feedback

Author: stevhliu

docs/source/en/tutorials/using_peft_for_inference.md

@@ -79,7 +79,7 @@ The [`~loaders.StableDiffusionLoraLoaderMixin.load_lora_weights`] method is the
 - the LoRA weights don't have separate UNet and text encoder identifiers
 - the LoRA weights have separate UNet and text encoder identifiers
 
-The [`~loaders.PeftAdapterMixin.load_lora_adapter`] method is used to directly load a LoRA adapter at the *model-level*. It builds and prepares the necessary model configuration for the adapter. This method can also load the LoRA adapter into the UNet and text encoder.
+The [`~loaders.PeftAdapterMixin.load_lora_adapter`] method is used to directly load a LoRA adapter at the *model-level*, as long as the model is a Diffusers model that is a subclass of [`PeftAdapterMixin`]. It builds and prepares the necessary model configuration for the adapter. This method also loads the LoRA adapter into the UNet.
 
 For example, if you're only loading a LoRA into the UNet, [`~loaders.PeftAdapterMixin.load_lora_adapter`] ignores the text encoder keys. Use the `prefix` parameter to filter and load the appropriate state dicts, `"unet"` to load.
 
@@ -629,4 +629,6 @@ Browse the [LoRA Studio](https://lorastudio.co/models) for different LoRAs to us
 	frameborder="0"
 	width="850"
 	height="450"
-></iframe>
+></iframe>
+
+You can find additional LoRAs in the [FLUX LoRA the Explorer](https://huggingface.co/spaces/multimodalart/flux-lora-the-explorer) and [LoRA the Explorer](https://huggingface.co/spaces/multimodalart/LoraTheExplorer) Spaces.

docs/source/en/using-diffusers/controlnet.md

@@ -14,7 +14,10 @@ specific language governing permissions and limitations under the License.
 
 [ControlNet](https://huggingface.co/papers/2302.05543) is an adapter that enables controllable generation such as generating an image of a cat in a *specific pose* or following the lines in a sketch of a *specific* cat. It works by adding a smaller network of "zero convolution" layers and progressively training these to avoid disrupting with the original model. The original model parameters are frozen to avoid retraining it.
 
-A ControlNet is conditioned on extra visual information or "controls" (canny edge, depth maps, human pose, etc.) that can be combined with text prompts to generate images that are guided by the visual input.
+A ControlNet is conditioned on extra visual information or "structural controls" (canny edge, depth maps, human pose, etc.) that can be combined with text prompts to generate images that are guided by the visual input.
+
+> [!TIP]
+> ControlNets are available to many models such as [Flux](../api/pipelines/controlnet_flux), [Hunyuan-DiT](../api/pipelines/controlnet_hunyuandit), [Stable Diffusion 3](../api/pipelines/controlnet_sd3), and more. The examples in this guide use Flux and Stable Diffusion XL.
 
 Load a ControlNet conditioned on a specific control, such as canny edge, and pass it to the pipeline in [`~DiffusionPipeline.from_pretrained`].
 

docs/source/en/using-diffusers/ip_adapter.md

@@ -16,6 +16,9 @@ specific language governing permissions and limitations under the License.
 
 IP-Adapter files are typically ~100MBs because they only contain the image embeddings. This means you need to load a model first, and then load the IP-Adapter with [`~loaders.IPAdapterMixin.load_ip_adapter`].
 
+> [!TIP]
+> IP-Adapters are available to many models such as [Flux](../api/pipelines/flux#ip-adapter) and [Stable Diffusion 3](../api/pipelines/stable_diffusion/stable_diffusion_3), and more. The examples in this guide use Stable Diffusion and Stable Diffusion XL.
+
 Use the [`~loaders.IPAdapterMixin.set_ip_adapter_scale`] parameter to scale the influence of the IP-Adapter during generation. A value of `1.0` means the model is only conditioned on the image prompt, and `0.5` typically produces balanced results between the text and image prompt.
 
 ```py