AtomicBot-ai/woodcock
Lossless speculative decoding for Apple Silicon (MLX): 2× faster Qwen3 generation on a Mac with a trained hidden-state drafter. MIT.
GitHub repository with 11 stars and 2 forks.
Language: Python
Topics: apple-silicon, draft-model, eagle, llm-inference, macos, mlx, qwen3, speculative-decoding