cleanup: Remove manual offload from Depth Anything Processor (#5812)

## What type of PR is this? (check all applicable)

- [ ] Refactor
- [ ] Feature
- [ ] Bug Fix
- [ ] Optimization
- [ ] Documentation Update
- [ ] Community Node Submission


## Have you discussed this change with the InvokeAI team?
- [ ] Yes
- [ ] No, because:

      
## Have you updated all relevant documentation?
- [ ] Yes
- [ ] No


## Description


## Related Tickets & Documents

<!--
For pull requests that relate or close an issue, please include them
below. 

For example having the text: "closes #1234" would connect the current
pull
request to issue 1234.  And when we merge the pull request, Github will
automatically close the issue.
-->

- Related Issue #
- Closes #

## QA Instructions, Screenshots, Recordings

<!-- 
Please provide steps on how to test changes, any hardware or 
software specifications as well as any other pertinent information. 
-->

## Merge Plan

<!--
A merge plan describes how this PR should be handled after it is
approved.

Example merge plans:
- "This PR can be merged when approved"
- "This must be squash-merged when approved"
- "DO NOT MERGE - I will rebase and tidy commits before merging"
- "#dev-chat on discord needs to be advised of this change when it is
merged"

A merge plan is particularly important for large PRs or PRs that touch
the
database in any way.
-->

## Added/updated tests?

- [ ] Yes
- [ ] No : _please replace this line with details on why tests
      have not been included_

## [optional] Are there any post deployment tasks we need to perform?

docs/nodes/INVOCATION_API.md

@@ -0,0 +1,45 @@
+# Invocation API
+
+Each invocation's `invoke` method is provided a single arg - the Invocation
+Context.
+
+This object provides access to various methods, used to interact with the
+application. Loading and saving images, logging messages, etc.
+
+!!! warning ""
+
+    This API may shift slightly until the release of v4.0.0 as we work through a few final updates to the Model Manager.
+
+```py
+class MyInvocation(BaseInvocation):
+  ...
+  def invoke(self, context: InvocationContext) -> ImageOutput:
+      image_pil = context.images.get_pil(image_name)
+      # Do something to the image
+      image_dto = context.images.save(image_pil)
+      # Log a message
+      context.logger.info(f"Did something cool, image saved!")
+      ...
+```
+
+<!-- prettier-ignore-start -->
+::: invokeai.app.services.shared.invocation_context.InvocationContext
+    options:
+        members: false
+
+::: invokeai.app.services.shared.invocation_context.ImagesInterface
+
+::: invokeai.app.services.shared.invocation_context.TensorsInterface
+
+::: invokeai.app.services.shared.invocation_context.ConditioningInterface
+
+::: invokeai.app.services.shared.invocation_context.ModelsInterface
+
+::: invokeai.app.services.shared.invocation_context.LoggerInterface
+
+::: invokeai.app.services.shared.invocation_context.ConfigInterface
+
+::: invokeai.app.services.shared.invocation_context.UtilInterface
+
+::: invokeai.app.services.shared.invocation_context.BoardsInterface
+<!-- prettier-ignore-end -->

docs/nodes/NODES_MIGRATION_V3_V4.md

@@ -0,0 +1,148 @@
+# Invoke v4.0.0 Nodes API Migration guide
+
+Invoke v4.0.0 is versioned as such due to breaking changes to the API utilized
+by nodes, both core and custom.
+
+## Motivation
+
+Prior to v4.0.0, the `invokeai` python package has not be set up to be utilized
+as a library. That is to say, it didn't have any explicitly public API, and node
+authors had to work with the unstable internal application API.
+
+v4.0.0 introduces a stable public API for nodes.
+
+## Changes
+
+There are two node-author-facing changes:
+
+1. Import Paths
+1. Invocation Context API
+
+### Import Paths
+
+All public objects are now exported from `invokeai.invocation_api`:
+
+```py
+# Old
+from invokeai.app.invocations.baseinvocation import (
+    BaseInvocation,
+    InputField,
+    InvocationContext,
+    invocation,
+)
+from invokeai.app.invocations.primitives import ImageField
+
+# New
+from invokeai.invocation_api import (
+    BaseInvocation,
+    ImageField,
+    InputField,
+    InvocationContext,
+    invocation,
+)
+```
+
+It's possible that we've missed some classes you need in your node. Please let
+us know if that's the case.
+
+### Invocation Context API
+
+Most nodes utilize the Invocation Context, an object that is passed to the
+`invoke` that provides access to data and services a node may need.
+
+Until now, that object and the services it exposed were internal. Exposing them
+to nodes means that changes to our internal implementation could break nodes.
+The methods on the services are also often fairly complicated and allowed nodes
+to footgun.
+
+In v4.0.0, this object has been refactored to be much simpler.
+
+See [INVOCATION_API](./INVOCATION_API.md) for full details of the API.
+
+!!! warning ""
+
+    This API may shift slightly until the release of v4.0.0 as we work through a few final updates to the Model Manager.
+
+#### Improved Service Methods
+
+The biggest offender was the image save method:
+
+```py
+# Old
+image_dto = context.services.images.create(
+    image=image,
+    image_origin=ResourceOrigin.INTERNAL,
+    image_category=ImageCategory.GENERAL,
+    node_id=self.id,
+    session_id=context.graph_execution_state_id,
+    is_intermediate=self.is_intermediate,
+    metadata=self.metadata,
+    workflow=context.workflow,
+)
+
+# New
+image_dto = context.images.save(image=image)
+```
+
+Other methods are simplified, or enhanced with additional functionality:
+
+```py
+# Old
+image = context.services.images.get_pil_image(image_name)
+
+# New
+image = context.images.get_pil(image_name)
+image_cmyk = context.images.get_pil(image_name, "CMYK")
+```
+
+We also had some typing issues around tensors:
+
+```py
+# Old
+# `latents` typed as `torch.Tensor`, but could be `ConditioningFieldData`
+latents = context.services.latents.get(self.latents.latents_name)
+# `data` typed as `torch.Tenssor,` but could be `ConditioningFieldData`
+context.services.latents.save(latents_name, data)
+
+# New - separate methods for tensors and conditioning data w/ correct typing
+# Also, the service generates the names
+tensor_name = context.tensors.save(tensor)
+tensor = context.tensors.load(tensor_name)
+# For conditioning
+cond_name = context.conditioning.save(cond_data)
+cond_data = context.conditioning.load(cond_name)
+```
+
+#### Output Construction
+
+Core Outputs have builder functions right on them - no need to manually
+construct these objects, or use an extra utility:
+
+```py
+# Old
+image_output = ImageOutput(
+    image=ImageField(image_name=image_dto.image_name),
+    width=image_dto.width,
+    height=image_dto.height,
+)
+latents_output = build_latents_output(latents_name=name, latents=latents, seed=None)
+noise_output = NoiseOutput(
+    noise=LatentsField(latents_name=latents_name, seed=seed),
+    width=latents.size()[3] * 8,
+    height=latents.size()[2] * 8,
+)
+cond_output = ConditioningOutput(
+    conditioning=ConditioningField(
+        conditioning_name=conditioning_name,
+    ),
+)
+
+# New
+image_output = ImageOutput.build(image_dto)
+latents_output = LatentsOutput.build(latents_name=name, latents=noise, seed=self.seed)
+noise_output = NoiseOutput.build(latents_name=name, latents=noise, seed=self.seed)
+cond_output = ConditioningOutput.build(conditioning_name)
+```
+
+You can still create the objects using constructors if you want, but we suggest
+using the builder methods.

docs/requirements-mkdocs.txt

@@ -1,5 +0,0 @@
-mkdocs
-mkdocs-material>=8, <9
-mkdocs-git-revision-date-localized-plugin
-mkdocs-redirects==1.2.0
-

docs/stylesheets/extra.css

@@ -1,5 +0,0 @@
-:root {
-    --md-primary-fg-color:        #35A4DB;
-    --md-primary-fg-color--light: #35A4DB;
-    --md-primary-fg-color--dark:  #35A4DB;
-  }

invokeai/app/api_app.py

@@ -2,6 +2,7 @@
 # which are imported/used before parse_args() is called will get the default config values instead of the
 # values from the command line or config file.
 import sys
+from contextlib import asynccontextmanager
 
 from invokeai.app.api.no_cache_staticfiles import NoCacheStaticFiles
 from invokeai.version.invokeai_version import __version__
@@ -71,9 +72,25 @@ logger = InvokeAILogger.get_logger(config=app_config)
 mimetypes.add_type("application/javascript", ".js")
 mimetypes.add_type("text/css", ".css")
 
+
+@asynccontextmanager
+async def lifespan(app: FastAPI):
+    # Add startup event to load dependencies
+    ApiDependencies.initialize(config=app_config, event_handler_id=event_handler_id, logger=logger)
+    yield
+    # Shut down threads
+    ApiDependencies.shutdown()
+
+
 # Create the app
 # TODO: create this all in a method so configuration/etc. can be passed in?
-app = FastAPI(title="Invoke - Community Edition", docs_url=None, redoc_url=None, separate_input_output_schemas=False)
+app = FastAPI(
+    title="Invoke - Community Edition",
+    docs_url=None,
+    redoc_url=None,
+    separate_input_output_schemas=False,
+    lifespan=lifespan,
+)
 
 # Add event handler
 event_handler_id: int = id(app)
@@ -96,18 +113,6 @@ app.add_middleware(
 app.add_middleware(GZipMiddleware, minimum_size=1000)
 
 
-# Add startup event to load dependencies
-@app.on_event("startup")
-async def startup_event() -> None:
-    ApiDependencies.initialize(config=app_config, event_handler_id=event_handler_id, logger=logger)
-
-
-# Shut down threads
-@app.on_event("shutdown")
-async def shutdown_event() -> None:
-    ApiDependencies.shutdown()
-
-
 # Include all routers
 app.include_router(utilities.utilities_router, prefix="/api")
 app.include_router(model_manager.model_manager_router, prefix="/api")

invokeai/app/invocations/compel.py

@@ -1,17 +1,11 @@
-from typing import Iterator, List, Optional, Tuple, Union
+from typing import Iterator, List, Optional, Tuple, Union, cast
 
 import torch
 from compel import Compel, ReturnedEmbeddingsType
 from compel.prompt_parser import Blend, Conjunction, CrossAttentionControlSubstitute, FlattenedPrompt, Fragment
-from transformers import CLIPTextModel, CLIPTokenizer
+from transformers import CLIPTextModel, CLIPTextModelWithProjection, CLIPTokenizer
 
-from invokeai.app.invocations.fields import (
-    FieldDescriptions,
-    Input,
-    InputField,
-    OutputField,
-    UIComponent,
-)
+from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField, UIComponent
 from invokeai.app.invocations.primitives import ConditioningOutput
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.util.ti_utils import generate_ti_list
@@ -25,12 +19,7 @@ from invokeai.backend.stable_diffusion.diffusion.conditioning_data import (
 )
 from invokeai.backend.util.devices import torch_dtype
 
-from .baseinvocation import (
-    BaseInvocation,
-    BaseInvocationOutput,
-    invocation,
-    invocation_output,
-)
+from .baseinvocation import BaseInvocation, BaseInvocationOutput, invocation, invocation_output
 from .model import ClipField
 
 # unconditioned: Optional[torch.Tensor]
@@ -149,7 +138,7 @@ class SDXLPromptInvocationBase:
         assert isinstance(tokenizer_model, CLIPTokenizer)
         text_encoder_info = context.models.load(**clip_field.text_encoder.model_dump())
         text_encoder_model = text_encoder_info.model
-        assert isinstance(text_encoder_model, CLIPTextModel)
+        assert isinstance(text_encoder_model, (CLIPTextModel, CLIPTextModelWithProjection))
 
         # return zero on empty
         if prompt == "" and zero_on_empty:
@@ -196,7 +185,8 @@ class SDXLPromptInvocationBase:
             # Apply CLIP Skip after LoRA to prevent LoRA application from failing on skipped layers.
             ModelPatcher.apply_clip_skip(text_encoder_model, clip_field.skipped_layers),
         ):
-            assert isinstance(text_encoder, CLIPTextModel)
+            assert isinstance(text_encoder, (CLIPTextModel, CLIPTextModelWithProjection))
+            text_encoder = cast(CLIPTextModel, text_encoder)
             compel = Compel(
                 tokenizer=tokenizer,
                 text_encoder=text_encoder,

invokeai/app/invocations/controlnet_image_processors.py

@@ -576,7 +576,7 @@ DEPTH_ANYTHING_MODEL_SIZES = Literal["large", "base", "small"]
     title="Depth Anything Processor",
     tags=["controlnet", "depth", "depth anything"],
     category="controlnet",
-    version="1.0.0",
+    version="1.0.1",
 )
 class DepthAnythingImageProcessorInvocation(ImageProcessorInvocation):
     """Generates a depth map based on the Depth Anything algorithm"""
@@ -585,13 +585,12 @@ class DepthAnythingImageProcessorInvocation(ImageProcessorInvocation):
         default="small", description="The size of the depth model to use"
     )
     resolution: int = InputField(default=512, ge=64, multiple_of=64, description=FieldDescriptions.image_res)
-    offload: bool = InputField(default=False)
 
     def run_processor(self, image: Image.Image):
         depth_anything_detector = DepthAnythingDetector()
         depth_anything_detector.load_model(model_size=self.model_size)
 
-        processed_image = depth_anything_detector(image=image, resolution=self.resolution, offload=self.offload)
+        processed_image = depth_anything_detector(image=image, resolution=self.resolution)
         return processed_image
 
 

invokeai/app/invocations/ip_adapter.py

@@ -93,7 +93,7 @@ class IPAdapterInvocation(BaseInvocation):
         image_encoder_model_id = ip_adapter_info.image_encoder_model_id
         image_encoder_model_name = image_encoder_model_id.split("/")[-1].strip()
         image_encoder_models = context.models.search_by_attrs(
-            model_name=image_encoder_model_name, base_model=BaseModelType.Any, model_type=ModelType.CLIPVision
+            name=image_encoder_model_name, base=BaseModelType.Any, type=ModelType.CLIPVision
         )
         assert len(image_encoder_models) == 1
         image_encoder_model = CLIPVisionModelField(key=image_encoder_models[0].key)

invokeai/app/services/download/download_default.py

@@ -228,16 +228,10 @@ class DownloadQueueService(DownloadQueueServiceBase):
             except (OSError, HTTPError) as excp:
                 job.error_type = excp.__class__.__name__ + f"({str(excp)})"
                 job.error = traceback.format_exc()
-                try:
                 self._signal_job_error(job, excp)
-                except:
-                    pass
             except DownloadJobCancelledException:
-                try:
                 self._signal_job_cancelled(job)
                 self._cleanup_cancelled_job(job)
-                except:
-                    pass
 
             finally:
                 job.job_ended = get_iso_timestamp()

invokeai/app/services/shared/invocation_context.py

@@ -65,75 +65,86 @@ class InvocationContextInterface:
 
 class BoardsInterface(InvocationContextInterface):
     def create(self, board_name: str) -> BoardDTO:
-        """
-        Creates a board.
+        """Creates a board.
 
-        :param board_name: The name of the board to create.
+        Args:
+            board_name: The name of the board to create.
+
+        Returns:
+            The created board DTO.
         """
         return self._services.boards.create(board_name)
 
     def get_dto(self, board_id: str) -> BoardDTO:
-        """
-        Gets a board DTO.
+        """Gets a board DTO.
 
-        :param board_id: The ID of the board to get.
+        Args:
+            board_id: The ID of the board to get.
+
+        Returns:
+            The board DTO.
         """
         return self._services.boards.get_dto(board_id)
 
     def get_all(self) -> list[BoardDTO]:
-        """
-        Gets all boards.
+        """Gets all boards.
+
+        Returns:
+            A list of all boards.
         """
         return self._services.boards.get_all()
 
     def add_image_to_board(self, board_id: str, image_name: str) -> None:
-        """
-        Adds an image to a board.
+        """Adds an image to a board.
 
-        :param board_id: The ID of the board to add the image to.
-        :param image_name: The name of the image to add to the board.
+        Args:
+            board_id: The ID of the board to add the image to.
+            image_name: The name of the image to add to the board.
         """
         return self._services.board_images.add_image_to_board(board_id, image_name)
 
     def get_all_image_names_for_board(self, board_id: str) -> list[str]:
-        """
-        Gets all image names for a board.
+        """Gets all image names for a board.
 
-        :param board_id: The ID of the board to get the image names for.
+        Args:
+            board_id: The ID of the board to get the image names for.
+
+        Returns:
+            A list of all image names for the board.
         """
         return self._services.board_images.get_all_board_image_names_for_board(board_id)
 
 
 class LoggerInterface(InvocationContextInterface):
     def debug(self, message: str) -> None:
-        """
-        Logs a debug message.
+        """Logs a debug message.
 
-        :param message: The message to log.
+        Args:
+            message: The message to log.
         """
         self._services.logger.debug(message)
 
     def info(self, message: str) -> None:
-        """
-        Logs an info message.
+        """Logs an info message.
 
-        :param message: The message to log.
+        Args:
+            message: The message to log.
         """
         self._services.logger.info(message)
 
     def warning(self, message: str) -> None:
-        """
-        Logs a warning message.
+        """Logs a warning message.
 
-        :param message: The message to log.
+        Args:
+            message: The message to log.
         """
         self._services.logger.warning(message)
 
     def error(self, message: str) -> None:
-        """
-        Logs an error message.
+        """Logs an error message.
 
-        :param message: The message to log.
+        Args:
+            message: The message to log.
         """
         self._services.logger.error(message)
 
@@ -146,20 +157,23 @@ class ImagesInterface(InvocationContextInterface):
         image_category: ImageCategory = ImageCategory.GENERAL,
         metadata: Optional[MetadataField] = None,
     ) -> ImageDTO:
-        """
-        Saves an image, returning its DTO.
+        """Saves an image, returning its DTO.
 
         If the current queue item has a workflow or metadata, it is automatically saved with the image.
 
-        :param image: The image to save, as a PIL image.
-        :param board_id: The board ID to add the image to, if it should be added. It the invocation \
+        Args:
+            image: The image to save, as a PIL image.
+            board_id: The board ID to add the image to, if it should be added. It the invocation \
             inherits from `WithBoard`, that board will be used automatically. **Use this only if \
             you want to override or provide a board manually!**
-        :param image_category: The category of the image. Only the GENERAL category is added \
+            image_category: The category of the image. Only the GENERAL category is added \
             to the gallery.
-        :param metadata: The metadata to save with the image, if it should have any. If the \
+            metadata: The metadata to save with the image, if it should have any. If the \
             invocation inherits from `WithMetadata`, that metadata will be used automatically. \
             **Use this only if you want to override or provide metadata manually!**
+
+        Returns:
+            The saved image DTO.
         """
 
         # If `metadata` is provided directly, use that. Else, use the metadata provided by `WithMetadata`, falling back to None.
@@ -189,11 +203,14 @@ class ImagesInterface(InvocationContextInterface):
         )
 
     def get_pil(self, image_name: str, mode: IMAGE_MODES | None = None) -> Image:
-        """
-        Gets an image as a PIL Image object.
+        """Gets an image as a PIL Image object.
 
-        :param image_name: The name of the image to get.
-        :param mode: The color mode to convert the image to. If None, the original mode is used.
+        Args:
+            image_name: The name of the image to get.
+            mode: The color mode to convert the image to. If None, the original mode is used.
+
+        Returns:
+            The image as a PIL Image object.
         """
         image = self._services.images.get_pil_image(image_name)
         if mode and mode != image.mode:
@@ -206,58 +223,76 @@ class ImagesInterface(InvocationContextInterface):
         return image
 
     def get_metadata(self, image_name: str) -> Optional[MetadataField]:
-        """
-        Gets an image's metadata, if it has any.
+        """Gets an image's metadata, if it has any.
 
-        :param image_name: The name of the image to get the metadata for.
+        Args:
+            image_name: The name of the image to get the metadata for.
+
+        Returns:
+            The image's metadata, if it has any.
         """
         return self._services.images.get_metadata(image_name)
 
     def get_dto(self, image_name: str) -> ImageDTO:
-        """
-        Gets an image as an ImageDTO object.
+        """Gets an image as an ImageDTO object.
 
-        :param image_name: The name of the image to get.
+        Args:
+            image_name: The name of the image to get.
+
+        Returns:
+            The image as an ImageDTO object.
         """
         return self._services.images.get_dto(image_name)
 
 
 class TensorsInterface(InvocationContextInterface):
     def save(self, tensor: Tensor) -> str:
-        """
-        Saves a tensor, returning its name.
+        """Saves a tensor, returning its name.
 
-        :param tensor: The tensor to save.
+        Args:
+            tensor: The tensor to save.
+
+        Returns:
+            The name of the saved tensor.
         """
 
         name = self._services.tensors.save(obj=tensor)
         return name
 
     def load(self, name: str) -> Tensor:
-        """
-        Loads a tensor by name.
+        """Loads a tensor by name.
 
-        :param name: The name of the tensor to load.
+        Args:
+            name: The name of the tensor to load.
+
+        Returns:
+            The loaded tensor.
         """
         return self._services.tensors.load(name)
 
 
 class ConditioningInterface(InvocationContextInterface):
     def save(self, conditioning_data: ConditioningFieldData) -> str:
-        """
-        Saves a conditioning data object, returning its name.
+        """Saves a conditioning data object, returning its name.
 
-        :param conditioning_data: The conditioning data to save.
+        Args:
+            conditioning_data: The conditioning data to save.
+
+        Returns:
+            The name of the saved conditioning data.
         """
 
         name = self._services.conditioning.save(obj=conditioning_data)
         return name
 
     def load(self, name: str) -> ConditioningFieldData:
-        """
-        Loads conditioning data by name.
+        """Loads conditioning data by name.
 
-        :param name: The name of the conditioning data to load.
+        Args:
+            name: The name of the conditioning data to load.
+
+        Returns:
+            The loaded conditioning data.
         """
 
         return self._services.conditioning.load(name)
@@ -265,20 +300,25 @@ class ConditioningInterface(InvocationContextInterface):
 
 class ModelsInterface(InvocationContextInterface):
     def exists(self, key: str) -> bool:
-        """
-        Checks if a model exists.
+        """Checks if a model exists.
 
-        :param key: The key of the model.
+        Args:
+            key: The key of the model.
+
+        Returns:
+            True if the model exists, False if not.
         """
         return self._services.model_manager.store.exists(key)
 
     def load(self, key: str, submodel_type: Optional[SubModelType] = None) -> LoadedModel:
-        """
-        Loads a model.
+        """Loads a model.
 
-        :param key: The key of the model.
-        :param submodel_type: The submodel of the model to get.
-        :returns: An object representing the loaded model.
+        Args:
+            key: The key of the model.
+            submodel_type: The submodel of the model to get.
+
+        Returns:
+            An object representing the loaded model.
         """
 
         # The model manager emits events as it loads the model. It needs the context data to build
@@ -289,75 +329,95 @@ class ModelsInterface(InvocationContextInterface):
         )
 
     def load_by_attrs(
-        self, model_name: str, base_model: BaseModelType, model_type: ModelType, submodel: Optional[SubModelType] = None
+        self, name: str, base: BaseModelType, type: ModelType, submodel_type: Optional[SubModelType] = None
     ) -> LoadedModel:
-        """
-        Loads a model by its attributes.
+        """Loads a model by its attributes.
 
-        :param model_name: Name of to be fetched.
-        :param base_model: Base model
-        :param model_type: Type of the model
-        :param submodel: For main (pipeline models), the submodel to fetch
+        Args:
+            name: Name of the model.
+            base: The models' base type, e.g. `BaseModelType.StableDiffusion1`, `BaseModelType.StableDiffusionXL`, etc.
+            type: Type of the model, e.g. `ModelType.Main`, `ModelType.Vae`, etc.
+            submodel_type: The type of submodel to load, e.g. `SubModelType.UNet`, `SubModelType.TextEncoder`, etc. Only main
+            models have submodels.
+
+        Returns:
+            An object representing the loaded model.
         """
         return self._services.model_manager.load_model_by_attr(
-            model_name=model_name,
-            base_model=base_model,
-            model_type=model_type,
-            submodel=submodel,
+            model_name=name,
+            base_model=base,
+            model_type=type,
+            submodel=submodel_type,
             context_data=self._data,
         )
 
     def get_config(self, key: str) -> AnyModelConfig:
-        """
-        Gets a model's info, an dict-like object.
+        """Gets a model's config.
 
-        :param key: The key of the model.
+        Args:
+            key: The key of the model.
+
+        Returns:
+            The model's config.
         """
         return self._services.model_manager.store.get_model(key=key)
 
     def get_metadata(self, key: str) -> Optional[AnyModelRepoMetadata]:
-        """
-        Gets a model's metadata, if it has any.
+        """Gets a model's metadata, if it has any.
 
-        :param key: The key of the model.
+        Args:
+            key: The key of the model.
+
+        Returns:
+            The model's metadata, if it has any.
         """
         return self._services.model_manager.store.get_metadata(key=key)
 
     def search_by_path(self, path: Path) -> list[AnyModelConfig]:
-        """
-        Searches for models by path.
+        """Searches for models by path.
 
-        :param path: The path to search for.
+        Args:
+            path: The path to search for.
+
+        Returns:
+            A list of models that match the path.
         """
         return self._services.model_manager.store.search_by_path(path)
 
     def search_by_attrs(
         self,
-        model_name: Optional[str] = None,
-        base_model: Optional[BaseModelType] = None,
-        model_type: Optional[ModelType] = None,
-        model_format: Optional[ModelFormat] = None,
+        name: Optional[str] = None,
+        base: Optional[BaseModelType] = None,
+        type: Optional[ModelType] = None,
+        format: Optional[ModelFormat] = None,
     ) -> list[AnyModelConfig]:
-        """
-        Searches for models by attributes.
+        """Searches for models by attributes.
 
-        :param model_name: Name of to be fetched.
-        :param base_model: Base model
-        :param model_type: Type of the model
-        :param submodel: For main (pipeline models), the submodel to fetch
+        Args:
+            name: The name to search for (exact match).
+            base: The base to search for, e.g. `BaseModelType.StableDiffusion1`, `BaseModelType.StableDiffusionXL`, etc.
+            type: Type type of model to search for, e.g. `ModelType.Main`, `ModelType.Vae`, etc.
+            format: The format of model to search for, e.g. `ModelFormat.Checkpoint`, `ModelFormat.Diffusers`, etc.
+
+        Returns:
+            A list of models that match the attributes.
         """
 
         return self._services.model_manager.store.search_by_attr(
-            model_name=model_name,
-            base_model=base_model,
-            model_type=model_type,
-            model_format=model_format,
+            model_name=name,
+            base_model=base,
+            model_type=type,
+            model_format=format,
         )
 
 
 class ConfigInterface(InvocationContextInterface):
     def get(self) -> InvokeAIAppConfig:
-        """Gets the app's config."""
+        """Gets the app's config.
+
+        Returns:
+            The app's config.
+        """
 
         return self._services.configuration.get_config()
 
@@ -370,7 +430,11 @@ class UtilInterface(InvocationContextInterface):
         self._cancel_event = cancel_event
 
     def is_canceled(self) -> bool:
-        """Checks if the current invocation has been canceled."""
+        """Checks if the current session has been canceled.
+
+        Returns:
+            True if the current session has been canceled, False if not.
+        """
         return self._cancel_event.is_set()
 
     def sd_step_callback(self, intermediate_state: PipelineIntermediateState, base_model: BaseModelType) -> None:
@@ -380,8 +444,9 @@ class UtilInterface(InvocationContextInterface):
 
         This should be called after each denoising step.
 
-        :param intermediate_state: The intermediate state of the diffusion pipeline.
-        :param base_model: The base model for the current denoising step.
+        Args:
+            intermediate_state: The intermediate state of the diffusion pipeline.
+            base_model: The base model for the current denoising step.
         """
 
         stable_diffusion_step_callback(
@@ -394,8 +459,17 @@ class UtilInterface(InvocationContextInterface):
 
 
 class InvocationContext:
-    """
-    The `InvocationContext` provides access to various services and data for the current invocation.
+    """Provides access to various services and data for the current invocation.
+
+    Attributes:
+        images (ImagesInterface): Methods to save, get and update images and their metadata.
+        tensors (TensorsInterface): Methods to save and get tensors, including image, noise, masks, and masked images.
+        conditioning (ConditioningInterface): Methods to save and get conditioning data.
+        models (ModelsInterface): Methods to check if a model exists, get a model, and get a model's info.
+        logger (LoggerInterface): The app logger.
+        config (ConfigInterface): The app config.
+        util (UtilInterface): Utility methods, including a method to check if an invocation was canceled and step callbacks.
+        boards (BoardsInterface): Methods to interact with boards.
     """
 
     def __init__(
@@ -438,11 +512,14 @@ def build_invocation_context(
     data: InvocationContextData,
     cancel_event: threading.Event,
 ) -> InvocationContext:
-    """
-    Builds the invocation context for a specific invocation execution.
+    """Builds the invocation context for a specific invocation execution.
 
-    :param services: The invocation services to wrap.
-    :param data: The invocation context data.
+    Args:
+        services: The invocation services to wrap.
+        data: The invocation context data.
+
+    Returns:
+        The invocation context.
     """
 
     logger = LoggerInterface(services=services, data=data)

invokeai/app/util/ti_utils.py

@@ -30,7 +30,7 @@ def generate_ti_list(
         except UnknownModelException:
             try:
                 loaded_model = context.models.load_by_attrs(
-                    model_name=name_or_key, base_model=base, model_type=ModelType.TextualInversion
+                    name=name_or_key, base=base, type=ModelType.TextualInversion
                 )
                 model = loaded_model.model
                 assert isinstance(model, TextualInversionModelRaw)

invokeai/backend/image_util/depth_anything/__init__.py

@@ -17,6 +17,8 @@ from invokeai.backend.util.util import download_with_progress_bar
 
 config = InvokeAIAppConfig.get_config()
 
+DEPTH_ANYTHING_MODEL_SIZES = Literal["large", "base", "small"]
+
 DEPTH_ANYTHING_MODELS = {
     "large": {
         "url": "https://huggingface.co/spaces/LiheYoung/Depth-Anything/resolve/main/checkpoints/depth_anything_vitl14.pth?download=true",
@@ -53,9 +55,9 @@ transform = Compose(
 class DepthAnythingDetector:
     def __init__(self) -> None:
         self.model = None
-        self.model_size: Union[Literal["large", "base", "small"], None] = None
+        self.model_size: Union[DEPTH_ANYTHING_MODEL_SIZES, None] = None
 
-    def load_model(self, model_size=Literal["large", "base", "small"]):
+    def load_model(self, model_size: DEPTH_ANYTHING_MODEL_SIZES = "small"):
         DEPTH_ANYTHING_MODEL_PATH = pathlib.Path(config.models_path / DEPTH_ANYTHING_MODELS[model_size]["local"])
         if not DEPTH_ANYTHING_MODEL_PATH.exists():
             download_with_progress_bar(DEPTH_ANYTHING_MODELS[model_size]["url"], DEPTH_ANYTHING_MODEL_PATH)
@@ -84,16 +86,19 @@ class DepthAnythingDetector:
         self.model.to(device)
         return self
 
-    def __call__(self, image, resolution=512, offload=False):
-        image = np.array(image, dtype=np.uint8)
-        image = image[:, :, ::-1] / 255.0
+    def __call__(self, image: Image.Image, resolution: int = 512):
+        if self.model is None:
+            raise Exception("Depth Anything Model not loaded")
 
-        image_height, image_width = image.shape[:2]
-        image = transform({"image": image})["image"]
-        image = torch.from_numpy(image).unsqueeze(0).to(choose_torch_device())
+        np_image = np.array(image, dtype=np.uint8)
+        np_image = np_image[:, :, ::-1] / 255.0
+
+        image_height, image_width = np_image.shape[:2]
+        np_image = transform({"image": image})["image"]
+        tensor_image = torch.from_numpy(np_image).unsqueeze(0).to(choose_torch_device())
 
         with torch.no_grad():
-            depth = self.model(image)
+            depth = self.model(tensor_image)
             depth = F.interpolate(depth[None], (image_height, image_width), mode="bilinear", align_corners=False)[0, 0]
             depth = (depth - depth.min()) / (depth.max() - depth.min()) * 255.0
 
@@ -103,7 +108,4 @@ class DepthAnythingDetector:
         new_height = int(image_height * (resolution / image_width))
         depth_map = depth_map.resize((resolution, new_height))
 
-        if offload:
-            del self.model
-
         return depth_map

invokeai/backend/model_manager/search.py

@@ -28,6 +28,7 @@ from typing import Callable, Optional, Set, Union
 
 from pydantic import BaseModel, Field
 
+from invokeai.app.services.config import InvokeAIAppConfig
 from invokeai.backend.util.logging import InvokeAILogger
 
 default_logger: Logger = InvokeAILogger.get_logger()
@@ -117,13 +118,10 @@ class ModelSearch(ModelSearchBase):
     """
 
     models_found: Set[Path] = Field(default_factory=set)
-    scanned_dirs: Set[Path] = Field(default_factory=set)
-    pruned_paths: Set[Path] = Field(default_factory=set)
+    config: InvokeAIAppConfig = InvokeAIAppConfig.get_config()
 
     def search_started(self) -> None:
         self.models_found = set()
-        self.scanned_dirs = set()
-        self.pruned_paths = set()
         if self.on_search_started:
             self.on_search_started(self._directory)
 
@@ -139,29 +137,28 @@ class ModelSearch(ModelSearchBase):
 
     def search(self, directory: Union[Path, str]) -> Set[Path]:
         self._directory = Path(directory)
+        if not self._directory.is_absolute():
+            self._directory = self.config.models_path / self._directory
         self.stats = SearchStats()  # zero out
         self.search_started()  # This will initialize _models_found to empty
-        self._walk_directory(directory)
+        self._walk_directory(self._directory)
         self.search_completed()
         return self.models_found
 
-    def _walk_directory(self, path: Union[Path, str]) -> None:
-        for root, dirs, files in os.walk(path, followlinks=True):
-            # don't descend into directories that start with a "."
-            # to avoid the Mac .DS_STORE issue.
-            if str(Path(root).name).startswith("."):
-                self.pruned_paths.add(Path(root))
-            if any(Path(root).is_relative_to(x) for x in self.pruned_paths):
-                continue
-
-            self.stats.items_scanned += len(dirs) + len(files)
-            for d in dirs:
-                path = Path(root) / d
-                if path.parent in self.scanned_dirs:
-                    self.scanned_dirs.add(path)
-                    continue
+    def _walk_directory(self, path: Union[Path, str], max_depth: int = 20) -> None:
+        absolute_path = Path(path)
+        if (
+            len(absolute_path.parts) - len(self._directory.parts) > max_depth
+            or not absolute_path.exists()
+            or absolute_path.parent in self.models_found
+        ):
+            return
+        entries = os.scandir(absolute_path.as_posix())
+        entries = [entry for entry in entries if not entry.name.startswith(".")]
+        dirs = [entry for entry in entries if entry.is_dir()]
+        file_names = [entry.name for entry in entries if entry.is_file()]
         if any(
-                    (path / x).exists()
+            x in file_names
             for x in [
                 "config.json",
                 "model_index.json",
@@ -170,22 +167,23 @@ class ModelSearch(ModelSearchBase):
                 "image_encoder.txt",
             ]
         ):
-                    self.scanned_dirs.add(path)
             try:
-                        self.model_found(path)
+                self.model_found(absolute_path)
+                return
+            except KeyboardInterrupt:
+                raise
+            except Exception as e:
+                self.logger.warning(str(e))
+                return
+
+        for n in file_names:
+            if n.endswith((".ckpt", ".bin", ".pth", ".safetensors", ".pt")):
+                try:
+                    self.model_found(absolute_path / n)
                 except KeyboardInterrupt:
                     raise
                 except Exception as e:
                     self.logger.warning(str(e))
 
-            for f in files:
-                path = Path(root) / f
-                if path.parent in self.scanned_dirs:
-                    continue
-                if path.suffix in {".ckpt", ".bin", ".pth", ".safetensors", ".pt"}:
-                    try:
-                        self.model_found(path)
-                    except KeyboardInterrupt:
-                        raise
-                    except Exception as e:
-                        self.logger.warning(str(e))
+        for d in dirs:
+            self._walk_directory(absolute_path / d)

invokeai/backend/model_patcher.py

@@ -4,12 +4,12 @@ from __future__ import annotations
 
 import pickle
 from contextlib import contextmanager
-from typing import Any, Dict, Iterator, List, Optional, Tuple
+from typing import Any, Dict, Iterator, List, Optional, Tuple, Union
 
 import numpy as np
 import torch
 from diffusers import OnnxRuntimeModel, UNet2DConditionModel
-from transformers import CLIPTextModel, CLIPTokenizer
+from transformers import CLIPTextModel, CLIPTextModelWithProjection, CLIPTokenizer
 
 from invokeai.app.shared.models import FreeUConfig
 from invokeai.backend.model_manager import AnyModel
@@ -168,7 +168,7 @@ class ModelPatcher:
     def apply_ti(
         cls,
         tokenizer: CLIPTokenizer,
-        text_encoder: CLIPTextModel,
+        text_encoder: Union[CLIPTextModel, CLIPTextModelWithProjection],
         ti_list: List[Tuple[str, TextualInversionModelRaw]],
     ) -> Iterator[Tuple[CLIPTokenizer, TextualInversionManager]]:
         init_tokens_count = None
@@ -265,7 +265,7 @@ class ModelPatcher:
     @contextmanager
     def apply_clip_skip(
         cls,
-        text_encoder: CLIPTextModel,
+        text_encoder: Union[CLIPTextModel, CLIPTextModelWithProjection],
         clip_skip: int,
     ) -> None:
         skipped_layers = []

mkdocs.yml

@@ -25,18 +25,7 @@ theme:
     edit: material/file-document-edit-outline
   favicon: img/favicon.ico
   palette:
-    # - media: '(prefers-color-scheme: light)'
-    #   scheme: default
-    #   primary: custom
-    #   toggle:
-    #     icon: material/lightbulb
-    #     name: Switch to dark mode
-    - media: '(prefers-color-scheme: dark)'
     scheme: slate
-      primary: custom
-      # toggle:
-      #   icon: material/lightbulb-outline
-      #   name: Switch to light mode
   features:
     - navigation.instant
     - navigation.tabs
@@ -48,7 +37,6 @@ theme:
     - search.suggest
     - toc.integrate
 
-
 # Extensions
 markdown_extensions:
   - abbr
@@ -66,8 +54,8 @@ markdown_extensions:
   - pymdownx.caret
   - pymdownx.details
   - pymdownx.emoji:
-      emoji_index: !!python/name:materialx.emoji.twemoji
-      emoji_generator: !!python/name:materialx.emoji.to_svg
+      emoji_index: !!python/name:material.extensions.emoji.twemoji
+      emoji_generator: !!python/name:material.extensions.emoji.to_svg
   - pymdownx.highlight:
       anchor_linenums: true
   - pymdownx.inlinehilite
@@ -103,6 +91,19 @@ plugins:
         'installation/INSTALL_DOCKER.md': 'installation/040_INSTALL_DOCKER.md'
         'installation/INSTALLING_MODELS.md': 'installation/050_INSTALLING_MODELS.md'
         'installation/INSTALL_PATCHMATCH.md': 'installation/060_INSTALL_PATCHMATCH.md'
+  - mkdocstrings:
+      handlers:
+        python:
+          options:
+            separate_signature: true
+            show_signature_annotations: true
+            parameter_headings: false
+            signature_crossrefs: true
+            show_source: false
+            summary: true
+            show_root_heading: true
+            show_root_full_path: false
+            show_bases: false
 
 extra_javascript:
   - https://unpkg.com/tablesort@5.3.0/dist/tablesort.min.js
@@ -142,6 +143,8 @@ nav:
       - ComfyUI to InvokeAI: 'nodes/comfyToInvoke.md'
       - Facetool Node: 'nodes/detailedNodes/faceTools.md'
       - Contributing Nodes: 'nodes/contributingNodes.md'
+      - Migrating from v3 to v4: 'nodes/NODES_MIGRATION_V3_V4.md'
+      - Invocation API: 'nodes/INVOCATION_API.md'
   - Community Nodes: 'nodes/communityNodes.md'
   - Features:
       - Overview: 'features/index.md'
@@ -161,8 +164,8 @@ nav:
           - Textual Inversion Training: 'features/TRAINING.md'
       - Unified Canvas: 'features/UNIFIED_CANVAS.md'
       - InvokeAI Web Server: 'features/WEB.md'
-      - WebUI Hotkeys: "features/WEBUIHOTKEYS.md"
-      - Maintenance Utilities: "features/UTILITIES.md"
+      - WebUI Hotkeys: 'features/WEBUIHOTKEYS.md'
+      - Maintenance Utilities: 'features/UTILITIES.md'
       - Other: 'features/OTHER.md'
   - Contributing:
       - How to Contribute: 'contributing/CONTRIBUTING.md'

pyproject.toml

@@ -105,9 +105,10 @@ dependencies = [
 "onnx-directml" = ["onnxruntime-directml"]
 "dist" = ["pip-tools", "pipdeptree", "twine"]
 "docs" = [
-  "mkdocs-material<9.0",
+  "mkdocs-material",
   "mkdocs-git-revision-date-localized-plugin",
-  "mkdocs-redirects==1.2.0",
+  "mkdocs-redirects",
+  "mkdocstrings[python]",
 ]
 "dev" = ["jurigged", "pudb", "snakeviz", "gprof2dot"]
 "test" = [