Draft& Verify: Lossless Large Language Model Acceleration via Self-Speculative Decoding
- Award ID(s):
- 2008993
- PAR ID:
- 10562500
- Publisher / Repository:
- Association for Computational Linguistics
- Date Published:
- Page Range / eLocation ID:
- 11263 to 11282
- Format(s):
- Medium: X
- Location:
- Bangkok, Thailand
- Sponsoring Org:
- National Science Foundation
More Like this
No document suggestions found
An official website of the United States government

