diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,3452 @@ +{ + "best_metric": 63.38577580698511, + "best_model_checkpoint": "./results_train_sae/roberta-base/coqa/checkpoint-26000", + "epoch": 9.999933007302204, + "global_step": 74630, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.07, + "learning_rate": 2.233139794551139e-06, + "loss": 5.815, + "step": 500 + }, + { + "epoch": 0.07, + "eval_HasAns_exact": 1.7049760040414246, + "eval_HasAns_f1": 13.104130654770286, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 1.5384615384615385, + "eval_NoAns_f1": 1.5384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 1.7036201929099337, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 12.997432860387208, + "eval_best_f1_thresh": 0.0, + "eval_exact": 1.7036201929099337, + "eval_f1": 13.009959479452721, + "eval_total": 7983, + "step": 500 + }, + { + "epoch": 0.13, + "learning_rate": 4.466279589102278e-06, + "loss": 4.3293, + "step": 1000 + }, + { + "epoch": 0.13, + "eval_HasAns_exact": 15.597373074008589, + "eval_HasAns_f1": 45.01124257660805, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 4.615384615384615, + "eval_NoAns_f1": 4.615384615384615, + "eval_NoAns_total": 65, + "eval_best_exact": 15.507954403106602, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 44.68232728567989, + "eval_best_f1_thresh": 0.0, + "eval_exact": 15.507954403106602, + "eval_f1": 44.68232728567989, + "eval_total": 7983, + "step": 1000 + }, + { + "epoch": 0.2, + "learning_rate": 6.699419383653417e-06, + "loss": 3.2185, + "step": 1500 + }, + { + "epoch": 0.2, + "eval_HasAns_exact": 20.35867643344279, + "eval_HasAns_f1": 52.56211238254352, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 4.615384615384615, + "eval_NoAns_f1": 4.615384615384615, + "eval_NoAns_total": 65, + "eval_best_exact": 20.23048979080546, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 52.171715626328414, + "eval_best_f1_thresh": 0.0, + "eval_exact": 20.23048979080546, + "eval_f1": 52.171715626328414, + "eval_total": 7983, + "step": 1500 + }, + { + "epoch": 0.27, + "learning_rate": 8.932559178204556e-06, + "loss": 2.8743, + "step": 2000 + }, + { + "epoch": 0.27, + "eval_HasAns_exact": 22.076281889366, + "eval_HasAns_f1": 53.668176491958135, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 10.76923076923077, + "eval_NoAns_f1": 10.76923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 21.984216459977453, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 53.31888030356063, + "eval_best_f1_thresh": 0.0, + "eval_exact": 21.984216459977453, + "eval_f1": 53.31888030356063, + "eval_total": 7983, + "step": 2000 + }, + { + "epoch": 0.33, + "learning_rate": 1.1165698972755696e-05, + "loss": 2.7084, + "step": 2500 + }, + { + "epoch": 0.33, + "eval_HasAns_exact": 22.960343521091186, + "eval_HasAns_f1": 56.52506128098628, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 0.0, + "eval_NoAns_f1": 0.0, + "eval_NoAns_total": 65, + "eval_best_exact": 22.77339346110485, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 56.064817139277096, + "eval_best_f1_thresh": 0.0, + "eval_exact": 22.77339346110485, + "eval_f1": 56.06481713927713, + "eval_total": 7983, + "step": 2500 + }, + { + "epoch": 0.4, + "learning_rate": 1.3398838767306834e-05, + "loss": 2.6029, + "step": 3000 + }, + { + "epoch": 0.4, + "eval_HasAns_exact": 24.74109623642334, + "eval_HasAns_f1": 58.33421249287827, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 1.5384615384615385, + "eval_NoAns_f1": 1.5384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 24.552173368407868, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 57.871764313993495, + "eval_best_f1_thresh": 0.0, + "eval_exact": 24.552173368407868, + "eval_f1": 57.87176431399351, + "eval_total": 7983, + "step": 3000 + }, + { + "epoch": 0.47, + "learning_rate": 1.5631978561857976e-05, + "loss": 2.5266, + "step": 3500 + }, + { + "epoch": 0.47, + "eval_HasAns_exact": 25.599898964384945, + "eval_HasAns_f1": 59.657728243287856, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 3.076923076923077, + "eval_NoAns_f1": 3.076923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 25.416510083928348, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 59.1970302180074, + "eval_best_f1_thresh": 0.0, + "eval_exact": 25.416510083928348, + "eval_f1": 59.19703021800742, + "eval_total": 7983, + "step": 3500 + }, + { + "epoch": 0.54, + "learning_rate": 1.786511835640911e-05, + "loss": 2.4933, + "step": 4000 + }, + { + "epoch": 0.54, + "eval_HasAns_exact": 25.978782520838596, + "eval_HasAns_f1": 58.7303362247268, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 3.076923076923077, + "eval_NoAns_f1": 3.076923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 25.792308655893773, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 58.27718930569793, + "eval_best_f1_thresh": 0.0, + "eval_exact": 25.792308655893773, + "eval_f1": 58.27718930569795, + "eval_total": 7983, + "step": 4000 + }, + { + "epoch": 0.6, + "learning_rate": 1.9993727905120314e-05, + "loss": 2.4258, + "step": 4500 + }, + { + "epoch": 0.6, + "eval_HasAns_exact": 26.799696893154838, + "eval_HasAns_f1": 59.664563121943544, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 10.76923076923077, + "eval_NoAns_f1": 10.76923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 26.669171990479768, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 59.266442540341856, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.669171990479768, + "eval_f1": 59.26644254034185, + "eval_total": 7983, + "step": 4500 + }, + { + "epoch": 0.67, + "learning_rate": 1.9851180294218272e-05, + "loss": 2.3915, + "step": 5000 + }, + { + "epoch": 0.67, + "eval_HasAns_exact": 27.68375852488002, + "eval_HasAns_f1": 60.812781351304864, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 6.153846153846154, + "eval_NoAns_f1": 6.153846153846154, + "eval_NoAns_total": 65, + "eval_best_exact": 27.508455467869222, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.367731772470485, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.508455467869222, + "eval_f1": 60.367731772470485, + "eval_total": 7983, + "step": 5000 + }, + { + "epoch": 0.74, + "learning_rate": 1.9708632683316228e-05, + "loss": 2.3553, + "step": 5500 + }, + { + "epoch": 0.74, + "eval_HasAns_exact": 27.44379893912604, + "eval_HasAns_f1": 60.88065108225846, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 6.153846153846154, + "eval_NoAns_f1": 6.153846153846154, + "eval_NoAns_total": 65, + "eval_best_exact": 27.27044970562445, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.43504888755135, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.27044970562445, + "eval_f1": 60.43504888755135, + "eval_total": 7983, + "step": 5500 + }, + { + "epoch": 0.8, + "learning_rate": 1.9566085072414186e-05, + "loss": 2.3258, + "step": 6000 + }, + { + "epoch": 0.8, + "eval_HasAns_exact": 27.36802222783531, + "eval_HasAns_f1": 59.84269145774818, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 7.6923076923076925, + "eval_NoAns_f1": 7.6923076923076925, + "eval_NoAns_total": 65, + "eval_best_exact": 27.20781661029688, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 59.41806726324064, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.20781661029688, + "eval_f1": 59.41806726324065, + "eval_total": 7983, + "step": 6000 + }, + { + "epoch": 0.87, + "learning_rate": 1.9423537461512148e-05, + "loss": 2.3185, + "step": 6500 + }, + { + "epoch": 0.87, + "eval_HasAns_exact": 27.69638797676181, + "eval_HasAns_f1": 60.38949697159613, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 7.6923076923076925, + "eval_NoAns_f1": 7.6923076923076925, + "eval_NoAns_total": 65, + "eval_best_exact": 27.53350870600025, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 59.96042052124492, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.53350870600025, + "eval_f1": 59.960420521244934, + "eval_total": 7983, + "step": 6500 + }, + { + "epoch": 0.94, + "learning_rate": 1.9280989850610107e-05, + "loss": 2.3023, + "step": 7000 + }, + { + "epoch": 0.94, + "eval_HasAns_exact": 28.06264208133367, + "eval_HasAns_f1": 61.52188273125215, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 13.846153846153847, + "eval_NoAns_f1": 13.846153846153847, + "eval_NoAns_total": 65, + "eval_best_exact": 27.94688713516222, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.13369252988283, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.94688713516222, + "eval_f1": 61.13369252988285, + "eval_total": 7983, + "step": 7000 + }, + { + "epoch": 1.0, + "learning_rate": 1.9138442239708065e-05, + "loss": 2.2535, + "step": 7500 + }, + { + "epoch": 1.0, + "eval_HasAns_exact": 27.69638797676181, + "eval_HasAns_f1": 60.61431606118244, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 27.608668420393336, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.25856878021324, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.608668420393336, + "eval_f1": 60.25856878021324, + "eval_total": 7983, + "step": 7500 + }, + { + "epoch": 1.07, + "learning_rate": 1.899589462880602e-05, + "loss": 2.0785, + "step": 8000 + }, + { + "epoch": 1.07, + "eval_HasAns_exact": 29.50239959585754, + "eval_HasAns_f1": 62.435922189166604, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 1.5384615384615385, + "eval_NoAns_f1": 1.5384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 29.274708756106726, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.94007664960809, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.274708756106726, + "eval_f1": 61.940076649608066, + "eval_total": 7983, + "step": 8000 + }, + { + "epoch": 1.14, + "learning_rate": 1.8853347017903983e-05, + "loss": 2.1234, + "step": 8500 + }, + { + "epoch": 1.14, + "eval_HasAns_exact": 28.73200303106845, + "eval_HasAns_f1": 60.71174380961002, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 7.6923076923076925, + "eval_NoAns_f1": 7.6923076923076925, + "eval_NoAns_total": 65, + "eval_best_exact": 28.560691469372415, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.28004352805865, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.560691469372415, + "eval_f1": 60.28004352805865, + "eval_total": 7983, + "step": 8500 + }, + { + "epoch": 1.21, + "learning_rate": 1.871079940700194e-05, + "loss": 2.1137, + "step": 9000 + }, + { + "epoch": 1.21, + "eval_HasAns_exact": 28.30260166708765, + "eval_HasAns_f1": 61.55707249426806, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 4.615384615384615, + "eval_NoAns_f1": 4.615384615384615, + "eval_NoAns_total": 65, + "eval_best_exact": 28.109733183013905, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.09343605281405, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.109733183013905, + "eval_f1": 61.09343605281405, + "eval_total": 7983, + "step": 9000 + }, + { + "epoch": 1.27, + "learning_rate": 1.85682517960999e-05, + "loss": 2.1102, + "step": 9500 + }, + { + "epoch": 1.27, + "eval_HasAns_exact": 28.83303864612276, + "eval_HasAns_f1": 62.037918749762206, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 4.615384615384615, + "eval_NoAns_f1": 4.615384615384615, + "eval_NoAns_total": 65, + "eval_best_exact": 28.635851183765503, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.57036711269161, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.635851183765503, + "eval_f1": 61.57036711269161, + "eval_total": 7983, + "step": 9500 + }, + { + "epoch": 1.34, + "learning_rate": 1.8425704185197858e-05, + "loss": 2.0702, + "step": 10000 + }, + { + "epoch": 1.34, + "eval_HasAns_exact": 29.022480424349585, + "eval_HasAns_f1": 62.00584905483293, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 3.076923076923077, + "eval_NoAns_f1": 3.076923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 28.8112238506827, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.526031919850574, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.8112238506827, + "eval_f1": 61.526031919850574, + "eval_total": 7983, + "step": 10000 + }, + { + "epoch": 1.41, + "learning_rate": 1.8283156574295817e-05, + "loss": 2.0792, + "step": 10500 + }, + { + "epoch": 1.41, + "eval_HasAns_exact": 29.401363980803232, + "eval_HasAns_f1": 61.87601216131859, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 10.76923076923077, + "eval_NoAns_f1": 10.76923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 29.249655517975697, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.459885292912475, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.249655517975697, + "eval_f1": 61.45988529291251, + "eval_total": 7983, + "step": 10500 + }, + { + "epoch": 1.47, + "learning_rate": 1.8140608963393775e-05, + "loss": 2.0611, + "step": 11000 + }, + { + "epoch": 1.47, + "eval_HasAns_exact": 30.18438999747411, + "eval_HasAns_f1": 63.41954494271191, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 6.153846153846154, + "eval_NoAns_f1": 6.153846153846154, + "eval_NoAns_total": 65, + "eval_best_exact": 29.988726042841037, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.953270306450314, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.988726042841037, + "eval_f1": 62.953270306450314, + "eval_total": 7983, + "step": 11000 + }, + { + "epoch": 1.54, + "learning_rate": 1.7998061352491734e-05, + "loss": 2.0993, + "step": 11500 + }, + { + "epoch": 1.54, + "eval_HasAns_exact": 28.51730234907805, + "eval_HasAns_f1": 61.39637957288941, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 4.615384615384615, + "eval_NoAns_f1": 4.615384615384615, + "eval_NoAns_total": 65, + "eval_best_exact": 28.322685707127647, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.93405154179362, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.322685707127647, + "eval_f1": 60.93405154179361, + "eval_total": 7983, + "step": 11500 + }, + { + "epoch": 1.61, + "learning_rate": 1.7855513741589693e-05, + "loss": 2.0957, + "step": 12000 + }, + { + "epoch": 1.61, + "eval_HasAns_exact": 28.769891386713816, + "eval_HasAns_f1": 61.95532272922007, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 3.076923076923077, + "eval_NoAns_f1": 3.076923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 28.560691469372415, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.475916994859624, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.560691469372415, + "eval_f1": 61.47591699485965, + "eval_total": 7983, + "step": 12000 + }, + { + "epoch": 1.67, + "learning_rate": 1.771296613068765e-05, + "loss": 2.045, + "step": 12500 + }, + { + "epoch": 1.67, + "eval_HasAns_exact": 29.50239959585754, + "eval_HasAns_f1": 62.67004766472256, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 4.615384615384615, + "eval_NoAns_f1": 4.615384615384615, + "eval_NoAns_total": 65, + "eval_best_exact": 29.299761994237755, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.19734904287529, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.299761994237755, + "eval_f1": 62.197349042875274, + "eval_total": 7983, + "step": 12500 + }, + { + "epoch": 1.74, + "learning_rate": 1.757041851978561e-05, + "loss": 2.0545, + "step": 13000 + }, + { + "epoch": 1.74, + "eval_HasAns_exact": 29.325587269512503, + "eval_HasAns_f1": 61.454527824893695, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 9.23076923076923, + "eval_NoAns_f1": 9.23076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 29.161969184517098, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.02930619034303, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.161969184517098, + "eval_f1": 61.02930619034303, + "eval_total": 7983, + "step": 13000 + }, + { + "epoch": 1.81, + "learning_rate": 1.742787090888357e-05, + "loss": 2.0459, + "step": 13500 + }, + { + "epoch": 1.81, + "eval_HasAns_exact": 28.74463248295024, + "eval_HasAns_f1": 60.669764383172534, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 12.307692307692308, + "eval_NoAns_f1": 12.307692307692308, + "eval_NoAns_total": 65, + "eval_best_exact": 28.610797945634474, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.275985768002, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.610797945634474, + "eval_f1": 60.275985768002016, + "eval_total": 7983, + "step": 13500 + }, + { + "epoch": 1.88, + "learning_rate": 1.7285323297981527e-05, + "loss": 2.0593, + "step": 14000 + }, + { + "epoch": 1.88, + "eval_HasAns_exact": 29.12351603940389, + "eval_HasAns_f1": 62.16330627308465, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 10.76923076923077, + "eval_NoAns_f1": 10.76923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 28.974069898534385, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.74484016914496, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.974069898534385, + "eval_f1": 61.744840169144965, + "eval_total": 7983, + "step": 14000 + }, + { + "epoch": 1.94, + "learning_rate": 1.7142775687079486e-05, + "loss": 2.0704, + "step": 14500 + }, + { + "epoch": 1.94, + "eval_HasAns_exact": 29.679211922202576, + "eval_HasAns_f1": 62.73972470457179, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 7.6923076923076925, + "eval_NoAns_f1": 7.6923076923076925, + "eval_NoAns_total": 65, + "eval_best_exact": 29.500187899285983, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.29151198932729, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.500187899285983, + "eval_f1": 62.29151198932727, + "eval_total": 7983, + "step": 14500 + }, + { + "epoch": 2.01, + "learning_rate": 1.7000228076177444e-05, + "loss": 2.0229, + "step": 15000 + }, + { + "epoch": 2.01, + "eval_HasAns_exact": 29.666582470320787, + "eval_HasAns_f1": 62.5911632254183, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 9.23076923076923, + "eval_NoAns_f1": 9.23076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 29.500187899285983, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.15668676172642, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.500187899285983, + "eval_f1": 62.156686761726434, + "eval_total": 7983, + "step": 15000 + }, + { + "epoch": 2.08, + "learning_rate": 1.6857680465275403e-05, + "loss": 1.8093, + "step": 15500 + }, + { + "epoch": 2.08, + "eval_HasAns_exact": 29.780247537256884, + "eval_HasAns_f1": 62.61910716735094, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 9.23076923076923, + "eval_NoAns_f1": 9.23076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 29.61292747087561, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.18440317563384, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.61292747087561, + "eval_f1": 62.18440317563382, + "eval_total": 7983, + "step": 15500 + }, + { + "epoch": 2.14, + "learning_rate": 1.671513285437336e-05, + "loss": 1.8153, + "step": 16000 + }, + { + "epoch": 2.14, + "eval_HasAns_exact": 29.527658499621115, + "eval_HasAns_f1": 62.33254202102029, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 10.76923076923077, + "eval_NoAns_f1": 10.76923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 29.37492170863084, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.91269794844527, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.37492170863084, + "eval_f1": 61.91269794844528, + "eval_total": 7983, + "step": 16000 + }, + { + "epoch": 2.21, + "learning_rate": 1.657258524347132e-05, + "loss": 1.8511, + "step": 16500 + }, + { + "epoch": 2.21, + "eval_HasAns_exact": 30.209648901237685, + "eval_HasAns_f1": 62.630251022506265, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 7.6923076923076925, + "eval_NoAns_f1": 7.6923076923076925, + "eval_NoAns_total": 65, + "eval_best_exact": 30.02630590003758, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.182929675085134, + "eval_best_f1_thresh": 0.0, + "eval_exact": 30.02630590003758, + "eval_f1": 62.18292967508515, + "eval_total": 7983, + "step": 16500 + }, + { + "epoch": 2.28, + "learning_rate": 1.643003763256928e-05, + "loss": 1.8308, + "step": 17000 + }, + { + "epoch": 2.28, + "eval_HasAns_exact": 29.919171507956555, + "eval_HasAns_f1": 62.44215649796323, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 29.83840661405487, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.09657962556345, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.83840661405487, + "eval_f1": 62.09657962556343, + "eval_total": 7983, + "step": 17000 + }, + { + "epoch": 2.34, + "learning_rate": 1.6287490021667237e-05, + "loss": 1.8285, + "step": 17500 + }, + { + "epoch": 2.34, + "eval_HasAns_exact": 29.12351603940389, + "eval_HasAns_f1": 61.87050711385868, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 7.6923076923076925, + "eval_NoAns_f1": 7.6923076923076925, + "eval_NoAns_total": 65, + "eval_best_exact": 28.949016660403355, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.42937183108268, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.949016660403355, + "eval_f1": 61.429371831082676, + "eval_total": 7983, + "step": 17500 + }, + { + "epoch": 2.41, + "learning_rate": 1.6144942410765196e-05, + "loss": 1.825, + "step": 18000 + }, + { + "epoch": 2.41, + "eval_HasAns_exact": 30.133872189946956, + "eval_HasAns_f1": 63.075006423503396, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 3.076923076923077, + "eval_NoAns_f1": 3.076923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 29.913566328447953, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.58648388592007, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.913566328447953, + "eval_f1": 62.58648388592007, + "eval_total": 7983, + "step": 18000 + }, + { + "epoch": 2.48, + "learning_rate": 1.6002394799863154e-05, + "loss": 1.8206, + "step": 18500 + }, + { + "epoch": 2.48, + "eval_HasAns_exact": 29.540287951502904, + "eval_HasAns_f1": 62.238845295007174, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 12.307692307692308, + "eval_NoAns_f1": 12.307692307692308, + "eval_NoAns_total": 65, + "eval_best_exact": 29.39997494676187, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.83229074857407, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.39997494676187, + "eval_f1": 61.832290748574046, + "eval_total": 7983, + "step": 18500 + }, + { + "epoch": 2.55, + "learning_rate": 1.5859847188961116e-05, + "loss": 1.8535, + "step": 19000 + }, + { + "epoch": 2.55, + "eval_HasAns_exact": 30.525385198282393, + "eval_HasAns_f1": 62.69640983742568, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 12.307692307692308, + "eval_NoAns_f1": 12.307692307692308, + "eval_NoAns_total": 65, + "eval_best_exact": 30.377051233871978, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.28612966212413, + "eval_best_f1_thresh": 0.0, + "eval_exact": 30.377051233871978, + "eval_f1": 62.28612966212409, + "eval_total": 7983, + "step": 19000 + }, + { + "epoch": 2.61, + "learning_rate": 1.571729957805907e-05, + "loss": 1.8286, + "step": 19500 + }, + { + "epoch": 2.61, + "eval_HasAns_exact": 30.272796160646628, + "eval_HasAns_f1": 63.614261536130094, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 10.76923076923077, + "eval_NoAns_f1": 10.76923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 30.11399223349618, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 63.1839813156806, + "eval_best_f1_thresh": 0.0, + "eval_exact": 30.11399223349618, + "eval_f1": 63.1839813156806, + "eval_total": 7983, + "step": 19500 + }, + { + "epoch": 2.68, + "learning_rate": 1.557475196715703e-05, + "loss": 1.8518, + "step": 20000 + }, + { + "epoch": 2.68, + "eval_HasAns_exact": 30.601161909573126, + "eval_HasAns_f1": 62.78914687552758, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 9.23076923076923, + "eval_NoAns_f1": 9.23076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 30.427157710134036, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.353058369087755, + "eval_best_f1_thresh": 0.0, + "eval_exact": 30.427157710134036, + "eval_f1": 62.35305836908774, + "eval_total": 7983, + "step": 20000 + }, + { + "epoch": 2.75, + "learning_rate": 1.543220435625499e-05, + "loss": 1.807, + "step": 20500 + }, + { + "epoch": 2.75, + "eval_HasAns_exact": 30.26016670876484, + "eval_HasAns_f1": 63.007803537175604, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 9.23076923076923, + "eval_NoAns_f1": 9.23076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 30.08893899536515, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.56993466207649, + "eval_best_f1_thresh": 0.0, + "eval_exact": 30.08893899536515, + "eval_f1": 62.56993466207647, + "eval_total": 7983, + "step": 20500 + }, + { + "epoch": 2.81, + "learning_rate": 1.528965674535295e-05, + "loss": 1.8148, + "step": 21000 + }, + { + "epoch": 2.81, + "eval_HasAns_exact": 30.33594342005557, + "eval_HasAns_f1": 62.23993269669792, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 15.384615384615385, + "eval_NoAns_f1": 15.384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 30.214205186020294, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.85842253444247, + "eval_best_f1_thresh": 0.0, + "eval_exact": 30.214205186020294, + "eval_f1": 61.85842253444246, + "eval_total": 7983, + "step": 21000 + }, + { + "epoch": 2.88, + "learning_rate": 1.5147109134450907e-05, + "loss": 1.812, + "step": 21500 + }, + { + "epoch": 2.88, + "eval_HasAns_exact": 30.967416014144987, + "eval_HasAns_f1": 63.42882011828027, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 9.23076923076923, + "eval_NoAns_f1": 9.23076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 30.790429663033947, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.98752319886554, + "eval_best_f1_thresh": 0.0, + "eval_exact": 30.790429663033947, + "eval_f1": 62.9875231988655, + "eval_total": 7983, + "step": 21500 + }, + { + "epoch": 2.95, + "learning_rate": 1.5004561523548866e-05, + "loss": 1.8223, + "step": 22000 + }, + { + "epoch": 2.95, + "eval_HasAns_exact": 30.083354382419802, + "eval_HasAns_f1": 62.2735874535647, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 7.6923076923076925, + "eval_NoAns_f1": 7.6923076923076925, + "eval_NoAns_total": 65, + "eval_best_exact": 29.901039709382438, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.8291701687743, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.901039709382438, + "eval_f1": 61.82917016877431, + "eval_total": 7983, + "step": 22000 + }, + { + "epoch": 3.01, + "learning_rate": 1.4862013912646825e-05, + "loss": 1.7384, + "step": 22500 + }, + { + "epoch": 3.01, + "eval_HasAns_exact": 31.232634503662542, + "eval_HasAns_f1": 63.69478284476774, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 13.846153846153847, + "eval_NoAns_f1": 13.846153846153847, + "eval_NoAns_total": 65, + "eval_best_exact": 31.09106852060629, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 63.28890023360531, + "eval_best_f1_thresh": 0.0, + "eval_exact": 31.09106852060629, + "eval_f1": 63.288900233605304, + "eval_total": 7983, + "step": 22500 + }, + { + "epoch": 3.08, + "learning_rate": 1.4719466301744785e-05, + "loss": 1.6147, + "step": 23000 + }, + { + "epoch": 3.08, + "eval_HasAns_exact": 30.550644102045972, + "eval_HasAns_f1": 62.705334193629774, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 9.23076923076923, + "eval_NoAns_f1": 9.23076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 30.377051233871978, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.2699281153903, + "eval_best_f1_thresh": 0.0, + "eval_exact": 30.377051233871978, + "eval_f1": 62.26992811539028, + "eval_total": 7983, + "step": 23000 + }, + { + "epoch": 3.15, + "learning_rate": 1.4576918690842744e-05, + "loss": 1.607, + "step": 23500 + }, + { + "epoch": 3.15, + "eval_HasAns_exact": 30.298055064410203, + "eval_HasAns_f1": 62.13669507615566, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 10.76923076923077, + "eval_NoAns_f1": 10.76923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 30.139045471627206, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.718445648628396, + "eval_best_f1_thresh": 0.0, + "eval_exact": 30.139045471627206, + "eval_f1": 61.718445648628375, + "eval_total": 7983, + "step": 23500 + }, + { + "epoch": 3.22, + "learning_rate": 1.44343710799407e-05, + "loss": 1.632, + "step": 24000 + }, + { + "epoch": 3.22, + "eval_HasAns_exact": 30.777974235918162, + "eval_HasAns_f1": 63.47221823426325, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 15.384615384615385, + "eval_NoAns_f1": 15.384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 30.65263685331329, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 63.080674430526926, + "eval_best_f1_thresh": 0.0, + "eval_exact": 30.65263685331329, + "eval_f1": 63.080674430526926, + "eval_total": 7983, + "step": 24000 + }, + { + "epoch": 3.28, + "learning_rate": 1.4291823469038659e-05, + "loss": 1.619, + "step": 24500 + }, + { + "epoch": 3.28, + "eval_HasAns_exact": 30.39909067946451, + "eval_HasAns_f1": 63.232701395640525, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 10.76923076923077, + "eval_NoAns_f1": 10.76923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 30.23925842415132, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.80552795323582, + "eval_best_f1_thresh": 0.0, + "eval_exact": 30.23925842415132, + "eval_f1": 62.80552795323583, + "eval_total": 7983, + "step": 24500 + }, + { + "epoch": 3.35, + "learning_rate": 1.414927585813662e-05, + "loss": 1.6286, + "step": 25000 + }, + { + "epoch": 3.35, + "eval_HasAns_exact": 30.09598383430159, + "eval_HasAns_f1": 63.15857502527025, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 10.76923076923077, + "eval_NoAns_f1": 10.76923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 29.938619566578982, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.73200514218837, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.938619566578982, + "eval_f1": 62.73200514218839, + "eval_total": 7983, + "step": 25000 + }, + { + "epoch": 3.42, + "learning_rate": 1.4006728247234578e-05, + "loss": 1.6041, + "step": 25500 + }, + { + "epoch": 3.42, + "eval_HasAns_exact": 28.795150290477395, + "eval_HasAns_f1": 60.893375347584964, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 13.846153846153847, + "eval_NoAns_f1": 13.846153846153847, + "eval_NoAns_total": 65, + "eval_best_exact": 28.673431040962043, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.510302643389444, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.673431040962043, + "eval_f1": 60.510302643389444, + "eval_total": 7983, + "step": 25500 + }, + { + "epoch": 3.48, + "learning_rate": 1.3864180636332536e-05, + "loss": 1.6267, + "step": 26000 + }, + { + "epoch": 3.48, + "eval_HasAns_exact": 30.954786562263198, + "eval_HasAns_f1": 63.80508313553448, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 12.307692307692308, + "eval_NoAns_f1": 12.307692307692308, + "eval_NoAns_total": 65, + "eval_best_exact": 30.802956282099462, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 63.38577580698511, + "eval_best_f1_thresh": 0.0, + "eval_exact": 30.802956282099462, + "eval_f1": 63.38577580698511, + "eval_total": 7983, + "step": 26000 + }, + { + "epoch": 3.55, + "learning_rate": 1.3721633025430493e-05, + "loss": 1.6306, + "step": 26500 + }, + { + "epoch": 3.55, + "eval_HasAns_exact": 29.616064662793633, + "eval_HasAns_f1": 61.97374766852039, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 29.512714518351498, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.606931484322224, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.512714518351498, + "eval_f1": 61.606931484322224, + "eval_total": 7983, + "step": 26500 + }, + { + "epoch": 3.62, + "learning_rate": 1.3579085414528455e-05, + "loss": 1.6283, + "step": 27000 + }, + { + "epoch": 3.62, + "eval_HasAns_exact": 29.376105077039657, + "eval_HasAns_f1": 61.19213550940523, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 29.274708756106726, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.831683447760305, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.274708756106726, + "eval_f1": 60.83168344776032, + "eval_total": 7983, + "step": 27000 + }, + { + "epoch": 3.68, + "learning_rate": 1.3436537803626412e-05, + "loss": 1.5994, + "step": 27500 + }, + { + "epoch": 3.68, + "eval_HasAns_exact": 30.133872189946956, + "eval_HasAns_f1": 63.16981457364355, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 23.076923076923077, + "eval_NoAns_f1": 23.076923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 30.076412376299636, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.84336612728418, + "eval_best_f1_thresh": 0.0, + "eval_exact": 30.076412376299636, + "eval_f1": 62.84336612728418, + "eval_total": 7983, + "step": 27500 + }, + { + "epoch": 3.75, + "learning_rate": 1.329399019272437e-05, + "loss": 1.6278, + "step": 28000 + }, + { + "epoch": 3.75, + "eval_HasAns_exact": 30.1970194493559, + "eval_HasAns_f1": 62.47879079785238, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 30.101465614430666, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.12038901883942, + "eval_best_f1_thresh": 0.0, + "eval_exact": 30.101465614430666, + "eval_f1": 62.1203890188394, + "eval_total": 7983, + "step": 28000 + }, + { + "epoch": 3.82, + "learning_rate": 1.315144258182233e-05, + "loss": 1.6349, + "step": 28500 + }, + { + "epoch": 3.82, + "eval_HasAns_exact": 30.361202323819146, + "eval_HasAns_f1": 62.83391227199402, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 15.384615384615385, + "eval_NoAns_f1": 15.384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 30.23925842415132, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.44756574842154, + "eval_best_f1_thresh": 0.0, + "eval_exact": 30.23925842415132, + "eval_f1": 62.4475657484215, + "eval_total": 7983, + "step": 28500 + }, + { + "epoch": 3.89, + "learning_rate": 1.300889497092029e-05, + "loss": 1.6193, + "step": 29000 + }, + { + "epoch": 3.89, + "eval_HasAns_exact": 29.856024248547612, + "eval_HasAns_f1": 62.14444702264984, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 24.615384615384617, + "eval_NoAns_f1": 24.615384615384617, + "eval_NoAns_total": 65, + "eval_best_exact": 29.81335337592384, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.83887404801973, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.81335337592384, + "eval_f1": 61.83887404801975, + "eval_total": 7983, + "step": 29000 + }, + { + "epoch": 3.95, + "learning_rate": 1.2866347360018248e-05, + "loss": 1.6105, + "step": 29500 + }, + { + "epoch": 3.95, + "eval_HasAns_exact": 29.628694114675422, + "eval_HasAns_f1": 61.29193297196648, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 29.537767756482527, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.94319494826893, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.537767756482527, + "eval_f1": 60.94319494826892, + "eval_total": 7983, + "step": 29500 + }, + { + "epoch": 4.02, + "learning_rate": 1.2723799749116205e-05, + "loss": 1.5694, + "step": 30000 + }, + { + "epoch": 4.02, + "eval_HasAns_exact": 29.26244001010356, + "eval_HasAns_f1": 62.08659690195635, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 15.384615384615385, + "eval_NoAns_f1": 15.384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 29.149442565451583, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.70633524610929, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.149442565451583, + "eval_f1": 61.706335246109276, + "eval_total": 7983, + "step": 30000 + }, + { + "epoch": 4.09, + "learning_rate": 1.2581252138214164e-05, + "loss": 1.393, + "step": 30500 + }, + { + "epoch": 4.09, + "eval_HasAns_exact": 28.84566809800455, + "eval_HasAns_f1": 60.84476292028231, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 27.692307692307693, + "eval_NoAns_f1": 27.692307692307693, + "eval_NoAns_total": 65, + "eval_best_exact": 28.83627708881373, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.574825604759546, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.83627708881373, + "eval_f1": 60.57482560475953, + "eval_total": 7983, + "step": 30500 + }, + { + "epoch": 4.15, + "learning_rate": 1.2438704527312124e-05, + "loss": 1.4271, + "step": 31000 + }, + { + "epoch": 4.15, + "eval_HasAns_exact": 29.312957817630714, + "eval_HasAns_f1": 61.54559872571804, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 10.76923076923077, + "eval_NoAns_f1": 10.76923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 29.161969184517098, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.1321621834192, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.161969184517098, + "eval_f1": 61.1321621834192, + "eval_total": 7983, + "step": 31000 + }, + { + "epoch": 4.22, + "learning_rate": 1.2296156916410083e-05, + "loss": 1.3853, + "step": 31500 + }, + { + "epoch": 4.22, + "eval_HasAns_exact": 29.50239959585754, + "eval_HasAns_f1": 62.97737162382639, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 29.412501565827384, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.61491024896122, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.412501565827384, + "eval_f1": 62.61491024896122, + "eval_total": 7983, + "step": 31500 + }, + { + "epoch": 4.29, + "learning_rate": 1.2153609305508041e-05, + "loss": 1.4217, + "step": 32000 + }, + { + "epoch": 4.29, + "eval_HasAns_exact": 30.058095478656227, + "eval_HasAns_f1": 62.53589980633206, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 10.76923076923077, + "eval_NoAns_f1": 10.76923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 29.901039709382438, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.114399933175186, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.901039709382438, + "eval_f1": 62.11439993317516, + "eval_total": 7983, + "step": 32000 + }, + { + "epoch": 4.35, + "learning_rate": 1.2011061694605998e-05, + "loss": 1.4187, + "step": 32500 + }, + { + "epoch": 4.35, + "eval_HasAns_exact": 28.719373579186662, + "eval_HasAns_f1": 61.3851264651957, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 13.846153846153847, + "eval_NoAns_f1": 13.846153846153847, + "eval_NoAns_total": 65, + "eval_best_exact": 28.59827132656896, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.998049774698686, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.59827132656896, + "eval_f1": 60.998049774698686, + "eval_total": 7983, + "step": 32500 + }, + { + "epoch": 4.42, + "learning_rate": 1.1868514083703958e-05, + "loss": 1.4188, + "step": 33000 + }, + { + "epoch": 4.42, + "eval_HasAns_exact": 29.085627683758524, + "eval_HasAns_f1": 60.907411553755075, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 21.53846153846154, + "eval_NoAns_f1": 21.53846153846154, + "eval_NoAns_total": 65, + "eval_best_exact": 29.024176374796443, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.586857657852036, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.024176374796443, + "eval_f1": 60.58685765785204, + "eval_total": 7983, + "step": 33000 + }, + { + "epoch": 4.49, + "learning_rate": 1.1725966472801917e-05, + "loss": 1.4111, + "step": 33500 + }, + { + "epoch": 4.49, + "eval_HasAns_exact": 28.95933316494064, + "eval_HasAns_f1": 61.442457819826295, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 28.861330326944756, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.079967558234344, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.861330326944756, + "eval_f1": 61.07996755823433, + "eval_total": 7983, + "step": 33500 + }, + { + "epoch": 4.56, + "learning_rate": 1.1583418861899876e-05, + "loss": 1.4481, + "step": 34000 + }, + { + "epoch": 4.56, + "eval_HasAns_exact": 29.022480424349585, + "eval_HasAns_f1": 61.57168144088351, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 28.949016660403355, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.23319223962364, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.949016660403355, + "eval_f1": 61.23319223962366, + "eval_total": 7983, + "step": 34000 + }, + { + "epoch": 4.62, + "learning_rate": 1.1440871250997834e-05, + "loss": 1.4128, + "step": 34500 + }, + { + "epoch": 4.62, + "eval_HasAns_exact": 29.590805759030058, + "eval_HasAns_f1": 62.65998240295622, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 15.384615384615385, + "eval_NoAns_f1": 15.384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 29.475134661154954, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.275052068972485, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.475134661154954, + "eval_f1": 62.275052068972485, + "eval_total": 7983, + "step": 34500 + }, + { + "epoch": 4.69, + "learning_rate": 1.1298323640095791e-05, + "loss": 1.4304, + "step": 35000 + }, + { + "epoch": 4.69, + "eval_HasAns_exact": 28.36574892649659, + "eval_HasAns_f1": 60.973142910638266, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 13.846153846153847, + "eval_NoAns_f1": 13.846153846153847, + "eval_NoAns_total": 65, + "eval_best_exact": 28.247525992734563, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.58942071482324, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.247525992734563, + "eval_f1": 60.58942071482324, + "eval_total": 7983, + "step": 35000 + }, + { + "epoch": 4.76, + "learning_rate": 1.1155776029193753e-05, + "loss": 1.39, + "step": 35500 + }, + { + "epoch": 4.76, + "eval_HasAns_exact": 28.656226319777723, + "eval_HasAns_f1": 60.36715986425469, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 28.560691469372415, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.01342500377909, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.560691469372415, + "eval_f1": 60.01342500377909, + "eval_total": 7983, + "step": 35500 + }, + { + "epoch": 4.82, + "learning_rate": 1.101322841829171e-05, + "loss": 1.4253, + "step": 36000 + }, + { + "epoch": 4.82, + "eval_HasAns_exact": 28.97196261682243, + "eval_HasAns_f1": 62.01226582184054, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 13.846153846153847, + "eval_NoAns_f1": 13.846153846153847, + "eval_NoAns_total": 65, + "eval_best_exact": 28.84880370787924, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.620082773059465, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.84880370787924, + "eval_f1": 61.62008277305945, + "eval_total": 7983, + "step": 36000 + }, + { + "epoch": 4.89, + "learning_rate": 1.0870680807389669e-05, + "loss": 1.4483, + "step": 36500 + }, + { + "epoch": 4.89, + "eval_HasAns_exact": 29.161404395049257, + "eval_HasAns_f1": 61.70284974349661, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 12.307692307692308, + "eval_NoAns_f1": 12.307692307692308, + "eval_NoAns_total": 65, + "eval_best_exact": 29.024176374796443, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.30065943492496, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.024176374796443, + "eval_f1": 61.30065943492498, + "eval_total": 7983, + "step": 36500 + }, + { + "epoch": 4.96, + "learning_rate": 1.0728133196487627e-05, + "loss": 1.4238, + "step": 37000 + }, + { + "epoch": 4.96, + "eval_HasAns_exact": 29.527658499621115, + "eval_HasAns_f1": 62.567053116696016, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 29.450081423023924, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.22045929825868, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.450081423023924, + "eval_f1": 62.22045929825868, + "eval_total": 7983, + "step": 37000 + }, + { + "epoch": 5.02, + "learning_rate": 1.0585585585585587e-05, + "loss": 1.3436, + "step": 37500 + }, + { + "epoch": 5.02, + "eval_HasAns_exact": 29.04773932811316, + "eval_HasAns_f1": 61.950658410403086, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 28.96154327946887, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.59655684499205, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.96154327946887, + "eval_f1": 61.59655684499204, + "eval_total": 7983, + "step": 37500 + }, + { + "epoch": 5.09, + "learning_rate": 1.0443037974683546e-05, + "loss": 1.1974, + "step": 38000 + }, + { + "epoch": 5.09, + "eval_HasAns_exact": 28.74463248295024, + "eval_HasAns_f1": 61.519208674273074, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 10.76923076923077, + "eval_NoAns_f1": 10.76923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 28.59827132656896, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.10598700775327, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.59827132656896, + "eval_f1": 61.10598700775327, + "eval_total": 7983, + "step": 38000 + }, + { + "epoch": 5.16, + "learning_rate": 1.0300490363781503e-05, + "loss": 1.2054, + "step": 38500 + }, + { + "epoch": 5.16, + "eval_HasAns_exact": 28.80777974235918, + "eval_HasAns_f1": 61.8546952173498, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 28.736064136289617, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.51390163234069, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.736064136289617, + "eval_f1": 61.51390163234068, + "eval_total": 7983, + "step": 38500 + }, + { + "epoch": 5.23, + "learning_rate": 1.0157942752879461e-05, + "loss": 1.2236, + "step": 39000 + }, + { + "epoch": 5.23, + "eval_HasAns_exact": 28.934074261177066, + "eval_HasAns_f1": 62.09249599293992, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 28.84880370787924, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.73723954304126, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.84880370787924, + "eval_f1": 61.73723954304126, + "eval_total": 7983, + "step": 39000 + }, + { + "epoch": 5.29, + "learning_rate": 1.0015395141977422e-05, + "loss": 1.2453, + "step": 39500 + }, + { + "epoch": 5.29, + "eval_HasAns_exact": 27.039656478908814, + "eval_HasAns_f1": 59.47666176407304, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 26.957284228986598, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 59.13017760840914, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.957284228986598, + "eval_f1": 59.130177608409156, + "eval_total": 7983, + "step": 39500 + }, + { + "epoch": 5.36, + "learning_rate": 9.87284753107538e-06, + "loss": 1.2198, + "step": 40000 + }, + { + "epoch": 5.36, + "eval_HasAns_exact": 28.403637282141954, + "eval_HasAns_f1": 60.841241678751864, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 23.076923076923077, + "eval_NoAns_f1": 23.076923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 28.360265564324187, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.53375317704588, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.360265564324187, + "eval_f1": 60.53375317704588, + "eval_total": 7983, + "step": 40000 + }, + { + "epoch": 5.43, + "learning_rate": 9.730299920173339e-06, + "loss": 1.2317, + "step": 40500 + }, + { + "epoch": 5.43, + "eval_HasAns_exact": 28.327860570851225, + "eval_HasAns_f1": 61.33584674035582, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 28.260052611800074, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.99927777654232, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.260052611800074, + "eval_f1": 60.99927777654235, + "eval_total": 7983, + "step": 40500 + }, + { + "epoch": 5.49, + "learning_rate": 9.587752309271297e-06, + "loss": 1.2371, + "step": 41000 + }, + { + "epoch": 5.49, + "eval_HasAns_exact": 28.66885577165951, + "eval_HasAns_f1": 61.37009490052371, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 10.76923076923077, + "eval_NoAns_f1": 10.76923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 28.523111612175875, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.958087363440626, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.523111612175875, + "eval_f1": 60.958087363440654, + "eval_total": 7983, + "step": 41000 + }, + { + "epoch": 5.56, + "learning_rate": 9.445204698369256e-06, + "loss": 1.217, + "step": 41500 + }, + { + "epoch": 5.56, + "eval_HasAns_exact": 27.532205102298562, + "eval_HasAns_f1": 59.78731908137268, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 27.458348991607163, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 59.450832078956395, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.458348991607163, + "eval_f1": 59.45083207895641, + "eval_total": 7983, + "step": 41500 + }, + { + "epoch": 5.63, + "learning_rate": 9.302657087467215e-06, + "loss": 1.2183, + "step": 42000 + }, + { + "epoch": 5.63, + "eval_HasAns_exact": 27.44379893912604, + "eval_HasAns_f1": 60.57543204614133, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 12.307692307692308, + "eval_NoAns_f1": 12.307692307692308, + "eval_NoAns_total": 65, + "eval_best_exact": 27.32055618188651, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.182421513384305, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.32055618188651, + "eval_f1": 60.18242151338433, + "eval_total": 7983, + "step": 42000 + }, + { + "epoch": 5.69, + "learning_rate": 9.160109476565173e-06, + "loss": 1.217, + "step": 42500 + }, + { + "epoch": 5.69, + "eval_HasAns_exact": 27.342763324071736, + "eval_HasAns_f1": 60.53898331335163, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 27.27044970562445, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.19637603346089, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.27044970562445, + "eval_f1": 60.19637603346091, + "eval_total": 7983, + "step": 42500 + }, + { + "epoch": 5.76, + "learning_rate": 9.017561865663132e-06, + "loss": 1.22, + "step": 43000 + }, + { + "epoch": 5.76, + "eval_HasAns_exact": 28.264713311442282, + "eval_HasAns_f1": 61.49391622017227, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 15.384615384615385, + "eval_NoAns_f1": 15.384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 28.15983965927596, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.118480349658526, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.15983965927596, + "eval_f1": 61.11848034965854, + "eval_total": 7983, + "step": 43000 + }, + { + "epoch": 5.83, + "learning_rate": 8.87501425476109e-06, + "loss": 1.2229, + "step": 43500 + }, + { + "epoch": 5.83, + "eval_HasAns_exact": 27.69638797676181, + "eval_HasAns_f1": 60.97226471814465, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 15.384615384615385, + "eval_NoAns_f1": 15.384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 27.59614180132782, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.60107629190396, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.59614180132782, + "eval_f1": 60.60107629190396, + "eval_total": 7983, + "step": 43500 + }, + { + "epoch": 5.9, + "learning_rate": 8.73246664385905e-06, + "loss": 1.2165, + "step": 44000 + }, + { + "epoch": 5.9, + "eval_HasAns_exact": 28.44152563778732, + "eval_HasAns_f1": 61.72099334388884, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 21.53846153846154, + "eval_NoAns_f1": 21.53846153846154, + "eval_NoAns_total": 65, + "eval_best_exact": 28.385318802455217, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.39381501902939, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.385318802455217, + "eval_f1": 61.39381501902941, + "eval_total": 7983, + "step": 44000 + }, + { + "epoch": 5.96, + "learning_rate": 8.589919032957008e-06, + "loss": 1.2222, + "step": 44500 + }, + { + "epoch": 5.96, + "eval_HasAns_exact": 29.249810558221775, + "eval_HasAns_f1": 62.21338028810192, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 15.384615384615385, + "eval_NoAns_f1": 15.384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 29.136915946386072, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.83208632358649, + "eval_best_f1_thresh": 0.0, + "eval_exact": 29.136915946386072, + "eval_f1": 61.83208632358649, + "eval_total": 7983, + "step": 44500 + }, + { + "epoch": 6.03, + "learning_rate": 8.447371422054968e-06, + "loss": 1.1451, + "step": 45000 + }, + { + "epoch": 6.03, + "eval_HasAns_exact": 28.087900985097246, + "eval_HasAns_f1": 61.32313276181045, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 28.022046849555306, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.98666731905489, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.022046849555306, + "eval_f1": 60.98666731905489, + "eval_total": 7983, + "step": 45000 + }, + { + "epoch": 6.1, + "learning_rate": 8.304823811152925e-06, + "loss": 1.0617, + "step": 45500 + }, + { + "epoch": 6.1, + "eval_HasAns_exact": 28.567820156605205, + "eval_HasAns_f1": 62.15445512423891, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 10.76923076923077, + "eval_NoAns_f1": 10.76923076923077, + "eval_NoAns_total": 65, + "eval_best_exact": 28.42289865965176, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.73606108902966, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.42289865965176, + "eval_f1": 61.73606108902965, + "eval_total": 7983, + "step": 45500 + }, + { + "epoch": 6.16, + "learning_rate": 8.162276200250885e-06, + "loss": 1.0341, + "step": 46000 + }, + { + "epoch": 6.16, + "eval_HasAns_exact": 27.1659509977267, + "eval_HasAns_f1": 60.83537787468107, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 15.384615384615385, + "eval_NoAns_f1": 15.384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 27.070023800576223, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.46530402251345, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.070023800576223, + "eval_f1": 60.46530402251344, + "eval_total": 7983, + "step": 46000 + }, + { + "epoch": 6.23, + "learning_rate": 8.019728589348844e-06, + "loss": 1.0539, + "step": 46500 + }, + { + "epoch": 6.23, + "eval_HasAns_exact": 27.494316746653194, + "eval_HasAns_f1": 61.29569471460783, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 13.846153846153847, + "eval_NoAns_f1": 13.846153846153847, + "eval_NoAns_total": 65, + "eval_best_exact": 27.38318927721408, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.90934620446764, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.38318927721408, + "eval_f1": 60.90934620446762, + "eval_total": 7983, + "step": 46500 + }, + { + "epoch": 6.3, + "learning_rate": 7.877180978446802e-06, + "loss": 1.0382, + "step": 47000 + }, + { + "epoch": 6.3, + "eval_HasAns_exact": 28.28997221520586, + "eval_HasAns_f1": 62.86439578497228, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 12.307692307692308, + "eval_NoAns_f1": 12.307692307692308, + "eval_NoAns_total": 65, + "eval_best_exact": 28.15983965927596, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 62.45274781728808, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.15983965927596, + "eval_f1": 62.45274781728808, + "eval_total": 7983, + "step": 47000 + }, + { + "epoch": 6.36, + "learning_rate": 7.73463336754476e-06, + "loss": 1.0584, + "step": 47500 + }, + { + "epoch": 6.36, + "eval_HasAns_exact": 27.999494821924728, + "eval_HasAns_f1": 61.68090507422947, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 13.846153846153847, + "eval_NoAns_f1": 13.846153846153847, + "eval_NoAns_total": 65, + "eval_best_exact": 27.884254039834648, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.29142006485642, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.884254039834648, + "eval_f1": 61.29142006485643, + "eval_total": 7983, + "step": 47500 + }, + { + "epoch": 6.43, + "learning_rate": 7.592085756642719e-06, + "loss": 1.0485, + "step": 48000 + }, + { + "epoch": 6.43, + "eval_HasAns_exact": 27.67112907299823, + "eval_HasAns_f1": 61.60241941642461, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 27.583615182262307, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.23862669914194, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.583615182262307, + "eval_f1": 61.23862669914192, + "eval_total": 7983, + "step": 48000 + }, + { + "epoch": 6.5, + "learning_rate": 7.449538145740678e-06, + "loss": 1.0589, + "step": 48500 + }, + { + "epoch": 6.5, + "eval_HasAns_exact": 27.46905784288962, + "eval_HasAns_f1": 61.88739573740814, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 27.395715896279594, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.53380927581083, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.395715896279594, + "eval_f1": 61.533809275810825, + "eval_total": 7983, + "step": 48500 + }, + { + "epoch": 6.57, + "learning_rate": 7.3069905348386374e-06, + "loss": 1.0665, + "step": 49000 + }, + { + "epoch": 6.57, + "eval_HasAns_exact": 27.22909825713564, + "eval_HasAns_f1": 60.76848064123057, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 27.170236753100337, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.436531343763434, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.170236753100337, + "eval_f1": 60.436531343763455, + "eval_total": 7983, + "step": 49000 + }, + { + "epoch": 6.63, + "learning_rate": 7.164442923936595e-06, + "loss": 1.053, + "step": 49500 + }, + { + "epoch": 6.63, + "eval_HasAns_exact": 26.91336196009093, + "eval_HasAns_f1": 60.13083722243696, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 21.53846153846154, + "eval_NoAns_f1": 21.53846153846154, + "eval_NoAns_total": 65, + "eval_best_exact": 26.869597895527995, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 59.816606429569816, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.869597895527995, + "eval_f1": 59.81660642956982, + "eval_total": 7983, + "step": 49500 + }, + { + "epoch": 6.7, + "learning_rate": 7.021895313034555e-06, + "loss": 1.06, + "step": 50000 + }, + { + "epoch": 6.7, + "eval_HasAns_exact": 27.885829754988634, + "eval_HasAns_f1": 62.14884731940757, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 27.79656770637605, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.780605420903086, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.79656770637605, + "eval_f1": 61.780605420903065, + "eval_total": 7983, + "step": 50000 + }, + { + "epoch": 6.77, + "learning_rate": 6.879347702132513e-06, + "loss": 1.0512, + "step": 50500 + }, + { + "epoch": 6.77, + "eval_HasAns_exact": 28.06264208133367, + "eval_HasAns_f1": 61.61980422643438, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 13.846153846153847, + "eval_NoAns_f1": 13.846153846153847, + "eval_NoAns_total": 65, + "eval_best_exact": 27.94688713516222, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.23081671864056, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.94688713516222, + "eval_f1": 61.23081671864054, + "eval_total": 7983, + "step": 50500 + }, + { + "epoch": 6.83, + "learning_rate": 6.736800091230472e-06, + "loss": 1.0668, + "step": 51000 + }, + { + "epoch": 6.83, + "eval_HasAns_exact": 27.709017428643598, + "eval_HasAns_f1": 61.393100322120155, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 27.62119503945885, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.03101194419986, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.62119503945885, + "eval_f1": 61.03101194419984, + "eval_total": 7983, + "step": 51000 + }, + { + "epoch": 6.9, + "learning_rate": 6.59425248032843e-06, + "loss": 1.0614, + "step": 51500 + }, + { + "epoch": 6.9, + "eval_HasAns_exact": 28.07527153321546, + "eval_HasAns_f1": 61.74371990867002, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 15.384615384615385, + "eval_NoAns_f1": 15.384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 27.971940373293247, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.36625006098573, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.971940373293247, + "eval_f1": 61.36625006098573, + "eval_total": 7983, + "step": 51500 + }, + { + "epoch": 6.97, + "learning_rate": 6.451704869426388e-06, + "loss": 1.0667, + "step": 52000 + }, + { + "epoch": 6.97, + "eval_HasAns_exact": 28.113159888860825, + "eval_HasAns_f1": 61.571783458538675, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 28.022046849555306, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.20824018848921, + "eval_best_f1_thresh": 0.0, + "eval_exact": 28.022046849555306, + "eval_f1": 61.2082401884892, + "eval_total": 7983, + "step": 52000 + }, + { + "epoch": 7.03, + "learning_rate": 6.3091572585243475e-06, + "loss": 0.9903, + "step": 52500 + }, + { + "epoch": 7.03, + "eval_HasAns_exact": 26.951250315736296, + "eval_HasAns_f1": 61.03543304135232, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 13.846153846153847, + "eval_NoAns_f1": 13.846153846153847, + "eval_NoAns_total": 65, + "eval_best_exact": 26.84454465739697, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.6512036604569, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.84454465739697, + "eval_f1": 60.651203660456915, + "eval_total": 7983, + "step": 52500 + }, + { + "epoch": 7.1, + "learning_rate": 6.166609647622306e-06, + "loss": 0.9175, + "step": 53000 + }, + { + "epoch": 7.1, + "eval_HasAns_exact": 26.610255114928012, + "eval_HasAns_f1": 60.023096663321915, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 26.531379180759114, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 59.672163269470495, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.531379180759114, + "eval_f1": 59.67216326947051, + "eval_total": 7983, + "step": 53000 + }, + { + "epoch": 7.17, + "learning_rate": 6.024062036720265e-06, + "loss": 0.9086, + "step": 53500 + }, + { + "epoch": 7.17, + "eval_HasAns_exact": 26.723920181864106, + "eval_HasAns_f1": 61.377884403219866, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 26.644118752348742, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.01591991791246, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.644118752348742, + "eval_f1": 61.01591991791245, + "eval_total": 7983, + "step": 53500 + }, + { + "epoch": 7.24, + "learning_rate": 5.881514425818223e-06, + "loss": 0.9074, + "step": 54000 + }, + { + "epoch": 7.24, + "eval_HasAns_exact": 27.102803738317757, + "eval_HasAns_f1": 61.465795058045565, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 27.019917324314168, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.103114777603004, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.019917324314168, + "eval_f1": 61.10311477760302, + "eval_total": 7983, + "step": 54000 + }, + { + "epoch": 7.3, + "learning_rate": 5.738966814916183e-06, + "loss": 0.9207, + "step": 54500 + }, + { + "epoch": 7.3, + "eval_HasAns_exact": 26.24400101035615, + "eval_HasAns_f1": 60.521848057930576, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 26.19316046599023, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.19190691753654, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.19316046599023, + "eval_f1": 60.191906917536556, + "eval_total": 7983, + "step": 54500 + }, + { + "epoch": 7.37, + "learning_rate": 5.5964192040141405e-06, + "loss": 0.9353, + "step": 55000 + }, + { + "epoch": 7.37, + "eval_HasAns_exact": 27.01439757514524, + "eval_HasAns_f1": 61.02819548653803, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 15.384615384615385, + "eval_NoAns_f1": 15.384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 26.919704371790054, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.656551655068036, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.919704371790054, + "eval_f1": 60.656551655068036, + "eval_total": 7983, + "step": 55000 + }, + { + "epoch": 7.44, + "learning_rate": 5.4538715931121e-06, + "loss": 0.9489, + "step": 55500 + }, + { + "epoch": 7.44, + "eval_HasAns_exact": 27.15332154584491, + "eval_HasAns_f1": 61.258275393661116, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 13.846153846153847, + "eval_NoAns_f1": 13.846153846153847, + "eval_NoAns_total": 65, + "eval_best_exact": 27.044970562445197, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.872231562947356, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.044970562445197, + "eval_f1": 60.87223156294736, + "eval_total": 7983, + "step": 55500 + }, + { + "epoch": 7.5, + "learning_rate": 5.3113239822100585e-06, + "loss": 0.9305, + "step": 56000 + }, + { + "epoch": 7.5, + "eval_HasAns_exact": 27.279616064662793, + "eval_HasAns_f1": 61.65225517730881, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 13.846153846153847, + "eval_NoAns_f1": 13.846153846153847, + "eval_NoAns_total": 65, + "eval_best_exact": 27.170236753100337, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.26300344405998, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.170236753100337, + "eval_f1": 61.26300344406, + "eval_total": 7983, + "step": 56000 + }, + { + "epoch": 7.57, + "learning_rate": 5.168776371308017e-06, + "loss": 0.9303, + "step": 56500 + }, + { + "epoch": 7.57, + "eval_HasAns_exact": 26.521848951755494, + "eval_HasAns_f1": 60.37027080526025, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 26.443692847300515, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.016510614562264, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.443692847300515, + "eval_f1": 60.01651061456227, + "eval_total": 7983, + "step": 56500 + }, + { + "epoch": 7.64, + "learning_rate": 5.026228760405976e-06, + "loss": 0.9449, + "step": 57000 + }, + { + "epoch": 7.64, + "eval_HasAns_exact": 27.39328113159889, + "eval_HasAns_f1": 62.14238343281087, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 27.308029562820995, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.77419416522567, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.308029562820995, + "eval_f1": 61.77419416522567, + "eval_total": 7983, + "step": 57000 + }, + { + "epoch": 7.7, + "learning_rate": 4.883681149503935e-06, + "loss": 0.9141, + "step": 57500 + }, + { + "epoch": 7.7, + "eval_HasAns_exact": 26.534478403637284, + "eval_HasAns_f1": 60.29428221993628, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 24.615384615384617, + "eval_NoAns_f1": 24.615384615384617, + "eval_NoAns_total": 65, + "eval_best_exact": 26.5188525616936, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.003773846605945, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.5188525616936, + "eval_f1": 60.003773846605945, + "eval_total": 7983, + "step": 57500 + }, + { + "epoch": 7.77, + "learning_rate": 4.741133538601894e-06, + "loss": 0.9199, + "step": 58000 + }, + { + "epoch": 7.77, + "eval_HasAns_exact": 26.976509219499874, + "eval_HasAns_f1": 61.379243481659564, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 26.90717775272454, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.02979454939003, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.90717775272454, + "eval_f1": 61.02979454939003, + "eval_total": 7983, + "step": 58000 + }, + { + "epoch": 7.84, + "learning_rate": 4.598585927699852e-06, + "loss": 0.9338, + "step": 58500 + }, + { + "epoch": 7.84, + "eval_HasAns_exact": 26.307148269765094, + "eval_HasAns_f1": 60.72553048550691, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 26.243266942252287, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.38140428212999, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.243266942252287, + "eval_f1": 60.38140428213, + "eval_total": 7983, + "step": 58500 + }, + { + "epoch": 7.91, + "learning_rate": 4.456038316797811e-06, + "loss": 0.9204, + "step": 59000 + }, + { + "epoch": 7.91, + "eval_HasAns_exact": 27.36802222783531, + "eval_HasAns_f1": 62.021263208727596, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 27.29550294375548, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.66658675769827, + "eval_best_f1_thresh": 0.0, + "eval_exact": 27.29550294375548, + "eval_f1": 61.666586757698255, + "eval_total": 7983, + "step": 59000 + }, + { + "epoch": 7.97, + "learning_rate": 4.3134907058957695e-06, + "loss": 0.9351, + "step": 59500 + }, + { + "epoch": 7.97, + "eval_HasAns_exact": 26.799696893154838, + "eval_HasAns_f1": 61.2660477382399, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 26.73180508580734, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.917520479942816, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.73180508580734, + "eval_f1": 60.91752047994283, + "eval_total": 7983, + "step": 59500 + }, + { + "epoch": 8.04, + "learning_rate": 4.170943094993729e-06, + "loss": 0.8545, + "step": 60000 + }, + { + "epoch": 8.04, + "eval_HasAns_exact": 27.01439757514524, + "eval_HasAns_f1": 61.314199148141626, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 26.957284228986598, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.97780644557003, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.957284228986598, + "eval_f1": 60.977806445570046, + "eval_total": 7983, + "step": 60000 + }, + { + "epoch": 8.11, + "learning_rate": 4.028395484091687e-06, + "loss": 0.8234, + "step": 60500 + }, + { + "epoch": 8.11, + "eval_HasAns_exact": 26.761808537509474, + "eval_HasAns_f1": 62.35576302120297, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 15.384615384615385, + "eval_NoAns_f1": 15.384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 26.669171990479768, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.97330973341917, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.669171990479768, + "eval_f1": 61.97330973341915, + "eval_total": 7983, + "step": 60500 + }, + { + "epoch": 8.17, + "learning_rate": 3.885847873189645e-06, + "loss": 0.8427, + "step": 61000 + }, + { + "epoch": 8.17, + "eval_HasAns_exact": 26.09244758777469, + "eval_HasAns_f1": 60.43409077659736, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 26.01778779907303, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.07981094439411, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.01778779907303, + "eval_f1": 60.079810944394104, + "eval_total": 7983, + "step": 61000 + }, + { + "epoch": 8.24, + "learning_rate": 3.7433002622876042e-06, + "loss": 0.8261, + "step": 61500 + }, + { + "epoch": 8.24, + "eval_HasAns_exact": 26.47133114422834, + "eval_HasAns_f1": 61.219688212113844, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 26.393586371038456, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.85901180803175, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.393586371038456, + "eval_f1": 60.85901180803177, + "eval_total": 7983, + "step": 61500 + }, + { + "epoch": 8.31, + "learning_rate": 3.600752651385563e-06, + "loss": 0.8344, + "step": 62000 + }, + { + "epoch": 8.31, + "eval_HasAns_exact": 26.54710785551907, + "eval_HasAns_f1": 61.43442445472298, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 15.384615384615385, + "eval_NoAns_f1": 15.384615384615385, + "eval_NoAns_total": 65, + "eval_best_exact": 26.45621946636603, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.05947298415341, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.45621946636603, + "eval_f1": 61.059472984153395, + "eval_total": 7983, + "step": 62000 + }, + { + "epoch": 8.37, + "learning_rate": 3.4582050404835214e-06, + "loss": 0.813, + "step": 62500 + }, + { + "epoch": 8.37, + "eval_HasAns_exact": 27.00176812326345, + "eval_HasAns_f1": 62.09603511891683, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 26.919704371790054, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.72822323331875, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.919704371790054, + "eval_f1": 61.72822323331873, + "eval_total": 7983, + "step": 62500 + }, + { + "epoch": 8.44, + "learning_rate": 3.3156574295814804e-06, + "loss": 0.8116, + "step": 63000 + }, + { + "epoch": 8.44, + "eval_HasAns_exact": 26.761808537509474, + "eval_HasAns_f1": 61.72187754602157, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 26.681698609545283, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.357112164524445, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.681698609545283, + "eval_f1": 61.35711216452446, + "eval_total": 7983, + "step": 63000 + }, + { + "epoch": 8.51, + "learning_rate": 3.173109818679439e-06, + "loss": 0.8212, + "step": 63500 + }, + { + "epoch": 8.51, + "eval_HasAns_exact": 26.862844152563778, + "eval_HasAns_f1": 62.14075244477736, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 26.806964800200426, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.79762969532096, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.806964800200426, + "eval_f1": 61.79762969532096, + "eval_total": 7983, + "step": 63500 + }, + { + "epoch": 8.58, + "learning_rate": 3.0305622077773976e-06, + "loss": 0.8222, + "step": 64000 + }, + { + "epoch": 8.58, + "eval_HasAns_exact": 26.723920181864106, + "eval_HasAns_f1": 61.42653113262713, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 26.656645371414257, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.07669717000397, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.656645371414257, + "eval_f1": 61.07669717000398, + "eval_total": 7983, + "step": 64000 + }, + { + "epoch": 8.64, + "learning_rate": 2.8880145968753566e-06, + "loss": 0.8205, + "step": 64500 + }, + { + "epoch": 8.64, + "eval_HasAns_exact": 26.33240717352867, + "eval_HasAns_f1": 60.97293815916608, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 26.280846799448828, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.63932410676149, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.280846799448828, + "eval_f1": 60.63932410676149, + "eval_total": 7983, + "step": 64500 + }, + { + "epoch": 8.71, + "learning_rate": 2.745466985973315e-06, + "loss": 0.8301, + "step": 65000 + }, + { + "epoch": 8.71, + "eval_HasAns_exact": 26.572366759282648, + "eval_HasAns_f1": 61.023201280062096, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 26.506325942628084, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.67665135106249, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.506325942628084, + "eval_f1": 60.676651351062475, + "eval_total": 7983, + "step": 65000 + }, + { + "epoch": 8.78, + "learning_rate": 2.6029193750712738e-06, + "loss": 0.8147, + "step": 65500 + }, + { + "epoch": 8.78, + "eval_HasAns_exact": 26.597625663046223, + "eval_HasAns_f1": 61.61202867789058, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 26.5188525616936, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.248157719095246, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.5188525616936, + "eval_f1": 61.248157719095246, + "eval_total": 7983, + "step": 65500 + }, + { + "epoch": 8.84, + "learning_rate": 2.4603717641692328e-06, + "loss": 0.829, + "step": 66000 + }, + { + "epoch": 8.84, + "eval_HasAns_exact": 26.117706491538268, + "eval_HasAns_f1": 61.29633370563857, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 16.923076923076923, + "eval_NoAns_f1": 16.923076923076923, + "eval_NoAns_total": 65, + "eval_best_exact": 26.04284103720406, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.9350332307712, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.04284103720406, + "eval_f1": 60.93503323077118, + "eval_total": 7983, + "step": 66000 + }, + { + "epoch": 8.91, + "learning_rate": 2.3178241532671914e-06, + "loss": 0.8161, + "step": 66500 + }, + { + "epoch": 8.91, + "eval_HasAns_exact": 26.54710785551907, + "eval_HasAns_f1": 61.4951492974776, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 26.481272704497055, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.14475662500658, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.481272704497055, + "eval_f1": 61.14475662500659, + "eval_total": 7983, + "step": 66500 + }, + { + "epoch": 8.98, + "learning_rate": 2.17527654236515e-06, + "loss": 0.8365, + "step": 67000 + }, + { + "epoch": 8.98, + "eval_HasAns_exact": 26.572366759282648, + "eval_HasAns_f1": 61.82650791887594, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 26.506325942628084, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.473417224309124, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.506325942628084, + "eval_f1": 61.473417224309124, + "eval_total": 7983, + "step": 67000 + }, + { + "epoch": 9.04, + "learning_rate": 2.032728931463109e-06, + "loss": 0.7749, + "step": 67500 + }, + { + "epoch": 9.04, + "eval_HasAns_exact": 26.433442788582976, + "eval_HasAns_f1": 61.4165049966057, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 18.46153846153846, + "eval_NoAns_f1": 18.46153846153846, + "eval_NoAns_total": 65, + "eval_best_exact": 26.368533132907427, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.06675266981385, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.368533132907427, + "eval_f1": 61.06675266981385, + "eval_total": 7983, + "step": 67500 + }, + { + "epoch": 9.11, + "learning_rate": 1.8901813205610676e-06, + "loss": 0.7611, + "step": 68000 + }, + { + "epoch": 9.11, + "eval_HasAns_exact": 26.117706491538268, + "eval_HasAns_f1": 61.777210346593094, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 26.067894275335085, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.43704766683255, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.067894275335085, + "eval_f1": 61.43704766683255, + "eval_total": 7983, + "step": 68000 + }, + { + "epoch": 9.18, + "learning_rate": 1.7476337096590261e-06, + "loss": 0.7326, + "step": 68500 + }, + { + "epoch": 9.18, + "eval_HasAns_exact": 26.117706491538268, + "eval_HasAns_f1": 61.245581761717254, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 21.53846153846154, + "eval_NoAns_f1": 21.53846153846154, + "eval_NoAns_total": 65, + "eval_best_exact": 26.0804208944006, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.922274381720825, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.0804208944006, + "eval_f1": 60.922274381720825, + "eval_total": 7983, + "step": 68500 + }, + { + "epoch": 9.25, + "learning_rate": 1.605086098756985e-06, + "loss": 0.7493, + "step": 69000 + }, + { + "epoch": 9.25, + "eval_HasAns_exact": 26.193483202828997, + "eval_HasAns_f1": 61.03326897772823, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 26.143053989728173, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.69916369355534, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.143053989728173, + "eval_f1": 60.699163693555334, + "eval_total": 7983, + "step": 69000 + }, + { + "epoch": 9.31, + "learning_rate": 1.4625384878549437e-06, + "loss": 0.7394, + "step": 69500 + }, + { + "epoch": 9.31, + "eval_HasAns_exact": 26.307148269765094, + "eval_HasAns_f1": 61.2582740205435, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 26.2557935613178, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.922336677272156, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.2557935613178, + "eval_f1": 60.92233667727215, + "eval_total": 7983, + "step": 69500 + }, + { + "epoch": 9.38, + "learning_rate": 1.3199908769529023e-06, + "loss": 0.7467, + "step": 70000 + }, + { + "epoch": 9.38, + "eval_HasAns_exact": 25.928264713311442, + "eval_HasAns_f1": 61.08285196853556, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 25.879994989352372, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.74834296465798, + "eval_best_f1_thresh": 0.0, + "eval_exact": 25.879994989352372, + "eval_f1": 60.74834296465798, + "eval_total": 7983, + "step": 70000 + }, + { + "epoch": 9.45, + "learning_rate": 1.1774432660508611e-06, + "loss": 0.7446, + "step": 70500 + }, + { + "epoch": 9.45, + "eval_HasAns_exact": 26.130335943420057, + "eval_HasAns_f1": 61.32879120974875, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 26.0804208944006, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.99227969419903, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.0804208944006, + "eval_f1": 60.99227969419903, + "eval_total": 7983, + "step": 70500 + }, + { + "epoch": 9.51, + "learning_rate": 1.03489565514882e-06, + "loss": 0.7519, + "step": 71000 + }, + { + "epoch": 9.51, + "eval_HasAns_exact": 26.269259914119726, + "eval_HasAns_f1": 61.611796711729426, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 26.218213704121258, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.2729808798038, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.218213704121258, + "eval_f1": 61.2729808798038, + "eval_total": 7983, + "step": 71000 + }, + { + "epoch": 9.58, + "learning_rate": 8.923480442467784e-07, + "loss": 0.7437, + "step": 71500 + }, + { + "epoch": 9.58, + "eval_HasAns_exact": 26.446072240464765, + "eval_HasAns_f1": 61.45579299519043, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 26.393586371038456, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.11824739269922, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.393586371038456, + "eval_f1": 61.11824739269923, + "eval_total": 7983, + "step": 71500 + }, + { + "epoch": 9.65, + "learning_rate": 7.498004333447371e-07, + "loss": 0.7584, + "step": 72000 + }, + { + "epoch": 9.65, + "eval_HasAns_exact": 26.01667087648396, + "eval_HasAns_f1": 61.29681697148508, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 25.967681322810975, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.960565799852056, + "eval_best_f1_thresh": 0.0, + "eval_exact": 25.967681322810975, + "eval_f1": 60.960565799852056, + "eval_total": 7983, + "step": 72000 + }, + { + "epoch": 9.71, + "learning_rate": 6.072528224426959e-07, + "loss": 0.7528, + "step": 72500 + }, + { + "epoch": 9.71, + "eval_HasAns_exact": 26.33240717352867, + "eval_HasAns_f1": 61.431235411450984, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 26.280846799448828, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.09388976423263, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.280846799448828, + "eval_f1": 61.09388976423263, + "eval_total": 7983, + "step": 72500 + }, + { + "epoch": 9.78, + "learning_rate": 4.647052115406546e-07, + "loss": 0.7434, + "step": 73000 + }, + { + "epoch": 9.78, + "eval_HasAns_exact": 26.408183884819397, + "eval_HasAns_f1": 61.50068023326903, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 26.356006513841916, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.16276914531179, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.356006513841916, + "eval_f1": 61.16276914531181, + "eval_total": 7983, + "step": 73000 + }, + { + "epoch": 9.85, + "learning_rate": 3.2215760063861334e-07, + "loss": 0.7397, + "step": 73500 + }, + { + "epoch": 9.85, + "eval_HasAns_exact": 26.01667087648396, + "eval_HasAns_f1": 61.14120383462663, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 25.967681322810975, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.80621971221019, + "eval_best_f1_thresh": 0.0, + "eval_exact": 25.967681322810975, + "eval_f1": 60.806219712210186, + "eval_total": 7983, + "step": 73500 + }, + { + "epoch": 9.92, + "learning_rate": 1.7960998973657203e-07, + "loss": 0.7467, + "step": 74000 + }, + { + "epoch": 9.92, + "eval_HasAns_exact": 26.25663046223794, + "eval_HasAns_f1": 61.308967980407346, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 26.205687085055743, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 60.9726178715853, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.205687085055743, + "eval_f1": 60.9726178715853, + "eval_total": 7983, + "step": 74000 + }, + { + "epoch": 9.98, + "learning_rate": 3.706237883453074e-08, + "loss": 0.7369, + "step": 74500 + }, + { + "epoch": 9.98, + "eval_HasAns_exact": 26.294518817883304, + "eval_HasAns_f1": 61.406322256852434, + "eval_HasAns_total": 7918, + "eval_NoAns_exact": 20.0, + "eval_NoAns_f1": 20.0, + "eval_NoAns_total": 65, + "eval_best_exact": 26.243266942252287, + "eval_best_exact_thresh": 0.0, + "eval_best_f1": 61.06917946007236, + "eval_best_f1_thresh": 0.0, + "eval_exact": 26.243266942252287, + "eval_f1": 61.06917946007235, + "eval_total": 7983, + "step": 74500 + }, + { + "epoch": 10.0, + "step": 74630, + "total_flos": 3.120233413081866e+17, + "train_loss": 1.4580501551796152, + "train_runtime": 74852.5694, + "train_samples_per_second": 15.953, + "train_steps_per_second": 0.997 + } + ], + "max_steps": 74630, + "num_train_epochs": 10, + "total_flos": 3.120233413081866e+17, + "trial_name": null, + "trial_params": null +}