Commits · 805d9d735d491d3481db34f501fb12164264720d · Flatland / baselines

Jun 09, 2020
- Update render_agent_behavior.py · 805d9d73
  Erik Nygren authored 4 years ago
  
  805d9d73
Nov 07, 2019
- updated files to new malfunction behavior · e1ce4008
  Erik Nygren authored 5 years ago
  
  e1ce4008
Oct 24, 2019
- catchin error with None observation · 653126fe
  Erik Nygren authored 5 years ago
  
  653126fe
Oct 07, 2019
- added shortest path predictor to multi agent training and inference · 44844ffa
  Erik Nygren authored 5 years ago
  
  44844ffa
- removed "bug" with reward. Attention, currently it is cheaper for an agent to... · 3babf29d
  Erik Nygren authored 5 years ago
  
  removed "bug" with reward. Attention, currently it is cheaper for an agent to wait if we cummulate rewards between the different state!
  3babf29d
Oct 06, 2019
- updated training file · df0b0ef1
  Erik Nygren authored 5 years ago
  
  df0b0ef1
Oct 05, 2019
- updated training for single agent navigation to be compatible with new env · 67763383
  Erik Nygren authored 5 years ago
  
  67763383
- updated to new reset behavior · caada6f1
  Erik Nygren authored 5 years ago
  
  caada6f1
Sep 27, 2019
- clean up Agent and dueling_double_dqn · aafa76af
  u229589 authored 5 years ago
  
  aafa76af
Sep 26, 2019
- remove unused observation and prediction files · 81513d7b
  u229589 authored 5 years ago
  
  81513d7b
Sep 02, 2019
- fixed reward bug · 81e6e103
  Erik Nygren authored 5 years ago
  
  81e6e103
Sep 01, 2019
- updating training to handle stochastic events and different agent speeds. · 2ba03dda
  Erik Nygren authored 5 years ago
  
  Trying to create feasible sarsa packages
  2ba03dda
- fixed bug where all values are infinite · 9b16819f
  Erik Nygren authored 5 years ago
  
  9b16819f
- only store observations in buffer when you are allowed to take an action! · 23e98b59
  Erik Nygren authored 5 years ago
  
  23e98b59
- updated observation and training to handle multi-speed · 2c63e825
  Erik Nygren authored 5 years ago
  
  2c63e825
- updated training navigation and render_agent_behavior · cebde3d8
  Erik Nygren authored 5 years ago
  
  cebde3d8
- using new level generator for training and inference · b8e41dc1
  Erik Nygren authored 5 years ago
  
  b8e41dc1
Aug 28, 2019
- merge #141 renamed agent_generator* to schedule_generator* · 3e8a8194
  u214892 authored 5 years ago
  
  3e8a8194
Aug 27, 2019
- #141 different agent classes · ee91911e
  u214892 authored 5 years ago
  
  ee91911e
- #141 different agent classes · b8eab2fb
  u214892 authored 5 years ago
  
  b8eab2fb
Jul 16, 2019
- minor bugfix in inference files · bb778417
  Erik Nygren authored 5 years ago
  
  bb778417
Jul 15, 2019
- updated agent behavior rendering code · 4fc55df4
  Erik Nygren authored 5 years ago
  
  4fc55df4
- minor update · 99ff2dc8
  Erik Nygren authored 5 years ago
  
  99ff2dc8
- fixed number of features import · 831db99c
  Erik Nygren authored 5 years ago
  
  831db99c