HF RL Explorer

tff.program managers are awkward to use from a non-async script

tff.program managers are awkward to use from a non-async script: a task in MiMo-V2.6-RL-oss: Agentic RL Environments (MiMo RL release). tff.program managers are awkward to use from a non-async script I'm running a fairly standard FL simulation. A simplified version of my main loop looks like this…

The task

tff.program managers are awkward to use from a non-async script I'm running a fairly standard FL simulation. A simplified version of my main loop looks like this: The script is a plain top-level Python program — no async framework, no asyncio.run wrapper around main, just a…

Part of XiaomiMiMo/MiMo-V2.6-RL-oss.