FocusLLM: Scaling LLM's Context by Parallel Decoding

Re-assign community