Issues with lib/utils/run.sh for input files

[sarahesl]

Issues with lib/utils/run.sh for input files

[sarahESL]

Thanks for the comment and opening the issue. I will have a look as soon as I get the time. Meanwhile, feel free to make a pull request with your changes if you want.

[lycutter]

Hello, I'm having trouble running the script for generating the input files (dictionaries, and pickles). One problem in lib/utils/run.sh runs python create_dictionary.py "../../data/data_slake" instead of /data_rad/. However, after changing the directory to data_rad more issues pop up including df columns not matching and I understand that most likely the file was not modified perfectly for generating both datasets' inputs.

It took a while to get them all working, but it would be nice to see a fix.

Hello,
Could you please share the preprocess data on data_rad, I found some troubles in generating the files. Thanks!

[sarahESL]

Thanks @lycutter and @zyzkevin for reporting this issue.

Both create_dictionary.py and create_label.py should be fixed now and work for both VQA_RAD and VQA_SLAKE datasets.

Let me know if any other issues shall arise!

[lycutter]

Thanks @lycutter and @zyzkevin for reporting this issue.

Both create_dictionary.py and create_label.py should be fixed now and work for both VQA_RAD and VQA_SLAKE datasets.

Let me know if any other issues shall arise!
Thank for your replying, I update these two files and use them to generate the input files. However, it occured some bugs too, suah as key error 'answer[image_name] ' in the data_RAD.py, dimension mismatch when loading the type_classifer.pth. The train.json file and test.json file I use are from https://github.com/aioz-ai/MICCAI19-MedVQA. I don't know whether these bugs are related to these two .json files. Could you plz share all the completed input files.
Thanks!

[sarahESL]

@lycutter The scripts in this repo should work for the official datasets released in the original VQA_RAD paper. But maybe you can use the data from https://github.com/Awenbocc/med-vqa/tree/master/data?

If the error in data_RAD.py still appears, I would appreciate it if you could make another issue and write the error you get there. Let's keep this thread for issues arising from lib/utils/run.sh.

[lycutter]

@lycutter The scripts in this repo should work for the official datasets released in the original VQA_RAD paper. But maybe you can use the data from https://github.com/Awenbocc/med-vqa/tree/master/data?

If the error in data_RAD.py still appears, I would appreciate it if you could make another issue and write the error you get there. Let's keep this thread for issues arising from lib/utils/run.sh.

Thanks for your help! It works now.