Stable Diffusion XL (Img2Img)

Type: huggingface.image_to_image.StableDiffusionXLImg2Img

Namespace: huggingface.image_to_image

Description

Transforms existing images based on text prompts using Stable Diffusion XL. image, generation, image-to-image, SDXL, style-transfer

Use cases:
- Modifying existing images to fit a specific style or theme
- Enhancing or altering photographs
- Creating variations of existing artwork
- Applying text-guided edits to images

Properties

Property Type Description Default
model hf.stable_diffusion_xl The Stable Diffusion XL model to use for generation. {'type': 'hf.stable_diffusion_xl', 'repo_id': '', 'path': None, 'variant': None, 'allow_patterns': None, 'ignore_patterns': None}
variant Enum['default', 'fp16', 'fp32', 'bf16'] The variant of the model to use for generation. fp16
prompt str The prompt for image generation. ``
negative_prompt str The negative prompt to guide what should not appear in the generated image. ``
width int Width of the generated image. 1024
height int Height of the generated image 1024
seed int Seed for the random number generator. -1
num_inference_steps int Number of inference steps. 25
guidance_scale float Guidance scale for generation. 7.0
scheduler Enum['DPMSolverSDEScheduler', 'EulerDiscreteScheduler', 'LMSDiscreteScheduler', 'DDIMScheduler', 'DDPMScheduler', 'HeunDiscreteScheduler', 'DPMSolverMultistepScheduler', 'DEISMultistepScheduler', 'PNDMScheduler', 'EulerAncestralDiscreteScheduler', 'UniPCMultistepScheduler', 'KDPM2DiscreteScheduler', 'DPMSolverSinglestepScheduler', 'KDPM2AncestralDiscreteScheduler'] The scheduler to use for the diffusion process. EulerDiscreteScheduler
pag_scale float Scale of the Perturbed-Attention Guidance applied to the image. 3.0
loras List[hf.lora_sdxl_config] The LoRA models to use for image processing []
lora_scale float Strength of the LoRAs 0.5
ip_adapter_model hf.ip_adapter The IP adapter model to use for image processing {'type': 'hf.ip_adapter', 'repo_id': '', 'path': None, 'variant': None, 'allow_patterns': None, 'ignore_patterns': None}
ip_adapter_image image When provided the image will be fed into the IP adapter {'type': 'image', 'uri': '', 'asset_id': None, 'data': None}
ip_adapter_scale float Strength of the IP adapter image 0.5
enable_attention_slicing bool Enable attention slicing for the pipeline. This can reduce VRAM usage. True
enable_tiling bool Enable tiling for the VAE. This can reduce VRAM usage. False
enable_cpu_offload bool Enable CPU offload for the pipeline. This can reduce VRAM usage. False
output_type Enum['Image', 'Latent'] The type of output to generate. Image
init_image image The initial image for Image-to-Image generation. {'type': 'image', 'uri': '', 'asset_id': None, 'data': None}
strength float Strength for Image-to-Image generation. Higher values allow for more deviation from the original image. 0.8

Outputs

Output Type Description
image image  
latent torch_tensor  

Metadata

Browse other nodes in the huggingface.image_to_image namespace.