HF RL Explorer

Found by @thomashampson

Found by @thomashampson: a task in MiMo-V2.6-RL-harbor-code: MiMo-V2.6-RL Code (Harbor) (Harbor dataset). (Note: as part of the fix, I'd expect a new module-level constant like SAVE STATE VERSION to be introduced in workbench.config and threaded through the relevant saveState / restoreState calls.)

The task

(Note: as part of the fix, I'd expect a new module-level constant like `SAVE_STATE_VERSION` to be introduced in `workbench.config` and threaded through the relevant `saveState` / `restoreState` calls.)

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