[e12e]

I was qurious about the classifier, and as expected it appears to be based on crawling to get the unsafe images - I doubt it's quite in compliance with neither copyright or gdpr..

The classifier lib:

https://github.com/alex000kim/nsfw_data_scraper

Announcement points to this data source (while the readme hints at a "premium" classifier):

https://github.com/infinitered/nsfwjs