FastSeq
FastSeq provides efficient implementation of popular sequence models (e.g. Bart, ProphetNet) for text generation, summarization, translation tasks etc. It automatically optimizes inference speed based on popular NLP toolkits (e.g. FairSeq and HuggingFace-Transformers) without accuracy loss.
2021 MSR Asia Theory Workshop
Theory research has been the foundation and the driving force for the advancement of computer science and technology. The rapid technological development in the new era of big data and artificial intelligence, such as deep…
MSR Asia Theory Center
The mission of this center is to promote in-depth integration of theoretical research, big data, and artificial intelligence technology by building international academic exchange and cooperation hubs. While driving progress of theoretical research, the center…