mila-iqia/babyai

Issues

• Get demonstration for 'MiniGrid-MultiRoom-N4-S5-v0'

  #123 opened 9 months ago by Abc11c
  1

• What's the package version of gym-minigrid and gym used for the baseline results?

  #110 opened 2 years ago by NagisaZj
  2

• When I run 'BabyAI-PutNext-v0', it did not work after 24 hours' training...

  #111 opened 2 years ago by Amanda2024
  1

• Error during install: "pip install --editable ." in babyai folder

  #116 opened 2 years ago by quinlanp
  2

• What does the agent see, in 7x7? -- possible bug in the task?

  #117 opened 2 years ago by sunchipsster1
  1

• Mission not accomplished in the last observation

  #115 opened 2 years ago by Valaybundele
  0

• Sampling unsolvable demonstrations

  #113 opened 2 years ago by tttyuntian
  0

• Unable to run PPO agent with arch=bow_endpool_res_pixels

  #109 opened 4 years ago by ayushchakravarthy
  2

• You cannot pick up or put down objects in manual-mode in enjoy.py

  #100 opened 4 years ago by jsyedidia
  6

• PutNextLocal

  #105 opened 4 years ago by wcarvalho
  8

• Bot bugs (attribute error: prev_carrying)

  #106 opened 4 years ago by Bihaqo
  2

• Question about performance

  #104 opened 4 years ago by wcarvalho
  2

• Document how to create new levels

  #33 opened 4 years ago by maximecb
  0

• Publish pip package

  #95 opened 4 years ago by luckeciano
  1

• Impossible missions are sampled?

  #103 opened 4 years ago by wcarvalho
  2

• Know full vocabulary apriori.

  #102 opened 4 years ago by wcarvalho
  2

• enjoy.py script seems to expect old PyQT version of gym-minigrid

  #98 opened 4 years ago by jsyedidia
  3

• Integration / training tests

  #46 opened 4 years ago by maximecb
  0

• Visualizing stored "images"

  #94 opened 5 years ago by zynos
  2

• AttGRU model is broken + my quick Fix

  #91 opened 5 years ago by zynos
  5

• Missing dependency for scikit-build

  #89 opened 5 years ago by dyth
  4

• Error in docs/train-eval.md?

  #78 opened 5 years ago by ricvolpe
  2

• default hyperparameters don't make sense for both IL and RL

  #50 opened 5 years ago by rizar
  2

• Bug in BOT Stack for PickupLoc

  #80 opened 5 years ago by chitwansaharia
  12

• Headless rendering issue for environments with doors

  #86 opened 5 years ago by mbchang
  5

• > Thanks

  #85 opened 5 years ago by chitwansaharia
  0

• Revised article TODO

  #4 opened 5 years ago by maximecb
  3

• Google Colab

  #76 opened 5 years ago by jordiae
  2

• Potential initialisation error of self.mask in base.py

  #77 opened 5 years ago by chitwansaharia
  2

• Build-essential missing from Linux instructions

  #71 opened 5 years ago by cicicice
  1

• How to measure RL sample efficiency

  #74 opened 5 years ago by rickyloynd-microsoft
  1

• GoNextToSubgoal explore bug

  #53 opened 6 years ago by dyth
  4

• Not getting the success rate mentioned in the paper using the pretrained models.

  #55 opened 6 years ago by akshay-sharma1995
  6

• Scrap cnn1, cnn2, etc?

  #34 opened 6 years ago by maximecb
  1

• Make train_intelligent_expert.py resumable

  #37 opened 6 years ago by maximecb
  3

• go straight for the key when you head for locked door

  #29 opened 6 years ago by rizar
  3

• support Python reference as an argument in GoNextTo

  #28 opened 6 years ago by rizar
  2

• refactor GoToObj and GoNextTo

  #15 opened 6 years ago by rizar
  3

• looking for very very blocked objects

  #14 opened 6 years ago by rizar
  3

• reseed the rng for every mission when generating demos

  #23 opened 6 years ago by rizar
  8

• get rid of GoToAdjPos

  #16 opened 6 years ago by rizar
  1

• bot spawned in a room without a key

  #22 opened 6 years ago by rizar
  1

• broken bonus levels

  #20 opened 6 years ago by rizar
  4

• bot failing some missions

  #11 opened 6 years ago by rizar
  2

• do we really need both test_mission_gen.py and gui.py?

  #5 opened 6 years ago by rizar
  3

• bot crash

  #7 opened 6 years ago by rizar
  15