The docs for register type strategy say that instead of registering tuple int , I can register tuple …
The docs for register type strategy say that instead of registering tuple int , I can register tuple …: a task in MiMo-V2.6-RL-oss: Agentic RL Environments (MiMo RL release). The docs for register type strategy say that instead of registering tuple int , I can register tuple with a function that…
The task
The docs for register_type_strategy say that instead of registering tuple[int], I can register tuple with a function that takes the type and return the strategy. I may be misunderstanding, but I take this to mean that I can register st.register_type_strategy(tuple, lambda typ…
Part of XiaomiMiMo/MiMo-V2.6-RL-oss.