MCTS Fixes
* Reimplemented UCT MCTS * Fixed softmax * Merged multiple branches into this branch, all of which should now be in master * Added reuse of tree functionality * Added the ability to expand nodes based on q values rather than at random * Refactored everything, deleted non necessary MCTS classes and files, and mcts.py can evaluate newer MCTS
Showing
- backends/__init__.py 1 addition, 1 deletionbackends/__init__.py
- backends/kerasrl_learner.py 18 additions, 12 deletionsbackends/kerasrl_learner.py
- backends/mcts_controller.py 107 additions, 0 deletionsbackends/mcts_controller.py
- backends/mcts_learner.py 350 additions, 302 deletionsbackends/mcts_learner.py
- backends/trained_policies/highlevel/highlevel_weights.h5f 0 additions, 0 deletionsbackends/trained_policies/highlevel/highlevel_weights.h5f
- backends/trained_policies/wait/wait_weights_actor.h5f 0 additions, 0 deletionsbackends/trained_policies/wait/wait_weights_actor.h5f
- backends/trained_policies/wait/wait_weights_critic.h5f 0 additions, 0 deletionsbackends/trained_policies/wait/wait_weights_critic.h5f
- high_level_policy_main.py 1 addition, 0 deletionshigh_level_policy_main.py
- mcts.py 68 additions, 246 deletionsmcts.py
- mcts_config.json 5 additions, 5 deletionsmcts_config.json
- options/options_loader.py 4 additions, 11 deletionsoptions/options_loader.py
- options/simple_intersection/maneuver_base.py 1 addition, 1 deletionoptions/simple_intersection/maneuver_base.py
- options/simple_intersection/maneuvers.py 89 additions, 14 deletionsoptions/simple_intersection/maneuvers.py
- options/simple_intersection/mcts_maneuvers.py 0 additions, 315 deletionsoptions/simple_intersection/mcts_maneuvers.py
Loading
Please register or sign in to comment