…ST models trained with an addition of the corpus obtain new state-of-the-art results on the MuST-C English-German benchmark test set.
5 PAPERS • NO BENCHMARKS YET