shrimai / Style-Transfer-Through-Back-Translation

162 stars 32 forks source link

Datasets links not working #29

Open hadi-ibra opened 1 week ago

hadi-ibra commented 1 week ago

I was exploring the repository and found that the dataset link for political slant is no longer working:

http://tts.speech.cs.cmu.edu/style_models/political_data.tar

It seems that the links return a "site can't be reached". Could you please provide updated links or suggest an alternative source to access these datasets?

thien commented 1 week ago

I also have the same issue. I have an older copy of this dataset that I've uploaded to huggingface: https://huggingface.co/datasets/thien/political

Since these are parquets, to use them with the codebase in this repository, you'll likely want to convert them back into text files.

@shrimai, if this is not okay, let me know and I can take it down. Thanks!

hadi-ibra commented 1 week ago

Thank you, and do you also have the gender dataset since its link is also not working:

http://tts.speech.cs.cmu.edu/style_models/gender_data.tar

thien commented 1 week ago

I don't, sorry :/

hadi-ibra commented 1 week ago

no worries, thank you for your help and providing the political slant dataset