Stable Diffusion V1 Model Plugin for dashAI

This plugin integrates CompVis' Stable Diffusion 1.x text-to-image models into the dashAI platform using the Hugging Face diffusers backend. It enables flexible image generation from text prompts, with configurable inference and sampling parameters.

Included Models

1. Stable Diffusion v1-1

Original 1.x checkpoint from CompVis, based on CompVis/stable-diffusion-v1-1.

2. Stable Diffusion v1-2

Continued training over v1-1, based on CompVis/stable-diffusion-v1-2.

3. Stable Diffusion v1-3

Continued training with classifier-free guidance, based on CompVis/stable-diffusion-v1-3.

4. Stable Diffusion v1-4

Most refined 1.x checkpoint, based on CompVis/stable-diffusion-v1-4.

All models are latent text-to-image diffusion models, designed for high-quality image generation and compatible with CPU or GPU inference through diffusers.

About Stable Diffusion

Stable Diffusion is a latent text-to-image diffusion model developed by CompVis, Stability AI, and Runway. It generates photo-realistic images from natural language descriptions.

Key features of Stable Diffusion 1.x models:

• Latent diffusion: generation runs in a compressed latent space for efficiency
• Text-conditioned: images are guided by natural language prompts
• Negative prompting: steer the model away from unwanted elements
• Open weights: freely available checkpoints
• Configurable resolution: supports a range of output dimensions

Stable Diffusion is designed for deployment on laptops, desktops, and cloud infrastructure, making advanced image generation more accessible.

Features

• Image generation from text prompts
• Multiple 1.x checkpoints selectable at runtime
• Reproducible generation via fixed random seeds
• Configurable generation parameters:
  • negative_prompt: elements to avoid in the image
  • num_inference_steps: number of denoising steps
  • guidance_scale: how strongly the model follows the prompt
  • width / height: output image dimensions
  • num_images_per_prompt: number of images per prompt
  • device: detected GPU (e.g. "GPU 0: NVIDIA RTX 3090 - Compute Capability 8.6") or "CPU"

Model Parameters

Parameter	Description	Default
model_name	Stable Diffusion 1.x checkpoint from Hugging Face	"CompVis/stable-diffusion-v1-1"
negative_prompt	Text prompt for elements to avoid in the image	"" (optional)
num_inference_steps	Number of denoising steps (higher = better quality, slower)	15
guidance_scale	How strongly the model follows the prompt	3.5
seed	Random seed for reproducibility (negative = random)	-1
width	Width of the generated image (multiple of 8)	512
height	Height of the generated image (multiple of 8)	512
num_images_per_prompt	Number of images to generate per prompt	1
device	Inference device — a detected GPU by name (e.g. "GPU 0: NVIDIA RTX 3090 - Compute Capability 8.6") or "CPU"	First detected GPU if available, else "CPU"

Requirements

• diffusers
• torch
• Model files from Hugging Face:
  • CompVis/stable-diffusion-v1-1
  • CompVis/stable-diffusion-v1-2
  • CompVis/stable-diffusion-v1-3
  • CompVis/stable-diffusion-v1-4

Notes

This plugin uses the Hugging Face diffusers pipeline for text-to-image generation.

The models are pretrained for inference and are not designed for fine-tuning through this plugin. Image dimensions must be multiples of 8.