Update BabyAI to use MiniGrid NumPy-based renderer (mila-iqia#87)

* Started refactoring to eliminate PyQT dependency

* Try to fix travis dependency issue

* Travis debugging mila-iqia#2

* Replaced test_mission_gen.py by manual_control.py

* Ask people to install pyqt5 if they run gui.py

* Remove gui.py since we also have manual_control.py

.travis.yml

@@ -9,9 +9,9 @@ before_install:
 # command to install dependencies
 install:
     - pip3 install http://download.pytorch.org/whl/cpu/torch-0.4.1-cp35-cp35m-linux_x86_64.whl
-    - pip3 
-    - pip3 install --editable .
     - pip3 install flake8
+    - pip3 install scikit-build
+    - pip3 install --editable .
 
 # command to run tests
 script:

README.md

@@ -40,8 +40,8 @@ Requirements:
 - Python 3.5+
 - OpenAI Gym
 - NumPy
-- PyQT5
 - PyTorch 0.4.1+
+- blosc
 
 Start by manually installing PyTorch. See the [PyTorch website](http://pytorch.org/)
 for installation instructions specific to your platform.
@@ -98,13 +98,13 @@ Models, logs and demos will be produced in this directory, in the folders `model
 To run the interactive GUI application that illustrates the platform:
 
 ```
-scripts/gui.py
+scripts/manual_control.py
 ```
 
 The level being run can be selected with the `--env` option, eg:
 
 ```
-scripts/gui.py --env BabyAI-UnlockPickup-v0
+scripts/manual_control.py --env BabyAI-UnlockPickup-v0
 ```
 
 ### The Levels

docs/codebase.md

@@ -14,4 +14,4 @@ In `scripts`:
 - use `train_intelligent_expert.py` to train an agent with an interactive imitation learning algorithm that incrementally grows the training set by adding demonstrations for the missions that the agent currently fails
 - use `evaluate.py` to evaluate a trained agent
 - use `enjoy.py` to visualze an agent's behavior
-- use `gui.py` or `test_mission_gen.py` to see example missions from BabyAI levels
+- use `manual_control.py` to visualize example missions from BabyAI levels

environment.yaml

@@ -5,8 +5,6 @@ channels:
 dependencies:
     - python=3.6
     - pytorch=0.4.1
-    - torchvision
-    - pyqt
     - numpy
     - blosc
     - pip: