Thank you for this most excellent project !
I am surveying the state of the art for RAG and I just discovered LLMLingua which seems like an obviously good idea to me, to make the most of an LLM's context.
Do you think it would fit LLMware's scope to provide LLMLingua's context compression ?
Thank you for this most excellent project ! I am surveying the state of the art for RAG and I just discovered LLMLingua which seems like an obviously good idea to me, to make the most of an LLM's context.
Do you think it would fit LLMware's scope to provide LLMLingua's context compression ?
Best Regards