[add pipeline]: A new Pixart-based inpainting pipeline.

docs/source/en/api/pipelines/pixart_inpaint.md

@@ -0,0 +1,29 @@
+# PixArt Inpainting
+The inpainting method based on the PixArt model is similar to the usage of the Stable Diffusion model for inpainting.
+
+```python
+import torch
+
+import diffusers
+from diffusers import PixArtAlphaInpaintPipeline
+
+img_url = "https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo.png"
+mask_url = "https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo_mask.png"
+
+init_image = diffusers.utils.load_image(img_url).resize((1024, 1024))
+mask_image = diffusers.utils.load_image(mask_url).resize((1024, 1024))
+
+# You can replace the checkpoint id with "PixArt-alpha/PixArt-XL-2-512x512" too
+pipe = PixArtAlphaInpaintPipeline.from_pretrained(
+    "PixArt-alpha/PixArt-XL-2-1024-MS", torch_dtype=torch.float16
+     )
+pipe = pipe.to("cuda")
+
+prompt = "Face of a yellow cat, high resolution, sitting on a park bench"
+image = pipe(prompt=prompt, image=init_image, mask_image=mask_image).images[0]
+```
+## PixArtAlphaPipeline
+
+[[autodoc]] PixArtAlphaPipeline
+	- all
+	- __call__

src/diffusers/__init__.py

@@ -263,6 +263,7 @@
             "PaintByExamplePipeline",
             "PIAPipeline",
             "PixArtAlphaPipeline",
+            "PixArtAlphaInpaintPipeline",
             "PixArtSigmaPipeline",
             "SemanticStableDiffusionPipeline",
             "ShapEImg2ImgPipeline",
@@ -640,6 +641,7 @@
             MusicLDMPipeline,
             PaintByExamplePipeline,
             PIAPipeline,
+            PixArtAlphaInpaintPipeline,
             PixArtAlphaPipeline,
             PixArtSigmaPipeline,
             SemanticStableDiffusionPipeline,

src/diffusers/pipelines/__init__.py

@@ -188,7 +188,7 @@
     _import_structure["musicldm"] = ["MusicLDMPipeline"]
     _import_structure["paint_by_example"] = ["PaintByExamplePipeline"]
     _import_structure["pia"] = ["PIAPipeline"]
-    _import_structure["pixart_alpha"] = ["PixArtAlphaPipeline", "PixArtSigmaPipeline"]
+    _import_structure["pixart_alpha"] = ["PixArtAlphaPipeline", "PixArtSigmaPipeline", "PixArtAlphaInpaintPipeline"]
     _import_structure["semantic_stable_diffusion"] = ["SemanticStableDiffusionPipeline"]
     _import_structure["shap_e"] = ["ShapEImg2ImgPipeline", "ShapEPipeline"]
     _import_structure["stable_cascade"] = [
@@ -451,7 +451,7 @@
         from .musicldm import MusicLDMPipeline
         from .paint_by_example import PaintByExamplePipeline
         from .pia import PIAPipeline
-        from .pixart_alpha import PixArtAlphaPipeline, PixArtSigmaPipeline
+        from .pixart_alpha import PixArtAlphaInpaintPipeline, PixArtAlphaPipeline, PixArtSigmaPipeline
         from .semantic_stable_diffusion import SemanticStableDiffusionPipeline
         from .shap_e import ShapEImg2ImgPipeline, ShapEPipeline
         from .stable_cascade import (

src/diffusers/pipelines/pixart_alpha/__init__.py

@@ -24,6 +24,7 @@
 else:
     _import_structure["pipeline_pixart_alpha"] = ["PixArtAlphaPipeline"]
     _import_structure["pipeline_pixart_sigma"] = ["PixArtSigmaPipeline"]
+    _import_structure["pipeline_pixart_inpaint"] = ["PixArtAlphaInpaintPipeline"]
 
 if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
     try:
@@ -39,6 +40,7 @@
             ASPECT_RATIO_1024_BIN,
             PixArtAlphaPipeline,
         )
+        from .pipeline_pixart_inpaint import PixArtAlphaInpaintPipeline
         from .pipeline_pixart_sigma import ASPECT_RATIO_2048_BIN, PixArtSigmaPipeline
 
 else: