[willing to PR] Add Lookahead speculative decoding

[jjjjohnson]

Checklist

• 1. If the issue you raised is not a feature but a question, please raise a discussion at https://github.com/sgl-project/sglang/discussions/new/choose Otherwise, it will be closed.
• 2. Please use English, otherwise it will be closed.

Motivation

n-gram based speculative is very effective in retrieval augmented generation(RAG). The cost of generating draft tokens is relatively low compared to eagle and has a great potential for accelerating token generation in RAG. Ant group has proposed the Trie-based retrieval and verification mechanism. I want to adopt it to SGLang.

Related resources

Lookahead: An Inference Acceleration Framework for Large Language Model with Lossless Generation Accuracy

[jjjjohnson]

@zhyncs #2790