Medusa: Simple Framework for Accelerating LLM Generation with Multiple Decoding Heads
No reviews for this project.