InFeeo
Language

PuzzleMoE: Efficient Compression of Large Mixture-of-Experts Models(github.com)

×
Link preview GitHub - Supercomputing-System-AI-Lab/Supercomputing-System-AI-Lab.github.io: Website URL Website URL. Contribute to Supercomputing-System-AI-Lab/Supercomputing-System-AI-Lab.github.io development by creating an account on GitHub. GitHub · github.com
PuzzleMoE: Efficient Compression of Large Mixture-of-Experts Models via Sparse Expert Merging and Bit-packed inference

Comments

Log in Log in to comment.

No comments yet.