██FR█████ █INTELL███████████
frenchintelligence.org
self-speculative
Meta LayerSkip Llama3.2 1B: Achieving Fast LLM Inference with Self-Speculative Decoding locally
October 31, 2024
