Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/google-deepmind/mctx

Monte Carlo tree search in JAX
https://github.com/google-deepmind/mctx

jax monte-carlo-tree-search reinforcement-learning

Last synced: 11 days ago
JSON representation

Monte Carlo tree search in JAX

Awesome Lists containing this project

README 

        
# Mctx: MCTS-in-JAX



Mctx is a library with a [JAX](https://github.com/google/jax)-native

implementation of Monte Carlo tree search (MCTS) algorithms such as

[AlphaZero](https://deepmind.com/blog/article/alphazero-shedding-new-light-grand-games-chess-shogi-and-go),

[MuZero](https://deepmind.com/blog/article/muzero-mastering-go-chess-shogi-and-atari-without-rules), and

[Gumbel MuZero](https://openreview.net/forum?id=bERaNdoegnO). For computation

speed up, the implementation fully supports JIT-compilation. Search algorithms

in Mctx are defined for and operate on batches of inputs, in parallel. This

allows to make the most of the accelerators and enables the algorithms to work

with large learned environment models parameterized by deep neural networks.



## Installation



You can install the latest released version of Mctx from PyPI via:



```sh

pip install mctx

```



or you can install the latest development version from GitHub:



```sh

pip install git+https://github.com/google-deepmind/mctx.git

```



## Motivation



Learning and search have been important topics since the early days of AI

research. In the [words of Rich Sutton](http://www.incompleteideas.net/IncIdeas/BitterLesson.html):



> One thing that should be learned [...] is the great power of general purpose

> methods, of methods that continue to scale with increased computation even as

> the available computation becomes very great. The two methods that seem to

> scale arbitrarily in this way are *search* and *learning*.



Recently, search algorithms have been successfully combined with learned models

parameterized by deep neural networks, resulting in some of the most powerful

and general reinforcement learning algorithms to date (e.g. MuZero).

However, using search algorithms in combination with deep neural networks

requires efficient implementations, typically written in fast compiled

languages; this can come at the expense of usability and hackability,

especially for researchers that are not familiar with C++. In turn, this limits

adoption and further research on this critical topic.



Through this library, we hope to help researchers everywhere to contribute to

such an exciting area of research. We provide JAX-native implementations of core

search algorithms such as MCTS, that we believe strike a good balance between

performance and usability for researchers that want to investigate search-based

algorithms in Python. The search methods provided by Mctx are

heavily configurable to allow researchers to explore a variety of ideas in

this space, and contribute to the next generation of search based agents.



## Search in Reinforcement Learning



In Reinforcement Learning the *agent* must learn to interact with the

*environment* in order to maximize a scalar *reward* signal. On each step the

agent must select an action and receives in exchange an observation and a

reward. We may call whatever mechanism the agent uses to select the action the

agent's *policy*.



Classically, policies are parameterized directly by a function approximator (as

in REINFORCE), or policies are inferred by inspecting a set of learned estimates

of the value of each action (as in Q-learning). Alternatively, search allows to

select actions by constructing on the fly, in each state, a policy or a value

function local to the current state, by *searching* using a learned *model* of

the environment.



Exhaustive search over all possible future courses of actions is computationally

prohibitive in any non trivial environment, hence we need search algorithms

that can make the best use of a finite computational budget. Typically priors

are needed to guide which nodes in the search tree to expand (to reduce the

*breadth* of the tree that we construct), and value functions are used to

estimate the value of incomplete paths in the tree that don't reach an episode

termination (to reduce the *depth* of the search tree).



## Quickstart



Mctx provides a low-level generic `search` function and high-level concrete

policies: `muzero_policy` and `gumbel_muzero_policy`.



The user needs to provide several learned components to specify the

representation, dynamics and prediction used by [MuZero](https://deepmind.com/blog/article/muzero-mastering-go-chess-shogi-and-atari-without-rules).

In the context of the Mctx library, the representation of the *root* state is

specified by a `RootFnOutput`. The `RootFnOutput` contains the `prior_logits`

from a policy network, the estimated `value` of the root state, and any

`embedding` suitable to represent the root state for the environment model.



The dynamics environment model needs to be specified by a `recurrent_fn`.

A `recurrent_fn(params, rng_key, action, embedding)` call takes an `action` and

a state `embedding`. The call should return a tuple `(recurrent_fn_output,

new_embedding)` with a `RecurrentFnOutput` and the embedding of the next state.

The `RecurrentFnOutput` contains the `reward` and `discount` for the transition,

and `prior_logits` and `value` for the new state.



In [`examples/visualization_demo.py`](https://github.com/google-deepmind/mctx/blob/main/examples/visualization_demo.py), you can

see calls to a policy:



```python

policy_output = mctx.gumbel_muzero_policy(params, rng_key, root, recurrent_fn,

                                          num_simulations=32)

```



The `policy_output.action` contains the action proposed by the search. That

action can be passed to the environment. To improve the policy, the

`policy_output.action_weights` contain targets usable to train the policy

probabilities.



We recommend to use the `gumbel_muzero_policy`.

[Gumbel MuZero](https://openreview.net/forum?id=bERaNdoegnO) guarantees a policy

improvement if the action values are correctly evaluated. The policy improvement

is demonstrated in

[`examples/policy_improvement_demo.py`](https://github.com/google-deepmind/mctx/blob/main/examples/policy_improvement_demo.py).



### Example projects

The following projects demonstrate the Mctx usage:



- [Pgx](https://github.com/sotetsuk/pgx) — A collection of 20+ vectorized

  JAX environments, including backgammon, chess, shogi, Go, and an AlphaZero

  example.

- [Basic Learning Demo with Mctx](https://github.com/kenjyoung/mctx_learning_demo) —

  AlphaZero on random mazes.

- [a0-jax](https://github.com/NTT123/a0-jax) — AlphaZero on Connect Four,

  Gomoku, and Go.

- [muax](https://github.com/bwfbowen/muax) — MuZero on gym-style environments

(CartPole, LunarLander).

- [Classic MCTS](https://github.com/Carbon225/mctx-classic) — A simple example on Connect Four.

- [mctx-az](https://github.com/lowrollr/mctx-az) — Mctx with AlphaZero subtree persistence.



Tell us about your project.



## Citing Mctx



This repository is part of the DeepMind JAX Ecosystem, to cite Mctx

please use the citation:



```bibtex

@software{deepmind2020jax,

  title = {The {D}eep{M}ind {JAX} {E}cosystem},

  author = {DeepMind and Babuschkin, Igor and Baumli, Kate and Bell, Alison and Bhupatiraju, Surya and Bruce, Jake and Buchlovsky, Peter and Budden, David and Cai, Trevor and Clark, Aidan and Danihelka, Ivo and Dedieu, Antoine and Fantacci, Claudio and Godwin, Jonathan and Jones, Chris and Hemsley, Ross and Hennigan, Tom and Hessel, Matteo and Hou, Shaobo and Kapturowski, Steven and Keck, Thomas and Kemaev, Iurii and King, Michael and Kunesch, Markus and Martens, Lena and Merzic, Hamza and Mikulik, Vladimir and Norman, Tamara and Papamakarios, George and Quan, John and Ring, Roman and Ruiz, Francisco and Sanchez, Alvaro and Sartran, Laurent and Schneider, Rosalia and Sezener, Eren and Spencer, Stephen and Srinivasan, Srivatsan and Stanojevi\'{c}, Milo\v{s} and Stokowiec, Wojciech and Wang, Luyu and Zhou, Guangyao and Viola, Fabio},

  url = {http://github.com/deepmind},

  year = {2020},

}

```