Archiving YouTube dislikes (and other video information) with YouTube's api!
- Make sure curl, parallel, grep, dos2unix is installed.
- Add your video ids to ids.txt, or use the default ones from sponsorblock, however I've already archived them. Recommended to run
dos2unix ids.txtso you don't have dos newlines, the program does not work correctly with them. - Run
bash start.sh - Tell YouTube to fuck off.
Please create a pull request/issue if you can find any more datasets. The more videos the better! We only have a few weeks left.
- ids.txt - This file contains every video from SponsorBlock's database
- https://archive.org/details/archived_annotations_video_ids.csv
https://archive.org/details/ytdislikes
- Make sure you only add the id part of the video url, I have not added support for long links or channels and don't plan to.
- Running lots of downloads at once (50) by default WILL use a lot of bandwidth (Up to 500mbps or so!), aswell as cpu power if you are on a lowend system/vps. Change the amount of downloads in start.sh. However if you are on a unmetered server and don't need to worry about bandwidth charges/cpu usage then you can use 100 or higher. This will speed up the download significantly.