devgmstudios/nsfw-filter-adjustable | Run with an API on Replicate

Adjustable Stable Diffusion NSFW safety checker. Higher threshold = less aggressive (fewer flags); 0 = most aggressive (stock). Fork of m1guelpf/nsfw-filter.

Public

320 runs

License

GitHub

Playground API Examples README Versions

Run time and cost

This model runs on Nvidia T4 GPU hardware. We don't yet have enough runs of this model to provide performance information.

Readme

An isolated version of Stable Diffusion’s content filter, which lets you run it against arbitrary images.

Input	Default	Meaning
`threshold`	`0`	NSFW sensitivity. Flag when concept score exceeds this value.
`special_threshold`	`0`	Special-care sensitivity. Same scale as `threshold`.

Scale (same for both):

Value	Behavior
`0` (default)	Stock / strictest
`0.02`	More lenient
`0.04`	Even more lenient

Higher = more lenient; lower toward 0 = stricter.

Model created 3 days, 20 hours ago

Model updated 5 hours ago