https://civitai.com/models/926992?modelVersionId=1274366
Trained on noobai1.1 with a larger dataset featuring different tags. For example, I didn’t use 'disembodied' for severed limbs and tried to differentiate between (double/triple/quadruple amputees), or (bloody/flesh/smooth arm/leg/neck stumps). You can find more tags in the txtsets.

As I mentioned on Civitai, the results are a bit disappointing. Some tags don’t work properly.

Also, I’m not sure if it’s due to the noobai1.1 model, but the outputs with noobai1.1 look terrible.

If you’re interested, I can upload different epochs.