Mctx Versions Save

Monte Carlo tree search in JAX

v0.0.5

5 months ago

What's Changed

  • Automatically determine num_actions and num_chance_outcomes in stochastic_muzero_policy (thanks Carlos Martin).
  • Explicitly use int32 for the argmax output even when using jax_enable_x64.

Full Changelog: https://github.com/google-deepmind/mctx/compare/v0.0.3...v0.0.5

v0.0.4

5 months ago

What's Changed

  • Automatically determine num_actions and num_chance_outcomes in stochastic_muzero_policy (thanks Carlos Martin).
  • Explicitly use int32 for the argmax output even when using jax_enable_x64.

Full Changelog: https://github.com/google-deepmind/mctx/compare/v0.0.3...v0.0.4

v0.0.3

9 months ago

What's Changed

  • Add experimental support for Stochastic MuZero.
  • Fix pylint and pytype warnings.

Full Changelog: https://github.com/deepmind/mctx/compare/v0.0.2...v0.0.3

v0.0.2

1 year ago

What's Changed

  • Add a link to a0-jax on Connect Four, Gomoku, and Go.
  • Improve debugging by avoiding an unused NaN.
  • Use jax.tree_util to avoid deprecation warnings.

Full Changelog: https://github.com/deepmind/mctx/commits/v0.0.2