This is nuts. #huggingface have just scraped the entirety of #github. https://huggingface.co/spaces/HuggingFaceCode/in-the-stack
7
Replies
·
0
Boosts
@concretedog the motherfuckers make you list each one of the repositories. Does anyone has a script to get all the repos I made commits or comments to?
@mdione @concretedog you can put “- all” as the list, apparently. Not that that makes it *better*, but at least slightly more convenient.
@concretedog "We want to give developers agency over their source code". Then maybe ask before stealing it?
@concretedog Well they needed to update their models after their partnership with Software Heritage in 2023 I suppose.
I wasn't particularly pleased then either.
@onepict
I didn't realise that Software Heritage scraped everything regardless of license. I see lot of my "all rights reserved" material in there. I guess at this point it is futile to ask for removal.
@wim_v12e @concretedog As I recall they had version one and two for hugging face.
Librecast code was in both. I was able in 2023 to get it removed from version two, but there were crickets when I asked for the same for version one.
So possibly it is too late.
I think Software Heritages habit of scraping without curation is wasteful and there was a bit of debate on Mastodon in 2023.
