View of Sparse Experts Scale Better in Efficient Mixture Architectures for Trillion Parameter Models

Return to Article Details Sparse Experts Scale Better in Efficient Mixture Architectures for Trillion Parameter Models Download Download PDF