experiment: fix types

This provides useful context for subsequent logs during queue item execution.

.github/pull_request_template.md

@@ -19,3 +19,4 @@
 - [ ] _The PR has a short but descriptive title, suitable for a changelog_
 - [ ] _Tests added / updated (if applicable)_
 - [ ] _Documentation added / updated (if applicable)_
+- [ ] _Updated `What's New` copy (if doing a release after this PR)_

.github/workflows/typegen-checks.yml

@@ -0,0 +1,85 @@
+# Runs typegen schema quality checks.
+# Frontend types should match the server.
+#
+# Checks for changes to files before running the checks.
+# If always_run is true, always runs the checks.
+
+name: 'typegen checks'
+
+on:
+  push:
+    branches:
+      - 'main'
+  pull_request:
+    types:
+      - 'ready_for_review'
+      - 'opened'
+      - '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:
+  typegen-checks:
+    runs-on: ubuntu-22.04
+    timeout-minutes: 15 # expected run time: <5 min
+    steps:
+      - name: checkout
+        uses: actions/checkout@v4
+
+      - name: check for changed files
+        if: ${{ inputs.always_run != true }}
+        id: changed-files
+        uses: tj-actions/changed-files@v42
+        with:
+          files_yaml: |
+            src:
+              - 'pyproject.toml'
+              - 'invokeai/**'
+
+      - name: setup python
+        if: ${{ steps.changed-files.outputs.src_any_changed == 'true' || inputs.always_run == true }}
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.10'
+          cache: pip
+          cache-dependency-path: pyproject.toml
+
+      - name: install python dependencies
+        if: ${{ steps.changed-files.outputs.src_any_changed == 'true' || inputs.always_run == true }}
+        run: pip3 install --use-pep517 --editable="."
+
+      - name: install frontend dependencies
+        if: ${{ steps.changed-files.outputs.src_any_changed == 'true' || inputs.always_run == true }}
+        uses: ./.github/actions/install-frontend-deps
+
+      - name: copy schema
+        if: ${{ steps.changed-files.outputs.src_any_changed == 'true' || inputs.always_run == true }}
+        run: cp invokeai/frontend/web/src/services/api/schema.ts invokeai/frontend/web/src/services/api/schema_orig.ts
+        shell: bash
+
+      - name: generate schema
+        if: ${{ steps.changed-files.outputs.src_any_changed == 'true' || inputs.always_run == true }}
+        run: make frontend-typegen
+        shell: bash
+
+      - name: compare files
+        if: ${{ steps.changed-files.outputs.src_any_changed == 'true' || inputs.always_run == true }}
+        run: |
+          if ! diff invokeai/frontend/web/src/services/api/schema.ts invokeai/frontend/web/src/services/api/schema_orig.ts; then
+            echo "Files are different!";
+            exit 1;
+          fi
+        shell: bash

SECURITY.md

@@ -0,0 +1,14 @@
+# Security Policy
+
+## Supported Versions
+
+Only the latest version of Invoke will receive security updates. 
+We do not currently maintain multiple versions of the application with updates.
+
+## Reporting a Vulnerability
+
+To report a vulnerability, contact the Invoke team directly at security@invoke.ai
+
+At this time, we do not maintain a formal bug bounty program. 
+
+You can also share identified security issues with our team on huntr.com

docker/Dockerfile

@@ -2,29 +2,42 @@
 
 ## Builder stage
 
-FROM library/ubuntu:23.04 AS builder
+FROM library/ubuntu:24.04 AS builder
 
 ARG DEBIAN_FRONTEND=noninteractive
 RUN rm -f /etc/apt/apt.conf.d/docker-clean; echo 'Binary::apt::APT::Keep-Downloaded-Packages "true";' > /etc/apt/apt.conf.d/keep-cache
 RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
     --mount=type=cache,target=/var/lib/apt,sharing=locked \
     apt update && apt-get install -y \
-        git \
-        python3-venv \
-        python3-pip \
-        build-essential
+        build-essential \
+        git
 
-ENV INVOKEAI_SRC=/opt/invokeai
-ENV VIRTUAL_ENV=/opt/venv/invokeai
+# Install `uv` for package management
+COPY --from=ghcr.io/astral-sh/uv:0.5.5 /uv /uvx /bin/
 
+ENV VIRTUAL_ENV=/opt/venv
 ENV PATH="$VIRTUAL_ENV/bin:$PATH"
+ENV INVOKEAI_SRC=/opt/invokeai
+ENV PYTHON_VERSION=3.11
+ENV UV_COMPILE_BYTECODE=1
+ENV UV_LINK_MODE=copy
+
 ARG GPU_DRIVER=cuda
 ARG TARGETPLATFORM="linux/amd64"
 # unused but available
 ARG BUILDPLATFORM
 
-WORKDIR ${INVOKEAI_SRC}
+# Switch to the `ubuntu` user to work around dependency issues with uv-installed python
+RUN mkdir -p ${VIRTUAL_ENV} && \
+    mkdir -p ${INVOKEAI_SRC} && \
+    chmod -R a+w /opt
+USER ubuntu
 
+# Install python and create the venv
+RUN uv python install ${PYTHON_VERSION} && \
+    uv venv --relocatable --prompt "invoke" --python ${PYTHON_VERSION} ${VIRTUAL_ENV}
+
+WORKDIR ${INVOKEAI_SRC}
 COPY invokeai ./invokeai
 COPY pyproject.toml ./
 
@@ -32,25 +45,18 @@ COPY pyproject.toml ./
 # the local working copy can be bind-mounted into the image
 # at path defined by ${INVOKEAI_SRC}
 # NOTE: there are no pytorch builds for arm64 + cuda, only cpu
-# x86_64/CUDA is default
-RUN --mount=type=cache,target=/root/.cache/pip \
-    python3 -m venv ${VIRTUAL_ENV} &&\
+# x86_64/CUDA is the default
+RUN --mount=type=cache,target=/home/ubuntu/.cache/uv,uid=1000,gid=1000 \
     if [ "$TARGETPLATFORM" = "linux/arm64" ] || [ "$GPU_DRIVER" = "cpu" ]; then \
         extra_index_url_arg="--extra-index-url https://download.pytorch.org/whl/cpu"; \
     elif [ "$GPU_DRIVER" = "rocm" ]; then \
         extra_index_url_arg="--extra-index-url https://download.pytorch.org/whl/rocm6.1"; \
     else \
         extra_index_url_arg="--extra-index-url https://download.pytorch.org/whl/cu124"; \
-    fi &&\
+    fi && \
+    uv pip install --python ${PYTHON_VERSION} $extra_index_url_arg -e "."
 
-    # xformers + triton fails to install on arm64
-    if [ "$GPU_DRIVER" = "cuda" ] && [ "$TARGETPLATFORM" = "linux/amd64" ]; then \
-        pip install $extra_index_url_arg -e ".[xformers]"; \
-    else \
-        pip install $extra_index_url_arg -e "."; \
-    fi
-
-# #### Build the Web UI ------------------------------------
+#### Build the Web UI ------------------------------------
 
 FROM node:20-slim AS web-builder
 ENV PNPM_HOME="/pnpm"
@@ -66,7 +72,7 @@ RUN npx vite build
 
 #### Runtime stage ---------------------------------------
 
-FROM library/ubuntu:23.04 AS runtime
+FROM library/ubuntu:24.04 AS runtime
 
 ARG DEBIAN_FRONTEND=noninteractive
 ENV PYTHONUNBUFFERED=1
@@ -83,17 +89,16 @@ RUN apt update && apt install -y --no-install-recommends \
         gosu \
         magic-wormhole \
         libglib2.0-0 \
-        libgl1-mesa-glx \
-        python3-venv \
-        python3-pip \
+        libgl1 \
+        libglx-mesa0 \
         build-essential \
         libopencv-dev \
         libstdc++-10-dev &&\
     apt-get clean && apt-get autoclean
 
-
 ENV INVOKEAI_SRC=/opt/invokeai
-ENV VIRTUAL_ENV=/opt/venv/invokeai
+ENV VIRTUAL_ENV=/opt/venv
+ENV PYTHON_VERSION=3.11
 ENV INVOKEAI_ROOT=/invokeai
 ENV INVOKEAI_HOST=0.0.0.0
 ENV INVOKEAI_PORT=9090
@@ -101,6 +106,14 @@ ENV PATH="$VIRTUAL_ENV/bin:$INVOKEAI_SRC:$PATH"
 ENV CONTAINER_UID=${CONTAINER_UID:-1000}
 ENV CONTAINER_GID=${CONTAINER_GID:-1000}
 
+# Install `uv` for package management
+# and install python for the ubuntu user (expected to exist on ubuntu >=24.x)
+# this is too tiny to optimize with multi-stage builds, but maybe we'll come back to it
+COPY --from=ghcr.io/astral-sh/uv:0.5.5 /uv /uvx /bin/
+USER ubuntu
+RUN uv python install ${PYTHON_VERSION}
+USER root
+
 # --link requires buldkit w/ dockerfile syntax 1.4
 COPY --link --from=builder ${INVOKEAI_SRC} ${INVOKEAI_SRC}
 COPY --link --from=builder ${VIRTUAL_ENV} ${VIRTUAL_ENV}
@@ -115,7 +128,7 @@ WORKDIR ${INVOKEAI_SRC}
 
 # build patchmatch
 RUN cd /usr/lib/$(uname -p)-linux-gnu/pkgconfig/ && ln -sf opencv4.pc opencv.pc
-RUN python3 -c "from patchmatch import patch_match"
+RUN python -c "from patchmatch import patch_match"
 
 RUN mkdir -p ${INVOKEAI_ROOT} && chown -R ${CONTAINER_UID}:${CONTAINER_GID} ${INVOKEAI_ROOT}
 

docker/docker-entrypoint.sh

@@ -16,6 +16,9 @@ set -e -o pipefail
 
 USER_ID=${CONTAINER_UID:-1000}
 USER=ubuntu
+# if the user does not exist, create it. It is expected to be present on ubuntu >=24.x
+_=$(id ${USER} 2>&1) || useradd -u ${USER_ID} ${USER}
+# ensure the UID is correct
 usermod -u ${USER_ID} ${USER} 1>/dev/null
 
 ### Set the $PUBLIC_KEY env var to enable SSH access.
@@ -36,6 +39,8 @@ fi
 mkdir -p "${INVOKEAI_ROOT}"
 chown --recursive ${USER} "${INVOKEAI_ROOT}" || true
 cd "${INVOKEAI_ROOT}"
+export HF_HOME=${HF_HOME:-$INVOKEAI_ROOT/.cache/huggingface}
+export MPLCONFIGDIR=${MPLCONFIGDIR:-$INVOKEAI_ROOT/.matplotlib}
 
 # Run the CMD as the Container User (not root).
 exec gosu ${USER} "$@"

docs/contributing/ARCHITECTURE.md

@@ -50,7 +50,7 @@ Applications are built on top of the invoke framework. They should construct `in
 
 ### Web UI
 
-The Web UI is built on top of an HTTP API built with [FastAPI](https://fastapi.tiangolo.com/) and [Socket.IO](https://socket.io/). The frontend code is found in `/frontend` and the backend code is found in `/ldm/invoke/app/api_app.py` and `/ldm/invoke/app/api/`. The code is further organized as such:
+The Web UI is built on top of an HTTP API built with [FastAPI](https://fastapi.tiangolo.com/) and [Socket.IO](https://socket.io/). The frontend code is found in `/invokeai/frontend` and the backend code is found in `/invokeai/app/api_app.py` and `/invokeai/app/api/`. The code is further organized as such:
 
 | Component | Description |
 | --- | --- |
@@ -62,7 +62,7 @@ The Web UI is built on top of an HTTP API built with [FastAPI](https://fastapi.t
 
 ### CLI
 
-The CLI is built automatically from invocation metadata, and also supports invocation piping and auto-linking. Code is available in `/ldm/invoke/app/cli_app.py`.
+The CLI is built automatically from invocation metadata, and also supports invocation piping and auto-linking. Code is available in `/invokeai/frontend/cli`.
 
 ## Invoke
 
@@ -70,7 +70,7 @@ The Invoke framework provides the interface to the underlying AI systems and is
 
 ### Invoker
 
-The invoker (`/ldm/invoke/app/services/invoker.py`) is the primary interface through which applications interact with the framework. Its primary purpose is to create, manage, and invoke sessions. It also maintains two sets of services:
+The invoker (`/invokeai/app/services/invoker.py`) is the primary interface through which applications interact with the framework. Its primary purpose is to create, manage, and invoke sessions. It also maintains two sets of services:
 - **invocation services**, which are used by invocations to interact with core functionality.
 - **invoker services**, which are used by the invoker to manage sessions and manage the invocation queue.
 
@@ -82,12 +82,12 @@ The session graph does not support looping. This is left as an application probl
 
 ### Invocations
 
-Invocations represent individual units of execution, with inputs and outputs. All invocations are located in `/ldm/invoke/app/invocations`, and are all automatically discovered and made available in the applications. These are the primary way to expose new functionality in Invoke.AI, and the [implementation guide](INVOCATIONS.md) explains how to add new invocations.
+Invocations represent individual units of execution, with inputs and outputs. All invocations are located in `/invokeai/app/invocations`, and are all automatically discovered and made available in the applications. These are the primary way to expose new functionality in Invoke.AI, and the [implementation guide](INVOCATIONS.md) explains how to add new invocations.
 
 ### Services
 
-Services provide invocations access AI Core functionality and other necessary functionality (e.g. image storage). These are available in `/ldm/invoke/app/services`. As a general rule, new services should provide an interface as an abstract base class, and may provide a lightweight local implementation by default in their module. The goal for all services should be to enable the usage of different implementations (e.g. using cloud storage for image storage), but should not load any module dependencies unless that implementation has been used (i.e. don't import anything that won't be used, especially if it's expensive to import).
+Services provide invocations access AI Core functionality and other necessary functionality (e.g. image storage). These are available in `/invokeai/app/services`. As a general rule, new services should provide an interface as an abstract base class, and may provide a lightweight local implementation by default in their module. The goal for all services should be to enable the usage of different implementations (e.g. using cloud storage for image storage), but should not load any module dependencies unless that implementation has been used (i.e. don't import anything that won't be used, especially if it's expensive to import).
 
 ## AI Core
 
-The AI Core is represented by the rest of the code base (i.e. the code outside of `/ldm/invoke/app/`).
+The AI Core is represented by the rest of the code base (i.e. the code outside of `/invokeai/app/`).

docs/contributing/INVOCATIONS.md

@@ -287,8 +287,8 @@ new Invocation ready to be used.
 
 Once you've created a Node, the next step is to share it with the community! The
 best way to do this is to submit a Pull Request to add the Node to the
-[Community Nodes](nodes/communityNodes) list. If you're not sure how to do that,
-take a look a at our [contributing nodes overview](contributingNodes).
+[Community Nodes](../nodes/communityNodes.md) list. If you're not sure how to do that,
+take a look a at our [contributing nodes overview](../nodes/contributingNodes.md).
 
 ## Advanced
 

docs/contributing/MODEL_MANAGER.md

@@ -9,20 +9,20 @@ model. These are the:
   configuration information. Among other things, the record service
   tracks the type of the model, its provenance, and where it can be
   found on disk.
-  
+
 * _ModelInstallServiceBase_ A service for installing models to
   disk. It uses `DownloadQueueServiceBase` to download models and
   their metadata, and `ModelRecordServiceBase` to store that
   information. It is also responsible for managing the InvokeAI
   `models` directory and its contents.
-  
+
 * _DownloadQueueServiceBase_
   A multithreaded downloader responsible
   for downloading models from a remote source to disk. The download
   queue has special methods for downloading repo_id folders from
   Hugging Face, as well as discriminating among model versions in
   Civitai, but can be used for arbitrary content.
-  
+
   * _ModelLoadServiceBase_
   Responsible for loading a model from disk
   into RAM and VRAM and getting it ready for inference.
@@ -207,9 +207,9 @@ for use in the InvokeAI web server. Its signature is:
 
 ```
 def open(
-       cls, 
-    config: InvokeAIAppConfig, 
-    conn: Optional[sqlite3.Connection] = None, 
+       cls,
+    config: InvokeAIAppConfig,
+    conn: Optional[sqlite3.Connection] = None,
     lock: Optional[threading.Lock] = None
     ) -> Union[ModelRecordServiceSQL, ModelRecordServiceFile]:
 ```
@@ -363,7 +363,7 @@ functionality:
 
 * Registering a model config record for a model already located on the
   local filesystem, without moving it or changing its path.
-  
+
 * Installing a model alreadiy located on the local filesystem, by
   moving it into the InvokeAI root directory under the
   `models` folder (or wherever config parameter `models_dir`
@@ -371,21 +371,21 @@ functionality:
 
 * Probing of models to determine their type, base type and other key
   information.
-  
+
 * Interface with the InvokeAI event bus to provide status updates on
   the download, installation and registration process.
-  
+
 * Downloading a model from an arbitrary URL and installing it in
   `models_dir`.
 
 * 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 HuggingFace).
-  
+
 ### Initializing the installer
 
 A default installer is created at InvokeAI api startup time and stored
@@ -461,7 +461,7 @@ revision.
 `config` is an optional dict of values that will override the
 autoprobed values for model type, base, scheduler prediction type, and
 so forth. See [Model configuration and
-probing](#Model-configuration-and-probing) for details.
+probing](#model-configuration-and-probing) for details.
 
 `access_token` is an optional access token for accessing resources
 that need authentication.
@@ -494,7 +494,7 @@ source8 = URLModelSource(url='https://civitai.com/api/download/models/63006', ac
 
 for source in [source1, source2, source3, source4, source5, source6, source7]:
    install_job = installer.install_model(source)
-   
+
 source2job = installer.wait_for_installs(timeout=120)
 for source in sources:
     job = source2job[source]
@@ -504,7 +504,7 @@ for source in sources:
   print(f"{source} installed as {model_key}")
  elif job.errored:
      print(f"{source}: {job.error_type}.\nStack trace:\n{job.error}")
- 
+
 ```
 
 As shown here, the `import_model()` method accepts a variety of

docs/contributing/TESTS.md

@@ -1,6 +1,6 @@
 # InvokeAI Backend Tests
 
-We use `pytest` to run the backend python tests. (See [pyproject.toml](/pyproject.toml) for the default `pytest` options.)
+We use `pytest` to run the backend python tests. (See [pyproject.toml](https://github.com/invoke-ai/InvokeAI/blob/main/pyproject.toml) for the default `pytest` options.)
 
 ## Fast vs. Slow
 All tests are categorized as either 'fast' (no test annotation) or 'slow' (annotated with the `@pytest.mark.slow` decorator).
@@ -33,7 +33,7 @@ pytest tests -m ""
 
 ## Test Organization
 
-All backend tests are in the [`tests/`](/tests/) directory. This directory mirrors the organization of the `invokeai/` directory. For example, tests for `invokeai/model_management/model_manager.py` would be found in `tests/model_management/test_model_manager.py`.
+All backend tests are in the [`tests/`](https://github.com/invoke-ai/InvokeAI/tree/main/tests) directory. This directory mirrors the organization of the `invokeai/` directory. For example, tests for `invokeai/model_management/model_manager.py` would be found in `tests/model_management/test_model_manager.py`.
 
 TODO: The above statement is aspirational. A re-organization of legacy tests is required to make it true.
 

docs/contributing/contribution_guides/development.md

@@ -2,7 +2,7 @@
 
 ## **What do I need to know to help?**
 
-If you are looking to help with a code contribution, InvokeAI uses several different technologies under the hood: Python (Pydantic, FastAPI, diffusers) and Typescript (React, Redux Toolkit, ChakraUI, Mantine, Konva). Familiarity with StableDiffusion and image generation concepts is helpful, but not essential. 
+If you are looking to help with a code contribution, InvokeAI uses several different technologies under the hood: Python (Pydantic, FastAPI, diffusers) and Typescript (React, Redux Toolkit, ChakraUI, Mantine, Konva). Familiarity with StableDiffusion and image generation concepts is helpful, but not essential.
 
 
 ## **Get Started**
@@ -12,7 +12,7 @@ To get started, take a look at our [new contributors checklist](newContributorCh
 Once you're setup, for more information, you can review the documentation specific to your area of interest:
 
 * #### [InvokeAI Architecure](../ARCHITECTURE.md)
-* #### [Frontend Documentation](https://github.com/invoke-ai/InvokeAI/tree/main/invokeai/frontend/web)
+* #### [Frontend Documentation](../frontend/index.md)
 * #### [Node Documentation](../INVOCATIONS.md)
 * #### [Local Development](../LOCAL_DEVELOPMENT.md)
 
@@ -20,15 +20,15 @@ Once you're setup, for more information, you can review the documentation specif
 
 If you don't feel ready to make a code contribution yet, no problem! You can also help out in other ways, such as [documentation](documentation.md), [translation](translation.md) or helping support other users and triage issues as they're reported in GitHub.
 
-There are two paths to making a development contribution: 
+There are two paths to making a development contribution:
 
 1. Choosing an open issue to address. Open issues can be found in the [Issues](https://github.com/invoke-ai/InvokeAI/issues?q=is%3Aissue+is%3Aopen) section of the InvokeAI repository. These are tagged by the issue type (bug, enhancement, etc.) along with the “good first issues” tag denoting if they are suitable for first time contributors.
-    1. Additional items can be found on our [roadmap](https://github.com/orgs/invoke-ai/projects/7). The roadmap is organized in terms of priority, and contains features of varying size and complexity. If there is an inflight item you’d like to help with, reach out to the contributor assigned to the item to see how you can help. 
+    1. Additional items can be found on our [roadmap](https://github.com/orgs/invoke-ai/projects/7). The roadmap is organized in terms of priority, and contains features of varying size and complexity. If there is an inflight item you’d like to help with, reach out to the contributor assigned to the item to see how you can help.
 2. Opening a new issue or feature to add. **Please make sure you have searched through existing issues before creating new ones.**
 
 *Regardless of what you choose, please post in the  [#dev-chat](https://discord.com/channels/1020123559063990373/1049495067846524939) channel of the Discord before you start development in order to confirm that the issue or feature is aligned with the current direction of the project. We value our contributors time and effort and want to ensure that no one’s time is being misspent.*
 
-## Best Practices: 
+## Best Practices:
 * Keep your pull requests small. Smaller pull requests are more likely to be accepted and merged
 * Comments! Commenting your code helps reviewers easily understand your contribution
 * Use Python and Typescript’s typing systems, and consider using an editor with [LSP](https://microsoft.github.io/language-server-protocol/) support to streamline development
@@ -38,7 +38,7 @@ There are two paths to making a development contribution:
 
 If you need help, you can ask questions in the [#dev-chat](https://discord.com/channels/1020123559063990373/1049495067846524939) channel of the Discord.
 
-For frontend related work, **@psychedelicious** is the best person to reach out to. 
+For frontend related work, **@psychedelicious** is the best person to reach out to.
 
 For backend related work, please reach out to **@blessedcoolant**, **@lstein**, **@StAlKeR7779** or **@psychedelicious**.
 

docs/contributing/contribution_guides/newContributorChecklist.md

@@ -5,7 +5,7 @@ If you're a new contributor to InvokeAI or Open Source Projects, this is the gui
 ## New Contributor Checklist
 
 - [x] Set up your local development environment & fork of InvokAI by following [the steps outlined here](../dev-environment.md)
-- [x] Set up your local tooling with [this guide](InvokeAI/contributing/LOCAL_DEVELOPMENT/#developing-invokeai-in-vscode). Feel free to skip this step if you already have tooling you're comfortable with.
+- [x] Set up your local tooling with [this guide](../LOCAL_DEVELOPMENT.md). Feel free to skip this step if you already have tooling you're comfortable with.
 - [x] Familiarize yourself with [Git](https://www.atlassian.com/git) & our project structure by reading through the [development documentation](development.md)
 - [x] Join the [#dev-chat](https://discord.com/channels/1020123559063990373/1049495067846524939) channel of the Discord
 - [x] Choose an issue to work on! This can be achieved by asking in the #dev-chat channel, tackling a [good first issue](https://github.com/invoke-ai/InvokeAI/contribute) or finding an item on the [roadmap](https://github.com/orgs/invoke-ai/projects/7). If nothing in any of those places catches your eye, feel free to work on something of interest to you!
@@ -22,15 +22,15 @@ Before starting these steps, ensure you have your local environment [configured
 2. Fork the [InvokeAI](https://github.com/invoke-ai/InvokeAI) repository to your GitHub profile. This means that you will have a copy of the repository under **your-GitHub-username/InvokeAI**.
 3. Clone the repository to your local machine using:
 
-   ```bash
-   git clone https://github.com/your-GitHub-username/InvokeAI.git
-   ```
+    ```bash
+    git clone https://github.com/your-GitHub-username/InvokeAI.git
+    ```
 
 If you're unfamiliar with using Git through the commandline, [GitHub Desktop](https://desktop.github.com) is a easy-to-use alternative with a UI. You can do all the same steps listed here, but through the interface. 4. Create a new branch for your fix using:
 
-    ```bash
-    git checkout -b branch-name-here
-    ```
+  ```bash
+  git checkout -b branch-name-here
+  ```
 
 5. Make the appropriate changes for the issue you are trying to address or the feature that you want to add.
 6. Add the file contents of the changed files to the "snapshot" git uses to manage the state of the project, also known as the index:

docs/contributing/dev-environment.md

@@ -27,9 +27,9 @@ If you just want to use Invoke, you should use the [installer][installer link].
 
 5. Activate the venv (you'll need to do this every time you want to run the app):
 
-        ```sh
-        source .venv/bin/activate
-        ```
+      ```sh
+      source .venv/bin/activate
+      ```
 
 6. Install the repo as an [editable install][editable install link]:
 
@@ -37,7 +37,7 @@ If you just want to use Invoke, you should use the [installer][installer link].
       pip install -e ".[dev,test,xformers]" --use-pep517 --extra-index-url https://download.pytorch.org/whl/cu121
       ```
 
-      Refer to the [manual installation][manual install link]] instructions for more determining the correct install options. `xformers` is optional, but `dev` and `test` are not.
+      Refer to the [manual installation][manual install link] instructions for more determining the correct install options. `xformers` is optional, but `dev` and `test` are not.
 
 7. Install the frontend dev toolchain:
 

docs/contributing/index.md

@@ -34,11 +34,11 @@ Please reach out to @hipsterusername on [Discord](https://discord.gg/ZmtBAhwWhy)
 
 ## Contributors
 
-This project is a combined effort of dedicated people from across the world. [Check out the list of all these amazing people](https://invoke-ai.github.io/InvokeAI/other/CONTRIBUTORS/). We thank them for their time, hard work and effort.
+This project is a combined effort of dedicated people from across the world. [Check out the list of all these amazing people](contributors.md). We thank them for their time, hard work and effort.
 
 ## Code of Conduct
 
-The InvokeAI community is a welcoming place, and we want your help in maintaining that. Please review our [Code of Conduct](https://github.com/invoke-ai/InvokeAI/blob/main/CODE_OF_CONDUCT.md) to learn more - it's essential to maintaining a respectful and inclusive environment.
+The InvokeAI community is a welcoming place, and we want your help in maintaining that. Please review our [Code of Conduct](../CODE_OF_CONDUCT.md) to learn more - it's essential to maintaining a respectful and inclusive environment.
 
 By making a contribution to this project, you certify that:
 

docs/faq.md

@@ -209,7 +209,7 @@ checkpoint models.
 
 To solve this, go to the Model Manager tab (the cube), select the
 checkpoint model that's giving you trouble, and press the "Convert"
-button in the upper right of your browser window. This will conver the
+button in the upper right of your browser window. This will convert the
 checkpoint into a diffusers model, after which loading should be
 faster and less memory-intensive.
 

docs/installation/patchmatch.md

@@ -97,16 +97,16 @@ Prior to installing PyPatchMatch, you need to take the following steps:
    sudo pacman -S --needed base-devel
    ```
 
-2. Install `opencv` and `blas`:
+2. Install `opencv`, `blas`, and required dependencies:
 
    ```sh
-   sudo pacman -S opencv blas
+   sudo pacman -S opencv blas fmt glew vtk hdf5
    ```
 
    or for CUDA support
 
    ```sh
-   sudo pacman -S opencv-cuda blas
+   sudo pacman -S opencv-cuda blas fmt glew vtk hdf5
    ```
 
 3. Fix the naming of the `opencv` package configuration file:

docs/nodes/defaultNodes.md

@@ -99,7 +99,6 @@ their descriptions.
 | Scale Latents                                                 | Scales latents by a given factor.                                                                                                                    |
 | Segment Anything Processor                                    | Applies segment anything processing to image                                                                                                         |
 | Show Image                                                    | Displays a provided image, and passes it forward in the pipeline.                                                                                    |
-| Step Param Easing                                             | Experimental per-step parameter easing for denoising steps                                                                                           |
 | String Primitive Collection                                   | A collection of string primitive values                                                                                                              |
 | String Primitive                                              | A string primitive value                                                                                                                             |
 | Subtract Integers                                             | Subtracts two numbers                                                                                                                                |

installer/lib/messages.py

@@ -259,7 +259,7 @@ def select_gpu() -> GpuType:
                     [
                         f"Detected the [gold1]{OS}-{ARCH}[/] platform",
                         "",
-                        "See [deep_sky_blue1]https://invoke-ai.github.io/InvokeAI/#system[/] to ensure your system meets the minimum requirements.",
+                        "See [deep_sky_blue1]https://invoke-ai.github.io/InvokeAI/installation/requirements/[/] to ensure your system meets the minimum requirements.",
                         "",
                         "[red3]🠶[/] [b]Your GPU drivers must be correctly installed before using InvokeAI![/] [red3]🠴[/]",
                     ]

installer/templates/invoke.sh.in

@@ -68,7 +68,7 @@ do_line_input() {
     printf "2: Open the developer console\n"
     printf "3: Command-line help\n"
     printf "Q: Quit\n\n"
-    printf "To update, download and run the installer from https://github.com/invoke-ai/InvokeAI/releases/latest.\n\n"
+    printf "To update, download and run the installer from https://github.com/invoke-ai/InvokeAI/releases/latest\n\n"
     read -p "Please enter 1-4, Q: [1] " yn
     choice=${yn:='1'}
     do_choice $choice

invokeai/app/api/routers/app_info.py

@@ -40,6 +40,8 @@ class AppVersion(BaseModel):
 
     version: str = Field(description="App version")
 
+    highlights: Optional[list[str]] = Field(default=None, description="Highlights of release")
+
 
 class AppDependencyVersions(BaseModel):
     """App depencency Versions Response"""

invokeai/app/api/routers/model_manager.py

@@ -1,6 +1,7 @@
 # Copyright (c) 2023 Lincoln D. Stein
 """FastAPI route for model configuration records."""
 
+import contextlib
 import io
 import pathlib
 import shutil
@@ -10,6 +11,7 @@ from enum import Enum
 from tempfile import TemporaryDirectory
 from typing import List, Optional, Type
 
+import huggingface_hub
 from fastapi import Body, Path, Query, Response, UploadFile
 from fastapi.responses import FileResponse, HTMLResponse
 from fastapi.routing import APIRouter
@@ -27,6 +29,7 @@ from invokeai.app.services.model_records import (
     ModelRecordChanges,
     UnknownModelException,
 )
+from invokeai.app.util.suppress_output import SuppressOutput
 from invokeai.backend.model_manager.config import (
     AnyModelConfig,
     BaseModelType,
@@ -923,3 +926,51 @@ async def get_stats() -> Optional[CacheStats]:
     """Return performance statistics on the model manager's RAM cache. Will return null if no models have been loaded."""
 
     return ApiDependencies.invoker.services.model_manager.load.ram_cache.stats
+
+
+class HFTokenStatus(str, Enum):
+    VALID = "valid"
+    INVALID = "invalid"
+    UNKNOWN = "unknown"
+
+
+class HFTokenHelper:
+    @classmethod
+    def get_status(cls) -> HFTokenStatus:
+        try:
+            if huggingface_hub.get_token_permission(huggingface_hub.get_token()):
+                # Valid token!
+                return HFTokenStatus.VALID
+            # No token set
+            return HFTokenStatus.INVALID
+        except Exception:
+            return HFTokenStatus.UNKNOWN
+
+    @classmethod
+    def set_token(cls, token: str) -> HFTokenStatus:
+        with SuppressOutput(), contextlib.suppress(Exception):
+            huggingface_hub.login(token=token, add_to_git_credential=False)
+        return cls.get_status()
+
+
+@model_manager_router.get("/hf_login", operation_id="get_hf_login_status", response_model=HFTokenStatus)
+async def get_hf_login_status() -> HFTokenStatus:
+    token_status = HFTokenHelper.get_status()
+
+    if token_status is HFTokenStatus.UNKNOWN:
+        ApiDependencies.invoker.services.logger.warning("Unable to verify HF token")
+
+    return token_status
+
+
+@model_manager_router.post("/hf_login", operation_id="do_hf_login", response_model=HFTokenStatus)
+async def do_hf_login(
+    token: str = Body(description="Hugging Face token to use for login", embed=True),
+) -> HFTokenStatus:
+    HFTokenHelper.set_token(token)
+    token_status = HFTokenHelper.get_status()
+
+    if token_status is HFTokenStatus.UNKNOWN:
+        ApiDependencies.invoker.services.logger.warning("Unable to verify HF token")
+
+    return token_status

invokeai/app/api/routers/session_queue.py

@@ -110,7 +110,7 @@ async def cancel_by_batch_ids(
 @session_queue_router.put(
     "/{queue_id}/cancel_by_destination",
     operation_id="cancel_by_destination",
-    responses={200: {"model": CancelByBatchIDsResult}},
+    responses={200: {"model": CancelByDestinationResult}},
 )
 async def cancel_by_destination(
     queue_id: str = Path(description="The queue id to perform this operation on"),

invokeai/app/api_app.py

@@ -59,11 +59,32 @@ logger.info(f"Using torch device: {torch_device_name}")
 
 loop = asyncio.new_event_loop()
 
+# We may change the port if the default is in use, this global variable is used to store the port so that we can log
+# the correct port when the server starts in the lifespan handler.
+port = app_config.port
+
 
 @asynccontextmanager
 async def lifespan(app: FastAPI):
     # Add startup event to load dependencies
     ApiDependencies.initialize(config=app_config, event_handler_id=event_handler_id, loop=loop, logger=logger)
+
+    # Log the server address when it starts - in case the network log level is not high enough to see the startup log
+    proto = "https" if app_config.ssl_certfile else "http"
+    msg = f"Invoke running on {proto}://{app_config.host}:{port} (Press CTRL+C to quit)"
+
+    # Logging this way ignores the logger's log level and _always_ logs the message
+    record = logger.makeRecord(
+        name=logger.name,
+        level=logging.INFO,
+        fn="",
+        lno=0,
+        msg=msg,
+        args=(),
+        exc_info=None,
+    )
+    logger.handle(record)
+
     yield
     # Shut down threads
     ApiDependencies.shutdown()
@@ -206,6 +227,7 @@ def invoke_api() -> None:
         else:
             jurigged.watch(logger=InvokeAILogger.get_logger(name="jurigged").info)
 
+    global port
     port = find_port(app_config.port)
     if port != app_config.port:
         logger.warn(f"Port {app_config.port} in use, using port {port}")
@@ -217,18 +239,17 @@ def invoke_api() -> None:
         host=app_config.host,
         port=port,
         loop="asyncio",
-        log_level=app_config.log_level,
+        log_level=app_config.log_level_network,
         ssl_certfile=app_config.ssl_certfile,
         ssl_keyfile=app_config.ssl_keyfile,
     )
     server = uvicorn.Server(config)
 
     # replace uvicorn's loggers with InvokeAI's for consistent appearance
-    for logname in ["uvicorn.access", "uvicorn"]:
-        log = InvokeAILogger.get_logger(logname)
-        log.handlers.clear()
-        for ch in logger.handlers:
-            log.addHandler(ch)
+    uvicorn_logger = InvokeAILogger.get_logger("uvicorn")
+    uvicorn_logger.handlers.clear()
+    for hdlr in logger.handlers:
+        uvicorn_logger.addHandler(hdlr)
 
     loop.run_until_complete(server.serve())
 

invokeai/app/invocations/__init__.py

@@ -15,6 +15,11 @@ custom_nodes_readme_path = str(custom_nodes_path / "README.md")
 shutil.copy(Path(__file__).parent / "custom_nodes/init.py", custom_nodes_init_path)
 shutil.copy(Path(__file__).parent / "custom_nodes/README.md", custom_nodes_readme_path)
 
+# set the same permissions as the destination directory, in case our source is read-only,
+# so that the files are user-writable
+for p in custom_nodes_path.glob("**/*"):
+    p.chmod(custom_nodes_path.stat().st_mode)
+
 # Import custom nodes, see https://docs.python.org/3/library/importlib.html#importing-programmatically
 spec = spec_from_file_location("custom_nodes", custom_nodes_init_path)
 if spec is None or spec.loader is None:

invokeai/app/invocations/baseinvocation.py

@@ -4,6 +4,7 @@ from __future__ import annotations
 
 import inspect
 import re
+import sys
 import warnings
 from abc import ABC, abstractmethod
 from enum import Enum
@@ -62,6 +63,7 @@ class Classification(str, Enum, metaclass=MetaEnum):
     - `Prototype`: The invocation is not yet stable and may be removed from the application at any time. Workflows built around this invocation may break, and we are *not* committed to supporting this invocation.
     - `Deprecated`: The invocation is deprecated and may be removed in a future version.
     - `Internal`: The invocation is not intended for use by end-users. It may be changed or removed at any time, but is exposed for users to play with.
+    - `Special`: The invocation is a special case and does not fit into any of the other classifications.
     """
 
     Stable = "stable"
@@ -69,6 +71,7 @@ class Classification(str, Enum, metaclass=MetaEnum):
     Prototype = "prototype"
     Deprecated = "deprecated"
     Internal = "internal"
+    Special = "special"
 
 
 class UIConfigBase(BaseModel):
@@ -192,12 +195,19 @@ class BaseInvocation(ABC, BaseModel):
         """Gets a pydantc TypeAdapter for the union of all invocation types."""
         if not cls._typeadapter or cls._typeadapter_needs_update:
             AnyInvocation = TypeAliasType(
-                "AnyInvocation", Annotated[Union[tuple(cls._invocation_classes)], Field(discriminator="type")]
+                "AnyInvocation", Annotated[Union[tuple(cls.get_invocations())], Field(discriminator="type")]
             )
             cls._typeadapter = TypeAdapter(AnyInvocation)
             cls._typeadapter_needs_update = False
         return cls._typeadapter
 
+    @classmethod
+    def invalidate_typeadapter(cls) -> None:
+        """Invalidates the typeadapter, forcing it to be rebuilt on next access. If the invocation allowlist or
+        denylist is changed, this should be called to ensure the typeadapter is updated and validation respects
+        the updated allowlist and denylist."""
+        cls._typeadapter_needs_update = True
+
     @classmethod
     def get_invocations(cls) -> Iterable[BaseInvocation]:
         """Gets all invocations, respecting the allowlist and denylist."""
@@ -479,6 +489,26 @@ def invocation(
             title="type", default=invocation_type, json_schema_extra={"field_kind": FieldKind.NodeAttribute}
         )
 
+        # Validate the `invoke()` method is implemented
+        if "invoke" in cls.__abstractmethods__:
+            raise ValueError(f'Invocation "{invocation_type}" must implement the "invoke" method')
+
+        # And validate that `invoke()` returns a subclass of `BaseInvocationOutput
+        invoke_return_annotation = signature(cls.invoke).return_annotation
+
+        try:
+            # TODO(psyche): If `invoke()` is not defined, `return_annotation` ends up as the string "BaseInvocationOutput"
+            # instead of the class `BaseInvocationOutput`. This may be a pydantic bug: https://github.com/pydantic/pydantic/issues/7978
+            if isinstance(invoke_return_annotation, str):
+                invoke_return_annotation = getattr(sys.modules[cls.__module__], invoke_return_annotation)
+
+            assert invoke_return_annotation is not BaseInvocationOutput
+            assert issubclass(invoke_return_annotation, BaseInvocationOutput)
+        except Exception:
+            raise ValueError(
+                f'Invocation "{invocation_type}" must have a return annotation of a subclass of BaseInvocationOutput (got "{invoke_return_annotation}")'
+            )
+
         docstring = cls.__doc__
         cls = create_model(
             cls.__qualname__,

invokeai/app/invocations/blend_latents.py

@@ -1,98 +1,120 @@
-from typing import Any, Union
+from typing import Optional, Union
 
 import numpy as np
-import numpy.typing as npt
 import torch
+import torchvision.transforms as T
+from PIL import Image
+from torchvision.transforms.functional import resize as tv_resize
 
 from invokeai.app.invocations.baseinvocation import BaseInvocation, invocation
-from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, LatentsField
+from invokeai.app.invocations.fields import FieldDescriptions, ImageField, Input, InputField, LatentsField
 from invokeai.app.invocations.primitives import LatentsOutput
 from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.stable_diffusion.diffusers_pipeline import image_resized_to_grid_as_tensor
 from invokeai.backend.util.devices import TorchDevice
 
 
+def slerp(
+    t: Union[float, np.ndarray],
+    v0: Union[torch.Tensor, np.ndarray],
+    v1: Union[torch.Tensor, np.ndarray],
+    device: torch.device,
+    DOT_THRESHOLD: float = 0.9995,
+):
+    """
+    Spherical linear interpolation
+    Args:
+        t (float/np.ndarray): Float value between 0.0 and 1.0
+        v0 (np.ndarray): Starting vector
+        v1 (np.ndarray): Final vector
+        DOT_THRESHOLD (float): Threshold for considering the two vectors as
+                            colineal. Not recommended to alter this.
+    Returns:
+        v2 (np.ndarray): Interpolation vector between v0 and v1
+    """
+    inputs_are_torch = False
+    if not isinstance(v0, np.ndarray):
+        inputs_are_torch = True
+        v0 = v0.detach().cpu().numpy()
+    if not isinstance(v1, np.ndarray):
+        inputs_are_torch = True
+        v1 = v1.detach().cpu().numpy()
+
+    dot = np.sum(v0 * v1 / (np.linalg.norm(v0) * np.linalg.norm(v1)))
+    if np.abs(dot) > DOT_THRESHOLD:
+        v2 = (1 - t) * v0 + t * v1
+    else:
+        theta_0 = np.arccos(dot)
+        sin_theta_0 = np.sin(theta_0)
+        theta_t = theta_0 * t
+        sin_theta_t = np.sin(theta_t)
+        s0 = np.sin(theta_0 - theta_t) / sin_theta_0
+        s1 = sin_theta_t / sin_theta_0
+        v2 = s0 * v0 + s1 * v1
+
+    if inputs_are_torch:
+        v2 = torch.from_numpy(v2).to(device)
+
+    return v2
+
+
 @invocation(
     "lblend",
     title="Blend Latents",
-    tags=["latents", "blend"],
+    tags=["latents", "blend", "mask"],
     category="latents",
-    version="1.0.3",
+    version="1.1.0",
 )
 class BlendLatentsInvocation(BaseInvocation):
-    """Blend two latents using a given alpha. Latents must have same size."""
+    """Blend two latents using a given alpha. If a mask is provided, the second latents will be masked before blending.
+    Latents must have same size. Masking functionality added by @dwringer."""
 
-    latents_a: LatentsField = InputField(
-        description=FieldDescriptions.latents,
-        input=Input.Connection,
-    )
-    latents_b: LatentsField = InputField(
-        description=FieldDescriptions.latents,
-        input=Input.Connection,
-    )
-    alpha: float = InputField(default=0.5, description=FieldDescriptions.blend_alpha)
+    latents_a: LatentsField = InputField(description=FieldDescriptions.latents, input=Input.Connection)
+    latents_b: LatentsField = InputField(description=FieldDescriptions.latents, input=Input.Connection)
+    mask: Optional[ImageField] = InputField(default=None, description="Mask for blending in latents B")
+    alpha: float = InputField(ge=0, default=0.5, description=FieldDescriptions.blend_alpha)
+
+    def prep_mask_tensor(self, mask_image: Image.Image) -> torch.Tensor:
+        if mask_image.mode != "L":
+            mask_image = mask_image.convert("L")
+        mask_tensor = image_resized_to_grid_as_tensor(mask_image, normalize=False)
+        if mask_tensor.dim() == 3:
+            mask_tensor = mask_tensor.unsqueeze(0)
+        return mask_tensor
+
+    def replace_tensor_from_masked_tensor(
+        self, tensor: torch.Tensor, other_tensor: torch.Tensor, mask_tensor: torch.Tensor
+    ):
+        output = tensor.clone()
+        mask_tensor = mask_tensor.expand(output.shape)
+        if output.dtype != torch.float16:
+            output = torch.add(output, mask_tensor * torch.sub(other_tensor, tensor))
+        else:
+            output = torch.add(output, mask_tensor.half() * torch.sub(other_tensor, tensor))
+        return output
 
     def invoke(self, context: InvocationContext) -> LatentsOutput:
         latents_a = context.tensors.load(self.latents_a.latents_name)
         latents_b = context.tensors.load(self.latents_b.latents_name)
+        if self.mask is None:
+            mask_tensor = torch.zeros(latents_a.shape[-2:])
+        else:
+            mask_tensor = self.prep_mask_tensor(context.images.get_pil(self.mask.image_name))
+            mask_tensor = tv_resize(mask_tensor, latents_a.shape[-2:], T.InterpolationMode.BILINEAR, antialias=False)
+
+        latents_b = self.replace_tensor_from_masked_tensor(latents_b, latents_a, mask_tensor)
 
         if latents_a.shape != latents_b.shape:
-            raise Exception("Latents to blend must be the same size.")
+            raise ValueError("Latents to blend must be the same size.")
 
         device = TorchDevice.choose_torch_device()
 
-        def slerp(
-            t: Union[float, npt.NDArray[Any]],  # FIXME: maybe use np.float32 here?
-            v0: Union[torch.Tensor, npt.NDArray[Any]],
-            v1: Union[torch.Tensor, npt.NDArray[Any]],
-            DOT_THRESHOLD: float = 0.9995,
-        ) -> Union[torch.Tensor, npt.NDArray[Any]]:
-            """
-            Spherical linear interpolation
-            Args:
-                t (float/np.ndarray): Float value between 0.0 and 1.0
-                v0 (np.ndarray): Starting vector
-                v1 (np.ndarray): Final vector
-                DOT_THRESHOLD (float): Threshold for considering the two vectors as
-                                    colineal. Not recommended to alter this.
-            Returns:
-                v2 (np.ndarray): Interpolation vector between v0 and v1
-            """
-            inputs_are_torch = False
-            if not isinstance(v0, np.ndarray):
-                inputs_are_torch = True
-                v0 = v0.detach().cpu().numpy()
-            if not isinstance(v1, np.ndarray):
-                inputs_are_torch = True
-                v1 = v1.detach().cpu().numpy()
-
-            dot = np.sum(v0 * v1 / (np.linalg.norm(v0) * np.linalg.norm(v1)))
-            if np.abs(dot) > DOT_THRESHOLD:
-                v2 = (1 - t) * v0 + t * v1
-            else:
-                theta_0 = np.arccos(dot)
-                sin_theta_0 = np.sin(theta_0)
-                theta_t = theta_0 * t
-                sin_theta_t = np.sin(theta_t)
-                s0 = np.sin(theta_0 - theta_t) / sin_theta_0
-                s1 = sin_theta_t / sin_theta_0
-                v2 = s0 * v0 + s1 * v1
-
-            if inputs_are_torch:
-                v2_torch: torch.Tensor = torch.from_numpy(v2).to(device)
-                return v2_torch
-            else:
-                assert isinstance(v2, np.ndarray)
-                return v2
-
         # blend
-        bl = slerp(self.alpha, latents_a, latents_b)
-        assert isinstance(bl, torch.Tensor)
-        blended_latents: torch.Tensor = bl  # for type checking convenience
+        blended_latents = slerp(self.alpha, latents_a, latents_b, device)
 
         # https://discuss.huggingface.co/t/memory-usage-by-later-pipeline-stages/23699
         blended_latents = blended_latents.to("cpu")
-
-        TorchDevice.empty_cache()
+        torch.cuda.empty_cache()
 
         name = context.tensors.save(tensor=blended_latents)
-        return LatentsOutput.build(latents_name=name, latents=blended_latents, seed=self.latents_a.seed)
+        return LatentsOutput.build(latents_name=name, latents=blended_latents)

invokeai/app/invocations/compel.py

@@ -19,9 +19,9 @@ from invokeai.app.invocations.model import CLIPField
 from invokeai.app.invocations.primitives import ConditioningOutput
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.util.ti_utils import generate_ti_list
-from invokeai.backend.lora.lora_model_raw import LoRAModelRaw
-from invokeai.backend.lora.lora_patcher import LoRAPatcher
 from invokeai.backend.model_patcher import ModelPatcher
+from invokeai.backend.patches.model_patch_raw import ModelPatchRaw
+from invokeai.backend.patches.model_patcher import LayerPatcher
 from invokeai.backend.stable_diffusion.diffusion.conditioning_data import (
     BasicConditioningInfo,
     ConditioningFieldData,
@@ -66,10 +66,10 @@ class CompelInvocation(BaseInvocation):
         tokenizer_info = context.models.load(self.clip.tokenizer)
         text_encoder_info = context.models.load(self.clip.text_encoder)
 
-        def _lora_loader() -> Iterator[Tuple[LoRAModelRaw, float]]:
+        def _lora_loader() -> Iterator[Tuple[ModelPatchRaw, float]]:
             for lora in self.clip.loras:
                 lora_info = context.models.load(lora.lora)
-                assert isinstance(lora_info.model, LoRAModelRaw)
+                assert isinstance(lora_info.model, ModelPatchRaw)
                 yield (lora_info.model, lora.weight)
                 del lora_info
             return
@@ -82,7 +82,7 @@ class CompelInvocation(BaseInvocation):
             # apply all patches while the model is on the target device
             text_encoder_info.model_on_device() as (cached_weights, text_encoder),
             tokenizer_info as tokenizer,
-            LoRAPatcher.apply_lora_patches(
+            LayerPatcher.apply_model_patches(
                 model=text_encoder,
                 patches=_lora_loader(),
                 prefix="lora_te_",
@@ -95,6 +95,7 @@ class CompelInvocation(BaseInvocation):
                 ti_manager,
             ),
         ):
+            context.util.signal_progress("Building conditioning")
             assert isinstance(text_encoder, CLIPTextModel)
             assert isinstance(tokenizer, CLIPTokenizer)
             compel = Compel(
@@ -161,11 +162,11 @@ class SDXLPromptInvocationBase:
                 c_pooled = None
             return c, c_pooled
 
-        def _lora_loader() -> Iterator[Tuple[LoRAModelRaw, float]]:
+        def _lora_loader() -> Iterator[Tuple[ModelPatchRaw, float]]:
             for lora in clip_field.loras:
                 lora_info = context.models.load(lora.lora)
                 lora_model = lora_info.model
-                assert isinstance(lora_model, LoRAModelRaw)
+                assert isinstance(lora_model, ModelPatchRaw)
                 yield (lora_model, lora.weight)
                 del lora_info
             return
@@ -178,7 +179,7 @@ class SDXLPromptInvocationBase:
             # apply all patches while the model is on the target device
             text_encoder_info.model_on_device() as (cached_weights, text_encoder),
             tokenizer_info as tokenizer,
-            LoRAPatcher.apply_lora_patches(
+            LayerPatcher.apply_model_patches(
                 text_encoder,
                 patches=_lora_loader(),
                 prefix=lora_prefix,
@@ -191,6 +192,7 @@ class SDXLPromptInvocationBase:
                 ti_manager,
             ),
         ):
+            context.util.signal_progress("Building conditioning")
             assert isinstance(text_encoder, (CLIPTextModel, CLIPTextModelWithProjection))
             assert isinstance(tokenizer, CLIPTokenizer)
 

invokeai/app/invocations/create_denoise_mask.py

@@ -6,7 +6,6 @@ from PIL import Image
 from torchvision.transforms.functional import resize as tv_resize
 
 from invokeai.app.invocations.baseinvocation import BaseInvocation, invocation
-from invokeai.app.invocations.constants import DEFAULT_PRECISION
 from invokeai.app.invocations.fields import FieldDescriptions, ImageField, Input, InputField
 from invokeai.app.invocations.image_to_latents import ImageToLatentsInvocation
 from invokeai.app.invocations.model import VAEField
@@ -29,11 +28,7 @@ class CreateDenoiseMaskInvocation(BaseInvocation):
     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)
-    fp32: bool = InputField(
-        default=DEFAULT_PRECISION == torch.float32,
-        description=FieldDescriptions.fp32,
-        ui_order=4,
-    )
+    fp32: bool = InputField(default=False, description=FieldDescriptions.fp32, ui_order=4)
 
     def prep_mask_tensor(self, mask_image: Image.Image) -> torch.Tensor:
         if mask_image.mode != "L":
@@ -65,6 +60,7 @@ class CreateDenoiseMaskInvocation(BaseInvocation):
             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)
             # TODO:
+            context.util.signal_progress("Running VAE encoder")
             masked_latents = ImageToLatentsInvocation.vae_encode(vae_info, self.fp32, self.tiled, masked_image.clone())
 
             masked_latents_name = context.tensors.save(tensor=masked_latents)

invokeai/app/invocations/create_gradient_mask.py

@@ -7,7 +7,6 @@ from PIL import Image, ImageFilter
 from torchvision.transforms.functional import resize as tv_resize
 
 from invokeai.app.invocations.baseinvocation import BaseInvocation, BaseInvocationOutput, invocation, invocation_output
-from invokeai.app.invocations.constants import DEFAULT_PRECISION
 from invokeai.app.invocations.fields import (
     DenoiseMaskField,
     FieldDescriptions,
@@ -76,11 +75,7 @@ class CreateGradientMaskInvocation(BaseInvocation):
         ui_order=7,
     )
     tiled: bool = InputField(default=False, description=FieldDescriptions.tiled, ui_order=8)
-    fp32: bool = InputField(
-        default=DEFAULT_PRECISION == torch.float32,
-        description=FieldDescriptions.fp32,
-        ui_order=9,
-    )
+    fp32: bool = InputField(default=False, description=FieldDescriptions.fp32, ui_order=9)
 
     @torch.no_grad()
     def invoke(self, context: InvocationContext) -> GradientMaskOutput:
@@ -131,6 +126,7 @@ class CreateGradientMaskInvocation(BaseInvocation):
                     image_tensor = image_tensor.unsqueeze(0)
                 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)
+                context.util.signal_progress("Running VAE encoder")
                 masked_latents = ImageToLatentsInvocation.vae_encode(
                     vae_info, self.fp32, self.tiled, masked_image.clone()
                 )

invokeai/app/invocations/denoise_latents.py

@@ -13,6 +13,7 @@ from diffusers.models.unets.unet_2d_condition import UNet2DConditionModel
 from diffusers.schedulers.scheduling_dpmsolver_sde import DPMSolverSDEScheduler
 from diffusers.schedulers.scheduling_tcd import TCDScheduler
 from diffusers.schedulers.scheduling_utils import SchedulerMixin as Scheduler
+from PIL import Image
 from pydantic import field_validator
 from torchvision.transforms.functional import resize as tv_resize
 from transformers import CLIPVisionModelWithProjection
@@ -36,10 +37,10 @@ from invokeai.app.invocations.t2i_adapter import T2IAdapterField
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.util.controlnet_utils import prepare_control_image
 from invokeai.backend.ip_adapter.ip_adapter import IPAdapter
-from invokeai.backend.lora.lora_model_raw import LoRAModelRaw
-from invokeai.backend.lora.lora_patcher import LoRAPatcher
 from invokeai.backend.model_manager import BaseModelType, ModelVariantType
 from invokeai.backend.model_patcher import ModelPatcher
+from invokeai.backend.patches.model_patch_raw import ModelPatchRaw
+from invokeai.backend.patches.model_patcher import LayerPatcher
 from invokeai.backend.stable_diffusion import PipelineIntermediateState
 from invokeai.backend.stable_diffusion.denoise_context import DenoiseContext, DenoiseInputs
 from invokeai.backend.stable_diffusion.diffusers_pipeline import (
@@ -510,6 +511,7 @@ class DenoiseLatentsInvocation(BaseInvocation):
         context: InvocationContext,
         t2i_adapters: Optional[Union[T2IAdapterField, list[T2IAdapterField]]],
         ext_manager: ExtensionsManager,
+        bgr_mode: bool = False,
     ) -> None:
         if t2i_adapters is None:
             return
@@ -519,6 +521,10 @@ class DenoiseLatentsInvocation(BaseInvocation):
             t2i_adapters = [t2i_adapters]
 
         for t2i_adapter_field in t2i_adapters:
+            image = context.images.get_pil(t2i_adapter_field.image.image_name)
+            if bgr_mode:  # SDXL t2i trained on cv2's BGR outputs, but PIL won't convert straight to BGR
+                r, g, b = image.split()
+                image = Image.merge("RGB", (b, g, r))
             ext_manager.add_extension(
                 T2IAdapterExt(
                     node_context=context,
@@ -616,13 +622,17 @@ class DenoiseLatentsInvocation(BaseInvocation):
         for t2i_adapter_field in t2i_adapter:
             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)
+            image = context.images.get_pil(t2i_adapter_field.image.image_name, mode="RGB")
 
             # The max_unet_downscale is the maximum amount that the UNet model downscales the latent image internally.
             if t2i_adapter_model_config.base == BaseModelType.StableDiffusion1:
                 max_unet_downscale = 8
             elif t2i_adapter_model_config.base == BaseModelType.StableDiffusionXL:
                 max_unet_downscale = 4
+
+                # SDXL adapters are trained on cv2's BGR outputs
+                r, g, b = image.split()
+                image = Image.merge("RGB", (b, g, r))
             else:
                 raise ValueError(f"Unexpected T2I-Adapter base model type: '{t2i_adapter_model_config.base}'.")
 
@@ -630,29 +640,39 @@ class DenoiseLatentsInvocation(BaseInvocation):
             with t2i_adapter_loaded_model as t2i_adapter_model:
                 total_downscale_factor = t2i_adapter_model.total_downscale_factor
 
-                # Resize the T2I-Adapter input image.
-                # We select the resize dimensions so that after the T2I-Adapter's total_downscale_factor is applied, the
-                # result will match the latent image's dimensions after max_unet_downscale is applied.
-                t2i_input_height = latents_shape[2] // max_unet_downscale * total_downscale_factor
-                t2i_input_width = latents_shape[3] // max_unet_downscale * total_downscale_factor
-
                 # Note: We have hard-coded `do_classifier_free_guidance=False`. This is because we only want to prepare
                 # a single image. If CFG is enabled, we will duplicate the resultant tensor after applying the
                 # T2I-Adapter model.
                 #
                 # Note: We re-use the `prepare_control_image(...)` from ControlNet for T2I-Adapter, because it has many
                 # of the same requirements (e.g. preserving binary masks during resize).
+
+                # Assuming fixed dimensional scaling of LATENT_SCALE_FACTOR.
+                _, _, latent_height, latent_width = latents_shape
+                control_height_resize = latent_height * LATENT_SCALE_FACTOR
+                control_width_resize = latent_width * LATENT_SCALE_FACTOR
                 t2i_image = prepare_control_image(
                     image=image,
                     do_classifier_free_guidance=False,
-                    width=t2i_input_width,
-                    height=t2i_input_height,
+                    width=control_width_resize,
+                    height=control_height_resize,
                     num_channels=t2i_adapter_model.config["in_channels"],  # mypy treats this as a FrozenDict
                     device=t2i_adapter_model.device,
                     dtype=t2i_adapter_model.dtype,
                     resize_mode=t2i_adapter_field.resize_mode,
                 )
 
+                # Resize the T2I-Adapter input image.
+                # We select the resize dimensions so that after the T2I-Adapter's total_downscale_factor is applied, the
+                # result will match the latent image's dimensions after max_unet_downscale is applied.
+                # We crop the image to this size so that the positions match the input image on non-standard resolutions
+                t2i_input_height = latents_shape[2] // max_unet_downscale * total_downscale_factor
+                t2i_input_width = latents_shape[3] // max_unet_downscale * total_downscale_factor
+                if t2i_image.shape[2] > t2i_input_height or t2i_image.shape[3] > t2i_input_width:
+                    t2i_image = t2i_image[
+                        :, :, : min(t2i_image.shape[2], t2i_input_height), : min(t2i_image.shape[3], t2i_input_width)
+                    ]
+
                 adapter_state = t2i_adapter_model(t2i_image)
 
             if do_classifier_free_guidance:
@@ -900,7 +920,8 @@ class DenoiseLatentsInvocation(BaseInvocation):
             #    ext = extension_field.to_extension(exit_stack, context, ext_manager)
             #    ext_manager.add_extension(ext)
             self.parse_controlnet_field(exit_stack, context, self.control, ext_manager)
-            self.parse_t2i_adapter_field(exit_stack, context, self.t2i_adapter, ext_manager)
+            bgr_mode = self.unet.unet.base == BaseModelType.StableDiffusionXL
+            self.parse_t2i_adapter_field(exit_stack, context, self.t2i_adapter, ext_manager, bgr_mode)
 
             # ext: t2i/ip adapter
             ext_manager.run_callback(ExtensionCallbackType.SETUP, denoise_ctx)
@@ -966,10 +987,10 @@ class DenoiseLatentsInvocation(BaseInvocation):
         def step_callback(state: PipelineIntermediateState) -> None:
             context.util.sd_step_callback(state, unet_config.base)
 
-        def _lora_loader() -> Iterator[Tuple[LoRAModelRaw, float]]:
+        def _lora_loader() -> Iterator[Tuple[ModelPatchRaw, float]]:
             for lora in self.unet.loras:
                 lora_info = context.models.load(lora.lora)
-                assert isinstance(lora_info.model, LoRAModelRaw)
+                assert isinstance(lora_info.model, ModelPatchRaw)
                 yield (lora_info.model, lora.weight)
                 del lora_info
             return
@@ -982,7 +1003,7 @@ class DenoiseLatentsInvocation(BaseInvocation):
             ModelPatcher.apply_freeu(unet, self.unet.freeu_config),
             SeamlessExt.static_patch_model(unet, self.unet.seamless_axes),  # FIXME
             # Apply the LoRA after unet has been moved to its target device for faster patching.
-            LoRAPatcher.apply_lora_patches(
+            LayerPatcher.apply_model_patches(
                 model=unet,
                 patches=_lora_loader(),
                 prefix="lora_unet_",

invokeai/app/invocations/fields.py

@@ -41,6 +41,7 @@ class UIType(str, Enum, metaclass=MetaEnum):
     # region Model Field Types
     MainModel = "MainModelField"
     FluxMainModel = "FluxMainModelField"
+    SD3MainModel = "SD3MainModelField"
     SDXLMainModel = "SDXLMainModelField"
     SDXLRefinerModel = "SDXLRefinerModelField"
     ONNXModel = "ONNXModelField"
@@ -52,7 +53,10 @@ class UIType(str, Enum, metaclass=MetaEnum):
     T2IAdapterModel = "T2IAdapterModelField"
     T5EncoderModel = "T5EncoderModelField"
     CLIPEmbedModel = "CLIPEmbedModelField"
+    CLIPLEmbedModel = "CLIPLEmbedModelField"
+    CLIPGEmbedModel = "CLIPGEmbedModelField"
     SpandrelImageToImageModel = "SpandrelImageToImageModelField"
+    ControlLoRAModel = "ControlLoRAModelField"
     # endregion
 
     # region Misc Field Types
@@ -131,15 +135,19 @@ class FieldDescriptions:
     clip = "CLIP (tokenizer, text encoder, LoRAs) and skipped layer count"
     t5_encoder = "T5 tokenizer and text encoder"
     clip_embed_model = "CLIP Embed loader"
+    clip_g_model = "CLIP-G Embed loader"
     unet = "UNet (scheduler, LoRAs)"
     transformer = "Transformer"
+    mmditx = "MMDiTX"
     vae = "VAE"
     cond = "Conditioning tensor"
     controlnet_model = "ControlNet model to load"
     vae_model = "VAE model to load"
     lora_model = "LoRA model to load"
+    control_lora_model = "Control LoRA model to load"
     main_model = "Main model (UNet, VAE, CLIP) to load"
     flux_model = "Flux model (Transformer) to load"
+    sd3_model = "SD3 model (MMDiTX) to load"
     sdxl_main_model = "SDXL Main model (UNet, VAE, CLIP1, CLIP2) to load"
     sdxl_refiner_model = "SDXL Refiner Main Modde (UNet, VAE, CLIP2) to load"
     onnx_main_model = "ONNX Main model (UNet, VAE, CLIP) to load"
@@ -244,6 +252,17 @@ class FluxConditioningField(BaseModel):
     """A conditioning tensor primitive value"""
 
     conditioning_name: str = Field(description="The name of conditioning tensor")
+    mask: Optional[TensorField] = Field(
+        default=None,
+        description="The mask associated with this conditioning tensor. Excluded regions should be set to False, "
+        "included regions should be set to True.",
+    )
+
+
+class SD3ConditioningField(BaseModel):
+    """A conditioning tensor primitive value"""
+
+    conditioning_name: str = Field(description="The name of conditioning tensor")
 
 
 class ConditioningField(BaseModel):

invokeai/app/invocations/flux_control_lora_loader.py

@@ -0,0 +1,49 @@
+from invokeai.app.invocations.baseinvocation import (
+    BaseInvocation,
+    BaseInvocationOutput,
+    Classification,
+    invocation,
+    invocation_output,
+)
+from invokeai.app.invocations.fields import FieldDescriptions, ImageField, InputField, OutputField, UIType
+from invokeai.app.invocations.model import ControlLoRAField, ModelIdentifierField
+from invokeai.app.services.shared.invocation_context import InvocationContext
+
+
+@invocation_output("flux_control_lora_loader_output")
+class FluxControlLoRALoaderOutput(BaseInvocationOutput):
+    """Flux Control LoRA Loader Output"""
+
+    control_lora: ControlLoRAField = OutputField(
+        title="Flux Control LoRA", description="Control LoRAs to apply on model loading", default=None
+    )
+
+
+@invocation(
+    "flux_control_lora_loader",
+    title="Flux Control LoRA",
+    tags=["lora", "model", "flux"],
+    category="model",
+    version="1.1.0",
+    classification=Classification.Prototype,
+)
+class FluxControlLoRALoaderInvocation(BaseInvocation):
+    """LoRA model and Image to use with FLUX transformer generation."""
+
+    lora: ModelIdentifierField = InputField(
+        description=FieldDescriptions.control_lora_model, title="Control LoRA", ui_type=UIType.ControlLoRAModel
+    )
+    image: ImageField = InputField(description="The image to encode.")
+    weight: float = InputField(description="The weight of the LoRA.", default=1.0)
+
+    def invoke(self, context: InvocationContext) -> FluxControlLoRALoaderOutput:
+        if not context.models.exists(self.lora.key):
+            raise ValueError(f"Unknown lora: {self.lora.key}!")
+
+        return FluxControlLoRALoaderOutput(
+            control_lora=ControlLoRAField(
+                lora=self.lora,
+                img=self.image,
+                weight=self.weight,
+            )
+        )

invokeai/app/invocations/flux_denoise.py

@@ -1,10 +1,12 @@
 from contextlib import ExitStack
-from typing import Callable, Iterator, Optional, Tuple
+from typing import Callable, Iterator, Optional, Tuple, Union
 
+import einops
 import numpy as np
 import numpy.typing as npt
 import torch
 import torchvision.transforms as tv_transforms
+from PIL import Image
 from torchvision.transforms.functional import resize as tv_resize
 from transformers import CLIPImageProcessor, CLIPVisionModelWithProjection
 
@@ -21,8 +23,9 @@ from invokeai.app.invocations.fields import (
     WithMetadata,
 )
 from invokeai.app.invocations.flux_controlnet import FluxControlNetField
+from invokeai.app.invocations.flux_vae_encode import FluxVaeEncodeInvocation
 from invokeai.app.invocations.ip_adapter import IPAdapterField
-from invokeai.app.invocations.model import TransformerField, VAEField
+from invokeai.app.invocations.model import ControlLoRAField, LoRAField, TransformerField, VAEField
 from invokeai.app.invocations.primitives import LatentsOutput
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.backend.flux.controlnet.instantx_controlnet_flux import InstantXControlNetFlux
@@ -30,6 +33,7 @@ from invokeai.backend.flux.controlnet.xlabs_controlnet_flux import XLabsControlN
 from invokeai.backend.flux.denoise import denoise
 from invokeai.backend.flux.extensions.inpaint_extension import InpaintExtension
 from invokeai.backend.flux.extensions.instantx_controlnet_extension import InstantXControlNetExtension
+from invokeai.backend.flux.extensions.regional_prompting_extension import RegionalPromptingExtension
 from invokeai.backend.flux.extensions.xlabs_controlnet_extension import XLabsControlNetExtension
 from invokeai.backend.flux.extensions.xlabs_ip_adapter_extension import XLabsIPAdapterExtension
 from invokeai.backend.flux.ip_adapter.xlabs_ip_adapter_flux import XlabsIpAdapterFlux
@@ -42,10 +46,11 @@ from invokeai.backend.flux.sampling_utils import (
     pack,
     unpack,
 )
-from invokeai.backend.lora.conversions.flux_lora_constants import FLUX_LORA_TRANSFORMER_PREFIX
-from invokeai.backend.lora.lora_model_raw import LoRAModelRaw
-from invokeai.backend.lora.lora_patcher import LoRAPatcher
+from invokeai.backend.flux.text_conditioning import FluxTextConditioning
 from invokeai.backend.model_manager.config import ModelFormat
+from invokeai.backend.patches.lora_conversions.flux_lora_constants import FLUX_LORA_TRANSFORMER_PREFIX
+from invokeai.backend.patches.model_patch_raw import ModelPatchRaw
+from invokeai.backend.patches.model_patcher import LayerPatcher
 from invokeai.backend.stable_diffusion.diffusers_pipeline import PipelineIntermediateState
 from invokeai.backend.stable_diffusion.diffusion.conditioning_data import FLUXConditioningInfo
 from invokeai.backend.util.devices import TorchDevice
@@ -56,7 +61,7 @@ from invokeai.backend.util.devices import TorchDevice
     title="FLUX Denoise",
     tags=["image", "flux"],
     category="image",
-    version="3.2.0",
+    version="3.2.2",
     classification=Classification.Prototype,
 )
 class FluxDenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
@@ -81,15 +86,19 @@ class FluxDenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
         description=FieldDescriptions.denoising_start,
     )
     denoising_end: float = InputField(default=1.0, ge=0, le=1, description=FieldDescriptions.denoising_end)
+    add_noise: bool = InputField(default=True, description="Add noise based on denoising start.")
     transformer: TransformerField = InputField(
         description=FieldDescriptions.flux_model,
         input=Input.Connection,
         title="Transformer",
     )
-    positive_text_conditioning: FluxConditioningField = InputField(
+    control_lora: Optional[ControlLoRAField] = InputField(
+        description=FieldDescriptions.control_lora_model, input=Input.Connection, title="Control LoRA", default=None
+    )
+    positive_text_conditioning: FluxConditioningField | list[FluxConditioningField] = InputField(
         description=FieldDescriptions.positive_cond, input=Input.Connection
     )
-    negative_text_conditioning: FluxConditioningField | None = InputField(
+    negative_text_conditioning: FluxConditioningField | list[FluxConditioningField] | None = InputField(
         default=None,
         description="Negative conditioning tensor. Can be None if cfg_scale is 1.0.",
         input=Input.Connection,
@@ -138,36 +147,12 @@ class FluxDenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
         name = context.tensors.save(tensor=latents)
         return LatentsOutput.build(latents_name=name, latents=latents, seed=None)
 
-    def _load_text_conditioning(
-        self, context: InvocationContext, conditioning_name: str, dtype: torch.dtype
-    ) -> Tuple[torch.Tensor, torch.Tensor]:
-        # Load the conditioning data.
-        cond_data = context.conditioning.load(conditioning_name)
-        assert len(cond_data.conditionings) == 1
-        flux_conditioning = cond_data.conditionings[0]
-        assert isinstance(flux_conditioning, FLUXConditioningInfo)
-        flux_conditioning = flux_conditioning.to(dtype=dtype)
-        t5_embeddings = flux_conditioning.t5_embeds
-        clip_embeddings = flux_conditioning.clip_embeds
-        return t5_embeddings, clip_embeddings
-
     def _run_diffusion(
         self,
         context: InvocationContext,
     ):
         inference_dtype = torch.bfloat16
 
-        # Load the conditioning data.
-        pos_t5_embeddings, pos_clip_embeddings = self._load_text_conditioning(
-            context, self.positive_text_conditioning.conditioning_name, inference_dtype
-        )
-        neg_t5_embeddings: torch.Tensor | None = None
-        neg_clip_embeddings: torch.Tensor | None = None
-        if self.negative_text_conditioning is not None:
-            neg_t5_embeddings, neg_clip_embeddings = self._load_text_conditioning(
-                context, self.negative_text_conditioning.conditioning_name, inference_dtype
-            )
-
         # Load the input latents, if provided.
         init_latents = context.tensors.load(self.latents.latents_name) if self.latents else None
         if init_latents is not None:
@@ -182,15 +167,45 @@ class FluxDenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
             dtype=inference_dtype,
             seed=self.seed,
         )
+        b, _c, latent_h, latent_w = noise.shape
+        packed_h = latent_h // 2
+        packed_w = latent_w // 2
+
+        # Load the conditioning data.
+        pos_text_conditionings = self._load_text_conditioning(
+            context=context,
+            cond_field=self.positive_text_conditioning,
+            packed_height=packed_h,
+            packed_width=packed_w,
+            dtype=inference_dtype,
+            device=TorchDevice.choose_torch_device(),
+        )
+        neg_text_conditionings: list[FluxTextConditioning] | None = None
+        if self.negative_text_conditioning is not None:
+            neg_text_conditionings = self._load_text_conditioning(
+                context=context,
+                cond_field=self.negative_text_conditioning,
+                packed_height=packed_h,
+                packed_width=packed_w,
+                dtype=inference_dtype,
+                device=TorchDevice.choose_torch_device(),
+            )
+        pos_regional_prompting_extension = RegionalPromptingExtension.from_text_conditioning(
+            pos_text_conditionings, img_seq_len=packed_h * packed_w
+        )
+        neg_regional_prompting_extension = (
+            RegionalPromptingExtension.from_text_conditioning(neg_text_conditionings, img_seq_len=packed_h * packed_w)
+            if neg_text_conditionings
+            else None
+        )
 
         transformer_info = context.models.load(self.transformer.transformer)
-        is_schnell = "schnell" in transformer_info.config.config_path
+        is_schnell = "schnell" in getattr(transformer_info.config, "config_path", "")
 
         # Calculate the timestep schedule.
-        image_seq_len = noise.shape[-1] * noise.shape[-2] // 4
         timesteps = get_schedule(
             num_steps=self.num_steps,
-            image_seq_len=image_seq_len,
+            image_seq_len=packed_h * packed_w,
             shift=not is_schnell,
         )
 
@@ -207,9 +222,12 @@ class FluxDenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
                     "to be poor. Consider using a FLUX dev model instead."
                 )
 
-            # Noise the orig_latents by the appropriate amount for the first timestep.
-            t_0 = timesteps[0]
-            x = t_0 * noise + (1.0 - t_0) * init_latents
+            if self.add_noise:
+                # Noise the orig_latents by the appropriate amount for the first timestep.
+                t_0 = timesteps[0]
+                x = t_0 * noise + (1.0 - t_0) * init_latents
+            else:
+                x = init_latents
         else:
             # init_latents are not provided, so we are not doing image-to-image (i.e. we are starting from pure noise).
             if self.denoising_start > 1e-5:
@@ -222,30 +240,26 @@ class FluxDenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
         if len(timesteps) <= 1:
             return x
 
+        if is_schnell and self.control_lora:
+            raise ValueError("Control LoRAs cannot be used with FLUX Schnell")
+
+        # Prepare the extra image conditioning tensor if a FLUX structural control image is provided.
+        img_cond = self._prep_structural_control_img_cond(context)
+
         inpaint_mask = self._prep_inpaint_mask(context, x)
 
-        b, _c, latent_h, latent_w = x.shape
         img_ids = generate_img_ids(h=latent_h, w=latent_w, batch_size=b, device=x.device, dtype=x.dtype)
 
-        pos_bs, pos_t5_seq_len, _ = pos_t5_embeddings.shape
-        pos_txt_ids = torch.zeros(
-            pos_bs, pos_t5_seq_len, 3, dtype=inference_dtype, device=TorchDevice.choose_torch_device()
-        )
-        neg_txt_ids: torch.Tensor | None = None
-        if neg_t5_embeddings is not None:
-            neg_bs, neg_t5_seq_len, _ = neg_t5_embeddings.shape
-            neg_txt_ids = torch.zeros(
-                neg_bs, neg_t5_seq_len, 3, dtype=inference_dtype, device=TorchDevice.choose_torch_device()
-            )
-
         # Pack all latent tensors.
         init_latents = pack(init_latents) if init_latents is not None else None
         inpaint_mask = pack(inpaint_mask) if inpaint_mask is not None else None
+        img_cond = pack(img_cond) if img_cond is not None else None
         noise = pack(noise)
         x = pack(x)
 
-        # Now that we have 'packed' the latent tensors, verify that we calculated the image_seq_len correctly.
-        assert image_seq_len == x.shape[1]
+        # Now that we have 'packed' the latent tensors, verify that we calculated the image_seq_len, packed_h, and
+        # packed_w correctly.
+        assert packed_h * packed_w == x.shape[1]
 
         # Prepare inpaint extension.
         inpaint_extension: InpaintExtension | None = None
@@ -295,7 +309,7 @@ class FluxDenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
             if config.format in [ModelFormat.Checkpoint]:
                 # The model is non-quantized, so we can apply the LoRA weights directly into the model.
                 exit_stack.enter_context(
-                    LoRAPatcher.apply_lora_patches(
+                    LayerPatcher.apply_model_patches(
                         model=transformer,
                         patches=self._lora_iterator(context),
                         prefix=FLUX_LORA_TRANSFORMER_PREFIX,
@@ -310,7 +324,7 @@ class FluxDenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
                 # The model is quantized, so apply the LoRA weights as sidecar layers. This results in slower inference,
                 # than directly patching the weights, but is agnostic to the quantization format.
                 exit_stack.enter_context(
-                    LoRAPatcher.apply_lora_sidecar_patches(
+                    LayerPatcher.apply_model_sidecar_patches(
                         model=transformer,
                         patches=self._lora_iterator(context),
                         prefix=FLUX_LORA_TRANSFORMER_PREFIX,
@@ -334,12 +348,8 @@ class FluxDenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
                 model=transformer,
                 img=x,
                 img_ids=img_ids,
-                txt=pos_t5_embeddings,
-                txt_ids=pos_txt_ids,
-                vec=pos_clip_embeddings,
-                neg_txt=neg_t5_embeddings,
-                neg_txt_ids=neg_txt_ids,
-                neg_vec=neg_clip_embeddings,
+                pos_regional_prompting_extension=pos_regional_prompting_extension,
+                neg_regional_prompting_extension=neg_regional_prompting_extension,
                 timesteps=timesteps,
                 step_callback=self._build_step_callback(context),
                 guidance=self.guidance,
@@ -348,11 +358,49 @@ class FluxDenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
                 controlnet_extensions=controlnet_extensions,
                 pos_ip_adapter_extensions=pos_ip_adapter_extensions,
                 neg_ip_adapter_extensions=neg_ip_adapter_extensions,
+                img_cond=img_cond,
             )
 
         x = unpack(x.float(), self.height, self.width)
         return x
 
+    def _load_text_conditioning(
+        self,
+        context: InvocationContext,
+        cond_field: FluxConditioningField | list[FluxConditioningField],
+        packed_height: int,
+        packed_width: int,
+        dtype: torch.dtype,
+        device: torch.device,
+    ) -> list[FluxTextConditioning]:
+        """Load text conditioning data from a FluxConditioningField or a list of FluxConditioningFields."""
+        # Normalize to a list of FluxConditioningFields.
+        cond_list = [cond_field] if isinstance(cond_field, FluxConditioningField) else cond_field
+
+        text_conditionings: list[FluxTextConditioning] = []
+        for cond_field in cond_list:
+            # Load the text embeddings.
+            cond_data = context.conditioning.load(cond_field.conditioning_name)
+            assert len(cond_data.conditionings) == 1
+            flux_conditioning = cond_data.conditionings[0]
+            assert isinstance(flux_conditioning, FLUXConditioningInfo)
+            flux_conditioning = flux_conditioning.to(dtype=dtype, device=device)
+            t5_embeddings = flux_conditioning.t5_embeds
+            clip_embeddings = flux_conditioning.clip_embeds
+
+            # Load the mask, if provided.
+            mask: Optional[torch.Tensor] = None
+            if cond_field.mask is not None:
+                mask = context.tensors.load(cond_field.mask.tensor_name)
+                mask = mask.to(device=device)
+                mask = RegionalPromptingExtension.preprocess_regional_prompt_mask(
+                    mask, packed_height, packed_width, dtype, device
+                )
+
+            text_conditionings.append(FluxTextConditioning(t5_embeddings, clip_embeddings, mask))
+
+        return text_conditionings
+
     @classmethod
     def prep_cfg_scale(
         cls, cfg_scale: float | list[float], timesteps: list[float], cfg_scale_start_step: int, cfg_scale_end_step: int
@@ -541,6 +589,29 @@ class FluxDenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
 
         return controlnet_extensions
 
+    def _prep_structural_control_img_cond(self, context: InvocationContext) -> torch.Tensor | None:
+        if self.control_lora is None:
+            return None
+
+        if not self.controlnet_vae:
+            raise ValueError("controlnet_vae must be set when using a FLUX Control LoRA.")
+
+        # Load the conditioning image and resize it to the target image size.
+        cond_img = context.images.get_pil(self.control_lora.img.image_name)
+        cond_img = cond_img.convert("RGB")
+        cond_img = cond_img.resize((self.width, self.height), Image.Resampling.BICUBIC)
+        cond_img = np.array(cond_img)
+
+        # Normalize the conditioning image to the range [-1, 1].
+        # This normalization is based on the original implementations here:
+        # https://github.com/black-forest-labs/flux/blob/805da8571a0b49b6d4043950bd266a65328c243b/src/flux/modules/image_embedders.py#L34
+        # https://github.com/black-forest-labs/flux/blob/805da8571a0b49b6d4043950bd266a65328c243b/src/flux/modules/image_embedders.py#L60
+        img_cond = torch.from_numpy(cond_img).float() / 127.5 - 1.0
+        img_cond = einops.rearrange(img_cond, "h w c -> 1 c h w")
+
+        vae_info = context.models.load(self.controlnet_vae.vae)
+        return FluxVaeEncodeInvocation.vae_encode(vae_info=vae_info, image_tensor=img_cond)
+
     def _normalize_ip_adapter_fields(self) -> list[IPAdapterField]:
         if self.ip_adapter is None:
             return []
@@ -647,10 +718,15 @@ class FluxDenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
 
         return pos_ip_adapter_extensions, neg_ip_adapter_extensions
 
-    def _lora_iterator(self, context: InvocationContext) -> Iterator[Tuple[LoRAModelRaw, float]]:
-        for lora in self.transformer.loras:
+    def _lora_iterator(self, context: InvocationContext) -> Iterator[Tuple[ModelPatchRaw, float]]:
+        loras: list[Union[LoRAField, ControlLoRAField]] = [*self.transformer.loras]
+        if self.control_lora:
+            # Note: Since FLUX structural control LoRAs modify the shape of some weights, it is important that they are
+            # applied last.
+            loras.append(self.control_lora)
+        for lora in loras:
             lora_info = context.models.load(lora.lora)
-            assert isinstance(lora_info.model, LoRAModelRaw)
+            assert isinstance(lora_info.model, ModelPatchRaw)
             yield (lora_info.model, lora.weight)
             del lora_info
 

invokeai/app/invocations/flux_model_loader.py

@@ -0,0 +1,89 @@
+from typing import Literal
+
+from invokeai.app.invocations.baseinvocation import (
+    BaseInvocation,
+    BaseInvocationOutput,
+    Classification,
+    invocation,
+    invocation_output,
+)
+from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField, UIType
+from invokeai.app.invocations.model import CLIPField, ModelIdentifierField, T5EncoderField, TransformerField, VAEField
+from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.flux.util import max_seq_lengths
+from invokeai.backend.model_manager.config import (
+    CheckpointConfigBase,
+    SubModelType,
+)
+
+
+@invocation_output("flux_model_loader_output")
+class FluxModelLoaderOutput(BaseInvocationOutput):
+    """Flux base model loader output"""
+
+    transformer: TransformerField = OutputField(description=FieldDescriptions.transformer, title="Transformer")
+    clip: CLIPField = OutputField(description=FieldDescriptions.clip, title="CLIP")
+    t5_encoder: T5EncoderField = OutputField(description=FieldDescriptions.t5_encoder, title="T5 Encoder")
+    vae: VAEField = OutputField(description=FieldDescriptions.vae, title="VAE")
+    max_seq_len: Literal[256, 512] = OutputField(
+        description="The max sequence length to used for the T5 encoder. (256 for schnell transformer, 512 for dev transformer)",
+        title="Max Seq Length",
+    )
+
+
+@invocation(
+    "flux_model_loader",
+    title="Flux Main Model",
+    tags=["model", "flux"],
+    category="model",
+    version="1.0.4",
+    classification=Classification.Prototype,
+)
+class FluxModelLoaderInvocation(BaseInvocation):
+    """Loads a flux base model, outputting its submodels."""
+
+    model: ModelIdentifierField = InputField(
+        description=FieldDescriptions.flux_model,
+        ui_type=UIType.FluxMainModel,
+        input=Input.Direct,
+    )
+
+    t5_encoder_model: ModelIdentifierField = InputField(
+        description=FieldDescriptions.t5_encoder, ui_type=UIType.T5EncoderModel, input=Input.Direct, title="T5 Encoder"
+    )
+
+    clip_embed_model: ModelIdentifierField = InputField(
+        description=FieldDescriptions.clip_embed_model,
+        ui_type=UIType.CLIPEmbedModel,
+        input=Input.Direct,
+        title="CLIP Embed",
+    )
+
+    vae_model: ModelIdentifierField = InputField(
+        description=FieldDescriptions.vae_model, ui_type=UIType.FluxVAEModel, title="VAE"
+    )
+
+    def invoke(self, context: InvocationContext) -> FluxModelLoaderOutput:
+        for key in [self.model.key, self.t5_encoder_model.key, self.clip_embed_model.key, self.vae_model.key]:
+            if not context.models.exists(key):
+                raise ValueError(f"Unknown model: {key}")
+
+        transformer = self.model.model_copy(update={"submodel_type": SubModelType.Transformer})
+        vae = self.vae_model.model_copy(update={"submodel_type": SubModelType.VAE})
+
+        tokenizer = self.clip_embed_model.model_copy(update={"submodel_type": SubModelType.Tokenizer})
+        clip_encoder = self.clip_embed_model.model_copy(update={"submodel_type": SubModelType.TextEncoder})
+
+        tokenizer2 = self.t5_encoder_model.model_copy(update={"submodel_type": SubModelType.Tokenizer2})
+        t5_encoder = self.t5_encoder_model.model_copy(update={"submodel_type": SubModelType.TextEncoder2})
+
+        transformer_config = context.models.get_config(transformer)
+        assert isinstance(transformer_config, CheckpointConfigBase)
+
+        return FluxModelLoaderOutput(
+            transformer=TransformerField(transformer=transformer, loras=[]),
+            clip=CLIPField(tokenizer=tokenizer, text_encoder=clip_encoder, loras=[], skipped_layers=0),
+            t5_encoder=T5EncoderField(tokenizer=tokenizer2, text_encoder=t5_encoder),
+            vae=VAEField(vae=vae),
+            max_seq_len=max_seq_lengths[transformer_config.config_path],
+        )

invokeai/app/invocations/flux_text_encoder.py

@@ -1,19 +1,26 @@
 from contextlib import ExitStack
-from typing import Iterator, Literal, Tuple
+from typing import Iterator, Literal, Optional, Tuple
 
 import torch
 from transformers import CLIPTextModel, CLIPTokenizer, T5EncoderModel, T5Tokenizer
 
 from invokeai.app.invocations.baseinvocation import BaseInvocation, Classification, invocation
-from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField
+from invokeai.app.invocations.fields import (
+    FieldDescriptions,
+    FluxConditioningField,
+    Input,
+    InputField,
+    TensorField,
+    UIComponent,
+)
 from invokeai.app.invocations.model import CLIPField, T5EncoderField
 from invokeai.app.invocations.primitives import FluxConditioningOutput
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.backend.flux.modules.conditioner import HFEncoder
-from invokeai.backend.lora.conversions.flux_lora_constants import FLUX_LORA_CLIP_PREFIX
-from invokeai.backend.lora.lora_model_raw import LoRAModelRaw
-from invokeai.backend.lora.lora_patcher import LoRAPatcher
 from invokeai.backend.model_manager.config import ModelFormat
+from invokeai.backend.patches.lora_conversions.flux_lora_constants import FLUX_LORA_CLIP_PREFIX
+from invokeai.backend.patches.model_patch_raw import ModelPatchRaw
+from invokeai.backend.patches.model_patcher import LayerPatcher
 from invokeai.backend.stable_diffusion.diffusion.conditioning_data import ConditioningFieldData, FLUXConditioningInfo
 
 
@@ -22,7 +29,7 @@ from invokeai.backend.stable_diffusion.diffusion.conditioning_data import Condit
     title="FLUX Text Encoding",
     tags=["prompt", "conditioning", "flux"],
     category="conditioning",
-    version="1.1.0",
+    version="1.1.1",
     classification=Classification.Prototype,
 )
 class FluxTextEncoderInvocation(BaseInvocation):
@@ -41,7 +48,10 @@ class FluxTextEncoderInvocation(BaseInvocation):
     t5_max_seq_len: Literal[256, 512] = InputField(
         description="Max sequence length for the T5 encoder. Expected to be 256 for FLUX schnell models and 512 for FLUX dev models."
     )
-    prompt: str = InputField(description="Text prompt to encode.")
+    prompt: str = InputField(description="Text prompt to encode.", ui_component=UIComponent.Textarea)
+    mask: Optional[TensorField] = InputField(
+        default=None, description="A mask defining the region that this conditioning prompt applies to."
+    )
 
     @torch.no_grad()
     def invoke(self, context: InvocationContext) -> FluxConditioningOutput:
@@ -54,7 +64,9 @@ class FluxTextEncoderInvocation(BaseInvocation):
         )
 
         conditioning_name = context.conditioning.save(conditioning_data)
-        return FluxConditioningOutput.build(conditioning_name)
+        return FluxConditioningOutput(
+            conditioning=FluxConditioningField(conditioning_name=conditioning_name, mask=self.mask)
+        )
 
     def _t5_encode(self, context: InvocationContext) -> torch.Tensor:
         t5_tokenizer_info = context.models.load(self.t5_encoder.tokenizer)
@@ -71,6 +83,7 @@ class FluxTextEncoderInvocation(BaseInvocation):
 
             t5_encoder = HFEncoder(t5_text_encoder, t5_tokenizer, False, self.t5_max_seq_len)
 
+            context.util.signal_progress("Running T5 encoder")
             prompt_embeds = t5_encoder(prompt)
 
         assert isinstance(prompt_embeds, torch.Tensor)
@@ -98,7 +111,7 @@ class FluxTextEncoderInvocation(BaseInvocation):
             if clip_text_encoder_config.format in [ModelFormat.Diffusers]:
                 # The model is non-quantized, so we can apply the LoRA weights directly into the model.
                 exit_stack.enter_context(
-                    LoRAPatcher.apply_lora_patches(
+                    LayerPatcher.apply_model_patches(
                         model=clip_text_encoder,
                         patches=self._clip_lora_iterator(context),
                         prefix=FLUX_LORA_CLIP_PREFIX,
@@ -111,14 +124,15 @@ class FluxTextEncoderInvocation(BaseInvocation):
 
             clip_encoder = HFEncoder(clip_text_encoder, clip_tokenizer, True, 77)
 
+            context.util.signal_progress("Running CLIP encoder")
             pooled_prompt_embeds = clip_encoder(prompt)
 
         assert isinstance(pooled_prompt_embeds, torch.Tensor)
         return pooled_prompt_embeds
 
-    def _clip_lora_iterator(self, context: InvocationContext) -> Iterator[Tuple[LoRAModelRaw, float]]:
+    def _clip_lora_iterator(self, context: InvocationContext) -> Iterator[Tuple[ModelPatchRaw, float]]:
         for lora in self.clip.loras:
             lora_info = context.models.load(lora.lora)
-            assert isinstance(lora_info.model, LoRAModelRaw)
+            assert isinstance(lora_info.model, ModelPatchRaw)
             yield (lora_info.model, lora.weight)
             del lora_info

invokeai/app/invocations/flux_vae_decode.py

@@ -41,7 +41,8 @@ class FluxVaeDecodeInvocation(BaseInvocation, WithMetadata, WithBoard):
     def _vae_decode(self, vae_info: LoadedModel, latents: torch.Tensor) -> Image.Image:
         with vae_info as vae:
             assert isinstance(vae, AutoEncoder)
-            latents = latents.to(device=TorchDevice.choose_torch_device(), dtype=TorchDevice.choose_torch_dtype())
+            vae_dtype = next(iter(vae.parameters())).dtype
+            latents = latents.to(device=TorchDevice.choose_torch_device(), dtype=vae_dtype)
             img = vae.decode(latents)
 
         img = img.clamp(-1, 1)
@@ -53,6 +54,7 @@ class FluxVaeDecodeInvocation(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)
+        context.util.signal_progress("Running VAE")
         image = self._vae_decode(vae_info=vae_info, latents=latents)
 
         TorchDevice.empty_cache()

invokeai/app/invocations/flux_vae_encode.py

@@ -44,9 +44,8 @@ class FluxVaeEncodeInvocation(BaseInvocation):
         generator = torch.Generator(device=TorchDevice.choose_torch_device()).manual_seed(0)
         with vae_info as vae:
             assert isinstance(vae, AutoEncoder)
-            image_tensor = image_tensor.to(
-                device=TorchDevice.choose_torch_device(), dtype=TorchDevice.choose_torch_dtype()
-            )
+            vae_dtype = next(iter(vae.parameters())).dtype
+            image_tensor = image_tensor.to(device=TorchDevice.choose_torch_device(), dtype=vae_dtype)
             latents = vae.encode(image_tensor, sample=True, generator=generator)
             return latents
 
@@ -60,6 +59,7 @@ class FluxVaeEncodeInvocation(BaseInvocation):
         if image_tensor.dim() == 3:
             image_tensor = einops.rearrange(image_tensor, "c h w -> 1 c h w")
 
+        context.util.signal_progress("Running VAE")
         latents = self.vae_encode(vae_info=vae_info, image_tensor=image_tensor)
 
         latents = latents.to("cpu")

invokeai/app/invocations/image.py

@@ -1055,3 +1055,19 @@ class CanvasV2MaskAndCropInvocation(BaseInvocation, WithMetadata, WithBoard):
             image_dto = context.images.save(image=generated_image)
 
         return ImageOutput.build(image_dto)
+
+
+@invocation(
+    "test_typegen_invocation",
+    title="Test Typegen Invocation",
+    tags=["image"],
+    category="image",
+    version="1.2.2",
+)
+class TestTypegenInvocation(BaseInvocation):
+    """Test typegen ci"""
+
+    def invoke(self, context: InvocationContext) -> ImageOutput:
+        image = Image.new("RGBA", (512, 512), (255, 0, 0, 255))
+        image_dto = context.images.save(image=image)
+        return ImageOutput.build(image_dto)

invokeai/app/invocations/image_panels.py

@@ -0,0 +1,59 @@
+from pydantic import ValidationInfo, field_validator
+
+from invokeai.app.invocations.baseinvocation import (
+    BaseInvocation,
+    BaseInvocationOutput,
+    Classification,
+    invocation,
+    invocation_output,
+)
+from invokeai.app.invocations.fields import InputField, OutputField
+from invokeai.app.services.shared.invocation_context import InvocationContext
+
+
+@invocation_output("image_panel_coordinate_output")
+class ImagePanelCoordinateOutput(BaseInvocationOutput):
+    x_left: int = OutputField(description="The left x-coordinate of the panel.")
+    y_top: int = OutputField(description="The top y-coordinate of the panel.")
+    width: int = OutputField(description="The width of the panel.")
+    height: int = OutputField(description="The height of the panel.")
+
+
+@invocation(
+    "image_panel_layout",
+    title="Image Panel Layout",
+    tags=["image", "panel", "layout"],
+    category="image",
+    version="1.0.0",
+    classification=Classification.Prototype,
+)
+class ImagePanelLayoutInvocation(BaseInvocation):
+    """Get the coordinates of a single panel in a grid. (If the full image shape cannot be divided evenly into panels,
+    then the grid may not cover the entire image.)
+    """
+
+    width: int = InputField(description="The width of the entire grid.")
+    height: int = InputField(description="The height of the entire grid.")
+    num_cols: int = InputField(ge=1, default=1, description="The number of columns in the grid.")
+    num_rows: int = InputField(ge=1, default=1, description="The number of rows in the grid.")
+    panel_col_idx: int = InputField(ge=0, default=0, description="The column index of the panel to be processed.")
+    panel_row_idx: int = InputField(ge=0, default=0, description="The row index of the panel to be processed.")
+
+    @field_validator("panel_col_idx")
+    def validate_panel_col_idx(cls, v: int, info: ValidationInfo) -> int:
+        if v < 0 or v >= info.data["num_cols"]:
+            raise ValueError(f"panel_col_idx must be between 0 and {info.data['num_cols'] - 1}")
+        return v
+
+    @field_validator("panel_row_idx")
+    def validate_panel_row_idx(cls, v: int, info: ValidationInfo) -> int:
+        if v < 0 or v >= info.data["num_rows"]:
+            raise ValueError(f"panel_row_idx must be between 0 and {info.data['num_rows'] - 1}")
+        return v
+
+    def invoke(self, context: InvocationContext) -> ImagePanelCoordinateOutput:
+        x_left = self.panel_col_idx * (self.width // self.num_cols)
+        y_top = self.panel_row_idx * (self.height // self.num_rows)
+        width = self.width // self.num_cols
+        height = self.height // self.num_rows
+        return ImagePanelCoordinateOutput(x_left=x_left, y_top=y_top, width=width, height=height)

invokeai/app/invocations/image_to_latents.py

@@ -13,7 +13,7 @@ from diffusers.models.autoencoders.autoencoder_kl import AutoencoderKL
 from diffusers.models.autoencoders.autoencoder_tiny import AutoencoderTiny
 
 from invokeai.app.invocations.baseinvocation import BaseInvocation, invocation
-from invokeai.app.invocations.constants import DEFAULT_PRECISION, LATENT_SCALE_FACTOR
+from invokeai.app.invocations.constants import LATENT_SCALE_FACTOR
 from invokeai.app.invocations.fields import (
     FieldDescriptions,
     ImageField,
@@ -49,7 +49,7 @@ class ImageToLatentsInvocation(BaseInvocation):
     # NOTE: tile_size = 0 is a special value. We use this rather than `int | None`, because the workflow UI does not
     # offer a way to directly set None values.
     tile_size: int = InputField(default=0, multiple_of=8, description=FieldDescriptions.vae_tile_size)
-    fp32: bool = InputField(default=DEFAULT_PRECISION == torch.float32, description=FieldDescriptions.fp32)
+    fp32: bool = InputField(default=False, description=FieldDescriptions.fp32)
 
     @staticmethod
     def vae_encode(
@@ -117,6 +117,7 @@ class ImageToLatentsInvocation(BaseInvocation):
         if image_tensor.dim() == 3:
             image_tensor = einops.rearrange(image_tensor, "c h w -> 1 c h w")
 
+        context.util.signal_progress("Running VAE encoder")
         latents = self.vae_encode(
             vae_info=vae_info, upcast=self.fp32, tiled=self.tiled, image_tensor=image_tensor, tile_size=self.tile_size
         )

invokeai/app/invocations/latents_to_image.py

@@ -12,7 +12,7 @@ from diffusers.models.autoencoders.autoencoder_kl import AutoencoderKL
 from diffusers.models.autoencoders.autoencoder_tiny import AutoencoderTiny
 
 from invokeai.app.invocations.baseinvocation import BaseInvocation, invocation
-from invokeai.app.invocations.constants import DEFAULT_PRECISION, LATENT_SCALE_FACTOR
+from invokeai.app.invocations.constants import LATENT_SCALE_FACTOR
 from invokeai.app.invocations.fields import (
     FieldDescriptions,
     Input,
@@ -51,7 +51,7 @@ class LatentsToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
     # NOTE: tile_size = 0 is a special value. We use this rather than `int | None`, because the workflow UI does not
     # offer a way to directly set None values.
     tile_size: int = InputField(default=0, multiple_of=8, description=FieldDescriptions.vae_tile_size)
-    fp32: bool = InputField(default=DEFAULT_PRECISION == torch.float32, description=FieldDescriptions.fp32)
+    fp32: bool = InputField(default=False, description=FieldDescriptions.fp32)
 
     @torch.no_grad()
     def invoke(self, context: InvocationContext) -> ImageOutput:
@@ -60,6 +60,7 @@ class LatentsToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
         vae_info = context.models.load(self.vae.vae)
         assert isinstance(vae_info.model, (AutoencoderKL, AutoencoderTiny))
         with SeamlessExt.static_patch_model(vae_info.model, self.vae.seamless_axes), vae_info as vae:
+            context.util.signal_progress("Running VAE decoder")
             assert isinstance(vae, (AutoencoderKL, AutoencoderTiny))
             latents = latents.to(vae.device)
             if self.fp32:

invokeai/app/invocations/metadata.py

@@ -147,6 +147,10 @@ GENERATION_MODES = Literal[
     "flux_img2img",
     "flux_inpaint",
     "flux_outpaint",
+    "sd3_txt2img",
+    "sd3_img2img",
+    "sd3_inpaint",
+    "sd3_outpaint",
 ]
 
 

invokeai/app/invocations/model.py

@@ -1,5 +1,5 @@
 import copy
-from typing import List, Literal, Optional
+from typing import List, Optional
 
 from pydantic import BaseModel, Field
 
@@ -10,14 +10,12 @@ from invokeai.app.invocations.baseinvocation import (
     invocation,
     invocation_output,
 )
-from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField, UIType
+from invokeai.app.invocations.fields import FieldDescriptions, ImageField, Input, InputField, OutputField, UIType
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.shared.models import FreeUConfig
-from invokeai.backend.flux.util import max_seq_lengths
 from invokeai.backend.model_manager.config import (
     AnyModelConfig,
     BaseModelType,
-    CheckpointConfigBase,
     ModelType,
     SubModelType,
 )
@@ -67,11 +65,6 @@ class CLIPField(BaseModel):
     loras: List[LoRAField] = Field(description="LoRAs to apply on model loading")
 
 
-class TransformerField(BaseModel):
-    transformer: ModelIdentifierField = Field(description="Info to load Transformer submodel")
-    loras: List[LoRAField] = Field(description="LoRAs to apply on model loading")
-
-
 class T5EncoderField(BaseModel):
     tokenizer: ModelIdentifierField = Field(description="Info to load tokenizer submodel")
     text_encoder: ModelIdentifierField = Field(description="Info to load text_encoder submodel")
@@ -82,6 +75,15 @@ class VAEField(BaseModel):
     seamless_axes: List[str] = Field(default_factory=list, description='Axes("x" and "y") to which apply seamless')
 
 
+class ControlLoRAField(LoRAField):
+    img: ImageField = Field(description="Image to use in structural conditioning")
+
+
+class TransformerField(BaseModel):
+    transformer: ModelIdentifierField = Field(description="Info to load Transformer submodel")
+    loras: List[LoRAField] = Field(description="LoRAs to apply on model loading")
+
+
 @invocation_output("unet_output")
 class UNetOutput(BaseInvocationOutput):
     """Base class for invocations that output a UNet field."""
@@ -139,78 +141,6 @@ class ModelIdentifierInvocation(BaseInvocation):
         return ModelIdentifierOutput(model=self.model)
 
 
-@invocation_output("flux_model_loader_output")
-class FluxModelLoaderOutput(BaseInvocationOutput):
-    """Flux base model loader output"""
-
-    transformer: TransformerField = OutputField(description=FieldDescriptions.transformer, title="Transformer")
-    clip: CLIPField = OutputField(description=FieldDescriptions.clip, title="CLIP")
-    t5_encoder: T5EncoderField = OutputField(description=FieldDescriptions.t5_encoder, title="T5 Encoder")
-    vae: VAEField = OutputField(description=FieldDescriptions.vae, title="VAE")
-    max_seq_len: Literal[256, 512] = OutputField(
-        description="The max sequence length to used for the T5 encoder. (256 for schnell transformer, 512 for dev transformer)",
-        title="Max Seq Length",
-    )
-
-
-@invocation(
-    "flux_model_loader",
-    title="Flux Main Model",
-    tags=["model", "flux"],
-    category="model",
-    version="1.0.4",
-    classification=Classification.Prototype,
-)
-class FluxModelLoaderInvocation(BaseInvocation):
-    """Loads a flux base model, outputting its submodels."""
-
-    model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.flux_model,
-        ui_type=UIType.FluxMainModel,
-        input=Input.Direct,
-    )
-
-    t5_encoder_model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.t5_encoder, ui_type=UIType.T5EncoderModel, input=Input.Direct, title="T5 Encoder"
-    )
-
-    clip_embed_model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.clip_embed_model,
-        ui_type=UIType.CLIPEmbedModel,
-        input=Input.Direct,
-        title="CLIP Embed",
-    )
-
-    vae_model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.vae_model, ui_type=UIType.FluxVAEModel, title="VAE"
-    )
-
-    def invoke(self, context: InvocationContext) -> FluxModelLoaderOutput:
-        for key in [self.model.key, self.t5_encoder_model.key, self.clip_embed_model.key, self.vae_model.key]:
-            if not context.models.exists(key):
-                raise ValueError(f"Unknown model: {key}")
-
-        transformer = self.model.model_copy(update={"submodel_type": SubModelType.Transformer})
-        vae = self.vae_model.model_copy(update={"submodel_type": SubModelType.VAE})
-
-        tokenizer = self.clip_embed_model.model_copy(update={"submodel_type": SubModelType.Tokenizer})
-        clip_encoder = self.clip_embed_model.model_copy(update={"submodel_type": SubModelType.TextEncoder})
-
-        tokenizer2 = self.t5_encoder_model.model_copy(update={"submodel_type": SubModelType.Tokenizer2})
-        t5_encoder = self.t5_encoder_model.model_copy(update={"submodel_type": SubModelType.TextEncoder2})
-
-        transformer_config = context.models.get_config(transformer)
-        assert isinstance(transformer_config, CheckpointConfigBase)
-
-        return FluxModelLoaderOutput(
-            transformer=TransformerField(transformer=transformer, loras=[]),
-            clip=CLIPField(tokenizer=tokenizer, text_encoder=clip_encoder, loras=[], skipped_layers=0),
-            t5_encoder=T5EncoderField(tokenizer=tokenizer2, text_encoder=t5_encoder),
-            vae=VAEField(vae=vae),
-            max_seq_len=max_seq_lengths[transformer_config.config_path],
-        )
-
-
 @invocation(
     "main_model_loader",
     title="Main Model",

invokeai/app/invocations/param_easing.py

@@ -1,43 +1,4 @@
-import io
-from typing import Literal, Optional
-
-import matplotlib.pyplot as plt
 import numpy as np
-import PIL.Image
-from easing_functions import (
-    BackEaseIn,
-    BackEaseInOut,
-    BackEaseOut,
-    BounceEaseIn,
-    BounceEaseInOut,
-    BounceEaseOut,
-    CircularEaseIn,
-    CircularEaseInOut,
-    CircularEaseOut,
-    CubicEaseIn,
-    CubicEaseInOut,
-    CubicEaseOut,
-    ElasticEaseIn,
-    ElasticEaseInOut,
-    ElasticEaseOut,
-    ExponentialEaseIn,
-    ExponentialEaseInOut,
-    ExponentialEaseOut,
-    LinearInOut,
-    QuadEaseIn,
-    QuadEaseInOut,
-    QuadEaseOut,
-    QuarticEaseIn,
-    QuarticEaseInOut,
-    QuarticEaseOut,
-    QuinticEaseIn,
-    QuinticEaseInOut,
-    QuinticEaseOut,
-    SineEaseIn,
-    SineEaseInOut,
-    SineEaseOut,
-)
-from matplotlib.ticker import MaxNLocator
 
 from invokeai.app.invocations.baseinvocation import BaseInvocation, invocation
 from invokeai.app.invocations.fields import InputField
@@ -65,191 +26,3 @@ class FloatLinearRangeInvocation(BaseInvocation):
     def invoke(self, context: InvocationContext) -> FloatCollectionOutput:
         param_list = list(np.linspace(self.start, self.stop, self.steps))
         return FloatCollectionOutput(collection=param_list)
-
-
-EASING_FUNCTIONS_MAP = {
-    "Linear": LinearInOut,
-    "QuadIn": QuadEaseIn,
-    "QuadOut": QuadEaseOut,
-    "QuadInOut": QuadEaseInOut,
-    "CubicIn": CubicEaseIn,
-    "CubicOut": CubicEaseOut,
-    "CubicInOut": CubicEaseInOut,
-    "QuarticIn": QuarticEaseIn,
-    "QuarticOut": QuarticEaseOut,
-    "QuarticInOut": QuarticEaseInOut,
-    "QuinticIn": QuinticEaseIn,
-    "QuinticOut": QuinticEaseOut,
-    "QuinticInOut": QuinticEaseInOut,
-    "SineIn": SineEaseIn,
-    "SineOut": SineEaseOut,
-    "SineInOut": SineEaseInOut,
-    "CircularIn": CircularEaseIn,
-    "CircularOut": CircularEaseOut,
-    "CircularInOut": CircularEaseInOut,
-    "ExponentialIn": ExponentialEaseIn,
-    "ExponentialOut": ExponentialEaseOut,
-    "ExponentialInOut": ExponentialEaseInOut,
-    "ElasticIn": ElasticEaseIn,
-    "ElasticOut": ElasticEaseOut,
-    "ElasticInOut": ElasticEaseInOut,
-    "BackIn": BackEaseIn,
-    "BackOut": BackEaseOut,
-    "BackInOut": BackEaseInOut,
-    "BounceIn": BounceEaseIn,
-    "BounceOut": BounceEaseOut,
-    "BounceInOut": BounceEaseInOut,
-}
-
-EASING_FUNCTION_KEYS = Literal[tuple(EASING_FUNCTIONS_MAP.keys())]
-
-
-# actually I think for now could just use CollectionOutput (which is list[Any]
-@invocation(
-    "step_param_easing",
-    title="Step Param Easing",
-    tags=["step", "easing"],
-    category="step",
-    version="1.0.2",
-)
-class StepParamEasingInvocation(BaseInvocation):
-    """Experimental per-step parameter easing for denoising steps"""
-
-    easing: EASING_FUNCTION_KEYS = InputField(default="Linear", description="The easing function to use")
-    num_steps: int = InputField(default=20, description="number of denoising steps")
-    start_value: float = InputField(default=0.0, description="easing starting value")
-    end_value: float = InputField(default=1.0, description="easing ending value")
-    start_step_percent: float = InputField(default=0.0, description="fraction of steps at which to start easing")
-    end_step_percent: float = InputField(default=1.0, description="fraction of steps after which to end easing")
-    # if None, then start_value is used prior to easing start
-    pre_start_value: Optional[float] = InputField(default=None, description="value before easing start")
-    # if None, then end value is used prior to easing end
-    post_end_value: Optional[float] = InputField(default=None, description="value after easing end")
-    mirror: bool = InputField(default=False, description="include mirror of easing function")
-    # FIXME: add alt_mirror option (alternative to default or mirror), or remove entirely
-    # alt_mirror: bool = InputField(default=False, description="alternative mirroring by dual easing")
-    show_easing_plot: bool = InputField(default=False, description="show easing plot")
-
-    def invoke(self, context: InvocationContext) -> FloatCollectionOutput:
-        log_diagnostics = False
-        # convert from start_step_percent to nearest step <= (steps * start_step_percent)
-        # start_step = int(np.floor(self.num_steps * self.start_step_percent))
-        start_step = int(np.round(self.num_steps * self.start_step_percent))
-        # convert from end_step_percent to nearest step >= (steps * end_step_percent)
-        # end_step = int(np.ceil((self.num_steps - 1) * self.end_step_percent))
-        end_step = int(np.round((self.num_steps - 1) * self.end_step_percent))
-
-        # end_step = int(np.ceil(self.num_steps * self.end_step_percent))
-        num_easing_steps = end_step - start_step + 1
-
-        # num_presteps = max(start_step - 1, 0)
-        num_presteps = start_step
-        num_poststeps = self.num_steps - (num_presteps + num_easing_steps)
-        prelist = list(num_presteps * [self.pre_start_value])
-        postlist = list(num_poststeps * [self.post_end_value])
-
-        if log_diagnostics:
-            context.logger.debug("start_step: " + str(start_step))
-            context.logger.debug("end_step: " + str(end_step))
-            context.logger.debug("num_easing_steps: " + str(num_easing_steps))
-            context.logger.debug("num_presteps: " + str(num_presteps))
-            context.logger.debug("num_poststeps: " + str(num_poststeps))
-            context.logger.debug("prelist size: " + str(len(prelist)))
-            context.logger.debug("postlist size: " + str(len(postlist)))
-            context.logger.debug("prelist: " + str(prelist))
-            context.logger.debug("postlist: " + str(postlist))
-
-        easing_class = EASING_FUNCTIONS_MAP[self.easing]
-        if log_diagnostics:
-            context.logger.debug("easing class: " + str(easing_class))
-        easing_list = []
-        if self.mirror:  # "expected" mirroring
-            # if number of steps is even, squeeze duration down to (number_of_steps)/2
-            # and create reverse copy of list to append
-            # if number of steps is odd, squeeze duration down to ceil(number_of_steps/2)
-            # and create reverse copy of list[1:end-1]
-            # but if even then number_of_steps/2 === ceil(number_of_steps/2), so can just use ceil always
-
-            base_easing_duration = int(np.ceil(num_easing_steps / 2.0))
-            if log_diagnostics:
-                context.logger.debug("base easing duration: " + str(base_easing_duration))
-            even_num_steps = num_easing_steps % 2 == 0  # even number of steps
-            easing_function = easing_class(
-                start=self.start_value,
-                end=self.end_value,
-                duration=base_easing_duration - 1,
-            )
-            base_easing_vals = []
-            for step_index in range(base_easing_duration):
-                easing_val = easing_function.ease(step_index)
-                base_easing_vals.append(easing_val)
-                if log_diagnostics:
-                    context.logger.debug("step_index: " + str(step_index) + ", easing_val: " + str(easing_val))
-            if even_num_steps:
-                mirror_easing_vals = list(reversed(base_easing_vals))
-            else:
-                mirror_easing_vals = list(reversed(base_easing_vals[0:-1]))
-            if log_diagnostics:
-                context.logger.debug("base easing vals: " + str(base_easing_vals))
-                context.logger.debug("mirror easing vals: " + str(mirror_easing_vals))
-            easing_list = base_easing_vals + mirror_easing_vals
-
-        # FIXME: add alt_mirror option (alternative to default or mirror), or remove entirely
-        # elif self.alt_mirror:  # function mirroring (unintuitive behavior (at least to me))
-        #     # half_ease_duration = round(num_easing_steps - 1 / 2)
-        #     half_ease_duration = round((num_easing_steps - 1) / 2)
-        #     easing_function = easing_class(start=self.start_value,
-        #                                    end=self.end_value,
-        #                                    duration=half_ease_duration,
-        #                                    )
-        #
-        #     mirror_function = easing_class(start=self.end_value,
-        #                                    end=self.start_value,
-        #                                    duration=half_ease_duration,
-        #                                    )
-        #     for step_index in range(num_easing_steps):
-        #         if step_index <= half_ease_duration:
-        #             step_val = easing_function.ease(step_index)
-        #         else:
-        #             step_val = mirror_function.ease(step_index - half_ease_duration)
-        #         easing_list.append(step_val)
-        #         if log_diagnostics: logger.debug(step_index, step_val)
-        #
-
-        else:  # no mirroring (default)
-            easing_function = easing_class(
-                start=self.start_value,
-                end=self.end_value,
-                duration=num_easing_steps - 1,
-            )
-            for step_index in range(num_easing_steps):
-                step_val = easing_function.ease(step_index)
-                easing_list.append(step_val)
-                if log_diagnostics:
-                    context.logger.debug("step_index: " + str(step_index) + ", easing_val: " + str(step_val))
-
-        if log_diagnostics:
-            context.logger.debug("prelist size: " + str(len(prelist)))
-            context.logger.debug("easing_list size: " + str(len(easing_list)))
-            context.logger.debug("postlist size: " + str(len(postlist)))
-
-        param_list = prelist + easing_list + postlist
-
-        if self.show_easing_plot:
-            plt.figure()
-            plt.xlabel("Step")
-            plt.ylabel("Param Value")
-            plt.title("Per-Step Values Based On Easing: " + self.easing)
-            plt.bar(range(len(param_list)), param_list)
-            # plt.plot(param_list)
-            ax = plt.gca()
-            ax.xaxis.set_major_locator(MaxNLocator(integer=True))
-            buf = io.BytesIO()
-            plt.savefig(buf, format="png")
-            buf.seek(0)
-            im = PIL.Image.open(buf)
-            im.show()
-            buf.close()
-
-        # output array of size steps, each entry list[i] is param value for step i
-        return FloatCollectionOutput(collection=param_list)

invokeai/app/invocations/primitives.py

@@ -4,7 +4,13 @@ from typing import Optional
 
 import torch
 
-from invokeai.app.invocations.baseinvocation import BaseInvocation, BaseInvocationOutput, invocation, invocation_output
+from invokeai.app.invocations.baseinvocation import (
+    BaseInvocation,
+    BaseInvocationOutput,
+    Classification,
+    invocation,
+    invocation_output,
+)
 from invokeai.app.invocations.constants import LATENT_SCALE_FACTOR
 from invokeai.app.invocations.fields import (
     BoundingBoxField,
@@ -18,6 +24,7 @@ from invokeai.app.invocations.fields import (
     InputField,
     LatentsField,
     OutputField,
+    SD3ConditioningField,
     TensorField,
     UIComponent,
 )
@@ -426,6 +433,17 @@ class FluxConditioningOutput(BaseInvocationOutput):
         return cls(conditioning=FluxConditioningField(conditioning_name=conditioning_name))
 
 
+@invocation_output("sd3_conditioning_output")
+class SD3ConditioningOutput(BaseInvocationOutput):
+    """Base class for nodes that output a single SD3 conditioning tensor"""
+
+    conditioning: SD3ConditioningField = OutputField(description=FieldDescriptions.cond)
+
+    @classmethod
+    def build(cls, conditioning_name: str) -> "SD3ConditioningOutput":
+        return cls(conditioning=SD3ConditioningField(conditioning_name=conditioning_name))
+
+
 @invocation_output("conditioning_output")
 class ConditioningOutput(BaseInvocationOutput):
     """Base class for nodes that output a single conditioning tensor"""
@@ -521,3 +539,23 @@ class BoundingBoxInvocation(BaseInvocation):
 
 
 # endregion
+
+
+@invocation(
+    "image_batch",
+    title="Image Batch",
+    tags=["primitives", "image", "batch", "internal"],
+    category="primitives",
+    version="1.0.0",
+    classification=Classification.Special,
+)
+class ImageBatchInvocation(BaseInvocation):
+    """Create a batched generation, where the workflow is executed once for each image in the batch."""
+
+    images: list[ImageField] = InputField(min_length=1, description="The images to batch over", input=Input.Direct)
+
+    def __init__(self):
+        raise NotImplementedError("This class should never be executed or instantiated directly.")
+
+    def invoke(self, context: InvocationContext) -> ImageOutput:
+        raise NotImplementedError("This class should never be executed or instantiated directly.")

invokeai/app/invocations/sd3_denoise.py

@@ -0,0 +1,338 @@
+from typing import Callable, Optional, Tuple
+
+import torch
+import torchvision.transforms as tv_transforms
+from diffusers.models.transformers.transformer_sd3 import SD3Transformer2DModel
+from torchvision.transforms.functional import resize as tv_resize
+from tqdm import tqdm
+
+from invokeai.app.invocations.baseinvocation import BaseInvocation, Classification, invocation
+from invokeai.app.invocations.constants import LATENT_SCALE_FACTOR
+from invokeai.app.invocations.fields import (
+    DenoiseMaskField,
+    FieldDescriptions,
+    Input,
+    InputField,
+    LatentsField,
+    SD3ConditioningField,
+    WithBoard,
+    WithMetadata,
+)
+from invokeai.app.invocations.model import TransformerField
+from invokeai.app.invocations.primitives import LatentsOutput
+from invokeai.app.invocations.sd3_text_encoder import SD3_T5_MAX_SEQ_LEN
+from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.flux.sampling_utils import clip_timestep_schedule_fractional
+from invokeai.backend.model_manager.config import BaseModelType
+from invokeai.backend.sd3.extensions.inpaint_extension import InpaintExtension
+from invokeai.backend.stable_diffusion.diffusers_pipeline import PipelineIntermediateState
+from invokeai.backend.stable_diffusion.diffusion.conditioning_data import SD3ConditioningInfo
+from invokeai.backend.util.devices import TorchDevice
+
+
+@invocation(
+    "sd3_denoise",
+    title="SD3 Denoise",
+    tags=["image", "sd3"],
+    category="image",
+    version="1.1.0",
+    classification=Classification.Prototype,
+)
+class SD3DenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
+    """Run denoising process with a SD3 model."""
+
+    # If latents is provided, this means we are doing image-to-image.
+    latents: Optional[LatentsField] = InputField(
+        default=None, description=FieldDescriptions.latents, input=Input.Connection
+    )
+    # denoise_mask is used for image-to-image inpainting. Only the masked region is modified.
+    denoise_mask: Optional[DenoiseMaskField] = InputField(
+        default=None, description=FieldDescriptions.denoise_mask, input=Input.Connection
+    )
+    denoising_start: float = InputField(default=0.0, ge=0, le=1, description=FieldDescriptions.denoising_start)
+    denoising_end: float = InputField(default=1.0, ge=0, le=1, description=FieldDescriptions.denoising_end)
+    transformer: TransformerField = InputField(
+        description=FieldDescriptions.sd3_model, input=Input.Connection, title="Transformer"
+    )
+    positive_conditioning: SD3ConditioningField = InputField(
+        description=FieldDescriptions.positive_cond, input=Input.Connection
+    )
+    negative_conditioning: SD3ConditioningField = InputField(
+        description=FieldDescriptions.negative_cond, input=Input.Connection
+    )
+    cfg_scale: float | list[float] = InputField(default=3.5, description=FieldDescriptions.cfg_scale, title="CFG Scale")
+    width: int = InputField(default=1024, multiple_of=16, description="Width of the generated image.")
+    height: int = InputField(default=1024, multiple_of=16, description="Height of the generated image.")
+    steps: int = InputField(default=10, gt=0, description=FieldDescriptions.steps)
+    seed: int = InputField(default=0, description="Randomness seed for reproducibility.")
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> LatentsOutput:
+        latents = self._run_diffusion(context)
+        latents = latents.detach().to("cpu")
+
+        name = context.tensors.save(tensor=latents)
+        return LatentsOutput.build(latents_name=name, latents=latents, seed=None)
+
+    def _prep_inpaint_mask(self, context: InvocationContext, latents: torch.Tensor) -> torch.Tensor | None:
+        """Prepare the inpaint mask.
+        - Loads the mask
+        - Resizes if necessary
+        - Casts to same device/dtype as latents
+
+        Args:
+            context (InvocationContext): The invocation context, for loading the inpaint mask.
+            latents (torch.Tensor): A latent image tensor. Used to determine the target shape, device, and dtype for the
+                inpaint mask.
+
+        Returns:
+            torch.Tensor | None: Inpaint mask. Values of 0.0 represent the regions to be fully denoised, and 1.0
+                represent the regions to be preserved.
+        """
+        if self.denoise_mask is None:
+            return None
+        mask = context.tensors.load(self.denoise_mask.mask_name)
+
+        # The input denoise_mask contains values in [0, 1], where 0.0 represents the regions to be fully denoised, and
+        # 1.0 represents the regions to be preserved.
+        # We invert the mask so that the regions to be preserved are 0.0 and the regions to be denoised are 1.0.
+        mask = 1.0 - mask
+
+        _, _, latent_height, latent_width = latents.shape
+        mask = tv_resize(
+            img=mask,
+            size=[latent_height, latent_width],
+            interpolation=tv_transforms.InterpolationMode.BILINEAR,
+            antialias=False,
+        )
+
+        mask = mask.to(device=latents.device, dtype=latents.dtype)
+        return mask
+
+    def _load_text_conditioning(
+        self,
+        context: InvocationContext,
+        conditioning_name: str,
+        joint_attention_dim: int,
+        dtype: torch.dtype,
+        device: torch.device,
+    ) -> Tuple[torch.Tensor, torch.Tensor]:
+        # Load the conditioning data.
+        cond_data = context.conditioning.load(conditioning_name)
+        assert len(cond_data.conditionings) == 1
+        sd3_conditioning = cond_data.conditionings[0]
+        assert isinstance(sd3_conditioning, SD3ConditioningInfo)
+        sd3_conditioning = sd3_conditioning.to(dtype=dtype, device=device)
+
+        t5_embeds = sd3_conditioning.t5_embeds
+        if t5_embeds is None:
+            t5_embeds = torch.zeros(
+                (1, SD3_T5_MAX_SEQ_LEN, joint_attention_dim),
+                device=device,
+                dtype=dtype,
+            )
+
+        clip_prompt_embeds = torch.cat([sd3_conditioning.clip_l_embeds, sd3_conditioning.clip_g_embeds], dim=-1)
+        clip_prompt_embeds = torch.nn.functional.pad(
+            clip_prompt_embeds, (0, t5_embeds.shape[-1] - clip_prompt_embeds.shape[-1])
+        )
+
+        prompt_embeds = torch.cat([clip_prompt_embeds, t5_embeds], dim=-2)
+        pooled_prompt_embeds = torch.cat(
+            [sd3_conditioning.clip_l_pooled_embeds, sd3_conditioning.clip_g_pooled_embeds], dim=-1
+        )
+
+        return prompt_embeds, pooled_prompt_embeds
+
+    def _get_noise(
+        self,
+        num_samples: int,
+        num_channels_latents: int,
+        height: int,
+        width: int,
+        dtype: torch.dtype,
+        device: torch.device,
+        seed: int,
+    ) -> torch.Tensor:
+        # We always generate noise on the same device and dtype then cast to ensure consistency across devices/dtypes.
+        rand_device = "cpu"
+        rand_dtype = torch.float16
+
+        return torch.randn(
+            num_samples,
+            num_channels_latents,
+            int(height) // LATENT_SCALE_FACTOR,
+            int(width) // LATENT_SCALE_FACTOR,
+            device=rand_device,
+            dtype=rand_dtype,
+            generator=torch.Generator(device=rand_device).manual_seed(seed),
+        ).to(device=device, dtype=dtype)
+
+    def _prepare_cfg_scale(self, num_timesteps: int) -> list[float]:
+        """Prepare the CFG scale list.
+
+        Args:
+            num_timesteps (int): The number of timesteps in the scheduler. Could be different from num_steps depending
+            on the scheduler used (e.g. higher order schedulers).
+
+        Returns:
+            list[float]: _description_
+        """
+        if isinstance(self.cfg_scale, float):
+            cfg_scale = [self.cfg_scale] * num_timesteps
+        elif isinstance(self.cfg_scale, list):
+            assert len(self.cfg_scale) == num_timesteps
+            cfg_scale = self.cfg_scale
+        else:
+            raise ValueError(f"Invalid CFG scale type: {type(self.cfg_scale)}")
+
+        return cfg_scale
+
+    def _run_diffusion(
+        self,
+        context: InvocationContext,
+    ):
+        inference_dtype = TorchDevice.choose_torch_dtype()
+        device = TorchDevice.choose_torch_device()
+
+        transformer_info = context.models.load(self.transformer.transformer)
+
+        # Load/process the conditioning data.
+        # TODO(ryand): Make CFG optional.
+        do_classifier_free_guidance = True
+        pos_prompt_embeds, pos_pooled_prompt_embeds = self._load_text_conditioning(
+            context=context,
+            conditioning_name=self.positive_conditioning.conditioning_name,
+            joint_attention_dim=transformer_info.model.config.joint_attention_dim,
+            dtype=inference_dtype,
+            device=device,
+        )
+        neg_prompt_embeds, neg_pooled_prompt_embeds = self._load_text_conditioning(
+            context=context,
+            conditioning_name=self.negative_conditioning.conditioning_name,
+            joint_attention_dim=transformer_info.model.config.joint_attention_dim,
+            dtype=inference_dtype,
+            device=device,
+        )
+        # TODO(ryand): Support both sequential and batched CFG inference.
+        prompt_embeds = torch.cat([neg_prompt_embeds, pos_prompt_embeds], dim=0)
+        pooled_prompt_embeds = torch.cat([neg_pooled_prompt_embeds, pos_pooled_prompt_embeds], dim=0)
+
+        # Prepare the timestep schedule.
+        # We add an extra step to the end to account for the final timestep of 0.0.
+        timesteps: list[float] = torch.linspace(1, 0, self.steps + 1).tolist()
+        # Clip the timesteps schedule based on denoising_start and denoising_end.
+        timesteps = clip_timestep_schedule_fractional(timesteps, self.denoising_start, self.denoising_end)
+        total_steps = len(timesteps) - 1
+
+        # Prepare the CFG scale list.
+        cfg_scale = self._prepare_cfg_scale(total_steps)
+
+        # Load the input latents, if provided.
+        init_latents = context.tensors.load(self.latents.latents_name) if self.latents else None
+        if init_latents is not None:
+            init_latents = init_latents.to(device=device, dtype=inference_dtype)
+
+        # Generate initial latent noise.
+        num_channels_latents = transformer_info.model.config.in_channels
+        assert isinstance(num_channels_latents, int)
+        noise = self._get_noise(
+            num_samples=1,
+            num_channels_latents=num_channels_latents,
+            height=self.height,
+            width=self.width,
+            dtype=inference_dtype,
+            device=device,
+            seed=self.seed,
+        )
+
+        # Prepare input latent image.
+        if init_latents is not None:
+            # Noise the init_latents by the appropriate amount for the first timestep.
+            t_0 = timesteps[0]
+            latents = t_0 * noise + (1.0 - t_0) * init_latents
+        else:
+            # init_latents are not provided, so we are not doing image-to-image (i.e. we are starting from pure noise).
+            if self.denoising_start > 1e-5:
+                raise ValueError("denoising_start should be 0 when initial latents are not provided.")
+            latents = noise
+
+        # If len(timesteps) == 1, then short-circuit. We are just noising the input latents, but not taking any
+        # denoising steps.
+        if len(timesteps) <= 1:
+            return latents
+
+        # Prepare inpaint extension.
+        inpaint_mask = self._prep_inpaint_mask(context, latents)
+        inpaint_extension: InpaintExtension | None = None
+        if inpaint_mask is not None:
+            assert init_latents is not None
+            inpaint_extension = InpaintExtension(
+                init_latents=init_latents,
+                inpaint_mask=inpaint_mask,
+                noise=noise,
+            )
+
+        step_callback = self._build_step_callback(context)
+
+        step_callback(
+            PipelineIntermediateState(
+                step=0,
+                order=1,
+                total_steps=total_steps,
+                timestep=int(timesteps[0]),
+                latents=latents,
+            ),
+        )
+
+        with transformer_info.model_on_device() as (cached_weights, transformer):
+            assert isinstance(transformer, SD3Transformer2DModel)
+
+            # 6. Denoising loop
+            for step_idx, (t_curr, t_prev) in tqdm(list(enumerate(zip(timesteps[:-1], timesteps[1:], strict=True)))):
+                # Expand the latents if we are doing CFG.
+                latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents
+                # Expand the timestep to match the latent model input.
+                # Multiply by 1000 to match the default FlowMatchEulerDiscreteScheduler num_train_timesteps.
+                timestep = torch.tensor([t_curr * 1000], device=device).expand(latent_model_input.shape[0])
+
+                noise_pred = transformer(
+                    hidden_states=latent_model_input,
+                    timestep=timestep,
+                    encoder_hidden_states=prompt_embeds,
+                    pooled_projections=pooled_prompt_embeds,
+                    joint_attention_kwargs=None,
+                    return_dict=False,
+                )[0]
+
+                # Apply CFG.
+                if do_classifier_free_guidance:
+                    noise_pred_uncond, noise_pred_cond = noise_pred.chunk(2)
+                    noise_pred = noise_pred_uncond + cfg_scale[step_idx] * (noise_pred_cond - noise_pred_uncond)
+
+                # Compute the previous noisy sample x_t -> x_t-1.
+                latents_dtype = latents.dtype
+                latents = latents.to(dtype=torch.float32)
+                latents = latents + (t_prev - t_curr) * noise_pred
+                latents = latents.to(dtype=latents_dtype)
+
+                if inpaint_extension is not None:
+                    latents = inpaint_extension.merge_intermediate_latents_with_init_latents(latents, t_prev)
+
+                step_callback(
+                    PipelineIntermediateState(
+                        step=step_idx + 1,
+                        order=1,
+                        total_steps=total_steps,
+                        timestep=int(t_curr),
+                        latents=latents,
+                    ),
+                )
+
+        return latents
+
+    def _build_step_callback(self, context: InvocationContext) -> Callable[[PipelineIntermediateState], None]:
+        def step_callback(state: PipelineIntermediateState) -> None:
+            context.util.sd_step_callback(state, BaseModelType.StableDiffusion3)
+
+        return step_callback

invokeai/app/invocations/sd3_image_to_latents.py

@@ -0,0 +1,65 @@
+import einops
+import torch
+from diffusers.models.autoencoders.autoencoder_kl import AutoencoderKL
+
+from invokeai.app.invocations.baseinvocation import BaseInvocation, Classification, invocation
+from invokeai.app.invocations.fields import (
+    FieldDescriptions,
+    ImageField,
+    Input,
+    InputField,
+    WithBoard,
+    WithMetadata,
+)
+from invokeai.app.invocations.model import VAEField
+from invokeai.app.invocations.primitives import LatentsOutput
+from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.model_manager.load.load_base import LoadedModel
+from invokeai.backend.stable_diffusion.diffusers_pipeline import image_resized_to_grid_as_tensor
+
+
+@invocation(
+    "sd3_i2l",
+    title="SD3 Image to Latents",
+    tags=["image", "latents", "vae", "i2l", "sd3"],
+    category="image",
+    version="1.0.0",
+    classification=Classification.Prototype,
+)
+class SD3ImageToLatentsInvocation(BaseInvocation, WithMetadata, WithBoard):
+    """Generates latents from an image."""
+
+    image: ImageField = InputField(description="The image to encode")
+    vae: VAEField = InputField(description=FieldDescriptions.vae, input=Input.Connection)
+
+    @staticmethod
+    def vae_encode(vae_info: LoadedModel, image_tensor: torch.Tensor) -> torch.Tensor:
+        with vae_info as vae:
+            assert isinstance(vae, AutoencoderKL)
+
+            vae.disable_tiling()
+
+            image_tensor = image_tensor.to(device=vae.device, dtype=vae.dtype)
+            with torch.inference_mode():
+                image_tensor_dist = vae.encode(image_tensor).latent_dist
+                # TODO: Use seed to make sampling reproducible.
+                latents: torch.Tensor = image_tensor_dist.sample().to(dtype=vae.dtype)
+
+            latents = vae.config.scaling_factor * latents
+
+        return latents
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> LatentsOutput:
+        image = context.images.get_pil(self.image.image_name)
+
+        image_tensor = image_resized_to_grid_as_tensor(image.convert("RGB"))
+        if image_tensor.dim() == 3:
+            image_tensor = einops.rearrange(image_tensor, "c h w -> 1 c h w")
+
+        vae_info = context.models.load(self.vae.vae)
+        latents = self.vae_encode(vae_info=vae_info, image_tensor=image_tensor)
+
+        latents = latents.to("cpu")
+        name = context.tensors.save(tensor=latents)
+        return LatentsOutput.build(latents_name=name, latents=latents, seed=None)

invokeai/app/invocations/sd3_latents_to_image.py

@@ -0,0 +1,74 @@
+from contextlib import nullcontext
+
+import torch
+from diffusers.models.autoencoders.autoencoder_kl import AutoencoderKL
+from einops import rearrange
+from PIL import Image
+
+from invokeai.app.invocations.baseinvocation import BaseInvocation, invocation
+from invokeai.app.invocations.fields import (
+    FieldDescriptions,
+    Input,
+    InputField,
+    LatentsField,
+    WithBoard,
+    WithMetadata,
+)
+from invokeai.app.invocations.model import VAEField
+from invokeai.app.invocations.primitives import ImageOutput
+from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.stable_diffusion.extensions.seamless import SeamlessExt
+from invokeai.backend.util.devices import TorchDevice
+
+
+@invocation(
+    "sd3_l2i",
+    title="SD3 Latents to Image",
+    tags=["latents", "image", "vae", "l2i", "sd3"],
+    category="latents",
+    version="1.3.0",
+)
+class SD3LatentsToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
+    """Generates an image from latents."""
+
+    latents: LatentsField = InputField(
+        description=FieldDescriptions.latents,
+        input=Input.Connection,
+    )
+    vae: VAEField = InputField(
+        description=FieldDescriptions.vae,
+        input=Input.Connection,
+    )
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> ImageOutput:
+        latents = context.tensors.load(self.latents.latents_name)
+
+        vae_info = context.models.load(self.vae.vae)
+        assert isinstance(vae_info.model, (AutoencoderKL))
+        with SeamlessExt.static_patch_model(vae_info.model, self.vae.seamless_axes), vae_info as vae:
+            context.util.signal_progress("Running VAE")
+            assert isinstance(vae, (AutoencoderKL))
+            latents = latents.to(vae.device)
+
+            vae.disable_tiling()
+
+            tiling_context = nullcontext()
+
+            # clear memory as vae decode can request a lot
+            TorchDevice.empty_cache()
+
+            with torch.inference_mode(), tiling_context:
+                # copied from diffusers pipeline
+                latents = latents / vae.config.scaling_factor
+                img = vae.decode(latents, return_dict=False)[0]
+
+            img = img.clamp(-1, 1)
+            img = rearrange(img[0], "c h w -> h w c")  # noqa: F821
+            img_pil = Image.fromarray((127.5 * (img + 1.0)).byte().cpu().numpy())
+
+        TorchDevice.empty_cache()
+
+        image_dto = context.images.save(image=img_pil)
+
+        return ImageOutput.build(image_dto)

invokeai/app/invocations/sd3_model_loader.py

@@ -0,0 +1,108 @@
+from typing import Optional
+
+from invokeai.app.invocations.baseinvocation import (
+    BaseInvocation,
+    BaseInvocationOutput,
+    Classification,
+    invocation,
+    invocation_output,
+)
+from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField, UIType
+from invokeai.app.invocations.model import CLIPField, ModelIdentifierField, T5EncoderField, TransformerField, VAEField
+from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.model_manager.config import SubModelType
+
+
+@invocation_output("sd3_model_loader_output")
+class Sd3ModelLoaderOutput(BaseInvocationOutput):
+    """SD3 base model loader output."""
+
+    transformer: TransformerField = OutputField(description=FieldDescriptions.transformer, title="Transformer")
+    clip_l: CLIPField = OutputField(description=FieldDescriptions.clip, title="CLIP L")
+    clip_g: CLIPField = OutputField(description=FieldDescriptions.clip, title="CLIP G")
+    t5_encoder: T5EncoderField = OutputField(description=FieldDescriptions.t5_encoder, title="T5 Encoder")
+    vae: VAEField = OutputField(description=FieldDescriptions.vae, title="VAE")
+
+
+@invocation(
+    "sd3_model_loader",
+    title="SD3 Main Model",
+    tags=["model", "sd3"],
+    category="model",
+    version="1.0.0",
+    classification=Classification.Prototype,
+)
+class Sd3ModelLoaderInvocation(BaseInvocation):
+    """Loads a SD3 base model, outputting its submodels."""
+
+    model: ModelIdentifierField = InputField(
+        description=FieldDescriptions.sd3_model,
+        ui_type=UIType.SD3MainModel,
+        input=Input.Direct,
+    )
+
+    t5_encoder_model: Optional[ModelIdentifierField] = InputField(
+        description=FieldDescriptions.t5_encoder,
+        ui_type=UIType.T5EncoderModel,
+        input=Input.Direct,
+        title="T5 Encoder",
+        default=None,
+    )
+
+    clip_l_model: Optional[ModelIdentifierField] = InputField(
+        description=FieldDescriptions.clip_embed_model,
+        ui_type=UIType.CLIPLEmbedModel,
+        input=Input.Direct,
+        title="CLIP L Encoder",
+        default=None,
+    )
+
+    clip_g_model: Optional[ModelIdentifierField] = InputField(
+        description=FieldDescriptions.clip_g_model,
+        ui_type=UIType.CLIPGEmbedModel,
+        input=Input.Direct,
+        title="CLIP G Encoder",
+        default=None,
+    )
+
+    vae_model: Optional[ModelIdentifierField] = InputField(
+        description=FieldDescriptions.vae_model, ui_type=UIType.VAEModel, title="VAE", default=None
+    )
+
+    def invoke(self, context: InvocationContext) -> Sd3ModelLoaderOutput:
+        transformer = self.model.model_copy(update={"submodel_type": SubModelType.Transformer})
+        vae = (
+            self.vae_model.model_copy(update={"submodel_type": SubModelType.VAE})
+            if self.vae_model
+            else self.model.model_copy(update={"submodel_type": SubModelType.VAE})
+        )
+        tokenizer_l = self.model.model_copy(update={"submodel_type": SubModelType.Tokenizer})
+        clip_encoder_l = (
+            self.clip_l_model.model_copy(update={"submodel_type": SubModelType.TextEncoder})
+            if self.clip_l_model
+            else self.model.model_copy(update={"submodel_type": SubModelType.TextEncoder})
+        )
+        tokenizer_g = self.model.model_copy(update={"submodel_type": SubModelType.Tokenizer2})
+        clip_encoder_g = (
+            self.clip_g_model.model_copy(update={"submodel_type": SubModelType.TextEncoder2})
+            if self.clip_g_model
+            else self.model.model_copy(update={"submodel_type": SubModelType.TextEncoder2})
+        )
+        tokenizer_t5 = (
+            self.t5_encoder_model.model_copy(update={"submodel_type": SubModelType.Tokenizer3})
+            if self.t5_encoder_model
+            else self.model.model_copy(update={"submodel_type": SubModelType.Tokenizer3})
+        )
+        t5_encoder = (
+            self.t5_encoder_model.model_copy(update={"submodel_type": SubModelType.TextEncoder3})
+            if self.t5_encoder_model
+            else self.model.model_copy(update={"submodel_type": SubModelType.TextEncoder3})
+        )
+
+        return Sd3ModelLoaderOutput(
+            transformer=TransformerField(transformer=transformer, loras=[]),
+            clip_l=CLIPField(tokenizer=tokenizer_l, text_encoder=clip_encoder_l, loras=[], skipped_layers=0),
+            clip_g=CLIPField(tokenizer=tokenizer_g, text_encoder=clip_encoder_g, loras=[], skipped_layers=0),
+            t5_encoder=T5EncoderField(tokenizer=tokenizer_t5, text_encoder=t5_encoder),
+            vae=VAEField(vae=vae),
+        )

invokeai/app/invocations/sd3_text_encoder.py

@@ -0,0 +1,201 @@
+from contextlib import ExitStack
+from typing import Iterator, Tuple
+
+import torch
+from transformers import (
+    CLIPTextModel,
+    CLIPTextModelWithProjection,
+    CLIPTokenizer,
+    T5EncoderModel,
+    T5Tokenizer,
+    T5TokenizerFast,
+)
+
+from invokeai.app.invocations.baseinvocation import BaseInvocation, Classification, invocation
+from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField
+from invokeai.app.invocations.model import CLIPField, T5EncoderField
+from invokeai.app.invocations.primitives import SD3ConditioningOutput
+from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.model_manager.config import ModelFormat
+from invokeai.backend.patches.lora_conversions.flux_lora_constants import FLUX_LORA_CLIP_PREFIX
+from invokeai.backend.patches.model_patch_raw import ModelPatchRaw
+from invokeai.backend.patches.model_patcher import LayerPatcher
+from invokeai.backend.stable_diffusion.diffusion.conditioning_data import ConditioningFieldData, SD3ConditioningInfo
+
+# The SD3 T5 Max Sequence Length set based on the default in diffusers.
+SD3_T5_MAX_SEQ_LEN = 256
+
+
+@invocation(
+    "sd3_text_encoder",
+    title="SD3 Text Encoding",
+    tags=["prompt", "conditioning", "sd3"],
+    category="conditioning",
+    version="1.0.0",
+    classification=Classification.Prototype,
+)
+class Sd3TextEncoderInvocation(BaseInvocation):
+    """Encodes and preps a prompt for a SD3 image."""
+
+    clip_l: CLIPField = InputField(
+        title="CLIP L",
+        description=FieldDescriptions.clip,
+        input=Input.Connection,
+    )
+    clip_g: CLIPField = InputField(
+        title="CLIP G",
+        description=FieldDescriptions.clip,
+        input=Input.Connection,
+    )
+
+    # The SD3 models were trained with text encoder dropout, so the T5 encoder can be omitted to save time/memory.
+    t5_encoder: T5EncoderField | None = InputField(
+        title="T5Encoder",
+        default=None,
+        description=FieldDescriptions.t5_encoder,
+        input=Input.Connection,
+    )
+    prompt: str = InputField(description="Text prompt to encode.")
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> SD3ConditioningOutput:
+        # Note: The text encoding model are run in separate functions to ensure that all model references are locally
+        # scoped. This ensures that earlier models can be freed and gc'd before loading later models (if necessary).
+
+        clip_l_embeddings, clip_l_pooled_embeddings = self._clip_encode(context, self.clip_l)
+        clip_g_embeddings, clip_g_pooled_embeddings = self._clip_encode(context, self.clip_g)
+
+        t5_embeddings: torch.Tensor | None = None
+        if self.t5_encoder is not None:
+            t5_embeddings = self._t5_encode(context, SD3_T5_MAX_SEQ_LEN)
+
+        conditioning_data = ConditioningFieldData(
+            conditionings=[
+                SD3ConditioningInfo(
+                    clip_l_embeds=clip_l_embeddings,
+                    clip_l_pooled_embeds=clip_l_pooled_embeddings,
+                    clip_g_embeds=clip_g_embeddings,
+                    clip_g_pooled_embeds=clip_g_pooled_embeddings,
+                    t5_embeds=t5_embeddings,
+                )
+            ]
+        )
+
+        conditioning_name = context.conditioning.save(conditioning_data)
+        return SD3ConditioningOutput.build(conditioning_name)
+
+    def _t5_encode(self, context: InvocationContext, max_seq_len: int) -> torch.Tensor:
+        assert self.t5_encoder is not None
+        t5_tokenizer_info = context.models.load(self.t5_encoder.tokenizer)
+        t5_text_encoder_info = context.models.load(self.t5_encoder.text_encoder)
+
+        prompt = [self.prompt]
+
+        with (
+            t5_text_encoder_info as t5_text_encoder,
+            t5_tokenizer_info as t5_tokenizer,
+        ):
+            context.util.signal_progress("Running T5 encoder")
+            assert isinstance(t5_text_encoder, T5EncoderModel)
+            assert isinstance(t5_tokenizer, (T5Tokenizer, T5TokenizerFast))
+
+            text_inputs = t5_tokenizer(
+                prompt,
+                padding="max_length",
+                max_length=max_seq_len,
+                truncation=True,
+                add_special_tokens=True,
+                return_tensors="pt",
+            )
+            text_input_ids = text_inputs.input_ids
+            untruncated_ids = t5_tokenizer(prompt, padding="longest", return_tensors="pt").input_ids
+            assert isinstance(text_input_ids, torch.Tensor)
+            assert isinstance(untruncated_ids, torch.Tensor)
+            if untruncated_ids.shape[-1] >= text_input_ids.shape[-1] and not torch.equal(
+                text_input_ids, untruncated_ids
+            ):
+                removed_text = t5_tokenizer.batch_decode(untruncated_ids[:, max_seq_len - 1 : -1])
+                context.logger.warning(
+                    "The following part of your input was truncated because `max_sequence_length` is set to "
+                    f" {max_seq_len} tokens: {removed_text}"
+                )
+
+            prompt_embeds = t5_text_encoder(text_input_ids.to(t5_text_encoder.device))[0]
+
+        assert isinstance(prompt_embeds, torch.Tensor)
+        return prompt_embeds
+
+    def _clip_encode(
+        self, context: InvocationContext, clip_model: CLIPField, tokenizer_max_length: int = 77
+    ) -> Tuple[torch.Tensor, torch.Tensor]:
+        clip_tokenizer_info = context.models.load(clip_model.tokenizer)
+        clip_text_encoder_info = context.models.load(clip_model.text_encoder)
+
+        prompt = [self.prompt]
+
+        with (
+            clip_text_encoder_info.model_on_device() as (cached_weights, clip_text_encoder),
+            clip_tokenizer_info as clip_tokenizer,
+            ExitStack() as exit_stack,
+        ):
+            context.util.signal_progress("Running CLIP encoder")
+            assert isinstance(clip_text_encoder, (CLIPTextModel, CLIPTextModelWithProjection))
+            assert isinstance(clip_tokenizer, CLIPTokenizer)
+
+            clip_text_encoder_config = clip_text_encoder_info.config
+            assert clip_text_encoder_config is not None
+
+            # Apply LoRA models to the CLIP encoder.
+            # Note: We apply the LoRA after the transformer has been moved to its target device for faster patching.
+            if clip_text_encoder_config.format in [ModelFormat.Diffusers]:
+                # The model is non-quantized, so we can apply the LoRA weights directly into the model.
+                exit_stack.enter_context(
+                    LayerPatcher.apply_model_patches(
+                        model=clip_text_encoder,
+                        patches=self._clip_lora_iterator(context, clip_model),
+                        prefix=FLUX_LORA_CLIP_PREFIX,
+                        cached_weights=cached_weights,
+                    )
+                )
+            else:
+                # There are currently no supported CLIP quantized models. Add support here if needed.
+                raise ValueError(f"Unsupported model format: {clip_text_encoder_config.format}")
+
+            clip_text_encoder = clip_text_encoder.eval().requires_grad_(False)
+
+            text_inputs = clip_tokenizer(
+                prompt,
+                padding="max_length",
+                max_length=tokenizer_max_length,
+                truncation=True,
+                return_tensors="pt",
+            )
+
+            text_input_ids = text_inputs.input_ids
+            untruncated_ids = clip_tokenizer(prompt, padding="longest", return_tensors="pt").input_ids
+            assert isinstance(text_input_ids, torch.Tensor)
+            assert isinstance(untruncated_ids, torch.Tensor)
+            if untruncated_ids.shape[-1] >= text_input_ids.shape[-1] and not torch.equal(
+                text_input_ids, untruncated_ids
+            ):
+                removed_text = clip_tokenizer.batch_decode(untruncated_ids[:, tokenizer_max_length - 1 : -1])
+                context.logger.warning(
+                    "The following part of your input was truncated because CLIP can only handle sequences up to"
+                    f" {tokenizer_max_length} tokens: {removed_text}"
+                )
+            prompt_embeds = clip_text_encoder(
+                input_ids=text_input_ids.to(clip_text_encoder.device), output_hidden_states=True
+            )
+            pooled_prompt_embeds = prompt_embeds[0]
+            prompt_embeds = prompt_embeds.hidden_states[-2]
+
+            return prompt_embeds, pooled_prompt_embeds
+
+    def _clip_lora_iterator(
+        self, context: InvocationContext, clip_model: CLIPField
+    ) -> Iterator[Tuple[ModelPatchRaw, float]]:
+        for lora in clip_model.loras:
+            lora_info = context.models.load(lora.lora)
+            assert isinstance(lora_info.model, ModelPatchRaw)
+            yield (lora_info.model, lora.weight)
+            del lora_info

invokeai/app/invocations/segment_anything.py

@@ -5,7 +5,7 @@ from typing import Literal
 import numpy as np
 import torch
 from PIL import Image
-from pydantic import BaseModel, Field, model_validator
+from pydantic import BaseModel, Field
 from transformers import AutoModelForMaskGeneration, AutoProcessor
 from transformers.models.sam import SamModel
 from transformers.models.sam.processing_sam import SamProcessor
@@ -77,19 +77,14 @@ class SegmentAnythingInvocation(BaseInvocation):
         default="all",
     )
 
-    @model_validator(mode="after")
-    def check_point_lists_or_bounding_box(self):
-        if self.point_lists is None and self.bounding_boxes is None:
-            raise ValueError("Either point_lists or bounding_box must be provided.")
-        elif self.point_lists is not None and self.bounding_boxes is not None:
-            raise ValueError("Only one of point_lists or bounding_box can be provided.")
-        return self
-
     @torch.no_grad()
     def invoke(self, context: InvocationContext) -> MaskOutput:
         # The models expect a 3-channel RGB image.
         image_pil = context.images.get_pil(self.image.image_name, mode="RGB")
 
+        if self.point_lists is not None and self.bounding_boxes is not None:
+            raise ValueError("Only one of point_lists or bounding_box can be provided.")
+
         if (not self.bounding_boxes or len(self.bounding_boxes) == 0) and (
             not self.point_lists or len(self.point_lists) == 0
         ):

invokeai/app/invocations/tiled_multi_diffusion_denoise_latents.py

@@ -22,8 +22,8 @@ from invokeai.app.invocations.fields import (
 from invokeai.app.invocations.model import UNetField
 from invokeai.app.invocations.primitives import LatentsOutput
 from invokeai.app.services.shared.invocation_context import InvocationContext
-from invokeai.backend.lora.lora_model_raw import LoRAModelRaw
-from invokeai.backend.lora.lora_patcher import LoRAPatcher
+from invokeai.backend.patches.model_patch_raw import ModelPatchRaw
+from invokeai.backend.patches.model_patcher import LayerPatcher
 from invokeai.backend.stable_diffusion.diffusers_pipeline import ControlNetData, PipelineIntermediateState
 from invokeai.backend.stable_diffusion.multi_diffusion_pipeline import (
     MultiDiffusionPipeline,
@@ -194,10 +194,10 @@ class TiledMultiDiffusionDenoiseLatents(BaseInvocation):
             context.util.sd_step_callback(state, unet_config.base)
 
         # Prepare an iterator that yields the UNet's LoRA models and their weights.
-        def _lora_loader() -> Iterator[Tuple[LoRAModelRaw, float]]:
+        def _lora_loader() -> Iterator[Tuple[ModelPatchRaw, float]]:
             for lora in self.unet.loras:
                 lora_info = context.models.load(lora.lora)
-                assert isinstance(lora_info.model, LoRAModelRaw)
+                assert isinstance(lora_info.model, ModelPatchRaw)
                 yield (lora_info.model, lora.weight)
                 del lora_info
 
@@ -207,7 +207,7 @@ class TiledMultiDiffusionDenoiseLatents(BaseInvocation):
         with (
             ExitStack() as exit_stack,
             unet_info as unet,
-            LoRAPatcher.apply_lora_patches(model=unet, patches=_lora_loader(), prefix="lora_unet_"),
+            LayerPatcher.apply_model_patches(model=unet, patches=_lora_loader(), prefix="lora_unet_"),
         ):
             assert isinstance(unet, UNet2DConditionModel)
             latents = latents.to(device=unet.device, dtype=unet.dtype)

invokeai/app/services/config/config_default.py

@@ -4,6 +4,7 @@
 from __future__ import annotations
 
 import copy
+import filecmp
 import locale
 import os
 import re
@@ -96,6 +97,7 @@ class InvokeAIAppConfig(BaseSettings):
         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.
+        log_level_network: Log level for network-related messages. 'info' and 'debug' are very verbose.<br>Valid values: `debug`, `info`, `warning`, `error`, `critical`
         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`.
@@ -162,6 +164,7 @@ class InvokeAIAppConfig(BaseSettings):
     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.")
+    log_level_network:        LOG_LEVEL = Field(default='warning',          description="Log level for network-related messages. 'info' and 'debug' are very verbose.")
 
     # Development
     use_memory_db:                 bool = Field(default=False,              description="Use in-memory database. Useful for development.")
@@ -525,9 +528,35 @@ def get_config() -> InvokeAIAppConfig:
     ]
     example_config.write_file(config.config_file_path.with_suffix(".example.yaml"), as_example=True)
 
-    # Copy all legacy configs - We know `__path__[0]` is correct here
+    # Copy all legacy configs only if needed
+    # We know `__path__[0]` is correct here
     configs_src = Path(model_configs.__path__[0])  # pyright: ignore [reportUnknownMemberType, reportUnknownArgumentType, reportAttributeAccessIssue]
-    shutil.copytree(configs_src, config.legacy_conf_path, dirs_exist_ok=True)
+    dest_path = config.legacy_conf_path
+
+    # Create destination (we don't need to check for existence)
+    dest_path.mkdir(parents=True, exist_ok=True)
+
+    # Compare directories recursively
+    comparison = filecmp.dircmp(configs_src, dest_path)
+    need_copy = any(
+        [
+            comparison.left_only,  # Files exist only in source
+            comparison.diff_files,  # Files that differ
+            comparison.common_funny,  # Files that couldn't be compared
+        ]
+    )
+
+    if need_copy:
+        # Get permissions from destination directory
+        dest_mode = dest_path.stat().st_mode
+
+        # Copy directory tree
+        shutil.copytree(configs_src, dest_path, dirs_exist_ok=True)
+
+        # Set permissions on copied files to match destination directory
+        dest_path.chmod(dest_mode)
+        for p in dest_path.glob("**/*"):
+            p.chmod(dest_mode)
 
     if config.config_file_path.exists():
         config_from_file = load_and_migrate_config(config.config_file_path)

invokeai/app/services/model_install/model_install_default.py

@@ -438,9 +438,10 @@ class ModelInstallService(ModelInstallServiceBase):
         variants = "|".join(ModelRepoVariant.__members__.values())
         hf_repoid_re = f"^([^/:]+/[^/:]+)(?::({variants})?(?::/?([^:]+))?)?$"
         source_obj: Optional[StringLikeSource] = None
+        source_stripped = source.strip('"')
 
-        if Path(source).exists():  # A local file or directory
-            source_obj = LocalModelSource(path=Path(source))
+        if Path(source_stripped).exists():  # A local file or directory
+            source_obj = LocalModelSource(path=Path(source_stripped))
         elif match := re.match(hf_repoid_re, source):
             source_obj = HFModelSource(
                 repo_id=match.group(1),

invokeai/app/services/model_load/model_load_default.py

@@ -86,7 +86,7 @@ class ModelLoadService(ModelLoadServiceBase):
 
         def torch_load_file(checkpoint: Path) -> AnyModel:
             scan_result = scan_file_path(checkpoint)
-            if scan_result.infected_files != 0:
+            if scan_result.infected_files != 0 or scan_result.scan_err:
                 raise Exception("The model at {checkpoint} is potentially infected by malware. Aborting load.")
             result = torch_load(checkpoint, map_location="cpu")
             return result

invokeai/app/services/model_records/model_records_base.py

@@ -15,6 +15,7 @@ from invokeai.app.util.model_exclude_null import BaseModelExcludeNull
 from invokeai.backend.model_manager.config import (
     AnyModelConfig,
     BaseModelType,
+    ClipVariantType,
     ControlAdapterDefaultSettings,
     MainModelDefaultSettings,
     ModelFormat,
@@ -85,7 +86,7 @@ class ModelRecordChanges(BaseModelExcludeNull):
 
     # Checkpoint-specific changes
     # TODO(MM2): Should we expose these? Feels footgun-y...
-    variant: Optional[ModelVariantType] = Field(description="The variant of the model.", default=None)
+    variant: Optional[ModelVariantType | ClipVariantType] = Field(description="The variant of the model.", default=None)
     prediction_type: Optional[SchedulerPredictionType] = Field(
         description="The prediction type of the model.", default=None
     )

invokeai/app/services/session_processor/session_processor_default.py

@@ -378,6 +378,9 @@ class DefaultSessionProcessor(SessionProcessorBase):
         self._poll_now()
 
     async def _on_queue_item_status_changed(self, event: FastAPIEvent[QueueItemStatusChangedEvent]) -> None:
+        # Make sure the cancel event is for the currently processing queue item
+        if self._queue_item and self._queue_item.item_id != event[1].item_id:
+            return
         if self._queue_item and event[1].status in ["completed", "failed", "canceled"]:
             # When the queue item is canceled via HTTP, the queue item status is set to `"canceled"` and this event is
             # emitted. We need to respond to this event and stop graph execution. This is done by setting the cancel
@@ -436,7 +439,9 @@ class DefaultSessionProcessor(SessionProcessorBase):
                         poll_now_event.wait(self._polling_interval)
                         continue
 
-                    self._invoker.services.logger.debug(f"Executing queue item {self._queue_item.item_id}")
+                    self._invoker.services.logger.info(
+                        f"Executing queue item {self._queue_item.item_id}, session {self._queue_item.session_id}"
+                    )
                     cancel_event.clear()
 
                     # Run the graph

invokeai/app/services/session_queue/session_queue_common.py

@@ -16,6 +16,7 @@ from pydantic import (
 from pydantic_core import to_jsonable_python
 
 from invokeai.app.invocations.baseinvocation import BaseInvocation
+from invokeai.app.invocations.fields import ImageField
 from invokeai.app.services.shared.graph import Graph, GraphExecutionState, NodeNotFoundError
 from invokeai.app.services.workflow_records.workflow_records_common import (
     WorkflowWithoutID,
@@ -51,11 +52,7 @@ class SessionQueueItemNotFoundError(ValueError):
 
 # region Batch
 
-BatchDataType = Union[
-    StrictStr,
-    float,
-    int,
-]
+BatchDataType = Union[StrictStr, float, int, ImageField]
 
 
 class NodeFieldValue(BaseModel):

invokeai/app/services/shared/invocation_context.py

@@ -160,6 +160,10 @@ class LoggerInterface(InvocationContextInterface):
 
 
 class ImagesInterface(InvocationContextInterface):
+    def __init__(self, services: InvocationServices, data: InvocationContextData, util: "UtilInterface") -> None:
+        super().__init__(services, data)
+        self._util = util
+
     def save(
         self,
         image: Image,
@@ -186,6 +190,8 @@ class ImagesInterface(InvocationContextInterface):
             The saved image DTO.
         """
 
+        self._util.signal_progress("Saving image")
+
         # If `metadata` is provided directly, use that. Else, use the metadata provided by `WithMetadata`, falling back to None.
         metadata_ = None
         if metadata:
@@ -336,6 +342,10 @@ class ConditioningInterface(InvocationContextInterface):
 class ModelsInterface(InvocationContextInterface):
     """Common API for loading, downloading and managing models."""
 
+    def __init__(self, services: InvocationServices, data: InvocationContextData, util: "UtilInterface") -> None:
+        super().__init__(services, data)
+        self._util = util
+
     def exists(self, identifier: Union[str, "ModelIdentifierField"]) -> bool:
         """Check if a model exists.
 
@@ -368,11 +378,15 @@ class ModelsInterface(InvocationContextInterface):
 
         if isinstance(identifier, str):
             model = self._services.model_manager.store.get_model(identifier)
-            return self._services.model_manager.load.load_model(model, submodel_type)
         else:
-            _submodel_type = submodel_type or identifier.submodel_type
+            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)
+
+        message = f"Loading model {model.name}"
+        if submodel_type:
+            message += f" ({submodel_type.value})"
+        self._util.signal_progress(message)
+        return self._services.model_manager.load.load_model(model, submodel_type)
 
     def load_by_attrs(
         self, name: str, base: BaseModelType, type: ModelType, submodel_type: Optional[SubModelType] = None
@@ -397,6 +411,10 @@ class ModelsInterface(InvocationContextInterface):
         if len(configs) > 1:
             raise ValueError(f"More than one model found with name {name}, base {base}, and type {type}")
 
+        message = f"Loading model {name}"
+        if submodel_type:
+            message += f" ({submodel_type.value})"
+        self._util.signal_progress(message)
         return self._services.model_manager.load.load_model(configs[0], submodel_type)
 
     def get_config(self, identifier: Union[str, "ModelIdentifierField"]) -> AnyModelConfig:
@@ -467,6 +485,7 @@ class ModelsInterface(InvocationContextInterface):
         Returns:
             Path to the downloaded model
         """
+        self._util.signal_progress(f"Downloading model {source}")
         return self._services.model_manager.install.download_and_cache_model(source=source)
 
     def load_local_model(
@@ -489,6 +508,8 @@ class ModelsInterface(InvocationContextInterface):
         Returns:
             A LoadedModelWithoutConfig object.
         """
+
+        self._util.signal_progress(f"Loading model {model_path.name}")
         return self._services.model_manager.load.load_model_from_path(model_path=model_path, loader=loader)
 
     def load_remote_model(
@@ -514,6 +535,8 @@ class ModelsInterface(InvocationContextInterface):
             A LoadedModelWithoutConfig object.
         """
         model_path = self._services.model_manager.install.download_and_cache_model(source=str(source))
+
+        self._util.signal_progress(f"Loading model {source}")
         return self._services.model_manager.load.load_model_from_path(model_path=model_path, loader=loader)
 
 
@@ -707,12 +730,12 @@ def build_invocation_context(
     """
 
     logger = LoggerInterface(services=services, data=data)
-    images = ImagesInterface(services=services, data=data)
     tensors = TensorsInterface(services=services, data=data)
-    models = ModelsInterface(services=services, data=data)
     config = ConfigInterface(services=services, data=data)
     util = UtilInterface(services=services, data=data, is_canceled=is_canceled)
     conditioning = ConditioningInterface(services=services, data=data)
+    models = ModelsInterface(services=services, data=data, util=util)
+    images = ImagesInterface(services=services, data=data, util=util)
     boards = BoardsInterface(services=services, data=data)
 
     ctx = InvocationContext(

invokeai/app/services/shared/sqlite_migrator/migrations/migration_11.py

@@ -35,7 +35,7 @@ class Migration11Callback:
 
     def _remove_convert_cache(self) -> None:
         """Rename models/.cache to models/.convert_cache."""
-        self._logger.info("Removing .cache directory. Converted models will now be cached in .convert_cache.")
+        self._logger.info("Removing models/.cache directory. Converted models will now be cached in .convert_cache.")
         legacy_convert_path = self._app_config.root_path / "models" / ".cache"
         shutil.rmtree(legacy_convert_path, ignore_errors=True)
 

invokeai/app/services/workflow_records/default_workflows/SD3.5 Text to Image.json

@@ -0,0 +1,382 @@
+{
+    "name": "SD3.5 Text to Image",
+   "author": "InvokeAI",
+    "description": "Sample text to image workflow for Stable Diffusion 3.5",
+    "version": "1.0.0",
+    "contact": "invoke@invoke.ai",
+    "tags": "text2image, SD3.5, default",
+  "notes": "",
+    "exposedFields": [
+      {
+        "nodeId": "3f22f668-0e02-4fde-a2bb-c339586ceb4c",
+        "fieldName": "model"
+      },
+      {
+        "nodeId": "e17d34e7-6ed1-493c-9a85-4fcd291cb084",
+        "fieldName": "prompt"
+      }
+    ],
+    "meta": {
+      "version": "3.0.0",
+      "category": "default"
+    },
+    "id": "e3a51d6b-8208-4d6d-b187-fcfe8b32934c",
+    "nodes": [
+      {
+        "id": "3f22f668-0e02-4fde-a2bb-c339586ceb4c",
+        "type": "invocation",
+        "data": {
+          "id": "3f22f668-0e02-4fde-a2bb-c339586ceb4c",
+          "type": "sd3_model_loader",
+          "version": "1.0.0",
+          "label": "",
+          "notes": "",
+          "isOpen": true,
+          "isIntermediate": true,
+          "useCache": true,
+          "nodePack": "invokeai",
+          "inputs": {
+            "model": {
+              "name": "model",
+              "label": "",
+              "value": {
+                "key": "f7b20be9-92a8-4cfb-bca4-6c3b5535c10b",
+                "hash": "placeholder",
+                "name": "stable-diffusion-3.5-medium",
+                "base": "sd-3",
+                "type": "main"
+              }
+            },
+            "t5_encoder_model": {
+              "name": "t5_encoder_model",
+              "label": ""
+            },
+            "clip_l_model": {
+              "name": "clip_l_model",
+              "label": ""
+            },
+            "clip_g_model": {
+              "name": "clip_g_model",
+              "label": ""
+            },
+            "vae_model": {
+              "name": "vae_model",
+              "label": ""
+            }
+          }
+        },
+        "position": {
+          "x": -55.58689609637031,
+          "y": -111.53602444662268
+        }
+      },
+      {
+        "id": "f7e394ac-6394-4096-abcb-de0d346506b3",
+        "type": "invocation",
+        "data": {
+          "id": "f7e394ac-6394-4096-abcb-de0d346506b3",
+          "type": "rand_int",
+          "version": "1.0.1",
+          "label": "",
+          "notes": "",
+          "isOpen": true,
+          "isIntermediate": true,
+          "useCache": false,
+          "nodePack": "invokeai",
+          "inputs": {
+            "low": {
+              "name": "low",
+              "label": "",
+              "value": 0
+            },
+            "high": {
+              "name": "high",
+              "label": "",
+              "value": 2147483647
+            }
+          }
+        },
+        "position": {
+          "x": 470.45870147220353,
+          "y": 350.3141781644303
+        }
+      },
+      {
+        "id": "9eb72af0-dd9e-4ec5-ad87-d65e3c01f48b",
+        "type": "invocation",
+        "data": {
+          "id": "9eb72af0-dd9e-4ec5-ad87-d65e3c01f48b",
+          "type": "sd3_l2i",
+          "version": "1.3.0",
+          "label": "",
+          "notes": "",
+          "isOpen": true,
+          "isIntermediate": false,
+          "useCache": true,
+          "nodePack": "invokeai",
+          "inputs": {
+            "board": {
+              "name": "board",
+              "label": ""
+            },
+            "metadata": {
+              "name": "metadata",
+              "label": ""
+            },
+            "latents": {
+              "name": "latents",
+              "label": ""
+            },
+            "vae": {
+              "name": "vae",
+              "label": ""
+            }
+          }
+        },
+        "position": {
+          "x": 1192.3097009334897,
+          "y": -366.0994675072209
+        }
+      },
+      {
+        "id": "3b4f7f27-cfc0-4373-a009-99c5290d0cd6",
+        "type": "invocation",
+        "data": {
+          "id": "3b4f7f27-cfc0-4373-a009-99c5290d0cd6",
+          "type": "sd3_text_encoder",
+          "version": "1.0.0",
+          "label": "",
+          "notes": "",
+          "isOpen": true,
+          "isIntermediate": true,
+          "useCache": true,
+          "nodePack": "invokeai",
+          "inputs": {
+            "clip_l": {
+              "name": "clip_l",
+              "label": ""
+            },
+            "clip_g": {
+              "name": "clip_g",
+              "label": ""
+            },
+            "t5_encoder": {
+              "name": "t5_encoder",
+              "label": ""
+            },
+            "prompt": {
+              "name": "prompt",
+              "label": "",
+              "value": ""
+            }
+          }
+        },
+        "position": {
+          "x": 408.16054647924784,
+          "y": 65.06415352118786
+        }
+      },
+      {
+        "id": "e17d34e7-6ed1-493c-9a85-4fcd291cb084",
+        "type": "invocation",
+        "data": {
+          "id": "e17d34e7-6ed1-493c-9a85-4fcd291cb084",
+          "type": "sd3_text_encoder",
+          "version": "1.0.0",
+          "label": "",
+          "notes": "",
+          "isOpen": true,
+          "isIntermediate": true,
+          "useCache": true,
+          "nodePack": "invokeai",
+          "inputs": {
+            "clip_l": {
+              "name": "clip_l",
+              "label": ""
+            },
+            "clip_g": {
+              "name": "clip_g",
+              "label": ""
+            },
+            "t5_encoder": {
+              "name": "t5_encoder",
+              "label": ""
+            },
+            "prompt": {
+              "name": "prompt",
+              "label": "",
+              "value": ""
+            }
+          }
+        },
+        "position": {
+          "x": 378.9283412440941,
+          "y": -302.65777497352553
+        }
+      },
+      {
+        "id": "c7539f7b-7ac5-49b9-93eb-87ede611409f",
+        "type": "invocation",
+        "data": {
+          "id": "c7539f7b-7ac5-49b9-93eb-87ede611409f",
+          "type": "sd3_denoise",
+          "version": "1.0.0",
+          "label": "",
+          "notes": "",
+          "isOpen": true,
+          "isIntermediate": true,
+          "useCache": true,
+          "nodePack": "invokeai",
+          "inputs": {
+            "board": {
+              "name": "board",
+              "label": ""
+            },
+            "metadata": {
+              "name": "metadata",
+              "label": ""
+            },
+            "transformer": {
+              "name": "transformer",
+              "label": ""
+            },
+            "positive_conditioning": {
+              "name": "positive_conditioning",
+              "label": ""
+            },
+            "negative_conditioning": {
+              "name": "negative_conditioning",
+              "label": ""
+            },
+            "cfg_scale": {
+              "name": "cfg_scale",
+              "label": "",
+              "value": 3.5
+            },
+            "width": {
+              "name": "width",
+              "label": "",
+              "value": 1024
+            },
+            "height": {
+              "name": "height",
+              "label": "",
+              "value": 1024
+            },
+            "steps": {
+              "name": "steps",
+              "label": "",
+              "value": 30
+            },
+            "seed": {
+              "name": "seed",
+              "label": "",
+              "value": 0
+            }
+          }
+        },
+        "position": {
+          "x": 813.7814762740603,
+          "y": -142.20529727605867
+        }
+      }
+    ],
+    "edges": [
+      {
+        "id": "reactflow__edge-3f22f668-0e02-4fde-a2bb-c339586ceb4cvae-9eb72af0-dd9e-4ec5-ad87-d65e3c01f48bvae",
+        "type": "default",
+        "source": "3f22f668-0e02-4fde-a2bb-c339586ceb4c",
+        "target": "9eb72af0-dd9e-4ec5-ad87-d65e3c01f48b",
+        "sourceHandle": "vae",
+        "targetHandle": "vae"
+      },
+      {
+        "id": "reactflow__edge-3f22f668-0e02-4fde-a2bb-c339586ceb4ct5_encoder-3b4f7f27-cfc0-4373-a009-99c5290d0cd6t5_encoder",
+        "type": "default",
+        "source": "3f22f668-0e02-4fde-a2bb-c339586ceb4c",
+        "target": "3b4f7f27-cfc0-4373-a009-99c5290d0cd6",
+        "sourceHandle": "t5_encoder",
+        "targetHandle": "t5_encoder"
+      },
+      {
+        "id": "reactflow__edge-3f22f668-0e02-4fde-a2bb-c339586ceb4ct5_encoder-e17d34e7-6ed1-493c-9a85-4fcd291cb084t5_encoder",
+        "type": "default",
+        "source": "3f22f668-0e02-4fde-a2bb-c339586ceb4c",
+        "target": "e17d34e7-6ed1-493c-9a85-4fcd291cb084",
+        "sourceHandle": "t5_encoder",
+        "targetHandle": "t5_encoder"
+      },
+      {
+        "id": "reactflow__edge-3f22f668-0e02-4fde-a2bb-c339586ceb4cclip_g-3b4f7f27-cfc0-4373-a009-99c5290d0cd6clip_g",
+        "type": "default",
+        "source": "3f22f668-0e02-4fde-a2bb-c339586ceb4c",
+        "target": "3b4f7f27-cfc0-4373-a009-99c5290d0cd6",
+        "sourceHandle": "clip_g",
+        "targetHandle": "clip_g"
+      },
+      {
+        "id": "reactflow__edge-3f22f668-0e02-4fde-a2bb-c339586ceb4cclip_g-e17d34e7-6ed1-493c-9a85-4fcd291cb084clip_g",
+        "type": "default",
+        "source": "3f22f668-0e02-4fde-a2bb-c339586ceb4c",
+        "target": "e17d34e7-6ed1-493c-9a85-4fcd291cb084",
+        "sourceHandle": "clip_g",
+        "targetHandle": "clip_g"
+      },
+      {
+        "id": "reactflow__edge-3f22f668-0e02-4fde-a2bb-c339586ceb4cclip_l-3b4f7f27-cfc0-4373-a009-99c5290d0cd6clip_l",
+        "type": "default",
+        "source": "3f22f668-0e02-4fde-a2bb-c339586ceb4c",
+        "target": "3b4f7f27-cfc0-4373-a009-99c5290d0cd6",
+        "sourceHandle": "clip_l",
+        "targetHandle": "clip_l"
+      },
+      {
+        "id": "reactflow__edge-3f22f668-0e02-4fde-a2bb-c339586ceb4cclip_l-e17d34e7-6ed1-493c-9a85-4fcd291cb084clip_l",
+        "type": "default",
+        "source": "3f22f668-0e02-4fde-a2bb-c339586ceb4c",
+        "target": "e17d34e7-6ed1-493c-9a85-4fcd291cb084",
+        "sourceHandle": "clip_l",
+        "targetHandle": "clip_l"
+      },
+      {
+        "id": "reactflow__edge-3f22f668-0e02-4fde-a2bb-c339586ceb4ctransformer-c7539f7b-7ac5-49b9-93eb-87ede611409ftransformer",
+        "type": "default",
+        "source": "3f22f668-0e02-4fde-a2bb-c339586ceb4c",
+        "target": "c7539f7b-7ac5-49b9-93eb-87ede611409f",
+        "sourceHandle": "transformer",
+        "targetHandle": "transformer"
+      },
+      {
+        "id": "reactflow__edge-f7e394ac-6394-4096-abcb-de0d346506b3value-c7539f7b-7ac5-49b9-93eb-87ede611409fseed",
+        "type": "default",
+        "source": "f7e394ac-6394-4096-abcb-de0d346506b3",
+        "target": "c7539f7b-7ac5-49b9-93eb-87ede611409f",
+        "sourceHandle": "value",
+        "targetHandle": "seed"
+      },
+      {
+        "id": "reactflow__edge-c7539f7b-7ac5-49b9-93eb-87ede611409flatents-9eb72af0-dd9e-4ec5-ad87-d65e3c01f48blatents",
+        "type": "default",
+        "source": "c7539f7b-7ac5-49b9-93eb-87ede611409f",
+        "target": "9eb72af0-dd9e-4ec5-ad87-d65e3c01f48b",
+        "sourceHandle": "latents",
+        "targetHandle": "latents"
+      },
+      {
+        "id": "reactflow__edge-e17d34e7-6ed1-493c-9a85-4fcd291cb084conditioning-c7539f7b-7ac5-49b9-93eb-87ede611409fpositive_conditioning",
+        "type": "default",
+        "source": "e17d34e7-6ed1-493c-9a85-4fcd291cb084",
+        "target": "c7539f7b-7ac5-49b9-93eb-87ede611409f",
+        "sourceHandle": "conditioning",
+        "targetHandle": "positive_conditioning"
+      },
+      {
+        "id": "reactflow__edge-3b4f7f27-cfc0-4373-a009-99c5290d0cd6conditioning-c7539f7b-7ac5-49b9-93eb-87ede611409fnegative_conditioning",
+        "type": "default",
+        "source": "3b4f7f27-cfc0-4373-a009-99c5290d0cd6",
+        "target": "c7539f7b-7ac5-49b9-93eb-87ede611409f",
+        "sourceHandle": "conditioning",
+        "targetHandle": "negative_conditioning"
+      }
+    ]
+  }

invokeai/app/util/step_callback.py

@@ -34,6 +34,25 @@ SD1_5_LATENT_RGB_FACTORS = [
     [-0.1307, -0.1874, -0.7445],  # L4
 ]
 
+SD3_5_LATENT_RGB_FACTORS = [
+    [-0.05240681, 0.03251581, 0.0749016],
+    [-0.0580572, 0.00759826, 0.05729818],
+    [0.16144888, 0.01270368, -0.03768577],
+    [0.14418615, 0.08460266, 0.15941818],
+    [0.04894035, 0.0056485, -0.06686988],
+    [0.05187166, 0.19222395, 0.06261094],
+    [0.1539433, 0.04818359, 0.07103094],
+    [-0.08601796, 0.09013458, 0.10893912],
+    [-0.12398469, -0.06766567, 0.0033688],
+    [-0.0439737, 0.07825329, 0.02258823],
+    [0.03101129, 0.06382551, 0.07753657],
+    [-0.01315361, 0.08554491, -0.08772475],
+    [0.06464487, 0.05914605, 0.13262741],
+    [-0.07863674, -0.02261737, -0.12761454],
+    [-0.09923835, -0.08010759, -0.06264447],
+    [-0.03392309, -0.0804029, -0.06078822],
+]
+
 FLUX_LATENT_RGB_FACTORS = [
     [-0.0412, 0.0149, 0.0521],
     [0.0056, 0.0291, 0.0768],
@@ -110,6 +129,9 @@ def stable_diffusion_step_callback(
         sdxl_latent_rgb_factors = torch.tensor(SDXL_LATENT_RGB_FACTORS, dtype=sample.dtype, device=sample.device)
         sdxl_smooth_matrix = torch.tensor(SDXL_SMOOTH_MATRIX, dtype=sample.dtype, device=sample.device)
         image = sample_to_lowres_estimated_image(sample, sdxl_latent_rgb_factors, sdxl_smooth_matrix)
+    elif base_model == BaseModelType.StableDiffusion3:
+        sd3_latent_rgb_factors = torch.tensor(SD3_5_LATENT_RGB_FACTORS, dtype=sample.dtype, device=sample.device)
+        image = sample_to_lowres_estimated_image(sample, sd3_latent_rgb_factors)
     else:
         v1_5_latent_rgb_factors = torch.tensor(SD1_5_LATENT_RGB_FACTORS, dtype=sample.dtype, device=sample.device)
         image = sample_to_lowres_estimated_image(sample, v1_5_latent_rgb_factors)

invokeai/backend/flux/custom_block_processor.py

@@ -1,9 +1,10 @@
 import einops
 import torch
 
+from invokeai.backend.flux.extensions.regional_prompting_extension import RegionalPromptingExtension
 from invokeai.backend.flux.extensions.xlabs_ip_adapter_extension import XLabsIPAdapterExtension
 from invokeai.backend.flux.math import attention
-from invokeai.backend.flux.modules.layers import DoubleStreamBlock
+from invokeai.backend.flux.modules.layers import DoubleStreamBlock, SingleStreamBlock
 
 
 class CustomDoubleStreamBlockProcessor:
@@ -13,7 +14,12 @@ class CustomDoubleStreamBlockProcessor:
 
     @staticmethod
     def _double_stream_block_forward(
-        block: DoubleStreamBlock, img: torch.Tensor, txt: torch.Tensor, vec: torch.Tensor, pe: torch.Tensor
+        block: DoubleStreamBlock,
+        img: torch.Tensor,
+        txt: torch.Tensor,
+        vec: torch.Tensor,
+        pe: torch.Tensor,
+        attn_mask: torch.Tensor | None = None,
     ) -> tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
         """This function is a direct copy of DoubleStreamBlock.forward(), but it returns some of the intermediate
         values.
@@ -40,7 +46,7 @@ class CustomDoubleStreamBlockProcessor:
         k = torch.cat((txt_k, img_k), dim=2)
         v = torch.cat((txt_v, img_v), dim=2)
 
-        attn = attention(q, k, v, pe=pe)
+        attn = attention(q, k, v, pe=pe, attn_mask=attn_mask)
         txt_attn, img_attn = attn[:, : txt.shape[1]], attn[:, txt.shape[1] :]
 
         # calculate the img bloks
@@ -63,11 +69,15 @@ class CustomDoubleStreamBlockProcessor:
         vec: torch.Tensor,
         pe: torch.Tensor,
         ip_adapter_extensions: list[XLabsIPAdapterExtension],
+        regional_prompting_extension: RegionalPromptingExtension,
     ) -> tuple[torch.Tensor, torch.Tensor]:
         """A custom implementation of DoubleStreamBlock.forward() with additional features:
         - IP-Adapter support
         """
-        img, txt, img_q = CustomDoubleStreamBlockProcessor._double_stream_block_forward(block, img, txt, vec, pe)
+        attn_mask = regional_prompting_extension.get_double_stream_attn_mask(block_index)
+        img, txt, img_q = CustomDoubleStreamBlockProcessor._double_stream_block_forward(
+            block, img, txt, vec, pe, attn_mask=attn_mask
+        )
 
         # Apply IP-Adapter conditioning.
         for ip_adapter_extension in ip_adapter_extensions:
@@ -81,3 +91,48 @@ class CustomDoubleStreamBlockProcessor:
             )
 
         return img, txt
+
+
+class CustomSingleStreamBlockProcessor:
+    """A class containing a custom implementation of SingleStreamBlock.forward() with additional features (masking,
+    etc.)
+    """
+
+    @staticmethod
+    def _single_stream_block_forward(
+        block: SingleStreamBlock,
+        x: torch.Tensor,
+        vec: torch.Tensor,
+        pe: torch.Tensor,
+        attn_mask: torch.Tensor | None = None,
+    ) -> torch.Tensor:
+        """This function is a direct copy of SingleStreamBlock.forward()."""
+        mod, _ = block.modulation(vec)
+        x_mod = (1 + mod.scale) * block.pre_norm(x) + mod.shift
+        qkv, mlp = torch.split(block.linear1(x_mod), [3 * block.hidden_size, block.mlp_hidden_dim], dim=-1)
+
+        q, k, v = einops.rearrange(qkv, "B L (K H D) -> K B H L D", K=3, H=block.num_heads)
+        q, k = block.norm(q, k, v)
+
+        # compute attention
+        attn = attention(q, k, v, pe=pe, attn_mask=attn_mask)
+        # compute activation in mlp stream, cat again and run second linear layer
+        output = block.linear2(torch.cat((attn, block.mlp_act(mlp)), 2))
+        return x + mod.gate * output
+
+    @staticmethod
+    def custom_single_block_forward(
+        timestep_index: int,
+        total_num_timesteps: int,
+        block_index: int,
+        block: SingleStreamBlock,
+        img: torch.Tensor,
+        vec: torch.Tensor,
+        pe: torch.Tensor,
+        regional_prompting_extension: RegionalPromptingExtension,
+    ) -> torch.Tensor:
+        """A custom implementation of SingleStreamBlock.forward() with additional features:
+        - Masking
+        """
+        attn_mask = regional_prompting_extension.get_single_stream_attn_mask(block_index)
+        return CustomSingleStreamBlockProcessor._single_stream_block_forward(block, img, vec, pe, attn_mask=attn_mask)

invokeai/backend/flux/denoise.py

@@ -7,6 +7,7 @@ from tqdm import tqdm
 from invokeai.backend.flux.controlnet.controlnet_flux_output import ControlNetFluxOutput, sum_controlnet_flux_outputs
 from invokeai.backend.flux.extensions.inpaint_extension import InpaintExtension
 from invokeai.backend.flux.extensions.instantx_controlnet_extension import InstantXControlNetExtension
+from invokeai.backend.flux.extensions.regional_prompting_extension import RegionalPromptingExtension
 from invokeai.backend.flux.extensions.xlabs_controlnet_extension import XLabsControlNetExtension
 from invokeai.backend.flux.extensions.xlabs_ip_adapter_extension import XLabsIPAdapterExtension
 from invokeai.backend.flux.model import Flux
@@ -18,14 +19,8 @@ def denoise(
     # model input
     img: torch.Tensor,
     img_ids: torch.Tensor,
-    # positive text conditioning
-    txt: torch.Tensor,
-    txt_ids: torch.Tensor,
-    vec: torch.Tensor,
-    # negative text conditioning
-    neg_txt: torch.Tensor | None,
-    neg_txt_ids: torch.Tensor | None,
-    neg_vec: torch.Tensor | None,
+    pos_regional_prompting_extension: RegionalPromptingExtension,
+    neg_regional_prompting_extension: RegionalPromptingExtension | None,
     # sampling parameters
     timesteps: list[float],
     step_callback: Callable[[PipelineIntermediateState], None],
@@ -35,6 +30,8 @@ def denoise(
     controlnet_extensions: list[XLabsControlNetExtension | InstantXControlNetExtension],
     pos_ip_adapter_extensions: list[XLabsIPAdapterExtension],
     neg_ip_adapter_extensions: list[XLabsIPAdapterExtension],
+    # extra img tokens
+    img_cond: torch.Tensor | None,
 ):
     # step 0 is the initial state
     total_steps = len(timesteps) - 1
@@ -61,9 +58,9 @@ def denoise(
                     total_num_timesteps=total_steps,
                     img=img,
                     img_ids=img_ids,
-                    txt=txt,
-                    txt_ids=txt_ids,
-                    y=vec,
+                    txt=pos_regional_prompting_extension.regional_text_conditioning.t5_embeddings,
+                    txt_ids=pos_regional_prompting_extension.regional_text_conditioning.t5_txt_ids,
+                    y=pos_regional_prompting_extension.regional_text_conditioning.clip_embeddings,
                     timesteps=t_vec,
                     guidance=guidance_vec,
                 )
@@ -74,13 +71,13 @@ def denoise(
         # controlnet_residuals datastructure is efficient in that it likely contains multiple references to the same
         # tensors. Calculating the sum materializes each tensor into its own instance.
         merged_controlnet_residuals = sum_controlnet_flux_outputs(controlnet_residuals)
-
+        pred_img = torch.cat((img, img_cond), dim=-1) if img_cond is not None else img
         pred = model(
-            img=img,
+            img=pred_img,
             img_ids=img_ids,
-            txt=txt,
-            txt_ids=txt_ids,
-            y=vec,
+            txt=pos_regional_prompting_extension.regional_text_conditioning.t5_embeddings,
+            txt_ids=pos_regional_prompting_extension.regional_text_conditioning.t5_txt_ids,
+            y=pos_regional_prompting_extension.regional_text_conditioning.clip_embeddings,
             timesteps=t_vec,
             guidance=guidance_vec,
             timestep_index=step_index,
@@ -88,6 +85,7 @@ def denoise(
             controlnet_double_block_residuals=merged_controlnet_residuals.double_block_residuals,
             controlnet_single_block_residuals=merged_controlnet_residuals.single_block_residuals,
             ip_adapter_extensions=pos_ip_adapter_extensions,
+            regional_prompting_extension=pos_regional_prompting_extension,
         )
 
         step_cfg_scale = cfg_scale[step_index]
@@ -97,15 +95,15 @@ def denoise(
             # TODO(ryand): Add option to run positive and negative predictions in a single batch for better performance
             # on systems with sufficient VRAM.
 
-            if neg_txt is None or neg_txt_ids is None or neg_vec is None:
+            if neg_regional_prompting_extension is None:
                 raise ValueError("Negative text conditioning is required when cfg_scale is not 1.0.")
 
             neg_pred = model(
                 img=img,
                 img_ids=img_ids,
-                txt=neg_txt,
-                txt_ids=neg_txt_ids,
-                y=neg_vec,
+                txt=neg_regional_prompting_extension.regional_text_conditioning.t5_embeddings,
+                txt_ids=neg_regional_prompting_extension.regional_text_conditioning.t5_txt_ids,
+                y=neg_regional_prompting_extension.regional_text_conditioning.clip_embeddings,
                 timesteps=t_vec,
                 guidance=guidance_vec,
                 timestep_index=step_index,
@@ -113,6 +111,7 @@ def denoise(
                 controlnet_double_block_residuals=None,
                 controlnet_single_block_residuals=None,
                 ip_adapter_extensions=neg_ip_adapter_extensions,
+                regional_prompting_extension=neg_regional_prompting_extension,
             )
             pred = neg_pred + step_cfg_scale * (pred - neg_pred)
 

invokeai/backend/flux/extensions/regional_prompting_extension.py

@@ -0,0 +1,276 @@
+from typing import Optional
+
+import torch
+import torchvision
+
+from invokeai.backend.flux.text_conditioning import FluxRegionalTextConditioning, FluxTextConditioning
+from invokeai.backend.stable_diffusion.diffusion.conditioning_data import Range
+from invokeai.backend.util.devices import TorchDevice
+from invokeai.backend.util.mask import to_standard_float_mask
+
+
+class RegionalPromptingExtension:
+    """A class for managing regional prompting with FLUX.
+
+    This implementation is inspired by https://arxiv.org/pdf/2411.02395 (though there are significant differences).
+    """
+
+    def __init__(
+        self,
+        regional_text_conditioning: FluxRegionalTextConditioning,
+        restricted_attn_mask: torch.Tensor | None = None,
+    ):
+        self.regional_text_conditioning = regional_text_conditioning
+        self.restricted_attn_mask = restricted_attn_mask
+
+    def get_double_stream_attn_mask(self, block_index: int) -> torch.Tensor | None:
+        order = [self.restricted_attn_mask, None]
+        return order[block_index % len(order)]
+
+    def get_single_stream_attn_mask(self, block_index: int) -> torch.Tensor | None:
+        order = [self.restricted_attn_mask, None]
+        return order[block_index % len(order)]
+
+    @classmethod
+    def from_text_conditioning(cls, text_conditioning: list[FluxTextConditioning], img_seq_len: int):
+        """Create a RegionalPromptingExtension from a list of text conditionings.
+
+        Args:
+            text_conditioning (list[FluxTextConditioning]): The text conditionings to use for regional prompting.
+            img_seq_len (int): The image sequence length (i.e. packed_height * packed_width).
+        """
+        regional_text_conditioning = cls._concat_regional_text_conditioning(text_conditioning)
+        attn_mask_with_restricted_img_self_attn = cls._prepare_restricted_attn_mask(
+            regional_text_conditioning, img_seq_len
+        )
+        return cls(
+            regional_text_conditioning=regional_text_conditioning,
+            restricted_attn_mask=attn_mask_with_restricted_img_self_attn,
+        )
+
+    # Keeping _prepare_unrestricted_attn_mask for reference as an alternative masking strategy:
+    #
+    # @classmethod
+    # def _prepare_unrestricted_attn_mask(
+    #     cls,
+    #     regional_text_conditioning: FluxRegionalTextConditioning,
+    #     img_seq_len: int,
+    # ) -> torch.Tensor:
+    #     """Prepare an 'unrestricted' attention mask. In this context, 'unrestricted' means that:
+    #     - img self-attention is not masked.
+    #     - img regions attend to both txt within their own region and to global prompts.
+    #     """
+    #     device = TorchDevice.choose_torch_device()
+
+    #     # Infer txt_seq_len from the t5_embeddings tensor.
+    #     txt_seq_len = regional_text_conditioning.t5_embeddings.shape[1]
+
+    #     # In the attention blocks, the txt seq and img seq are concatenated and then attention is applied.
+    #     # Concatenation happens in the following order: [txt_seq, img_seq].
+    #     # There are 4 portions of the attention mask to consider as we prepare it:
+    #     # 1. txt attends to itself
+    #     # 2. txt attends to corresponding regional img
+    #     # 3. regional img attends to corresponding txt
+    #     # 4. regional img attends to itself
+
+    #     # Initialize empty attention mask.
+    #     regional_attention_mask = torch.zeros(
+    #         (txt_seq_len + img_seq_len, txt_seq_len + img_seq_len), device=device, dtype=torch.float16
+    #     )
+
+    #     for image_mask, t5_embedding_range in zip(
+    #         regional_text_conditioning.image_masks, regional_text_conditioning.t5_embedding_ranges, strict=True
+    #     ):
+    #         # 1. txt attends to itself
+    #         regional_attention_mask[
+    #             t5_embedding_range.start : t5_embedding_range.end, t5_embedding_range.start : t5_embedding_range.end
+    #         ] = 1.0
+
+    #         # 2. txt attends to corresponding regional img
+    #         # Note that we reshape to (1, img_seq_len) to ensure broadcasting works as desired.
+    #         fill_value = image_mask.view(1, img_seq_len) if image_mask is not None else 1.0
+    #         regional_attention_mask[t5_embedding_range.start : t5_embedding_range.end, txt_seq_len:] = fill_value
+
+    #         # 3. regional img attends to corresponding txt
+    #         # Note that we reshape to (img_seq_len, 1) to ensure broadcasting works as desired.
+    #         fill_value = image_mask.view(img_seq_len, 1) if image_mask is not None else 1.0
+    #         regional_attention_mask[txt_seq_len:, t5_embedding_range.start : t5_embedding_range.end] = fill_value
+
+    #     # 4. regional img attends to itself
+    #     # Allow unrestricted img self attention.
+    #     regional_attention_mask[txt_seq_len:, txt_seq_len:] = 1.0
+
+    #     # Convert attention mask to boolean.
+    #     regional_attention_mask = regional_attention_mask > 0.5
+
+    #     return regional_attention_mask
+
+    @classmethod
+    def _prepare_restricted_attn_mask(
+        cls,
+        regional_text_conditioning: FluxRegionalTextConditioning,
+        img_seq_len: int,
+    ) -> torch.Tensor | None:
+        """Prepare a 'restricted' attention mask. In this context, 'restricted' means that:
+        - img self-attention is only allowed within regions.
+        - img regions only attend to txt within their own region, not to global prompts.
+        """
+        # Identify background region. I.e. the region that is not covered by any region masks.
+        background_region_mask: None | torch.Tensor = None
+        for image_mask in regional_text_conditioning.image_masks:
+            if image_mask is not None:
+                if background_region_mask is None:
+                    background_region_mask = torch.ones_like(image_mask)
+                background_region_mask *= 1 - image_mask
+
+        if background_region_mask is None:
+            # There are no region masks, short-circuit and return None.
+            # TODO(ryand): We could restrict txt-txt attention across multiple global prompts, but this would
+            # is a rare use case and would make the logic here significantly more complicated.
+            return None
+
+        device = TorchDevice.choose_torch_device()
+
+        # Infer txt_seq_len from the t5_embeddings tensor.
+        txt_seq_len = regional_text_conditioning.t5_embeddings.shape[1]
+
+        # In the attention blocks, the txt seq and img seq are concatenated and then attention is applied.
+        # Concatenation happens in the following order: [txt_seq, img_seq].
+        # There are 4 portions of the attention mask to consider as we prepare it:
+        # 1. txt attends to itself
+        # 2. txt attends to corresponding regional img
+        # 3. regional img attends to corresponding txt
+        # 4. regional img attends to itself
+
+        # Initialize empty attention mask.
+        regional_attention_mask = torch.zeros(
+            (txt_seq_len + img_seq_len, txt_seq_len + img_seq_len), device=device, dtype=torch.float16
+        )
+
+        for image_mask, t5_embedding_range in zip(
+            regional_text_conditioning.image_masks, regional_text_conditioning.t5_embedding_ranges, strict=True
+        ):
+            # 1. txt attends to itself
+            regional_attention_mask[
+                t5_embedding_range.start : t5_embedding_range.end, t5_embedding_range.start : t5_embedding_range.end
+            ] = 1.0
+
+            if image_mask is not None:
+                # 2. txt attends to corresponding regional img
+                # Note that we reshape to (1, img_seq_len) to ensure broadcasting works as desired.
+                regional_attention_mask[t5_embedding_range.start : t5_embedding_range.end, txt_seq_len:] = (
+                    image_mask.view(1, img_seq_len)
+                )
+
+                # 3. regional img attends to corresponding txt
+                # Note that we reshape to (img_seq_len, 1) to ensure broadcasting works as desired.
+                regional_attention_mask[txt_seq_len:, t5_embedding_range.start : t5_embedding_range.end] = (
+                    image_mask.view(img_seq_len, 1)
+                )
+
+                # 4. regional img attends to itself
+                image_mask = image_mask.view(img_seq_len, 1)
+                regional_attention_mask[txt_seq_len:, txt_seq_len:] += image_mask @ image_mask.T
+            else:
+                # We don't allow attention between non-background image regions and global prompts. This helps to ensure
+                # that regions focus on their local prompts. We do, however, allow attention between background regions
+                # and global prompts. If we didn't do this, then the background regions would not attend to any txt
+                # embeddings, which we found experimentally to cause artifacts.
+
+                # 2. global txt attends to background region
+                # Note that we reshape to (1, img_seq_len) to ensure broadcasting works as desired.
+                regional_attention_mask[t5_embedding_range.start : t5_embedding_range.end, txt_seq_len:] = (
+                    background_region_mask.view(1, img_seq_len)
+                )
+
+                # 3. background region attends to global txt
+                # Note that we reshape to (img_seq_len, 1) to ensure broadcasting works as desired.
+                regional_attention_mask[txt_seq_len:, t5_embedding_range.start : t5_embedding_range.end] = (
+                    background_region_mask.view(img_seq_len, 1)
+                )
+
+        # Allow background regions to attend to themselves.
+        regional_attention_mask[txt_seq_len:, txt_seq_len:] += background_region_mask.view(img_seq_len, 1)
+        regional_attention_mask[txt_seq_len:, txt_seq_len:] += background_region_mask.view(1, img_seq_len)
+
+        # Convert attention mask to boolean.
+        regional_attention_mask = regional_attention_mask > 0.5
+
+        return regional_attention_mask
+
+    @classmethod
+    def _concat_regional_text_conditioning(
+        cls,
+        text_conditionings: list[FluxTextConditioning],
+    ) -> FluxRegionalTextConditioning:
+        """Concatenate regional text conditioning data into a single conditioning tensor (with associated masks)."""
+        concat_t5_embeddings: list[torch.Tensor] = []
+        concat_t5_embedding_ranges: list[Range] = []
+        image_masks: list[torch.Tensor | None] = []
+
+        # Choose global CLIP embedding.
+        # Use the first global prompt's CLIP embedding as the global CLIP embedding. If there is no global prompt, use
+        # the first prompt's CLIP embedding.
+        global_clip_embedding: torch.Tensor = text_conditionings[0].clip_embeddings
+        for text_conditioning in text_conditionings:
+            if text_conditioning.mask is None:
+                global_clip_embedding = text_conditioning.clip_embeddings
+                break
+
+        cur_t5_embedding_len = 0
+        for text_conditioning in text_conditionings:
+            concat_t5_embeddings.append(text_conditioning.t5_embeddings)
+
+            concat_t5_embedding_ranges.append(
+                Range(start=cur_t5_embedding_len, end=cur_t5_embedding_len + text_conditioning.t5_embeddings.shape[1])
+            )
+
+            image_masks.append(text_conditioning.mask)
+
+            cur_t5_embedding_len += text_conditioning.t5_embeddings.shape[1]
+
+        t5_embeddings = torch.cat(concat_t5_embeddings, dim=1)
+
+        # Initialize the txt_ids tensor.
+        pos_bs, pos_t5_seq_len, _ = t5_embeddings.shape
+        t5_txt_ids = torch.zeros(
+            pos_bs, pos_t5_seq_len, 3, dtype=t5_embeddings.dtype, device=TorchDevice.choose_torch_device()
+        )
+
+        return FluxRegionalTextConditioning(
+            t5_embeddings=t5_embeddings,
+            clip_embeddings=global_clip_embedding,
+            t5_txt_ids=t5_txt_ids,
+            image_masks=image_masks,
+            t5_embedding_ranges=concat_t5_embedding_ranges,
+        )
+
+    @staticmethod
+    def preprocess_regional_prompt_mask(
+        mask: Optional[torch.Tensor], packed_height: int, packed_width: int, dtype: torch.dtype, device: torch.device
+    ) -> torch.Tensor:
+        """Preprocess a regional prompt mask to match the target height and width.
+        If mask is None, returns a mask of all ones with the target height and width.
+        If mask is not None, resizes the mask to the target height and width using 'nearest' interpolation.
+
+        packed_height and packed_width are the target height and width of the mask in the 'packed' latent space.
+
+        Returns:
+            torch.Tensor: The processed mask. shape: (1, 1, packed_height * packed_width).
+        """
+
+        if mask is None:
+            return torch.ones((1, 1, packed_height * packed_width), dtype=dtype, device=device)
+
+        mask = to_standard_float_mask(mask, out_dtype=dtype)
+
+        tf = torchvision.transforms.Resize(
+            (packed_height, packed_width), interpolation=torchvision.transforms.InterpolationMode.NEAREST
+        )
+
+        # Add a batch dimension to the mask, because torchvision expects shape (batch, channels, h, w).
+        mask = mask.unsqueeze(0)  # Shape: (1, h, w) -> (1, 1, h, w)
+        resized_mask = tf(mask)
+
+        # Flatten the height and width dimensions into a single image_seq_len dimension.
+        return resized_mask.flatten(start_dim=2)

invokeai/backend/flux/ip_adapter/state_dict_utils.py

@@ -41,10 +41,12 @@ def infer_xlabs_ip_adapter_params_from_state_dict(state_dict: dict[str, torch.Te
     hidden_dim = state_dict["double_blocks.0.processor.ip_adapter_double_stream_k_proj.weight"].shape[0]
     context_dim = state_dict["double_blocks.0.processor.ip_adapter_double_stream_k_proj.weight"].shape[1]
     clip_embeddings_dim = state_dict["ip_adapter_proj_model.proj.weight"].shape[1]
+    clip_extra_context_tokens = state_dict["ip_adapter_proj_model.proj.weight"].shape[0] // context_dim
 
     return XlabsIpAdapterParams(
         num_double_blocks=num_double_blocks,
         context_dim=context_dim,
         hidden_dim=hidden_dim,
         clip_embeddings_dim=clip_embeddings_dim,
+        clip_extra_context_tokens=clip_extra_context_tokens,
     )

invokeai/backend/flux/ip_adapter/xlabs_ip_adapter_flux.py

@@ -31,13 +31,16 @@ class XlabsIpAdapterParams:
     hidden_dim: int
 
     clip_embeddings_dim: int
+    clip_extra_context_tokens: int
 
 
 class XlabsIpAdapterFlux(torch.nn.Module):
     def __init__(self, params: XlabsIpAdapterParams):
         super().__init__()
         self.image_proj = ImageProjModel(
-            cross_attention_dim=params.context_dim, clip_embeddings_dim=params.clip_embeddings_dim
+            cross_attention_dim=params.context_dim,
+            clip_embeddings_dim=params.clip_embeddings_dim,
+            clip_extra_context_tokens=params.clip_extra_context_tokens,
         )
         self.ip_adapter_double_blocks = IPAdapterDoubleBlocks(
             num_double_blocks=params.num_double_blocks, context_dim=params.context_dim, hidden_dim=params.hidden_dim

invokeai/backend/flux/math.py

@@ -5,10 +5,10 @@ from einops import rearrange
 from torch import Tensor
 
 
-def attention(q: Tensor, k: Tensor, v: Tensor, pe: Tensor) -> Tensor:
+def attention(q: Tensor, k: Tensor, v: Tensor, pe: Tensor, attn_mask: Tensor | None = None) -> Tensor:
     q, k = apply_rope(q, k, pe)
 
-    x = torch.nn.functional.scaled_dot_product_attention(q, k, v)
+    x = torch.nn.functional.scaled_dot_product_attention(q, k, v, attn_mask=attn_mask)
     x = rearrange(x, "B H L D -> B L (H D)")
 
     return x
@@ -24,12 +24,12 @@ def rope(pos: Tensor, dim: int, theta: int) -> Tensor:
     out = torch.einsum("...n,d->...nd", pos, omega)
     out = torch.stack([torch.cos(out), -torch.sin(out), torch.sin(out), torch.cos(out)], dim=-1)
     out = rearrange(out, "b n d (i j) -> b n d i j", i=2, j=2)
-    return out.float()
+    return out.to(dtype=pos.dtype, device=pos.device)
 
 
 def apply_rope(xq: Tensor, xk: Tensor, freqs_cis: Tensor) -> tuple[Tensor, Tensor]:
-    xq_ = xq.float().reshape(*xq.shape[:-1], -1, 1, 2)
-    xk_ = xk.float().reshape(*xk.shape[:-1], -1, 1, 2)
+    xq_ = xq.view(*xq.shape[:-1], -1, 1, 2)
+    xk_ = xk.view(*xk.shape[:-1], -1, 1, 2)
     xq_out = freqs_cis[..., 0] * xq_[..., 0] + freqs_cis[..., 1] * xq_[..., 1]
     xk_out = freqs_cis[..., 0] * xk_[..., 0] + freqs_cis[..., 1] * xk_[..., 1]
-    return xq_out.reshape(*xq.shape).type_as(xq), xk_out.reshape(*xk.shape).type_as(xk)
+    return xq_out.view(*xq.shape).type_as(xq), xk_out.view(*xk.shape).type_as(xk)

invokeai/backend/flux/model.py

@@ -1,11 +1,16 @@
 # Initially pulled from https://github.com/black-forest-labs/flux
 
 from dataclasses import dataclass
+from typing import Optional
 
 import torch
 from torch import Tensor, nn
 
-from invokeai.backend.flux.custom_block_processor import CustomDoubleStreamBlockProcessor
+from invokeai.backend.flux.custom_block_processor import (
+    CustomDoubleStreamBlockProcessor,
+    CustomSingleStreamBlockProcessor,
+)
+from invokeai.backend.flux.extensions.regional_prompting_extension import RegionalPromptingExtension
 from invokeai.backend.flux.extensions.xlabs_ip_adapter_extension import XLabsIPAdapterExtension
 from invokeai.backend.flux.modules.layers import (
     DoubleStreamBlock,
@@ -31,6 +36,7 @@ class FluxParams:
     theta: int
     qkv_bias: bool
     guidance_embed: bool
+    out_channels: Optional[int] = None
 
 
 class Flux(nn.Module):
@@ -43,7 +49,7 @@ class Flux(nn.Module):
 
         self.params = params
         self.in_channels = params.in_channels
-        self.out_channels = self.in_channels
+        self.out_channels = params.out_channels or self.in_channels
         if params.hidden_size % params.num_heads != 0:
             raise ValueError(f"Hidden size {params.hidden_size} must be divisible by num_heads {params.num_heads}")
         pe_dim = params.hidden_size // params.num_heads
@@ -95,6 +101,7 @@ class Flux(nn.Module):
         controlnet_double_block_residuals: list[Tensor] | None,
         controlnet_single_block_residuals: list[Tensor] | None,
         ip_adapter_extensions: list[XLabsIPAdapterExtension],
+        regional_prompting_extension: RegionalPromptingExtension,
     ) -> Tensor:
         if img.ndim != 3 or txt.ndim != 3:
             raise ValueError("Input img and txt tensors must have 3 dimensions.")
@@ -117,7 +124,6 @@ class Flux(nn.Module):
             assert len(controlnet_double_block_residuals) == len(self.double_blocks)
         for block_index, block in enumerate(self.double_blocks):
             assert isinstance(block, DoubleStreamBlock)
-
             img, txt = CustomDoubleStreamBlockProcessor.custom_double_block_forward(
                 timestep_index=timestep_index,
                 total_num_timesteps=total_num_timesteps,
@@ -128,6 +134,7 @@ class Flux(nn.Module):
                 vec=vec,
                 pe=pe,
                 ip_adapter_extensions=ip_adapter_extensions,
+                regional_prompting_extension=regional_prompting_extension,
             )
 
             if controlnet_double_block_residuals is not None:
@@ -140,7 +147,17 @@ class Flux(nn.Module):
             assert len(controlnet_single_block_residuals) == len(self.single_blocks)
 
         for block_index, block in enumerate(self.single_blocks):
-            img = block(img, vec=vec, pe=pe)
+            assert isinstance(block, SingleStreamBlock)
+            img = CustomSingleStreamBlockProcessor.custom_single_block_forward(
+                timestep_index=timestep_index,
+                total_num_timesteps=total_num_timesteps,
+                block_index=block_index,
+                block=block,
+                img=img,
+                vec=vec,
+                pe=pe,
+                regional_prompting_extension=regional_prompting_extension,
+            )
 
             if controlnet_single_block_residuals is not None:
                 img[:, txt.shape[1] :, ...] += controlnet_single_block_residuals[block_index]

invokeai/backend/flux/modules/layers.py

@@ -66,10 +66,7 @@ class RMSNorm(torch.nn.Module):
         self.scale = nn.Parameter(torch.ones(dim))
 
     def forward(self, x: Tensor):
-        x_dtype = x.dtype
-        x = x.float()
-        rrms = torch.rsqrt(torch.mean(x**2, dim=-1, keepdim=True) + 1e-6)
-        return (x * rrms).to(dtype=x_dtype) * self.scale
+        return torch.nn.functional.rms_norm(x, self.scale.shape, self.scale, eps=1e-6)
 
 
 class QKNorm(torch.nn.Module):

invokeai/backend/flux/text_conditioning.py

@@ -0,0 +1,36 @@
+from dataclasses import dataclass
+
+import torch
+
+from invokeai.backend.stable_diffusion.diffusion.conditioning_data import Range
+
+
+@dataclass
+class FluxTextConditioning:
+    t5_embeddings: torch.Tensor
+    clip_embeddings: torch.Tensor
+    # If mask is None, the prompt is a global prompt.
+    mask: torch.Tensor | None
+
+
+@dataclass
+class FluxRegionalTextConditioning:
+    # Concatenated text embeddings.
+    # Shape: (1, concatenated_txt_seq_len, 4096)
+    t5_embeddings: torch.Tensor
+    # Shape: (1, concatenated_txt_seq_len, 3)
+    t5_txt_ids: torch.Tensor
+
+    # Global CLIP embeddings.
+    # Shape: (1, 768)
+    clip_embeddings: torch.Tensor
+
+    # A binary mask indicating the regions of the image that the prompt should be applied to. If None, the prompt is a
+    # global prompt.
+    # image_masks[i] is the mask for the ith prompt.
+    # image_masks[i] has shape (1, image_seq_len) and dtype torch.bool.
+    image_masks: list[torch.Tensor | None]
+
+    # List of ranges that represent the embedding ranges for each mask.
+    # t5_embedding_ranges[i] contains the range of the t5 embeddings that correspond to image_masks[i].
+    t5_embedding_ranges: list[Range]

invokeai/backend/lora/layers/any_lora_layer.py

@@ -1,11 +0,0 @@
-from typing import Union
-
-from invokeai.backend.lora.layers.concatenated_lora_layer import ConcatenatedLoRALayer
-from invokeai.backend.lora.layers.full_layer import FullLayer
-from invokeai.backend.lora.layers.ia3_layer import IA3Layer
-from invokeai.backend.lora.layers.loha_layer import LoHALayer
-from invokeai.backend.lora.layers.lokr_layer import LoKRLayer
-from invokeai.backend.lora.layers.lora_layer import LoRALayer
-from invokeai.backend.lora.layers.norm_layer import NormLayer
-
-AnyLoRALayer = Union[LoRALayer, LoHALayer, LoKRLayer, FullLayer, IA3Layer, NormLayer, ConcatenatedLoRALayer]

invokeai/backend/lora/sidecar_layers/concatenated_lora/concatenated_lora_linear_sidecar_layer.py

@@ -1,34 +0,0 @@
-import torch
-
-from invokeai.backend.lora.layers.concatenated_lora_layer import ConcatenatedLoRALayer
-
-
-class ConcatenatedLoRALinearSidecarLayer(torch.nn.Module):
-    def __init__(
-        self,
-        concatenated_lora_layer: ConcatenatedLoRALayer,
-        weight: float,
-    ):
-        super().__init__()
-
-        self._concatenated_lora_layer = concatenated_lora_layer
-        self._weight = weight
-
-    def forward(self, input: torch.Tensor) -> torch.Tensor:
-        x_chunks: list[torch.Tensor] = []
-        for lora_layer in self._concatenated_lora_layer.lora_layers:
-            x_chunk = torch.nn.functional.linear(input, lora_layer.down)
-            if lora_layer.mid is not None:
-                x_chunk = torch.nn.functional.linear(x_chunk, lora_layer.mid)
-            x_chunk = torch.nn.functional.linear(x_chunk, lora_layer.up, bias=lora_layer.bias)
-            x_chunk *= self._weight * lora_layer.scale()
-            x_chunks.append(x_chunk)
-
-        # TODO(ryand): Generalize to support concat_axis != 0.
-        assert self._concatenated_lora_layer.concat_axis == 0
-        x = torch.cat(x_chunks, dim=-1)
-        return x
-
-    def to(self, device: torch.device | None = None, dtype: torch.dtype | None = None):
-        self._concatenated_lora_layer.to(device=device, dtype=dtype)
-        return self

invokeai/backend/lora/sidecar_layers/lora/lora_linear_sidecar_layer.py

@@ -1,27 +0,0 @@
-import torch
-
-from invokeai.backend.lora.layers.lora_layer import LoRALayer
-
-
-class LoRALinearSidecarLayer(torch.nn.Module):
-    def __init__(
-        self,
-        lora_layer: LoRALayer,
-        weight: float,
-    ):
-        super().__init__()
-
-        self._lora_layer = lora_layer
-        self._weight = weight
-
-    def forward(self, x: torch.Tensor) -> torch.Tensor:
-        x = torch.nn.functional.linear(x, self._lora_layer.down)
-        if self._lora_layer.mid is not None:
-            x = torch.nn.functional.linear(x, self._lora_layer.mid)
-        x = torch.nn.functional.linear(x, self._lora_layer.up, bias=self._lora_layer.bias)
-        x *= self._weight * self._lora_layer.scale()
-        return x
-
-    def to(self, device: torch.device | None = None, dtype: torch.dtype | None = None):
-        self._lora_layer.to(device=device, dtype=dtype)
-        return self

invokeai/backend/lora/sidecar_layers/lora_sidecar_module.py

@@ -1,24 +0,0 @@
-import torch
-
-
-class LoRASidecarModule(torch.nn.Module):
-    """A LoRA sidecar module that wraps an original module and adds LoRA layers to it."""
-
-    def __init__(self, orig_module: torch.nn.Module, lora_layers: list[torch.nn.Module]):
-        super().__init__()
-        self.orig_module = orig_module
-        self._lora_layers = lora_layers
-
-    def add_lora_layer(self, lora_layer: torch.nn.Module):
-        self._lora_layers.append(lora_layer)
-
-    def forward(self, input: torch.Tensor) -> torch.Tensor:
-        x = self.orig_module(input)
-        for lora_layer in self._lora_layers:
-            x += lora_layer(input)
-        return x
-
-    def to(self, device: torch.device | None = None, dtype: torch.dtype | None = None):
-        self._orig_module.to(device=device, dtype=dtype)
-        for lora_layer in self._lora_layers:
-            lora_layer.to(device=device, dtype=dtype)

invokeai/backend/model_manager/config.py

@@ -53,6 +53,7 @@ class BaseModelType(str, Enum):
     Any = "any"
     StableDiffusion1 = "sd-1"
     StableDiffusion2 = "sd-2"
+    StableDiffusion3 = "sd-3"
     StableDiffusionXL = "sdxl"
     StableDiffusionXLRefiner = "sdxl-refiner"
     Flux = "flux"
@@ -66,6 +67,7 @@ class ModelType(str, Enum):
     Main = "main"
     VAE = "vae"
     LoRA = "lora"
+    ControlLoRa = "control_lora"
     ControlNet = "controlnet"  # used by model_probe
     TextualInversion = "embedding"
     IPAdapter = "ip_adapter"
@@ -83,8 +85,10 @@ class SubModelType(str, Enum):
     Transformer = "transformer"
     TextEncoder = "text_encoder"
     TextEncoder2 = "text_encoder_2"
+    TextEncoder3 = "text_encoder_3"
     Tokenizer = "tokenizer"
     Tokenizer2 = "tokenizer_2"
+    Tokenizer3 = "tokenizer_3"
     VAE = "vae"
     VAEDecoder = "vae_decoder"
     VAEEncoder = "vae_encoder"
@@ -92,6 +96,13 @@ class SubModelType(str, Enum):
     SafetyChecker = "safety_checker"
 
 
+class ClipVariantType(str, Enum):
+    """Variant type."""
+
+    L = "large"
+    G = "gigantic"
+
+
 class ModelVariantType(str, Enum):
     """Variant type."""
 
@@ -147,6 +158,17 @@ class ModelSourceType(str, Enum):
 DEFAULTS_PRECISION = Literal["fp16", "fp32"]
 
 
+AnyVariant: TypeAlias = Union[ModelVariantType, ClipVariantType, None]
+
+
+class SubmodelDefinition(BaseModel):
+    path_or_prefix: str
+    model_type: ModelType
+    variant: AnyVariant = None
+
+    model_config = ConfigDict(protected_namespaces=())
+
+
 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")
@@ -193,6 +215,9 @@ class ModelConfigBase(BaseModel):
         schema["required"].extend(["key", "type", "format"])
 
     model_config = ConfigDict(validate_assignment=True, json_schema_extra=json_schema_extra)
+    submodels: Optional[Dict[SubModelType, SubmodelDefinition]] = Field(
+        description="Loadable submodels in this model", default=None
+    )
 
 
 class CheckpointConfigBase(ModelConfigBase):
@@ -249,6 +274,36 @@ class LoRALyCORISConfig(LoRAConfigBase):
         return Tag(f"{ModelType.LoRA.value}.{ModelFormat.LyCORIS.value}")
 
 
+class ControlAdapterConfigBase(BaseModel):
+    default_settings: Optional[ControlAdapterDefaultSettings] = Field(
+        description="Default settings for this model", default=None
+    )
+
+
+class ControlLoRALyCORISConfig(ModelConfigBase, ControlAdapterConfigBase):
+    """Model config for Control LoRA models."""
+
+    type: Literal[ModelType.ControlLoRa] = ModelType.ControlLoRa
+    trigger_phrases: Optional[set[str]] = Field(description="Set of trigger phrases for this model", default=None)
+    format: Literal[ModelFormat.LyCORIS] = ModelFormat.LyCORIS
+
+    @staticmethod
+    def get_tag() -> Tag:
+        return Tag(f"{ModelType.ControlLoRa.value}.{ModelFormat.LyCORIS.value}")
+
+
+class ControlLoRADiffusersConfig(ModelConfigBase, ControlAdapterConfigBase):
+    """Model config for Control LoRA models."""
+
+    type: Literal[ModelType.ControlLoRa] = ModelType.ControlLoRa
+    trigger_phrases: Optional[set[str]] = Field(description="Set of trigger phrases for this model", default=None)
+    format: Literal[ModelFormat.Diffusers] = ModelFormat.Diffusers
+
+    @staticmethod
+    def get_tag() -> Tag:
+        return Tag(f"{ModelType.ControlLoRa.value}.{ModelFormat.Diffusers.value}")
+
+
 class LoRADiffusersConfig(LoRAConfigBase):
     """Model config for LoRA/Diffusers models."""
 
@@ -280,12 +335,6 @@ class VAEDiffusersConfig(ModelConfigBase):
         return Tag(f"{ModelType.VAE.value}.{ModelFormat.Diffusers.value}")
 
 
-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)."""
 
@@ -335,7 +384,7 @@ class MainConfigBase(ModelConfigBase):
     default_settings: Optional[MainModelDefaultSettings] = Field(
         description="Default settings for this model", default=None
     )
-    variant: ModelVariantType = ModelVariantType.Normal
+    variant: AnyVariant = ModelVariantType.Normal
 
 
 class MainCheckpointConfig(CheckpointConfigBase, MainConfigBase):
@@ -419,12 +468,33 @@ class CLIPEmbedDiffusersConfig(DiffusersConfigBase):
 
     type: Literal[ModelType.CLIPEmbed] = ModelType.CLIPEmbed
     format: Literal[ModelFormat.Diffusers] = ModelFormat.Diffusers
+    variant: ClipVariantType = ClipVariantType.L
 
     @staticmethod
     def get_tag() -> Tag:
         return Tag(f"{ModelType.CLIPEmbed.value}.{ModelFormat.Diffusers.value}")
 
 
+class CLIPGEmbedDiffusersConfig(CLIPEmbedDiffusersConfig):
+    """Model config for CLIP-G Embeddings."""
+
+    variant: ClipVariantType = ClipVariantType.G
+
+    @staticmethod
+    def get_tag() -> Tag:
+        return Tag(f"{ModelType.CLIPEmbed.value}.{ModelFormat.Diffusers.value}.{ClipVariantType.G}")
+
+
+class CLIPLEmbedDiffusersConfig(CLIPEmbedDiffusersConfig):
+    """Model config for CLIP-L Embeddings."""
+
+    variant: ClipVariantType = ClipVariantType.L
+
+    @staticmethod
+    def get_tag() -> Tag:
+        return Tag(f"{ModelType.CLIPEmbed.value}.{ModelFormat.Diffusers.value}.{ClipVariantType.L}")
+
+
 class CLIPVisionDiffusersConfig(DiffusersConfigBase):
     """Model config for CLIPVision."""
 
@@ -490,6 +560,8 @@ AnyModelConfig = Annotated[
         Annotated[ControlNetDiffusersConfig, ControlNetDiffusersConfig.get_tag()],
         Annotated[ControlNetCheckpointConfig, ControlNetCheckpointConfig.get_tag()],
         Annotated[LoRALyCORISConfig, LoRALyCORISConfig.get_tag()],
+        Annotated[ControlLoRALyCORISConfig, ControlLoRALyCORISConfig.get_tag()],
+        Annotated[ControlLoRADiffusersConfig, ControlLoRADiffusersConfig.get_tag()],
         Annotated[LoRADiffusersConfig, LoRADiffusersConfig.get_tag()],
         Annotated[T5EncoderConfig, T5EncoderConfig.get_tag()],
         Annotated[T5EncoderBnbQuantizedLlmInt8bConfig, T5EncoderBnbQuantizedLlmInt8bConfig.get_tag()],
@@ -501,6 +573,8 @@ AnyModelConfig = Annotated[
         Annotated[SpandrelImageToImageConfig, SpandrelImageToImageConfig.get_tag()],
         Annotated[CLIPVisionDiffusersConfig, CLIPVisionDiffusersConfig.get_tag()],
         Annotated[CLIPEmbedDiffusersConfig, CLIPEmbedDiffusersConfig.get_tag()],
+        Annotated[CLIPLEmbedDiffusersConfig, CLIPLEmbedDiffusersConfig.get_tag()],
+        Annotated[CLIPGEmbedDiffusersConfig, CLIPGEmbedDiffusersConfig.get_tag()],
     ],
     Discriminator(get_model_discriminator_value),
 ]

invokeai/backend/model_manager/load/load_default.py

@@ -35,6 +35,7 @@ class ModelLoader(ModelLoaderBase):
         self._logger = logger
         self._ram_cache = ram_cache
         self._torch_dtype = TorchDevice.choose_torch_dtype()
+        self._torch_device = TorchDevice.choose_torch_device()
 
     def load_model(self, model_config: AnyModelConfig, submodel_type: Optional[SubModelType] = None) -> LoadedModel:
         """

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

@@ -84,7 +84,15 @@ class FluxVAELoader(ModelLoader):
             model = AutoEncoder(ae_params[config.config_path])
             sd = load_file(model_path)
             model.load_state_dict(sd, assign=True)
-            model.to(dtype=self._torch_dtype)
+            # VAE is broken in float16, which mps defaults to
+            if self._torch_dtype == torch.float16:
+                try:
+                    vae_dtype = torch.tensor([1.0], dtype=torch.bfloat16, device=self._torch_device).dtype
+                except TypeError:
+                    vae_dtype = torch.float32
+            else:
+                vae_dtype = self._torch_dtype
+            model.to(vae_dtype)
 
         return model
 
@@ -128,9 +136,9 @@ class BnbQuantizedLlmInt8bCheckpointModel(ModelLoader):
                 "The bnb modules are not available. Please install bitsandbytes if available on your platform."
             )
         match submodel_type:
-            case SubModelType.Tokenizer2:
+            case SubModelType.Tokenizer2 | SubModelType.Tokenizer3:
                 return T5Tokenizer.from_pretrained(Path(config.path) / "tokenizer_2", max_length=512)
-            case SubModelType.TextEncoder2:
+            case SubModelType.TextEncoder2 | SubModelType.TextEncoder3:
                 te2_model_path = Path(config.path) / "text_encoder_2"
                 model_config = AutoConfig.from_pretrained(te2_model_path)
                 with accelerate.init_empty_weights():
@@ -172,9 +180,9 @@ class T5EncoderCheckpointModel(ModelLoader):
             raise ValueError("Only T5EncoderConfig models are currently supported here.")
 
         match submodel_type:
-            case SubModelType.Tokenizer2:
+            case SubModelType.Tokenizer2 | SubModelType.Tokenizer3:
                 return T5Tokenizer.from_pretrained(Path(config.path) / "tokenizer_2", max_length=512)
-            case SubModelType.TextEncoder2:
+            case SubModelType.TextEncoder2 | SubModelType.TextEncoder3:
                 return T5EncoderModel.from_pretrained(Path(config.path) / "text_encoder_2", torch_dtype="auto")
 
         raise ValueError(

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

@@ -9,14 +9,6 @@ import torch
 from safetensors.torch import load_file
 
 from invokeai.app.services.config import InvokeAIAppConfig
-from invokeai.backend.lora.conversions.flux_diffusers_lora_conversion_utils import (
-    lora_model_from_flux_diffusers_state_dict,
-)
-from invokeai.backend.lora.conversions.flux_kohya_lora_conversion_utils import (
-    lora_model_from_flux_kohya_state_dict,
-)
-from invokeai.backend.lora.conversions.sd_lora_conversion_utils import lora_model_from_sd_state_dict
-from invokeai.backend.lora.conversions.sdxl_lora_conversion_utils import convert_sdxl_keys_to_diffusers_format
 from invokeai.backend.model_manager import (
     AnyModel,
     AnyModelConfig,
@@ -28,10 +20,25 @@ from invokeai.backend.model_manager import (
 from invokeai.backend.model_manager.load.load_default import ModelLoader
 from invokeai.backend.model_manager.load.model_cache.model_cache_base import ModelCacheBase
 from invokeai.backend.model_manager.load.model_loader_registry import ModelLoaderRegistry
+from invokeai.backend.patches.lora_conversions.flux_control_lora_utils import (
+    is_state_dict_likely_flux_control,
+    lora_model_from_flux_control_state_dict,
+)
+from invokeai.backend.patches.lora_conversions.flux_diffusers_lora_conversion_utils import (
+    lora_model_from_flux_diffusers_state_dict,
+)
+from invokeai.backend.patches.lora_conversions.flux_kohya_lora_conversion_utils import (
+    is_state_dict_likely_in_flux_kohya_format,
+    lora_model_from_flux_kohya_state_dict,
+)
+from invokeai.backend.patches.lora_conversions.sd_lora_conversion_utils import lora_model_from_sd_state_dict
+from invokeai.backend.patches.lora_conversions.sdxl_lora_conversion_utils import convert_sdxl_keys_to_diffusers_format
 
 
 @ModelLoaderRegistry.register(base=BaseModelType.Any, type=ModelType.LoRA, format=ModelFormat.Diffusers)
 @ModelLoaderRegistry.register(base=BaseModelType.Any, type=ModelType.LoRA, format=ModelFormat.LyCORIS)
+@ModelLoaderRegistry.register(base=BaseModelType.Flux, type=ModelType.ControlLoRa, format=ModelFormat.LyCORIS)
+@ModelLoaderRegistry.register(base=BaseModelType.Flux, type=ModelType.ControlLoRa, format=ModelFormat.Diffusers)
 class LoRALoader(ModelLoader):
     """Class to load LoRA models."""
 
@@ -75,7 +82,10 @@ class LoRALoader(ModelLoader):
                 # https://github.com/huggingface/diffusers/blob/main/examples/dreambooth/train_dreambooth_lora_flux.py#L1194
                 model = lora_model_from_flux_diffusers_state_dict(state_dict=state_dict, alpha=None)
             elif config.format == ModelFormat.LyCORIS:
-                model = lora_model_from_flux_kohya_state_dict(state_dict=state_dict)
+                if is_state_dict_likely_in_flux_kohya_format(state_dict=state_dict):
+                    model = lora_model_from_flux_kohya_state_dict(state_dict=state_dict)
+                elif is_state_dict_likely_flux_control(state_dict=state_dict):
+                    model = lora_model_from_flux_control_state_dict(state_dict=state_dict)
             else:
                 raise ValueError(f"LoRA model is in unsupported FLUX format: {config.format}")
         elif self._model_base in [BaseModelType.StableDiffusion1, BaseModelType.StableDiffusion2]:

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

@@ -42,6 +42,7 @@ VARIANT_TO_IN_CHANNEL_MAP = {
 @ModelLoaderRegistry.register(
     base=BaseModelType.StableDiffusionXLRefiner, type=ModelType.Main, format=ModelFormat.Diffusers
 )
+@ModelLoaderRegistry.register(base=BaseModelType.StableDiffusion3, type=ModelType.Main, format=ModelFormat.Diffusers)
 @ModelLoaderRegistry.register(base=BaseModelType.StableDiffusion1, type=ModelType.Main, format=ModelFormat.Checkpoint)
 @ModelLoaderRegistry.register(base=BaseModelType.StableDiffusion2, type=ModelType.Main, format=ModelFormat.Checkpoint)
 @ModelLoaderRegistry.register(base=BaseModelType.StableDiffusionXL, type=ModelType.Main, format=ModelFormat.Checkpoint)
@@ -51,13 +52,6 @@ VARIANT_TO_IN_CHANNEL_MAP = {
 class StableDiffusionDiffusersModel(GenericDiffusersLoader):
     """Class to load main models."""
 
-    model_base_to_model_type = {
-        BaseModelType.StableDiffusion1: "FrozenCLIPEmbedder",
-        BaseModelType.StableDiffusion2: "FrozenOpenCLIPEmbedder",
-        BaseModelType.StableDiffusionXL: "SDXL",
-        BaseModelType.StableDiffusionXLRefiner: "SDXL-Refiner",
-    }
-
     def _load_model(
         self,
         config: AnyModelConfig,
@@ -117,8 +111,6 @@ class StableDiffusionDiffusersModel(GenericDiffusersLoader):
             load_class = load_classes[config.base][config.variant]
         except KeyError as e:
             raise Exception(f"No diffusers pipeline known for base={config.base}, variant={config.variant}") from e
-        prediction_type = config.prediction_type.value
-        upcast_attention = config.upcast_attention
 
         # Without SilenceWarnings we get log messages like this:
         # site-packages/huggingface_hub/file_download.py:1132: FutureWarning: `resume_download` is deprecated and will be removed in version 1.0.0. Downloads always resume when possible. If you want to force a new download, use `force_download=True`.
@@ -129,13 +121,7 @@ class StableDiffusionDiffusersModel(GenericDiffusersLoader):
         # ['text_model.embeddings.position_ids']
 
         with SilenceWarnings():
-            pipeline = load_class.from_single_file(
-                config.path,
-                torch_dtype=self._torch_dtype,
-                prediction_type=prediction_type,
-                upcast_attention=upcast_attention,
-                load_safety_checker=False,
-            )
+            pipeline = load_class.from_single_file(config.path, torch_dtype=self._torch_dtype)
 
         if not submodel_type:
             return pipeline

invokeai/backend/model_manager/load/model_util.py

@@ -15,9 +15,9 @@ from invokeai.backend.image_util.depth_anything.depth_anything_pipeline import D
 from invokeai.backend.image_util.grounding_dino.grounding_dino_pipeline import GroundingDinoPipeline
 from invokeai.backend.image_util.segment_anything.segment_anything_pipeline import SegmentAnythingPipeline
 from invokeai.backend.ip_adapter.ip_adapter import IPAdapter
-from invokeai.backend.lora.lora_model_raw import LoRAModelRaw
 from invokeai.backend.model_manager.config import AnyModel
 from invokeai.backend.onnx.onnx_runtime import IAIOnnxRuntimeModel
+from invokeai.backend.patches.model_patch_raw import ModelPatchRaw
 from invokeai.backend.spandrel_image_to_image_model import SpandrelImageToImageModel
 from invokeai.backend.textual_inversion import TextualInversionModelRaw
 from invokeai.backend.util.calc_tensor_size import calc_tensor_size
@@ -43,7 +43,7 @@ def calc_model_size_by_data(logger: logging.Logger, model: AnyModel) -> int:
         (
             TextualInversionModelRaw,
             IPAdapter,
-            LoRAModelRaw,
+            ModelPatchRaw,
             SpandrelImageToImageModel,
             GroundingDinoPipeline,
             SegmentAnythingPipeline,

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

@@ -20,7 +20,7 @@ from typing import Optional
 
 import requests
 from huggingface_hub import HfApi, configure_http_backend, hf_hub_url
-from huggingface_hub.utils._errors import RepositoryNotFoundError, RevisionNotFoundError
+from huggingface_hub.errors import RepositoryNotFoundError, RevisionNotFoundError
 from pydantic.networks import AnyHttpUrl
 from requests.sessions import Session
 

invokeai/backend/model_manager/probe.py

@@ -1,7 +1,7 @@
 import json
 import re
 from pathlib import Path
-from typing import Any, Dict, Literal, Optional, Union
+from typing import Any, Callable, Dict, Literal, Optional, Union
 
 import safetensors.torch
 import spandrel
@@ -15,13 +15,10 @@ from invokeai.backend.flux.controlnet.state_dict_utils import (
     is_state_dict_xlabs_controlnet,
 )
 from invokeai.backend.flux.ip_adapter.state_dict_utils import is_state_dict_xlabs_ip_adapter
-from invokeai.backend.lora.conversions.flux_diffusers_lora_conversion_utils import (
-    is_state_dict_likely_in_flux_diffusers_format,
-)
-from invokeai.backend.lora.conversions.flux_kohya_lora_conversion_utils import is_state_dict_likely_in_flux_kohya_format
 from invokeai.backend.model_hash.model_hash import HASHING_ALGORITHMS, ModelHash
 from invokeai.backend.model_manager.config import (
     AnyModelConfig,
+    AnyVariant,
     BaseModelType,
     ControlAdapterDefaultSettings,
     InvalidModelConfigException,
@@ -33,8 +30,22 @@ from invokeai.backend.model_manager.config import (
     ModelType,
     ModelVariantType,
     SchedulerPredictionType,
+    SubmodelDefinition,
+    SubModelType,
+)
+from invokeai.backend.model_manager.load.model_loaders.generic_diffusers import ConfigLoader
+from invokeai.backend.model_manager.util.model_util import (
+    get_clip_variant_type,
+    lora_token_vector_length,
+    read_checkpoint_meta,
+)
+from invokeai.backend.patches.lora_conversions.flux_control_lora_utils import is_state_dict_likely_flux_control
+from invokeai.backend.patches.lora_conversions.flux_diffusers_lora_conversion_utils import (
+    is_state_dict_likely_in_flux_diffusers_format,
+)
+from invokeai.backend.patches.lora_conversions.flux_kohya_lora_conversion_utils import (
+    is_state_dict_likely_in_flux_kohya_format,
 )
-from invokeai.backend.model_manager.util.model_util import lora_token_vector_length, read_checkpoint_meta
 from invokeai.backend.quantization.gguf.ggml_tensor import GGMLTensor
 from invokeai.backend.quantization.gguf.loaders import gguf_sd_loader
 from invokeai.backend.spandrel_image_to_image_model import SpandrelImageToImageModel
@@ -112,6 +123,7 @@ class ModelProbe(object):
         "StableDiffusionXLPipeline": ModelType.Main,
         "StableDiffusionXLImg2ImgPipeline": ModelType.Main,
         "StableDiffusionXLInpaintPipeline": ModelType.Main,
+        "StableDiffusion3Pipeline": ModelType.Main,
         "LatentConsistencyModelPipeline": ModelType.Main,
         "AutoencoderKL": ModelType.VAE,
         "AutoencoderTiny": ModelType.VAE,
@@ -122,8 +134,12 @@ class ModelProbe(object):
         "CLIPTextModel": ModelType.CLIPEmbed,
         "T5EncoderModel": ModelType.T5Encoder,
         "FluxControlNetModel": ModelType.ControlNet,
+        "SD3Transformer2DModel": ModelType.Main,
+        "CLIPTextModelWithProjection": ModelType.CLIPEmbed,
     }
 
+    TYPE2VARIANT: Dict[ModelType, Callable[[str], Optional[AnyVariant]]] = {ModelType.CLIPEmbed: get_clip_variant_type}
+
     @classmethod
     def register_probe(
         cls, format: Literal["diffusers", "checkpoint", "onnx"], model_type: ModelType, probe_class: type[ProbeBase]
@@ -170,7 +186,10 @@ class ModelProbe(object):
         fields["path"] = model_path.as_posix()
         fields["type"] = fields.get("type") or model_type
         fields["base"] = fields.get("base") or probe.get_base_type()
-        fields["variant"] = fields.get("variant") or probe.get_variant_type()
+        variant_func = cls.TYPE2VARIANT.get(fields["type"], None)
+        fields["variant"] = (
+            fields.get("variant") or (variant_func and variant_func(model_path.as_posix())) or probe.get_variant_type()
+        )
         fields["prediction_type"] = fields.get("prediction_type") or probe.get_scheduler_prediction_type()
         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)
@@ -183,8 +202,8 @@ class ModelProbe(object):
         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"])
+            if fields["type"] in {ModelType.ControlNet, ModelType.T2IAdapter, ModelType.ControlLoRa}:
+                fields["default_settings"] = get_default_settings_control_adapters(fields["name"])
             elif fields["type"] is ModelType.Main:
                 fields["default_settings"] = get_default_settings_main(fields["base"])
 
@@ -217,6 +236,10 @@ class ModelProbe(object):
                 and fields["prediction_type"] == SchedulerPredictionType.VPrediction
             )
 
+        get_submodels = getattr(probe, "get_submodels", None)
+        if fields["base"] == BaseModelType.StableDiffusion3 and callable(get_submodels):
+            fields["submodels"] = get_submodels()
+
         model_info = ModelConfigFactory.make_config(fields)  # , key=fields.get("key", None))
         return model_info
 
@@ -238,6 +261,9 @@ class ModelProbe(object):
         ckpt = checkpoint if checkpoint else read_checkpoint_meta(model_path, scan=True)
         ckpt = ckpt.get("state_dict", ckpt)
 
+        if isinstance(ckpt, dict) and is_state_dict_likely_flux_control(ckpt):
+            return ModelType.ControlLoRa
+
         for key in [str(k) for k in ckpt.keys()]:
             if key.startswith(
                 (
@@ -449,7 +475,7 @@ class ModelProbe(object):
         """
         # scan model
         scan_result = scan_file_path(checkpoint)
-        if scan_result.infected_files != 0:
+        if scan_result.infected_files != 0 or scan_result.scan_err:
             raise Exception("The model {model_name} is potentially infected by malware. Aborting import.")
 
 
@@ -465,6 +491,7 @@ MODEL_NAME_TO_PREPROCESSOR = {
     "lineart anime": "lineart_anime_image_processor",
     "lineart_anime": "lineart_anime_image_processor",
     "lineart": "lineart_image_processor",
+    "soft": "hed_image_processor",
     "softedge": "hed_image_processor",
     "hed": "hed_image_processor",
     "shuffle": "content_shuffle_image_processor",
@@ -476,7 +503,7 @@ MODEL_NAME_TO_PREPROCESSOR = {
 }
 
 
-def get_default_settings_controlnet_t2i_adapter(model_name: str) -> Optional[ControlAdapterDefaultSettings]:
+def get_default_settings_control_adapters(model_name: str) -> Optional[ControlAdapterDefaultSettings]:
     for k, v in MODEL_NAME_TO_PREPROCESSOR.items():
         model_name_lower = model_name.lower()
         if k in model_name_lower:
@@ -603,8 +630,10 @@ class LoRACheckpointProbe(CheckpointProbeBase):
         return ModelFormat.LyCORIS
 
     def get_base_type(self) -> BaseModelType:
-        if is_state_dict_likely_in_flux_kohya_format(self.checkpoint) or is_state_dict_likely_in_flux_diffusers_format(
-            self.checkpoint
+        if (
+            is_state_dict_likely_in_flux_kohya_format(self.checkpoint)
+            or is_state_dict_likely_in_flux_diffusers_format(self.checkpoint)
+            or is_state_dict_likely_flux_control(self.checkpoint)
         ):
             return BaseModelType.Flux
 
@@ -747,18 +776,33 @@ class FolderProbeBase(ProbeBase):
 
 class PipelineFolderProbe(FolderProbeBase):
     def get_base_type(self) -> BaseModelType:
-        with open(self.model_path / "unet" / "config.json", "r") as file:
-            unet_conf = json.load(file)
-        if unet_conf["cross_attention_dim"] == 768:
-            return BaseModelType.StableDiffusion1
-        elif unet_conf["cross_attention_dim"] == 1024:
-            return BaseModelType.StableDiffusion2
-        elif unet_conf["cross_attention_dim"] == 1280:
-            return BaseModelType.StableDiffusionXLRefiner
-        elif unet_conf["cross_attention_dim"] == 2048:
-            return BaseModelType.StableDiffusionXL
-        else:
-            raise InvalidModelConfigException(f"Unknown base model for {self.model_path}")
+        # Handle pipelines with a UNet (i.e SD 1.x, SD2, SDXL).
+        config_path = self.model_path / "unet" / "config.json"
+        if config_path.exists():
+            with open(config_path) as file:
+                unet_conf = json.load(file)
+            if unet_conf["cross_attention_dim"] == 768:
+                return BaseModelType.StableDiffusion1
+            elif unet_conf["cross_attention_dim"] == 1024:
+                return BaseModelType.StableDiffusion2
+            elif unet_conf["cross_attention_dim"] == 1280:
+                return BaseModelType.StableDiffusionXLRefiner
+            elif unet_conf["cross_attention_dim"] == 2048:
+                return BaseModelType.StableDiffusionXL
+            else:
+                raise InvalidModelConfigException(f"Unknown base model for {self.model_path}")
+
+        # Handle pipelines with a transformer (i.e. SD3).
+        config_path = self.model_path / "transformer" / "config.json"
+        if config_path.exists():
+            with open(config_path) as file:
+                transformer_conf = json.load(file)
+            if transformer_conf["_class_name"] == "SD3Transformer2DModel":
+                return BaseModelType.StableDiffusion3
+            else:
+                raise InvalidModelConfigException(f"Unknown base model for {self.model_path}")
+
+        raise InvalidModelConfigException(f"Unknown base model for {self.model_path}")
 
     def get_scheduler_prediction_type(self) -> SchedulerPredictionType:
         with open(self.model_path / "scheduler" / "scheduler_config.json", "r") as file:
@@ -770,6 +814,23 @@ class PipelineFolderProbe(FolderProbeBase):
         else:
             raise InvalidModelConfigException("Unknown scheduler prediction type: {scheduler_conf['prediction_type']}")
 
+    def get_submodels(self) -> Dict[SubModelType, SubmodelDefinition]:
+        config = ConfigLoader.load_config(self.model_path, config_name="model_index.json")
+        submodels: Dict[SubModelType, SubmodelDefinition] = {}
+        for key, value in config.items():
+            if key.startswith("_") or not (isinstance(value, list) and len(value) == 2):
+                continue
+            model_loader = str(value[1])
+            if model_type := ModelProbe.CLASS2TYPE.get(model_loader):
+                variant_func = ModelProbe.TYPE2VARIANT.get(model_type, None)
+                submodels[SubModelType(key)] = SubmodelDefinition(
+                    path_or_prefix=(self.model_path / key).resolve().as_posix(),
+                    model_type=model_type,
+                    variant=variant_func and variant_func((self.model_path / key).as_posix()),
+                )
+
+        return submodels
+
     def get_variant_type(self) -> ModelVariantType:
         # This only works for pipelines! Any kind of
         # exception results in our returning the
@@ -981,6 +1042,7 @@ class T2IAdapterFolderProbe(FolderProbeBase):
 ModelProbe.register_probe("diffusers", ModelType.Main, PipelineFolderProbe)
 ModelProbe.register_probe("diffusers", ModelType.VAE, VaeFolderProbe)
 ModelProbe.register_probe("diffusers", ModelType.LoRA, LoRAFolderProbe)
+ModelProbe.register_probe("diffusers", ModelType.ControlLoRa, LoRAFolderProbe)
 ModelProbe.register_probe("diffusers", ModelType.TextualInversion, TextualInversionFolderProbe)
 ModelProbe.register_probe("diffusers", ModelType.T5Encoder, T5EncoderFolderProbe)
 ModelProbe.register_probe("diffusers", ModelType.ControlNet, ControlNetFolderProbe)
@@ -993,6 +1055,7 @@ ModelProbe.register_probe("diffusers", ModelType.SpandrelImageToImage, SpandrelI
 ModelProbe.register_probe("checkpoint", ModelType.Main, PipelineCheckpointProbe)
 ModelProbe.register_probe("checkpoint", ModelType.VAE, VaeCheckpointProbe)
 ModelProbe.register_probe("checkpoint", ModelType.LoRA, LoRACheckpointProbe)
+ModelProbe.register_probe("checkpoint", ModelType.ControlLoRa, LoRACheckpointProbe)
 ModelProbe.register_probe("checkpoint", ModelType.TextualInversion, TextualInversionCheckpointProbe)
 ModelProbe.register_probe("checkpoint", ModelType.ControlNet, ControlNetCheckpointProbe)
 ModelProbe.register_probe("checkpoint", ModelType.IPAdapter, IPAdapterCheckpointProbe)

invokeai/backend/model_manager/starter_models.py

@@ -140,6 +140,22 @@ flux_dev = StarterModel(
     type=ModelType.Main,
     dependencies=[t5_base_encoder, flux_vae, clip_l_encoder],
 )
+sd35_medium = StarterModel(
+    name="SD3.5 Medium",
+    base=BaseModelType.StableDiffusion3,
+    source="stabilityai/stable-diffusion-3.5-medium",
+    description="Medium SD3.5 Model: ~15GB",
+    type=ModelType.Main,
+    dependencies=[],
+)
+sd35_large = StarterModel(
+    name="SD3.5 Large",
+    base=BaseModelType.StableDiffusion3,
+    source="stabilityai/stable-diffusion-3.5-large",
+    description="Large SD3.5 Model: ~19G",
+    type=ModelType.Main,
+    dependencies=[],
+)
 cyberrealistic_sd1 = StarterModel(
     name="CyberRealistic v4.1",
     base=BaseModelType.StableDiffusion1,
@@ -282,13 +298,12 @@ ip_adapter_sdxl = StarterModel(
     previous_names=["IP Adapter SDXL"],
 )
 ip_adapter_flux = StarterModel(
-    name="Standard Reference (XLabs FLUX IP-Adapter)",
+    name="Standard Reference (XLabs FLUX IP-Adapter v2)",
     base=BaseModelType.Flux,
-    source="https://huggingface.co/XLabs-AI/flux-ip-adapter/resolve/main/flux-ip-adapter.safetensors",
+    source="https://huggingface.co/XLabs-AI/flux-ip-adapter-v2/resolve/main/ip_adapter.safetensors",
     description="References images with a more generalized/looser degree of precision.",
     type=ModelType.IPAdapter,
     dependencies=[clip_vit_l_image_encoder],
-    previous_names=["XLabs FLUX IP-Adapter"],
 )
 # endregion
 # region ControlNet
@@ -473,6 +488,22 @@ union_cnet_flux = StarterModel(
     type=ModelType.ControlNet,
 )
 # endregion
+# region Control LoRA
+flux_canny_control_lora = StarterModel(
+    name="Hard Edge Detection (Canny)",
+    base=BaseModelType.Flux,
+    source="black-forest-labs/FLUX.1-Canny-dev-lora::flux1-canny-dev-lora.safetensors",
+    description="Uses detected edges in the image to control composition.",
+    type=ModelType.ControlLoRa,
+)
+flux_depth_control_lora = StarterModel(
+    name="Depth Map",
+    base=BaseModelType.Flux,
+    source="black-forest-labs/FLUX.1-Depth-dev-lora::flux1-depth-dev-lora.safetensors",
+    description="Uses depth information in the image to control the depth in the generation.",
+    type=ModelType.ControlLoRa,
+)
+# endregion
 # region T2I Adapter
 t2i_canny_sd1 = StarterModel(
     name="Hard Edge Detection (canny)",
@@ -570,6 +601,8 @@ STARTER_MODELS: list[StarterModel] = [
     flux_dev_quantized,
     flux_schnell,
     flux_dev,
+    sd35_medium,
+    sd35_large,
     cyberrealistic_sd1,
     rev_animated_sd1,
     dreamshaper_8_sd1,
@@ -613,6 +646,8 @@ STARTER_MODELS: list[StarterModel] = [
     tile_sdxl,
     union_cnet_sdxl,
     union_cnet_flux,
+    flux_canny_control_lora,
+    flux_depth_control_lora,
     t2i_canny_sd1,
     t2i_sketch_sd1,
     t2i_depth_sd1,
@@ -671,6 +706,8 @@ flux_bundle: list[StarterModel] = [
     clip_l_encoder,
     union_cnet_flux,
     ip_adapter_flux,
+    flux_canny_control_lora,
+    flux_depth_control_lora,
 ]
 
 STARTER_BUNDLES: dict[str, list[StarterModel]] = {

invokeai/backend/model_manager/util/model_util.py

@@ -8,6 +8,7 @@ import safetensors
 import torch
 from picklescan.scanner import scan_file_path
 
+from invokeai.backend.model_manager.config import ClipVariantType
 from invokeai.backend.quantization.gguf.loaders import gguf_sd_loader
 
 
@@ -43,7 +44,7 @@ def _fast_safetensors_reader(path: str) -> Dict[str, torch.Tensor]:
     return checkpoint
 
 
-def read_checkpoint_meta(path: Union[str, Path], scan: bool = False) -> Dict[str, torch.Tensor]:
+def read_checkpoint_meta(path: Union[str, Path], scan: bool = True) -> Dict[str, torch.Tensor]:
     if str(path).endswith(".safetensors"):
         try:
             path_str = path.as_posix() if isinstance(path, Path) else path
@@ -51,16 +52,15 @@ def read_checkpoint_meta(path: Union[str, Path], scan: bool = False) -> Dict[str
         except Exception:
             # TODO: create issue for support "meta"?
             checkpoint = safetensors.torch.load_file(path, device="cpu")
+    elif str(path).endswith(".gguf"):
+        # The GGUF reader used here uses numpy memmap, so these tensors are not loaded into memory during this function
+        checkpoint = gguf_sd_loader(Path(path), compute_dtype=torch.float32)
     else:
         if scan:
             scan_result = scan_file_path(path)
-            if scan_result.infected_files != 0:
+            if scan_result.infected_files != 0 or scan_result.scan_err:
                 raise Exception(f'The model file "{path}" is potentially infected by malware. Aborting import.')
-        if str(path).endswith(".gguf"):
-            # The GGUF reader used here uses numpy memmap, so these tensors are not loaded into memory during this function
-            checkpoint = gguf_sd_loader(Path(path), compute_dtype=torch.float32)
-        else:
-            checkpoint = torch.load(path, map_location=torch.device("meta"))
+        checkpoint = torch.load(path, map_location=torch.device("meta"))
     return checkpoint
 
 
@@ -165,3 +165,25 @@ def convert_bundle_to_flux_transformer_checkpoint(
         del transformer_state_dict[k]
 
     return original_state_dict
+
+
+def get_clip_variant_type(location: str) -> Optional[ClipVariantType]:
+    try:
+        path = Path(location)
+        config_path = path / "config.json"
+        if not config_path.exists():
+            config_path = path / "text_encoder" / "config.json"
+        if not config_path.exists():
+            return ClipVariantType.L
+        with open(config_path) as file:
+            clip_conf = json.load(file)
+            hidden_size = clip_conf.get("hidden_size", -1)
+            match hidden_size:
+                case 1280:
+                    return ClipVariantType.G
+                case 768:
+                    return ClipVariantType.L
+                case _:
+                    return ClipVariantType.L
+    except Exception:
+        return ClipVariantType.L

invokeai/backend/model_manager/util/select_hf_files.py

@@ -85,6 +85,7 @@ def _filter_by_variant(files: List[Path], variant: ModelRepoVariant) -> Set[Path
     """Select the proper variant files from a list of HuggingFace repo_id paths."""
     result: set[Path] = set()
     subfolder_weights: dict[Path, list[SubfolderCandidate]] = {}
+    safetensors_detected = False
     for path in files:
         if path.suffix in [".onnx", ".pb", ".onnx_data"]:
             if variant == ModelRepoVariant.ONNX:
@@ -119,19 +120,27 @@ def _filter_by_variant(files: List[Path], variant: ModelRepoVariant) -> Set[Path
             # We prefer safetensors over other file formats and an exact variant match. We'll score each file based on
             # variant and format and select the best one.
 
+            if safetensors_detected and path.suffix == ".bin":
+                continue
+
             parent = path.parent
             score = 0
 
             if path.suffix == ".safetensors":
+                safetensors_detected = True
+                if parent in subfolder_weights:
+                    subfolder_weights[parent] = [sfc for sfc in subfolder_weights[parent] if sfc.path.suffix != ".bin"]
                 score += 1
 
             candidate_variant_label = path.suffixes[0] if len(path.suffixes) == 2 else None
 
             # Some special handling is needed here if there is not an exact match and if we cannot infer the variant
             # from the file name. In this case, we only give this file a point if the requested variant is FP32 or DEFAULT.
-            if candidate_variant_label == f".{variant}" or (
-                not candidate_variant_label and variant in [ModelRepoVariant.FP32, ModelRepoVariant.Default]
-            ):
+            if (
+                variant is not ModelRepoVariant.Default
+                and candidate_variant_label
+                and candidate_variant_label.startswith(f".{variant.value}")
+            ) or (not candidate_variant_label and variant in [ModelRepoVariant.FP32, ModelRepoVariant.Default]):
                 score += 1
 
             if parent not in subfolder_weights:
@@ -146,7 +155,7 @@ def _filter_by_variant(files: List[Path], variant: ModelRepoVariant) -> Set[Path
         # Check if at least one of the files has the explicit fp16 variant.
         at_least_one_fp16 = False
         for candidate in candidate_list:
-            if len(candidate.path.suffixes) == 2 and candidate.path.suffixes[0] == ".fp16":
+            if len(candidate.path.suffixes) == 2 and candidate.path.suffixes[0].startswith(".fp16"):
                 at_least_one_fp16 = True
                 break
 
@@ -162,7 +171,16 @@ def _filter_by_variant(files: List[Path], variant: ModelRepoVariant) -> Set[Path
         # candidate.
         highest_score_candidate = max(candidate_list, key=lambda candidate: candidate.score)
         if highest_score_candidate:
-            result.add(highest_score_candidate.path)
+            pattern = r"^(.*?)-\d+-of-\d+(\.\w+)$"
+            match = re.match(pattern, highest_score_candidate.path.as_posix())
+            if match:
+                for candidate in candidate_list:
+                    if candidate.path.as_posix().startswith(match.group(1)) and candidate.path.as_posix().endswith(
+                        match.group(2)
+                    ):
+                        result.add(candidate.path)
+            else:
+                result.add(highest_score_candidate.path)
 
     # If one of the architecture-related variants was specified and no files matched other than
     # config and text files then we return an empty list

invokeai/backend/model_patcher.py

@@ -5,17 +5,14 @@ from __future__ import annotations
 
 import pickle
 from contextlib import contextmanager
-from typing import Any, Dict, Iterator, List, Optional, Tuple, Type, Union
+from typing import Any, Iterator, List, Optional, Tuple, Type, Union
 
-import numpy as np
 import torch
 from diffusers import UNet2DConditionModel
 from transformers import CLIPTextModel, CLIPTextModelWithProjection, CLIPTokenizer
 
 from invokeai.app.shared.models import FreeUConfig
-from invokeai.backend.lora.lora_model_raw import LoRAModelRaw
 from invokeai.backend.model_manager.load.optimizations import skip_torch_weight_init
-from invokeai.backend.onnx.onnx_runtime import IAIOnnxRuntimeModel
 from invokeai.backend.textual_inversion import TextualInversionManager, TextualInversionModelRaw
 
 
@@ -176,180 +173,3 @@ class ModelPatcher:
             assert hasattr(unet, "disable_freeu")  # mypy doesn't pick up this attribute?
             if did_apply_freeu:
                 unet.disable_freeu()
-
-
-class ONNXModelPatcher:
-    # based on
-    # https://github.com/ssube/onnx-web/blob/ca2e436f0623e18b4cfe8a0363fcfcf10508acf7/api/onnx_web/convert/diffusion/lora.py#L323
-    @classmethod
-    @contextmanager
-    def apply_lora(
-        cls,
-        model: IAIOnnxRuntimeModel,
-        loras: List[Tuple[LoRAModelRaw, float]],
-        prefix: str,
-    ) -> None:
-        from invokeai.backend.models.base import IAIOnnxRuntimeModel
-
-        if not isinstance(model, IAIOnnxRuntimeModel):
-            raise Exception("Only IAIOnnxRuntimeModel models supported")
-
-        orig_weights = {}
-
-        try:
-            blended_loras: Dict[str, torch.Tensor] = {}
-
-            for lora, lora_weight in loras:
-                for layer_key, layer in lora.layers.items():
-                    if not layer_key.startswith(prefix):
-                        continue
-
-                    layer.to(dtype=torch.float32)
-                    layer_key = layer_key.replace(prefix, "")
-                    # TODO: rewrite to pass original tensor weight(required by ia3)
-                    layer_weight = layer.get_weight(None).detach().cpu().numpy() * lora_weight
-                    if layer_key in blended_loras:
-                        blended_loras[layer_key] += layer_weight
-                    else:
-                        blended_loras[layer_key] = layer_weight
-
-            node_names = {}
-            for node in model.nodes.values():
-                node_names[node.name.replace("/", "_").replace(".", "_").lstrip("_")] = node.name
-
-            for layer_key, lora_weight in blended_loras.items():
-                conv_key = layer_key + "_Conv"
-                gemm_key = layer_key + "_Gemm"
-                matmul_key = layer_key + "_MatMul"
-
-                if conv_key in node_names or gemm_key in node_names:
-                    if conv_key in node_names:
-                        conv_node = model.nodes[node_names[conv_key]]
-                    else:
-                        conv_node = model.nodes[node_names[gemm_key]]
-
-                    weight_name = [n for n in conv_node.input if ".weight" in n][0]
-                    orig_weight = model.tensors[weight_name]
-
-                    if orig_weight.shape[-2:] == (1, 1):
-                        if lora_weight.shape[-2:] == (1, 1):
-                            new_weight = orig_weight.squeeze((3, 2)) + lora_weight.squeeze((3, 2))
-                        else:
-                            new_weight = orig_weight.squeeze((3, 2)) + lora_weight
-
-                        new_weight = np.expand_dims(new_weight, (2, 3))
-                    else:
-                        if orig_weight.shape != lora_weight.shape:
-                            new_weight = orig_weight + lora_weight.reshape(orig_weight.shape)
-                        else:
-                            new_weight = orig_weight + lora_weight
-
-                    orig_weights[weight_name] = orig_weight
-                    model.tensors[weight_name] = new_weight.astype(orig_weight.dtype)
-
-                elif matmul_key in node_names:
-                    weight_node = model.nodes[node_names[matmul_key]]
-                    matmul_name = [n for n in weight_node.input if "MatMul" in n][0]
-
-                    orig_weight = model.tensors[matmul_name]
-                    new_weight = orig_weight + lora_weight.transpose()
-
-                    orig_weights[matmul_name] = orig_weight
-                    model.tensors[matmul_name] = new_weight.astype(orig_weight.dtype)
-
-                else:
-                    # warn? err?
-                    pass
-
-            yield
-
-        finally:
-            # restore original weights
-            for name, orig_weight in orig_weights.items():
-                model.tensors[name] = orig_weight
-
-    @classmethod
-    @contextmanager
-    def apply_ti(
-        cls,
-        tokenizer: CLIPTokenizer,
-        text_encoder: IAIOnnxRuntimeModel,
-        ti_list: List[Tuple[str, Any]],
-    ) -> Iterator[Tuple[CLIPTokenizer, TextualInversionManager]]:
-        from invokeai.backend.models.base import IAIOnnxRuntimeModel
-
-        if not isinstance(text_encoder, IAIOnnxRuntimeModel):
-            raise Exception("Only IAIOnnxRuntimeModel models supported")
-
-        orig_embeddings = None
-
-        try:
-            # HACK: The CLIPTokenizer API does not include a way to remove tokens after calling add_tokens(...). As a
-            # workaround, we create a full copy of `tokenizer` so that its original behavior can be restored after
-            # exiting this `apply_ti(...)` context manager.
-            #
-            # In a previous implementation, the deep copy was obtained with `ti_tokenizer = copy.deepcopy(tokenizer)`,
-            # but a pickle roundtrip was found to be much faster (1 sec vs. 0.05 secs).
-            ti_tokenizer = pickle.loads(pickle.dumps(tokenizer))
-            ti_manager = TextualInversionManager(ti_tokenizer)
-
-            def _get_trigger(ti_name: str, index: int) -> str:
-                trigger = ti_name
-                if index > 0:
-                    trigger += f"-!pad-{i}"
-                return f"<{trigger}>"
-
-            # modify text_encoder
-            orig_embeddings = text_encoder.tensors["text_model.embeddings.token_embedding.weight"]
-
-            # modify tokenizer
-            new_tokens_added = 0
-            for ti_name, ti in ti_list:
-                if ti.embedding_2 is not None:
-                    ti_embedding = (
-                        ti.embedding_2 if ti.embedding_2.shape[1] == orig_embeddings.shape[0] else ti.embedding
-                    )
-                else:
-                    ti_embedding = ti.embedding
-
-                for i in range(ti_embedding.shape[0]):
-                    new_tokens_added += ti_tokenizer.add_tokens(_get_trigger(ti_name, i))
-
-            embeddings = np.concatenate(
-                (np.copy(orig_embeddings), np.zeros((new_tokens_added, orig_embeddings.shape[1]))),
-                axis=0,
-            )
-
-            for ti_name, _ in ti_list:
-                ti_tokens = []
-                for i in range(ti_embedding.shape[0]):
-                    embedding = ti_embedding[i].detach().numpy()
-                    trigger = _get_trigger(ti_name, i)
-
-                    token_id = ti_tokenizer.convert_tokens_to_ids(trigger)
-                    if token_id == ti_tokenizer.unk_token_id:
-                        raise RuntimeError(f"Unable to find token id for token '{trigger}'")
-
-                    if embeddings[token_id].shape != embedding.shape:
-                        raise ValueError(
-                            f"Cannot load embedding for {trigger}. It was trained on a model with token dimension"
-                            f" {embedding.shape[0]}, but the current model has token dimension"
-                            f" {embeddings[token_id].shape[0]}."
-                        )
-
-                    embeddings[token_id] = embedding
-                    ti_tokens.append(token_id)
-
-                if len(ti_tokens) > 1:
-                    ti_manager.pad_tokens[ti_tokens[0]] = ti_tokens[1:]
-
-            text_encoder.tensors["text_model.embeddings.token_embedding.weight"] = embeddings.astype(
-                orig_embeddings.dtype
-            )
-
-            yield ti_tokenizer, ti_manager
-
-        finally:
-            # restore
-            if orig_embeddings is not None:
-                text_encoder.tensors["text_model.embeddings.token_embedding.weight"] = orig_embeddings

invokeai/backend/patches/layers/base_layer_patch.py

@@ -0,0 +1,22 @@
+from abc import ABC, abstractmethod
+
+import torch
+
+
+class BaseLayerPatch(ABC):
+    @abstractmethod
+    def get_parameters(self, orig_module: torch.nn.Module, weight: float) -> dict[str, torch.Tensor]:
+        """Get the parameter residual updates that should be applied to the original parameters. Parameters omitted
+        from the returned dict are not updated.
+        """
+        ...
+
+    @abstractmethod
+    def to(self, device: torch.device | None = None, dtype: torch.dtype | None = None):
+        """Move all internal tensors to the specified device and dtype."""
+        ...
+
+    @abstractmethod
+    def calc_size(self) -> int:
+        """Calculate the total size of all internal tensors in bytes."""
+        ...

invokeai/backend/patches/layers/concatenated_lora_layer.py

@@ -2,8 +2,8 @@ from typing import Optional, Sequence
 
 import torch
 
-from invokeai.backend.lora.layers.lora_layer import LoRALayer
-from invokeai.backend.lora.layers.lora_layer_base import LoRALayerBase
+from invokeai.backend.patches.layers.lora_layer import LoRALayer
+from invokeai.backend.patches.layers.lora_layer_base import LoRALayerBase
 
 
 class ConcatenatedLoRALayer(LoRALayerBase):
@@ -20,7 +20,7 @@ class ConcatenatedLoRALayer(LoRALayerBase):
         self.lora_layers = lora_layers
         self.concat_axis = concat_axis
 
-    def rank(self) -> int | None:
+    def _rank(self) -> int | None:
         return None
 
     def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:

invokeai/backend/patches/layers/flux_control_lora_layer.py

@@ -0,0 +1,19 @@
+import torch
+
+from invokeai.backend.patches.layers.lora_layer import LoRALayer
+
+
+class FluxControlLoRALayer(LoRALayer):
+    """A special case of LoRALayer for use with FLUX Control LoRAs that pads the target parameter with zeros if the
+    shapes don't match.
+    """
+
+    def get_parameters(self, orig_module: torch.nn.Module, weight: float) -> dict[str, torch.Tensor]:
+        """This overrides the base class behavior to skip the reshaping step."""
+        scale = self.scale()
+        params = {"weight": self.get_weight(orig_module.weight) * (weight * scale)}
+        bias = self.get_bias(orig_module.bias)
+        if bias is not None:
+            params["bias"] = bias * (weight * scale)
+
+        return params

invokeai/backend/patches/layers/full_layer.py

@@ -2,7 +2,7 @@ from typing import Dict, Optional
 
 import torch
 
-from invokeai.backend.lora.layers.lora_layer_base import LoRALayerBase
+from invokeai.backend.patches.layers.lora_layer_base import LoRALayerBase
 from invokeai.backend.util.calc_tensor_size import calc_tensor_size
 
 
@@ -20,7 +20,7 @@ class FullLayer(LoRALayerBase):
         cls.warn_on_unhandled_keys(values=values, handled_keys={"diff", "diff_b"})
         return layer
 
-    def rank(self) -> int | None:
+    def _rank(self) -> int | None:
         return None
 
     def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:

invokeai/backend/patches/layers/ia3_layer.py

@@ -2,7 +2,7 @@ from typing import Dict, Optional
 
 import torch
 
-from invokeai.backend.lora.layers.lora_layer_base import LoRALayerBase
+from invokeai.backend.patches.layers.lora_layer_base import LoRALayerBase
 
 
 class IA3Layer(LoRALayerBase):
@@ -16,7 +16,7 @@ class IA3Layer(LoRALayerBase):
         self.weight = weight
         self.on_input = on_input
 
-    def rank(self) -> int | None:
+    def _rank(self) -> int | None:
         return None
 
     @classmethod

invokeai/backend/patches/layers/loha_layer.py

@@ -2,7 +2,7 @@ from typing import Dict
 
 import torch
 
-from invokeai.backend.lora.layers.lora_layer_base import LoRALayerBase
+from invokeai.backend.patches.layers.lora_layer_base import LoRALayerBase
 from invokeai.backend.util.calc_tensor_size import calc_tensors_size
 
 
@@ -32,7 +32,7 @@ class LoHALayer(LoRALayerBase):
         self.t2 = t2
         assert (self.t1 is None) == (self.t2 is None)
 
-    def rank(self) -> int | None:
+    def _rank(self) -> int | None:
         return self.w1_b.shape[0]
 
     @classmethod

invokeai/backend/patches/layers/lokr_layer.py

@@ -2,7 +2,7 @@ from typing import Dict
 
 import torch
 
-from invokeai.backend.lora.layers.lora_layer_base import LoRALayerBase
+from invokeai.backend.patches.layers.lora_layer_base import LoRALayerBase
 from invokeai.backend.util.calc_tensor_size import calc_tensors_size
 
 
@@ -39,7 +39,7 @@ class LoKRLayer(LoRALayerBase):
         assert (self.w2 is None) != (self.w2_a is None)
         assert (self.w2_a is None) == (self.w2_b is None)
 
-    def rank(self) -> int | None:
+    def _rank(self) -> int | None:
         if self.w1_b is not None:
             return self.w1_b.shape[0]
         elif self.w2_b is not None:

invokeai/backend/patches/layers/lora_layer.py

@@ -2,7 +2,7 @@ from typing import Dict, Optional
 
 import torch
 
-from invokeai.backend.lora.layers.lora_layer_base import LoRALayerBase
+from invokeai.backend.patches.layers.lora_layer_base import LoRALayerBase
 from invokeai.backend.util.calc_tensor_size import calc_tensors_size
 
 
@@ -55,7 +55,7 @@ class LoRALayer(LoRALayerBase):
 
         return layer
 
-    def rank(self) -> int:
+    def _rank(self) -> int:
         return self.down.shape[0]
 
     def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor: