Merge branch 'main' into to-single-file/wan

Author: a-r-r-o-w

src/diffusers/modular_pipelines/modular_pipeline.py

@@ -323,6 +323,7 @@ class ModularPipelineBlocks(ConfigMixin, PushToHubMixin):
     """
 
     config_name = "config.json"
+    model_name = None
 
     @classmethod
     def _get_signature_keys(cls, obj):
@@ -333,6 +334,14 @@ def _get_signature_keys(cls, obj):
 
         return expected_modules, optional_parameters
 
+    @property
+    def expected_components(self) -> List[ComponentSpec]:
+        return []
+
+    @property
+    def expected_configs(self) -> List[ConfigSpec]:
+        return []
+
     @classmethod
     def from_pretrained(
         cls,
@@ -358,7 +367,9 @@ def from_pretrained(
             trust_remote_code, pretrained_model_name_or_path, has_remote_code
         )
         if not (has_remote_code and trust_remote_code):
-            raise ValueError("TODO")
+            raise ValueError(
+                "Selected model repository does not happear to have any custom code or does not have a valid `config.json` file."
+            )
 
         class_ref = config["auto_map"][cls.__name__]
         module_file, class_name = class_ref.split(".")
@@ -367,7 +378,6 @@ def from_pretrained(
             pretrained_model_name_or_path,
             module_file=module_file,
             class_name=class_name,
-            is_modular=True,
             **hub_kwargs,
             **kwargs,
         )

src/diffusers/modular_pipelines/modular_pipeline_utils.py

@@ -93,7 +93,7 @@ class ComponentSpec:
     config: Optional[FrozenDict] = None
     # YiYi Notes: should we change it to pretrained_model_name_or_path for consistency? a bit long for a field name
     repo: Optional[Union[str, List[str]]] = field(default=None, metadata={"loading": True})
-    subfolder: Optional[str] = field(default=None, metadata={"loading": True})
+    subfolder: Optional[str] = field(default="", metadata={"loading": True})
     variant: Optional[str] = field(default=None, metadata={"loading": True})
     revision: Optional[str] = field(default=None, metadata={"loading": True})
     default_creation_method: Literal["from_config", "from_pretrained"] = "from_pretrained"

src/diffusers/pipelines/dit/pipeline_dit.py

@@ -46,7 +46,9 @@ class DiTPipeline(DiffusionPipeline):
 
     Parameters:
         transformer ([`DiTTransformer2DModel`]):
-            A class conditioned `DiTTransformer2DModel` to denoise the encoded image latents.
+            A class conditioned `DiTTransformer2DModel` to denoise the encoded image latents. Initially published as
+            [`Transformer2DModel`](https://huggingface.co/facebook/DiT-XL-2-256/blob/main/transformer/config.json#L2)
+            in the config, but the mismatch can be ignored.
         vae ([`AutoencoderKL`]):
             Variational Auto-Encoder (VAE) model to encode and decode images to and from latent representations.
         scheduler ([`DDIMScheduler`]):

src/diffusers/pipelines/pixart_alpha/pipeline_pixart_alpha.py

@@ -256,7 +256,9 @@ class PixArtAlphaPipeline(DiffusionPipeline):
             Tokenizer of class
             [T5Tokenizer](https://huggingface.co/docs/transformers/model_doc/t5#transformers.T5Tokenizer).
         transformer ([`PixArtTransformer2DModel`]):
-            A text conditioned `PixArtTransformer2DModel` to denoise the encoded image latents.
+            A text conditioned `PixArtTransformer2DModel` to denoise the encoded image latents. Initially published as
+            [`Transformer2DModel`](https://huggingface.co/PixArt-alpha/PixArt-XL-2-1024-MS/blob/main/transformer/config.json#L2)
+            in the config, but the mismatch can be ignored.
         scheduler ([`SchedulerMixin`]):
             A scheduler to be used in combination with `transformer` to denoise the encoded image latents.
     """

src/diffusers/pipelines/pixart_alpha/pipeline_pixart_sigma.py

@@ -185,6 +185,26 @@ def retrieve_timesteps(
 class PixArtSigmaPipeline(DiffusionPipeline):
     r"""
     Pipeline for text-to-image generation using PixArt-Sigma.
+
+    This model inherits from [`DiffusionPipeline`]. Check the superclass documentation for the generic methods the
+    library implements for all the pipelines (such as downloading or saving, running on a particular device, etc.)
+
+    Args:
+        vae ([`AutoencoderKL`]):
+            Variational Auto-Encoder (VAE) Model to encode and decode images to and from latent representations.
+        text_encoder ([`T5EncoderModel`]):
+            Frozen text-encoder. PixArt-Alpha uses
+            [T5](https://huggingface.co/docs/transformers/model_doc/t5#transformers.T5EncoderModel), specifically the
+            [t5-v1_1-xxl](https://huggingface.co/PixArt-alpha/PixArt-alpha/tree/main/t5-v1_1-xxl) variant.
+        tokenizer (`T5Tokenizer`):
+            Tokenizer of class
+            [T5Tokenizer](https://huggingface.co/docs/transformers/model_doc/t5#transformers.T5Tokenizer).
+        transformer ([`PixArtTransformer2DModel`]):
+            A text conditioned `PixArtTransformer2DModel` to denoise the encoded image latents. Initially published as
+            [`Transformer2DModel`](https://huggingface.co/PixArt-alpha/PixArt-Sigma-XL-2-1024-MS/blob/main/transformer/config.json#L2)
+            in the config, but the mismatch can be ignored.
+        scheduler ([`SchedulerMixin`]):
+            A scheduler to be used in combination with `transformer` to denoise the encoded image latents.
     """
 
     bad_punct_regex = re.compile(