We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
No description provided.
The text was updated successfully, but these errors were encountered:
首先这个在主页写清楚了,在这里也重新明确表达一下,safety-checker 专用于 R15,nsfw-model 专用于 R18。两个模型可以适用于绝大多数场景。
下面讲讲为什么(检测数值我用 strength 表示):
nsfw-model 的结果虽然是 probs,但是我对其进行研究后,是无法判断的。对于特殊的检测,建议你自行训练它的模型,它用的就是原版 mobilenet v2 和官方的训练脚本,只是套用了 R18 图片来检测。
safety-checker 能做到的事情更多,因为它是基于 CLIP 的一种概念模型。但是关于它能够实现 strength 目前还有很大争议,事实上 stable diffusion 和 clip 那边都没有说它能使用 strength 参数(代码里写了但是算是 experimental)。
关于 safety-checker 的更多事情可以看 iyume/safety-checker#1
Sorry, something went wrong.
No branches or pull requests
No description provided.
The text was updated successfully, but these errors were encountered: