算力平台:
管道
管道提供了一种简单的方法，通过将所有必要的组件（多个独立训练的模型、调度器和处理器）捆绑到一个端到端的类中，来在推理过程中运行最先进的扩散模型。管道是灵活的，可以适应使用不同的调度器甚至模型组件。
所有管道都基于 [DiffusionPipeline] 类构建，该类提供了加载、下载和保存所有组件的基本功能。特定的管道类型（例如 [StableDiffusionPipeline]）使用 [~DiffusionPipeline.from_pretrained] 加载时，会自动检测并加载管道组件，并传递给管道的 __init__ 函数。
下表列出了 🤗 Diffusers 中目前可用的所有管道及其支持的任务。点击一个管道以查看其摘要和已发表的论文。
| Pipeline | Tasks |
|---|---|
| aMUSEd | text2image |
| AnimateDiff | text2video |
| Attend-and-Excite | text2image |
| AudioLDM | text2audio |
| AudioLDM2 | text2audio |
| AuraFlow | text2image |
| BLIP Diffusion | text2image |
| CogVideoX | text2video |
| Consistency Models | unconditional image generation |
| ControlNet | text2image, image2image, inpainting |
| ControlNet with Flux.1 | text2image |
| ControlNet with Hunyuan-DiT | text2image |
| ControlNet with Stable Diffusion 3 | text2image |
| ControlNet with Stable Diffusion XL | text2image |
| ControlNet-XS | text2image |
| ControlNet-XS with Stable Diffusion XL | text2image |
| Dance Diffusion | unconditional audio generation |
| DDIM | unconditional image generation |
| DDPM | unconditional image generation |
| DeepFloyd IF | text2image, image2image, inpainting, super-resolution |
| DiffEdit | inpainting |
| DiT | text2image |
| Flux | text2image |
| Hunyuan-DiT | text2image |
| I2VGen-XL | text2video |
| InstructPix2Pix | image editing |
| Kandinsky 2.1 | text2image, image2image, inpainting, interpolation |
| Kandinsky 2.2 | text2image, image2image, inpainting |
| Kandinsky 3 | text2image, image2image |
| Kolors | text2image |
| Latent Consistency Models | text2image |
| Latent Diffusion | text2image, super-resolution |
| Latte | text2image |
| LEDITS++ | image editing |
| Lumina-T2X | text2image |
| Marigold | depth |
| MultiDiffusion | text2image |
| MusicLDM | text2audio |
| PAG | text2image |
| Paint by Example | inpainting |
| PIA | image2video |
| PixArt-α | text2image |
| PixArt-Σ | text2image |
| Self-Attention Guidance | text2image |
| Semantic Guidance | text2image |
| Shap-E | text-to-3D, image-to-3D |
| Stable Audio | text2audio |
| Stable Cascade | text2image |
| Stable Diffusion | text2image, image2image, depth2image, inpainting, image variation, latent upscaler, super-resolution |
| Stable Diffusion XL | text2image, image2image, inpainting |
| Stable Diffusion XL Turbo | text2image, image2image, inpainting |
| Stable unCLIP | text2image, image variation |
| T2I-Adapter | text2image |
| Text2Video | text2video, video2video |
| Text2Video-Zero | text2video |
| unCLIP | text2image, image variation |
| UniDiffuser | text2image, image2text, image variation, text variation, unconditional image generation, unconditional audio generation |
| Value-guided planning | value guided sampling |
| Wuerstchen | text2image |
DiffusionPipeline
[[autodoc]] DiffusionPipeline - all - call - device - to - components
[[autodoc]] pipelines.StableDiffusionMixin.enable_freeu
[[autodoc]] pipelines.StableDiffusionMixin.disable_freeu
FlaxDiffusionPipeline
[[autodoc]] pipelines.pipeline_flax_utils.FlaxDiffusionPipeline
PushToHubMixin
[[autodoc]] utils.PushToHubMixin