[PEFT / docs] Add a note about torch.compile (#6864)

* Update using_peft_for_inference.md

* add more explanation

Author: younesbelkada

docs/source/en/tutorials/using_peft_for_inference.md

@@ -165,6 +165,25 @@ list_adapters_component_wise
 {"text_encoder": ["toy", "pixel"], "unet": ["toy", "pixel"], "text_encoder_2": ["toy", "pixel"]}
 ```
 
+## Compatibility with `torch.compile`
+
+If you want to compile your model with `torch.compile` make sure to first fuse the LoRA weights into the base model and unload them.
+
+```py
+pipe.load_lora_weights("nerijs/pixel-art-xl", weight_name="pixel-art-xl.safetensors", adapter_name="pixel")
+pipe.load_lora_weights("CiroN2022/toy-face", weight_name="toy_face_sdxl.safetensors", adapter_name="toy")
+
+pipe.set_adapters(["pixel", "toy"], adapter_weights=[0.5, 1.0])
+# Fuses the LoRAs into the Unet
+pipe.fuse_lora()
+pipe.unload_lora_weights()
+
+pipe = torch.compile(pipe)
+
+prompt = "toy_face of a hacker with a hoodie, pixel art"
+image = pipe(prompt, num_inference_steps=30, generator=torch.manual_seed(0)).images[0]
+```
+
 ## Fusing adapters into the model
 
 You can use PEFT to easily fuse/unfuse multiple adapters directly into the model weights (both UNet and text encoder) using the [`~diffusers.loaders.LoraLoaderMixin.fuse_lora`] method, which can lead to a speed-up in inference and lower VRAM usage.