| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.2458036734335179, |
| "eval_steps": 500, |
| "global_step": 2797, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0008788118463836892, |
| "grad_norm": 0.7183610796928406, |
| "learning_rate": 1.1061797752808989e-06, |
| "loss": 0.5856, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.0017576236927673784, |
| "grad_norm": 0.6834720373153687, |
| "learning_rate": 2.2241573033707866e-06, |
| "loss": 0.5868, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.002636435539151068, |
| "grad_norm": 0.6469134092330933, |
| "learning_rate": 3.342134831460675e-06, |
| "loss": 0.5883, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.003515247385534757, |
| "grad_norm": 0.5836713910102844, |
| "learning_rate": 4.4601123595505624e-06, |
| "loss": 0.6023, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.004394059231918446, |
| "grad_norm": 0.6964038610458374, |
| "learning_rate": 5.5780898876404495e-06, |
| "loss": 0.5766, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.005272871078302136, |
| "grad_norm": 1.2740495204925537, |
| "learning_rate": 6.6960674157303374e-06, |
| "loss": 0.5805, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.006151682924685825, |
| "grad_norm": 0.6438924670219421, |
| "learning_rate": 7.814044943820226e-06, |
| "loss": 0.5624, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.007030494771069514, |
| "grad_norm": 0.6378642320632935, |
| "learning_rate": 8.932022471910112e-06, |
| "loss": 0.561, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.007909306617453203, |
| "grad_norm": 0.6773518323898315, |
| "learning_rate": 1.005e-05, |
| "loss": 0.5738, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.008788118463836892, |
| "grad_norm": 0.5750642418861389, |
| "learning_rate": 1.1167977528089888e-05, |
| "loss": 0.565, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.009666930310220583, |
| "grad_norm": 0.6147071719169617, |
| "learning_rate": 1.2285955056179776e-05, |
| "loss": 0.5776, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.010545742156604272, |
| "grad_norm": 0.5244885087013245, |
| "learning_rate": 1.3403932584269664e-05, |
| "loss": 0.5569, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.01142455400298796, |
| "grad_norm": 0.7714810371398926, |
| "learning_rate": 1.4521910112359552e-05, |
| "loss": 0.5731, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.01230336584937165, |
| "grad_norm": 0.681827187538147, |
| "learning_rate": 1.563988764044944e-05, |
| "loss": 0.5487, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.013182177695755339, |
| "grad_norm": 0.6038880944252014, |
| "learning_rate": 1.6757865168539328e-05, |
| "loss": 0.5655, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.014060989542139028, |
| "grad_norm": 0.5889508128166199, |
| "learning_rate": 1.7875842696629214e-05, |
| "loss": 0.5804, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.014939801388522717, |
| "grad_norm": 0.6750523447990417, |
| "learning_rate": 1.89938202247191e-05, |
| "loss": 0.5946, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.015818613234906406, |
| "grad_norm": 0.5245691537857056, |
| "learning_rate": 2e-05, |
| "loss": 0.5724, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.016697425081290095, |
| "grad_norm": 0.6065133810043335, |
| "learning_rate": 2e-05, |
| "loss": 0.5716, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.017576236927673784, |
| "grad_norm": 0.5226805806159973, |
| "learning_rate": 2e-05, |
| "loss": 0.5864, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.018455048774057473, |
| "grad_norm": 0.5066975355148315, |
| "learning_rate": 2e-05, |
| "loss": 0.564, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.019333860620441165, |
| "grad_norm": 0.5714482069015503, |
| "learning_rate": 2e-05, |
| "loss": 0.5555, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.020212672466824854, |
| "grad_norm": 0.4541548788547516, |
| "learning_rate": 2e-05, |
| "loss": 0.5641, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.021091484313208543, |
| "grad_norm": 0.4934563934803009, |
| "learning_rate": 2e-05, |
| "loss": 0.5667, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.021970296159592232, |
| "grad_norm": 0.4963897168636322, |
| "learning_rate": 2e-05, |
| "loss": 0.5721, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.02284910800597592, |
| "grad_norm": 0.45243293046951294, |
| "learning_rate": 2e-05, |
| "loss": 0.5924, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.02372791985235961, |
| "grad_norm": 0.4372263550758362, |
| "learning_rate": 2e-05, |
| "loss": 0.6031, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.0246067316987433, |
| "grad_norm": 0.4153901934623718, |
| "learning_rate": 2e-05, |
| "loss": 0.5682, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.025485543545126988, |
| "grad_norm": 0.43816617131233215, |
| "learning_rate": 2e-05, |
| "loss": 0.5886, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.026364355391510677, |
| "grad_norm": 0.4889500141143799, |
| "learning_rate": 2e-05, |
| "loss": 0.577, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.027243167237894366, |
| "grad_norm": 0.38636863231658936, |
| "learning_rate": 2e-05, |
| "loss": 0.6108, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.028121979084278055, |
| "grad_norm": 0.46534135937690735, |
| "learning_rate": 2e-05, |
| "loss": 0.5867, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.029000790930661744, |
| "grad_norm": 0.9078437685966492, |
| "learning_rate": 2e-05, |
| "loss": 0.5642, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.029879602777045433, |
| "grad_norm": 0.5279733538627625, |
| "learning_rate": 2e-05, |
| "loss": 0.5767, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.030758414623429122, |
| "grad_norm": 0.4053654372692108, |
| "learning_rate": 2e-05, |
| "loss": 0.5714, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.03163722646981281, |
| "grad_norm": 0.3655475676059723, |
| "learning_rate": 2e-05, |
| "loss": 0.5658, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.032516038316196504, |
| "grad_norm": 0.42354220151901245, |
| "learning_rate": 2e-05, |
| "loss": 0.5735, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.03339485016258019, |
| "grad_norm": 0.4039507806301117, |
| "learning_rate": 2e-05, |
| "loss": 0.5713, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.03427366200896388, |
| "grad_norm": 0.3883609473705292, |
| "learning_rate": 2e-05, |
| "loss": 0.5736, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.03515247385534757, |
| "grad_norm": 0.43263545632362366, |
| "learning_rate": 2e-05, |
| "loss": 0.6093, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.03603128570173126, |
| "grad_norm": 0.4262617528438568, |
| "learning_rate": 2e-05, |
| "loss": 0.57, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.036910097548114945, |
| "grad_norm": 0.39129018783569336, |
| "learning_rate": 2e-05, |
| "loss": 0.5798, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.03778890939449864, |
| "grad_norm": 0.4544897973537445, |
| "learning_rate": 2e-05, |
| "loss": 0.5747, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.03866772124088233, |
| "grad_norm": 0.4992966651916504, |
| "learning_rate": 2e-05, |
| "loss": 0.6086, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.039546533087266016, |
| "grad_norm": 0.411885142326355, |
| "learning_rate": 2e-05, |
| "loss": 0.5822, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.04042534493364971, |
| "grad_norm": 0.3853900134563446, |
| "learning_rate": 2e-05, |
| "loss": 0.5449, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.041304156780033394, |
| "grad_norm": 0.810398519039154, |
| "learning_rate": 2e-05, |
| "loss": 0.5477, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.042182968626417086, |
| "grad_norm": 0.45868226885795593, |
| "learning_rate": 2e-05, |
| "loss": 0.5688, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.04306178047280077, |
| "grad_norm": 0.378400593996048, |
| "learning_rate": 2e-05, |
| "loss": 0.5556, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.043940592319184464, |
| "grad_norm": 0.37122344970703125, |
| "learning_rate": 2e-05, |
| "loss": 0.5601, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.04481940416556815, |
| "grad_norm": 0.390835165977478, |
| "learning_rate": 2e-05, |
| "loss": 0.5872, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.04569821601195184, |
| "grad_norm": 0.41930919885635376, |
| "learning_rate": 2e-05, |
| "loss": 0.5501, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.04657702785833553, |
| "grad_norm": 0.4113750755786896, |
| "learning_rate": 2e-05, |
| "loss": 0.5878, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.04745583970471922, |
| "grad_norm": 0.6010010838508606, |
| "learning_rate": 2e-05, |
| "loss": 0.5784, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.048334651551102906, |
| "grad_norm": 0.3737349510192871, |
| "learning_rate": 2e-05, |
| "loss": 0.583, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.0492134633974866, |
| "grad_norm": 0.4933265745639801, |
| "learning_rate": 2e-05, |
| "loss": 0.5771, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.05009227524387029, |
| "grad_norm": 0.4226865768432617, |
| "learning_rate": 2e-05, |
| "loss": 0.547, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.050971087090253976, |
| "grad_norm": 0.36813685297966003, |
| "learning_rate": 2e-05, |
| "loss": 0.5718, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.05184989893663767, |
| "grad_norm": 0.3392640948295593, |
| "learning_rate": 2e-05, |
| "loss": 0.5665, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.052728710783021354, |
| "grad_norm": 0.41001757979393005, |
| "learning_rate": 2e-05, |
| "loss": 0.5783, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.05360752262940505, |
| "grad_norm": 0.3661254346370697, |
| "learning_rate": 2e-05, |
| "loss": 0.5634, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.05448633447578873, |
| "grad_norm": 0.532596230506897, |
| "learning_rate": 2e-05, |
| "loss": 0.5944, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.055365146322172425, |
| "grad_norm": 0.3782620429992676, |
| "learning_rate": 2e-05, |
| "loss": 0.5559, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.05624395816855611, |
| "grad_norm": 0.36520808935165405, |
| "learning_rate": 2e-05, |
| "loss": 0.5781, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.0571227700149398, |
| "grad_norm": 0.422725111246109, |
| "learning_rate": 2e-05, |
| "loss": 0.5887, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.05800158186132349, |
| "grad_norm": 0.47393128275871277, |
| "learning_rate": 2e-05, |
| "loss": 0.5897, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.05888039370770718, |
| "grad_norm": 0.37145882844924927, |
| "learning_rate": 2e-05, |
| "loss": 0.6046, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.059759205554090866, |
| "grad_norm": 0.37511953711509705, |
| "learning_rate": 2e-05, |
| "loss": 0.5616, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.06063801740047456, |
| "grad_norm": 0.36933016777038574, |
| "learning_rate": 2e-05, |
| "loss": 0.575, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.061516829246858244, |
| "grad_norm": 0.36136773228645325, |
| "learning_rate": 2e-05, |
| "loss": 0.5674, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.06239564109324194, |
| "grad_norm": 0.387931227684021, |
| "learning_rate": 2e-05, |
| "loss": 0.552, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.06327445293962562, |
| "grad_norm": 0.357898473739624, |
| "learning_rate": 2e-05, |
| "loss": 0.5733, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.06415326478600931, |
| "grad_norm": 0.45435261726379395, |
| "learning_rate": 2e-05, |
| "loss": 0.5475, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.06503207663239301, |
| "grad_norm": 0.3749592900276184, |
| "learning_rate": 2e-05, |
| "loss": 0.5828, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.0659108884787767, |
| "grad_norm": 0.552718997001648, |
| "learning_rate": 2e-05, |
| "loss": 0.5765, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.06678970032516038, |
| "grad_norm": 0.3859593868255615, |
| "learning_rate": 2e-05, |
| "loss": 0.5642, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.06766851217154407, |
| "grad_norm": 0.3709048628807068, |
| "learning_rate": 2e-05, |
| "loss": 0.5861, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.06854732401792776, |
| "grad_norm": 0.36113089323043823, |
| "learning_rate": 2e-05, |
| "loss": 0.5777, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.06942613586431146, |
| "grad_norm": 0.355157732963562, |
| "learning_rate": 2e-05, |
| "loss": 0.5779, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.07030494771069513, |
| "grad_norm": 0.34941405057907104, |
| "learning_rate": 2e-05, |
| "loss": 0.5589, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.07118375955707883, |
| "grad_norm": 0.37520429491996765, |
| "learning_rate": 2e-05, |
| "loss": 0.5619, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.07206257140346252, |
| "grad_norm": 0.38500913977622986, |
| "learning_rate": 2e-05, |
| "loss": 0.5532, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.07294138324984621, |
| "grad_norm": 0.35824254155158997, |
| "learning_rate": 2e-05, |
| "loss": 0.5778, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.07382019509622989, |
| "grad_norm": 0.3841072618961334, |
| "learning_rate": 2e-05, |
| "loss": 0.5733, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.07469900694261358, |
| "grad_norm": 0.38087451457977295, |
| "learning_rate": 2e-05, |
| "loss": 0.5752, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.07557781878899728, |
| "grad_norm": 0.3607955873012543, |
| "learning_rate": 2e-05, |
| "loss": 0.5509, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.07645663063538097, |
| "grad_norm": 0.36730197072029114, |
| "learning_rate": 2e-05, |
| "loss": 0.5664, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.07733544248176466, |
| "grad_norm": 0.35436132550239563, |
| "learning_rate": 2e-05, |
| "loss": 0.5599, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.07821425432814834, |
| "grad_norm": 0.3620469570159912, |
| "learning_rate": 2e-05, |
| "loss": 0.5821, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.07909306617453203, |
| "grad_norm": 0.3576803505420685, |
| "learning_rate": 2e-05, |
| "loss": 0.5859, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.07997187802091572, |
| "grad_norm": 0.3401060104370117, |
| "learning_rate": 2e-05, |
| "loss": 0.5649, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.08085068986729942, |
| "grad_norm": 0.35253942012786865, |
| "learning_rate": 2e-05, |
| "loss": 0.5624, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.0817295017136831, |
| "grad_norm": 0.3855127692222595, |
| "learning_rate": 2e-05, |
| "loss": 0.5648, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.08260831356006679, |
| "grad_norm": 0.3717329502105713, |
| "learning_rate": 2e-05, |
| "loss": 0.5602, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.08348712540645048, |
| "grad_norm": 0.6942209601402283, |
| "learning_rate": 2e-05, |
| "loss": 0.5947, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.08436593725283417, |
| "grad_norm": 0.39715272188186646, |
| "learning_rate": 2e-05, |
| "loss": 0.5669, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.08524474909921785, |
| "grad_norm": 0.35998284816741943, |
| "learning_rate": 2e-05, |
| "loss": 0.5785, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.08612356094560154, |
| "grad_norm": 0.4870074391365051, |
| "learning_rate": 2e-05, |
| "loss": 0.6038, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.08700237279198524, |
| "grad_norm": 0.35796302556991577, |
| "learning_rate": 2e-05, |
| "loss": 0.5861, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.08788118463836893, |
| "grad_norm": 0.3613901436328888, |
| "learning_rate": 2e-05, |
| "loss": 0.5718, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.08875999648475262, |
| "grad_norm": 0.3208099603652954, |
| "learning_rate": 2e-05, |
| "loss": 0.5484, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.0896388083311363, |
| "grad_norm": 0.46393319964408875, |
| "learning_rate": 2e-05, |
| "loss": 0.5745, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.09051762017751999, |
| "grad_norm": 0.39247334003448486, |
| "learning_rate": 2e-05, |
| "loss": 0.554, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.09139643202390368, |
| "grad_norm": 0.3445221185684204, |
| "learning_rate": 2e-05, |
| "loss": 0.5677, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.09227524387028738, |
| "grad_norm": 0.3665354549884796, |
| "learning_rate": 2e-05, |
| "loss": 0.5512, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.09315405571667106, |
| "grad_norm": 0.4386295974254608, |
| "learning_rate": 2e-05, |
| "loss": 0.5789, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.09403286756305475, |
| "grad_norm": 0.33483144640922546, |
| "learning_rate": 2e-05, |
| "loss": 0.5744, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.09491167940943844, |
| "grad_norm": 0.36662590503692627, |
| "learning_rate": 2e-05, |
| "loss": 0.5612, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.09579049125582213, |
| "grad_norm": 0.3931642770767212, |
| "learning_rate": 2e-05, |
| "loss": 0.5816, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.09666930310220581, |
| "grad_norm": 0.37220466136932373, |
| "learning_rate": 2e-05, |
| "loss": 0.5612, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.0975481149485895, |
| "grad_norm": 0.40914490818977356, |
| "learning_rate": 2e-05, |
| "loss": 0.574, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.0984269267949732, |
| "grad_norm": 0.3663410246372223, |
| "learning_rate": 2e-05, |
| "loss": 0.5576, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.09930573864135689, |
| "grad_norm": 0.3547806143760681, |
| "learning_rate": 2e-05, |
| "loss": 0.5779, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.10018455048774058, |
| "grad_norm": 0.3315262794494629, |
| "learning_rate": 2e-05, |
| "loss": 0.5607, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.10106336233412426, |
| "grad_norm": 0.34696242213249207, |
| "learning_rate": 2e-05, |
| "loss": 0.5796, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.10194217418050795, |
| "grad_norm": 0.38509389758110046, |
| "learning_rate": 2e-05, |
| "loss": 0.5713, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.10282098602689164, |
| "grad_norm": 0.37624210119247437, |
| "learning_rate": 2e-05, |
| "loss": 0.5464, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.10369979787327534, |
| "grad_norm": 0.34956589341163635, |
| "learning_rate": 2e-05, |
| "loss": 0.5578, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.10457860971965902, |
| "grad_norm": 0.35008957982063293, |
| "learning_rate": 2e-05, |
| "loss": 0.5441, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.10545742156604271, |
| "grad_norm": 0.34270432591438293, |
| "learning_rate": 2e-05, |
| "loss": 0.5645, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.1063362334124264, |
| "grad_norm": 0.4176119863986969, |
| "learning_rate": 2e-05, |
| "loss": 0.5866, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.1072150452588101, |
| "grad_norm": 0.4271017014980316, |
| "learning_rate": 2e-05, |
| "loss": 0.5854, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.10809385710519377, |
| "grad_norm": 0.3339524269104004, |
| "learning_rate": 2e-05, |
| "loss": 0.5748, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.10897266895157746, |
| "grad_norm": 0.4067085385322571, |
| "learning_rate": 2e-05, |
| "loss": 0.5779, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.10985148079796116, |
| "grad_norm": 0.3834608197212219, |
| "learning_rate": 2e-05, |
| "loss": 0.5552, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.11073029264434485, |
| "grad_norm": 0.38505619764328003, |
| "learning_rate": 2e-05, |
| "loss": 0.5785, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.11160910449072854, |
| "grad_norm": 0.3317438066005707, |
| "learning_rate": 2e-05, |
| "loss": 0.5673, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.11248791633711222, |
| "grad_norm": 0.37222227454185486, |
| "learning_rate": 2e-05, |
| "loss": 0.5762, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.11336672818349591, |
| "grad_norm": 0.40460169315338135, |
| "learning_rate": 2e-05, |
| "loss": 0.5747, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.1142455400298796, |
| "grad_norm": 0.3579246997833252, |
| "learning_rate": 2e-05, |
| "loss": 0.5895, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.1151243518762633, |
| "grad_norm": 0.34237009286880493, |
| "learning_rate": 2e-05, |
| "loss": 0.5574, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.11600316372264698, |
| "grad_norm": 0.3422780930995941, |
| "learning_rate": 2e-05, |
| "loss": 0.5913, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.11688197556903067, |
| "grad_norm": 0.3603161573410034, |
| "learning_rate": 2e-05, |
| "loss": 0.5465, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.11776078741541436, |
| "grad_norm": 0.34054920077323914, |
| "learning_rate": 2e-05, |
| "loss": 0.5698, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.11863959926179805, |
| "grad_norm": 0.35030534863471985, |
| "learning_rate": 2e-05, |
| "loss": 0.5953, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.11951841110818173, |
| "grad_norm": 0.4028947353363037, |
| "learning_rate": 2e-05, |
| "loss": 0.5675, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.12039722295456542, |
| "grad_norm": 0.3338577151298523, |
| "learning_rate": 2e-05, |
| "loss": 0.5659, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.12127603480094912, |
| "grad_norm": 0.3435846269130707, |
| "learning_rate": 2e-05, |
| "loss": 0.5833, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.12215484664733281, |
| "grad_norm": 0.6601398587226868, |
| "learning_rate": 2e-05, |
| "loss": 0.5554, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.12303365849371649, |
| "grad_norm": 0.37921780347824097, |
| "learning_rate": 2e-05, |
| "loss": 0.5565, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.12391247034010018, |
| "grad_norm": 0.34458211064338684, |
| "learning_rate": 2e-05, |
| "loss": 0.5595, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.12479128218648387, |
| "grad_norm": 0.4461982548236847, |
| "learning_rate": 2e-05, |
| "loss": 0.5616, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.12567009403286755, |
| "grad_norm": 0.4057040512561798, |
| "learning_rate": 2e-05, |
| "loss": 0.5754, |
| "step": 1430 |
| }, |
| { |
| "epoch": 0.12654890587925124, |
| "grad_norm": 0.35918596386909485, |
| "learning_rate": 2e-05, |
| "loss": 0.5781, |
| "step": 1440 |
| }, |
| { |
| "epoch": 0.12742771772563494, |
| "grad_norm": 0.33282655477523804, |
| "learning_rate": 2e-05, |
| "loss": 0.5608, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.12830652957201863, |
| "grad_norm": 0.4065190553665161, |
| "learning_rate": 2e-05, |
| "loss": 0.5608, |
| "step": 1460 |
| }, |
| { |
| "epoch": 0.12918534141840232, |
| "grad_norm": 0.3624435365200043, |
| "learning_rate": 2e-05, |
| "loss": 0.5518, |
| "step": 1470 |
| }, |
| { |
| "epoch": 0.13006415326478601, |
| "grad_norm": 0.37889572978019714, |
| "learning_rate": 2e-05, |
| "loss": 0.5649, |
| "step": 1480 |
| }, |
| { |
| "epoch": 0.1309429651111697, |
| "grad_norm": 0.39061930775642395, |
| "learning_rate": 2e-05, |
| "loss": 0.5631, |
| "step": 1490 |
| }, |
| { |
| "epoch": 0.1318217769575534, |
| "grad_norm": 0.3646930754184723, |
| "learning_rate": 2e-05, |
| "loss": 0.5544, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.13270058880393706, |
| "grad_norm": 0.3735094964504242, |
| "learning_rate": 2e-05, |
| "loss": 0.5566, |
| "step": 1510 |
| }, |
| { |
| "epoch": 0.13357940065032076, |
| "grad_norm": 0.32499170303344727, |
| "learning_rate": 2e-05, |
| "loss": 0.5561, |
| "step": 1520 |
| }, |
| { |
| "epoch": 0.13445821249670445, |
| "grad_norm": 0.3363400399684906, |
| "learning_rate": 2e-05, |
| "loss": 0.5578, |
| "step": 1530 |
| }, |
| { |
| "epoch": 0.13533702434308814, |
| "grad_norm": 0.3458295464515686, |
| "learning_rate": 2e-05, |
| "loss": 0.6095, |
| "step": 1540 |
| }, |
| { |
| "epoch": 0.13621583618947183, |
| "grad_norm": 0.3249399662017822, |
| "learning_rate": 2e-05, |
| "loss": 0.5599, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.13709464803585553, |
| "grad_norm": 0.33804813027381897, |
| "learning_rate": 2e-05, |
| "loss": 0.5561, |
| "step": 1560 |
| }, |
| { |
| "epoch": 0.13797345988223922, |
| "grad_norm": 0.34222960472106934, |
| "learning_rate": 2e-05, |
| "loss": 0.6034, |
| "step": 1570 |
| }, |
| { |
| "epoch": 0.1388522717286229, |
| "grad_norm": 0.33628180623054504, |
| "learning_rate": 2e-05, |
| "loss": 0.5703, |
| "step": 1580 |
| }, |
| { |
| "epoch": 0.1397310835750066, |
| "grad_norm": 0.36276009678840637, |
| "learning_rate": 2e-05, |
| "loss": 0.5843, |
| "step": 1590 |
| }, |
| { |
| "epoch": 0.14060989542139027, |
| "grad_norm": 0.355282723903656, |
| "learning_rate": 2e-05, |
| "loss": 0.5719, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.14148870726777396, |
| "grad_norm": 0.3517269194126129, |
| "learning_rate": 2e-05, |
| "loss": 0.5629, |
| "step": 1610 |
| }, |
| { |
| "epoch": 0.14236751911415765, |
| "grad_norm": 0.35511642694473267, |
| "learning_rate": 2e-05, |
| "loss": 0.5458, |
| "step": 1620 |
| }, |
| { |
| "epoch": 0.14324633096054135, |
| "grad_norm": 0.3451294004917145, |
| "learning_rate": 2e-05, |
| "loss": 0.572, |
| "step": 1630 |
| }, |
| { |
| "epoch": 0.14412514280692504, |
| "grad_norm": 0.3705539107322693, |
| "learning_rate": 2e-05, |
| "loss": 0.5335, |
| "step": 1640 |
| }, |
| { |
| "epoch": 0.14500395465330873, |
| "grad_norm": 0.43997931480407715, |
| "learning_rate": 2e-05, |
| "loss": 0.5767, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.14588276649969242, |
| "grad_norm": 0.35617420077323914, |
| "learning_rate": 2e-05, |
| "loss": 0.57, |
| "step": 1660 |
| }, |
| { |
| "epoch": 0.14676157834607612, |
| "grad_norm": 0.37875717878341675, |
| "learning_rate": 2e-05, |
| "loss": 0.5734, |
| "step": 1670 |
| }, |
| { |
| "epoch": 0.14764039019245978, |
| "grad_norm": 0.3402183949947357, |
| "learning_rate": 2e-05, |
| "loss": 0.5483, |
| "step": 1680 |
| }, |
| { |
| "epoch": 0.14851920203884347, |
| "grad_norm": 0.3516283333301544, |
| "learning_rate": 2e-05, |
| "loss": 0.5653, |
| "step": 1690 |
| }, |
| { |
| "epoch": 0.14939801388522717, |
| "grad_norm": 0.854839563369751, |
| "learning_rate": 2e-05, |
| "loss": 0.5392, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.15027682573161086, |
| "grad_norm": 0.3747766315937042, |
| "learning_rate": 2e-05, |
| "loss": 0.5819, |
| "step": 1710 |
| }, |
| { |
| "epoch": 0.15115563757799455, |
| "grad_norm": 0.370029479265213, |
| "learning_rate": 2e-05, |
| "loss": 0.5952, |
| "step": 1720 |
| }, |
| { |
| "epoch": 0.15203444942437824, |
| "grad_norm": 0.3330126404762268, |
| "learning_rate": 2e-05, |
| "loss": 0.5299, |
| "step": 1730 |
| }, |
| { |
| "epoch": 0.15291326127076194, |
| "grad_norm": 0.36191222071647644, |
| "learning_rate": 2e-05, |
| "loss": 0.5552, |
| "step": 1740 |
| }, |
| { |
| "epoch": 0.15379207311714563, |
| "grad_norm": 0.3550053536891937, |
| "learning_rate": 2e-05, |
| "loss": 0.5432, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.15467088496352932, |
| "grad_norm": 0.37736421823501587, |
| "learning_rate": 2e-05, |
| "loss": 0.5751, |
| "step": 1760 |
| }, |
| { |
| "epoch": 0.15554969680991299, |
| "grad_norm": 0.42097368836402893, |
| "learning_rate": 2e-05, |
| "loss": 0.5758, |
| "step": 1770 |
| }, |
| { |
| "epoch": 0.15642850865629668, |
| "grad_norm": 0.4581955671310425, |
| "learning_rate": 2e-05, |
| "loss": 0.5722, |
| "step": 1780 |
| }, |
| { |
| "epoch": 0.15730732050268037, |
| "grad_norm": 0.329158216714859, |
| "learning_rate": 2e-05, |
| "loss": 0.5497, |
| "step": 1790 |
| }, |
| { |
| "epoch": 0.15818613234906406, |
| "grad_norm": 0.474506139755249, |
| "learning_rate": 2e-05, |
| "loss": 0.5526, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.15906494419544775, |
| "grad_norm": 0.3612702488899231, |
| "learning_rate": 2e-05, |
| "loss": 0.5653, |
| "step": 1810 |
| }, |
| { |
| "epoch": 0.15994375604183145, |
| "grad_norm": 0.38358527421951294, |
| "learning_rate": 2e-05, |
| "loss": 0.5698, |
| "step": 1820 |
| }, |
| { |
| "epoch": 0.16082256788821514, |
| "grad_norm": 0.6245725750923157, |
| "learning_rate": 2e-05, |
| "loss": 0.5551, |
| "step": 1830 |
| }, |
| { |
| "epoch": 0.16170137973459883, |
| "grad_norm": 0.47649702429771423, |
| "learning_rate": 2e-05, |
| "loss": 0.586, |
| "step": 1840 |
| }, |
| { |
| "epoch": 0.16258019158098252, |
| "grad_norm": 0.3675776422023773, |
| "learning_rate": 2e-05, |
| "loss": 0.5589, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.1634590034273662, |
| "grad_norm": 0.498647004365921, |
| "learning_rate": 2e-05, |
| "loss": 0.5741, |
| "step": 1860 |
| }, |
| { |
| "epoch": 0.16433781527374988, |
| "grad_norm": 0.3350846767425537, |
| "learning_rate": 2e-05, |
| "loss": 0.5476, |
| "step": 1870 |
| }, |
| { |
| "epoch": 0.16521662712013357, |
| "grad_norm": 0.3603364825248718, |
| "learning_rate": 2e-05, |
| "loss": 0.5652, |
| "step": 1880 |
| }, |
| { |
| "epoch": 0.16609543896651727, |
| "grad_norm": 0.35095348954200745, |
| "learning_rate": 2e-05, |
| "loss": 0.5746, |
| "step": 1890 |
| }, |
| { |
| "epoch": 0.16697425081290096, |
| "grad_norm": 0.33752840757369995, |
| "learning_rate": 2e-05, |
| "loss": 0.5308, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.16785306265928465, |
| "grad_norm": 0.3529576361179352, |
| "learning_rate": 2e-05, |
| "loss": 0.5671, |
| "step": 1910 |
| }, |
| { |
| "epoch": 0.16873187450566834, |
| "grad_norm": 0.39925357699394226, |
| "learning_rate": 2e-05, |
| "loss": 0.5751, |
| "step": 1920 |
| }, |
| { |
| "epoch": 0.16961068635205204, |
| "grad_norm": 0.34715285897254944, |
| "learning_rate": 2e-05, |
| "loss": 0.5669, |
| "step": 1930 |
| }, |
| { |
| "epoch": 0.1704894981984357, |
| "grad_norm": 0.35600221157073975, |
| "learning_rate": 2e-05, |
| "loss": 0.5748, |
| "step": 1940 |
| }, |
| { |
| "epoch": 0.1713683100448194, |
| "grad_norm": 0.33379316329956055, |
| "learning_rate": 2e-05, |
| "loss": 0.5785, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.1722471218912031, |
| "grad_norm": 0.3700233995914459, |
| "learning_rate": 2e-05, |
| "loss": 0.5509, |
| "step": 1960 |
| }, |
| { |
| "epoch": 0.17312593373758678, |
| "grad_norm": 0.3257662355899811, |
| "learning_rate": 2e-05, |
| "loss": 0.5586, |
| "step": 1970 |
| }, |
| { |
| "epoch": 0.17400474558397047, |
| "grad_norm": 0.3797135055065155, |
| "learning_rate": 2e-05, |
| "loss": 0.5652, |
| "step": 1980 |
| }, |
| { |
| "epoch": 0.17488355743035416, |
| "grad_norm": 0.42362040281295776, |
| "learning_rate": 2e-05, |
| "loss": 0.5493, |
| "step": 1990 |
| }, |
| { |
| "epoch": 0.17576236927673786, |
| "grad_norm": 0.3780948519706726, |
| "learning_rate": 2e-05, |
| "loss": 0.5569, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.17664118112312155, |
| "grad_norm": 0.3641783893108368, |
| "learning_rate": 2e-05, |
| "loss": 0.6153, |
| "step": 2010 |
| }, |
| { |
| "epoch": 0.17751999296950524, |
| "grad_norm": 0.4090346097946167, |
| "learning_rate": 2e-05, |
| "loss": 0.5754, |
| "step": 2020 |
| }, |
| { |
| "epoch": 0.1783988048158889, |
| "grad_norm": 0.38258033990859985, |
| "learning_rate": 2e-05, |
| "loss": 0.5628, |
| "step": 2030 |
| }, |
| { |
| "epoch": 0.1792776166622726, |
| "grad_norm": 0.35445868968963623, |
| "learning_rate": 2e-05, |
| "loss": 0.5542, |
| "step": 2040 |
| }, |
| { |
| "epoch": 0.1801564285086563, |
| "grad_norm": 0.31840959191322327, |
| "learning_rate": 2e-05, |
| "loss": 0.5689, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.18103524035503998, |
| "grad_norm": 0.3384089767932892, |
| "learning_rate": 2e-05, |
| "loss": 0.5553, |
| "step": 2060 |
| }, |
| { |
| "epoch": 0.18191405220142368, |
| "grad_norm": 0.4153691828250885, |
| "learning_rate": 2e-05, |
| "loss": 0.5831, |
| "step": 2070 |
| }, |
| { |
| "epoch": 0.18279286404780737, |
| "grad_norm": 0.3473172187805176, |
| "learning_rate": 2e-05, |
| "loss": 0.5285, |
| "step": 2080 |
| }, |
| { |
| "epoch": 0.18367167589419106, |
| "grad_norm": 1.0200769901275635, |
| "learning_rate": 2e-05, |
| "loss": 0.5793, |
| "step": 2090 |
| }, |
| { |
| "epoch": 0.18455048774057475, |
| "grad_norm": 0.34419798851013184, |
| "learning_rate": 2e-05, |
| "loss": 0.5602, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.18542929958695845, |
| "grad_norm": 0.3456267714500427, |
| "learning_rate": 2e-05, |
| "loss": 0.5545, |
| "step": 2110 |
| }, |
| { |
| "epoch": 0.1863081114333421, |
| "grad_norm": 0.32470136880874634, |
| "learning_rate": 2e-05, |
| "loss": 0.5664, |
| "step": 2120 |
| }, |
| { |
| "epoch": 0.1871869232797258, |
| "grad_norm": 0.34200549125671387, |
| "learning_rate": 2e-05, |
| "loss": 0.549, |
| "step": 2130 |
| }, |
| { |
| "epoch": 0.1880657351261095, |
| "grad_norm": 0.3530149757862091, |
| "learning_rate": 2e-05, |
| "loss": 0.5596, |
| "step": 2140 |
| }, |
| { |
| "epoch": 0.1889445469724932, |
| "grad_norm": 0.4087691605091095, |
| "learning_rate": 2e-05, |
| "loss": 0.563, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.18982335881887688, |
| "grad_norm": 0.35635480284690857, |
| "learning_rate": 2e-05, |
| "loss": 0.5535, |
| "step": 2160 |
| }, |
| { |
| "epoch": 0.19070217066526057, |
| "grad_norm": 0.35621872544288635, |
| "learning_rate": 2e-05, |
| "loss": 0.5818, |
| "step": 2170 |
| }, |
| { |
| "epoch": 0.19158098251164427, |
| "grad_norm": 0.3401750922203064, |
| "learning_rate": 2e-05, |
| "loss": 0.5909, |
| "step": 2180 |
| }, |
| { |
| "epoch": 0.19245979435802796, |
| "grad_norm": 0.3132881820201874, |
| "learning_rate": 2e-05, |
| "loss": 0.5579, |
| "step": 2190 |
| }, |
| { |
| "epoch": 0.19333860620441162, |
| "grad_norm": 0.3682171702384949, |
| "learning_rate": 2e-05, |
| "loss": 0.5912, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.19421741805079532, |
| "grad_norm": 0.4018304646015167, |
| "learning_rate": 2e-05, |
| "loss": 0.5898, |
| "step": 2210 |
| }, |
| { |
| "epoch": 0.195096229897179, |
| "grad_norm": 0.36482760310173035, |
| "learning_rate": 2e-05, |
| "loss": 0.5752, |
| "step": 2220 |
| }, |
| { |
| "epoch": 0.1959750417435627, |
| "grad_norm": 0.29210537672042847, |
| "learning_rate": 2e-05, |
| "loss": 0.556, |
| "step": 2230 |
| }, |
| { |
| "epoch": 0.1968538535899464, |
| "grad_norm": 0.35293927788734436, |
| "learning_rate": 2e-05, |
| "loss": 0.5914, |
| "step": 2240 |
| }, |
| { |
| "epoch": 0.19773266543633009, |
| "grad_norm": 0.34668824076652527, |
| "learning_rate": 2e-05, |
| "loss": 0.577, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.19861147728271378, |
| "grad_norm": 0.360667884349823, |
| "learning_rate": 2e-05, |
| "loss": 0.5497, |
| "step": 2260 |
| }, |
| { |
| "epoch": 0.19949028912909747, |
| "grad_norm": 0.3482297956943512, |
| "learning_rate": 2e-05, |
| "loss": 0.5465, |
| "step": 2270 |
| }, |
| { |
| "epoch": 0.20036910097548116, |
| "grad_norm": 0.339021772146225, |
| "learning_rate": 2e-05, |
| "loss": 0.5473, |
| "step": 2280 |
| }, |
| { |
| "epoch": 0.20124791282186483, |
| "grad_norm": 0.36110997200012207, |
| "learning_rate": 2e-05, |
| "loss": 0.5538, |
| "step": 2290 |
| }, |
| { |
| "epoch": 0.20212672466824852, |
| "grad_norm": 0.35237032175064087, |
| "learning_rate": 2e-05, |
| "loss": 0.5548, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.2030055365146322, |
| "grad_norm": 0.4096246063709259, |
| "learning_rate": 2e-05, |
| "loss": 0.553, |
| "step": 2310 |
| }, |
| { |
| "epoch": 0.2038843483610159, |
| "grad_norm": 0.41096365451812744, |
| "learning_rate": 2e-05, |
| "loss": 0.5379, |
| "step": 2320 |
| }, |
| { |
| "epoch": 0.2047631602073996, |
| "grad_norm": 0.3804115355014801, |
| "learning_rate": 2e-05, |
| "loss": 0.5508, |
| "step": 2330 |
| }, |
| { |
| "epoch": 0.2056419720537833, |
| "grad_norm": 0.3373338580131531, |
| "learning_rate": 2e-05, |
| "loss": 0.5659, |
| "step": 2340 |
| }, |
| { |
| "epoch": 0.20652078390016698, |
| "grad_norm": 0.37081179022789, |
| "learning_rate": 2e-05, |
| "loss": 0.5592, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.20739959574655067, |
| "grad_norm": 0.34369438886642456, |
| "learning_rate": 2e-05, |
| "loss": 0.6028, |
| "step": 2360 |
| }, |
| { |
| "epoch": 0.20827840759293434, |
| "grad_norm": 0.3445272147655487, |
| "learning_rate": 2e-05, |
| "loss": 0.536, |
| "step": 2370 |
| }, |
| { |
| "epoch": 0.20915721943931803, |
| "grad_norm": 0.36156561970710754, |
| "learning_rate": 2e-05, |
| "loss": 0.5625, |
| "step": 2380 |
| }, |
| { |
| "epoch": 0.21003603128570172, |
| "grad_norm": 0.3855374753475189, |
| "learning_rate": 2e-05, |
| "loss": 0.5516, |
| "step": 2390 |
| }, |
| { |
| "epoch": 0.21091484313208542, |
| "grad_norm": 0.36775392293930054, |
| "learning_rate": 2e-05, |
| "loss": 0.5893, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.2117936549784691, |
| "grad_norm": 0.3330851197242737, |
| "learning_rate": 2e-05, |
| "loss": 0.5767, |
| "step": 2410 |
| }, |
| { |
| "epoch": 0.2126724668248528, |
| "grad_norm": 0.34514012932777405, |
| "learning_rate": 2e-05, |
| "loss": 0.5727, |
| "step": 2420 |
| }, |
| { |
| "epoch": 0.2135512786712365, |
| "grad_norm": 0.47892290353775024, |
| "learning_rate": 2e-05, |
| "loss": 0.5875, |
| "step": 2430 |
| }, |
| { |
| "epoch": 0.2144300905176202, |
| "grad_norm": 0.36140429973602295, |
| "learning_rate": 2e-05, |
| "loss": 0.5522, |
| "step": 2440 |
| }, |
| { |
| "epoch": 0.21530890236400388, |
| "grad_norm": 0.33918988704681396, |
| "learning_rate": 2e-05, |
| "loss": 0.5562, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.21618771421038754, |
| "grad_norm": 0.3281526565551758, |
| "learning_rate": 2e-05, |
| "loss": 0.5565, |
| "step": 2460 |
| }, |
| { |
| "epoch": 0.21706652605677124, |
| "grad_norm": 0.3529514968395233, |
| "learning_rate": 2e-05, |
| "loss": 0.5498, |
| "step": 2470 |
| }, |
| { |
| "epoch": 0.21794533790315493, |
| "grad_norm": 0.339227557182312, |
| "learning_rate": 2e-05, |
| "loss": 0.5498, |
| "step": 2480 |
| }, |
| { |
| "epoch": 0.21882414974953862, |
| "grad_norm": 0.3628411293029785, |
| "learning_rate": 2e-05, |
| "loss": 0.557, |
| "step": 2490 |
| }, |
| { |
| "epoch": 0.2197029615959223, |
| "grad_norm": 0.34096869826316833, |
| "learning_rate": 2e-05, |
| "loss": 0.5719, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.220581773442306, |
| "grad_norm": 0.32310160994529724, |
| "learning_rate": 2e-05, |
| "loss": 0.5729, |
| "step": 2510 |
| }, |
| { |
| "epoch": 0.2214605852886897, |
| "grad_norm": 0.3401024043560028, |
| "learning_rate": 2e-05, |
| "loss": 0.555, |
| "step": 2520 |
| }, |
| { |
| "epoch": 0.2223393971350734, |
| "grad_norm": 0.46229088306427, |
| "learning_rate": 2e-05, |
| "loss": 0.5514, |
| "step": 2530 |
| }, |
| { |
| "epoch": 0.22321820898145708, |
| "grad_norm": 0.3563999831676483, |
| "learning_rate": 2e-05, |
| "loss": 0.5748, |
| "step": 2540 |
| }, |
| { |
| "epoch": 0.22409702082784075, |
| "grad_norm": 0.31088465452194214, |
| "learning_rate": 1.997261400182901e-05, |
| "loss": 0.5634, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.22497583267422444, |
| "grad_norm": 0.37515947222709656, |
| "learning_rate": 1.9805800746810423e-05, |
| "loss": 0.5409, |
| "step": 2560 |
| }, |
| { |
| "epoch": 0.22585464452060813, |
| "grad_norm": 0.3521416187286377, |
| "learning_rate": 1.9489934220110274e-05, |
| "loss": 0.5712, |
| "step": 2570 |
| }, |
| { |
| "epoch": 0.22673345636699183, |
| "grad_norm": 0.8562304377555847, |
| "learning_rate": 1.9029840364820392e-05, |
| "loss": 0.5728, |
| "step": 2580 |
| }, |
| { |
| "epoch": 0.22761226821337552, |
| "grad_norm": 0.32249194383621216, |
| "learning_rate": 1.8432548690479908e-05, |
| "loss": 0.5702, |
| "step": 2590 |
| }, |
| { |
| "epoch": 0.2284910800597592, |
| "grad_norm": 0.36118215322494507, |
| "learning_rate": 1.7707184873239772e-05, |
| "loss": 0.5451, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.2293698919061429, |
| "grad_norm": 0.3505563735961914, |
| "learning_rate": 1.6864831329902565e-05, |
| "loss": 0.5427, |
| "step": 2610 |
| }, |
| { |
| "epoch": 0.2302487037525266, |
| "grad_norm": 0.3535621464252472, |
| "learning_rate": 1.591835789604667e-05, |
| "loss": 0.5502, |
| "step": 2620 |
| }, |
| { |
| "epoch": 0.23112751559891026, |
| "grad_norm": 0.3735027015209198, |
| "learning_rate": 1.4882225195206459e-05, |
| "loss": 0.5559, |
| "step": 2630 |
| }, |
| { |
| "epoch": 0.23200632744529395, |
| "grad_norm": 0.32464268803596497, |
| "learning_rate": 1.3772263703317695e-05, |
| "loss": 0.5449, |
| "step": 2640 |
| }, |
| { |
| "epoch": 0.23288513929167765, |
| "grad_norm": 0.3927508592605591, |
| "learning_rate": 1.2605431883975547e-05, |
| "loss": 0.5441, |
| "step": 2650 |
| }, |
| { |
| "epoch": 0.23376395113806134, |
| "grad_norm": 0.3734285235404968, |
| "learning_rate": 1.1399557089817574e-05, |
| "loss": 0.5555, |
| "step": 2660 |
| }, |
| { |
| "epoch": 0.23464276298444503, |
| "grad_norm": 0.3142608404159546, |
| "learning_rate": 1.0173063188650463e-05, |
| "loss": 0.5619, |
| "step": 2670 |
| }, |
| { |
| "epoch": 0.23552157483082872, |
| "grad_norm": 0.3565467894077301, |
| "learning_rate": 8.944689075764396e-06, |
| "loss": 0.5592, |
| "step": 2680 |
| }, |
| { |
| "epoch": 0.23640038667721242, |
| "grad_norm": 0.35128548741340637, |
| "learning_rate": 7.73320237312362e-06, |
| "loss": 0.5435, |
| "step": 2690 |
| }, |
| { |
| "epoch": 0.2372791985235961, |
| "grad_norm": 0.36121687293052673, |
| "learning_rate": 6.557112689658629e-06, |
| "loss": 0.5634, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.2381580103699798, |
| "grad_norm": 0.356099009513855, |
| "learning_rate": 5.434388823591213e-06, |
| "loss": 0.5536, |
| "step": 2710 |
| }, |
| { |
| "epoch": 0.23903682221636346, |
| "grad_norm": 0.6229901313781738, |
| "learning_rate": 4.382184227495424e-06, |
| "loss": 0.5636, |
| "step": 2720 |
| }, |
| { |
| "epoch": 0.23991563406274716, |
| "grad_norm": 0.3396807312965393, |
| "learning_rate": 3.4165749305560203e-06, |
| "loss": 0.5322, |
| "step": 2730 |
| }, |
| { |
| "epoch": 0.24079444590913085, |
| "grad_norm": 0.34962546825408936, |
| "learning_rate": 2.5523139221595556e-06, |
| "loss": 0.5607, |
| "step": 2740 |
| }, |
| { |
| "epoch": 0.24167325775551454, |
| "grad_norm": 0.35136163234710693, |
| "learning_rate": 1.8026057494499821e-06, |
| "loss": 0.555, |
| "step": 2750 |
| }, |
| { |
| "epoch": 0.24255206960189823, |
| "grad_norm": 0.3626517653465271, |
| "learning_rate": 1.178904772643223e-06, |
| "loss": 0.5367, |
| "step": 2760 |
| }, |
| { |
| "epoch": 0.24343088144828193, |
| "grad_norm": 0.32543617486953735, |
| "learning_rate": 6.90740160441754e-07, |
| "loss": 0.5386, |
| "step": 2770 |
| }, |
| { |
| "epoch": 0.24430969329466562, |
| "grad_norm": 0.4628903269767761, |
| "learning_rate": 3.45570299343632e-07, |
| "loss": 0.55, |
| "step": 2780 |
| }, |
| { |
| "epoch": 0.2451885051410493, |
| "grad_norm": 0.6712743639945984, |
| "learning_rate": 1.486688412425152e-07, |
| "loss": 0.5659, |
| "step": 2790 |
| }, |
| { |
| "epoch": 0.2458036734335179, |
| "step": 2797, |
| "total_flos": 1.2485883762455871e+19, |
| "train_loss": 0.05006915874297081, |
| "train_runtime": 2879.3078, |
| "train_samples_per_second": 11.657, |
| "train_steps_per_second": 0.971 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 2797, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "save_steps": 2543, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1.2485883762455871e+19, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
