feat(api): chore: pydantic & fastapi upgrade

Upgrade pydantic and fastapi to latest.

- pydantic~=2.4.2
- fastapi~=103.2
- fastapi-events~=0.9.1

**Big Changes**

There are a number of logic changes needed to support pydantic v2. Most changes are very simple, like using the new methods to serialized and deserialize models, but there are a few more complex changes.

**Invocations**

The biggest change relates to invocation creation, instantiation and validation.

Because pydantic v2 moves all validation logic into the rust pydantic-core, we may no longer directly stick our fingers into the validation pie.

Previously, we (ab)used models and fields to allow invocation fields to be optional at instantiation, but required when `invoke()` is called. We directly manipulated the fields and invocation models when calling `invoke()`.

With pydantic v2, this is much more involved. Changes to the python wrapper do not propagate down to the rust validation logic - you have to rebuild the model. This causes problem with concurrent access to the invocation classes and is not a free operation.

This logic has been totally refactored and we do not need to change the model any more. The details are in `baseinvocation.py`, in the `InputField` function and `BaseInvocation.invoke_internal()` method.

In the end, this implementation is cleaner.

**Invocation Fields**

In pydantic v2, you can no longer directly add or remove fields from a model.

Previously, we did this to add the `type` field to invocations.

**Invocation Decorators**

With pydantic v2, we instead use the imperative `create_model()` API to create a new model with the additional field. This is done in `baseinvocation.py` in the `invocation()` wrapper.

A similar technique is used for `invocation_output()`.

**Minor Changes**

There are a number of minor changes around the pydantic v2 models API.

**Protected `model_` Namespace**

All models' pydantic-provided methods and attributes are prefixed with `model_` and this is considered a protected namespace. This causes some conflict, because "model" means something to us, and we have a ton of pydantic models with attributes starting with "model_".

Forunately, there are no direct conflicts. However, in any pydantic model where we define an attribute or method that starts with "model_", we must tell set the protected namespaces to an empty tuple.

```py
class IPAdapterModelField(BaseModel):
    model_name: str = Field(description="Name of the IP-Adapter model")
    base_model: BaseModelType = Field(description="Base model")

    model_config = ConfigDict(protected_namespaces=())
```

**Model Serialization**

Pydantic models no longer have `Model.dict()` or `Model.json()`.

Instead, we use `Model.model_dump()` or `Model.model_dump_json()`.

**Model Deserialization**

Pydantic models no longer have `Model.parse_obj()` or `Model.parse_raw()`, and there are no `parse_raw_as()` or `parse_obj_as()` functions.

Instead, you need to create a `TypeAdapter` object to parse python objects or JSON into a model.

```py
adapter_graph = TypeAdapter(Graph)
deserialized_graph_from_json = adapter_graph.validate_json(graph_json)
deserialized_graph_from_dict = adapter_graph.validate_python(graph_dict)
```

**Field Customisation**

Pydantic `Field`s no longer accept arbitrary args.

Now, you must put all additional arbitrary args in a `json_schema_extra` arg on the field.

**Schema Customisation**

FastAPI and pydantic schema generation now follows the OpenAPI version 3.1 spec.

This necessitates two changes:
- Our schema customization logic has been revised
- Schema parsing to build node templates has been revised

The specific aren't important, but this does present additional surface area for bugs.

**Performance Improvements**

Pydantic v2 is a full rewrite with a rust backend. This offers a substantial performance improvement (pydantic claims 5x to 50x depending on the task). We'll notice this the most during serialization and deserialization of sessions/graphs, which happens very very often - a couple times per node.

I haven't done any benchmarks, but anecdotally, graph execution is much faster. Also, very larges graphs - like with massive iterators - are much, much faster.

invokeai/app/invocations/image.py

@@ -36,7 +36,13 @@ class ShowImageInvocation(BaseInvocation):
         )
 
 
-@invocation("blank_image", title="Blank Image", tags=["image"], category="image", version="1.0.0")
+@invocation(
+    "blank_image",
+    title="Blank Image",
+    tags=["image"],
+    category="image",
+    version="1.0.0",
+)
 class BlankImageInvocation(BaseInvocation):
     """Creates a blank image and forwards it to the pipeline"""
 
@@ -65,7 +71,13 @@ class BlankImageInvocation(BaseInvocation):
         )
 
 
-@invocation("img_crop", title="Crop Image", tags=["image", "crop"], category="image", version="1.0.0")
+@invocation(
+    "img_crop",
+    title="Crop Image",
+    tags=["image", "crop"],
+    category="image",
+    version="1.0.0",
+)
 class ImageCropInvocation(BaseInvocation):
     """Crops an image to a specified box. The box can be outside of the image."""
 
@@ -98,7 +110,13 @@ class ImageCropInvocation(BaseInvocation):
         )
 
 
-@invocation("img_paste", title="Paste Image", tags=["image", "paste"], category="image", version="1.0.1")
+@invocation(
+    "img_paste",
+    title="Paste Image",
+    tags=["image", "paste"],
+    category="image",
+    version="1.0.1",
+)
 class ImagePasteInvocation(BaseInvocation):
     """Pastes an image into another image."""
 
@@ -151,7 +169,13 @@ class ImagePasteInvocation(BaseInvocation):
         )
 
 
-@invocation("tomask", title="Mask from Alpha", tags=["image", "mask"], category="image", version="1.0.0")
+@invocation(
+    "tomask",
+    title="Mask from Alpha",
+    tags=["image", "mask"],
+    category="image",
+    version="1.0.0",
+)
 class MaskFromAlphaInvocation(BaseInvocation):
     """Extracts the alpha channel of an image as a mask."""
 
@@ -182,7 +206,13 @@ class MaskFromAlphaInvocation(BaseInvocation):
         )
 
 
-@invocation("img_mul", title="Multiply Images", tags=["image", "multiply"], category="image", version="1.0.0")
+@invocation(
+    "img_mul",
+    title="Multiply Images",
+    tags=["image", "multiply"],
+    category="image",
+    version="1.0.0",
+)
 class ImageMultiplyInvocation(BaseInvocation):
     """Multiplies two images together using `PIL.ImageChops.multiply()`."""
 
@@ -215,7 +245,13 @@ class ImageMultiplyInvocation(BaseInvocation):
 IMAGE_CHANNELS = Literal["A", "R", "G", "B"]
 
 
-@invocation("img_chan", title="Extract Image Channel", tags=["image", "channel"], category="image", version="1.0.0")
+@invocation(
+    "img_chan",
+    title="Extract Image Channel",
+    tags=["image", "channel"],
+    category="image",
+    version="1.0.0",
+)
 class ImageChannelInvocation(BaseInvocation):
     """Gets a channel from an image."""
 
@@ -247,7 +283,13 @@ class ImageChannelInvocation(BaseInvocation):
 IMAGE_MODES = Literal["L", "RGB", "RGBA", "CMYK", "YCbCr", "LAB", "HSV", "I", "F"]
 
 
-@invocation("img_conv", title="Convert Image Mode", tags=["image", "convert"], category="image", version="1.0.0")
+@invocation(
+    "img_conv",
+    title="Convert Image Mode",
+    tags=["image", "convert"],
+    category="image",
+    version="1.0.0",
+)
 class ImageConvertInvocation(BaseInvocation):
     """Converts an image to a different mode."""
 
@@ -276,7 +318,13 @@ class ImageConvertInvocation(BaseInvocation):
         )
 
 
-@invocation("img_blur", title="Blur Image", tags=["image", "blur"], category="image", version="1.0.0")
+@invocation(
+    "img_blur",
+    title="Blur Image",
+    tags=["image", "blur"],
+    category="image",
+    version="1.0.0",
+)
 class ImageBlurInvocation(BaseInvocation):
     """Blurs an image"""
 
@@ -330,7 +378,13 @@ PIL_RESAMPLING_MAP = {
 }
 
 
-@invocation("img_resize", title="Resize Image", tags=["image", "resize"], category="image", version="1.0.0")
+@invocation(
+    "img_resize",
+    title="Resize Image",
+    tags=["image", "resize"],
+    category="image",
+    version="1.0.0",
+)
 class ImageResizeInvocation(BaseInvocation):
     """Resizes an image to specific dimensions"""
 
@@ -359,7 +413,7 @@ class ImageResizeInvocation(BaseInvocation):
             node_id=self.id,
             session_id=context.graph_execution_state_id,
             is_intermediate=self.is_intermediate,
-            metadata=self.metadata.dict() if self.metadata else None,
+            metadata=self.metadata.model_dump() if self.metadata else None,
             workflow=self.workflow,
         )
 
@@ -370,7 +424,13 @@ class ImageResizeInvocation(BaseInvocation):
         )
 
 
-@invocation("img_scale", title="Scale Image", tags=["image", "scale"], category="image", version="1.0.0")
+@invocation(
+    "img_scale",
+    title="Scale Image",
+    tags=["image", "scale"],
+    category="image",
+    version="1.0.0",
+)
 class ImageScaleInvocation(BaseInvocation):
     """Scales an image by a factor"""
 
@@ -411,7 +471,13 @@ class ImageScaleInvocation(BaseInvocation):
         )
 
 
-@invocation("img_lerp", title="Lerp Image", tags=["image", "lerp"], category="image", version="1.0.0")
+@invocation(
+    "img_lerp",
+    title="Lerp Image",
+    tags=["image", "lerp"],
+    category="image",
+    version="1.0.0",
+)
 class ImageLerpInvocation(BaseInvocation):
     """Linear interpolation of all pixels of an image"""
 
@@ -444,7 +510,13 @@ class ImageLerpInvocation(BaseInvocation):
         )
 
 
-@invocation("img_ilerp", title="Inverse Lerp Image", tags=["image", "ilerp"], category="image", version="1.0.0")
+@invocation(
+    "img_ilerp",
+    title="Inverse Lerp Image",
+    tags=["image", "ilerp"],
+    category="image",
+    version="1.0.0",
+)
 class ImageInverseLerpInvocation(BaseInvocation):
     """Inverse linear interpolation of all pixels of an image"""
 
@@ -456,7 +528,7 @@ class ImageInverseLerpInvocation(BaseInvocation):
         image = context.services.images.get_pil_image(self.image.image_name)
 
         image_arr = numpy.asarray(image, dtype=numpy.float32)
-        image_arr = numpy.minimum(numpy.maximum(image_arr - self.min, 0) / float(self.max - self.min), 1) * 255
+        image_arr = numpy.minimum(numpy.maximum(image_arr - self.min, 0) / float(self.max - self.min), 1) * 255  # type: ignore [assignment]
 
         ilerp_image = Image.fromarray(numpy.uint8(image_arr))
 
@@ -477,7 +549,13 @@ class ImageInverseLerpInvocation(BaseInvocation):
         )
 
 
-@invocation("img_nsfw", title="Blur NSFW Image", tags=["image", "nsfw"], category="image", version="1.0.0")
+@invocation(
+    "img_nsfw",
+    title="Blur NSFW Image",
+    tags=["image", "nsfw"],
+    category="image",
+    version="1.0.0",
+)
 class ImageNSFWBlurInvocation(BaseInvocation):
     """Add blur to NSFW-flagged images"""
 
@@ -505,7 +583,7 @@ class ImageNSFWBlurInvocation(BaseInvocation):
             node_id=self.id,
             session_id=context.graph_execution_state_id,
             is_intermediate=self.is_intermediate,
-            metadata=self.metadata.dict() if self.metadata else None,
+            metadata=self.metadata.model_dump() if self.metadata else None,
             workflow=self.workflow,
         )
 
@@ -515,7 +593,7 @@ class ImageNSFWBlurInvocation(BaseInvocation):
             height=image_dto.height,
         )
 
-    def _get_caution_img(self) -> Image:
+    def _get_caution_img(self) -> Image.Image:
         import invokeai.app.assets.images as image_assets
 
         caution = Image.open(Path(image_assets.__path__[0]) / "caution.png")
@@ -523,7 +601,11 @@ class ImageNSFWBlurInvocation(BaseInvocation):
 
 
 @invocation(
-    "img_watermark", title="Add Invisible Watermark", tags=["image", "watermark"], category="image", version="1.0.0"
+    "img_watermark",
+    title="Add Invisible Watermark",
+    tags=["image", "watermark"],
+    category="image",
+    version="1.0.0",
 )
 class ImageWatermarkInvocation(BaseInvocation):
     """Add an invisible watermark to an image"""
@@ -544,7 +626,7 @@ class ImageWatermarkInvocation(BaseInvocation):
             node_id=self.id,
             session_id=context.graph_execution_state_id,
             is_intermediate=self.is_intermediate,
-            metadata=self.metadata.dict() if self.metadata else None,
+            metadata=self.metadata.model_dump() if self.metadata else None,
             workflow=self.workflow,
         )
 
@@ -555,7 +637,13 @@ class ImageWatermarkInvocation(BaseInvocation):
         )
 
 
-@invocation("mask_edge", title="Mask Edge", tags=["image", "mask", "inpaint"], category="image", version="1.0.0")
+@invocation(
+    "mask_edge",
+    title="Mask Edge",
+    tags=["image", "mask", "inpaint"],
+    category="image",
+    version="1.0.0",
+)
 class MaskEdgeInvocation(BaseInvocation):
     """Applies an edge mask to an image"""
 
@@ -601,7 +689,11 @@ class MaskEdgeInvocation(BaseInvocation):
 
 
 @invocation(
-    "mask_combine", title="Combine Masks", tags=["image", "mask", "multiply"], category="image", version="1.0.0"
+    "mask_combine",
+    title="Combine Masks",
+    tags=["image", "mask", "multiply"],
+    category="image",
+    version="1.0.0",
 )
 class MaskCombineInvocation(BaseInvocation):
     """Combine two masks together by multiplying them using `PIL.ImageChops.multiply()`."""
@@ -632,7 +724,13 @@ class MaskCombineInvocation(BaseInvocation):
         )
 
 
-@invocation("color_correct", title="Color Correct", tags=["image", "color"], category="image", version="1.0.0")
+@invocation(
+    "color_correct",
+    title="Color Correct",
+    tags=["image", "color"],
+    category="image",
+    version="1.0.0",
+)
 class ColorCorrectInvocation(BaseInvocation):
     """
     Shifts the colors of a target image to match the reference image, optionally
@@ -742,7 +840,13 @@ class ColorCorrectInvocation(BaseInvocation):
         )
 
 
-@invocation("img_hue_adjust", title="Adjust Image Hue", tags=["image", "hue"], category="image", version="1.0.0")
+@invocation(
+    "img_hue_adjust",
+    title="Adjust Image Hue",
+    tags=["image", "hue"],
+    category="image",
+    version="1.0.0",
+)
 class ImageHueAdjustmentInvocation(BaseInvocation):
     """Adjusts the Hue of an image."""
 
@@ -980,7 +1084,7 @@ class SaveImageInvocation(BaseInvocation):
 
     image: ImageField = InputField(description=FieldDescriptions.image)
     board: Optional[BoardField] = InputField(default=None, description=FieldDescriptions.board, input=Input.Direct)
-    metadata: CoreMetadata = InputField(
+    metadata: Optional[CoreMetadata] = InputField(
         default=None,
         description=FieldDescriptions.core_metadata,
         ui_hidden=True,
@@ -997,7 +1101,7 @@ class SaveImageInvocation(BaseInvocation):
             node_id=self.id,
             session_id=context.graph_execution_state_id,
             is_intermediate=self.is_intermediate,
-            metadata=self.metadata.dict() if self.metadata else None,
+            metadata=self.metadata.model_dump() if self.metadata else None,
             workflow=self.workflow,
         )