Last released Sep 11, 2023
Simple Framework for Accelerating LLM Generation with Multiple Decoding Heads
Supported by