Remove optional lists from configs in favor of defaulting to empty lists

Currently translated at 99.5% (1091 of 1096 strings)

Co-authored-by: Васянатор <ilabulanov339@gmail.com>
Translate-URL: https://hosted.weblate.org/projects/invokeai/web-ui/ru/
Translation: InvokeAI/Web UI

.github/workflows/frontend-checks.yml

@@ -1,7 +1,7 @@
 # Runs frontend code quality checks.
 #
 # Checks for changes to frontend files before running the checks.
-# When manually triggered or when called from another workflow, always runs the checks.
+# If always_run is true, always runs the checks.
 
 name: 'frontend checks'
 
@@ -16,7 +16,19 @@ on:
       - 'synchronize'
   merge_group:
   workflow_dispatch:
+    inputs:
+      always_run:
+        description: 'Always run the checks'
+        required: true
+        type: boolean
+        default: true
   workflow_call:
+    inputs:
+      always_run:
+        description: 'Always run the checks'
+        required: true
+        type: boolean
+        default: true
 
 defaults:
   run:
@@ -30,7 +42,7 @@ jobs:
       - uses: actions/checkout@v4
 
       - name: check for changed frontend files
-        if: ${{ github.event_name != 'workflow_dispatch' && github.event_name != 'workflow_call' }}
+        if: ${{ inputs.always_run != true }}
         id: changed-files
         uses: tj-actions/changed-files@v42
         with:
@@ -39,30 +51,30 @@ jobs:
               - 'invokeai/frontend/web/**'
 
       - name: install dependencies
-        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call' }}
+        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || inputs.always_run == true }}
         uses: ./.github/actions/install-frontend-deps
 
       - name: tsc
-        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call' }}
+        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || inputs.always_run == true }}
         run: 'pnpm lint:tsc'
         shell: bash
 
       - name: dpdm
-        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call' }}
+        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || inputs.always_run == true }}
         run: 'pnpm lint:dpdm'
         shell: bash
 
       - name: eslint
-        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call' }}
+        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || inputs.always_run == true }}
         run: 'pnpm lint:eslint'
         shell: bash
 
       - name: prettier
-        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call' }}
+        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || inputs.always_run == true }}
         run: 'pnpm lint:prettier'
         shell: bash
 
       - name: knip
-        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call' }}
+        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || inputs.always_run == true }}
         run: 'pnpm lint:knip'
         shell: bash

.github/workflows/frontend-tests.yml

@@ -1,7 +1,7 @@
 # Runs frontend tests.
 #
 # Checks for changes to frontend files before running the tests.
-# When manually triggered or called from another workflow, always runs the tests.
+# If always_run is true, always runs the tests.
 
 name: 'frontend tests'
 
@@ -16,7 +16,19 @@ on:
       - 'synchronize'
   merge_group:
   workflow_dispatch:
+    inputs:
+      always_run:
+        description: 'Always run the tests'
+        required: true
+        type: boolean
+        default: true
   workflow_call:
+    inputs:
+      always_run:
+        description: 'Always run the tests'
+        required: true
+        type: boolean
+        default: true
 
 defaults:
   run:
@@ -30,7 +42,7 @@ jobs:
       - uses: actions/checkout@v4
 
       - name: check for changed frontend files
-        if: ${{ github.event_name != 'workflow_dispatch' && github.event_name != 'workflow_call' }}
+        if: ${{ inputs.always_run != true }}
         id: changed-files
         uses: tj-actions/changed-files@v42
         with:
@@ -39,10 +51,10 @@ jobs:
               - 'invokeai/frontend/web/**'
 
       - name: install dependencies
-        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call' }}
+        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || inputs.always_run == true }}
         uses: ./.github/actions/install-frontend-deps
 
       - name: vitest
-        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call' }}
+        if: ${{ steps.changed-files.outputs.frontend_any_changed == 'true' || inputs.always_run == true }}
         run: 'pnpm test:no-watch'
         shell: bash

.github/workflows/python-checks.yml

@@ -1,7 +1,7 @@
 # Runs python code quality checks.
 #
 # Checks for changes to python files before running the checks.
-# When manually triggered or called from another workflow, always runs the tests.
+# If always_run is true, always runs the checks.
 #
 # TODO: Add mypy or pyright to the checks.
 
@@ -18,7 +18,19 @@ on:
       - 'synchronize'
   merge_group:
   workflow_dispatch:
+    inputs:
+      always_run:
+        description: 'Always run the checks'
+        required: true
+        type: boolean
+        default: true
   workflow_call:
+    inputs:
+      always_run:
+        description: 'Always run the checks'
+        required: true
+        type: boolean
+        default: true
 
 jobs:
   python-checks:
@@ -29,7 +41,7 @@ jobs:
         uses: actions/checkout@v4
 
       - name: check for changed python files
-        if: ${{ github.event_name != 'workflow_dispatch' && github.event_name != 'workflow_call' }}
+        if: ${{ inputs.always_run != true }}
         id: changed-files
         uses: tj-actions/changed-files@v42
         with:
@@ -41,7 +53,7 @@ jobs:
               - 'tests/**'
 
       - name: setup python
-        if: ${{ steps.changed-files.outputs.python_any_changed == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call' }}
+        if: ${{ steps.changed-files.outputs.python_any_changed == 'true' || inputs.always_run == true }}
         uses: actions/setup-python@v5
         with:
           python-version: '3.10'
@@ -49,16 +61,16 @@ jobs:
           cache-dependency-path: pyproject.toml
 
       - name: install ruff
-        if: ${{ steps.changed-files.outputs.python_any_changed == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call' }}
+        if: ${{ steps.changed-files.outputs.python_any_changed == 'true' || inputs.always_run == true }}
         run: pip install ruff
         shell: bash
 
       - name: ruff check
-        if: ${{ steps.changed-files.outputs.python_any_changed == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call' }}
+        if: ${{ steps.changed-files.outputs.python_any_changed == 'true' || inputs.always_run == true }}
         run: ruff check --output-format=github .
         shell: bash
 
       - name: ruff format
-        if: ${{ steps.changed-files.outputs.python_any_changed == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call' }}
+        if: ${{ steps.changed-files.outputs.python_any_changed == 'true' || inputs.always_run == true }}
         run: ruff format --check .
         shell: bash

.github/workflows/python-tests.yml

@@ -1,7 +1,7 @@
 # Runs python tests on a matrix of python versions and platforms.
 #
 # Checks for changes to python files before running the tests.
-# When manually triggered or called from another workflow, always runs the tests.
+# If always_run is true, always runs the tests.
 
 name: 'python tests'
 
@@ -16,7 +16,19 @@ on:
       - 'synchronize'
   merge_group:
   workflow_dispatch:
+    inputs:
+      always_run:
+        description: 'Always run the tests'
+        required: true
+        type: boolean
+        default: true
   workflow_call:
+    inputs:
+      always_run:
+        description: 'Always run the tests'
+        required: true
+        type: boolean
+        default: true
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
@@ -63,7 +75,7 @@ jobs:
         uses: actions/checkout@v4
 
       - name: check for changed python files
-        if: ${{ github.event_name != 'workflow_dispatch' && github.event_name != 'workflow_call' }}
+        if: ${{ inputs.always_run != true }}
         id: changed-files
         uses: tj-actions/changed-files@v42
         with:
@@ -75,7 +87,7 @@ jobs:
               - 'tests/**'
 
       - name: setup python
-        if: ${{ steps.changed-files.outputs.python_any_changed == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call' }}
+        if: ${{ steps.changed-files.outputs.python_any_changed == 'true' || inputs.always_run == true }}
         uses: actions/setup-python@v5
         with:
           python-version: ${{ matrix.python-version }}
@@ -83,12 +95,12 @@ jobs:
           cache-dependency-path: pyproject.toml
 
       - name: install dependencies
-        if: ${{ steps.changed-files.outputs.python_any_changed == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call' }}
+        if: ${{ steps.changed-files.outputs.python_any_changed == 'true' || inputs.always_run == true }}
         env:
           PIP_EXTRA_INDEX_URL: ${{ matrix.extra-index-url }}
         run: >
           pip3 install --editable=".[test]"
 
       - name: run pytest
-        if: ${{ steps.changed-files.outputs.python_any_changed == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call' }}
+        if: ${{ steps.changed-files.outputs.python_any_changed == 'true' || inputs.always_run == true }}
         run: pytest

.github/workflows/release.yml

@@ -30,15 +30,23 @@ jobs:
 
   frontend-checks:
     uses: ./.github/workflows/frontend-checks.yml
+    with:
+      always_run: true
 
   frontend-tests:
     uses: ./.github/workflows/frontend-tests.yml
+    with:
+      always_run: true
 
   python-checks:
     uses: ./.github/workflows/python-checks.yml
+    with:
+      always_run: true
 
   python-tests:
     uses: ./.github/workflows/python-tests.yml
+    with:
+      always_run: true
 
   build:
     uses: ./.github/workflows/build-installer.yml
@@ -58,6 +66,8 @@ jobs:
     environment:
       name: testpypi
       url: https://test.pypi.org/p/invokeai
+    permissions:
+      id-token: write
     steps:
       - name: download distribution from build job
         uses: actions/download-artifact@v4
@@ -85,6 +95,8 @@ jobs:
     environment:
       name: pypi
       url: https://pypi.org/p/invokeai
+    permissions:
+      id-token: write
     steps:
       - name: download distribution from build job
         uses: actions/download-artifact@v4

Makefile

@@ -6,16 +6,18 @@ default: help
 help:
 	@echo Developer commands:
 	@echo
-	@echo "ruff              Run ruff, fixing any safely-fixable errors and formatting"
-	@echo "ruff-unsafe       Run ruff, fixing all fixable errors and formatting"
-	@echo "mypy              Run mypy using the config in pyproject.toml to identify type mismatches and other coding errors"
-	@echo "mypy-all          Run mypy ignoring the config in pyproject.tom but still ignoring missing imports"
-	@echo "test"             Run the unit tests.
-	@echo "frontend-install" Install the pnpm modules needed for the front end
-	@echo "frontend-build    Build the frontend in order to run on localhost:9090"
-	@echo "frontend-dev      Run the frontend in developer mode on localhost:5173"
-	@echo "installer-zip     Build the installer .zip file for the current version"
-	@echo "tag-release       Tag the GitHub repository with the current version (use at release time only!)"
+	@echo "ruff                     Run ruff, fixing any safely-fixable errors and formatting"
+	@echo "ruff-unsafe              Run ruff, fixing all fixable errors and formatting"
+	@echo "mypy                     Run mypy using the config in pyproject.toml to identify type mismatches and other coding errors"
+	@echo "mypy-all                 Run mypy ignoring the config in pyproject.tom but still ignoring missing imports"
+	@echo "test                     Run the unit tests."
+	@echo "update-config-docstring  Update the app's config docstring so mkdocs can autogenerate it correctly."
+	@echo "frontend-install         Install the pnpm modules needed for the front end"
+	@echo "frontend-build           Build the frontend in order to run on localhost:9090"
+	@echo "frontend-dev             Run the frontend in developer mode on localhost:5173"
+	@echo "frontend-typegen         Generate types for the frontend from the OpenAPI schema"
+	@echo "installer-zip            Build the installer .zip file for the current version"
+	@echo "tag-release              Tag the GitHub repository with the current version (use at release time only!)"
 
 # Runs ruff, fixing any safely-fixable errors and formatting
 ruff:
@@ -40,6 +42,10 @@ mypy-all:
 test:
 	pytest ./tests
 
+# Update config docstring
+update-config-docstring:
+	python scripts/update_config_docstring.py
+
 # Install the pnpm modules needed for the front end
 frontend-install:
 	rm -rf invokeai/frontend/web/node_modules
@@ -53,6 +59,9 @@ frontend-build:
 frontend-dev:
 	cd invokeai/frontend/web && pnpm dev
 
+frontend-typegen:
+	cd invokeai/frontend/web && python ../../../scripts/generate_openapi_schema.py | pnpm typegen
+
 # Installer zip file
 installer-zip:
 	cd installer && ./create_installer.sh

docs/contributing/MODEL_MANAGER.md

@@ -16,11 +16,6 @@ model. These are the:
   information. It is also responsible for managing the InvokeAI
   `models` directory and its contents.
   
-* _ModelMetadataStore_ and _ModelMetaDataFetch_ Backend modules that
-  are able to retrieve metadata from online model repositories,
-  transform them into Pydantic models, and cache them to the InvokeAI
-  SQL database.
-  
 * _DownloadQueueServiceBase_
   A multithreaded downloader responsible
   for downloading models from a remote source to disk. The download
@@ -382,17 +377,14 @@ functionality:
   
 * Downloading a model from an arbitrary URL and installing it in
   `models_dir`.
-  
-* Special handling for Civitai model URLs which allow the user to
-  paste in a model page's URL or download link
-  
+
 * Special handling for HuggingFace repo_ids to recursively download
   the contents of the repository, paying attention to alternative
   variants such as fp16.
   
 * Saving tags and other metadata about the model into the invokeai database
   when fetching from a repo that provides that type of information,
-  (currently only Civitai and HuggingFace).
+  (currently only HuggingFace).
   
 ### Initializing the installer
 
@@ -436,7 +428,6 @@ required parameters:
 | `app_config`         | InvokeAIAppConfig       | InvokeAI app configuration object |
 | `record_store`   | ModelRecordServiceBase  | Config record storage database |
 | `download_queue`   | DownloadQueueServiceBase  | Download queue object |
-| `metadata_store`   | Optional[ModelMetadataStore]  | Metadata storage object |
 |`session`           | Optional[requests.Session]    | Swap in a different Session object (usually for debugging) |
 
 Once initialized, the installer will provide the following methods:
@@ -580,33 +571,7 @@ The `AnyHttpUrl` class can be imported from `pydantic.networks`.
 
 Ordinarily, no metadata is retrieved from these sources. However,
 there is special-case code in the installer that looks for HuggingFace
-and Civitai URLs and fetches the corresponding model metadata from
-the corresponding repo.
-
-#### CivitaiModelSource
-
-This is used for a model that is hosted by the Civitai web site.
-
-| **Argument**     | **Type**                     | **Default** | **Description**                           |
-|------------------|------------------------------|-------------|-------------------------------------------|
-| `version_id`     | int                          | None        | The ID of the particular version of the desired model. |
-| `access_token`   | str                          | None        | An access token needed to gain access to a subscriber's-only model. |
-
-Civitai has two model IDs, both of which are integers. The `model_id`
-corresponds to a collection of model versions that may different in
-arbitrary ways, such as derivation from different checkpoint training
-steps, SFW vs NSFW generation, pruned vs non-pruned, etc. The
-`version_id` points to a specific version. Please use the latter.
-
-Some Civitai models require an access token to download. These can be
-generated from the Civitai profile page of a logged-in
-account. Somewhat annoyingly, if you fail to provide the access token
-when downloading a model that needs it, Civitai generates a redirect
-to a login page rather than a 403 Forbidden error. The installer
-attempts to catch this event and issue an informative error
-message. Otherwise you will get an "unrecognized model suffix" error
-when the model prober tries to identify the type of the HTML login
-page.
+and fetches the corresponding model metadata from the corresponding repo.
 
 #### HFModelSource
 
@@ -1253,9 +1218,9 @@ queue and have not yet reached a terminal state.
 
 The modules found under `invokeai.backend.model_manager.metadata`
 provide a straightforward API for fetching model metadatda from online
-repositories. Currently two repositories are supported: HuggingFace
-and Civitai. However, the modules are easily extended for additional
-repos, provided that they have defined APIs for metadata access.
+repositories. Currently only HuggingFace is supported. However, the
+modules are easily extended for additional repos, provided that they
+have defined APIs for metadata access.
 
 Metadata comprises any descriptive information that is not essential
 for getting the model to run. For example "author" is metadata, while
@@ -1267,37 +1232,16 @@ model's config, as defined in `invokeai.backend.model_manager.config`.
 ```
 from invokeai.backend.model_manager.metadata import (
    AnyModelRepoMetadata,
-   CivitaiMetadataFetch,
-   CivitaiMetadata
-   ModelMetadataStore,
 )
 # to access the initialized sql database
 from invokeai.app.api.dependencies import ApiDependencies
 
-civitai = CivitaiMetadataFetch()
+hf = HuggingFaceMetadataFetch()
 
 # fetch the metadata
-model_metadata = civitai.from_url("https://civitai.com/models/215796")
+model_metadata = hf.from_id("<repo_id>")
 
-# get some common metadata fields
-author = model_metadata.author
-tags = model_metadata.tags
-
-# get some Civitai-specific fields
-assert isinstance(model_metadata, CivitaiMetadata)
-
-trained_words = model_metadata.trained_words
-base_model = model_metadata.base_model_trained_on
-thumbnail = model_metadata.thumbnail_url
-
-# cache the metadata to the database using the key corresponding to
-# an existing model config record in the `model_config` table
-sql_cache = ModelMetadataStore(ApiDependencies.invoker.services.db)
-sql_cache.add_metadata('fb237ace520b6716adc98bcb16e8462c', model_metadata)
-
-# now we can search the database by tag, author or model name
-# matches will contain a list of model keys that match the search
-matches = sql_cache.search_by_tag({"tool", "turbo"})
+assert isinstance(model_metadata, HuggingFaceMetadata)
 ```
 
 ### Structure of the Metadata objects
@@ -1334,52 +1278,14 @@ This descends from `ModelMetadataBase` and adds the following fields:
 | `last_modified`| datetime         | Date of last commit of this model to the repo |
 | `files`        | List[Path]       | List of the files in the model repo |
 
-#### `CivitaiMetadata`
-
-This descends from `ModelMetadataBase` and adds the following fields:
-
-| **Field Name** | **Type**        |  **Description** |
-|----------------|-----------------|------------------|
-| `type`         | Literal["civitai"]   | Used for the discriminated union of metadata classes|
-| `id`           | int                  | Civitai model id |
-| `version_name` | str                  | Name of this version of the model (distinct from model name) |
-| `version_id`   | int                  | Civitai model version id (distinct from model id) |
-| `created`      | datetime             | Date this version of the model was created |
-| `updated`      | datetime             | Date this version of the model was last updated |
-| `published`    | datetime             | Date this version of the model was published to Civitai |
-| `description`  | str                  | Model description. Quite verbose and contains HTML tags |
-| `version_description` | str           | Model version description, usually describes changes to the model |
-| `nsfw`         | bool                 | Whether the model tends to generate NSFW content |
-| `restrictions` | LicenseRestrictions  | An object that describes what is and isn't allowed with this model |
-| `trained_words`| Set[str]             | Trigger words for this model, if any |
-| `download_url` | AnyHttpUrl           | URL for downloading this version of the model |
-| `base_model_trained_on` | str         | Name of the model that this version was trained on |
-| `thumbnail_url` | AnyHttpUrl          | URL to access a representative thumbnail image of the model's output |
-| `weight_min`    | int                 | For LoRA sliders, the minimum suggested weight to apply |
-| `weight_max`    | int                 | For LoRA sliders, the maximum suggested weight to apply |
-
-Note that `weight_min` and `weight_max` are not currently populated
-and take the default values of (-1.0, +2.0). The issue is that these
-values aren't part of the structured data but appear in the text
-description. Some regular expression or LLM coding may be able to
-extract these values.
-
-Also be aware that `base_model_trained_on` is free text and doesn't
-correspond to our `ModelType` enum.
-
-`CivitaiMetadata` also defines some convenience properties relating to
-licensing restrictions: `credit_required`, `allow_commercial_use`,
-`allow_derivatives` and `allow_different_license`.
-
 #### `AnyModelRepoMetadata`
 
-This is a discriminated Union of `CivitaiMetadata` and
-`HuggingFaceMetadata`.
+This is a discriminated Union of `HuggingFaceMetadata`.
 
 ### Fetching Metadata from Online Repos
 
-The `HuggingFaceMetadataFetch` and `CivitaiMetadataFetch` classes will
-retrieve metadata from their corresponding repositories and return
+The `HuggingFaceMetadataFetch` class will
+retrieve metadata from its corresponding repository and return
 `AnyModelRepoMetadata` objects. Their base class
 `ModelMetadataFetchBase` is an abstract class that defines two
 methods: `from_url()` and `from_id()`. The former accepts the type of
@@ -1397,96 +1303,17 @@ provide a `requests.Session` argument. This allows you to customize
 the low-level HTTP fetch requests and is used, for instance, in the
 testing suite to avoid hitting the internet.
 
-The HuggingFace and Civitai fetcher subclasses add additional
-repo-specific fetching methods:
+The HuggingFace fetcher subclass add additional repo-specific fetching methods:
 
 #### HuggingFaceMetadataFetch
 
 This overrides its base class `from_json()` method to return a
 `HuggingFaceMetadata` object directly.
 
-#### CivitaiMetadataFetch
-
-This adds the following methods:
-
-`from_civitai_modelid()` This takes the ID of a model, finds the
-default version of the model, and then retrieves the metadata for
-that version, returning a `CivitaiMetadata` object directly.
-
-`from_civitai_versionid()` This takes the ID of a model version and
-retrieves its metadata. Functionally equivalent to `from_id()`, the
-only difference is that it returna a `CivitaiMetadata` object rather
-than an `AnyModelRepoMetadata`.
-
 ### Metadata Storage
 
-The `ModelMetadataStore` provides a simple facility to store model
-metadata in the `invokeai.db` database. The data is stored as a JSON
-blob, with a few common fields (`name`, `author`, `tags`) broken out
-to be searchable.
-
-When a metadata object is saved to the database, it is identified
-using the model key, _and this key must correspond to an existing
-model key in the model_config table_. There is a foreign key integrity
-constraint between the `model_config.id` field and the
-`model_metadata.id` field such that if you attempt to save metadata
-under an unknown key, the attempt will result in an
-`UnknownModelException`. Likewise, when a model is deleted from
-`model_config`, the deletion of the corresponding metadata record will
-be triggered.
-
-Tags are stored in a normalized fashion in the tables `model_tags` and
-`tags`. Triggers keep the tag table in sync with the `model_metadata`
-table.
-
-To create the storage object, initialize it with the InvokeAI
-`SqliteDatabase` object. This is often done this way:
-
-```
-from invokeai.app.api.dependencies import ApiDependencies
-metadata_store = ModelMetadataStore(ApiDependencies.invoker.services.db)
-```
-
-You can then access the storage with the following methods:
-
-#### `add_metadata(key, metadata)`
-
-Add the metadata using a previously-defined model key.
-
-There is currently no `delete_metadata()` method. The metadata will
-persist until the matching config is deleted from the `model_config`
-table.
-
-#### `get_metadata(key) -> AnyModelRepoMetadata`
-
-Retrieve the metadata corresponding to the model key.
-
-#### `update_metadata(key, new_metadata)`
-
-Update an existing metadata record with new metadata.
-
-#### `search_by_tag(tags: Set[str]) -> Set[str]`
-
-Given a set of tags, find models that are tagged with them. If
-multiple tags are provided then a matching model must be tagged with
-*all* the tags in the set. This method returns a set of model keys and
-is intended to be used in conjunction with the `ModelRecordService`:
-
-```
-model_config_store = ApiDependencies.invoker.services.model_records
-matches = metadata_store.search_by_tag({'license:other'})
-models = [model_config_store.get(x) for x in matches]
-```
-
-#### `search_by_name(name: str) -> Set[str]
-
-Find all model metadata records that have the given name and return a
-set of keys to the corresponding model config objects.
-
-#### `search_by_author(author: str) -> Set[str]
-
-Find all model metadata records that have the given author and return
-a set of keys to the corresponding model config objects.
+The `ModelConfigBase` stores this response in the `source_api_response` field
+as a JSON blob.
 
 ***
 

docs/contributing/frontend/OVERVIEW.md

@@ -0,0 +1,133 @@
+# Invoke UI
+
+Invoke's UI is made possible by many contributors and open-source libraries. Thank you!
+
+## Dev environment
+
+### Setup
+
+1. Install [node] and [pnpm].
+1. Run `pnpm i` to install all packages.
+
+#### Run in dev mode
+
+1. From `invokeai/frontend/web/`, run `pnpm dev`.
+1. From repo root, run `python scripts/invokeai-web.py`.
+1. Point your browser to the dev server address, e.g. <http://localhost:5173/>
+
+### Package scripts
+
+- `dev`: run the frontend in dev mode, enabling hot reloading
+- `build`: run all checks (madge, eslint, prettier, tsc) and then build the frontend
+- `typegen`: generate types from the OpenAPI schema (see [Type generation])
+- `lint:dpdm`: check circular dependencies
+- `lint:eslint`: check code quality
+- `lint:prettier`: check code formatting
+- `lint:tsc`: check type issues
+- `lint:knip`: check for unused exports or objects (failures here are just suggestions, not hard fails)
+- `lint`: run all checks concurrently
+- `fix`: run `eslint` and `prettier`, fixing fixable issues
+
+### Type generation
+
+We use [openapi-typescript] to generate types from the app's OpenAPI schema.
+
+The generated types are committed to the repo in [schema.ts].
+
+```sh
+# from the repo root, start the server
+python scripts/invokeai-web.py
+# from invokeai/frontend/web/, run the script
+pnpm typegen
+```
+
+### Localization
+
+We use [i18next] for localization, but translation to languages other than English happens on our [Weblate] project.
+
+Only the English source strings should be changed on this repo.
+
+### VSCode
+
+#### Example debugger config
+
+```jsonc
+{
+  "version": "0.2.0",
+  "configurations": [
+    {
+      "type": "chrome",
+      "request": "launch",
+      "name": "Invoke UI",
+      "url": "http://localhost:5173",
+      "webRoot": "${workspaceFolder}/invokeai/frontend/web"
+    }
+  ]
+}
+```
+
+#### Remote dev
+
+We've noticed an intermittent timeout issue with the VSCode remote dev port forwarding.
+
+We suggest disabling the editor's port forwarding feature and doing it manually via SSH:
+
+```sh
+ssh -L 9090:localhost:9090 -L 5173:localhost:5173 user@host
+```
+
+## Contributing Guidelines
+
+Thanks for your interest in contributing to the Invoke Web UI!
+
+Please follow these guidelines when contributing.
+
+### Check in before investing your time
+
+Please check in before you invest your time on anything besides a trivial fix, in case it conflicts with ongoing work or isn't aligned with the vision for the app.
+
+If a feature request or issue doesn't already exist for the thing you want to work on, please create one.
+
+Ping `@psychedelicious` on [discord] in the `#frontend-dev` channel or in the feature request / issue you want to work on - we're happy to chat.
+
+### Code conventions
+
+- This is a fairly complex app with a deep component tree. Please use memoization (`useCallback`, `useMemo`, `memo`) with enthusiasm.
+- If you need to add some global, ephemeral state, please use [nanostores] if possible.
+- Be careful with your redux selectors. If they need to be parameterized, consider creating them inside a `useMemo`.
+- Feel free to use `lodash` (via `lodash-es`) to make the intent of your code clear.
+- Please add comments describing the "why", not the "how" (unless it is really arcane).
+
+### Commit format
+
+Please use the [conventional commits] spec for the web UI, with a scope of "ui":
+
+- `chore(ui): bump deps`
+- `chore(ui): lint`
+- `feat(ui): add some cool new feature`
+- `fix(ui): fix some bug`
+
+### Submitting a PR
+
+- Ensure your branch is tidy. Use an interactive rebase to clean up the commit history and reword the commit messages if they are not descriptive.
+- Run `pnpm lint`. Some issues are auto-fixable with `pnpm fix`.
+- Fill out the PR form when creating the PR.
+  - It doesn't need to be super detailed, but a screenshot or video is nice if you changed something visually.
+  - If a section isn't relevant, delete it. There are no UI tests at this time.
+
+## Other docs
+
+- [Workflows - Design and Implementation]
+- [State Management]
+
+[node]: https://nodejs.org/en/download/
+[pnpm]: https://github.com/pnpm/pnpm
+[discord]: https://discord.gg/ZmtBAhwWhy
+[i18next]: https://github.com/i18next/react-i18next
+[Weblate]: https://hosted.weblate.org/engage/invokeai/
+[openapi-typescript]: https://github.com/drwpow/openapi-typescript
+[Type generation]: #type-generation
+[schema.ts]: https://github.com/invoke-ai/InvokeAI/blob/main/invokeai/frontend/web/src/services/api/schema.ts
+[conventional commits]: https://www.conventionalcommits.org/en/v1.0.0/
+[Workflows - Design and Implementation]: ./WORKFLOWS.md
+[State Management]: ./STATE_MGMT.md

docs/contributing/frontend/WORKFLOWS.md

@@ -1,40 +1,5 @@
 # Workflows - Design and Implementation
 
-<!-- @import "[TOC]" {cmd="toc" depthFrom=1 depthTo=6 orderedList=false} -->
-
-<!-- code_chunk_output -->
-
-- [Workflows - Design and Implementation](#workflows---design-and-implementation)
-  - [Design](#design)
-    - [Linear UI](#linear-ui)
-    - [Workflow Editor](#workflow-editor)
-      - [Workflows](#workflows)
-        - [Workflow -> reactflow state -> InvokeAI graph](#workflow---reactflow-state---invokeai-graph)
-        - [Nodes vs Invocations](#nodes-vs-invocations)
-        - [Workflow Linear View](#workflow-linear-view)
-      - [OpenAPI Schema](#openapi-schema)
-        - [Field Instances and Templates](#field-instances-and-templates)
-        - [Stateful vs Stateless Fields](#stateful-vs-stateless-fields)
-        - [Collection and Polymorphic Fields](#collection-and-polymorphic-fields)
-  - [Implementation](#implementation)
-    - [zod Schemas and Types](#zod-schemas-and-types)
-    - [OpenAPI Schema Parsing](#openapi-schema-parsing)
-      - [Parsing Field Types](#parsing-field-types)
-        - [Primitive Types](#primitive-types)
-        - [Complex Types](#complex-types)
-        - [Collection Types](#collection-types)
-        - [Collection or Scalar Types](#collection-or-scalar-types)
-        - [Optional Fields](#optional-fields)
-      - [Building Field Input Templates](#building-field-input-templates)
-      - [Building Field Output Templates](#building-field-output-templates)
-    - [Managing reactflow State](#managing-reactflow-state)
-      - [Building Nodes and Edges](#building-nodes-and-edges)
-      - [Building a Workflow](#building-a-workflow)
-      - [Loading a Workflow](#loading-a-workflow)
-    - [Workflow Migrations](#workflow-migrations)
-
-<!-- /code_chunk_output -->
-
 > This document describes, at a high level, the design and implementation of workflows in the InvokeAI frontend. There are a substantial number of implementation details not included, but which are hopefully clear from the code.
 
 InvokeAI's backend uses graphs, composed of **nodes** and **edges**, to process data and generate images.
@@ -152,13 +117,13 @@ Stateless fields do not store their value in the node, so their field instances
 
 "Custom" fields will always be treated as stateless fields.
 
-##### Collection and Polymorphic Fields
+##### Collection and Scalar Fields
 
-Field types have a name and two flags which may identify it as a **collection** or **polymorphic** field.
+Field types have a name and two flags which may identify it as a **collection** or **collection or scalar** field.
 
-If a field is annotated in python as a list, its field type is parsed and flagged as a collection type (e.g. `list[int]`).
+If a field is annotated in python as a list, its field type is parsed and flagged as a **collection** type (e.g. `list[int]`).
 
-If it is annotated as a union of a type and list, the type will be flagged as a polymorphic type (e.g. `Union[int, list[int]]`). Fields may not be unions of different types (e.g. `Union[int, list[str]]` and `Union[int, str]` are not allowed).
+If it is annotated as a union of a type and list, the type will be flagged as a **collection or scalar** type (e.g. `Union[int, list[int]]`). Fields may not be unions of different types (e.g. `Union[int, list[str]]` and `Union[int, str]` are not allowed).
 
 ## Implementation
 
@@ -338,13 +303,13 @@ Migration logic is in [migrations.ts].
 [reactflow]: https://github.com/xyflow/xyflow 'reactflow'
 [reactflow-concepts]: https://reactflow.dev/learn/concepts/terms-and-definitions
 [reactflow-events]: https://reactflow.dev/api-reference/react-flow#event-handlers
-[buildWorkflow.ts]: ../src/features/nodes/util/workflow/buildWorkflow.ts
-[nodesSlice.ts]: ../src/features/nodes/store/nodesSlice.ts
-[buildLinearTextToImageGraph.ts]: ../src/features/nodes/util/graph/buildLinearTextToImageGraph.ts
-[buildNodesGraph.ts]: ../src/features/nodes/util/graph/buildNodesGraph.ts
-[buildInvocationNode.ts]: ../src/features/nodes/util/node/buildInvocationNode.ts
-[validateWorkflow.ts]: ../src/features/nodes/util/workflow/validateWorkflow.ts
-[migrations.ts]: ../src/features/nodes/util/workflow/migrations.ts
-[parseSchema.ts]: ../src/features/nodes/util/schema/parseSchema.ts
-[buildFieldInputTemplate.ts]: ../src/features/nodes/util/schema/buildFieldInputTemplate.ts
-[buildFieldOutputTemplate.ts]: ../src/features/nodes/util/schema/buildFieldOutputTemplate.ts
+[buildWorkflow.ts]: https://github.com/invoke-ai/InvokeAI/blob/main/invokeai/frontend/web/src/features/nodes/util/workflow/buildWorkflow.ts
+[nodesSlice.ts]: https://github.com/invoke-ai/InvokeAI/blob/main/invokeai/frontend/web/src/features/nodes/store/nodesSlice.ts
+[buildLinearTextToImageGraph.ts]: https://github.com/invoke-ai/InvokeAI/blob/main/invokeai/frontend/web/src/features/nodes/util/graph/buildLinearTextToImageGraph.ts
+[buildNodesGraph.ts]: https://github.com/invoke-ai/InvokeAI/blob/main/invokeai/frontend/web/src/features/nodes/util/graph/buildNodesGraph.ts
+[buildInvocationNode.ts]: https://github.com/invoke-ai/InvokeAI/blob/main/invokeai/frontend/web/src/features/nodes/util/node/buildInvocationNode.ts
+[validateWorkflow.ts]: https://github.com/invoke-ai/InvokeAI/blob/main/invokeai/frontend/web/src/features/nodes/util/workflow/validateWorkflow.ts
+[migrations.ts]: https://github.com/invoke-ai/InvokeAI/blob/main/invokeai/frontend/web/src/features/nodes/util/workflow/migrations.ts
+[parseSchema.ts]: https://github.com/invoke-ai/InvokeAI/blob/main/invokeai/frontend/web/src/features/nodes/util/schema/parseSchema.ts
+[buildFieldInputTemplate.ts]: https://github.com/invoke-ai/InvokeAI/blob/main/invokeai/frontend/web/src/features/nodes/util/schema/buildFieldInputTemplate.ts
+[buildFieldOutputTemplate.ts]: https://github.com/invoke-ai/InvokeAI/blob/main/invokeai/frontend/web/src/features/nodes/util/schema/buildFieldOutputTemplate.ts

docs/features/CONFIGURATION.md

@@ -6,259 +6,162 @@ title: Configuration
 
 ## Intro
 
-InvokeAI has numerous runtime settings which can be used to adjust
-many aspects of its operations, including the location of files and
-directories, memory usage, and performance. These settings can be
-viewed and customized in several ways:
+Runtime settings, including the location of files and
+directories, memory usage, and performance, are managed via the
+`invokeai.yaml` config file or environment variables. A subset
+of settings may be set via commandline arguments.
 
-1. By editing settings in the `invokeai.yaml` file.
-2. By setting environment variables.
-3. On the command-line, when InvokeAI is launched.
+Settings sources are used in this order:
 
-In addition, the most commonly changed settings are accessible
+- CLI args
+- Environment variables
+- `invokeai.yaml` settings
+- Fallback: defaults
+
+The most commonly changed settings are also accessible
 graphically via the `invokeai-configure` script.
 
-### How the Configuration System Works
+### InvokeAI Root Directory
 
-When InvokeAI is launched, the very first thing it needs to do is to
-find its "root" directory, which contains its configuration files,
-installed models, its database of images, and the folder(s) of
-generated images themselves. In this document, the root directory will
-be referred to as ROOT.
+On startup, InvokeAI searches for its "root" directory. This is the directory
+that contains models, images, the database, and so on. It also contains
+a configuration file called `invokeai.yaml`.
 
-#### Finding the Root Directory
+InvokeAI searches for the root directory in this order:
 
-To find its root directory, InvokeAI uses the following recipe:
+1. The `--root <path>` CLI arg.
+2. The environment variable INVOKEAI_ROOT.
+3. The directory containing the currently active virtual environment.
+4. Fallback: a directory in the current user's home directory named `invokeai`.
 
-1. It first looks for the argument `--root <path>` on the command line
-it was launched from, and uses the indicated path if present.
+### InvokeAI Configuration File
 
-2. Next it looks for the environment variable INVOKEAI_ROOT, and uses
-the directory path found there if present.
+Inside the root directory, we read settings from the `invokeai.yaml` file.
 
-3. If neither of these are present, then InvokeAI looks for the
-folder containing the `.venv` Python virtual environment directory for
-the currently active environment. This directory is checked for files
-expected inside the InvokeAI root before it is used.
+It has two sections - one for internal use and one for user settings:
 
-4. Finally, InvokeAI looks for a directory in the current user's home
-directory named `invokeai`.
+```yaml
+# Internal metadata - do not edit:
+meta:
+  schema_version: 4
 
-#### Reading the InvokeAI Configuration File
-
-Once the root directory has been located, InvokeAI looks for a file
-named `ROOT/invokeai.yaml`, and if present reads configuration values
-from it. The top of this file looks like this:
-
-```
-InvokeAI:
-  Web Server:
-    host: localhost
-    port: 9090
-    allow_origins: []
-    allow_credentials: true
-    allow_methods:
-    - '*'
-    allow_headers:
-    - '*'
-  Features:
-    esrgan: true
-    internet_available: true
-    log_tokenization: false
-    patchmatch: true
-    restore: true
-...
+# Put user settings here:
+host: 0.0.0.0 # serve the app on your local network
+models_dir: D:\invokeai\models # store models on an external drive
+precision: float16 # always use fp16 precision
 ```
 
-This lines in this file are used to establish default values for
-Invoke's settings. In the above fragment, the Web Server's listening
-port is set to 9090 by the `port` setting.
+The settings in this file will override the defaults. You only need
+to change this file if the default for a particular setting doesn't
+work for you.
 
-You can edit this file with a text editor such as "Notepad" (do not
-use Word or any other word processor). When editing, be careful to
-maintain the indentation, and do not add extraneous text, as syntax
-errors will prevent InvokeAI from launching. A basic guide to the
-format of YAML files can be found
-[here](https://circleci.com/blog/what-is-yaml-a-beginner-s-guide/).
+Some settings, like [Model Marketplace API Keys], require the YAML
+to be formatted correctly. Here is a [basic guide to YAML files].
 
 You can fix a broken `invokeai.yaml` by deleting it and running the
 configuration script again -- option [6] in the launcher, "Re-run the
 configure script".
 
-#### Reading Environment Variables
+### Environment Variables
 
-Next InvokeAI looks for defined environment variables in the format
-`INVOKEAI_<setting_name>`, for example `INVOKEAI_port`. Environment
-variable values take precedence over configuration file variables. On
-a Macintosh system, for example, you could change the port that the
-web server listens on by setting the environment variable this way:
+All settings may be set via environment variables by prefixing `INVOKEAI_`
+to the variable name. For example, `INVOKEAI_HOST` would set the `host`
+setting.
 
-```
-export INVOKEAI_port=8000
-invokeai-web
+For non-primitive values, pass a JSON-encoded string:
+
+```sh
+export INVOKEAI_REMOTE_API_TOKENS='[{"url_regex":"modelmarketplace", "token": "12345"}]'
 ```
 
-Please check out these
-[Macintosh](https://phoenixnap.com/kb/set-environment-variable-mac)
-and
-[Windows](https://phoenixnap.com/kb/windows-set-environment-variable)
-guides for setting temporary and permanent environment variables.
+We suggest using `invokeai.yaml`, as it is more user-friendly.
 
-#### Reading the Command Line
+### CLI Args
 
-Lastly, InvokeAI takes settings from the command line, which override
-everything else. The command-line settings have the same name as the
-corresponding configuration file settings, preceded by a `--`, for
-example `--port 8000`.
+A subset of settings may be specified using CLI args:
 
-If you are using the launcher (`invoke.sh` or `invoke.bat`) to launch
-InvokeAI, then just pass the command-line arguments to the launcher:
+- `--root`: specify the root directory
+- `--ignore_missing_core-models`: if set, do not check for models needed
+  to convert checkpoint/safetensor models to diffusers
 
-```
-invoke.bat --port 8000 --host 0.0.0.0
+### All Settings
+
+The config is managed by the `InvokeAIAppConfig` class. The below docs are autogenerated from the class.
+
+Following the table are additional explanations for certain settings.
+
+<!-- prettier-ignore-start -->
+::: invokeai.app.services.config.config_default.InvokeAIAppConfig
+    options:
+        heading_level: 4
+        members: false
+        show_docstring_description: false
+        group_by_category: true
+        show_category_heading: false
+<!-- prettier-ignore-end -->
+
+#### Model Marketplace API Keys
+
+Some model marketplaces require an API key to download models. You can provide a URL pattern and appropriate token in your `invokeai.yaml` file to provide that API key.
+
+The pattern can be any valid regex (you may need to surround the pattern with quotes):
+
+```yaml
+remote_api_tokens:
+  # Any URL containing `models.com` will automatically use `your_models_com_token`
+  - url_regex: models.com
+    token: your_models_com_token
+  # Any URL matching this contrived regex will use `some_other_token`
+  - url_regex: '^[a-z]{3}whatever.*\.com$'
+    token: some_other_token
 ```
 
-The arguments will be applied when you select the web server option
-(and the other options as well).
+The provided token will be added as a `Bearer` token to the network requests to download the model files. As far as we know, this works for all model marketplaces that require authorization.
 
-If, on the other hand, you prefer to launch InvokeAI directly from the
-command line, you would first activate the virtual environment (known
-as the "developer's console" in the launcher), and run `invokeai-web`:
+#### Model Hashing
 
-```
-> C:\Users\Fred\invokeai\.venv\scripts\activate
-(.venv) > invokeai-web --port 8000 --host 0.0.0.0
+Models are hashed during installation, providing a stable identifier for models across all platforms. The default algorithm is `blake3`, with a multi-threaded implementation.
+
+If your models are stored on a spinning hard drive, we suggest using `blake3_single`, the single-threaded implementation. The hashes are the same, but it's much faster on spinning disks.
+
+```yaml
+hashing_algorithm: blake3_single
 ```
 
-You can get a listing and brief instructions for each of the
-command-line options by giving the `--help` argument:
+Model hashing is a one-time operation, but it may take a couple minutes to hash a large model collection. You may opt out of model hashing entirely by setting the algorithm to `random`.
 
-```
-(.venv) > invokeai-web --help
-usage: InvokeAI [-h] [--host HOST] [--port PORT] [--allow_origins [ALLOW_ORIGINS ...]] [--allow_credentials | --no-allow_credentials] [--allow_methods [ALLOW_METHODS ...]]
-                [--allow_headers [ALLOW_HEADERS ...]] [--esrgan | --no-esrgan] [--internet_available | --no-internet_available] [--log_tokenization | --no-log_tokenization]
-                [--patchmatch | --no-patchmatch] [--restore | --no-restore]
-                [--always_use_cpu | --no-always_use_cpu] [--free_gpu_mem | --no-free_gpu_mem] [--max_loaded_models MAX_LOADED_MODELS] [--max_cache_size MAX_CACHE_SIZE]
-                [--max_vram_cache_size MAX_VRAM_CACHE_SIZE] [--gpu_mem_reserved GPU_MEM_RESERVED] [--precision {auto,float16,float32,autocast}]
-                [--sequential_guidance | --no-sequential_guidance] [--xformers_enabled | --no-xformers_enabled] [--tiled_decode | --no-tiled_decode] [--root ROOT]
-                [--autoimport_dir AUTOIMPORT_DIR] [--lora_dir LORA_DIR] [--embedding_dir EMBEDDING_DIR] [--controlnet_dir CONTROLNET_DIR] [--conf_path CONF_PATH]
-                [--models_dir MODELS_DIR] [--legacy_conf_dir LEGACY_CONF_DIR] [--db_dir DB_DIR] [--outdir OUTDIR] [--from_file FROM_FILE]
-                [--use_memory_db | --no-use_memory_db] [--model MODEL] [--log_handlers [LOG_HANDLERS ...]] [--log_format {plain,color,syslog,legacy}]
-                [--log_level {debug,info,warning,error,critical}] [--version | --no-version]
+```yaml
+hashing_algorithm: random
 ```
 
-## The Configuration Settings
+Most common algorithms are supported, like `md5`, `sha256`, and `sha512`. These are typically much, much slower than `blake3`.
 
-The configuration settings are divided into several distinct
-groups in `invokeia.yaml`:
-
-### Web Server
-
-| Setting             | Default Value | Description                                                                                                                |
-|---------------------|---------------|----------------------------------------------------------------------------------------------------------------------------|
-| `host`              | `localhost`   | Name or IP address of the network interface that the web server will listen on                                             |
-| `port`              | `9090`        | Network port number that the web server will listen on                                                                     |
-| `allow_origins`     | `[]`          | A list of host names or IP addresses that are allowed to connect to the InvokeAI API in the format `['host1','host2',...]` |
-| `allow_credentials` | `true`        | Require credentials for a foreign host to access the InvokeAI API (don't change this)                                      |
-| `allow_methods`     | `*`           | List of HTTP methods ("GET", "POST") that the web server is allowed to use when accessing the API                          |
-| `allow_headers`     | `*`           | List of HTTP headers that the web server will accept when accessing the API                                                |
-| `ssl_certfile`      | null          | Path to an SSL certificate file, used to enable HTTPS.                                                                     |
-| `ssl_keyfile`       | null          | Path to an SSL keyfile, if the key is not included in the certificate file.                                                |
-
-The documentation for InvokeAI's API can be accessed by browsing to the following URL: [http://localhost:9090/docs].
-
-### Features
-
-These configuration settings allow you to enable and disable various InvokeAI features:
-
-| Setting  | Default Value  |  Description |
-|----------|----------------|--------------|
-| `esrgan`     | `true`      | Activate the ESRGAN upscaling options|
-| `internet_available` | `true`     | When a resource is not available locally, try to fetch it via the internet |
-| `log_tokenization` | `false`      | Before each text2image generation, print a color-coded representation of the prompt to the console; this can help understand why a prompt is not working as expected |
-| `patchmatch` | `true`     | Activate the "patchmatch" algorithm for improved inpainting |
-
-### Generation
-
-These options tune InvokeAI's memory and performance characteristics.
-
-| Setting               | Default Value | Description                                                                                                                                                                                                                                                      |
-|-----------------------|---------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
-| `sequential_guidance` | `false`       | Calculate guidance in serial rather than in parallel, lowering memory requirements at the cost of some performance loss                                                                                                                                          |
-| `attention_type`      | `auto`        | Select the type of attention to use. One of `auto`,`normal`,`xformers`,`sliced`, or `torch-sdp`                                                                                                                                                                  |
-| `attention_slice_size` | `auto`       | When "sliced" attention is selected, set the slice size. One of `auto`, `balanced`, `max` or the integers 1-8|
-| `force_tiled_decode`  | `false`       | Force the VAE step to decode in tiles, reducing memory consumption at the cost of performance |
-
-### Device
-
-These options configure the generation execution device.
-
-| Setting               | Default Value | Description                                                                                                                                                                                                                                                      |
-|-----------------------|---------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
-| `device`           | `auto`        | Preferred execution device. One of `auto`, `cpu`, `cuda`, `cuda:1`, `mps`. `auto` will choose the device depending on the hardware platform and the installed torch capabilities. |
-| `precision`           | `auto`        | Floating point precision. One of `auto`, `float16` or `float32`. `float16` will consume half the memory of `float32` but produce slightly lower-quality images. The `auto` setting will guess the proper precision based on your video card and operating system |
-
-
-### Paths
+#### Path Settings
 
 These options set the paths of various directories and files used by
-InvokeAI. Relative paths are interpreted relative to INVOKEAI_ROOT, so
-if INVOKEAI_ROOT is `/home/fred/invokeai` and the path is
+InvokeAI. Relative paths are interpreted relative to the root directory, so
+if root is `/home/fred/invokeai` and the path is
 `autoimport/main`, then the corresponding directory will be located at
 `/home/fred/invokeai/autoimport/main`.
 
-| Setting  | Default Value  |  Description |
-|----------|----------------|--------------|
-| `autoimport_dir` | `autoimport/main`     | At startup time, read and import any main model files found in this directory |
-| `lora_dir` | `autoimport/lora`     | At startup time, read and import any LoRA/LyCORIS models found in this directory |
-| `embedding_dir` | `autoimport/embedding`  | At startup time, read and import any textual inversion (embedding) models found in this directory |
-| `controlnet_dir` | `autoimport/controlnet`  | At startup time, read and import any ControlNet models found in this directory |
-| `conf_path` | `configs/models.yaml`  | Location of the `models.yaml` model configuration file |
-| `models_dir` | `models`  | Location of the directory containing models installed by InvokeAI's model manager |
-| `legacy_conf_dir` | `configs/stable-diffusion`  | Location of the directory containing the .yaml configuration files for legacy checkpoint models |
-| `db_dir` | `databases`  | Location of the directory containing InvokeAI's image, schema and session database |
-| `outdir` | `outputs`  | Location of the directory in which the gallery of generated and uploaded images will be stored |
-| `use_memory_db` | `false`  | Keep database information in memory rather than on disk; this will not preserve image gallery information across restarts |
-
-Note that the autoimport directories will be searched recursively,
+Note that the autoimport directory will be searched recursively,
 allowing you to organize the models into folders and subfolders in any
-way you wish. In addition, while we have split up autoimport
-directories by the type of model they contain, this isn't
-necessary. You can combine different model types in the same folder
-and InvokeAI will figure out what they are. So you can easily use just
-one autoimport directory by commenting out the unneeded paths:
+way you wish.
 
-```
-Paths:
-  autoimport_dir: autoimport
-#  lora_dir: null
-#  embedding_dir: null
-#  controlnet_dir: null
-```
-
-### Logging
-
-These settings control the information, warning, and debugging
-messages printed to the console log while InvokeAI is running:
-
-| Setting  | Default Value  |  Description |
-|----------|----------------|--------------|
-| `log_handlers` | `console` | This controls where log messages are sent, and can be a list of one or more destinations. Values include `console`, `file`, `syslog` and `http`. These are described in more detail below |
-| `log_format` | `color` | This controls the formatting of the log messages. Values are `plain`, `color`, `legacy` and `syslog` |
-| `log_level`  | `debug` | This filters messages according to the level of severity and can be one of `debug`, `info`, `warning`, `error` and `critical`. For example, setting to `warning` will display all messages at the warning level or higher, but won't display "debug" or "info" messages |
+#### Logging
 
 Several different log handler destinations are available, and multiple destinations are supported by providing a list:
 
-```
-  log_handlers:
-     - console
-     - syslog=localhost
-     - file=/var/log/invokeai.log
+```yaml
+log_handlers:
+  - console
+  - syslog=localhost
+  - file=/var/log/invokeai.log
 ```
 
-* `console` is the default. It prints log messages to the command-line window from which InvokeAI was launched.
+- `console` is the default. It prints log messages to the command-line window from which InvokeAI was launched.
 
-* `syslog` is only available on Linux and Macintosh systems. It uses
+- `syslog` is only available on Linux and Macintosh systems. It uses
   the operating system's "syslog" facility to write log file entries
   locally or to a remote logging machine. `syslog` offers a variety
   of configuration options:
@@ -271,7 +174,7 @@ Several different log handler destinations are available, and multiple destinati
                         - Log to LAN-connected server "fredserver" using the facility LOG_USER and datagram packets.
 ```
 
-* `http` can be used to log to a remote web server. The server must be
+- `http` can be used to log to a remote web server. The server must be
   properly configured to receive and act on log messages. The option
   accepts the URL to the web server, and a `method` argument
   indicating whether the message should be submitted using the GET or
@@ -283,7 +186,10 @@ Several different log handler destinations are available, and multiple destinati
 
 The `log_format` option provides several alternative formats:
 
-* `color`    - default format providing time, date and a message, using text colors to distinguish different log severities
-* `plain`    - same as above, but monochrome text only
-* `syslog`   - the log level and error message only, allowing the syslog system to attach the time and date
-* `legacy`   - a format similar to the one used by the legacy 2.3 InvokeAI releases.
+- `color` - default format providing time, date and a message, using text colors to distinguish different log severities
+- `plain` - same as above, but monochrome text only
+- `syslog` - the log level and error message only, allowing the syslog system to attach the time and date
+- `legacy` - a format similar to the one used by the legacy 2.3 InvokeAI releases.
+
+[basic guide to yaml files]: https://circleci.com/blog/what-is-yaml-a-beginner-s-guide/
+[Model Marketplace API Keys]: #model-marketplace-api-keys

docs/features/DATABASE.md

@@ -0,0 +1,35 @@
+---
+title: Database
+---
+
+# Invoke's SQLite Database
+
+Invoke uses a SQLite database to store image, workflow, model, and execution data.
+
+We take great care to ensure your data is safe, by utilizing transactions and a database migration system.
+
+Even so, when testing an prerelease version of the app, we strongly suggest either backing up your database or using an in-memory database. This ensures any prelease hiccups or databases schema changes will not cause problems for your data.
+
+## Database Backup
+
+Backing up your database is very simple. Invoke's data is stored in an `$INVOKEAI_ROOT` directory - where your `invoke.sh`/`invoke.bat` and `invokeai.yaml` files live.
+
+To back up your database, copy the `invokeai.db` file from `$INVOKEAI_ROOT/databases/invokeai.db` to somewhere safe.
+
+If anything comes up during prelease testing, you can simply copy your backup back into `$INVOKEAI_ROOT/databases/`.
+
+## In-Memory Database
+
+SQLite can run on an in-memory database. Your existing database is untouched when this mode is enabled, but your existing data won't be accessible.
+
+This is very useful for testing, as there is no chance of a database change modifying your "physical" database.
+
+To run Invoke with a memory database, edit your `invokeai.yaml` file, and add `use_memory_db: true` to the `Paths:` stanza:
+
+```yaml
+InvokeAI:
+  Development:
+    use_memory_db: true
+```
+
+Delete this line (or set it to `false`) to use your main database.

docs/nodes/INVOCATION_API.md

@@ -22,6 +22,24 @@ class MyInvocation(BaseInvocation):
       ...
 ```
 
+The full API is documented below.
+
+## Invocation Mixins
+
+Two important mixins are provided to facilitate working with metadata and gallery boards.
+
+### `WithMetadata`
+
+Inherit from this class (in addition to `BaseInvocation`) to add a `metadata` input to your node. When you do this, you can access the metadata dict from `self.metadata` in the `invoke()` function.
+
+The dict will be populated via the node's input, and you can add any metadata you'd like to it. When you call `context.images.save()`, if the metadata dict has any data, it be automatically embedded in the image.
+
+### `WithBoard`
+
+Inherit from this class (in addition to `BaseInvocation`) to add a `board` input to your node. This renders as a drop-down to select a board. The user's selection will be accessible from `self.board` in the `invoke()` function.
+
+When you call `context.images.save()`, if a board was selected, the image will added to that board as it is saved.
+
 <!-- prettier-ignore-start -->
 ::: invokeai.app.services.shared.invocation_context.InvocationContext
     options:

invokeai/app/api/dependencies.py

@@ -25,6 +25,7 @@ from ..services.invocation_cache.invocation_cache_memory import MemoryInvocation
 from ..services.invocation_services import InvocationServices
 from ..services.invocation_stats.invocation_stats_default import InvocationStatsService
 from ..services.invoker import Invoker
+from ..services.model_images.model_images_default import ModelImageFileStorageDisk
 from ..services.model_manager.model_manager_default import ModelManagerService
 from ..services.model_records import ModelRecordServiceSQL
 from ..services.names.names_default import SimpleNameService
@@ -63,14 +64,15 @@ class ApiDependencies:
     def initialize(config: InvokeAIAppConfig, event_handler_id: int, logger: Logger = logger) -> None:
         logger.info(f"InvokeAI version {__version__}")
         logger.info(f"Root directory = {str(config.root_path)}")
-        logger.debug(f"Internet connectivity is {config.internet_available}")
 
-        output_folder = config.output_path
+        output_folder = config.outputs_path
         if output_folder is None:
             raise ValueError("Output folder is not set")
 
         image_files = DiskImageFileStorage(f"{output_folder}/images")
 
+        model_images_folder = config.models_path
+
         db = init_db(config=config, logger=logger, image_files=image_files)
 
         configuration = config
@@ -92,6 +94,7 @@ class ApiDependencies:
             ObjectSerializerDisk[ConditioningFieldData](output_folder / "conditioning", ephemeral=True)
         )
         download_queue_service = DownloadQueueService(event_bus=events)
+        model_images_service = ModelImageFileStorageDisk(model_images_folder / "model_images")
         model_manager = ModelManagerService.build_model_manager(
             app_config=configuration,
             model_record_service=ModelRecordServiceSQL(db=db),
@@ -118,6 +121,7 @@ class ApiDependencies:
             images=images,
             invocation_cache=invocation_cache,
             logger=logger,
+            model_images=model_images_service,
             model_manager=model_manager,
             download_queue=download_queue_service,
             names=names,

invokeai/app/api/routers/app_info.py

@@ -12,7 +12,6 @@ from pydantic import BaseModel, Field
 
 from invokeai.app.invocations.upscale import ESRGAN_MODELS
 from invokeai.app.services.invocation_cache.invocation_cache_common import InvocationCacheStatus
-from invokeai.backend.image_util.invisible_watermark import InvisibleWatermark
 from invokeai.backend.image_util.patchmatch import PatchMatch
 from invokeai.backend.image_util.safety_checker import SafetyChecker
 from invokeai.backend.util.logging import logging
@@ -114,9 +113,7 @@ async def get_config() -> AppConfig:
     if SafetyChecker.safety_checker_available():
         nsfw_methods.append("nsfw_checker")
 
-    watermarking_methods = []
-    if InvisibleWatermark.invisible_watermark_available():
-        watermarking_methods.append("invisible_watermark")
+    watermarking_methods = ["invisible_watermark"]
 
     return AppConfig(
         infill_methods=infill_methods,

invokeai/app/api/routers/model_manager.py

@@ -1,13 +1,17 @@
 # Copyright (c) 2023 Lincoln D. Stein
 """FastAPI route for model configuration records."""
 
+import io
 import pathlib
 import shutil
+import traceback
 from typing import Any, Dict, List, Optional
 
-from fastapi import Body, Path, Query, Response
+from fastapi import Body, Path, Query, Response, UploadFile
+from fastapi.responses import FileResponse
 from fastapi.routing import APIRouter
-from pydantic import BaseModel, ConfigDict, Field
+from PIL import Image
+from pydantic import AnyHttpUrl, BaseModel, ConfigDict, Field
 from starlette.exceptions import HTTPException
 from typing_extensions import Annotated
 
@@ -25,12 +29,17 @@ from invokeai.backend.model_manager.config import (
     ModelType,
     SubModelType,
 )
+from invokeai.backend.model_manager.metadata.fetch.huggingface import HuggingFaceMetadataFetch
+from invokeai.backend.model_manager.metadata.metadata_base import ModelMetadataWithFiles, UnknownMetadataException
 from invokeai.backend.model_manager.search import ModelSearch
 
 from ..dependencies import ApiDependencies
 
 model_manager_router = APIRouter(prefix="/v2/models", tags=["model_manager"])
 
+# images are immutable; set a high max-age
+IMAGE_MAX_AGE = 31536000
+
 
 class ModelsList(BaseModel):
     """Return list of configs."""
@@ -105,6 +114,9 @@ async def list_model_records(
         found_models.extend(
             record_store.search_by_attr(model_type=model_type, model_name=model_name, model_format=model_format)
         )
+    for model in found_models:
+        cover_image = ApiDependencies.invoker.services.model_images.get_url(model.key)
+        model.cover_image = cover_image
     return ModelsList(models=found_models)
 
 
@@ -148,6 +160,8 @@ async def get_model_record(
     record_store = ApiDependencies.invoker.services.model_manager.store
     try:
         config: AnyModelConfig = record_store.get_model(key)
+        cover_image = ApiDependencies.invoker.services.model_images.get_url(key)
+        config.cover_image = cover_image
         return config
     except UnknownModelException as e:
         raise HTTPException(status_code=404, detail=str(e))
@@ -234,6 +248,40 @@ async def scan_for_models(
     return scan_results
 
 
+class HuggingFaceModels(BaseModel):
+    urls: List[AnyHttpUrl] | None = Field(description="URLs for all checkpoint format models in the metadata")
+    is_diffusers: bool = Field(description="Whether the metadata is for a Diffusers format model")
+
+
+@model_manager_router.get(
+    "/hugging_face",
+    operation_id="get_hugging_face_models",
+    responses={
+        200: {"description": "Hugging Face repo scanned successfully"},
+        400: {"description": "Invalid hugging face repo"},
+    },
+    status_code=200,
+    response_model=HuggingFaceModels,
+)
+async def get_hugging_face_models(
+    hugging_face_repo: str = Query(description="Hugging face repo to search for models", default=None),
+) -> HuggingFaceModels:
+    try:
+        metadata = HuggingFaceMetadataFetch().from_id(hugging_face_repo)
+    except UnknownMetadataException:
+        raise HTTPException(
+            status_code=400,
+            detail="No HuggingFace repository found",
+        )
+
+    assert isinstance(metadata, ModelMetadataWithFiles)
+
+    return HuggingFaceModels(
+        urls=metadata.ckpt_urls,
+        is_diffusers=metadata.is_diffusers,
+    )
+
+
 @model_manager_router.patch(
     "/i/{key}",
     operation_id="update_model_record",
@@ -266,6 +314,75 @@ async def update_model_record(
     return model_response
 
 
+@model_manager_router.get(
+    "/i/{key}/image",
+    operation_id="get_model_image",
+    responses={
+        200: {
+            "description": "The model image was fetched successfully",
+        },
+        400: {"description": "Bad request"},
+        404: {"description": "The model image could not be found"},
+    },
+    status_code=200,
+)
+async def get_model_image(
+    key: str = Path(description="The name of model image file to get"),
+) -> FileResponse:
+    """Gets an image file that previews the model"""
+
+    try:
+        path = ApiDependencies.invoker.services.model_images.get_path(key)
+
+        response = FileResponse(
+            path,
+            media_type="image/png",
+            filename=key + ".png",
+            content_disposition_type="inline",
+        )
+        response.headers["Cache-Control"] = f"max-age={IMAGE_MAX_AGE}"
+        return response
+    except Exception:
+        raise HTTPException(status_code=404)
+
+
+@model_manager_router.patch(
+    "/i/{key}/image",
+    operation_id="update_model_image",
+    responses={
+        200: {
+            "description": "The model image was updated successfully",
+        },
+        400: {"description": "Bad request"},
+    },
+    status_code=200,
+)
+async def update_model_image(
+    key: Annotated[str, Path(description="Unique key of model")],
+    image: UploadFile,
+) -> None:
+    if not image.content_type or not image.content_type.startswith("image"):
+        raise HTTPException(status_code=415, detail="Not an image")
+
+    contents = await image.read()
+    try:
+        pil_image = Image.open(io.BytesIO(contents))
+
+    except Exception:
+        ApiDependencies.invoker.services.logger.error(traceback.format_exc())
+        raise HTTPException(status_code=415, detail="Failed to read image")
+
+    logger = ApiDependencies.invoker.services.logger
+    model_images = ApiDependencies.invoker.services.model_images
+    try:
+        model_images.save(pil_image, key)
+        logger.info(f"Updated image for model: {key}")
+    except ValueError as e:
+        logger.error(str(e))
+        raise HTTPException(status_code=409, detail=str(e))
+    return
+
+
 @model_manager_router.delete(
     "/i/{key}",
     operation_id="delete_model",
@@ -296,6 +413,29 @@ async def delete_model(
         raise HTTPException(status_code=404, detail=str(e))
 
 
+@model_manager_router.delete(
+    "/i/{key}/image",
+    operation_id="delete_model_image",
+    responses={
+        204: {"description": "Model image deleted successfully"},
+        404: {"description": "Model image not found"},
+    },
+    status_code=204,
+)
+async def delete_model_image(
+    key: str = Path(description="Unique key of model image to remove from model_images directory."),
+) -> None:
+    logger = ApiDependencies.invoker.services.logger
+    model_images = ApiDependencies.invoker.services.model_images
+    try:
+        model_images.delete(key)
+        logger.info(f"Deleted model image: {key}")
+        return
+    except UnknownModelException as e:
+        logger.error(str(e))
+        raise HTTPException(status_code=404, detail=str(e))
+
+
 # @model_manager_router.post(
 #     "/i/",
 #     operation_id="add_model_record",
@@ -539,7 +679,7 @@ async def convert_model(
         raise HTTPException(400, f"The model with key {key} is not a main checkpoint model.")
 
     # loading the model will convert it into a cached diffusers file
-    model_manager.load_model_by_config(model_config, submodel_type=SubModelType.Scheduler)
+    model_manager.load.load_model(model_config, submodel_type=SubModelType.Scheduler)
 
     # Get the path of the converted model from the loader
     cache_path = loader.convert_cache.cache_path(key)

invokeai/app/api_app.py

@@ -1,71 +1,59 @@
-# parse_args() must be called before any other imports. if it is not called first, consumers of the config
-# 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
+import asyncio
+import mimetypes
+import socket
 from contextlib import asynccontextmanager
+from inspect import signature
+from pathlib import Path
+from typing import Any
 
+import uvicorn
+from fastapi import FastAPI
+from fastapi.middleware.cors import CORSMiddleware
+from fastapi.middleware.gzip import GZipMiddleware
+from fastapi.openapi.docs import get_redoc_html, get_swagger_ui_html
+from fastapi.openapi.utils import get_openapi
+from fastapi.responses import HTMLResponse
+from fastapi_events.handlers.local import local_handler
+from fastapi_events.middleware import EventHandlerASGIMiddleware
+from pydantic.json_schema import models_json_schema
+from torch.backends.mps import is_available as is_mps_available
+
+# for PyCharm:
+# noinspection PyUnresolvedReferences
+import invokeai.backend.util.hotfixes  # noqa: F401 (monkeypatching on import)
+import invokeai.frontend.web as web_dir
 from invokeai.app.api.no_cache_staticfiles import NoCacheStaticFiles
-from invokeai.version.invokeai_version import __version__
+from invokeai.app.invocations.model import ModelIdentifierField
+from invokeai.app.services.config.config_default import get_config
+from invokeai.app.services.session_processor.session_processor_common import ProgressImage
 
+from ..backend.util.logging import InvokeAILogger
+from .api.dependencies import ApiDependencies
+from .api.routers import (
+    app_info,
+    board_images,
+    boards,
+    download_queue,
+    images,
+    model_manager,
+    session_queue,
+    utilities,
+    workflows,
+)
+from .api.sockets import SocketIO
+from .invocations.baseinvocation import (
+    BaseInvocation,
+    UIConfigBase,
+)
 from .invocations.fields import InputFieldJSONSchemaExtra, OutputFieldJSONSchemaExtra
-from .services.config import InvokeAIAppConfig
 
-app_config = InvokeAIAppConfig.get_config()
-app_config.parse_args()
-if app_config.version:
-    print(f"InvokeAI version {__version__}")
-    sys.exit(0)
-
-if True:  # hack to make flake8 happy with imports coming after setting up the config
-    import asyncio
-    import mimetypes
-    import socket
-    from inspect import signature
-    from pathlib import Path
-    from typing import Any
-
-    import uvicorn
-    from fastapi import FastAPI
-    from fastapi.middleware.cors import CORSMiddleware
-    from fastapi.middleware.gzip import GZipMiddleware
-    from fastapi.openapi.docs import get_redoc_html, get_swagger_ui_html
-    from fastapi.openapi.utils import get_openapi
-    from fastapi.responses import HTMLResponse
-    from fastapi_events.handlers.local import local_handler
-    from fastapi_events.middleware import EventHandlerASGIMiddleware
-    from pydantic.json_schema import models_json_schema
-    from torch.backends.mps import is_available as is_mps_available
-
-    # for PyCharm:
-    # noinspection PyUnresolvedReferences
-    import invokeai.backend.util.hotfixes  # noqa: F401 (monkeypatching on import)
-    import invokeai.frontend.web as web_dir
-
-    from ..backend.util.logging import InvokeAILogger
-    from .api.dependencies import ApiDependencies
-    from .api.routers import (
-        app_info,
-        board_images,
-        boards,
-        download_queue,
-        images,
-        model_manager,
-        session_queue,
-        utilities,
-        workflows,
-    )
-    from .api.sockets import SocketIO
-    from .invocations.baseinvocation import (
-        BaseInvocation,
-        UIConfigBase,
-    )
-
-    if is_mps_available():
-        import invokeai.backend.util.mps_fixes  # noqa: F401 (monkeypatching on import)
+app_config = get_config()
+
+
+if is_mps_available():
+    import invokeai.backend.util.mps_fixes  # noqa: F401 (monkeypatching on import)
 
 
-app_config = InvokeAIAppConfig.get_config()
-app_config.parse_args()
 logger = InvokeAILogger.get_logger(config=app_config)
 # fix for windows mimetypes registry entries being borked
 # see https://github.com/invoke-ai/InvokeAI/discussions/3684#discussioncomment-6391352
@@ -156,17 +144,19 @@ def custom_openapi() -> dict[str, Any]:
         openapi_schema["components"]["schemas"][schema_key] = output_schema
         openapi_schema["components"]["schemas"][schema_key]["class"] = "output"
 
-    # Add Node Editor UI helper schemas
-    ui_config_schemas = models_json_schema(
+    # Some models don't end up in the schemas as standalone definitions
+    additional_schemas = models_json_schema(
         [
             (UIConfigBase, "serialization"),
             (InputFieldJSONSchemaExtra, "serialization"),
             (OutputFieldJSONSchemaExtra, "serialization"),
+            (ModelIdentifierField, "serialization"),
+            (ProgressImage, "serialization"),
         ],
         ref_template="#/components/schemas/{model}",
     )
-    for schema_key, ui_config_schema in ui_config_schemas[1]["$defs"].items():
-        openapi_schema["components"]["schemas"][schema_key] = ui_config_schema
+    for schema_key, schema_json in additional_schemas[1]["$defs"].items():
+        openapi_schema["components"]["schemas"][schema_key] = schema_json
 
     # Add a reference to the output type to additionalProperties of the invoker schema
     for invoker in all_invocations:
@@ -243,9 +233,9 @@ def invoke_api() -> None:
             else:
                 return port
 
-    from invokeai.backend.install.check_root import check_invokeai_root
+    from invokeai.backend.install.check_directories import check_directories
 
-    check_invokeai_root(app_config)  # note, may exit with an exception if root not set up
+    check_directories(app_config)  # note, may exit with an exception if root not set up
 
     if app_config.dev_reload:
         try:

invokeai/app/invocations/__init__.py

@@ -3,9 +3,9 @@ import sys
 from importlib.util import module_from_spec, spec_from_file_location
 from pathlib import Path
 
-from invokeai.app.services.config.config_default import InvokeAIAppConfig
+from invokeai.app.services.config.config_default import get_config
 
-custom_nodes_path = Path(InvokeAIAppConfig.get_config().custom_nodes_path.resolve())
+custom_nodes_path = Path(get_config().custom_nodes_path)
 custom_nodes_path.mkdir(parents=True, exist_ok=True)
 
 custom_nodes_init_path = str(custom_nodes_path / "__init__.py")

invokeai/app/invocations/baseinvocation.py

@@ -33,7 +33,7 @@ from invokeai.app.invocations.fields import (
     FieldKind,
     Input,
 )
-from invokeai.app.services.config.config_default import InvokeAIAppConfig
+from invokeai.app.services.config.config_default import get_config
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.util.metaenum import MetaEnum
 from invokeai.app.util.misc import uuid_string
@@ -191,7 +191,7 @@ class BaseInvocation(ABC, BaseModel):
     @classmethod
     def get_invocations(cls) -> Iterable[BaseInvocation]:
         """Gets all invocations, respecting the allowlist and denylist."""
-        app_config = InvokeAIAppConfig.get_config()
+        app_config = get_config()
         allowed_invocations: set[BaseInvocation] = set()
         for sc in cls._invocation_classes:
             invocation_type = sc.get_type()

invokeai/app/invocations/compel.py

@@ -20,7 +20,7 @@ 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 .model import ClipField
+from .model import CLIPField
 
 # unconditioned: Optional[torch.Tensor]
 
@@ -46,7 +46,7 @@ class CompelInvocation(BaseInvocation):
         description=FieldDescriptions.compel_prompt,
         ui_component=UIComponent.Textarea,
     )
-    clip: ClipField = InputField(
+    clip: CLIPField = InputField(
         title="CLIP",
         description=FieldDescriptions.clip,
         input=Input.Connection,
@@ -54,16 +54,16 @@ class CompelInvocation(BaseInvocation):
 
     @torch.no_grad()
     def invoke(self, context: InvocationContext) -> ConditioningOutput:
-        tokenizer_info = context.models.load(**self.clip.tokenizer.model_dump())
+        tokenizer_info = context.models.load(self.clip.tokenizer)
         tokenizer_model = tokenizer_info.model
         assert isinstance(tokenizer_model, CLIPTokenizer)
-        text_encoder_info = context.models.load(**self.clip.text_encoder.model_dump())
+        text_encoder_info = context.models.load(self.clip.text_encoder)
         text_encoder_model = text_encoder_info.model
         assert isinstance(text_encoder_model, CLIPTextModel)
 
         def _lora_loader() -> Iterator[Tuple[LoRAModelRaw, float]]:
             for lora in self.clip.loras:
-                lora_info = context.models.load(**lora.model_dump(exclude={"weight"}))
+                lora_info = context.models.load(lora.lora)
                 assert isinstance(lora_info.model, LoRAModelRaw)
                 yield (lora_info.model, lora.weight)
                 del lora_info
@@ -127,16 +127,16 @@ class SDXLPromptInvocationBase:
     def run_clip_compel(
         self,
         context: InvocationContext,
-        clip_field: ClipField,
+        clip_field: CLIPField,
         prompt: str,
         get_pooled: bool,
         lora_prefix: str,
         zero_on_empty: bool,
     ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[ExtraConditioningInfo]]:
-        tokenizer_info = context.models.load(**clip_field.tokenizer.model_dump())
+        tokenizer_info = context.models.load(clip_field.tokenizer)
         tokenizer_model = tokenizer_info.model
         assert isinstance(tokenizer_model, CLIPTokenizer)
-        text_encoder_info = context.models.load(**clip_field.text_encoder.model_dump())
+        text_encoder_info = context.models.load(clip_field.text_encoder)
         text_encoder_model = text_encoder_info.model
         assert isinstance(text_encoder_model, (CLIPTextModel, CLIPTextModelWithProjection))
 
@@ -163,7 +163,7 @@ class SDXLPromptInvocationBase:
 
         def _lora_loader() -> Iterator[Tuple[LoRAModelRaw, float]]:
             for lora in clip_field.loras:
-                lora_info = context.models.load(**lora.model_dump(exclude={"weight"}))
+                lora_info = context.models.load(lora.lora)
                 lora_model = lora_info.model
                 assert isinstance(lora_model, LoRAModelRaw)
                 yield (lora_model, lora.weight)
@@ -253,8 +253,8 @@ class SDXLCompelPromptInvocation(BaseInvocation, SDXLPromptInvocationBase):
     crop_left: int = InputField(default=0, description="")
     target_width: int = InputField(default=1024, description="")
     target_height: int = InputField(default=1024, description="")
-    clip: ClipField = InputField(description=FieldDescriptions.clip, input=Input.Connection, title="CLIP 1")
-    clip2: ClipField = InputField(description=FieldDescriptions.clip, input=Input.Connection, title="CLIP 2")
+    clip: CLIPField = InputField(description=FieldDescriptions.clip, input=Input.Connection, title="CLIP 1")
+    clip2: CLIPField = InputField(description=FieldDescriptions.clip, input=Input.Connection, title="CLIP 2")
 
     @torch.no_grad()
     def invoke(self, context: InvocationContext) -> ConditioningOutput:
@@ -340,7 +340,7 @@ class SDXLRefinerCompelPromptInvocation(BaseInvocation, SDXLPromptInvocationBase
     crop_top: int = InputField(default=0, description="")
     crop_left: int = InputField(default=0, description="")
     aesthetic_score: float = InputField(default=6.0, description=FieldDescriptions.sdxl_aesthetic)
-    clip2: ClipField = InputField(description=FieldDescriptions.clip, input=Input.Connection)
+    clip2: CLIPField = InputField(description=FieldDescriptions.clip, input=Input.Connection)
 
     @torch.no_grad()
     def invoke(self, context: InvocationContext) -> ConditioningOutput:
@@ -370,10 +370,10 @@ class SDXLRefinerCompelPromptInvocation(BaseInvocation, SDXLPromptInvocationBase
 
 
 @invocation_output("clip_skip_output")
-class ClipSkipInvocationOutput(BaseInvocationOutput):
-    """Clip skip node output"""
+class CLIPSkipInvocationOutput(BaseInvocationOutput):
+    """CLIP skip node output"""
 
-    clip: Optional[ClipField] = OutputField(default=None, description=FieldDescriptions.clip, title="CLIP")
+    clip: Optional[CLIPField] = OutputField(default=None, description=FieldDescriptions.clip, title="CLIP")
 
 
 @invocation(
@@ -383,15 +383,15 @@ class ClipSkipInvocationOutput(BaseInvocationOutput):
     category="conditioning",
     version="1.0.0",
 )
-class ClipSkipInvocation(BaseInvocation):
+class CLIPSkipInvocation(BaseInvocation):
     """Skip layers in clip text_encoder model."""
 
-    clip: ClipField = InputField(description=FieldDescriptions.clip, input=Input.Connection, title="CLIP")
+    clip: CLIPField = InputField(description=FieldDescriptions.clip, input=Input.Connection, title="CLIP")
     skipped_layers: int = InputField(default=0, ge=0, description=FieldDescriptions.skipped_layers)
 
-    def invoke(self, context: InvocationContext) -> ClipSkipInvocationOutput:
+    def invoke(self, context: InvocationContext) -> CLIPSkipInvocationOutput:
         self.clip.skipped_layers += self.skipped_layers
-        return ClipSkipInvocationOutput(
+        return CLIPSkipInvocationOutput(
             clip=self.clip,
         )
 

invokeai/app/invocations/controlnet_image_processors.py

@@ -31,9 +31,11 @@ from invokeai.app.invocations.fields import (
     Input,
     InputField,
     OutputField,
+    UIType,
     WithBoard,
     WithMetadata,
 )
+from invokeai.app.invocations.model import ModelIdentifierField
 from invokeai.app.invocations.primitives import ImageOutput
 from invokeai.app.invocations.util import validate_begin_end_step, validate_weights
 from invokeai.app.services.shared.invocation_context import InvocationContext
@@ -51,15 +53,9 @@ CONTROLNET_RESIZE_VALUES = Literal[
 ]
 
 
-class ControlNetModelField(BaseModel):
-    """ControlNet model field"""
-
-    key: str = Field(description="Model config record key for the ControlNet model")
-
-
 class ControlField(BaseModel):
     image: ImageField = Field(description="The control image")
-    control_model: ControlNetModelField = Field(description="The ControlNet model to use")
+    control_model: ModelIdentifierField = Field(description="The ControlNet model to use")
     control_weight: Union[float, List[float]] = Field(default=1, description="The weight given to the ControlNet")
     begin_step_percent: float = Field(
         default=0, ge=0, le=1, description="When the ControlNet is first applied (% of total steps)"
@@ -95,7 +91,9 @@ class ControlNetInvocation(BaseInvocation):
     """Collects ControlNet info to pass to other nodes"""
 
     image: ImageField = InputField(description="The control image")
-    control_model: ControlNetModelField = InputField(description=FieldDescriptions.controlnet_model, input=Input.Direct)
+    control_model: ModelIdentifierField = InputField(
+        description=FieldDescriptions.controlnet_model, input=Input.Direct, ui_type=UIType.ControlNetModel
+    )
     control_weight: Union[float, List[float]] = InputField(
         default=1.0, ge=-1, le=2, description="The weight given to the ControlNet"
     )
@@ -178,6 +176,7 @@ class ImageProcessorInvocation(BaseInvocation, WithMetadata, WithBoard):
 class CannyImageProcessorInvocation(ImageProcessorInvocation):
     """Canny edge detection for ControlNet"""
 
+    image_resolution: int = InputField(default=512, ge=0, description=FieldDescriptions.image_res)
     low_threshold: int = InputField(
         default=100, ge=0, le=255, description="The low threshold of the Canny pixel gradient (0-255)"
     )
@@ -191,7 +190,12 @@ class CannyImageProcessorInvocation(ImageProcessorInvocation):
 
     def run_processor(self, image):
         canny_processor = CannyDetector()
-        processed_image = canny_processor(image, self.low_threshold, self.high_threshold)
+        processed_image = canny_processor(
+            image,
+            self.low_threshold,
+            self.high_threshold,
+            image_resolution=self.image_resolution,
+        )
         return processed_image
 
 
@@ -281,6 +285,7 @@ class MidasDepthImageProcessorInvocation(ImageProcessorInvocation):
 
     a_mult: float = InputField(default=2.0, ge=0, description="Midas parameter `a_mult` (a = a_mult * PI)")
     bg_th: float = InputField(default=0.1, ge=0, description="Midas parameter `bg_th`")
+    image_resolution: int = InputField(default=512, ge=0, description=FieldDescriptions.image_res)
     # depth_and_normal not supported in controlnet_aux v0.0.3
     # depth_and_normal: bool = InputField(default=False, description="whether to use depth and normal mode")
 
@@ -290,6 +295,7 @@ class MidasDepthImageProcessorInvocation(ImageProcessorInvocation):
             image,
             a=np.pi * self.a_mult,
             bg_th=self.bg_th,
+            image_resolution=self.image_resolution,
             # dept_and_normal not supported in controlnet_aux v0.0.3
             # depth_and_normal=self.depth_and_normal,
         )
@@ -421,10 +427,13 @@ class MediapipeFaceProcessorInvocation(ImageProcessorInvocation):
 
     max_faces: int = InputField(default=1, ge=1, description="Maximum number of faces to detect")
     min_confidence: float = InputField(default=0.5, ge=0, le=1, description="Minimum confidence for face detection")
+    image_resolution: int = InputField(default=512, ge=0, description=FieldDescriptions.image_res)
 
     def run_processor(self, image):
         mediapipe_face_processor = MediapipeFaceDetector()
-        processed_image = mediapipe_face_processor(image, max_faces=self.max_faces, min_confidence=self.min_confidence)
+        processed_image = mediapipe_face_processor(
+            image, max_faces=self.max_faces, min_confidence=self.min_confidence, image_resolution=self.image_resolution
+        )
         return processed_image
 
 
@@ -507,13 +516,15 @@ class TileResamplerProcessorInvocation(ImageProcessorInvocation):
 class SegmentAnythingProcessorInvocation(ImageProcessorInvocation):
     """Applies segment anything processing to image"""
 
+    image_resolution: int = InputField(default=512, ge=0, description=FieldDescriptions.image_res)
+
     def run_processor(self, image):
         # segment_anything_processor = SamDetector.from_pretrained("ybelkada/segment-anything", subfolder="checkpoints")
         segment_anything_processor = SamDetectorReproducibleColors.from_pretrained(
             "ybelkada/segment-anything", subfolder="checkpoints"
         )
         np_img = np.array(image, dtype=np.uint8)
-        processed_image = segment_anything_processor(np_img)
+        processed_image = segment_anything_processor(np_img, image_resolution=self.image_resolution)
         return processed_image
 
 
@@ -576,7 +587,7 @@ 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 +596,12 @@ 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/fields.py

@@ -39,13 +39,15 @@ class UIType(str, Enum, metaclass=MetaEnum):
     """
 
     # region Model Field Types
+    MainModel = "MainModelField"
     SDXLMainModel = "SDXLMainModelField"
     SDXLRefinerModel = "SDXLRefinerModelField"
     ONNXModel = "ONNXModelField"
-    VaeModel = "VAEModelField"
+    VAEModel = "VAEModelField"
     LoRAModel = "LoRAModelField"
     ControlNetModel = "ControlNetModelField"
     IPAdapterModel = "IPAdapterModelField"
+    T2IAdapterModel = "T2IAdapterModelField"
     # endregion
 
     # region Misc Field Types
@@ -86,7 +88,6 @@ class UIType(str, Enum, metaclass=MetaEnum):
     IntegerPolymorphic = "DEPRECATED_IntegerPolymorphic"
     LatentsPolymorphic = "DEPRECATED_LatentsPolymorphic"
     StringPolymorphic = "DEPRECATED_StringPolymorphic"
-    MainModel = "DEPRECATED_MainModel"
     UNet = "DEPRECATED_UNet"
     Vae = "DEPRECATED_Vae"
     CLIP = "DEPRECATED_CLIP"
@@ -228,7 +229,7 @@ class ConditioningField(BaseModel):
     # endregion
 
 
-class MetadataField(RootModel):
+class MetadataField(RootModel[dict[str, Any]]):
     """
     Pydantic model for metadata with custom root of type dict[str, Any].
     Metadata is stored without a strict schema.

invokeai/app/invocations/ip_adapter.py

@@ -10,26 +10,18 @@ from invokeai.app.invocations.baseinvocation import (
     invocation,
     invocation_output,
 )
-from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField
+from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField, UIType
+from invokeai.app.invocations.model import ModelIdentifierField
 from invokeai.app.invocations.primitives import ImageField
 from invokeai.app.invocations.util import validate_begin_end_step, validate_weights
 from invokeai.app.services.shared.invocation_context import InvocationContext
-from invokeai.backend.model_manager.config import BaseModelType, ModelType
-
-
-# LS: Consider moving these two classes into model.py
-class IPAdapterModelField(BaseModel):
-    key: str = Field(description="Key to the IP-Adapter model")
-
-
-class CLIPVisionModelField(BaseModel):
-    key: str = Field(description="Key to the CLIP Vision image encoder model")
+from invokeai.backend.model_manager.config import AnyModelConfig, BaseModelType, IPAdapterConfig, ModelType
 
 
 class IPAdapterField(BaseModel):
     image: Union[ImageField, List[ImageField]] = Field(description="The IP-Adapter image prompt(s).")
-    ip_adapter_model: IPAdapterModelField = Field(description="The IP-Adapter model to use.")
-    image_encoder_model: CLIPVisionModelField = Field(description="The name of the CLIP image encoder model.")
+    ip_adapter_model: ModelIdentifierField = Field(description="The IP-Adapter model to use.")
+    image_encoder_model: ModelIdentifierField = Field(description="The name of the CLIP image encoder model.")
     weight: Union[float, List[float]] = Field(default=1, description="The weight given to the ControlNet")
     begin_step_percent: float = Field(
         default=0, ge=0, le=1, description="When the IP-Adapter is first applied (% of total steps)"
@@ -62,8 +54,12 @@ class IPAdapterInvocation(BaseInvocation):
 
     # Inputs
     image: Union[ImageField, List[ImageField]] = InputField(description="The IP-Adapter image prompt(s).")
-    ip_adapter_model: IPAdapterModelField = InputField(
-        description="The IP-Adapter model.", title="IP-Adapter Model", input=Input.Direct, ui_order=-1
+    ip_adapter_model: ModelIdentifierField = InputField(
+        description="The IP-Adapter model.",
+        title="IP-Adapter Model",
+        input=Input.Direct,
+        ui_order=-1,
+        ui_type=UIType.IPAdapterModel,
     )
 
     weight: Union[float, List[float]] = InputField(
@@ -90,20 +86,35 @@ class IPAdapterInvocation(BaseInvocation):
     def invoke(self, context: InvocationContext) -> IPAdapterOutput:
         # Lookup the CLIP Vision encoder that is intended to be used with the IP-Adapter model.
         ip_adapter_info = context.models.get_config(self.ip_adapter_model.key)
+        assert isinstance(ip_adapter_info, IPAdapterConfig)
         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(
-            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)
+        image_encoder_model = self._get_image_encoder(context, image_encoder_model_name)
         return IPAdapterOutput(
             ip_adapter=IPAdapterField(
                 image=self.image,
                 ip_adapter_model=self.ip_adapter_model,
-                image_encoder_model=image_encoder_model,
+                image_encoder_model=ModelIdentifierField.from_config(image_encoder_model),
                 weight=self.weight,
                 begin_step_percent=self.begin_step_percent,
                 end_step_percent=self.end_step_percent,
             ),
         )
+
+    def _get_image_encoder(self, context: InvocationContext, image_encoder_model_name: str) -> AnyModelConfig:
+        found = False
+        while not found:
+            image_encoder_models = context.models.search_by_attrs(
+                name=image_encoder_model_name, base=BaseModelType.Any, type=ModelType.CLIPVision
+            )
+            found = len(image_encoder_models) > 0
+            if not found:
+                context.logger.warning(
+                    f"The image encoder required by this IP Adapter ({image_encoder_model_name}) is not installed."
+                )
+                context.logger.warning("Downloading and installing now. This may take a while.")
+                installer = context._services.model_manager.install
+                job = installer.heuristic_import(f"InvokeAI/{image_encoder_model_name}")
+                installer.wait_for_job(job, timeout=600)  # wait up to 10 minutes - then raise a TimeoutException
+        assert len(image_encoder_models) == 1
+        return image_encoder_models[0]

invokeai/app/invocations/latent.py

@@ -26,6 +26,7 @@ from diffusers.schedulers import SchedulerMixin as Scheduler
 from PIL import Image, ImageFilter
 from pydantic import field_validator
 from torchvision.transforms.functional import resize as tv_resize
+from transformers import CLIPVisionModelWithProjection
 
 from invokeai.app.invocations.constants import LATENT_SCALE_FACTOR, SCHEDULER_NAME_VALUES
 from invokeai.app.invocations.fields import (
@@ -65,7 +66,6 @@ from ...backend.stable_diffusion.diffusers_pipeline import (
     T2IAdapterData,
     image_resized_to_grid_as_tensor,
 )
-from ...backend.stable_diffusion.diffusion.shared_invokeai_diffusion import PostprocessingSettings
 from ...backend.stable_diffusion.schedulers import SCHEDULER_MAP
 from ...backend.util.devices import choose_precision, choose_torch_device
 from .baseinvocation import (
@@ -75,7 +75,7 @@ from .baseinvocation import (
     invocation_output,
 )
 from .controlnet_image_processors import ControlField
-from .model import ModelInfo, UNetField, VaeField
+from .model import ModelIdentifierField, UNetField, VAEField
 
 if choose_torch_device() == torch.device("mps"):
     from torch import mps
@@ -118,7 +118,7 @@ class SchedulerInvocation(BaseInvocation):
 class CreateDenoiseMaskInvocation(BaseInvocation):
     """Creates mask for denoising model run."""
 
-    vae: VaeField = InputField(description=FieldDescriptions.vae, input=Input.Connection, ui_order=0)
+    vae: VAEField = InputField(description=FieldDescriptions.vae, input=Input.Connection, ui_order=0)
     image: Optional[ImageField] = InputField(default=None, description="Image which will be masked", ui_order=1)
     mask: ImageField = InputField(description="The mask to use when pasting", ui_order=2)
     tiled: bool = InputField(default=False, description=FieldDescriptions.tiled, ui_order=3)
@@ -153,7 +153,7 @@ class CreateDenoiseMaskInvocation(BaseInvocation):
         )
 
         if image_tensor is not None:
-            vae_info = context.models.load(**self.vae.vae.model_dump())
+            vae_info = context.models.load(self.vae.vae)
 
             img_mask = tv_resize(mask, image_tensor.shape[-2:], T.InterpolationMode.BILINEAR, antialias=False)
             masked_image = image_tensor * torch.where(img_mask < 0.5, 0.0, 1.0)
@@ -244,12 +244,12 @@ class CreateGradientMaskInvocation(BaseInvocation):
 
 def get_scheduler(
     context: InvocationContext,
-    scheduler_info: ModelInfo,
+    scheduler_info: ModelIdentifierField,
     scheduler_name: str,
     seed: int,
 ) -> Scheduler:
     scheduler_class, scheduler_extra_config = SCHEDULER_MAP.get(scheduler_name, SCHEDULER_MAP["ddim"])
-    orig_scheduler_info = context.models.load(**scheduler_info.model_dump())
+    orig_scheduler_info = context.models.load(scheduler_info)
     with orig_scheduler_info as orig_scheduler:
         scheduler_config = orig_scheduler.config
 
@@ -383,12 +383,6 @@ class DenoiseLatentsInvocation(BaseInvocation):
             text_embeddings=c,
             guidance_scale=self.cfg_scale,
             guidance_rescale_multiplier=self.cfg_rescale_multiplier,
-            postprocessing_settings=PostprocessingSettings(
-                threshold=0.0,  # threshold,
-                warmup=0.2,  # warmup,
-                h_symmetry_time_pct=None,  # h_symmetry_time_pct,
-                v_symmetry_time_pct=None,  # v_symmetry_time_pct,
-            ),
         )
 
         conditioning_data = conditioning_data.add_scheduler_args_if_applicable(  # FIXME
@@ -461,7 +455,7 @@ class DenoiseLatentsInvocation(BaseInvocation):
         #        and if weight is None, populate with default 1.0?
         controlnet_data = []
         for control_info in control_list:
-            control_model = exit_stack.enter_context(context.models.load(key=control_info.control_model.key))
+            control_model = exit_stack.enter_context(context.models.load(control_info.control_model))
 
             # control_models.append(control_model)
             control_image_field = control_info.image
@@ -523,11 +517,10 @@ class DenoiseLatentsInvocation(BaseInvocation):
         conditioning_data.ip_adapter_conditioning = []
         for single_ip_adapter in ip_adapter:
             ip_adapter_model: Union[IPAdapter, IPAdapterPlus] = exit_stack.enter_context(
-                context.models.load(key=single_ip_adapter.ip_adapter_model.key)
+                context.models.load(single_ip_adapter.ip_adapter_model)
             )
 
-            image_encoder_model_info = context.models.load(key=single_ip_adapter.image_encoder_model.key)
-
+            image_encoder_model_info = context.models.load(single_ip_adapter.image_encoder_model)
             # `single_ip_adapter.image` could be a list or a single ImageField. Normalize to a list here.
             single_ipa_image_fields = single_ip_adapter.image
             if not isinstance(single_ipa_image_fields, list):
@@ -538,6 +531,7 @@ class DenoiseLatentsInvocation(BaseInvocation):
             # TODO(ryand): With some effort, the step of running the CLIP Vision encoder could be done before any other
             # models are needed in memory. This would help to reduce peak memory utilization in low-memory environments.
             with image_encoder_model_info as image_encoder_model:
+                assert isinstance(image_encoder_model, CLIPVisionModelWithProjection)
                 # Get image embeddings from CLIP and ImageProjModel.
                 image_prompt_embeds, uncond_image_prompt_embeds = ip_adapter_model.get_image_embeds(
                     single_ipa_images, image_encoder_model
@@ -577,8 +571,8 @@ class DenoiseLatentsInvocation(BaseInvocation):
 
         t2i_adapter_data = []
         for t2i_adapter_field in t2i_adapter:
-            t2i_adapter_model_config = context.models.get_config(key=t2i_adapter_field.t2i_adapter_model.key)
-            t2i_adapter_loaded_model = context.models.load(key=t2i_adapter_field.t2i_adapter_model.key)
+            t2i_adapter_model_config = context.models.get_config(t2i_adapter_field.t2i_adapter_model.key)
+            t2i_adapter_loaded_model = context.models.load(t2i_adapter_field.t2i_adapter_model)
             image = context.images.get_pil(t2i_adapter_field.image.image_name)
 
             # The max_unet_downscale is the maximum amount that the UNet model downscales the latent image internally.
@@ -683,7 +677,7 @@ class DenoiseLatentsInvocation(BaseInvocation):
         if self.denoise_mask.masked_latents_name is not None:
             masked_latents = context.tensors.load(self.denoise_mask.masked_latents_name)
         else:
-            masked_latents = None
+            masked_latents = torch.where(mask < 0.5, 0.0, latents)
 
         return 1 - mask, masked_latents, self.denoise_mask.gradient
 
@@ -731,12 +725,13 @@ class DenoiseLatentsInvocation(BaseInvocation):
 
             def _lora_loader() -> Iterator[Tuple[LoRAModelRaw, float]]:
                 for lora in self.unet.loras:
-                    lora_info = context.models.load(**lora.model_dump(exclude={"weight"}))
+                    lora_info = context.models.load(lora.lora)
+                    assert isinstance(lora_info.model, LoRAModelRaw)
                     yield (lora_info.model, lora.weight)
                     del lora_info
                 return
 
-            unet_info = context.models.load(**self.unet.unet.model_dump())
+            unet_info = context.models.load(self.unet.unet)
             assert isinstance(unet_info.model, UNet2DConditionModel)
             with (
                 ExitStack() as exit_stack,
@@ -830,7 +825,7 @@ class LatentsToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
         description=FieldDescriptions.latents,
         input=Input.Connection,
     )
-    vae: VaeField = InputField(
+    vae: VAEField = InputField(
         description=FieldDescriptions.vae,
         input=Input.Connection,
     )
@@ -841,15 +836,15 @@ class LatentsToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
     def invoke(self, context: InvocationContext) -> ImageOutput:
         latents = context.tensors.load(self.latents.latents_name)
 
-        vae_info = context.models.load(**self.vae.vae.model_dump())
-
+        vae_info = context.models.load(self.vae.vae)
+        assert isinstance(vae_info.model, (UNet2DConditionModel, AutoencoderKL, AutoencoderTiny))
         with set_seamless(vae_info.model, self.vae.seamless_axes), vae_info as vae:
             assert isinstance(vae, torch.nn.Module)
             latents = latents.to(vae.device)
             if self.fp32:
                 vae.to(dtype=torch.float32)
 
-                use_torch_2_0_or_xformers = isinstance(
+                use_torch_2_0_or_xformers = hasattr(vae.decoder, "mid_block") and isinstance(
                     vae.decoder.mid_block.attentions[0].processor,
                     (
                         AttnProcessor2_0,
@@ -871,7 +866,7 @@ class LatentsToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
                 vae.to(dtype=torch.float16)
                 latents = latents.half()
 
-            if self.tiled or context.config.get().tiled_decode:
+            if self.tiled or context.config.get().force_tiled_decode:
                 vae.enable_tiling()
             else:
                 vae.disable_tiling()
@@ -1008,7 +1003,7 @@ class ImageToLatentsInvocation(BaseInvocation):
     image: ImageField = InputField(
         description="The image to encode",
     )
-    vae: VaeField = InputField(
+    vae: VAEField = InputField(
         description=FieldDescriptions.vae,
         input=Input.Connection,
     )
@@ -1023,7 +1018,7 @@ class ImageToLatentsInvocation(BaseInvocation):
             if upcast:
                 vae.to(dtype=torch.float32)
 
-                use_torch_2_0_or_xformers = isinstance(
+                use_torch_2_0_or_xformers = hasattr(vae.decoder, "mid_block") and isinstance(
                     vae.decoder.mid_block.attentions[0].processor,
                     (
                         AttnProcessor2_0,
@@ -1064,7 +1059,7 @@ class ImageToLatentsInvocation(BaseInvocation):
     def invoke(self, context: InvocationContext) -> LatentsOutput:
         image = context.images.get_pil(self.image.image_name)
 
-        vae_info = context.models.load(**self.vae.vae.model_dump())
+        vae_info = context.models.load(self.vae.vae)
 
         image_tensor = image_resized_to_grid_as_tensor(image.convert("RGB"))
         if image_tensor.dim() == 3:

invokeai/app/invocations/metadata.py

@@ -8,7 +8,10 @@ from invokeai.app.invocations.baseinvocation import (
     invocation,
     invocation_output,
 )
-from invokeai.app.invocations.controlnet_image_processors import ControlField
+from invokeai.app.invocations.controlnet_image_processors import (
+    CONTROLNET_MODE_VALUES,
+    CONTROLNET_RESIZE_VALUES,
+)
 from invokeai.app.invocations.fields import (
     FieldDescriptions,
     ImageField,
@@ -17,9 +20,7 @@ from invokeai.app.invocations.fields import (
     OutputField,
     UIType,
 )
-from invokeai.app.invocations.ip_adapter import IPAdapterModelField
-from invokeai.app.invocations.model import LoRAModelField, MainModelField, VAEModelField
-from invokeai.app.invocations.t2i_adapter import T2IAdapterField
+from invokeai.app.invocations.model import ModelIdentifierField
 from invokeai.app.services.shared.invocation_context import InvocationContext
 
 from ...version import __version__
@@ -33,7 +34,7 @@ class MetadataItemField(BaseModel):
 class LoRAMetadataField(BaseModel):
     """LoRA Metadata Field"""
 
-    model: LoRAModelField = Field(description=FieldDescriptions.lora_model)
+    model: ModelIdentifierField = Field(description=FieldDescriptions.lora_model)
     weight: float = Field(description=FieldDescriptions.lora_weight)
 
 
@@ -41,16 +42,41 @@ class IPAdapterMetadataField(BaseModel):
     """IP Adapter Field, minus the CLIP Vision Encoder model"""
 
     image: ImageField = Field(description="The IP-Adapter image prompt.")
-    ip_adapter_model: IPAdapterModelField = Field(
-        description="The IP-Adapter model.",
-    )
-    weight: Union[float, list[float]] = Field(
-        description="The weight given to the IP-Adapter",
-    )
+    ip_adapter_model: ModelIdentifierField = Field(description="The IP-Adapter model.")
+    weight: Union[float, list[float]] = Field(description="The weight given to the IP-Adapter")
     begin_step_percent: float = Field(description="When the IP-Adapter is first applied (% of total steps)")
     end_step_percent: float = Field(description="When the IP-Adapter is last applied (% of total steps)")
 
 
+class T2IAdapterMetadataField(BaseModel):
+    image: ImageField = Field(description="The control image.")
+    processed_image: Optional[ImageField] = Field(default=None, description="The control image, after processing.")
+    t2i_adapter_model: ModelIdentifierField = Field(description="The T2I-Adapter model to use.")
+    weight: Union[float, list[float]] = Field(default=1, description="The weight given to the T2I-Adapter")
+    begin_step_percent: float = Field(
+        default=0, ge=0, le=1, description="When the T2I-Adapter is first applied (% of total steps)"
+    )
+    end_step_percent: float = Field(
+        default=1, ge=0, le=1, description="When the T2I-Adapter is last applied (% of total steps)"
+    )
+    resize_mode: CONTROLNET_RESIZE_VALUES = Field(default="just_resize", description="The resize mode to use")
+
+
+class ControlNetMetadataField(BaseModel):
+    image: ImageField = Field(description="The control image")
+    processed_image: Optional[ImageField] = Field(default=None, description="The control image, after processing.")
+    control_model: ModelIdentifierField = Field(description="The ControlNet model to use")
+    control_weight: Union[float, list[float]] = Field(default=1, description="The weight given to the ControlNet")
+    begin_step_percent: float = Field(
+        default=0, ge=0, le=1, description="When the ControlNet is first applied (% of total steps)"
+    )
+    end_step_percent: float = Field(
+        default=1, ge=0, le=1, description="When the ControlNet is last applied (% of total steps)"
+    )
+    control_mode: CONTROLNET_MODE_VALUES = Field(default="balanced", description="The control mode to use")
+    resize_mode: CONTROLNET_RESIZE_VALUES = Field(default="just_resize", description="The resize mode to use")
+
+
 @invocation_output("metadata_item_output")
 class MetadataItemOutput(BaseInvocationOutput):
     """Metadata Item Output"""
@@ -140,14 +166,14 @@ class CoreMetadataInvocation(BaseInvocation):
         default=None,
         description="The number of skipped CLIP layers",
     )
-    model: Optional[MainModelField] = InputField(default=None, description="The main model used for inference")
-    controlnets: Optional[list[ControlField]] = InputField(
+    model: Optional[ModelIdentifierField] = InputField(default=None, description="The main model used for inference")
+    controlnets: Optional[list[ControlNetMetadataField]] = InputField(
         default=None, description="The ControlNets used for inference"
     )
     ipAdapters: Optional[list[IPAdapterMetadataField]] = InputField(
         default=None, description="The IP Adapters used for inference"
     )
-    t2iAdapters: Optional[list[T2IAdapterField]] = InputField(
+    t2iAdapters: Optional[list[T2IAdapterMetadataField]] = InputField(
         default=None, description="The IP Adapters used for inference"
     )
     loras: Optional[list[LoRAMetadataField]] = InputField(default=None, description="The LoRAs used for inference")
@@ -159,7 +185,7 @@ class CoreMetadataInvocation(BaseInvocation):
         default=None,
         description="The name of the initial image",
     )
-    vae: Optional[VAEModelField] = InputField(
+    vae: Optional[ModelIdentifierField] = InputField(
         default=None,
         description="The VAE used for decoding, if the main model's default was not used",
     )
@@ -190,7 +216,7 @@ class CoreMetadataInvocation(BaseInvocation):
     )
 
     # SDXL Refiner
-    refiner_model: Optional[MainModelField] = InputField(
+    refiner_model: Optional[ModelIdentifierField] = InputField(
         default=None,
         description="The SDXL Refiner model used",
     )
@@ -222,10 +248,9 @@ class CoreMetadataInvocation(BaseInvocation):
     def invoke(self, context: InvocationContext) -> MetadataOutput:
         """Collects and outputs a CoreMetadata object"""
 
-        return MetadataOutput(
-            metadata=MetadataField.model_validate(
-                self.model_dump(exclude_none=True, exclude={"id", "type", "is_intermediate", "use_cache"})
-            )
-        )
+        as_dict = self.model_dump(exclude_none=True, exclude={"id", "type", "is_intermediate", "use_cache"})
+        as_dict["app_version"] = __version__
+
+        return MetadataOutput(metadata=MetadataField.model_validate(as_dict))
 
     model_config = ConfigDict(extra="allow")

invokeai/app/invocations/model.py

@@ -3,11 +3,11 @@ from typing import List, Optional
 
 from pydantic import BaseModel, Field
 
-from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField
+from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField, UIType
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.shared.models import FreeUConfig
+from invokeai.backend.model_manager.config import AnyModelConfig, BaseModelType, ModelType, SubModelType
 
-from ...backend.model_manager import SubModelType
 from .baseinvocation import (
     BaseInvocation,
     BaseInvocationOutput,
@@ -16,33 +16,52 @@ from .baseinvocation import (
 )
 
 
-class ModelInfo(BaseModel):
-    key: str = Field(description="Key of model as returned by ModelRecordServiceBase.get_model()")
-    submodel_type: Optional[SubModelType] = Field(default=None, description="Info to load submodel")
+class ModelIdentifierField(BaseModel):
+    key: str = Field(description="The model's unique key")
+    hash: str = Field(description="The model's BLAKE3 hash")
+    name: str = Field(description="The model's name")
+    base: BaseModelType = Field(description="The model's base model type")
+    type: ModelType = Field(description="The model's type")
+    submodel_type: Optional[SubModelType] = Field(
+        description="The submodel to load, if this is a main model", default=None
+    )
+
+    @classmethod
+    def from_config(
+        cls, config: "AnyModelConfig", submodel_type: Optional[SubModelType] = None
+    ) -> "ModelIdentifierField":
+        return cls(
+            key=config.key,
+            hash=config.hash,
+            name=config.name,
+            base=config.base,
+            type=config.type,
+            submodel_type=submodel_type,
+        )
 
 
-class LoraInfo(ModelInfo):
-    weight: float = Field(description="Lora's weight which to use when apply to model")
+class LoRAField(BaseModel):
+    lora: ModelIdentifierField = Field(description="Info to load lora model")
+    weight: float = Field(description="Weight to apply to lora model")
 
 
 class UNetField(BaseModel):
-    unet: ModelInfo = Field(description="Info to load unet submodel")
-    scheduler: ModelInfo = Field(description="Info to load scheduler submodel")
-    loras: List[LoraInfo] = Field(description="Loras to apply on model loading")
+    unet: ModelIdentifierField = Field(description="Info to load unet submodel")
+    scheduler: ModelIdentifierField = Field(description="Info to load scheduler submodel")
+    loras: List[LoRAField] = Field(description="LoRAs to apply on model loading")
     seamless_axes: List[str] = Field(default_factory=list, description='Axes("x" and "y") to which apply seamless')
     freeu_config: Optional[FreeUConfig] = Field(default=None, description="FreeU configuration")
 
 
-class ClipField(BaseModel):
-    tokenizer: ModelInfo = Field(description="Info to load tokenizer submodel")
-    text_encoder: ModelInfo = Field(description="Info to load text_encoder submodel")
+class CLIPField(BaseModel):
+    tokenizer: ModelIdentifierField = Field(description="Info to load tokenizer submodel")
+    text_encoder: ModelIdentifierField = Field(description="Info to load text_encoder submodel")
     skipped_layers: int = Field(description="Number of skipped layers in text_encoder")
-    loras: List[LoraInfo] = Field(description="Loras to apply on model loading")
+    loras: List[LoRAField] = Field(description="LoRAs to apply on model loading")
 
 
-class VaeField(BaseModel):
-    # TODO: better naming?
-    vae: ModelInfo = Field(description="Info to load vae submodel")
+class VAEField(BaseModel):
+    vae: ModelIdentifierField = Field(description="Info to load vae submodel")
     seamless_axes: List[str] = Field(default_factory=list, description='Axes("x" and "y") to which apply seamless')
 
 
@@ -57,14 +76,14 @@ class UNetOutput(BaseInvocationOutput):
 class VAEOutput(BaseInvocationOutput):
     """Base class for invocations that output a VAE field"""
 
-    vae: VaeField = OutputField(description=FieldDescriptions.vae, title="VAE")
+    vae: VAEField = OutputField(description=FieldDescriptions.vae, title="VAE")
 
 
 @invocation_output("clip_output")
 class CLIPOutput(BaseInvocationOutput):
     """Base class for invocations that output a CLIP field"""
 
-    clip: ClipField = OutputField(description=FieldDescriptions.clip, title="CLIP")
+    clip: CLIPField = OutputField(description=FieldDescriptions.clip, title="CLIP")
 
 
 @invocation_output("model_loader_output")
@@ -74,18 +93,6 @@ class ModelLoaderOutput(UNetOutput, CLIPOutput, VAEOutput):
     pass
 
 
-class MainModelField(BaseModel):
-    """Main model field"""
-
-    key: str = Field(description="Model key")
-
-
-class LoRAModelField(BaseModel):
-    """LoRA model field"""
-
-    key: str = Field(description="LoRA model key")
-
-
 @invocation(
     "main_model_loader",
     title="Main Model",
@@ -96,62 +103,44 @@ class LoRAModelField(BaseModel):
 class MainModelLoaderInvocation(BaseInvocation):
     """Loads a main model, outputting its submodels."""
 
-    model: MainModelField = InputField(description=FieldDescriptions.main_model, input=Input.Direct)
+    model: ModelIdentifierField = InputField(
+        description=FieldDescriptions.main_model, input=Input.Direct, ui_type=UIType.MainModel
+    )
     # TODO: precision?
 
     def invoke(self, context: InvocationContext) -> ModelLoaderOutput:
-        key = self.model.key
-
         # TODO: not found exceptions
-        if not context.models.exists(key):
-            raise Exception(f"Unknown model {key}")
+        if not context.models.exists(self.model.key):
+            raise Exception(f"Unknown model {self.model.key}")
+
+        unet = self.model.model_copy(update={"submodel_type": SubModelType.UNet})
+        scheduler = self.model.model_copy(update={"submodel_type": SubModelType.Scheduler})
+        tokenizer = self.model.model_copy(update={"submodel_type": SubModelType.Tokenizer})
+        text_encoder = self.model.model_copy(update={"submodel_type": SubModelType.TextEncoder})
+        vae = self.model.model_copy(update={"submodel_type": SubModelType.VAE})
 
         return ModelLoaderOutput(
-            unet=UNetField(
-                unet=ModelInfo(
-                    key=key,
-                    submodel_type=SubModelType.UNet,
-                ),
-                scheduler=ModelInfo(
-                    key=key,
-                    submodel_type=SubModelType.Scheduler,
-                ),
-                loras=[],
-            ),
-            clip=ClipField(
-                tokenizer=ModelInfo(
-                    key=key,
-                    submodel_type=SubModelType.Tokenizer,
-                ),
-                text_encoder=ModelInfo(
-                    key=key,
-                    submodel_type=SubModelType.TextEncoder,
-                ),
-                loras=[],
-                skipped_layers=0,
-            ),
-            vae=VaeField(
-                vae=ModelInfo(
-                    key=key,
-                    submodel_type=SubModelType.VAE,
-                ),
-            ),
+            unet=UNetField(unet=unet, scheduler=scheduler, loras=[]),
+            clip=CLIPField(tokenizer=tokenizer, text_encoder=text_encoder, loras=[], skipped_layers=0),
+            vae=VAEField(vae=vae),
         )
 
 
 @invocation_output("lora_loader_output")
-class LoraLoaderOutput(BaseInvocationOutput):
+class LoRALoaderOutput(BaseInvocationOutput):
     """Model loader output"""
 
     unet: Optional[UNetField] = OutputField(default=None, description=FieldDescriptions.unet, title="UNet")
-    clip: Optional[ClipField] = OutputField(default=None, description=FieldDescriptions.clip, title="CLIP")
+    clip: Optional[CLIPField] = OutputField(default=None, description=FieldDescriptions.clip, title="CLIP")
 
 
 @invocation("lora_loader", title="LoRA", tags=["model"], category="model", version="1.0.1")
-class LoraLoaderInvocation(BaseInvocation):
+class LoRALoaderInvocation(BaseInvocation):
     """Apply selected lora to unet and text_encoder."""
 
-    lora: LoRAModelField = InputField(description=FieldDescriptions.lora_model, input=Input.Direct, title="LoRA")
+    lora: ModelIdentifierField = InputField(
+        description=FieldDescriptions.lora_model, input=Input.Direct, title="LoRA", ui_type=UIType.LoRAModel
+    )
     weight: float = InputField(default=0.75, description=FieldDescriptions.lora_weight)
     unet: Optional[UNetField] = InputField(
         default=None,
@@ -159,46 +148,41 @@ class LoraLoaderInvocation(BaseInvocation):
         input=Input.Connection,
         title="UNet",
     )
-    clip: Optional[ClipField] = InputField(
+    clip: Optional[CLIPField] = InputField(
         default=None,
         description=FieldDescriptions.clip,
         input=Input.Connection,
         title="CLIP",
     )
 
-    def invoke(self, context: InvocationContext) -> LoraLoaderOutput:
-        if self.lora is None:
-            raise Exception("No LoRA provided")
-
+    def invoke(self, context: InvocationContext) -> LoRALoaderOutput:
         lora_key = self.lora.key
 
         if not context.models.exists(lora_key):
             raise Exception(f"Unkown lora: {lora_key}!")
 
-        if self.unet is not None and any(lora.key == lora_key for lora in self.unet.loras):
-            raise Exception(f'Lora "{lora_key}" already applied to unet')
+        if self.unet is not None and any(lora.lora.key == lora_key for lora in self.unet.loras):
+            raise Exception(f'LoRA "{lora_key}" already applied to unet')
 
-        if self.clip is not None and any(lora.key == lora_key for lora in self.clip.loras):
-            raise Exception(f'Lora "{lora_key}" already applied to clip')
+        if self.clip is not None and any(lora.lora.key == lora_key for lora in self.clip.loras):
+            raise Exception(f'LoRA "{lora_key}" already applied to clip')
 
-        output = LoraLoaderOutput()
+        output = LoRALoaderOutput()
 
         if self.unet is not None:
-            output.unet = copy.deepcopy(self.unet)
+            output.unet = self.unet.model_copy(deep=True)
             output.unet.loras.append(
-                LoraInfo(
-                    key=lora_key,
-                    submodel_type=None,
+                LoRAField(
+                    lora=self.lora,
                     weight=self.weight,
                 )
             )
 
         if self.clip is not None:
-            output.clip = copy.deepcopy(self.clip)
+            output.clip = self.clip.model_copy(deep=True)
             output.clip.loras.append(
-                LoraInfo(
-                    key=lora_key,
-                    submodel_type=None,
+                LoRAField(
+                    lora=self.lora,
                     weight=self.weight,
                 )
             )
@@ -207,12 +191,12 @@ class LoraLoaderInvocation(BaseInvocation):
 
 
 @invocation_output("sdxl_lora_loader_output")
-class SDXLLoraLoaderOutput(BaseInvocationOutput):
+class SDXLLoRALoaderOutput(BaseInvocationOutput):
     """SDXL LoRA Loader Output"""
 
     unet: Optional[UNetField] = OutputField(default=None, description=FieldDescriptions.unet, title="UNet")
-    clip: Optional[ClipField] = OutputField(default=None, description=FieldDescriptions.clip, title="CLIP 1")
-    clip2: Optional[ClipField] = OutputField(default=None, description=FieldDescriptions.clip, title="CLIP 2")
+    clip: Optional[CLIPField] = OutputField(default=None, description=FieldDescriptions.clip, title="CLIP 1")
+    clip2: Optional[CLIPField] = OutputField(default=None, description=FieldDescriptions.clip, title="CLIP 2")
 
 
 @invocation(
@@ -222,10 +206,12 @@ class SDXLLoraLoaderOutput(BaseInvocationOutput):
     category="model",
     version="1.0.1",
 )
-class SDXLLoraLoaderInvocation(BaseInvocation):
+class SDXLLoRALoaderInvocation(BaseInvocation):
     """Apply selected lora to unet and text_encoder."""
 
-    lora: LoRAModelField = InputField(description=FieldDescriptions.lora_model, input=Input.Direct, title="LoRA")
+    lora: ModelIdentifierField = InputField(
+        description=FieldDescriptions.lora_model, input=Input.Direct, title="LoRA", ui_type=UIType.LoRAModel
+    )
     weight: float = InputField(default=0.75, description=FieldDescriptions.lora_weight)
     unet: Optional[UNetField] = InputField(
         default=None,
@@ -233,65 +219,59 @@ class SDXLLoraLoaderInvocation(BaseInvocation):
         input=Input.Connection,
         title="UNet",
     )
-    clip: Optional[ClipField] = InputField(
+    clip: Optional[CLIPField] = InputField(
         default=None,
         description=FieldDescriptions.clip,
         input=Input.Connection,
         title="CLIP 1",
     )
-    clip2: Optional[ClipField] = InputField(
+    clip2: Optional[CLIPField] = InputField(
         default=None,
         description=FieldDescriptions.clip,
         input=Input.Connection,
         title="CLIP 2",
     )
 
-    def invoke(self, context: InvocationContext) -> SDXLLoraLoaderOutput:
-        if self.lora is None:
-            raise Exception("No LoRA provided")
-
+    def invoke(self, context: InvocationContext) -> SDXLLoRALoaderOutput:
         lora_key = self.lora.key
 
         if not context.models.exists(lora_key):
             raise Exception(f"Unknown lora: {lora_key}!")
 
-        if self.unet is not None and any(lora.key == lora_key for lora in self.unet.loras):
-            raise Exception(f'Lora "{lora_key}" already applied to unet')
+        if self.unet is not None and any(lora.lora.key == lora_key for lora in self.unet.loras):
+            raise Exception(f'LoRA "{lora_key}" already applied to unet')
 
-        if self.clip is not None and any(lora.key == lora_key for lora in self.clip.loras):
-            raise Exception(f'Lora "{lora_key}" already applied to clip')
+        if self.clip is not None and any(lora.lora.key == lora_key for lora in self.clip.loras):
+            raise Exception(f'LoRA "{lora_key}" already applied to clip')
 
-        if self.clip2 is not None and any(lora.key == lora_key for lora in self.clip2.loras):
-            raise Exception(f'Lora "{lora_key}" already applied to clip2')
+        if self.clip2 is not None and any(lora.lora.key == lora_key for lora in self.clip2.loras):
+            raise Exception(f'LoRA "{lora_key}" already applied to clip2')
 
-        output = SDXLLoraLoaderOutput()
+        output = SDXLLoRALoaderOutput()
 
         if self.unet is not None:
-            output.unet = copy.deepcopy(self.unet)
+            output.unet = self.unet.model_copy(deep=True)
             output.unet.loras.append(
-                LoraInfo(
-                    key=lora_key,
-                    submodel_type=None,
+                LoRAField(
+                    lora=self.lora,
                     weight=self.weight,
                 )
             )
 
         if self.clip is not None:
-            output.clip = copy.deepcopy(self.clip)
+            output.clip = self.clip.model_copy(deep=True)
             output.clip.loras.append(
-                LoraInfo(
-                    key=lora_key,
-                    submodel_type=None,
+                LoRAField(
+                    lora=self.lora,
                     weight=self.weight,
                 )
             )
 
         if self.clip2 is not None:
-            output.clip2 = copy.deepcopy(self.clip2)
+            output.clip2 = self.clip2.model_copy(deep=True)
             output.clip2.loras.append(
-                LoraInfo(
-                    key=lora_key,
-                    submodel_type=None,
+                LoRAField(
+                    lora=self.lora,
                     weight=self.weight,
                 )
             )
@@ -299,20 +279,12 @@ class SDXLLoraLoaderInvocation(BaseInvocation):
         return output
 
 
-class VAEModelField(BaseModel):
-    """Vae model field"""
-
-    key: str = Field(description="Model's key")
-
-
 @invocation("vae_loader", title="VAE", tags=["vae", "model"], category="model", version="1.0.1")
-class VaeLoaderInvocation(BaseInvocation):
+class VAELoaderInvocation(BaseInvocation):
     """Loads a VAE model, outputting a VaeLoaderOutput"""
 
-    vae_model: VAEModelField = InputField(
-        description=FieldDescriptions.vae_model,
-        input=Input.Direct,
-        title="VAE",
+    vae_model: ModelIdentifierField = InputField(
+        description=FieldDescriptions.vae_model, input=Input.Direct, title="VAE", ui_type=UIType.VAEModel
     )
 
     def invoke(self, context: InvocationContext) -> VAEOutput:
@@ -321,7 +293,7 @@ class VaeLoaderInvocation(BaseInvocation):
         if not context.models.exists(key):
             raise Exception(f"Unkown vae: {key}!")
 
-        return VAEOutput(vae=VaeField(vae=ModelInfo(key=key)))
+        return VAEOutput(vae=VAEField(vae=self.vae_model))
 
 
 @invocation_output("seamless_output")
@@ -329,7 +301,7 @@ class SeamlessModeOutput(BaseInvocationOutput):
     """Modified Seamless Model output"""
 
     unet: Optional[UNetField] = OutputField(default=None, description=FieldDescriptions.unet, title="UNet")
-    vae: Optional[VaeField] = OutputField(default=None, description=FieldDescriptions.vae, title="VAE")
+    vae: Optional[VAEField] = OutputField(default=None, description=FieldDescriptions.vae, title="VAE")
 
 
 @invocation(
@@ -348,7 +320,7 @@ class SeamlessModeInvocation(BaseInvocation):
         input=Input.Connection,
         title="UNet",
     )
-    vae: Optional[VaeField] = InputField(
+    vae: Optional[VAEField] = InputField(
         default=None,
         description=FieldDescriptions.vae_model,
         input=Input.Connection,

invokeai/app/invocations/sdxl.py

@@ -8,7 +8,7 @@ from .baseinvocation import (
     invocation,
     invocation_output,
 )
-from .model import ClipField, MainModelField, ModelInfo, UNetField, VaeField
+from .model import CLIPField, ModelIdentifierField, UNetField, VAEField
 
 
 @invocation_output("sdxl_model_loader_output")
@@ -16,9 +16,9 @@ class SDXLModelLoaderOutput(BaseInvocationOutput):
     """SDXL base model loader output"""
 
     unet: UNetField = OutputField(description=FieldDescriptions.unet, title="UNet")
-    clip: ClipField = OutputField(description=FieldDescriptions.clip, title="CLIP 1")
-    clip2: ClipField = OutputField(description=FieldDescriptions.clip, title="CLIP 2")
-    vae: VaeField = OutputField(description=FieldDescriptions.vae, title="VAE")
+    clip: CLIPField = OutputField(description=FieldDescriptions.clip, title="CLIP 1")
+    clip2: CLIPField = OutputField(description=FieldDescriptions.clip, title="CLIP 2")
+    vae: VAEField = OutputField(description=FieldDescriptions.vae, title="VAE")
 
 
 @invocation_output("sdxl_refiner_model_loader_output")
@@ -26,15 +26,15 @@ class SDXLRefinerModelLoaderOutput(BaseInvocationOutput):
     """SDXL refiner model loader output"""
 
     unet: UNetField = OutputField(description=FieldDescriptions.unet, title="UNet")
-    clip2: ClipField = OutputField(description=FieldDescriptions.clip, title="CLIP 2")
-    vae: VaeField = OutputField(description=FieldDescriptions.vae, title="VAE")
+    clip2: CLIPField = OutputField(description=FieldDescriptions.clip, title="CLIP 2")
+    vae: VAEField = OutputField(description=FieldDescriptions.vae, title="VAE")
 
 
 @invocation("sdxl_model_loader", title="SDXL Main Model", tags=["model", "sdxl"], category="model", version="1.0.1")
 class SDXLModelLoaderInvocation(BaseInvocation):
     """Loads an sdxl base model, outputting its submodels."""
 
-    model: MainModelField = InputField(
+    model: ModelIdentifierField = InputField(
         description=FieldDescriptions.sdxl_main_model, input=Input.Direct, ui_type=UIType.SDXLMainModel
     )
     # TODO: precision?
@@ -46,48 +46,19 @@ class SDXLModelLoaderInvocation(BaseInvocation):
         if not context.models.exists(model_key):
             raise Exception(f"Unknown model: {model_key}")
 
+        unet = self.model.model_copy(update={"submodel_type": SubModelType.UNet})
+        scheduler = self.model.model_copy(update={"submodel_type": SubModelType.Scheduler})
+        tokenizer = self.model.model_copy(update={"submodel_type": SubModelType.Tokenizer})
+        text_encoder = self.model.model_copy(update={"submodel_type": SubModelType.TextEncoder})
+        tokenizer2 = self.model.model_copy(update={"submodel_type": SubModelType.Tokenizer2})
+        text_encoder2 = self.model.model_copy(update={"submodel_type": SubModelType.TextEncoder2})
+        vae = self.model.model_copy(update={"submodel_type": SubModelType.VAE})
+
         return SDXLModelLoaderOutput(
-            unet=UNetField(
-                unet=ModelInfo(
-                    key=model_key,
-                    submodel_type=SubModelType.UNet,
-                ),
-                scheduler=ModelInfo(
-                    key=model_key,
-                    submodel_type=SubModelType.Scheduler,
-                ),
-                loras=[],
-            ),
-            clip=ClipField(
-                tokenizer=ModelInfo(
-                    key=model_key,
-                    submodel_type=SubModelType.Tokenizer,
-                ),
-                text_encoder=ModelInfo(
-                    key=model_key,
-                    submodel_type=SubModelType.TextEncoder,
-                ),
-                loras=[],
-                skipped_layers=0,
-            ),
-            clip2=ClipField(
-                tokenizer=ModelInfo(
-                    key=model_key,
-                    submodel_type=SubModelType.Tokenizer2,
-                ),
-                text_encoder=ModelInfo(
-                    key=model_key,
-                    submodel_type=SubModelType.TextEncoder2,
-                ),
-                loras=[],
-                skipped_layers=0,
-            ),
-            vae=VaeField(
-                vae=ModelInfo(
-                    key=model_key,
-                    submodel_type=SubModelType.VAE,
-                ),
-            ),
+            unet=UNetField(unet=unet, scheduler=scheduler, loras=[]),
+            clip=CLIPField(tokenizer=tokenizer, text_encoder=text_encoder, loras=[], skipped_layers=0),
+            clip2=CLIPField(tokenizer=tokenizer2, text_encoder=text_encoder2, loras=[], skipped_layers=0),
+            vae=VAEField(vae=vae),
         )
 
 
@@ -101,10 +72,8 @@ class SDXLModelLoaderInvocation(BaseInvocation):
 class SDXLRefinerModelLoaderInvocation(BaseInvocation):
     """Loads an sdxl refiner model, outputting its submodels."""
 
-    model: MainModelField = InputField(
-        description=FieldDescriptions.sdxl_refiner_model,
-        input=Input.Direct,
-        ui_type=UIType.SDXLRefinerModel,
+    model: ModelIdentifierField = InputField(
+        description=FieldDescriptions.sdxl_refiner_model, input=Input.Direct, ui_type=UIType.SDXLRefinerModel
     )
     # TODO: precision?
 
@@ -115,34 +84,14 @@ class SDXLRefinerModelLoaderInvocation(BaseInvocation):
         if not context.models.exists(model_key):
             raise Exception(f"Unknown model: {model_key}")
 
+        unet = self.model.model_copy(update={"submodel_type": SubModelType.UNet})
+        scheduler = self.model.model_copy(update={"submodel_type": SubModelType.Scheduler})
+        tokenizer2 = self.model.model_copy(update={"submodel_type": SubModelType.Tokenizer2})
+        text_encoder2 = self.model.model_copy(update={"submodel_type": SubModelType.TextEncoder2})
+        vae = self.model.model_copy(update={"submodel_type": SubModelType.VAE})
+
         return SDXLRefinerModelLoaderOutput(
-            unet=UNetField(
-                unet=ModelInfo(
-                    key=model_key,
-                    submodel_type=SubModelType.UNet,
-                ),
-                scheduler=ModelInfo(
-                    key=model_key,
-                    submodel_type=SubModelType.Scheduler,
-                ),
-                loras=[],
-            ),
-            clip2=ClipField(
-                tokenizer=ModelInfo(
-                    key=model_key,
-                    submodel_type=SubModelType.Tokenizer2,
-                ),
-                text_encoder=ModelInfo(
-                    key=model_key,
-                    submodel_type=SubModelType.TextEncoder2,
-                ),
-                loras=[],
-                skipped_layers=0,
-            ),
-            vae=VaeField(
-                vae=ModelInfo(
-                    key=model_key,
-                    submodel_type=SubModelType.VAE,
-                ),
-            ),
+            unet=UNetField(unet=unet, scheduler=scheduler, loras=[]),
+            clip2=CLIPField(tokenizer=tokenizer2, text_encoder=text_encoder2, loras=[], skipped_layers=0),
+            vae=VAEField(vae=vae),
         )

invokeai/app/invocations/t2i_adapter.py

@@ -9,18 +9,15 @@ from invokeai.app.invocations.baseinvocation import (
     invocation_output,
 )
 from invokeai.app.invocations.controlnet_image_processors import CONTROLNET_RESIZE_VALUES
-from invokeai.app.invocations.fields import FieldDescriptions, ImageField, Input, InputField, OutputField
+from invokeai.app.invocations.fields import FieldDescriptions, ImageField, Input, InputField, OutputField, UIType
+from invokeai.app.invocations.model import ModelIdentifierField
 from invokeai.app.invocations.util import validate_begin_end_step, validate_weights
 from invokeai.app.services.shared.invocation_context import InvocationContext
 
 
-class T2IAdapterModelField(BaseModel):
-    key: str = Field(description="Model record key for the T2I-Adapter model")
-
-
 class T2IAdapterField(BaseModel):
     image: ImageField = Field(description="The T2I-Adapter image prompt.")
-    t2i_adapter_model: T2IAdapterModelField = Field(description="The T2I-Adapter model to use.")
+    t2i_adapter_model: ModelIdentifierField = Field(description="The T2I-Adapter model to use.")
     weight: Union[float, list[float]] = Field(default=1, description="The weight given to the T2I-Adapter")
     begin_step_percent: float = Field(
         default=0, ge=0, le=1, description="When the T2I-Adapter is first applied (% of total steps)"
@@ -55,11 +52,12 @@ class T2IAdapterInvocation(BaseInvocation):
 
     # Inputs
     image: ImageField = InputField(description="The IP-Adapter image prompt.")
-    t2i_adapter_model: T2IAdapterModelField = InputField(
+    t2i_adapter_model: ModelIdentifierField = InputField(
         description="The T2I-Adapter model.",
         title="T2I-Adapter Model",
         input=Input.Direct,
         ui_order=-1,
+        ui_type=UIType.T2IAdapterModel,
     )
     weight: Union[float, list[float]] = InputField(
         default=1, ge=0, description="The weight given to the T2I-Adapter", title="Weight"

invokeai/app/run_app.py

@@ -0,0 +1,12 @@
+"""This is a wrapper around the main app entrypoint, to allow for CLI args to be parsed before running the app."""
+
+
+def run_app() -> None:
+    # Before doing _anything_, parse CLI args!
+    from invokeai.frontend.cli.arg_parser import InvokeAIArgs
+
+    InvokeAIArgs.parse_args()
+
+    from invokeai.app.api_app import invoke_api
+
+    invoke_api()

invokeai/app/services/config/__init__.py

@@ -2,6 +2,6 @@
 
 from invokeai.app.services.config.config_common import PagingArgumentParser
 
-from .config_default import InvokeAIAppConfig, get_invokeai_config
+from .config_default import InvokeAIAppConfig, get_config
 
-__all__ = ["InvokeAIAppConfig", "get_invokeai_config", "PagingArgumentParser"]
+__all__ = ["InvokeAIAppConfig", "get_config", "PagingArgumentParser"]

invokeai/app/services/config/config_base.py

@@ -1,224 +0,0 @@
-# Copyright (c) 2023 Lincoln Stein (https://github.com/lstein) and the InvokeAI Development Team
-
-"""
-Base class for the InvokeAI configuration system.
-It defines a type of pydantic BaseSettings object that
-is able to read and write from an omegaconf-based config file,
-with overriding of settings from environment variables and/or
-the command line.
-"""
-
-from __future__ import annotations
-
-import argparse
-import os
-import sys
-from argparse import ArgumentParser
-from pathlib import Path
-from typing import Any, ClassVar, Dict, List, Literal, Optional, Union, get_args, get_origin, get_type_hints
-
-from omegaconf import DictConfig, ListConfig, OmegaConf
-from pydantic_settings import BaseSettings, SettingsConfigDict
-
-from invokeai.app.services.config.config_common import PagingArgumentParser, int_or_float_or_str
-
-
-class InvokeAISettings(BaseSettings):
-    """Runtime configuration settings in which default values are read from an omegaconf .yaml file."""
-
-    initconf: ClassVar[Optional[DictConfig]] = None
-    argparse_groups: ClassVar[Dict[str, Any]] = {}
-
-    model_config = SettingsConfigDict(env_file_encoding="utf-8", arbitrary_types_allowed=True, case_sensitive=True)
-
-    def parse_args(self, argv: Optional[List[str]] = sys.argv[1:]) -> None:
-        """Call to parse command-line arguments."""
-        parser = self.get_parser()
-        opt, unknown_opts = parser.parse_known_args(argv)
-        if len(unknown_opts) > 0:
-            print("Unknown args:", unknown_opts)
-        for name in self.model_fields:
-            if name not in self._excluded():
-                value = getattr(opt, name)
-                if isinstance(value, ListConfig):
-                    value = list(value)
-                elif isinstance(value, DictConfig):
-                    value = dict(value)
-                setattr(self, name, value)
-
-    def to_yaml(self) -> str:
-        """Return a YAML string representing our settings. This can be used as the contents of `invokeai.yaml` to restore settings later."""
-        cls = self.__class__
-        type = get_args(get_type_hints(cls)["type"])[0]
-        field_dict: Dict[str, Dict[str, Any]] = {type: {}}
-        for name, field in self.model_fields.items():
-            if name in cls._excluded_from_yaml():
-                continue
-            assert isinstance(field.json_schema_extra, dict)
-            category = (
-                field.json_schema_extra.get("category", "Uncategorized") if field.json_schema_extra else "Uncategorized"
-            )
-            value = getattr(self, name)
-            assert isinstance(category, str)
-            if category not in field_dict[type]:
-                field_dict[type][category] = {}
-            # keep paths as strings to make it easier to read
-            field_dict[type][category][name] = str(value) if isinstance(value, Path) else value
-        conf = OmegaConf.create(field_dict)
-        return OmegaConf.to_yaml(conf)
-
-    @classmethod
-    def add_parser_arguments(cls, parser: ArgumentParser) -> None:
-        """Dynamically create arguments for a settings parser."""
-        if "type" in get_type_hints(cls):
-            settings_stanza = get_args(get_type_hints(cls)["type"])[0]
-        else:
-            settings_stanza = "Uncategorized"
-
-        env_prefix = getattr(cls.model_config, "env_prefix", None)
-        env_prefix = env_prefix if env_prefix is not None else settings_stanza.upper()
-
-        initconf = (
-            cls.initconf.get(settings_stanza)
-            if cls.initconf and settings_stanza in cls.initconf
-            else OmegaConf.create()
-        )
-
-        # create an upcase version of the environment in
-        # order to achieve case-insensitive environment
-        # variables (the way Windows does)
-        upcase_environ = {}
-        for key, value in os.environ.items():
-            upcase_environ[key.upper()] = value
-
-        fields = cls.model_fields
-        cls.argparse_groups = {}
-
-        for name, field in fields.items():
-            if name not in cls._excluded():
-                current_default = field.default
-
-                category = (
-                    field.json_schema_extra.get("category", "Uncategorized")
-                    if field.json_schema_extra
-                    else "Uncategorized"
-                )
-                env_name = env_prefix + "_" + name
-                if category in initconf and name in initconf.get(category):
-                    field.default = initconf.get(category).get(name)
-                if env_name.upper() in upcase_environ:
-                    field.default = upcase_environ[env_name.upper()]
-                cls.add_field_argument(parser, name, field)
-
-                field.default = current_default
-
-    @classmethod
-    def cmd_name(cls, command_field: str = "type") -> str:
-        """Return the category of a setting."""
-        hints = get_type_hints(cls)
-        if command_field in hints:
-            result: str = get_args(hints[command_field])[0]
-            return result
-        else:
-            return "Uncategorized"
-
-    @classmethod
-    def get_parser(cls) -> ArgumentParser:
-        """Get the command-line parser for a setting."""
-        parser = PagingArgumentParser(
-            prog=cls.cmd_name(),
-            description=cls.__doc__,
-        )
-        cls.add_parser_arguments(parser)
-        return parser
-
-    @classmethod
-    def _excluded(cls) -> List[str]:
-        # internal fields that shouldn't be exposed as command line options
-        return ["type", "initconf"]
-
-    @classmethod
-    def _excluded_from_yaml(cls) -> List[str]:
-        # combination of deprecated parameters and internal ones that shouldn't be exposed as invokeai.yaml options
-        return [
-            "type",
-            "initconf",
-            "version",
-            "from_file",
-            "model",
-            "root",
-            "max_cache_size",
-            "max_vram_cache_size",
-            "always_use_cpu",
-            "free_gpu_mem",
-            "xformers_enabled",
-            "tiled_decode",
-            "lora_dir",
-            "embedding_dir",
-            "controlnet_dir",
-            "conf_path",
-        ]
-
-    @classmethod
-    def add_field_argument(cls, command_parser, name: str, field, default_override=None) -> None:
-        """Add the argparse arguments for a setting parser."""
-        field_type = get_type_hints(cls).get(name)
-        default = (
-            default_override
-            if default_override is not None
-            else field.default
-            if field.default_factory is None
-            else field.default_factory()
-        )
-        if category := (field.json_schema_extra.get("category", None) if field.json_schema_extra else None):
-            if category not in cls.argparse_groups:
-                cls.argparse_groups[category] = command_parser.add_argument_group(category)
-            argparse_group = cls.argparse_groups[category]
-        else:
-            argparse_group = command_parser
-
-        if get_origin(field_type) == Literal:
-            allowed_values = get_args(field.annotation)
-            allowed_types = set()
-            for val in allowed_values:
-                allowed_types.add(type(val))
-            allowed_types_list = list(allowed_types)
-            field_type = allowed_types_list[0] if len(allowed_types) == 1 else int_or_float_or_str
-
-            argparse_group.add_argument(
-                f"--{name}",
-                dest=name,
-                type=field_type,
-                default=default,
-                choices=allowed_values,
-                help=field.description,
-            )
-
-        elif get_origin(field_type) == Union:
-            argparse_group.add_argument(
-                f"--{name}",
-                dest=name,
-                type=int_or_float_or_str,
-                default=default,
-                help=field.description,
-            )
-
-        elif get_origin(field_type) == list:
-            argparse_group.add_argument(
-                f"--{name}",
-                dest=name,
-                nargs="*",
-                type=field.annotation,
-                default=default,
-                action=argparse.BooleanOptionalAction if field.annotation == bool else "store",
-                help=field.description,
-            )
-        else:
-            argparse_group.add_argument(
-                f"--{name}",
-                dest=name,
-                type=field.annotation,
-                default=default,
-                action=argparse.BooleanOptionalAction if field.annotation == bool else "store",
-                help=field.description,
-            )

invokeai/app/services/config/config_common.py

@@ -12,7 +12,6 @@ from __future__ import annotations
 
 import argparse
 import pydoc
-from typing import Union
 
 
 class PagingArgumentParser(argparse.ArgumentParser):
@@ -24,18 +23,3 @@ class PagingArgumentParser(argparse.ArgumentParser):
     def print_help(self, file=None) -> None:
         text = self.format_help()
         pydoc.pager(text)
-
-
-def int_or_float_or_str(value: str) -> Union[int, float, str]:
-    """
-    Workaround for argparse type checking.
-    """
-    try:
-        return int(value)
-    except Exception as e:  # noqa F841
-        pass
-    try:
-        return float(value)
-    except Exception as e:  # noqa F841
-        pass
-    return str(value)

invokeai/app/services/config/config_default.py

@@ -1,184 +1,21 @@
-# Copyright (c) 2023 Lincoln Stein (https://github.com/lstein) and the InvokeAI Development Team
-
-"""Invokeai configuration system.
-
-Arguments and fields are taken from the pydantic definition of the
-model.  Defaults can be set by creating a yaml configuration file that
-has a top-level key of "InvokeAI" and subheadings for each of the
-categories returned by `invokeai --help`. The file looks like this:
-
-[file: invokeai.yaml]
-
-InvokeAI:
-  Web Server:
-    host: 127.0.0.1
-    port: 9090
-    allow_origins: []
-    allow_credentials: true
-    allow_methods:
-    - '*'
-    allow_headers:
-    - '*'
-  Features:
-    esrgan: true
-    internet_available: true
-    log_tokenization: false
-    patchmatch: true
-    ignore_missing_core_models: false
-  Paths:
-    autoimport_dir: autoimport
-    lora_dir: null
-    embedding_dir: null
-    controlnet_dir: null
-    models_dir: models
-    legacy_conf_dir: configs/stable-diffusion
-    db_dir: databases
-    outdir: /home/lstein/invokeai-main/outputs
-    use_memory_db: false
-  Logging:
-    log_handlers:
-    - console
-    log_format: plain
-    log_level: info
-  Model Cache:
-    ram: 13.5
-    vram: 0.25
-    lazy_offload: true
-    log_memory_usage: false
-  Device:
-    device: auto
-    precision: auto
-  Generation:
-    sequential_guidance: false
-    attention_type: xformers
-    attention_slice_size: auto
-    force_tiled_decode: false
-
-The default name of the configuration file is `invokeai.yaml`, located
-in INVOKEAI_ROOT. You can replace supersede this by providing any
-OmegaConf dictionary object initialization time:
-
- omegaconf = OmegaConf.load('/tmp/init.yaml')
- conf = InvokeAIAppConfig()
- conf.parse_args(conf=omegaconf)
-
-InvokeAIAppConfig.parse_args() will parse the contents of `sys.argv`
-at initialization time. You may pass a list of strings in the optional
-`argv` argument to use instead of the system argv:
-
- conf.parse_args(argv=['--log_tokenization'])
-
-It is also possible to set a value at initialization time. However, if
-you call parse_args() it may be overwritten.
-
- conf = InvokeAIAppConfig(log_tokenization=True)
- conf.parse_args(argv=['--no-log_tokenization'])
- conf.log_tokenization
- # False
-
-To avoid this, use `get_config()` to retrieve the application-wide
-configuration object. This will retain any properties set at object
-creation time:
-
- conf = InvokeAIAppConfig.get_config(log_tokenization=True)
- conf.parse_args(argv=['--no-log_tokenization'])
- conf.log_tokenization
- # True
-
-Any setting can be overwritten by setting an environment variable of
-form: "INVOKEAI_<setting>", as in:
-
-  export INVOKEAI_port=8080
-
-Order of precedence (from highest):
-   1) initialization options
-   2) command line options
-   3) environment variable options
-   4) config file options
-   5) pydantic defaults
-
-Typical usage at the top level file:
-
- from invokeai.app.services.config import InvokeAIAppConfig
-
- # get global configuration and print its cache size
- conf = InvokeAIAppConfig.get_config()
- conf.parse_args()
- print(conf.ram_cache_size)
-
-Typical usage in a backend module:
-
- from invokeai.app.services.config import InvokeAIAppConfig
-
- # get global configuration and print its cache size value
- conf = InvokeAIAppConfig.get_config()
- print(conf.ram_cache_size)
-
-Computed properties:
-
-The InvokeAIAppConfig object has a series of properties that
-resolve paths relative to the runtime root directory. They each return
-a Path object:
-
- root_path          - path to InvokeAI root
- output_path        - path to default outputs directory
- conf               - alias for the above
- embedding_path     - path to the embeddings directory
- lora_path          - path to the LoRA directory
-
-In most cases, you will want to create a single InvokeAIAppConfig
-object for the entire application. The InvokeAIAppConfig.get_config() function
-does this:
-
-  config = InvokeAIAppConfig.get_config()
-  config.parse_args()   # read values from the command line/config file
-  print(config.root)
-
-# Subclassing
-
-If you wish to create a similar class, please subclass the
-`InvokeAISettings` class and define a Literal field named "type",
-which is set to the desired top-level name.  For example, to create a
-"InvokeBatch" configuration, define like this:
-
-  class InvokeBatch(InvokeAISettings):
-     type: Literal["InvokeBatch"] = "InvokeBatch"
-     node_count : int = Field(default=1, description="Number of nodes to run on", json_schema_extra=dict(category='Resources'))
-     cpu_count  : int = Field(default=8, description="Number of GPUs to run on per node", json_schema_extra=dict(category='Resources'))
-
-This will now read and write from the "InvokeBatch" section of the
-config file, look for environment variables named INVOKEBATCH_*, and
-accept the command-line arguments `--node_count` and `--cpu_count`. The
-two configs are kept in separate sections of the config file:
-
-  # invokeai.yaml
-
-  InvokeBatch:
-     Resources:
-        node_count: 1
-        cpu_count: 8
-
-  InvokeAI:
-     Paths:
-        root: /home/lstein/invokeai-main
-        legacy_conf_dir: configs/stable-diffusion
-        outdir: outputs
-     ...
-
-"""
+# TODO(psyche): pydantic-settings supports YAML settings sources. If we can figure out a way to integrate the YAML
+# migration logic, we could use that for simpler config loading.
 
 from __future__ import annotations
 
 import os
+import re
+import shutil
+from functools import lru_cache
 from pathlib import Path
-from typing import Any, ClassVar, Dict, List, Literal, Optional
+from typing import Any, Literal, Optional
 
-from omegaconf import DictConfig, OmegaConf
-from pydantic import Field
-from pydantic.config import JsonDict
-from pydantic_settings import SettingsConfigDict
+import yaml
+from pydantic import BaseModel, Field, PrivateAttr, field_validator
+from pydantic_settings import BaseSettings, SettingsConfigDict
 
-from .config_base import InvokeAISettings
+from invokeai.backend.model_hash.model_hash import HASHING_ALGORITHMS
+from invokeai.frontend.cli.arg_parser import InvokeAIArgs
 
 INIT_FILE = Path("invokeai.yaml")
 DB_FILE = Path("invokeai.db")
@@ -186,309 +23,409 @@ LEGACY_INIT_FILE = Path("invokeai.init")
 DEFAULT_RAM_CACHE = 10.0
 DEFAULT_VRAM_CACHE = 0.25
 DEFAULT_CONVERT_CACHE = 20.0
+DEVICE = Literal["auto", "cpu", "cuda", "cuda:1", "mps"]
+PRECISION = Literal["auto", "float16", "bfloat16", "float32", "autocast"]
+ATTENTION_TYPE = Literal["auto", "normal", "xformers", "sliced", "torch-sdp"]
+ATTENTION_SLICE_SIZE = Literal["auto", "balanced", "max", 1, 2, 3, 4, 5, 6, 7, 8]
+LOG_FORMAT = Literal["plain", "color", "syslog", "legacy"]
+LOG_LEVEL = Literal["debug", "info", "warning", "error", "critical"]
+CONFIG_SCHEMA_VERSION = 4
 
 
-class Categories(object):
-    """Category headers for configuration variable groups."""
+class URLRegexTokenPair(BaseModel):
+    url_regex: str = Field(description="Regular expression to match against the URL")
+    token: str = Field(description="Token to use when the URL matches the regex")
 
-    WebServer: JsonDict = {"category": "Web Server"}
-    Features: JsonDict = {"category": "Features"}
-    Paths: JsonDict = {"category": "Paths"}
-    Logging: JsonDict = {"category": "Logging"}
-    Development: JsonDict = {"category": "Development"}
-    Other: JsonDict = {"category": "Other"}
-    ModelCache: JsonDict = {"category": "Model Cache"}
-    Device: JsonDict = {"category": "Device"}
-    Generation: JsonDict = {"category": "Generation"}
-    Queue: JsonDict = {"category": "Queue"}
-    Nodes: JsonDict = {"category": "Nodes"}
-    MemoryPerformance: JsonDict = {"category": "Memory/Performance"}
+    @field_validator("url_regex")
+    @classmethod
+    def validate_url_regex(cls, v: str) -> str:
+        """Validate that the value is a valid regex."""
+        try:
+            re.compile(v)
+        except re.error as e:
+            raise ValueError(f"Invalid regex: {e}")
+        return v
 
 
-class InvokeAIAppConfig(InvokeAISettings):
-    """Configuration object for InvokeAI App."""
+class InvokeAIAppConfig(BaseSettings):
+    """Invoke's global app configuration.
 
-    singleton_config: ClassVar[Optional[InvokeAIAppConfig]] = None
-    singleton_init: ClassVar[Optional[Dict[str, Any]]] = None
+    Typically, you won't need to interact with this class directly. Instead, use the `get_config` function from `invokeai.app.services.config` to get a singleton config object.
+
+    Attributes:
+        host: IP address to bind to. Use `0.0.0.0` to serve to your local network.
+        port: Port to bind to.
+        allow_origins: Allowed CORS origins.
+        allow_credentials: Allow CORS credentials.
+        allow_methods: Methods allowed for CORS.
+        allow_headers: Headers allowed for CORS.
+        ssl_certfile: SSL certificate file for HTTPS. See https://www.uvicorn.org/settings/#https.
+        ssl_keyfile: SSL key file for HTTPS. See https://www.uvicorn.org/settings/#https.
+        log_tokenization: Enable logging of parsed prompt tokens.
+        patchmatch: Enable patchmatch inpaint code.
+        ignore_missing_core_models: Ignore missing core models on startup. If `True`, the app will attempt to download missing models on startup.
+        autoimport_dir: Path to a directory of models files to be imported on startup.
+        models_dir: Path to the models directory.
+        convert_cache_dir: Path to the converted models cache directory. When loading a non-diffusers model, it will be converted and store on disk at this location.
+        legacy_conf_dir: Path to directory of legacy checkpoint config files.
+        db_dir: Path to InvokeAI databases directory.
+        outputs_dir: Path to directory for outputs.
+        custom_nodes_dir: Path to directory for custom nodes.
+        log_handlers: Log handler. Valid options are "console", "file=<path>", "syslog=path|address:host:port", "http=<url>".
+        log_format: Log format. Use "plain" for text-only, "color" for colorized output, "legacy" for 2.3-style logging and "syslog" for syslog-style.<br>Valid values: `plain`, `color`, `syslog`, `legacy`
+        log_level: Emit logging messages at this level or higher.<br>Valid values: `debug`, `info`, `warning`, `error`, `critical`
+        log_sql: Log SQL queries. `log_level` must be `debug` for this to do anything. Extremely verbose.
+        use_memory_db: Use in-memory database. Useful for development.
+        dev_reload: Automatically reload when Python sources are changed. Does not reload node definitions.
+        profile_graphs: Enable graph profiling using `cProfile`.
+        profile_prefix: An optional prefix for profile output files.
+        profiles_dir: Path to profiles output directory.
+        ram: Maximum memory amount used by memory model cache for rapid switching (GB).
+        vram: Amount of VRAM reserved for model storage (GB).
+        convert_cache: Maximum size of on-disk converted models cache (GB).
+        lazy_offload: Keep models in VRAM until their space is needed.
+        log_memory_usage: If True, a memory snapshot will be captured before and after every model cache operation, and the result will be logged (at debug level). There is a time cost to capturing the memory snapshots, so it is recommended to only enable this feature if you are actively inspecting the model cache's behaviour.
+        device: Preferred execution device. `auto` will choose the device depending on the hardware platform and the installed torch capabilities.<br>Valid values: `auto`, `cpu`, `cuda`, `cuda:1`, `mps`
+        precision: Floating point precision. `float16` will consume half the memory of `float32` but produce slightly lower-quality images. The `auto` setting will guess the proper precision based on your video card and operating system.<br>Valid values: `auto`, `float16`, `bfloat16`, `float32`, `autocast`
+        sequential_guidance: Whether to calculate guidance in serial instead of in parallel, lowering memory requirements.
+        attention_type: Attention type.<br>Valid values: `auto`, `normal`, `xformers`, `sliced`, `torch-sdp`
+        attention_slice_size: Slice size, valid when attention_type=="sliced".<br>Valid values: `auto`, `balanced`, `max`, `1`, `2`, `3`, `4`, `5`, `6`, `7`, `8`
+        force_tiled_decode: Whether to enable tiled VAE decode (reduces memory consumption with some performance penalty).
+        pil_compress_level: The compress_level setting of PIL.Image.save(), used for PNG encoding. All settings are lossless. 0 = no compression, 1 = fastest with slightly larger filesize, 9 = slowest with smallest filesize. 1 is typically the best setting.
+        max_queue_size: Maximum number of items in the session queue.
+        allow_nodes: List of nodes to allow. Omit to allow all.
+        deny_nodes: List of nodes to deny. Omit to deny none.
+        node_cache_size: How many cached nodes to keep in memory.
+        hashing_algorithm: Model hashing algorthim for model installs. 'blake3' is best for SSDs. 'blake3_single' is best for spinning disk HDDs. 'random' disables hashing, instead assigning a UUID to models. Useful when using a memory db to reduce model installation time, or if you don't care about storing stable hashes for models. Alternatively, any other hashlib algorithm is accepted, though these are not nearly as performant as blake3.<br>Valid values: `md5`, `sha1`, `sha224`, `sha256`, `sha384`, `sha512`, `blake2b`, `blake2s`, `sha3_224`, `sha3_256`, `sha3_384`, `sha3_512`, `shake_128`, `shake_256`, `blake3`, `blake3_single`, `random`
+        remote_api_tokens: List of regular expression and token pairs used when downloading models from URLs. The download URL is tested against the regex, and if it matches, the token is provided in as a Bearer token.
+    """
+
+    _root: Optional[Path] = PrivateAttr(default=None)
 
     # fmt: off
-    type: Literal["InvokeAI"] = "InvokeAI"
+
+    # INTERNAL
+    schema_version:                 int = Field(default=CONFIG_SCHEMA_VERSION, description="Schema version of the config file. This is not a user-configurable setting.")
+    legacy_models_yaml_path: Optional[Path] = Field(default=None,           description="Path to the legacy models.yaml file. This is not a user-configurable setting.")
 
     # WEB
-    host                : str = Field(default="127.0.0.1", description="IP address to bind to", json_schema_extra=Categories.WebServer)
-    port                : int = Field(default=9090, description="Port to bind to", json_schema_extra=Categories.WebServer)
-    allow_origins       : List[str] = Field(default=[], description="Allowed CORS origins", json_schema_extra=Categories.WebServer)
-    allow_credentials   : bool = Field(default=True, description="Allow CORS credentials", json_schema_extra=Categories.WebServer)
-    allow_methods       : List[str] = Field(default=["*"], description="Methods allowed for CORS", json_schema_extra=Categories.WebServer)
-    allow_headers       : List[str] = Field(default=["*"], description="Headers allowed for CORS", json_schema_extra=Categories.WebServer)
-    # SSL options correspond to https://www.uvicorn.org/settings/#https
-    ssl_certfile        : Optional[Path] = Field(default=None, description="SSL certificate file (for HTTPS)", json_schema_extra=Categories.WebServer)
-    ssl_keyfile         : Optional[Path] = Field(default=None, description="SSL key file", json_schema_extra=Categories.WebServer)
+    host:                           str = Field(default="127.0.0.1",        description="IP address to bind to. Use `0.0.0.0` to serve to your local network.")
+    port:                           int = Field(default=9090,               description="Port to bind to.")
+    allow_origins:            list[str] = Field(default=[],                 description="Allowed CORS origins.")
+    allow_credentials:             bool = Field(default=True,               description="Allow CORS credentials.")
+    allow_methods:            list[str] = Field(default=["*"],              description="Methods allowed for CORS.")
+    allow_headers:            list[str] = Field(default=["*"],              description="Headers allowed for CORS.")
+    ssl_certfile:        Optional[Path] = Field(default=None,               description="SSL certificate file for HTTPS. See https://www.uvicorn.org/settings/#https.")
+    ssl_keyfile:         Optional[Path] = Field(default=None,               description="SSL key file for HTTPS. See https://www.uvicorn.org/settings/#https.")
 
-    # FEATURES
-    esrgan              : bool = Field(default=True, description="Enable/disable upscaling code", json_schema_extra=Categories.Features)
-    internet_available  : bool = Field(default=True, description="If true, attempt to download models on the fly; otherwise only use local models", json_schema_extra=Categories.Features)
-    log_tokenization    : bool = Field(default=False, description="Enable logging of parsed prompt tokens.", json_schema_extra=Categories.Features)
-    patchmatch          : bool = Field(default=True, description="Enable/disable patchmatch inpaint code", json_schema_extra=Categories.Features)
-    ignore_missing_core_models : bool = Field(default=False, description='Ignore missing models in models/core/convert', json_schema_extra=Categories.Features)
+    # MISC FEATURES
+    log_tokenization:              bool = Field(default=False,              description="Enable logging of parsed prompt tokens.")
+    patchmatch:                    bool = Field(default=True,               description="Enable patchmatch inpaint code.")
+    ignore_missing_core_models:    bool = Field(default=False,              description="Ignore missing core models on startup. If `True`, the app will attempt to download missing models on startup.")
 
     # PATHS
-    root                : Optional[Path] = Field(default=None, description='InvokeAI runtime root directory', json_schema_extra=Categories.Paths)
-    autoimport_dir      : Path = Field(default=Path('autoimport'), description='Path to a directory of models files to be imported on startup.', json_schema_extra=Categories.Paths)
-    models_dir          : Path = Field(default=Path('models'), description='Path to the models directory', json_schema_extra=Categories.Paths)
-    convert_cache_dir   : Path = Field(default=Path('models/.cache'), description='Path to the converted models cache directory', json_schema_extra=Categories.Paths)
-    legacy_conf_dir     : Path = Field(default=Path('configs/stable-diffusion'), description='Path to directory of legacy checkpoint config files', json_schema_extra=Categories.Paths)
-    db_dir              : Path = Field(default=Path('databases'), description='Path to InvokeAI databases directory', json_schema_extra=Categories.Paths)
-    outdir              : Path = Field(default=Path('outputs'), description='Default folder for output images', json_schema_extra=Categories.Paths)
-    use_memory_db       : bool = Field(default=False, description='Use in-memory database for storing image metadata', json_schema_extra=Categories.Paths)
-    custom_nodes_dir    : Path = Field(default=Path('nodes'), description='Path to directory for custom nodes', json_schema_extra=Categories.Paths)
-    from_file           : Optional[Path] = Field(default=None, description='Take command input from the indicated file (command-line client only)', json_schema_extra=Categories.Paths)
+    autoimport_dir:                Path = Field(default=Path("autoimport"), description="Path to a directory of models files to be imported on startup.")
+    models_dir:                    Path = Field(default=Path("models"),     description="Path to the models directory.")
+    convert_cache_dir:             Path = Field(default=Path("models/.cache"), description="Path to the converted models cache directory. When loading a non-diffusers model, it will be converted and store on disk at this location.")
+    legacy_conf_dir:               Path = Field(default=Path("configs/stable-diffusion"), description="Path to directory of legacy checkpoint config files.")
+    db_dir:                        Path = Field(default=Path("databases"),  description="Path to InvokeAI databases directory.")
+    outputs_dir:                   Path = Field(default=Path("outputs"),    description="Path to directory for outputs.")
+    custom_nodes_dir:              Path = Field(default=Path("nodes"),      description="Path to directory for custom nodes.")
 
     # LOGGING
-    log_handlers        : List[str] = Field(default=["console"], description='Log handler. Valid options are "console", "file=<path>", "syslog=path|address:host:port", "http=<url>"', json_schema_extra=Categories.Logging)
+    log_handlers:             list[str] = Field(default=["console"],        description='Log handler. Valid options are "console", "file=<path>", "syslog=path|address:host:port", "http=<url>".')
     # note - would be better to read the log_format values from logging.py, but this creates circular dependencies issues
-    log_format          : Literal['plain', 'color', 'syslog', 'legacy'] = Field(default="color", description='Log format. Use "plain" for text-only, "color" for colorized output, "legacy" for 2.3-style logging and "syslog" for syslog-style', json_schema_extra=Categories.Logging)
-    log_level           : Literal["debug", "info", "warning", "error", "critical"] = Field(default="info", description="Emit logging messages at this level or  higher", json_schema_extra=Categories.Logging)
-    log_sql             : bool = Field(default=False, description="Log SQL queries", json_schema_extra=Categories.Logging)
+    log_format:              LOG_FORMAT = Field(default="color",            description='Log format. Use "plain" for text-only, "color" for colorized output, "legacy" for 2.3-style logging and "syslog" for syslog-style.')
+    log_level:                LOG_LEVEL = Field(default="info",             description="Emit logging messages at this level or higher.")
+    log_sql:                       bool = Field(default=False,              description="Log SQL queries. `log_level` must be `debug` for this to do anything. Extremely verbose.")
 
     # Development
-    dev_reload          : bool = Field(default=False, description="Automatically reload when Python sources are changed.", json_schema_extra=Categories.Development)
-    profile_graphs      : bool = Field(default=False, description="Enable graph profiling", json_schema_extra=Categories.Development)
-    profile_prefix      : Optional[str] = Field(default=None, description="An optional prefix for profile output files.", json_schema_extra=Categories.Development)
-    profiles_dir        : Path = Field(default=Path('profiles'), description="Directory for graph profiles", json_schema_extra=Categories.Development)
-    skip_model_hash     : bool = Field(default=False, description="Skip model hashing, instead assigning a UUID to models. Useful when using a memory db to reduce startup time.", json_schema_extra=Categories.Development)
-
-    version             : bool = Field(default=False, description="Show InvokeAI version and exit", json_schema_extra=Categories.Other)
+    use_memory_db:                 bool = Field(default=False,              description="Use in-memory database. Useful for development.")
+    dev_reload:                    bool = Field(default=False,              description="Automatically reload when Python sources are changed. Does not reload node definitions.")
+    profile_graphs:                bool = Field(default=False,              description="Enable graph profiling using `cProfile`.")
+    profile_prefix:       Optional[str] = Field(default=None,               description="An optional prefix for profile output files.")
+    profiles_dir:                  Path = Field(default=Path("profiles"),   description="Path to profiles output directory.")
 
     # CACHE
-    ram                 : float = Field(default=DEFAULT_RAM_CACHE, gt=0, description="Maximum memory amount used by model cache for rapid switching (floating point number, GB)", json_schema_extra=Categories.ModelCache, )
-    vram                : float = Field(default=DEFAULT_VRAM_CACHE, ge=0, description="Amount of VRAM reserved for model storage (floating point number, GB)", json_schema_extra=Categories.ModelCache, )
-    convert_cache       : float = Field(default=DEFAULT_CONVERT_CACHE, ge=0, description="Maximum size of on-disk converted models cache (GB)", json_schema_extra=Categories.ModelCache)
-
-    lazy_offload        : bool = Field(default=True, description="Keep models in VRAM until their space is needed", json_schema_extra=Categories.ModelCache, )
-    log_memory_usage    : bool = Field(default=False, description="If True, a memory snapshot will be captured before and after every model cache operation, and the result will be logged (at debug level). There is a time cost to capturing the memory snapshots, so it is recommended to only enable this feature if you are actively inspecting the model cache's behaviour.", json_schema_extra=Categories.ModelCache)
+    ram:                          float = Field(default=DEFAULT_RAM_CACHE, gt=0, description="Maximum memory amount used by memory model cache for rapid switching (GB).")
+    vram:                         float = Field(default=DEFAULT_VRAM_CACHE, ge=0, description="Amount of VRAM reserved for model storage (GB).")
+    convert_cache:                float = Field(default=DEFAULT_CONVERT_CACHE, ge=0, description="Maximum size of on-disk converted models cache (GB).")
+    lazy_offload:                  bool = Field(default=True,               description="Keep models in VRAM until their space is needed.")
+    log_memory_usage:              bool = Field(default=False,              description="If True, a memory snapshot will be captured before and after every model cache operation, and the result will be logged (at debug level). There is a time cost to capturing the memory snapshots, so it is recommended to only enable this feature if you are actively inspecting the model cache's behaviour.")
 
     # DEVICE
-    device              : Literal["auto", "cpu", "cuda", "cuda:1", "mps"] = Field(default="auto", description="Generation device", json_schema_extra=Categories.Device)
-    precision           : Literal["auto", "float16", "bfloat16", "float32", "autocast"] = Field(default="auto", description="Floating point precision", json_schema_extra=Categories.Device)
+    device:                      DEVICE = Field(default="auto",             description="Preferred execution device. `auto` will choose the device depending on the hardware platform and the installed torch capabilities.")
+    precision:                PRECISION = Field(default="auto",             description="Floating point precision. `float16` will consume half the memory of `float32` but produce slightly lower-quality images. The `auto` setting will guess the proper precision based on your video card and operating system.")
 
     # GENERATION
-    sequential_guidance : bool = Field(default=False, description="Whether to calculate guidance in serial instead of in parallel, lowering memory requirements", json_schema_extra=Categories.Generation)
-    attention_type      : Literal["auto", "normal", "xformers", "sliced", "torch-sdp"] = Field(default="auto", description="Attention type", json_schema_extra=Categories.Generation)
-    attention_slice_size: Literal["auto", "balanced", "max", 1, 2, 3, 4, 5, 6, 7, 8] = Field(default="auto", description='Slice size, valid when attention_type=="sliced"', json_schema_extra=Categories.Generation)
-    force_tiled_decode  : bool = Field(default=False, description="Whether to enable tiled VAE decode (reduces memory consumption with some performance penalty)", json_schema_extra=Categories.Generation)
-    png_compress_level  : int = Field(default=1, description="The compress_level setting of PIL.Image.save(), used for PNG encoding. All settings are lossless. 0 = fastest, largest filesize, 9 = slowest, smallest filesize", json_schema_extra=Categories.Generation)
-
-    # QUEUE
-    max_queue_size      : int = Field(default=10000, gt=0, description="Maximum number of items in the session queue", json_schema_extra=Categories.Queue)
+    sequential_guidance:           bool = Field(default=False,              description="Whether to calculate guidance in serial instead of in parallel, lowering memory requirements.")
+    attention_type:      ATTENTION_TYPE = Field(default="auto",             description="Attention type.")
+    attention_slice_size: ATTENTION_SLICE_SIZE = Field(default="auto",      description='Slice size, valid when attention_type=="sliced".')
+    force_tiled_decode:            bool = Field(default=False,              description="Whether to enable tiled VAE decode (reduces memory consumption with some performance penalty).")
+    pil_compress_level:             int = Field(default=1,                  description="The compress_level setting of PIL.Image.save(), used for PNG encoding. All settings are lossless. 0 = no compression, 1 = fastest with slightly larger filesize, 9 = slowest with smallest filesize. 1 is typically the best setting.")
+    max_queue_size:                 int = Field(default=10000, gt=0,        description="Maximum number of items in the session queue.")
 
     # NODES
-    allow_nodes         : Optional[List[str]] = Field(default=None, description="List of nodes to allow. Omit to allow all.", json_schema_extra=Categories.Nodes)
-    deny_nodes          : Optional[List[str]] = Field(default=None, description="List of nodes to deny. Omit to deny none.", json_schema_extra=Categories.Nodes)
-    node_cache_size     : int = Field(default=512, description="How many cached nodes to keep in memory", json_schema_extra=Categories.Nodes)
+    allow_nodes:              list[str] = Field(default=[],                 description="List of nodes to allow. Omit to allow all.")
+    deny_nodes:               list[str] = Field(default=[],                 description="List of nodes to deny. Omit to deny none.")
+    node_cache_size:                int = Field(default=512,                description="How many cached nodes to keep in memory.")
 
-    # MODEL IMPORT
-    civitai_api_key       : Optional[str] = Field(default=os.environ.get("CIVITAI_API_KEY"), description="API key for CivitAI", json_schema_extra=Categories.Other)
+    # MODEL INSTALL
+    hashing_algorithm: HASHING_ALGORITHMS = Field(default="blake3",         description="Model hashing algorthim for model installs. 'blake3' is best for SSDs. 'blake3_single' is best for spinning disk HDDs. 'random' disables hashing, instead assigning a UUID to models. Useful when using a memory db to reduce model installation time, or if you don't care about storing stable hashes for models. Alternatively, any other hashlib algorithm is accepted, though these are not nearly as performant as blake3.")
+    remote_api_tokens: list[URLRegexTokenPair] = Field(default=[],          description="List of regular expression and token pairs used when downloading models from URLs. The download URL is tested against the regex, and if it matches, the token is provided in as a Bearer token.")
 
-    # DEPRECATED FIELDS - STILL HERE IN ORDER TO OBTAN VALUES FROM PRE-3.1 CONFIG FILES
-    always_use_cpu      : bool = Field(default=False, description="If true, use the CPU for rendering even if a GPU is available.", json_schema_extra=Categories.MemoryPerformance)
-    max_cache_size      : Optional[float] = Field(default=None, gt=0, description="Maximum memory amount used by model cache for rapid switching", json_schema_extra=Categories.MemoryPerformance)
-    max_vram_cache_size : Optional[float] = Field(default=None, ge=0, description="Amount of VRAM reserved for model storage", json_schema_extra=Categories.MemoryPerformance)
-    xformers_enabled    : bool = Field(default=True, description="Enable/disable memory-efficient attention", json_schema_extra=Categories.MemoryPerformance)
-    tiled_decode        : bool = Field(default=False, description="Whether to enable tiled VAE decode (reduces memory consumption with some performance penalty)", json_schema_extra=Categories.MemoryPerformance)
-    lora_dir            : Optional[Path] = Field(default=None, description='Path to a directory of LoRA/LyCORIS models to be imported on startup.', json_schema_extra=Categories.Paths)
-    embedding_dir       : Optional[Path] = Field(default=None, description='Path to a directory of Textual Inversion embeddings to be imported on startup.', json_schema_extra=Categories.Paths)
-    controlnet_dir      : Optional[Path] = Field(default=None, description='Path to a directory of ControlNet embeddings to be imported on startup.', json_schema_extra=Categories.Paths)
-    conf_path           : Path = Field(default=Path('configs/models.yaml'), description='Path to models definition file', json_schema_extra=Categories.Paths)
-
-    # this is not referred to in the source code and can be removed entirely
-    #free_gpu_mem        : Optional[bool] = Field(default=None, description="If true, purge model from GPU after each generation.", json_schema_extra=Categories.MemoryPerformance)
-
-    # See InvokeAIAppConfig subclass below for CACHE and DEVICE categories
     # fmt: on
 
-    model_config = SettingsConfigDict(validate_assignment=True, env_prefix="INVOKEAI")
+    model_config = SettingsConfigDict(env_prefix="INVOKEAI_", env_ignore_empty=True)
 
-    def parse_args(
-        self,
-        argv: Optional[list[str]] = None,
-        conf: Optional[DictConfig] = None,
-        clobber: Optional[bool] = False,
-    ) -> None:
+    def update_config(self, config: dict[str, Any] | InvokeAIAppConfig, clobber: bool = True) -> None:
+        """Updates the config, overwriting existing values.
+
+        Args:
+            config: A dictionary of config settings, or instance of `InvokeAIAppConfig`. If an instance of \
+                `InvokeAIAppConfig`, only the explicitly set fields will be merged into the singleton config.
+            clobber: If `True`, overwrite existing values. If `False`, only update fields that are not already set.
         """
-        Update settings with contents of init file, environment, and command-line settings.
 
-        :param conf: alternate Omegaconf dictionary object
-        :param argv: aternate sys.argv list
-        :param clobber: ovewrite any initialization parameters passed during initialization
+        if isinstance(config, dict):
+            new_config = self.model_validate(config)
+        else:
+            new_config = config
+
+        for field_name in new_config.model_fields_set:
+            new_value = getattr(new_config, field_name)
+            current_value = getattr(self, field_name)
+
+            if field_name in self.model_fields_set and not clobber:
+                continue
+
+            if new_value != current_value:
+                setattr(self, field_name, new_value)
+
+    def write_file(self, dest_path: Path) -> None:
+        """Write the current configuration to file. This will overwrite the existing file.
+
+        A `meta` stanza is added to the top of the file, containing metadata about the config file. This is not stored in the config object.
+
+        Args:
+            dest_path: Path to write the config to.
         """
-        # Set the runtime root directory. We parse command-line switches here
-        # in order to pick up the --root_dir option.
-        super().parse_args(argv)
-        loaded_conf = None
-        if conf is None:
-            try:
-                loaded_conf = OmegaConf.load(self.root_dir / INIT_FILE)
-            except Exception:
-                pass
-        if isinstance(loaded_conf, DictConfig):
-            InvokeAISettings.initconf = loaded_conf
-        else:
-            InvokeAISettings.initconf = conf
+        with open(dest_path, "w") as file:
+            # Meta fields should be written in a separate stanza
+            meta_dict = self.model_dump(mode="json", include={"schema_version"})
+            # Only include the legacy_models_yaml_path if it's set
+            if self.legacy_models_yaml_path:
+                meta_dict.update(self.model_dump(mode="json", include={"legacy_models_yaml_path"}))
 
-        # parse args again in order to pick up settings in configuration file
-        super().parse_args(argv)
+            # User settings
+            config_dict = self.model_dump(
+                mode="json",
+                exclude_unset=True,
+                exclude_defaults=True,
+                exclude={"schema_version", "legacy_models_yaml_path"},
+            )
 
-        if self.singleton_init and not clobber:
-            # When setting values in this way, set validate_assignment to true if you want to validate the value.
-            for k, v in self.singleton_init.items():
-                setattr(self, k, v)
+            file.write("# Internal metadata - do not edit:\n")
+            file.write(yaml.dump(meta_dict, sort_keys=False))
+            file.write("\n")
+            file.write("# Put user settings here:\n")
+            if len(config_dict) > 0:
+                file.write(yaml.dump(config_dict, sort_keys=False))
 
-    @classmethod
-    def get_config(cls, **kwargs: Any) -> InvokeAIAppConfig:
-        """Return a singleton InvokeAIAppConfig configuration object."""
-        if (
-            cls.singleton_config is None
-            or type(cls.singleton_config) is not cls
-            or (kwargs and cls.singleton_init != kwargs)
-        ):
-            cls.singleton_config = cls(**kwargs)
-            cls.singleton_init = kwargs
-        return cls.singleton_config
+    def merge_from_file(self, source_path: Optional[Path] = None) -> None:
+        """Read the config from the `invokeai.yaml` file, migrating it if necessary and merging it into the singleton config.
 
-    @property
-    def root_path(self) -> Path:
-        """Path to the runtime root directory."""
-        if self.root:
-            root = Path(self.root).expanduser().absolute()
-        else:
-            root = self.find_root().expanduser().absolute()
-        self.root = root  # insulate ourselves from relative paths that may change
-        return root.resolve()
+        This function will write to the `invokeai.yaml` file if the config is migrated.
 
-    @property
-    def root_dir(self) -> Path:
-        """Alias for above."""
-        return self.root_path
+        Args:
+            source_path: Path to the config file. If not provided, the default path is used.
+        """
+        path = source_path or self.init_file_path
+        config_from_file = load_and_migrate_config(path)
+        # Clobbering here will overwrite any settings that were set via environment variables
+        self.update_config(config_from_file, clobber=False)
+
+    def set_root(self, root: Path) -> None:
+        """Set the runtime root directory. This is typically set using a CLI arg."""
+        assert isinstance(root, Path)
+        self._root = root
 
     def _resolve(self, partial_path: Path) -> Path:
         return (self.root_path / partial_path).resolve()
 
+    @property
+    def root_path(self) -> Path:
+        """Path to the runtime root directory, resolved to an absolute path."""
+        if self._root:
+            root = Path(self._root).expanduser().absolute()
+        else:
+            root = self.find_root().expanduser().absolute()
+        self._root = root  # insulate ourselves from relative paths that may change
+        return root.resolve()
+
     @property
     def init_file_path(self) -> Path:
-        """Path to invokeai.yaml."""
+        """Path to invokeai.yaml, resolved to an absolute path.."""
         resolved_path = self._resolve(INIT_FILE)
         assert resolved_path is not None
         return resolved_path
 
     @property
-    def output_path(self) -> Optional[Path]:
-        """Path to defaults outputs directory."""
-        return self._resolve(self.outdir)
+    def autoimport_path(self) -> Path:
+        """Path to the autoimports directory, resolved to an absolute path.."""
+        return self._resolve(self.autoimport_dir)
+
+    @property
+    def outputs_path(self) -> Optional[Path]:
+        """Path to the outputs directory, resolved to an absolute path.."""
+        return self._resolve(self.outputs_dir)
 
     @property
     def db_path(self) -> Path:
-        """Path to the invokeai.db file."""
+        """Path to the invokeai.db file, resolved to an absolute path.."""
         db_dir = self._resolve(self.db_dir)
         assert db_dir is not None
         return db_dir / DB_FILE
 
-    @property
-    def model_conf_path(self) -> Path:
-        """Path to models configuration file."""
-        return self._resolve(self.conf_path)
-
     @property
     def legacy_conf_path(self) -> Path:
-        """Path to directory of legacy configuration files (e.g. v1-inference.yaml)."""
+        """Path to directory of legacy configuration files (e.g. v1-inference.yaml), resolved to an absolute path.."""
         return self._resolve(self.legacy_conf_dir)
 
     @property
     def models_path(self) -> Path:
-        """Path to the models directory."""
+        """Path to the models directory, resolved to an absolute path.."""
         return self._resolve(self.models_dir)
 
     @property
-    def models_convert_cache_path(self) -> Path:
-        """Path to the converted cache models directory."""
+    def convert_cache_path(self) -> Path:
+        """Path to the converted cache models directory, resolved to an absolute path.."""
         return self._resolve(self.convert_cache_dir)
 
     @property
     def custom_nodes_path(self) -> Path:
-        """Path to the custom nodes directory."""
+        """Path to the custom nodes directory, resolved to an absolute path.."""
         custom_nodes_path = self._resolve(self.custom_nodes_dir)
         assert custom_nodes_path is not None
         return custom_nodes_path
 
-    # the following methods support legacy calls leftover from the Globals era
-    @property
-    def full_precision(self) -> bool:
-        """Return true if precision set to float32."""
-        return self.precision == "float32"
-
-    @property
-    def try_patchmatch(self) -> bool:
-        """Return true if patchmatch true."""
-        return self.patchmatch
-
-    @property
-    def nsfw_checker(self) -> bool:
-        """Return value for NSFW checker. The NSFW node is always active and disabled from Web UI."""
-        return True
-
-    @property
-    def invisible_watermark(self) -> bool:
-        """Return value of invisible watermark. It is always active and disabled from Web UI."""
-        return True
-
-    @property
-    def ram_cache_size(self) -> float:
-        """Return the ram cache size using the legacy or modern setting (GB)."""
-        return self.max_cache_size or self.ram
-
-    @property
-    def vram_cache_size(self) -> float:
-        """Return the vram cache size using the legacy or modern setting (GB)."""
-        return self.max_vram_cache_size or self.vram
-
-    @property
-    def convert_cache_size(self) -> float:
-        """Return the convert cache size on disk (GB)."""
-        return self.convert_cache
-
-    @property
-    def use_cpu(self) -> bool:
-        """Return true if the device is set to CPU or the always_use_cpu flag is set."""
-        return self.always_use_cpu or self.device == "cpu"
-
-    @property
-    def disable_xformers(self) -> bool:
-        """Return true if enable_xformers is false (reversed logic) and attention type is not set to xformers."""
-        disabled_in_config = not self.xformers_enabled
-        return disabled_in_config and self.attention_type != "xformers"
-
     @property
     def profiles_path(self) -> Path:
-        """Path to the graph profiles directory."""
+        """Path to the graph profiles directory, resolved to an absolute path.."""
         return self._resolve(self.profiles_dir)
 
     @staticmethod
     def find_root() -> Path:
         """Choose the runtime root directory when not specified on command line or init file."""
-        return _find_root()
+        venv = Path(os.environ.get("VIRTUAL_ENV") or ".")
+        if os.environ.get("INVOKEAI_ROOT"):
+            root = Path(os.environ["INVOKEAI_ROOT"])
+        elif any((venv.parent / x).exists() for x in [INIT_FILE, LEGACY_INIT_FILE]):
+            root = (venv.parent).resolve()
+        else:
+            root = Path("~/invokeai").expanduser().resolve()
+        return root
 
 
-def get_invokeai_config(**kwargs: Any) -> InvokeAIAppConfig:
-    """Legacy function which returns InvokeAIAppConfig.get_config()."""
-    return InvokeAIAppConfig.get_config(**kwargs)
+def migrate_v3_config_dict(config_dict: dict[str, Any]) -> InvokeAIAppConfig:
+    """Migrate a v3 config dictionary to a current config object.
+
+    Args:
+        config_dict: A dictionary of settings from a v3 config file.
+
+    Returns:
+        An instance of `InvokeAIAppConfig` with the migrated settings.
+
+    """
+    parsed_config_dict: dict[str, Any] = {}
+    for _category_name, category_dict in config_dict["InvokeAI"].items():
+        for k, v in category_dict.items():
+            # `outdir` was renamed to `outputs_dir` in v4
+            if k == "outdir":
+                parsed_config_dict["outputs_dir"] = v
+            # `max_cache_size` was renamed to `ram` some time in v3, but both names were used
+            if k == "max_cache_size" and "ram" not in category_dict:
+                parsed_config_dict["ram"] = v
+            # `max_vram_cache_size` was renamed to `vram` some time in v3, but both names were used
+            if k == "max_vram_cache_size" and "vram" not in category_dict:
+                parsed_config_dict["vram"] = v
+            if k == "conf_path":
+                parsed_config_dict["legacy_models_yaml_path"] = v
+            elif k in InvokeAIAppConfig.model_fields:
+                # skip unknown fields
+                parsed_config_dict[k] = v
+    config = InvokeAIAppConfig.model_validate(parsed_config_dict)
+
+    return config
 
 
-def _find_root() -> Path:
-    venv = Path(os.environ.get("VIRTUAL_ENV") or ".")
-    if os.environ.get("INVOKEAI_ROOT"):
-        root = Path(os.environ["INVOKEAI_ROOT"])
-    elif any((venv.parent / x).exists() for x in [INIT_FILE, LEGACY_INIT_FILE]):
-        root = (venv.parent).resolve()
+def load_and_migrate_config(config_path: Path) -> InvokeAIAppConfig:
+    """Load and migrate a config file to the latest version.
+
+    Args:
+        config_path: Path to the config file.
+
+    Returns:
+        An instance of `InvokeAIAppConfig` with the loaded and migrated settings.
+    """
+    assert config_path.suffix == ".yaml"
+    with open(config_path) as file:
+        loaded_config_dict = yaml.safe_load(file)
+
+    assert isinstance(loaded_config_dict, dict)
+
+    if "InvokeAI" in loaded_config_dict:
+        # This is a v3 config file, attempt to migrate it
+        shutil.copy(config_path, config_path.with_suffix(".yaml.bak"))
+        try:
+            config = migrate_v3_config_dict(loaded_config_dict)
+        except Exception as e:
+            shutil.copy(config_path.with_suffix(".yaml.bak"), config_path)
+            raise RuntimeError(f"Failed to load and migrate v3 config file {config_path}: {e}") from e
+        # By excluding defaults, we ensure that the new config file only contains the settings that were explicitly set
+        config.write_file(config_path)
+        return config
     else:
-        root = Path("~/invokeai").expanduser().resolve()
-    return root
+        # Attempt to load as a v4 config file
+        try:
+            # Meta is not included in the model fields, so we need to validate it separately
+            config = InvokeAIAppConfig.model_validate(loaded_config_dict)
+            assert (
+                config.schema_version == CONFIG_SCHEMA_VERSION
+            ), f"Invalid schema version, expected {CONFIG_SCHEMA_VERSION}: {config.schema_version}"
+            return config
+        except Exception as e:
+            raise RuntimeError(f"Failed to load config file {config_path}: {e}") from e
+
+
+@lru_cache(maxsize=1)
+def get_config() -> InvokeAIAppConfig:
+    """Return the global singleton app config.
+
+    When called, this function will parse the CLI args and merge in config from the `invokeai.yaml` config file.
+    """
+    config = InvokeAIAppConfig()
+
+    args = InvokeAIArgs.args
+
+    # CLI args trump environment variables
+    if root := getattr(args, "root", None):
+        config.set_root(Path(root))
+    if ignore_missing_core_models := getattr(args, "ignore_missing_core_models", None):
+        config.ignore_missing_core_models = ignore_missing_core_models
+
+    # TODO(psyche): This shouldn't be wrapped in a try/catch. The configuration script imports a number of classes
+    # from throughout the app, which in turn call get_config(). At that time, there may not be a config file to
+    # read from, and this raises.
+    #
+    # Once we move all* model installation to the web app, the configure script will not be reaching into the main app
+    # and we can make this an unhandled error, which feels correct.
+    #
+    # *all user-facing models. Core model installation will still be handled by the configure/install script.
+
+    try:
+        config.merge_from_file()
+    except FileNotFoundError:
+        pass
+
+    return config

invokeai/app/services/events/events_base.py

@@ -12,6 +12,7 @@ from invokeai.app.services.session_queue.session_queue_common import (
 )
 from invokeai.app.util.misc import get_timestamp
 from invokeai.backend.model_manager import AnyModelConfig
+from invokeai.backend.model_manager.config import SubModelType
 
 
 class EventServiceBase:
@@ -80,7 +81,7 @@ class EventServiceBase:
                 "graph_execution_state_id": graph_execution_state_id,
                 "node_id": node_id,
                 "source_node_id": source_node_id,
-                "progress_image": progress_image.model_dump() if progress_image is not None else None,
+                "progress_image": progress_image.model_dump(mode="json") if progress_image is not None else None,
                 "step": step,
                 "order": order,
                 "total_steps": total_steps,
@@ -180,6 +181,7 @@ class EventServiceBase:
         queue_batch_id: str,
         graph_execution_state_id: str,
         model_config: AnyModelConfig,
+        submodel_type: Optional[SubModelType] = None,
     ) -> None:
         """Emitted when a model is requested"""
         self.__emit_queue_event(
@@ -189,7 +191,8 @@ class EventServiceBase:
                 "queue_item_id": queue_item_id,
                 "queue_batch_id": queue_batch_id,
                 "graph_execution_state_id": graph_execution_state_id,
-                "model_config": model_config.model_dump(),
+                "model_config": model_config.model_dump(mode="json"),
+                "submodel_type": submodel_type,
             },
         )
 
@@ -200,6 +203,7 @@ class EventServiceBase:
         queue_batch_id: str,
         graph_execution_state_id: str,
         model_config: AnyModelConfig,
+        submodel_type: Optional[SubModelType] = None,
     ) -> None:
         """Emitted when a model is correctly loaded (returns model info)"""
         self.__emit_queue_event(
@@ -209,7 +213,8 @@ class EventServiceBase:
                 "queue_item_id": queue_item_id,
                 "queue_batch_id": queue_batch_id,
                 "graph_execution_state_id": graph_execution_state_id,
-                "model_config": model_config.model_dump(),
+                "model_config": model_config.model_dump(mode="json"),
+                "submodel_type": submodel_type,
             },
         )
 
@@ -254,8 +259,8 @@ class EventServiceBase:
                     "started_at": str(session_queue_item.started_at) if session_queue_item.started_at else None,
                     "completed_at": str(session_queue_item.completed_at) if session_queue_item.completed_at else None,
                 },
-                "batch_status": batch_status.model_dump(),
-                "queue_status": queue_status.model_dump(),
+                "batch_status": batch_status.model_dump(mode="json"),
+                "queue_status": queue_status.model_dump(mode="json"),
             },
         )
 
@@ -405,7 +410,7 @@ class EventServiceBase:
             payload={"source": source, "total_bytes": total_bytes, "key": key, "id": id},
         )
 
-    def emit_model_install_cancelled(self, source: str) -> None:
+    def emit_model_install_cancelled(self, source: str, id: int) -> None:
         """
         Emit when an install job is cancelled.
 
@@ -413,7 +418,7 @@ class EventServiceBase:
         """
         self.__emit_model_event(
             event_name="model_install_cancelled",
-            payload={"source": source},
+            payload={"source": source, "id": id},
         )
 
     def emit_model_install_error(self, source: str, error_type: str, error: str, id: int) -> None:

invokeai/app/services/image_files/image_files_disk.py

@@ -82,7 +82,7 @@ class DiskImageFileStorage(ImageFileStorageBase):
                 image_path,
                 "PNG",
                 pnginfo=pnginfo,
-                compress_level=self.__invoker.services.configuration.png_compress_level,
+                compress_level=self.__invoker.services.configuration.pil_compress_level,
             )
 
             thumbnail_name = get_thumbnail_name(image_name)

invokeai/app/services/invocation_cache/invocation_cache_base.py

@@ -41,8 +41,9 @@ class InvocationCacheBase(ABC):
         """Clears the cache"""
         pass
 
+    @staticmethod
     @abstractmethod
-    def create_key(self, invocation: BaseInvocation) -> int:
+    def create_key(invocation: BaseInvocation) -> int:
         """Gets the key for the invocation's cache item"""
         pass
 

invokeai/app/services/invocation_cache/invocation_cache_memory.py

@@ -61,9 +61,7 @@ class MemoryInvocationCache(InvocationCacheBase):
             self._delete_oldest_access(number_to_delete)
             self._cache[key] = CachedItem(
                 invocation_output,
-                invocation_output.model_dump_json(
-                    warnings=False, exclude_defaults=True, exclude_unset=True, include={"type"}
-                ),
+                invocation_output.model_dump_json(warnings=False, exclude_defaults=True, exclude_unset=True),
             )
 
     def _delete_oldest_access(self, number_to_delete: int) -> None:
@@ -81,7 +79,7 @@ class MemoryInvocationCache(InvocationCacheBase):
         with self._lock:
             return self._delete(key)
 
-    def clear(self, *args, **kwargs) -> None:
+    def clear(self) -> None:
         with self._lock:
             if self._max_cache_size == 0:
                 return

invokeai/app/services/invocation_services.py

@@ -25,6 +25,7 @@ if TYPE_CHECKING:
     from .images.images_base import ImageServiceABC
     from .invocation_cache.invocation_cache_base import InvocationCacheBase
     from .invocation_stats.invocation_stats_base import InvocationStatsServiceBase
+    from .model_images.model_images_base import ModelImageFileStorageBase
     from .model_manager.model_manager_base import ModelManagerServiceBase
     from .names.names_base import NameServiceBase
     from .session_processor.session_processor_base import SessionProcessorBase
@@ -49,6 +50,7 @@ class InvocationServices:
         image_files: "ImageFileStorageBase",
         image_records: "ImageRecordStorageBase",
         logger: "Logger",
+        model_images: "ModelImageFileStorageBase",
         model_manager: "ModelManagerServiceBase",
         download_queue: "DownloadQueueServiceBase",
         performance_statistics: "InvocationStatsServiceBase",
@@ -72,6 +74,7 @@ class InvocationServices:
         self.image_files = image_files
         self.image_records = image_records
         self.logger = logger
+        self.model_images = model_images
         self.model_manager = model_manager
         self.download_queue = download_queue
         self.performance_statistics = performance_statistics

invokeai/app/services/model_images/model_images_base.py

@@ -0,0 +1,33 @@
+from abc import ABC, abstractmethod
+from pathlib import Path
+
+from PIL.Image import Image as PILImageType
+
+
+class ModelImageFileStorageBase(ABC):
+    """Low-level service responsible for storing and retrieving image files."""
+
+    @abstractmethod
+    def get(self, model_key: str) -> PILImageType:
+        """Retrieves a model image as PIL Image."""
+        pass
+
+    @abstractmethod
+    def get_path(self, model_key: str) -> Path:
+        """Gets the internal path to a model image."""
+        pass
+
+    @abstractmethod
+    def get_url(self, model_key: str) -> str | None:
+        """Gets the URL to fetch a model image."""
+        pass
+
+    @abstractmethod
+    def save(self, image: PILImageType, model_key: str) -> None:
+        """Saves a model image."""
+        pass
+
+    @abstractmethod
+    def delete(self, model_key: str) -> None:
+        """Deletes a model image."""
+        pass

invokeai/app/services/model_images/model_images_common.py

@@ -0,0 +1,20 @@
+# TODO: Should these excpetions subclass existing python exceptions?
+class ModelImageFileNotFoundException(Exception):
+    """Raised when an image file is not found in storage."""
+
+    def __init__(self, message="Model image file not found"):
+        super().__init__(message)
+
+
+class ModelImageFileSaveException(Exception):
+    """Raised when an image cannot be saved."""
+
+    def __init__(self, message="Model image file not saved"):
+        super().__init__(message)
+
+
+class ModelImageFileDeleteException(Exception):
+    """Raised when an image cannot be deleted."""
+
+    def __init__(self, message="Model image file not deleted"):
+        super().__init__(message)

invokeai/app/services/model_images/model_images_default.py

@@ -0,0 +1,85 @@
+from pathlib import Path
+
+from PIL import Image
+from PIL.Image import Image as PILImageType
+from send2trash import send2trash
+
+from invokeai.app.services.invoker import Invoker
+from invokeai.app.util.misc import uuid_string
+from invokeai.app.util.thumbnails import make_thumbnail
+
+from .model_images_base import ModelImageFileStorageBase
+from .model_images_common import (
+    ModelImageFileDeleteException,
+    ModelImageFileNotFoundException,
+    ModelImageFileSaveException,
+)
+
+
+class ModelImageFileStorageDisk(ModelImageFileStorageBase):
+    """Stores images on disk"""
+
+    def __init__(self, model_images_folder: Path):
+        self._model_images_folder = model_images_folder
+        self._validate_storage_folders()
+
+    def start(self, invoker: Invoker) -> None:
+        self._invoker = invoker
+
+    def get(self, model_key: str) -> PILImageType:
+        try:
+            path = self.get_path(model_key)
+
+            if not self._validate_path(path):
+                raise ModelImageFileNotFoundException
+
+            return Image.open(path)
+        except FileNotFoundError as e:
+            raise ModelImageFileNotFoundException from e
+
+    def save(self, image: PILImageType, model_key: str) -> None:
+        try:
+            self._validate_storage_folders()
+            image_path = self._model_images_folder / (model_key + ".webp")
+            thumbnail = make_thumbnail(image, 256)
+            thumbnail.save(image_path, format="webp")
+
+        except Exception as e:
+            raise ModelImageFileSaveException from e
+
+    def get_path(self, model_key: str) -> Path:
+        path = self._model_images_folder / (model_key + ".webp")
+
+        return path
+
+    def get_url(self, model_key: str) -> str | None:
+        path = self.get_path(model_key)
+        if not self._validate_path(path):
+            return
+
+        url = self._invoker.services.urls.get_model_image_url(model_key)
+
+        # The image URL never changes, so we must add random query string to it to prevent caching
+        url += f"?{uuid_string()}"
+
+        return url
+
+    def delete(self, model_key: str) -> None:
+        try:
+            path = self.get_path(model_key)
+
+            if not self._validate_path(path):
+                raise ModelImageFileNotFoundException
+
+            send2trash(path)
+
+        except Exception as e:
+            raise ModelImageFileDeleteException from e
+
+    def _validate_path(self, path: Path) -> bool:
+        """Validates the path given for an image."""
+        return path.exists()
+
+    def _validate_storage_folders(self) -> None:
+        """Checks if the required folders exist and create them if they don't"""
+        self._model_images_folder.mkdir(parents=True, exist_ok=True)

invokeai/app/services/model_install/__init__.py

@@ -1,7 +1,6 @@
 """Initialization file for model install service package."""
 
 from .model_install_base import (
-    CivitaiModelSource,
     HFModelSource,
     InstallStatus,
     LocalModelSource,
@@ -23,5 +22,4 @@ __all__ = [
     "LocalModelSource",
     "HFModelSource",
     "URLModelSource",
-    "CivitaiModelSource",
 ]

invokeai/app/services/model_install/model_install_base.py

@@ -91,21 +91,6 @@ class LocalModelSource(StringLikeSource):
         return Path(self.path).as_posix()
 
 
-class CivitaiModelSource(StringLikeSource):
-    """A Civitai version id, with optional variant and access token."""
-
-    version_id: int
-    variant: Optional[ModelRepoVariant] = None
-    access_token: Optional[str] = None
-    type: Literal["civitai"] = "civitai"
-
-    def __str__(self) -> str:
-        """Return string version of repoid when string rep needed."""
-        base: str = str(self.version_id)
-        base += f" ({self.variant})" if self.variant else ""
-        return base
-
-
 class HFModelSource(StringLikeSource):
     """
     A HuggingFace repo_id with optional variant, sub-folder and access token.
@@ -146,14 +131,11 @@ class URLModelSource(StringLikeSource):
         return str(self.url)
 
 
-ModelSource = Annotated[
-    Union[LocalModelSource, HFModelSource, CivitaiModelSource, URLModelSource], Field(discriminator="type")
-]
+ModelSource = Annotated[Union[LocalModelSource, HFModelSource, URLModelSource], Field(discriminator="type")]
 
 MODEL_SOURCE_TO_TYPE_MAP = {
     URLModelSource: ModelSourceType.Url,
     HFModelSource: ModelSourceType.HFRepoID,
-    CivitaiModelSource: ModelSourceType.CivitAI,
     LocalModelSource: ModelSourceType.Path,
 }
 

invokeai/app/services/model_install/model_install_default.py

@@ -11,6 +11,7 @@ from shutil import copyfile, copytree, move, rmtree
 from tempfile import mkdtemp
 from typing import Any, Dict, List, Optional, Set, Union
 
+import yaml
 from huggingface_hub import HfFolder
 from pydantic.networks import AnyHttpUrl
 from requests import Session
@@ -21,7 +22,6 @@ from invokeai.app.services.events.events_base import EventServiceBase
 from invokeai.app.services.invoker import Invoker
 from invokeai.app.services.model_records import DuplicateModelException, ModelRecordServiceBase
 from invokeai.app.services.model_records.model_records_base import ModelRecordChanges
-from invokeai.app.util.misc import uuid_string
 from invokeai.backend.model_manager.config import (
     AnyModelConfig,
     BaseModelType,
@@ -33,12 +33,11 @@ from invokeai.backend.model_manager.config import (
 )
 from invokeai.backend.model_manager.metadata import (
     AnyModelRepoMetadata,
-    CivitaiMetadataFetch,
     HuggingFaceMetadataFetch,
     ModelMetadataWithFiles,
     RemoteModelFile,
 )
-from invokeai.backend.model_manager.metadata.metadata_base import CivitaiMetadata, HuggingFaceMetadata
+from invokeai.backend.model_manager.metadata.metadata_base import HuggingFaceMetadata
 from invokeai.backend.model_manager.probe import ModelProbe
 from invokeai.backend.model_manager.search import ModelSearch
 from invokeai.backend.util import Chdir, InvokeAILogger
@@ -46,7 +45,6 @@ from invokeai.backend.util.devices import choose_precision, choose_torch_device
 
 from .model_install_base import (
     MODEL_SOURCE_TO_TYPE_MAP,
-    CivitaiModelSource,
     HFModelSource,
     InstallStatus,
     LocalModelSource,
@@ -117,6 +115,7 @@ class ModelInstallService(ModelInstallServiceBase):
                 raise Exception("Attempt to start the installer service twice")
             self._start_installer_thread()
             self._remove_dangling_install_dirs()
+            self._migrate_yaml()
             self.sync_to_config()
 
     def stop(self, invoker: Optional[Invoker] = None) -> None:
@@ -154,10 +153,7 @@ class ModelInstallService(ModelInstallServiceBase):
         model_path = Path(model_path)
         config = config or {}
 
-        if self._app_config.skip_model_hash:
-            config["hash"] = uuid_string()
-
-        info: AnyModelConfig = ModelProbe.probe(Path(model_path), config)
+        info: AnyModelConfig = ModelProbe.probe(Path(model_path), config, hash_algo=self._app_config.hashing_algorithm)
 
         if preferred_name := config.get("name"):
             preferred_name = Path(preferred_name).with_suffix(model_path.suffix)
@@ -199,9 +195,16 @@ class ModelInstallService(ModelInstallServiceBase):
                 access_token=access_token,
             )
         elif re.match(r"^https?://[^/]+", source):
+            # Pull the token from config if it exists and matches the URL
+            _token = access_token
+            if _token is None:
+                for pair in self.app_config.remote_api_tokens or []:
+                    if re.search(pair.url_regex, source):
+                        _token = pair.token
+                        break
             source_obj = URLModelSource(
                 url=AnyHttpUrl(source),
-                access_token=access_token,
+                access_token=_token,
             )
         else:
             raise ValueError(f"Unsupported model source: '{source}'")
@@ -216,8 +219,6 @@ class ModelInstallService(ModelInstallServiceBase):
         if isinstance(source, LocalModelSource):
             install_job = self._import_local_model(source, config)
             self._install_queue.put(install_job)  # synchronously install
-        elif isinstance(source, CivitaiModelSource):
-            install_job = self._import_from_civitai(source, config)
         elif isinstance(source, HFModelSource):
             install_job = self._import_from_hf(source, config)
         elif isinstance(source, URLModelSource):
@@ -278,16 +279,64 @@ class ModelInstallService(ModelInstallServiceBase):
     def sync_to_config(self) -> None:
         """Synchronize models on disk to those in the config record store database."""
         self._scan_models_directory()
-        if autoimport := self._app_config.autoimport_dir:
+        if self._app_config.autoimport_path:
             self._logger.info("Scanning autoimport directory for new models")
-            installed = self.scan_directory(self._app_config.root_path / autoimport)
+            installed = self.scan_directory(self._app_config.autoimport_path)
             self._logger.info(f"{len(installed)} new models registered")
         self._logger.info("Model installer (re)initialized")
 
+    def _migrate_yaml(self) -> None:
+        db_models = self.record_store.all_models()
+
+        legacy_models_yaml_path = (
+            self._app_config.legacy_models_yaml_path or self._app_config.root_path / "configs" / "models.yaml"
+        )
+
+        if legacy_models_yaml_path.exists():
+            legacy_models_yaml = yaml.safe_load(legacy_models_yaml_path.read_text())
+
+            yaml_metadata = legacy_models_yaml.pop("__metadata__")
+            yaml_version = yaml_metadata.get("version")
+
+            if yaml_version != "3.0.0":
+                raise ValueError(
+                    f"Attempted migration of unsupported `models.yaml` v{yaml_version}. Only v3.0.0 is supported. Exiting."
+                )
+
+            self._logger.info(
+                f"Starting one-time migration of {len(legacy_models_yaml.items())} models from {str(legacy_models_yaml_path)}. This may take a few minutes."
+            )
+
+            if len(db_models) == 0 and len(legacy_models_yaml.items()) != 0:
+                for model_key, stanza in legacy_models_yaml.items():
+                    _, _, model_name = str(model_key).split("/")
+                    model_path = Path(stanza["path"])
+                    if not model_path.is_absolute():
+                        model_path = self._app_config.models_path / model_path
+                    model_path = model_path.resolve()
+
+                    config: dict[str, Any] = {}
+                    config["name"] = model_name
+                    config["description"] = stanza.get("description")
+                    config["config_path"] = stanza.get("config")
+
+                    try:
+                        id = self.register_path(model_path=model_path, config=config)
+                        self._logger.info(f"Migrated {model_name} with id {id}")
+                    except Exception as e:
+                        self._logger.warning(f"Model at {model_path} could not be migrated: {e}")
+
+            # Rename `models.yaml` to `models.yaml.bak` to prevent re-migration
+            legacy_models_yaml_path.rename(legacy_models_yaml_path.with_suffix(".yaml.bak"))
+
+        # Remove `legacy_models_yaml_path` from the config file - we are done with it either way
+        self._app_config.legacy_models_yaml_path = None
+        self._app_config.write_file(self._app_config.init_file_path)
+
     def scan_directory(self, scan_dir: Path, install: bool = False) -> List[str]:  # noqa D102
-        self._cached_model_paths = {Path(x.path).absolute() for x in self.record_store.all_models()}
+        self._cached_model_paths = {Path(x.path).resolve() for x in self.record_store.all_models()}
         callback = self._scan_install if install else self._scan_register
-        search = ModelSearch(on_model_found=callback, config=self._app_config)
+        search = ModelSearch(on_model_found=callback)
         self._models_installed.clear()
         search.search(scan_dir)
         return list(self._models_installed)
@@ -299,7 +348,7 @@ class ModelInstallService(ModelInstallServiceBase):
         """Unregister the model. Delete its files only if they are within our models directory."""
         model = self.record_store.get_model(key)
         models_dir = self.app_config.models_path
-        model_path = models_dir / model.path
+        model_path = models_dir / Path(model.path)  # handle legacy relative model paths
         if model_path.is_relative_to(models_dir):
             self.unconditionally_delete(key)
         else:
@@ -307,11 +356,11 @@ class ModelInstallService(ModelInstallServiceBase):
 
     def unconditionally_delete(self, key: str) -> None:  # noqa D102
         model = self.record_store.get_model(key)
-        path = self.app_config.models_path / model.path
-        if path.is_dir():
-            rmtree(path)
+        model_path = self.app_config.models_path / model.path
+        if model_path.is_dir():
+            rmtree(model_path)
         else:
-            path.unlink()
+            model_path.unlink()
         self.unregister(key)
 
     def download_and_cache(
@@ -322,7 +371,7 @@ class ModelInstallService(ModelInstallServiceBase):
     ) -> Path:
         """Download the model file located at source to the models cache and return its Path."""
         model_hash = sha256(str(source).encode("utf-8")).hexdigest()[0:32]
-        model_path = self._app_config.models_convert_cache_path / model_hash
+        model_path = self._app_config.convert_cache_path / model_hash
 
         # We expect the cache directory to contain one and only one downloaded file.
         # We don't know the file's name in advance, as it is set by the download
@@ -381,10 +430,8 @@ class ModelInstallService(ModelInstallServiceBase):
                     job.config_in["source"] = str(job.source)
                     job.config_in["source_type"] = MODEL_SOURCE_TO_TYPE_MAP[job.source.__class__]
                     # enter the metadata, if there is any
-                    if isinstance(job.source_metadata, (CivitaiMetadata, HuggingFaceMetadata)):
+                    if isinstance(job.source_metadata, (HuggingFaceMetadata)):
                         job.config_in["source_api_response"] = job.source_metadata.api_response
-                    if isinstance(job.source_metadata, CivitaiMetadata) and job.source_metadata.trigger_phrases:
-                        job.config_in["trigger_phrases"] = job.source_metadata.trigger_phrases
 
                     if job.inplace:
                         key = self.register_path(job.local_path, job.config_in)
@@ -450,7 +497,9 @@ class ModelInstallService(ModelInstallServiceBase):
             self._logger.info(f"Scanning {self._app_config.models_path} for new and orphaned models")
             for cur_base_model in BaseModelType:
                 for cur_model_type in ModelType:
-                    models_dir = Path(cur_base_model.value, cur_model_type.value)
+                    models_dir = self._app_config.models_path / Path(cur_base_model.value, cur_model_type.value)
+                    if not models_dir.exists():
+                        continue
                     installed.update(self.scan_directory(models_dir))
             self._logger.info(f"{len(installed)} new models registered; {len(defunct_models)} unregistered")
 
@@ -469,13 +518,20 @@ class ModelInstallService(ModelInstallServiceBase):
         old_path = Path(model.path)
         models_dir = self.app_config.models_path
 
-        if not old_path.is_relative_to(models_dir):
+        try:
+            old_path.relative_to(models_dir)
+            return model
+        except ValueError:
+            pass
+
+        new_path = models_dir / model.base.value / model.type.value / old_path.name
+
+        if old_path == new_path or new_path.exists() and old_path == new_path.resolve():
             return model
 
-        new_path = models_dir / model.base.value / model.type.value / model.name
         self._logger.info(f"Moving {model.name} to {new_path}.")
         new_path = self._move_model(old_path, new_path)
-        model.path = new_path.relative_to(models_dir).as_posix()
+        model.path = new_path.as_posix()
         self.record_store.update_model(key, ModelRecordChanges(path=model.path))
         return model
 
@@ -533,22 +589,16 @@ class ModelInstallService(ModelInstallServiceBase):
     ) -> str:
         config = config or {}
 
-        if self._app_config.skip_model_hash:
-            config["hash"] = uuid_string()
+        info = info or ModelProbe.probe(model_path, config, hash_algo=self._app_config.hashing_algorithm)
 
-        info = info or ModelProbe.probe(model_path, config)
-
-        model_path = model_path.absolute()
-        if model_path.is_relative_to(self.app_config.models_path):
-            model_path = model_path.relative_to(self.app_config.models_path)
+        model_path = model_path.resolve()
 
         info.path = model_path.as_posix()
 
         # add 'main' specific fields
         if isinstance(info, CheckpointConfigBase):
-            # make config relative to our root
-            legacy_conf = (self.app_config.root_dir / self.app_config.legacy_conf_dir / info.config_path).resolve()
-            info.config_path = legacy_conf.relative_to(self.app_config.root_dir).as_posix()
+            legacy_conf = (self.app_config.legacy_conf_path / info.config_path).resolve()
+            info.config_path = legacy_conf.as_posix()
         self.record_store.add_model(info)
         return info.key
 
@@ -573,16 +623,6 @@ class ModelInstallService(ModelInstallServiceBase):
             inplace=source.inplace or False,
         )
 
-    def _import_from_civitai(self, source: CivitaiModelSource, config: Optional[Dict[str, Any]]) -> ModelInstallJob:
-        if not source.access_token:
-            self._logger.info("No Civitai access token provided; some models may not be downloadable.")
-        metadata = CivitaiMetadataFetch(self._session, self.app_config.get_config().civitai_api_key).from_id(
-            str(source.version_id)
-        )
-        assert isinstance(metadata, ModelMetadataWithFiles)
-        remote_files = metadata.download_urls(session=self._session)
-        return self._import_remote_model(source=source, config=config, metadata=metadata, remote_files=remote_files)
-
     def _import_from_hf(self, source: HFModelSource, config: Optional[Dict[str, Any]]) -> ModelInstallJob:
         # Add user's cached access token to HuggingFace requests
         source.access_token = source.access_token or HfFolder.get_token()
@@ -605,7 +645,7 @@ class ModelInstallService(ModelInstallServiceBase):
         )
 
     def _import_from_url(self, source: URLModelSource, config: Optional[Dict[str, Any]]) -> ModelInstallJob:
-        # URLs from Civitai or HuggingFace will be handled specially
+        # URLs from HuggingFace will be handled specially
         metadata = None
         fetcher = None
         try:
@@ -613,8 +653,6 @@ class ModelInstallService(ModelInstallServiceBase):
         except ValueError:
             pass
         kwargs: dict[str, Any] = {"session": self._session}
-        if fetcher is CivitaiMetadataFetch:
-            kwargs["api_key"] = self._app_config.get_config().civitai_api_key
         if fetcher is not None:
             metadata = fetcher(**kwargs).from_url(source.url)
         self._logger.debug(f"metadata={metadata}")
@@ -631,7 +669,7 @@ class ModelInstallService(ModelInstallServiceBase):
 
     def _import_remote_model(
         self,
-        source: HFModelSource | CivitaiModelSource | URLModelSource,
+        source: HFModelSource | URLModelSource,
         remote_files: List[RemoteModelFile],
         metadata: Optional[AnyModelRepoMetadata],
         config: Optional[Dict[str, Any]],
@@ -845,12 +883,10 @@ class ModelInstallService(ModelInstallServiceBase):
     def _signal_job_cancelled(self, job: ModelInstallJob) -> None:
         self._logger.info(f"{job.source}: model installation was cancelled")
         if self._event_bus:
-            self._event_bus.emit_model_install_cancelled(str(job.source))
+            self._event_bus.emit_model_install_cancelled(str(job.source), id=job.id)
 
     @staticmethod
     def get_fetcher_from_url(url: str):
-        if re.match(r"^https?://civitai.com/", url.lower()):
-            return CivitaiMetadataFetch
-        elif re.match(r"^https?://huggingface.co/[^/]+/[^/]+$", url.lower()):
+        if re.match(r"^https?://huggingface.co/[^/]+/[^/]+$", url.lower()):
             return HuggingFaceMetadataFetch
         raise ValueError(f"Unsupported model source: '{url}'")

invokeai/app/services/model_load/model_load_default.py

@@ -68,6 +68,7 @@ class ModelLoadService(ModelLoadServiceBase):
             self._emit_load_event(
                 context_data=context_data,
                 model_config=model_config,
+                submodel_type=submodel_type,
             )
 
         implementation, model_config, submodel_type = self._registry.get_implementation(model_config, submodel_type)  # type: ignore
@@ -82,6 +83,7 @@ class ModelLoadService(ModelLoadServiceBase):
             self._emit_load_event(
                 context_data=context_data,
                 model_config=model_config,
+                submodel_type=submodel_type,
                 loaded=True,
             )
         return loaded_model
@@ -91,6 +93,7 @@ class ModelLoadService(ModelLoadServiceBase):
         context_data: InvocationContextData,
         model_config: AnyModelConfig,
         loaded: Optional[bool] = False,
+        submodel_type: Optional[SubModelType] = None,
     ) -> None:
         if not self._invoker:
             return
@@ -102,6 +105,7 @@ class ModelLoadService(ModelLoadServiceBase):
                 queue_batch_id=context_data.queue_item.batch_id,
                 graph_execution_state_id=context_data.queue_item.session_id,
                 model_config=model_config,
+                submodel_type=submodel_type,
             )
         else:
             self._invoker.services.events.emit_model_load_completed(
@@ -110,4 +114,5 @@ class ModelLoadService(ModelLoadServiceBase):
                 queue_batch_id=context_data.queue_item.batch_id,
                 graph_execution_state_id=context_data.queue_item.session_id,
                 model_config=model_config,
+                submodel_type=submodel_type,
             )

invokeai/app/services/model_manager/model_manager_base.py

@@ -1,15 +1,11 @@
 # Copyright (c) 2023 Lincoln D. Stein and the InvokeAI Team
 
 from abc import ABC, abstractmethod
-from typing import Optional
 
 import torch
 from typing_extensions import Self
 
 from invokeai.app.services.invoker import Invoker
-from invokeai.app.services.shared.invocation_context import InvocationContextData
-from invokeai.backend.model_manager.config import AnyModelConfig, BaseModelType, ModelType, SubModelType
-from invokeai.backend.model_manager.load.load_base import LoadedModel
 
 from ..config import InvokeAIAppConfig
 from ..download import DownloadQueueServiceBase
@@ -70,32 +66,3 @@ class ModelManagerServiceBase(ABC):
     @abstractmethod
     def stop(self, invoker: Invoker) -> None:
         pass
-
-    @abstractmethod
-    def load_model_by_config(
-        self,
-        model_config: AnyModelConfig,
-        submodel_type: Optional[SubModelType] = None,
-        context_data: Optional[InvocationContextData] = None,
-    ) -> LoadedModel:
-        pass
-
-    @abstractmethod
-    def load_model_by_key(
-        self,
-        key: str,
-        submodel_type: Optional[SubModelType] = None,
-        context_data: Optional[InvocationContextData] = None,
-    ) -> LoadedModel:
-        pass
-
-    @abstractmethod
-    def load_model_by_attr(
-        self,
-        model_name: str,
-        base_model: BaseModelType,
-        model_type: ModelType,
-        submodel: Optional[SubModelType] = None,
-        context_data: Optional[InvocationContextData] = None,
-    ) -> LoadedModel:
-        pass

invokeai/app/services/model_manager/model_manager_default.py

@@ -1,14 +1,10 @@
 # Copyright (c) 2023 Lincoln D. Stein and the InvokeAI Team
 """Implementation of ModelManagerServiceBase."""
 
-from typing import Optional
-
 import torch
 from typing_extensions import Self
 
 from invokeai.app.services.invoker import Invoker
-from invokeai.app.services.shared.invocation_context import InvocationContextData
-from invokeai.backend.model_manager import AnyModelConfig, BaseModelType, LoadedModel, ModelType, SubModelType
 from invokeai.backend.model_manager.load import ModelCache, ModelConvertCache, ModelLoaderRegistry
 from invokeai.backend.util.devices import choose_torch_device
 from invokeai.backend.util.logging import InvokeAILogger
@@ -18,7 +14,7 @@ from ..download import DownloadQueueServiceBase
 from ..events.events_base import EventServiceBase
 from ..model_install import ModelInstallService, ModelInstallServiceBase
 from ..model_load import ModelLoadService, ModelLoadServiceBase
-from ..model_records import ModelRecordServiceBase, UnknownModelException
+from ..model_records import ModelRecordServiceBase
 from .model_manager_base import ModelManagerServiceBase
 
 
@@ -64,56 +60,6 @@ class ModelManagerService(ModelManagerServiceBase):
             if hasattr(service, "stop"):
                 service.stop(invoker)
 
-    def load_model_by_config(
-        self,
-        model_config: AnyModelConfig,
-        submodel_type: Optional[SubModelType] = None,
-        context_data: Optional[InvocationContextData] = None,
-    ) -> LoadedModel:
-        return self.load.load_model(model_config, submodel_type, context_data)
-
-    def load_model_by_key(
-        self,
-        key: str,
-        submodel_type: Optional[SubModelType] = None,
-        context_data: Optional[InvocationContextData] = None,
-    ) -> LoadedModel:
-        config = self.store.get_model(key)
-        return self.load.load_model(config, submodel_type, context_data)
-
-    def load_model_by_attr(
-        self,
-        model_name: str,
-        base_model: BaseModelType,
-        model_type: ModelType,
-        submodel: Optional[SubModelType] = None,
-        context_data: Optional[InvocationContextData] = None,
-    ) -> LoadedModel:
-        """
-        Given a model's attributes, search the database for it, and if found, load and return the LoadedModel object.
-
-        This is provided for API compatability with the get_model() method
-        in the original model manager. However, note that LoadedModel is
-        not the same as the original ModelInfo that ws returned.
-
-        :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
-        :param context: The invocation context.
-
-        Exceptions: UnknownModelException -- model with this key not known
-                    NotImplementedException -- a model loader was not provided at initialization time
-                    ValueError -- more than one model matches this combination
-        """
-        configs = self.store.search_by_attr(model_name, base_model, model_type)
-        if len(configs) == 0:
-            raise UnknownModelException(f"{base_model}/{model_type}/{model_name}: Unknown model")
-        elif len(configs) > 1:
-            raise ValueError(f"{base_model}/{model_type}/{model_name}: More than one model matches.")
-        else:
-            return self.load.load_model(configs[0], submodel, context_data)
-
     @classmethod
     def build_model_manager(
         cls,
@@ -132,14 +78,12 @@ class ModelManagerService(ModelManagerServiceBase):
         logger.setLevel(app_config.log_level.upper())
 
         ram_cache = ModelCache(
-            max_cache_size=app_config.ram_cache_size,
-            max_vram_cache_size=app_config.vram_cache_size,
+            max_cache_size=app_config.ram,
+            max_vram_cache_size=app_config.vram,
             logger=logger,
             execution_device=execution_device,
         )
-        convert_cache = ModelConvertCache(
-            cache_path=app_config.models_convert_cache_path, max_size=app_config.convert_cache_size
-        )
+        convert_cache = ModelConvertCache(cache_path=app_config.convert_cache_path, max_size=app_config.convert_cache)
         loader = ModelLoadService(
             app_config=app_config,
             ram_cache=ram_cache,

invokeai/app/services/model_records/model_records_base.py

@@ -18,7 +18,12 @@ from invokeai.backend.model_manager import (
     ModelFormat,
     ModelType,
 )
-from invokeai.backend.model_manager.config import ModelDefaultSettings, ModelVariantType, SchedulerPredictionType
+from invokeai.backend.model_manager.config import (
+    ControlAdapterDefaultSettings,
+    MainModelDefaultSettings,
+    ModelVariantType,
+    SchedulerPredictionType,
+)
 
 
 class DuplicateModelException(Exception):
@@ -68,7 +73,7 @@ class ModelRecordChanges(BaseModelExcludeNull):
     description: Optional[str] = Field(description="Model description", default=None)
     base: Optional[BaseModelType] = Field(description="The base model.", default=None)
     trigger_phrases: Optional[set[str]] = Field(description="Set of trigger phrases for this model", default=None)
-    default_settings: Optional[ModelDefaultSettings] = Field(
+    default_settings: Optional[MainModelDefaultSettings | ControlAdapterDefaultSettings] = Field(
         description="Default settings for this model", default=None
     )
 
@@ -79,6 +84,7 @@ class ModelRecordChanges(BaseModelExcludeNull):
         description="The prediction type of the model.", default=None
     )
     upcast_attention: Optional[bool] = Field(description="Whether to upcast attention.", default=None)
+    config_path: Optional[str] = Field(description="Path to config file for model", default=None)
 
 
 class ModelRecordServiceBase(ABC):
@@ -129,6 +135,17 @@ class ModelRecordServiceBase(ABC):
         """
         pass
 
+    @abstractmethod
+    def get_model_by_hash(self, hash: str) -> AnyModelConfig:
+        """
+        Retrieve the configuration for the indicated model.
+
+        :param hash: Hash of model config to be fetched.
+
+        Exceptions: UnknownModelException
+        """
+        pass
+
     @abstractmethod
     def list_models(
         self, page: int = 0, per_page: int = 10, order_by: ModelRecordOrderBy = ModelRecordOrderBy.Default

invokeai/app/services/model_records/model_records_sql.py

@@ -203,6 +203,21 @@ class ModelRecordServiceSQL(ModelRecordServiceBase):
             model = ModelConfigFactory.make_config(json.loads(rows[0]), timestamp=rows[1])
         return model
 
+    def get_model_by_hash(self, hash: str) -> AnyModelConfig:
+        with self._db.lock:
+            self._cursor.execute(
+                """--sql
+                SELECT config, strftime('%s',updated_at) FROM models
+                WHERE hash=?;
+                """,
+                (hash,),
+            )
+            rows = self._cursor.fetchone()
+            if not rows:
+                raise UnknownModelException("model not found")
+            model = ModelConfigFactory.make_config(json.loads(rows[0]), timestamp=rows[1])
+        return model
+
     def exists(self, key: str) -> bool:
         """
         Return True if a model with the indicated key exists in the databse.
@@ -227,6 +242,7 @@ class ModelRecordServiceSQL(ModelRecordServiceBase):
         base_model: Optional[BaseModelType] = None,
         model_type: Optional[ModelType] = None,
         model_format: Optional[ModelFormat] = None,
+        order_by: ModelRecordOrderBy = ModelRecordOrderBy.Default,
     ) -> List[AnyModelConfig]:
         """
         Return models matching name, base and/or type.
@@ -235,10 +251,21 @@ class ModelRecordServiceSQL(ModelRecordServiceBase):
         :param base_model: Filter by base model (optional)
         :param model_type: Filter by type of model (optional)
         :param model_format: Filter by model format (e.g. "diffusers") (optional)
+        :param order_by: Result order
 
         If none of the optional filters are passed, will return all
         models in the database.
         """
+
+        assert isinstance(order_by, ModelRecordOrderBy)
+        ordering = {
+            ModelRecordOrderBy.Default: "type, base, name, format",
+            ModelRecordOrderBy.Type: "type",
+            ModelRecordOrderBy.Base: "base",
+            ModelRecordOrderBy.Name: "name",
+            ModelRecordOrderBy.Format: "format",
+        }
+
         where_clause: list[str] = []
         bindings: list[str] = []
         if model_name:
@@ -257,8 +284,10 @@ class ModelRecordServiceSQL(ModelRecordServiceBase):
         with self._db.lock:
             self._cursor.execute(
                 f"""--sql
-                SELECT config, strftime('%s',updated_at) FROM models
-                {where};
+                SELECT config, strftime('%s',updated_at)
+                FROM models
+                {where}
+                ORDER BY {ordering[order_by]} -- using ? to bind doesn't work here for some reason;
                 """,
                 tuple(bindings),
             )
@@ -304,7 +333,7 @@ class ModelRecordServiceSQL(ModelRecordServiceBase):
         """Return a paginated summary listing of each model in the database."""
         assert isinstance(order_by, ModelRecordOrderBy)
         ordering = {
-            ModelRecordOrderBy.Default: "type, base, format, name",
+            ModelRecordOrderBy.Default: "type, base, name, format",
             ModelRecordOrderBy.Type: "type",
             ModelRecordOrderBy.Base: "base",
             ModelRecordOrderBy.Name: "name",

invokeai/app/services/session_queue/session_queue_sqlite.py

@@ -151,7 +151,7 @@ class SqliteSessionQueue(SessionQueueBase):
 
             # TODO: how does this work in a multi-user scenario?
             current_queue_size = self._get_current_queue_size(queue_id)
-            max_queue_size = self.__invoker.services.configuration.get_config().max_queue_size
+            max_queue_size = self.__invoker.services.configuration.max_queue_size
             max_new_queue_items = max_queue_size - current_queue_size
 
             priority = 0

invokeai/app/services/shared/invocation_context.py

@@ -1,7 +1,7 @@
 import threading
 from dataclasses import dataclass
 from pathlib import Path
-from typing import TYPE_CHECKING, Optional
+from typing import TYPE_CHECKING, Optional, Union
 
 from PIL.Image import Image
 from torch import Tensor
@@ -13,15 +13,16 @@ from invokeai.app.services.config.config_default import InvokeAIAppConfig
 from invokeai.app.services.image_records.image_records_common import ImageCategory, ResourceOrigin
 from invokeai.app.services.images.images_common import ImageDTO
 from invokeai.app.services.invocation_services import InvocationServices
+from invokeai.app.services.model_records.model_records_base import UnknownModelException
 from invokeai.app.util.step_callback import stable_diffusion_step_callback
 from invokeai.backend.model_manager.config import AnyModelConfig, BaseModelType, ModelFormat, ModelType, SubModelType
 from invokeai.backend.model_manager.load.load_base import LoadedModel
-from invokeai.backend.model_manager.metadata.metadata_base import AnyModelRepoMetadata
 from invokeai.backend.stable_diffusion.diffusers_pipeline import PipelineIntermediateState
 from invokeai.backend.stable_diffusion.diffusion.conditioning_data import ConditioningFieldData
 
 if TYPE_CHECKING:
     from invokeai.app.invocations.baseinvocation import BaseInvocation
+    from invokeai.app.invocations.model import ModelIdentifierField
     from invokeai.app.services.session_queue.session_queue_common import SessionQueueItem
 
 """
@@ -299,22 +300,27 @@ class ConditioningInterface(InvocationContextInterface):
 
 
 class ModelsInterface(InvocationContextInterface):
-    def exists(self, key: str) -> bool:
+    def exists(self, identifier: Union[str, "ModelIdentifierField"]) -> bool:
         """Checks if a model exists.
 
         Args:
-            key: The key of the model.
+            identifier: The key or ModelField representing the model.
 
         Returns:
             True if the model exists, False if not.
         """
-        return self._services.model_manager.store.exists(key)
+        if isinstance(identifier, str):
+            return self._services.model_manager.store.exists(identifier)
 
-    def load(self, key: str, submodel_type: Optional[SubModelType] = None) -> LoadedModel:
+        return self._services.model_manager.store.exists(identifier.key)
+
+    def load(
+        self, identifier: Union[str, "ModelIdentifierField"], submodel_type: Optional[SubModelType] = None
+    ) -> LoadedModel:
         """Loads a model.
 
         Args:
-            key: The key of the model.
+            identifier: The key or ModelField representing the model.
             submodel_type: The submodel of the model to get.
 
         Returns:
@@ -324,9 +330,13 @@ class ModelsInterface(InvocationContextInterface):
         # The model manager emits events as it loads the model. It needs the context data to build
         # the event payloads.
 
-        return self._services.model_manager.load_model_by_key(
-            key=key, submodel_type=submodel_type, context_data=self._data
-        )
+        if isinstance(identifier, str):
+            model = self._services.model_manager.store.get_model(identifier)
+            return self._services.model_manager.load.load_model(model, submodel_type, self._data)
+        else:
+            _submodel_type = submodel_type or identifier.submodel_type
+            model = self._services.model_manager.store.get_model(identifier.key)
+            return self._services.model_manager.load.load_model(model, _submodel_type, self._data)
 
     def load_by_attrs(
         self, name: str, base: BaseModelType, type: ModelType, submodel_type: Optional[SubModelType] = None
@@ -343,35 +353,29 @@ class ModelsInterface(InvocationContextInterface):
         Returns:
             An object representing the loaded model.
         """
-        return self._services.model_manager.load_model_by_attr(
-            model_name=name,
-            base_model=base,
-            model_type=type,
-            submodel=submodel_type,
-            context_data=self._data,
-        )
 
-    def get_config(self, key: str) -> AnyModelConfig:
+        configs = self._services.model_manager.store.search_by_attr(model_name=name, base_model=base, model_type=type)
+        if len(configs) == 0:
+            raise UnknownModelException(f"No model found with name {name}, base {base}, and type {type}")
+
+        if len(configs) > 1:
+            raise ValueError(f"More than one model found with name {name}, base {base}, and type {type}")
+
+        return self._services.model_manager.load.load_model(configs[0], submodel_type, self._data)
+
+    def get_config(self, identifier: Union[str, "ModelIdentifierField"]) -> AnyModelConfig:
         """Gets a model's config.
 
         Args:
-            key: The key of the model.
+            identifier: The key or ModelField representing the model.
 
         Returns:
             The model's config.
         """
-        return self._services.model_manager.store.get_model(key=key)
+        if isinstance(identifier, str):
+            return self._services.model_manager.store.get_model(identifier)
 
-    def get_metadata(self, key: str) -> Optional[AnyModelRepoMetadata]:
-        """Gets a model's metadata, if it has any.
-
-        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)
+        return self._services.model_manager.store.get_model(identifier.key)
 
     def search_by_path(self, path: Path) -> list[AnyModelConfig]:
         """Searches for models by path.
@@ -419,7 +423,7 @@ class ConfigInterface(InvocationContextInterface):
             The app's config.
         """
 
-        return self._services.configuration.get_config()
+        return self._services.configuration
 
 
 class UtilInterface(InvocationContextInterface):

invokeai/app/services/shared/sqlite_migrator/migrations/migration_3.py

@@ -4,8 +4,6 @@ from logging import Logger
 from invokeai.app.services.config import InvokeAIAppConfig
 from invokeai.app.services.shared.sqlite_migrator.sqlite_migrator_common import Migration
 
-from .util.migrate_yaml_config_1 import MigrateModelYamlToDb1
-
 
 class Migration3Callback:
     def __init__(self, app_config: InvokeAIAppConfig, logger: Logger) -> None:
@@ -15,7 +13,6 @@ class Migration3Callback:
     def __call__(self, cursor: sqlite3.Cursor) -> None:
         self._drop_model_manager_metadata(cursor)
         self._recreate_model_config(cursor)
-        self._migrate_model_config_records(cursor)
 
     def _drop_model_manager_metadata(self, cursor: sqlite3.Cursor) -> None:
         """Drops the `model_manager_metadata` table."""
@@ -55,12 +52,6 @@ class Migration3Callback:
             """
         )
 
-    def _migrate_model_config_records(self, cursor: sqlite3.Cursor) -> None:
-        """After updating the model config table, we repopulate it."""
-        self._logger.info("Migrating model config records from models.yaml to database")
-        model_record_migrator = MigrateModelYamlToDb1(self._app_config, self._logger, cursor)
-        model_record_migrator.migrate()
-
 
 def build_migration_3(app_config: InvokeAIAppConfig, logger: Logger) -> Migration:
     """

invokeai/app/services/shared/sqlite_migrator/migrations/util/migrate_yaml_config_1.py

@@ -1,163 +0,0 @@
-# Copyright (c) 2023 Lincoln D. Stein
-"""Migrate from the InvokeAI v2 models.yaml format to the v3 sqlite format."""
-
-import json
-import sqlite3
-from logging import Logger
-from pathlib import Path
-from typing import Optional
-
-from omegaconf import DictConfig, OmegaConf
-from pydantic import TypeAdapter
-
-from invokeai.app.services.config import InvokeAIAppConfig
-from invokeai.app.services.model_records import (
-    DuplicateModelException,
-    UnknownModelException,
-)
-from invokeai.backend.model_manager.config import (
-    AnyModelConfig,
-    BaseModelType,
-    ModelConfigFactory,
-    ModelType,
-)
-from invokeai.backend.model_manager.hash import ModelHash
-
-ModelsValidator = TypeAdapter(AnyModelConfig)
-
-
-class MigrateModelYamlToDb1:
-    """
-    Migrate the InvokeAI models.yaml format (VERSION 3.0.0) to SQL3 database format (VERSION 3.5.0).
-
-    The class has one externally useful method, migrate(), which scans the
-    currently models.yaml file and imports all its entries into invokeai.db.
-
-    Use this way:
-
-      from invokeai.backend.model_manager/migrate_to_db import MigrateModelYamlToDb
-      MigrateModelYamlToDb().migrate()
-
-    """
-
-    config: InvokeAIAppConfig
-    logger: Logger
-    cursor: sqlite3.Cursor
-
-    def __init__(self, config: InvokeAIAppConfig, logger: Logger, cursor: sqlite3.Cursor = None) -> None:
-        self.config = config
-        self.logger = logger
-        self.cursor = cursor
-
-    def get_yaml(self) -> DictConfig:
-        """Fetch the models.yaml DictConfig for this installation."""
-        yaml_path = self.config.model_conf_path
-        omegaconf = OmegaConf.load(yaml_path)
-        assert isinstance(omegaconf, DictConfig)
-        return omegaconf
-
-    def migrate(self) -> None:
-        """Do the migration from models.yaml to invokeai.db."""
-        try:
-            yaml = self.get_yaml()
-        except OSError:
-            return
-
-        for model_key, stanza in yaml.items():
-            if model_key == "__metadata__":
-                assert (
-                    stanza["version"] == "3.0.0"
-                ), f"This script works on version 3.0.0 yaml files, but your configuration points to a {stanza['version']} version"
-                continue
-
-            base_type, model_type, model_name = str(model_key).split("/")
-            try:
-                hash = ModelHash().hash(self.config.models_path / stanza.path)
-            except OSError:
-                self.logger.warning(f"The model at {stanza.path} is not a valid file or directory. Skipping migration.")
-                continue
-
-            stanza["base"] = BaseModelType(base_type)
-            stanza["type"] = ModelType(model_type)
-            stanza["name"] = model_name
-            stanza["original_hash"] = hash
-            stanza["current_hash"] = hash
-            new_key = hash  # deterministic key assignment
-
-            # special case for ip adapters, which need the new `image_encoder_model_id` field
-            if stanza["type"] == ModelType.IPAdapter:
-                try:
-                    stanza["image_encoder_model_id"] = self._get_image_encoder_model_id(
-                        self.config.models_path / stanza.path
-                    )
-                except OSError:
-                    self.logger.warning(f"Could not determine image encoder for {stanza.path}. Skipping.")
-                    continue
-
-            new_config: AnyModelConfig = ModelsValidator.validate_python(stanza)  # type: ignore # see https://github.com/pydantic/pydantic/discussions/7094
-
-            try:
-                if original_record := self._search_by_path(stanza.path):
-                    key = original_record.key
-                    self.logger.info(f"Updating model {model_name} with information from models.yaml using key {key}")
-                    self._update_model(key, new_config)
-                else:
-                    self.logger.info(f"Adding model {model_name} with key {new_key}")
-                    self._add_model(new_key, new_config)
-            except DuplicateModelException:
-                self.logger.warning(f"Model {model_name} is already in the database")
-            except UnknownModelException:
-                self.logger.warning(f"Model at {stanza.path} could not be found in database")
-
-    def _search_by_path(self, path: Path) -> Optional[AnyModelConfig]:
-        self.cursor.execute(
-            """--sql
-            SELECT config FROM model_config
-            WHERE path=?;
-            """,
-            (str(path),),
-        )
-        results = [ModelConfigFactory.make_config(json.loads(x[0])) for x in self.cursor.fetchall()]
-        return results[0] if results else None
-
-    def _update_model(self, key: str, config: AnyModelConfig) -> None:
-        record = ModelConfigFactory.make_config(config, key=key)  # ensure it is a valid config obect
-        json_serialized = record.model_dump_json()  # and turn it into a json string.
-        self.cursor.execute(
-            """--sql
-            UPDATE model_config
-            SET
-                config=?
-            WHERE id=?;
-            """,
-            (json_serialized, key),
-        )
-        if self.cursor.rowcount == 0:
-            raise UnknownModelException("model not found")
-
-    def _add_model(self, key: str, config: AnyModelConfig) -> None:
-        record = ModelConfigFactory.make_config(config, key=key)  # ensure it is a valid config obect.
-        json_serialized = record.model_dump_json()  # and turn it into a json string.
-        try:
-            self.cursor.execute(
-                """--sql
-                INSERT INTO model_config (
-                   id,
-                   original_hash,
-                   config
-                  )
-                VALUES (?,?,?);
-                """,
-                (
-                    key,
-                    record.hash,
-                    json_serialized,
-                ),
-            )
-        except sqlite3.IntegrityError as exc:
-            raise DuplicateModelException(f"{record.name}: model is already in database") from exc
-
-    def _get_image_encoder_model_id(self, model_path: Path) -> str:
-        with open(model_path / "image_encoder.txt") as f:
-            encoder = f.read()
-        return encoder.strip()

invokeai/app/services/urls/urls_base.py

@@ -8,3 +8,8 @@ class UrlServiceBase(ABC):
     def get_image_url(self, image_name: str, thumbnail: bool = False) -> str:
         """Gets the URL for an image or thumbnail."""
         pass
+
+    @abstractmethod
+    def get_model_image_url(self, model_key: str) -> str:
+        """Gets the URL for a model image"""
+        pass

invokeai/app/services/urls/urls_default.py

@@ -4,8 +4,9 @@ from .urls_base import UrlServiceBase
 
 
 class LocalUrlService(UrlServiceBase):
-    def __init__(self, base_url: str = "api/v1"):
+    def __init__(self, base_url: str = "api/v1", base_url_v2: str = "api/v2"):
         self._base_url = base_url
+        self._base_url_v2 = base_url_v2
 
     def get_image_url(self, image_name: str, thumbnail: bool = False) -> str:
         image_basename = os.path.basename(image_name)
@@ -15,3 +16,6 @@ class LocalUrlService(UrlServiceBase):
             return f"{self._base_url}/images/i/{image_basename}/thumbnail"
 
         return f"{self._base_url}/images/i/{image_basename}/full"
+
+    def get_model_image_url(self, model_key: str) -> str:
+        return f"{self._base_url_v2}/models/i/{model_key}/image"

invokeai/app/util/ti_utils.py

@@ -22,7 +22,7 @@ def generate_ti_list(
     for trigger in extract_ti_triggers_from_prompt(prompt):
         name_or_key = trigger[1:-1]
         try:
-            loaded_model = context.models.load(key=name_or_key)
+            loaded_model = context.models.load(name_or_key)
             model = loaded_model.model
             assert isinstance(model, TextualInversionModelRaw)
             assert loaded_model.config.base == base

invokeai/backend/image_util/depth_anything/__init__.py

@@ -9,13 +9,15 @@ from einops import repeat
 from PIL import Image
 from torchvision.transforms import Compose
 
-from invokeai.app.services.config.config_default import InvokeAIAppConfig
+from invokeai.app.services.config.config_default import get_config
 from invokeai.backend.image_util.depth_anything.model.dpt import DPT_DINOv2
 from invokeai.backend.image_util.depth_anything.utilities.util import NormalizeImage, PrepareForNet, Resize
 from invokeai.backend.util.devices import choose_torch_device
+from invokeai.backend.util.logging import InvokeAILogger
 from invokeai.backend.util.util import download_with_progress_bar
 
-config = InvokeAIAppConfig.get_config()
+config = get_config()
+logger = InvokeAILogger.get_logger(config=config)
 
 DEPTH_ANYTHING_MODELS = {
     "large": {
@@ -54,8 +56,9 @@ class DepthAnythingDetector:
     def __init__(self) -> None:
         self.model = None
         self.model_size: Union[Literal["large", "base", "small"], None] = None
+        self.device = choose_torch_device()
 
-    def load_model(self, model_size=Literal["large", "base", "small"]):
+    def load_model(self, model_size: Literal["large", "base", "small"] = "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)
@@ -71,8 +74,6 @@ class DepthAnythingDetector:
                     self.model = DPT_DINOv2(encoder="vitb", features=128, out_channels=[96, 192, 384, 768])
                 case "large":
                     self.model = DPT_DINOv2(encoder="vitl", features=256, out_channels=[256, 512, 1024, 1024])
-                case _:
-                    raise TypeError("Not a supported model")
 
             self.model.load_state_dict(torch.load(DEPTH_ANYTHING_MODEL_PATH.as_posix(), map_location="cpu"))
             self.model.eval()
@@ -80,20 +81,20 @@ class DepthAnythingDetector:
         self.model.to(choose_torch_device())
         return self.model
 
-    def to(self, device):
-        self.model.to(device)
-        return self
+    def __call__(self, image: Image.Image, resolution: int = 512) -> Image.Image:
+        if not self.model:
+            logger.warn("DepthAnything model was not loaded. Returning original image")
+            return image
 
-    def __call__(self, image, resolution=512, offload=False):
-        image = np.array(image, dtype=np.uint8)
-        image = image[:, :, ::-1] / 255.0
+        np_image = np.array(image, dtype=np.uint8)
+        np_image = np_image[:, :, ::-1] / 255.0
 
-        image_height, image_width = image.shape[:2]
-        image = transform({"image": image})["image"]
-        image = torch.from_numpy(image).unsqueeze(0).to(choose_torch_device())
+        image_height, image_width = np_image.shape[:2]
+        np_image = transform({"image": np_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 +104,4 @@ 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/image_util/dw_openpose/wholebody.py

@@ -6,7 +6,7 @@ import pathlib
 import numpy as np
 import onnxruntime as ort
 
-from invokeai.app.services.config.config_default import InvokeAIAppConfig
+from invokeai.app.services.config.config_default import get_config
 from invokeai.backend.util.devices import choose_torch_device
 from invokeai.backend.util.util import download_with_progress_bar
 
@@ -24,7 +24,7 @@ DWPOSE_MODELS = {
     },
 }
 
-config = InvokeAIAppConfig.get_config()
+config = get_config
 
 
 class Wholebody:

invokeai/backend/image_util/invisible_watermark.py

@@ -10,9 +10,9 @@ from imwatermark import WatermarkEncoder
 from PIL import Image
 
 import invokeai.backend.util.logging as logger
-from invokeai.app.services.config import InvokeAIAppConfig
+from invokeai.app.services.config.config_default import get_config
 
-config = InvokeAIAppConfig.get_config()
+config = get_config()
 
 
 class InvisibleWatermark:
@@ -20,14 +20,8 @@ class InvisibleWatermark:
     Wrapper around InvisibleWatermark module.
     """
 
-    @classmethod
-    def invisible_watermark_available(cls) -> bool:
-        return config.invisible_watermark
-
     @classmethod
     def add_watermark(cls, image: Image.Image, watermark_text: str) -> Image.Image:
-        if not cls.invisible_watermark_available():
-            return image
         logger.debug(f'Applying invisible watermark "{watermark_text}"')
         bgr = cv2.cvtColor(np.array(image.convert("RGB")), cv2.COLOR_RGB2BGR)
         encoder = WatermarkEncoder()

invokeai/backend/image_util/lama.py

@@ -6,7 +6,7 @@ import torch
 from PIL import Image
 
 import invokeai.backend.util.logging as logger
-from invokeai.app.services.config import get_invokeai_config
+from invokeai.app.services.config.config_default import get_config
 from invokeai.backend.util.devices import choose_torch_device
 
 
@@ -29,7 +29,7 @@ def load_jit_model(url_or_path, device):
 class LaMA:
     def __call__(self, input_image: Image.Image, *args: Any, **kwds: Any) -> Any:
         device = choose_torch_device()
-        model_location = get_invokeai_config().models_path / "core/misc/lama/lama.pt"
+        model_location = get_config().models_path / "core/misc/lama/lama.pt"
         model = load_jit_model(model_location, device)
 
         image = np.asarray(input_image.convert("RGB"))

invokeai/backend/image_util/patchmatch.py

@@ -8,9 +8,7 @@ be suppressed or deferred
 import numpy as np
 
 import invokeai.backend.util.logging as logger
-from invokeai.app.services.config import InvokeAIAppConfig
-
-config = InvokeAIAppConfig.get_config()
+from invokeai.app.services.config.config_default import get_config
 
 
 class PatchMatch:
@@ -28,7 +26,7 @@ class PatchMatch:
     def _load_patch_match(self):
         if self.tried_load:
             return
-        if config.try_patchmatch:
+        if get_config().patchmatch:
             from patchmatch import patch_match as pm
 
             if pm.patchmatch_available:

invokeai/backend/image_util/safety_checker.py

@@ -5,15 +5,15 @@ configuration variable, that allows the checker to be supressed.
 """
 
 import numpy as np
+from diffusers.pipelines.stable_diffusion.safety_checker import StableDiffusionSafetyChecker
 from PIL import Image
+from transformers import AutoFeatureExtractor
 
 import invokeai.backend.util.logging as logger
-from invokeai.app.services.config import InvokeAIAppConfig
+from invokeai.app.services.config.config_default import get_config
 from invokeai.backend.util.devices import choose_torch_device
 from invokeai.backend.util.silence_warnings import SilenceWarnings
 
-config = InvokeAIAppConfig.get_config()
-
 CHECKER_PATH = "core/convert/stable-diffusion-safety-checker"
 
 
@@ -31,18 +31,12 @@ class SafetyChecker:
         if cls.tried_load:
             return
 
-        if config.nsfw_checker:
-            try:
-                from diffusers.pipelines.stable_diffusion.safety_checker import StableDiffusionSafetyChecker
-                from transformers import AutoFeatureExtractor
-
-                cls.safety_checker = StableDiffusionSafetyChecker.from_pretrained(config.models_path / CHECKER_PATH)
-                cls.feature_extractor = AutoFeatureExtractor.from_pretrained(config.models_path / CHECKER_PATH)
-                logger.info("NSFW checker initialized")
-            except Exception as e:
-                logger.warning(f"Could not load NSFW checker: {str(e)}")
-        else:
-            logger.info("NSFW checker loading disabled")
+        try:
+            cls.safety_checker = StableDiffusionSafetyChecker.from_pretrained(get_config().models_path / CHECKER_PATH)
+            cls.feature_extractor = AutoFeatureExtractor.from_pretrained(get_config().models_path / CHECKER_PATH)
+            logger.info("NSFW checker initialized")
+        except Exception as e:
+            logger.warning(f"Could not load NSFW checker: {str(e)}")
         cls.tried_load = True
 
     @classmethod
@@ -54,7 +48,8 @@ class SafetyChecker:
     def has_nsfw_concept(cls, image: Image.Image) -> bool:
         if not cls.safety_checker_available():
             return False
-
+        assert cls.safety_checker is not None
+        assert cls.feature_extractor is not None
         device = choose_torch_device()
         features = cls.feature_extractor([image], return_tensors="pt")
         features.to(device)

invokeai/backend/image_util/txt2mask.py

@@ -33,11 +33,10 @@ from PIL import Image, ImageOps
 from transformers import AutoProcessor, CLIPSegForImageSegmentation
 
 import invokeai.backend.util.logging as logger
-from invokeai.app.services.config import InvokeAIAppConfig
+from invokeai.app.services.config.config_default import get_config
 
 CLIPSEG_MODEL = "CIDAS/clipseg-rd64-refined"
 CLIPSEG_SIZE = 352
-config = InvokeAIAppConfig.get_config()
 
 
 class SegmentedGrayscale(object):
@@ -78,8 +77,8 @@ class Txt2Mask(object):
 
         # BUG: we are not doing anything with the device option at this time
         self.device = device
-        self.processor = AutoProcessor.from_pretrained(CLIPSEG_MODEL, cache_dir=config.cache_dir)
-        self.model = CLIPSegForImageSegmentation.from_pretrained(CLIPSEG_MODEL, cache_dir=config.cache_dir)
+        self.processor = AutoProcessor.from_pretrained(CLIPSEG_MODEL, cache_dir=get_config().cache_dir)
+        self.model = CLIPSegForImageSegmentation.from_pretrained(CLIPSEG_MODEL, cache_dir=get_config().cache_dir)
 
     @torch.no_grad()
     def segment(self, image: Image.Image, prompt: str) -> SegmentedGrayscale:

invokeai/backend/install/check_directories.py

@@ -0,0 +1,30 @@
+"""
+Check that the invokeai_root is correctly configured and exit if not.
+"""
+
+import sys
+
+from invokeai.app.services.config import InvokeAIAppConfig
+
+
+# TODO(psyche): Should this also check for things like ESRGAN models, database, etc?
+def validate_directories(config: InvokeAIAppConfig) -> None:
+    assert config.db_path.parent.exists(), f"{config.db_path.parent} not found"
+    assert config.models_path.exists(), f"{config.models_path} not found"
+
+
+def check_directories(config: InvokeAIAppConfig):
+    try:
+        validate_directories(config)
+    except Exception as e:
+        print()
+        print(f"An exception has occurred: {str(e)}")
+        print("== STARTUP ABORTED ==")
+        print("** One or more necessary files is missing from your InvokeAI directories **")
+        print("** Please rerun the configuration script to fix this problem. **")
+        print("** From the launcher, selection option [6]. **")
+        print(
+            '** From the command line, activate the virtual environment and run "invokeai-configure --yes --skip-sd-weights" **'
+        )
+        input("Press any key to continue...")
+        sys.exit(0)

invokeai/backend/install/check_root.py

@@ -1,41 +0,0 @@
-"""
-Check that the invokeai_root is correctly configured and exit if not.
-"""
-
-import sys
-
-from invokeai.app.services.config import InvokeAIAppConfig
-
-
-def check_invokeai_root(config: InvokeAIAppConfig):
-    try:
-        assert config.db_path.parent.exists(), f"{config.db_path.parent} not found"
-        assert config.models_path.exists(), f"{config.models_path} not found"
-        if not config.ignore_missing_core_models:
-            for model in [
-                "CLIP-ViT-bigG-14-laion2B-39B-b160k",
-                "bert-base-uncased",
-                "clip-vit-large-patch14",
-                "sd-vae-ft-mse",
-                "stable-diffusion-2-clip",
-                "stable-diffusion-safety-checker",
-            ]:
-                path = config.models_path / f"core/convert/{model}"
-                assert path.exists(), f"{path} is missing"
-    except Exception as e:
-        print()
-        print(f"An exception has occurred: {str(e)}")
-        print("== STARTUP ABORTED ==")
-        print("** One or more necessary files is missing from your InvokeAI root directory **")
-        print("** Please rerun the configuration script to fix this problem. **")
-        print("** From the launcher, selection option [6]. **")
-        print(
-            '** From the command line, activate the virtual environment and run "invokeai-configure --yes --skip-sd-weights" **'
-        )
-        print(
-            '** (To skip this check completely, add "--ignore_missing_core_models" to your CLI args. Not installing '
-            "these core models will prevent the loading of some or all .safetensors and .ckpt files. However, you can "
-            "always come back and install these core models in the future.)"
-        )
-        input("Press any key to continue...")
-        sys.exit(0)

invokeai/backend/install/install_helper.py

@@ -19,7 +19,6 @@ from invokeai.app.services.model_install import (
     ModelInstallService,
     ModelInstallServiceBase,
 )
-from invokeai.app.services.model_metadata import ModelMetadataStoreSQL
 from invokeai.app.services.model_records import ModelRecordServiceBase, ModelRecordServiceSQL
 from invokeai.app.services.shared.sqlite.sqlite_util import init_db
 from invokeai.backend.model_manager import (
@@ -37,9 +36,9 @@ INITIAL_MODELS = "INITIAL_MODELS.yaml"
 def initialize_record_store(app_config: InvokeAIAppConfig) -> ModelRecordServiceBase:
     """Return an initialized ModelConfigRecordServiceBase object."""
     logger = InvokeAILogger.get_logger(config=app_config)
-    image_files = DiskImageFileStorage(f"{app_config.output_path}/images")
+    image_files = DiskImageFileStorage(f"{app_config.outputs_path}/images")
     db = init_db(config=app_config, logger=logger, image_files=image_files)
-    obj: ModelRecordServiceBase = ModelRecordServiceSQL(db, ModelMetadataStoreSQL(db))
+    obj: ModelRecordServiceBase = ModelRecordServiceSQL(db)
     return obj
 
 
@@ -150,7 +149,7 @@ class InstallHelper(object):
         """
         # previously-installed models
         for model in self._installer.record_store.all_models():
-            info = UnifiedModelInfo.parse_obj(model.dict())
+            info = UnifiedModelInfo.model_validate(model.model_dump())
             info.installed = True
             model_key = f"{model.base.value}/{model.type.value}/{model.name}"
             self.all_models[model_key] = info
@@ -184,7 +183,7 @@ class InstallHelper(object):
 
         # previously-installed models
         for model in self._installer.record_store.all_models():
-            info = UnifiedModelInfo.parse_obj(model.dict())
+            info = UnifiedModelInfo.model_validate(model.model_dump())
             info.installed = True
             model_key = f"{model.base.value}/{model.type.value}/{model.name}"
             self.all_models[model_key] = info

invokeai/backend/install/invokeai_configure.py

@@ -6,7 +6,6 @@
 #
 # Coauthor: Kevin Turner http://github.com/keturn
 #
-import argparse
 import io
 import os
 import shutil
@@ -17,28 +16,25 @@ import warnings
 from argparse import Namespace
 from enum import Enum
 from pathlib import Path
-from shutil import get_terminal_size
-from typing import Any, Optional, Set, Tuple, Type, get_args, get_type_hints
+from shutil import copy, get_terminal_size, move
+from typing import Any, Optional, Tuple, Type, get_args, get_type_hints
 from urllib import request
 
 import npyscreen
 import psutil
 import torch
 import transformers
-from diffusers import AutoencoderKL, ModelMixin
+from diffusers import ModelMixin
 from diffusers.pipelines.stable_diffusion.safety_checker import StableDiffusionSafetyChecker
 from huggingface_hub import HfFolder
 from huggingface_hub import login as hf_hub_login
-from omegaconf import DictConfig, OmegaConf
-from pydantic.error_wrappers import ValidationError
 from tqdm import tqdm
-from transformers import AutoFeatureExtractor, BertTokenizerFast, CLIPTextConfig, CLIPTextModel, CLIPTokenizer
+from transformers import AutoFeatureExtractor
 
-import invokeai.configs as configs
-from invokeai.app.services.config import InvokeAIAppConfig
+import invokeai.configs as model_configs
+from invokeai.app.services.config import InvokeAIAppConfig, get_config
 from invokeai.backend.install.install_helper import InstallHelper, InstallSelections
-from invokeai.backend.install.legacy_arg_parsing import legacy_parser
-from invokeai.backend.model_manager import BaseModelType, ModelType
+from invokeai.backend.model_manager import ModelType
 from invokeai.backend.util import choose_precision, choose_torch_device
 from invokeai.backend.util.logging import InvokeAILogger
 from invokeai.frontend.install.model_install import addModelsForm
@@ -65,12 +61,7 @@ def get_literal_fields(field: str) -> Tuple[Any]:
 
 
 # --------------------------globals-----------------------
-
-config = InvokeAIAppConfig.get_config()
-
-Model_dir = "models"
-Default_config_file = config.model_conf_path
-SD_Configs = config.legacy_conf_path
+config = None
 
 PRECISION_CHOICES = get_literal_fields("precision")
 DEVICE_CHOICES = get_literal_fields("device")
@@ -104,7 +95,7 @@ class DummyWidgetValue(Enum):
 
 
 # --------------------------------------------
-def postscript(errors: Set[str]) -> None:
+def postscript(errors: set[str]) -> None:
     if not any(errors):
         message = f"""
 ** INVOKEAI INSTALLATION SUCCESSFUL **
@@ -195,7 +186,7 @@ def hf_download_from_pretrained(model_class: Type[ModelMixin], model_name: str,
 
 
 # ---------------------------------------------
-def download_with_progress_bar(model_url: str, model_dest: str, label: str = "the"):
+def download_with_progress_bar(model_url: str, model_dest: str | Path, label: str = "the"):
     try:
         logger.info(f"Installing {label} model file {model_url}...")
         if not os.path.exists(model_dest):
@@ -210,51 +201,15 @@ def download_with_progress_bar(model_url: str, model_dest: str, label: str = "th
         print(traceback.format_exc(), file=sys.stderr)
 
 
-def download_conversion_models():
+def download_safety_checker():
     target_dir = config.models_path / "core/convert"
     kwargs = {}  # for future use
     try:
-        logger.info("Downloading core tokenizers and text encoders")
-
-        # bert
-        with warnings.catch_warnings():
-            warnings.filterwarnings("ignore", category=DeprecationWarning)
-            bert = BertTokenizerFast.from_pretrained("bert-base-uncased", **kwargs)
-            bert.save_pretrained(target_dir / "bert-base-uncased", safe_serialization=True)
-
-        # sd-1
-        repo_id = "openai/clip-vit-large-patch14"
-        hf_download_from_pretrained(CLIPTokenizer, repo_id, target_dir / "clip-vit-large-patch14")
-        hf_download_from_pretrained(CLIPTextModel, repo_id, target_dir / "clip-vit-large-patch14")
-
-        # sd-2
-        repo_id = "stabilityai/stable-diffusion-2"
-        pipeline = CLIPTokenizer.from_pretrained(repo_id, subfolder="tokenizer", **kwargs)
-        pipeline.save_pretrained(target_dir / "stable-diffusion-2-clip" / "tokenizer", safe_serialization=True)
-
-        pipeline = CLIPTextModel.from_pretrained(repo_id, subfolder="text_encoder", **kwargs)
-        pipeline.save_pretrained(target_dir / "stable-diffusion-2-clip" / "text_encoder", safe_serialization=True)
-
-        # sd-xl - tokenizer_2
-        repo_id = "laion/CLIP-ViT-bigG-14-laion2B-39B-b160k"
-        _, model_name = repo_id.split("/")
-        pipeline = CLIPTokenizer.from_pretrained(repo_id, **kwargs)
-        pipeline.save_pretrained(target_dir / model_name, safe_serialization=True)
-
-        pipeline = CLIPTextConfig.from_pretrained(repo_id, **kwargs)
-        pipeline.save_pretrained(target_dir / model_name, safe_serialization=True)
-
-        # VAE
-        logger.info("Downloading stable diffusion VAE")
-        vae = AutoencoderKL.from_pretrained("stabilityai/sd-vae-ft-mse", **kwargs)
-        vae.save_pretrained(target_dir / "sd-vae-ft-mse", safe_serialization=True)
-
         # safety checking
         logger.info("Downloading safety checker")
         repo_id = "CompVis/stable-diffusion-safety-checker"
         pipeline = AutoFeatureExtractor.from_pretrained(repo_id, **kwargs)
         pipeline.save_pretrained(target_dir / "stable-diffusion-safety-checker", safe_serialization=True)
-
         pipeline = StableDiffusionSafetyChecker.from_pretrained(repo_id, **kwargs)
         pipeline.save_pretrained(target_dir / "stable-diffusion-safety-checker", safe_serialization=True)
     except KeyboardInterrupt:
@@ -307,7 +262,7 @@ def download_lama():
 def download_support_models() -> None:
     download_realesrgan()
     download_lama()
-    download_conversion_models()
+    download_safety_checker()
 
 
 # -------------------------------------
@@ -328,7 +283,7 @@ class editOptsForm(CyclingForm, npyscreen.FormMultiPage):
 
     def create(self):
         program_opts = self.parentApp.program_opts
-        old_opts = self.parentApp.invokeai_opts
+        old_opts: InvokeAIAppConfig = self.parentApp.invokeai_opts
         first_time = not (config.root_path / "invokeai.yaml").exists()
         access_token = HfFolder.get_token()
         window_width, window_height = get_terminal_size()
@@ -502,7 +457,7 @@ Use cursor arrows to make a checkbox selection, and space to toggle.
         self.nextrely -= 1
         self.ram = self.add_widget_intelligent(
             npyscreen.Slider,
-            value=clip(old_opts.ram_cache_size, range=(3.0, MAX_RAM), step=0.5),
+            value=clip(old_opts.ram, range=(3.0, MAX_RAM), step=0.5),
             out_of=round(MAX_RAM),
             lowest=0.0,
             step=0.5,
@@ -522,7 +477,7 @@ Use cursor arrows to make a checkbox selection, and space to toggle.
             self.nextrely -= 1
             self.vram = self.add_widget_intelligent(
                 npyscreen.Slider,
-                value=clip(old_opts.vram_cache_size, range=(0, MAX_VRAM), step=0.25),
+                value=clip(old_opts.vram, range=(0, MAX_VRAM), step=0.25),
                 out_of=round(MAX_VRAM * 2) / 2,
                 lowest=0.0,
                 relx=8,
@@ -557,7 +512,7 @@ Use cursor arrows to make a checkbox selection, and space to toggle.
         self.autoimport_dirs["autoimport_dir"] = self.add_widget_intelligent(
             FileBox,
             name="Optional folder to scan for new checkpoints, ControlNets, LoRAs and TI models",
-            value=str(config.root_path / config.autoimport_dir) if config.autoimport_dir else "",
+            value=str(config.autoimport_path),
             select_dir=True,
             must_exist=False,
             use_two_lines=False,
@@ -694,7 +649,7 @@ class EditOptApplication(npyscreen.NPSAppManaged):
             )
 
 
-def default_ramcache() -> float:
+def get_default_ram_cache_size() -> float:
     """Run a heuristic for the default RAM cache based on installed RAM."""
 
     # Note that on my 64 GB machine, psutil.virtual_memory().total gives 62 GB,
@@ -704,11 +659,12 @@ def default_ramcache() -> float:
     )  # 2.1 is just large enough for sd 1.5 ;-)
 
 
-def default_startup_options(init_file: Path) -> InvokeAIAppConfig:
-    opts = InvokeAIAppConfig.get_config()
-    opts.ram = default_ramcache()
-    opts.precision = "float32" if FORCE_FULL_PRECISION else choose_precision(torch.device(choose_torch_device()))
-    return opts
+def get_default_config() -> InvokeAIAppConfig:
+    """Builds a new config object, setting the ram and precision using the appropriate heuristic."""
+    config = InvokeAIAppConfig()
+    config.ram = get_default_ram_cache_size()
+    config.precision = "float32" if FORCE_FULL_PRECISION else choose_precision(torch.device(choose_torch_device()))
+    return config
 
 
 def default_user_selections(program_opts: Namespace, install_helper: InstallHelper) -> InstallSelections:
@@ -738,34 +694,30 @@ def initialize_rootdir(root: Path, yes_to_all: bool = False):
     for model_type in ModelType:
         Path(root, "autoimport", model_type.value).mkdir(parents=True, exist_ok=True)
 
-    configs_src = Path(configs.__path__[0])
+    configs_src = Path(model_configs.__path__[0])
     configs_dest = root / "configs"
     if not os.path.samefile(configs_src, configs_dest):
         shutil.copytree(configs_src, configs_dest, dirs_exist_ok=True)
 
     dest = root / "models"
-    for model_base in BaseModelType:
-        for model_type in ModelType:
-            path = dest / model_base.value / model_type.value
-            path.mkdir(parents=True, exist_ok=True)
-    path = dest / "core"
-    path.mkdir(parents=True, exist_ok=True)
+    dest.mkdir(parents=True, exist_ok=True)
 
 
 # -------------------------------------
 def run_console_ui(
-    program_opts: Namespace, initfile: Path, install_helper: InstallHelper
+    program_opts: Namespace, install_helper: InstallHelper
 ) -> Tuple[Optional[Namespace], Optional[InstallSelections]]:
-    first_time = not (config.root_path / "invokeai.yaml").exists()
-    invokeai_opts = default_startup_options(initfile) if first_time else config
-    invokeai_opts.root = program_opts.root
+    first_time = not config.init_file_path.exists()
+    config_opts = get_default_config() if first_time else config
+    if program_opts.root:
+        config_opts.set_root(Path(program_opts.root))
 
     if not set_min_terminal_size(MIN_COLS, MIN_LINES):
         raise WindowTooSmallException(
             "Could not increase terminal size. Try running again with a larger window or smaller font size."
         )
 
-    editApp = EditOptApplication(program_opts, invokeai_opts, install_helper)
+    editApp = EditOptApplication(program_opts, config_opts, install_helper)
     editApp.run()
     if editApp.user_cancelled:
         return (None, None)
@@ -773,167 +725,54 @@ def run_console_ui(
         return (editApp.new_opts, editApp.install_selections)
 
 
-# -------------------------------------
-def write_opts(opts: InvokeAIAppConfig, init_file: Path) -> None:
-    """
-    Update the invokeai.yaml file with values from current settings.
-    """
-    # this will load current settings
-    new_config = InvokeAIAppConfig.get_config()
-    new_config.root = config.root
-
-    for key, value in vars(opts).items():
-        if hasattr(new_config, key):
-            setattr(new_config, key, value)
-
-    with open(init_file, "w", encoding="utf-8") as file:
-        file.write(new_config.to_yaml())
-
-    if hasattr(opts, "hf_token") and opts.hf_token:
-        HfLogin(opts.hf_token)
-
-
 # -------------------------------------
 def default_output_dir() -> Path:
     return config.root_path / "outputs"
 
 
-# -------------------------------------
-def write_default_options(program_opts: Namespace, initfile: Path) -> None:
-    opt = default_startup_options(initfile)
-    write_opts(opt, initfile)
-
-
-# -------------------------------------
-# Here we bring in
-# the legacy Args object in order to parse
-# the old init file and write out the new
-# yaml format.
-def migrate_init_file(legacy_format: Path) -> None:
-    old = legacy_parser.parse_args([f"@{str(legacy_format)}"])
-    new = InvokeAIAppConfig.get_config()
-
-    for attr in InvokeAIAppConfig.model_fields.keys():
-        if hasattr(old, attr):
-            try:
-                setattr(new, attr, getattr(old, attr))
-            except ValidationError as e:
-                print(f"* Ignoring incompatible value for field {attr}:\n  {str(e)}")
-
-    # a few places where the field names have changed and we have to
-    # manually add in the new names/values
-    new.xformers_enabled = old.xformers
-    new.conf_path = old.conf
-    new.root = legacy_format.parent.resolve()
-
-    invokeai_yaml = legacy_format.parent / "invokeai.yaml"
-    with open(invokeai_yaml, "w", encoding="utf-8") as outfile:
-        outfile.write(new.to_yaml())
-
-    legacy_format.replace(legacy_format.parent / "invokeai.init.orig")
-
-
-# -------------------------------------
-def migrate_models(root: Path) -> None:
-    from invokeai.backend.install.migrate_to_3 import do_migrate
-
-    do_migrate(root, root)
-
-
-def migrate_if_needed(opt: Namespace, root: Path) -> bool:
+def is_v2_install(root: Path) -> bool:
     # We check for to see if the runtime directory is correctly initialized.
     old_init_file = root / "invokeai.init"
     new_init_file = root / "invokeai.yaml"
     old_hub = root / "models/hub"
-    migration_needed = (old_init_file.exists() and not new_init_file.exists()) and old_hub.exists()
-
-    if migration_needed:
-        if opt.yes_to_all or yes_or_no(
-            f"{str(config.root_path)} appears to be a 2.3 format root directory. Convert to version 3.0?"
-        ):
-            logger.info("** Migrating invokeai.init to invokeai.yaml")
-            migrate_init_file(old_init_file)
-            omegaconf = OmegaConf.load(new_init_file)
-            assert isinstance(omegaconf, DictConfig)
-            config.parse_args(argv=[], conf=omegaconf)
-
-            if old_hub.exists():
-                migrate_models(config.root_path)
-        else:
-            print("Cannot continue without conversion. Aborting.")
-
-    return migration_needed
+    is_v2 = (old_init_file.exists() and not new_init_file.exists()) and old_hub.exists()
+    return is_v2
 
 
 # -------------------------------------
-def main() -> None:
+def main(opt: Namespace) -> None:
     global FORCE_FULL_PRECISION  # FIXME
-    parser = argparse.ArgumentParser(description="InvokeAI model downloader")
-    parser.add_argument(
-        "--skip-sd-weights",
-        dest="skip_sd_weights",
-        action=argparse.BooleanOptionalAction,
-        default=False,
-        help="skip downloading the large Stable Diffusion weight files",
-    )
-    parser.add_argument(
-        "--skip-support-models",
-        dest="skip_support_models",
-        action=argparse.BooleanOptionalAction,
-        default=False,
-        help="skip downloading the support models",
-    )
-    parser.add_argument(
-        "--full-precision",
-        dest="full_precision",
-        action=argparse.BooleanOptionalAction,
-        type=bool,
-        default=False,
-        help="use 32-bit weights instead of faster 16-bit weights",
-    )
-    parser.add_argument(
-        "--yes",
-        "-y",
-        dest="yes_to_all",
-        action="store_true",
-        help='answer "yes" to all prompts',
-    )
-    parser.add_argument(
-        "--default_only",
-        action="store_true",
-        help="when --yes specified, only install the default model",
-    )
-    parser.add_argument(
-        "--config_file",
-        "-c",
-        dest="config_file",
-        type=str,
-        default=None,
-        help="path to configuration file to create",
-    )
-    parser.add_argument(
-        "--root_dir",
-        dest="root",
-        type=str,
-        default=None,
-        help="path to root of install directory",
-    )
-    opt = parser.parse_args()
-    invoke_args = []
-    if opt.root:
-        invoke_args.extend(["--root", opt.root])
-    if opt.full_precision:
-        invoke_args.extend(["--precision", "float32"])
-    config.parse_args(invoke_args)
-    logger = InvokeAILogger().get_logger(config=config)
+    global config
 
-    errors = set()
-    FORCE_FULL_PRECISION = opt.full_precision  # FIXME global
+    updates: dict[str, Any] = {}
+
+    config = get_config()
+    if opt.full_precision:
+        updates["precision"] = "float32"
 
     try:
-        # if we do a root migration/upgrade, then we are keeping previous
-        # configuration and we are done.
-        if migrate_if_needed(opt, config.root_path):
+        # Attempt to read the config file into the config object
+        config.merge_from_file()
+    except FileNotFoundError:
+        # No config file, first time running the app
+        pass
+
+    config.update_config(updates)
+    logger = InvokeAILogger().get_logger(config=config)
+
+    errors: set[str] = set()
+    FORCE_FULL_PRECISION = opt.full_precision  # FIXME global
+
+    # Before we write anything else, make a backup of the existing init file
+    new_init_file = config.init_file_path
+    backup_init_file = new_init_file.with_suffix(".bak")
+    if new_init_file.exists():
+        copy(new_init_file, backup_init_file)
+
+    try:
+        # v2.3 -> v4.0.0 upgrade is no longer supported
+        if is_v2_install(config.root_path):
+            logger.error("Migration from v2.3 to v4.0.0 is no longer supported. Please install a fresh copy.")
             sys.exit(0)
 
         # run this unconditionally in case new directories need to be added
@@ -943,16 +782,22 @@ def main() -> None:
         install_helper = InstallHelper(config, logger)
 
         models_to_download = default_user_selections(opt, install_helper)
-        new_init_file = config.root_path / "invokeai.yaml"
 
         if opt.yes_to_all:
-            write_default_options(opt, new_init_file)
-            init_options = Namespace(precision="float32" if opt.full_precision else "float16")
-
+            # We will not show the UI - just write the default config to the file and move on to installing models.
+            get_default_config().write_file(new_init_file)
         else:
-            init_options, models_to_download = run_console_ui(opt, new_init_file, install_helper)
-            if init_options:
-                write_opts(init_options, new_init_file)
+            # Run the UI to get the user's options & model choices
+            user_opts, models_to_download = run_console_ui(opt, install_helper)
+            if user_opts:
+                # Create a dict of the user's opts, omitting any fields that are not config settings (like `hf_token`)
+                user_opts_dict = {k: v for k, v in vars(user_opts).items() if k in config.model_fields}
+                # Merge the user's opts back into the config object & write it
+                config.update_config(user_opts_dict)
+                config.write_file(config.init_file_path)
+
+                if hasattr(user_opts, "hf_token") and user_opts.hf_token:
+                    HfLogin(user_opts.hf_token)
             else:
                 logger.info('\n** CANCELLED AT USER\'S REQUEST. USE THE "invoke.sh" LAUNCHER TO RUN LATER **\n')
                 sys.exit(0)
@@ -965,7 +810,6 @@ def main() -> None:
 
         if opt.skip_sd_weights:
             logger.warning("Skipping diffusion weights download per user request")
-
         elif models_to_download:
             install_helper.add_or_delete(models_to_download)
 
@@ -975,8 +819,17 @@ def main() -> None:
             input("Press any key to continue...")
     except WindowTooSmallException as e:
         logger.error(str(e))
+        if backup_init_file.exists():
+            move(backup_init_file, new_init_file)
     except KeyboardInterrupt:
         print("\nGoodbye! Come back soon.")
+        if backup_init_file.exists():
+            move(backup_init_file, new_init_file)
+    except Exception:
+        print("An error occurred during installation.")
+        if backup_init_file.exists():
+            move(backup_init_file, new_init_file)
+        print(traceback.format_exc(), file=sys.stderr)
 
 
 # -------------------------------------

invokeai/backend/model_hash/model_hash.py

@@ -1,12 +1,4 @@
 # Copyright (c) 2023 Lincoln D. Stein and the InvokeAI Development Team
-"""
-Fast hashing of diffusers and checkpoint-style models.
-
-Usage:
-from invokeai.backend.model_managre.model_hash import FastModelHash
->>> FastModelHash.hash('/home/models/stable-diffusion-v1.5')
-'a8e693a126ea5b831c96064dc569956f'
-"""
 
 import hashlib
 import os
@@ -15,9 +7,9 @@ from typing import Callable, Literal, Optional, Union
 
 from blake3 import blake3
 
-MODEL_FILE_EXTENSIONS = (".ckpt", ".safetensors", ".bin", ".pt", ".pth")
+from invokeai.app.util.misc import uuid_string
 
-ALGORITHM = Literal[
+HASHING_ALGORITHMS = Literal[
     "md5",
     "sha1",
     "sha224",
@@ -33,12 +25,15 @@ ALGORITHM = Literal[
     "shake_128",
     "shake_256",
     "blake3",
+    "blake3_single",
+    "random",
 ]
+MODEL_FILE_EXTENSIONS = (".ckpt", ".safetensors", ".bin", ".pt", ".pth")
 
 
 class ModelHash:
     """
-    Creates a hash of a model using a specified algorithm.
+    Creates a hash of a model using a specified algorithm. The hash is prefixed by the algorithm used.
 
     Args:
         algorithm: Hashing algorithm to use. Defaults to BLAKE3.
@@ -53,20 +48,29 @@ class ModelHash:
     The final hash is computed by hashing the hashes of all model files in the directory using BLAKE3, ensuring
     that directory hashes are never weaker than the file hashes.
 
+    A convenience algorithm choice of "random" is also available, which returns a random string. This is not a hash.
+
     Usage:
         ```py
         # BLAKE3 hash
-        ModelHash().hash("path/to/some/model.safetensors")
+        ModelHash().hash("path/to/some/model.safetensors") # "blake3:ce3f0c5f3c05d119f4a5dcaf209b50d3149046a0d3a9adee9fed4c83cad6b4d0"
         # MD5
-        ModelHash("md5").hash("path/to/model/dir/")
+        ModelHash("md5").hash("path/to/model/dir/") # "md5:a0cd925fc063f98dbf029eee315060c3"
         ```
     """
 
-    def __init__(self, algorithm: ALGORITHM = "blake3", file_filter: Optional[Callable[[str], bool]] = None) -> None:
+    def __init__(
+        self, algorithm: HASHING_ALGORITHMS = "blake3", file_filter: Optional[Callable[[str], bool]] = None
+    ) -> None:
+        self.algorithm: HASHING_ALGORITHMS = algorithm
         if algorithm == "blake3":
             self._hash_file = self._blake3
+        elif algorithm == "blake3_single":
+            self._hash_file = self._blake3_single
         elif algorithm in hashlib.algorithms_available:
             self._hash_file = self._get_hashlib(algorithm)
+        elif algorithm == "random":
+            self._hash_file = self._random
         else:
             raise ValueError(f"Algorithm {algorithm} not available")
 
@@ -87,10 +91,12 @@ class ModelHash:
         """
 
         model_path = Path(model_path)
+        # blake3_single is a single-threaded version of blake3, prefix should still be "blake3:"
+        prefix = self._get_prefix(self.algorithm)
         if model_path.is_file():
-            return self._hash_file(model_path)
+            return prefix + self._hash_file(model_path)
         elif model_path.is_dir():
-            return self._hash_dir(model_path)
+            return prefix + self._hash_dir(model_path)
         else:
             raise OSError(f"Not a valid file or directory: {model_path}")
 
@@ -114,6 +120,7 @@ class ModelHash:
         composite_hasher = blake3()
         for h in component_hashes:
             composite_hasher.update(h.encode("utf-8"))
+
         return composite_hasher.hexdigest()
 
     @staticmethod
@@ -137,7 +144,7 @@ class ModelHash:
 
     @staticmethod
     def _blake3(file_path: Path) -> str:
-        """Hashes a file using BLAKE3
+        """Hashes a file using BLAKE3, using parallelized and memory-mapped I/O to avoid reading the entire file into memory.
 
         Args:
             file_path: Path to the file to hash
@@ -150,7 +157,21 @@ class ModelHash:
         return file_hasher.hexdigest()
 
     @staticmethod
-    def _get_hashlib(algorithm: ALGORITHM) -> Callable[[Path], str]:
+    def _blake3_single(file_path: Path) -> str:
+        """Hashes a file using BLAKE3, without parallelism. Suitable for spinning hard drives.
+
+        Args:
+            file_path: Path to the file to hash
+
+        Returns:
+            Hexdigest of the hash of the file
+        """
+        file_hasher = blake3()
+        file_hasher.update_mmap(file_path)
+        return file_hasher.hexdigest()
+
+    @staticmethod
+    def _get_hashlib(algorithm: HASHING_ALGORITHMS) -> Callable[[Path], str]:
         """Factory function that returns a function to hash a file with the given algorithm.
 
         Args:
@@ -172,6 +193,13 @@ class ModelHash:
 
         return hashlib_hasher
 
+    @staticmethod
+    def _random(_file_path: Path) -> str:
+        """Returns a random string. This is not a hash.
+
+        The string is a UUID, hashed with BLAKE3 to ensure that it is unique."""
+        return blake3(uuid_string().encode()).hexdigest()
+
     @staticmethod
     def _default_file_filter(file_path: str) -> bool:
         """A default file filter that only includes files with the following extensions: .ckpt, .safetensors, .bin, .pt, .pth
@@ -183,3 +211,9 @@ class ModelHash:
             True if the file matches the given extensions, otherwise False
         """
         return file_path.endswith(MODEL_FILE_EXTENSIONS)
+
+    @staticmethod
+    def _get_prefix(algorithm: HASHING_ALGORITHMS) -> str:
+        """Return the prefix for the given algorithm, e.g. \"blake3:\" or \"md5:\"."""
+        # blake3_single is a single-threaded version of blake3, prefix should still be "blake3:"
+        return "blake3:" if algorithm == "blake3_single" else f"{algorithm}:"

invokeai/backend/model_manager/config.py

@@ -22,7 +22,7 @@ Validation errors will raise an InvalidModelConfigException error.
 
 import time
 from enum import Enum
-from typing import Literal, Optional, Type, Union
+from typing import Literal, Optional, Type, TypeAlias, Union
 
 import torch
 from diffusers.models.modeling_utils import ModelMixin
@@ -129,16 +129,27 @@ class ModelSourceType(str, Enum):
     Path = "path"
     Url = "url"
     HFRepoID = "hf_repo_id"
-    CivitAI = "civitai"
 
 
-class ModelDefaultSettings(BaseModel):
-    vae: str | None
-    vae_precision: str | None
-    scheduler: SCHEDULER_NAME_VALUES | None
-    steps: int | None
-    cfg_scale: float | None
-    cfg_rescale_multiplier: float | None
+DEFAULTS_PRECISION = Literal["fp16", "fp32"]
+
+
+class MainModelDefaultSettings(BaseModel):
+    vae: str | None = Field(default=None, description="Default VAE for this model (model key)")
+    vae_precision: DEFAULTS_PRECISION | None = Field(default=None, description="Default VAE precision for this model")
+    scheduler: SCHEDULER_NAME_VALUES | None = Field(default=None, description="Default scheduler for this model")
+    steps: int | None = Field(default=None, gt=0, description="Default number of steps for this model")
+    cfg_scale: float | None = Field(default=None, ge=1, description="Default CFG Scale for this model")
+    cfg_rescale_multiplier: float | None = Field(
+        default=None, ge=0, lt=1, description="Default CFG Rescale Multiplier for this model"
+    )
+    width: int | None = Field(default=None, multiple_of=8, ge=64, description="Default width for this model")
+    height: int | None = Field(default=None, multiple_of=8, ge=64, description="Default height for this model")
+
+
+class ControlAdapterDefaultSettings(BaseModel):
+    # This could be narrowed to controlnet processor nodes, but they change. Leaving this a string is safer.
+    preprocessor: str | None
 
 
 class ModelConfigBase(BaseModel):
@@ -157,10 +168,7 @@ class ModelConfigBase(BaseModel):
     source_api_response: Optional[str] = Field(
         description="The original API response from the source, as stringified JSON.", default=None
     )
-    trigger_phrases: Optional[set[str]] = Field(description="Set of trigger phrases for this model", default=None)
-    default_settings: Optional[ModelDefaultSettings] = Field(
-        description="Default settings for this model", default=None
-    )
+    cover_image: Optional[str] = Field(description="Url for image to preview model", default=None)
 
     @staticmethod
     def json_schema_extra(schema: dict[str, Any], model_class: Type[BaseModel]) -> None:
@@ -186,10 +194,14 @@ class DiffusersConfigBase(ModelConfigBase):
     repo_variant: Optional[ModelRepoVariant] = ModelRepoVariant.Default
 
 
-class LoRALyCORISConfig(ModelConfigBase):
+class LoRAConfigBase(ModelConfigBase):
+    type: Literal[ModelType.LoRA] = ModelType.LoRA
+    trigger_phrases: Optional[set[str]] = Field(description="Set of trigger phrases for this model", default=None)
+
+
+class LoRALyCORISConfig(LoRAConfigBase):
     """Model config for LoRA/Lycoris models."""
 
-    type: Literal[ModelType.LoRA] = ModelType.LoRA
     format: Literal[ModelFormat.LyCORIS] = ModelFormat.LyCORIS
 
     @staticmethod
@@ -197,10 +209,9 @@ class LoRALyCORISConfig(ModelConfigBase):
         return Tag(f"{ModelType.LoRA.value}.{ModelFormat.LyCORIS.value}")
 
 
-class LoRADiffusersConfig(ModelConfigBase):
+class LoRADiffusersConfig(LoRAConfigBase):
     """Model config for LoRA/Diffusers models."""
 
-    type: Literal[ModelType.LoRA] = ModelType.LoRA
     format: Literal[ModelFormat.Diffusers] = ModelFormat.Diffusers
 
     @staticmethod
@@ -230,7 +241,13 @@ class VAEDiffusersConfig(ModelConfigBase):
         return Tag(f"{ModelType.VAE.value}.{ModelFormat.Diffusers.value}")
 
 
-class ControlNetDiffusersConfig(DiffusersConfigBase):
+class ControlAdapterConfigBase(BaseModel):
+    default_settings: Optional[ControlAdapterDefaultSettings] = Field(
+        description="Default settings for this model", default=None
+    )
+
+
+class ControlNetDiffusersConfig(DiffusersConfigBase, ControlAdapterConfigBase):
     """Model config for ControlNet models (diffusers version)."""
 
     type: Literal[ModelType.ControlNet] = ModelType.ControlNet
@@ -241,7 +258,7 @@ class ControlNetDiffusersConfig(DiffusersConfigBase):
         return Tag(f"{ModelType.ControlNet.value}.{ModelFormat.Diffusers.value}")
 
 
-class ControlNetCheckpointConfig(CheckpointConfigBase):
+class ControlNetCheckpointConfig(CheckpointConfigBase, ControlAdapterConfigBase):
     """Model config for ControlNet models (diffusers version)."""
 
     type: Literal[ModelType.ControlNet] = ModelType.ControlNet
@@ -274,10 +291,17 @@ class TextualInversionFolderConfig(ModelConfigBase):
         return Tag(f"{ModelType.TextualInversion.value}.{ModelFormat.EmbeddingFolder.value}")
 
 
-class MainCheckpointConfig(CheckpointConfigBase):
+class MainConfigBase(ModelConfigBase):
+    type: Literal[ModelType.Main] = ModelType.Main
+    trigger_phrases: Optional[set[str]] = Field(description="Set of trigger phrases for this model", default=None)
+    default_settings: Optional[MainModelDefaultSettings] = Field(
+        description="Default settings for this model", default=None
+    )
+
+
+class MainCheckpointConfig(CheckpointConfigBase, MainConfigBase):
     """Model config for main checkpoint models."""
 
-    type: Literal[ModelType.Main] = ModelType.Main
     variant: ModelVariantType = ModelVariantType.Normal
     prediction_type: SchedulerPredictionType = SchedulerPredictionType.Epsilon
     upcast_attention: bool = False
@@ -287,11 +311,9 @@ class MainCheckpointConfig(CheckpointConfigBase):
         return Tag(f"{ModelType.Main.value}.{ModelFormat.Checkpoint.value}")
 
 
-class MainDiffusersConfig(DiffusersConfigBase):
+class MainDiffusersConfig(DiffusersConfigBase, MainConfigBase):
     """Model config for main diffusers models."""
 
-    type: Literal[ModelType.Main] = ModelType.Main
-
     @staticmethod
     def get_tag() -> Tag:
         return Tag(f"{ModelType.Main.value}.{ModelFormat.Diffusers.value}")
@@ -310,7 +332,7 @@ class IPAdapterConfig(ModelConfigBase):
 
 
 class CLIPVisionDiffusersConfig(ModelConfigBase):
-    """Model config for ClipVision."""
+    """Model config for CLIPVision."""
 
     type: Literal[ModelType.CLIPVision] = ModelType.CLIPVision
     format: Literal[ModelFormat.Diffusers]
@@ -320,7 +342,7 @@ class CLIPVisionDiffusersConfig(ModelConfigBase):
         return Tag(f"{ModelType.CLIPVision.value}.{ModelFormat.Diffusers.value}")
 
 
-class T2IAdapterConfig(ModelConfigBase):
+class T2IAdapterConfig(ModelConfigBase, ControlAdapterConfigBase):
     """Model config for T2I."""
 
     type: Literal[ModelType.T2IAdapter] = ModelType.T2IAdapter
@@ -372,6 +394,7 @@ AnyModelConfig = Annotated[
 ]
 
 AnyModelConfigValidator = TypeAdapter(AnyModelConfig)
+AnyDefaultSettings: TypeAlias = Union[MainModelDefaultSettings, ControlAdapterDefaultSettings]
 
 
 class ModelConfigFactory(object):

invokeai/backend/model_manager/load/model_cache/model_cache_default.py

@@ -19,7 +19,6 @@ context. Use like this:
 """
 
 import gc
-import logging
 import math
 import sys
 import time
@@ -92,8 +91,7 @@ class ModelCache(ModelCacheBase[AnyModel]):
         self._execution_device: torch.device = execution_device
         self._storage_device: torch.device = storage_device
         self._logger = logger or InvokeAILogger.get_logger(self.__class__.__name__)
-        self._log_memory_usage = log_memory_usage or self._logger.level == logging.DEBUG
-        # used for stats collection
+        self._log_memory_usage = log_memory_usage
         self._stats: Optional[CacheStats] = None
 
         self._cached_models: Dict[str, CacheRecord[AnyModel]] = {}

invokeai/backend/model_manager/load/model_loader_registry.py

@@ -60,7 +60,7 @@ class ModelLoaderRegistryBase(ABC):
 TModelLoader = TypeVar("TModelLoader", bound=ModelLoaderBase)
 
 
-class ModelLoaderRegistry:
+class ModelLoaderRegistry(ModelLoaderRegistryBase):
     """
     This class allows model loaders to register their type, base and format.
     """

invokeai/backend/model_manager/load/model_loaders/controlnet.py

@@ -3,9 +3,6 @@
 
 from pathlib import Path
 
-import torch
-from safetensors.torch import load_file as safetensors_load_file
-
 from invokeai.backend.model_manager import (
     AnyModelConfig,
     BaseModelType,
@@ -37,27 +34,25 @@ class ControlNetLoader(GenericDiffusersLoader):
             return True
 
     def _convert_model(self, config: AnyModelConfig, model_path: Path, output_path: Path) -> Path:
-        if config.base not in {BaseModelType.StableDiffusion1, BaseModelType.StableDiffusion2}:
-            raise Exception(f"ControlNet conversion not supported for model type: {config.base}")
-        else:
-            assert isinstance(config, CheckpointConfigBase)
-            config_file = config.config_path
+        assert isinstance(config, CheckpointConfigBase)
+        config_file = config.config_path
 
-        if model_path.suffix == ".safetensors":
-            checkpoint = safetensors_load_file(model_path, device="cpu")
-        else:
-            checkpoint = torch.load(model_path, map_location="cpu")
-
-        # sometimes weights are hidden under "state_dict", and sometimes not
-        if "state_dict" in checkpoint:
-            checkpoint = checkpoint["state_dict"]
-
-        convert_controlnet_to_diffusers(
-            model_path,
-            output_path,
-            original_config_file=self._app_config.root_path / config_file,
-            image_size=512,
-            scan_needed=True,
-            from_safetensors=model_path.suffix == ".safetensors",
+        image_size = (
+            512
+            if config.base == BaseModelType.StableDiffusion1
+            else 768
+            if config.base == BaseModelType.StableDiffusion2
+            else 1024
         )
+
+        self._logger.info(f"Converting {model_path} to diffusers format")
+        with open(self._app_config.root_path / config_file, "r") as config_stream:
+            convert_controlnet_to_diffusers(
+                model_path,
+                output_path,
+                original_config_file=config_stream,
+                image_size=image_size,
+                precision=self._torch_dtype,
+                from_safetensors=model_path.suffix == ".safetensors",
+            )
         return output_path

invokeai/backend/model_manager/load/model_loaders/lora.py

@@ -24,7 +24,7 @@ from .. import ModelLoader, ModelLoaderRegistry
 
 @ModelLoaderRegistry.register(base=BaseModelType.Any, type=ModelType.LoRA, format=ModelFormat.Diffusers)
 @ModelLoaderRegistry.register(base=BaseModelType.Any, type=ModelType.LoRA, format=ModelFormat.LyCORIS)
-class LoraLoader(ModelLoader):
+class LoRALoader(ModelLoader):
     """Class to load LoRA models."""
 
     # We cheat a little bit to get access to the model base

invokeai/backend/model_manager/load/model_loaders/stable_diffusion.py

@@ -4,9 +4,6 @@
 from pathlib import Path
 from typing import Optional
 
-from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion import StableDiffusionPipeline
-from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion_inpaint import StableDiffusionInpaintPipeline
-
 from invokeai.backend.model_manager import (
     AnyModel,
     AnyModelConfig,
@@ -14,7 +11,7 @@ from invokeai.backend.model_manager import (
     ModelFormat,
     ModelRepoVariant,
     ModelType,
-    ModelVariantType,
+    SchedulerPredictionType,
     SubModelType,
 )
 from invokeai.backend.model_manager.config import CheckpointConfigBase, MainCheckpointConfig
@@ -68,27 +65,31 @@ class StableDiffusionDiffusersModel(GenericDiffusersLoader):
 
     def _convert_model(self, config: AnyModelConfig, model_path: Path, output_path: Path) -> Path:
         assert isinstance(config, MainCheckpointConfig)
-        variant = config.variant
         base = config.base
-        pipeline_class = (
-            StableDiffusionInpaintPipeline if variant == ModelVariantType.Inpaint else StableDiffusionPipeline
-        )
 
         config_file = config.config_path
+        prediction_type = config.prediction_type.value
+        upcast_attention = config.upcast_attention
+        image_size = (
+            1024
+            if base == BaseModelType.StableDiffusionXL
+            else 768
+            if config.prediction_type == SchedulerPredictionType.VPrediction and base == BaseModelType.StableDiffusion2
+            else 512
+        )
 
         self._logger.info(f"Converting {model_path} to diffusers format")
         convert_ckpt_to_diffusers(
             model_path,
             output_path,
             model_type=self.model_base_to_model_type[base],
-            model_version=base,
-            model_variant=variant,
             original_config_file=self._app_config.root_path / config_file,
             extract_ema=True,
-            scan_needed=True,
-            pipeline_class=pipeline_class,
             from_safetensors=model_path.suffix == ".safetensors",
             precision=self._torch_dtype,
+            prediction_type=prediction_type,
+            image_size=image_size,
+            upcast_attention=upcast_attention,
             load_safety_checker=False,
         )
         return output_path

invokeai/backend/model_manager/load/model_loaders/vae.py

@@ -23,7 +23,7 @@ from .generic_diffusers import GenericDiffusersLoader
 @ModelLoaderRegistry.register(base=BaseModelType.Any, type=ModelType.VAE, format=ModelFormat.Diffusers)
 @ModelLoaderRegistry.register(base=BaseModelType.StableDiffusion1, type=ModelType.VAE, format=ModelFormat.Checkpoint)
 @ModelLoaderRegistry.register(base=BaseModelType.StableDiffusion2, type=ModelType.VAE, format=ModelFormat.Checkpoint)
-class VaeLoader(GenericDiffusersLoader):
+class VAELoader(GenericDiffusersLoader):
     """Class to load VAE models."""
 
     def _needs_conversion(self, config: AnyModelConfig, model_path: Path, dest_path: Path) -> bool:
@@ -57,12 +57,12 @@ class VaeLoader(GenericDiffusersLoader):
 
         ckpt_config = OmegaConf.load(self._app_config.root_path / config_file)
         assert isinstance(ckpt_config, DictConfig)
-
+        self._logger.info(f"Converting {model_path} to diffusers format")
         vae_model = convert_ldm_vae_to_diffusers(
             checkpoint=checkpoint,
             vae_config=ckpt_config,
             image_size=512,
+            precision=self._torch_dtype,
         )
-        vae_model.to(self._torch_dtype)  # set precision appropriately
         vae_model.save_pretrained(output_path, safe_serialization=True)
         return output_path

invokeai/backend/model_manager/merge.py

@@ -118,7 +118,7 @@ class ModelMerger(object):
         config = self._installer.app_config
         store = self._installer.record_store
         base_models: Set[BaseModelType] = set()
-        variant = None if self._installer.app_config.full_precision else "fp16"
+        variant = None if self._installer.app_config.precision == "float32" else "fp16"
 
         assert (
             len(model_keys) <= 2 or interp == MergeInterpolationMethod.AddDifference

invokeai/backend/model_manager/metadata/__init__.py

@@ -8,23 +8,19 @@ from invokeai.backend.model_manager.metadata import(
    CommercialUsage,
    LicenseRestrictions,
    HuggingFaceMetadata,
-   CivitaiMetadata,
 )
 
-from invokeai.backend.model_manager.metadata.fetch import CivitaiMetadataFetch
+from invokeai.backend.model_manager.metadata.fetch import HuggingFaceMetadataFetch
 
-data = CivitaiMetadataFetch().from_url("https://civitai.com/models/206883/split")
-assert isinstance(data, CivitaiMetadata)
-if data.allow_commercial_use:
-   print("Commercial use of this model is allowed")
+data = HuggingFaceMetadataFetch().from_id("<REPO_ID>")
+assert isinstance(data, HuggingFaceMetadata)
 """
 
-from .fetch import CivitaiMetadataFetch, HuggingFaceMetadataFetch, ModelMetadataFetchBase
+from .fetch import HuggingFaceMetadataFetch, ModelMetadataFetchBase
 from .metadata_base import (
     AnyModelRepoMetadata,
     AnyModelRepoMetadataValidator,
     BaseMetadata,
-    CivitaiMetadata,
     HuggingFaceMetadata,
     ModelMetadataWithFiles,
     RemoteModelFile,
@@ -34,8 +30,6 @@ from .metadata_base import (
 __all__ = [
     "AnyModelRepoMetadata",
     "AnyModelRepoMetadataValidator",
-    "CivitaiMetadata",
-    "CivitaiMetadataFetch",
     "HuggingFaceMetadata",
     "HuggingFaceMetadataFetch",
     "ModelMetadataFetchBase",

invokeai/backend/model_manager/metadata/fetch/__init__.py

@@ -3,19 +3,14 @@ Initialization file for invokeai.backend.model_manager.metadata.fetch
 
 Usage:
 from invokeai.backend.model_manager.metadata.fetch import (
-    CivitaiMetadataFetch,
     HuggingFaceMetadataFetch,
 )
-from invokeai.backend.model_manager.metadata import CivitaiMetadata
 
-data = CivitaiMetadataFetch().from_url("https://civitai.com/models/206883/split")
-assert isinstance(data, CivitaiMetadata)
-if data.allow_commercial_use:
-   print("Commercial use of this model is allowed")
+data = HuggingFaceMetadataFetch().from_id("<repo_id>")
+assert isinstance(data, HuggingFaceMetadata)
 """
 
-from .civitai import CivitaiMetadataFetch
 from .fetch_base import ModelMetadataFetchBase
 from .huggingface import HuggingFaceMetadataFetch
 
-__all__ = ["ModelMetadataFetchBase", "CivitaiMetadataFetch", "HuggingFaceMetadataFetch"]
+__all__ = ["ModelMetadataFetchBase", "HuggingFaceMetadataFetch"]

invokeai/backend/model_manager/metadata/fetch/civitai.py

@@ -1,188 +0,0 @@
-# Copyright (c) 2023 Lincoln D. Stein and the InvokeAI Development Team
-
-"""
-This module fetches model metadata objects from the Civitai model repository.
-In addition to the `from_url()` and `from_id()` methods inherited from the
-`ModelMetadataFetchBase` base class.
-
-Civitai has two separate ID spaces: a model ID and a version ID. The
-version ID corresponds to a specific model, and is the ID accepted by
-`from_id()`. The model ID corresponds to a family of related models,
-such as different training checkpoints or 16 vs 32-bit versions. The
-`from_civitai_modelid()` method will accept a model ID and return the
-metadata from the default version within this model set. The default
-version is the same as what the user sees when they click on a model's
-thumbnail.
-
-Usage:
-
-from invokeai.backend.model_manager.metadata.fetch import CivitaiMetadataFetch
-
-fetcher = CivitaiMetadataFetch()
-metadata = fetcher.from_url("https://civitai.com/models/206883/split")
-print(metadata.trained_words)
-"""
-
-import json
-import re
-from pathlib import Path
-from typing import Any, Optional
-
-import requests
-from pydantic import TypeAdapter, ValidationError
-from pydantic.networks import AnyHttpUrl
-from requests.sessions import Session
-
-from invokeai.backend.model_manager.config import ModelRepoVariant
-
-from ..metadata_base import (
-    AnyModelRepoMetadata,
-    CivitaiMetadata,
-    RemoteModelFile,
-    UnknownMetadataException,
-)
-from .fetch_base import ModelMetadataFetchBase
-
-CIVITAI_MODEL_PAGE_RE = r"https?://civitai.com/models/(\d+)"
-CIVITAI_VERSION_PAGE_RE = r"https?://civitai.com/models/(\d+)\?modelVersionId=(\d+)"
-CIVITAI_DOWNLOAD_RE = r"https?://civitai.com/api/download/models/(\d+)"
-
-CIVITAI_VERSION_ENDPOINT = "https://civitai.com/api/v1/model-versions/"
-CIVITAI_MODEL_ENDPOINT = "https://civitai.com/api/v1/models/"
-
-
-StringSetAdapter = TypeAdapter(set[str])
-
-
-class CivitaiMetadataFetch(ModelMetadataFetchBase):
-    """Fetch model metadata from Civitai."""
-
-    def __init__(self, session: Optional[Session] = None, api_key: Optional[str] = None):
-        """
-        Initialize the fetcher with an optional requests.sessions.Session object.
-
-        By providing a configurable Session object, we can support unit tests on
-        this module without an internet connection.
-        """
-        self._requests = session or requests.Session()
-        self._api_key = api_key
-
-    def from_url(self, url: AnyHttpUrl) -> AnyModelRepoMetadata:
-        """
-        Given a URL to a CivitAI model or version page, return a ModelMetadata object.
-
-        In the event that the URL points to a model page without the particular version
-        indicated, the default model version is returned. Otherwise, the requested version
-        is returned.
-        """
-        if match := re.match(CIVITAI_VERSION_PAGE_RE, str(url), re.IGNORECASE):
-            model_id = match.group(1)
-            version_id = match.group(2)
-            return self.from_civitai_versionid(int(version_id), int(model_id))
-        elif match := re.match(CIVITAI_MODEL_PAGE_RE, str(url), re.IGNORECASE):
-            model_id = match.group(1)
-            return self.from_civitai_modelid(int(model_id))
-        elif match := re.match(CIVITAI_DOWNLOAD_RE, str(url), re.IGNORECASE):
-            version_id = match.group(1)
-            return self.from_civitai_versionid(int(version_id))
-        raise UnknownMetadataException("The url '{url}' does not match any known Civitai URL patterns")
-
-    def from_id(self, id: str, variant: Optional[ModelRepoVariant] = None) -> AnyModelRepoMetadata:
-        """
-        Given a Civitai model version ID, return a ModelRepoMetadata object.
-
-        :param id: An ID.
-        :param variant: A model variant from the ModelRepoVariant enum (currently ignored)
-
-        May raise an `UnknownMetadataException`.
-        """
-        return self.from_civitai_versionid(int(id))
-
-    def from_civitai_modelid(self, model_id: int) -> CivitaiMetadata:
-        """
-        Return metadata from the default version of the indicated model.
-
-        May raise an `UnknownMetadataException`.
-        """
-        model_url = CIVITAI_MODEL_ENDPOINT + str(model_id)
-        model_json = self._requests.get(self._get_url_with_api_key(model_url)).json()
-        return self._from_api_response(model_json)
-
-    def _from_api_response(self, api_response: dict[str, Any], version_id: Optional[int] = None) -> CivitaiMetadata:
-        try:
-            version_id = version_id or api_response["modelVersions"][0]["id"]
-        except TypeError as excp:
-            raise UnknownMetadataException from excp
-
-        # loop till we find the section containing the version requested
-        version_sections = [x for x in api_response["modelVersions"] if x["id"] == version_id]
-        if not version_sections:
-            raise UnknownMetadataException(f"Version {version_id} not found in model metadata")
-
-        version_json = version_sections[0]
-
-        # Civitai has one "primary" file plus others such as VAEs. We only fetch the primary.
-        primary = [x for x in version_json["files"] if x.get("primary")]
-        assert len(primary) == 1
-        primary_file = primary[0]
-
-        url = primary_file["downloadUrl"]
-        if "?" not in url:  # work around apparent bug in civitai api
-            metadata_string = ""
-            for key, value in primary_file["metadata"].items():
-                if not value:
-                    continue
-                metadata_string += f"&{key}={value}"
-            url = url + f"?type={primary_file['type']}{metadata_string}"
-        model_files = [
-            RemoteModelFile(
-                url=self._get_url_with_api_key(url),
-                path=Path(primary_file["name"]),
-                size=int(primary_file["sizeKB"] * 1024),
-                sha256=primary_file["hashes"]["SHA256"],
-            )
-        ]
-
-        try:
-            trigger_phrases = StringSetAdapter.validate_python(version_json.get("trainedWords"))
-        except ValidationError:
-            trigger_phrases: set[str] = set()
-
-        return CivitaiMetadata(
-            name=version_json["name"],
-            files=model_files,
-            trigger_phrases=trigger_phrases,
-            api_response=json.dumps(version_json),
-        )
-
-    def from_civitai_versionid(self, version_id: int, model_id: Optional[int] = None) -> CivitaiMetadata:
-        """
-        Return a CivitaiMetadata object given a model version id.
-
-        May raise an `UnknownMetadataException`.
-        """
-        if model_id is None:
-            version_url = CIVITAI_VERSION_ENDPOINT + str(version_id)
-            version = self._requests.get(self._get_url_with_api_key(version_url)).json()
-            if error := version.get("error"):
-                raise UnknownMetadataException(error)
-            model_id = version["modelId"]
-
-        model_url = CIVITAI_MODEL_ENDPOINT + str(model_id)
-        model_json = self._requests.get(self._get_url_with_api_key(model_url)).json()
-        return self._from_api_response(model_json, version_id)
-
-    @classmethod
-    def from_json(cls, json: str) -> CivitaiMetadata:
-        """Given the JSON representation of the metadata, return the corresponding Pydantic object."""
-        metadata = CivitaiMetadata.model_validate_json(json)
-        return metadata
-
-    def _get_url_with_api_key(self, url: str) -> str:
-        if not self._api_key:
-            return url
-
-        if "?" in url:
-            return f"{url}&token={self._api_key}"
-
-        return f"{url}?token={self._api_key}"

invokeai/backend/model_manager/metadata/fetch/fetch_base.py

@@ -5,11 +5,10 @@ This module is the base class for subclasses that fetch metadata from model repo
 
 Usage:
 
-from invokeai.backend.model_manager.metadata.fetch import CivitAIMetadataFetch
+from invokeai.backend.model_manager.metadata.fetch import HuggingFaceMetadataFetch
 
-fetcher = CivitaiMetadataFetch()
-metadata = fetcher.from_url("https://civitai.com/models/206883/split")
-print(metadata.trained_words)
+data = HuggingFaceMetadataFetch().from_id("<REPO_ID>")
+assert isinstance(data, HuggingFaceMetadata)
 """
 
 from abc import ABC, abstractmethod

invokeai/backend/model_manager/metadata/fetch/huggingface.py

@@ -90,8 +90,35 @@ class HuggingFaceMetadataFetch(ModelMetadataFetchBase):
                 )
             )
 
+        # diffusers models have a `model_index.json` or `config.json` file
+        is_diffusers = any(str(f.url).endswith(("model_index.json", "config.json")) for f in files)
+
+        # These URLs will be exposed to the user - I think these are the only file types we fully support
+        ckpt_urls = (
+            None
+            if is_diffusers
+            else [
+                f.url
+                for f in files
+                if str(f.url).endswith(
+                    (
+                        ".safetensors",
+                        ".bin",
+                        ".pth",
+                        ".pt",
+                        ".ckpt",
+                    )
+                )
+            ]
+        )
+
         return HuggingFaceMetadata(
-            id=model_info.id, name=name, files=files, api_response=json.dumps(model_info.__dict__, default=str)
+            id=model_info.id,
+            name=name,
+            files=files,
+            api_response=json.dumps(model_info.__dict__, default=str),
+            is_diffusers=is_diffusers,
+            ckpt_urls=ckpt_urls,
         )
 
     def from_url(self, url: AnyHttpUrl) -> AnyModelRepoMetadata:

invokeai/backend/model_manager/metadata/metadata_base.py

@@ -78,20 +78,16 @@ class ModelMetadataWithFiles(ModelMetadataBase):
         return self.files
 
 
-class CivitaiMetadata(ModelMetadataWithFiles):
-    """Extended metadata fields provided by Civitai."""
-
-    type: Literal["civitai"] = "civitai"
-    trigger_phrases: set[str] = Field(description="Trigger phrases extracted from the API response")
-    api_response: Optional[str] = Field(description="Response from the Civitai API as stringified JSON", default=None)
-
-
 class HuggingFaceMetadata(ModelMetadataWithFiles):
     """Extended metadata fields provided by HuggingFace."""
 
     type: Literal["huggingface"] = "huggingface"
     id: str = Field(description="The HF model id")
     api_response: Optional[str] = Field(description="Response from the HF API as stringified JSON", default=None)
+    is_diffusers: bool = Field(description="Whether the metadata is for a Diffusers format model", default=False)
+    ckpt_urls: Optional[List[AnyHttpUrl]] = Field(
+        description="URLs for all checkpoint format models in the metadata", default=None
+    )
 
     def download_urls(
         self,
@@ -130,5 +126,5 @@ class HuggingFaceMetadata(ModelMetadataWithFiles):
         return [x for x in self.files if x.path in paths]
 
 
-AnyModelRepoMetadata = Annotated[Union[BaseMetadata, HuggingFaceMetadata, CivitaiMetadata], Field(discriminator="type")]
+AnyModelRepoMetadata = Annotated[Union[BaseMetadata, HuggingFaceMetadata], Field(discriminator="type")]
 AnyModelRepoMetadataValidator = TypeAdapter(AnyModelRepoMetadata)

invokeai/backend/model_manager/probe.py

@@ -9,12 +9,15 @@ from picklescan.scanner import scan_file_path
 
 import invokeai.backend.util.logging as logger
 from invokeai.app.util.misc import uuid_string
+from invokeai.backend.model_hash.model_hash import HASHING_ALGORITHMS, ModelHash
 from invokeai.backend.util.util import SilenceWarnings
 
 from .config import (
     AnyModelConfig,
     BaseModelType,
+    ControlAdapterDefaultSettings,
     InvalidModelConfigException,
+    MainModelDefaultSettings,
     ModelConfigFactory,
     ModelFormat,
     ModelRepoVariant,
@@ -23,7 +26,6 @@ from .config import (
     ModelVariantType,
     SchedulerPredictionType,
 )
-from .hash import ModelHash
 from .util.model_util import lora_token_vector_length, read_checkpoint_meta
 
 CkptType = Dict[str, Any]
@@ -112,9 +114,7 @@ class ModelProbe(object):
 
     @classmethod
     def probe(
-        cls,
-        model_path: Path,
-        fields: Optional[Dict[str, Any]] = None,
+        cls, model_path: Path, fields: Optional[Dict[str, Any]] = None, hash_algo: HASHING_ALGORITHMS = "blake3"
     ) -> AnyModelConfig:
         """
         Probe the model at model_path and return its configuration record.
@@ -128,13 +128,16 @@ class ModelProbe(object):
         if fields is None:
             fields = {}
 
+        model_path = model_path.resolve()
+
         format_type = ModelFormat.Diffusers if model_path.is_dir() else ModelFormat.Checkpoint
         model_info = None
-        model_type = None
-        if format_type is ModelFormat.Diffusers:
-            model_type = cls.get_model_type_from_folder(model_path)
-        else:
-            model_type = cls.get_model_type_from_checkpoint(model_path)
+        model_type = ModelType(fields["type"]) if "type" in fields and fields["type"] else None
+        if not model_type:
+            if format_type is ModelFormat.Diffusers:
+                model_type = cls.get_model_type_from_folder(model_path)
+            else:
+                model_type = cls.get_model_type_from_checkpoint(model_path)
         format_type = ModelFormat.ONNX if model_type == ModelType.ONNX else format_type
 
         probe_class = cls.PROBES[format_type].get(model_type)
@@ -154,10 +157,18 @@ class ModelProbe(object):
         fields["image_encoder_model_id"] = fields.get("image_encoder_model_id") or probe.get_image_encoder_model_id()
         fields["name"] = fields.get("name") or cls.get_model_name(model_path)
         fields["description"] = (
-            fields.get("description") or f"{fields['base'].value} {fields['type'].value} model {fields['name']}"
+            fields.get("description") or f"{fields['base'].value} {model_type.value} model {fields['name']}"
         )
         fields["format"] = fields.get("format") or probe.get_format()
-        fields["hash"] = fields.get("hash") or ModelHash().hash(model_path)
+        fields["hash"] = fields.get("hash") or ModelHash(algorithm=hash_algo).hash(model_path)
+
+        fields["default_settings"] = fields.get("default_settings")
+
+        if not fields["default_settings"]:
+            if fields["type"] in {ModelType.ControlNet, ModelType.T2IAdapter}:
+                fields["default_settings"] = get_default_settings_controlnet_t2i_adapter(fields["name"])
+            elif fields["type"] is ModelType.Main:
+                fields["default_settings"] = get_default_settings_main(fields["base"])
 
         if format_type == ModelFormat.Diffusers and isinstance(probe, FolderProbeBase):
             fields["repo_variant"] = fields.get("repo_variant") or probe.get_repo_variant()
@@ -309,7 +320,7 @@ class ModelProbe(object):
     @classmethod
     def _scan_and_load_checkpoint(cls, model_path: Path) -> CkptType:
         with SilenceWarnings():
-            if model_path.suffix.endswith((".ckpt", ".pt", ".bin")):
+            if model_path.suffix.endswith((".ckpt", ".pt", ".pth", ".bin")):
                 cls._scan_model(model_path.name, model_path)
                 model = torch.load(model_path)
                 assert isinstance(model, dict)
@@ -329,6 +340,43 @@ class ModelProbe(object):
             raise Exception("The model {model_name} is potentially infected by malware. Aborting import.")
 
 
+# Probing utilities
+MODEL_NAME_TO_PREPROCESSOR = {
+    "canny": "canny_image_processor",
+    "mlsd": "mlsd_image_processor",
+    "depth": "depth_anything_image_processor",
+    "bae": "normalbae_image_processor",
+    "normal": "normalbae_image_processor",
+    "sketch": "pidi_image_processor",
+    "scribble": "lineart_image_processor",
+    "lineart": "lineart_image_processor",
+    "lineart_anime": "lineart_anime_image_processor",
+    "softedge": "hed_image_processor",
+    "shuffle": "content_shuffle_image_processor",
+    "pose": "dw_openpose_image_processor",
+    "mediapipe": "mediapipe_face_processor",
+    "pidi": "pidi_image_processor",
+    "zoe": "zoe_depth_image_processor",
+    "color": "color_map_image_processor",
+}
+
+
+def get_default_settings_controlnet_t2i_adapter(model_name: str) -> Optional[ControlAdapterDefaultSettings]:
+    for k, v in MODEL_NAME_TO_PREPROCESSOR.items():
+        if k in model_name:
+            return ControlAdapterDefaultSettings(preprocessor=v)
+    return None
+
+
+def get_default_settings_main(model_base: BaseModelType) -> Optional[MainModelDefaultSettings]:
+    if model_base is BaseModelType.StableDiffusion1 or model_base is BaseModelType.StableDiffusion2:
+        return MainModelDefaultSettings(width=512, height=512)
+    elif model_base is BaseModelType.StableDiffusionXL:
+        return MainModelDefaultSettings(width=1024, height=1024)
+    # We don't provide defaults for BaseModelType.StableDiffusionXLRefiner, as they are not standalone models.
+    return None
+
+
 # ##################################################3
 # Checkpoint probing
 # ##################################################3

invokeai/backend/model_manager/search.py

@@ -4,121 +4,75 @@ Abstract base class and implementation for recursive directory search for models
 
 Example usage:
 ```
-  from invokeai.backend.model_manager import ModelSearch, ModelProbe
+    from invokeai.backend.model_manager import ModelSearch, ModelProbe
 
-  def find_main_models(model: Path) -> bool:
-    info = ModelProbe.probe(model)
-    if info.model_type == 'main' and info.base_type == 'sd-1':
-        return True
-    else:
-        return False
+    def find_main_models(model: Path) -> bool:
+        info = ModelProbe.probe(model)
+        if info.model_type == 'main' and info.base_type == 'sd-1':
+            return True
+        else:
+            return False
 
-  search = ModelSearch(on_model_found=report_it)
-  found = search.search('/tmp/models')
-  print(found)   #  list of matching model paths
-  print(search.stats)  #  search stats
+    search = ModelSearch(on_model_found=report_it)
+    found = search.search('/tmp/models')
+    print(found)   #  list of matching model paths
+    print(search.stats)  #  search stats
 ```
 """
 
 import os
-from abc import ABC, abstractmethod
-from logging import Logger
+from dataclasses import dataclass
 from pathlib import Path
-from typing import Callable, Optional, Set, Union
+from typing import Callable, Optional
 
-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()
 
+@dataclass
+class SearchStats:
+    """Statistics about the search.
 
-class SearchStats(BaseModel):
-    items_scanned: int = 0
-    models_found: int = 0
-    models_filtered: int = 0
-
-
-class ModelSearchBase(ABC, BaseModel):
+    Attributes:
+        items_scanned: number of items scanned
+        models_found: number of models found
+        models_filtered: number of models that passed the filter
     """
-    Abstract directory traversal model search class
+
+    items_scanned = 0
+    models_found = 0
+    models_filtered = 0
+
+
+class ModelSearch:
+    """Searches a directory tree for models, using a callback to filter the results.
 
     Usage:
-       search = ModelSearchBase(
-            on_search_started = search_started_callback,
-            on_search_completed = search_completed_callback,
-            on_model_found = model_found_callback,
-       )
-       models_found = search.search('/path/to/directory')
+        search = ModelSearch()
+        search.model_found = lambda path : 'anime' in path.as_posix()
+        found = search.list_models(['/tmp/models1','/tmp/models2'])
+        # returns all models that have 'anime' in the path
     """
 
-    # fmt: off
-    on_search_started   : Optional[Callable[[Path], None]]      = Field(default=None, description="Called just before the search starts.")  # noqa E221
-    on_model_found      : Optional[Callable[[Path], bool]]      = Field(default=None, description="Called when a model is found.")          # noqa E221
-    on_search_completed : Optional[Callable[[Set[Path]], None]] = Field(default=None, description="Called when search is complete.")        # noqa E221
-    stats               : SearchStats                           = Field(default_factory=SearchStats, description="Summary statistics after search")  # noqa E221
-    logger              : Logger                                = Field(default=default_logger, description="Logger instance.")     # noqa E221
-    # fmt: on
+    def __init__(
+        self,
+        on_search_started: Optional[Callable[[Path], None]] = None,
+        on_model_found: Optional[Callable[[Path], bool]] = None,
+        on_search_completed: Optional[Callable[[set[Path]], None]] = None,
+    ) -> None:
+        """Create a new ModelSearch object.
 
-    class Config:
-        arbitrary_types_allowed = True
-
-    @abstractmethod
-    def search_started(self) -> None:
+        Args:
+            on_search_started: callback to be invoked when the search starts
+            on_model_found: callback to be invoked when a model is found. The callback should return True if the model
+                should be included in the results.
+            on_search_completed: callback to be invoked when the search is completed
         """
-        Called before the scan starts.
-
-        Passes the root search directory to the Callable `on_search_started`.
-        """
-        pass
-
-    @abstractmethod
-    def model_found(self, model: Path) -> None:
-        """
-        Called when a model is found during search.
-
-        :param model: Model to process - could be a directory or checkpoint.
-
-        Passes the model's Path to the Callable `on_model_found`.
-        This Callable receives the path to the model and returns a boolean
-        to indicate whether the model should be returned in the search
-        results.
-        """
-        pass
-
-    @abstractmethod
-    def search_completed(self) -> None:
-        """
-        Called before the scan starts.
-
-        Passes the Set of found model Paths to the Callable `on_search_completed`.
-        """
-        pass
-
-    @abstractmethod
-    def search(self, directory: Union[Path, str]) -> Set[Path]:
-        """
-        Recursively search for models in `directory` and return a set of model paths.
-
-        If provided, the `on_search_started`, `on_model_found` and `on_search_completed`
-        Callables will be invoked during the search.
-        """
-        pass
-
-
-class ModelSearch(ModelSearchBase):
-    """
-    Implementation of ModelSearch with callbacks.
-    Usage:
-       search = ModelSearch()
-       search.model_found = lambda path : 'anime' in path.as_posix()
-       found = search.list_models(['/tmp/models1','/tmp/models2'])
-       # returns all models that have 'anime' in the path
-    """
-
-    models_found: Set[Path] = Field(default_factory=set)
-    config: InvokeAIAppConfig = InvokeAIAppConfig.get_config()
+        self.stats = SearchStats()
+        self.logger = InvokeAILogger.get_logger()
+        self.on_search_started = on_search_started
+        self.on_model_found = on_model_found
+        self.on_search_completed = on_search_completed
+        self.models_found: set[Path] = set()
 
     def search_started(self) -> None:
         self.models_found = set()
@@ -135,17 +89,17 @@ class ModelSearch(ModelSearchBase):
         if self.on_search_completed is not None:
             self.on_search_completed(self.models_found)
 
-    def search(self, directory: Union[Path, str]) -> Set[Path]:
+    def search(self, directory: Path) -> set[Path]:
         self._directory = Path(directory)
-        if not self._directory.is_absolute():
-            self._directory = self.config.models_path / self._directory
+        self._directory = self._directory.resolve()
         self.stats = SearchStats()  # zero out
         self.search_started()  # This will initialize _models_found to empty
         self._walk_directory(self._directory)
         self.search_completed()
         return self.models_found
 
-    def _walk_directory(self, path: Union[Path, str], max_depth: int = 20) -> None:
+    def _walk_directory(self, path: Path, max_depth: int = 20) -> None:
+        """Recursively walk the directory tree, looking for models."""
         absolute_path = Path(path)
         if (
             len(absolute_path.parts) - len(self._directory.parts) > max_depth

invokeai/backend/stable_diffusion/diffusers_pipeline.py

@@ -21,7 +21,7 @@ from diffusers.utils.outputs import BaseOutput
 from pydantic import Field
 from transformers import CLIPFeatureExtractor, CLIPTextModel, CLIPTokenizer
 
-from invokeai.app.services.config import InvokeAIAppConfig
+from invokeai.app.services.config.config_default import get_config
 from invokeai.backend.ip_adapter.ip_adapter import IPAdapter
 from invokeai.backend.ip_adapter.unet_patcher import UNetPatcher
 from invokeai.backend.stable_diffusion.diffusion.conditioning_data import ConditioningData
@@ -251,7 +251,7 @@ class StableDiffusionGeneratorPipeline(StableDiffusionPipeline):
         """
         if xformers is available, use it, otherwise use sliced attention.
         """
-        config = InvokeAIAppConfig.get_config()
+        config = get_config()
         if config.attention_type == "xformers":
             self.enable_xformers_memory_efficient_attention()
             return
@@ -275,7 +275,7 @@ class StableDiffusionGeneratorPipeline(StableDiffusionPipeline):
 
         # the remainder if this code is called when attention_type=='auto'
         if self.unet.device.type == "cuda":
-            if is_xformers_available() and not config.disable_xformers:
+            if is_xformers_available():
                 self.enable_xformers_memory_efficient_attention()
                 return
             elif hasattr(torch.nn.functional, "scaled_dot_product_attention"):
@@ -455,15 +455,6 @@ class StableDiffusionGeneratorPipeline(StableDiffusionPipeline):
                     ip_adapter_unet_patcher=ip_adapter_unet_patcher,
                 )
                 latents = step_output.prev_sample
-
-                latents = self.invokeai_diffuser.do_latent_postprocessing(
-                    postprocessing_settings=conditioning_data.postprocessing_settings,
-                    latents=latents,
-                    sigma=batched_t,
-                    step_index=i,
-                    total_step_count=len(timesteps),
-                )
-
                 predicted_original = getattr(step_output, "pred_original_sample", None)
 
                 if callback is not None:

invokeai/backend/stable_diffusion/diffusion/conditioning_data.py

@@ -44,14 +44,6 @@ class SDXLConditioningInfo(BasicConditioningInfo):
         return super().to(device=device, dtype=dtype)
 
 
-@dataclass(frozen=True)
-class PostprocessingSettings:
-    threshold: float
-    warmup: float
-    h_symmetry_time_pct: Optional[float]
-    v_symmetry_time_pct: Optional[float]
-
-
 @dataclass
 class IPAdapterConditioningInfo:
     cond_image_prompt_embeds: torch.Tensor
@@ -80,10 +72,6 @@ class ConditioningData:
     """
     guidance_rescale_multiplier: float = 0
     scheduler_args: dict[str, Any] = field(default_factory=dict)
-    """
-    Additional arguments to pass to invokeai_diffuser.do_latent_postprocessing().
-    """
-    postprocessing_settings: Optional[PostprocessingSettings] = None
 
     ip_adapter_conditioning: Optional[list[IPAdapterConditioningInfo]] = None
 

invokeai/backend/stable_diffusion/diffusion/shared_invokeai_diffusion.py

@@ -8,11 +8,10 @@ import torch
 from diffusers import UNet2DConditionModel
 from typing_extensions import TypeAlias
 
-from invokeai.app.services.config import InvokeAIAppConfig
+from invokeai.app.services.config.config_default import get_config
 from invokeai.backend.stable_diffusion.diffusion.conditioning_data import (
     ConditioningData,
     ExtraConditioningInfo,
-    PostprocessingSettings,
     SDXLConditioningInfo,
 )
 
@@ -55,7 +54,7 @@ class InvokeAIDiffuserComponent:
         :param model: the unet model to pass through to cross attention control
         :param model_forward_callback: a lambda with arguments (x, sigma, conditioning_to_apply). will be called repeatedly. most likely, this should simply call model.forward(x, sigma, conditioning)
         """
-        config = InvokeAIAppConfig.get_config()
+        config = get_config()
         self.conditioning = None
         self.model = model
         self.model_forward_callback = model_forward_callback
@@ -244,19 +243,6 @@ class InvokeAIDiffuserComponent:
 
         return unconditioned_next_x, conditioned_next_x
 
-    def do_latent_postprocessing(
-        self,
-        postprocessing_settings: PostprocessingSettings,
-        latents: torch.Tensor,
-        sigma,
-        step_index,
-        total_step_count,
-    ) -> torch.Tensor:
-        if postprocessing_settings is not None:
-            percent_through = step_index / total_step_count
-            latents = self.apply_symmetry(postprocessing_settings, latents, percent_through)
-        return latents
-
     def _concat_conditionings_for_batch(self, unconditioning, conditioning):
         def _pad_conditioning(cond, target_len, encoder_attention_mask):
             conditioning_attention_mask = torch.ones(
@@ -506,64 +492,3 @@ class InvokeAIDiffuserComponent:
         scaled_delta = (conditioned_next_x - unconditioned_next_x) * guidance_scale
         combined_next_x = unconditioned_next_x + scaled_delta
         return combined_next_x
-
-    def apply_symmetry(
-        self,
-        postprocessing_settings: PostprocessingSettings,
-        latents: torch.Tensor,
-        percent_through: float,
-    ) -> torch.Tensor:
-        # Reset our last percent through if this is our first step.
-        if percent_through == 0.0:
-            self.last_percent_through = 0.0
-
-        if postprocessing_settings is None:
-            return latents
-
-        # Check for out of bounds
-        h_symmetry_time_pct = postprocessing_settings.h_symmetry_time_pct
-        if h_symmetry_time_pct is not None and (h_symmetry_time_pct <= 0.0 or h_symmetry_time_pct > 1.0):
-            h_symmetry_time_pct = None
-
-        v_symmetry_time_pct = postprocessing_settings.v_symmetry_time_pct
-        if v_symmetry_time_pct is not None and (v_symmetry_time_pct <= 0.0 or v_symmetry_time_pct > 1.0):
-            v_symmetry_time_pct = None
-
-        dev = latents.device.type
-
-        latents.to(device="cpu")
-
-        if (
-            h_symmetry_time_pct is not None
-            and self.last_percent_through < h_symmetry_time_pct
-            and percent_through >= h_symmetry_time_pct
-        ):
-            # Horizontal symmetry occurs on the 3rd dimension of the latent
-            width = latents.shape[3]
-            x_flipped = torch.flip(latents, dims=[3])
-            latents = torch.cat(
-                [
-                    latents[:, :, :, 0 : int(width / 2)],
-                    x_flipped[:, :, :, int(width / 2) : int(width)],
-                ],
-                dim=3,
-            )
-
-        if (
-            v_symmetry_time_pct is not None
-            and self.last_percent_through < v_symmetry_time_pct
-            and percent_through >= v_symmetry_time_pct
-        ):
-            # Vertical symmetry occurs on the 2nd dimension of the latent
-            height = latents.shape[2]
-            y_flipped = torch.flip(latents, dims=[2])
-            latents = torch.cat(
-                [
-                    latents[:, :, 0 : int(height / 2)],
-                    y_flipped[:, :, int(height / 2) : int(height)],
-                ],
-                dim=2,
-            )
-
-        self.last_percent_through = percent_through
-        return latents.to(device=dev)

invokeai/backend/stable_diffusion/seamless.py

@@ -5,6 +5,7 @@ from typing import Callable, List, Union
 
 import torch.nn as nn
 from diffusers.models.autoencoders.autoencoder_kl import AutoencoderKL
+from diffusers.models.autoencoders.autoencoder_tiny import AutoencoderTiny
 from diffusers.models.unets.unet_2d_condition import UNet2DConditionModel
 
 
@@ -26,7 +27,7 @@ def _conv_forward_asymmetric(self, input, weight, bias):
 
 
 @contextmanager
-def set_seamless(model: Union[UNet2DConditionModel, AutoencoderKL], seamless_axes: List[str]):
+def set_seamless(model: Union[UNet2DConditionModel, AutoencoderKL, AutoencoderTiny], seamless_axes: List[str]):
     # Callable: (input: Tensor, weight: Tensor, bias: Optional[Tensor]) -> Tensor
     to_restore: list[tuple[nn.Conv2d | nn.ConvTranspose2d, Callable]] = []
     try:

invokeai/backend/training/textual_inversion_training.py

@@ -40,6 +40,7 @@ from transformers import CLIPTextModel, CLIPTokenizer
 
 # invokeai stuff
 from invokeai.app.services.config import InvokeAIAppConfig, PagingArgumentParser
+from invokeai.app.services.config.config_default import get_config
 from invokeai.backend.install.install_helper import initialize_record_store
 from invokeai.backend.model_manager import BaseModelType, ModelType
 
@@ -77,7 +78,7 @@ def save_progress(text_encoder, placeholder_token_id, accelerator, placeholder_t
 
 
 def parse_args() -> Namespace:
-    config = InvokeAIAppConfig.get_config()
+    config = get_config()
     parser = PagingArgumentParser(description="Textual inversion training")
     general_group = parser.add_argument_group("General")
     model_group = parser.add_argument_group("Models and Paths")
@@ -98,7 +99,7 @@ def parse_args() -> Namespace:
         "--root_dir",
         "--root",
         type=Path,
-        default=config.root,
+        default=config.root_path,
         help="Path to the invokeai runtime directory",
     )
     general_group.add_argument(
@@ -113,7 +114,7 @@ def parse_args() -> Namespace:
     general_group.add_argument(
         "--output_dir",
         type=Path,
-        default=f"{config.root}/text-inversion-model",
+        default=f"{config.root_path}/text-inversion-model",
         help="The output directory where the model predictions and checkpoints will be written.",
     )
     model_group.add_argument(
@@ -549,7 +550,7 @@ def do_textual_inversion_training(
 
     # setting up things the way invokeai expects them
     if not os.path.isabs(output_dir):
-        output_dir = os.path.join(config.root, output_dir)
+        output_dir = config.root_path / output_dir
 
     logging_dir = output_dir / logging_dir
 

invokeai/backend/util/devices.py

@@ -7,18 +7,17 @@ import torch
 from torch import autocast
 
 from invokeai.app.services.config import InvokeAIAppConfig
+from invokeai.app.services.config.config_default import get_config
 
 CPU_DEVICE = torch.device("cpu")
 CUDA_DEVICE = torch.device("cuda")
 MPS_DEVICE = torch.device("mps")
-config = InvokeAIAppConfig.get_config()
 
 
 def choose_torch_device() -> torch.device:
     """Convenience routine for guessing which GPU device to run model on"""
-    if config.use_cpu:  # legacy setting - force CPU
-        return CPU_DEVICE
-    elif config.device == "auto":
+    config = get_config()
+    if config.device == "auto":
         if torch.cuda.is_available():
             return torch.device("cuda")
         if hasattr(torch.backends, "mps") and torch.backends.mps.is_available():
@@ -35,7 +34,7 @@ def choose_precision(
     device: torch.device, app_config: Optional[InvokeAIAppConfig] = None
 ) -> Literal["float32", "float16", "bfloat16"]:
     """Return an appropriate precision for the given torch device."""
-    app_config = app_config or config
+    app_config = app_config or get_config()
     if device.type == "cuda":
         device_name = torch.cuda.get_device_name(device)
         if not ("GeForce GTX 1660" in device_name or "GeForce GTX 1650" in device_name):

invokeai/backend/util/logging.py

@@ -181,6 +181,7 @@ from pathlib import Path
 from typing import Any, Dict, Optional
 
 from invokeai.app.services.config import InvokeAIAppConfig
+from invokeai.app.services.config.config_default import get_config
 
 try:
     import syslog
@@ -338,9 +339,8 @@ class InvokeAILogger(object):  # noqa D102
     loggers: Dict[str, logging.Logger] = {}
 
     @classmethod
-    def get_logger(
-        cls, name: str = "InvokeAI", config: InvokeAIAppConfig = InvokeAIAppConfig.get_config()
-    ) -> logging.Logger:  # noqa D102
+    def get_logger(cls, name: str = "InvokeAI", config: Optional[InvokeAIAppConfig] = None) -> logging.Logger:  # noqa D102
+        config = config or get_config()
         if name in cls.loggers:
             return cls.loggers[name]
 

invokeai/backend/util/test_utils.py

@@ -42,9 +42,10 @@ def install_and_load_model(
     # If the requested model is already installed, return its LoadedModel
     with contextlib.suppress(UnknownModelException):
         # TODO: Replace with wrapper call
-        loaded_model: LoadedModel = model_manager.load_model_by_attr(
+        configs = model_manager.store.search_by_attr(
             model_name=model_name, base_model=base_model, model_type=model_type
         )
+        loaded_model: LoadedModel = model_manager.load.load_model(configs[0])
         return loaded_model
 
     # Install the requested model.
@@ -53,7 +54,7 @@ def install_and_load_model(
     assert job.complete
 
     try:
-        loaded_model = model_manager.load_model_by_config(job.config_out)
+        loaded_model = model_manager.load.load_model(job.config_out)
         return loaded_model
     except UnknownModelException as e:
         raise Exception(

invokeai/configs/INITIAL_MODELS.yaml

@@ -62,40 +62,72 @@ sd-1/main/trinart_stable_diffusion_v2:
    recommended: False
 sd-1/controlnet/qrcode_monster:
    source: monster-labs/control_v1p_sd15_qrcode_monster
+   description: Controlnet model that generates scannable creative QR codes
    subfolder: v2
 sd-1/controlnet/canny:
+   description: Controlnet weights trained on sd-1.5 with canny conditioning.
    source: lllyasviel/control_v11p_sd15_canny
    recommended: True
 sd-1/controlnet/inpaint:
    source: lllyasviel/control_v11p_sd15_inpaint
+   description: Controlnet weights trained on sd-1.5 with canny conditioning, inpaint version
 sd-1/controlnet/mlsd:
+   description: Controlnet weights trained on sd-1.5 with canny conditioning, MLSD version
    source: lllyasviel/control_v11p_sd15_mlsd
 sd-1/controlnet/depth:
+   description: Controlnet weights trained on sd-1.5 with depth conditioning
    source: lllyasviel/control_v11f1p_sd15_depth
    recommended: True
 sd-1/controlnet/normal_bae:
+   description: Controlnet weights trained on sd-1.5 with normalbae image conditioning
    source: lllyasviel/control_v11p_sd15_normalbae
 sd-1/controlnet/seg:
+   description: Controlnet weights trained on sd-1.5 with seg image conditioning
    source: lllyasviel/control_v11p_sd15_seg
 sd-1/controlnet/lineart:
+   description: Controlnet weights trained on sd-1.5 with lineart image conditioning
    source: lllyasviel/control_v11p_sd15_lineart
    recommended: True
 sd-1/controlnet/lineart_anime:
+   description: Controlnet weights trained on sd-1.5 with anime image conditioning
    source: lllyasviel/control_v11p_sd15s2_lineart_anime
 sd-1/controlnet/openpose:
+   description: Controlnet weights trained on sd-1.5 with openpose image conditioning
    source: lllyasviel/control_v11p_sd15_openpose
    recommended: True
 sd-1/controlnet/scribble:
    source: lllyasviel/control_v11p_sd15_scribble
+   description: Controlnet weights trained on sd-1.5 with scribble image conditioning
    recommended: False
 sd-1/controlnet/softedge:
    source: lllyasviel/control_v11p_sd15_softedge
+   description: Controlnet weights trained on sd-1.5 with soft edge conditioning
 sd-1/controlnet/shuffle:
    source: lllyasviel/control_v11e_sd15_shuffle
+   description: Controlnet weights trained on sd-1.5 with shuffle image conditioning
 sd-1/controlnet/tile:
    source: lllyasviel/control_v11f1e_sd15_tile
+   description: Controlnet weights trained on sd-1.5 with tiled image conditioning
 sd-1/controlnet/ip2p:
    source: lllyasviel/control_v11e_sd15_ip2p
+   description: Controlnet weights trained on sd-1.5 with ip2p conditioning.
+sdxl/controlnet/canny-sdxl:
+   description: Controlnet weights trained on sdxl-1.0 with canny conditioning.
+   source: diffusers/controlnet-canny-sdxl-1.0
+   recommended: True
+sdxl/controlnet/depth-sdxl:
+   description: Controlnet weights trained on sdxl-1.0 with depth conditioning.
+   source: diffusers/controlnet-depth-sdxl-1.0
+   recommended: True
+sdxl/controlnet/softedge-dexined-sdxl:
+   description: Controlnet weights trained on sdxl-1.0 with dexined soft edge preprocessing.
+   source: SargeZT/controlnet-sd-xl-1.0-softedge-dexined
+sdxl/controlnet/depth-16bit-zoe-sdxl:
+   description: Controlnet weights trained on sdxl-1.0 with Zoe's preprocessor (16 bits).
+   source: SargeZT/controlnet-sd-xl-1.0-depth-16bit-zoe
+sdxl/controlnet/depth-zoe-sdxl:
+   description: Controlnet weights trained on sdxl-1.0 with Zoe's preprocessor (32 bits).
+   source: diffusers/controlnet-zoe-depth-sdxl-1.0
 sd-1/t2i_adapter/canny-sd15:
    source: TencentARC/t2iadapter_canny_sd15v2
 sd-1/t2i_adapter/sketch-sd15:

invokeai/frontend/CLI/__init__.py

@@ -1,5 +0,0 @@
-"""
-Initialization file for invokeai.frontend.CLI
-"""
-
-from .CLI import main as invokeai_command_line_interface  # noqa: F401

invokeai/frontend/cli/arg_parser.py

@@ -0,0 +1,47 @@
+from argparse import ArgumentParser, Namespace, RawTextHelpFormatter
+from typing import Optional
+
+from invokeai.version import __version__
+
+_root_help = r"""Sets a root directory for the app.
+If omitted, the app will search for the root directory in the following order:
+- The `$INVOKEAI_ROOT` environment variable
+- The currently active virtual environment's parent directory
+- `$HOME/invokeai`"""
+
+_ignore_missing_core_models_help = r"""If set, the app will ignore missing core diffusers conversion models.
+These are required to use checkpoint/safetensors models.
+If you only use diffusers models, you can safely enable this."""
+
+_parser = ArgumentParser(description="Invoke Studio", formatter_class=RawTextHelpFormatter)
+_parser.add_argument("--root", type=str, help=_root_help)
+_parser.add_argument("--ignore_missing_core_models", action="store_true", help=_ignore_missing_core_models_help)
+_parser.add_argument("--version", action="version", version=__version__, help="Displays the version and exits.")
+
+
+class InvokeAIArgs:
+    """Helper class for parsing CLI args.
+
+    Args should never be parsed within the application code, only in the CLI entrypoints. Parsing args within the
+    application creates conflicts when running tests or when using application modules directly.
+
+    If the args are needed within the application, the consumer should access them from this class.
+
+    Example:
+    ```
+    # In a CLI wrapper
+    from invokeai.frontend.cli.arg_parser import InvokeAIArgs
+    InvokeAIArgs.parse_args()
+
+    # In the application
+    from invokeai.frontend.cli.arg_parser import InvokeAIArgs
+    args = InvokeAIArgs.args
+    """
+
+    args: Optional[Namespace] = None
+
+    @staticmethod
+    def parse_args() -> Optional[Namespace]:
+        """Parse CLI args and store the result."""
+        InvokeAIArgs.args = _parser.parse_args()
+        return InvokeAIArgs.args

invokeai/frontend/install/import_images.py

@@ -25,10 +25,10 @@ from prompt_toolkit.completion import PathCompleter
 from prompt_toolkit.key_binding import KeyBindings
 from prompt_toolkit.shortcuts import message_dialog
 
-from invokeai.app.services.config import InvokeAIAppConfig
+from invokeai.app.services.config.config_default import get_config
 from invokeai.app.util.misc import uuid_string
 
-app_config = InvokeAIAppConfig.get_config()
+app_config = get_config()
 
 bindings = KeyBindings()
 

invokeai/frontend/install/invokeai_configure.py

@@ -2,6 +2,59 @@
 Wrapper for invokeai.backend.configure.invokeai_configure
 """
 
-from ...backend.install.invokeai_configure import main as invokeai_configure  # noqa: F401
+import argparse
 
-__all__ = ["invokeai_configure"]
+
+def run_configure() -> None:
+    # Before doing _anything_, parse CLI args!
+    from invokeai.frontend.cli.arg_parser import InvokeAIArgs
+
+    parser = argparse.ArgumentParser(description="InvokeAI model downloader")
+    parser.add_argument(
+        "--skip-sd-weights",
+        dest="skip_sd_weights",
+        action=argparse.BooleanOptionalAction,
+        default=False,
+        help="skip downloading the large Stable Diffusion weight files",
+    )
+    parser.add_argument(
+        "--skip-support-models",
+        dest="skip_support_models",
+        action=argparse.BooleanOptionalAction,
+        default=False,
+        help="skip downloading the support models",
+    )
+    parser.add_argument(
+        "--full-precision",
+        dest="full_precision",
+        action=argparse.BooleanOptionalAction,
+        type=bool,
+        default=False,
+        help="use 32-bit weights instead of faster 16-bit weights",
+    )
+    parser.add_argument(
+        "--yes",
+        "-y",
+        dest="yes_to_all",
+        action="store_true",
+        help='answer "yes" to all prompts',
+    )
+    parser.add_argument(
+        "--default_only",
+        action="store_true",
+        help="when --yes specified, only install the default model",
+    )
+    parser.add_argument(
+        "--root_dir",
+        dest="root",
+        type=str,
+        default=None,
+        help="path to root of install directory",
+    )
+
+    opt = parser.parse_args()
+    InvokeAIArgs.args = opt
+
+    from invokeai.backend.install.invokeai_configure import main as invokeai_configure
+
+    invokeai_configure(opt)

invokeai/frontend/install/model_install.py

@@ -11,6 +11,7 @@ It is currently named model_install2.py, but will ultimately replace model_insta
 
 import argparse
 import curses
+import pathlib
 import sys
 import traceback
 import warnings
@@ -22,8 +23,9 @@ import npyscreen
 import torch
 from npyscreen import widget
 
-from invokeai.app.services.config import InvokeAIAppConfig
+from invokeai.app.services.config.config_default import get_config
 from invokeai.app.services.model_install import ModelInstallServiceBase
+from invokeai.backend.install.check_directories import validate_directories
 from invokeai.backend.install.install_helper import InstallHelper, InstallSelections, UnifiedModelInfo
 from invokeai.backend.model_manager import ModelType
 from invokeai.backend.util import choose_precision, choose_torch_device
@@ -41,8 +43,8 @@ from invokeai.frontend.install.widgets import (
 )
 
 warnings.filterwarnings("ignore", category=UserWarning)  # noqa: E402
-config = InvokeAIAppConfig.get_config()
-logger = InvokeAILogger.get_logger("ModelInstallService")
+config = get_config()
+logger = InvokeAILogger.get_logger("ModelInstallService", config=config)
 # logger.setLevel("WARNING")
 # logger.setLevel('DEBUG')
 
@@ -512,7 +514,6 @@ def list_models(installer: ModelInstallServiceBase, model_type: ModelType):
 def select_and_download_models(opt: Namespace) -> None:
     """Prompt user for install/delete selections and execute."""
     precision = "float32" if opt.full_precision else choose_precision(torch.device(choose_torch_device()))
-    # unsure how to avoid a typing complaint in the next line: config.precision is an enumerated Literal
     config.precision = precision
     install_helper = InstallHelper(config, logger)
     installer = install_helper.installer
@@ -594,22 +595,26 @@ def main() -> None:
     parser.add_argument(
         "--root_dir",
         dest="root",
-        type=str,
+        type=pathlib.Path,
         default=None,
         help="path to root of install directory",
     )
     opt = parser.parse_args()
 
-    invoke_args = []
-    if opt.root:
-        invoke_args.extend(["--root", opt.root])
+    invoke_args: dict[str, Any] = {}
     if opt.full_precision:
-        invoke_args.extend(["--precision", "float32"])
-    config.parse_args(invoke_args)
+        invoke_args["precision"] = "float32"
+    config.update_config(invoke_args)
+    if opt.root:
+        config.set_root(opt.root)
+
     logger = InvokeAILogger().get_logger(config=config)
 
-    if not config.model_conf_path.exists():
+    try:
+        validate_directories(config)
+    except AssertionError:
         logger.info("Your InvokeAI root directory is not set up. Calling invokeai-configure.")
+        sys.argv = ["invokeai_configure", "--yes", "--skip-sd-weights"]
         from invokeai.frontend.install.invokeai_configure import invokeai_configure
 
         invokeai_configure()