rizavelioglu / challenge-aido_LF-baseline-IL-sim-tensorflow

VGG16 based approach to AI Driving Olympics competition (AIDO-3)

challenge-aido_LF-baseline-IL-sim-tensorflow

VGG16 based approach to AI Driving Olympics competition (AIDO-3)

Steps

1. Download required packages

cd challenge-aido_LF-baseline-IL-sim-tensorflow/learning
pip install -r requirements.txt
pip install -e git://github.com/duckietown/gym-duckietown.git@daffy#egg=gym-duckietown

2. Run "/learning/log.py" to create training data:

python log.py

3. Train VGG16 based CNN:

python train.py

4. Copy the final model from the "learning/" directory to the "submission/tf_models/" one.

5. Evaluate!:

dts challenges evaluate

OR make an official submission:

dts challenges submit

Note:

You might want to run train.py on the cloud because VGG16 is a huge model and it takes long time to train. Therefore, a jupyter notebook instance is created: train.ipynb, that can be run on Google Colab!