This week I completely abandoned the prospect of vgg16, and shifted focus entirely to AlexNet. Perhaps most importantly, By the end of the week I managed to get enough data from the BAIR dataset sorted out between rainy and nonrainy. It was still likely that I would have to grab other images since some are incredibly blurry and only adds to confusion – especially for such a small dataset of 200 images. However, the goal is to depend mostly on AlexNet’s already fairly good “intuition” that came from its previous training. After all this is finetuning and not a brand new training session. After some small tweaks to AlexNet to bring it up to date with my version of tensorflow, we should be ready to get the ball rolling for training.