jpmorganchase/Phantom

Issues

• Add support for bounded rationality for strategic agents

  #134 opened 7 months ago by benjaminpatrickevans
  1

• Make the supertype `_managed` feature configurable

  #133 opened 8 months ago by leoardon
  0

• Errors when running the "digital_ads_market" and "simple_market" sample environments

  #111 opened a year ago by leejasper851
  1

• Feature Request: Training plans

  #54 opened 2 years ago by jaredvann
  0

• Better logging of message flows in Mercury

  #20 opened 3 years ago by jaredvann
  0

• Remove unused dependencies and update version pinning for everything else where possible

  #42 opened 3 years ago by jaredvann
  0

• Track recent Ray/RLlib version

  #27 opened 3 years ago by jaredvann
  2

• Better support for post-rollout analysis

  #19 opened 3 years ago by jaredvann
  0

• Switch Mercury to use NetworkX

  #21 opened 3 years ago by jaredvann
  0

• Environment level types and supertypes

  #3 opened 3 years ago by jaredvann
  0

• 'phantom' command causes segmentation faults when used in Conda environment

  #37 opened 3 years ago by jaredvann
  0

• FSM Environment - allow agents to take multiple action types

  #22 opened 3 years ago by jaredvann
  0

• Experiments with multiple 'sub-steps' need better handling or design guidance

  #4 opened 3 years ago by jaredvann
  2

• Need to be able to easily create and use fixed/custom policies

  #18 opened 3 years ago by jaredvann
  0

• Support for Weights & Biases

  #28 opened 3 years ago by jaredvann
  0

• Use Dict instead of Tuple for `observation_space` and `action_space`

  #10 opened 3 years ago by leoardon
  0

• Add more Phantom environments

  #1 opened 3 years ago by jaredvann
  0