Fix ray & TF v2.10 dependency

ray v2.3.1 lists tensorboardX as a dependency. tensorboardX upgraded to v2.6.1 which lists protobuf >v4.2 as a dependency. That breaks ray's import of tensorflow v2.10 which we're currently using. Fixed by downgrading tensorboardX to v2.6.

Fix another warning

Fixes rllib's warning that the reset function can't set the seed

v0.2.4

Skip the warning message of choosing a masked child state from rllib's precheck module. I believe it is caused because rllib's precheck function chooses a random action, many of which could be masked out for the first state.

Update to use Ray v2.3.1

What's Changed

• Update to use ray v2.3 by @jlaw9 in #54

Full Changelog: v0.2.2...v0.2.3

Pin ray to v2.2.0

Removed Python 3.7

Doesn't allow some of the dependencies

Update to use Ray v2.2.0

Ray v2.2.0 uses new algorithm config objects instead of config dicts.

v0.1.3

trying zenodo again

v0.1.2

Additional zenodo json edits

v0.1.1

Fix zenodo JSON

v0.1.0

What's Changed

• UPDATE: Fix typos, PEP8 python format, correct declaration of return … by @iammix in #47
• Joss paper edits by @pstjohn in #51

New Contributors

• @iammix made their first contribution in #47

Full Changelog: v0.0.9...v0.1.0