Finspire13/pytorch-policy-gradient-example

TODO List

Closed this issue · 1 comments

  • Set up / find good tree data structure for doing checks and traversal
  • Create and store dictionary of tokens for each expression
  • Implement tree updating from distribution over nodes
  • Implement tree traversal for recovering parent and sibling node tokens
  • Define node characteristics (argument counts, type, terminal?)
  • Implement fitting constants
  • Implement policy gradient/ figure out how

sorry lol