[Feature Request] MCTS Issue tracker

[vmoens]

• break_when_all_done in env.rollout() [Feature Request] break_when_all_done in env.rollout() #2355
• Partial steps in env [Feature Request] Partial steps in env #2356
  • BatchedEnv: pass the indices of envs where a step should be done
  • BatchedEnv: index a BatchedEnv
  • Stateless env: discard done parts of the TD
• Scores [Feature] MCTS Scoring functions #2358
  • PUCT (missing tests)
  • UCB (missing tests)
  • UCB1_TUNED
  • EXP3
  • PUCT_VARIANT
• TensorSpec.enumerate()
• Storage
  • Hashing functions [Feature] TensorDictMap hashing functions #2304
  • Query modules [Feature] TensorDictMap Query module #2305
  • Map [Feature] TensorDictMap #2306
  • MCTSForest [Feature] MCTSForest #2307
• Policy classes [Feature] MCTS policy #2359
  • MCTSPolicyBase
  • MCTSPolicy
  • AlphaGoPolicy
  • AlphaStarPolicy
  • MuZeroPolicy

cc @dtsaras @mjlaali

[ParamThakkar123]

@vmoens I would like to work on this issue if it's open for contributions. Looks quite interesting!

[vmoens]

There's a PR open but I never really had the time to properly review it.

Feel free to branch out and pick it up! #2796

[ParamThakkar123]

Sure @vmoens ! Thank you!