AI Models - skier233/nsfw_ai_model_server GitHub Wiki
Here is where you can find the best AI Models currently available to use with this AI server. To install/use an AI model, unzip the file and copy the folders into the folder where the AI Server is installed. (ie copy the config and models into the top-level server folder)
Tagging Models
| Model Name | Patron Tier | Tag Category | Tag Count | Purpose | Link |
|---|---|---|---|---|---|
| vivid galaxy | Free | Sexual Actions | 10 | Free Model with 10 Sexual Action Tags | https://mega.nz/file/xG41na4R#TJ1U_fl866sv9S4WMlNc8gDZISvp-6gHIdd8jmRRCHM |
| actions_30 | Member | Sexual Actions | 36 | Highest Accuracy Sexual Action Model | https://www.patreon.com/Skier/posts/actions-ai-model-164208669 |
| bdsm_20 | Member | BDSM | 50 | BDSM Model | https://www.patreon.com/Skier/posts/bdsm-ai-model-164208699 |
| body_20 | Member | Body Parts | 37 | Body Parts Model | https://www.patreon.com/Skier/posts/body-parts-ai-164208713 |
| positions_20 | VIP | Positions Model Beta | 28 | Highest Accuracy Positions Model | https://www.patreon.com/Skier/posts/positions-ai-164208724 |
Other Models
| Model Name | Patron Tier | Purpose | Link |
|---|---|---|---|
| face_rec_detector & face_rec_embedder | Free | Facial recognition models (for use with AI Faces in Cove) | https://mega.nz/file/tPICHCSI#5MeQOFYDkVy9kaPLi6Jds3h-ck6Fp9ZgJQq_ofTCf-M |
| audioclass & audioembed | Member | Audio Intelligence models to be able to classify audio for use with recommender systems or finding similar content by audio | https://www.patreon.com/Skier/posts/audio-ai-models-164208849 |
| visual | Member | Visual Intelligence model for recommenders & to be able to find similar visual content | https://www.patreon.com/Skier/posts/visual-ai-models-164208815 |
| semvisual | Member | Semantic Visual Intelligence model for recommenders & to be able to search through content with words and phrases | https://www.patreon.com/Skier/posts/visual-ai-models-164208815 |
Face Packs
These files contain pre-computed face data for online sources of performer data and can be imported with the AI Faces extension in Cove to improve facial recognition capabilities.
| Face Source | Performer Count | Url |
|---|---|---|
| StashDB | 98,134 | https://mega.nz/file/kfpAiZ4a#O1mfDrExGIJGtaQYJyb_xDZySwWcJIxQsyrfzJSK1TU |
| FansDB | 14,486 | https://mega.nz/file/QawXEJoZ#Z3mlSVaTl2XYMIgVXeVOPke9Wf61E8zrO-tdrfo1YLo |
| TPDB | 4,453 | https://mega.nz/file/hW5UQCSB#duy4DCbl4haK99U6TDset4-0ADiYfwzD4P7kWjCcktk |
Switching the currently enabled models
To switch which AI models are currently active, go to the Settings->Extensions->AI and you can load or unload models.
If you are still using stash and haven't migrated to Cove yet, run (windows) .\select_models.ps1 or (linux) ./select_models.sh and choose which models you want currently active.