Closed statsu1990 closed 4 years ago
https://www.kaggle.com/seesee/faster-2x-tf-roberta
training time become about 2/3.
Model_v1_4_0 score 0.54476 (only posi and nega), implement remove_excessive_padding train only positive and negative label smoothing 0.05 lr 1e-5 different learning rate (x30)
https://www.kaggle.com/seesee/faster-2x-tf-roberta