Open apuder opened 6 years ago
I figured out what happened. In a recent commit 26cceb4 the scaling of the pixel values to range [0-1] was added. This is inconsistent with the pre-trained model. Reverting commit 26cceb4 makes the pre-trained model work again. The model should be updated with the change that was introduced in that commit.
Ascott02's model solved this issue for me
I'm using Tensorflow 1.7 and Keras 2.1.6. Doing a git clone followed by "python qlearn.py -m Run" should run the pre-trained model, right? However, the result does not look good. Flappy Bird immediately crashes at the first pipe.