I have tested neural style transfer on google colab. Second time around.
This is the second time I test a Neural Style Transfer implementation on google colab (python notebook). If you don't know what Neural Style Transfer is, well, you take a photo (that's your content) and a picture of some painting, some texture, another random photo, whatever (that's your style) and you end up with a synthesized output image that looks like your content but with the style of your style image. It's pretty cool.
It uses Lucid which has been ported to PyTorch and that port of Lucid is called Lucent. Lucid only works on TensorFlow 1.x.
Link to google colab python notebook:
https://colab.research.google.com/git...