🚩 Report: Illegal or restricted content
Trained on illegal content
Nope, you're the one that trained a model on illegal content: https://civitai.com/models/639094/rnormalnudes-for-flux
This is a merge, not a finetune either way :)
@lunarflu If this mass harassment continues, The companies knowingly and willingly allowing and encouraging the illegal harassment will be held liable. Done with this, especially on HF. This is very unprofessional behavior that is harming the huggingface name as a whole. It's been reported directly to the offices, yet you let this mentally unstable guy run around dragging peoples names through the dirt and taking repos down. it's time to boot your buddy. There will be no more communications about this from me period.
It's frankly embarrassing that you allow this sort of abuse to happen @lunarflu . Huggingface clearly continues to have issues with failing to appropriately handle bad faith actors like bghira . You give us no solution to report users themselves on their profiles, but allow users to spam reports on as many repos as they like without any restrictions, or clearly, and sort of through investigation to determine if the user in question is acting in good faith, or has ulterior motives. bghira has and continues to spread false information about other users and models due to his disagreements with them. He is vindictive, petty and feels himself above reproach, by not acting to address it, you are basically signing off of this sort of behavior as acceptable.
If his hearsay is enough to go off, then I formally accuse all of bghira's models of containing illegal content, violating model and code licenses, copyright violation, trademark violation, training on individuals without consent, and as well as being nudify capable. Just trust me.
the model reports are simply because they're not in accordance with hugging face's policy, it has no other reason behind it. discovered this model purely because of research done by others on the person who is behind the organised harassment campaign against me and huggingface.
the models that were removed did one or more of these things:
- declared their purpose on their model card, eg. removing or ripping clothes
- had their purpose in their filename
- matched known file hashes of otherwise-removed models
how do you think they're even discovered? because they advertise their purpose. i'm sure if no one advertised what the model was for, no one would know. in the case of one of today's reuploads, they called it "desired LoRA" but they also included the username who originally uploaded it and the entire original model card. pretty poor covering of tracks. it's bold. it's breaking the terms of service very publicly, and in a way that's out of spite toward the people who own and operate this platform.
coming along here to try and mislead others who don't have the full context is a pretty scummy move.
To be honest. I'm not one of your little friends, and you won't intimidate, threaten, or attempt to dox me a single bit. I'll hand that straight to my law team to investigate you for leaking/doxing since huggingface allows you to threaten and harass users at will because of your buddies. And you are in an extradition country, good luck with that.
i can be more clear here about your deliberate twisting of reality.
you use terms like "mentally unstable," throwing around baseless threats of extradition (for posting a screenshot from a fully public, Google-indexed Discord channel?), and invoke imaginary "law teams" to intimidate and harass me as well as our shared colleagues at Hugging Face. It's very strange behaviour.
everything I've posted is publicly accessible information. there's no "leaking," there's no "doxing," just screenshots that anyone with an internet connection can find (without even agreeing to your own server's rules).
you're accusing me of harassment while openly harassing me, spreading malicious, defamatory statements without evidence, and inventing fictional scenarios to frighten people away from holding you accountable.
if you genuinely held legal standing you wouldn't blurt crap out that would hurt your case. you're just lashing out because your attempts at intimidating me into silence have failed, again.
i'm going to continue filing the policy violations when i find them, just as i have for more than a year. not sure why it's all of a sudden making people upset that rules exist.
"i'm going to continue filing the policy violations when i find them, just as i have for more than a year. not sure why it's all of a sudden making people upset that rules exist."
Funnily enough, you did participate, not initiate, in these reporting activities, like a couple of times for 2 years, suddenly, you started filing bogus reports against a variety of users, people did not suddenly get upset about these reports, you suddenly started acting like a Karen and reporting a bunch of people.
Anyone can check your recent activity and see it with their own eyes.
Man, go outside and get some air, stop eroding the little amount of reputation you still have.
the kontext model that only existed recently introduced some changes to the Flux non-commercial license which is going to take people some time to adapt to.
the kontext model that only existed recently introduced some changes to the Flux non-commercial license which is going to take people some time to adapt to.
@bghira I wonder if you got consent of everyone you trained your https://civitai.com/models/639094/rnormalnudes-for-flux on (which funny enough, is no longer up, I wonder why?), and did you check to verify it can't generate nude children? You know Flux can generate children right? You wouldn't be enabling CSAM, would you? Did you verify it can't be used to nudify people via image to image or inpainting?
If the issue is capability, then you best watch out it doesn't come back on you, as models tend to generalize past what any of us want, regardless of advertising such functionality. You are still responsible if your models can do such things, by your own stance on the matter.
the model you linked to is still available, go ahead and give that all a try yourself.
@bghira So did you or didn't you do those things?
Edit: Didn't ask for consent, did you? Didn't make sure the models aren't capable of things that violate licenses or TOS?
Are you suggesting it is the responsibility for OTHERS to make sure release models can't do those things, not the one responsible for creating and releasing the model?
The model got "updated" overnight, interesting. No changes in the model description. Nothing new.
The model had been taken down, and is now up again.
Unlike Chroma, which you filed a bogus report against and is based on Flux Schnell(under apache 2.0 license), your lora seems to be trained on Dev, which has different licensing conditions
Guys i don't think there is any use talking to a mentally ill person. Even comfyanonymous , the creator of comfyui said that this dude @bghira is mentally unstable.
It's not about who they are as a person, it's about their behavior. Being an asshole and going around and trying to get other people's work removed out of vindictiveness (or clearing the field for his own work, who can say), especially being a hypocrite about it, is not acceptable. Huggingface should NOT enable and allow such behavior.
He is not making an effort to clean all of Huggingface or anything so absurd, there is plenty of stuff all over Huggingface that would supposedly violate some TOS, law, ethics/morals, or license somewhere in the world...it would be a full time job to sort through it all, instead he is being targeted in his reporting, making it even more suspect.
It's probable EnforcementMan is his alt, similar timing, behavior/attitude, way of reasoning, and presence of particular individuals in discussions leads to enough circumstantial evidence to support the possibility, and I can understand the fear of backlash on his main would lead to trying to hide who he is. Only he would know for sure, but any statement to the contrary could also be a lie, rumors suggest he has a past of using alts to evade direct accountability. Edit: If this is him (or maybe someone close to him), he is using bad faith arguments that furry = beastiality, and he has, when asked about supporting Chroma on his trainer in the past, stated something to the effect, emphatically, that he won't because of what it can generate, or what it was trained on, so there is a connection there to feeling Chroma is "wrong" in some way.
I highly suspect he doesn't:
- Get consent from any people, license/trademark/intellectual property holders, or artists in training his models, and doesn't verify the state of the licensing of any such data he pulls from a source or it's legality.
- Verify that any models he trains do not introduce capabilities or are used in ways that violates TOS, laws, or licenses.
- Actually check if the models he is reporting are in violation of anything, making assumptions based on descriptions, or just fabricating hearsay.
He can provide any retort he likes, he will know if he is lying, and he will be committing to a statement that can be corroborated or found inconsistent. But for me? It seems likely he is doing these reports in bad faith for reasons not completely clear, but seemingly out of personal disagreement or distain for certain content or users*. His behavior in response to people actually doing their homework and digging into his possible own violations has been met with flippancy, dismissal, and vindictive responses to said users, and notably, no evidence or direct statements answering the users in those regards, perhaps because he is well aware he has no good answer that doesn't commit him to lying, or exposing him as a hypocrite.
Edit 4: He blocked my user, so I thought he deleted the post and comments, no, he is still committing social suicide actively. If any of his stuff shows up as deleted, he probably blocked you, use incognito or an alt. Or could check archival services.
EDIT 3: He has since deleted the reddit post, all his comments on it, and the HF post, you'll have to see if they were archived anywhere.
https://archive.is/hZJi0 for his reddit post
The saga continues, enjoy his hypocrisy, holier than thou attitude, and his ability to come up reasons why what he does is different and not a problem from his comments. https://old.reddit.com/r/StableDiffusion/comments/1lx164s/tensorart_no_longer_allowing_nudity_or_celebrity
And no doubt he made that post to gloat.
https://huggingface.co/posts/bghira/418004548344151
EDIT: Here he is being dishonest and straw manning. Most people would be fine preventing NSFW generations of real people, but Tensorart is nuking all NSFW, don't let bghira dishonestly frame the discussion.
EDIT2: This breakdown mirror post, it's comments being different than the original, has some additional behavior that indicates bghira lacks any integrity, cares about having power over others while imposing his personal morals on them, while exempting himself from them: https://old.reddit.com/r/comfyui/comments/1lsfodj/full_breakdown_the_bghirasimpletuner_situation/
https://huggingface.co/posts/bghira/418004548344151
This one 404'd
What I find incredible is how someone can get over 100 downvotes in a single comment and still be convinced to be in the right.
@YuLexuan30 Yeah, nothing really special other than him saying NSFW celebrity was banned on Tensorart, neglecting to mention all NSFW was banned, and setting up a strawman that anyone who is upset/disagrees/etc has some serious explaining to do. He, rightfully, realized rather quickly once I linked the original post that he was just going to look like a moron, so deleted it.
It looks like he deleted his reddit account (https://www.reddit.com/user/terminusresearchorg/)
It also appears his NSFW nudes lora is no longer up (https://civitai.com/models/639094/rnormalnudes-for-flux)
It looks like he deleted his reddit account (https://www.reddit.com/user/terminusresearchorg/)
Last time I checked, some of his comments got downvoted nearly 150 times. That is MASSIVE, I don't think I've ever seen a Reddit user being so heavily downvoted that quickly. By attacking Chroma, which is being supported by so many people and does not violate anything, he pissed off so many people.
It also appears his NSFW nudes lora is no longer up (https://civitai.com/models/639094/rnormalnudes-for-flux)
Because it was really funny to see him witch hunting people for uploading supposedly illegal NSFW stuff when he himself was making the same content, maybe this is why he deleted that lora, you probably can't reach it by the wayback machine, but I posted some images here proving it existed. The entire drama is absolutely illogical, claiming that he was fighting in our best interest by filling bogus reports on a model that the community is putting a lot of collective effort into is insane. Kudos for Lodestones for handling it well and to the HF staff for not accepting some false reporting right away.
no one expected HF to just simply act on reports and close models down 🤦🏽 the purpose of the reports are for things to be reviewed. and quite a lot is permitted on HF, but very specific things are not. it seems that loli and shota models like this one are ok, but nudify LoRAs are not.