Type: huggingface.text_to_image.Flux
Namespace: huggingface.text_to_image
Description
Generates images using FLUX models with support for GGUF quantization for memory efficiency. image, generation, AI, text-to-image, flux, quantization
Use cases:
- High-quality image generation with FLUX models
- Memory-efficient generation using GGUF quantization
- Fast generation with FLUX.1-schnell
- High-fidelity generation with FLUX.1-dev
- Controlled generation with Fill, Canny, or Depth variants
Properties
| Property | Type | Description | Default |
|---|---|---|---|
| model | hf.flux |
The FLUX model to use for text-to-image generation. | {'type': 'hf.flux', 'repo_id': '', 'path': None, 'variant': None, 'allow_patterns': None, 'ignore_patterns': None} |
| prompt | str |
A text prompt describing the desired image. | A cat holding a sign that says hello world |
| guidance_scale | float |
The scale for classifier-free guidance. Use 0.0 for schnell, 3.5 for dev. | 3.5 |
| width | int |
The width of the generated image. | 1024 |
| height | int |
The height of the generated image. | 1024 |
| num_inference_steps | int |
The number of denoising steps. 4 steps is forced for schnell models. | 20 |
| max_sequence_length | int |
Maximum sequence length for the prompt. Use 256 for schnell, 512 for dev. | 512 |
| seed | int |
Seed for the random number generator. Use -1 for a random seed. | -1 |
| enable_cpu_offload | bool |
Enable CPU offload to reduce VRAM usage. | True |
Outputs
| Output | Type | Description |
|---|---|---|
| output | image |
Metadata
Related Nodes
Browse other nodes in the huggingface.text_to_image namespace.