diff --git "a/multi_student_4/trainer_state.json" "b/multi_student_4/trainer_state.json" new file mode 100644--- /dev/null +++ "b/multi_student_4/trainer_state.json" @@ -0,0 +1,7491 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 7.0, + "eval_steps": 500, + "global_step": 10640, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.006578947368421052, + "grad_norm": 19.201007081270827, + "learning_rate": 8.458646616541353e-07, + "loss": 8.2522, + "step": 10 + }, + { + "epoch": 0.013157894736842105, + "grad_norm": 18.114759300569585, + "learning_rate": 1.7857142857142857e-06, + "loss": 8.1346, + "step": 20 + }, + { + "epoch": 0.019736842105263157, + "grad_norm": 9.479492840003324, + "learning_rate": 2.725563909774436e-06, + "loss": 7.5119, + "step": 30 + }, + { + "epoch": 0.02631578947368421, + "grad_norm": 4.976331973381574, + "learning_rate": 3.6654135338345866e-06, + "loss": 6.7788, + "step": 40 + }, + { + "epoch": 0.03289473684210526, + "grad_norm": 3.2804593438080754, + "learning_rate": 4.605263157894737e-06, + "loss": 6.2892, + "step": 50 + }, + { + "epoch": 0.039473684210526314, + "grad_norm": 2.8583967169035027, + "learning_rate": 5.5451127819548875e-06, + "loss": 5.9528, + "step": 60 + }, + { + "epoch": 0.046052631578947366, + "grad_norm": 3.5512069733934597, + "learning_rate": 6.484962406015038e-06, + "loss": 5.5755, + "step": 70 + }, + { + "epoch": 0.05263157894736842, + "grad_norm": 3.731444843289522, + "learning_rate": 7.424812030075189e-06, + "loss": 4.9914, + "step": 80 + }, + { + "epoch": 0.05921052631578947, + "grad_norm": 5.60127901105382, + "learning_rate": 8.364661654135339e-06, + "loss": 4.3233, + "step": 90 + }, + { + "epoch": 0.06578947368421052, + "grad_norm": 3.7839180004348996, + "learning_rate": 9.30451127819549e-06, + "loss": 3.7305, + "step": 100 + }, + { + "epoch": 0.07236842105263158, + "grad_norm": 5.421177691842472, + "learning_rate": 1.024436090225564e-05, + "loss": 3.3248, + "step": 110 + }, + { + "epoch": 0.07894736842105263, + "grad_norm": 3.7158079343320694, + "learning_rate": 1.118421052631579e-05, + "loss": 3.092, + "step": 120 + }, + { + "epoch": 0.08552631578947369, + "grad_norm": 5.042680618782553, + "learning_rate": 1.212406015037594e-05, + "loss": 2.9383, + "step": 130 + }, + { + "epoch": 0.09210526315789473, + "grad_norm": 5.8029975922287615, + "learning_rate": 1.3063909774436093e-05, + "loss": 2.8108, + "step": 140 + }, + { + "epoch": 0.09868421052631579, + "grad_norm": 7.443887556478386, + "learning_rate": 1.4003759398496241e-05, + "loss": 2.7256, + "step": 150 + }, + { + "epoch": 0.10526315789473684, + "grad_norm": 6.942055828457074, + "learning_rate": 1.4943609022556393e-05, + "loss": 2.6465, + "step": 160 + }, + { + "epoch": 0.1118421052631579, + "grad_norm": 20.211444588304172, + "learning_rate": 1.5883458646616542e-05, + "loss": 2.5756, + "step": 170 + }, + { + "epoch": 0.11842105263157894, + "grad_norm": 23.482614308148037, + "learning_rate": 1.6823308270676694e-05, + "loss": 2.5262, + "step": 180 + }, + { + "epoch": 0.125, + "grad_norm": 13.648396628618737, + "learning_rate": 1.7763157894736842e-05, + "loss": 2.4699, + "step": 190 + }, + { + "epoch": 0.13157894736842105, + "grad_norm": 15.205800265082752, + "learning_rate": 1.8703007518796993e-05, + "loss": 2.4115, + "step": 200 + }, + { + "epoch": 0.13815789473684212, + "grad_norm": 16.093588917694532, + "learning_rate": 1.9642857142857145e-05, + "loss": 2.3548, + "step": 210 + }, + { + "epoch": 0.14473684210526316, + "grad_norm": 19.61654573509106, + "learning_rate": 2.0582706766917297e-05, + "loss": 2.3042, + "step": 220 + }, + { + "epoch": 0.1513157894736842, + "grad_norm": 12.268305710278183, + "learning_rate": 2.1522556390977445e-05, + "loss": 2.2494, + "step": 230 + }, + { + "epoch": 0.15789473684210525, + "grad_norm": 12.399788763360808, + "learning_rate": 2.2462406015037596e-05, + "loss": 2.1853, + "step": 240 + }, + { + "epoch": 0.16447368421052633, + "grad_norm": 7.463420275410263, + "learning_rate": 2.3402255639097744e-05, + "loss": 2.0998, + "step": 250 + }, + { + "epoch": 0.17105263157894737, + "grad_norm": 12.087535187997679, + "learning_rate": 2.4342105263157896e-05, + "loss": 2.0075, + "step": 260 + }, + { + "epoch": 0.17763157894736842, + "grad_norm": 6.62910919540051, + "learning_rate": 2.5281954887218044e-05, + "loss": 1.8974, + "step": 270 + }, + { + "epoch": 0.18421052631578946, + "grad_norm": 8.989470792407845, + "learning_rate": 2.62218045112782e-05, + "loss": 1.7928, + "step": 280 + }, + { + "epoch": 0.19078947368421054, + "grad_norm": 6.928524369383665, + "learning_rate": 2.716165413533835e-05, + "loss": 1.7057, + "step": 290 + }, + { + "epoch": 0.19736842105263158, + "grad_norm": 10.311746680652767, + "learning_rate": 2.8101503759398495e-05, + "loss": 1.6251, + "step": 300 + }, + { + "epoch": 0.20394736842105263, + "grad_norm": 7.107313244759928, + "learning_rate": 2.9041353383458647e-05, + "loss": 1.5522, + "step": 310 + }, + { + "epoch": 0.21052631578947367, + "grad_norm": 5.8857302431981156, + "learning_rate": 2.9981203007518798e-05, + "loss": 1.4974, + "step": 320 + }, + { + "epoch": 0.21710526315789475, + "grad_norm": 6.775917238582609, + "learning_rate": 3.092105263157895e-05, + "loss": 1.4355, + "step": 330 + }, + { + "epoch": 0.2236842105263158, + "grad_norm": 9.938119729849578, + "learning_rate": 3.1860902255639094e-05, + "loss": 1.3895, + "step": 340 + }, + { + "epoch": 0.23026315789473684, + "grad_norm": 4.870796794495703, + "learning_rate": 3.2800751879699246e-05, + "loss": 1.3432, + "step": 350 + }, + { + "epoch": 0.23684210526315788, + "grad_norm": 5.353011232164181, + "learning_rate": 3.37406015037594e-05, + "loss": 1.3002, + "step": 360 + }, + { + "epoch": 0.24342105263157895, + "grad_norm": 7.078794515741611, + "learning_rate": 3.468045112781955e-05, + "loss": 1.2641, + "step": 370 + }, + { + "epoch": 0.25, + "grad_norm": 5.419600483372193, + "learning_rate": 3.56203007518797e-05, + "loss": 1.209, + "step": 380 + }, + { + "epoch": 0.2565789473684211, + "grad_norm": 3.5327527407457926, + "learning_rate": 3.656015037593985e-05, + "loss": 1.1644, + "step": 390 + }, + { + "epoch": 0.2631578947368421, + "grad_norm": 5.894523313050658, + "learning_rate": 3.7500000000000003e-05, + "loss": 1.1319, + "step": 400 + }, + { + "epoch": 0.26973684210526316, + "grad_norm": 3.9863193805342316, + "learning_rate": 3.8439849624060155e-05, + "loss": 1.094, + "step": 410 + }, + { + "epoch": 0.27631578947368424, + "grad_norm": 4.225678742387179, + "learning_rate": 3.9379699248120307e-05, + "loss": 1.0651, + "step": 420 + }, + { + "epoch": 0.28289473684210525, + "grad_norm": 5.586389531542015, + "learning_rate": 4.031954887218045e-05, + "loss": 1.0247, + "step": 430 + }, + { + "epoch": 0.2894736842105263, + "grad_norm": 7.0928476432315675, + "learning_rate": 4.12593984962406e-05, + "loss": 0.9973, + "step": 440 + }, + { + "epoch": 0.29605263157894735, + "grad_norm": 5.797737179453841, + "learning_rate": 4.2199248120300754e-05, + "loss": 0.9738, + "step": 450 + }, + { + "epoch": 0.3026315789473684, + "grad_norm": 5.436160406452568, + "learning_rate": 4.3139097744360906e-05, + "loss": 0.9539, + "step": 460 + }, + { + "epoch": 0.3092105263157895, + "grad_norm": 4.940473111611711, + "learning_rate": 4.407894736842105e-05, + "loss": 0.9197, + "step": 470 + }, + { + "epoch": 0.3157894736842105, + "grad_norm": 5.844247327503255, + "learning_rate": 4.50187969924812e-05, + "loss": 0.8862, + "step": 480 + }, + { + "epoch": 0.3223684210526316, + "grad_norm": 4.744124381908569, + "learning_rate": 4.5958646616541354e-05, + "loss": 0.8668, + "step": 490 + }, + { + "epoch": 0.32894736842105265, + "grad_norm": 4.689028836800058, + "learning_rate": 4.6898496240601505e-05, + "loss": 0.8476, + "step": 500 + }, + { + "epoch": 0.3355263157894737, + "grad_norm": 4.398308995209079, + "learning_rate": 4.783834586466166e-05, + "loss": 0.82, + "step": 510 + }, + { + "epoch": 0.34210526315789475, + "grad_norm": 5.554811082819198, + "learning_rate": 4.877819548872181e-05, + "loss": 0.7974, + "step": 520 + }, + { + "epoch": 0.34868421052631576, + "grad_norm": 3.7243976034524584, + "learning_rate": 4.971804511278196e-05, + "loss": 0.7715, + "step": 530 + }, + { + "epoch": 0.35526315789473684, + "grad_norm": 4.879577435611285, + "learning_rate": 5.0657894736842104e-05, + "loss": 0.7514, + "step": 540 + }, + { + "epoch": 0.3618421052631579, + "grad_norm": 5.516075231521318, + "learning_rate": 5.1597744360902256e-05, + "loss": 0.7291, + "step": 550 + }, + { + "epoch": 0.3684210526315789, + "grad_norm": 5.494655015332771, + "learning_rate": 5.253759398496241e-05, + "loss": 0.6974, + "step": 560 + }, + { + "epoch": 0.375, + "grad_norm": 5.0610034411984826, + "learning_rate": 5.347744360902256e-05, + "loss": 0.6779, + "step": 570 + }, + { + "epoch": 0.3815789473684211, + "grad_norm": 6.241132985448258, + "learning_rate": 5.441729323308271e-05, + "loss": 0.6723, + "step": 580 + }, + { + "epoch": 0.3881578947368421, + "grad_norm": 4.582816961569543, + "learning_rate": 5.535714285714286e-05, + "loss": 0.652, + "step": 590 + }, + { + "epoch": 0.39473684210526316, + "grad_norm": 3.5845708789821042, + "learning_rate": 5.6296992481203013e-05, + "loss": 0.6154, + "step": 600 + }, + { + "epoch": 0.40131578947368424, + "grad_norm": 4.328768044098044, + "learning_rate": 5.723684210526316e-05, + "loss": 0.6042, + "step": 610 + }, + { + "epoch": 0.40789473684210525, + "grad_norm": 3.5473602851751793, + "learning_rate": 5.817669172932331e-05, + "loss": 0.5803, + "step": 620 + }, + { + "epoch": 0.4144736842105263, + "grad_norm": 4.233181045605189, + "learning_rate": 5.911654135338346e-05, + "loss": 0.5688, + "step": 630 + }, + { + "epoch": 0.42105263157894735, + "grad_norm": 4.0831509210740835, + "learning_rate": 6.005639097744361e-05, + "loss": 0.5385, + "step": 640 + }, + { + "epoch": 0.4276315789473684, + "grad_norm": 3.8705204303765814, + "learning_rate": 6.0996240601503764e-05, + "loss": 0.5201, + "step": 650 + }, + { + "epoch": 0.4342105263157895, + "grad_norm": 3.8269827275761337, + "learning_rate": 6.193609022556392e-05, + "loss": 0.504, + "step": 660 + }, + { + "epoch": 0.4407894736842105, + "grad_norm": 3.7042333223437476, + "learning_rate": 6.287593984962407e-05, + "loss": 0.485, + "step": 670 + }, + { + "epoch": 0.4473684210526316, + "grad_norm": 3.1454584376141725, + "learning_rate": 6.381578947368422e-05, + "loss": 0.4692, + "step": 680 + }, + { + "epoch": 0.45394736842105265, + "grad_norm": 3.6715281367823698, + "learning_rate": 6.475563909774437e-05, + "loss": 0.4474, + "step": 690 + }, + { + "epoch": 0.4605263157894737, + "grad_norm": 3.3520165299765203, + "learning_rate": 6.569548872180451e-05, + "loss": 0.4335, + "step": 700 + }, + { + "epoch": 0.46710526315789475, + "grad_norm": 3.4830856151143554, + "learning_rate": 6.663533834586466e-05, + "loss": 0.4156, + "step": 710 + }, + { + "epoch": 0.47368421052631576, + "grad_norm": 3.4130338264227835, + "learning_rate": 6.757518796992481e-05, + "loss": 0.3984, + "step": 720 + }, + { + "epoch": 0.48026315789473684, + "grad_norm": 3.6699720598135834, + "learning_rate": 6.851503759398496e-05, + "loss": 0.3884, + "step": 730 + }, + { + "epoch": 0.4868421052631579, + "grad_norm": 3.2598730307422192, + "learning_rate": 6.945488721804511e-05, + "loss": 0.3727, + "step": 740 + }, + { + "epoch": 0.4934210526315789, + "grad_norm": 3.0065146921519097, + "learning_rate": 7.039473684210527e-05, + "loss": 0.3521, + "step": 750 + }, + { + "epoch": 0.5, + "grad_norm": 2.3801132893266956, + "learning_rate": 7.133458646616542e-05, + "loss": 0.3415, + "step": 760 + }, + { + "epoch": 0.506578947368421, + "grad_norm": 2.5040976105308466, + "learning_rate": 7.227443609022557e-05, + "loss": 0.3303, + "step": 770 + }, + { + "epoch": 0.5131578947368421, + "grad_norm": 2.3095873713136004, + "learning_rate": 7.321428571428571e-05, + "loss": 0.3155, + "step": 780 + }, + { + "epoch": 0.5197368421052632, + "grad_norm": 2.221909098699267, + "learning_rate": 7.415413533834586e-05, + "loss": 0.3, + "step": 790 + }, + { + "epoch": 0.5263157894736842, + "grad_norm": 2.3650008130107008, + "learning_rate": 7.509398496240601e-05, + "loss": 0.2881, + "step": 800 + }, + { + "epoch": 0.5328947368421053, + "grad_norm": 2.92228181953594, + "learning_rate": 7.603383458646618e-05, + "loss": 0.2759, + "step": 810 + }, + { + "epoch": 0.5394736842105263, + "grad_norm": 2.39576316770683, + "learning_rate": 7.697368421052633e-05, + "loss": 0.2611, + "step": 820 + }, + { + "epoch": 0.5460526315789473, + "grad_norm": 1.9345517046210943, + "learning_rate": 7.791353383458648e-05, + "loss": 0.2472, + "step": 830 + }, + { + "epoch": 0.5526315789473685, + "grad_norm": 2.321485582474392, + "learning_rate": 7.885338345864663e-05, + "loss": 0.2409, + "step": 840 + }, + { + "epoch": 0.5592105263157895, + "grad_norm": 1.8690657258212422, + "learning_rate": 7.979323308270678e-05, + "loss": 0.2308, + "step": 850 + }, + { + "epoch": 0.5657894736842105, + "grad_norm": 1.901049780468097, + "learning_rate": 8.073308270676692e-05, + "loss": 0.2171, + "step": 860 + }, + { + "epoch": 0.5723684210526315, + "grad_norm": 2.2229311039801596, + "learning_rate": 8.167293233082707e-05, + "loss": 0.2072, + "step": 870 + }, + { + "epoch": 0.5789473684210527, + "grad_norm": 2.213221418124993, + "learning_rate": 8.261278195488722e-05, + "loss": 0.1965, + "step": 880 + }, + { + "epoch": 0.5855263157894737, + "grad_norm": 1.870540842703205, + "learning_rate": 8.355263157894737e-05, + "loss": 0.1877, + "step": 890 + }, + { + "epoch": 0.5921052631578947, + "grad_norm": 1.311680351399902, + "learning_rate": 8.449248120300753e-05, + "loss": 0.1792, + "step": 900 + }, + { + "epoch": 0.5986842105263158, + "grad_norm": 1.5798663811679752, + "learning_rate": 8.543233082706768e-05, + "loss": 0.1716, + "step": 910 + }, + { + "epoch": 0.6052631578947368, + "grad_norm": 7.169134004434136, + "learning_rate": 8.637218045112783e-05, + "loss": 0.1721, + "step": 920 + }, + { + "epoch": 0.6118421052631579, + "grad_norm": 1.5044155665130965, + "learning_rate": 8.731203007518798e-05, + "loss": 0.193, + "step": 930 + }, + { + "epoch": 0.618421052631579, + "grad_norm": 1.305309283144098, + "learning_rate": 8.825187969924813e-05, + "loss": 0.1566, + "step": 940 + }, + { + "epoch": 0.625, + "grad_norm": 1.23686939558907, + "learning_rate": 8.919172932330827e-05, + "loss": 0.1445, + "step": 950 + }, + { + "epoch": 0.631578947368421, + "grad_norm": 1.0445490067006857, + "learning_rate": 9.013157894736842e-05, + "loss": 0.1356, + "step": 960 + }, + { + "epoch": 0.6381578947368421, + "grad_norm": 1.3136634908844078, + "learning_rate": 9.107142857142857e-05, + "loss": 0.1308, + "step": 970 + }, + { + "epoch": 0.6447368421052632, + "grad_norm": 1.7716809252302665, + "learning_rate": 9.201127819548872e-05, + "loss": 0.124, + "step": 980 + }, + { + "epoch": 0.6513157894736842, + "grad_norm": 1.0772175167269311, + "learning_rate": 9.295112781954888e-05, + "loss": 0.1202, + "step": 990 + }, + { + "epoch": 0.6578947368421053, + "grad_norm": 1.1878390185089365, + "learning_rate": 9.389097744360903e-05, + "loss": 0.114, + "step": 1000 + }, + { + "epoch": 0.6644736842105263, + "grad_norm": 0.8915150933559234, + "learning_rate": 9.483082706766918e-05, + "loss": 0.1054, + "step": 1010 + }, + { + "epoch": 0.6710526315789473, + "grad_norm": 0.7818448157917527, + "learning_rate": 9.577067669172933e-05, + "loss": 0.0996, + "step": 1020 + }, + { + "epoch": 0.6776315789473685, + "grad_norm": 0.990430969229284, + "learning_rate": 9.671052631578947e-05, + "loss": 0.0974, + "step": 1030 + }, + { + "epoch": 0.6842105263157895, + "grad_norm": 0.9276518666370865, + "learning_rate": 9.765037593984962e-05, + "loss": 0.0933, + "step": 1040 + }, + { + "epoch": 0.6907894736842105, + "grad_norm": 0.6846215892309722, + "learning_rate": 9.859022556390977e-05, + "loss": 0.0874, + "step": 1050 + }, + { + "epoch": 0.6973684210526315, + "grad_norm": 0.714553023496577, + "learning_rate": 9.953007518796992e-05, + "loss": 0.0814, + "step": 1060 + }, + { + "epoch": 0.7039473684210527, + "grad_norm": 0.7546346945965402, + "learning_rate": 9.999993273155467e-05, + "loss": 0.0774, + "step": 1070 + }, + { + "epoch": 0.7105263157894737, + "grad_norm": 0.571192300635228, + "learning_rate": 9.999939458507816e-05, + "loss": 0.0762, + "step": 1080 + }, + { + "epoch": 0.7171052631578947, + "grad_norm": 0.8584193322720889, + "learning_rate": 9.999831829791713e-05, + "loss": 0.0756, + "step": 1090 + }, + { + "epoch": 0.7236842105263158, + "grad_norm": 1.3952479453359226, + "learning_rate": 9.999670388165563e-05, + "loss": 0.075, + "step": 1100 + }, + { + "epoch": 0.7302631578947368, + "grad_norm": 0.751465926189691, + "learning_rate": 9.999455135366951e-05, + "loss": 0.069, + "step": 1110 + }, + { + "epoch": 0.7368421052631579, + "grad_norm": 0.4873110858150168, + "learning_rate": 9.999186073712635e-05, + "loss": 0.064, + "step": 1120 + }, + { + "epoch": 0.743421052631579, + "grad_norm": 0.4495642582063532, + "learning_rate": 9.998863206098507e-05, + "loss": 0.0623, + "step": 1130 + }, + { + "epoch": 0.75, + "grad_norm": 0.45684969056385805, + "learning_rate": 9.998486535999573e-05, + "loss": 0.059, + "step": 1140 + }, + { + "epoch": 0.756578947368421, + "grad_norm": 0.4304228051311271, + "learning_rate": 9.998056067469915e-05, + "loss": 0.0574, + "step": 1150 + }, + { + "epoch": 0.7631578947368421, + "grad_norm": 0.38692754350368835, + "learning_rate": 9.997571805142639e-05, + "loss": 0.0568, + "step": 1160 + }, + { + "epoch": 0.7697368421052632, + "grad_norm": 0.44697102978905795, + "learning_rate": 9.997033754229835e-05, + "loss": 0.054, + "step": 1170 + }, + { + "epoch": 0.7763157894736842, + "grad_norm": 0.3475319362475707, + "learning_rate": 9.996441920522512e-05, + "loss": 0.0512, + "step": 1180 + }, + { + "epoch": 0.7828947368421053, + "grad_norm": 0.3776964171490181, + "learning_rate": 9.995796310390547e-05, + "loss": 0.0496, + "step": 1190 + }, + { + "epoch": 0.7894736842105263, + "grad_norm": 0.3599383764053137, + "learning_rate": 9.995096930782604e-05, + "loss": 0.049, + "step": 1200 + }, + { + "epoch": 0.7960526315789473, + "grad_norm": 0.4004701605242826, + "learning_rate": 9.994343789226066e-05, + "loss": 0.0468, + "step": 1210 + }, + { + "epoch": 0.8026315789473685, + "grad_norm": 0.3822329385724864, + "learning_rate": 9.993536893826954e-05, + "loss": 0.0457, + "step": 1220 + }, + { + "epoch": 0.8092105263157895, + "grad_norm": 0.3620248875103048, + "learning_rate": 9.992676253269837e-05, + "loss": 0.0446, + "step": 1230 + }, + { + "epoch": 0.8157894736842105, + "grad_norm": 0.28358957764617276, + "learning_rate": 9.991761876817743e-05, + "loss": 0.0429, + "step": 1240 + }, + { + "epoch": 0.8223684210526315, + "grad_norm": 0.32929903528232773, + "learning_rate": 9.990793774312053e-05, + "loss": 0.0421, + "step": 1250 + }, + { + "epoch": 0.8289473684210527, + "grad_norm": 0.34141322063560897, + "learning_rate": 9.9897719561724e-05, + "loss": 0.0418, + "step": 1260 + }, + { + "epoch": 0.8355263157894737, + "grad_norm": 0.28296527941759625, + "learning_rate": 9.988696433396557e-05, + "loss": 0.0403, + "step": 1270 + }, + { + "epoch": 0.8421052631578947, + "grad_norm": 0.3686897659419288, + "learning_rate": 9.987567217560314e-05, + "loss": 0.0395, + "step": 1280 + }, + { + "epoch": 0.8486842105263158, + "grad_norm": 0.3066999773216274, + "learning_rate": 9.986384320817359e-05, + "loss": 0.0395, + "step": 1290 + }, + { + "epoch": 0.8552631578947368, + "grad_norm": 0.33225917789321974, + "learning_rate": 9.985147755899142e-05, + "loss": 0.0385, + "step": 1300 + }, + { + "epoch": 0.8618421052631579, + "grad_norm": 0.3004075266809671, + "learning_rate": 9.983857536114743e-05, + "loss": 0.0377, + "step": 1310 + }, + { + "epoch": 0.868421052631579, + "grad_norm": 0.2698302147467991, + "learning_rate": 9.982513675350724e-05, + "loss": 0.0367, + "step": 1320 + }, + { + "epoch": 0.875, + "grad_norm": 0.26807280624783797, + "learning_rate": 9.981116188070984e-05, + "loss": 0.0356, + "step": 1330 + }, + { + "epoch": 0.881578947368421, + "grad_norm": 0.3127969801843736, + "learning_rate": 9.979665089316602e-05, + "loss": 0.035, + "step": 1340 + }, + { + "epoch": 0.8881578947368421, + "grad_norm": 0.28653732054861003, + "learning_rate": 9.978160394705668e-05, + "loss": 0.0343, + "step": 1350 + }, + { + "epoch": 0.8947368421052632, + "grad_norm": 0.2504478087548026, + "learning_rate": 9.976602120433133e-05, + "loss": 0.0341, + "step": 1360 + }, + { + "epoch": 0.9013157894736842, + "grad_norm": 0.24532062538161709, + "learning_rate": 9.97499028327061e-05, + "loss": 0.0331, + "step": 1370 + }, + { + "epoch": 0.9078947368421053, + "grad_norm": 0.5595503828448865, + "learning_rate": 9.973324900566213e-05, + "loss": 0.041, + "step": 1380 + }, + { + "epoch": 0.9144736842105263, + "grad_norm": 0.32103277633584737, + "learning_rate": 9.971605990244365e-05, + "loss": 0.0396, + "step": 1390 + }, + { + "epoch": 0.9210526315789473, + "grad_norm": 0.39243105455264066, + "learning_rate": 9.969833570805599e-05, + "loss": 0.0345, + "step": 1400 + }, + { + "epoch": 0.9276315789473685, + "grad_norm": 0.24537350120004972, + "learning_rate": 9.968007661326369e-05, + "loss": 0.0323, + "step": 1410 + }, + { + "epoch": 0.9342105263157895, + "grad_norm": 0.21282115870790244, + "learning_rate": 9.966128281458832e-05, + "loss": 0.0309, + "step": 1420 + }, + { + "epoch": 0.9407894736842105, + "grad_norm": 0.22792100873990018, + "learning_rate": 9.964195451430653e-05, + "loss": 0.0302, + "step": 1430 + }, + { + "epoch": 0.9473684210526315, + "grad_norm": 0.2513211642537743, + "learning_rate": 9.962209192044772e-05, + "loss": 0.0298, + "step": 1440 + }, + { + "epoch": 0.9539473684210527, + "grad_norm": 0.25009559541845094, + "learning_rate": 9.960169524679187e-05, + "loss": 0.0293, + "step": 1450 + }, + { + "epoch": 0.9605263157894737, + "grad_norm": 0.23293995789061867, + "learning_rate": 9.958076471286724e-05, + "loss": 0.0289, + "step": 1460 + }, + { + "epoch": 0.9671052631578947, + "grad_norm": 0.22361624312785255, + "learning_rate": 9.955930054394798e-05, + "loss": 0.0282, + "step": 1470 + }, + { + "epoch": 0.9736842105263158, + "grad_norm": 0.23906361425178743, + "learning_rate": 9.953730297105175e-05, + "loss": 0.0281, + "step": 1480 + }, + { + "epoch": 0.9802631578947368, + "grad_norm": 0.23143391534241473, + "learning_rate": 9.951477223093722e-05, + "loss": 0.0275, + "step": 1490 + }, + { + "epoch": 0.9868421052631579, + "grad_norm": 0.2147052689574608, + "learning_rate": 9.949170856610144e-05, + "loss": 0.0271, + "step": 1500 + }, + { + "epoch": 0.993421052631579, + "grad_norm": 0.20377711124649098, + "learning_rate": 9.946811222477737e-05, + "loss": 0.0263, + "step": 1510 + }, + { + "epoch": 1.0, + "grad_norm": 0.2135903768894096, + "learning_rate": 9.944398346093109e-05, + "loss": 0.0262, + "step": 1520 + }, + { + "epoch": 1.006578947368421, + "grad_norm": 0.20254766138767108, + "learning_rate": 9.941932253425917e-05, + "loss": 0.0248, + "step": 1530 + }, + { + "epoch": 1.013157894736842, + "grad_norm": 0.2197030932035537, + "learning_rate": 9.939412971018574e-05, + "loss": 0.0246, + "step": 1540 + }, + { + "epoch": 1.019736842105263, + "grad_norm": 0.1933628837351623, + "learning_rate": 9.936840525985978e-05, + "loss": 0.0242, + "step": 1550 + }, + { + "epoch": 1.0263157894736843, + "grad_norm": 0.19721213696918533, + "learning_rate": 9.934214946015213e-05, + "loss": 0.0238, + "step": 1560 + }, + { + "epoch": 1.0328947368421053, + "grad_norm": 0.20774618691091581, + "learning_rate": 9.931536259365244e-05, + "loss": 0.0234, + "step": 1570 + }, + { + "epoch": 1.0394736842105263, + "grad_norm": 0.19785742831408606, + "learning_rate": 9.928804494866629e-05, + "loss": 0.0228, + "step": 1580 + }, + { + "epoch": 1.0460526315789473, + "grad_norm": 0.2035358859489929, + "learning_rate": 9.926019681921196e-05, + "loss": 0.0227, + "step": 1590 + }, + { + "epoch": 1.0526315789473684, + "grad_norm": 0.20629560858059492, + "learning_rate": 9.92318185050173e-05, + "loss": 0.0227, + "step": 1600 + }, + { + "epoch": 1.0592105263157894, + "grad_norm": 0.17943582514083126, + "learning_rate": 9.920291031151652e-05, + "loss": 0.0223, + "step": 1610 + }, + { + "epoch": 1.0657894736842106, + "grad_norm": 0.2069047937737935, + "learning_rate": 9.917347254984688e-05, + "loss": 0.0219, + "step": 1620 + }, + { + "epoch": 1.0723684210526316, + "grad_norm": 0.18149990279040348, + "learning_rate": 9.91435055368454e-05, + "loss": 0.0215, + "step": 1630 + }, + { + "epoch": 1.0789473684210527, + "grad_norm": 0.21588259501861692, + "learning_rate": 9.91130095950453e-05, + "loss": 0.0216, + "step": 1640 + }, + { + "epoch": 1.0855263157894737, + "grad_norm": 0.19248955273518795, + "learning_rate": 9.908198505267273e-05, + "loss": 0.0211, + "step": 1650 + }, + { + "epoch": 1.0921052631578947, + "grad_norm": 0.21323425974286558, + "learning_rate": 9.90504322436431e-05, + "loss": 0.0212, + "step": 1660 + }, + { + "epoch": 1.0986842105263157, + "grad_norm": 0.19659430552879165, + "learning_rate": 9.901835150755754e-05, + "loss": 0.0204, + "step": 1670 + }, + { + "epoch": 1.1052631578947367, + "grad_norm": 0.18723287056746327, + "learning_rate": 9.898574318969918e-05, + "loss": 0.0202, + "step": 1680 + }, + { + "epoch": 1.111842105263158, + "grad_norm": 0.18512625335755417, + "learning_rate": 9.895260764102956e-05, + "loss": 0.0198, + "step": 1690 + }, + { + "epoch": 1.118421052631579, + "grad_norm": 0.19124114649058113, + "learning_rate": 9.89189452181847e-05, + "loss": 0.0201, + "step": 1700 + }, + { + "epoch": 1.125, + "grad_norm": 0.18895302686535878, + "learning_rate": 9.888475628347142e-05, + "loss": 0.0195, + "step": 1710 + }, + { + "epoch": 1.131578947368421, + "grad_norm": 0.1772267382799342, + "learning_rate": 9.885004120486326e-05, + "loss": 0.0191, + "step": 1720 + }, + { + "epoch": 1.138157894736842, + "grad_norm": 0.17908235775562148, + "learning_rate": 9.881480035599667e-05, + "loss": 0.0189, + "step": 1730 + }, + { + "epoch": 1.1447368421052633, + "grad_norm": 0.17580586627136524, + "learning_rate": 9.877903411616697e-05, + "loss": 0.0189, + "step": 1740 + }, + { + "epoch": 1.1513157894736843, + "grad_norm": 0.18518601084678493, + "learning_rate": 9.874274287032417e-05, + "loss": 0.0186, + "step": 1750 + }, + { + "epoch": 1.1578947368421053, + "grad_norm": 0.18048608002452143, + "learning_rate": 9.870592700906892e-05, + "loss": 0.0183, + "step": 1760 + }, + { + "epoch": 1.1644736842105263, + "grad_norm": 0.15777383982627308, + "learning_rate": 9.866858692864829e-05, + "loss": 0.018, + "step": 1770 + }, + { + "epoch": 1.1710526315789473, + "grad_norm": 0.20607208990685497, + "learning_rate": 9.863072303095144e-05, + "loss": 0.0175, + "step": 1780 + }, + { + "epoch": 1.1776315789473684, + "grad_norm": 0.1611318675087161, + "learning_rate": 9.859233572350541e-05, + "loss": 0.0171, + "step": 1790 + }, + { + "epoch": 1.1842105263157894, + "grad_norm": 0.17589801094486104, + "learning_rate": 9.855342541947065e-05, + "loss": 0.0171, + "step": 1800 + }, + { + "epoch": 1.1907894736842106, + "grad_norm": 0.17121384019730954, + "learning_rate": 9.851399253763656e-05, + "loss": 0.017, + "step": 1810 + }, + { + "epoch": 1.1973684210526316, + "grad_norm": 0.16253563174923477, + "learning_rate": 9.847403750241704e-05, + "loss": 0.0166, + "step": 1820 + }, + { + "epoch": 1.2039473684210527, + "grad_norm": 0.17851024747091337, + "learning_rate": 9.84335607438459e-05, + "loss": 0.0166, + "step": 1830 + }, + { + "epoch": 1.2105263157894737, + "grad_norm": 0.16919035493728743, + "learning_rate": 9.839256269757223e-05, + "loss": 0.0161, + "step": 1840 + }, + { + "epoch": 1.2171052631578947, + "grad_norm": 0.1939807706315173, + "learning_rate": 9.835104380485572e-05, + "loss": 0.0163, + "step": 1850 + }, + { + "epoch": 1.2236842105263157, + "grad_norm": 0.17147592181772123, + "learning_rate": 9.830900451256187e-05, + "loss": 0.0157, + "step": 1860 + }, + { + "epoch": 1.2302631578947367, + "grad_norm": 0.17218945397214414, + "learning_rate": 9.826644527315722e-05, + "loss": 0.0156, + "step": 1870 + }, + { + "epoch": 1.236842105263158, + "grad_norm": 0.16873620481355664, + "learning_rate": 9.82233665447045e-05, + "loss": 0.0151, + "step": 1880 + }, + { + "epoch": 1.243421052631579, + "grad_norm": 0.16469742855514352, + "learning_rate": 9.817976879085764e-05, + "loss": 0.0153, + "step": 1890 + }, + { + "epoch": 1.25, + "grad_norm": 0.1692877712959756, + "learning_rate": 9.813565248085686e-05, + "loss": 0.0148, + "step": 1900 + }, + { + "epoch": 1.256578947368421, + "grad_norm": 0.16625631046909073, + "learning_rate": 9.809101808952348e-05, + "loss": 0.0148, + "step": 1910 + }, + { + "epoch": 1.263157894736842, + "grad_norm": 0.18262802890291954, + "learning_rate": 9.804586609725499e-05, + "loss": 0.0147, + "step": 1920 + }, + { + "epoch": 1.2697368421052633, + "grad_norm": 0.18607861180613644, + "learning_rate": 9.800019699001974e-05, + "loss": 0.0144, + "step": 1930 + }, + { + "epoch": 1.2763157894736843, + "grad_norm": 0.16285676137508193, + "learning_rate": 9.795401125935177e-05, + "loss": 0.0143, + "step": 1940 + }, + { + "epoch": 1.2828947368421053, + "grad_norm": 0.1707670608457863, + "learning_rate": 9.790730940234552e-05, + "loss": 0.0143, + "step": 1950 + }, + { + "epoch": 1.2894736842105263, + "grad_norm": 0.15525731103278195, + "learning_rate": 9.786009192165048e-05, + "loss": 0.0137, + "step": 1960 + }, + { + "epoch": 1.2960526315789473, + "grad_norm": 0.16347793214903164, + "learning_rate": 9.781235932546574e-05, + "loss": 0.0133, + "step": 1970 + }, + { + "epoch": 1.3026315789473684, + "grad_norm": 0.15933544137667022, + "learning_rate": 9.776411212753453e-05, + "loss": 0.0134, + "step": 1980 + }, + { + "epoch": 1.3092105263157894, + "grad_norm": 0.13814532510818728, + "learning_rate": 9.771535084713879e-05, + "loss": 0.0129, + "step": 1990 + }, + { + "epoch": 1.3157894736842106, + "grad_norm": 0.16523793061268452, + "learning_rate": 9.766607600909341e-05, + "loss": 0.0128, + "step": 2000 + }, + { + "epoch": 1.3223684210526316, + "grad_norm": 0.14022908820622706, + "learning_rate": 9.761628814374073e-05, + "loss": 0.0126, + "step": 2010 + }, + { + "epoch": 1.3289473684210527, + "grad_norm": 0.2069333200433914, + "learning_rate": 9.756598778694477e-05, + "loss": 0.0125, + "step": 2020 + }, + { + "epoch": 1.3355263157894737, + "grad_norm": 0.17334974425963282, + "learning_rate": 9.751517548008542e-05, + "loss": 0.0123, + "step": 2030 + }, + { + "epoch": 1.3421052631578947, + "grad_norm": 0.1505657722626437, + "learning_rate": 9.746385177005272e-05, + "loss": 0.0121, + "step": 2040 + }, + { + "epoch": 1.3486842105263157, + "grad_norm": 0.168192073561971, + "learning_rate": 9.741201720924089e-05, + "loss": 0.0117, + "step": 2050 + }, + { + "epoch": 1.3552631578947367, + "grad_norm": 0.15074266703733896, + "learning_rate": 9.73596723555424e-05, + "loss": 0.0119, + "step": 2060 + }, + { + "epoch": 1.361842105263158, + "grad_norm": 0.17348324277964447, + "learning_rate": 9.730681777234196e-05, + "loss": 0.0115, + "step": 2070 + }, + { + "epoch": 1.368421052631579, + "grad_norm": 0.14638399471630661, + "learning_rate": 9.725345402851053e-05, + "loss": 0.0118, + "step": 2080 + }, + { + "epoch": 1.375, + "grad_norm": 0.15352208983536622, + "learning_rate": 9.719958169839906e-05, + "loss": 0.0113, + "step": 2090 + }, + { + "epoch": 1.381578947368421, + "grad_norm": 0.16084962439952788, + "learning_rate": 9.714520136183243e-05, + "loss": 0.0115, + "step": 2100 + }, + { + "epoch": 1.388157894736842, + "grad_norm": 0.1344941180659645, + "learning_rate": 9.709031360410318e-05, + "loss": 0.011, + "step": 2110 + }, + { + "epoch": 1.3947368421052633, + "grad_norm": 0.1452614934065403, + "learning_rate": 9.703491901596516e-05, + "loss": 0.0109, + "step": 2120 + }, + { + "epoch": 1.4013157894736843, + "grad_norm": 0.13843890039006693, + "learning_rate": 9.697901819362721e-05, + "loss": 0.0107, + "step": 2130 + }, + { + "epoch": 1.4078947368421053, + "grad_norm": 0.13684375615554292, + "learning_rate": 9.692261173874675e-05, + "loss": 0.0105, + "step": 2140 + }, + { + "epoch": 1.4144736842105263, + "grad_norm": 0.14083447160331208, + "learning_rate": 9.68657002584233e-05, + "loss": 0.0102, + "step": 2150 + }, + { + "epoch": 1.4210526315789473, + "grad_norm": 0.13789659384952696, + "learning_rate": 9.680828436519196e-05, + "loss": 0.0104, + "step": 2160 + }, + { + "epoch": 1.4276315789473684, + "grad_norm": 0.14516526202810717, + "learning_rate": 9.675036467701674e-05, + "loss": 0.0101, + "step": 2170 + }, + { + "epoch": 1.4342105263157894, + "grad_norm": 0.12293537720506266, + "learning_rate": 9.669194181728403e-05, + "loss": 0.0099, + "step": 2180 + }, + { + "epoch": 1.4407894736842106, + "grad_norm": 0.1259249188196858, + "learning_rate": 9.663301641479579e-05, + "loss": 0.0097, + "step": 2190 + }, + { + "epoch": 1.4473684210526316, + "grad_norm": 0.12557298712724865, + "learning_rate": 9.65735891037628e-05, + "loss": 0.0097, + "step": 2200 + }, + { + "epoch": 1.4539473684210527, + "grad_norm": 0.143039377177587, + "learning_rate": 9.65136605237979e-05, + "loss": 0.0094, + "step": 2210 + }, + { + "epoch": 1.4605263157894737, + "grad_norm": 0.13504613441385002, + "learning_rate": 9.645323131990908e-05, + "loss": 0.0096, + "step": 2220 + }, + { + "epoch": 1.4671052631578947, + "grad_norm": 0.13228183618353204, + "learning_rate": 9.639230214249242e-05, + "loss": 0.0093, + "step": 2230 + }, + { + "epoch": 1.4736842105263157, + "grad_norm": 0.1480027541710592, + "learning_rate": 9.633087364732527e-05, + "loss": 0.0093, + "step": 2240 + }, + { + "epoch": 1.4802631578947367, + "grad_norm": 0.1293455699844313, + "learning_rate": 9.626894649555909e-05, + "loss": 0.0094, + "step": 2250 + }, + { + "epoch": 1.486842105263158, + "grad_norm": 0.1393940674869785, + "learning_rate": 9.620652135371232e-05, + "loss": 0.0091, + "step": 2260 + }, + { + "epoch": 1.493421052631579, + "grad_norm": 0.15121559409028545, + "learning_rate": 9.61435988936633e-05, + "loss": 0.0091, + "step": 2270 + }, + { + "epoch": 1.5, + "grad_norm": 0.1301327368552193, + "learning_rate": 9.608017979264292e-05, + "loss": 0.0091, + "step": 2280 + }, + { + "epoch": 1.506578947368421, + "grad_norm": 0.1421460799492161, + "learning_rate": 9.601626473322744e-05, + "loss": 0.0089, + "step": 2290 + }, + { + "epoch": 1.513157894736842, + "grad_norm": 0.17041691269286594, + "learning_rate": 9.595185440333103e-05, + "loss": 0.0092, + "step": 2300 + }, + { + "epoch": 1.5197368421052633, + "grad_norm": 0.14376606305315906, + "learning_rate": 9.588694949619851e-05, + "loss": 0.0086, + "step": 2310 + }, + { + "epoch": 1.526315789473684, + "grad_norm": 0.13694156071766578, + "learning_rate": 9.582155071039771e-05, + "loss": 0.0087, + "step": 2320 + }, + { + "epoch": 1.5328947368421053, + "grad_norm": 0.12727756352863123, + "learning_rate": 9.575565874981214e-05, + "loss": 0.0087, + "step": 2330 + }, + { + "epoch": 1.5394736842105263, + "grad_norm": 0.12785472470651071, + "learning_rate": 9.568927432363327e-05, + "loss": 0.0082, + "step": 2340 + }, + { + "epoch": 1.5460526315789473, + "grad_norm": 0.12091981259643964, + "learning_rate": 9.562239814635294e-05, + "loss": 0.0083, + "step": 2350 + }, + { + "epoch": 1.5526315789473686, + "grad_norm": 0.12758853427180586, + "learning_rate": 9.555503093775575e-05, + "loss": 0.008, + "step": 2360 + }, + { + "epoch": 1.5592105263157894, + "grad_norm": 0.12354223372657464, + "learning_rate": 9.548717342291115e-05, + "loss": 0.0081, + "step": 2370 + }, + { + "epoch": 1.5657894736842106, + "grad_norm": 0.12076843350000686, + "learning_rate": 9.541882633216579e-05, + "loss": 0.0079, + "step": 2380 + }, + { + "epoch": 1.5723684210526314, + "grad_norm": 0.1197614697335571, + "learning_rate": 9.534999040113559e-05, + "loss": 0.008, + "step": 2390 + }, + { + "epoch": 1.5789473684210527, + "grad_norm": 0.12153418093253635, + "learning_rate": 9.52806663706978e-05, + "loss": 0.0077, + "step": 2400 + }, + { + "epoch": 1.5855263157894737, + "grad_norm": 0.13543828010220124, + "learning_rate": 9.521085498698309e-05, + "loss": 0.0079, + "step": 2410 + }, + { + "epoch": 1.5921052631578947, + "grad_norm": 0.10948806505136653, + "learning_rate": 9.514055700136747e-05, + "loss": 0.0077, + "step": 2420 + }, + { + "epoch": 1.598684210526316, + "grad_norm": 0.10980266602887555, + "learning_rate": 9.506977317046424e-05, + "loss": 0.0074, + "step": 2430 + }, + { + "epoch": 1.6052631578947367, + "grad_norm": 0.12485717268206407, + "learning_rate": 9.49985042561158e-05, + "loss": 0.0075, + "step": 2440 + }, + { + "epoch": 1.611842105263158, + "grad_norm": 0.11541304858534408, + "learning_rate": 9.492675102538547e-05, + "loss": 0.0072, + "step": 2450 + }, + { + "epoch": 1.618421052631579, + "grad_norm": 0.12122451397522918, + "learning_rate": 9.485451425054926e-05, + "loss": 0.0072, + "step": 2460 + }, + { + "epoch": 1.625, + "grad_norm": 0.4210179348904546, + "learning_rate": 9.478179470908756e-05, + "loss": 0.0075, + "step": 2470 + }, + { + "epoch": 1.631578947368421, + "grad_norm": 0.5896074678764591, + "learning_rate": 9.470859318367669e-05, + "loss": 0.0086, + "step": 2480 + }, + { + "epoch": 1.638157894736842, + "grad_norm": 3.7823641198616387, + "learning_rate": 9.463491046218058e-05, + "loss": 0.0186, + "step": 2490 + }, + { + "epoch": 1.6447368421052633, + "grad_norm": 5.014692495425624, + "learning_rate": 9.456074733764224e-05, + "loss": 0.0827, + "step": 2500 + }, + { + "epoch": 1.651315789473684, + "grad_norm": 0.30328471004984053, + "learning_rate": 9.448610460827522e-05, + "loss": 0.0328, + "step": 2510 + }, + { + "epoch": 1.6578947368421053, + "grad_norm": 0.17555996108962088, + "learning_rate": 9.441098307745507e-05, + "loss": 0.0119, + "step": 2520 + }, + { + "epoch": 1.6644736842105263, + "grad_norm": 0.12830924181654307, + "learning_rate": 9.433538355371057e-05, + "loss": 0.0096, + "step": 2530 + }, + { + "epoch": 1.6710526315789473, + "grad_norm": 0.1306978986667964, + "learning_rate": 9.425930685071517e-05, + "loss": 0.0083, + "step": 2540 + }, + { + "epoch": 1.6776315789473686, + "grad_norm": 0.1507331325454933, + "learning_rate": 9.418275378727819e-05, + "loss": 0.0083, + "step": 2550 + }, + { + "epoch": 1.6842105263157894, + "grad_norm": 0.11592835345613971, + "learning_rate": 9.410572518733596e-05, + "loss": 0.0079, + "step": 2560 + }, + { + "epoch": 1.6907894736842106, + "grad_norm": 0.10808538914723062, + "learning_rate": 9.402822187994298e-05, + "loss": 0.0073, + "step": 2570 + }, + { + "epoch": 1.6973684210526314, + "grad_norm": 0.11722096305528684, + "learning_rate": 9.395024469926302e-05, + "loss": 0.0073, + "step": 2580 + }, + { + "epoch": 1.7039473684210527, + "grad_norm": 0.10458331688166168, + "learning_rate": 9.387179448456009e-05, + "loss": 0.0072, + "step": 2590 + }, + { + "epoch": 1.7105263157894737, + "grad_norm": 0.11291592102103867, + "learning_rate": 9.379287208018945e-05, + "loss": 0.0067, + "step": 2600 + }, + { + "epoch": 1.7171052631578947, + "grad_norm": 0.08646582566807197, + "learning_rate": 9.371347833558856e-05, + "loss": 0.0065, + "step": 2610 + }, + { + "epoch": 1.723684210526316, + "grad_norm": 0.10553066382224695, + "learning_rate": 9.363361410526783e-05, + "loss": 0.0067, + "step": 2620 + }, + { + "epoch": 1.7302631578947367, + "grad_norm": 0.09811144117737546, + "learning_rate": 9.35532802488015e-05, + "loss": 0.0065, + "step": 2630 + }, + { + "epoch": 1.736842105263158, + "grad_norm": 0.09975550040113049, + "learning_rate": 9.347247763081835e-05, + "loss": 0.0064, + "step": 2640 + }, + { + "epoch": 1.743421052631579, + "grad_norm": 0.11209372785035113, + "learning_rate": 9.339120712099245e-05, + "loss": 0.0065, + "step": 2650 + }, + { + "epoch": 1.75, + "grad_norm": 0.09335458400468914, + "learning_rate": 9.330946959403375e-05, + "loss": 0.0061, + "step": 2660 + }, + { + "epoch": 1.756578947368421, + "grad_norm": 0.0975073527993663, + "learning_rate": 9.322726592967868e-05, + "loss": 0.006, + "step": 2670 + }, + { + "epoch": 1.763157894736842, + "grad_norm": 0.09366196020997528, + "learning_rate": 9.314459701268065e-05, + "loss": 0.006, + "step": 2680 + }, + { + "epoch": 1.7697368421052633, + "grad_norm": 0.09562732934182246, + "learning_rate": 9.306146373280061e-05, + "loss": 0.0059, + "step": 2690 + }, + { + "epoch": 1.776315789473684, + "grad_norm": 0.09966326146415878, + "learning_rate": 9.297786698479739e-05, + "loss": 0.006, + "step": 2700 + }, + { + "epoch": 1.7828947368421053, + "grad_norm": 0.10122383105599052, + "learning_rate": 9.289380766841811e-05, + "loss": 0.0059, + "step": 2710 + }, + { + "epoch": 1.7894736842105263, + "grad_norm": 0.10095110671102905, + "learning_rate": 9.280928668838848e-05, + "loss": 0.0057, + "step": 2720 + }, + { + "epoch": 1.7960526315789473, + "grad_norm": 0.091482492267746, + "learning_rate": 9.27243049544031e-05, + "loss": 0.0057, + "step": 2730 + }, + { + "epoch": 1.8026315789473686, + "grad_norm": 0.08891652643763134, + "learning_rate": 9.263886338111559e-05, + "loss": 0.0055, + "step": 2740 + }, + { + "epoch": 1.8092105263157894, + "grad_norm": 0.09263144069565209, + "learning_rate": 9.25529628881288e-05, + "loss": 0.0055, + "step": 2750 + }, + { + "epoch": 1.8157894736842106, + "grad_norm": 0.11137233685500829, + "learning_rate": 9.246660439998497e-05, + "loss": 0.0055, + "step": 2760 + }, + { + "epoch": 1.8223684210526314, + "grad_norm": 0.11114880917458508, + "learning_rate": 9.237978884615565e-05, + "loss": 0.0054, + "step": 2770 + }, + { + "epoch": 1.8289473684210527, + "grad_norm": 0.09841409593938921, + "learning_rate": 9.229251716103179e-05, + "loss": 0.0055, + "step": 2780 + }, + { + "epoch": 1.8355263157894737, + "grad_norm": 0.11223577944844708, + "learning_rate": 9.220479028391364e-05, + "loss": 0.0054, + "step": 2790 + }, + { + "epoch": 1.8421052631578947, + "grad_norm": 0.09110216546608073, + "learning_rate": 9.211660915900069e-05, + "loss": 0.0053, + "step": 2800 + }, + { + "epoch": 1.848684210526316, + "grad_norm": 0.09506804998323148, + "learning_rate": 9.202797473538143e-05, + "loss": 0.0053, + "step": 2810 + }, + { + "epoch": 1.8552631578947367, + "grad_norm": 0.09167591406498078, + "learning_rate": 9.193888796702321e-05, + "loss": 0.0052, + "step": 2820 + }, + { + "epoch": 1.861842105263158, + "grad_norm": 0.0910480942341495, + "learning_rate": 9.184934981276192e-05, + "loss": 0.0053, + "step": 2830 + }, + { + "epoch": 1.868421052631579, + "grad_norm": 0.08389589500624173, + "learning_rate": 9.175936123629171e-05, + "loss": 0.0053, + "step": 2840 + }, + { + "epoch": 1.875, + "grad_norm": 0.08751212493753141, + "learning_rate": 9.16689232061546e-05, + "loss": 0.0052, + "step": 2850 + }, + { + "epoch": 1.881578947368421, + "grad_norm": 0.08925947798147087, + "learning_rate": 9.157803669573e-05, + "loss": 0.005, + "step": 2860 + }, + { + "epoch": 1.888157894736842, + "grad_norm": 0.0896619316816315, + "learning_rate": 9.148670268322438e-05, + "loss": 0.005, + "step": 2870 + }, + { + "epoch": 1.8947368421052633, + "grad_norm": 0.09175226088849178, + "learning_rate": 9.139492215166059e-05, + "loss": 0.0048, + "step": 2880 + }, + { + "epoch": 1.901315789473684, + "grad_norm": 0.08885389457486895, + "learning_rate": 9.130269608886735e-05, + "loss": 0.005, + "step": 2890 + }, + { + "epoch": 1.9078947368421053, + "grad_norm": 0.09883908696946943, + "learning_rate": 9.12100254874686e-05, + "loss": 0.0051, + "step": 2900 + }, + { + "epoch": 1.9144736842105263, + "grad_norm": 0.09233528269274262, + "learning_rate": 9.111691134487286e-05, + "loss": 0.005, + "step": 2910 + }, + { + "epoch": 1.9210526315789473, + "grad_norm": 0.08544980345629175, + "learning_rate": 9.102335466326241e-05, + "loss": 0.0049, + "step": 2920 + }, + { + "epoch": 1.9276315789473686, + "grad_norm": 0.10003966542752647, + "learning_rate": 9.092935644958257e-05, + "loss": 0.0048, + "step": 2930 + }, + { + "epoch": 1.9342105263157894, + "grad_norm": 0.08168055005388938, + "learning_rate": 9.083491771553089e-05, + "loss": 0.0047, + "step": 2940 + }, + { + "epoch": 1.9407894736842106, + "grad_norm": 0.0992409697866901, + "learning_rate": 9.074003947754615e-05, + "loss": 0.0047, + "step": 2950 + }, + { + "epoch": 1.9473684210526314, + "grad_norm": 0.08492119209203541, + "learning_rate": 9.06447227567975e-05, + "loss": 0.0048, + "step": 2960 + }, + { + "epoch": 1.9539473684210527, + "grad_norm": 0.08129120975425032, + "learning_rate": 9.054896857917349e-05, + "loss": 0.0048, + "step": 2970 + }, + { + "epoch": 1.9605263157894737, + "grad_norm": 0.08471582707226462, + "learning_rate": 9.045277797527096e-05, + "loss": 0.0047, + "step": 2980 + }, + { + "epoch": 1.9671052631578947, + "grad_norm": 0.08915074149762475, + "learning_rate": 9.035615198038401e-05, + "loss": 0.0047, + "step": 2990 + }, + { + "epoch": 1.973684210526316, + "grad_norm": 0.08922804724560857, + "learning_rate": 9.02590916344928e-05, + "loss": 0.0046, + "step": 3000 + }, + { + "epoch": 1.9802631578947367, + "grad_norm": 0.08357488105287375, + "learning_rate": 9.016159798225239e-05, + "loss": 0.0045, + "step": 3010 + }, + { + "epoch": 1.986842105263158, + "grad_norm": 0.09089197660267276, + "learning_rate": 9.006367207298152e-05, + "loss": 0.0045, + "step": 3020 + }, + { + "epoch": 1.993421052631579, + "grad_norm": 0.07872067497811534, + "learning_rate": 8.996531496065128e-05, + "loss": 0.0046, + "step": 3030 + }, + { + "epoch": 2.0, + "grad_norm": 0.08439906072392114, + "learning_rate": 8.986652770387372e-05, + "loss": 0.0044, + "step": 3040 + }, + { + "epoch": 2.0065789473684212, + "grad_norm": 0.06961910527253773, + "learning_rate": 8.976731136589059e-05, + "loss": 0.0033, + "step": 3050 + }, + { + "epoch": 2.013157894736842, + "grad_norm": 0.07154791151160267, + "learning_rate": 8.966766701456177e-05, + "loss": 0.0033, + "step": 3060 + }, + { + "epoch": 2.0197368421052633, + "grad_norm": 0.07772175093467776, + "learning_rate": 8.956759572235383e-05, + "loss": 0.0031, + "step": 3070 + }, + { + "epoch": 2.026315789473684, + "grad_norm": 0.0733292062504736, + "learning_rate": 8.946709856632848e-05, + "loss": 0.0031, + "step": 3080 + }, + { + "epoch": 2.0328947368421053, + "grad_norm": 0.07783736387463817, + "learning_rate": 8.9366176628131e-05, + "loss": 0.003, + "step": 3090 + }, + { + "epoch": 2.039473684210526, + "grad_norm": 0.06633555943399455, + "learning_rate": 8.926483099397855e-05, + "loss": 0.003, + "step": 3100 + }, + { + "epoch": 2.0460526315789473, + "grad_norm": 0.07782294604433761, + "learning_rate": 8.916306275464852e-05, + "loss": 0.0029, + "step": 3110 + }, + { + "epoch": 2.0526315789473686, + "grad_norm": 0.07670901138238165, + "learning_rate": 8.906087300546675e-05, + "loss": 0.0031, + "step": 3120 + }, + { + "epoch": 2.0592105263157894, + "grad_norm": 0.07524758423152544, + "learning_rate": 8.895826284629582e-05, + "loss": 0.0029, + "step": 3130 + }, + { + "epoch": 2.0657894736842106, + "grad_norm": 0.07695970302800449, + "learning_rate": 8.885523338152313e-05, + "loss": 0.0031, + "step": 3140 + }, + { + "epoch": 2.0723684210526314, + "grad_norm": 0.0745426942514364, + "learning_rate": 8.875178572004903e-05, + "loss": 0.0032, + "step": 3150 + }, + { + "epoch": 2.0789473684210527, + "grad_norm": 0.0723930470222858, + "learning_rate": 8.864792097527489e-05, + "loss": 0.0031, + "step": 3160 + }, + { + "epoch": 2.085526315789474, + "grad_norm": 0.08361406797403782, + "learning_rate": 8.854364026509117e-05, + "loss": 0.0031, + "step": 3170 + }, + { + "epoch": 2.0921052631578947, + "grad_norm": 0.06955666676258866, + "learning_rate": 8.843894471186529e-05, + "loss": 0.0031, + "step": 3180 + }, + { + "epoch": 2.098684210526316, + "grad_norm": 0.07227264207927699, + "learning_rate": 8.833383544242966e-05, + "loss": 0.0031, + "step": 3190 + }, + { + "epoch": 2.1052631578947367, + "grad_norm": 0.06914653302815033, + "learning_rate": 8.822831358806942e-05, + "loss": 0.0029, + "step": 3200 + }, + { + "epoch": 2.111842105263158, + "grad_norm": 0.07323401762292578, + "learning_rate": 8.812238028451043e-05, + "loss": 0.003, + "step": 3210 + }, + { + "epoch": 2.1184210526315788, + "grad_norm": 0.07393279181869614, + "learning_rate": 8.801603667190686e-05, + "loss": 0.003, + "step": 3220 + }, + { + "epoch": 2.125, + "grad_norm": 0.07096487949532827, + "learning_rate": 8.790928389482909e-05, + "loss": 0.003, + "step": 3230 + }, + { + "epoch": 2.1315789473684212, + "grad_norm": 0.0728089853059884, + "learning_rate": 8.780212310225126e-05, + "loss": 0.0031, + "step": 3240 + }, + { + "epoch": 2.138157894736842, + "grad_norm": 0.0695501122917682, + "learning_rate": 8.7694555447539e-05, + "loss": 0.0031, + "step": 3250 + }, + { + "epoch": 2.1447368421052633, + "grad_norm": 0.07705485434544583, + "learning_rate": 8.758658208843694e-05, + "loss": 0.0031, + "step": 3260 + }, + { + "epoch": 2.151315789473684, + "grad_norm": 0.0845273469079357, + "learning_rate": 8.74782041870563e-05, + "loss": 0.0031, + "step": 3270 + }, + { + "epoch": 2.1578947368421053, + "grad_norm": 0.08068281928940782, + "learning_rate": 8.736942290986239e-05, + "loss": 0.0031, + "step": 3280 + }, + { + "epoch": 2.1644736842105265, + "grad_norm": 0.07961859812048024, + "learning_rate": 8.726023942766196e-05, + "loss": 0.003, + "step": 3290 + }, + { + "epoch": 2.1710526315789473, + "grad_norm": 0.08758099933665693, + "learning_rate": 8.715065491559076e-05, + "loss": 0.0029, + "step": 3300 + }, + { + "epoch": 2.1776315789473686, + "grad_norm": 0.07188296495338724, + "learning_rate": 8.704067055310072e-05, + "loss": 0.0031, + "step": 3310 + }, + { + "epoch": 2.1842105263157894, + "grad_norm": 0.07663306624675915, + "learning_rate": 8.693028752394741e-05, + "loss": 0.003, + "step": 3320 + }, + { + "epoch": 2.1907894736842106, + "grad_norm": 0.07462874387461102, + "learning_rate": 8.681950701617715e-05, + "loss": 0.003, + "step": 3330 + }, + { + "epoch": 2.1973684210526314, + "grad_norm": 0.07336225552850846, + "learning_rate": 8.67083302221144e-05, + "loss": 0.003, + "step": 3340 + }, + { + "epoch": 2.2039473684210527, + "grad_norm": 0.0752519347418371, + "learning_rate": 8.659675833834872e-05, + "loss": 0.0029, + "step": 3350 + }, + { + "epoch": 2.2105263157894735, + "grad_norm": 0.0727931975094885, + "learning_rate": 8.648479256572208e-05, + "loss": 0.003, + "step": 3360 + }, + { + "epoch": 2.2171052631578947, + "grad_norm": 0.07511952648108541, + "learning_rate": 8.637243410931582e-05, + "loss": 0.0029, + "step": 3370 + }, + { + "epoch": 2.223684210526316, + "grad_norm": 0.07882019908319284, + "learning_rate": 8.625968417843769e-05, + "loss": 0.0029, + "step": 3380 + }, + { + "epoch": 2.2302631578947367, + "grad_norm": 0.08265112815093757, + "learning_rate": 8.614654398660889e-05, + "loss": 0.0029, + "step": 3390 + }, + { + "epoch": 2.236842105263158, + "grad_norm": 0.0797172772654838, + "learning_rate": 8.603301475155098e-05, + "loss": 0.003, + "step": 3400 + }, + { + "epoch": 2.2434210526315788, + "grad_norm": 0.07058513328217333, + "learning_rate": 8.591909769517277e-05, + "loss": 0.0031, + "step": 3410 + }, + { + "epoch": 2.25, + "grad_norm": 0.0806856826321758, + "learning_rate": 8.580479404355714e-05, + "loss": 0.003, + "step": 3420 + }, + { + "epoch": 2.2565789473684212, + "grad_norm": 0.08816650076302643, + "learning_rate": 8.56901050269479e-05, + "loss": 0.003, + "step": 3430 + }, + { + "epoch": 2.263157894736842, + "grad_norm": 0.08576523991244484, + "learning_rate": 8.557503187973651e-05, + "loss": 0.0031, + "step": 3440 + }, + { + "epoch": 2.2697368421052633, + "grad_norm": 0.07817935240746017, + "learning_rate": 8.545957584044882e-05, + "loss": 0.003, + "step": 3450 + }, + { + "epoch": 2.276315789473684, + "grad_norm": 0.07132102691438429, + "learning_rate": 8.534373815173173e-05, + "loss": 0.0028, + "step": 3460 + }, + { + "epoch": 2.2828947368421053, + "grad_norm": 0.06907666862127639, + "learning_rate": 8.522752006033977e-05, + "loss": 0.0029, + "step": 3470 + }, + { + "epoch": 2.2894736842105265, + "grad_norm": 0.07888763534903219, + "learning_rate": 8.511092281712174e-05, + "loss": 0.0028, + "step": 3480 + }, + { + "epoch": 2.2960526315789473, + "grad_norm": 0.07104911749539242, + "learning_rate": 8.499394767700728e-05, + "loss": 0.0028, + "step": 3490 + }, + { + "epoch": 2.3026315789473686, + "grad_norm": 0.07161589717548758, + "learning_rate": 8.487659589899325e-05, + "loss": 0.003, + "step": 3500 + }, + { + "epoch": 2.3092105263157894, + "grad_norm": 0.07545738438626039, + "learning_rate": 8.475886874613027e-05, + "loss": 0.003, + "step": 3510 + }, + { + "epoch": 2.3157894736842106, + "grad_norm": 0.06960029822263444, + "learning_rate": 8.464076748550911e-05, + "loss": 0.0028, + "step": 3520 + }, + { + "epoch": 2.3223684210526314, + "grad_norm": 0.09053565638501836, + "learning_rate": 8.4522293388247e-05, + "loss": 0.003, + "step": 3530 + }, + { + "epoch": 2.3289473684210527, + "grad_norm": 0.08111502757412685, + "learning_rate": 8.440344772947403e-05, + "loss": 0.0029, + "step": 3540 + }, + { + "epoch": 2.3355263157894735, + "grad_norm": 0.07907419444465598, + "learning_rate": 8.428423178831937e-05, + "loss": 0.0029, + "step": 3550 + }, + { + "epoch": 2.3421052631578947, + "grad_norm": 0.07530055852137314, + "learning_rate": 8.416464684789751e-05, + "loss": 0.0033, + "step": 3560 + }, + { + "epoch": 2.348684210526316, + "grad_norm": 0.08578659904895668, + "learning_rate": 8.404469419529448e-05, + "loss": 0.0032, + "step": 3570 + }, + { + "epoch": 2.3552631578947367, + "grad_norm": 0.07749694254198063, + "learning_rate": 8.392437512155397e-05, + "loss": 0.003, + "step": 3580 + }, + { + "epoch": 2.361842105263158, + "grad_norm": 0.09613810033382267, + "learning_rate": 8.380369092166343e-05, + "loss": 0.0031, + "step": 3590 + }, + { + "epoch": 2.3684210526315788, + "grad_norm": 0.10312166758776871, + "learning_rate": 8.368264289454012e-05, + "loss": 0.003, + "step": 3600 + }, + { + "epoch": 2.375, + "grad_norm": 0.06867498813233379, + "learning_rate": 8.35612323430172e-05, + "loss": 0.003, + "step": 3610 + }, + { + "epoch": 2.3815789473684212, + "grad_norm": 0.07705859736599766, + "learning_rate": 8.343946057382967e-05, + "loss": 0.0031, + "step": 3620 + }, + { + "epoch": 2.388157894736842, + "grad_norm": 0.07041390906154826, + "learning_rate": 8.33173288976002e-05, + "loss": 0.0029, + "step": 3630 + }, + { + "epoch": 2.3947368421052633, + "grad_norm": 0.08776967253298393, + "learning_rate": 8.319483862882523e-05, + "loss": 0.003, + "step": 3640 + }, + { + "epoch": 2.401315789473684, + "grad_norm": 0.07556100987543657, + "learning_rate": 8.307199108586064e-05, + "loss": 0.0028, + "step": 3650 + }, + { + "epoch": 2.4078947368421053, + "grad_norm": 0.07071854707231168, + "learning_rate": 8.294878759090769e-05, + "loss": 0.003, + "step": 3660 + }, + { + "epoch": 2.4144736842105265, + "grad_norm": 0.06851373158159291, + "learning_rate": 8.282522946999865e-05, + "loss": 0.0028, + "step": 3670 + }, + { + "epoch": 2.4210526315789473, + "grad_norm": 0.07586987705553193, + "learning_rate": 8.270131805298269e-05, + "loss": 0.003, + "step": 3680 + }, + { + "epoch": 2.4276315789473686, + "grad_norm": 0.0711029870103729, + "learning_rate": 8.257705467351145e-05, + "loss": 0.0028, + "step": 3690 + }, + { + "epoch": 2.4342105263157894, + "grad_norm": 0.07353488997496402, + "learning_rate": 8.245244066902472e-05, + "loss": 0.0027, + "step": 3700 + }, + { + "epoch": 2.4407894736842106, + "grad_norm": 0.08069559598721139, + "learning_rate": 8.232747738073606e-05, + "loss": 0.0027, + "step": 3710 + }, + { + "epoch": 2.4473684210526314, + "grad_norm": 0.08032004169661551, + "learning_rate": 8.220216615361833e-05, + "loss": 0.0028, + "step": 3720 + }, + { + "epoch": 2.4539473684210527, + "grad_norm": 0.06368051417844219, + "learning_rate": 8.207650833638926e-05, + "loss": 0.0026, + "step": 3730 + }, + { + "epoch": 2.4605263157894735, + "grad_norm": 0.0674086770746194, + "learning_rate": 8.195050528149693e-05, + "loss": 0.0029, + "step": 3740 + }, + { + "epoch": 2.4671052631578947, + "grad_norm": 0.07830285476547033, + "learning_rate": 8.18241583451051e-05, + "loss": 0.0027, + "step": 3750 + }, + { + "epoch": 2.473684210526316, + "grad_norm": 0.0866857145186936, + "learning_rate": 8.169746888707884e-05, + "loss": 0.0027, + "step": 3760 + }, + { + "epoch": 2.4802631578947367, + "grad_norm": 0.06911558660644061, + "learning_rate": 8.157043827096964e-05, + "loss": 0.0028, + "step": 3770 + }, + { + "epoch": 2.486842105263158, + "grad_norm": 0.07259825247456937, + "learning_rate": 8.144306786400096e-05, + "loss": 0.0028, + "step": 3780 + }, + { + "epoch": 2.4934210526315788, + "grad_norm": 0.06844850269700717, + "learning_rate": 8.131535903705332e-05, + "loss": 0.0029, + "step": 3790 + }, + { + "epoch": 2.5, + "grad_norm": 0.08278983177211563, + "learning_rate": 8.118731316464967e-05, + "loss": 0.0027, + "step": 3800 + }, + { + "epoch": 2.5065789473684212, + "grad_norm": 0.06219877650762257, + "learning_rate": 8.10589316249406e-05, + "loss": 0.0027, + "step": 3810 + }, + { + "epoch": 2.513157894736842, + "grad_norm": 0.06824657148655913, + "learning_rate": 8.093021579968941e-05, + "loss": 0.0028, + "step": 3820 + }, + { + "epoch": 2.5197368421052633, + "grad_norm": 0.07996281509026834, + "learning_rate": 8.080116707425732e-05, + "loss": 0.003, + "step": 3830 + }, + { + "epoch": 2.526315789473684, + "grad_norm": 0.06422551748134994, + "learning_rate": 8.067178683758855e-05, + "loss": 0.0027, + "step": 3840 + }, + { + "epoch": 2.5328947368421053, + "grad_norm": 0.09323022007000262, + "learning_rate": 8.054207648219534e-05, + "loss": 0.0029, + "step": 3850 + }, + { + "epoch": 2.5394736842105265, + "grad_norm": 0.07497257348678213, + "learning_rate": 8.041203740414298e-05, + "loss": 0.0028, + "step": 3860 + }, + { + "epoch": 2.5460526315789473, + "grad_norm": 0.08301851491054957, + "learning_rate": 8.028167100303476e-05, + "loss": 0.0028, + "step": 3870 + }, + { + "epoch": 2.5526315789473686, + "grad_norm": 0.06474154045528839, + "learning_rate": 8.015097868199699e-05, + "loss": 0.0028, + "step": 3880 + }, + { + "epoch": 2.5592105263157894, + "grad_norm": 0.06874706823653907, + "learning_rate": 8.001996184766381e-05, + "loss": 0.0027, + "step": 3890 + }, + { + "epoch": 2.5657894736842106, + "grad_norm": 0.08107437503239583, + "learning_rate": 7.988862191016205e-05, + "loss": 0.0026, + "step": 3900 + }, + { + "epoch": 2.5723684210526314, + "grad_norm": 0.07597780656490642, + "learning_rate": 7.975696028309611e-05, + "loss": 0.0026, + "step": 3910 + }, + { + "epoch": 2.5789473684210527, + "grad_norm": 0.07357713377704023, + "learning_rate": 7.96249783835327e-05, + "loss": 0.0026, + "step": 3920 + }, + { + "epoch": 2.5855263157894735, + "grad_norm": 0.06225901705415947, + "learning_rate": 7.949267763198563e-05, + "loss": 0.0027, + "step": 3930 + }, + { + "epoch": 2.5921052631578947, + "grad_norm": 0.0703385334924199, + "learning_rate": 7.936005945240051e-05, + "loss": 0.0026, + "step": 3940 + }, + { + "epoch": 2.598684210526316, + "grad_norm": 0.0736999092196569, + "learning_rate": 7.92271252721393e-05, + "loss": 0.0026, + "step": 3950 + }, + { + "epoch": 2.6052631578947367, + "grad_norm": 0.07729069940862861, + "learning_rate": 7.909387652196522e-05, + "loss": 0.0026, + "step": 3960 + }, + { + "epoch": 2.611842105263158, + "grad_norm": 0.07481211359608593, + "learning_rate": 7.896031463602706e-05, + "loss": 0.0024, + "step": 3970 + }, + { + "epoch": 2.6184210526315788, + "grad_norm": 0.06814464123536673, + "learning_rate": 7.882644105184395e-05, + "loss": 0.0025, + "step": 3980 + }, + { + "epoch": 2.625, + "grad_norm": 0.07129054851729393, + "learning_rate": 7.869225721028978e-05, + "loss": 0.0026, + "step": 3990 + }, + { + "epoch": 2.6315789473684212, + "grad_norm": 0.08376454356231268, + "learning_rate": 7.855776455557769e-05, + "loss": 0.0026, + "step": 4000 + }, + { + "epoch": 2.638157894736842, + "grad_norm": 0.06662673919070906, + "learning_rate": 7.842296453524463e-05, + "loss": 0.0025, + "step": 4010 + }, + { + "epoch": 2.6447368421052633, + "grad_norm": 0.06020388907136684, + "learning_rate": 7.828785860013564e-05, + "loss": 0.0025, + "step": 4020 + }, + { + "epoch": 2.651315789473684, + "grad_norm": 0.06444105223285773, + "learning_rate": 7.815244820438836e-05, + "loss": 0.0023, + "step": 4030 + }, + { + "epoch": 2.6578947368421053, + "grad_norm": 0.06050467090500708, + "learning_rate": 7.801673480541728e-05, + "loss": 0.0024, + "step": 4040 + }, + { + "epoch": 2.6644736842105265, + "grad_norm": 0.06527207049067266, + "learning_rate": 7.788071986389814e-05, + "loss": 0.0024, + "step": 4050 + }, + { + "epoch": 2.6710526315789473, + "grad_norm": 0.06979889616926872, + "learning_rate": 7.774440484375212e-05, + "loss": 0.0023, + "step": 4060 + }, + { + "epoch": 2.6776315789473686, + "grad_norm": 0.06743094271741353, + "learning_rate": 7.760779121213013e-05, + "loss": 0.0023, + "step": 4070 + }, + { + "epoch": 2.6842105263157894, + "grad_norm": 0.07204637868870485, + "learning_rate": 7.747088043939709e-05, + "loss": 0.0024, + "step": 4080 + }, + { + "epoch": 2.6907894736842106, + "grad_norm": 0.06415928404465751, + "learning_rate": 7.733367399911593e-05, + "loss": 0.0023, + "step": 4090 + }, + { + "epoch": 2.6973684210526314, + "grad_norm": 0.07280188098779707, + "learning_rate": 7.719617336803194e-05, + "loss": 0.0026, + "step": 4100 + }, + { + "epoch": 2.7039473684210527, + "grad_norm": 0.05998059047850096, + "learning_rate": 7.705838002605665e-05, + "loss": 0.0024, + "step": 4110 + }, + { + "epoch": 2.7105263157894735, + "grad_norm": 0.06400082013104166, + "learning_rate": 7.692029545625213e-05, + "loss": 0.0024, + "step": 4120 + }, + { + "epoch": 2.7171052631578947, + "grad_norm": 0.07352802986302866, + "learning_rate": 7.678192114481487e-05, + "loss": 0.0024, + "step": 4130 + }, + { + "epoch": 2.723684210526316, + "grad_norm": 0.0702258918716786, + "learning_rate": 7.664325858105981e-05, + "loss": 0.0024, + "step": 4140 + }, + { + "epoch": 2.7302631578947367, + "grad_norm": 0.06641871081534573, + "learning_rate": 7.650430925740438e-05, + "loss": 0.0023, + "step": 4150 + }, + { + "epoch": 2.736842105263158, + "grad_norm": 0.06546274261192724, + "learning_rate": 7.636507466935237e-05, + "loss": 0.0023, + "step": 4160 + }, + { + "epoch": 2.7434210526315788, + "grad_norm": 0.06644783924036698, + "learning_rate": 7.622555631547784e-05, + "loss": 0.0023, + "step": 4170 + }, + { + "epoch": 2.75, + "grad_norm": 0.05362886039562456, + "learning_rate": 7.608575569740906e-05, + "loss": 0.0022, + "step": 4180 + }, + { + "epoch": 2.7565789473684212, + "grad_norm": 0.06903071092819318, + "learning_rate": 7.59456743198122e-05, + "loss": 0.0023, + "step": 4190 + }, + { + "epoch": 2.763157894736842, + "grad_norm": 0.05789722056728834, + "learning_rate": 7.580531369037533e-05, + "loss": 0.0021, + "step": 4200 + }, + { + "epoch": 2.7697368421052633, + "grad_norm": 0.0643558212306574, + "learning_rate": 7.566467531979202e-05, + "loss": 0.0023, + "step": 4210 + }, + { + "epoch": 2.776315789473684, + "grad_norm": 0.07052044594959722, + "learning_rate": 7.55237607217452e-05, + "loss": 0.0023, + "step": 4220 + }, + { + "epoch": 2.7828947368421053, + "grad_norm": 0.064534269417758, + "learning_rate": 7.538257141289079e-05, + "loss": 0.0022, + "step": 4230 + }, + { + "epoch": 2.7894736842105265, + "grad_norm": 0.060914750483076874, + "learning_rate": 7.52411089128414e-05, + "loss": 0.0021, + "step": 4240 + }, + { + "epoch": 2.7960526315789473, + "grad_norm": 0.06975178617845787, + "learning_rate": 7.509937474415001e-05, + "loss": 0.0021, + "step": 4250 + }, + { + "epoch": 2.8026315789473686, + "grad_norm": 0.05801976261020139, + "learning_rate": 7.495737043229356e-05, + "loss": 0.0022, + "step": 4260 + }, + { + "epoch": 2.8092105263157894, + "grad_norm": 0.07312109565555201, + "learning_rate": 7.481509750565647e-05, + "loss": 0.0022, + "step": 4270 + }, + { + "epoch": 2.8157894736842106, + "grad_norm": 0.06731904434882949, + "learning_rate": 7.467255749551432e-05, + "loss": 0.0023, + "step": 4280 + }, + { + "epoch": 2.8223684210526314, + "grad_norm": 0.06865526567597074, + "learning_rate": 7.452975193601723e-05, + "loss": 0.0023, + "step": 4290 + }, + { + "epoch": 2.8289473684210527, + "grad_norm": 0.08582709819548066, + "learning_rate": 7.438668236417346e-05, + "loss": 0.0024, + "step": 4300 + }, + { + "epoch": 2.8355263157894735, + "grad_norm": 0.0739094089893243, + "learning_rate": 7.424335031983278e-05, + "loss": 0.0023, + "step": 4310 + }, + { + "epoch": 2.8421052631578947, + "grad_norm": 0.05891595007709077, + "learning_rate": 7.409975734566998e-05, + "loss": 0.0022, + "step": 4320 + }, + { + "epoch": 2.848684210526316, + "grad_norm": 0.0721809677663778, + "learning_rate": 7.39559049871682e-05, + "loss": 0.0023, + "step": 4330 + }, + { + "epoch": 2.8552631578947367, + "grad_norm": 0.055924960105078504, + "learning_rate": 7.381179479260233e-05, + "loss": 0.0023, + "step": 4340 + }, + { + "epoch": 2.861842105263158, + "grad_norm": 0.07924382043010371, + "learning_rate": 7.36674283130223e-05, + "loss": 0.0022, + "step": 4350 + }, + { + "epoch": 2.8684210526315788, + "grad_norm": 0.07736302121317264, + "learning_rate": 7.352280710223645e-05, + "loss": 0.0025, + "step": 4360 + }, + { + "epoch": 2.875, + "grad_norm": 0.07607651273281507, + "learning_rate": 7.337793271679478e-05, + "loss": 0.0024, + "step": 4370 + }, + { + "epoch": 2.8815789473684212, + "grad_norm": 0.06346291574478992, + "learning_rate": 7.323280671597219e-05, + "loss": 0.0023, + "step": 4380 + }, + { + "epoch": 2.888157894736842, + "grad_norm": 0.061302508622908966, + "learning_rate": 7.308743066175172e-05, + "loss": 0.0022, + "step": 4390 + }, + { + "epoch": 2.8947368421052633, + "grad_norm": 0.0609505149341482, + "learning_rate": 7.294180611880766e-05, + "loss": 0.0022, + "step": 4400 + }, + { + "epoch": 2.901315789473684, + "grad_norm": 0.06332339525555587, + "learning_rate": 7.279593465448884e-05, + "loss": 0.0023, + "step": 4410 + }, + { + "epoch": 2.9078947368421053, + "grad_norm": 0.06389140242733114, + "learning_rate": 7.264981783880166e-05, + "loss": 0.0022, + "step": 4420 + }, + { + "epoch": 2.9144736842105265, + "grad_norm": 0.05995795028450876, + "learning_rate": 7.250345724439323e-05, + "loss": 0.0022, + "step": 4430 + }, + { + "epoch": 2.9210526315789473, + "grad_norm": 0.07002238446710278, + "learning_rate": 7.235685444653444e-05, + "loss": 0.0022, + "step": 4440 + }, + { + "epoch": 2.9276315789473686, + "grad_norm": 0.06460285769189916, + "learning_rate": 7.221001102310298e-05, + "loss": 0.0023, + "step": 4450 + }, + { + "epoch": 2.9342105263157894, + "grad_norm": 0.06995475688793039, + "learning_rate": 7.206292855456638e-05, + "loss": 0.002, + "step": 4460 + }, + { + "epoch": 2.9407894736842106, + "grad_norm": 0.07124928971946126, + "learning_rate": 7.191560862396505e-05, + "loss": 0.0023, + "step": 4470 + }, + { + "epoch": 2.9473684210526314, + "grad_norm": 0.07629357364809666, + "learning_rate": 7.176805281689514e-05, + "loss": 0.0022, + "step": 4480 + }, + { + "epoch": 2.9539473684210527, + "grad_norm": 0.08193349418976627, + "learning_rate": 7.162026272149152e-05, + "loss": 0.0021, + "step": 4490 + }, + { + "epoch": 2.9605263157894735, + "grad_norm": 0.07303326416003306, + "learning_rate": 7.147223992841074e-05, + "loss": 0.0022, + "step": 4500 + }, + { + "epoch": 2.9671052631578947, + "grad_norm": 0.08529692109248832, + "learning_rate": 7.132398603081382e-05, + "loss": 0.0022, + "step": 4510 + }, + { + "epoch": 2.973684210526316, + "grad_norm": 0.061074529913310756, + "learning_rate": 7.117550262434917e-05, + "loss": 0.0021, + "step": 4520 + }, + { + "epoch": 2.9802631578947367, + "grad_norm": 0.0652049352287244, + "learning_rate": 7.102679130713537e-05, + "loss": 0.0021, + "step": 4530 + }, + { + "epoch": 2.986842105263158, + "grad_norm": 0.05954432016331547, + "learning_rate": 7.087785367974402e-05, + "loss": 0.002, + "step": 4540 + }, + { + "epoch": 2.9934210526315788, + "grad_norm": 0.061906863190638615, + "learning_rate": 7.072869134518243e-05, + "loss": 0.002, + "step": 4550 + }, + { + "epoch": 3.0, + "grad_norm": 0.06269104180471789, + "learning_rate": 7.057930590887648e-05, + "loss": 0.0021, + "step": 4560 + }, + { + "epoch": 3.0065789473684212, + "grad_norm": 0.044136760952164476, + "learning_rate": 7.042969897865325e-05, + "loss": 0.0013, + "step": 4570 + }, + { + "epoch": 3.013157894736842, + "grad_norm": 0.04755011035898616, + "learning_rate": 7.027987216472377e-05, + "loss": 0.0014, + "step": 4580 + }, + { + "epoch": 3.0197368421052633, + "grad_norm": 0.04504389965591228, + "learning_rate": 7.012982707966561e-05, + "loss": 0.0012, + "step": 4590 + }, + { + "epoch": 3.026315789473684, + "grad_norm": 0.039326385477169795, + "learning_rate": 6.997956533840565e-05, + "loss": 0.0012, + "step": 4600 + }, + { + "epoch": 3.0328947368421053, + "grad_norm": 0.050717328095234285, + "learning_rate": 6.98290885582026e-05, + "loss": 0.001, + "step": 4610 + }, + { + "epoch": 3.039473684210526, + "grad_norm": 0.07042333565934958, + "learning_rate": 6.967839835862962e-05, + "loss": 0.0011, + "step": 4620 + }, + { + "epoch": 3.0460526315789473, + "grad_norm": 0.041419088284560174, + "learning_rate": 6.952749636155686e-05, + "loss": 0.001, + "step": 4630 + }, + { + "epoch": 3.0526315789473686, + "grad_norm": 0.047502535032076625, + "learning_rate": 6.93763841911341e-05, + "loss": 0.001, + "step": 4640 + }, + { + "epoch": 3.0592105263157894, + "grad_norm": 0.04278077403078802, + "learning_rate": 6.922506347377317e-05, + "loss": 0.001, + "step": 4650 + }, + { + "epoch": 3.0657894736842106, + "grad_norm": 0.038900292014083175, + "learning_rate": 6.907353583813045e-05, + "loss": 0.0009, + "step": 4660 + }, + { + "epoch": 3.0723684210526314, + "grad_norm": 0.055076300918611376, + "learning_rate": 6.892180291508944e-05, + "loss": 0.001, + "step": 4670 + }, + { + "epoch": 3.0789473684210527, + "grad_norm": 0.04627844007982353, + "learning_rate": 6.876986633774304e-05, + "loss": 0.001, + "step": 4680 + }, + { + "epoch": 3.085526315789474, + "grad_norm": 0.04320535511746515, + "learning_rate": 6.861772774137618e-05, + "loss": 0.0011, + "step": 4690 + }, + { + "epoch": 3.0921052631578947, + "grad_norm": 0.05504211999563014, + "learning_rate": 6.846538876344802e-05, + "loss": 0.001, + "step": 4700 + }, + { + "epoch": 3.098684210526316, + "grad_norm": 0.05448192261747016, + "learning_rate": 6.831285104357447e-05, + "loss": 0.001, + "step": 4710 + }, + { + "epoch": 3.1052631578947367, + "grad_norm": 0.04552769740470275, + "learning_rate": 6.816011622351046e-05, + "loss": 0.001, + "step": 4720 + }, + { + "epoch": 3.111842105263158, + "grad_norm": 0.04380094534401855, + "learning_rate": 6.80071859471323e-05, + "loss": 0.0011, + "step": 4730 + }, + { + "epoch": 3.1184210526315788, + "grad_norm": 0.05316132876867928, + "learning_rate": 6.785406186041999e-05, + "loss": 0.001, + "step": 4740 + }, + { + "epoch": 3.125, + "grad_norm": 0.06748335057222496, + "learning_rate": 6.770074561143953e-05, + "loss": 0.001, + "step": 4750 + }, + { + "epoch": 3.1315789473684212, + "grad_norm": 0.05077013593653331, + "learning_rate": 6.754723885032509e-05, + "loss": 0.001, + "step": 4760 + }, + { + "epoch": 3.138157894736842, + "grad_norm": 0.04785887103202659, + "learning_rate": 6.739354322926136e-05, + "loss": 0.001, + "step": 4770 + }, + { + "epoch": 3.1447368421052633, + "grad_norm": 0.04683825413816567, + "learning_rate": 6.723966040246572e-05, + "loss": 0.001, + "step": 4780 + }, + { + "epoch": 3.151315789473684, + "grad_norm": 0.04117391629226033, + "learning_rate": 6.708559202617038e-05, + "loss": 0.001, + "step": 4790 + }, + { + "epoch": 3.1578947368421053, + "grad_norm": 0.04553304662726201, + "learning_rate": 6.693133975860469e-05, + "loss": 0.0009, + "step": 4800 + }, + { + "epoch": 3.1644736842105265, + "grad_norm": 0.049513375189675204, + "learning_rate": 6.677690525997714e-05, + "loss": 0.001, + "step": 4810 + }, + { + "epoch": 3.1710526315789473, + "grad_norm": 0.04391216195471761, + "learning_rate": 6.66222901924576e-05, + "loss": 0.001, + "step": 4820 + }, + { + "epoch": 3.1776315789473686, + "grad_norm": 0.044387416555585354, + "learning_rate": 6.64674962201594e-05, + "loss": 0.001, + "step": 4830 + }, + { + "epoch": 3.1842105263157894, + "grad_norm": 0.04438285777547525, + "learning_rate": 6.631252500912138e-05, + "loss": 0.0011, + "step": 4840 + }, + { + "epoch": 3.1907894736842106, + "grad_norm": 0.047101474036633194, + "learning_rate": 6.615737822729e-05, + "loss": 0.001, + "step": 4850 + }, + { + "epoch": 3.1973684210526314, + "grad_norm": 0.040121280292295405, + "learning_rate": 6.60020575445014e-05, + "loss": 0.001, + "step": 4860 + }, + { + "epoch": 3.2039473684210527, + "grad_norm": 0.044535989576363004, + "learning_rate": 6.584656463246342e-05, + "loss": 0.001, + "step": 4870 + }, + { + "epoch": 3.2105263157894735, + "grad_norm": 0.07776632409312663, + "learning_rate": 6.569090116473751e-05, + "loss": 0.0011, + "step": 4880 + }, + { + "epoch": 3.2171052631578947, + "grad_norm": 0.04682875202534155, + "learning_rate": 6.553506881672088e-05, + "loss": 0.0011, + "step": 4890 + }, + { + "epoch": 3.223684210526316, + "grad_norm": 0.05497866325358715, + "learning_rate": 6.537906926562837e-05, + "loss": 0.001, + "step": 4900 + }, + { + "epoch": 3.2302631578947367, + "grad_norm": 0.06474765645613281, + "learning_rate": 6.522290419047444e-05, + "loss": 0.001, + "step": 4910 + }, + { + "epoch": 3.236842105263158, + "grad_norm": 0.04305292992711723, + "learning_rate": 6.506657527205501e-05, + "loss": 0.001, + "step": 4920 + }, + { + "epoch": 3.2434210526315788, + "grad_norm": 0.045387566372484936, + "learning_rate": 6.49100841929295e-05, + "loss": 0.001, + "step": 4930 + }, + { + "epoch": 3.25, + "grad_norm": 0.1120983923723266, + "learning_rate": 6.475343263740261e-05, + "loss": 0.0035, + "step": 4940 + }, + { + "epoch": 3.2565789473684212, + "grad_norm": 0.07941930777156028, + "learning_rate": 6.459662229150629e-05, + "loss": 0.0024, + "step": 4950 + }, + { + "epoch": 3.263157894736842, + "grad_norm": 0.07966562157041894, + "learning_rate": 6.44396548429815e-05, + "loss": 0.0016, + "step": 4960 + }, + { + "epoch": 3.2697368421052633, + "grad_norm": 0.07346284814536075, + "learning_rate": 6.428253198126009e-05, + "loss": 0.0015, + "step": 4970 + }, + { + "epoch": 3.276315789473684, + "grad_norm": 0.12406992455137089, + "learning_rate": 6.412525539744664e-05, + "loss": 0.0028, + "step": 4980 + }, + { + "epoch": 3.2828947368421053, + "grad_norm": 0.12753701081458718, + "learning_rate": 6.396782678430021e-05, + "loss": 0.002, + "step": 4990 + }, + { + "epoch": 3.2894736842105265, + "grad_norm": 0.06701592063617125, + "learning_rate": 6.381024783621614e-05, + "loss": 0.0016, + "step": 5000 + }, + { + "epoch": 3.2960526315789473, + "grad_norm": 0.05171422173018302, + "learning_rate": 6.365252024920784e-05, + "loss": 0.0013, + "step": 5010 + }, + { + "epoch": 3.3026315789473686, + "grad_norm": 0.05818564557929872, + "learning_rate": 6.349464572088849e-05, + "loss": 0.0013, + "step": 5020 + }, + { + "epoch": 3.3092105263157894, + "grad_norm": 0.073006934280184, + "learning_rate": 6.333662595045281e-05, + "loss": 0.0014, + "step": 5030 + }, + { + "epoch": 3.3157894736842106, + "grad_norm": 0.054778073601371255, + "learning_rate": 6.317846263865872e-05, + "loss": 0.0014, + "step": 5040 + }, + { + "epoch": 3.3223684210526314, + "grad_norm": 0.05659243950533705, + "learning_rate": 6.302015748780909e-05, + "loss": 0.0012, + "step": 5050 + }, + { + "epoch": 3.3289473684210527, + "grad_norm": 0.04770376014919815, + "learning_rate": 6.286171220173343e-05, + "loss": 0.0012, + "step": 5060 + }, + { + "epoch": 3.3355263157894735, + "grad_norm": 0.06283432984369176, + "learning_rate": 6.270312848576944e-05, + "loss": 0.0013, + "step": 5070 + }, + { + "epoch": 3.3421052631578947, + "grad_norm": 0.0570672191700207, + "learning_rate": 6.25444080467448e-05, + "loss": 0.0012, + "step": 5080 + }, + { + "epoch": 3.348684210526316, + "grad_norm": 0.05506455230381448, + "learning_rate": 6.238555259295872e-05, + "loss": 0.0012, + "step": 5090 + }, + { + "epoch": 3.3552631578947367, + "grad_norm": 0.04071882369132935, + "learning_rate": 6.222656383416352e-05, + "loss": 0.0011, + "step": 5100 + }, + { + "epoch": 3.361842105263158, + "grad_norm": 0.04561858668173625, + "learning_rate": 6.206744348154635e-05, + "loss": 0.0012, + "step": 5110 + }, + { + "epoch": 3.3684210526315788, + "grad_norm": 0.04919438581324567, + "learning_rate": 6.190819324771062e-05, + "loss": 0.0011, + "step": 5120 + }, + { + "epoch": 3.375, + "grad_norm": 0.05418076160498796, + "learning_rate": 6.174881484665768e-05, + "loss": 0.0011, + "step": 5130 + }, + { + "epoch": 3.3815789473684212, + "grad_norm": 0.056960371113187074, + "learning_rate": 6.158930999376838e-05, + "loss": 0.0011, + "step": 5140 + }, + { + "epoch": 3.388157894736842, + "grad_norm": 0.04255944372616425, + "learning_rate": 6.142968040578449e-05, + "loss": 0.0011, + "step": 5150 + }, + { + "epoch": 3.3947368421052633, + "grad_norm": 0.048340947040949475, + "learning_rate": 6.126992780079031e-05, + "loss": 0.0011, + "step": 5160 + }, + { + "epoch": 3.401315789473684, + "grad_norm": 0.0442107482386399, + "learning_rate": 6.111005389819422e-05, + "loss": 0.0011, + "step": 5170 + }, + { + "epoch": 3.4078947368421053, + "grad_norm": 0.04136846778037803, + "learning_rate": 6.0950060418710065e-05, + "loss": 0.0011, + "step": 5180 + }, + { + "epoch": 3.4144736842105265, + "grad_norm": 0.04323795038841515, + "learning_rate": 6.078994908433872e-05, + "loss": 0.0011, + "step": 5190 + }, + { + "epoch": 3.4210526315789473, + "grad_norm": 0.05416639442819582, + "learning_rate": 6.062972161834948e-05, + "loss": 0.001, + "step": 5200 + }, + { + "epoch": 3.4276315789473686, + "grad_norm": 0.04010090748408715, + "learning_rate": 6.0469379745261624e-05, + "loss": 0.0011, + "step": 5210 + }, + { + "epoch": 3.4342105263157894, + "grad_norm": 0.05141200227401958, + "learning_rate": 6.0308925190825725e-05, + "loss": 0.0011, + "step": 5220 + }, + { + "epoch": 3.4407894736842106, + "grad_norm": 0.05960335396630123, + "learning_rate": 6.014835968200521e-05, + "loss": 0.0011, + "step": 5230 + }, + { + "epoch": 3.4473684210526314, + "grad_norm": 0.03996319019239445, + "learning_rate": 5.9987684946957625e-05, + "loss": 0.0009, + "step": 5240 + }, + { + "epoch": 3.4539473684210527, + "grad_norm": 0.03754917546922171, + "learning_rate": 5.9826902715016154e-05, + "loss": 0.001, + "step": 5250 + }, + { + "epoch": 3.4605263157894735, + "grad_norm": 0.044709261031248435, + "learning_rate": 5.966601471667096e-05, + "loss": 0.001, + "step": 5260 + }, + { + "epoch": 3.4671052631578947, + "grad_norm": 0.05394706576681355, + "learning_rate": 5.950502268355057e-05, + "loss": 0.001, + "step": 5270 + }, + { + "epoch": 3.473684210526316, + "grad_norm": 0.04640630644633933, + "learning_rate": 5.934392834840321e-05, + "loss": 0.001, + "step": 5280 + }, + { + "epoch": 3.4802631578947367, + "grad_norm": 0.03996275338024238, + "learning_rate": 5.91827334450782e-05, + "loss": 0.001, + "step": 5290 + }, + { + "epoch": 3.486842105263158, + "grad_norm": 0.04513935016051976, + "learning_rate": 5.902143970850725e-05, + "loss": 0.0009, + "step": 5300 + }, + { + "epoch": 3.4934210526315788, + "grad_norm": 0.04727412105314611, + "learning_rate": 5.8860048874685836e-05, + "loss": 0.001, + "step": 5310 + }, + { + "epoch": 3.5, + "grad_norm": 0.041740001551287126, + "learning_rate": 5.8698562680654436e-05, + "loss": 0.0009, + "step": 5320 + }, + { + "epoch": 3.5065789473684212, + "grad_norm": 0.04711270749810497, + "learning_rate": 5.8536982864479925e-05, + "loss": 0.001, + "step": 5330 + }, + { + "epoch": 3.513157894736842, + "grad_norm": 0.04576375416018754, + "learning_rate": 5.837531116523682e-05, + "loss": 0.0009, + "step": 5340 + }, + { + "epoch": 3.5197368421052633, + "grad_norm": 0.038452015974273716, + "learning_rate": 5.821354932298858e-05, + "loss": 0.0009, + "step": 5350 + }, + { + "epoch": 3.526315789473684, + "grad_norm": 0.03906108461777927, + "learning_rate": 5.805169907876884e-05, + "loss": 0.0009, + "step": 5360 + }, + { + "epoch": 3.5328947368421053, + "grad_norm": 0.05168966022643398, + "learning_rate": 5.788976217456274e-05, + "loss": 0.001, + "step": 5370 + }, + { + "epoch": 3.5394736842105265, + "grad_norm": 0.0445000606377978, + "learning_rate": 5.772774035328809e-05, + "loss": 0.0009, + "step": 5380 + }, + { + "epoch": 3.5460526315789473, + "grad_norm": 0.03328912347538213, + "learning_rate": 5.756563535877668e-05, + "loss": 0.001, + "step": 5390 + }, + { + "epoch": 3.5526315789473686, + "grad_norm": 0.044808425570984904, + "learning_rate": 5.74034489357555e-05, + "loss": 0.001, + "step": 5400 + }, + { + "epoch": 3.5592105263157894, + "grad_norm": 0.04045168861133601, + "learning_rate": 5.7241182829827934e-05, + "loss": 0.0009, + "step": 5410 + }, + { + "epoch": 3.5657894736842106, + "grad_norm": 0.04027670154872098, + "learning_rate": 5.707883878745499e-05, + "loss": 0.0009, + "step": 5420 + }, + { + "epoch": 3.5723684210526314, + "grad_norm": 0.04494753709166914, + "learning_rate": 5.691641855593655e-05, + "loss": 0.0009, + "step": 5430 + }, + { + "epoch": 3.5789473684210527, + "grad_norm": 0.05017171190789273, + "learning_rate": 5.67539238833924e-05, + "loss": 0.0011, + "step": 5440 + }, + { + "epoch": 3.5855263157894735, + "grad_norm": 0.048350190780472826, + "learning_rate": 5.659135651874365e-05, + "loss": 0.001, + "step": 5450 + }, + { + "epoch": 3.5921052631578947, + "grad_norm": 0.05725707071908558, + "learning_rate": 5.642871821169374e-05, + "loss": 0.0011, + "step": 5460 + }, + { + "epoch": 3.598684210526316, + "grad_norm": 0.052194793467222785, + "learning_rate": 5.626601071270966e-05, + "loss": 0.001, + "step": 5470 + }, + { + "epoch": 3.6052631578947367, + "grad_norm": 0.05323296981030075, + "learning_rate": 5.6103235773003105e-05, + "loss": 0.001, + "step": 5480 + }, + { + "epoch": 3.611842105263158, + "grad_norm": 0.051406214461841945, + "learning_rate": 5.594039514451164e-05, + "loss": 0.001, + "step": 5490 + }, + { + "epoch": 3.6184210526315788, + "grad_norm": 0.050452329568075015, + "learning_rate": 5.577749057987982e-05, + "loss": 0.0009, + "step": 5500 + }, + { + "epoch": 3.625, + "grad_norm": 0.048405495025027306, + "learning_rate": 5.561452383244039e-05, + "loss": 0.0009, + "step": 5510 + }, + { + "epoch": 3.6315789473684212, + "grad_norm": 0.047492109120741345, + "learning_rate": 5.545149665619528e-05, + "loss": 0.001, + "step": 5520 + }, + { + "epoch": 3.638157894736842, + "grad_norm": 0.04996527735381986, + "learning_rate": 5.5288410805796895e-05, + "loss": 0.0009, + "step": 5530 + }, + { + "epoch": 3.6447368421052633, + "grad_norm": 0.044505318984903314, + "learning_rate": 5.51252680365291e-05, + "loss": 0.001, + "step": 5540 + }, + { + "epoch": 3.651315789473684, + "grad_norm": 0.038117214544184895, + "learning_rate": 5.49620701042884e-05, + "loss": 0.0009, + "step": 5550 + }, + { + "epoch": 3.6578947368421053, + "grad_norm": 0.047586657041494784, + "learning_rate": 5.479881876556498e-05, + "loss": 0.0009, + "step": 5560 + }, + { + "epoch": 3.6644736842105265, + "grad_norm": 0.05115085388996902, + "learning_rate": 5.463551577742385e-05, + "loss": 0.001, + "step": 5570 + }, + { + "epoch": 3.6710526315789473, + "grad_norm": 0.049097466404171355, + "learning_rate": 5.4472162897485956e-05, + "loss": 0.001, + "step": 5580 + }, + { + "epoch": 3.6776315789473686, + "grad_norm": 0.04286708176427824, + "learning_rate": 5.430876188390921e-05, + "loss": 0.001, + "step": 5590 + }, + { + "epoch": 3.6842105263157894, + "grad_norm": 0.046545423756860486, + "learning_rate": 5.4145314495369536e-05, + "loss": 0.0009, + "step": 5600 + }, + { + "epoch": 3.6907894736842106, + "grad_norm": 0.05298747670683033, + "learning_rate": 5.3981822491042034e-05, + "loss": 0.001, + "step": 5610 + }, + { + "epoch": 3.6973684210526314, + "grad_norm": 0.043286652522166486, + "learning_rate": 5.381828763058201e-05, + "loss": 0.001, + "step": 5620 + }, + { + "epoch": 3.7039473684210527, + "grad_norm": 0.05004897178098679, + "learning_rate": 5.3654711674106004e-05, + "loss": 0.0009, + "step": 5630 + }, + { + "epoch": 3.7105263157894735, + "grad_norm": 0.049263781024319955, + "learning_rate": 5.3491096382172866e-05, + "loss": 0.0011, + "step": 5640 + }, + { + "epoch": 3.7171052631578947, + "grad_norm": 0.045534509263611286, + "learning_rate": 5.332744351576483e-05, + "loss": 0.001, + "step": 5650 + }, + { + "epoch": 3.723684210526316, + "grad_norm": 0.04425416625940554, + "learning_rate": 5.316375483626854e-05, + "loss": 0.001, + "step": 5660 + }, + { + "epoch": 3.7302631578947367, + "grad_norm": 0.041689090987499504, + "learning_rate": 5.300003210545608e-05, + "loss": 0.0009, + "step": 5670 + }, + { + "epoch": 3.736842105263158, + "grad_norm": 0.04243716615026122, + "learning_rate": 5.2836277085466036e-05, + "loss": 0.0009, + "step": 5680 + }, + { + "epoch": 3.7434210526315788, + "grad_norm": 0.04150831742868238, + "learning_rate": 5.267249153878452e-05, + "loss": 0.001, + "step": 5690 + }, + { + "epoch": 3.75, + "grad_norm": 0.045313010765497444, + "learning_rate": 5.250867722822621e-05, + "loss": 0.0009, + "step": 5700 + }, + { + "epoch": 3.7565789473684212, + "grad_norm": 0.04405757046560163, + "learning_rate": 5.234483591691536e-05, + "loss": 0.001, + "step": 5710 + }, + { + "epoch": 3.763157894736842, + "grad_norm": 0.05258313821024352, + "learning_rate": 5.218096936826681e-05, + "loss": 0.0009, + "step": 5720 + }, + { + "epoch": 3.7697368421052633, + "grad_norm": 0.04056893648743933, + "learning_rate": 5.201707934596706e-05, + "loss": 0.001, + "step": 5730 + }, + { + "epoch": 3.776315789473684, + "grad_norm": 0.06398291065162262, + "learning_rate": 5.185316761395523e-05, + "loss": 0.001, + "step": 5740 + }, + { + "epoch": 3.7828947368421053, + "grad_norm": 0.042027727690221034, + "learning_rate": 5.1689235936404155e-05, + "loss": 0.001, + "step": 5750 + }, + { + "epoch": 3.7894736842105265, + "grad_norm": 0.056335909713406476, + "learning_rate": 5.152528607770126e-05, + "loss": 0.0009, + "step": 5760 + }, + { + "epoch": 3.7960526315789473, + "grad_norm": 0.033910785390483186, + "learning_rate": 5.136131980242972e-05, + "loss": 0.0009, + "step": 5770 + }, + { + "epoch": 3.8026315789473686, + "grad_norm": 0.05262608327262604, + "learning_rate": 5.1197338875349356e-05, + "loss": 0.0011, + "step": 5780 + }, + { + "epoch": 3.8092105263157894, + "grad_norm": 0.07069273119784056, + "learning_rate": 5.103334506137772e-05, + "loss": 0.0011, + "step": 5790 + }, + { + "epoch": 3.8157894736842106, + "grad_norm": 0.040558779266260746, + "learning_rate": 5.086934012557104e-05, + "loss": 0.0011, + "step": 5800 + }, + { + "epoch": 3.8223684210526314, + "grad_norm": 0.0663477079873889, + "learning_rate": 5.0705325833105265e-05, + "loss": 0.0011, + "step": 5810 + }, + { + "epoch": 3.8289473684210527, + "grad_norm": 0.051595354465206265, + "learning_rate": 5.0541303949257035e-05, + "loss": 0.001, + "step": 5820 + }, + { + "epoch": 3.8355263157894735, + "grad_norm": 0.05957291484709742, + "learning_rate": 5.037727623938472e-05, + "loss": 0.001, + "step": 5830 + }, + { + "epoch": 3.8421052631578947, + "grad_norm": 0.04110902409285192, + "learning_rate": 5.021324446890936e-05, + "loss": 0.0011, + "step": 5840 + }, + { + "epoch": 3.848684210526316, + "grad_norm": 0.05756925936636771, + "learning_rate": 5.004921040329571e-05, + "loss": 0.001, + "step": 5850 + }, + { + "epoch": 3.8552631578947367, + "grad_norm": 0.03971678943527698, + "learning_rate": 4.988517580803324e-05, + "loss": 0.001, + "step": 5860 + }, + { + "epoch": 3.861842105263158, + "grad_norm": 0.04477846099186899, + "learning_rate": 4.972114244861713e-05, + "loss": 0.0009, + "step": 5870 + }, + { + "epoch": 3.8684210526315788, + "grad_norm": 0.05152382595601969, + "learning_rate": 4.955711209052923e-05, + "loss": 0.001, + "step": 5880 + }, + { + "epoch": 3.875, + "grad_norm": 0.04497320087802862, + "learning_rate": 4.939308649921912e-05, + "loss": 0.001, + "step": 5890 + }, + { + "epoch": 3.8815789473684212, + "grad_norm": 0.05059532785773457, + "learning_rate": 4.9229067440085016e-05, + "loss": 0.0008, + "step": 5900 + }, + { + "epoch": 3.888157894736842, + "grad_norm": 0.04880705607784876, + "learning_rate": 4.9065056678454904e-05, + "loss": 0.0009, + "step": 5910 + }, + { + "epoch": 3.8947368421052633, + "grad_norm": 0.05875285184712623, + "learning_rate": 4.89010559795674e-05, + "loss": 0.0009, + "step": 5920 + }, + { + "epoch": 3.901315789473684, + "grad_norm": 0.06297099680443237, + "learning_rate": 4.8737067108552864e-05, + "loss": 0.001, + "step": 5930 + }, + { + "epoch": 3.9078947368421053, + "grad_norm": 0.05100327809910756, + "learning_rate": 4.857309183041433e-05, + "loss": 0.001, + "step": 5940 + }, + { + "epoch": 3.9144736842105265, + "grad_norm": 0.042218196077337525, + "learning_rate": 4.8409131910008534e-05, + "loss": 0.001, + "step": 5950 + }, + { + "epoch": 3.9210526315789473, + "grad_norm": 0.04706136014405962, + "learning_rate": 4.824518911202691e-05, + "loss": 0.0009, + "step": 5960 + }, + { + "epoch": 3.9276315789473686, + "grad_norm": 0.05066604983210616, + "learning_rate": 4.808126520097665e-05, + "loss": 0.001, + "step": 5970 + }, + { + "epoch": 3.9342105263157894, + "grad_norm": 0.05714030630918534, + "learning_rate": 4.791736194116158e-05, + "loss": 0.001, + "step": 5980 + }, + { + "epoch": 3.9407894736842106, + "grad_norm": 0.04603109156621843, + "learning_rate": 4.775348109666333e-05, + "loss": 0.001, + "step": 5990 + }, + { + "epoch": 3.9473684210526314, + "grad_norm": 0.06952657713346984, + "learning_rate": 4.758962443132226e-05, + "loss": 0.0009, + "step": 6000 + }, + { + "epoch": 3.9539473684210527, + "grad_norm": 0.04466220069363302, + "learning_rate": 4.742579370871849e-05, + "loss": 0.0009, + "step": 6010 + }, + { + "epoch": 3.9605263157894735, + "grad_norm": 0.045787404434420205, + "learning_rate": 4.726199069215289e-05, + "loss": 0.001, + "step": 6020 + }, + { + "epoch": 3.9671052631578947, + "grad_norm": 0.041204532696627126, + "learning_rate": 4.709821714462818e-05, + "loss": 0.0008, + "step": 6030 + }, + { + "epoch": 3.973684210526316, + "grad_norm": 0.03600439485923749, + "learning_rate": 4.6934474828829875e-05, + "loss": 0.0008, + "step": 6040 + }, + { + "epoch": 3.9802631578947367, + "grad_norm": 0.04087884556954519, + "learning_rate": 4.677076550710737e-05, + "loss": 0.0008, + "step": 6050 + }, + { + "epoch": 3.986842105263158, + "grad_norm": 0.058167832413074244, + "learning_rate": 4.660709094145489e-05, + "loss": 0.0009, + "step": 6060 + }, + { + "epoch": 3.9934210526315788, + "grad_norm": 0.038488978904163695, + "learning_rate": 4.644345289349264e-05, + "loss": 0.0009, + "step": 6070 + }, + { + "epoch": 4.0, + "grad_norm": 0.04305944112119989, + "learning_rate": 4.627985312444777e-05, + "loss": 0.0008, + "step": 6080 + }, + { + "epoch": 4.006578947368421, + "grad_norm": 0.048786900069551595, + "learning_rate": 4.611629339513543e-05, + "loss": 0.0006, + "step": 6090 + }, + { + "epoch": 4.0131578947368425, + "grad_norm": 0.035076766391848965, + "learning_rate": 4.595277546593984e-05, + "loss": 0.0006, + "step": 6100 + }, + { + "epoch": 4.019736842105263, + "grad_norm": 0.023256725564357015, + "learning_rate": 4.57893010967953e-05, + "loss": 0.0005, + "step": 6110 + }, + { + "epoch": 4.026315789473684, + "grad_norm": 0.02457501793226784, + "learning_rate": 4.5625872047167304e-05, + "loss": 0.0005, + "step": 6120 + }, + { + "epoch": 4.032894736842105, + "grad_norm": 0.026288211965478626, + "learning_rate": 4.546249007603359e-05, + "loss": 0.0004, + "step": 6130 + }, + { + "epoch": 4.0394736842105265, + "grad_norm": 0.023070988474870147, + "learning_rate": 4.5299156941865104e-05, + "loss": 0.0004, + "step": 6140 + }, + { + "epoch": 4.046052631578948, + "grad_norm": 0.03549730164036071, + "learning_rate": 4.513587440260726e-05, + "loss": 0.0004, + "step": 6150 + }, + { + "epoch": 4.052631578947368, + "grad_norm": 0.027024080097335302, + "learning_rate": 4.497264421566089e-05, + "loss": 0.0004, + "step": 6160 + }, + { + "epoch": 4.059210526315789, + "grad_norm": 0.027407217963954806, + "learning_rate": 4.480946813786334e-05, + "loss": 0.0004, + "step": 6170 + }, + { + "epoch": 4.065789473684211, + "grad_norm": 0.02948654063181369, + "learning_rate": 4.464634792546961e-05, + "loss": 0.0003, + "step": 6180 + }, + { + "epoch": 4.072368421052632, + "grad_norm": 0.02906096035731952, + "learning_rate": 4.4483285334133404e-05, + "loss": 0.0004, + "step": 6190 + }, + { + "epoch": 4.078947368421052, + "grad_norm": 0.02275711987090776, + "learning_rate": 4.4320282118888285e-05, + "loss": 0.0003, + "step": 6200 + }, + { + "epoch": 4.0855263157894735, + "grad_norm": 0.02556423402594296, + "learning_rate": 4.415734003412873e-05, + "loss": 0.0003, + "step": 6210 + }, + { + "epoch": 4.092105263157895, + "grad_norm": 0.0271151070508472, + "learning_rate": 4.3994460833591243e-05, + "loss": 0.0004, + "step": 6220 + }, + { + "epoch": 4.098684210526316, + "grad_norm": 0.02036373809346978, + "learning_rate": 4.383164627033556e-05, + "loss": 0.0003, + "step": 6230 + }, + { + "epoch": 4.105263157894737, + "grad_norm": 0.027292522196807834, + "learning_rate": 4.366889809672571e-05, + "loss": 0.0003, + "step": 6240 + }, + { + "epoch": 4.1118421052631575, + "grad_norm": 0.022674806731559142, + "learning_rate": 4.3506218064411166e-05, + "loss": 0.0004, + "step": 6250 + }, + { + "epoch": 4.118421052631579, + "grad_norm": 0.017884630922516506, + "learning_rate": 4.3343607924308016e-05, + "loss": 0.0004, + "step": 6260 + }, + { + "epoch": 4.125, + "grad_norm": 0.029365848813510394, + "learning_rate": 4.318106942658009e-05, + "loss": 0.0004, + "step": 6270 + }, + { + "epoch": 4.131578947368421, + "grad_norm": 0.04132370576722313, + "learning_rate": 4.301860432062014e-05, + "loss": 0.0004, + "step": 6280 + }, + { + "epoch": 4.1381578947368425, + "grad_norm": 0.02442250234606405, + "learning_rate": 4.285621435503101e-05, + "loss": 0.0004, + "step": 6290 + }, + { + "epoch": 4.144736842105263, + "grad_norm": 0.03261061440724285, + "learning_rate": 4.26939012776068e-05, + "loss": 0.0004, + "step": 6300 + }, + { + "epoch": 4.151315789473684, + "grad_norm": 0.034532554355897395, + "learning_rate": 4.253166683531407e-05, + "loss": 0.0003, + "step": 6310 + }, + { + "epoch": 4.157894736842105, + "grad_norm": 0.03022539568330769, + "learning_rate": 4.236951277427305e-05, + "loss": 0.0004, + "step": 6320 + }, + { + "epoch": 4.1644736842105265, + "grad_norm": 0.018853994253276635, + "learning_rate": 4.2207440839738816e-05, + "loss": 0.0003, + "step": 6330 + }, + { + "epoch": 4.171052631578948, + "grad_norm": 0.031102303786766313, + "learning_rate": 4.2045452776082525e-05, + "loss": 0.0004, + "step": 6340 + }, + { + "epoch": 4.177631578947368, + "grad_norm": 0.027654360572464114, + "learning_rate": 4.188355032677263e-05, + "loss": 0.0003, + "step": 6350 + }, + { + "epoch": 4.184210526315789, + "grad_norm": 0.02304357686978954, + "learning_rate": 4.172173523435614e-05, + "loss": 0.0004, + "step": 6360 + }, + { + "epoch": 4.190789473684211, + "grad_norm": 0.03054730475669186, + "learning_rate": 4.156000924043985e-05, + "loss": 0.0004, + "step": 6370 + }, + { + "epoch": 4.197368421052632, + "grad_norm": 0.02221856030969314, + "learning_rate": 4.139837408567153e-05, + "loss": 0.0004, + "step": 6380 + }, + { + "epoch": 4.203947368421052, + "grad_norm": 0.02050409446114795, + "learning_rate": 4.1236831509721344e-05, + "loss": 0.0003, + "step": 6390 + }, + { + "epoch": 4.2105263157894735, + "grad_norm": 0.030468813406657563, + "learning_rate": 4.1075383251262956e-05, + "loss": 0.0003, + "step": 6400 + }, + { + "epoch": 4.217105263157895, + "grad_norm": 0.03159159409993661, + "learning_rate": 4.0914031047954967e-05, + "loss": 0.0003, + "step": 6410 + }, + { + "epoch": 4.223684210526316, + "grad_norm": 0.02214285520270982, + "learning_rate": 4.075277663642208e-05, + "loss": 0.0003, + "step": 6420 + }, + { + "epoch": 4.230263157894737, + "grad_norm": 0.03238420436453702, + "learning_rate": 4.059162175223652e-05, + "loss": 0.0004, + "step": 6430 + }, + { + "epoch": 4.2368421052631575, + "grad_norm": 0.021715578291728287, + "learning_rate": 4.043056812989927e-05, + "loss": 0.0004, + "step": 6440 + }, + { + "epoch": 4.243421052631579, + "grad_norm": 0.021468368768439196, + "learning_rate": 4.0269617502821476e-05, + "loss": 0.0004, + "step": 6450 + }, + { + "epoch": 4.25, + "grad_norm": 0.029636103291823517, + "learning_rate": 4.0108771603305687e-05, + "loss": 0.0003, + "step": 6460 + }, + { + "epoch": 4.256578947368421, + "grad_norm": 0.03511721305855742, + "learning_rate": 3.994803216252732e-05, + "loss": 0.0003, + "step": 6470 + }, + { + "epoch": 4.2631578947368425, + "grad_norm": 0.027942214006821824, + "learning_rate": 3.978740091051599e-05, + "loss": 0.0004, + "step": 6480 + }, + { + "epoch": 4.269736842105263, + "grad_norm": 0.027628316183672477, + "learning_rate": 3.962687957613686e-05, + "loss": 0.0004, + "step": 6490 + }, + { + "epoch": 4.276315789473684, + "grad_norm": 0.025411695699700733, + "learning_rate": 3.946646988707208e-05, + "loss": 0.0004, + "step": 6500 + }, + { + "epoch": 4.282894736842105, + "grad_norm": 0.023670983253881096, + "learning_rate": 3.930617356980214e-05, + "loss": 0.0004, + "step": 6510 + }, + { + "epoch": 4.2894736842105265, + "grad_norm": 0.025425487544451265, + "learning_rate": 3.914599234958732e-05, + "loss": 0.0003, + "step": 6520 + }, + { + "epoch": 4.296052631578947, + "grad_norm": 0.035276922020601745, + "learning_rate": 3.8985927950449154e-05, + "loss": 0.0004, + "step": 6530 + }, + { + "epoch": 4.302631578947368, + "grad_norm": 0.023596466787555136, + "learning_rate": 3.8825982095151786e-05, + "loss": 0.0004, + "step": 6540 + }, + { + "epoch": 4.309210526315789, + "grad_norm": 0.027524549712108343, + "learning_rate": 3.866615650518349e-05, + "loss": 0.0003, + "step": 6550 + }, + { + "epoch": 4.315789473684211, + "grad_norm": 0.027992489747583418, + "learning_rate": 3.850645290073816e-05, + "loss": 0.0003, + "step": 6560 + }, + { + "epoch": 4.322368421052632, + "grad_norm": 0.025282029229742646, + "learning_rate": 3.8346873000696745e-05, + "loss": 0.0003, + "step": 6570 + }, + { + "epoch": 4.328947368421053, + "grad_norm": 0.019455703144644388, + "learning_rate": 3.818741852260877e-05, + "loss": 0.0003, + "step": 6580 + }, + { + "epoch": 4.3355263157894735, + "grad_norm": 0.022056024718997747, + "learning_rate": 3.802809118267386e-05, + "loss": 0.0003, + "step": 6590 + }, + { + "epoch": 4.342105263157895, + "grad_norm": 0.028135395943504917, + "learning_rate": 3.7868892695723255e-05, + "loss": 0.0003, + "step": 6600 + }, + { + "epoch": 4.348684210526316, + "grad_norm": 0.027073786441011695, + "learning_rate": 3.770982477520137e-05, + "loss": 0.0003, + "step": 6610 + }, + { + "epoch": 4.355263157894737, + "grad_norm": 0.024579308202388312, + "learning_rate": 3.7550889133147294e-05, + "loss": 0.0003, + "step": 6620 + }, + { + "epoch": 4.3618421052631575, + "grad_norm": 0.022084420409421886, + "learning_rate": 3.739208748017647e-05, + "loss": 0.0003, + "step": 6630 + }, + { + "epoch": 4.368421052631579, + "grad_norm": 0.03212548367708638, + "learning_rate": 3.723342152546219e-05, + "loss": 0.0003, + "step": 6640 + }, + { + "epoch": 4.375, + "grad_norm": 0.029802709230858277, + "learning_rate": 3.707489297671723e-05, + "loss": 0.0004, + "step": 6650 + }, + { + "epoch": 4.381578947368421, + "grad_norm": 0.02774135744578668, + "learning_rate": 3.691650354017551e-05, + "loss": 0.0003, + "step": 6660 + }, + { + "epoch": 4.3881578947368425, + "grad_norm": 0.020585678499088, + "learning_rate": 3.675825492057364e-05, + "loss": 0.0003, + "step": 6670 + }, + { + "epoch": 4.394736842105263, + "grad_norm": 0.030162513963475055, + "learning_rate": 3.6600148821132665e-05, + "loss": 0.0003, + "step": 6680 + }, + { + "epoch": 4.401315789473684, + "grad_norm": 0.024545017831199844, + "learning_rate": 3.644218694353971e-05, + "loss": 0.0003, + "step": 6690 + }, + { + "epoch": 4.407894736842105, + "grad_norm": 0.017424944165336666, + "learning_rate": 3.628437098792957e-05, + "loss": 0.0003, + "step": 6700 + }, + { + "epoch": 4.4144736842105265, + "grad_norm": 0.032084281740095255, + "learning_rate": 3.6126702652866554e-05, + "loss": 0.0004, + "step": 6710 + }, + { + "epoch": 4.421052631578947, + "grad_norm": 0.03561053047585465, + "learning_rate": 3.596918363532614e-05, + "loss": 0.0004, + "step": 6720 + }, + { + "epoch": 4.427631578947368, + "grad_norm": 0.021429271124819484, + "learning_rate": 3.581181563067668e-05, + "loss": 0.0004, + "step": 6730 + }, + { + "epoch": 4.434210526315789, + "grad_norm": 0.027495944265479043, + "learning_rate": 3.5654600332661195e-05, + "loss": 0.0003, + "step": 6740 + }, + { + "epoch": 4.440789473684211, + "grad_norm": 0.03342992427304297, + "learning_rate": 3.5497539433379146e-05, + "loss": 0.0005, + "step": 6750 + }, + { + "epoch": 4.447368421052632, + "grad_norm": 0.09760473451788312, + "learning_rate": 3.53406346232682e-05, + "loss": 0.0009, + "step": 6760 + }, + { + "epoch": 4.453947368421053, + "grad_norm": 0.048069732833877554, + "learning_rate": 3.518388759108606e-05, + "loss": 0.0007, + "step": 6770 + }, + { + "epoch": 4.4605263157894735, + "grad_norm": 0.027125781350855974, + "learning_rate": 3.502730002389221e-05, + "loss": 0.0005, + "step": 6780 + }, + { + "epoch": 4.467105263157895, + "grad_norm": 0.02797243173049197, + "learning_rate": 3.4870873607029887e-05, + "loss": 0.0005, + "step": 6790 + }, + { + "epoch": 4.473684210526316, + "grad_norm": 0.032332463135765675, + "learning_rate": 3.4714610024107874e-05, + "loss": 0.0005, + "step": 6800 + }, + { + "epoch": 4.480263157894737, + "grad_norm": 0.03964874448218963, + "learning_rate": 3.455851095698235e-05, + "loss": 0.0005, + "step": 6810 + }, + { + "epoch": 4.4868421052631575, + "grad_norm": 0.07706367807342288, + "learning_rate": 3.440257808573884e-05, + "loss": 0.0004, + "step": 6820 + }, + { + "epoch": 4.493421052631579, + "grad_norm": 0.0389318932442531, + "learning_rate": 3.4246813088674105e-05, + "loss": 0.0005, + "step": 6830 + }, + { + "epoch": 4.5, + "grad_norm": 0.01947947445378604, + "learning_rate": 3.409121764227809e-05, + "loss": 0.0004, + "step": 6840 + }, + { + "epoch": 4.506578947368421, + "grad_norm": 0.031085781178681444, + "learning_rate": 3.393579342121589e-05, + "loss": 0.0004, + "step": 6850 + }, + { + "epoch": 4.5131578947368425, + "grad_norm": 0.04037587331280131, + "learning_rate": 3.3780542098309654e-05, + "loss": 0.0004, + "step": 6860 + }, + { + "epoch": 4.519736842105263, + "grad_norm": 0.03738208282511506, + "learning_rate": 3.36254653445207e-05, + "loss": 0.0004, + "step": 6870 + }, + { + "epoch": 4.526315789473684, + "grad_norm": 0.02607151315125573, + "learning_rate": 3.347056482893145e-05, + "loss": 0.0004, + "step": 6880 + }, + { + "epoch": 4.532894736842105, + "grad_norm": 0.05038774162308774, + "learning_rate": 3.331584221872748e-05, + "loss": 0.0004, + "step": 6890 + }, + { + "epoch": 4.5394736842105265, + "grad_norm": 0.028765450514947763, + "learning_rate": 3.3161299179179575e-05, + "loss": 0.0003, + "step": 6900 + }, + { + "epoch": 4.546052631578947, + "grad_norm": 0.025925649257290538, + "learning_rate": 3.300693737362581e-05, + "loss": 0.0003, + "step": 6910 + }, + { + "epoch": 4.552631578947368, + "grad_norm": 0.027039604255583594, + "learning_rate": 3.285275846345368e-05, + "loss": 0.0004, + "step": 6920 + }, + { + "epoch": 4.559210526315789, + "grad_norm": 0.023818680904592604, + "learning_rate": 3.269876410808215e-05, + "loss": 0.0003, + "step": 6930 + }, + { + "epoch": 4.565789473684211, + "grad_norm": 0.02545063605331823, + "learning_rate": 3.254495596494383e-05, + "loss": 0.0003, + "step": 6940 + }, + { + "epoch": 4.572368421052632, + "grad_norm": 0.032246784394968434, + "learning_rate": 3.239133568946716e-05, + "loss": 0.0003, + "step": 6950 + }, + { + "epoch": 4.578947368421053, + "grad_norm": 0.0340242701880849, + "learning_rate": 3.2237904935058586e-05, + "loss": 0.0003, + "step": 6960 + }, + { + "epoch": 4.5855263157894735, + "grad_norm": 0.02694969387910795, + "learning_rate": 3.20846653530847e-05, + "loss": 0.0003, + "step": 6970 + }, + { + "epoch": 4.592105263157895, + "grad_norm": 0.019700280388903126, + "learning_rate": 3.193161859285455e-05, + "loss": 0.0003, + "step": 6980 + }, + { + "epoch": 4.598684210526316, + "grad_norm": 0.026431352463759935, + "learning_rate": 3.177876630160186e-05, + "loss": 0.0003, + "step": 6990 + }, + { + "epoch": 4.605263157894737, + "grad_norm": 0.025468470332503778, + "learning_rate": 3.162611012446727e-05, + "loss": 0.0003, + "step": 7000 + }, + { + "epoch": 4.6118421052631575, + "grad_norm": 0.0210363424444781, + "learning_rate": 3.1473651704480686e-05, + "loss": 0.0003, + "step": 7010 + }, + { + "epoch": 4.618421052631579, + "grad_norm": 0.021208692965199404, + "learning_rate": 3.132139268254352e-05, + "loss": 0.0003, + "step": 7020 + }, + { + "epoch": 4.625, + "grad_norm": 0.029317605606678097, + "learning_rate": 3.116933469741111e-05, + "loss": 0.0003, + "step": 7030 + }, + { + "epoch": 4.631578947368421, + "grad_norm": 0.023236339710598755, + "learning_rate": 3.1017479385675034e-05, + "loss": 0.0003, + "step": 7040 + }, + { + "epoch": 4.6381578947368425, + "grad_norm": 0.026552155530906087, + "learning_rate": 3.086582838174551e-05, + "loss": 0.0004, + "step": 7050 + }, + { + "epoch": 4.644736842105263, + "grad_norm": 0.030451983662230033, + "learning_rate": 3.071438331783382e-05, + "loss": 0.0003, + "step": 7060 + }, + { + "epoch": 4.651315789473684, + "grad_norm": 0.028173409808831724, + "learning_rate": 3.0563145823934686e-05, + "loss": 0.0003, + "step": 7070 + }, + { + "epoch": 4.657894736842105, + "grad_norm": 0.026747225279310057, + "learning_rate": 3.041211752780879e-05, + "loss": 0.0003, + "step": 7080 + }, + { + "epoch": 4.6644736842105265, + "grad_norm": 0.025847986739271123, + "learning_rate": 3.0261300054965246e-05, + "loss": 0.0003, + "step": 7090 + }, + { + "epoch": 4.671052631578947, + "grad_norm": 0.025649631102021397, + "learning_rate": 3.0110695028644065e-05, + "loss": 0.0003, + "step": 7100 + }, + { + "epoch": 4.677631578947368, + "grad_norm": 0.0412048258668028, + "learning_rate": 2.9960304069798696e-05, + "loss": 0.0003, + "step": 7110 + }, + { + "epoch": 4.684210526315789, + "grad_norm": 0.027616712338665625, + "learning_rate": 2.9810128797078608e-05, + "loss": 0.0003, + "step": 7120 + }, + { + "epoch": 4.690789473684211, + "grad_norm": 0.027901637947525558, + "learning_rate": 2.966017082681186e-05, + "loss": 0.0003, + "step": 7130 + }, + { + "epoch": 4.697368421052632, + "grad_norm": 0.02473206630993768, + "learning_rate": 2.95104317729877e-05, + "loss": 0.0003, + "step": 7140 + }, + { + "epoch": 4.703947368421053, + "grad_norm": 0.021348965557535116, + "learning_rate": 2.936091324723917e-05, + "loss": 0.0003, + "step": 7150 + }, + { + "epoch": 4.7105263157894735, + "grad_norm": 0.0266877583753156, + "learning_rate": 2.9211616858825798e-05, + "loss": 0.0003, + "step": 7160 + }, + { + "epoch": 4.717105263157895, + "grad_norm": 0.027448984050407993, + "learning_rate": 2.906254421461624e-05, + "loss": 0.0003, + "step": 7170 + }, + { + "epoch": 4.723684210526316, + "grad_norm": 0.024000080095674052, + "learning_rate": 2.8913696919071043e-05, + "loss": 0.0003, + "step": 7180 + }, + { + "epoch": 4.730263157894737, + "grad_norm": 0.02102552080183026, + "learning_rate": 2.8765076574225307e-05, + "loss": 0.0002, + "step": 7190 + }, + { + "epoch": 4.7368421052631575, + "grad_norm": 0.018805690011454896, + "learning_rate": 2.8616684779671456e-05, + "loss": 0.0003, + "step": 7200 + }, + { + "epoch": 4.743421052631579, + "grad_norm": 0.020714622547958205, + "learning_rate": 2.8468523132542103e-05, + "loss": 0.0002, + "step": 7210 + }, + { + "epoch": 4.75, + "grad_norm": 0.027118166045878994, + "learning_rate": 2.832059322749272e-05, + "loss": 0.0002, + "step": 7220 + }, + { + "epoch": 4.756578947368421, + "grad_norm": 0.02649463264612147, + "learning_rate": 2.8172896656684646e-05, + "loss": 0.0003, + "step": 7230 + }, + { + "epoch": 4.7631578947368425, + "grad_norm": 0.031368963020435355, + "learning_rate": 2.8025435009767747e-05, + "loss": 0.0003, + "step": 7240 + }, + { + "epoch": 4.769736842105263, + "grad_norm": 0.03415277446285844, + "learning_rate": 2.7878209873863527e-05, + "loss": 0.0003, + "step": 7250 + }, + { + "epoch": 4.776315789473684, + "grad_norm": 0.02283882947145523, + "learning_rate": 2.7731222833547844e-05, + "loss": 0.0003, + "step": 7260 + }, + { + "epoch": 4.782894736842105, + "grad_norm": 0.021919040454945967, + "learning_rate": 2.7584475470834e-05, + "loss": 0.0003, + "step": 7270 + }, + { + "epoch": 4.7894736842105265, + "grad_norm": 0.019380219417494217, + "learning_rate": 2.7437969365155615e-05, + "loss": 0.0003, + "step": 7280 + }, + { + "epoch": 4.796052631578947, + "grad_norm": 0.021022723800318548, + "learning_rate": 2.7291706093349744e-05, + "loss": 0.0003, + "step": 7290 + }, + { + "epoch": 4.802631578947368, + "grad_norm": 0.026396525187615566, + "learning_rate": 2.714568722963975e-05, + "loss": 0.0002, + "step": 7300 + }, + { + "epoch": 4.809210526315789, + "grad_norm": 0.029282012131448747, + "learning_rate": 2.6999914345618527e-05, + "loss": 0.0003, + "step": 7310 + }, + { + "epoch": 4.815789473684211, + "grad_norm": 0.026146646421774805, + "learning_rate": 2.6854389010231428e-05, + "loss": 0.0003, + "step": 7320 + }, + { + "epoch": 4.822368421052632, + "grad_norm": 0.03600236240180555, + "learning_rate": 2.6709112789759528e-05, + "loss": 0.0003, + "step": 7330 + }, + { + "epoch": 4.828947368421053, + "grad_norm": 0.02447767163371457, + "learning_rate": 2.6564087247802626e-05, + "loss": 0.0002, + "step": 7340 + }, + { + "epoch": 4.8355263157894735, + "grad_norm": 0.02567474175665851, + "learning_rate": 2.6419313945262513e-05, + "loss": 0.0003, + "step": 7350 + }, + { + "epoch": 4.842105263157895, + "grad_norm": 0.022975015661936476, + "learning_rate": 2.6274794440326156e-05, + "loss": 0.0003, + "step": 7360 + }, + { + "epoch": 4.848684210526316, + "grad_norm": 0.02674281355005045, + "learning_rate": 2.6130530288448863e-05, + "loss": 0.0003, + "step": 7370 + }, + { + "epoch": 4.855263157894737, + "grad_norm": 0.03471272991096315, + "learning_rate": 2.5986523042337663e-05, + "loss": 0.0003, + "step": 7380 + }, + { + "epoch": 4.8618421052631575, + "grad_norm": 0.028880004017695318, + "learning_rate": 2.5842774251934444e-05, + "loss": 0.0003, + "step": 7390 + }, + { + "epoch": 4.868421052631579, + "grad_norm": 0.03147217508155407, + "learning_rate": 2.5699285464399426e-05, + "loss": 0.0003, + "step": 7400 + }, + { + "epoch": 4.875, + "grad_norm": 0.020875864408497622, + "learning_rate": 2.5556058224094376e-05, + "loss": 0.0003, + "step": 7410 + }, + { + "epoch": 4.881578947368421, + "grad_norm": 0.022120005201240307, + "learning_rate": 2.5413094072566067e-05, + "loss": 0.0003, + "step": 7420 + }, + { + "epoch": 4.8881578947368425, + "grad_norm": 0.028876359011806824, + "learning_rate": 2.527039454852963e-05, + "loss": 0.0003, + "step": 7430 + }, + { + "epoch": 4.894736842105263, + "grad_norm": 0.033771472279775235, + "learning_rate": 2.51279611878521e-05, + "loss": 0.0003, + "step": 7440 + }, + { + "epoch": 4.901315789473684, + "grad_norm": 0.022941541925073892, + "learning_rate": 2.498579552353572e-05, + "loss": 0.0003, + "step": 7450 + }, + { + "epoch": 4.907894736842105, + "grad_norm": 0.027242414832660027, + "learning_rate": 2.4843899085701616e-05, + "loss": 0.0003, + "step": 7460 + }, + { + "epoch": 4.9144736842105265, + "grad_norm": 0.022656949108881823, + "learning_rate": 2.4702273401573162e-05, + "loss": 0.0002, + "step": 7470 + }, + { + "epoch": 4.921052631578947, + "grad_norm": 0.02393381615843426, + "learning_rate": 2.456091999545971e-05, + "loss": 0.0002, + "step": 7480 + }, + { + "epoch": 4.927631578947368, + "grad_norm": 0.022044682836246252, + "learning_rate": 2.4419840388740007e-05, + "loss": 0.0003, + "step": 7490 + }, + { + "epoch": 4.934210526315789, + "grad_norm": 0.03822308893576144, + "learning_rate": 2.4279036099846032e-05, + "loss": 0.0003, + "step": 7500 + }, + { + "epoch": 4.940789473684211, + "grad_norm": 0.017200625526351394, + "learning_rate": 2.4138508644246376e-05, + "loss": 0.0002, + "step": 7510 + }, + { + "epoch": 4.947368421052632, + "grad_norm": 0.030473270945807145, + "learning_rate": 2.399825953443024e-05, + "loss": 0.0003, + "step": 7520 + }, + { + "epoch": 4.953947368421053, + "grad_norm": 0.03803385482337547, + "learning_rate": 2.3858290279890898e-05, + "loss": 0.0003, + "step": 7530 + }, + { + "epoch": 4.9605263157894735, + "grad_norm": 0.02584750781463227, + "learning_rate": 2.371860238710964e-05, + "loss": 0.0002, + "step": 7540 + }, + { + "epoch": 4.967105263157895, + "grad_norm": 0.018563054866135756, + "learning_rate": 2.3579197359539395e-05, + "loss": 0.0002, + "step": 7550 + }, + { + "epoch": 4.973684210526316, + "grad_norm": 0.01965296619151117, + "learning_rate": 2.344007669758871e-05, + "loss": 0.0003, + "step": 7560 + }, + { + "epoch": 4.980263157894737, + "grad_norm": 0.022243738803198618, + "learning_rate": 2.330124189860544e-05, + "loss": 0.0002, + "step": 7570 + }, + { + "epoch": 4.9868421052631575, + "grad_norm": 0.024230269193611945, + "learning_rate": 2.3162694456860818e-05, + "loss": 0.0003, + "step": 7580 + }, + { + "epoch": 4.993421052631579, + "grad_norm": 0.03770804060720148, + "learning_rate": 2.302443586353311e-05, + "loss": 0.0003, + "step": 7590 + }, + { + "epoch": 5.0, + "grad_norm": 0.015523299726247955, + "learning_rate": 2.288646760669189e-05, + "loss": 0.0002, + "step": 7600 + }, + { + "epoch": 5.006578947368421, + "grad_norm": 0.01308674669628331, + "learning_rate": 2.274879117128173e-05, + "loss": 0.0001, + "step": 7610 + }, + { + "epoch": 5.0131578947368425, + "grad_norm": 0.018835858337500996, + "learning_rate": 2.261140803910644e-05, + "loss": 0.0002, + "step": 7620 + }, + { + "epoch": 5.019736842105263, + "grad_norm": 0.021019811103633688, + "learning_rate": 2.2474319688812944e-05, + "loss": 0.0001, + "step": 7630 + }, + { + "epoch": 5.026315789473684, + "grad_norm": 0.024031107920490243, + "learning_rate": 2.233752759587552e-05, + "loss": 0.0001, + "step": 7640 + }, + { + "epoch": 5.032894736842105, + "grad_norm": 0.009517224332994808, + "learning_rate": 2.2201033232579753e-05, + "loss": 0.0001, + "step": 7650 + }, + { + "epoch": 5.0394736842105265, + "grad_norm": 0.013684410654666345, + "learning_rate": 2.206483806800689e-05, + "loss": 0.0001, + "step": 7660 + }, + { + "epoch": 5.046052631578948, + "grad_norm": 0.010244717454875643, + "learning_rate": 2.1928943568017825e-05, + "loss": 0.0001, + "step": 7670 + }, + { + "epoch": 5.052631578947368, + "grad_norm": 0.019525292666762634, + "learning_rate": 2.179335119523745e-05, + "loss": 0.0001, + "step": 7680 + }, + { + "epoch": 5.059210526315789, + "grad_norm": 0.0056361428706676745, + "learning_rate": 2.1658062409038903e-05, + "loss": 0.0001, + "step": 7690 + }, + { + "epoch": 5.065789473684211, + "grad_norm": 0.01297753430716979, + "learning_rate": 2.1523078665527797e-05, + "loss": 0.0001, + "step": 7700 + }, + { + "epoch": 5.072368421052632, + "grad_norm": 0.005174394102366024, + "learning_rate": 2.1388401417526625e-05, + "loss": 0.0001, + "step": 7710 + }, + { + "epoch": 5.078947368421052, + "grad_norm": 0.007310662083654455, + "learning_rate": 2.1254032114559042e-05, + "loss": 0.0001, + "step": 7720 + }, + { + "epoch": 5.0855263157894735, + "grad_norm": 0.013504841510499545, + "learning_rate": 2.1119972202834366e-05, + "loss": 0.0001, + "step": 7730 + }, + { + "epoch": 5.092105263157895, + "grad_norm": 0.008304238082242036, + "learning_rate": 2.0986223125231902e-05, + "loss": 0.0001, + "step": 7740 + }, + { + "epoch": 5.098684210526316, + "grad_norm": 0.007003361758782009, + "learning_rate": 2.085278632128548e-05, + "loss": 0.0001, + "step": 7750 + }, + { + "epoch": 5.105263157894737, + "grad_norm": 0.014767756139081512, + "learning_rate": 2.0719663227167934e-05, + "loss": 0.0001, + "step": 7760 + }, + { + "epoch": 5.1118421052631575, + "grad_norm": 0.008879341776595509, + "learning_rate": 2.058685527567571e-05, + "loss": 0.0001, + "step": 7770 + }, + { + "epoch": 5.118421052631579, + "grad_norm": 0.011232828377090194, + "learning_rate": 2.0454363896213317e-05, + "loss": 0.0001, + "step": 7780 + }, + { + "epoch": 5.125, + "grad_norm": 0.01177512821821725, + "learning_rate": 2.0322190514778093e-05, + "loss": 0.0001, + "step": 7790 + }, + { + "epoch": 5.131578947368421, + "grad_norm": 0.0034367512892905815, + "learning_rate": 2.0190336553944705e-05, + "loss": 0.0001, + "step": 7800 + }, + { + "epoch": 5.1381578947368425, + "grad_norm": 0.013524709582938698, + "learning_rate": 2.0058803432849987e-05, + "loss": 0.0001, + "step": 7810 + }, + { + "epoch": 5.144736842105263, + "grad_norm": 0.011627182786304667, + "learning_rate": 1.9927592567177538e-05, + "loss": 0.0001, + "step": 7820 + }, + { + "epoch": 5.151315789473684, + "grad_norm": 0.0076237965763939, + "learning_rate": 1.9796705369142556e-05, + "loss": 0.0001, + "step": 7830 + }, + { + "epoch": 5.157894736842105, + "grad_norm": 0.009790101547454817, + "learning_rate": 1.966614324747661e-05, + "loss": 0.0001, + "step": 7840 + }, + { + "epoch": 5.1644736842105265, + "grad_norm": 0.015266035990866829, + "learning_rate": 1.953590760741254e-05, + "loss": 0.0001, + "step": 7850 + }, + { + "epoch": 5.171052631578948, + "grad_norm": 0.012349145659134126, + "learning_rate": 1.9405999850669216e-05, + "loss": 0.0001, + "step": 7860 + }, + { + "epoch": 5.177631578947368, + "grad_norm": 0.010576627455789814, + "learning_rate": 1.927642137543658e-05, + "loss": 0.0001, + "step": 7870 + }, + { + "epoch": 5.184210526315789, + "grad_norm": 0.015284759015632338, + "learning_rate": 1.9147173576360484e-05, + "loss": 0.0001, + "step": 7880 + }, + { + "epoch": 5.190789473684211, + "grad_norm": 0.0046992693752646785, + "learning_rate": 1.901825784452777e-05, + "loss": 0.0001, + "step": 7890 + }, + { + "epoch": 5.197368421052632, + "grad_norm": 0.023375867751267324, + "learning_rate": 1.8889675567451232e-05, + "loss": 0.0001, + "step": 7900 + }, + { + "epoch": 5.203947368421052, + "grad_norm": 0.008777824124232628, + "learning_rate": 1.8761428129054725e-05, + "loss": 0.0001, + "step": 7910 + }, + { + "epoch": 5.2105263157894735, + "grad_norm": 0.006277829504136982, + "learning_rate": 1.863351690965822e-05, + "loss": 0.0001, + "step": 7920 + }, + { + "epoch": 5.217105263157895, + "grad_norm": 0.013841979474173404, + "learning_rate": 1.8505943285963052e-05, + "loss": 0.0001, + "step": 7930 + }, + { + "epoch": 5.223684210526316, + "grad_norm": 0.01205465510885825, + "learning_rate": 1.8378708631036956e-05, + "loss": 0.0001, + "step": 7940 + }, + { + "epoch": 5.230263157894737, + "grad_norm": 0.011830037477491624, + "learning_rate": 1.8251814314299427e-05, + "loss": 0.0001, + "step": 7950 + }, + { + "epoch": 5.2368421052631575, + "grad_norm": 0.008421734666409964, + "learning_rate": 1.8125261701506863e-05, + "loss": 0.0001, + "step": 7960 + }, + { + "epoch": 5.243421052631579, + "grad_norm": 0.008873300004629535, + "learning_rate": 1.799905215473798e-05, + "loss": 0.0001, + "step": 7970 + }, + { + "epoch": 5.25, + "grad_norm": 0.010278328623451821, + "learning_rate": 1.7873187032379052e-05, + "loss": 0.0001, + "step": 7980 + }, + { + "epoch": 5.256578947368421, + "grad_norm": 0.013812955331331695, + "learning_rate": 1.7747667689109322e-05, + "loss": 0.0001, + "step": 7990 + }, + { + "epoch": 5.2631578947368425, + "grad_norm": 0.0070521083618710085, + "learning_rate": 1.7622495475886487e-05, + "loss": 0.0001, + "step": 8000 + }, + { + "epoch": 5.269736842105263, + "grad_norm": 0.008322069016369617, + "learning_rate": 1.7497671739932036e-05, + "loss": 0.0001, + "step": 8010 + }, + { + "epoch": 5.276315789473684, + "grad_norm": 0.004885255623185907, + "learning_rate": 1.7373197824716885e-05, + "loss": 0.0001, + "step": 8020 + }, + { + "epoch": 5.282894736842105, + "grad_norm": 0.003027551697093571, + "learning_rate": 1.724907506994679e-05, + "loss": 0.0001, + "step": 8030 + }, + { + "epoch": 5.2894736842105265, + "grad_norm": 0.006147449906894669, + "learning_rate": 1.712530481154803e-05, + "loss": 0.0001, + "step": 8040 + }, + { + "epoch": 5.296052631578947, + "grad_norm": 0.010372441417200305, + "learning_rate": 1.7001888381652942e-05, + "loss": 0.0001, + "step": 8050 + }, + { + "epoch": 5.302631578947368, + "grad_norm": 0.006745515718462745, + "learning_rate": 1.6878827108585693e-05, + "loss": 0.0001, + "step": 8060 + }, + { + "epoch": 5.309210526315789, + "grad_norm": 0.006254813716049743, + "learning_rate": 1.6756122316847816e-05, + "loss": 0.0001, + "step": 8070 + }, + { + "epoch": 5.315789473684211, + "grad_norm": 0.0036128306826948313, + "learning_rate": 1.6633775327104163e-05, + "loss": 0.0001, + "step": 8080 + }, + { + "epoch": 5.322368421052632, + "grad_norm": 0.0055879568762191014, + "learning_rate": 1.6511787456168497e-05, + "loss": 0.0001, + "step": 8090 + }, + { + "epoch": 5.328947368421053, + "grad_norm": 0.006794088332004271, + "learning_rate": 1.6390160016989487e-05, + "loss": 0.0001, + "step": 8100 + }, + { + "epoch": 5.3355263157894735, + "grad_norm": 0.014056713440581896, + "learning_rate": 1.6268894318636403e-05, + "loss": 0.0001, + "step": 8110 + }, + { + "epoch": 5.342105263157895, + "grad_norm": 0.019460325770739945, + "learning_rate": 1.6147991666285207e-05, + "loss": 0.0001, + "step": 8120 + }, + { + "epoch": 5.348684210526316, + "grad_norm": 0.014908479642357846, + "learning_rate": 1.6027453361204342e-05, + "loss": 0.0001, + "step": 8130 + }, + { + "epoch": 5.355263157894737, + "grad_norm": 0.003220634441216755, + "learning_rate": 1.590728070074089e-05, + "loss": 0.0001, + "step": 8140 + }, + { + "epoch": 5.3618421052631575, + "grad_norm": 0.016590188051423045, + "learning_rate": 1.5787474978306405e-05, + "loss": 0.0001, + "step": 8150 + }, + { + "epoch": 5.368421052631579, + "grad_norm": 0.0059269669038211365, + "learning_rate": 1.5668037483363208e-05, + "loss": 0.0001, + "step": 8160 + }, + { + "epoch": 5.375, + "grad_norm": 0.010408346371898379, + "learning_rate": 1.5548969501410337e-05, + "loss": 0.0001, + "step": 8170 + }, + { + "epoch": 5.381578947368421, + "grad_norm": 0.003014499862618881, + "learning_rate": 1.5430272313969862e-05, + "loss": 0.0001, + "step": 8180 + }, + { + "epoch": 5.3881578947368425, + "grad_norm": 0.007824521063835533, + "learning_rate": 1.531194719857292e-05, + "loss": 0.0001, + "step": 8190 + }, + { + "epoch": 5.394736842105263, + "grad_norm": 0.01683072515718024, + "learning_rate": 1.5193995428746143e-05, + "loss": 0.0001, + "step": 8200 + }, + { + "epoch": 5.401315789473684, + "grad_norm": 0.015099533106096437, + "learning_rate": 1.5076418273997789e-05, + "loss": 0.0001, + "step": 8210 + }, + { + "epoch": 5.407894736842105, + "grad_norm": 0.009018818807405016, + "learning_rate": 1.4959216999804238e-05, + "loss": 0.0001, + "step": 8220 + }, + { + "epoch": 5.4144736842105265, + "grad_norm": 0.005165022918863679, + "learning_rate": 1.4842392867596183e-05, + "loss": 0.0001, + "step": 8230 + }, + { + "epoch": 5.421052631578947, + "grad_norm": 0.007247023522818596, + "learning_rate": 1.4725947134745265e-05, + "loss": 0.0001, + "step": 8240 + }, + { + "epoch": 5.427631578947368, + "grad_norm": 0.011678600650129461, + "learning_rate": 1.4609881054550357e-05, + "loss": 0.0001, + "step": 8250 + }, + { + "epoch": 5.434210526315789, + "grad_norm": 0.011929320335915653, + "learning_rate": 1.4494195876224204e-05, + "loss": 0.0001, + "step": 8260 + }, + { + "epoch": 5.440789473684211, + "grad_norm": 0.0045916259217581475, + "learning_rate": 1.4378892844879881e-05, + "loss": 0.0001, + "step": 8270 + }, + { + "epoch": 5.447368421052632, + "grad_norm": 0.003668471587814517, + "learning_rate": 1.4263973201517489e-05, + "loss": 0.0001, + "step": 8280 + }, + { + "epoch": 5.453947368421053, + "grad_norm": 0.013570058420265816, + "learning_rate": 1.4149438183010683e-05, + "loss": 0.0001, + "step": 8290 + }, + { + "epoch": 5.4605263157894735, + "grad_norm": 0.0053056688149639725, + "learning_rate": 1.4035289022093478e-05, + "loss": 0.0001, + "step": 8300 + }, + { + "epoch": 5.467105263157895, + "grad_norm": 0.010667443872923121, + "learning_rate": 1.3921526947346902e-05, + "loss": 0.0001, + "step": 8310 + }, + { + "epoch": 5.473684210526316, + "grad_norm": 0.004656989307097169, + "learning_rate": 1.3808153183185768e-05, + "loss": 0.0001, + "step": 8320 + }, + { + "epoch": 5.480263157894737, + "grad_norm": 0.011255826496223033, + "learning_rate": 1.3695168949845582e-05, + "loss": 0.0001, + "step": 8330 + }, + { + "epoch": 5.4868421052631575, + "grad_norm": 0.007466616630931639, + "learning_rate": 1.3582575463369302e-05, + "loss": 0.0001, + "step": 8340 + }, + { + "epoch": 5.493421052631579, + "grad_norm": 0.009712528932142932, + "learning_rate": 1.3470373935594343e-05, + "loss": 0.0001, + "step": 8350 + }, + { + "epoch": 5.5, + "grad_norm": 0.005720607642427896, + "learning_rate": 1.3358565574139432e-05, + "loss": 0.0001, + "step": 8360 + }, + { + "epoch": 5.506578947368421, + "grad_norm": 0.005484323558557781, + "learning_rate": 1.3247151582391743e-05, + "loss": 0.0001, + "step": 8370 + }, + { + "epoch": 5.5131578947368425, + "grad_norm": 0.011071460964179453, + "learning_rate": 1.3136133159493802e-05, + "loss": 0.0001, + "step": 8380 + }, + { + "epoch": 5.519736842105263, + "grad_norm": 0.0168732221218483, + "learning_rate": 1.3025511500330694e-05, + "loss": 0.0001, + "step": 8390 + }, + { + "epoch": 5.526315789473684, + "grad_norm": 0.005605518836973013, + "learning_rate": 1.2915287795517118e-05, + "loss": 0.0001, + "step": 8400 + }, + { + "epoch": 5.532894736842105, + "grad_norm": 0.016742385699511076, + "learning_rate": 1.2805463231384673e-05, + "loss": 0.0001, + "step": 8410 + }, + { + "epoch": 5.5394736842105265, + "grad_norm": 0.01107643414188672, + "learning_rate": 1.2696038989968962e-05, + "loss": 0.0001, + "step": 8420 + }, + { + "epoch": 5.546052631578947, + "grad_norm": 0.007249589366749116, + "learning_rate": 1.2587016248996997e-05, + "loss": 0.0001, + "step": 8430 + }, + { + "epoch": 5.552631578947368, + "grad_norm": 0.009918958856801371, + "learning_rate": 1.2478396181874403e-05, + "loss": 0.0001, + "step": 8440 + }, + { + "epoch": 5.559210526315789, + "grad_norm": 0.005743646230044239, + "learning_rate": 1.2370179957672917e-05, + "loss": 0.0001, + "step": 8450 + }, + { + "epoch": 5.565789473684211, + "grad_norm": 0.0029604989722085092, + "learning_rate": 1.2262368741117686e-05, + "loss": 0.0001, + "step": 8460 + }, + { + "epoch": 5.572368421052632, + "grad_norm": 0.012890537200679111, + "learning_rate": 1.2154963692574794e-05, + "loss": 0.0001, + "step": 8470 + }, + { + "epoch": 5.578947368421053, + "grad_norm": 0.004698208321657866, + "learning_rate": 1.2047965968038749e-05, + "loss": 0.0001, + "step": 8480 + }, + { + "epoch": 5.5855263157894735, + "grad_norm": 0.006389493226712366, + "learning_rate": 1.1941376719120101e-05, + "loss": 0.0001, + "step": 8490 + }, + { + "epoch": 5.592105263157895, + "grad_norm": 0.0028033342479427893, + "learning_rate": 1.183519709303294e-05, + "loss": 0.0001, + "step": 8500 + }, + { + "epoch": 5.598684210526316, + "grad_norm": 0.0080953761400738, + "learning_rate": 1.1729428232582667e-05, + "loss": 0.0001, + "step": 8510 + }, + { + "epoch": 5.605263157894737, + "grad_norm": 0.004957654730737493, + "learning_rate": 1.162407127615357e-05, + "loss": 0.0001, + "step": 8520 + }, + { + "epoch": 5.6118421052631575, + "grad_norm": 0.005707287356253592, + "learning_rate": 1.1519127357696708e-05, + "loss": 0.0001, + "step": 8530 + }, + { + "epoch": 5.618421052631579, + "grad_norm": 0.008121159493824286, + "learning_rate": 1.141459760671758e-05, + "loss": 0.0001, + "step": 8540 + }, + { + "epoch": 5.625, + "grad_norm": 0.01957665996113406, + "learning_rate": 1.1310483148264034e-05, + "loss": 0.0001, + "step": 8550 + }, + { + "epoch": 5.631578947368421, + "grad_norm": 0.012707168992640206, + "learning_rate": 1.120678510291418e-05, + "loss": 0.0001, + "step": 8560 + }, + { + "epoch": 5.6381578947368425, + "grad_norm": 0.003044955736496986, + "learning_rate": 1.1103504586764263e-05, + "loss": 0.0001, + "step": 8570 + }, + { + "epoch": 5.644736842105263, + "grad_norm": 0.006873499084550594, + "learning_rate": 1.1000642711416682e-05, + "loss": 0.0001, + "step": 8580 + }, + { + "epoch": 5.651315789473684, + "grad_norm": 0.007215778665166686, + "learning_rate": 1.0898200583968072e-05, + "loss": 0.0001, + "step": 8590 + }, + { + "epoch": 5.657894736842105, + "grad_norm": 0.013392635258067134, + "learning_rate": 1.0796179306997279e-05, + "loss": 0.0001, + "step": 8600 + }, + { + "epoch": 5.6644736842105265, + "grad_norm": 0.013825895059446676, + "learning_rate": 1.0694579978553632e-05, + "loss": 0.0001, + "step": 8610 + }, + { + "epoch": 5.671052631578947, + "grad_norm": 0.02166554611350375, + "learning_rate": 1.0593403692144998e-05, + "loss": 0.0001, + "step": 8620 + }, + { + "epoch": 5.677631578947368, + "grad_norm": 0.006393439385029049, + "learning_rate": 1.0492651536726067e-05, + "loss": 0.0001, + "step": 8630 + }, + { + "epoch": 5.684210526315789, + "grad_norm": 0.011818417331531703, + "learning_rate": 1.0392324596686676e-05, + "loss": 0.0001, + "step": 8640 + }, + { + "epoch": 5.690789473684211, + "grad_norm": 0.0032546461666155316, + "learning_rate": 1.0292423951840035e-05, + "loss": 0.0001, + "step": 8650 + }, + { + "epoch": 5.697368421052632, + "grad_norm": 0.011404441083432902, + "learning_rate": 1.0192950677411217e-05, + "loss": 0.0001, + "step": 8660 + }, + { + "epoch": 5.703947368421053, + "grad_norm": 0.004176671934655934, + "learning_rate": 1.0093905844025481e-05, + "loss": 0.0001, + "step": 8670 + }, + { + "epoch": 5.7105263157894735, + "grad_norm": 0.010304584324367016, + "learning_rate": 9.995290517696854e-06, + "loss": 0.0001, + "step": 8680 + }, + { + "epoch": 5.717105263157895, + "grad_norm": 0.010604220961249872, + "learning_rate": 9.897105759816532e-06, + "loss": 0.0001, + "step": 8690 + }, + { + "epoch": 5.723684210526316, + "grad_norm": 0.01652772631244675, + "learning_rate": 9.799352627141622e-06, + "loss": 0.0001, + "step": 8700 + }, + { + "epoch": 5.730263157894737, + "grad_norm": 0.002838488676642155, + "learning_rate": 9.702032171783554e-06, + "loss": 0.0001, + "step": 8710 + }, + { + "epoch": 5.7368421052631575, + "grad_norm": 0.011431321594461334, + "learning_rate": 9.605145441196989e-06, + "loss": 0.0001, + "step": 8720 + }, + { + "epoch": 5.743421052631579, + "grad_norm": 0.0033197716204990857, + "learning_rate": 9.508693478168346e-06, + "loss": 0.0001, + "step": 8730 + }, + { + "epoch": 5.75, + "grad_norm": 0.010815902149616368, + "learning_rate": 9.412677320804741e-06, + "loss": 0.0001, + "step": 8740 + }, + { + "epoch": 5.756578947368421, + "grad_norm": 0.007968676598176684, + "learning_rate": 9.317098002522661e-06, + "loss": 0.0001, + "step": 8750 + }, + { + "epoch": 5.7631578947368425, + "grad_norm": 0.004264311349098634, + "learning_rate": 9.221956552036992e-06, + "loss": 0.0001, + "step": 8760 + }, + { + "epoch": 5.769736842105263, + "grad_norm": 0.007659009384104305, + "learning_rate": 9.127253993349816e-06, + "loss": 0.0001, + "step": 8770 + }, + { + "epoch": 5.776315789473684, + "grad_norm": 0.012195776259417656, + "learning_rate": 9.0329913457395e-06, + "loss": 0.0001, + "step": 8780 + }, + { + "epoch": 5.782894736842105, + "grad_norm": 0.007081287777946904, + "learning_rate": 8.939169623749593e-06, + "loss": 0.0001, + "step": 8790 + }, + { + "epoch": 5.7894736842105265, + "grad_norm": 0.00267533304324417, + "learning_rate": 8.845789837178076e-06, + "loss": 0.0001, + "step": 8800 + }, + { + "epoch": 5.796052631578947, + "grad_norm": 0.014725309022525966, + "learning_rate": 8.75285299106634e-06, + "loss": 0.0001, + "step": 8810 + }, + { + "epoch": 5.802631578947368, + "grad_norm": 0.014879863127954874, + "learning_rate": 8.660360085688479e-06, + "loss": 0.0001, + "step": 8820 + }, + { + "epoch": 5.809210526315789, + "grad_norm": 0.004634183050818591, + "learning_rate": 8.568312116540427e-06, + "loss": 0.0001, + "step": 8830 + }, + { + "epoch": 5.815789473684211, + "grad_norm": 0.017299568890365276, + "learning_rate": 8.476710074329352e-06, + "loss": 0.0, + "step": 8840 + }, + { + "epoch": 5.822368421052632, + "grad_norm": 0.018095881033860166, + "learning_rate": 8.38555494496288e-06, + "loss": 0.0001, + "step": 8850 + }, + { + "epoch": 5.828947368421053, + "grad_norm": 0.0034932964218642045, + "learning_rate": 8.29484770953859e-06, + "loss": 0.0001, + "step": 8860 + }, + { + "epoch": 5.8355263157894735, + "grad_norm": 0.005201021728839843, + "learning_rate": 8.204589344333363e-06, + "loss": 0.0001, + "step": 8870 + }, + { + "epoch": 5.842105263157895, + "grad_norm": 0.008580919024950527, + "learning_rate": 8.114780820792923e-06, + "loss": 0.0001, + "step": 8880 + }, + { + "epoch": 5.848684210526316, + "grad_norm": 0.010511235385955836, + "learning_rate": 8.025423105521395e-06, + "loss": 0.0001, + "step": 8890 + }, + { + "epoch": 5.855263157894737, + "grad_norm": 0.008141005305355695, + "learning_rate": 7.93651716027085e-06, + "loss": 0.0001, + "step": 8900 + }, + { + "epoch": 5.8618421052631575, + "grad_norm": 0.01085247124881257, + "learning_rate": 7.848063941931028e-06, + "loss": 0.0001, + "step": 8910 + }, + { + "epoch": 5.868421052631579, + "grad_norm": 0.0094965838697217, + "learning_rate": 7.760064402518935e-06, + "loss": 0.0001, + "step": 8920 + }, + { + "epoch": 5.875, + "grad_norm": 0.0064223214244996255, + "learning_rate": 7.672519489168695e-06, + "loss": 0.0001, + "step": 8930 + }, + { + "epoch": 5.881578947368421, + "grad_norm": 0.013687216803899342, + "learning_rate": 7.585430144121319e-06, + "loss": 0.0001, + "step": 8940 + }, + { + "epoch": 5.8881578947368425, + "grad_norm": 0.006225396329115645, + "learning_rate": 7.498797304714544e-06, + "loss": 0.0, + "step": 8950 + }, + { + "epoch": 5.894736842105263, + "grad_norm": 0.0021317816032665367, + "learning_rate": 7.4126219033727575e-06, + "loss": 0.0001, + "step": 8960 + }, + { + "epoch": 5.901315789473684, + "grad_norm": 0.008906579601796973, + "learning_rate": 7.326904867597007e-06, + "loss": 0.0001, + "step": 8970 + }, + { + "epoch": 5.907894736842105, + "grad_norm": 0.006689917654711315, + "learning_rate": 7.241647119954919e-06, + "loss": 0.0001, + "step": 8980 + }, + { + "epoch": 5.9144736842105265, + "grad_norm": 0.004820515455663672, + "learning_rate": 7.15684957807089e-06, + "loss": 0.0001, + "step": 8990 + }, + { + "epoch": 5.921052631578947, + "grad_norm": 0.00665429533877142, + "learning_rate": 7.072513154616084e-06, + "loss": 0.0001, + "step": 9000 + }, + { + "epoch": 5.927631578947368, + "grad_norm": 0.009128697692733848, + "learning_rate": 6.988638757298732e-06, + "loss": 0.0001, + "step": 9010 + }, + { + "epoch": 5.934210526315789, + "grad_norm": 0.006740095845324385, + "learning_rate": 6.905227288854266e-06, + "loss": 0.0001, + "step": 9020 + }, + { + "epoch": 5.940789473684211, + "grad_norm": 0.0076823089411141654, + "learning_rate": 6.822279647035645e-06, + "loss": 0.0, + "step": 9030 + }, + { + "epoch": 5.947368421052632, + "grad_norm": 0.006823553697779683, + "learning_rate": 6.739796724603692e-06, + "loss": 0.0001, + "step": 9040 + }, + { + "epoch": 5.953947368421053, + "grad_norm": 0.0021846053611632595, + "learning_rate": 6.6577794093174935e-06, + "loss": 0.0001, + "step": 9050 + }, + { + "epoch": 5.9605263157894735, + "grad_norm": 0.01440404931798736, + "learning_rate": 6.576228583924804e-06, + "loss": 0.0001, + "step": 9060 + }, + { + "epoch": 5.967105263157895, + "grad_norm": 0.003704123175739426, + "learning_rate": 6.495145126152619e-06, + "loss": 0.0001, + "step": 9070 + }, + { + "epoch": 5.973684210526316, + "grad_norm": 0.002183821847086016, + "learning_rate": 6.4145299086976305e-06, + "loss": 0.0001, + "step": 9080 + }, + { + "epoch": 5.980263157894737, + "grad_norm": 0.01550696360833518, + "learning_rate": 6.334383799216925e-06, + "loss": 0.0001, + "step": 9090 + }, + { + "epoch": 5.9868421052631575, + "grad_norm": 0.009749806743884233, + "learning_rate": 6.254707660318593e-06, + "loss": 0.0, + "step": 9100 + }, + { + "epoch": 5.993421052631579, + "grad_norm": 0.011301237856781943, + "learning_rate": 6.175502349552442e-06, + "loss": 0.0001, + "step": 9110 + }, + { + "epoch": 6.0, + "grad_norm": 0.006900129120631107, + "learning_rate": 6.096768719400797e-06, + "loss": 0.0, + "step": 9120 + }, + { + "epoch": 6.006578947368421, + "grad_norm": 0.009048255078987696, + "learning_rate": 6.018507617269331e-06, + "loss": 0.0, + "step": 9130 + }, + { + "epoch": 6.0131578947368425, + "grad_norm": 0.0013163258278890133, + "learning_rate": 5.94071988547788e-06, + "loss": 0.0, + "step": 9140 + }, + { + "epoch": 6.019736842105263, + "grad_norm": 0.0020211890073067143, + "learning_rate": 5.863406361251472e-06, + "loss": 0.0, + "step": 9150 + }, + { + "epoch": 6.026315789473684, + "grad_norm": 0.001341796114379377, + "learning_rate": 5.786567876711213e-06, + "loss": 0.0, + "step": 9160 + }, + { + "epoch": 6.032894736842105, + "grad_norm": 0.0010257940933172413, + "learning_rate": 5.710205258865431e-06, + "loss": 0.0, + "step": 9170 + }, + { + "epoch": 6.0394736842105265, + "grad_norm": 0.0011074478926877968, + "learning_rate": 5.634319329600701e-06, + "loss": 0.0, + "step": 9180 + }, + { + "epoch": 6.046052631578948, + "grad_norm": 0.002544635558174862, + "learning_rate": 5.558910905673015e-06, + "loss": 0.0, + "step": 9190 + }, + { + "epoch": 6.052631578947368, + "grad_norm": 0.0014766302559603302, + "learning_rate": 5.48398079869904e-06, + "loss": 0.0, + "step": 9200 + }, + { + "epoch": 6.059210526315789, + "grad_norm": 0.0011389759666169234, + "learning_rate": 5.409529815147307e-06, + "loss": 0.0, + "step": 9210 + }, + { + "epoch": 6.065789473684211, + "grad_norm": 0.0014258070804333942, + "learning_rate": 5.335558756329595e-06, + "loss": 0.0, + "step": 9220 + }, + { + "epoch": 6.072368421052632, + "grad_norm": 0.0015232771543967783, + "learning_rate": 5.262068418392257e-06, + "loss": 0.0, + "step": 9230 + }, + { + "epoch": 6.078947368421052, + "grad_norm": 0.0013681239639624226, + "learning_rate": 5.189059592307699e-06, + "loss": 0.0, + "step": 9240 + }, + { + "epoch": 6.0855263157894735, + "grad_norm": 0.0013451393119743322, + "learning_rate": 5.1165330638658085e-06, + "loss": 0.0, + "step": 9250 + }, + { + "epoch": 6.092105263157895, + "grad_norm": 0.0016189180992600058, + "learning_rate": 5.0444896136655875e-06, + "loss": 0.0, + "step": 9260 + }, + { + "epoch": 6.098684210526316, + "grad_norm": 0.0010207689384299117, + "learning_rate": 4.972930017106609e-06, + "loss": 0.0, + "step": 9270 + }, + { + "epoch": 6.105263157894737, + "grad_norm": 0.0037121856436361437, + "learning_rate": 4.9018550443808375e-06, + "loss": 0.0, + "step": 9280 + }, + { + "epoch": 6.1118421052631575, + "grad_norm": 0.0027577187813681635, + "learning_rate": 4.8312654604642015e-06, + "loss": 0.0, + "step": 9290 + }, + { + "epoch": 6.118421052631579, + "grad_norm": 0.003859972725302638, + "learning_rate": 4.761162025108468e-06, + "loss": 0.0, + "step": 9300 + }, + { + "epoch": 6.125, + "grad_norm": 0.000993502205738445, + "learning_rate": 4.6915454928329605e-06, + "loss": 0.0, + "step": 9310 + }, + { + "epoch": 6.131578947368421, + "grad_norm": 0.0012937324022856465, + "learning_rate": 4.622416612916541e-06, + "loss": 0.0, + "step": 9320 + }, + { + "epoch": 6.1381578947368425, + "grad_norm": 0.003644135475761256, + "learning_rate": 4.5537761293894535e-06, + "loss": 0.0, + "step": 9330 + }, + { + "epoch": 6.144736842105263, + "grad_norm": 0.0008599937966376883, + "learning_rate": 4.485624781025405e-06, + "loss": 0.0, + "step": 9340 + }, + { + "epoch": 6.151315789473684, + "grad_norm": 0.0019715998938395485, + "learning_rate": 4.417963301333511e-06, + "loss": 0.0, + "step": 9350 + }, + { + "epoch": 6.157894736842105, + "grad_norm": 0.004231686150859265, + "learning_rate": 4.350792418550509e-06, + "loss": 0.0, + "step": 9360 + }, + { + "epoch": 6.1644736842105265, + "grad_norm": 0.0014001932475934556, + "learning_rate": 4.284112855632833e-06, + "loss": 0.0, + "step": 9370 + }, + { + "epoch": 6.171052631578948, + "grad_norm": 0.0014327925071072583, + "learning_rate": 4.217925330248907e-06, + "loss": 0.0, + "step": 9380 + }, + { + "epoch": 6.177631578947368, + "grad_norm": 0.0009429020806426565, + "learning_rate": 4.1522305547713416e-06, + "loss": 0.0, + "step": 9390 + }, + { + "epoch": 6.184210526315789, + "grad_norm": 0.0011260014983553402, + "learning_rate": 4.087029236269346e-06, + "loss": 0.0, + "step": 9400 + }, + { + "epoch": 6.190789473684211, + "grad_norm": 0.0015173799667341779, + "learning_rate": 4.022322076501039e-06, + "loss": 0.0, + "step": 9410 + }, + { + "epoch": 6.197368421052632, + "grad_norm": 0.0009049508723918266, + "learning_rate": 3.9581097719059875e-06, + "loss": 0.0, + "step": 9420 + }, + { + "epoch": 6.203947368421052, + "grad_norm": 0.002127961358655321, + "learning_rate": 3.894393013597597e-06, + "loss": 0.0, + "step": 9430 + }, + { + "epoch": 6.2105263157894735, + "grad_norm": 0.003052205115392126, + "learning_rate": 3.831172487355794e-06, + "loss": 0.0, + "step": 9440 + }, + { + "epoch": 6.217105263157895, + "grad_norm": 0.0015358468951456854, + "learning_rate": 3.7684488736195457e-06, + "loss": 0.0, + "step": 9450 + }, + { + "epoch": 6.223684210526316, + "grad_norm": 0.0009435573925978957, + "learning_rate": 3.7062228474796122e-06, + "loss": 0.0, + "step": 9460 + }, + { + "epoch": 6.230263157894737, + "grad_norm": 0.0009936403846155727, + "learning_rate": 3.644495078671212e-06, + "loss": 0.0, + "step": 9470 + }, + { + "epoch": 6.2368421052631575, + "grad_norm": 0.0010158270691015962, + "learning_rate": 3.5832662315668786e-06, + "loss": 0.0, + "step": 9480 + }, + { + "epoch": 6.243421052631579, + "grad_norm": 0.001047193069930328, + "learning_rate": 3.522536965169254e-06, + "loss": 0.0, + "step": 9490 + }, + { + "epoch": 6.25, + "grad_norm": 0.0007087328006936424, + "learning_rate": 3.4623079331040476e-06, + "loss": 0.0, + "step": 9500 + }, + { + "epoch": 6.256578947368421, + "grad_norm": 0.002089193063851605, + "learning_rate": 3.402579783612947e-06, + "loss": 0.0, + "step": 9510 + }, + { + "epoch": 6.2631578947368425, + "grad_norm": 0.001696414816560037, + "learning_rate": 3.343353159546675e-06, + "loss": 0.0, + "step": 9520 + }, + { + "epoch": 6.269736842105263, + "grad_norm": 0.000950000080180676, + "learning_rate": 3.2846286983580887e-06, + "loss": 0.0, + "step": 9530 + }, + { + "epoch": 6.276315789473684, + "grad_norm": 0.011507235616655264, + "learning_rate": 3.226407032095252e-06, + "loss": 0.0, + "step": 9540 + }, + { + "epoch": 6.282894736842105, + "grad_norm": 0.0008557439834521571, + "learning_rate": 3.16868878739473e-06, + "loss": 0.0, + "step": 9550 + }, + { + "epoch": 6.2894736842105265, + "grad_norm": 0.0007251611783064778, + "learning_rate": 3.1114745854747275e-06, + "loss": 0.0001, + "step": 9560 + }, + { + "epoch": 6.296052631578947, + "grad_norm": 0.005183223351755799, + "learning_rate": 3.054765042128521e-06, + "loss": 0.0, + "step": 9570 + }, + { + "epoch": 6.302631578947368, + "grad_norm": 0.02345361462883932, + "learning_rate": 2.9985607677177343e-06, + "loss": 0.0, + "step": 9580 + }, + { + "epoch": 6.309210526315789, + "grad_norm": 0.0009284684684224283, + "learning_rate": 2.9428623671658497e-06, + "loss": 0.0, + "step": 9590 + }, + { + "epoch": 6.315789473684211, + "grad_norm": 0.0011856587902449377, + "learning_rate": 2.8876704399515953e-06, + "loss": 0.0, + "step": 9600 + }, + { + "epoch": 6.322368421052632, + "grad_norm": 0.0008500403623564375, + "learning_rate": 2.832985580102626e-06, + "loss": 0.0, + "step": 9610 + }, + { + "epoch": 6.328947368421053, + "grad_norm": 0.0008263301757097466, + "learning_rate": 2.7788083761890072e-06, + "loss": 0.0, + "step": 9620 + }, + { + "epoch": 6.3355263157894735, + "grad_norm": 0.0014390985588880602, + "learning_rate": 2.7251394113169626e-06, + "loss": 0.0, + "step": 9630 + }, + { + "epoch": 6.342105263157895, + "grad_norm": 0.001489422202682345, + "learning_rate": 2.671979263122548e-06, + "loss": 0.0, + "step": 9640 + }, + { + "epoch": 6.348684210526316, + "grad_norm": 0.001656999905895154, + "learning_rate": 2.619328503765478e-06, + "loss": 0.0, + "step": 9650 + }, + { + "epoch": 6.355263157894737, + "grad_norm": 0.0019313702763058484, + "learning_rate": 2.567187699922902e-06, + "loss": 0.0, + "step": 9660 + }, + { + "epoch": 6.3618421052631575, + "grad_norm": 0.00170177406927297, + "learning_rate": 2.515557412783409e-06, + "loss": 0.0, + "step": 9670 + }, + { + "epoch": 6.368421052631579, + "grad_norm": 0.0005225273661022363, + "learning_rate": 2.4644381980408472e-06, + "loss": 0.0, + "step": 9680 + }, + { + "epoch": 6.375, + "grad_norm": 0.0009083190780930485, + "learning_rate": 2.4138306058884897e-06, + "loss": 0.0, + "step": 9690 + }, + { + "epoch": 6.381578947368421, + "grad_norm": 0.0007151588940599993, + "learning_rate": 2.363735181013005e-06, + "loss": 0.0, + "step": 9700 + }, + { + "epoch": 6.3881578947368425, + "grad_norm": 0.002344362431300476, + "learning_rate": 2.314152462588659e-06, + "loss": 0.0, + "step": 9710 + }, + { + "epoch": 6.394736842105263, + "grad_norm": 0.003673299833566869, + "learning_rate": 2.265082984271455e-06, + "loss": 0.0, + "step": 9720 + }, + { + "epoch": 6.401315789473684, + "grad_norm": 0.005580901358083729, + "learning_rate": 2.216527274193464e-06, + "loss": 0.0, + "step": 9730 + }, + { + "epoch": 6.407894736842105, + "grad_norm": 0.0024896171705098644, + "learning_rate": 2.168485854957053e-06, + "loss": 0.0, + "step": 9740 + }, + { + "epoch": 6.4144736842105265, + "grad_norm": 0.006034031006575288, + "learning_rate": 2.1209592436293503e-06, + "loss": 0.0, + "step": 9750 + }, + { + "epoch": 6.421052631578947, + "grad_norm": 0.001257888117785242, + "learning_rate": 2.073947951736588e-06, + "loss": 0.0, + "step": 9760 + }, + { + "epoch": 6.427631578947368, + "grad_norm": 0.0011316429907789183, + "learning_rate": 2.027452485258685e-06, + "loss": 0.0, + "step": 9770 + }, + { + "epoch": 6.434210526315789, + "grad_norm": 0.00130207664846368, + "learning_rate": 1.9814733446237354e-06, + "loss": 0.0, + "step": 9780 + }, + { + "epoch": 6.440789473684211, + "grad_norm": 0.001012695925762098, + "learning_rate": 1.9360110247026608e-06, + "loss": 0.0, + "step": 9790 + }, + { + "epoch": 6.447368421052632, + "grad_norm": 0.0005676853492370789, + "learning_rate": 1.8910660148038607e-06, + "loss": 0.0, + "step": 9800 + }, + { + "epoch": 6.453947368421053, + "grad_norm": 0.0008825402023503085, + "learning_rate": 1.8466387986679712e-06, + "loss": 0.0, + "step": 9810 + }, + { + "epoch": 6.4605263157894735, + "grad_norm": 0.0010622566658100676, + "learning_rate": 1.8027298544626149e-06, + "loss": 0.0, + "step": 9820 + }, + { + "epoch": 6.467105263157895, + "grad_norm": 0.0009578695139693795, + "learning_rate": 1.7593396547773144e-06, + "loss": 0.0, + "step": 9830 + }, + { + "epoch": 6.473684210526316, + "grad_norm": 0.01470044647652375, + "learning_rate": 1.7164686666183538e-06, + "loss": 0.0, + "step": 9840 + }, + { + "epoch": 6.480263157894737, + "grad_norm": 0.007381575301002969, + "learning_rate": 1.6741173514037757e-06, + "loss": 0.0, + "step": 9850 + }, + { + "epoch": 6.4868421052631575, + "grad_norm": 0.0011181381303234043, + "learning_rate": 1.6322861649584186e-06, + "loss": 0.0, + "step": 9860 + }, + { + "epoch": 6.493421052631579, + "grad_norm": 0.0009430106837510958, + "learning_rate": 1.5909755575090001e-06, + "loss": 0.0, + "step": 9870 + }, + { + "epoch": 6.5, + "grad_norm": 0.00118307270711618, + "learning_rate": 1.5501859736792745e-06, + "loss": 0.0, + "step": 9880 + }, + { + "epoch": 6.506578947368421, + "grad_norm": 0.0012521523923892414, + "learning_rate": 1.509917852485243e-06, + "loss": 0.0, + "step": 9890 + }, + { + "epoch": 6.5131578947368425, + "grad_norm": 0.001954500278285609, + "learning_rate": 1.4701716273304521e-06, + "loss": 0.0, + "step": 9900 + }, + { + "epoch": 6.519736842105263, + "grad_norm": 0.0012097851605845577, + "learning_rate": 1.4309477260012916e-06, + "loss": 0.0, + "step": 9910 + }, + { + "epoch": 6.526315789473684, + "grad_norm": 0.0015024927833277085, + "learning_rate": 1.392246570662409e-06, + "loss": 0.0, + "step": 9920 + }, + { + "epoch": 6.532894736842105, + "grad_norm": 0.0014780250352941372, + "learning_rate": 1.3540685778521744e-06, + "loss": 0.0, + "step": 9930 + }, + { + "epoch": 6.5394736842105265, + "grad_norm": 0.0008950104866338849, + "learning_rate": 1.3164141584781964e-06, + "loss": 0.0, + "step": 9940 + }, + { + "epoch": 6.546052631578947, + "grad_norm": 0.0023444458580515774, + "learning_rate": 1.2792837178128737e-06, + "loss": 0.0, + "step": 9950 + }, + { + "epoch": 6.552631578947368, + "grad_norm": 0.0007041151313729805, + "learning_rate": 1.2426776554890662e-06, + "loss": 0.0, + "step": 9960 + }, + { + "epoch": 6.559210526315789, + "grad_norm": 0.0059479080467208165, + "learning_rate": 1.2065963654957602e-06, + "loss": 0.0, + "step": 9970 + }, + { + "epoch": 6.565789473684211, + "grad_norm": 0.001069492550234626, + "learning_rate": 1.1710402361738814e-06, + "loss": 0.0, + "step": 9980 + }, + { + "epoch": 6.572368421052632, + "grad_norm": 0.000985644847670968, + "learning_rate": 1.1360096502120387e-06, + "loss": 0.0, + "step": 9990 + }, + { + "epoch": 6.578947368421053, + "grad_norm": 0.0020192809569499603, + "learning_rate": 1.101504984642482e-06, + "loss": 0.0, + "step": 10000 + }, + { + "epoch": 6.5855263157894735, + "grad_norm": 0.002703232151186019, + "learning_rate": 1.067526610836983e-06, + "loss": 0.0, + "step": 10010 + }, + { + "epoch": 6.592105263157895, + "grad_norm": 0.0011477701802225793, + "learning_rate": 1.0340748945028845e-06, + "loss": 0.0, + "step": 10020 + }, + { + "epoch": 6.598684210526316, + "grad_norm": 0.0011586580766907656, + "learning_rate": 1.0011501956791291e-06, + "loss": 0.0, + "step": 10030 + }, + { + "epoch": 6.605263157894737, + "grad_norm": 0.0009117709025104098, + "learning_rate": 9.687528687324144e-07, + "loss": 0.0, + "step": 10040 + }, + { + "epoch": 6.6118421052631575, + "grad_norm": 0.002991881435549838, + "learning_rate": 9.368832623533552e-07, + "loss": 0.0, + "step": 10050 + }, + { + "epoch": 6.618421052631579, + "grad_norm": 0.001899653803877392, + "learning_rate": 9.055417195527382e-07, + "loss": 0.0, + "step": 10060 + }, + { + "epoch": 6.625, + "grad_norm": 0.0007544065030484288, + "learning_rate": 8.747285776578351e-07, + "loss": 0.0, + "step": 10070 + }, + { + "epoch": 6.631578947368421, + "grad_norm": 0.0014084077019379283, + "learning_rate": 8.444441683087722e-07, + "loss": 0.0, + "step": 10080 + }, + { + "epoch": 6.6381578947368425, + "grad_norm": 0.0009232437173826582, + "learning_rate": 8.146888174549339e-07, + "loss": 0.0, + "step": 10090 + }, + { + "epoch": 6.644736842105263, + "grad_norm": 0.0017301066008694962, + "learning_rate": 7.8546284535152e-07, + "loss": 0.0, + "step": 10100 + }, + { + "epoch": 6.651315789473684, + "grad_norm": 0.0009812772784686134, + "learning_rate": 7.567665665560109e-07, + "loss": 0.0, + "step": 10110 + }, + { + "epoch": 6.657894736842105, + "grad_norm": 0.0036026944527775915, + "learning_rate": 7.286002899248579e-07, + "loss": 0.0, + "step": 10120 + }, + { + "epoch": 6.6644736842105265, + "grad_norm": 0.0007918627023736293, + "learning_rate": 7.0096431861012e-07, + "loss": 0.0, + "step": 10130 + }, + { + "epoch": 6.671052631578947, + "grad_norm": 0.0008410888110147981, + "learning_rate": 6.738589500562109e-07, + "loss": 0.0, + "step": 10140 + }, + { + "epoch": 6.677631578947368, + "grad_norm": 0.013805294085786955, + "learning_rate": 6.472844759966901e-07, + "loss": 0.0, + "step": 10150 + }, + { + "epoch": 6.684210526315789, + "grad_norm": 0.0009541232503700829, + "learning_rate": 6.212411824511322e-07, + "loss": 0.0, + "step": 10160 + }, + { + "epoch": 6.690789473684211, + "grad_norm": 0.0009490992244571239, + "learning_rate": 5.957293497220518e-07, + "loss": 0.0, + "step": 10170 + }, + { + "epoch": 6.697368421052632, + "grad_norm": 0.0008394158586557938, + "learning_rate": 5.707492523918778e-07, + "loss": 0.0, + "step": 10180 + }, + { + "epoch": 6.703947368421053, + "grad_norm": 0.0009550627472560924, + "learning_rate": 5.463011593199951e-07, + "loss": 0.0, + "step": 10190 + }, + { + "epoch": 6.7105263157894735, + "grad_norm": 0.0015541750924644652, + "learning_rate": 5.223853336398632e-07, + "loss": 0.0, + "step": 10200 + }, + { + "epoch": 6.717105263157895, + "grad_norm": 0.0009729740867904749, + "learning_rate": 4.990020327561739e-07, + "loss": 0.0, + "step": 10210 + }, + { + "epoch": 6.723684210526316, + "grad_norm": 0.000794139125183638, + "learning_rate": 4.76151508342082e-07, + "loss": 0.0, + "step": 10220 + }, + { + "epoch": 6.730263157894737, + "grad_norm": 0.002224072057115781, + "learning_rate": 4.5383400633651206e-07, + "loss": 0.0, + "step": 10230 + }, + { + "epoch": 6.7368421052631575, + "grad_norm": 0.0035553632152483562, + "learning_rate": 4.3204976694147247e-07, + "loss": 0.0, + "step": 10240 + }, + { + "epoch": 6.743421052631579, + "grad_norm": 0.0010919219633369732, + "learning_rate": 4.107990246195292e-07, + "loss": 0.0, + "step": 10250 + }, + { + "epoch": 6.75, + "grad_norm": 0.0008806556246415422, + "learning_rate": 3.9008200809122485e-07, + "loss": 0.0, + "step": 10260 + }, + { + "epoch": 6.756578947368421, + "grad_norm": 0.0011156606568035916, + "learning_rate": 3.6989894033264695e-07, + "loss": 0.0, + "step": 10270 + }, + { + "epoch": 6.7631578947368425, + "grad_norm": 0.0008354587151455838, + "learning_rate": 3.50250038573019e-07, + "loss": 0.0, + "step": 10280 + }, + { + "epoch": 6.769736842105263, + "grad_norm": 0.0019032873369056202, + "learning_rate": 3.311355142923855e-07, + "loss": 0.0, + "step": 10290 + }, + { + "epoch": 6.776315789473684, + "grad_norm": 0.0010059703761624691, + "learning_rate": 3.125555732192864e-07, + "loss": 0.0, + "step": 10300 + }, + { + "epoch": 6.782894736842105, + "grad_norm": 0.0013654808290868468, + "learning_rate": 2.945104153286027e-07, + "loss": 0.0, + "step": 10310 + }, + { + "epoch": 6.7894736842105265, + "grad_norm": 0.002719460783555821, + "learning_rate": 2.7700023483933634e-07, + "loss": 0.0, + "step": 10320 + }, + { + "epoch": 6.796052631578947, + "grad_norm": 0.0009491326130550004, + "learning_rate": 2.600252202126008e-07, + "loss": 0.0, + "step": 10330 + }, + { + "epoch": 6.802631578947368, + "grad_norm": 0.0010992607182720912, + "learning_rate": 2.4358555414951137e-07, + "loss": 0.0, + "step": 10340 + }, + { + "epoch": 6.809210526315789, + "grad_norm": 0.0022424626156541502, + "learning_rate": 2.2768141358928686e-07, + "loss": 0.0, + "step": 10350 + }, + { + "epoch": 6.815789473684211, + "grad_norm": 0.001029424407377985, + "learning_rate": 2.1231296970729564e-07, + "loss": 0.0, + "step": 10360 + }, + { + "epoch": 6.822368421052632, + "grad_norm": 0.000931287609630211, + "learning_rate": 1.9748038791323476e-07, + "loss": 0.0, + "step": 10370 + }, + { + "epoch": 6.828947368421053, + "grad_norm": 0.0019619446556149106, + "learning_rate": 1.8318382784935918e-07, + "loss": 0.0, + "step": 10380 + }, + { + "epoch": 6.8355263157894735, + "grad_norm": 0.0010415239386734165, + "learning_rate": 1.6942344338874428e-07, + "loss": 0.0, + "step": 10390 + }, + { + "epoch": 6.842105263157895, + "grad_norm": 0.0008420710398536633, + "learning_rate": 1.5619938263363164e-07, + "loss": 0.0, + "step": 10400 + }, + { + "epoch": 6.848684210526316, + "grad_norm": 0.0007561509749144412, + "learning_rate": 1.43511787913847e-07, + "loss": 0.0, + "step": 10410 + }, + { + "epoch": 6.855263157894737, + "grad_norm": 0.001082477090539546, + "learning_rate": 1.3136079578526805e-07, + "loss": 0.0, + "step": 10420 + }, + { + "epoch": 6.8618421052631575, + "grad_norm": 0.009062543301082302, + "learning_rate": 1.1974653702834237e-07, + "loss": 0.0, + "step": 10430 + }, + { + "epoch": 6.868421052631579, + "grad_norm": 0.0012818477749738891, + "learning_rate": 1.0866913664667744e-07, + "loss": 0.0, + "step": 10440 + }, + { + "epoch": 6.875, + "grad_norm": 0.0010660035685472302, + "learning_rate": 9.812871386571943e-08, + "loss": 0.0, + "step": 10450 + }, + { + "epoch": 6.881578947368421, + "grad_norm": 0.0018244941399890708, + "learning_rate": 8.812538213145427e-08, + "loss": 0.0, + "step": 10460 + }, + { + "epoch": 6.8881578947368425, + "grad_norm": 0.0008715384263849906, + "learning_rate": 7.865924910916977e-08, + "loss": 0.0, + "step": 10470 + }, + { + "epoch": 6.894736842105263, + "grad_norm": 0.0009966546785373075, + "learning_rate": 6.973041668233982e-08, + "loss": 0.0, + "step": 10480 + }, + { + "epoch": 6.901315789473684, + "grad_norm": 0.002128136793964254, + "learning_rate": 6.133898095146973e-08, + "loss": 0.0, + "step": 10490 + }, + { + "epoch": 6.907894736842105, + "grad_norm": 0.0014725600691340081, + "learning_rate": 5.3485032233130396e-08, + "loss": 0.0, + "step": 10500 + }, + { + "epoch": 6.9144736842105265, + "grad_norm": 0.0012527567537013562, + "learning_rate": 4.616865505892021e-08, + "loss": 0.0, + "step": 10510 + }, + { + "epoch": 6.921052631578947, + "grad_norm": 0.0009202347273378457, + "learning_rate": 3.9389928174604626e-08, + "loss": 0.0, + "step": 10520 + }, + { + "epoch": 6.927631578947368, + "grad_norm": 0.0009687718588937938, + "learning_rate": 3.314892453923357e-08, + "loss": 0.0, + "step": 10530 + }, + { + "epoch": 6.934210526315789, + "grad_norm": 0.0015585041930432386, + "learning_rate": 2.7445711324386448e-08, + "loss": 0.0, + "step": 10540 + }, + { + "epoch": 6.940789473684211, + "grad_norm": 0.0006957847207254594, + "learning_rate": 2.228034991342276e-08, + "loss": 0.0, + "step": 10550 + }, + { + "epoch": 6.947368421052632, + "grad_norm": 0.0010824703157524445, + "learning_rate": 1.7652895900843735e-08, + "loss": 0.0, + "step": 10560 + }, + { + "epoch": 6.953947368421053, + "grad_norm": 0.002329043577342889, + "learning_rate": 1.3563399091676143e-08, + "loss": 0.0, + "step": 10570 + }, + { + "epoch": 6.9605263157894735, + "grad_norm": 0.0007253364207954188, + "learning_rate": 1.001190350093939e-08, + "loss": 0.0, + "step": 10580 + }, + { + "epoch": 6.967105263157895, + "grad_norm": 0.002017700399902469, + "learning_rate": 6.998447353190329e-09, + "loss": 0.0, + "step": 10590 + }, + { + "epoch": 6.973684210526316, + "grad_norm": 0.0023433687131813766, + "learning_rate": 4.523063082084722e-09, + "loss": 0.0, + "step": 10600 + }, + { + "epoch": 6.980263157894737, + "grad_norm": 0.0008649734346166388, + "learning_rate": 2.5857773300441702e-09, + "loss": 0.0, + "step": 10610 + }, + { + "epoch": 6.9868421052631575, + "grad_norm": 0.00113742632817922, + "learning_rate": 1.1866109479674592e-09, + "loss": 0.0, + "step": 10620 + }, + { + "epoch": 6.993421052631579, + "grad_norm": 0.001963089912233135, + "learning_rate": 3.25578995002962e-10, + "loss": 0.0, + "step": 10630 + }, + { + "epoch": 7.0, + "grad_norm": 0.0014959587663241605, + "learning_rate": 2.690738393207326e-12, + "loss": 0.0, + "step": 10640 + }, + { + "epoch": 7.0, + "step": 10640, + "total_flos": 847436332400640.0, + "train_loss": 0.14730279242647062, + "train_runtime": 56644.8539, + "train_samples_per_second": 96.159, + "train_steps_per_second": 0.188 + } + ], + "logging_steps": 10, + "max_steps": 10640, + "num_input_tokens_seen": 0, + "num_train_epochs": 7, + "save_steps": 1800, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 847436332400640.0, + "train_batch_size": 128, + "trial_name": null, + "trial_params": null +}