Missing dataset files and general improvements

[Gamapro]

Hello authors!

I'm very interested in your work and I want to reproduce your results to understand your method in detail. However I find that it is not possible due to some missing files, resulting in errors as follow.

FileNotFoundError: [Errno 2] No such file or directory: '../dataplace/data/TrimmedTestScene1_graph/val.json'

In this previous issue you mentioned you used this dataset, but I couldn't find the missing files in that data.

Can you please help me provide, or in case, explain where to get this files?

Other than this, there are some improvements I recommend you to improve your repository.

• First I kindly suggest update the README as the steps to follow are unclear, just run the run.py seems not enough to understand the whole pipeline.
• Also your current implementation relies in text-davinci-003, which is discontinued, as well as an old version of the OpenAI api, I suggest you to update it to the new API and possibly with gpt-3.5-turbo
• And last but not least there might be other issues we still don't know, it will be great if you can take some time to try reproduce your repo in a fresh environment so you notice what data/steps are missing and help the community adapt your method faster.

Thank you very much for your work!

[Richardhrz]

hello author，it spent me a lot of time to reproduce your work and finally get stucked in the missing file: val.json. can you please help to give a link for this file? thx.