Topic recognition

[leahvillr]

Hiya,
This pull request is for an implementation of problem 8 (Create a generative model of one of the OASIS brain, ADNI brain or the OAI AKOA knee data set (see Appendix for links) using a VQVAE [11] or VQVAE2 [12] that has a “reasonably clear image” and a Structured Similarity (SSIM) of over 0.6)

Please find contained within
interface.ipynb containing the main interface for interacting with my models
README.md containing information on how the models work together including how to use the models yourself
dataset.py containing the code used to implement the various datasets and data loaders used throughout
predict.py containing the code for generating new images
train.py containing the code used for the training loops for the various models
environment.yml containing the environment used during the creation of this project
images/*.png containing the images used in the README
modules/*.py containing the various models used in this project

• modules.decoder.py decoder implementation
• modules.encoder.py encoder implementation
• modules.quantizer.py vector quantizer implementation
• modules.stack.py implementation for the stacked residual layers used by other modules
• modules.vqvae.py VQ-VAE implementation
• modules.pixelcnn.py pixel cnn implementation

Thank you for your time taken to review and hope for feedback.

Alex Viller,
45375325

[SiyuLiu0329]

This is an initial inspection, no action is required at this point

• Reconstruction: OK
• Generation: code seem largely OK. I can see the model code and image generation code. Logic seems correct. No results due to some errors.

[shakes76]

Good Practice (Design/Commenting, TF/Torch Usage)

Adequate use and implementation (some errors prevented it from working, partially working) -3
Good spacing and comments
Header blocks missing -1

Recognition Problem

Solves problem (some outputs missing) -2
Driver Script present
File structure present
Shows Usage & Demo & Visualisation & Data usage (no plots, no full output) -2
Module present
Commenting
No Data leakage
Difficulty: Hard

Commit Log

Meaningful commit messages
Progressive commits used

Documentation

ReadMe acceptable
Good Description and Comments
Markdown used PDF submitted

Pull Request

Successful Pull Request (Working Algorithm Delivered on Time in Correct Branch)
Feedback required, undo/revert changes to other student files and conflicts -2
Request Description good

[leahvillr]

I see that I need to undo/revert changes to other student files. I don't remember making those changes I think it might have happened from having some kind of inconsistency between my version and the public version when I submitted the pull request. For fixing these changes should I just edit the files so they are identical to the other students ones?

[leahvillr]

I've attempted to get rid of my changes to other students read me and even after copying and pasting from origin and attempting git checkout origin/topic-recognition -- recognition/ISICs_UNet/README.md and git checkout origin/topic-recognition -- recognition/XUE4645768/Readme.md and then commiting after that it seems like I cannot get rid of these changes. Any tips?

[shakes76]

Seems the student files still showing up as changed. Can you check again. I will grant the Feedback marks, because you did attempt to solve it, but can't merge until you do fix it.