Hello,

I am trying to reproduce some of the results in your paper.

I am focused for now on the Diabetic Retinopathy results, and in particular the file ssl_mocov2_mse.py

It seems there are many bugs in this code, including syntax errors and the use or passing of undefined variables, such that it doesn't run.

Just one example - in your main function here: https://github.com/aakashrkaku/intermdiate_layer_matter_ssl/blob/main/diabetic_retinopathy/ssl_mocov2_mse.py#L521-L544

You define dm = BaselineDataModule ... but then you are looking for a variable called data_module

I can try to fix the errors like this (and also make a Dockerised version of the training to make reproducability easier). But if you could provide a script that actually runs and produces your results, that would be very useful!

Thanks a lot, and thanks for sharing your research code!

Hi there,

It is unclear from your respository how exactly to recreate the training dataset for Diabetic Retinopathy used in your paper.

I downloaded the Diabetic Retinopathy dataset from Kaggle, then unzipped it and ran it through the script available here as it seems to me that this is what you have done.

I then get several different outputs from this script, and it seems the one you have used for training is that output to /jama16-retina-replication/data/eyepacs/bin2

However, when I run the above-mentioned script, the bin2/validation folder is empty. How exactly did you partition the training set into train and validation?

Thanks in advance!

I tried to review your results, but a lot of packages in requirements.txt were loaded from files which I had no access to. For example, scipy @ file:///tmp/build/80754af9/scipy_1597686620742/work seaborn @ file:///tmp/build/80754af9/seaborn_1600553570093/work SecretStorage @ file:///tmp/build/80754af9/secretstorage_1594419219833/work Would you mind providing the versions of all the packages mentioned (and not mentioned) above? It would be really helpful.