Gotta read more about the approaches for tokenization / training of #LLM tech in Chinese. Seems like there's enough difference that there'd be some clever techniques. Anyone have links to stuff they've read? (any language)