AI Image Gen · Checkpoint
InvictusRedmond - Stable Cascade Stage C Generalist Finetune
Generalist Stable Cascade Stage C finetune for realistic, artistic, car, and character images.
About this model
InvictusRedmond is a generalist checkpoint by ArtificialGuyBR, fine-tuned on Stable Cascade rather than SDXL 1.0. The author describes it as a versatile model with high capacity for realistic and artistic images, cars, people, and a wide variety of themes. It is a full Stable Cascade Stage C (prior) checkpoint, so it must be used inside a Stable Cascade three-stage pipeline with the Stage B decoder and the previewer VAE/text encoder, not as an SDXL drop-in.
The recommended version is v1.1 - Less training, published March 1, 2024. The author notes this version shows better results than the base model and v1.0, at the cost of being a little closer to the original Stable Cascade look; the model turns out to be extremely sensitive to longer training runs even at low learning rates, which is why the lighter fine-tune is preferred. The download is a 6.69 GB BF16 SafeTensor ("InvictusRedmond-StageC-Updated.safetensors") - a single Stage C checkpoint with no trigger words or instance prompt required.
In ComfyUI, place the Stage C checkpoint in the models/unet folder and build a Stable Cascade workflow (Stage C prior -> Stage B decoder -> previewer/VAE). No special prompt syntax is needed - standard descriptive prompts work, as expected from a generalist model. License details were not stated on the published page.
Base model
SDXL 1.0
Example outputs
Learn more
Read the in-depth guide about InvictusRedmond - Stable Cascade Stage C Generalist Finetune
Project signals
- 3,547 Civitai downloads
Topics
base model · anime · realistic · female · woman · generalist · game character · photorealistic