HF RL Explorer

I want from nlg yongzhuo import mmr to expose a summarizer that returns a non-redundant ranking of…

I want from nlg yongzhuo import mmr to expose a summarizer that returns a non-redundant ranking of…: a task in MiMo-V2.6-RL-harbor-code: MiMo-V2.6-RL Code (Harbor) (Harbor dataset). The alpha weight should control the tradeoff between sentence importance and redundancy, so higher values favor…

The task

The `alpha` weight should control the tradeoff between sentence importance and redundancy, so higher values favor relevance and lower values favor diversity. If the input is not a string or a sentence list, it should raise `RuntimeError` instead of coercing it. Repeated sentence text should not appear more than once…

Part of FineEnvs/MiMo-V2.6-RL-harbor-code.