Potential performance issue: to_csv regression in pandas below 1.4 version

[TendouArisu]

Issue Description:

Hello.
I have discovered a performance degradation in the .to_csv function of pandas version 1.2.0. And I notice the repository depends on pandas 1.2.0 in requirements.txt. I am not sure whether this performance problem in pandas will affect this repository. I found some discussions on pandas GitHub related to this issue, including #39413 and #44908.
I also found that 03-process-unsplash-dataset.ipynb used the influenced api. There may be more files using the influenced api.

Suggestion

I would recommend considering an upgrade to a different version of pandas >= 1.4 or exploring other solutions to optimize the performance of .to_csv.
Any other workarounds or solutions would be greatly appreciated.
Thank you!

[haltakov]

Thanks @TendouArisu, this is a valid point. I don't have much time to maintain this project right now, so not sure when I can get to it. I'm happy to merge a PR though