no code implementations • 17 Oct 2020 • Çağla Aksoy, Alper Ahmetoğlu, Tunga Güngör
In this work, we adopt hierarchical multitask learning approaches for BERT pre-training.
1 code implementation • 5 Nov 2019 • Alper Ahmetoğlu, Ethem Alpaydin
There is also a discriminator that is trained to discriminate such fake samples from true samples of the distribution; at the same time, the generator is trained to generate fakes that the discriminator cannot tell apart from the true samples.