loss": 0.0129, "step": 35220 }, { "grad_norm": 0.06144655868411064, "learning_rate": 2.202177583524926e-05, "loss": 0.0132, "step": 35230 }, { "grad_norm": 0.10029655694961548, "learning_rate": 2.1994374476796343e-05, "loss": 0.014, "step": 35240 }, { "grad_norm": 0.11264881491661072, "learning_rate": 2.1966985368955477e-05, "loss": 0.015, "step": 35250 }, { "grad_norm": 0.11741034686565399, "learning_rate": 2.193960852370762e-05, "loss": 0.016, "step": 35260 }, { "grad_norm": 0.10891994088888168, "learning_rate": 2.1912243953028317e-05, "loss": 0.0169, "step": 35270 }, { "grad_norm": 0.12765248119831085, "learning_rate": 2.1884891668887776e-05, "loss": 0.0199, "step": 35280 }, { "grad_norm": 0.1391124427318573, "learning_rate": 2.1857551683250787e-05, "loss": 0.0149, "step": 35290 }, { "grad_norm": 0.07118666917085648, "learning_rate": 2.1830224008076827e-05, "loss": 0.0161, "step": 35300 }, { "grad_norm": 0.092554472386837, "learning_rate": 2.1802908655319924e-05, "loss": 0.0168, "step": 35310 }, { "grad_norm": 0.10319339483976364, "learning_rate": 2.1775605636928738e-05, "loss": 0.0154, "step": 35320 }, { "grad_norm": 0.06669821590185165, "learning_rate": 2.1748314964846555e-05, "loss": 0.0153, "step": 35330 }, { "grad_norm": 0.09066233783960342, "learning_rate": 2.172103665101124e-05, "loss": 0.0129, "step": 35340 }, { "grad_norm": 0.07902524620294571, "learning_rate": 2.1693770707355226e-05, "loss": 0.0145, "step": 35350 }, { "grad_norm": 0.12539276480674744, "learning_rate": 2.1666517145805605e-05, "loss": 0.0156, "step": 35360 }, { "grad_norm": 0.09636663645505905, "learning_rate": 2.1639275978283985e-05, "loss": 0.0126, "step": 35370 }, { "grad_norm": 0.1181192398071289, "learning_rate": 2.1612047216706567e-05, "loss": 0.0152, "step": 35380 }, { "grad_norm": 0.08618928492069244, "learning_rate": 2.158483087298417e-05, "loss": 0.0185, "step": 35390 }, { "grad_norm": 0.07587522268295288, "learning_rate": 2.155762695902213e-05, "loss": 0.0176, "step": 35400 }, { "grad_norm": 0.14456476271152496, "learning_rate": 2.1530435486720363e-05, "loss": 0.0147, "step": 35410 }, { "grad_norm": 0.07680168002843857, "learning_rate": 2.1503256467973327e-05, "loss": 0.0165, "step": 35420 }, { "grad_norm": 0.1571665108203888, "learning_rate": 2.147608991467008e-05, "loss": 0.017, "step": 35430 }, { "grad_norm": 0.11174748837947845, "learning_rate": 2.144893583869418e-05, "loss": 0.014, "step": 35440 }, { "grad_norm": 0.09594614803791046, "learning_rate": 2.142179425192372e-05, "loss": 0.0139, "step": 35450 }, { "grad_norm": 0.1103348508477211, "learning_rate": 2.1394665166231393e-05, "loss": 0.0161, "step": 35460 }, { "grad_norm": 0.10563479363918304, "learning_rate": 2.1367548593484364e-05, "loss": 0.0126, "step": 35470 }, { "grad_norm": 0.11313625425100327, "learning_rate": 2.134044454554433e-05, "loss": 0.0137, "step": 35480 }, { "grad_norm": 0.12458949536085129, "learning_rate": 2.1313353034267548e-05, "loss": 0.0138, "step": 35490 }, { "grad_norm": 0.10962171852588654, "learning_rate": 2.1286274071504735e-05, "loss": 0.014, "step": 35500 }, { "grad_norm": 0.09774384647607803, "learning_rate": 2.125920766910119e-05, "loss": 0.0127, "step": 35510 }, { "grad_norm": 0.11103505641222, "learning_rate": 2.1232153838896628e-05, "loss": 0.0166, "step": 35520 }, { "grad_norm": 0.09832818806171417, "learning_rate": 2.1205112592725366e-05, "loss": 0.0143, "step": 35530 }, { "grad_norm": 0.07925006747245789, "learning_rate": 2.1178083942416114e-05, "loss": 0.0151, "step": 35540 }, { "grad_norm": 0.07250972837209702, "learning_rate": 2.1151067899792164e-05, "loss": 0.0131, "step": 35550 }, { "grad_norm": 0.11938099563121796, "learning_rate": 2.1124064476671235e-05, "loss": 0.013, "step": 35560 }, { "grad_norm": 0.1245737373828888, "learning_rate": 2.1097073684865516e-05, "loss": 0.0137, "step": 35570 }, { "grad_norm": 0.07125645875930786, "learning_rate": 2.107009553618174e-05, "loss": 0.0149, "step": 35580 }, { "grad_norm": 0.2648867666721344, "learning_rate": 2.1043130042421044e-05, "loss": 0.0174, "step": 35590 }, { "grad_norm": 0.09646587818861008, "learning_rate": 2.1016177215379053e-05, "loss": 0.0214, "step": 35600 }, { "grad_norm": 0.12493997812271118, "learning_rate": 2.0989237066845824e-05, "loss": 0.0157, "step": 35610 }, { "grad_norm": 0.08723662048578262, "learning_rate": 2.096230960860594e-05, "loss": 0.0126, "step": 35620 }, { "grad_norm": 0.151747927069664, "learning_rate": 2.0935394852438366e-05, "loss": 0.0139, "step": 35630 }, { "grad_norm": 0.13574931025505066, "learning_rate": 2.090849281011651e-05, "loss": 0.0155, "step": 35640 }, { "grad_norm": 0.09744877368211746, "learning_rate": 2.088160349340827e-05, "loss": 0.0138, "step": 35650 }, { "grad_norm": 0.07280749827623367, "learning_rate": 2.085472691407594e-05, "loss": 0.0129, "step": 35660 }, { "grad_norm": 0.09083867818117142, "learning_rate": 2.082786308387622e-05, "loss": 0.0153, "step": 35670 }, { "grad_norm": 0.18607079982757568, "learning_rate": 2.0801012014560306e-05, "loss": 0.0139, "step": 35680 }, { "grad_norm": 0.0922609493136406, "learning_rate": 2.0774173717873746e-05, "loss": 0.0133, "step": 35690 }, { "grad_norm": 0.13832904398441315, "learning_rate": 2.074734820555651e-05, "loss": 0.0136, "step": 35700 }, { "grad_norm": 0.06082800030708313, "learning_rate": 2.0720535489343014e-05, "loss": 0.0135, "step": 35710 }, { "grad_norm": 0.07981713116168976, "learning_rate": 2.0693735580962044e-05, "loss": 0.0119, "step": 35720 }, { "grad_norm": 0.09308809041976929, "learning_rate": 2.0666948492136766e-05, "loss": 0.0151, "step": 35730 }, { "grad_norm": 0.08392054587602615, "learning_rate": 2.0640174234584797e-05, "loss": 0.0141, "step": 35740 }, { "grad_norm": 0.0720025897026062, "learning_rate": 2.061341282001808e-05, "loss": 0.0152, "step": 35750 }, { "grad_norm": 0.10248713195323944, "learning_rate": 2.0586664260142997e-05, "loss": 0.0141, "step": 35760 }, { "grad_norm": 0.09226291626691818, "learning_rate": 2.0559928566660237e-05, "loss": 0.0128, "step": 35770 }, { "grad_norm": 0.08261563628911972, "learning_rate": 2.0533205751264945e-05, "loss": 0.0143, "step": 35780 }, { "grad_norm": 0.04991210997104645, "learning_rate": 2.050649582564656e-05, "loss": 0.0133, "step": 35790 }, { "grad_norm": 0.08582762628793716, "learning_rate": 2.04797988014889e-05, "loss": 0.0156, "step": 35800 }, { "grad_norm": 0.05939561873674393, "learning_rate": 2.0453114690470194e-05, "loss": 0.0155, "step": 35810 }, { "grad_norm": 0.11950133740901947, "learning_rate": 2.0426443504262944e-05, "loss": 0.0138, "step": 35820 }, { "grad_norm": 0.12608541548252106, "learning_rate": 2.0399785254534027e-05, "loss": 0.0134, "step": 35830 }, { "grad_norm": 0.11751116067171097, "learning_rate": 2.0373139952944697e-05, "loss": 0.0148, "step": 35840 }, { "grad_norm": 0.1035522073507309, "learning_rate": 2.03465076111505e-05, "loss": 0.0134, "step": 35850 }, { "grad_norm": 0.12510155141353607, "learning_rate": 2.0319888240801305e-05, "loss": 0.0135, "step": 35860 }, { "grad_norm": 0.06963233649730682, "learning_rate": 2.0293281853541375e-05, "loss": 0.0111, "step": 35870 }, { "grad_norm": 0.08859703689813614, "learning_rate": 2.0266688461009216e-05, "loss": 0.0144, "step": 35880 }, { "grad_norm": 0.08597966283559799, "learning_rate": 2.0240108074837666e-05, "loss": 0.0116, "step": 35890 }, { "grad_norm": 0.12318187206983566, "learning_rate": 2.021354070665393e-05, "loss": 0.0111, "step": 35900 }, { "grad_norm": 0.13420893251895905, "learning_rate": 2.0186986368079465e-05, "loss": 0.0292, "step": 35910 }, { "grad_norm": 0.09099278599023819, "learning_rate": 2.016044507073001e-05, "loss": 0.0131, "step": 35920 }, { "grad_norm": 0.07101541012525558, "learning_rate": 2.013391682621567e-05, "loss": 0.0133, "step": 35930 }, { "grad_norm": 0.14940352737903595, "learning_rate": 2.0107401646140787e-05, "loss": 0.0163, "step": 35940 }, { "grad_norm": 0.06321157515048981, "learning_rate": 2.0080899542104003e-05, "loss": 0.0139, "step": 35950 }, { "grad_norm": 0.11718635261058807, "learning_rate": 2.0054410525698215e-05, "loss": 0.012, "step": 35960 }, { "grad_norm": 0.10589990764856339, "learning_rate": 2.0027934608510653e-05, "loss": 0.0157, "step": 35970 }, { "grad_norm": 0.10038421303033829, "learning_rate": 2.000147180212275e-05, "loss": 0.0157, "step": 35980 }, { "grad_norm": 0.09005310386419296, "learning_rate": 1.9975022118110275e-05, "loss": 0.0138, "step": 35990 }, { "grad_norm": 0.17217744886875153, "learning_rate": 1.994858556804318e-05, "loss": 0.0154, "step": 36000 }, { "grad_norm": 0.06426405161619186, "learning_rate": 1.9922162163485748e-05, "loss": 0.0102, "step": 36010 }, { "grad_norm": 0.08826125413179398, "learning_rate": 1.989575191599643e-05, "loss": 0.0129, "step": 36020 }, { "grad_norm": 0.04485726356506348, "learning_rate": 1.9869354837128012e-05, "loss": 0.0095, "step": 36030 }, { "grad_norm": 0.1184317022562027, "learning_rate": 1.984297093842745e-05, "loss": 0.0129, "step": 36040 }, { "grad_norm": 0.08656446635723114, "learning_rate": 1.981660023143594e-05, "loss": 0.0114, "step": 36050 }, { "grad_norm": 0.08014104515314102, "learning_rate": 1.9790242727688963e-05, "loss": 0.0122, "step": 36060 }, { "grad_norm": 0.12088526785373688, "learning_rate": 1.9763898438716165e-05, "loss": 0.0119, "step": 36070 }, { "grad_norm": 0.07126446813344955, "learning_rate": 1.973756737604142e-05, "loss": 0.0116, "step": 36080 }, { "grad_norm": 0.09052559733390808, "learning_rate": 1.9711249551182858e-05, "loss": 0.0119, "step": 36090 }, { "grad_norm": 0.1144745796918869, "learning_rate": 1.968494497565278e-05, "loss": 0.0142, "step": 36100 }, { "grad_norm": 0.0644993707537651, "learning_rate": 1.9658653660957698e-05, "loss": 0.0087, "step": 36110 }, { "grad_norm": 0.05780060589313507, "learning_rate": 1.9632375618598302e-05, "loss": 0.011, "step": 36120 }, { "grad_norm": 0.06656183302402496, "learning_rate": 1.960611086006955e-05, "loss": 0.0125, "step": 36130 }, { "grad_norm": 0.12205240875482559, "learning_rate": 1.9579859396860513e-05, "loss": 0.0117, "step": 36140 }, { "grad_norm": 0.08800026774406433, "learning_rate": 1.9553621240454452e-05, "loss": 0.0121, "step": 36150 }, { "grad_norm": 0.06195502355694771, "learning_rate": 1.9527396402328864e-05, "loss": 0.0144, "step": 36160 }, { "grad_norm": 0.07610782980918884, "learning_rate": 1.9501184893955376e-05, "loss": 0.0118, "step": 36170 }, { "grad_norm": 0.08679954707622528, "learning_rate": 1.9474986726799765e-05, "loss": 0.012, "step": 36180 }, { "grad_norm": 0.06790617108345032, "learning_rate": 1.944880191232204e-05, "loss": 0.0126, "step": 36190 }, { "grad_norm": 0.1101951152086258, "learning_rate": 1.9422630461976304e-05, "loss": 0.0129, "step": 36200 }, { "grad_norm": 0.247132807970047, "learning_rate": 1.9396472387210835e-05, "loss": 0.013, "step": 36210 }, { "grad_norm": 0.10169372707605362, "learning_rate": 1.9370327699468082e-05, "loss": 0.0146, "step": 36220 }, { "grad_norm": 0.09462057799100876, "learning_rate": 1.9344196410184594e-05, "loss": 0.0121, "step": 36230 }, { "grad_norm": 0.07312800735235214, "learning_rate": 1.9318078530791123e-05, "loss": 0.0089, "step": 36240 }, { "grad_norm": 0.09368734806776047, "learning_rate": 1.9291974072712473e-05, "loss": 0.0116, "step": 36250 }, { "grad_norm": 0.07664746046066284, "learning_rate": 1.9265883047367656e-05, "loss": 0.0124, "step": 36260 }, { "grad_norm": 0.11685261130332947, "learning_rate": 1.9239805466169748e-05, "loss": 0.0132, "step": 36270 }, { "grad_norm": 0.13267172873020172, "learning_rate": 1.9213741340525992e-05, "loss": 0.0152, "step": 36280 }, { "grad_norm": 0.09354468435049057, "learning_rate": 1.9187690681837713e-05, "loss": 0.0166, "step": 36290 }, { "grad_norm": 0.09589816629886627, "learning_rate": 1.916165350150035e-05, "loss": 0.0107, "step": 36300 }, { "grad_norm": 0.10421444475650787, "learning_rate": 1.9135629810903434e-05, "loss": 0.0116, "step": 36310 }, { "grad_norm": 0.09591148048639297, "learning_rate": 1.9109619621430648e-05, "loss": 0.0117, "step": 36320 }, { "grad_norm": 0.117729052901268, "learning_rate": 1.9083622944459712e-05, "loss": 0.0136, "step": 36330 }, { "grad_norm": 0.10222963988780975, "learning_rate": 1.9057639791362437e-05, "loss": 0.0117, "step": 36340 }, { "grad_norm": 0.07011421024799347, "learning_rate": 1.903167017350478e-05, "loss": 0.0148, "step": 36350 }, { "grad_norm": 0.08639261871576309, "learning_rate": 1.9005714102246707e-05, "loss": 0.0116, "step": 36360 }, { "grad_norm": 0.11018253862857819, "learning_rate": 1.897977158894227e-05, "loss": 0.0123, "step": 36370 }, { "grad_norm": 0.07811347395181656, "learning_rate": 1.8953842644939644e-05, "loss": 0.015, "step": 36380 }, { "grad_norm": 0.0901326835155487, "learning_rate": 1.8927927281581015e-05, "loss": 0.0159, "step": 36390 }, { "grad_norm": 0.08428441733121872, "learning_rate": 1.890202551020262e-05, "loss": 0.0127, "step": 36400 }, { "grad_norm": 0.08489116281270981, "learning_rate": 1.8876137342134813e-05, "loss": 0.0132, "step": 36410 }, { "grad_norm": 0.09746373444795609, "learning_rate": 1.885026278870194e-05, "loss": 0.0131, "step": 36420 }, { "grad_norm": 0.10150062292814255, "learning_rate": 1.8824401861222384e-05, "loss": 0.0144, "step": 36430 }, { "grad_norm": 0.09640156477689743, "learning_rate": 1.879855457100864e-05, "loss": 0.0153, "step": 36440 }, { "grad_norm": 0.1028960570693016, "learning_rate": 1.877272092936717e-05, "loss": 0.014, "step": 36450 }, { "grad_norm": 0.09887297451496124, "learning_rate": 1.8746900947598467e-05, "loss": 0.0145, "step": 36460 }, { "grad_norm": 0.08919759094715118, "learning_rate": 1.87210946369971e-05, "loss": 0.0104, "step": 36470 }, { "grad_norm": 0.06770538538694382, "learning_rate": 1.8695302008851602e-05, "loss": 0.0142, "step": 36480 }, { "grad_norm": 0.05308077111840248, "learning_rate": 1.8669523074444568e-05, "loss": 0.0112, "step": 36490 }, { "grad_norm": 0.06367208808660507, "learning_rate": 1.8643757845052545e-05, "loss": 0.0104, "step": 36500 }, { "grad_norm": 0.057876504957675934, "learning_rate": 1.8618006331946164e-05, "loss": 0.0116, "step": 36510 }, { "grad_norm": 0.11029105633497238, "learning_rate": 1.859226854638999e-05, "loss": 0.0124, "step": 36520 }, { "grad_norm": 0.08025723695755005, "learning_rate": 1.8566544499642587e-05, "loss": 0.0137, "step": 36530 }, { "grad_norm": 0.13898725807666779, "learning_rate": 1.854083420295656e-05, "loss": 0.0126, "step": 36540 }, { "grad_norm": 0.08382653445005417, "learning_rate": 1.8515137667578453e-05, "loss": 0.0114, "step": 36550 }, { "grad_norm": 0.0890081375837326, "learning_rate": 1.8489454904748784e-05, "loss": 0.0147, "step": 36560 }, { "grad_norm": 0.08093296736478806, "learning_rate": 1.8463785925702094e-05, "loss": 0.011, "step": 36570 }, { "grad_norm": 0.055731840431690216, "learning_rate": 1.843813074166686e-05, "loss": 0.0155, "step": 36580 }, { "grad_norm": 0.12100949883460999, "learning_rate": 1.841248936386551e-05, "loss": 0.0155, "step": 36590 }, { "grad_norm": 0.0638415589928627, "learning_rate": 1.8386861803514483e-05, "loss": 0.0146, "step": 36600 }, { "grad_norm": 0.08791102468967438, "learning_rate": 1.836124807182414e-05, "loss": 0.0129, "step": 36610 }, { "grad_norm": 0.07037591934204102, "learning_rate": 1.8335648179998783e-05, "loss": 0.0112, "step": 36620 }, { "grad_norm": 0.19461633265018463, "learning_rate": 1.8310062139236667e-05, "loss": 0.0146, "step": 36630 }, { "grad_norm": 0.08427708595991135, "learning_rate": 1.828448996073002e-05, "loss": 0.0115, "step": 36640 }, { "grad_norm": 0.10557407140731812, "learning_rate": 1.825893165566498e-05, "loss": 0.0137, "step": 36650 }, { "grad_norm": 0.06648732721805573, "learning_rate": 1.823338723522159e-05, "loss": 0.0116, "step": 36660 }, { "grad_norm": 0.14383506774902344, "learning_rate": 1.8207856710573878e-05, "loss": 0.0148, "step": 36670 }, { "grad_norm": 0.1920340359210968, "learning_rate": 1.8182340092889756e-05, "loss": 0.0148, "step": 36680 }, { "grad_norm": 0.09964033961296082, "learning_rate": 1.8156837393331038e-05, "loss": 0.015, "step": 36690 }, { "grad_norm": 0.058069873601198196, "learning_rate": 1.81313486230535e-05, "loss": 0.0116, "step": 36700 }, { "grad_norm": 0.16274982690811157, "learning_rate": 1.8105873793206773e-05, "loss": 0.0144, "step": 36710 }, { "grad_norm": 0.17050644755363464, "learning_rate": 1.8080412914934436e-05, "loss": 0.0134, "step": 36720 }, { "grad_norm": 0.10602205991744995, "learning_rate": 1.805496599937392e-05, "loss": 0.0138, "step": 36730 }, { "grad_norm": 0.07514651119709015, "learning_rate": 1.802953305765659e-05, "loss": 0.0113, "step": 36740 }, { "grad_norm": 0.10877754539251328, "learning_rate": 1.8004114100907653e-05, "loss": 0.0107, "step": 36750 }, { "grad_norm": 0.0546819232404232, "learning_rate": 1.7978709140246263e-05, "loss": 0.013, "step": 36760 }, { "grad_norm": 0.13239945471286774, "learning_rate": 1.795331818678539e-05, "loss": 0.0117, "step": 36770 }, { "grad_norm": 0.09093985706567764, "learning_rate": 1.792794125163188e-05, "loss": 0.0139, "step": 36780 }, { "grad_norm": 0.054972317069768906, "learning_rate": 1.7902578345886507e-05, "loss": 0.014, "step": 36790 }, { "grad_norm": 0.09589000046253204, "learning_rate": 1.7877229480643843e-05, "loss": 0.012, "step": 36800 }, { "grad_norm": 0.08395534008741379, "learning_rate": 1.785189466699235e-05, "loss": 0.0118, "step": 36810 }, { "grad_norm": 0.11072669178247452, "learning_rate": 1.7826573916014317e-05, "loss": 0.0145, "step": 36820 }, { "grad_norm": 0.08671616017818451, "learning_rate": 1.7801267238785935e-05, "loss": 0.0129, "step": 36830 }, { "grad_norm": 0.15423588454723358, "learning_rate": 1.7775974646377187e-05, "loss": 0.0136, "step": 36840 }, { "grad_norm": 0.09163989871740341, "learning_rate": 1.775069614985189e-05, "loss": 0.0151, "step": 36850 }, { "grad_norm": 0.14402587711811066, "learning_rate": 1.7725431760267757e-05, "loss": 0.0148, "step": 36860 }, { "grad_norm": 0.07229046523571014, "learning_rate": 1.7700181488676277e-05, "loss": 0.0122, "step": 36870 }, { "grad_norm": 0.08818549662828445, "learning_rate": 1.767494534612275e-05, "loss": 0.0129, "step": 36880 }, { "grad_norm": 0.12669546902179718, "learning_rate": 1.7649723343646367e-05, "loss": 0.0115, "step": 36890 }, { "grad_norm": 0.08995263278484344, "learning_rate": 1.7624515492280063e-05, "loss": 0.0131, "step": 36900 }, { "grad_norm": 0.1107286587357521, "learning_rate": 1.7599321803050596e-05, "loss": 0.0141, "step": 36910 }, { "grad_norm": 0.09046443551778793, "learning_rate": 1.7574142286978574e-05, "loss": 0.015, "step": 36920 }, { "grad_norm": 0.10693898797035217, "learning_rate": 1.7548976955078354e-05, "loss": 0.0149, "step": 36930 }, { "grad_norm": 0.10954488068819046, "learning_rate": 1.7523825818358092e-05, "loss": 0.0142, "step": 36940 }, { "grad_norm": 0.09664609283208847, "learning_rate": 1.7498688887819785e-05, "loss": 0.0096, "step": 36950 }, { "grad_norm": 0.0820503681898117, "learning_rate": 1.7473566174459143e-05, "loss": 0.014, "step": 36960 }, { "grad_norm": 0.0780118852853775, "learning_rate": 1.7448457689265735e-05, "loss": 0.0131, "step": 36970 }, { "grad_norm": 0.083038330078125, "learning_rate": 1.7423363443222823e-05, "loss": 0.0159, "step": 36980 }, { "grad_norm": 0.11850925534963608, "learning_rate": 1.7398283447307524e-05, "loss": 0.0139, "step": 36990 }, { "grad_norm": 0.08918622881174088, "learning_rate": 1.737321771249066e-05, "loss": 0.0139, "step": 37000 }, { "grad_norm": 0.08616393059492111, "learning_rate": 1.7348166249736814e-05, "loss": 0.015, "step": 37010 }, { "grad_norm": 0.07525176554918289, "learning_rate": 1.732312907000439e-05, "loss": 0.014, "step": 37020 }, { "grad_norm": 0.05545657128095627, "learning_rate": 1.7298106184245478e-05, "loss": 0.0118, "step": 37030 }, { "grad_norm": 0.2505204379558563, "learning_rate": 1.727309760340592e-05, "loss": 0.014, "step": 37040 }, { "grad_norm": 0.08703159540891647, "learning_rate": 1.724810333842536e-05, "loss": 0.0141, "step": 37050 }, { "grad_norm": 0.09879998117685318, "learning_rate": 1.7223123400237112e-05, "loss": 0.0124, "step": 37060 }, { "grad_norm": 0.13586805760860443, "learning_rate": 1.7198157799768238e-05, "loss": 0.0132, "step": 37070 }, { "grad_norm": 0.1260882318019867, "learning_rate": 1.7173206547939576e-05, "loss": 0.0114, "step": 37080 }, { "grad_norm": 0.19121378660202026, "learning_rate": 1.7148269655665628e-05, "loss": 0.0125, "step": 37090 }, { "grad_norm": 0.054971951991319656, "learning_rate": 1.712334713385463e-05, "loss": 0.0131, "step": 37100 }, { "grad_norm": 0.057814374566078186, "learning_rate": 1.709843899340856e-05, "loss": 0.0129, "step": 37110 }, { "grad_norm": 0.0627867579460144, "learning_rate": 1.7073545245223084e-05, "loss": 0.0119, "step": 37120 }, { "grad_norm": 0.11164313554763794, "learning_rate": 1.704866590018755e-05, "loss": 0.0136, "step": 37130 }, { "grad_norm": 0.0673481822013855, "learning_rate": 1.7023800969185054e-05, "loss": 0.0105, "step": 37140 }, { "grad_norm": 0.06206319481134415, "learning_rate": 1.6998950463092354e-05, "loss": 0.0112, "step": 37150 }, { "grad_norm": 0.12645183503627777, "learning_rate": 1.6974114392779907e-05, "loss": 0.0123, "step": 37160 }, { "grad_norm": 0.06537093967199326, "learning_rate": 1.694929276911183e-05, "loss": 0.0106, "step": 37170 }, { "grad_norm": 0.09216554462909698, "learning_rate": 1.6924485602945983e-05, "loss": 0.0175, "step": 37180 }, { "grad_norm": 0.06226283311843872, "learning_rate": 1.6899692905133825e-05, "loss": 0.013, "step": 37190 }, { "grad_norm": 0.08221084624528885, "learning_rate": 1.6874914686520565e-05, "loss": 0.0141, "step": 37200 }, { "grad_norm": 0.06674236804246902, "learning_rate": 1.6850150957945e-05, "loss": 0.0118, "step": 37210 }, { "grad_norm": 0.10321240872144699, "learning_rate": 1.682540173023966e-05, "loss": 0.0139, "step": 37220 }, { "grad_norm": 0.09318981319665909, "learning_rate": 1.6800667014230665e-05, "loss": 0.0161, "step": 37230 }, { "grad_norm": 0.16357770562171936, "learning_rate": 1.677594682073785e-05, "loss": 0.0139, "step": 37240 }, { "grad_norm": 0.09720191359519958, "learning_rate": 1.6751241160574653e-05, "loss": 0.0135, "step": 37250 }, { "grad_norm": 0.07022228837013245, "learning_rate": 1.6726550044548156e-05, "loss": 0.0131, "step": 37260 }, { "grad_norm": 0.14807404577732086, "learning_rate": 1.6701873483459125e-05, "loss": 0.0159, "step": 37270 }, { "grad_norm": 0.0929894670844078, "learning_rate": 1.6677211488101906e-05, "loss": 0.0138, "step": 37280 }, { "grad_norm": 0.09653991460800171, "learning_rate": 1.6652564069264475e-05, "loss": 0.0139, "step": 37290 }, { "grad_norm": 0.08781261742115021, "learning_rate": 1.662793123772849e-05, "loss": 0.0116, "step": 37300 }, { "grad_norm": 0.09142547100782394, "learning_rate": 1.6603313004269168e-05, "loss": 0.0161, "step": 37310 }, { "grad_norm": 0.17990313470363617, "learning_rate": 1.657870937965536e-05, "loss": 0.0173, "step": 37320 }, { "grad_norm": 0.11011884361505508, "learning_rate": 1.655412037464952e-05, "loss": 0.0139, "step": 37330 }, { "grad_norm": 0.10564382374286652, "learning_rate": 1.652954600000773e-05, "loss": 0.0154, "step": 37340 }, { "grad_norm": 0.12904725968837738, "learning_rate": 1.650498626647965e-05, "loss": 0.0147, "step": 37350 }, { "grad_norm": 0.08559811860322952, "learning_rate": 1.6480441184808526e-05, "loss": 0.0134, "step": 37360 }, { "grad_norm": 0.12361647188663483, "learning_rate": 1.6455910765731242e-05, "loss": 0.0133, "step": 37370 }, { "grad_norm": 0.06939207762479782, "learning_rate": 1.6431395019978217e-05, "loss": 0.0121, "step": 37380 }, { "grad_norm": 0.06242825835943222, "learning_rate": 1.6406893958273467e-05, "loss": 0.0115, "step": 37390 }, { "grad_norm": 0.08588084578514099, "learning_rate": 1.6382407591334602e-05, "loss": 0.0129, "step": 37400 }, { "grad_norm": 0.08906853944063187, "learning_rate": 1.635793592987279e-05, "loss": 0.0129, "step": 37410 }, { "grad_norm": 0.08978231996297836, "learning_rate": 1.633347898459275e-05, "loss": 0.013, "step": 37420 }, { "grad_norm": 0.10001571476459503, "learning_rate": 1.6309036766192805e-05, "loss": 0.0126, "step": 37430 }, { "grad_norm": 0.0765836089849472, "learning_rate": 1.628460928536478e-05, "loss": 0.0261, "step": 37440 }, { "grad_norm": 0.13733835518360138, "learning_rate": 1.6260196552794128e-05, "loss": 0.0168, "step": 37450 }, { "grad_norm": 0.08080688863992691, "learning_rate": 1.6235798579159766e-05, "loss": 0.0124, "step": 37460 }, { "grad_norm": 0.09864375740289688, "learning_rate": 1.6211415375134233e-05, "loss": 0.0159, "step": 37470 }, { "grad_norm": 0.11301559209823608, "learning_rate": 1.618704695138353e-05, "loss": 0.0109, "step": 37480 }, { "grad_norm": 0.17962224781513214, "learning_rate": 1.6162693318567273e-05, "loss": 0.0141, "step": 37490 }, { "grad_norm": 0.08684860169887543, "learning_rate": 1.613835448733856e-05, "loss": 0.0137, "step": 37500 }, { "grad_norm": 0.07779046893119812, "learning_rate": 1.6114030468344006e-05, "loss": 0.014, "step": 37510 }, { "grad_norm": 0.09073345363140106, "learning_rate": 1.6089721272223755e-05, "loss": 0.0152, "step": 37520 }, { "grad_norm": 0.13786964118480682, "learning_rate": 1.606542690961151e-05, "loss": 0.0113, "step": 37530 }, { "grad_norm": 0.10425374656915665, "learning_rate": 1.6041147391134438e-05, "loss": 0.0155, "step": 37540 }, { "grad_norm": 0.06640756875276566, "learning_rate": 1.6016882727413196e-05, "loss": 0.0134, "step": 37550 }, { "grad_norm": 0.24438293278217316, "learning_rate": 1.5992632929062014e-05, "loss": 0.0126, "step": 37560 }, { "grad_norm": 0.04697669297456741, "learning_rate": 1.5968398006688557e-05, "loss": 0.0121, "step": 37570 }, { "grad_norm": 0.07072485983371735, "learning_rate": 1.594417797089399e-05, "loss": 0.0132, "step": 37580 }, { "grad_norm": 0.0663643628358841, "learning_rate": 1.5919972832273015e-05, "loss": 0.0128, "step": 37590 }, { "grad_norm": 0.09561315923929214, "learning_rate": 1.5895782601413767e-05, "loss": 0.014, "step": 37600 }, { "grad_norm": 0.05944562330842018, "learning_rate": 1.587160728889785e-05, "loss": 0.0124, "step": 37610 }, { "grad_norm": 0.08550975471735, "learning_rate": 1.58474469053004e-05, "loss": 0.014, "step": 37620 }, { "grad_norm": 0.09291244298219681, "learning_rate": 1.5823301461189976e-05, "loss": 0.0151, "step": 37630 }, { "grad_norm": 0.09848286956548691, "learning_rate": 1.5799170967128606e-05, "loss": 0.013, "step": 37640 }, { "grad_norm": 0.06029438599944115, "learning_rate": 1.5775055433671814e-05, "loss": 0.014, "step": 37650 }, { "grad_norm": 0.08273753523826599, "learning_rate": 1.5750954871368535e-05, "loss": 0.0181, "step": 37660 }, { "grad_norm": 0.06897709518671036, "learning_rate": 1.5726869290761158e-05, "loss": 0.0123, "step": 37670 }, { "grad_norm": 0.09472766518592834, "learning_rate": 1.5702798702385574e-05, "loss": 0.0131, "step": 37680 }, { "grad_norm": 0.05194862186908722, "learning_rate": 1.5678743116771038e-05, "loss": 0.0136, "step": 37690 }, { "grad_norm": 0.0621529296040535, "learning_rate": 1.5654702544440313e-05, "loss": 0.0147, "step": 37700 }, { "grad_norm": 0.09111890941858292, "learning_rate": 1.5630676995909532e-05, "loss": 0.0108, "step": 37710 }, { "grad_norm": 0.0657653957605362, "learning_rate": 1.560666648168832e-05, "loss": 0.016, "step": 37720 }, { "grad_norm": 0.05962085351347923, "learning_rate": 1.5582671012279663e-05, "loss": 0.0112, "step": 37730 }, { "grad_norm": 0.06800738722085953, "learning_rate": 1.5558690598179997e-05, "loss": 0.0146, "step": 37740 }, { "grad_norm": 0.12390004843473434, "learning_rate": 1.553472524987919e-05, "loss": 0.0173, "step": 37750 }, { "grad_norm": 0.10781806707382202, "learning_rate": 1.5510774977860485e-05, "loss": 0.0182, "step": 37760 }, { "grad_norm": 0.0960540771484375, "learning_rate": 1.5486839792600532e-05, "loss": 0.015, "step": 37770 }, { "grad_norm": 0.16163037717342377, "learning_rate": 1.546291970456942e-05, "loss": 0.0153, "step": 37780 }, { "grad_norm": 0.07738208770751953, "learning_rate": 1.5439014724230606e-05, "loss": 0.0132, "step": 37790 }, { "grad_norm": 0.10243168473243713, "learning_rate": 1.5415124862040907e-05, "loss": 0.0146, "step": 37800 }, { "grad_norm": 0.10699627548456192, "learning_rate": 1.53912501284506e-05, "loss": 0.0153, "step": 37810 }, { "grad_norm": 0.09114550054073334, "learning_rate": 1.5367390533903292e-05, "loss": 0.0153, "step": 37820 }, { "grad_norm": 0.0816657766699791, "learning_rate": 1.5343546088835964e-05, "loss": 0.0137, "step": 37830 }, { "grad_norm": 0.06334838271141052, "learning_rate": 1.531971680367901e-05, "loss": 0.0159, "step": 37840 }, { "grad_norm": 0.11326760798692703, "learning_rate": 1.529590268885616e-05, "loss": 0.015, "step": 37850 }, { "grad_norm": 0.10218831151723862, "learning_rate": 1.5272103754784517e-05, "loss": 0.0166, "step": 37860 }, { "grad_norm": 0.0699816346168518, "learning_rate": 1.5248320011874523e-05, "loss": 0.0151, "step": 37870 }, { "grad_norm": 0.05650510638952255, "learning_rate": 1.5224551470530035e-05, "loss": 0.0165, "step": 37880 }, { "grad_norm": 0.11198611557483673, "learning_rate": 1.5200798141148193e-05, "loss": 0.0164, "step": 37890 }, { "grad_norm": 0.1590384542942047, "learning_rate": 1.5177060034119505e-05, "loss": 0.0149, "step": 37900 }, { "grad_norm": 0.07921655476093292, "learning_rate": 1.5153337159827835e-05, "loss": 0.017, "step": 37910 }, { "grad_norm": 0.06662149727344513, "learning_rate": 1.5129629528650397e-05, "loss": 0.015, "step": 37920 }, { "grad_norm": 0.0985463336110115, "learning_rate": 1.5105937150957671e-05, "loss": 0.0163, "step": 37930 }, { "grad_norm": 0.06579524278640747, "learning_rate": 1.5082260037113548e-05, "loss": 0.0179, "step": 37940 }, { "grad_norm": 0.16490185260772705, "learning_rate": 1.5058598197475187e-05, "loss": 0.0175, "step": 37950 }, { "grad_norm": 0.10476762801408768, "learning_rate": 1.5034951642393064e-05, "loss": 0.0181, "step": 37960 }, { "grad_norm": 0.06870236992835999, "learning_rate": 1.5011320382211013e-05, "loss": 0.0155, "step": 37970 }, { "grad_norm": 0.09344038367271423, "learning_rate": 1.4987704427266136e-05, "loss": 0.0143, "step": 37980 }, { "grad_norm": 0.1436249315738678, "learning_rate": 1.4964103787888839e-05, "loss": 0.0142, "step": 37990 }, { "grad_norm": 0.09021659940481186, "learning_rate": 1.4940518474402882e-05, "loss": 0.0318, "step": 38000 }, { "grad_norm": 0.09064048528671265, "learning_rate": 1.4916948497125249e-05, "loss": 0.0171, "step": 38010 }, { "grad_norm": 0.07483850419521332, "learning_rate": 1.4893393866366267e-05, "loss": 0.0142, "step": 38020 }, { "grad_norm": 0.09386090189218521, "learning_rate": 1.4869854592429506e-05, "loss": 0.0147, "step": 38030 }, { "grad_norm": 0.050602756440639496, "learning_rate": 1.4846330685611875e-05, "loss": 0.0146, "step": 38040 }, { "grad_norm": 0.11612150073051453, "learning_rate": 1.482282215620352e-05, "loss": 0.0173, "step": 38050 }, { "grad_norm": 0.08375230431556702, "learning_rate": 1.4799329014487857e-05, "loss": 0.0158, "step": 38060 }, { "grad_norm": 0.08023037016391754, "learning_rate": 1.4775851270741603e-05, "loss": 0.0147, "step": 38070 }, { "grad_norm": 0.24137328565120697, "learning_rate": 1.4752388935234719e-05, "loss": 0.0201, "step": 38080 }, { "grad_norm": 0.12144016474485397, "learning_rate": 1.4728942018230401e-05, "loss": 0.0148, "step": 38090 }, { "grad_norm": 0.07178225368261337, "learning_rate": 1.4705510529985167e-05, "loss": 0.0152, "step": 38100 }, { "grad_norm": 0.09518378973007202, "learning_rate": 1.4682094480748726e-05, "loss": 0.0141, "step": 38110 }, { "grad_norm": 0.07558343559503555, "learning_rate": 1.4658693880764035e-05, "loss": 0.0137, "step": 38120 }, { "grad_norm": 0.07516822963953018, "learning_rate": 1.463530874026735e-05, "loss": 0.0148, "step": 38130 }, { "grad_norm": 0.17970076203346252, "learning_rate": 1.461193906948809e-05, "loss": 0.0164, "step": 38140 }, { "grad_norm": 0.13277186453342438, "learning_rate": 1.4588584878648975e-05, "loss": 0.0114, "step": 38150 }, { "grad_norm": 0.1316787153482437, "learning_rate": 1.4565246177965896e-05, "loss": 0.0151, "step": 38160 }, { "grad_norm": 0.0866377130150795, "learning_rate": 1.4541922977648015e-05, "loss": 0.0149, "step": 38170 }, { "grad_norm": 0.06761042028665543, "learning_rate": 1.451861528789768e-05, "loss": 0.015, "step": 38180 }, { "grad_norm": 0.1013309434056282, "learning_rate": 1.4495323118910454e-05, "loss": 0.0152, "step": 38190 }, { "grad_norm": 0.07243230938911438, "learning_rate": 1.4472046480875145e-05, "loss": 0.0146, "step": 38200 }, { "grad_norm": 0.16710655391216278, "learning_rate": 1.4448785383973745e-05, "loss": 0.0134, "step": 38210 }, { "grad_norm": 0.09845857322216034, "learning_rate": 1.4425539838381418e-05, "loss": 0.0143, "step": 38220 }, { "grad_norm": 0.11244633048772812, "learning_rate": 1.4402309854266594e-05, "loss": 0.0176, "step": 38230 }, { "grad_norm": 0.09032439440488815, "learning_rate": 1.4379095441790846e-05, "loss": 0.0134, "step": 38240 }, { "grad_norm": 0.08362163603305817, "learning_rate": 1.4355896611108921e-05, "loss": 0.0128, "step": 38250 }, { "grad_norm": 0.059555016458034515, "learning_rate": 1.433271337236881e-05, "loss": 0.0122, "step": 38260 }, { "grad_norm": 0.12363806366920471, "learning_rate": 1.4309545735711638e-05, "loss": 0.0143, "step": 38270 }, { "grad_norm": 0.09536012262105942, "learning_rate": 1.4286393711271696e-05, "loss": 0.0147, "step": 38280 }, { "grad_norm": 0.06992332637310028, "learning_rate": 1.4263257309176497e-05, "loss": 0.0117, "step": 38290 }, { "grad_norm": 0.07250940054655075, "learning_rate": 1.4240136539546678e-05, "loss": 0.0135, "step": 38300 }, { "grad_norm": 0.08023668080568314, "learning_rate": 1.4217031412496029e-05, "loss": 0.0125, "step": 38310 }, { "grad_norm": 0.09545648843050003, "learning_rate": 1.4193941938131554e-05, "loss": 0.0142, "step": 38320 }, { "grad_norm": 0.08351179212331772, "learning_rate": 1.4170868126553356e-05, "loss": 0.0111, "step": 38330 }, { "grad_norm": 0.06405124068260193, "learning_rate": 1.4147809987854682e-05, "loss": 0.0138, "step": 38340 }, { "grad_norm": 0.06624399870634079, "learning_rate": 1.4124767532121985e-05, "loss": 0.0111, "step": 38350 }, { "grad_norm": 0.044047776609659195, "learning_rate": 1.41017407694348e-05, "loss": 0.0101, "step": 38360 }, { "grad_norm": 0.054298285394907, "learning_rate": 1.4078729709865801e-05, "loss": 0.0095, "step": 38370 }, { "grad_norm": 0.039004046469926834, "learning_rate": 1.4055734363480833e-05, "loss": 0.0112, "step": 38380 }, { "grad_norm": 0.05033409968018532, "learning_rate": 1.4032754740338815e-05, "loss": 0.0124, "step": 38390 }, { "grad_norm": 0.06253651529550552, "learning_rate": 1.4009790850491844e-05, "loss": 0.0121, "step": 38400 }, { "grad_norm": 0.1823539435863495, "learning_rate": 1.3986842703985076e-05, "loss": 0.0128, "step": 38410 }, { "grad_norm": 0.32615602016448975, "learning_rate": 1.3963910310856842e-05, "loss": 0.0144, "step": 38420 }, { "grad_norm": 0.11545213311910629, "learning_rate": 1.394099368113853e-05, "loss": 0.0148, "step": 38430 }, { "grad_norm": 0.08955011516809464, "learning_rate": 1.3918092824854644e-05, "loss": 0.0153, "step": 38440 }, { "grad_norm": 0.07666780054569244, "learning_rate": 1.3895207752022821e-05, "loss": 0.0146, "step": 38450 }, { "grad_norm": 0.06359293311834335, "learning_rate": 1.3872338472653756e-05, "loss": 0.0123, "step": 38460 }, { "grad_norm": 0.06866833567619324, "learning_rate": 1.3849484996751233e-05, "loss": 0.0134, "step": 38470 }, { "grad_norm": 0.07811441272497177, "learning_rate": 1.382664733431217e-05, "loss": 0.0137, "step": 38480 }, { "grad_norm": 0.05255579575896263, "learning_rate": 1.3803825495326522e-05, "loss": 0.0135, "step": 38490 }, { "grad_norm": 0.08888687193393707, "learning_rate": 1.3781019489777324e-05, "loss": 0.0137, "step": 38500 }, { "grad_norm": 0.08120469748973846, "learning_rate": 1.375822932764072e-05, "loss": 0.0107, "step": 38510 }, { "grad_norm": 0.06399614363908768, "learning_rate": 1.373545501888589e-05, "loss": 0.0132, "step": 38520 }, { "grad_norm": 0.0497673898935318, "learning_rate": 1.3712696573475092e-05, "loss": 0.0112, "step": 38530 }, { "grad_norm": 0.10645662993192673, "learning_rate": 1.368995400136363e-05, "loss": 0.0132, "step": 38540 }, { "grad_norm": 0.07501479238271713, "learning_rate": 1.3667227312499903e-05, "loss": 0.0114, "step": 38550 }, { "grad_norm": 0.08925170451402664, "learning_rate": 1.3644516516825323e-05, "loss": 0.0137, "step": 38560 }, { "grad_norm": 0.055429503321647644, "learning_rate": 1.3621821624274356e-05, "loss": 0.0152, "step": 38570 }, { "grad_norm": 0.11202526092529297, "learning_rate": 1.3599142644774537e-05, "loss": 0.0146, "step": 38580 }, { "grad_norm": 0.10380654036998749, "learning_rate": 1.357647958824641e-05, "loss": 0.0105, "step": 38590 }, { "grad_norm": 0.07377313822507858, "learning_rate": 1.3553832464603555e-05, "loss": 0.0138, "step": 38600 }, { "grad_norm": 0.04974083602428436, "learning_rate": 1.3531201283752625e-05, "loss": 0.0119, "step": 38610 }, { "grad_norm": 0.09491181373596191, "learning_rate": 1.350858605559323e-05, "loss": 0.0105, "step": 38620 }, { "grad_norm": 0.0706641897559166, "learning_rate": 1.348598679001808e-05, "loss": 0.0163, "step": 38630 }, { "grad_norm": 0.08597110956907272, "learning_rate": 1.3463403496912819e-05, "loss": 0.0123, "step": 38640 }, { "grad_norm": 0.058488357812166214, "learning_rate": 1.3440836186156187e-05, "loss": 0.0136, "step": 38650 }, { "grad_norm": 0.07382983714342117, "learning_rate": 1.3418284867619852e-05, "loss": 0.0101, "step": 38660 }, { "grad_norm": 0.06215815618634224, "learning_rate": 1.3395749551168567e-05, "loss": 0.013, "step": 38670 }, { "grad_norm": 0.08260812610387802, "learning_rate": 1.3373230246660024e-05, "loss": 0.0146, "step": 38680 }, { "grad_norm": 0.0702468678355217, "learning_rate": 1.335072696394492e-05, "loss": 0.0109, "step": 38690 }, { "grad_norm": 0.10016676783561707, "learning_rate": 1.3328239712866986e-05, "loss": 0.0132, "step": 38700 }, { "grad_norm": 0.11461073905229568, "learning_rate": 1.3305768503262889e-05, "loss": 0.0143, "step": 38710 }, { "grad_norm": 0.1343158781528473, "learning_rate": 1.3283313344962305e-05, "loss": 0.0161, "step": 38720 }, { "grad_norm": 0.053582653403282166, "learning_rate": 1.3260874247787863e-05, "loss": 0.0123, "step": 38730 }, { "grad_norm": 0.07445257157087326, "learning_rate": 1.3238451221555226e-05, "loss": 0.0132, "step": 38740 }, { "grad_norm": 0.08006022870540619, "learning_rate": 1.3216044276072965e-05, "loss": 0.0127, "step": 38750 }, { "grad_norm": 0.10676088184118271, "learning_rate": 1.3193653421142622e-05, "loss": 0.0143, "step": 38760 }, { "grad_norm": 0.10372733324766159, "learning_rate": 1.3171278666558756e-05, "loss": 0.0129, "step": 38770 }, { "grad_norm": 0.07859113812446594, "learning_rate": 1.3148920022108819e-05, "loss": 0.0148, "step": 38780 }, { "grad_norm": 0.09672097861766815, "learning_rate": 1.312657749757324e-05, "loss": 0.0125, "step": 38790 }, { "grad_norm": 0.07053747773170471, "learning_rate": 1.3104251102725412e-05, "loss": 0.0127, "step": 38800 }, { "grad_norm": 0.07408404350280762, "learning_rate": 1.3081940847331659e-05, "loss": 0.0152, "step": 38810 }, { "grad_norm": 0.043015286326408386, "learning_rate": 1.3059646741151222e-05, "loss": 0.0123, "step": 38820 }, { "grad_norm": 0.23877030611038208, "learning_rate": 1.3037368793936334e-05, "loss": 0.0132, "step": 38830 }, { "grad_norm": 0.08109124004840851, "learning_rate": 1.3015107015432104e-05, "loss": 0.0134, "step": 38840 }, { "grad_norm": 0.0787353664636612, "learning_rate": 1.2992861415376584e-05, "loss": 0.0158, "step": 38850 }, { "grad_norm": 0.10292460024356842, "learning_rate": 1.2970632003500782e-05, "loss": 0.0143, "step": 38860 }, { "grad_norm": 0.0762653723359108, "learning_rate": 1.294841878952856e-05, "loss": 0.0137, "step": 38870 }, { "grad_norm": 0.13310636579990387, "learning_rate": 1.2926221783176779e-05, "loss": 0.0138, "step": 38880 }, { "grad_norm": 0.15879805386066437, "learning_rate": 1.290404099415512e-05, "loss": 0.0156, "step": 38890 }, { "grad_norm": 0.07704561948776245, "learning_rate": 1.2881876432166246e-05, "loss": 0.016, "step": 38900 }, { "grad_norm": 0.1853618174791336, "learning_rate": 1.2859728106905678e-05, "loss": 0.0159, "step": 38910 }, { "grad_norm": 0.06972210854291916, "learning_rate": 1.2837596028061832e-05, "loss": 0.0166, "step": 38920 }, { "grad_norm": 0.121426522731781, "learning_rate": 1.2815480205316055e-05, "loss": 0.0144, "step": 38930 }, { "grad_norm": 0.057451289147138596, "learning_rate": 1.279338064834255e-05, "loss": 0.0167, "step": 38940 }, { "grad_norm": 0.11364437639713287, "learning_rate": 1.27712973668084e-05, "loss": 0.0136, "step": 38950 }, { "grad_norm": 0.081961490213871, "learning_rate": 1.2749230370373605e-05, "loss": 0.0135, "step": 38960 }, { "grad_norm": 0.11559350788593292, "learning_rate": 1.2727179668691009e-05, "loss": 0.0131, "step": 38970 }, { "grad_norm": 0.12512964010238647, "learning_rate": 1.2705145271406327e-05, "loss": 0.0173, "step": 38980 }, { "grad_norm": 0.0776362270116806, "learning_rate": 1.2683127188158172e-05, "loss": 0.0141, "step": 38990 }, { "grad_norm": 0.08010167628526688, "learning_rate": 1.2661125428577997e-05, "loss": 0.0116, "step": 39000 }, { "grad_norm": 0.0706515684723854, "learning_rate": 1.2639140002290101e-05, "loss": 0.0134, "step": 39010 }, { "grad_norm": 0.08601407706737518, "learning_rate": 1.261717091891168e-05, "loss": 0.0141, "step": 39020 }, { "grad_norm": 0.06868116557598114, "learning_rate": 1.2595218188052755e-05, "loss": 0.0153, "step": 39030 }, { "grad_norm": 0.10469871014356613, "learning_rate": 1.2573281819316174e-05, "loss": 0.0131, "step": 39040 }, { "grad_norm": 0.06366486847400665, "learning_rate": 1.2551361822297674e-05, "loss": 0.0146, "step": 39050 }, { "grad_norm": 0.07701673358678818, "learning_rate": 1.2529458206585804e-05, "loss": 0.0145, "step": 39060 }, { "grad_norm": 0.09253714978694916, "learning_rate": 1.250757098176194e-05, "loss": 0.0147, "step": 39070 }, { "grad_norm": 0.07232977449893951, "learning_rate": 1.2485700157400294e-05, "loss": 0.0121, "step": 39080 }, { "grad_norm": 0.07158397883176804, "learning_rate": 1.246384574306792e-05, "loss": 0.0153, "step": 39090 }, { "grad_norm": 0.11524984240531921, "learning_rate": 1.2442007748324664e-05, "loss": 0.0139, "step": 39100 }, { "grad_norm": 0.1022915467619896, "learning_rate": 1.2420186182723237e-05, "loss": 0.0129, "step": 39110 }, { "grad_norm": 0.07988275587558746, "learning_rate": 1.23983810558091e-05, "loss": 0.0099, "step": 39120 }, { "grad_norm": 0.06317409873008728, "learning_rate": 1.2376592377120582e-05, "loss": 0.0128, "step": 39130 }, { "grad_norm": 0.10925856977701187, "learning_rate": 1.2354820156188768e-05, "loss": 0.013, "step": 39140 }, { "grad_norm": 0.08623570948839188, "learning_rate": 1.2333064402537597e-05, "loss": 0.0156, "step": 39150 }, { "grad_norm": 0.046241387724876404, "learning_rate": 1.2311325125683754e-05, "loss": 0.0132, "step": 39160 }, { "grad_norm": 0.10277967154979706, "learning_rate": 1.228960233513673e-05, "loss": 0.0137, "step": 39170 }, { "grad_norm": 0.05096100643277168, "learning_rate": 1.2267896040398836e-05, "loss": 0.0115, "step": 39180 }, { "grad_norm": 0.09405959397554398, "learning_rate": 1.2246206250965125e-05, "loss": 0.0131, "step": 39190 }, { "grad_norm": 0.09208636730909348, "learning_rate": 1.2224532976323444e-05, "loss": 0.0134, "step": 39200 }, { "grad_norm": 0.10660763084888458, "learning_rate": 1.220287622595444e-05, "loss": 0.0157, "step": 39210 }, { "grad_norm": 0.08736369758844376, "learning_rate": 1.2181236009331497e-05, "loss": 0.0126, "step": 39220 }, { "grad_norm": 0.0904604122042656, "learning_rate": 1.2159612335920784e-05, "loss": 0.0139, "step": 39230 }, { "grad_norm": 0.09487153589725494, "learning_rate": 1.2138005215181219e-05, "loss": 0.0145, "step": 39240 }, { "grad_norm": 0.10433145612478256, "learning_rate": 1.2116414656564512e-05, "loss": 0.0117, "step": 39250 }, { "grad_norm": 0.046328816562891006, "learning_rate": 1.2094840669515095e-05, "loss": 0.0115, "step": 39260 }, { "grad_norm": 0.11020752042531967, "learning_rate": 1.2073283263470152e-05, "loss": 0.0128, "step": 39270 }, { "grad_norm": 0.09442154318094254, "learning_rate": 1.2051742447859653e-05, "loss": 0.0161, "step": 39280 }, { "grad_norm": 0.12056183069944382, "learning_rate": 1.203021823210626e-05, "loss": 0.0189, "step": 39290 }, { "grad_norm": 0.0867287665605545, "learning_rate": 1.2008710625625397e-05, "loss": 0.0152, "step": 39300 }, { "grad_norm": 0.0980941504240036, "learning_rate": 1.1987219637825236e-05, "loss": 0.012, "step": 39310 }, { "grad_norm": 0.08843007683753967, "learning_rate": 1.1965745278106656e-05, "loss": 0.0137, "step": 39320 }, { "grad_norm": 0.07520236819982529, "learning_rate": 1.1944287555863259e-05, "loss": 0.0126, "step": 39330 }, { "grad_norm": 0.1065099686384201, "learning_rate": 1.1922846480481409e-05, "loss": 0.0132, "step": 39340 }, { "grad_norm": 0.1179264560341835, "learning_rate": 1.1901422061340133e-05, "loss": 0.0164, "step": 39350 }, { "grad_norm": 0.06811780482530594, "learning_rate": 1.1880014307811227e-05, "loss": 0.0131, "step": 39360 }, { "grad_norm": 0.08087584376335144, "learning_rate": 1.1858623229259142e-05, "loss": 0.014, "step": 39370 }, { "grad_norm": 0.0633472427725792, "learning_rate": 1.1837248835041092e-05, "loss": 0.0132, "step": 39380 }, { "grad_norm": 0.07765570282936096, "learning_rate": 1.1815891134506935e-05, "loss": 0.0135, "step": 39390 }, { "grad_norm": 0.0710654929280281, "learning_rate": 1.1794550136999277e-05, "loss": 0.0154, "step": 39400 }, { "grad_norm": 0.12188739329576492, "learning_rate": 1.1773225851853387e-05, "loss": 0.0258, "step": 39410 }, { "grad_norm": 0.07887330651283264, "learning_rate": 1.1751918288397235e-05, "loss": 0.0108, "step": 39420 }, { "grad_norm": 0.10598951578140259, "learning_rate": 1.1730627455951449e-05, "loss": 0.0118, "step": 39430 }, { "grad_norm": 0.12017720937728882, "learning_rate": 1.17093533638294e-05, "loss": 0.0143, "step": 39440 }, { "grad_norm": 0.1300533413887024, "learning_rate": 1.168809602133708e-05, "loss": 0.0122, "step": 39450 }, { "grad_norm": 0.079989954829216, "learning_rate": 1.1666855437773162e-05, "loss": 0.0118, "step": 39460 }, { "grad_norm": 0.12201790511608124, "learning_rate": 1.1645631622429026e-05, "loss": 0.0118, "step": 39470 }, { "grad_norm": 0.09303494542837143, "learning_rate": 1.1624424584588673e-05, "loss": 0.0124, "step": 39480 }, { "grad_norm": 0.12783469259738922, "learning_rate": 1.1603234333528768e-05, "loss": 0.0158, "step": 39490 }, { "grad_norm": 0.07878249138593674, "learning_rate": 1.1582060878518674e-05, "loss": 0.0117, "step": 39500 }, { "grad_norm": 0.06587151437997818, "learning_rate": 1.1560904228820375e-05, "loss": 0.0126, "step": 39510 }, { "grad_norm": 0.12656429409980774, "learning_rate": 1.1539764393688478e-05, "loss": 0.0161, "step": 39520 }, { "grad_norm": 0.08011612296104431, "learning_rate": 1.151864138237031e-05, "loss": 0.0135, "step": 39530 }, { "grad_norm": 0.12758764624595642, "learning_rate": 1.1497535204105769e-05, "loss": 0.0149, "step": 39540 }, { "grad_norm": 0.07453663647174835, "learning_rate": 1.14764458681274e-05, "loss": 0.0153, "step": 39550 }, { "grad_norm": 0.17841893434524536, "learning_rate": 1.1455373383660434e-05, "loss": 0.015, "step": 39560 }, { "grad_norm": 0.08204679936170578, "learning_rate": 1.1434317759922664e-05, "loss": 0.01, "step": 39570 }, { "grad_norm": 0.11075787246227264, "learning_rate": 1.1413279006124528e-05, "loss": 0.0143, "step": 39580 }, { "grad_norm": 0.0911034494638443, "learning_rate": 1.1392257131469119e-05, "loss": 0.0168, "step": 39590 }, { "grad_norm": 0.09082074463367462, "learning_rate": 1.1371252145152095e-05, "loss": 0.0136, "step": 39600 }, { "grad_norm": 0.09433219581842422, "learning_rate": 1.1350264056361776e-05, "loss": 0.0123, "step": 39610 }, { "grad_norm": 0.06679043173789978, "learning_rate": 1.1329292874279029e-05, "loss": 0.0134, "step": 39620 }, { "grad_norm": 0.06529246270656586, "learning_rate": 1.1308338608077401e-05, "loss": 0.0129, "step": 39630 }, { "grad_norm": 0.07605665922164917, "learning_rate": 1.1287401266922981e-05, "loss": 0.0153, "step": 39640 }, { "grad_norm": 0.14427129924297333, "learning_rate": 1.126648085997446e-05, "loss": 0.0166, "step": 39650 }, { "grad_norm": 0.08680807054042816, "learning_rate": 1.124557739638316e-05, "loss": 0.0116, "step": 39660 }, { "grad_norm": 0.11112569272518158, "learning_rate": 1.1224690885292955e-05, "loss": 0.0124, "step": 39670 }, { "grad_norm": 0.12074945867061615, "learning_rate": 1.1203821335840303e-05, "loss": 0.0135, "step": 39680 }, { "grad_norm": 0.06283034384250641, "learning_rate": 1.1182968757154278e-05, "loss": 0.0114, "step": 39690 }, { "grad_norm": 0.08441851288080215, "learning_rate": 1.1162133158356492e-05, "loss": 0.0119, "step": 39700 }, { "grad_norm": 0.07896436005830765, "learning_rate": 1.1141314548561133e-05, "loss": 0.0134, "step": 39710 }, { "grad_norm": 0.12577266991138458, "learning_rate": 1.1120512936874994e-05, "loss": 0.0124, "step": 39720 }, { "grad_norm": 0.06761026382446289, "learning_rate": 1.109972833239739e-05, "loss": 0.0182, "step": 39730 }, { "grad_norm": 0.0912456214427948, "learning_rate": 1.1078960744220218e-05, "loss": 0.014, "step": 39740 }, { "grad_norm": 0.10725877434015274, "learning_rate": 1.1058210181427914e-05, "loss": 0.0136, "step": 39750 }, { "grad_norm": 0.0725492388010025, "learning_rate": 1.10374766530975e-05, "loss": 0.0149, "step": 39760 }, { "grad_norm": 0.07519302517175674, "learning_rate": 1.1016760168298518e-05, "loss": 0.0149, "step": 39770 }, { "grad_norm": 0.0752166137099266, "learning_rate": 1.0996060736093044e-05, "loss": 0.0148, "step": 39780 }, { "grad_norm": 0.11205363273620605, "learning_rate": 1.0975378365535749e-05, "loss": 0.013, "step": 39790 }, { "grad_norm": 0.11059927940368652, "learning_rate": 1.0954713065673777e-05, "loss": 0.0172, "step": 39800 }, { "grad_norm": 0.10956163704395294, "learning_rate": 1.0934064845546832e-05, "loss": 0.0123, "step": 39810 }, { "grad_norm": 0.06998936831951141, "learning_rate": 1.0913433714187166e-05, "loss": 0.013, "step": 39820 }, { "grad_norm": 0.13359221816062927, "learning_rate": 1.089281968061951e-05, "loss": 0.0129, "step": 39830 }, { "grad_norm": 0.10767863690853119, "learning_rate": 1.0872222753861172e-05, "loss": 0.0151, "step": 39840 }, { "grad_norm": 0.08455020934343338, "learning_rate": 1.0851642942921925e-05, "loss": 0.0135, "step": 39850 }, { "grad_norm": 0.12428200989961624, "learning_rate": 1.0831080256804093e-05, "loss": 0.0142, "step": 39860 }, { "grad_norm": 0.048102084547281265, "learning_rate": 1.0810534704502478e-05, "loss": 0.0154, "step": 39870 }, { "grad_norm": 0.08365139365196228, "learning_rate": 1.0790006295004423e-05, "loss": 0.0154, "step": 39880 }, { "grad_norm": 0.06194759160280228, "learning_rate": 1.076949503728974e-05, "loss": 0.0143, "step": 39890 }, { "grad_norm": 0.07926459610462189, "learning_rate": 1.074900094033074e-05, "loss": 0.0113, "step": 39900 }, { "grad_norm": 0.058454450219869614, "learning_rate": 1.0728524013092256e-05, "loss": 0.0149, "step": 39910 }, { "grad_norm": 0.15443843603134155, "learning_rate": 1.0708064264531581e-05, "loss": 0.0146, "step": 39920 }, { "grad_norm": 0.05748658627271652, "learning_rate": 1.068762170359851e-05, "loss": 0.0147, "step": 39930 }, { "grad_norm": 0.10460906475782394, "learning_rate": 1.0667196339235297e-05, "loss": 0.0136, "step": 39940 }, { "grad_norm": 0.07065197825431824, "learning_rate": 1.0646788180376716e-05, "loss": 0.0123, "step": 39950 }, { "grad_norm": 0.06509845703840256, "learning_rate": 1.0626397235949976e-05, "loss": 0.013, "step": 39960 }, { "grad_norm": 0.06683387607336044, "learning_rate": 1.0606023514874751e-05, "loss": 0.0145, "step": 39970 }, { "grad_norm": 0.1063448116183281, "learning_rate": 1.0585667026063233e-05, "loss": 0.0153, "step": 39980 }, { "grad_norm": 0.08093943446874619, "learning_rate": 1.0565327778420025e-05, "loss": 0.0116, "step": 39990 }, { "grad_norm": 0.08807086944580078, "learning_rate": 1.0545005780842193e-05, "loss": 0.0132, "step": 40000 }, { "grad_norm": 0.08023342490196228, "learning_rate": 1.0524701042219299e-05, "loss": 0.0134, "step": 40010 }, { "grad_norm": 0.0919448658823967, "learning_rate": 1.0504413571433303e-05, "loss": 0.0125, "step": 40020 }, { "grad_norm": 0.18101845681667328, "learning_rate": 1.0484143377358624e-05, "loss": 0.0138, "step": 40030 }, { "grad_norm": 0.08537650853395462, "learning_rate": 1.046389046886217e-05, "loss": 0.0126, "step": 40040 }, { "grad_norm": 0.103840172290802, "learning_rate": 1.044365485480322e-05, "loss": 0.0146, "step": 40050 }, { "grad_norm": 0.08164190500974655, "learning_rate": 1.0423436544033522e-05, "loss": 0.0122, "step": 40060 }, { "grad_norm": 0.12038128077983856, "learning_rate": 1.0403235545397272e-05, "loss": 0.015, "step": 40070 }, { "grad_norm": 0.10267791897058487, "learning_rate": 1.0383051867731043e-05, "loss": 0.012, "step": 40080 }, { "grad_norm": 0.11203201860189438, "learning_rate": 1.0362885519863896e-05, "loss": 0.0135, "step": 40090 }, { "grad_norm": 0.05561299994587898, "learning_rate": 1.034273651061724e-05, "loss": 0.0135, "step": 40100 }, { "grad_norm": 0.04633539915084839, "learning_rate": 1.0322604848804973e-05, "loss": 0.0124, "step": 40110 }, { "grad_norm": 0.07130225747823715, "learning_rate": 1.030249054323335e-05, "loss": 0.0133, "step": 40120 }, { "grad_norm": 0.042183768004179, "learning_rate": 1.0282393602701041e-05, "loss": 0.013, "step": 40130 }, { "grad_norm": 0.1272534430027008, "learning_rate": 1.026231403599915e-05, "loss": 0.0141, "step": 40140 }, { "grad_norm": 0.17826411128044128, "learning_rate": 1.0242251851911155e-05, "loss": 0.0149, "step": 40150 }, { "grad_norm": 0.05435154214501381, "learning_rate": 1.0222207059212918e-05, "loss": 0.0139, "step": 40160 }, { "grad_norm": 0.10198378562927246, "learning_rate": 1.0202179666672746e-05, "loss": 0.012, "step": 40170 }, { "grad_norm": 0.0630047619342804, "learning_rate": 1.0182169683051284e-05, "loss": 0.0129, "step": 40180 }, { "grad_norm": 0.11660720407962799, "learning_rate": 1.0162177117101567e-05, "loss": 0.0147, "step": 40190 }, { "grad_norm": 0.039264533668756485, "learning_rate": 1.0142201977569044e-05, "loss": 0.0115, "step": 40200 }, { "grad_norm": 0.08441238850355148, "learning_rate": 1.0122244273191517e-05, "loss": 0.0141, "step": 40210 }, { "grad_norm": 0.14233188331127167, "learning_rate": 1.0102304012699137e-05, "loss": 0.0135, "step": 40220 }, { "grad_norm": 0.06070315092802048, "learning_rate": 1.0082381204814484e-05, "loss": 0.0127, "step": 40230 }, { "grad_norm": 0.08139443397521973, "learning_rate": 1.0062475858252463e-05, "loss": 0.0141, "step": 40240 }, { "grad_norm": 0.06454426795244217, "learning_rate": 1.0042587981720326e-05, "loss": 0.0114, "step": 40250 }, { "grad_norm": 0.18190917372703552, "learning_rate": 1.002271758391774e-05, "loss": 0.0139, "step": 40260 }, { "grad_norm": 0.08103424310684204, "learning_rate": 1.0002864673536676e-05, "loss": 0.0149, "step": 40270 }, { "grad_norm": 0.06289529800415039, "learning_rate": 9.983029259261473e-06, "loss": 0.0125, "step": 40280 }, { "grad_norm": 0.12133010476827621, "learning_rate": 9.963211349768798e-06, "loss": 0.0144, "step": 40290 }, { "grad_norm": 0.054635412991046906, "learning_rate": 9.94341095372771e-06, "loss": 0.0125, "step": 40300 }, { "grad_norm": 0.08342691510915756, "learning_rate": 9.923628079799546e-06, "loss": 0.0135, "step": 40310 }, { "grad_norm": 0.06882734596729279, "learning_rate": 9.903862736638031e-06, "loss": 0.0114, "step": 40320 }, { "grad_norm": 0.07385941594839096, "learning_rate": 9.884114932889171e-06, "loss": 0.0128, "step": 40330 }, { "grad_norm": 0.11318623274564743, "learning_rate": 9.864384677191357e-06, "loss": 0.0113, "step": 40340 }, { "grad_norm": 0.04717416688799858, "learning_rate": 9.844671978175247e-06, "loss": 0.0119, "step": 40350 }, { "grad_norm": 0.07567162811756134, "learning_rate": 9.824976844463862e-06, "loss": 0.0111, "step": 40360 }, { "grad_norm": 0.06945750117301941, "learning_rate": 9.805299284672525e-06, "loss": 0.0119, "step": 40370 }, { "grad_norm": 0.07392650097608566, "learning_rate": 9.785639307408845e-06, "loss": 0.0126, "step": 40380 }, { "grad_norm": 0.07365402579307556, "learning_rate": 9.765996921272792e-06, "loss": 0.0107, "step": 40390 }, { "grad_norm": 0.059340883046388626, "learning_rate": 9.746372134856595e-06, "loss": 0.0116, "step": 40400 }, { "grad_norm": 0.08973202109336853, "learning_rate": 9.726764956744805e-06, "loss": 0.0111, "step": 40410 }, { "grad_norm": 0.06771227717399597, "learning_rate": 9.707175395514278e-06, "loss": 0.0143, "step": 40420 }, { "grad_norm": 0.08599057793617249, "learning_rate": 9.68760345973415e-06, "loss": 0.0147, "step": 40430 }, { "grad_norm": 0.10930511355400085, "learning_rate": 9.668049157965848e-06, "loss": 0.0116, "step": 40440 }, { "grad_norm": 0.07344004511833191, "learning_rate": 9.648512498763074e-06, "loss": 0.0118, "step": 40450 }, { "grad_norm": 0.046078070998191833, "learning_rate": 9.62899349067185e-06, "loss": 0.0136, "step": 40460 }, { "grad_norm": 0.06177305430173874, "learning_rate": 9.609492142230442e-06, "loss": 0.0128, "step": 40470 }, { "grad_norm": 0.08352632075548172, "learning_rate": 9.59000846196939e-06, "loss": 0.017, "step": 40480 }, { "grad_norm": 0.09007356315851212, "learning_rate": 9.570542458411552e-06, "loss": 0.013, "step": 40490 }, { "grad_norm": 0.08235818892717361, "learning_rate": 9.551094140071993e-06, "loss": 0.0123, "step": 40500 }, { "grad_norm": 0.06653550267219543, "learning_rate": 9.53166351545806e-06, "loss": 0.0143, "step": 40510 }, { "grad_norm": 0.11553967744112015, "learning_rate": 9.512250593069393e-06, "loss": 0.0139, "step": 40520 }, { "grad_norm": 0.07334969192743301, "learning_rate": 9.492855381397859e-06, "loss": 0.0141, "step": 40530 }, { "grad_norm": 0.06594078242778778, "learning_rate": 9.473477888927556e-06, "loss": 0.011, "step": 40540 }, { "grad_norm": 0.06073692440986633, "learning_rate": 9.454118124134892e-06, "loss": 0.0112, "step": 40550 }, { "grad_norm": 0.06980478763580322, "learning_rate": 9.434776095488463e-06, "loss": 0.0143, "step": 40560 }, { "grad_norm": 0.09828543663024902, "learning_rate": 9.41545181144915e-06, "loss": 0.0106, "step": 40570 }, { "grad_norm": 0.03668816760182381, "learning_rate": 9.396145280470025e-06, "loss": 0.013, "step": 40580 }, { "grad_norm": 0.0638221725821495, "learning_rate": 9.376856510996445e-06, "loss": 0.0127, "step": 40590 }, { "grad_norm": 0.07520058751106262, "learning_rate": 9.357585511465955e-06, "loss": 0.0124, "step": 40600 }, { "grad_norm": 0.06838450580835342, "learning_rate": 9.338332290308367e-06, "loss": 0.012, "step": 40610 }, { "grad_norm": 0.08456502854824066, "learning_rate": 9.319096855945675e-06, "loss": 0.0114, "step": 40620 }, { "grad_norm": 0.06918766349554062, "learning_rate": 9.299879216792118e-06, "loss": 0.0108, "step": 40630 }, { "grad_norm": 0.12580740451812744, "learning_rate": 9.280679381254132e-06, "loss": 0.0123, "step": 40640 }, { "grad_norm": 0.05989726632833481, "learning_rate": 9.261497357730397e-06, "loss": 0.0116, "step": 40650 }, { "grad_norm": 0.0870589092373848, "learning_rate": 9.242333154611777e-06, "loss": 0.0147, "step": 40660 }, { "grad_norm": 0.08053659647703171, "learning_rate": 9.223186780281323e-06, "loss": 0.0145, "step": 40670 }, { "grad_norm": 0.0517120398581028, "learning_rate": 9.20405824311435e-06, "loss": 0.0086, "step": 40680 }, { "grad_norm": 0.09440837800502777, "learning_rate": 9.184947551478301e-06, "loss": 0.0146, "step": 40690 }, { "grad_norm": 0.14211304485797882, "learning_rate": 9.165854713732835e-06, "loss": 0.012, "step": 40700 }, { "grad_norm": 0.08164086192846298, "learning_rate": 9.146779738229838e-06, "loss": 0.0111, "step": 40710 }, { "grad_norm": 0.08974028378725052, "learning_rate": 9.12772263331334e-06, "loss": 0.0118, "step": 40720 }, { "grad_norm": 0.08093539625406265, "learning_rate": 9.10868340731954e-06, "loss": 0.0126, "step": 40730 }, { "grad_norm": 0.08835148811340332, "learning_rate": 9.089662068576887e-06, "loss": 0.0146, "step": 40740 }, { "grad_norm": 0.08060979843139648, "learning_rate": 9.07065862540593e-06, "loss": 0.0144, "step": 40750 }, { "grad_norm": 0.06291349977254868, "learning_rate": 9.051673086119422e-06, "loss": 0.0114, "step": 40760 }, { "grad_norm": 0.048201512545347214, "learning_rate": 9.0327054590223e-06, "loss": 0.0126, "step": 40770 }, { "grad_norm": 0.14948008954524994, "learning_rate": 9.01375575241164e-06, "loss": 0.0156, "step": 40780 }, { "grad_norm": 0.12602680921554565, "learning_rate": 8.994823974576672e-06, "loss": 0.0119, "step": 40790 }, { "grad_norm": 0.09854954481124878, "learning_rate": 8.975910133798816e-06, "loss": 0.0136, "step": 40800 }, { "grad_norm": 0.08633670955896378, "learning_rate": 8.957014238351635e-06, "loss": 0.0153, "step": 40810 }, { "grad_norm": 0.10714085400104523, "learning_rate": 8.938136296500826e-06, "loss": 0.0129, "step": 40820 }, { "grad_norm": 0.09665337949991226, "learning_rate": 8.919276316504227e-06, "loss": 0.0123, "step": 40830 }, { "grad_norm": 0.07484319061040878, "learning_rate": 8.900434306611866e-06, "loss": 0.0154, "step": 40840 }, { "grad_norm": 0.07098758220672607, "learning_rate": 8.881610275065861e-06, "loss": 0.0116, "step": 40850 }, { "grad_norm": 0.0824514627456665, "learning_rate": 8.862804230100463e-06, "loss": 0.0128, "step": 40860 }, { "grad_norm": 0.07655462622642517, "learning_rate": 8.844016179942116e-06, "loss": 0.0134, "step": 40870 }, { "grad_norm": 0.15096943080425262, "learning_rate": 8.825246132809329e-06, "loss": 0.0138, "step": 40880 }, { "grad_norm": 0.055588796734809875, "learning_rate": 8.80649409691275e-06, "loss": 0.0154, "step": 40890 }, { "grad_norm": 0.07998518645763397, "learning_rate": 8.78776008045517e-06, "loss": 0.0143, "step": 40900 }, { "grad_norm": 0.0705200582742691, "learning_rate": 8.76904409163149e-06, "loss": 0.0145, "step": 40910 }, { "grad_norm": 0.06581218540668488, "learning_rate": 8.750346138628696e-06, "loss": 0.0129, "step": 40920 }, { "grad_norm": 0.08660058677196503, "learning_rate": 8.73166622962593e-06, "loss": 0.013, "step": 40930 }, { "grad_norm": 0.06365685164928436, "learning_rate": 8.713004372794415e-06, "loss": 0.011, "step": 40940 }, { "grad_norm": 0.06661161780357361, "learning_rate": 8.694360576297461e-06, "loss": 0.0133, "step": 40950 }, { "grad_norm": 0.06185680627822876, "learning_rate": 8.675734848290529e-06, "loss": 0.0122, "step": 40960 }, { "grad_norm": 0.06782277673482895, "learning_rate": 8.657127196921133e-06, "loss": 0.0135, "step": 40970 }, { "grad_norm": 0.05388281121850014, "learning_rate": 8.638537630328885e-06, "loss": 0.011, "step": 40980 }, { "grad_norm": 0.1676572561264038, "learning_rate": 8.619966156645487e-06, "loss": 0.0119, "step": 40990 }, { "grad_norm": 0.1561310738325119, "learning_rate": 8.601412783994756e-06, "loss": 0.0121, "step": 41000 }, { "grad_norm": 0.07881902158260345, "learning_rate": 8.582877520492538e-06, "loss": 0.0107, "step": 41010 }, { "grad_norm": 0.09606889635324478, "learning_rate": 8.564360374246821e-06, "loss": 0.0108, "step": 41020 }, { "grad_norm": 0.22008544206619263, "learning_rate": 8.545861353357597e-06, "loss": 0.0162, "step": 41030 }, { "grad_norm": 0.1189369484782219, "learning_rate": 8.527380465917001e-06, "loss": 0.0149, "step": 41040 }, { "grad_norm": 0.0799170583486557, "learning_rate": 8.508917720009175e-06, "loss": 0.0155, "step": 41050 }, { "grad_norm": 0.06840576976537704, "learning_rate": 8.490473123710374e-06, "loss": 0.0122, "step": 41060 }, { "grad_norm": 0.06590301543474197, "learning_rate": 8.472046685088874e-06, "loss": 0.0153, "step": 41070 }, { "grad_norm": 0.0820191353559494, "learning_rate": 8.453638412205023e-06, "loss": 0.0109, "step": 41080 }, { "grad_norm": 0.1054946631193161, "learning_rate": 8.435248313111243e-06, "loss": 0.0195, "step": 41090 }, { "grad_norm": 0.0850658044219017, "learning_rate": 8.416876395851974e-06, "loss": 0.0133, "step": 41100 }, { "grad_norm": 0.11681000143289566, "learning_rate": 8.39852266846371e-06, "loss": 0.0132, "step": 41110 }, { "grad_norm": 0.11120916903018951, "learning_rate": 8.380187138975021e-06, "loss": 0.0141, "step": 41120 }, { "grad_norm": 0.21805846691131592, "learning_rate": 8.36186981540647e-06, "loss": 0.0137, "step": 41130 }, { "grad_norm": 0.15620705485343933, "learning_rate": 8.343570705770681e-06, "loss": 0.0151, "step": 41140 }, { "grad_norm": 0.06294967979192734, "learning_rate": 8.325289818072296e-06, "loss": 0.01, "step": 41150 }, { "grad_norm": 0.08304492384195328, "learning_rate": 8.307027160308018e-06, "loss": 0.0158, "step": 41160 }, { "grad_norm": 0.17763009667396545, "learning_rate": 8.288782740466538e-06, "loss": 0.0128, "step": 41170 }, { "grad_norm": 0.08051881939172745, "learning_rate": 8.270556566528581e-06, "loss": 0.0117, "step": 41180 }, { "grad_norm": 0.061475180089473724, "learning_rate": 8.252348646466918e-06, "loss": 0.013, "step": 41190 }, { "grad_norm": 0.05022388696670532, "learning_rate": 8.234158988246294e-06, "loss": 0.0121, "step": 41200 }, { "grad_norm": 0.1302880495786667, "learning_rate": 8.215987599823476e-06, "loss": 0.011, "step": 41210 }, { "grad_norm": 0.0785575807094574, "learning_rate": 8.19783448914727e-06, "loss": 0.0119, "step": 41220 }, { "grad_norm": 0.07986260950565338, "learning_rate": 8.179699664158452e-06, "loss": 0.0161, "step": 41230 }, { "grad_norm": 0.09289270639419556, "learning_rate": 8.16158313278979e-06, "loss": 0.0124, "step": 41240 }, { "grad_norm": 0.11467690765857697, "learning_rate": 8.14348490296611e-06, "loss": 0.0112, "step": 41250 }, { "grad_norm": 0.09922505170106888, "learning_rate": 8.125404982604156e-06, "loss": 0.0136, "step": 41260 }, { "grad_norm": 0.11025633662939072, "learning_rate": 8.107343379612725e-06, "loss": 0.0148, "step": 41270 }, { "grad_norm": 0.07527876645326614, "learning_rate": 8.08930010189256e-06, "loss": 0.0104, "step": 41280 }, { "grad_norm": 0.06281867623329163, "learning_rate": 8.071275157336423e-06, "loss": 0.0103, "step": 41290 }, { "grad_norm": 0.07677346467971802, "learning_rate": 8.053268553829018e-06, "loss": 0.0099, "step": 41300 }, { "grad_norm": 0.09256012737751007, "learning_rate": 8.03528029924704e-06, "loss": 0.0154, "step": 41310 }, { "grad_norm": 0.06157543137669563, "learning_rate": 8.017310401459188e-06, "loss": 0.011, "step": 41320 }, { "grad_norm": 0.07340627163648605, "learning_rate": 7.999358868326095e-06, "loss": 0.0115, "step": 41330 }, { "grad_norm": 0.08293067663908005, "learning_rate": 7.981425707700352e-06, "loss": 0.0116, "step": 41340 }, { "grad_norm": 0.06484534591436386, "learning_rate": 7.96351092742656e-06, "loss": 0.0116, "step": 41350 }, { "grad_norm": 0.10532266646623611, "learning_rate": 7.945614535341246e-06, "loss": 0.011, "step": 41360 }, { "grad_norm": 0.1422448307275772, "learning_rate": 7.92773653927288e-06, "loss": 0.0141, "step": 41370 }, { "grad_norm": 0.08342869579792023, "learning_rate": 7.909876947041934e-06, "loss": 0.0119, "step": 41380 }, { "grad_norm": 0.08094995468854904, "learning_rate": 7.892035766460782e-06, "loss": 0.0112, "step": 41390 }, { "grad_norm": 0.11841030418872833, "learning_rate": 7.874213005333764e-06, "loss": 0.0122, "step": 41400 }, { "grad_norm": 0.09193816035985947, "learning_rate": 7.856408671457166e-06, "loss": 0.0088, "step": 41410 }, { "grad_norm": 0.09334911406040192, "learning_rate": 7.838622772619214e-06, "loss": 0.0116, "step": 41420 }, { "grad_norm": 0.06711819767951965, "learning_rate": 7.820855316600045e-06, "loss": 0.0131, "step": 41430 }, { "grad_norm": 0.1371445506811142, "learning_rate": 7.803106311171771e-06, "loss": 0.0104, "step": 41440 }, { "grad_norm": 0.05091569200158119, "learning_rate": 7.785375764098401e-06, "loss": 0.0129, "step": 41450 }, { "grad_norm": 0.053080569952726364, "learning_rate": 7.767663683135862e-06, "loss": 0.0148, "step": 41460 }, { "grad_norm": 0.08563516288995743, "learning_rate": 7.749970076032049e-06, "loss": 0.0104, "step": 41470 }, { "grad_norm": 0.23193246126174927, "learning_rate": 7.73229495052673e-06, "loss": 0.0149, "step": 41480 }, { "grad_norm": 0.06935961544513702, "learning_rate": 7.714638314351597e-06, "loss": 0.0125, "step": 41490 }, { "grad_norm": 0.06675230711698532, "learning_rate": 7.69700017523029e-06, "loss": 0.0094, "step": 41500 }, { "grad_norm": 0.0566752664744854, "learning_rate": 7.679380540878294e-06, "loss": 0.0106, "step": 41510 }, { "grad_norm": 0.12495072931051254, "learning_rate": 7.661779419003073e-06, "loss": 0.0127, "step": 41520 }, { "grad_norm": 0.08191555738449097, "learning_rate": 7.644196817303928e-06, "loss": 0.012, "step": 41530 }, { "grad_norm": 0.07984992861747742, "learning_rate": 7.626632743472106e-06, "loss": 0.0111, "step": 41540 }, { "grad_norm": 0.08333718031644821, "learning_rate": 7.609087205190724e-06, "loss": 0.0109, "step": 41550 }, { "grad_norm": 0.10031551867723465, "learning_rate": 7.591560210134774e-06, "loss": 0.0118, "step": 41560 }, { "grad_norm": 0.33296066522598267, "learning_rate": 7.574051765971196e-06, "loss": 0.0129, "step": 41570 }, { "grad_norm": 0.1074276939034462, "learning_rate": 7.556561880358754e-06, "loss": 0.015, "step": 41580 }, { "grad_norm": 0.12915879487991333, "learning_rate": 7.5390905609481196e-06, "loss": 0.0128, "step": 41590 }, { "grad_norm": 0.08384353667497635, "learning_rate": 7.521637815381855e-06, "loss": 0.0106, "step": 41600 }, { "grad_norm": 0.0807134360074997, "learning_rate": 7.504203651294373e-06, "loss": 0.0123, "step": 41610 }, { "grad_norm": 0.20480984449386597, "learning_rate": 7.486788076311963e-06, "loss": 0.0127, "step": 41620 }, { "grad_norm": 0.05256987735629082, "learning_rate": 7.469391098052808e-06, "loss": 0.0141, "step": 41630 }, { "grad_norm": 0.08452639728784561, "learning_rate": 7.452012724126928e-06, "loss": 0.0113, "step": 41640 }, { "grad_norm": 0.07126561552286148, "learning_rate": 7.434652962136213e-06, "loss": 0.0125, "step": 41650 }, { "grad_norm": 0.08420971781015396, "learning_rate": 7.4173118196744e-06, "loss": 0.0137, "step": 41660 }, { "grad_norm": 0.07271154224872589, "learning_rate": 7.39998930432712e-06, "loss": 0.0109, "step": 41670 }, { "grad_norm": 0.10196756571531296, "learning_rate": 7.38268542367182e-06, "loss": 0.0123, "step": 41680 }, { "grad_norm": 0.06709232926368713, "learning_rate": 7.365400185277782e-06, "loss": 0.011, "step": 41690 }, { "grad_norm": 0.07937761396169662, "learning_rate": 7.348133596706197e-06, "loss": 0.012, "step": 41700 }, { "grad_norm": 0.10089340806007385, "learning_rate": 7.3308856655100286e-06, "loss": 0.0117, "step": 41710 }, { "grad_norm": 0.10509032756090164, "learning_rate": 7.313656399234109e-06, "loss": 0.0143, "step": 41720 }, { "grad_norm": 0.1659092903137207, "learning_rate": 7.296445805415125e-06, "loss": 0.0142, "step": 41730 }, { "grad_norm": 0.16551214456558228, "learning_rate": 7.2792538915815504e-06, "loss": 0.0124, "step": 41740 }, { "grad_norm": 0.05584968626499176, "learning_rate": 7.262080665253729e-06, "loss": 0.0138, "step": 41750 }, { "grad_norm": 0.11375255137681961, "learning_rate": 7.244926133943803e-06, "loss": 0.0119, "step": 41760 }, { "grad_norm": 0.11349157989025116, "learning_rate": 7.227790305155763e-06, "loss": 0.0125, "step": 41770 }, { "grad_norm": 0.09935339540243149, "learning_rate": 7.210673186385381e-06, "loss": 0.0129, "step": 41780 }, { "grad_norm": 0.14275957643985748, "learning_rate": 7.193574785120288e-06, "loss": 0.0137, "step": 41790 }, { "grad_norm": 0.12709100544452667, "learning_rate": 7.176495108839893e-06, "loss": 0.0126, "step": 41800 }, { "grad_norm": 0.054393187165260315, "learning_rate": 7.159434165015417e-06, "loss": 0.0106, "step": 41810 }, { "grad_norm": 0.056455567479133606, "learning_rate": 7.142391961109923e-06, "loss": 0.0096, "step": 41820 }, { "grad_norm": 0.0778888463973999, "learning_rate": 7.1253685045782335e-06, "loss": 0.0126, "step": 41830 }, { "grad_norm": 0.1028815507888794, "learning_rate": 7.108363802866991e-06, "loss": 0.0151, "step": 41840 }, { "grad_norm": 0.07999451458454132, "learning_rate": 7.091377863414611e-06, "loss": 0.0113, "step": 41850 }, { "grad_norm": 0.08831703662872314, "learning_rate": 7.074410693651351e-06, "loss": 0.0155, "step": 41860 }, { "grad_norm": 0.05979784578084946, "learning_rate": 7.057462300999212e-06, "loss": 0.0125, "step": 41870 }, { "grad_norm": 0.06325878947973251, "learning_rate": 7.0405326928719876e-06, "loss": 0.0109, "step": 41880 }, { "grad_norm": 0.08660660684108734, "learning_rate": 7.0236218766752844e-06, "loss": 0.0106, "step": 41890 }, { "grad_norm": 0.14609508216381073, "learning_rate": 7.006729859806454e-06, "loss": 0.0137, "step": 41900 }, { "grad_norm": 0.060197729617357254, "learning_rate": 6.989856649654636e-06, "loss": 0.0128, "step": 41910 }, { "grad_norm": 0.08450326323509216, "learning_rate": 6.973002253600764e-06, "loss": 0.0107, "step": 41920 }, { "grad_norm": 0.09814087301492691, "learning_rate": 6.956166679017506e-06, "loss": 0.0123, "step": 41930 }, { "grad_norm": 0.08168792724609375, "learning_rate": 6.939349933269312e-06, "loss": 0.0098, "step": 41940 }, { "grad_norm": 0.055263541638851166, "learning_rate": 6.9225520237124146e-06, "loss": 0.0121, "step": 41950 }, { "grad_norm": 0.08254868537187576, "learning_rate": 6.905772957694784e-06, "loss": 0.0121, "step": 41960 }, { "grad_norm": 0.08477209508419037, "learning_rate": 6.889012742556133e-06, "loss": 0.0134, "step": 41970 }, { "grad_norm": 0.08383259922266006, "learning_rate": 6.872271385627982e-06, "loss": 0.0124, "step": 41980 }, { "grad_norm": 0.07217971235513687, "learning_rate": 6.8555488942335435e-06, "loss": 0.0123, "step": 41990 }, { "grad_norm": 0.07768049836158752, "learning_rate": 6.838845275687822e-06, "loss": 0.0109, "step": 42000 }, { "grad_norm": 0.05668661370873451, "learning_rate": 6.822160537297534e-06, "loss": 0.012, "step": 42010 }, { "grad_norm": 0.0993385761976242, "learning_rate": 6.805494686361164e-06, "loss": 0.0137, "step": 42020 }, { "grad_norm": 0.0838538110256195, "learning_rate": 6.788847730168912e-06, "loss": 0.0164, "step": 42030 }, { "grad_norm": 0.074896901845932, "learning_rate": 6.772219676002716e-06, "loss": 0.0233, "step": 42040 }, { "grad_norm": 0.15767528116703033, "learning_rate": 6.755610531136269e-06, "loss": 0.0185, "step": 42050 }, { "grad_norm": 0.07847974449396133, "learning_rate": 6.73902030283497e-06, "loss": 0.0113, "step": 42060 }, { "grad_norm": 0.05570325255393982, "learning_rate": 6.7224489983559295e-06, "loss": 0.0108, "step": 42070 }, { "grad_norm": 0.124453105032444, "learning_rate": 6.705896624948027e-06, "loss": 0.015, "step": 42080 }, { "grad_norm": 0.06593237817287445, "learning_rate": 6.689363189851816e-06, "loss": 0.0139, "step": 42090 }, { "grad_norm": 0.06683323532342911, "learning_rate": 6.672848700299583e-06, "loss": 0.013, "step": 42100 }, { "grad_norm": 0.06697674840688705, "learning_rate": 6.656353163515339e-06, "loss": 0.0102, "step": 42110 }, { "grad_norm": 0.05574244260787964, "learning_rate": 6.6398765867147824e-06, "loss": 0.0145, "step": 42120 }, { "grad_norm": 0.05677774176001549, "learning_rate": 6.623418977105322e-06, "loss": 0.0113, "step": 42130 }, { "grad_norm": 0.05794672295451164, "learning_rate": 6.606980341886087e-06, "loss": 0.0103, "step": 42140 }, { "grad_norm": 0.08859563618898392, "learning_rate": 6.590560688247898e-06, "loss": 0.0127, "step": 42150 }, { "grad_norm": 0.06293125450611115, "learning_rate": 6.574160023373249e-06, "loss": 0.0117, "step": 42160 }, { "grad_norm": 0.06695857644081116, "learning_rate": 6.557778354436367e-06, "loss": 0.0116, "step": 42170 }, { "grad_norm": 0.04666592553257942, "learning_rate": 6.5414156886031475e-06, "loss": 0.0116, "step": 42180 }, { "grad_norm": 0.08454198390245438, "learning_rate": 6.525072033031177e-06, "loss": 0.0125, "step": 42190 }, { "grad_norm": 0.064456507563591, "learning_rate": 6.508747394869708e-06, "loss": 0.0105, "step": 42200 }, { "grad_norm": 0.08605371415615082, "learning_rate": 6.492441781259717e-06, "loss": 0.0126, "step": 42210 }, { "grad_norm": 0.10513318330049515, "learning_rate": 6.476155199333811e-06, "loss": 0.0119, "step": 42220 }, { "grad_norm": 0.11185197532176971, "learning_rate": 6.459887656216318e-06, "loss": 0.0119, "step": 42230 }, { "grad_norm": 0.0764755979180336, "learning_rate": 6.443639159023191e-06, "loss": 0.0121, "step": 42240 }, { "grad_norm": 0.09594730287790298, "learning_rate": 6.427409714862092e-06, "loss": 0.0117, "step": 42250 }, { "grad_norm": 0.057191506028175354, "learning_rate": 6.411199330832318e-06, "loss": 0.0159, "step": 42260 }, { "grad_norm": 0.06929018348455429, "learning_rate": 6.3950080140248556e-06, "loss": 0.0124, "step": 42270 }, { "grad_norm": 0.08021893352270126, "learning_rate": 6.378835771522335e-06, "loss": 0.0119, "step": 42280 }, { "grad_norm": 0.10040401667356491, "learning_rate": 6.362682610399023e-06, "loss": 0.0128, "step": 42290 }, { "grad_norm": 0.05330561846494675, "learning_rate": 6.346548537720886e-06, "loss": 0.0094, "step": 42300 }, { "grad_norm": 0.08951075375080109, "learning_rate": 6.330433560545512e-06, "loss": 0.0127, "step": 42310 }, { "grad_norm": 0.088449627161026, "learning_rate": 6.3143376859221225e-06, "loss": 0.0113, "step": 42320 }, { "grad_norm": 0.08119135349988937, "learning_rate": 6.298260920891624e-06, "loss": 0.0132, "step": 42330 }, { "grad_norm": 0.09345356374979019, "learning_rate": 6.282203272486525e-06, "loss": 0.0151, "step": 42340 }, { "grad_norm": 0.07589007169008255, "learning_rate": 6.266164747730985e-06, "loss": 0.0146, "step": 42350 }, { "grad_norm": 0.07591290771961212, "learning_rate": 6.250145353640796e-06, "loss": 0.0128, "step": 42360 }, { "grad_norm": 0.07134699821472168, "learning_rate": 6.234145097223404e-06, "loss": 0.0128, "step": 42370 }, { "grad_norm": 0.06773052364587784, "learning_rate": 6.218163985477854e-06, "loss": 0.0108, "step": 42380 }, { "grad_norm": 0.08767600357532501, "learning_rate": 6.202202025394815e-06, "loss": 0.0133, "step": 42390 }, { "grad_norm": 0.09721200913190842, "learning_rate": 6.186259223956614e-06, "loss": 0.0113, "step": 42400 }, { "grad_norm": 0.06761650741100311, "learning_rate": 6.170335588137161e-06, "loss": 0.0123, "step": 42410 }, { "grad_norm": 0.04482118412852287, "learning_rate": 6.154431124901982e-06, "loss": 0.0123, "step": 42420 }, { "grad_norm": 0.08786734938621521, "learning_rate": 6.1385458412082595e-06, "loss": 0.014, "step": 42430 }, { "grad_norm": 0.15657992660999298, "learning_rate": 6.122679744004739e-06, "loss": 0.0139, "step": 42440 }, { "grad_norm": 0.09346558153629303, "learning_rate": 6.106832840231786e-06, "loss": 0.0123, "step": 42450 }, { "grad_norm": 0.0684761255979538, "learning_rate": 6.091005136821387e-06, "loss": 0.0123, "step": 42460 }, { "grad_norm": 0.060644444078207016, "learning_rate": 6.075196640697101e-06, "loss": 0.0126, "step": 42470 }, { "grad_norm": 0.09180520474910736, "learning_rate": 6.059407358774122e-06, "loss": 0.0092, "step": 42480 }, { "grad_norm": 0.1026988998055458, "learning_rate": 6.043637297959198e-06, "loss": 0.0127, "step": 42490 }, { "grad_norm": 0.058408163487911224, "learning_rate": 6.0278864651507065e-06, "loss": 0.013, "step": 42500 }, { "grad_norm": 0.06825146079063416, "learning_rate": 6.012154867238584e-06, "loss": 0.013, "step": 42510 }, { "grad_norm": 0.060345541685819626, "learning_rate": 5.996442511104378e-06, "loss": 0.012, "step": 42520 }, { "grad_norm": 0.08233707398176193, "learning_rate": 5.9807494036211995e-06, "loss": 0.0122, "step": 42530 }, { "grad_norm": 0.04281025379896164, "learning_rate": 5.965075551653748e-06, "loss": 0.0118, "step": 42540 }, { "grad_norm": 0.09352441132068634, "learning_rate": 5.949420962058294e-06, "loss": 0.0122, "step": 42550 }, { "grad_norm": 0.10246340185403824, "learning_rate": 5.933785641682693e-06, "loss": 0.0113, "step": 42560 }, { "grad_norm": 0.08550791442394257, "learning_rate": 5.918169597366369e-06, "loss": 0.0108, "step": 42570 }, { "grad_norm": 0.058818619698286057, "learning_rate": 5.902572835940295e-06, "loss": 0.0094, "step": 42580 }, { "grad_norm": 0.1452997773885727, "learning_rate": 5.886995364227044e-06, "loss": 0.0139, "step": 42590 }, { "grad_norm": 0.07886643707752228, "learning_rate": 5.871437189040718e-06, "loss": 0.0101, "step": 42600 }, { "grad_norm": 0.048166170716285706, "learning_rate": 5.855898317186992e-06, "loss": 0.0106, "step": 42610 }, { "grad_norm": 0.0825561136007309, "learning_rate": 5.840378755463105e-06, "loss": 0.013, "step": 42620 }, { "grad_norm": 0.11613915115594864, "learning_rate": 5.824878510657833e-06, "loss": 0.013, "step": 42630 }, { "grad_norm": 0.08796895295381546, "learning_rate": 5.809397589551502e-06, "loss": 0.0115, "step": 42640 }, { "grad_norm": 0.06907706707715988, "learning_rate": 5.793935998916017e-06, "loss": 0.0139, "step": 42650 }, { "grad_norm": 0.12679335474967957, "learning_rate": 5.778493745514785e-06, "loss": 0.0113, "step": 42660 }, { "grad_norm": 0.05770030990242958, "learning_rate": 5.763070836102768e-06, "loss": 0.0106, "step": 42670 }, { "grad_norm": 0.11436334997415543, "learning_rate": 5.7476672774264885e-06, "loss": 0.0124, "step": 42680 }, { "grad_norm": 0.09038861840963364, "learning_rate": 5.732283076223977e-06, "loss": 0.0139, "step": 42690 }, { "grad_norm": 0.06367217749357224, "learning_rate": 5.716918239224794e-06, "loss": 0.0104, "step": 42700 }, { "grad_norm": 0.1231161504983902, "learning_rate": 5.70157277315006e-06, "loss": 0.0134, "step": 42710 }, { "grad_norm": 0.05683391913771629, "learning_rate": 5.686246684712387e-06, "loss": 0.0122, "step": 42720 }, { "grad_norm": 0.0646800696849823, "learning_rate": 5.670939980615947e-06, "loss": 0.0106, "step": 42730 }, { "grad_norm": 0.043332576751708984, "learning_rate": 5.655652667556388e-06, "loss": 0.0114, "step": 42740 }, { "grad_norm": 0.058923691511154175, "learning_rate": 5.640384752220928e-06, "loss": 0.012, "step": 42750 }, { "grad_norm": 0.07323578000068665, "learning_rate": 5.625136241288253e-06, "loss": 0.0123, "step": 42760 }, { "grad_norm": 0.12544718384742737, "learning_rate": 5.609907141428572e-06, "loss": 0.0111, "step": 42770 }, { "grad_norm": 0.060635097324848175, "learning_rate": 5.594697459303633e-06, "loss": 0.0129, "step": 42780 }, { "grad_norm": 0.07765184342861176, "learning_rate": 5.579507201566658e-06, "loss": 0.0139, "step": 42790 }, { "grad_norm": 0.07260327786207199, "learning_rate": 5.564336374862372e-06, "loss": 0.0128, "step": 42800 }, { "grad_norm": 0.05421106889843941, "learning_rate": 5.549184985827027e-06, "loss": 0.013, "step": 42810 }, { "grad_norm": 0.1179877370595932, "learning_rate": 5.534053041088349e-06, "loss": 0.0114, "step": 42820 }, { "grad_norm": 0.09609463065862656, "learning_rate": 5.5189405472655495e-06, "loss": 0.0145, "step": 42830 }, { "grad_norm": 0.055474571883678436, "learning_rate": 5.503847510969374e-06, "loss": 0.0149, "step": 42840 }, { "grad_norm": 0.07453738152980804, "learning_rate": 5.488773938802011e-06, "loss": 0.0093, "step": 42850 }, { "grad_norm": 0.068412184715271, "learning_rate": 5.473719837357155e-06, "loss": 0.0123, "step": 42860 }, { "grad_norm": 0.06611385941505432, "learning_rate": 5.4586852132199825e-06, "loss": 0.0104, "step": 42870 }, { "grad_norm": 0.056331150233745575, "learning_rate": 5.443670072967155e-06, "loss": 0.013, "step": 42880 }, { "grad_norm": 0.17125454545021057, "learning_rate": 5.428674423166796e-06, "loss": 0.0135, "step": 42890 }, { "grad_norm": 0.08789560198783875, "learning_rate": 5.413698270378515e-06, "loss": 0.0116, "step": 42900 }, { "grad_norm": 0.07732050865888596, "learning_rate": 5.398741621153397e-06, "loss": 0.011, "step": 42910 }, { "grad_norm": 0.06275799125432968, "learning_rate": 5.383804482033983e-06, "loss": 0.0123, "step": 42920 }, { "grad_norm": 0.06338466703891754, "learning_rate": 5.368886859554273e-06, "loss": 0.0128, "step": 42930 }, { "grad_norm": 0.1187833845615387, "learning_rate": 5.353988760239764e-06, "loss": 0.011, "step": 42940 }, { "grad_norm": 0.05501749366521835, "learning_rate": 5.339110190607377e-06, "loss": 0.01, "step": 42950 }, { "grad_norm": 0.0671108290553093, "learning_rate": 5.324251157165511e-06, "loss": 0.0136, "step": 42960 }, { "grad_norm": 0.09946762770414352, "learning_rate": 5.309411666413999e-06, "loss": 0.0131, "step": 42970 }, { "grad_norm": 0.06896968185901642, "learning_rate": 5.294591724844161e-06, "loss": 0.0116, "step": 42980 }, { "grad_norm": 0.08387649059295654, "learning_rate": 5.279791338938716e-06, "loss": 0.0132, "step": 42990 }, { "grad_norm": 0.07154630124568939, "learning_rate": 5.265010515171886e-06, "loss": 0.0116, "step": 43000 }, { "grad_norm": 0.11401842534542084, "learning_rate": 5.250249260009282e-06, "loss": 0.0123, "step": 43010 }, { "grad_norm": 0.11708151549100876, "learning_rate": 5.2355075799079835e-06, "loss": 0.0115, "step": 43020 }, { "grad_norm": 0.18949946761131287, "learning_rate": 5.220785481316509e-06, "loss": 0.0148, "step": 43030 }, { "grad_norm": 0.07747366279363632, "learning_rate": 5.206082970674797e-06, "loss": 0.0133, "step": 43040 }, { "grad_norm": 0.07504931837320328, "learning_rate": 5.191400054414231e-06, "loss": 0.0147, "step": 43050 }, { "grad_norm": 0.0934320017695427, "learning_rate": 5.176736738957605e-06, "loss": 0.0134, "step": 43060 }, { "grad_norm": 0.09562680870294571, "learning_rate": 5.162093030719173e-06, "loss": 0.013, "step": 43070 }, { "grad_norm": 0.05384210869669914, "learning_rate": 5.1474689361045715e-06, "loss": 0.0134, "step": 43080 }, { "grad_norm": 0.11335650831460953, "learning_rate": 5.132864461510878e-06, "loss": 0.0158, "step": 43090 }, { "grad_norm": 0.07128454744815826, "learning_rate": 5.118279613326604e-06, "loss": 0.0134, "step": 43100 }, { "grad_norm": 0.06974861770868301, "learning_rate": 5.103714397931636e-06, "loss": 0.0142, "step": 43110 }, { "grad_norm": 0.07388997077941895, "learning_rate": 5.0891688216972985e-06, "loss": 0.0105, "step": 43120 }, { "grad_norm": 0.0838470458984375, "learning_rate": 5.074642890986336e-06, "loss": 0.0121, "step": 43130 }, { "grad_norm": 0.09384794533252716, "learning_rate": 5.060136612152871e-06, "loss": 0.0144, "step": 43140 }, { "grad_norm": 0.06322265416383743, "learning_rate": 5.045649991542434e-06, "loss": 0.0117, "step": 43150 }, { "grad_norm": 0.09377554059028625, "learning_rate": 5.031183035491982e-06, "loss": 0.0242, "step": 43160 }, { "grad_norm": 0.06800924986600876, "learning_rate": 5.016735750329854e-06, "loss": 0.0138, "step": 43170 }, { "grad_norm": 0.07843763381242752, "learning_rate": 5.002308142375761e-06, "loss": 0.0125, "step": 43180 }, { "grad_norm": 0.08415485918521881, "learning_rate": 4.987900217940855e-06, "loss": 0.0108, "step": 43190 }, { "grad_norm": 0.07295873761177063, "learning_rate": 4.973511983327633e-06, "loss": 0.0127, "step": 43200 }, { "grad_norm": 0.07175947725772858, "learning_rate": 4.959143444830011e-06, "loss": 0.0121, "step": 43210 }, { "grad_norm": 0.1203155666589737, "learning_rate": 4.944794608733267e-06, "loss": 0.0118, "step": 43220 }, { "grad_norm": 0.10378172248601913, "learning_rate": 4.930465481314078e-06, "loss": 0.0135, "step": 43230 }, { "grad_norm": 0.044393058866262436, "learning_rate": 4.916156068840483e-06, "loss": 0.012, "step": 43240 }, { "grad_norm": 0.08016838133335114, "learning_rate": 4.901866377571901e-06, "loss": 0.0147, "step": 43250 }, { "grad_norm": 0.07334733754396439, "learning_rate": 4.887596413759143e-06, "loss": 0.0136, "step": 43260 }, { "grad_norm": 0.11467938870191574, "learning_rate": 4.873346183644373e-06, "loss": 0.0128, "step": 43270 }, { "grad_norm": 0.06520161777734756, "learning_rate": 4.8591156934611085e-06, "loss": 0.0136, "step": 43280 }, { "grad_norm": 0.09060197323560715, "learning_rate": 4.8449049494342704e-06, "loss": 0.014, "step": 43290 }, { "grad_norm": 0.09262946993112564, "learning_rate": 4.830713957780114e-06, "loss": 0.013, "step": 43300 }, { "grad_norm": 0.1206948384642601, "learning_rate": 4.816542724706253e-06, "loss": 0.0152, "step": 43310 }, { "grad_norm": 0.07134894281625748, "learning_rate": 4.802391256411675e-06, "loss": 0.0138, "step": 43320 }, { "grad_norm": 0.09513837844133377, "learning_rate": 4.788259559086717e-06, "loss": 0.0136, "step": 43330 }, { "grad_norm": 0.08216099441051483, "learning_rate": 4.774147638913046e-06, "loss": 0.0126, "step": 43340 }, { "grad_norm": 0.06764119863510132, "learning_rate": 4.760055502063721e-06, "loss": 0.0117, "step": 43350 }, { "grad_norm": 0.05412289500236511, "learning_rate": 4.745983154703099e-06, "loss": 0.0113, "step": 43360 }, { "grad_norm": 0.09085645526647568, "learning_rate": 4.731930602986906e-06, "loss": 0.0144, "step": 43370 }, { "grad_norm": 0.0959145724773407, "learning_rate": 4.7178978530622196e-06, "loss": 0.0145, "step": 43380 }, { "grad_norm": 0.1157832071185112, "learning_rate": 4.70388491106743e-06, "loss": 0.0126, "step": 43390 }, { "grad_norm": 0.04965398833155632, "learning_rate": 4.689891783132278e-06, "loss": 0.0108, "step": 43400 }, { "grad_norm": 0.10478059947490692, "learning_rate": 4.675918475377822e-06, "loss": 0.0121, "step": 43410 }, { "grad_norm": 0.08351294696331024, "learning_rate": 4.6619649939164845e-06, "loss": 0.0137, "step": 43420 }, { "grad_norm": 0.08305381238460541, "learning_rate": 4.6480313448519665e-06, "loss": 0.0132, "step": 43430 }, { "grad_norm": 0.09241440892219543, "learning_rate": 4.634117534279337e-06, "loss": 0.0144, "step": 43440 }, { "grad_norm": 0.0533401183784008, "learning_rate": 4.620223568284959e-06, "loss": 0.0136, "step": 43450 }, { "grad_norm": 0.10839928686618805, "learning_rate": 4.60634945294654e-06, "loss": 0.0127, "step": 43460 }, { "grad_norm": 0.1059078574180603, "learning_rate": 4.592495194333074e-06, "loss": 0.014, "step": 43470 }, { "grad_norm": 0.03327491879463196, "learning_rate": 4.578660798504897e-06, "loss": 0.0112, "step": 43480 }, { "grad_norm": 0.11243324726819992, "learning_rate": 4.5648462715136346e-06, "loss": 0.0142, "step": 43490 }, { "grad_norm": 0.11136157065629959, "learning_rate": 4.551051619402225e-06, "loss": 0.0115, "step": 43500 }, { "grad_norm": 0.09358129650354385, "learning_rate": 4.537276848204936e-06, "loss": 0.0133, "step": 43510 }, { "grad_norm": 0.0860670953989029, "learning_rate": 4.523521963947308e-06, "loss": 0.0134, "step": 43520 }, { "grad_norm": 0.07944691181182861, "learning_rate": 4.509786972646191e-06, "loss": 0.0123, "step": 43530 }, { "grad_norm": 0.07710746675729752, "learning_rate": 4.496071880309749e-06, "loss": 0.0125, "step": 43540 }, { "grad_norm": 0.0884467288851738, "learning_rate": 4.4823766929374234e-06, "loss": 0.0129, "step": 43550 }, { "grad_norm": 0.11969202756881714, "learning_rate": 4.468701416519955e-06, "loss": 0.0131, "step": 43560 }, { "grad_norm": 0.09033498167991638, "learning_rate": 4.4550460570393635e-06, "loss": 0.0144, "step": 43570 }, { "grad_norm": 0.06724975258111954, "learning_rate": 4.441410620468989e-06, "loss": 0.0138, "step": 43580 }, { "grad_norm": 0.05075416341423988, "learning_rate": 4.427795112773419e-06, "loss": 0.01, "step": 43590 }, { "grad_norm": 0.08945634216070175, "learning_rate": 4.414199539908531e-06, "loss": 0.013, "step": 43600 }, { "grad_norm": 0.04694605618715286, "learning_rate": 4.400623907821516e-06, "loss": 0.0134, "step": 43610 }, { "grad_norm": 0.12389319390058517, "learning_rate": 4.387068222450802e-06, "loss": 0.0117, "step": 43620 }, { "grad_norm": 0.07153929024934769, "learning_rate": 4.3735324897260954e-06, "loss": 0.0127, "step": 43630 }, { "grad_norm": 0.06860889494419098, "learning_rate": 4.36001671556841e-06, "loss": 0.0141, "step": 43640 }, { "grad_norm": 0.131075918674469, "learning_rate": 4.346520905889989e-06, "loss": 0.0146, "step": 43650 }, { "grad_norm": 0.10360931605100632, "learning_rate": 4.333045066594354e-06, "loss": 0.0123, "step": 43660 }, { "grad_norm": 0.07463011145591736, "learning_rate": 4.319589203576313e-06, "loss": 0.0157, "step": 43670 }, { "grad_norm": 0.07992967963218689, "learning_rate": 4.306153322721895e-06, "loss": 0.0121, "step": 43680 }, { "grad_norm": 0.09483496099710464, "learning_rate": 4.292737429908422e-06, "loss": 0.0166, "step": 43690 }, { "grad_norm": 0.09802956134080887, "learning_rate": 4.279341531004471e-06, "loss": 0.0139, "step": 43700 }, { "grad_norm": 0.06567131727933884, "learning_rate": 4.26596563186985e-06, "loss": 0.012, "step": 43710 }, { "grad_norm": 0.06183163449168205, "learning_rate": 4.252609738355634e-06, "loss": 0.0134, "step": 43720 }, { "grad_norm": 0.07508756965398788, "learning_rate": 4.239273856304149e-06, "loss": 0.0116, "step": 43730 }, { "grad_norm": 0.08940863609313965, "learning_rate": 4.2259579915489625e-06, "loss": 0.0136, "step": 43740 }, { "grad_norm": 0.10727040469646454, "learning_rate": 4.212662149914886e-06, "loss": 0.0137, "step": 43750 }, { "grad_norm": 0.07039087265729904, "learning_rate": 4.199386337217958e-06, "loss": 0.0129, "step": 43760 }, { "grad_norm": 0.13942071795463562, "learning_rate": 4.186130559265494e-06, "loss": 0.0149, "step": 43770 }, { "grad_norm": 0.08526845276355743, "learning_rate": 4.172894821856011e-06, "loss": 0.0133, "step": 43780 }, { "grad_norm": 0.0688057467341423, "learning_rate": 4.159679130779254e-06, "loss": 0.0128, "step": 43790 }, { "grad_norm": 0.07410506904125214, "learning_rate": 4.1464834918162445e-06, "loss": 0.0139, "step": 43800 }, { "grad_norm": 0.0851108655333519, "learning_rate": 4.133307910739187e-06, "loss": 0.0133, "step": 43810 }, { "grad_norm": 0.0758635625243187, "learning_rate": 4.120152393311521e-06, "loss": 0.011, "step": 43820 }, { "grad_norm": 0.06296105682849884, "learning_rate": 4.107016945287933e-06, "loss": 0.013, "step": 43830 }, { "grad_norm": 0.06201896071434021, "learning_rate": 4.09390157241431e-06, "loss": 0.0138, "step": 43840 }, { "grad_norm": 0.11009002476930618, "learning_rate": 4.0808062804277515e-06, "loss": 0.0134, "step": 43850 }, { "grad_norm": 0.09312368929386139, "learning_rate": 4.067731075056596e-06, "loss": 0.0125, "step": 43860 }, { "grad_norm": 0.08254053443670273, "learning_rate": 4.054675962020376e-06, "loss": 0.0121, "step": 43870 }, { "grad_norm": 0.06692612171173096, "learning_rate": 4.04164094702984e-06, "loss": 0.0128, "step": 43880 }, { "grad_norm": 0.04559154435992241, "learning_rate": 4.028626035786953e-06, "loss": 0.0118, "step": 43890 }, { "grad_norm": 0.13256263732910156, "learning_rate": 4.0156312339848675e-06, "loss": 0.0134, "step": 43900 }, { "grad_norm": 0.09431388974189758, "learning_rate": 4.00265654730797e-06, "loss": 0.0136, "step": 43910 }, { "grad_norm": 0.07816233485937119, "learning_rate": 3.989701981431809e-06, "loss": 0.0117, "step": 43920 }, { "grad_norm": 0.07830566167831421, "learning_rate": 3.976767542023169e-06, "loss": 0.0133, "step": 43930 }, { "grad_norm": 0.09692902863025665, "learning_rate": 3.963853234740006e-06, "loss": 0.0123, "step": 43940 }, { "grad_norm": 0.05450575426220894, "learning_rate": 3.950959065231464e-06, "loss": 0.0123, "step": 43950 }, { "grad_norm": 0.11558815091848373, "learning_rate": 3.938085039137917e-06, "loss": 0.0137, "step": 43960 }, { "grad_norm": 0.0672549158334732, "learning_rate": 3.925231162090881e-06, "loss": 0.0137, "step": 43970 }, { "grad_norm": 0.13315744698047638, "learning_rate": 3.912397439713084e-06, "loss": 0.0145, "step": 43980 }, { "grad_norm": 0.09718213230371475, "learning_rate": 3.899583877618435e-06, "loss": 0.0171, "step": 43990 }, { "grad_norm": 0.06683185696601868, "learning_rate": 3.8867904814120225e-06, "loss": 0.0144, "step": 44000 }, { "grad_norm": 0.09915776550769806, "learning_rate": 3.8740172566901044e-06, "loss": 0.0114, "step": 44010 }, { "grad_norm": 0.058125581592321396, "learning_rate": 3.861264209040133e-06, "loss": 0.0146, "step": 44020 }, { "grad_norm": 0.06435991078615189, "learning_rate": 3.848531344040718e-06, "loss": 0.0118, "step": 44030 }, { "grad_norm": 0.06763908267021179, "learning_rate": 3.835818667261642e-06, "loss": 0.0124, "step": 44040 }, { "grad_norm": 0.06921260803937912, "learning_rate": 3.823126184263881e-06, "loss": 0.0144, "step": 44050 }, { "grad_norm": 0.14791616797447205, "learning_rate": 3.810453900599542e-06, "loss": 0.0142, "step": 44060 }, { "grad_norm": 0.08338169008493423, "learning_rate": 3.7978018218119137e-06, "loss": 0.0148, "step": 44070 }, { "grad_norm": 0.09090873599052429, "learning_rate": 3.7851699534354524e-06, "loss": 0.0145, "step": 44080 }, { "grad_norm": 0.062264323234558105, "learning_rate": 3.77255830099576e-06, "loss": 0.0169, "step": 44090 }, { "grad_norm": 0.11273957043886185, "learning_rate": 3.759966870009607e-06, "loss": 0.018, "step": 44100 }, { "grad_norm": 0.07997002452611923, "learning_rate": 3.7473956659848987e-06, "loss": 0.0147, "step": 44110 }, { "grad_norm": 0.11187828332185745, "learning_rate": 3.7348446944207305e-06, "loss": 0.0133, "step": 44120 }, { "grad_norm": 0.10749422758817673, "learning_rate": 3.7223139608073e-06, "loss": 0.0118, "step": 44130 }, { "grad_norm": 0.1003490537405014, "learning_rate": 3.7098034706259943e-06, "loss": 0.0116, "step": 44140 }, { "grad_norm": 0.07329006493091583, "learning_rate": 3.697313229349314e-06, "loss": 0.0108, "step": 44150 }, { "grad_norm": 0.07490543276071548, "learning_rate": 3.684843242440933e-06, "loss": 0.0119, "step": 44160 }, { "grad_norm": 0.13702069222927094, "learning_rate": 3.672393515355621e-06, "loss": 0.0124, "step": 44170 }, { "grad_norm": 0.0925806537270546, "learning_rate": 3.659964053539333e-06, "loss": 0.0128, "step": 44180 }, { "grad_norm": 0.07638248056173325, "learning_rate": 3.647554862429131e-06, "loss": 0.017, "step": 44190 }, { "grad_norm": 0.14413674175739288, "learning_rate": 3.635165947453212e-06, "loss": 0.0153, "step": 44200 }, { "grad_norm": 0.09467978775501251, "learning_rate": 3.622797314030918e-06, "loss": 0.0149, "step": 44210 }, { "grad_norm": 0.08182041347026825, "learning_rate": 3.610448967572705e-06, "loss": 0.0113, "step": 44220 }, { "grad_norm": 0.12071390450000763, "learning_rate": 3.5981209134801464e-06, "loss": 0.0149, "step": 44230 }, { "grad_norm": 0.09076068550348282, "learning_rate": 3.585813157145973e-06, "loss": 0.012, "step": 44240 }, { "grad_norm": 0.09447415173053741, "learning_rate": 3.5735257039540114e-06, "loss": 0.0168, "step": 44250 }, { "grad_norm": 0.10511034727096558, "learning_rate": 3.5612585592792014e-06, "loss": 0.0159, "step": 44260 }, { "grad_norm": 0.11511628329753876, "learning_rate": 3.5490117284876055e-06, "loss": 0.0133, "step": 44270 }, { "grad_norm": 0.11948255449533463, "learning_rate": 3.5367852169364224e-06, "loss": 0.0167, "step": 44280 }, { "grad_norm": 0.1424892097711563, "learning_rate": 3.524579029973929e-06, "loss": 0.0145, "step": 44290 }, { "grad_norm": 0.11668345332145691, "learning_rate": 3.5123931729395267e-06, "loss": 0.0144, "step": 44300 }, { "grad_norm": 0.11062653362751007, "learning_rate": 3.500227651163729e-06, "loss": 0.0141, "step": 44310 }, { "grad_norm": 0.0808158814907074, "learning_rate": 3.488082469968146e-06, "loss": 0.0161, "step": 44320 }, { "grad_norm": 0.13285882771015167, "learning_rate": 3.475957634665489e-06, "loss": 0.0125, "step": 44330 }, { "grad_norm": 0.09808893501758575, "learning_rate": 3.463853150559582e-06, "loss": 0.014, "step": 44340 }, { "grad_norm": 0.14603115618228912, "learning_rate": 3.451769022945334e-06, "loss": 0.0146, "step": 44350 }, { "grad_norm": 0.06865782290697098, "learning_rate": 3.4397052571087385e-06, "loss": 0.0132, "step": 44360 }, { "grad_norm": 0.08046642690896988, "learning_rate": 3.427661858326925e-06, "loss": 0.0123, "step": 44370 }, { "grad_norm": 0.09502113610506058, "learning_rate": 3.4156388318680566e-06, "loss": 0.0161, "step": 44380 }, { "grad_norm": 0.1269877403974533, "learning_rate": 3.4036361829914322e-06, "loss": 0.0133, "step": 44390 }, { "grad_norm": 0.08871833235025406, "learning_rate": 3.391653916947407e-06, "loss": 0.0132, "step": 44400 }, { "grad_norm": 0.16292032599449158, "learning_rate": 3.3796920389774444e-06, "loss": 0.0165, "step": 44410 }, { "grad_norm": 0.10403776913881302, "learning_rate": 3.3677505543140643e-06, "loss": 0.0157, "step": 44420 }, { "grad_norm": 0.0851910263299942, "learning_rate": 3.355829468180882e-06, "loss": 0.0126, "step": 44430 }, { "grad_norm": 0.051437858492136, "learning_rate": 3.343928785792588e-06, "loss": 0.0143, "step": 44440 }, { "grad_norm": 0.12315288931131363, "learning_rate": 3.3320485123549405e-06, "loss": 0.0132, "step": 44450 }, { "grad_norm": 0.07887852191925049, "learning_rate": 3.3201886530647707e-06, "loss": 0.0138, "step": 44460 }, { "grad_norm": 0.11531884223222733, "learning_rate": 3.3083492131100015e-06, "loss": 0.0151, "step": 44470 }, { "grad_norm": 0.052869997918605804, "learning_rate": 3.2965301976695896e-06, "loss": 0.0113, "step": 44480 }, { "grad_norm": 0.08090338855981827, "learning_rate": 3.284731611913577e-06, "loss": 0.0105, "step": 44490 }, { "grad_norm": 0.09724154323339462, "learning_rate": 3.27295346100307e-06, "loss": 0.0134, "step": 44500 }, { "grad_norm": 0.07462911307811737, "learning_rate": 3.2611957500902347e-06, "loss": 0.013, "step": 44510 }, { "grad_norm": 0.055669404566287994, "learning_rate": 3.2494584843182796e-06, "loss": 0.0095, "step": 44520 }, { "grad_norm": 0.0756765827536583, "learning_rate": 3.2377416688214978e-06, "loss": 0.01, "step": 44530 }, { "grad_norm": 0.0667685717344284, "learning_rate": 3.2260453087252175e-06, "loss": 0.0125, "step": 44540 }, { "grad_norm": 0.11087175458669662, "learning_rate": 3.2143694091458175e-06, "loss": 0.0144, "step": 44550 }, { "grad_norm": 0.08262477815151215, "learning_rate": 3.2027139751907466e-06, "loss": 0.0106, "step": 44560 }, { "grad_norm": 0.09242645651102066, "learning_rate": 3.191079011958481e-06, "loss": 0.0142, "step": 44570 }, { "grad_norm": 0.07614562660455704, "learning_rate": 3.1794645245385337e-06, "loss": 0.0137, "step": 44580 }, { "grad_norm": 0.11369063705205917, "learning_rate": 3.1678705180115063e-06, "loss": 0.0105, "step": 44590 }, { "grad_norm": 0.05044955387711525, "learning_rate": 3.1562969974489933e-06, "loss": 0.0125, "step": 44600 }, { "grad_norm": 0.16003669798374176, "learning_rate": 3.144743967913638e-06, "loss": 0.012, "step": 44610 }, { "grad_norm": 0.11659938097000122, "learning_rate": 3.1332114344591436e-06, "loss": 0.0118, "step": 44620 }, { "grad_norm": 0.15144670009613037, "learning_rate": 3.1216994021302235e-06, "loss": 0.0149, "step": 44630 }, { "grad_norm": 0.09937619417905807, "learning_rate": 3.1102078759626396e-06, "loss": 0.0122, "step": 44640 }, { "grad_norm": 0.0824909582734108, "learning_rate": 3.098736860983159e-06, "loss": 0.0125, "step": 44650 }, { "grad_norm": 0.07153870910406113, "learning_rate": 3.087286362209618e-06, "loss": 0.0124, "step": 44660 }, { "grad_norm": 0.06753794103860855, "learning_rate": 3.0758563846508383e-06, "loss": 0.0144, "step": 44670 }, { "grad_norm": 0.06541910022497177, "learning_rate": 3.0644469333066748e-06, "loss": 0.014, "step": 44680 }, { "grad_norm": 0.07393179833889008, "learning_rate": 3.0530580131680307e-06, "loss": 0.0132, "step": 44690 }, { "grad_norm": 0.08587953448295593, "learning_rate": 3.0416896292167875e-06, "loss": 0.0135, "step": 44700 }, { "grad_norm": 0.06891924887895584, "learning_rate": 3.030341786425861e-06, "loss": 0.0123, "step": 44710 }, { "grad_norm": 0.0769297182559967, "learning_rate": 3.0190144897591975e-06, "loss": 0.0087, "step": 44720 }, { "grad_norm": 0.12799333035945892, "learning_rate": 3.0077077441717395e-06, "loss": 0.0118, "step": 44730 }, { "grad_norm": 0.05250577628612518, "learning_rate": 2.9964215546094244e-06, "loss": 0.0117, "step": 44740 }, { "grad_norm": 0.08208472281694412, "learning_rate": 2.9851559260092367e-06, "loss": 0.0154, "step": 44750 }, { "grad_norm": 0.07330133765935898, "learning_rate": 2.973910863299134e-06, "loss": 0.0137, "step": 44760 }, { "grad_norm": 0.06275558471679688, "learning_rate": 2.962686371398088e-06, "loss": 0.0126, "step": 44770 }, { "grad_norm": 0.056696344166994095, "learning_rate": 2.95148245521607e-06, "loss": 0.0136, "step": 44780 }, { "grad_norm": 0.0565388984978199, "learning_rate": 2.9402991196540665e-06, "loss": 0.0153, "step": 44790 }, { "grad_norm": 0.06834081560373306, "learning_rate": 2.929136369604041e-06, "loss": 0.0117, "step": 44800 }, { "grad_norm": 0.2299293577671051, "learning_rate": 2.9179942099489487e-06, "loss": 0.0131, "step": 44810 }, { "grad_norm": 0.06607715040445328, "learning_rate": 2.9068726455627694e-06, "loss": 0.0148, "step": 44820 }, { "grad_norm": 0.06871280074119568, "learning_rate": 2.8957716813104384e-06, "loss": 0.0136, "step": 44830 }, { "grad_norm": 0.09141505509614944, "learning_rate": 2.8846913220479e-06, "loss": 0.0115, "step": 44840 }, { "grad_norm": 0.10283029824495316, "learning_rate": 2.873631572622082e-06, "loss": 0.0115, "step": 44850 }, { "grad_norm": 0.07787823677062988, "learning_rate": 2.862592437870881e-06, "loss": 0.0118, "step": 44860 }, { "grad_norm": 0.13171304762363434, "learning_rate": 2.8515739226232075e-06, "loss": 0.0146, "step": 44870 }, { "grad_norm": 0.08251690864562988, "learning_rate": 2.840576031698927e-06, "loss": 0.0119, "step": 44880 }, { "grad_norm": 0.11165642738342285, "learning_rate": 2.8295987699088923e-06, "loss": 0.0139, "step": 44890 }, { "grad_norm": 0.0664639100432396, "learning_rate": 2.8186421420549227e-06, "loss": 0.0109, "step": 44900 }, { "grad_norm": 0.08588813245296478, "learning_rate": 2.8077061529298343e-06, "loss": 0.0129, "step": 44910 }, { "grad_norm": 0.08214276283979416, "learning_rate": 2.79679080731739e-06, "loss": 0.0163, "step": 44920 }, { "grad_norm": 0.06990755349397659, "learning_rate": 2.7858961099923254e-06, "loss": 0.0106, "step": 44930 }, { "grad_norm": 0.037873003631830215, "learning_rate": 2.7750220657203727e-06, "loss": 0.012, "step": 44940 }, { "grad_norm": 0.09894590079784393, "learning_rate": 2.764168679258189e-06, "loss": 0.0108, "step": 44950 }, { "grad_norm": 0.054965656250715256, "learning_rate": 2.7533359553534254e-06, "loss": 0.0101, "step": 44960 }, { "grad_norm": 0.09167791157960892, "learning_rate": 2.7425238987446754e-06, "loss": 0.0143, "step": 44970 }, { "grad_norm": 0.04594491049647331, "learning_rate": 2.731732514161506e-06, "loss": 0.0137, "step": 44980 }, { "grad_norm": 0.06525512784719467, "learning_rate": 2.720961806324435e-06, "loss": 0.0118, "step": 44990 }, { "grad_norm": 0.08041693270206451, "learning_rate": 2.7102117799449266e-06, "loss": 0.0146, "step": 45000 }, { "grad_norm": 0.061671871691942215, "learning_rate": 2.6994824397254258e-06, "loss": 0.0101, "step": 45010 }, { "grad_norm": 0.06623886525630951, "learning_rate": 2.688773790359295e-06, "loss": 0.0142, "step": 45020 }, { "grad_norm": 0.0759635716676712, "learning_rate": 2.678085836530858e-06, "loss": 0.0129, "step": 45030 }, { "grad_norm": 0.06792957335710526, "learning_rate": 2.6674185829154096e-06, "loss": 0.0117, "step": 45040 }, { "grad_norm": 0.10201575607061386, "learning_rate": 2.6567720341791556e-06, "loss": 0.0139, "step": 45050 }, { "grad_norm": 0.06486080586910248, "learning_rate": 2.646146194979249e-06, "loss": 0.0121, "step": 45060 }, { "grad_norm": 0.09997846186161041, "learning_rate": 2.6355410699638094e-06, "loss": 0.0109, "step": 45070 }, { "grad_norm": 0.06902147084474564, "learning_rate": 2.6249566637718714e-06, "loss": 0.0129, "step": 45080 }, { "grad_norm": 0.06724291294813156, "learning_rate": 2.614392981033409e-06, "loss": 0.0122, "step": 45090 }, { "grad_norm": 0.08708584308624268, "learning_rate": 2.603850026369348e-06, "loss": 0.013, "step": 45100 }, { "grad_norm": 0.0819583609700203, "learning_rate": 2.593327804391521e-06, "loss": 0.0142, "step": 45110 }, { "grad_norm": 0.061995700001716614, "learning_rate": 2.5828263197027247e-06, "loss": 0.0142, "step": 45120 }, { "grad_norm": 0.15381364524364471, "learning_rate": 2.57234557689664e-06, "loss": 0.0139, "step": 45130 }, { "grad_norm": 0.05297038331627846, "learning_rate": 2.5618855805579213e-06, "loss": 0.0114, "step": 45140 }, { "grad_norm": 0.11979392170906067, "learning_rate": 2.5514463352621197e-06, "loss": 0.0129, "step": 45150 }, { "grad_norm": 0.05340486764907837, "learning_rate": 2.541027845575711e-06, "loss": 0.0149, "step": 45160 }, { "grad_norm": 0.10721472650766373, "learning_rate": 2.530630116056104e-06, "loss": 0.0137, "step": 45170 }, { "grad_norm": 0.051563020795583725, "learning_rate": 2.5202531512516114e-06, "loss": 0.0132, "step": 45180 }, { "grad_norm": 0.09133093059062958, "learning_rate": 2.509896955701474e-06, "loss": 0.0133, "step": 45190 }, { "grad_norm": 0.09806592017412186, "learning_rate": 2.4995615339358404e-06, "loss": 0.0118, "step": 45200 }, { "grad_norm": 0.08107036352157593, "learning_rate": 2.489246890475788e-06, "loss": 0.0111, "step": 45210 }, { "grad_norm": 0.09080179035663605, "learning_rate": 2.4789530298332696e-06, "loss": 0.0094, "step": 45220 }, { "grad_norm": 0.08867940306663513, "learning_rate": 2.4686799565111885e-06, "loss": 0.0121, "step": 45230 }, { "grad_norm": 0.07794011384248734, "learning_rate": 2.4584276750033332e-06, "loss": 0.0123, "step": 45240 }, { "grad_norm": 0.04847053810954094, "learning_rate": 2.448196189794383e-06, "loss": 0.0124, "step": 45250 }, { "grad_norm": 0.09617339074611664, "learning_rate": 2.437985505359963e-06, "loss": 0.011, "step": 45260 }, { "grad_norm": 0.10027403384447098, "learning_rate": 2.4277956261665626e-06, "loss": 0.0114, "step": 45270 }, { "grad_norm": 0.07229266315698624, "learning_rate": 2.4176265566715706e-06, "loss": 0.0109, "step": 45280 }, { "grad_norm": 0.07571446895599365, "learning_rate": 2.407478301323296e-06, "loss": 0.0134, "step": 45290 }, { "grad_norm": 0.06487390398979187, "learning_rate": 2.397350864560932e-06, "loss": 0.0146, "step": 45300 }, { "grad_norm": 0.10956956446170807, "learning_rate": 2.387244250814552e-06, "loss": 0.0128, "step": 45310 }, { "grad_norm": 0.11592189967632294, "learning_rate": 2.377158464505136e-06, "loss": 0.0109, "step": 45320 }, { "grad_norm": 0.0785035490989685, "learning_rate": 2.36709351004456e-06, "loss": 0.0129, "step": 45330 }, { "grad_norm": 0.09220388531684875, "learning_rate": 2.357049391835564e-06, "loss": 0.0134, "step": 45340 }, { "grad_norm": 0.07412773370742798, "learning_rate": 2.347026114271794e-06, "loss": 0.0126, "step": 45350 }, { "grad_norm": 0.2035852074623108, "learning_rate": 2.3370236817377653e-06, "loss": 0.0154, "step": 45360 }, { "grad_norm": 0.08187414705753326, "learning_rate": 2.3270420986088947e-06, "loss": 0.0115, "step": 45370 }, { "grad_norm": 0.09418494254350662, "learning_rate": 2.3170813692514505e-06, "loss": 0.0124, "step": 45380 }, { "grad_norm": 0.08071620017290115, "learning_rate": 2.3071414980226025e-06, "loss": 0.014, "step": 45390 }, { "grad_norm": 0.06587230414152145, "learning_rate": 2.29722248927039e-06, "loss": 0.0113, "step": 45400 }, { "grad_norm": 0.0666327178478241, "learning_rate": 2.287324347333719e-06, "loss": 0.0126, "step": 45410 }, { "grad_norm": 0.0910312831401825, "learning_rate": 2.2774470765423773e-06, "loss": 0.01, "step": 45420 }, { "grad_norm": 0.09146799147129059, "learning_rate": 2.2675906812170133e-06, "loss": 0.012, "step": 45430 }, { "grad_norm": 0.06905647367238998, "learning_rate": 2.257755165669151e-06, "loss": 0.0137, "step": 45440 }, { "grad_norm": 0.15200531482696533, "learning_rate": 2.2479405342011872e-06, "loss": 0.0152, "step": 45450 }, { "grad_norm": 0.15142840147018433, "learning_rate": 2.238146791106366e-06, "loss": 0.0134, "step": 45460 }, { "grad_norm": 0.09654604643583298, "learning_rate": 2.2283739406688096e-06, "loss": 0.0132, "step": 45470 }, { "grad_norm": 0.08822927623987198, "learning_rate": 2.2186219871634774e-06, "loss": 0.0157, "step": 45480 }, { "grad_norm": 0.07527465373277664, "learning_rate": 2.2088909348562335e-06, "loss": 0.0108, "step": 45490 }, { "grad_norm": 0.0940910056233406, "learning_rate": 2.1991807880037517e-06, "loss": 0.0117, "step": 45500 }, { "grad_norm": 0.07685248553752899, "learning_rate": 2.1894915508535784e-06, "loss": 0.0129, "step": 45510 }, { "grad_norm": 0.0686308890581131, "learning_rate": 2.179823227644123e-06, "loss": 0.0104, "step": 45520 }, { "grad_norm": 0.10978491604328156, "learning_rate": 2.170175822604642e-06, "loss": 0.0114, "step": 45530 }, { "grad_norm": 0.06637243926525116, "learning_rate": 2.1605493399552256e-06, "loss": 0.0107, "step": 45540 }, { "grad_norm": 0.10980456322431564, "learning_rate": 2.150943783906839e-06, "loss": 0.0142, "step": 45550 }, { "grad_norm": 0.03985992446541786, "learning_rate": 2.1413591586612712e-06, "loss": 0.0127, "step": 45560 }, { "grad_norm": 0.06938248127698898, "learning_rate": 2.1317954684111576e-06, "loss": 0.0118, "step": 45570 }, { "grad_norm": 0.07361506670713425, "learning_rate": 2.1222527173399896e-06, "loss": 0.0115, "step": 45580 }, { "grad_norm": 0.11395763605833054, "learning_rate": 2.112730909622085e-06, "loss": 0.0101, "step": 45590 }, { "grad_norm": 0.08824358880519867, "learning_rate": 2.103230049422611e-06, "loss": 0.011, "step": 45600 }, { "grad_norm": 0.08217060565948486, "learning_rate": 2.0937501408975603e-06, "loss": 0.0126, "step": 45610 }, { "grad_norm": 0.07410702854394913, "learning_rate": 2.084291188193782e-06, "loss": 0.0133, "step": 45620 }, { "grad_norm": 0.11157742887735367, "learning_rate": 2.0748531954489224e-06, "loss": 0.0126, "step": 45630 }, { "grad_norm": 0.059282176196575165, "learning_rate": 2.065436166791496e-06, "loss": 0.0114, "step": 45640 }, { "grad_norm": 0.12566980719566345, "learning_rate": 2.056040106340823e-06, "loss": 0.0135, "step": 45650 }, { "grad_norm": 0.0753290131688118, "learning_rate": 2.0466650182070667e-06, "loss": 0.0106, "step": 45660 }, { "grad_norm": 0.08588395267724991, "learning_rate": 2.0373109064911964e-06, "loss": 0.0118, "step": 45670 }, { "grad_norm": 0.09714297950267792, "learning_rate": 2.027977775285028e-06, "loss": 0.0129, "step": 45680 }, { "grad_norm": 0.0724717527627945, "learning_rate": 2.0186656286711958e-06, "loss": 0.0125, "step": 45690 }, { "grad_norm": 0.07470250874757767, "learning_rate": 2.00937447072313e-06, "loss": 0.012, "step": 45700 }, { "grad_norm": 0.10397017002105713, "learning_rate": 2.0001043055051183e-06, "loss": 0.014, "step": 45710 }, { "grad_norm": 0.0971086323261261, "learning_rate": 1.9908551370722393e-06, "loss": 0.0124, "step": 45720 }, { "grad_norm": 0.08505898714065552, "learning_rate": 1.981626969470385e-06, "loss": 0.0114, "step": 45730 }, { "grad_norm": 0.08585590869188309, "learning_rate": 1.9724198067362864e-06, "loss": 0.0142, "step": 45740 }, { "grad_norm": 0.07819481939077377, "learning_rate": 1.963233652897456e-06, "loss": 0.0143, "step": 45750 }, { "grad_norm": 0.09013769030570984, "learning_rate": 1.9540685119722234e-06, "loss": 0.014, "step": 45760 }, { "grad_norm": 0.144620880484581, "learning_rate": 1.9449243879697543e-06, "loss": 0.0124, "step": 45770 }, { "grad_norm": 0.06158514320850372, "learning_rate": 1.9358012848899874e-06, "loss": 0.0121, "step": 45780 }, { "grad_norm": 0.09092529118061066, "learning_rate": 1.92669920672367e-06, "loss": 0.0152, "step": 45790 }, { "grad_norm": 0.06336814910173416, "learning_rate": 1.917618157452372e-06, "loss": 0.0133, "step": 45800 }, { "grad_norm": 0.13231119513511658, "learning_rate": 1.9085581410484498e-06, "loss": 0.0157, "step": 45810 }, { "grad_norm": 0.08805546909570694, "learning_rate": 1.8995191614750607e-06, "loss": 0.0115, "step": 45820 }, { "grad_norm": 0.0962497666478157, "learning_rate": 1.890501222686164e-06, "loss": 0.0168, "step": 45830 }, { "grad_norm": 0.07537184655666351, "learning_rate": 1.8815043286265045e-06, "loss": 0.0168, "step": 45840 }, { "grad_norm": 0.11239447444677353, "learning_rate": 1.8725284832316448e-06, "loss": 0.0153, "step": 45850 }, { "grad_norm": 0.0644657090306282, "learning_rate": 1.8635736904279055e-06, "loss": 0.0132, "step": 45860 }, { "grad_norm": 0.12235561013221741, "learning_rate": 1.8546399541324367e-06, "loss": 0.0149, "step": 45870 }, { "grad_norm": 0.04473651573061943, "learning_rate": 1.8457272782531453e-06, "loss": 0.012, "step": 45880 }, { "grad_norm": 0.0743875727057457, "learning_rate": 1.836835666688741e-06, "loss": 0.0128, "step": 45890 }, { "grad_norm": 0.06054644286632538, "learning_rate": 1.8279651233287175e-06, "loss": 0.0113, "step": 45900 }, { "grad_norm": 0.07264158874750137, "learning_rate": 1.8191156520533547e-06, "loss": 0.0117, "step": 45910 }, { "grad_norm": 0.07198909670114517, "learning_rate": 1.8102872567337059e-06, "loss": 0.012, "step": 45920 }, { "grad_norm": 0.09494653344154358, "learning_rate": 1.8014799412316152e-06, "loss": 0.0107, "step": 45930 }, { "grad_norm": 0.06343734264373779, "learning_rate": 1.7926937093997065e-06, "loss": 0.0106, "step": 45940 }, { "grad_norm": 0.09689802676439285, "learning_rate": 1.7839285650813608e-06, "loss": 0.0145, "step": 45950 }, { "grad_norm": 0.07687104493379593, "learning_rate": 1.7751845121107668e-06, "loss": 0.0109, "step": 45960 }, { "grad_norm": 0.055294934660196304, "learning_rate": 1.7664615543128648e-06, "loss": 0.0121, "step": 45970 }, { "grad_norm": 0.05642994865775108, "learning_rate": 1.7577596955033692e-06, "loss": 0.0122, "step": 45980 }, { "grad_norm": 0.07863549143075943, "learning_rate": 1.7490789394887686e-06, "loss": 0.0164, "step": 45990 }, { "grad_norm": 0.06526515632867813, "learning_rate": 1.7404192900663252e-06, "loss": 0.0115, "step": 46000 }, { "grad_norm": 0.081745445728302, "learning_rate": 1.7317807510240647e-06, "loss": 0.0113, "step": 46010 }, { "grad_norm": 0.0650157555937767, "learning_rate": 1.7231633261407642e-06, "loss": 0.0099, "step": 46020 }, { "grad_norm": 0.06667453050613403, "learning_rate": 1.7145670191859974e-06, "loss": 0.0123, "step": 46030 }, { "grad_norm": 0.05218898132443428, "learning_rate": 1.7059918339200732e-06, "loss": 0.0116, "step": 46040 }, { "grad_norm": 0.10589827597141266, "learning_rate": 1.6974377740940629e-06, "loss": 0.0121, "step": 46050 }, { "grad_norm": 0.10726291686296463, "learning_rate": 1.6889048434498066e-06, "loss": 0.013, "step": 46060 }, { "grad_norm": 0.07907558232545853, "learning_rate": 1.6803930457199023e-06, "loss": 0.0148, "step": 46070 }, { "grad_norm": 0.11464876681566238, "learning_rate": 1.6719023846276938e-06, "loss": 0.012, "step": 46080 }, { "grad_norm": 0.21910814940929413, "learning_rate": 1.6634328638872877e-06, "loss": 0.0136, "step": 46090 }, { "grad_norm": 0.08690514415502548, "learning_rate": 1.6549844872035424e-06, "loss": 0.0096, "step": 46100 }, { "grad_norm": 0.088023841381073, "learning_rate": 1.6465572582720579e-06, "loss": 0.0136, "step": 46110 }, { "grad_norm": 0.066740021109581, "learning_rate": 1.6381511807792017e-06, "loss": 0.0117, "step": 46120 }, { "grad_norm": 0.06528142094612122, "learning_rate": 1.6297662584020656e-06, "loss": 0.0126, "step": 46130 }, { "grad_norm": 0.08077261596918106, "learning_rate": 1.621402494808505e-06, "loss": 0.0106, "step": 46140 }, { "grad_norm": 0.06880804896354675, "learning_rate": 1.613059893657115e-06, "loss": 0.0115, "step": 46150 }, { "grad_norm": 0.051824238151311874, "learning_rate": 1.604738458597238e-06, "loss": 0.0111, "step": 46160 }, { "grad_norm": 0.060967255383729935, "learning_rate": 1.5964381932689398e-06, "loss": 0.0123, "step": 46170 }, { "grad_norm": 0.08858536928892136, "learning_rate": 1.5881591013030384e-06, "loss": 0.0119, "step": 46180 }, { "grad_norm": 0.07629454880952835, "learning_rate": 1.5799011863211089e-06, "loss": 0.0116, "step": 46190 }, { "grad_norm": 0.10100612789392471, "learning_rate": 1.5716644519354285e-06, "loss": 0.012, "step": 46200 }, { "grad_norm": 0.06727284938097, "learning_rate": 1.5634489017490151e-06, "loss": 0.0113, "step": 46210 }, { "grad_norm": 0.06892146170139313, "learning_rate": 1.555254539355655e-06, "loss": 0.0092, "step": 46220 }, { "grad_norm": 0.09032171219587326, "learning_rate": 1.5470813683398255e-06, "loss": 0.0124, "step": 46230 }, { "grad_norm": 0.14479416608810425, "learning_rate": 1.5389293922767444e-06, "loss": 0.0115, "step": 46240 }, { "grad_norm": 0.09538638591766357, "learning_rate": 1.530798614732376e-06, "loss": 0.0118, "step": 46250 }, { "grad_norm": 0.08301331847906113, "learning_rate": 1.5226890392633919e-06, "loss": 0.0125, "step": 46260 }, { "grad_norm": 0.06105225533246994, "learning_rate": 1.5146006694171932e-06, "loss": 0.0093, "step": 46270 }, { "grad_norm": 0.09247642755508423, "learning_rate": 1.5065335087319166e-06, "loss": 0.0093, "step": 46280 }, { "grad_norm": 0.06703849136829376, "learning_rate": 1.4984875607364057e-06, "loss": 0.0082, "step": 46290 }, { "grad_norm": 0.061206016689538956, "learning_rate": 1.4904628289502287e-06, "loss": 0.0098, "step": 46300 }, { "grad_norm": 0.06553707271814346, "learning_rate": 1.482459316883683e-06, "loss": 0.0105, "step": 46310 }, { "grad_norm": 0.1695472151041031, "learning_rate": 1.474477028037763e-06, "loss": 0.0109, "step": 46320 }, { "grad_norm": 0.06365691870450974, "learning_rate": 1.4665159659042139e-06, "loss": 0.0117, "step": 46330 }, { "grad_norm": 0.304290771484375, "learning_rate": 1.4585761339654613e-06, "loss": 0.0132, "step": 46340 }, { "grad_norm": 0.06859178841114044, "learning_rate": 1.4506575356946606e-06, "loss": 0.0109, "step": 46350 }, { "grad_norm": 0.08721417188644409, "learning_rate": 1.4427601745556796e-06, "loss": 0.0107, "step": 46360 }, { "grad_norm": 0.06467794626951218, "learning_rate": 1.4348840540030828e-06, "loss": 0.0113, "step": 46370 }, { "grad_norm": 0.0758374035358429, "learning_rate": 1.4270291774821586e-06, "loss": 0.0121, "step": 46380 }, { "grad_norm": 0.09634756296873093, "learning_rate": 1.4191955484289032e-06, "loss": 0.0121, "step": 46390 }, { "grad_norm": 0.0851946622133255, "learning_rate": 1.4113831702699976e-06, "loss": 0.0105, "step": 46400 }, { "grad_norm": 0.08662772923707962, "learning_rate": 1.4035920464228526e-06, "loss": 0.0143, "step": 46410 }, { "grad_norm": 0.08106410503387451, "learning_rate": 1.3958221802955696e-06, "loss": 0.0122, "step": 46420 }, { "grad_norm": 0.041195038706064224, "learning_rate": 1.3880735752869412e-06, "loss": 0.0108, "step": 46430 }, { "grad_norm": 0.10195690393447876, "learning_rate": 1.3803462347864893e-06, "loss": 0.0123, "step": 46440 }, { "grad_norm": 0.0740080326795578, "learning_rate": 1.372640162174399e-06, "loss": 0.0112, "step": 46450 }, { "grad_norm": 0.11108174175024033, "learning_rate": 1.3649553608215738e-06, "loss": 0.0114, "step": 46460 }, { "grad_norm": 0.09141182154417038, "learning_rate": 1.357291834089608e-06, "loss": 0.0109, "step": 46470 }, { "grad_norm": 0.1066519170999527, "learning_rate": 1.3496495853307921e-06, "loss": 0.0126, "step": 46480 }, { "grad_norm": 0.07188583165407181, "learning_rate": 1.3420286178880914e-06, "loss": 0.0097, "step": 46490 }, { "grad_norm": 0.08489549160003662, "learning_rate": 1.3344289350951943e-06, "loss": 0.0107, "step": 46500 }, { "grad_norm": 0.06269391626119614, "learning_rate": 1.3268505402764475e-06, "loss": 0.0125, "step": 46510 }, { "grad_norm": 0.10798964649438858, "learning_rate": 1.3192934367469045e-06, "loss": 0.0106, "step": 46520 }, { "grad_norm": 0.08618542551994324, "learning_rate": 1.3117576278122934e-06, "loss": 0.0158, "step": 46530 }, { "grad_norm": 0.099432073533535, "learning_rate": 1.3042431167690382e-06, "loss": 0.0117, "step": 46540 }, { "grad_norm": 0.05297955498099327, "learning_rate": 1.2967499069042377e-06, "loss": 0.0096, "step": 46550 }, { "grad_norm": 0.0650169849395752, "learning_rate": 1.2892780014956862e-06, "loss": 0.0098, "step": 46560 }, { "grad_norm": 0.04386909678578377, "learning_rate": 1.2818274038118306e-06, "loss": 0.0104, "step": 46570 }, { "grad_norm": 0.11250429600477219, "learning_rate": 1.2743981171118357e-06, "loss": 0.0139, "step": 46580 }, { "grad_norm": 0.047371894121170044, "learning_rate": 1.2669901446455024e-06, "loss": 0.008, "step": 46590 }, { "grad_norm": 0.09963285177946091, "learning_rate": 1.2596034896533549e-06, "loss": 0.0112, "step": 46600 }, { "grad_norm": 0.05495532974600792, "learning_rate": 1.2522381553665474e-06, "loss": 0.0124, "step": 46610 }, { "grad_norm": 0.10133551061153412, "learning_rate": 1.2448941450069306e-06, "loss": 0.0105, "step": 46620 }, { "grad_norm": 0.051794037222862244, "learning_rate": 1.2375714617870293e-06, "loss": 0.0089, "step": 46630 }, { "grad_norm": 0.08083542436361313, "learning_rate": 1.2302701089100366e-06, "loss": 0.0102, "step": 46640 }, { "grad_norm": 0.05814748629927635, "learning_rate": 1.2229900895698032e-06, "loss": 0.011, "step": 46650 }, { "grad_norm": 0.07320768386125565, "learning_rate": 1.2157314069508596e-06, "loss": 0.011, "step": 46660 }, { "grad_norm": 0.07689418643712997, "learning_rate": 1.2084940642284048e-06, "loss": 0.0104, "step": 46670 }, { "grad_norm": 0.13465945422649384, "learning_rate": 1.2012780645682952e-06, "loss": 0.0108, "step": 46680 }, { "grad_norm": 0.06637213379144669, "learning_rate": 1.1940834111270505e-06, "loss": 0.0098, "step": 46690 }, { "grad_norm": 0.05827229842543602, "learning_rate": 1.1869101070518695e-06, "loss": 0.0111, "step": 46700 }, { "grad_norm": 0.18320031464099884, "learning_rate": 1.1797581554805813e-06, "loss": 0.0113, "step": 46710 }, { "grad_norm": 0.06119481101632118, "learning_rate": 1.1726275595417001e-06, "loss": 0.0091, "step": 46720 }, { "grad_norm": 0.07461850345134735, "learning_rate": 1.1655183223543976e-06, "loss": 0.0124, "step": 46730 }, { "grad_norm": 0.0808042511343956, "learning_rate": 1.1584304470284912e-06, "loss": 0.0098, "step": 46740 }, { "grad_norm": 0.03551483899354935, "learning_rate": 1.1513639366644513e-06, "loss": 0.0117, "step": 46750 }, { "grad_norm": 0.09081891179084778, "learning_rate": 1.1443187943534162e-06, "loss": 0.0117, "step": 46760 }, { "grad_norm": 0.05466216802597046, "learning_rate": 1.1372950231771707e-06, "loss": 0.0109, "step": 46770 }, { "grad_norm": 0.0686897337436676, "learning_rate": 1.1302926262081515e-06, "loss": 0.0118, "step": 46780 }, { "grad_norm": 0.10184678435325623, "learning_rate": 1.1233116065094362e-06, "loss": 0.0114, "step": 46790 }, { "grad_norm": 0.07412346452474594, "learning_rate": 1.1163519671347767e-06, "loss": 0.0104, "step": 46800 }, { "grad_norm": 0.053724054247140884, "learning_rate": 1.1094137111285375e-06, "loss": 0.0134, "step": 46810 }, { "grad_norm": 0.06047271937131882, "learning_rate": 1.102496841525763e-06, "loss": 0.0123, "step": 46820 }, { "grad_norm": 0.07605777680873871, "learning_rate": 1.095601361352122e-06, "loss": 0.0128, "step": 46830 }, { "grad_norm": 0.09064611792564392, "learning_rate": 1.088727273623924e-06, "loss": 0.0091, "step": 46840 }, { "grad_norm": 0.052510034292936325, "learning_rate": 1.0818745813481356e-06, "loss": 0.01, "step": 46850 }, { "grad_norm": 0.11230063438415527, "learning_rate": 1.0750432875223594e-06, "loss": 0.009, "step": 46860 }, { "grad_norm": 0.04220069944858551, "learning_rate": 1.0682333951348323e-06, "loss": 0.0105, "step": 46870 }, { "grad_norm": 0.09793628007173538, "learning_rate": 1.0614449071644272e-06, "loss": 0.0101, "step": 46880 }, { "grad_norm": 0.05227198079228401, "learning_rate": 1.054677826580669e-06, "loss": 0.0099, "step": 46890 }, { "grad_norm": 0.0519123412668705, "learning_rate": 1.0479321563437062e-06, "loss": 0.0092, "step": 46900 }, { "grad_norm": 0.09598100185394287, "learning_rate": 1.0412078994043172e-06, "loss": 0.0116, "step": 46910 }, { "grad_norm": 0.12203603237867355, "learning_rate": 1.0345050587039273e-06, "loss": 0.0121, "step": 46920 }, { "grad_norm": 0.0844954252243042, "learning_rate": 1.0278236371745854e-06, "loss": 0.0092, "step": 46930 }, { "grad_norm": 0.10809360444545746, "learning_rate": 1.02116363773897e-06, "loss": 0.0127, "step": 46940 }, { "grad_norm": 0.08233648538589478, "learning_rate": 1.0145250633103954e-06, "loss": 0.0109, "step": 46950 }, { "grad_norm": 0.09422240406274796, "learning_rate": 1.007907916792794e-06, "loss": 0.0102, "step": 46960 }, { "grad_norm": 0.06772050261497498, "learning_rate": 1.001312201080734e-06, "loss": 0.0116, "step": 46970 }, { "grad_norm": 0.12186328321695328, "learning_rate": 9.947379190594075e-07, "loss": 0.0116, "step": 46980 }, { "grad_norm": 0.07129186391830444, "learning_rate": 9.88185073604625e-07, "loss": 0.0115, "step": 46990 }, { "grad_norm": 0.07852427661418915, "learning_rate": 9.816536675828214e-07, "loss": 0.0131, "step": 47000 }, { "grad_norm": 0.11746501922607422, "learning_rate": 9.751437038510615e-07, "loss": 0.0131, "step": 47010 }, { "grad_norm": 0.053363099694252014, "learning_rate": 9.686551852570225e-07, "loss": 0.0109, "step": 47020 }, { "grad_norm": 0.05718264356255531, "learning_rate": 9.621881146390067e-07, "loss": 0.012, "step": 47030 }, { "grad_norm": 0.09721244126558304, "learning_rate": 9.557424948259175e-07, "loss": 0.0129, "step": 47040 }, { "grad_norm": 0.07205980271100998, "learning_rate": 9.493183286373053e-07, "loss": 0.01, "step": 47050 }, { "grad_norm": 0.05293399840593338, "learning_rate": 9.429156188833111e-07, "loss": 0.0107, "step": 47060 }, { "grad_norm": 0.04874548316001892, "learning_rate": 9.365343683646944e-07, "loss": 0.0115, "step": 47070 }, { "grad_norm": 0.05515584722161293, "learning_rate": 9.301745798728334e-07, "loss": 0.0101, "step": 47080 }, { "grad_norm": 0.07287522405385971, "learning_rate": 9.238362561897141e-07, "loss": 0.0117, "step": 47090 }, { "grad_norm": 0.07502900809049606, "learning_rate": 9.17519400087935e-07, "loss": 0.0127, "step": 47100 }, { "grad_norm": 0.1115260049700737, "learning_rate": 9.112240143307138e-07, "loss": 0.0126, "step": 47110 }, { "grad_norm": 0.05068874731659889, "learning_rate": 9.049501016718475e-07, "loss": 0.0112, "step": 47120 }, { "grad_norm": 0.05297445133328438, "learning_rate": 8.986976648557743e-07, "loss": 0.0103, "step": 47130 }, { "grad_norm": 0.08426675200462341, "learning_rate": 8.924667066175119e-07, "loss": 0.0148, "step": 47140 }, { "grad_norm": 0.10323990136384964, "learning_rate": 8.86257229682702e-07, "loss": 0.0108, "step": 47150 }, { "grad_norm": 0.10963299870491028, "learning_rate": 8.800692367675722e-07, "loss": 0.0159, "step": 47160 }, { "grad_norm": 0.08064355701208115, "learning_rate": 8.739027305789683e-07, "loss": 0.0145, "step": 47170 }, { "grad_norm": 0.0837382972240448, "learning_rate": 8.677577138143267e-07, "loss": 0.0131, "step": 47180 }, { "grad_norm": 0.10906544327735901, "learning_rate": 8.616341891616864e-07, "loss": 0.0124, "step": 47190 }, { "grad_norm": 0.08139335364103317, "learning_rate": 8.555321592996824e-07, "loss": 0.0099, "step": 47200 }, { "grad_norm": 0.051958419382572174, "learning_rate": 8.494516268975572e-07, "loss": 0.0117, "step": 47210 }, { "grad_norm": 0.1036590039730072, "learning_rate": 8.433925946151333e-07, "loss": 0.0126, "step": 47220 }, { "grad_norm": 0.06589127331972122, "learning_rate": 8.373550651028406e-07, "loss": 0.0133, "step": 47230 }, { "grad_norm": 0.06108089163899422, "learning_rate": 8.313390410017052e-07, "loss": 0.0099, "step": 47240 }, { "grad_norm": 0.061756569892168045, "learning_rate": 8.253445249433334e-07, "loss": 0.0109, "step": 47250 }, { "grad_norm": 0.30074983835220337, "learning_rate": 8.193715195499329e-07, "loss": 0.0158, "step": 47260 }, { "grad_norm": 0.05801782384514809, "learning_rate": 8.134200274342918e-07, "loss": 0.0103, "step": 47270 }, { "grad_norm": 0.0903700441122055, "learning_rate": 8.074900511998107e-07, "loss": 0.0125, "step": 47280 }, { "grad_norm": 0.045317478477954865, "learning_rate": 8.01581593440448e-07, "loss": 0.0132, "step": 47290 }, { "grad_norm": 0.06980827450752258, "learning_rate": 7.956946567407753e-07, "loss": 0.0113, "step": 47300 }, { "grad_norm": 0.04530477151274681, "learning_rate": 7.898292436759324e-07, "loss": 0.0133, "step": 47310 }, { "grad_norm": 0.07617703080177307, "learning_rate": 7.839853568116506e-07, "loss": 0.012, "step": 47320 }, { "grad_norm": 0.058311283588409424, "learning_rate": 7.78162998704246e-07, "loss": 0.011, "step": 47330 }, { "grad_norm": 0.08916760236024857, "learning_rate": 7.723621719006147e-07, "loss": 0.0092, "step": 47340 }, { "grad_norm": 0.05624203756451607, "learning_rate": 7.665828789382323e-07, "loss": 0.0105, "step": 47350 }, { "grad_norm": 0.055876992642879486, "learning_rate": 7.608251223451602e-07, "loss": 0.0121, "step": 47360 }, { "grad_norm": 0.03623562306165695, "learning_rate": 7.550889046400444e-07, "loss": 0.0233, "step": 47370 }, { "grad_norm": 0.0740228146314621, "learning_rate": 7.493742283320893e-07, "loss": 0.0102, "step": 47380 }, { "grad_norm": 0.048638965934515, "learning_rate": 7.43681095921095e-07, "loss": 0.0107, "step": 47390 }, { "grad_norm": 0.10663942247629166, "learning_rate": 7.380095098974305e-07, "loss": 0.0107, "step": 47400 }, { "grad_norm": 0.05744173750281334, "learning_rate": 7.323594727420446e-07, "loss": 0.011, "step": 47410 }, { "grad_norm": 0.1297844648361206, "learning_rate": 7.267309869264438e-07, "loss": 0.0147, "step": 47420 }, { "grad_norm": 0.05639788508415222, "learning_rate": 7.211240549127252e-07, "loss": 0.0117, "step": 47430 }, { "grad_norm": 0.07453762739896774, "learning_rate": 7.155386791535546e-07, "loss": 0.013, "step": 47440 }, { "grad_norm": 0.0402723103761673, "learning_rate": 7.09974862092161e-07, "loss": 0.011, "step": 47450 }, { "grad_norm": 0.08062425255775452, "learning_rate": 7.044326061623529e-07, "loss": 0.0128, "step": 47460 }, { "grad_norm": 0.08906246721744537, "learning_rate": 6.989119137884914e-07, "loss": 0.0143, "step": 47470 }, { "grad_norm": 0.08685782551765442, "learning_rate": 6.934127873855223e-07, "loss": 0.0131, "step": 47480 }, { "grad_norm": 0.0935058444738388, "learning_rate": 6.879352293589492e-07, "loss": 0.0102, "step": 47490 }, { "grad_norm": 0.08210422843694687, "learning_rate": 6.824792421048388e-07, "loss": 0.0134, "step": 47500 } ], "logging_steps": 10, "max_steps": 50000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 2500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }