Update to reflect an alpha release

Need to figure out why the rect isn't reset when the entity is reset. Probably just needs some special handling.

.github/workflows/build-container.yml

@@ -13,6 +13,12 @@ on:
     tags:
       - 'v*.*.*'
   workflow_dispatch:
+    inputs:
+      push-to-registry:
+        description: Push the built image to the container registry
+        required: false
+        type: boolean
+        default: false
 
 permissions:
   contents: write
@@ -50,16 +56,15 @@ jobs:
           df -h
 
       - name: Checkout
-        uses: actions/checkout@v3
+        uses: actions/checkout@v4
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@v4
+        uses: docker/metadata-action@v5
         with:
           github-token: ${{ secrets.GITHUB_TOKEN }}
           images: |
             ghcr.io/${{ github.repository }}
-            ${{ env.DOCKERHUB_REPOSITORY }}
           tags: |
             type=ref,event=branch
             type=ref,event=tag
@@ -72,49 +77,33 @@ jobs:
             suffix=-${{ matrix.gpu-driver }},onlatest=false
 
       - name: Set up QEMU
-        uses: docker/setup-qemu-action@v2
+        uses: docker/setup-qemu-action@v3
 
       - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v2
+        uses: docker/setup-buildx-action@v3
         with:
           platforms: ${{ env.PLATFORMS }}
 
       - name: Login to GitHub Container Registry
         if: github.event_name != 'pull_request'
-        uses: docker/login-action@v2
+        uses: docker/login-action@v3
         with:
           registry: ghcr.io
           username: ${{ github.repository_owner }}
           password: ${{ secrets.GITHUB_TOKEN }}
 
-      # - name: Login to Docker Hub
-      #   if: github.event_name != 'pull_request' && vars.DOCKERHUB_REPOSITORY != ''
-      #   uses: docker/login-action@v2
-      #   with:
-      #     username: ${{ secrets.DOCKERHUB_USERNAME }}
-      #     password: ${{ secrets.DOCKERHUB_TOKEN }}
-
       - name: Build container
         timeout-minutes: 40
         id: docker_build
-        uses: docker/build-push-action@v4
+        uses: docker/build-push-action@v6
         with:
           context: .
           file: docker/Dockerfile
           platforms: ${{ env.PLATFORMS }}
-          push: ${{ github.ref == 'refs/heads/main' || github.ref_type == 'tag' }}
+          push: ${{ github.ref == 'refs/heads/main' || github.ref_type == 'tag' || github.event.inputs.push-to-registry }}
           tags: ${{ steps.meta.outputs.tags }}
           labels: ${{ steps.meta.outputs.labels }}
           cache-from: |
             type=gha,scope=${{ github.ref_name }}-${{ matrix.gpu-driver }}
             type=gha,scope=main-${{ matrix.gpu-driver }}
           cache-to: type=gha,mode=max,scope=${{ github.ref_name }}-${{ matrix.gpu-driver }}
-
-      # - name: Docker Hub Description
-      #   if: github.ref == 'refs/heads/main' || github.ref == 'refs/tags/*' && vars.DOCKERHUB_REPOSITORY != ''
-      #   uses: peter-evans/dockerhub-description@v3
-      #   with:
-      #     username: ${{ secrets.DOCKERHUB_USERNAME }}
-      #     password: ${{ secrets.DOCKERHUB_TOKEN }}
-      #     repository: ${{ vars.DOCKERHUB_REPOSITORY }}
-      #     short-description: ${{ github.event.repository.description }}

.github/workflows/python-checks.yml

@@ -62,7 +62,7 @@ jobs:
 
       - name: install ruff
         if: ${{ steps.changed-files.outputs.python_any_changed == 'true' || inputs.always_run == true }}
-        run: pip install ruff
+        run: pip install ruff==0.6.0
         shell: bash
 
       - name: ruff check

.github/workflows/python-tests.yml

@@ -60,7 +60,7 @@ jobs:
             extra-index-url: 'https://download.pytorch.org/whl/cpu'
             github-env: $GITHUB_ENV
           - platform: macos-default
-            os: macOS-12
+            os: macOS-14
             github-env: $GITHUB_ENV
           - platform: windows-cpu
             os: windows-2022

docker/Dockerfile

@@ -55,6 +55,7 @@ RUN --mount=type=cache,target=/root/.cache/pip \
 FROM node:20-slim AS web-builder
 ENV PNPM_HOME="/pnpm"
 ENV PATH="$PNPM_HOME:$PATH"
+RUN corepack use pnpm@8.x
 RUN corepack enable
 
 WORKDIR /build

docker/README.md

@@ -1,20 +1,22 @@
 # Invoke in Docker
 
-- Ensure that Docker can use the GPU on your system
-- This documentation assumes Linux, but should work similarly under Windows with WSL2
+First things first:
+
+- Ensure that Docker can use your [NVIDIA][nvidia docker docs] or [AMD][amd docker docs] GPU.
+- This document assumes a Linux system, but should work similarly under Windows with WSL2.
 - We don't recommend running Invoke in Docker on macOS at this time. It works, but very slowly.
 
-## Quickstart :lightning:
+## Quickstart
 
-No `docker compose`, no persistence, just a simple one-liner using the official images:
+No `docker compose`, no persistence, single command, using the official images:
 
-**CUDA:**
+**CUDA (NVIDIA GPU):**
 
 ```bash
 docker run --runtime=nvidia --gpus=all --publish 9090:9090 ghcr.io/invoke-ai/invokeai
 ```
 
-**ROCm:**
+**ROCm (AMD GPU):**
 
 ```bash
 docker run --device /dev/kfd --device /dev/dri --publish 9090:9090 ghcr.io/invoke-ai/invokeai:main-rocm
@@ -22,12 +24,20 @@ docker run --device /dev/kfd --device /dev/dri --publish 9090:9090 ghcr.io/invok
 
 Open `http://localhost:9090` in your browser once the container finishes booting, install some models, and generate away!
 
-> [!TIP]
-> To persist your data (including downloaded models) outside of the container, add a `--volume/-v` flag to the above command, e.g.: `docker run --volume /some/local/path:/invokeai <...the rest of the command>`
+### Data persistence
+
+To persist your generated images and downloaded models outside of the container, add a `--volume/-v` flag to the above command, e.g.:
+
+```bash
+docker run --volume /some/local/path:/invokeai {...etc...}
+```
+
+`/some/local/path/invokeai` will contain all your data.
+It can *usually* be reused between different installs of Invoke. Tread with caution and read the release notes!
 
 ## Customize the container
 
-We ship the `run.sh` script, which is a convenient wrapper around `docker compose` for cases where custom image build args are needed. Alternatively, the familiar `docker compose` commands work just as well.
+The included `run.sh` script is a convenience wrapper around `docker compose`. It can be helpful for passing additional build arguments to `docker compose`. Alternatively, the familiar `docker compose` commands work just as well.
 
 ```bash
 cd docker
@@ -38,11 +48,14 @@ cp .env.sample .env
 
 It will take a few minutes to build the image the first time. Once the application starts up, open `http://localhost:9090` in your browser to invoke!
 
+>[!TIP]
+>When using the `run.sh` script, the container will continue running after Ctrl+C. To shut it down, use the `docker compose down` command.
+
 ## Docker setup in detail
 
 #### Linux
 
-1. Ensure builkit is enabled in the Docker daemon settings (`/etc/docker/daemon.json`)
+1. Ensure buildkit is enabled in the Docker daemon settings (`/etc/docker/daemon.json`)
 2. Install the `docker compose` plugin using your package manager, or follow a [tutorial](https://docs.docker.com/compose/install/linux/#install-using-the-repository).
     - The deprecated `docker-compose` (hyphenated) CLI probably won't work. Update to a recent version.
 3. Ensure docker daemon is able to access the GPU.
@@ -98,25 +111,7 @@ GPU_DRIVER=cuda
 
 Any environment variables supported by InvokeAI can be set here. See the [Configuration docs](https://invoke-ai.github.io/InvokeAI/features/CONFIGURATION/) for further detail.
 
-## Even More Customizing!
+---
 
-See the `docker-compose.yml` file. The `command` instruction can be uncommented and used to run arbitrary startup commands. Some examples below.
-
-### Reconfigure the runtime directory
-
-Can be used to download additional models from the supported model list
-
-In conjunction with `INVOKEAI_ROOT` can be also used to initialize a runtime directory
-
-```yaml
-command:
-  - invokeai-configure
-  - --yes
-```
-
-Or install models:
-
-```yaml
-command:
-  - invokeai-model-install
-```
+[nvidia docker docs]: https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html
+[amd docker docs]: https://rocm.docs.amd.com/projects/install-on-linux/en/latest/how-to/docker.html

docs/help/FAQ.md

@@ -196,6 +196,22 @@ tips to reduce the problem:
     === "12GB VRAM GPU"
 
         This should be sufficient to generate larger images up to about 1280x1280.
+		
+## Checkpoint Models Load Slowly or Use Too Much RAM
+
+The difference between diffusers models (a folder containing multiple
+subfolders) and checkpoint models (a file ending with .safetensors or
+.ckpt) is that InvokeAI is able to load diffusers models into memory
+incrementally, while checkpoint models must be loaded all at
+once. With very large models, or systems with limited RAM, you may
+experience slowdowns and other memory-related issues when loading
+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
+checkpoint into a diffusers model, after which loading should be
+faster and less memory-intensive.
 
 ## Memory Leak (Linux)
 

installer/templates/invoke.sh.in

@@ -17,7 +17,7 @@
 set -eu
 
 # Ensure we're in the correct folder in case user's CWD is somewhere else
-scriptdir=$(dirname "$0")
+scriptdir=$(dirname $(readlink -f "$0"))
 cd "$scriptdir"
 
 . .venv/bin/activate

invokeai/app/api/dependencies.py

@@ -1,5 +1,6 @@
 # Copyright (c) 2022 Kyle Schouviller (https://github.com/kyle0654)
 
+import asyncio
 from logging import Logger
 
 import torch
@@ -31,6 +32,8 @@ from invokeai.app.services.session_processor.session_processor_default import (
 )
 from invokeai.app.services.session_queue.session_queue_sqlite import SqliteSessionQueue
 from invokeai.app.services.shared.sqlite.sqlite_util import init_db
+from invokeai.app.services.style_preset_images.style_preset_images_disk import StylePresetImageFileStorageDisk
+from invokeai.app.services.style_preset_records.style_preset_records_sqlite import SqliteStylePresetRecordsStorage
 from invokeai.app.services.urls.urls_default import LocalUrlService
 from invokeai.app.services.workflow_records.workflow_records_sqlite import SqliteWorkflowRecordsStorage
 from invokeai.backend.stable_diffusion.diffusion.conditioning_data import ConditioningFieldData
@@ -63,7 +66,12 @@ class ApiDependencies:
     invoker: Invoker
 
     @staticmethod
-    def initialize(config: InvokeAIAppConfig, event_handler_id: int, logger: Logger = logger) -> None:
+    def initialize(
+        config: InvokeAIAppConfig,
+        event_handler_id: int,
+        loop: asyncio.AbstractEventLoop,
+        logger: Logger = logger,
+    ) -> None:
         logger.info(f"InvokeAI version {__version__}")
         logger.info(f"Root directory = {str(config.root_path)}")
 
@@ -74,6 +82,7 @@ class ApiDependencies:
         image_files = DiskImageFileStorage(f"{output_folder}/images")
 
         model_images_folder = config.models_path
+        style_presets_folder = config.style_presets_path
 
         db = init_db(config=config, logger=logger, image_files=image_files)
 
@@ -84,7 +93,7 @@ class ApiDependencies:
         board_images = BoardImagesService()
         board_records = SqliteBoardRecordStorage(db=db)
         boards = BoardService()
-        events = FastAPIEventService(event_handler_id)
+        events = FastAPIEventService(event_handler_id, loop=loop)
         bulk_download = BulkDownloadService()
         image_records = SqliteImageRecordStorage(db=db)
         images = ImageService()
@@ -109,6 +118,8 @@ class ApiDependencies:
         session_queue = SqliteSessionQueue(db=db)
         urls = LocalUrlService()
         workflow_records = SqliteWorkflowRecordsStorage(db=db)
+        style_preset_records = SqliteStylePresetRecordsStorage(db=db)
+        style_preset_image_files = StylePresetImageFileStorageDisk(style_presets_folder / "images")
 
         services = InvocationServices(
             board_image_records=board_image_records,
@@ -134,6 +145,8 @@ class ApiDependencies:
             workflow_records=workflow_records,
             tensors=tensors,
             conditioning=conditioning,
+            style_preset_records=style_preset_records,
+            style_preset_image_files=style_preset_image_files,
         )
 
         ApiDependencies.invoker = Invoker(services)

invokeai/app/api/routers/images.py

@@ -218,9 +218,8 @@ async def get_image_workflow(
         raise HTTPException(status_code=404)
 
 
-@images_router.api_route(
+@images_router.get(
     "/i/{image_name}/full",
-    methods=["GET", "HEAD"],
     operation_id="get_image_full",
     response_class=Response,
     responses={
@@ -231,6 +230,18 @@ async def get_image_workflow(
         404: {"description": "Image not found"},
     },
 )
+@images_router.head(
+    "/i/{image_name}/full",
+    operation_id="get_image_full_head",
+    response_class=Response,
+    responses={
+        200: {
+            "description": "Return the full-resolution image",
+            "content": {"image/png": {}},
+        },
+        404: {"description": "Image not found"},
+    },
+)
 async def get_image_full(
     image_name: str = Path(description="The name of full-resolution image file to get"),
 ) -> Response:
@@ -242,6 +253,7 @@ async def get_image_full(
             content = f.read()
         response = Response(content, media_type="image/png")
         response.headers["Cache-Control"] = f"max-age={IMAGE_MAX_AGE}"
+        response.headers["Content-Disposition"] = f'inline; filename="{image_name}"'
         return response
     except Exception:
         raise HTTPException(status_code=404)

invokeai/app/api/routers/model_manager.py

@@ -3,10 +3,12 @@
 
 import io
 import pathlib
+import shutil
 import traceback
 from copy import deepcopy
+from enum import Enum
 from tempfile import TemporaryDirectory
-from typing import Any, Dict, List, Optional, Type
+from typing import List, Optional, Type
 
 from fastapi import Body, Path, Query, Response, UploadFile
 from fastapi.responses import FileResponse, HTMLResponse
@@ -17,6 +19,7 @@ from starlette.exceptions import HTTPException
 from typing_extensions import Annotated
 
 from invokeai.app.api.dependencies import ApiDependencies
+from invokeai.app.services.config import get_config
 from invokeai.app.services.model_images.model_images_common import ModelImageFileNotFoundException
 from invokeai.app.services.model_install.model_install_common import ModelInstallJob
 from invokeai.app.services.model_records import (
@@ -31,6 +34,7 @@ from invokeai.backend.model_manager.config import (
     ModelFormat,
     ModelType,
 )
+from invokeai.backend.model_manager.load.model_cache.model_cache_base import CacheStats
 from invokeai.backend.model_manager.metadata.fetch.huggingface import HuggingFaceMetadataFetch
 from invokeai.backend.model_manager.metadata.metadata_base import ModelMetadataWithFiles, UnknownMetadataException
 from invokeai.backend.model_manager.search import ModelSearch
@@ -50,6 +54,13 @@ class ModelsList(BaseModel):
     model_config = ConfigDict(use_enum_values=True)
 
 
+class CacheType(str, Enum):
+    """Cache type - one of vram or ram."""
+
+    RAM = "RAM"
+    VRAM = "VRAM"
+
+
 def add_cover_image_to_model_config(config: AnyModelConfig, dependencies: Type[ApiDependencies]) -> AnyModelConfig:
     """Add a cover image URL to a model configuration."""
     cover_image = dependencies.invoker.services.model_images.get_url(config.key)
@@ -430,13 +441,11 @@ async def delete_model_image(
 async def install_model(
     source: str = Query(description="Model source to install, can be a local path, repo_id, or remote URL"),
     inplace: Optional[bool] = Query(description="Whether or not to install a local model in place", default=False),
-    # TODO(MM2): Can we type this?
-    config: Optional[Dict[str, Any]] = Body(
-        description="Dict of fields that override auto-probed values in the model config record, such as name, description and prediction_type ",
-        default=None,
+    access_token: Optional[str] = Query(description="access token for the remote resource", default=None),
+    config: ModelRecordChanges = Body(
+        description="Object containing fields that override auto-probed values in the model config record, such as name, description and prediction_type ",
         example={"name": "string", "description": "string"},
     ),
-    access_token: Optional[str] = None,
 ) -> ModelInstallJob:
     """Install a model using a string identifier.
 
@@ -451,8 +460,9 @@ async def install_model(
        - model/name:fp16:path/to/model.safetensors
        - model/name::path/to/model.safetensors
 
-    `config` is an optional dict containing model configuration values that will override
-    the ones that are probed automatically.
+    `config` is a ModelRecordChanges object. Fields in this object will override
+    the ones that are probed automatically. Pass an empty object to accept
+    all the defaults.
 
     `access_token` is an optional access token for use with Urls that require
     authentication.
@@ -737,7 +747,7 @@ async def convert_model(
         # write the converted file to the convert path
         raw_model = converted_model.model
         assert hasattr(raw_model, "save_pretrained")
-        raw_model.save_pretrained(convert_path)
+        raw_model.save_pretrained(convert_path)  # type: ignore
         assert convert_path.exists()
 
         # temporarily rename the original safetensors file so that there is no naming conflict
@@ -750,12 +760,12 @@ async def convert_model(
         try:
             new_key = installer.install_path(
                 convert_path,
-                config={
-                    "name": original_name,
-                    "description": model_config.description,
-                    "hash": model_config.hash,
-                    "source": model_config.source,
-                },
+                config=ModelRecordChanges(
+                    name=original_name,
+                    description=model_config.description,
+                    hash=model_config.hash,
+                    source=model_config.source,
+                ),
             )
         except Exception as e:
             logger.error(str(e))
@@ -798,3 +808,83 @@ async def get_starter_models() -> list[StarterModel]:
         model.dependencies = missing_deps
 
     return starter_models
+
+
+@model_manager_router.get(
+    "/model_cache",
+    operation_id="get_cache_size",
+    response_model=float,
+    summary="Get maximum size of model manager RAM or VRAM cache.",
+)
+async def get_cache_size(cache_type: CacheType = Query(description="The cache type", default=CacheType.RAM)) -> float:
+    """Return the current RAM or VRAM cache size setting (in GB)."""
+    cache = ApiDependencies.invoker.services.model_manager.load.ram_cache
+    value = 0.0
+    if cache_type == CacheType.RAM:
+        value = cache.max_cache_size
+    elif cache_type == CacheType.VRAM:
+        value = cache.max_vram_cache_size
+    return value
+
+
+@model_manager_router.put(
+    "/model_cache",
+    operation_id="set_cache_size",
+    response_model=float,
+    summary="Set maximum size of model manager RAM or VRAM cache, optionally writing new value out to invokeai.yaml config file.",
+)
+async def set_cache_size(
+    value: float = Query(description="The new value for the maximum cache size"),
+    cache_type: CacheType = Query(description="The cache type", default=CacheType.RAM),
+    persist: bool = Query(description="Write new value out to invokeai.yaml", default=False),
+) -> float:
+    """Set the current RAM or VRAM cache size setting (in GB). ."""
+    cache = ApiDependencies.invoker.services.model_manager.load.ram_cache
+    app_config = get_config()
+    # Record initial state.
+    vram_old = app_config.vram
+    ram_old = app_config.ram
+
+    # Prepare target state.
+    vram_new = vram_old
+    ram_new = ram_old
+    if cache_type == CacheType.RAM:
+        ram_new = value
+    elif cache_type == CacheType.VRAM:
+        vram_new = value
+    else:
+        raise ValueError(f"Unexpected {cache_type=}.")
+
+    config_path = app_config.config_file_path
+    new_config_path = config_path.with_suffix(".yaml.new")
+
+    try:
+        # Try to apply the target state.
+        cache.max_vram_cache_size = vram_new
+        cache.max_cache_size = ram_new
+        app_config.ram = ram_new
+        app_config.vram = vram_new
+        if persist:
+            app_config.write_file(new_config_path)
+            shutil.move(new_config_path, config_path)
+    except Exception as e:
+        # If there was a failure, restore the initial state.
+        cache.max_cache_size = ram_old
+        cache.max_vram_cache_size = vram_old
+        app_config.ram = ram_old
+        app_config.vram = vram_old
+
+        raise RuntimeError("Failed to update cache size") from e
+    return value
+
+
+@model_manager_router.get(
+    "/stats",
+    operation_id="get_stats",
+    response_model=Optional[CacheStats],
+    summary="Get model manager RAM cache performance statistics.",
+)
+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

invokeai/app/api/routers/session_queue.py

@@ -11,6 +11,7 @@ from invokeai.app.services.session_queue.session_queue_common import (
     Batch,
     BatchStatus,
     CancelByBatchIDsResult,
+    CancelByDestinationResult,
     ClearResult,
     EnqueueBatchResult,
     PruneResult,
@@ -105,6 +106,21 @@ async def cancel_by_batch_ids(
     return ApiDependencies.invoker.services.session_queue.cancel_by_batch_ids(queue_id=queue_id, batch_ids=batch_ids)
 
 
+@session_queue_router.put(
+    "/{queue_id}/cancel_by_destination",
+    operation_id="cancel_by_destination",
+    responses={200: {"model": CancelByBatchIDsResult}},
+)
+async def cancel_by_destination(
+    queue_id: str = Path(description="The queue id to perform this operation on"),
+    destination: str = Query(description="The destination to cancel all queue items for"),
+) -> CancelByDestinationResult:
+    """Immediately cancels all queue items with the given origin"""
+    return ApiDependencies.invoker.services.session_queue.cancel_by_destination(
+        queue_id=queue_id, destination=destination
+    )
+
+
 @session_queue_router.put(
     "/{queue_id}/clear",
     operation_id="clear",

invokeai/app/api/routers/style_presets.py

@@ -0,0 +1,274 @@
+import csv
+import io
+import json
+import traceback
+from typing import Optional
+
+import pydantic
+from fastapi import APIRouter, File, Form, HTTPException, Path, Response, UploadFile
+from fastapi.responses import FileResponse
+from PIL import Image
+from pydantic import BaseModel, Field
+
+from invokeai.app.api.dependencies import ApiDependencies
+from invokeai.app.api.routers.model_manager import IMAGE_MAX_AGE
+from invokeai.app.services.style_preset_images.style_preset_images_common import StylePresetImageFileNotFoundException
+from invokeai.app.services.style_preset_records.style_preset_records_common import (
+    InvalidPresetImportDataError,
+    PresetData,
+    PresetType,
+    StylePresetChanges,
+    StylePresetNotFoundError,
+    StylePresetRecordWithImage,
+    StylePresetWithoutId,
+    UnsupportedFileTypeError,
+    parse_presets_from_file,
+)
+
+
+class StylePresetFormData(BaseModel):
+    name: str = Field(description="Preset name")
+    positive_prompt: str = Field(description="Positive prompt")
+    negative_prompt: str = Field(description="Negative prompt")
+    type: PresetType = Field(description="Preset type")
+
+
+style_presets_router = APIRouter(prefix="/v1/style_presets", tags=["style_presets"])
+
+
+@style_presets_router.get(
+    "/i/{style_preset_id}",
+    operation_id="get_style_preset",
+    responses={
+        200: {"model": StylePresetRecordWithImage},
+    },
+)
+async def get_style_preset(
+    style_preset_id: str = Path(description="The style preset to get"),
+) -> StylePresetRecordWithImage:
+    """Gets a style preset"""
+    try:
+        image = ApiDependencies.invoker.services.style_preset_image_files.get_url(style_preset_id)
+        style_preset = ApiDependencies.invoker.services.style_preset_records.get(style_preset_id)
+        return StylePresetRecordWithImage(image=image, **style_preset.model_dump())
+    except StylePresetNotFoundError:
+        raise HTTPException(status_code=404, detail="Style preset not found")
+
+
+@style_presets_router.patch(
+    "/i/{style_preset_id}",
+    operation_id="update_style_preset",
+    responses={
+        200: {"model": StylePresetRecordWithImage},
+    },
+)
+async def update_style_preset(
+    image: Optional[UploadFile] = File(description="The image file to upload", default=None),
+    style_preset_id: str = Path(description="The id of the style preset to update"),
+    data: str = Form(description="The data of the style preset to update"),
+) -> StylePresetRecordWithImage:
+    """Updates a style preset"""
+    if image is not None:
+        if not image.content_type or not image.content_type.startswith("image"):
+            raise HTTPException(status_code=415, detail="Not an image")
+
+        contents = await image.read()
+        try:
+            pil_image = Image.open(io.BytesIO(contents))
+
+        except Exception:
+            ApiDependencies.invoker.services.logger.error(traceback.format_exc())
+            raise HTTPException(status_code=415, detail="Failed to read image")
+
+        try:
+            ApiDependencies.invoker.services.style_preset_image_files.save(style_preset_id, pil_image)
+        except ValueError as e:
+            raise HTTPException(status_code=409, detail=str(e))
+    else:
+        try:
+            ApiDependencies.invoker.services.style_preset_image_files.delete(style_preset_id)
+        except StylePresetImageFileNotFoundException:
+            pass
+
+    try:
+        parsed_data = json.loads(data)
+        validated_data = StylePresetFormData(**parsed_data)
+
+        name = validated_data.name
+        type = validated_data.type
+        positive_prompt = validated_data.positive_prompt
+        negative_prompt = validated_data.negative_prompt
+
+    except pydantic.ValidationError:
+        raise HTTPException(status_code=400, detail="Invalid preset data")
+
+    preset_data = PresetData(positive_prompt=positive_prompt, negative_prompt=negative_prompt)
+    changes = StylePresetChanges(name=name, preset_data=preset_data, type=type)
+
+    style_preset_image = ApiDependencies.invoker.services.style_preset_image_files.get_url(style_preset_id)
+    style_preset = ApiDependencies.invoker.services.style_preset_records.update(
+        style_preset_id=style_preset_id, changes=changes
+    )
+    return StylePresetRecordWithImage(image=style_preset_image, **style_preset.model_dump())
+
+
+@style_presets_router.delete(
+    "/i/{style_preset_id}",
+    operation_id="delete_style_preset",
+)
+async def delete_style_preset(
+    style_preset_id: str = Path(description="The style preset to delete"),
+) -> None:
+    """Deletes a style preset"""
+    try:
+        ApiDependencies.invoker.services.style_preset_image_files.delete(style_preset_id)
+    except StylePresetImageFileNotFoundException:
+        pass
+
+    ApiDependencies.invoker.services.style_preset_records.delete(style_preset_id)
+
+
+@style_presets_router.post(
+    "/",
+    operation_id="create_style_preset",
+    responses={
+        200: {"model": StylePresetRecordWithImage},
+    },
+)
+async def create_style_preset(
+    image: Optional[UploadFile] = File(description="The image file to upload", default=None),
+    data: str = Form(description="The data of the style preset to create"),
+) -> StylePresetRecordWithImage:
+    """Creates a style preset"""
+
+    try:
+        parsed_data = json.loads(data)
+        validated_data = StylePresetFormData(**parsed_data)
+
+        name = validated_data.name
+        type = validated_data.type
+        positive_prompt = validated_data.positive_prompt
+        negative_prompt = validated_data.negative_prompt
+
+    except pydantic.ValidationError:
+        raise HTTPException(status_code=400, detail="Invalid preset data")
+
+    preset_data = PresetData(positive_prompt=positive_prompt, negative_prompt=negative_prompt)
+    style_preset = StylePresetWithoutId(name=name, preset_data=preset_data, type=type)
+    new_style_preset = ApiDependencies.invoker.services.style_preset_records.create(style_preset=style_preset)
+
+    if image is not None:
+        if not image.content_type or not image.content_type.startswith("image"):
+            raise HTTPException(status_code=415, detail="Not an image")
+
+        contents = await image.read()
+        try:
+            pil_image = Image.open(io.BytesIO(contents))
+
+        except Exception:
+            ApiDependencies.invoker.services.logger.error(traceback.format_exc())
+            raise HTTPException(status_code=415, detail="Failed to read image")
+
+        try:
+            ApiDependencies.invoker.services.style_preset_image_files.save(new_style_preset.id, pil_image)
+        except ValueError as e:
+            raise HTTPException(status_code=409, detail=str(e))
+
+    preset_image = ApiDependencies.invoker.services.style_preset_image_files.get_url(new_style_preset.id)
+    return StylePresetRecordWithImage(image=preset_image, **new_style_preset.model_dump())
+
+
+@style_presets_router.get(
+    "/",
+    operation_id="list_style_presets",
+    responses={
+        200: {"model": list[StylePresetRecordWithImage]},
+    },
+)
+async def list_style_presets() -> list[StylePresetRecordWithImage]:
+    """Gets a page of style presets"""
+    style_presets_with_image: list[StylePresetRecordWithImage] = []
+    style_presets = ApiDependencies.invoker.services.style_preset_records.get_many()
+    for preset in style_presets:
+        image = ApiDependencies.invoker.services.style_preset_image_files.get_url(preset.id)
+        style_preset_with_image = StylePresetRecordWithImage(image=image, **preset.model_dump())
+        style_presets_with_image.append(style_preset_with_image)
+
+    return style_presets_with_image
+
+
+@style_presets_router.get(
+    "/i/{style_preset_id}/image",
+    operation_id="get_style_preset_image",
+    responses={
+        200: {
+            "description": "The style preset image was fetched successfully",
+        },
+        400: {"description": "Bad request"},
+        404: {"description": "The style preset image could not be found"},
+    },
+    status_code=200,
+)
+async def get_style_preset_image(
+    style_preset_id: str = Path(description="The id of the style preset image to get"),
+) -> FileResponse:
+    """Gets an image file that previews the model"""
+
+    try:
+        path = ApiDependencies.invoker.services.style_preset_image_files.get_path(style_preset_id)
+
+        response = FileResponse(
+            path,
+            media_type="image/png",
+            filename=style_preset_id + ".png",
+            content_disposition_type="inline",
+        )
+        response.headers["Cache-Control"] = f"max-age={IMAGE_MAX_AGE}"
+        return response
+    except Exception:
+        raise HTTPException(status_code=404)
+
+
+@style_presets_router.get(
+    "/export",
+    operation_id="export_style_presets",
+    responses={200: {"content": {"text/csv": {}}, "description": "A CSV file with the requested data."}},
+    status_code=200,
+)
+async def export_style_presets():
+    # Create an in-memory stream to store the CSV data
+    output = io.StringIO()
+    writer = csv.writer(output)
+
+    # Write the header
+    writer.writerow(["name", "prompt", "negative_prompt"])
+
+    style_presets = ApiDependencies.invoker.services.style_preset_records.get_many(type=PresetType.User)
+
+    for preset in style_presets:
+        writer.writerow([preset.name, preset.preset_data.positive_prompt, preset.preset_data.negative_prompt])
+
+    csv_data = output.getvalue()
+    output.close()
+
+    return Response(
+        content=csv_data,
+        media_type="text/csv",
+        headers={"Content-Disposition": "attachment; filename=prompt_templates.csv"},
+    )
+
+
+@style_presets_router.post(
+    "/import",
+    operation_id="import_style_presets",
+)
+async def import_style_presets(file: UploadFile = File(description="The file to import")):
+    try:
+        style_presets = await parse_presets_from_file(file)
+        ApiDependencies.invoker.services.style_preset_records.create_many(style_presets)
+    except InvalidPresetImportDataError as e:
+        ApiDependencies.invoker.services.logger.error(traceback.format_exc())
+        raise HTTPException(status_code=400, detail=str(e))
+    except UnsupportedFileTypeError as e:
+        ApiDependencies.invoker.services.logger.error(traceback.format_exc())
+        raise HTTPException(status_code=415, detail=str(e))

invokeai/app/api_app.py

@@ -30,6 +30,7 @@ from invokeai.app.api.routers import (
     images,
     model_manager,
     session_queue,
+    style_presets,
     utilities,
     workflows,
 )
@@ -55,11 +56,13 @@ mimetypes.add_type("text/css", ".css")
 torch_device_name = TorchDevice.get_torch_device_name()
 logger.info(f"Using torch device: {torch_device_name}")
 
+loop = asyncio.new_event_loop()
+
 
 @asynccontextmanager
 async def lifespan(app: FastAPI):
     # Add startup event to load dependencies
-    ApiDependencies.initialize(config=app_config, event_handler_id=event_handler_id, logger=logger)
+    ApiDependencies.initialize(config=app_config, event_handler_id=event_handler_id, loop=loop, logger=logger)
     yield
     # Shut down threads
     ApiDependencies.shutdown()
@@ -106,6 +109,7 @@ app.include_router(board_images.board_images_router, prefix="/api")
 app.include_router(app_info.app_router, prefix="/api")
 app.include_router(session_queue.session_queue_router, prefix="/api")
 app.include_router(workflows.workflows_router, prefix="/api")
+app.include_router(style_presets.style_presets_router, prefix="/api")
 
 app.openapi = get_openapi_func(app)
 
@@ -184,8 +188,6 @@ def invoke_api() -> None:
 
     check_cudnn(logger)
 
-    # Start our own event loop for eventing usage
-    loop = asyncio.new_event_loop()
     config = uvicorn.Config(
         app=app,
         host=app_config.host,

invokeai/app/invocations/baseinvocation.py

@@ -20,7 +20,6 @@ from typing import (
     Type,
     TypeVar,
     Union,
-    cast,
 )
 
 import semver
@@ -80,7 +79,7 @@ class UIConfigBase(BaseModel):
     version: str = Field(
         description='The node\'s version. Should be a valid semver string e.g. "1.0.0" or "3.8.13".',
     )
-    node_pack: Optional[str] = Field(default=None, description="Whether or not this is a custom node")
+    node_pack: str = Field(description="The node pack that this node belongs to, will be 'invokeai' for built-in nodes")
     classification: Classification = Field(default=Classification.Stable, description="The node's classification")
 
     model_config = ConfigDict(
@@ -230,18 +229,16 @@ class BaseInvocation(ABC, BaseModel):
     @staticmethod
     def json_schema_extra(schema: dict[str, Any], model_class: Type[BaseInvocation]) -> None:
         """Adds various UI-facing attributes to the invocation's OpenAPI schema."""
-        uiconfig = cast(UIConfigBase | None, getattr(model_class, "UIConfig", None))
-        if uiconfig is not None:
-            if uiconfig.title is not None:
-                schema["title"] = uiconfig.title
-            if uiconfig.tags is not None:
-                schema["tags"] = uiconfig.tags
-            if uiconfig.category is not None:
-                schema["category"] = uiconfig.category
-            if uiconfig.node_pack is not None:
-                schema["node_pack"] = uiconfig.node_pack
-            schema["classification"] = uiconfig.classification
-            schema["version"] = uiconfig.version
+        if title := model_class.UIConfig.title:
+            schema["title"] = title
+        if tags := model_class.UIConfig.tags:
+            schema["tags"] = tags
+        if category := model_class.UIConfig.category:
+            schema["category"] = category
+        if node_pack := model_class.UIConfig.node_pack:
+            schema["node_pack"] = node_pack
+        schema["classification"] = model_class.UIConfig.classification
+        schema["version"] = model_class.UIConfig.version
         if "required" not in schema or not isinstance(schema["required"], list):
             schema["required"] = []
         schema["class"] = "invocation"
@@ -312,7 +309,7 @@ class BaseInvocation(ABC, BaseModel):
         json_schema_extra={"field_kind": FieldKind.NodeAttribute},
     )
 
-    UIConfig: ClassVar[Type[UIConfigBase]]
+    UIConfig: ClassVar[UIConfigBase]
 
     model_config = ConfigDict(
         protected_namespaces=(),
@@ -441,30 +438,25 @@ def invocation(
         validate_fields(cls.model_fields, invocation_type)
 
         # Add OpenAPI schema extras
-        uiconfig_name = cls.__qualname__ + ".UIConfig"
-        if not hasattr(cls, "UIConfig") or cls.UIConfig.__qualname__ != uiconfig_name:
-            cls.UIConfig = type(uiconfig_name, (UIConfigBase,), {})
-        cls.UIConfig.title = title
-        cls.UIConfig.tags = tags
-        cls.UIConfig.category = category
-        cls.UIConfig.classification = classification
-
-        # Grab the node pack's name from the module name, if it's a custom node
-        is_custom_node = cls.__module__.rsplit(".", 1)[0] == "invokeai.app.invocations"
-        if is_custom_node:
-            cls.UIConfig.node_pack = cls.__module__.split(".")[0]
-        else:
-            cls.UIConfig.node_pack = None
+        uiconfig: dict[str, Any] = {}
+        uiconfig["title"] = title
+        uiconfig["tags"] = tags
+        uiconfig["category"] = category
+        uiconfig["classification"] = classification
+        # The node pack is the module name - will be "invokeai" for built-in nodes
+        uiconfig["node_pack"] = cls.__module__.split(".")[0]
 
         if version is not None:
             try:
                 semver.Version.parse(version)
             except ValueError as e:
                 raise InvalidVersionError(f'Invalid version string for node "{invocation_type}": "{version}"') from e
-            cls.UIConfig.version = version
+            uiconfig["version"] = version
         else:
             logger.warn(f'No version specified for node "{invocation_type}", using "1.0.0"')
-            cls.UIConfig.version = "1.0.0"
+            uiconfig["version"] = "1.0.0"
+
+        cls.UIConfig = UIConfigBase(**uiconfig)
 
         if use_cache is not None:
             cls.model_fields["use_cache"].default = use_cache

invokeai/app/invocations/compel.py

@@ -80,12 +80,12 @@ class CompelInvocation(BaseInvocation):
 
         with (
             # apply all patches while the model is on the target device
-            text_encoder_info.model_on_device() as (model_state_dict, text_encoder),
+            text_encoder_info.model_on_device() as (cached_weights, text_encoder),
             tokenizer_info as tokenizer,
             ModelPatcher.apply_lora_text_encoder(
                 text_encoder,
                 loras=_lora_loader(),
-                model_state_dict=model_state_dict,
+                cached_weights=cached_weights,
             ),
             # Apply CLIP Skip after LoRA to prevent LoRA application from failing on skipped layers.
             ModelPatcher.apply_clip_skip(text_encoder, self.clip.skipped_layers),
@@ -175,13 +175,13 @@ class SDXLPromptInvocationBase:
 
         with (
             # apply all patches while the model is on the target device
-            text_encoder_info.model_on_device() as (state_dict, text_encoder),
+            text_encoder_info.model_on_device() as (cached_weights, text_encoder),
             tokenizer_info as tokenizer,
             ModelPatcher.apply_lora(
                 text_encoder,
                 loras=_lora_loader(),
                 prefix=lora_prefix,
-                model_state_dict=state_dict,
+                cached_weights=cached_weights,
             ),
             # Apply CLIP Skip after LoRA to prevent LoRA application from failing on skipped layers.
             ModelPatcher.apply_clip_skip(text_encoder, clip_field.skipped_layers),

invokeai/app/invocations/controlnet_image_processors.py

@@ -21,6 +21,8 @@ from controlnet_aux import (
 from controlnet_aux.util import HWC3, ade_palette
 from PIL import Image
 from pydantic import BaseModel, Field, field_validator, model_validator
+from transformers import pipeline
+from transformers.pipelines import DepthEstimationPipeline
 
 from invokeai.app.invocations.baseinvocation import (
     BaseInvocation,
@@ -44,13 +46,12 @@ from invokeai.app.invocations.util import validate_begin_end_step, validate_weig
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.util.controlnet_utils import CONTROLNET_MODE_VALUES, CONTROLNET_RESIZE_VALUES, heuristic_resize
 from invokeai.backend.image_util.canny import get_canny_edges
-from invokeai.backend.image_util.depth_anything import DEPTH_ANYTHING_MODELS, DepthAnythingDetector
+from invokeai.backend.image_util.depth_anything.depth_anything_pipeline import DepthAnythingPipeline
 from invokeai.backend.image_util.dw_openpose import DWPOSE_MODELS, DWOpenposeDetector
 from invokeai.backend.image_util.hed import HEDProcessor
 from invokeai.backend.image_util.lineart import LineartProcessor
 from invokeai.backend.image_util.lineart_anime import LineartAnimeProcessor
 from invokeai.backend.image_util.util import np_to_pil, pil_to_np
-from invokeai.backend.util.devices import TorchDevice
 
 
 class ControlField(BaseModel):
@@ -592,7 +593,14 @@ class ColorMapImageProcessorInvocation(ImageProcessorInvocation):
         return color_map
 
 
-DEPTH_ANYTHING_MODEL_SIZES = Literal["large", "base", "small"]
+DEPTH_ANYTHING_MODEL_SIZES = Literal["large", "base", "small", "small_v2"]
+# DepthAnything V2 Small model is licensed under Apache 2.0 but not the base and large models.
+DEPTH_ANYTHING_MODELS = {
+    "large": "LiheYoung/depth-anything-large-hf",
+    "base": "LiheYoung/depth-anything-base-hf",
+    "small": "LiheYoung/depth-anything-small-hf",
+    "small_v2": "depth-anything/Depth-Anything-V2-Small-hf",
+}
 
 
 @invocation(
@@ -600,28 +608,33 @@ DEPTH_ANYTHING_MODEL_SIZES = Literal["large", "base", "small"]
     title="Depth Anything Processor",
     tags=["controlnet", "depth", "depth anything"],
     category="controlnet",
-    version="1.1.2",
+    version="1.1.3",
 )
 class DepthAnythingImageProcessorInvocation(ImageProcessorInvocation):
     """Generates a depth map based on the Depth Anything algorithm"""
 
     model_size: DEPTH_ANYTHING_MODEL_SIZES = InputField(
-        default="small", description="The size of the depth model to use"
+        default="small_v2", description="The size of the depth model to use"
     )
     resolution: int = InputField(default=512, ge=1, description=FieldDescriptions.image_res)
 
     def run_processor(self, image: Image.Image) -> Image.Image:
-        def loader(model_path: Path):
-            return DepthAnythingDetector.load_model(
-                model_path, model_size=self.model_size, device=TorchDevice.choose_torch_device()
-            )
+        def load_depth_anything(model_path: Path):
+            depth_anything_pipeline = pipeline(model=str(model_path), task="depth-estimation", local_files_only=True)
+            assert isinstance(depth_anything_pipeline, DepthEstimationPipeline)
+            return DepthAnythingPipeline(depth_anything_pipeline)
 
         with self._context.models.load_remote_model(
-            source=DEPTH_ANYTHING_MODELS[self.model_size], loader=loader
-        ) as model:
-            depth_anything_detector = DepthAnythingDetector(model, TorchDevice.choose_torch_device())
-            processed_image = depth_anything_detector(image=image, resolution=self.resolution)
-            return processed_image
+            source=DEPTH_ANYTHING_MODELS[self.model_size], loader=load_depth_anything
+        ) as depth_anything_detector:
+            assert isinstance(depth_anything_detector, DepthAnythingPipeline)
+            depth_map = depth_anything_detector.generate_depth(image)
+
+            # Resizing to user target specified size
+            new_height = int(image.size[1] * (self.resolution / image.size[0]))
+            depth_map = depth_map.resize((self.resolution, new_height))
+
+            return depth_map
 
 
 @invocation(

invokeai/app/invocations/create_gradient_mask.py

@@ -39,7 +39,7 @@ class GradientMaskOutput(BaseInvocationOutput):
     title="Create Gradient Mask",
     tags=["mask", "denoise"],
     category="latents",
-    version="1.1.0",
+    version="1.2.0",
 )
 class CreateGradientMaskInvocation(BaseInvocation):
     """Creates mask for denoising model run."""
@@ -93,6 +93,7 @@ class CreateGradientMaskInvocation(BaseInvocation):
 
             # redistribute blur so that the original edges are 0 and blur outwards to 1
             blur_tensor = (blur_tensor - 0.5) * 2
+            blur_tensor[blur_tensor < 0] = 0.0
 
             threshold = 1 - self.minimum_denoise
 

invokeai/app/invocations/denoise_latents.py

@@ -37,9 +37,9 @@ 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 import LoRAModelRaw
-from invokeai.backend.model_manager import BaseModelType
+from invokeai.backend.model_manager import BaseModelType, ModelVariantType
 from invokeai.backend.model_patcher import ModelPatcher
-from invokeai.backend.stable_diffusion import PipelineIntermediateState, set_seamless
+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 (
     ControlNetData,
@@ -58,7 +58,15 @@ from invokeai.backend.stable_diffusion.diffusion.conditioning_data import (
 from invokeai.backend.stable_diffusion.diffusion.custom_atttention import CustomAttnProcessor2_0
 from invokeai.backend.stable_diffusion.diffusion_backend import StableDiffusionBackend
 from invokeai.backend.stable_diffusion.extension_callback_type import ExtensionCallbackType
+from invokeai.backend.stable_diffusion.extensions.controlnet import ControlNetExt
+from invokeai.backend.stable_diffusion.extensions.freeu import FreeUExt
+from invokeai.backend.stable_diffusion.extensions.inpaint import InpaintExt
+from invokeai.backend.stable_diffusion.extensions.inpaint_model import InpaintModelExt
+from invokeai.backend.stable_diffusion.extensions.lora import LoRAExt
 from invokeai.backend.stable_diffusion.extensions.preview import PreviewExt
+from invokeai.backend.stable_diffusion.extensions.rescale_cfg import RescaleCFGExt
+from invokeai.backend.stable_diffusion.extensions.seamless import SeamlessExt
+from invokeai.backend.stable_diffusion.extensions.t2i_adapter import T2IAdapterExt
 from invokeai.backend.stable_diffusion.extensions_manager import ExtensionsManager
 from invokeai.backend.stable_diffusion.schedulers import SCHEDULER_MAP
 from invokeai.backend.stable_diffusion.schedulers.schedulers import SCHEDULER_NAME_VALUES
@@ -177,7 +185,7 @@ class DenoiseLatentsInvocation(BaseInvocation):
     )
     denoise_mask: Optional[DenoiseMaskField] = InputField(
         default=None,
-        description=FieldDescriptions.mask,
+        description=FieldDescriptions.denoise_mask,
         input=Input.Connection,
         ui_order=8,
     )
@@ -463,6 +471,65 @@ class DenoiseLatentsInvocation(BaseInvocation):
 
         return controlnet_data
 
+    @staticmethod
+    def parse_controlnet_field(
+        exit_stack: ExitStack,
+        context: InvocationContext,
+        control_input: ControlField | list[ControlField] | None,
+        ext_manager: ExtensionsManager,
+    ) -> None:
+        # Normalize control_input to a list.
+        control_list: list[ControlField]
+        if isinstance(control_input, ControlField):
+            control_list = [control_input]
+        elif isinstance(control_input, list):
+            control_list = control_input
+        elif control_input is None:
+            control_list = []
+        else:
+            raise ValueError(f"Unexpected control_input type: {type(control_input)}")
+
+        for control_info in control_list:
+            model = exit_stack.enter_context(context.models.load(control_info.control_model))
+            ext_manager.add_extension(
+                ControlNetExt(
+                    model=model,
+                    image=context.images.get_pil(control_info.image.image_name),
+                    weight=control_info.control_weight,
+                    begin_step_percent=control_info.begin_step_percent,
+                    end_step_percent=control_info.end_step_percent,
+                    control_mode=control_info.control_mode,
+                    resize_mode=control_info.resize_mode,
+                )
+            )
+
+    @staticmethod
+    def parse_t2i_adapter_field(
+        exit_stack: ExitStack,
+        context: InvocationContext,
+        t2i_adapters: Optional[Union[T2IAdapterField, list[T2IAdapterField]]],
+        ext_manager: ExtensionsManager,
+    ) -> None:
+        if t2i_adapters is None:
+            return
+
+        # Handle the possibility that t2i_adapters could be a list or a single T2IAdapterField.
+        if isinstance(t2i_adapters, T2IAdapterField):
+            t2i_adapters = [t2i_adapters]
+
+        for t2i_adapter_field in t2i_adapters:
+            ext_manager.add_extension(
+                T2IAdapterExt(
+                    node_context=context,
+                    model_id=t2i_adapter_field.t2i_adapter_model,
+                    image=context.images.get_pil(t2i_adapter_field.image.image_name),
+                    weight=t2i_adapter_field.weight,
+                    begin_step_percent=t2i_adapter_field.begin_step_percent,
+                    end_step_percent=t2i_adapter_field.end_step_percent,
+                    resize_mode=t2i_adapter_field.resize_mode,
+                )
+            )
+
     def prep_ip_adapter_image_prompts(
         self,
         context: InvocationContext,
@@ -672,7 +739,7 @@ class DenoiseLatentsInvocation(BaseInvocation):
         else:
             masked_latents = torch.where(mask < 0.5, 0.0, latents)
 
-        return 1 - mask, masked_latents, self.denoise_mask.gradient
+        return mask, masked_latents, self.denoise_mask.gradient
 
     @staticmethod
     def prepare_noise_and_latents(
@@ -730,10 +797,6 @@ class DenoiseLatentsInvocation(BaseInvocation):
         dtype = TorchDevice.choose_torch_dtype()
 
         seed, noise, latents = self.prepare_noise_and_latents(context, self.noise, self.latents)
-        latents = latents.to(device=device, dtype=dtype)
-        if noise is not None:
-            noise = noise.to(device=device, dtype=dtype)
-
         _, _, latent_height, latent_width = latents.shape
 
         conditioning_data = self.get_conditioning_data(
@@ -766,6 +829,52 @@ class DenoiseLatentsInvocation(BaseInvocation):
             denoising_end=self.denoising_end,
         )
 
+        # get the unet's config so that we can pass the base to sd_step_callback()
+        unet_config = context.models.get_config(self.unet.unet.key)
+
+        ### preview
+        def step_callback(state: PipelineIntermediateState) -> None:
+            context.util.sd_step_callback(state, unet_config.base)
+
+        ext_manager.add_extension(PreviewExt(step_callback))
+
+        ### cfg rescale
+        if self.cfg_rescale_multiplier > 0:
+            ext_manager.add_extension(RescaleCFGExt(self.cfg_rescale_multiplier))
+
+        ### freeu
+        if self.unet.freeu_config:
+            ext_manager.add_extension(FreeUExt(self.unet.freeu_config))
+
+        ### lora
+        if self.unet.loras:
+            for lora_field in self.unet.loras:
+                ext_manager.add_extension(
+                    LoRAExt(
+                        node_context=context,
+                        model_id=lora_field.lora,
+                        weight=lora_field.weight,
+                    )
+                )
+        ### seamless
+        if self.unet.seamless_axes:
+            ext_manager.add_extension(SeamlessExt(self.unet.seamless_axes))
+
+        ### inpaint
+        mask, masked_latents, is_gradient_mask = self.prep_inpaint_mask(context, latents)
+        # NOTE: We used to identify inpainting models by inpecting the shape of the loaded UNet model weights. Now we
+        # use the ModelVariantType config. During testing, there was a report of a user with models that had an
+        # incorrect ModelVariantType value. Re-installing the model fixed the issue. If this issue turns out to be
+        # prevalent, we will have to revisit how we initialize the inpainting extensions.
+        if unet_config.variant == ModelVariantType.Inpaint:
+            ext_manager.add_extension(InpaintModelExt(mask, masked_latents, is_gradient_mask))
+        elif mask is not None:
+            ext_manager.add_extension(InpaintExt(mask, is_gradient_mask))
+
+        # Initialize context for modular denoise
+        latents = latents.to(device=device, dtype=dtype)
+        if noise is not None:
+            noise = noise.to(device=device, dtype=dtype)
         denoise_ctx = DenoiseContext(
             inputs=DenoiseInputs(
                 orig_latents=latents,
@@ -781,31 +890,31 @@ class DenoiseLatentsInvocation(BaseInvocation):
             scheduler=scheduler,
         )
 
-        # get the unet's config so that we can pass the base to sd_step_callback()
-        unet_config = context.models.get_config(self.unet.unet.key)
+        # context for loading additional models
+        with ExitStack() as exit_stack:
+            # later should be smth like:
+            # for extension_field in self.extensions:
+            #    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)
 
-        ### preview
-        def step_callback(state: PipelineIntermediateState) -> None:
-            context.util.sd_step_callback(state, unet_config.base)
+            # ext: t2i/ip adapter
+            ext_manager.run_callback(ExtensionCallbackType.SETUP, denoise_ctx)
 
-        ext_manager.add_extension(PreviewExt(step_callback))
-
-        # ext: t2i/ip adapter
-        ext_manager.run_callback(ExtensionCallbackType.SETUP, denoise_ctx)
-
-        unet_info = context.models.load(self.unet.unet)
-        assert isinstance(unet_info.model, UNet2DConditionModel)
-        with (
-            unet_info.model_on_device() as (model_state_dict, unet),
-            ModelPatcher.patch_unet_attention_processor(unet, denoise_ctx.inputs.attention_processor_cls),
-            # ext: controlnet
-            ext_manager.patch_extensions(unet),
-            # ext: freeu, seamless, ip adapter, lora
-            ext_manager.patch_unet(model_state_dict, unet),
-        ):
-            sd_backend = StableDiffusionBackend(unet, scheduler)
-            denoise_ctx.unet = unet
-            result_latents = sd_backend.latents_from_embeddings(denoise_ctx, ext_manager)
+            unet_info = context.models.load(self.unet.unet)
+            assert isinstance(unet_info.model, UNet2DConditionModel)
+            with (
+                unet_info.model_on_device() as (cached_weights, unet),
+                ModelPatcher.patch_unet_attention_processor(unet, denoise_ctx.inputs.attention_processor_cls),
+                # ext: controlnet
+                ext_manager.patch_extensions(denoise_ctx),
+                # ext: freeu, seamless, ip adapter, lora
+                ext_manager.patch_unet(unet, cached_weights),
+            ):
+                sd_backend = StableDiffusionBackend(unet, scheduler)
+                denoise_ctx.unet = unet
+                result_latents = sd_backend.latents_from_embeddings(denoise_ctx, ext_manager)
 
         # https://discuss.huggingface.co/t/memory-usage-by-later-pipeline-stages/23699
         result_latents = result_latents.detach().to("cpu")
@@ -820,6 +929,10 @@ class DenoiseLatentsInvocation(BaseInvocation):
         seed, noise, latents = self.prepare_noise_and_latents(context, self.noise, self.latents)
 
         mask, masked_latents, gradient_mask = self.prep_inpaint_mask(context, latents)
+        # At this point, the mask ranges from 0 (leave unchanged) to 1 (inpaint).
+        # We invert the mask here for compatibility with the old backend implementation.
+        if mask is not None:
+            mask = 1 - mask
 
         # TODO(ryand): I have hard-coded `do_classifier_free_guidance=True` to mirror the behaviour of ControlNets,
         # below. Investigate whether this is appropriate.
@@ -862,14 +975,14 @@ class DenoiseLatentsInvocation(BaseInvocation):
         assert isinstance(unet_info.model, UNet2DConditionModel)
         with (
             ExitStack() as exit_stack,
-            unet_info.model_on_device() as (model_state_dict, unet),
+            unet_info.model_on_device() as (cached_weights, unet),
             ModelPatcher.apply_freeu(unet, self.unet.freeu_config),
-            set_seamless(unet, self.unet.seamless_axes),  # FIXME
+            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.
             ModelPatcher.apply_lora_unet(
                 unet,
                 loras=_lora_loader(),
-                model_state_dict=model_state_dict,
+                cached_weights=cached_weights,
             ),
         ):
             assert isinstance(unet, UNet2DConditionModel)

invokeai/app/invocations/fields.py

@@ -1,7 +1,7 @@
 from enum import Enum
 from typing import Any, Callable, Optional, Tuple
 
-from pydantic import BaseModel, ConfigDict, Field, RootModel, TypeAdapter
+from pydantic import BaseModel, ConfigDict, Field, RootModel, TypeAdapter, model_validator
 from pydantic.fields import _Unset
 from pydantic_core import PydanticUndefined
 
@@ -40,14 +40,18 @@ class UIType(str, Enum, metaclass=MetaEnum):
 
     # region Model Field Types
     MainModel = "MainModelField"
+    FluxMainModel = "FluxMainModelField"
     SDXLMainModel = "SDXLMainModelField"
     SDXLRefinerModel = "SDXLRefinerModelField"
     ONNXModel = "ONNXModelField"
     VAEModel = "VAEModelField"
+    FluxVAEModel = "FluxVAEModelField"
     LoRAModel = "LoRAModelField"
     ControlNetModel = "ControlNetModelField"
     IPAdapterModel = "IPAdapterModelField"
     T2IAdapterModel = "T2IAdapterModelField"
+    T5EncoderModel = "T5EncoderModelField"
+    CLIPEmbedModel = "CLIPEmbedModelField"
     SpandrelImageToImageModel = "SpandrelImageToImageModelField"
     # endregion
 
@@ -125,13 +129,17 @@ class FieldDescriptions:
     negative_cond = "Negative conditioning tensor"
     noise = "Noise tensor"
     clip = "CLIP (tokenizer, text encoder, LoRAs) and skipped layer count"
+    t5_encoder = "T5 tokenizer and text encoder"
+    clip_embed_model = "CLIP Embed loader"
     unet = "UNet (scheduler, LoRAs)"
+    transformer = "Transformer"
     vae = "VAE"
     cond = "Conditioning tensor"
     controlnet_model = "ControlNet model to load"
     vae_model = "VAE model to load"
     lora_model = "LoRA model to load"
     main_model = "Main model (UNet, VAE, CLIP) to load"
+    flux_model = "Flux model (Transformer) 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"
@@ -173,7 +181,7 @@ class FieldDescriptions:
     )
     num_1 = "The first number"
     num_2 = "The second number"
-    mask = "The mask to use for the operation"
+    denoise_mask = "A mask of the region to apply the denoising process to."
     board = "The board to save the image to"
     image = "The image to process"
     tile_size = "Tile size"
@@ -231,6 +239,12 @@ class ColorField(BaseModel):
         return (self.r, self.g, self.b, self.a)
 
 
+class FluxConditioningField(BaseModel):
+    """A conditioning tensor primitive value"""
+
+    conditioning_name: str = Field(description="The name of conditioning tensor")
+
+
 class ConditioningField(BaseModel):
     """A conditioning tensor primitive value"""
 
@@ -242,6 +256,31 @@ class ConditioningField(BaseModel):
     )
 
 
+class BoundingBoxField(BaseModel):
+    """A bounding box primitive value."""
+
+    x_min: int = Field(ge=0, description="The minimum x-coordinate of the bounding box (inclusive).")
+    x_max: int = Field(ge=0, description="The maximum x-coordinate of the bounding box (exclusive).")
+    y_min: int = Field(ge=0, description="The minimum y-coordinate of the bounding box (inclusive).")
+    y_max: int = Field(ge=0, description="The maximum y-coordinate of the bounding box (exclusive).")
+
+    score: Optional[float] = Field(
+        default=None,
+        ge=0.0,
+        le=1.0,
+        description="The score associated with the bounding box. In the range [0, 1]. This value is typically set "
+        "when the bounding box was produced by a detector and has an associated confidence score.",
+    )
+
+    @model_validator(mode="after")
+    def check_coords(self):
+        if self.x_min > self.x_max:
+            raise ValueError(f"x_min ({self.x_min}) is greater than x_max ({self.x_max}).")
+        if self.y_min > self.y_max:
+            raise ValueError(f"y_min ({self.y_min}) is greater than y_max ({self.y_max}).")
+        return self
+
+
 class MetadataField(RootModel[dict[str, Any]]):
     """
     Pydantic model for metadata with custom root of type dict[str, Any].

invokeai/app/invocations/flux_denoise.py

@@ -0,0 +1,249 @@
+from typing import Callable, Optional
+
+import torch
+import torchvision.transforms as tv_transforms
+from torchvision.transforms.functional import resize as tv_resize
+
+from invokeai.app.invocations.baseinvocation import BaseInvocation, Classification, invocation
+from invokeai.app.invocations.fields import (
+    DenoiseMaskField,
+    FieldDescriptions,
+    FluxConditioningField,
+    Input,
+    InputField,
+    LatentsField,
+    WithBoard,
+    WithMetadata,
+)
+from invokeai.app.invocations.model import TransformerField
+from invokeai.app.invocations.primitives import LatentsOutput
+from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.flux.denoise import denoise
+from invokeai.backend.flux.inpaint_extension import InpaintExtension
+from invokeai.backend.flux.model import Flux
+from invokeai.backend.flux.sampling_utils import (
+    clip_timestep_schedule,
+    generate_img_ids,
+    get_noise,
+    get_schedule,
+    pack,
+    unpack,
+)
+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
+
+
+@invocation(
+    "flux_denoise",
+    title="FLUX Denoise",
+    tags=["image", "flux"],
+    category="image",
+    version="1.0.0",
+    classification=Classification.Prototype,
+)
+class FluxDenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
+    """Run denoising process with a FLUX transformer 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.flux_model,
+        input=Input.Connection,
+        title="Transformer",
+    )
+    positive_text_conditioning: FluxConditioningField = InputField(
+        description=FieldDescriptions.positive_cond, input=Input.Connection
+    )
+    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.")
+    num_steps: int = InputField(
+        default=4, description="Number of diffusion steps. Recommended values are schnell: 4, dev: 50."
+    )
+    guidance: float = InputField(
+        default=4.0,
+        description="The guidance strength. Higher values adhere more strictly to the prompt, and will produce less diverse images. FLUX dev only, ignored for schnell.",
+    )
+    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 _run_diffusion(
+        self,
+        context: InvocationContext,
+    ):
+        inference_dtype = torch.bfloat16
+
+        # Load the conditioning data.
+        cond_data = context.conditioning.load(self.positive_text_conditioning.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=inference_dtype)
+        t5_embeddings = flux_conditioning.t5_embeds
+        clip_embeddings = flux_conditioning.clip_embeds
+
+        # 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=TorchDevice.choose_torch_device(), dtype=inference_dtype)
+
+        # Prepare input noise.
+        noise = get_noise(
+            num_samples=1,
+            height=self.height,
+            width=self.width,
+            device=TorchDevice.choose_torch_device(),
+            dtype=inference_dtype,
+            seed=self.seed,
+        )
+
+        transformer_info = context.models.load(self.transformer.transformer)
+        is_schnell = "schnell" in 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,
+            shift=not is_schnell,
+        )
+
+        # Clip the timesteps schedule based on denoising_start and denoising_end.
+        timesteps = clip_timestep_schedule(timesteps, self.denoising_start, self.denoising_end)
+
+        # Prepare input latent image.
+        if init_latents is not None:
+            # If init_latents is provided, we are doing image-to-image.
+
+            if is_schnell:
+                context.logger.warning(
+                    "Running image-to-image with a FLUX schnell model. This is not recommended. The results are likely "
+                    "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
+        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.")
+
+            x = 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 x
+
+        inpaint_mask = self._prep_inpaint_mask(context, x)
+
+        b, _c, h, w = x.shape
+        img_ids = generate_img_ids(h=h, w=w, batch_size=b, device=x.device, dtype=x.dtype)
+
+        bs, t5_seq_len, _ = t5_embeddings.shape
+        txt_ids = torch.zeros(bs, 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
+        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]
+
+        # Prepare inpaint extension.
+        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,
+            )
+
+        with transformer_info as transformer:
+            assert isinstance(transformer, Flux)
+
+            x = denoise(
+                model=transformer,
+                img=x,
+                img_ids=img_ids,
+                txt=t5_embeddings,
+                txt_ids=txt_ids,
+                vec=clip_embeddings,
+                timesteps=timesteps,
+                step_callback=self._build_step_callback(context),
+                guidance=self.guidance,
+                inpaint_extension=inpaint_extension,
+            )
+
+        x = unpack(x.float(), self.height, self.width)
+        return x
+
+    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
+        - Expands mask to the same shape as latents so that they line up after 'packing'
+
+        Args:
+            context (InvocationContext): The invocation context, for loading the inpaint mask.
+            latents (torch.Tensor): A latent image tensor. In 'unpacked' format. Used to determine the target shape,
+                device, and dtype for the inpaint mask.
+
+        Returns:
+            torch.Tensor | None: Inpaint mask.
+        """
+        if self.denoise_mask is None:
+            return None
+
+        mask = context.tensors.load(self.denoise_mask.mask_name)
+
+        _, _, 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)
+
+        # Expand the inpaint mask to the same shape as `latents` so that when we 'pack' `mask` it lines up with
+        # `latents`.
+        return mask.expand_as(latents)
+
+    def _build_step_callback(self, context: InvocationContext) -> Callable[[PipelineIntermediateState], None]:
+        def step_callback(state: PipelineIntermediateState) -> None:
+            state.latents = unpack(state.latents.float(), self.height, self.width).squeeze()
+            context.util.flux_step_callback(state)
+
+        return step_callback

invokeai/app/invocations/flux_text_encoder.py

@@ -0,0 +1,92 @@
+from typing import Literal
+
+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.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.stable_diffusion.diffusion.conditioning_data import ConditioningFieldData, FLUXConditioningInfo
+
+
+@invocation(
+    "flux_text_encoder",
+    title="FLUX Text Encoding",
+    tags=["prompt", "conditioning", "flux"],
+    category="conditioning",
+    version="1.0.0",
+    classification=Classification.Prototype,
+)
+class FluxTextEncoderInvocation(BaseInvocation):
+    """Encodes and preps a prompt for a flux image."""
+
+    clip: CLIPField = InputField(
+        title="CLIP",
+        description=FieldDescriptions.clip,
+        input=Input.Connection,
+    )
+    t5_encoder: T5EncoderField = InputField(
+        title="T5Encoder",
+        description=FieldDescriptions.t5_encoder,
+        input=Input.Connection,
+    )
+    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.")
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> FluxConditioningOutput:
+        # Note: The T5 and CLIP encoding are done in separate functions to ensure that all model references are locally
+        # scoped. This ensures that the T5 model can be freed and gc'd before loading the CLIP model (if necessary).
+        t5_embeddings = self._t5_encode(context)
+        clip_embeddings = self._clip_encode(context)
+        conditioning_data = ConditioningFieldData(
+            conditionings=[FLUXConditioningInfo(clip_embeds=clip_embeddings, t5_embeds=t5_embeddings)]
+        )
+
+        conditioning_name = context.conditioning.save(conditioning_data)
+        return FluxConditioningOutput.build(conditioning_name)
+
+    def _t5_encode(self, context: InvocationContext) -> torch.Tensor:
+        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,
+        ):
+            assert isinstance(t5_text_encoder, T5EncoderModel)
+            assert isinstance(t5_tokenizer, T5Tokenizer)
+
+            t5_encoder = HFEncoder(t5_text_encoder, t5_tokenizer, False, self.t5_max_seq_len)
+
+            prompt_embeds = t5_encoder(prompt)
+
+        assert isinstance(prompt_embeds, torch.Tensor)
+        return prompt_embeds
+
+    def _clip_encode(self, context: InvocationContext) -> torch.Tensor:
+        clip_tokenizer_info = context.models.load(self.clip.tokenizer)
+        clip_text_encoder_info = context.models.load(self.clip.text_encoder)
+
+        prompt = [self.prompt]
+
+        with (
+            clip_text_encoder_info as clip_text_encoder,
+            clip_tokenizer_info as clip_tokenizer,
+        ):
+            assert isinstance(clip_text_encoder, CLIPTextModel)
+            assert isinstance(clip_tokenizer, CLIPTokenizer)
+
+            clip_encoder = HFEncoder(clip_text_encoder, clip_tokenizer, True, 77)
+
+            pooled_prompt_embeds = clip_encoder(prompt)
+
+        assert isinstance(pooled_prompt_embeds, torch.Tensor)
+        return pooled_prompt_embeds

invokeai/app/invocations/flux_vae_decode.py

@@ -0,0 +1,60 @@
+import torch
+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.flux.modules.autoencoder import AutoEncoder
+from invokeai.backend.model_manager.load.load_base import LoadedModel
+from invokeai.backend.util.devices import TorchDevice
+
+
+@invocation(
+    "flux_vae_decode",
+    title="FLUX Latents to Image",
+    tags=["latents", "image", "vae", "l2i", "flux"],
+    category="latents",
+    version="1.0.0",
+)
+class FluxVaeDecodeInvocation(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,
+    )
+
+    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())
+            img = vae.decode(latents)
+
+        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())
+        return img_pil
+
+    @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)
+        image = self._vae_decode(vae_info=vae_info, latents=latents)
+
+        TorchDevice.empty_cache()
+        image_dto = context.images.save(image=image)
+        return ImageOutput.build(image_dto)

invokeai/app/invocations/flux_vae_encode.py

@@ -0,0 +1,67 @@
+import einops
+import torch
+
+from invokeai.app.invocations.baseinvocation import BaseInvocation, invocation
+from invokeai.app.invocations.fields import (
+    FieldDescriptions,
+    ImageField,
+    Input,
+    InputField,
+)
+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.flux.modules.autoencoder import AutoEncoder
+from invokeai.backend.model_manager import LoadedModel
+from invokeai.backend.stable_diffusion.diffusers_pipeline import image_resized_to_grid_as_tensor
+from invokeai.backend.util.devices import TorchDevice
+
+
+@invocation(
+    "flux_vae_encode",
+    title="FLUX Image to Latents",
+    tags=["latents", "image", "vae", "i2l", "flux"],
+    category="latents",
+    version="1.0.0",
+)
+class FluxVaeEncodeInvocation(BaseInvocation):
+    """Encodes an image into latents."""
+
+    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:
+        # TODO(ryand): Expose seed parameter at the invocation level.
+        # TODO(ryand): Write a util function for generating random tensors that is consistent across devices / dtypes.
+        # There's a starting point in get_noise(...), but it needs to be extracted and generalized. This function
+        # should be used for VAE encode sampling.
+        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()
+            )
+            latents = vae.encode(image_tensor, sample=True, generator=generator)
+            return latents
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> LatentsOutput:
+        image = context.images.get_pil(self.image.image_name)
+
+        vae_info = context.models.load(self.vae.vae)
+
+        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")
+
+        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/grounding_dino.py

@@ -0,0 +1,100 @@
+from pathlib import Path
+from typing import Literal
+
+import torch
+from PIL import Image
+from transformers import pipeline
+from transformers.pipelines import ZeroShotObjectDetectionPipeline
+
+from invokeai.app.invocations.baseinvocation import BaseInvocation, invocation
+from invokeai.app.invocations.fields import BoundingBoxField, ImageField, InputField
+from invokeai.app.invocations.primitives import BoundingBoxCollectionOutput
+from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.image_util.grounding_dino.detection_result import DetectionResult
+from invokeai.backend.image_util.grounding_dino.grounding_dino_pipeline import GroundingDinoPipeline
+
+GroundingDinoModelKey = Literal["grounding-dino-tiny", "grounding-dino-base"]
+GROUNDING_DINO_MODEL_IDS: dict[GroundingDinoModelKey, str] = {
+    "grounding-dino-tiny": "IDEA-Research/grounding-dino-tiny",
+    "grounding-dino-base": "IDEA-Research/grounding-dino-base",
+}
+
+
+@invocation(
+    "grounding_dino",
+    title="Grounding DINO (Text Prompt Object Detection)",
+    tags=["prompt", "object detection"],
+    category="image",
+    version="1.0.0",
+)
+class GroundingDinoInvocation(BaseInvocation):
+    """Runs a Grounding DINO model. Performs zero-shot bounding-box object detection from a text prompt."""
+
+    # Reference:
+    # - https://arxiv.org/pdf/2303.05499
+    # - https://huggingface.co/docs/transformers/v4.43.3/en/model_doc/grounding-dino#grounded-sam
+    # - https://github.com/NielsRogge/Transformers-Tutorials/blob/a39f33ac1557b02ebfb191ea7753e332b5ca933f/Grounding%20DINO/GroundingDINO_with_Segment_Anything.ipynb
+
+    model: GroundingDinoModelKey = InputField(description="The Grounding DINO model to use.")
+    prompt: str = InputField(description="The prompt describing the object to segment.")
+    image: ImageField = InputField(description="The image to segment.")
+    detection_threshold: float = InputField(
+        description="The detection threshold for the Grounding DINO model. All detected bounding boxes with scores above this threshold will be returned.",
+        ge=0.0,
+        le=1.0,
+        default=0.3,
+    )
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> BoundingBoxCollectionOutput:
+        # The model expects a 3-channel RGB image.
+        image_pil = context.images.get_pil(self.image.image_name, mode="RGB")
+
+        detections = self._detect(
+            context=context, image=image_pil, labels=[self.prompt], threshold=self.detection_threshold
+        )
+
+        # Convert detections to BoundingBoxCollectionOutput.
+        bounding_boxes: list[BoundingBoxField] = []
+        for detection in detections:
+            bounding_boxes.append(
+                BoundingBoxField(
+                    x_min=detection.box.xmin,
+                    x_max=detection.box.xmax,
+                    y_min=detection.box.ymin,
+                    y_max=detection.box.ymax,
+                    score=detection.score,
+                )
+            )
+        return BoundingBoxCollectionOutput(collection=bounding_boxes)
+
+    @staticmethod
+    def _load_grounding_dino(model_path: Path):
+        grounding_dino_pipeline = pipeline(
+            model=str(model_path),
+            task="zero-shot-object-detection",
+            local_files_only=True,
+            # TODO(ryand): Setting the torch_dtype here doesn't work. Investigate whether fp16 is supported by the
+            # model, and figure out how to make it work in the pipeline.
+            # torch_dtype=TorchDevice.choose_torch_dtype(),
+        )
+        assert isinstance(grounding_dino_pipeline, ZeroShotObjectDetectionPipeline)
+        return GroundingDinoPipeline(grounding_dino_pipeline)
+
+    def _detect(
+        self,
+        context: InvocationContext,
+        image: Image.Image,
+        labels: list[str],
+        threshold: float = 0.3,
+    ) -> list[DetectionResult]:
+        """Use Grounding DINO to detect bounding boxes for a set of labels in an image."""
+        # TODO(ryand): I copied this "."-handling logic from the transformers example code. Test it and see if it
+        # actually makes a difference.
+        labels = [label if label.endswith(".") else label + "." for label in labels]
+
+        with context.models.load_remote_model(
+            source=GROUNDING_DINO_MODEL_IDS[self.model], loader=GroundingDinoInvocation._load_grounding_dino
+        ) as detector:
+            assert isinstance(detector, GroundingDinoPipeline)
+            return detector.detect(image=image, candidate_labels=labels, threshold=threshold)

invokeai/app/invocations/image.py

@@ -6,13 +6,19 @@ import cv2
 import numpy
 from PIL import Image, ImageChops, ImageFilter, ImageOps
 
-from invokeai.app.invocations.baseinvocation import BaseInvocation, Classification, invocation
+from invokeai.app.invocations.baseinvocation import (
+    BaseInvocation,
+    Classification,
+    invocation,
+    invocation_output,
+)
 from invokeai.app.invocations.constants import IMAGE_MODES
 from invokeai.app.invocations.fields import (
     ColorField,
     FieldDescriptions,
     ImageField,
     InputField,
+    OutputField,
     WithBoard,
     WithMetadata,
 )
@@ -1007,3 +1013,62 @@ class MaskFromIDInvocation(BaseInvocation, WithMetadata, WithBoard):
         image_dto = context.images.save(image=mask, image_category=ImageCategory.MASK)
 
         return ImageOutput.build(image_dto)
+
+
+@invocation_output("canvas_v2_mask_and_crop_output")
+class CanvasV2MaskAndCropOutput(ImageOutput):
+    offset_x: int = OutputField(description="The x offset of the image, after cropping")
+    offset_y: int = OutputField(description="The y offset of the image, after cropping")
+
+
+@invocation(
+    "canvas_v2_mask_and_crop",
+    title="Canvas V2 Mask and Crop",
+    tags=["image", "mask", "id"],
+    category="image",
+    version="1.0.0",
+    classification=Classification.Prototype,
+)
+class CanvasV2MaskAndCropInvocation(BaseInvocation, WithMetadata, WithBoard):
+    """Handles Canvas V2 image output masking and cropping"""
+
+    source_image: ImageField | None = InputField(
+        default=None,
+        description="The source image onto which the masked generated image is pasted. If omitted, the masked generated image is returned with transparency.",
+    )
+    generated_image: ImageField = InputField(description="The image to apply the mask to")
+    mask: ImageField = InputField(description="The mask to apply")
+    mask_blur: int = InputField(default=0, ge=0, description="The amount to blur the mask by")
+
+    def _prepare_mask(self, mask: Image.Image) -> Image.Image:
+        mask_array = numpy.array(mask)
+        kernel = numpy.ones((self.mask_blur, self.mask_blur), numpy.uint8)
+        dilated_mask_array = cv2.erode(mask_array, kernel, iterations=3)
+        dilated_mask = Image.fromarray(dilated_mask_array)
+        if self.mask_blur > 0:
+            mask = dilated_mask.filter(ImageFilter.GaussianBlur(self.mask_blur))
+        return ImageOps.invert(mask.convert("L"))
+
+    def invoke(self, context: InvocationContext) -> CanvasV2MaskAndCropOutput:
+        mask = self._prepare_mask(context.images.get_pil(self.mask.image_name))
+
+        if self.source_image:
+            generated_image = context.images.get_pil(self.generated_image.image_name)
+            source_image = context.images.get_pil(self.source_image.image_name)
+            source_image.paste(generated_image, (0, 0), mask)
+            image_dto = context.images.save(image=source_image)
+        else:
+            generated_image = context.images.get_pil(self.generated_image.image_name)
+            generated_image.putalpha(mask)
+            image_dto = context.images.save(image=generated_image)
+
+        # bbox = image.getbbox()
+        # image = image.crop(bbox)
+
+        return CanvasV2MaskAndCropOutput(
+            image=ImageField(image_name=image_dto.image_name),
+            offset_x=0,
+            offset_y=0,
+            width=image_dto.width,
+            height=image_dto.height,
+        )

invokeai/app/invocations/latents_to_image.py

@@ -24,7 +24,7 @@ from invokeai.app.invocations.fields import (
 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 import set_seamless
+from invokeai.backend.stable_diffusion.extensions.seamless import SeamlessExt
 from invokeai.backend.stable_diffusion.vae_tiling import patch_vae_tiling_params
 from invokeai.backend.util.devices import TorchDevice
 
@@ -59,7 +59,7 @@ class LatentsToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
 
         vae_info = context.models.load(self.vae.vae)
         assert isinstance(vae_info.model, (AutoencoderKL, AutoencoderTiny))
-        with set_seamless(vae_info.model, self.vae.seamless_axes), vae_info as vae:
+        with SeamlessExt.static_patch_model(vae_info.model, self.vae.seamless_axes), vae_info as vae:
             assert isinstance(vae, (AutoencoderKL, AutoencoderTiny))
             latents = latents.to(vae.device)
             if self.fp32:

invokeai/app/invocations/mask.py

@@ -1,9 +1,10 @@
 import numpy as np
 import torch
+from PIL import Image
 
 from invokeai.app.invocations.baseinvocation import BaseInvocation, Classification, InvocationContext, invocation
-from invokeai.app.invocations.fields import ImageField, InputField, TensorField, WithMetadata
-from invokeai.app.invocations.primitives import MaskOutput
+from invokeai.app.invocations.fields import ImageField, InputField, TensorField, WithBoard, WithMetadata
+from invokeai.app.invocations.primitives import ImageOutput, MaskOutput
 
 
 @invocation(
@@ -118,3 +119,32 @@ class ImageMaskToTensorInvocation(BaseInvocation, WithMetadata):
             height=mask.shape[1],
             width=mask.shape[2],
         )
+
+
+@invocation(
+    "tensor_mask_to_image",
+    title="Tensor Mask to Image",
+    tags=["mask"],
+    category="mask",
+    version="1.1.0",
+)
+class MaskTensorToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
+    """Convert a mask tensor to an image."""
+
+    mask: TensorField = InputField(description="The mask tensor to convert.")
+
+    def invoke(self, context: InvocationContext) -> ImageOutput:
+        mask = context.tensors.load(self.mask.tensor_name)
+
+        # Squeeze the channel dimension if it exists.
+        if mask.dim() == 3:
+            mask = mask.squeeze(0)
+
+        # Ensure that the mask is binary.
+        if mask.dtype != torch.bool:
+            mask = mask > 0.5
+        mask_np = (mask.float() * 255).byte().cpu().numpy()
+
+        mask_pil = Image.fromarray(mask_np, mode="L")
+        image_dto = context.images.save(image=mask_pil)
+        return ImageOutput.build(image_dto)

invokeai/app/invocations/model.py

@@ -1,5 +1,5 @@
 import copy
-from typing import List, Optional
+from typing import List, Literal, Optional
 
 from pydantic import BaseModel, Field
 
@@ -13,7 +13,14 @@ from invokeai.app.invocations.baseinvocation import (
 from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField, UIType
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.shared.models import FreeUConfig
-from invokeai.backend.model_manager.config import AnyModelConfig, BaseModelType, ModelType, SubModelType
+from invokeai.backend.flux.util import max_seq_lengths
+from invokeai.backend.model_manager.config import (
+    AnyModelConfig,
+    BaseModelType,
+    CheckpointConfigBase,
+    ModelType,
+    SubModelType,
+)
 
 
 class ModelIdentifierField(BaseModel):
@@ -60,6 +67,15 @@ 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")
+
+
+class T5EncoderField(BaseModel):
+    tokenizer: ModelIdentifierField = Field(description="Info to load tokenizer submodel")
+    text_encoder: ModelIdentifierField = Field(description="Info to load text_encoder submodel")
+
+
 class VAEField(BaseModel):
     vae: ModelIdentifierField = Field(description="Info to load vae submodel")
     seamless_axes: List[str] = Field(default_factory=list, description='Axes("x" and "y") to which apply seamless')
@@ -122,6 +138,78 @@ 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),
+            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/primitives.py

@@ -7,10 +7,12 @@ import torch
 from invokeai.app.invocations.baseinvocation import BaseInvocation, BaseInvocationOutput, invocation, invocation_output
 from invokeai.app.invocations.constants import LATENT_SCALE_FACTOR
 from invokeai.app.invocations.fields import (
+    BoundingBoxField,
     ColorField,
     ConditioningField,
     DenoiseMaskField,
     FieldDescriptions,
+    FluxConditioningField,
     ImageField,
     Input,
     InputField,
@@ -413,6 +415,17 @@ class MaskOutput(BaseInvocationOutput):
     height: int = OutputField(description="The height of the mask in pixels.")
 
 
+@invocation_output("flux_conditioning_output")
+class FluxConditioningOutput(BaseInvocationOutput):
+    """Base class for nodes that output a single conditioning tensor"""
+
+    conditioning: FluxConditioningField = OutputField(description=FieldDescriptions.cond)
+
+    @classmethod
+    def build(cls, conditioning_name: str) -> "FluxConditioningOutput":
+        return cls(conditioning=FluxConditioningField(conditioning_name=conditioning_name))
+
+
 @invocation_output("conditioning_output")
 class ConditioningOutput(BaseInvocationOutput):
     """Base class for nodes that output a single conditioning tensor"""
@@ -469,3 +482,42 @@ class ConditioningCollectionInvocation(BaseInvocation):
 
 
 # endregion
+
+# region BoundingBox
+
+
+@invocation_output("bounding_box_output")
+class BoundingBoxOutput(BaseInvocationOutput):
+    """Base class for nodes that output a single bounding box"""
+
+    bounding_box: BoundingBoxField = OutputField(description="The output bounding box.")
+
+
+@invocation_output("bounding_box_collection_output")
+class BoundingBoxCollectionOutput(BaseInvocationOutput):
+    """Base class for nodes that output a collection of bounding boxes"""
+
+    collection: list[BoundingBoxField] = OutputField(description="The output bounding boxes.", title="Bounding Boxes")
+
+
+@invocation(
+    "bounding_box",
+    title="Bounding Box",
+    tags=["primitives", "segmentation", "collection", "bounding box"],
+    category="primitives",
+    version="1.0.0",
+)
+class BoundingBoxInvocation(BaseInvocation):
+    """Create a bounding box manually by supplying box coordinates"""
+
+    x_min: int = InputField(default=0, description="x-coordinate of the bounding box's top left vertex")
+    y_min: int = InputField(default=0, description="y-coordinate of the bounding box's top left vertex")
+    x_max: int = InputField(default=0, description="x-coordinate of the bounding box's bottom right vertex")
+    y_max: int = InputField(default=0, description="y-coordinate of the bounding box's bottom right vertex")
+
+    def invoke(self, context: InvocationContext) -> BoundingBoxOutput:
+        bounding_box = BoundingBoxField(x_min=self.x_min, y_min=self.y_min, x_max=self.x_max, y_max=self.y_max)
+        return BoundingBoxOutput(bounding_box=bounding_box)
+
+
+# endregion

invokeai/app/invocations/segment_anything.py

@@ -0,0 +1,161 @@
+from pathlib import Path
+from typing import Literal
+
+import numpy as np
+import torch
+from PIL import Image
+from transformers import AutoModelForMaskGeneration, AutoProcessor
+from transformers.models.sam import SamModel
+from transformers.models.sam.processing_sam import SamProcessor
+
+from invokeai.app.invocations.baseinvocation import BaseInvocation, invocation
+from invokeai.app.invocations.fields import BoundingBoxField, ImageField, InputField, TensorField
+from invokeai.app.invocations.primitives import MaskOutput
+from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.image_util.segment_anything.mask_refinement import mask_to_polygon, polygon_to_mask
+from invokeai.backend.image_util.segment_anything.segment_anything_pipeline import SegmentAnythingPipeline
+
+SegmentAnythingModelKey = Literal["segment-anything-base", "segment-anything-large", "segment-anything-huge"]
+SEGMENT_ANYTHING_MODEL_IDS: dict[SegmentAnythingModelKey, str] = {
+    "segment-anything-base": "facebook/sam-vit-base",
+    "segment-anything-large": "facebook/sam-vit-large",
+    "segment-anything-huge": "facebook/sam-vit-huge",
+}
+
+
+@invocation(
+    "segment_anything",
+    title="Segment Anything",
+    tags=["prompt", "segmentation"],
+    category="segmentation",
+    version="1.0.0",
+)
+class SegmentAnythingInvocation(BaseInvocation):
+    """Runs a Segment Anything Model."""
+
+    # Reference:
+    # - https://arxiv.org/pdf/2304.02643
+    # - https://huggingface.co/docs/transformers/v4.43.3/en/model_doc/grounding-dino#grounded-sam
+    # - https://github.com/NielsRogge/Transformers-Tutorials/blob/a39f33ac1557b02ebfb191ea7753e332b5ca933f/Grounding%20DINO/GroundingDINO_with_Segment_Anything.ipynb
+
+    model: SegmentAnythingModelKey = InputField(description="The Segment Anything model to use.")
+    image: ImageField = InputField(description="The image to segment.")
+    bounding_boxes: list[BoundingBoxField] = InputField(description="The bounding boxes to prompt the SAM model with.")
+    apply_polygon_refinement: bool = InputField(
+        description="Whether to apply polygon refinement to the masks. This will smooth the edges of the masks slightly and ensure that each mask consists of a single closed polygon (before merging).",
+        default=True,
+    )
+    mask_filter: Literal["all", "largest", "highest_box_score"] = InputField(
+        description="The filtering to apply to the detected masks before merging them into a final output.",
+        default="all",
+    )
+
+    @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 len(self.bounding_boxes) == 0:
+            combined_mask = torch.zeros(image_pil.size[::-1], dtype=torch.bool)
+        else:
+            masks = self._segment(context=context, image=image_pil)
+            masks = self._filter_masks(masks=masks, bounding_boxes=self.bounding_boxes)
+
+            # masks contains bool values, so we merge them via max-reduce.
+            combined_mask, _ = torch.stack(masks).max(dim=0)
+
+        mask_tensor_name = context.tensors.save(combined_mask)
+        height, width = combined_mask.shape
+        return MaskOutput(mask=TensorField(tensor_name=mask_tensor_name), width=width, height=height)
+
+    @staticmethod
+    def _load_sam_model(model_path: Path):
+        sam_model = AutoModelForMaskGeneration.from_pretrained(
+            model_path,
+            local_files_only=True,
+            # TODO(ryand): Setting the torch_dtype here doesn't work. Investigate whether fp16 is supported by the
+            # model, and figure out how to make it work in the pipeline.
+            # torch_dtype=TorchDevice.choose_torch_dtype(),
+        )
+        assert isinstance(sam_model, SamModel)
+
+        sam_processor = AutoProcessor.from_pretrained(model_path, local_files_only=True)
+        assert isinstance(sam_processor, SamProcessor)
+        return SegmentAnythingPipeline(sam_model=sam_model, sam_processor=sam_processor)
+
+    def _segment(
+        self,
+        context: InvocationContext,
+        image: Image.Image,
+    ) -> list[torch.Tensor]:
+        """Use Segment Anything (SAM) to generate masks given an image + a set of bounding boxes."""
+        # Convert the bounding boxes to the SAM input format.
+        sam_bounding_boxes = [[bb.x_min, bb.y_min, bb.x_max, bb.y_max] for bb in self.bounding_boxes]
+
+        with (
+            context.models.load_remote_model(
+                source=SEGMENT_ANYTHING_MODEL_IDS[self.model], loader=SegmentAnythingInvocation._load_sam_model
+            ) as sam_pipeline,
+        ):
+            assert isinstance(sam_pipeline, SegmentAnythingPipeline)
+            masks = sam_pipeline.segment(image=image, bounding_boxes=sam_bounding_boxes)
+
+        masks = self._process_masks(masks)
+        if self.apply_polygon_refinement:
+            masks = self._apply_polygon_refinement(masks)
+
+        return masks
+
+    def _process_masks(self, masks: torch.Tensor) -> list[torch.Tensor]:
+        """Convert the tensor output from the Segment Anything model from a tensor of shape
+        [num_masks, channels, height, width] to a list of tensors of shape [height, width].
+        """
+        assert masks.dtype == torch.bool
+        # [num_masks, channels, height, width] -> [num_masks, height, width]
+        masks, _ = masks.max(dim=1)
+        # Split the first dimension into a list of masks.
+        return list(masks.cpu().unbind(dim=0))
+
+    def _apply_polygon_refinement(self, masks: list[torch.Tensor]) -> list[torch.Tensor]:
+        """Apply polygon refinement to the masks.
+
+        Convert each mask to a polygon, then back to a mask. This has the following effect:
+        - Smooth the edges of the mask slightly.
+        - Ensure that each mask consists of a single closed polygon
+            - Removes small mask pieces.
+            - Removes holes from the mask.
+        """
+        # Convert tensor masks to np masks.
+        np_masks = [mask.cpu().numpy().astype(np.uint8) for mask in masks]
+
+        # Apply polygon refinement.
+        for idx, mask in enumerate(np_masks):
+            shape = mask.shape
+            assert len(shape) == 2  # Assert length to satisfy type checker.
+            polygon = mask_to_polygon(mask)
+            mask = polygon_to_mask(polygon, shape)
+            np_masks[idx] = mask
+
+        # Convert np masks back to tensor masks.
+        masks = [torch.tensor(mask, dtype=torch.bool) for mask in np_masks]
+
+        return masks
+
+    def _filter_masks(self, masks: list[torch.Tensor], bounding_boxes: list[BoundingBoxField]) -> list[torch.Tensor]:
+        """Filter the detected masks based on the specified mask filter."""
+        assert len(masks) == len(bounding_boxes)
+
+        if self.mask_filter == "all":
+            return masks
+        elif self.mask_filter == "largest":
+            # Find the largest mask.
+            return [max(masks, key=lambda x: float(x.sum()))]
+        elif self.mask_filter == "highest_box_score":
+            # Find the index of the bounding box with the highest score.
+            # Note that we fallback to -1.0 if the score is None. This is mainly to satisfy the type checker. In most
+            # cases the scores should all be non-None when using this filtering mode. That being said, -1.0 is a
+            # reasonable fallback since the expected score range is [0.0, 1.0].
+            max_score_idx = max(range(len(bounding_boxes)), key=lambda i: bounding_boxes[i].score or -1.0)
+            return [masks[max_score_idx]]
+        else:
+            raise ValueError(f"Invalid mask filter: {self.mask_filter}")

invokeai/app/invocations/spandrel_image_to_image.py

@@ -1,3 +1,5 @@
+from typing import Callable
+
 import numpy as np
 import torch
 from PIL import Image
@@ -21,7 +23,7 @@ from invokeai.backend.tiles.tiles import calc_tiles_min_overlap
 from invokeai.backend.tiles.utils import TBLR, Tile
 
 
-@invocation("spandrel_image_to_image", title="Image-to-Image", tags=["upscale"], category="upscale", version="1.1.0")
+@invocation("spandrel_image_to_image", title="Image-to-Image", tags=["upscale"], category="upscale", version="1.3.0")
 class SpandrelImageToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
     """Run any spandrel image-to-image model (https://github.com/chaiNNer-org/spandrel)."""
 
@@ -35,7 +37,8 @@ class SpandrelImageToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
         default=512, description="The tile size for tiled image-to-image. Set to 0 to disable tiling."
     )
 
-    def _scale_tile(self, tile: Tile, scale: int) -> Tile:
+    @classmethod
+    def scale_tile(cls, tile: Tile, scale: int) -> Tile:
         return Tile(
             coords=TBLR(
                 top=tile.coords.top * scale,
@@ -51,20 +54,22 @@ class SpandrelImageToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
             ),
         )
 
-    @torch.inference_mode()
-    def invoke(self, context: InvocationContext) -> ImageOutput:
-        # Images are converted to RGB, because most models don't support an alpha channel. In the future, we may want to
-        # revisit this.
-        image = context.images.get_pil(self.image.image_name, mode="RGB")
-
+    @classmethod
+    def upscale_image(
+        cls,
+        image: Image.Image,
+        tile_size: int,
+        spandrel_model: SpandrelImageToImageModel,
+        is_canceled: Callable[[], bool],
+    ) -> Image.Image:
         # Compute the image tiles.
-        if self.tile_size > 0:
+        if tile_size > 0:
             min_overlap = 20
             tiles = calc_tiles_min_overlap(
                 image_height=image.height,
                 image_width=image.width,
-                tile_height=self.tile_size,
-                tile_width=self.tile_size,
+                tile_height=tile_size,
+                tile_width=tile_size,
                 min_overlap=min_overlap,
             )
         else:
@@ -85,60 +90,164 @@ class SpandrelImageToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
         # Prepare input image for inference.
         image_tensor = SpandrelImageToImageModel.pil_to_tensor(image)
 
-        # Load the model.
-        spandrel_model_info = context.models.load(self.image_to_image_model)
+        # Scale the tiles for re-assembling the final image.
+        scale = spandrel_model.scale
+        scaled_tiles = [cls.scale_tile(tile, scale=scale) for tile in tiles]
+
+        # Prepare the output tensor.
+        _, channels, height, width = image_tensor.shape
+        output_tensor = torch.zeros(
+            (height * scale, width * scale, channels), dtype=torch.uint8, device=torch.device("cpu")
+        )
+
+        image_tensor = image_tensor.to(device=spandrel_model.device, dtype=spandrel_model.dtype)
 
         # Run the model on each tile.
-        with spandrel_model_info as spandrel_model:
-            assert isinstance(spandrel_model, SpandrelImageToImageModel)
+        for tile, scaled_tile in tqdm(list(zip(tiles, scaled_tiles, strict=True)), desc="Upscaling Tiles"):
+            # Exit early if the invocation has been canceled.
+            if is_canceled():
+                raise CanceledException
 
-            # Scale the tiles for re-assembling the final image.
-            scale = spandrel_model.scale
-            scaled_tiles = [self._scale_tile(tile, scale=scale) for tile in tiles]
+            # Extract the current tile from the input tensor.
+            input_tile = image_tensor[
+                :, :, tile.coords.top : tile.coords.bottom, tile.coords.left : tile.coords.right
+            ].to(device=spandrel_model.device, dtype=spandrel_model.dtype)
 
-            # Prepare the output tensor.
-            _, channels, height, width = image_tensor.shape
-            output_tensor = torch.zeros(
-                (height * scale, width * scale, channels), dtype=torch.uint8, device=torch.device("cpu")
-            )
+            # Run the model on the tile.
+            output_tile = spandrel_model.run(input_tile)
 
-            image_tensor = image_tensor.to(device=spandrel_model.device, dtype=spandrel_model.dtype)
+            # Convert the output tile into the output tensor's format.
+            # (N, C, H, W) -> (C, H, W)
+            output_tile = output_tile.squeeze(0)
+            # (C, H, W) -> (H, W, C)
+            output_tile = output_tile.permute(1, 2, 0)
+            output_tile = output_tile.clamp(0, 1)
+            output_tile = (output_tile * 255).to(dtype=torch.uint8, device=torch.device("cpu"))
 
-            for tile, scaled_tile in tqdm(list(zip(tiles, scaled_tiles, strict=True)), desc="Upscaling Tiles"):
-                # Exit early if the invocation has been canceled.
-                if context.util.is_canceled():
-                    raise CanceledException
-
-                # Extract the current tile from the input tensor.
-                input_tile = image_tensor[
-                    :, :, tile.coords.top : tile.coords.bottom, tile.coords.left : tile.coords.right
-                ].to(device=spandrel_model.device, dtype=spandrel_model.dtype)
-
-                # Run the model on the tile.
-                output_tile = spandrel_model.run(input_tile)
-
-                # Convert the output tile into the output tensor's format.
-                # (N, C, H, W) -> (C, H, W)
-                output_tile = output_tile.squeeze(0)
-                # (C, H, W) -> (H, W, C)
-                output_tile = output_tile.permute(1, 2, 0)
-                output_tile = output_tile.clamp(0, 1)
-                output_tile = (output_tile * 255).to(dtype=torch.uint8, device=torch.device("cpu"))
-
-                # Merge the output tile into the output tensor.
-                # We only keep half of the overlap on the top and left side of the tile. We do this in case there are
-                # edge artifacts. We don't bother with any 'blending' in the current implementation - for most upscalers
-                # it seems unnecessary, but we may find a need in the future.
-                top_overlap = scaled_tile.overlap.top // 2
-                left_overlap = scaled_tile.overlap.left // 2
-                output_tensor[
-                    scaled_tile.coords.top + top_overlap : scaled_tile.coords.bottom,
-                    scaled_tile.coords.left + left_overlap : scaled_tile.coords.right,
-                    :,
-                ] = output_tile[top_overlap:, left_overlap:, :]
+            # Merge the output tile into the output tensor.
+            # We only keep half of the overlap on the top and left side of the tile. We do this in case there are
+            # edge artifacts. We don't bother with any 'blending' in the current implementation - for most upscalers
+            # it seems unnecessary, but we may find a need in the future.
+            top_overlap = scaled_tile.overlap.top // 2
+            left_overlap = scaled_tile.overlap.left // 2
+            output_tensor[
+                scaled_tile.coords.top + top_overlap : scaled_tile.coords.bottom,
+                scaled_tile.coords.left + left_overlap : scaled_tile.coords.right,
+                :,
+            ] = output_tile[top_overlap:, left_overlap:, :]
 
         # Convert the output tensor to a PIL image.
         np_image = output_tensor.detach().numpy().astype(np.uint8)
         pil_image = Image.fromarray(np_image)
+
+        return pil_image
+
+    @torch.inference_mode()
+    def invoke(self, context: InvocationContext) -> ImageOutput:
+        # Images are converted to RGB, because most models don't support an alpha channel. In the future, we may want to
+        # revisit this.
+        image = context.images.get_pil(self.image.image_name, mode="RGB")
+
+        # Load the model.
+        spandrel_model_info = context.models.load(self.image_to_image_model)
+
+        # Do the upscaling.
+        with spandrel_model_info as spandrel_model:
+            assert isinstance(spandrel_model, SpandrelImageToImageModel)
+
+            # Upscale the image
+            pil_image = self.upscale_image(image, self.tile_size, spandrel_model, context.util.is_canceled)
+
+        image_dto = context.images.save(image=pil_image)
+        return ImageOutput.build(image_dto)
+
+
+@invocation(
+    "spandrel_image_to_image_autoscale",
+    title="Image-to-Image (Autoscale)",
+    tags=["upscale"],
+    category="upscale",
+    version="1.0.0",
+)
+class SpandrelImageToImageAutoscaleInvocation(SpandrelImageToImageInvocation):
+    """Run any spandrel image-to-image model (https://github.com/chaiNNer-org/spandrel) until the target scale is reached."""
+
+    scale: float = InputField(
+        default=4.0,
+        gt=0.0,
+        le=16.0,
+        description="The final scale of the output image. If the model does not upscale the image, this will be ignored.",
+    )
+    fit_to_multiple_of_8: bool = InputField(
+        default=False,
+        description="If true, the output image will be resized to the nearest multiple of 8 in both dimensions.",
+    )
+
+    @torch.inference_mode()
+    def invoke(self, context: InvocationContext) -> ImageOutput:
+        # Images are converted to RGB, because most models don't support an alpha channel. In the future, we may want to
+        # revisit this.
+        image = context.images.get_pil(self.image.image_name, mode="RGB")
+
+        # Load the model.
+        spandrel_model_info = context.models.load(self.image_to_image_model)
+
+        # The target size of the image, determined by the provided scale. We'll run the upscaler until we hit this size.
+        # Later, we may mutate this value if the model doesn't upscale the image or if the user requested a multiple of 8.
+        target_width = int(image.width * self.scale)
+        target_height = int(image.height * self.scale)
+
+        # Do the upscaling.
+        with spandrel_model_info as spandrel_model:
+            assert isinstance(spandrel_model, SpandrelImageToImageModel)
+
+            # First pass of upscaling. Note: `pil_image` will be mutated.
+            pil_image = self.upscale_image(image, self.tile_size, spandrel_model, context.util.is_canceled)
+
+            # Some models don't upscale the image, but we have no way to know this in advance. We'll check if the model
+            # upscaled the image and run the loop below if it did. We'll require the model to upscale both dimensions
+            # to be considered an upscale model.
+            is_upscale_model = pil_image.width > image.width and pil_image.height > image.height
+
+            if is_upscale_model:
+                # This is an upscale model, so we should keep upscaling until we reach the target size.
+                iterations = 1
+                while pil_image.width < target_width or pil_image.height < target_height:
+                    pil_image = self.upscale_image(pil_image, self.tile_size, spandrel_model, context.util.is_canceled)
+                    iterations += 1
+
+                    # Sanity check to prevent excessive or infinite loops. All known upscaling models are at least 2x.
+                    # Our max scale is 16x, so with a 2x model, we should never exceed 16x == 2^4 -> 4 iterations.
+                    # We'll allow one extra iteration "just in case" and bail at 5 upscaling iterations. In practice,
+                    # we should never reach this limit.
+                    if iterations >= 5:
+                        context.logger.warning(
+                            "Upscale loop reached maximum iteration count of 5, stopping upscaling early."
+                        )
+                        break
+            else:
+                # This model doesn't upscale the image. We should ignore the scale parameter, modifying the output size
+                # to be the same as the processed image size.
+
+                # The output size is now the size of the processed image.
+                target_width = pil_image.width
+                target_height = pil_image.height
+
+                # Warn the user if they requested a scale greater than 1.
+                if self.scale > 1:
+                    context.logger.warning(
+                        "Model does not increase the size of the image, but a greater scale than 1 was requested. Image will not be scaled."
+                    )
+
+        # We may need to resize the image to a multiple of 8. Use floor division to ensure we don't scale the image up
+        # in the final resize
+        if self.fit_to_multiple_of_8:
+            target_width = int(target_width // 8 * 8)
+            target_height = int(target_height // 8 * 8)
+
+        # Final resize. Per PIL documentation, Lanczos provides the best quality for both upscale and downscale.
+        # See: https://pillow.readthedocs.io/en/stable/handbook/concepts.html#filters-comparison-table
+        pil_image = pil_image.resize((target_width, target_height), resample=Image.Resampling.LANCZOS)
+
         image_dto = context.images.save(image=pil_image)
         return ImageOutput.build(image_dto)

invokeai/app/services/config/config_default.py

@@ -91,6 +91,7 @@ class InvokeAIAppConfig(BaseSettings):
         db_dir: Path to InvokeAI databases directory.
         outputs_dir: Path to directory for outputs.
         custom_nodes_dir: Path to directory for custom nodes.
+        style_presets_dir: Path to directory for style presets.
         log_handlers: Log handler. Valid options are "console", "file=<path>", "syslog=path|address:host:port", "http=<url>".
         log_format: Log format. Use "plain" for text-only, "color" for colorized output, "legacy" for 2.3-style logging and "syslog" for syslog-style.<br>Valid values: `plain`, `color`, `syslog`, `legacy`
         log_level: Emit logging messages at this level or higher.<br>Valid values: `debug`, `info`, `warning`, `error`, `critical`
@@ -153,6 +154,7 @@ class InvokeAIAppConfig(BaseSettings):
     db_dir:                        Path = Field(default=Path("databases"),  description="Path to InvokeAI databases directory.")
     outputs_dir:                   Path = Field(default=Path("outputs"),    description="Path to directory for outputs.")
     custom_nodes_dir:              Path = Field(default=Path("nodes"),      description="Path to directory for custom nodes.")
+    style_presets_dir:      Path = Field(default=Path("style_presets"),      description="Path to directory for style presets.")
 
     # LOGGING
     log_handlers:             list[str] = Field(default=["console"],        description='Log handler. Valid options are "console", "file=<path>", "syslog=path|address:host:port", "http=<url>".')
@@ -300,6 +302,11 @@ class InvokeAIAppConfig(BaseSettings):
         """Path to the models directory, resolved to an absolute path.."""
         return self._resolve(self.models_dir)
 
+    @property
+    def style_presets_path(self) -> Path:
+        """Path to the style presets directory, resolved to an absolute path.."""
+        return self._resolve(self.style_presets_dir)
+
     @property
     def convert_cache_path(self) -> Path:
         """Path to the converted cache models directory, resolved to an absolute path.."""

invokeai/app/services/events/events_common.py

@@ -88,6 +88,8 @@ class QueueItemEventBase(QueueEventBase):
 
     item_id: int = Field(description="The ID of the queue item")
     batch_id: str = Field(description="The ID of the queue batch")
+    origin: str | None = Field(default=None, description="The origin of the queue item")
+    destination: str | None = Field(default=None, description="The destination of the queue item")
 
 
 class InvocationEventBase(QueueItemEventBase):
@@ -95,8 +97,6 @@ class InvocationEventBase(QueueItemEventBase):
 
     session_id: str = Field(description="The ID of the session (aka graph execution state)")
     queue_id: str = Field(description="The ID of the queue")
-    item_id: int = Field(description="The ID of the queue item")
-    batch_id: str = Field(description="The ID of the queue batch")
     session_id: str = Field(description="The ID of the session (aka graph execution state)")
     invocation: AnyInvocation = Field(description="The ID of the invocation")
     invocation_source_id: str = Field(description="The ID of the prepared invocation's source node")
@@ -114,6 +114,8 @@ class InvocationStartedEvent(InvocationEventBase):
             queue_id=queue_item.queue_id,
             item_id=queue_item.item_id,
             batch_id=queue_item.batch_id,
+            origin=queue_item.origin,
+            destination=queue_item.destination,
             session_id=queue_item.session_id,
             invocation=invocation,
             invocation_source_id=queue_item.session.prepared_source_mapping[invocation.id],
@@ -147,6 +149,8 @@ class InvocationDenoiseProgressEvent(InvocationEventBase):
             queue_id=queue_item.queue_id,
             item_id=queue_item.item_id,
             batch_id=queue_item.batch_id,
+            origin=queue_item.origin,
+            destination=queue_item.destination,
             session_id=queue_item.session_id,
             invocation=invocation,
             invocation_source_id=queue_item.session.prepared_source_mapping[invocation.id],
@@ -184,6 +188,8 @@ class InvocationCompleteEvent(InvocationEventBase):
             queue_id=queue_item.queue_id,
             item_id=queue_item.item_id,
             batch_id=queue_item.batch_id,
+            origin=queue_item.origin,
+            destination=queue_item.destination,
             session_id=queue_item.session_id,
             invocation=invocation,
             invocation_source_id=queue_item.session.prepared_source_mapping[invocation.id],
@@ -216,6 +222,8 @@ class InvocationErrorEvent(InvocationEventBase):
             queue_id=queue_item.queue_id,
             item_id=queue_item.item_id,
             batch_id=queue_item.batch_id,
+            origin=queue_item.origin,
+            destination=queue_item.destination,
             session_id=queue_item.session_id,
             invocation=invocation,
             invocation_source_id=queue_item.session.prepared_source_mapping[invocation.id],
@@ -253,6 +261,8 @@ class QueueItemStatusChangedEvent(QueueItemEventBase):
             queue_id=queue_item.queue_id,
             item_id=queue_item.item_id,
             batch_id=queue_item.batch_id,
+            origin=queue_item.origin,
+            destination=queue_item.destination,
             session_id=queue_item.session_id,
             status=queue_item.status,
             error_type=queue_item.error_type,
@@ -279,12 +289,14 @@ class BatchEnqueuedEvent(QueueEventBase):
         description="The number of invocations initially requested to be enqueued (may be less than enqueued if queue was full)"
     )
     priority: int = Field(description="The priority of the batch")
+    origin: str | None = Field(default=None, description="The origin of the batch")
 
     @classmethod
     def build(cls, enqueue_result: EnqueueBatchResult) -> "BatchEnqueuedEvent":
         return cls(
             queue_id=enqueue_result.queue_id,
             batch_id=enqueue_result.batch.batch_id,
+            origin=enqueue_result.batch.origin,
             enqueued=enqueue_result.enqueued,
             requested=enqueue_result.requested,
             priority=enqueue_result.priority,

invokeai/app/services/events/events_fastapievents.py

@@ -1,46 +1,44 @@
-# Copyright (c) 2022 Kyle Schouviller (https://github.com/kyle0654)
-
 import asyncio
 import threading
-from queue import Empty, Queue
 
 from fastapi_events.dispatcher import dispatch
 
 from invokeai.app.services.events.events_base import EventServiceBase
-from invokeai.app.services.events.events_common import (
-    EventBase,
-)
+from invokeai.app.services.events.events_common import EventBase
 
 
 class FastAPIEventService(EventServiceBase):
-    def __init__(self, event_handler_id: int) -> None:
+    def __init__(self, event_handler_id: int, loop: asyncio.AbstractEventLoop) -> None:
         self.event_handler_id = event_handler_id
-        self._queue = Queue[EventBase | None]()
+        self._queue = asyncio.Queue[EventBase | None]()
         self._stop_event = threading.Event()
-        asyncio.create_task(self._dispatch_from_queue(stop_event=self._stop_event))
+        self._loop = loop
+
+        # We need to store a reference to the task so it doesn't get GC'd
+        # See: https://docs.python.org/3/library/asyncio-task.html#creating-tasks
+        self._background_tasks: set[asyncio.Task[None]] = set()
+        task = self._loop.create_task(self._dispatch_from_queue(stop_event=self._stop_event))
+        self._background_tasks.add(task)
+        task.add_done_callback(self._background_tasks.remove)
 
         super().__init__()
 
     def stop(self, *args, **kwargs):
         self._stop_event.set()
-        self._queue.put(None)
+        self._loop.call_soon_threadsafe(self._queue.put_nowait, None)
 
     def dispatch(self, event: EventBase) -> None:
-        self._queue.put(event)
+        self._loop.call_soon_threadsafe(self._queue.put_nowait, event)
 
     async def _dispatch_from_queue(self, stop_event: threading.Event):
         """Get events on from the queue and dispatch them, from the correct thread"""
         while not stop_event.is_set():
             try:
-                event = self._queue.get(block=False)
+                event = await self._queue.get()
                 if not event:  # Probably stopping
                     continue
                 # Leave the payloads as live pydantic models
                 dispatch(event, middleware_id=self.event_handler_id, payload_schema_dump=False)
 
-            except Empty:
-                await asyncio.sleep(0.1)
-                pass
-
             except asyncio.CancelledError as e:
                 raise e  # Raise a proper error

invokeai/app/services/image_files/image_files_disk.py

@@ -1,11 +1,10 @@
 # Copyright (c) 2022 Kyle Schouviller (https://github.com/kyle0654) and the InvokeAI Team
 from pathlib import Path
 from queue import Queue
-from typing import Dict, Optional, Union
+from typing import Optional, Union
 
 from PIL import Image, PngImagePlugin
 from PIL.Image import Image as PILImageType
-from send2trash import send2trash
 
 from invokeai.app.services.image_files.image_files_base import ImageFileStorageBase
 from invokeai.app.services.image_files.image_files_common import (
@@ -20,18 +19,12 @@ from invokeai.app.util.thumbnails import get_thumbnail_name, make_thumbnail
 class DiskImageFileStorage(ImageFileStorageBase):
     """Stores images on disk"""
 
-    __output_folder: Path
-    __cache_ids: Queue  # TODO: this is an incredibly naive cache
-    __cache: Dict[Path, PILImageType]
-    __max_cache_size: int
-    __invoker: Invoker
-
     def __init__(self, output_folder: Union[str, Path]):
-        self.__cache = {}
-        self.__cache_ids = Queue()
+        self.__cache: dict[Path, PILImageType] = {}
+        self.__cache_ids = Queue[Path]()
         self.__max_cache_size = 10  # TODO: get this from config
 
-        self.__output_folder: Path = output_folder if isinstance(output_folder, Path) else Path(output_folder)
+        self.__output_folder = output_folder if isinstance(output_folder, Path) else Path(output_folder)
         self.__thumbnails_folder = self.__output_folder / "thumbnails"
         # Validate required output folders at launch
         self.__validate_storage_folders()
@@ -103,7 +96,7 @@ class DiskImageFileStorage(ImageFileStorageBase):
             image_path = self.get_path(image_name)
 
             if image_path.exists():
-                send2trash(image_path)
+                image_path.unlink()
             if image_path in self.__cache:
                 del self.__cache[image_path]
 
@@ -111,7 +104,7 @@ class DiskImageFileStorage(ImageFileStorageBase):
             thumbnail_path = self.get_path(thumbnail_name, True)
 
             if thumbnail_path.exists():
-                send2trash(thumbnail_path)
+                thumbnail_path.unlink()
             if thumbnail_path in self.__cache:
                 del self.__cache[thumbnail_path]
         except Exception as e:

invokeai/app/services/invocation_services.py

@@ -4,6 +4,8 @@ from __future__ import annotations
 from typing import TYPE_CHECKING
 
 from invokeai.app.services.object_serializer.object_serializer_base import ObjectSerializerBase
+from invokeai.app.services.style_preset_images.style_preset_images_base import StylePresetImageFileStorageBase
+from invokeai.app.services.style_preset_records.style_preset_records_base import StylePresetRecordsStorageBase
 
 if TYPE_CHECKING:
     from logging import Logger
@@ -61,6 +63,8 @@ class InvocationServices:
         workflow_records: "WorkflowRecordsStorageBase",
         tensors: "ObjectSerializerBase[torch.Tensor]",
         conditioning: "ObjectSerializerBase[ConditioningFieldData]",
+        style_preset_records: "StylePresetRecordsStorageBase",
+        style_preset_image_files: "StylePresetImageFileStorageBase",
     ):
         self.board_images = board_images
         self.board_image_records = board_image_records
@@ -85,3 +89,5 @@ class InvocationServices:
         self.workflow_records = workflow_records
         self.tensors = tensors
         self.conditioning = conditioning
+        self.style_preset_records = style_preset_records
+        self.style_preset_image_files = style_preset_image_files

invokeai/app/services/model_images/model_images_default.py

@@ -2,7 +2,6 @@ from pathlib import Path
 
 from PIL import Image
 from PIL.Image import Image as PILImageType
-from send2trash import send2trash
 
 from invokeai.app.services.invoker import Invoker
 from invokeai.app.services.model_images.model_images_base import ModelImageFileStorageBase
@@ -70,7 +69,7 @@ class ModelImageFileStorageDisk(ModelImageFileStorageBase):
             if not self._validate_path(path):
                 raise ModelImageFileNotFoundException
 
-            send2trash(path)
+            path.unlink()
 
         except Exception as e:
             raise ModelImageFileDeleteException from e

invokeai/app/services/model_install/model_install_base.py

@@ -3,7 +3,7 @@
 
 from abc import ABC, abstractmethod
 from pathlib import Path
-from typing import Any, Dict, List, Optional, Union
+from typing import List, Optional, Union
 
 from pydantic.networks import AnyHttpUrl
 
@@ -12,7 +12,7 @@ from invokeai.app.services.download import DownloadQueueServiceBase
 from invokeai.app.services.events.events_base import EventServiceBase
 from invokeai.app.services.invoker import Invoker
 from invokeai.app.services.model_install.model_install_common import ModelInstallJob, ModelSource
-from invokeai.app.services.model_records import ModelRecordServiceBase
+from invokeai.app.services.model_records import ModelRecordChanges, ModelRecordServiceBase
 from invokeai.backend.model_manager import AnyModelConfig
 
 
@@ -64,7 +64,7 @@ class ModelInstallServiceBase(ABC):
     def register_path(
         self,
         model_path: Union[Path, str],
-        config: Optional[Dict[str, Any]] = None,
+        config: Optional[ModelRecordChanges] = None,
     ) -> str:
         """
         Probe and register the model at model_path.
@@ -72,7 +72,7 @@ class ModelInstallServiceBase(ABC):
         This keeps the model in its current location.
 
         :param model_path: Filesystem Path to the model.
-        :param config: Dict of attributes that will override autoassigned values.
+        :param config: ModelRecordChanges object that will override autoassigned model record values.
         :returns id: The string ID of the registered model.
         """
 
@@ -92,7 +92,7 @@ class ModelInstallServiceBase(ABC):
     def install_path(
         self,
         model_path: Union[Path, str],
-        config: Optional[Dict[str, Any]] = None,
+        config: Optional[ModelRecordChanges] = None,
     ) -> str:
         """
         Probe, register and install the model in the models directory.
@@ -101,7 +101,7 @@ class ModelInstallServiceBase(ABC):
         the models directory handled by InvokeAI.
 
         :param model_path: Filesystem Path to the model.
-        :param config: Dict of attributes that will override autoassigned values.
+        :param config: ModelRecordChanges object that will override autoassigned model record values.
         :returns id: The string ID of the registered model.
         """
 
@@ -109,14 +109,14 @@ class ModelInstallServiceBase(ABC):
     def heuristic_import(
         self,
         source: str,
-        config: Optional[Dict[str, Any]] = None,
+        config: Optional[ModelRecordChanges] = None,
         access_token: Optional[str] = None,
         inplace: Optional[bool] = False,
     ) -> ModelInstallJob:
         r"""Install the indicated model using heuristics to interpret user intentions.
 
         :param source: String source
-        :param config: Optional dict. Any fields in this dict
+        :param config: Optional ModelRecordChanges object. Any fields in this object
          will override corresponding autoassigned probe fields in the
          model's config record as described in `import_model()`.
         :param access_token: Optional access token for remote sources.
@@ -147,7 +147,7 @@ class ModelInstallServiceBase(ABC):
     def import_model(
         self,
         source: ModelSource,
-        config: Optional[Dict[str, Any]] = None,
+        config: Optional[ModelRecordChanges] = None,
     ) -> ModelInstallJob:
         """Install the indicated model.
 

invokeai/app/services/model_install/model_install_common.py

@@ -2,13 +2,14 @@ import re
 import traceback
 from enum import Enum
 from pathlib import Path
-from typing import Any, Dict, Literal, Optional, Set, Union
+from typing import Literal, Optional, Set, Union
 
 from pydantic import BaseModel, Field, PrivateAttr, field_validator
 from pydantic.networks import AnyHttpUrl
 from typing_extensions import Annotated
 
 from invokeai.app.services.download import DownloadJob, MultiFileDownloadJob
+from invokeai.app.services.model_records import ModelRecordChanges
 from invokeai.backend.model_manager import AnyModelConfig, ModelRepoVariant
 from invokeai.backend.model_manager.config import ModelSourceType
 from invokeai.backend.model_manager.metadata import AnyModelRepoMetadata
@@ -102,7 +103,7 @@ class HFModelSource(StringLikeSource):
         if self.variant:
             base += f":{self.variant or ''}"
         if self.subfolder:
-            base += f":{self.subfolder}"
+            base += f"::{self.subfolder.as_posix()}"
         return base
 
 
@@ -133,8 +134,9 @@ class ModelInstallJob(BaseModel):
     id: int = Field(description="Unique ID for this job")
     status: InstallStatus = Field(default=InstallStatus.WAITING, description="Current status of install process")
     error_reason: Optional[str] = Field(default=None, description="Information about why the job failed")
-    config_in: Dict[str, Any] = Field(
-        default_factory=dict, description="Configuration information (e.g. 'description') to apply to model."
+    config_in: ModelRecordChanges = Field(
+        default_factory=ModelRecordChanges,
+        description="Configuration information (e.g. 'description') to apply to model.",
     )
     config_out: Optional[AnyModelConfig] = Field(
         default=None, description="After successful installation, this will hold the configuration object."

invokeai/app/services/model_install/model_install_default.py

@@ -163,26 +163,27 @@ class ModelInstallService(ModelInstallServiceBase):
     def register_path(
         self,
         model_path: Union[Path, str],
-        config: Optional[Dict[str, Any]] = None,
+        config: Optional[ModelRecordChanges] = None,
     ) -> str:  # noqa D102
         model_path = Path(model_path)
-        config = config or {}
-        if not config.get("source"):
-            config["source"] = model_path.resolve().as_posix()
-        config["source_type"] = ModelSourceType.Path
+        config = config or ModelRecordChanges()
+        if not config.source:
+            config.source = model_path.resolve().as_posix()
+        config.source_type = ModelSourceType.Path
         return self._register(model_path, config)
 
     def install_path(
         self,
         model_path: Union[Path, str],
-        config: Optional[Dict[str, Any]] = None,
+        config: Optional[ModelRecordChanges] = None,
     ) -> str:  # noqa D102
         model_path = Path(model_path)
-        config = config or {}
+        config = config or ModelRecordChanges()
+        info: AnyModelConfig = ModelProbe.probe(
+            Path(model_path), config.model_dump(), hash_algo=self._app_config.hashing_algorithm
+        )  # type: ignore
 
-        info: AnyModelConfig = ModelProbe.probe(Path(model_path), config, hash_algo=self._app_config.hashing_algorithm)
-
-        if preferred_name := config.get("name"):
+        if preferred_name := config.name:
             preferred_name = Path(preferred_name).with_suffix(model_path.suffix)
 
         dest_path = (
@@ -204,7 +205,7 @@ class ModelInstallService(ModelInstallServiceBase):
     def heuristic_import(
         self,
         source: str,
-        config: Optional[Dict[str, Any]] = None,
+        config: Optional[ModelRecordChanges] = None,
         access_token: Optional[str] = None,
         inplace: Optional[bool] = False,
     ) -> ModelInstallJob:
@@ -216,7 +217,7 @@ class ModelInstallService(ModelInstallServiceBase):
             source_obj.access_token = access_token
         return self.import_model(source_obj, config)
 
-    def import_model(self, source: ModelSource, config: Optional[Dict[str, Any]] = None) -> ModelInstallJob:  # noqa D102
+    def import_model(self, source: ModelSource, config: Optional[ModelRecordChanges] = None) -> ModelInstallJob:  # noqa D102
         similar_jobs = [x for x in self.list_jobs() if x.source == source and not x.in_terminal_state]
         if similar_jobs:
             self._logger.warning(f"There is already an active install job for {source}. Not enqueuing.")
@@ -318,16 +319,17 @@ class ModelInstallService(ModelInstallServiceBase):
                         model_path = self._app_config.models_path / model_path
                     model_path = model_path.resolve()
 
-                    config: dict[str, Any] = {}
-                    config["name"] = model_name
-                    config["description"] = stanza.get("description")
+                    config = ModelRecordChanges(
+                        name=model_name,
+                        description=stanza.get("description"),
+                    )
                     legacy_config_path = stanza.get("config")
                     if legacy_config_path:
                         # In v3, these paths were relative to the root. Migrate them to be relative to the legacy_conf_dir.
                         legacy_config_path = self._app_config.root_path / legacy_config_path
                         if legacy_config_path.is_relative_to(self._app_config.legacy_conf_path):
                             legacy_config_path = legacy_config_path.relative_to(self._app_config.legacy_conf_path)
-                        config["config_path"] = str(legacy_config_path)
+                        config.config_path = str(legacy_config_path)
                     try:
                         id = self.register_path(model_path=model_path, config=config)
                         self._logger.info(f"Migrated {model_name} with id {id}")
@@ -500,11 +502,11 @@ class ModelInstallService(ModelInstallServiceBase):
         job.total_bytes = self._stat_size(job.local_path)
         job.bytes = job.total_bytes
         self._signal_job_running(job)
-        job.config_in["source"] = str(job.source)
-        job.config_in["source_type"] = MODEL_SOURCE_TO_TYPE_MAP[job.source.__class__]
+        job.config_in.source = str(job.source)
+        job.config_in.source_type = MODEL_SOURCE_TO_TYPE_MAP[job.source.__class__]
         # enter the metadata, if there is any
         if isinstance(job.source_metadata, (HuggingFaceMetadata)):
-            job.config_in["source_api_response"] = job.source_metadata.api_response
+            job.config_in.source_api_response = job.source_metadata.api_response
 
         if job.inplace:
             key = self.register_path(job.local_path, job.config_in)
@@ -639,11 +641,11 @@ class ModelInstallService(ModelInstallServiceBase):
         return new_path
 
     def _register(
-        self, model_path: Path, config: Optional[Dict[str, Any]] = None, info: Optional[AnyModelConfig] = None
+        self, model_path: Path, config: Optional[ModelRecordChanges] = None, info: Optional[AnyModelConfig] = None
     ) -> str:
-        config = config or {}
+        config = config or ModelRecordChanges()
 
-        info = info or ModelProbe.probe(model_path, config, hash_algo=self._app_config.hashing_algorithm)
+        info = info or ModelProbe.probe(model_path, config.model_dump(), hash_algo=self._app_config.hashing_algorithm)  # type: ignore
 
         model_path = model_path.resolve()
 
@@ -674,11 +676,13 @@ class ModelInstallService(ModelInstallServiceBase):
         precision = TorchDevice.choose_torch_dtype()
         return ModelRepoVariant.FP16 if precision == torch.float16 else None
 
-    def _import_local_model(self, source: LocalModelSource, config: Optional[Dict[str, Any]]) -> ModelInstallJob:
+    def _import_local_model(
+        self, source: LocalModelSource, config: Optional[ModelRecordChanges] = None
+    ) -> ModelInstallJob:
         return ModelInstallJob(
             id=self._next_id(),
             source=source,
-            config_in=config or {},
+            config_in=config or ModelRecordChanges(),
             local_path=Path(source.path),
             inplace=source.inplace or False,
         )
@@ -686,7 +690,7 @@ class ModelInstallService(ModelInstallServiceBase):
     def _import_from_hf(
         self,
         source: HFModelSource,
-        config: Optional[Dict[str, Any]] = None,
+        config: Optional[ModelRecordChanges] = None,
     ) -> ModelInstallJob:
         # Add user's cached access token to HuggingFace requests
         if source.access_token is None:
@@ -702,7 +706,7 @@ class ModelInstallService(ModelInstallServiceBase):
     def _import_from_url(
         self,
         source: URLModelSource,
-        config: Optional[Dict[str, Any]],
+        config: Optional[ModelRecordChanges] = None,
     ) -> ModelInstallJob:
         remote_files, metadata = self._remote_files_from_source(source)
         return self._import_remote_model(
@@ -717,7 +721,7 @@ class ModelInstallService(ModelInstallServiceBase):
         source: HFModelSource | URLModelSource,
         remote_files: List[RemoteModelFile],
         metadata: Optional[AnyModelRepoMetadata],
-        config: Optional[Dict[str, Any]],
+        config: Optional[ModelRecordChanges],
     ) -> ModelInstallJob:
         if len(remote_files) == 0:
             raise ValueError(f"{source}: No downloadable files found")
@@ -730,7 +734,7 @@ class ModelInstallService(ModelInstallServiceBase):
         install_job = ModelInstallJob(
             id=self._next_id(),
             source=source,
-            config_in=config or {},
+            config_in=config or ModelRecordChanges(),
             source_metadata=metadata,
             local_path=destdir,  # local path may change once the download has started due to content-disposition handling
             bytes=0,
@@ -779,8 +783,9 @@ class ModelInstallService(ModelInstallServiceBase):
         # So what we do is to synthesize a folder named "sdxl-turbo_vae" here.
         if subfolder:
             top = Path(remote_files[0].path.parts[0])  # e.g. "sdxl-turbo/"
-            path_to_remove = top / subfolder.parts[-1]  # sdxl-turbo/vae/
-            path_to_add = Path(f"{top}_{subfolder}")
+            path_to_remove = top / subfolder  # sdxl-turbo/vae/
+            subfolder_rename = subfolder.name.replace("/", "_").replace("\\", "_")
+            path_to_add = Path(f"{top}_{subfolder_rename}")
         else:
             path_to_remove = Path(".")
             path_to_add = Path(".")

invokeai/app/services/model_records/model_records_base.py

@@ -18,6 +18,7 @@ from invokeai.backend.model_manager.config import (
     ControlAdapterDefaultSettings,
     MainModelDefaultSettings,
     ModelFormat,
+    ModelSourceType,
     ModelType,
     ModelVariantType,
     SchedulerPredictionType,
@@ -66,10 +67,17 @@ class ModelRecordChanges(BaseModelExcludeNull):
     """A set of changes to apply to a model."""
 
     # Changes applicable to all models
+    source: Optional[str] = Field(description="original source of the model", default=None)
+    source_type: Optional[ModelSourceType] = Field(description="type of model source", default=None)
+    source_api_response: Optional[str] = Field(description="metadata from remote source", default=None)
     name: Optional[str] = Field(description="Name of the model.", default=None)
     path: Optional[str] = Field(description="Path to the model.", default=None)
     description: Optional[str] = Field(description="Model description", default=None)
     base: Optional[BaseModelType] = Field(description="The base model.", default=None)
+    type: Optional[ModelType] = Field(description="Type of model", default=None)
+    key: Optional[str] = Field(description="Database ID for this model", default=None)
+    hash: Optional[str] = Field(description="hash of model file", default=None)
+    format: Optional[str] = Field(description="format of model file", default=None)
     trigger_phrases: Optional[set[str]] = Field(description="Set of trigger phrases for this model", default=None)
     default_settings: Optional[MainModelDefaultSettings | ControlAdapterDefaultSettings] = Field(
         description="Default settings for this model", default=None

invokeai/app/services/session_queue/session_queue_base.py

@@ -6,6 +6,7 @@ from invokeai.app.services.session_queue.session_queue_common import (
     Batch,
     BatchStatus,
     CancelByBatchIDsResult,
+    CancelByDestinationResult,
     CancelByQueueIDResult,
     ClearResult,
     EnqueueBatchResult,
@@ -95,6 +96,11 @@ class SessionQueueBase(ABC):
         """Cancels all queue items with matching batch IDs"""
         pass
 
+    @abstractmethod
+    def cancel_by_destination(self, queue_id: str, destination: str) -> CancelByDestinationResult:
+        """Cancels all queue items with the given batch destination"""
+        pass
+
     @abstractmethod
     def cancel_by_queue_id(self, queue_id: str) -> CancelByQueueIDResult:
         """Cancels all queue items with matching queue ID"""

invokeai/app/services/session_queue/session_queue_common.py

@@ -77,6 +77,14 @@ BatchDataCollection: TypeAlias = list[list[BatchDatum]]
 
 class Batch(BaseModel):
     batch_id: str = Field(default_factory=uuid_string, description="The ID of the batch")
+    origin: str | None = Field(
+        default=None,
+        description="The origin of this queue item. This data is used by the frontend to determine how to handle results.",
+    )
+    destination: str | None = Field(
+        default=None,
+        description="The origin of this queue item. This data is used by the frontend to determine how to handle results",
+    )
     data: Optional[BatchDataCollection] = Field(default=None, description="The batch data collection.")
     graph: Graph = Field(description="The graph to initialize the session with")
     workflow: Optional[WorkflowWithoutID] = Field(
@@ -195,6 +203,14 @@ class SessionQueueItemWithoutGraph(BaseModel):
     status: QUEUE_ITEM_STATUS = Field(default="pending", description="The status of this queue item")
     priority: int = Field(default=0, description="The priority of this queue item")
     batch_id: str = Field(description="The ID of the batch associated with this queue item")
+    origin: str | None = Field(
+        default=None,
+        description="The origin of this queue item. This data is used by the frontend to determine how to handle results.",
+    )
+    destination: str | None = Field(
+        default=None,
+        description="The origin of this queue item. This data is used by the frontend to determine how to handle results",
+    )
     session_id: str = Field(
         description="The ID of the session associated with this queue item. The session doesn't exist in graph_executions until the queue item is executed."
     )
@@ -294,6 +310,8 @@ class SessionQueueStatus(BaseModel):
 class BatchStatus(BaseModel):
     queue_id: str = Field(..., description="The ID of the queue")
     batch_id: str = Field(..., description="The ID of the batch")
+    origin: str | None = Field(..., description="The origin of the batch")
+    destination: str | None = Field(..., description="The destination of the batch")
     pending: int = Field(..., description="Number of queue items with status 'pending'")
     in_progress: int = Field(..., description="Number of queue items with status 'in_progress'")
     completed: int = Field(..., description="Number of queue items with status 'complete'")
@@ -328,6 +346,12 @@ class CancelByBatchIDsResult(BaseModel):
     canceled: int = Field(..., description="Number of queue items canceled")
 
 
+class CancelByDestinationResult(CancelByBatchIDsResult):
+    """Result of canceling by a destination"""
+
+    pass
+
+
 class CancelByQueueIDResult(CancelByBatchIDsResult):
     """Result of canceling by queue id"""
 
@@ -433,6 +457,8 @@ class SessionQueueValueToInsert(NamedTuple):
     field_values: Optional[str]  # field_values json
     priority: int  # priority
     workflow: Optional[str]  # workflow json
+    origin: str | None
+    destination: str | None
 
 
 ValuesToInsert: TypeAlias = list[SessionQueueValueToInsert]
@@ -453,6 +479,8 @@ def prepare_values_to_insert(queue_id: str, batch: Batch, priority: int, max_new
                 json.dumps(field_values, default=to_jsonable_python) if field_values else None,  # field_values (json)
                 priority,  # priority
                 json.dumps(workflow, default=to_jsonable_python) if workflow else None,  # workflow (json)
+                batch.origin,  # origin
+                batch.destination,  # destination
             )
         )
     return values_to_insert

invokeai/app/services/session_queue/session_queue_sqlite.py

@@ -10,6 +10,7 @@ from invokeai.app.services.session_queue.session_queue_common import (
     Batch,
     BatchStatus,
     CancelByBatchIDsResult,
+    CancelByDestinationResult,
     CancelByQueueIDResult,
     ClearResult,
     EnqueueBatchResult,
@@ -127,8 +128,8 @@ class SqliteSessionQueue(SessionQueueBase):
 
             self.__cursor.executemany(
                 """--sql
-                INSERT INTO session_queue (queue_id, session, session_id, batch_id, field_values, priority, workflow)
-                VALUES (?, ?, ?, ?, ?, ?, ?)
+                INSERT INTO session_queue (queue_id, session, session_id, batch_id, field_values, priority, workflow, origin, destination)
+                VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)
                 """,
                 values_to_insert,
             )
@@ -417,11 +418,7 @@ class SqliteSessionQueue(SessionQueueBase):
             )
             self.__conn.commit()
             if current_queue_item is not None and current_queue_item.batch_id in batch_ids:
-                batch_status = self.get_batch_status(queue_id=queue_id, batch_id=current_queue_item.batch_id)
-                queue_status = self.get_queue_status(queue_id=queue_id)
-                self.__invoker.services.events.emit_queue_item_status_changed(
-                    current_queue_item, batch_status, queue_status
-                )
+                self._set_queue_item_status(current_queue_item.item_id, "canceled")
         except Exception:
             self.__conn.rollback()
             raise
@@ -429,6 +426,46 @@ class SqliteSessionQueue(SessionQueueBase):
             self.__lock.release()
         return CancelByBatchIDsResult(canceled=count)
 
+    def cancel_by_destination(self, queue_id: str, destination: str) -> CancelByDestinationResult:
+        try:
+            current_queue_item = self.get_current(queue_id)
+            self.__lock.acquire()
+            where = """--sql
+                WHERE
+                  queue_id == ?
+                  AND destination == ?
+                  AND status != 'canceled'
+                  AND status != 'completed'
+                  AND status != 'failed'
+                """
+            params = (queue_id, destination)
+            self.__cursor.execute(
+                f"""--sql
+                SELECT COUNT(*)
+                FROM session_queue
+                {where};
+                """,
+                params,
+            )
+            count = self.__cursor.fetchone()[0]
+            self.__cursor.execute(
+                f"""--sql
+                UPDATE session_queue
+                SET status = 'canceled'
+                {where};
+                """,
+                params,
+            )
+            self.__conn.commit()
+            if current_queue_item is not None and current_queue_item.destination == destination:
+                self._set_queue_item_status(current_queue_item.item_id, "canceled")
+        except Exception:
+            self.__conn.rollback()
+            raise
+        finally:
+            self.__lock.release()
+        return CancelByDestinationResult(canceled=count)
+
     def cancel_by_queue_id(self, queue_id: str) -> CancelByQueueIDResult:
         try:
             current_queue_item = self.get_current(queue_id)
@@ -541,7 +578,9 @@ class SqliteSessionQueue(SessionQueueBase):
                     started_at,
                     session_id,
                     batch_id,
-                    queue_id
+                    queue_id,
+                    origin,
+                    destination
                 FROM session_queue
                 WHERE queue_id = ?
             """
@@ -621,7 +660,7 @@ class SqliteSessionQueue(SessionQueueBase):
             self.__lock.acquire()
             self.__cursor.execute(
                 """--sql
-                SELECT status, count(*)
+                SELECT status, count(*), origin, destination
                 FROM session_queue
                 WHERE
                   queue_id = ?
@@ -633,6 +672,8 @@ class SqliteSessionQueue(SessionQueueBase):
             result = cast(list[sqlite3.Row], self.__cursor.fetchall())
             total = sum(row[1] for row in result)
             counts: dict[str, int] = {row[0]: row[1] for row in result}
+            origin = result[0]["origin"] if result else None
+            destination = result[0]["destination"] if result else None
         except Exception:
             self.__conn.rollback()
             raise
@@ -641,6 +682,8 @@ class SqliteSessionQueue(SessionQueueBase):
 
         return BatchStatus(
             batch_id=batch_id,
+            origin=origin,
+            destination=destination,
             queue_id=queue_id,
             pending=counts.get("pending", 0),
             in_progress=counts.get("in_progress", 0),

invokeai/app/services/shared/invocation_context.py

@@ -14,7 +14,7 @@ from invokeai.app.services.image_records.image_records_common import ImageCatego
 from invokeai.app.services.images.images_common import ImageDTO
 from invokeai.app.services.invocation_services import InvocationServices
 from invokeai.app.services.model_records.model_records_base import UnknownModelException
-from invokeai.app.util.step_callback import stable_diffusion_step_callback
+from invokeai.app.util.step_callback import flux_step_callback, stable_diffusion_step_callback
 from invokeai.backend.model_manager.config import (
     AnyModel,
     AnyModelConfig,
@@ -557,6 +557,24 @@ class UtilInterface(InvocationContextInterface):
             is_canceled=self.is_canceled,
         )
 
+    def flux_step_callback(self, intermediate_state: PipelineIntermediateState) -> None:
+        """
+        The step callback emits a progress event with the current step, the total number of
+        steps, a preview image, and some other internal metadata.
+
+        This should be called after each denoising step.
+
+        Args:
+            intermediate_state: The intermediate state of the diffusion pipeline.
+        """
+
+        flux_step_callback(
+            context_data=self._data,
+            intermediate_state=intermediate_state,
+            events=self._services.events,
+            is_canceled=self.is_canceled,
+        )
+
 
 class InvocationContext:
     """Provides access to various services and data for the current invocation.

invokeai/app/services/shared/sqlite/sqlite_util.py

@@ -16,6 +16,8 @@ from invokeai.app.services.shared.sqlite_migrator.migrations.migration_10 import
 from invokeai.app.services.shared.sqlite_migrator.migrations.migration_11 import build_migration_11
 from invokeai.app.services.shared.sqlite_migrator.migrations.migration_12 import build_migration_12
 from invokeai.app.services.shared.sqlite_migrator.migrations.migration_13 import build_migration_13
+from invokeai.app.services.shared.sqlite_migrator.migrations.migration_14 import build_migration_14
+from invokeai.app.services.shared.sqlite_migrator.migrations.migration_15 import build_migration_15
 from invokeai.app.services.shared.sqlite_migrator.sqlite_migrator_impl import SqliteMigrator
 
 
@@ -49,6 +51,8 @@ def init_db(config: InvokeAIAppConfig, logger: Logger, image_files: ImageFileSto
     migrator.register_migration(build_migration_11(app_config=config, logger=logger))
     migrator.register_migration(build_migration_12(app_config=config))
     migrator.register_migration(build_migration_13())
+    migrator.register_migration(build_migration_14())
+    migrator.register_migration(build_migration_15())
     migrator.run_migrations()
 
     return db

invokeai/app/services/shared/sqlite_migrator/migrations/migration_14.py

@@ -0,0 +1,61 @@
+import sqlite3
+
+from invokeai.app.services.shared.sqlite_migrator.sqlite_migrator_common import Migration
+
+
+class Migration14Callback:
+    def __call__(self, cursor: sqlite3.Cursor) -> None:
+        self._create_style_presets(cursor)
+
+    def _create_style_presets(self, cursor: sqlite3.Cursor) -> None:
+        """Create the table used to store style presets."""
+        tables = [
+            """--sql
+            CREATE TABLE IF NOT EXISTS style_presets (
+                id TEXT NOT NULL PRIMARY KEY,
+                name TEXT NOT NULL,
+                preset_data TEXT NOT NULL,
+                type TEXT NOT NULL DEFAULT "user",
+                created_at DATETIME NOT NULL DEFAULT(STRFTIME('%Y-%m-%d %H:%M:%f', 'NOW')),
+                -- Updated via trigger
+                updated_at DATETIME NOT NULL DEFAULT(STRFTIME('%Y-%m-%d %H:%M:%f', 'NOW'))
+            );
+            """
+        ]
+
+        # Add trigger for `updated_at`.
+        triggers = [
+            """--sql
+            CREATE TRIGGER IF NOT EXISTS style_presets
+            AFTER UPDATE
+            ON style_presets FOR EACH ROW
+            BEGIN
+                UPDATE style_presets SET updated_at = STRFTIME('%Y-%m-%d %H:%M:%f', 'NOW')
+                    WHERE id = old.id;
+            END;
+            """
+        ]
+
+        # Add indexes for searchable fields
+        indices = [
+            "CREATE INDEX IF NOT EXISTS idx_style_presets_name ON style_presets(name);",
+        ]
+
+        for stmt in tables + indices + triggers:
+            cursor.execute(stmt)
+
+
+def build_migration_14() -> Migration:
+    """
+    Build the migration from database version 13 to 14..
+
+    This migration does the following:
+    - Create the table used to store style presets.
+    """
+    migration_14 = Migration(
+        from_version=13,
+        to_version=14,
+        callback=Migration14Callback(),
+    )
+
+    return migration_14

invokeai/app/services/shared/sqlite_migrator/migrations/migration_15.py

@@ -0,0 +1,34 @@
+import sqlite3
+
+from invokeai.app.services.shared.sqlite_migrator.sqlite_migrator_common import Migration
+
+
+class Migration15Callback:
+    def __call__(self, cursor: sqlite3.Cursor) -> None:
+        self._add_origin_col(cursor)
+
+    def _add_origin_col(self, cursor: sqlite3.Cursor) -> None:
+        """
+        - Adds `origin` column to the session queue table.
+        - Adds `destination` column to the session queue table.
+        """
+
+        cursor.execute("ALTER TABLE session_queue ADD COLUMN origin TEXT;")
+        cursor.execute("ALTER TABLE session_queue ADD COLUMN destination TEXT;")
+
+
+def build_migration_15() -> Migration:
+    """
+    Build the migration from database version 14 to 15.
+
+    This migration does the following:
+        - Adds `origin` column to the session queue table.
+        - Adds `destination` column to the session queue table.
+    """
+    migration_15 = Migration(
+        from_version=14,
+        to_version=15,
+        callback=Migration15Callback(),
+    )
+
+    return migration_15

invokeai/app/services/style_preset_images/style_preset_images_base.py

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

invokeai/app/services/style_preset_images/style_preset_images_common.py

@@ -0,0 +1,19 @@
+class StylePresetImageFileNotFoundException(Exception):
+    """Raised when an image file is not found in storage."""
+
+    def __init__(self, message: str = "Style preset image file not found"):
+        super().__init__(message)
+
+
+class StylePresetImageFileSaveException(Exception):
+    """Raised when an image cannot be saved."""
+
+    def __init__(self, message: str = "Style preset image file not saved"):
+        super().__init__(message)
+
+
+class StylePresetImageFileDeleteException(Exception):
+    """Raised when an image cannot be deleted."""
+
+    def __init__(self, message: str = "Style preset image file not deleted"):
+        super().__init__(message)

invokeai/app/services/style_preset_images/style_preset_images_disk.py

@@ -0,0 +1,88 @@
+from pathlib import Path
+
+from PIL import Image
+from PIL.Image import Image as PILImageType
+
+from invokeai.app.services.invoker import Invoker
+from invokeai.app.services.style_preset_images.style_preset_images_base import StylePresetImageFileStorageBase
+from invokeai.app.services.style_preset_images.style_preset_images_common import (
+    StylePresetImageFileDeleteException,
+    StylePresetImageFileNotFoundException,
+    StylePresetImageFileSaveException,
+)
+from invokeai.app.services.style_preset_records.style_preset_records_common import PresetType
+from invokeai.app.util.misc import uuid_string
+from invokeai.app.util.thumbnails import make_thumbnail
+
+
+class StylePresetImageFileStorageDisk(StylePresetImageFileStorageBase):
+    """Stores images on disk"""
+
+    def __init__(self, style_preset_images_folder: Path):
+        self._style_preset_images_folder = style_preset_images_folder
+        self._validate_storage_folders()
+
+    def start(self, invoker: Invoker) -> None:
+        self._invoker = invoker
+
+    def get(self, style_preset_id: str) -> PILImageType:
+        try:
+            path = self.get_path(style_preset_id)
+
+            return Image.open(path)
+        except FileNotFoundError as e:
+            raise StylePresetImageFileNotFoundException from e
+
+    def save(self, style_preset_id: str, image: PILImageType) -> None:
+        try:
+            self._validate_storage_folders()
+            image_path = self._style_preset_images_folder / (style_preset_id + ".webp")
+            thumbnail = make_thumbnail(image, 256)
+            thumbnail.save(image_path, format="webp")
+
+        except Exception as e:
+            raise StylePresetImageFileSaveException from e
+
+    def get_path(self, style_preset_id: str) -> Path:
+        style_preset = self._invoker.services.style_preset_records.get(style_preset_id)
+        if style_preset.type is PresetType.Default:
+            default_images_dir = Path(__file__).parent / Path("default_style_preset_images")
+            path = default_images_dir / (style_preset.name + ".png")
+        else:
+            path = self._style_preset_images_folder / (style_preset_id + ".webp")
+
+        return path
+
+    def get_url(self, style_preset_id: str) -> str | None:
+        path = self.get_path(style_preset_id)
+        if not self._validate_path(path):
+            return
+
+        url = self._invoker.services.urls.get_style_preset_image_url(style_preset_id)
+
+        # The image URL never changes, so we must add random query string to it to prevent caching
+        url += f"?{uuid_string()}"
+
+        return url
+
+    def delete(self, style_preset_id: str) -> None:
+        try:
+            path = self.get_path(style_preset_id)
+
+            if not self._validate_path(path):
+                raise StylePresetImageFileNotFoundException
+
+            path.unlink()
+
+        except StylePresetImageFileNotFoundException as e:
+            raise StylePresetImageFileNotFoundException from e
+        except Exception as e:
+            raise StylePresetImageFileDeleteException from e
+
+    def _validate_path(self, path: Path) -> bool:
+        """Validates the path given for an image."""
+        return path.exists()
+
+    def _validate_storage_folders(self) -> None:
+        """Checks if the required folders exist and create them if they don't"""
+        self._style_preset_images_folder.mkdir(parents=True, exist_ok=True)

invokeai/app/services/style_preset_records/default_style_presets.json

@@ -0,0 +1,146 @@
+[
+  {
+    "name": "Photography (General)",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "{prompt}. photography. f/2.8 macro photo, bokeh, photorealism",
+      "negative_prompt": "painting, digital art. sketch, blurry"
+    }
+  },
+  {
+    "name": "Photography (Studio Lighting)",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "{prompt}, photography. f/8 photo. centered subject, studio lighting.",
+      "negative_prompt": "painting, digital art. sketch, blurry"
+    }
+  },
+  {
+    "name": "Photography (Landscape)",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "{prompt}, landscape photograph, f/12, lifelike,  highly detailed.",
+      "negative_prompt": "painting, digital art. sketch, blurry"
+    }
+  },
+  {
+    "name": "Photography (Portrait)",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "{prompt}. photography. portraiture. catch light in eyes. one flash. rembrandt lighting. Soft box. dark shadows. High contrast. 80mm lens. F2.8.",
+      "negative_prompt": "painting, digital art. sketch, blurry"
+    }
+  },
+  {
+    "name": "Photography (Black and White)",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "{prompt} photography. natural light. 80mm lens. F1.4. strong contrast, hard light. dark contrast. blurred background. black and white",
+      "negative_prompt": "painting, digital art. sketch, colour+"
+    }
+  },
+  {
+    "name": "Architectural Visualization",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "{prompt}. architectural photography, f/12, luxury, aesthetically pleasing form and function.",
+      "negative_prompt": "painting, digital art. sketch, blurry"
+    }
+  },
+  {
+    "name": "Concept Art (Fantasy)",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "concept artwork of a {prompt}. (digital painterly art style)++, mythological, (textured 2d dry media brushpack)++, glazed brushstrokes, otherworldly. painting+, illustration+",
+      "negative_prompt": "photo. distorted, blurry, out of focus. sketch. (cgi, 3d.)++"
+    }
+  },
+  {
+    "name": "Concept Art (Sci-Fi)",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "(concept art)++, {prompt}, (sleek futurism)++, (textured 2d dry media)++, metallic highlights, digital painting style",
+      "negative_prompt": "photo. distorted, blurry, out of focus. sketch. (cgi, 3d.)++"
+    }
+  },
+  {
+    "name": "Concept Art (Character)",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "(character concept art)++, stylized painterly digital painting of {prompt}, (painterly, impasto. Dry brush.)++",
+      "negative_prompt": "photo. distorted, blurry, out of focus. sketch. (cgi, 3d.)++"
+    }
+  },
+  {
+    "name": "Concept Art (Painterly)",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "{prompt} oil painting. high contrast. impasto. sfumato. chiaroscuro. Palette knife.",
+      "negative_prompt": "photo. smooth. border. frame"
+    }
+  },
+  {
+    "name": "Environment Art",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "{prompt} environment artwork, hyper-realistic digital painting style with cinematic composition, atmospheric, depth and detail, voluminous. textured dry brush 2d media",
+      "negative_prompt": "photo, distorted, blurry, out of focus. sketch."
+    }
+  },
+  {
+    "name": "Interior Design (Visualization)",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "{prompt} interior design photo, gentle shadows, light mid-tones, dimension, mix of smooth and textured surfaces, focus on negative space and clean lines, focus",
+      "negative_prompt": "photo, distorted. sketch."
+    }
+  },
+  {
+    "name": "Product Rendering",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "{prompt} high quality product photography, 3d rendering with key lighting, shallow depth of field, simple plain background, studio lighting.",
+      "negative_prompt": "blurry, sketch, messy, dirty. unfinished."
+    }
+  },
+  {
+    "name": "Sketch",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "{prompt} black and white pencil drawing, off-center composition, cross-hatching for shadows, bold strokes, textured paper. sketch+++",
+      "negative_prompt": "blurry, photo, painting, color. messy, dirty. unfinished. frame, borders."
+    }
+  },
+  {
+    "name": "Line Art",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "{prompt} Line art. bold outline. simplistic. white background. 2d",
+      "negative_prompt": "photo. digital art. greyscale. solid black. painting"
+    }
+  },
+  {
+    "name": "Anime",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "{prompt} anime++, bold outline, cel-shaded coloring, shounen, seinen",
+      "negative_prompt": "(photo)+++. greyscale. solid black. painting"
+    }
+  },
+  {
+    "name": "Illustration",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "{prompt} illustration, bold linework, illustrative details, vector art style, flat coloring",
+      "negative_prompt": "(photo)+++. greyscale. painting, black and white."
+    }
+  },
+  {
+    "name": "Vehicles",
+    "type": "default",
+    "preset_data": {
+      "positive_prompt": "A weird futuristic normal auto, {prompt} elegant design, nice color, nice wheels",
+      "negative_prompt": "sketch. digital art. greyscale. painting"
+    }
+  }
+]

invokeai/app/services/style_preset_records/style_preset_records_base.py

@@ -0,0 +1,42 @@
+from abc import ABC, abstractmethod
+
+from invokeai.app.services.style_preset_records.style_preset_records_common import (
+    PresetType,
+    StylePresetChanges,
+    StylePresetRecordDTO,
+    StylePresetWithoutId,
+)
+
+
+class StylePresetRecordsStorageBase(ABC):
+    """Base class for style preset storage services."""
+
+    @abstractmethod
+    def get(self, style_preset_id: str) -> StylePresetRecordDTO:
+        """Get style preset by id."""
+        pass
+
+    @abstractmethod
+    def create(self, style_preset: StylePresetWithoutId) -> StylePresetRecordDTO:
+        """Creates a style preset."""
+        pass
+
+    @abstractmethod
+    def create_many(self, style_presets: list[StylePresetWithoutId]) -> None:
+        """Creates many style presets."""
+        pass
+
+    @abstractmethod
+    def update(self, style_preset_id: str, changes: StylePresetChanges) -> StylePresetRecordDTO:
+        """Updates a style preset."""
+        pass
+
+    @abstractmethod
+    def delete(self, style_preset_id: str) -> None:
+        """Deletes a style preset."""
+        pass
+
+    @abstractmethod
+    def get_many(self, type: PresetType | None = None) -> list[StylePresetRecordDTO]:
+        """Gets many workflows."""
+        pass

invokeai/app/services/style_preset_records/style_preset_records_common.py

@@ -0,0 +1,139 @@
+import codecs
+import csv
+import json
+from enum import Enum
+from typing import Any, Optional
+
+import pydantic
+from fastapi import UploadFile
+from pydantic import AliasChoices, BaseModel, ConfigDict, Field, TypeAdapter
+
+from invokeai.app.util.metaenum import MetaEnum
+
+
+class StylePresetNotFoundError(Exception):
+    """Raised when a style preset is not found"""
+
+
+class PresetData(BaseModel, extra="forbid"):
+    positive_prompt: str = Field(description="Positive prompt")
+    negative_prompt: str = Field(description="Negative prompt")
+
+
+PresetDataValidator = TypeAdapter(PresetData)
+
+
+class PresetType(str, Enum, metaclass=MetaEnum):
+    User = "user"
+    Default = "default"
+    Project = "project"
+
+
+class StylePresetChanges(BaseModel, extra="forbid"):
+    name: Optional[str] = Field(default=None, description="The style preset's new name.")
+    preset_data: Optional[PresetData] = Field(default=None, description="The updated data for style preset.")
+    type: Optional[PresetType] = Field(description="The updated type of the style preset")
+
+
+class StylePresetWithoutId(BaseModel):
+    name: str = Field(description="The name of the style preset.")
+    preset_data: PresetData = Field(description="The preset data")
+    type: PresetType = Field(description="The type of style preset")
+
+
+class StylePresetRecordDTO(StylePresetWithoutId):
+    id: str = Field(description="The style preset ID.")
+
+    @classmethod
+    def from_dict(cls, data: dict[str, Any]) -> "StylePresetRecordDTO":
+        data["preset_data"] = PresetDataValidator.validate_json(data.get("preset_data", ""))
+        return StylePresetRecordDTOValidator.validate_python(data)
+
+
+StylePresetRecordDTOValidator = TypeAdapter(StylePresetRecordDTO)
+
+
+class StylePresetRecordWithImage(StylePresetRecordDTO):
+    image: Optional[str] = Field(description="The path for image")
+
+
+class StylePresetImportRow(BaseModel):
+    name: str = Field(min_length=1, description="The name of the preset.")
+    positive_prompt: str = Field(
+        default="",
+        description="The positive prompt for the preset.",
+        validation_alias=AliasChoices("positive_prompt", "prompt"),
+    )
+    negative_prompt: str = Field(default="", description="The negative prompt for the preset.")
+
+    model_config = ConfigDict(str_strip_whitespace=True, extra="forbid")
+
+
+StylePresetImportList = list[StylePresetImportRow]
+StylePresetImportListTypeAdapter = TypeAdapter(StylePresetImportList)
+
+
+class UnsupportedFileTypeError(ValueError):
+    """Raised when an unsupported file type is encountered"""
+
+    pass
+
+
+class InvalidPresetImportDataError(ValueError):
+    """Raised when invalid preset import data is encountered"""
+
+    pass
+
+
+async def parse_presets_from_file(file: UploadFile) -> list[StylePresetWithoutId]:
+    """Parses style presets from a file. The file must be a CSV or JSON file.
+
+    If CSV, the file must have the following columns:
+    - name
+    - prompt (or positive_prompt)
+    - negative_prompt
+
+    If JSON, the file must be a list of objects with the following keys:
+    - name
+    - prompt (or positive_prompt)
+    - negative_prompt
+
+    Args:
+        file (UploadFile): The file to parse.
+
+    Returns:
+        list[StylePresetWithoutId]: The parsed style presets.
+
+    Raises:
+        UnsupportedFileTypeError: If the file type is not supported.
+        InvalidPresetImportDataError: If the data in the file is invalid.
+    """
+    if file.content_type not in ["text/csv", "application/json"]:
+        raise UnsupportedFileTypeError()
+
+    if file.content_type == "text/csv":
+        csv_reader = csv.DictReader(codecs.iterdecode(file.file, "utf-8"))
+        data = list(csv_reader)
+    else:  # file.content_type == "application/json":
+        json_data = await file.read()
+        data = json.loads(json_data)
+
+    try:
+        imported_presets = StylePresetImportListTypeAdapter.validate_python(data)
+
+        style_presets: list[StylePresetWithoutId] = []
+
+        for imported in imported_presets:
+            preset_data = PresetData(positive_prompt=imported.positive_prompt, negative_prompt=imported.negative_prompt)
+            style_preset = StylePresetWithoutId(name=imported.name, preset_data=preset_data, type=PresetType.User)
+            style_presets.append(style_preset)
+    except pydantic.ValidationError as e:
+        if file.content_type == "text/csv":
+            msg = "Invalid CSV format: must include columns 'name', 'prompt', and 'negative_prompt' and name cannot be blank"
+        else:  # file.content_type == "application/json":
+            msg = "Invalid JSON format: must be a list of objects with keys 'name', 'prompt', and 'negative_prompt' and name cannot be blank"
+        raise InvalidPresetImportDataError(msg) from e
+    finally:
+        file.file.close()
+
+    return style_presets

invokeai/app/services/style_preset_records/style_preset_records_sqlite.py

@@ -0,0 +1,215 @@
+import json
+from pathlib import Path
+
+from invokeai.app.services.invoker import Invoker
+from invokeai.app.services.shared.sqlite.sqlite_database import SqliteDatabase
+from invokeai.app.services.style_preset_records.style_preset_records_base import StylePresetRecordsStorageBase
+from invokeai.app.services.style_preset_records.style_preset_records_common import (
+    PresetType,
+    StylePresetChanges,
+    StylePresetNotFoundError,
+    StylePresetRecordDTO,
+    StylePresetWithoutId,
+)
+from invokeai.app.util.misc import uuid_string
+
+
+class SqliteStylePresetRecordsStorage(StylePresetRecordsStorageBase):
+    def __init__(self, db: SqliteDatabase) -> None:
+        super().__init__()
+        self._lock = db.lock
+        self._conn = db.conn
+        self._cursor = self._conn.cursor()
+
+    def start(self, invoker: Invoker) -> None:
+        self._invoker = invoker
+        self._sync_default_style_presets()
+
+    def get(self, style_preset_id: str) -> StylePresetRecordDTO:
+        """Gets a style preset by ID."""
+        try:
+            self._lock.acquire()
+            self._cursor.execute(
+                """--sql
+                SELECT *
+                FROM style_presets
+                WHERE id = ?;
+                """,
+                (style_preset_id,),
+            )
+            row = self._cursor.fetchone()
+            if row is None:
+                raise StylePresetNotFoundError(f"Style preset with id {style_preset_id} not found")
+            return StylePresetRecordDTO.from_dict(dict(row))
+        except Exception:
+            self._conn.rollback()
+            raise
+        finally:
+            self._lock.release()
+
+    def create(self, style_preset: StylePresetWithoutId) -> StylePresetRecordDTO:
+        style_preset_id = uuid_string()
+        try:
+            self._lock.acquire()
+            self._cursor.execute(
+                """--sql
+                INSERT OR IGNORE INTO style_presets (
+                    id,
+                    name,
+                    preset_data,
+                    type
+                )
+                VALUES (?, ?, ?, ?);
+                """,
+                (
+                    style_preset_id,
+                    style_preset.name,
+                    style_preset.preset_data.model_dump_json(),
+                    style_preset.type,
+                ),
+            )
+            self._conn.commit()
+        except Exception:
+            self._conn.rollback()
+            raise
+        finally:
+            self._lock.release()
+        return self.get(style_preset_id)
+
+    def create_many(self, style_presets: list[StylePresetWithoutId]) -> None:
+        style_preset_ids = []
+        try:
+            self._lock.acquire()
+            for style_preset in style_presets:
+                style_preset_id = uuid_string()
+                style_preset_ids.append(style_preset_id)
+                self._cursor.execute(
+                    """--sql
+                    INSERT OR IGNORE INTO style_presets (
+                        id,
+                        name,
+                        preset_data,
+                        type
+                    )
+                    VALUES (?, ?, ?, ?);
+                    """,
+                    (
+                        style_preset_id,
+                        style_preset.name,
+                        style_preset.preset_data.model_dump_json(),
+                        style_preset.type,
+                    ),
+                )
+            self._conn.commit()
+        except Exception:
+            self._conn.rollback()
+            raise
+        finally:
+            self._lock.release()
+
+        return None
+
+    def update(self, style_preset_id: str, changes: StylePresetChanges) -> StylePresetRecordDTO:
+        try:
+            self._lock.acquire()
+            # Change the name of a style preset
+            if changes.name is not None:
+                self._cursor.execute(
+                    """--sql
+                    UPDATE style_presets
+                    SET name = ?
+                    WHERE id = ?;
+                    """,
+                    (changes.name, style_preset_id),
+                )
+
+            # Change the preset data for a style preset
+            if changes.preset_data is not None:
+                self._cursor.execute(
+                    """--sql
+                    UPDATE style_presets
+                    SET preset_data = ?
+                    WHERE id = ?;
+                    """,
+                    (changes.preset_data.model_dump_json(), style_preset_id),
+                )
+
+            self._conn.commit()
+        except Exception:
+            self._conn.rollback()
+            raise
+        finally:
+            self._lock.release()
+        return self.get(style_preset_id)
+
+    def delete(self, style_preset_id: str) -> None:
+        try:
+            self._lock.acquire()
+            self._cursor.execute(
+                """--sql
+                DELETE from style_presets
+                WHERE id = ?;
+                """,
+                (style_preset_id,),
+            )
+            self._conn.commit()
+        except Exception:
+            self._conn.rollback()
+            raise
+        finally:
+            self._lock.release()
+        return None
+
+    def get_many(self, type: PresetType | None = None) -> list[StylePresetRecordDTO]:
+        try:
+            self._lock.acquire()
+            main_query = """
+                SELECT
+                    *
+                FROM style_presets
+                """
+
+            if type is not None:
+                main_query += "WHERE type = ? "
+
+            main_query += "ORDER BY LOWER(name) ASC"
+
+            if type is not None:
+                self._cursor.execute(main_query, (type,))
+            else:
+                self._cursor.execute(main_query)
+
+            rows = self._cursor.fetchall()
+            style_presets = [StylePresetRecordDTO.from_dict(dict(row)) for row in rows]
+
+            return style_presets
+        except Exception:
+            self._conn.rollback()
+            raise
+        finally:
+            self._lock.release()
+
+    def _sync_default_style_presets(self) -> None:
+        """Syncs default style presets to the database. Internal use only."""
+
+        # First delete all existing default style presets
+        try:
+            self._lock.acquire()
+            self._cursor.execute(
+                """--sql
+                DELETE FROM style_presets
+                WHERE type = "default";
+                """
+            )
+            self._conn.commit()
+        except Exception:
+            self._conn.rollback()
+            raise
+        finally:
+            self._lock.release()
+        # Next, parse and create the default style presets
+        with self._lock, open(Path(__file__).parent / Path("default_style_presets.json"), "r") as file:
+            presets = json.load(file)
+            for preset in presets:
+                style_preset = StylePresetWithoutId.model_validate(preset)
+                self.create(style_preset)

invokeai/app/services/urls/urls_base.py

@@ -13,3 +13,8 @@ class UrlServiceBase(ABC):
     def get_model_image_url(self, model_key: str) -> str:
         """Gets the URL for a model image"""
         pass
+
+    @abstractmethod
+    def get_style_preset_image_url(self, style_preset_id: str) -> str:
+        """Gets the URL for a style preset image"""
+        pass

invokeai/app/services/urls/urls_default.py

@@ -19,3 +19,6 @@ class LocalUrlService(UrlServiceBase):
 
     def get_model_image_url(self, model_key: str) -> str:
         return f"{self._base_url_v2}/models/i/{model_key}/image"
+
+    def get_style_preset_image_url(self, style_preset_id: str) -> str:
+        return f"{self._base_url}/style_presets/i/{style_preset_id}/image"

invokeai/app/services/workflow_records/default_workflows/FLUX Image to Image.json

@@ -0,0 +1,407 @@
+{
+  "name": "FLUX Image to Image",
+  "author": "InvokeAI",
+  "description": "A simple image-to-image workflow using a FLUX dev model. ",
+  "version": "1.0.4",
+  "contact": "",
+  "tags": "image2image, flux, image-to-image",
+  "notes": "Prerequisite model downloads: T5 Encoder, CLIP-L Encoder, and FLUX VAE. Quantized and un-quantized versions can be found in the starter models tab within your Model Manager. We recommend using FLUX dev models for image-to-image workflows. The image-to-image performance with FLUX schnell models is poor.",
+  "exposedFields": [
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "model"
+    },
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "t5_encoder_model"
+    },
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "clip_embed_model"
+    },
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "vae_model"
+    },
+    {
+      "nodeId": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "fieldName": "denoising_start"
+    },
+    {
+      "nodeId": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "fieldName": "prompt"
+    },
+    {
+      "nodeId": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "fieldName": "num_steps"
+    }
+  ],
+  "meta": {
+    "version": "3.0.0",
+    "category": "default"
+  },
+  "nodes": [
+    {
+      "id": "2981a67c-480f-4237-9384-26b68dbf912b",
+      "type": "invocation",
+      "data": {
+        "id": "2981a67c-480f-4237-9384-26b68dbf912b",
+        "type": "flux_vae_encode",
+        "version": "1.0.0",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": true,
+        "inputs": {
+          "image": {
+            "name": "image",
+            "label": "",
+            "value": {
+              "image_name": "8a5c62aa-9335-45d2-9c71-89af9fc1f8d4.png"
+            }
+          },
+          "vae": {
+            "name": "vae",
+            "label": ""
+          }
+        }
+      },
+      "position": {
+        "x": 732.7680166609682,
+        "y": -24.37398171806909
+      }
+    },
+    {
+      "id": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "type": "invocation",
+      "data": {
+        "id": "ace0258f-67d7-4eee-a218-6fff27065214",
+        "type": "flux_denoise",
+        "version": "1.0.0",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": true,
+        "inputs": {
+          "board": {
+            "name": "board",
+            "label": ""
+          },
+          "metadata": {
+            "name": "metadata",
+            "label": ""
+          },
+          "latents": {
+            "name": "latents",
+            "label": ""
+          },
+          "denoise_mask": {
+            "name": "denoise_mask",
+            "label": ""
+          },
+          "denoising_start": {
+            "name": "denoising_start",
+            "label": "",
+            "value": 0.04
+          },
+          "denoising_end": {
+            "name": "denoising_end",
+            "label": "",
+            "value": 1
+          },
+          "transformer": {
+            "name": "transformer",
+            "label": ""
+          },
+          "positive_text_conditioning": {
+            "name": "positive_text_conditioning",
+            "label": ""
+          },
+          "width": {
+            "name": "width",
+            "label": "",
+            "value": 1024
+          },
+          "height": {
+            "name": "height",
+            "label": "",
+            "value": 1024
+          },
+          "num_steps": {
+            "name": "num_steps",
+            "label": "Steps (Recommend 30 for Dev, 4 for Schnell)",
+            "value": 30
+          },
+          "guidance": {
+            "name": "guidance",
+            "label": "",
+            "value": 4
+          },
+          "seed": {
+            "name": "seed",
+            "label": "",
+            "value": 0
+          }
+        }
+      },
+      "position": {
+        "x": 1182.8836633018684,
+        "y": -251.38882958913183
+      }
+    },
+    {
+      "id": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+      "type": "invocation",
+      "data": {
+        "id": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+        "type": "flux_vae_decode",
+        "version": "1.0.0",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": false,
+        "useCache": true,
+        "inputs": {
+          "board": {
+            "name": "board",
+            "label": ""
+          },
+          "metadata": {
+            "name": "metadata",
+            "label": ""
+          },
+          "latents": {
+            "name": "latents",
+            "label": ""
+          },
+          "vae": {
+            "name": "vae",
+            "label": ""
+          }
+        }
+      },
+      "position": {
+        "x": 1575.5797431839133,
+        "y": -209.00150975507415
+      }
+    },
+    {
+      "id": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "type": "invocation",
+      "data": {
+        "id": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+        "type": "flux_model_loader",
+        "version": "1.0.4",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": false,
+        "inputs": {
+          "model": {
+            "name": "model",
+            "label": "Model (dev variant recommended for Image-to-Image)"
+          },
+          "t5_encoder_model": {
+            "name": "t5_encoder_model",
+            "label": ""
+          },
+          "clip_embed_model": {
+            "name": "clip_embed_model",
+            "label": "",
+            "value": {
+              "key": "fa23a584-b623-415d-832a-21b5098ff1a1",
+              "hash": "blake3:17c19f0ef941c3b7609a9c94a659ca5364de0be364a91d4179f0e39ba17c3b70",
+              "name": "clip-vit-large-patch14",
+              "base": "any",
+              "type": "clip_embed"
+            }
+          },
+          "vae_model": {
+            "name": "vae_model",
+            "label": "",
+            "value": {
+              "key": "74fc82ba-c0a8-479d-a890-2126f82da758",
+              "hash": "blake3:ce21cb76364aa6e2421311cf4a4b5eb052a76c4f1cd207b50703d8978198a068",
+              "name": "FLUX.1-schnell_ae",
+              "base": "flux",
+              "type": "vae"
+            }
+          }
+        }
+      },
+      "position": {
+        "x": 328.1809894659957,
+        "y": -90.2241133566946
+      }
+    },
+    {
+      "id": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "type": "invocation",
+      "data": {
+        "id": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+        "type": "flux_text_encoder",
+        "version": "1.0.0",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": true,
+        "inputs": {
+          "clip": {
+            "name": "clip",
+            "label": ""
+          },
+          "t5_encoder": {
+            "name": "t5_encoder",
+            "label": ""
+          },
+          "t5_max_seq_len": {
+            "name": "t5_max_seq_len",
+            "label": "T5 Max Seq Len",
+            "value": 256
+          },
+          "prompt": {
+            "name": "prompt",
+            "label": "",
+            "value": "a cat wearing a birthday hat"
+          }
+        }
+      },
+      "position": {
+        "x": 745.8823365057267,
+        "y": -299.60249175851914
+      }
+    },
+    {
+      "id": "4754c534-a5f3-4ad0-9382-7887985e668c",
+      "type": "invocation",
+      "data": {
+        "id": "4754c534-a5f3-4ad0-9382-7887985e668c",
+        "type": "rand_int",
+        "version": "1.0.1",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": false,
+        "inputs": {
+          "low": {
+            "name": "low",
+            "label": "",
+            "value": 0
+          },
+          "high": {
+            "name": "high",
+            "label": "",
+            "value": 2147483647
+          }
+        }
+      },
+      "position": {
+        "x": 725.834098928012,
+        "y": 496.2710031089931
+      }
+    }
+  ],
+  "edges": [
+    {
+      "id": "reactflow__edge-2981a67c-480f-4237-9384-26b68dbf912bheight-ace0258f-67d7-4eee-a218-6fff27065214height",
+      "type": "default",
+      "source": "2981a67c-480f-4237-9384-26b68dbf912b",
+      "target": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "sourceHandle": "height",
+      "targetHandle": "height"
+    },
+    {
+      "id": "reactflow__edge-2981a67c-480f-4237-9384-26b68dbf912bwidth-ace0258f-67d7-4eee-a218-6fff27065214width",
+      "type": "default",
+      "source": "2981a67c-480f-4237-9384-26b68dbf912b",
+      "target": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "sourceHandle": "width",
+      "targetHandle": "width"
+    },
+    {
+      "id": "reactflow__edge-2981a67c-480f-4237-9384-26b68dbf912blatents-ace0258f-67d7-4eee-a218-6fff27065214latents",
+      "type": "default",
+      "source": "2981a67c-480f-4237-9384-26b68dbf912b",
+      "target": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "sourceHandle": "latents",
+      "targetHandle": "latents"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90vae-2981a67c-480f-4237-9384-26b68dbf912bvae",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "2981a67c-480f-4237-9384-26b68dbf912b",
+      "sourceHandle": "vae",
+      "targetHandle": "vae"
+    },
+    {
+      "id": "reactflow__edge-ace0258f-67d7-4eee-a218-6fff27065214latents-7e5172eb-48c1-44db-a770-8fd83e1435d1latents",
+      "type": "default",
+      "source": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "target": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+      "sourceHandle": "latents",
+      "targetHandle": "latents"
+    },
+    {
+      "id": "reactflow__edge-4754c534-a5f3-4ad0-9382-7887985e668cvalue-ace0258f-67d7-4eee-a218-6fff27065214seed",
+      "type": "default",
+      "source": "4754c534-a5f3-4ad0-9382-7887985e668c",
+      "target": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "sourceHandle": "value",
+      "targetHandle": "seed"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90transformer-ace0258f-67d7-4eee-a218-6fff27065214transformer",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "sourceHandle": "transformer",
+      "targetHandle": "transformer"
+    },
+    {
+      "id": "reactflow__edge-01f674f8-b3d1-4df1-acac-6cb8e0bfb63cconditioning-ace0258f-67d7-4eee-a218-6fff27065214positive_text_conditioning",
+      "type": "default",
+      "source": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "target": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "sourceHandle": "conditioning",
+      "targetHandle": "positive_text_conditioning"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90vae-7e5172eb-48c1-44db-a770-8fd83e1435d1vae",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+      "sourceHandle": "vae",
+      "targetHandle": "vae"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90max_seq_len-01f674f8-b3d1-4df1-acac-6cb8e0bfb63ct5_max_seq_len",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "sourceHandle": "max_seq_len",
+      "targetHandle": "t5_max_seq_len"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90t5_encoder-01f674f8-b3d1-4df1-acac-6cb8e0bfb63ct5_encoder",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "sourceHandle": "t5_encoder",
+      "targetHandle": "t5_encoder"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90clip-01f674f8-b3d1-4df1-acac-6cb8e0bfb63cclip",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "sourceHandle": "clip",
+      "targetHandle": "clip"
+    }
+  ]
+}

invokeai/app/services/workflow_records/default_workflows/Flux Text to Image.json

@@ -0,0 +1,326 @@
+{
+  "name": "FLUX Text to Image",
+  "author": "InvokeAI",
+  "description": "A simple text-to-image workflow using FLUX dev or schnell models.",
+  "version": "1.0.4",
+  "contact": "",
+  "tags": "text2image, flux",
+  "notes": "Prerequisite model downloads: T5 Encoder, CLIP-L Encoder, and FLUX VAE. Quantized and un-quantized versions can be found in the starter models tab within your Model Manager. We recommend 4 steps for FLUX schnell models and 30 steps for FLUX dev models.",
+  "exposedFields": [
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "model"
+    },
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "t5_encoder_model"
+    },
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "clip_embed_model"
+    },
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "vae_model"
+    },
+    {
+      "nodeId": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "fieldName": "prompt"
+    },
+    {
+      "nodeId": "4fe24f07-f906-4f55-ab2c-9beee56ef5bd",
+      "fieldName": "num_steps"
+    }
+  ],
+  "meta": {
+    "version": "3.0.0",
+    "category": "default"
+  },
+  "nodes": [
+    {
+      "id": "4fe24f07-f906-4f55-ab2c-9beee56ef5bd",
+      "type": "invocation",
+      "data": {
+        "id": "4fe24f07-f906-4f55-ab2c-9beee56ef5bd",
+        "type": "flux_denoise",
+        "version": "1.0.0",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": true,
+        "inputs": {
+          "board": {
+            "name": "board",
+            "label": ""
+          },
+          "metadata": {
+            "name": "metadata",
+            "label": ""
+          },
+          "latents": {
+            "name": "latents",
+            "label": ""
+          },
+          "denoise_mask": {
+            "name": "denoise_mask",
+            "label": ""
+          },
+          "denoising_start": {
+            "name": "denoising_start",
+            "label": "",
+            "value": 0
+          },
+          "denoising_end": {
+            "name": "denoising_end",
+            "label": "",
+            "value": 1
+          },
+          "transformer": {
+            "name": "transformer",
+            "label": ""
+          },
+          "positive_text_conditioning": {
+            "name": "positive_text_conditioning",
+            "label": ""
+          },
+          "width": {
+            "name": "width",
+            "label": "",
+            "value": 1024
+          },
+          "height": {
+            "name": "height",
+            "label": "",
+            "value": 1024
+          },
+          "num_steps": {
+            "name": "num_steps",
+            "label": "Steps (Recommend 30 for Dev, 4 for Schnell)",
+            "value": 30
+          },
+          "guidance": {
+            "name": "guidance",
+            "label": "",
+            "value": 4
+          },
+          "seed": {
+            "name": "seed",
+            "label": "",
+            "value": 0
+          }
+        }
+      },
+      "position": {
+        "x": 1186.1868226120378,
+        "y": -214.9459927686657
+      }
+    },
+    {
+      "id": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+      "type": "invocation",
+      "data": {
+        "id": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+        "type": "flux_vae_decode",
+        "version": "1.0.0",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": false,
+        "useCache": true,
+        "inputs": {
+          "board": {
+            "name": "board",
+            "label": ""
+          },
+          "metadata": {
+            "name": "metadata",
+            "label": ""
+          },
+          "latents": {
+            "name": "latents",
+            "label": ""
+          },
+          "vae": {
+            "name": "vae",
+            "label": ""
+          }
+        }
+      },
+      "position": {
+        "x": 1575.5797431839133,
+        "y": -209.00150975507415
+      }
+    },
+    {
+      "id": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "type": "invocation",
+      "data": {
+        "id": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+        "type": "flux_model_loader",
+        "version": "1.0.4",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": false,
+        "inputs": {
+          "model": {
+            "name": "model",
+            "label": ""
+          },
+          "t5_encoder_model": {
+            "name": "t5_encoder_model",
+            "label": ""
+          },
+          "clip_embed_model": {
+            "name": "clip_embed_model",
+            "label": ""
+          },
+          "vae_model": {
+            "name": "vae_model",
+            "label": ""
+          }
+        }
+      },
+      "position": {
+        "x": 381.1882713063478,
+        "y": -95.89663532854017
+      }
+    },
+    {
+      "id": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "type": "invocation",
+      "data": {
+        "id": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+        "type": "flux_text_encoder",
+        "version": "1.0.0",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": true,
+        "inputs": {
+          "clip": {
+            "name": "clip",
+            "label": ""
+          },
+          "t5_encoder": {
+            "name": "t5_encoder",
+            "label": ""
+          },
+          "t5_max_seq_len": {
+            "name": "t5_max_seq_len",
+            "label": "T5 Max Seq Len",
+            "value": 256
+          },
+          "prompt": {
+            "name": "prompt",
+            "label": "",
+            "value": "a cat"
+          }
+        }
+      },
+      "position": {
+        "x": 778.4899149328337,
+        "y": -100.36469216659502
+      }
+    },
+    {
+      "id": "4754c534-a5f3-4ad0-9382-7887985e668c",
+      "type": "invocation",
+      "data": {
+        "id": "4754c534-a5f3-4ad0-9382-7887985e668c",
+        "type": "rand_int",
+        "version": "1.0.1",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": false,
+        "inputs": {
+          "low": {
+            "name": "low",
+            "label": "",
+            "value": 0
+          },
+          "high": {
+            "name": "high",
+            "label": "",
+            "value": 2147483647
+          }
+        }
+      },
+      "position": {
+        "x": 800.9667463219505,
+        "y": 285.8297267547506
+      }
+    }
+  ],
+  "edges": [
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90transformer-4fe24f07-f906-4f55-ab2c-9beee56ef5bdtransformer",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "4fe24f07-f906-4f55-ab2c-9beee56ef5bd",
+      "sourceHandle": "transformer",
+      "targetHandle": "transformer"
+    },
+    {
+      "id": "reactflow__edge-01f674f8-b3d1-4df1-acac-6cb8e0bfb63cconditioning-4fe24f07-f906-4f55-ab2c-9beee56ef5bdpositive_text_conditioning",
+      "type": "default",
+      "source": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "target": "4fe24f07-f906-4f55-ab2c-9beee56ef5bd",
+      "sourceHandle": "conditioning",
+      "targetHandle": "positive_text_conditioning"
+    },
+    {
+      "id": "reactflow__edge-4754c534-a5f3-4ad0-9382-7887985e668cvalue-4fe24f07-f906-4f55-ab2c-9beee56ef5bdseed",
+      "type": "default",
+      "source": "4754c534-a5f3-4ad0-9382-7887985e668c",
+      "target": "4fe24f07-f906-4f55-ab2c-9beee56ef5bd",
+      "sourceHandle": "value",
+      "targetHandle": "seed"
+    },
+    {
+      "id": "reactflow__edge-4fe24f07-f906-4f55-ab2c-9beee56ef5bdlatents-7e5172eb-48c1-44db-a770-8fd83e1435d1latents",
+      "type": "default",
+      "source": "4fe24f07-f906-4f55-ab2c-9beee56ef5bd",
+      "target": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+      "sourceHandle": "latents",
+      "targetHandle": "latents"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90vae-7e5172eb-48c1-44db-a770-8fd83e1435d1vae",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+      "sourceHandle": "vae",
+      "targetHandle": "vae"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90max_seq_len-01f674f8-b3d1-4df1-acac-6cb8e0bfb63ct5_max_seq_len",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "sourceHandle": "max_seq_len",
+      "targetHandle": "t5_max_seq_len"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90t5_encoder-01f674f8-b3d1-4df1-acac-6cb8e0bfb63ct5_encoder",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "sourceHandle": "t5_encoder",
+      "targetHandle": "t5_encoder"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90clip-01f674f8-b3d1-4df1-acac-6cb8e0bfb63cclip",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "sourceHandle": "clip",
+      "targetHandle": "clip"
+    }
+  ]
+}

invokeai/app/util/custom_openapi.py

@@ -81,7 +81,7 @@ def get_openapi_func(
         # Add the output map to the schema
         openapi_schema["components"]["schemas"]["InvocationOutputMap"] = {
             "type": "object",
-            "properties": invocation_output_map_properties,
+            "properties": dict(sorted(invocation_output_map_properties.items())),
             "required": invocation_output_map_required,
         }
 

invokeai/app/util/step_callback.py

@@ -38,6 +38,25 @@ SD1_5_LATENT_RGB_FACTORS = [
     [-0.1307, -0.1874, -0.7445],  # L4
 ]
 
+FLUX_LATENT_RGB_FACTORS = [
+    [-0.0412, 0.0149, 0.0521],
+    [0.0056, 0.0291, 0.0768],
+    [0.0342, -0.0681, -0.0427],
+    [-0.0258, 0.0092, 0.0463],
+    [0.0863, 0.0784, 0.0547],
+    [-0.0017, 0.0402, 0.0158],
+    [0.0501, 0.1058, 0.1152],
+    [-0.0209, -0.0218, -0.0329],
+    [-0.0314, 0.0083, 0.0896],
+    [0.0851, 0.0665, -0.0472],
+    [-0.0534, 0.0238, -0.0024],
+    [0.0452, -0.0026, 0.0048],
+    [0.0892, 0.0831, 0.0881],
+    [-0.1117, -0.0304, -0.0789],
+    [0.0027, -0.0479, -0.0043],
+    [-0.1146, -0.0827, -0.0598],
+]
+
 
 def sample_to_lowres_estimated_image(
     samples: torch.Tensor, latent_rgb_factors: torch.Tensor, smooth_matrix: Optional[torch.Tensor] = None
@@ -94,3 +113,32 @@ def stable_diffusion_step_callback(
         intermediate_state,
         ProgressImage(dataURL=dataURL, width=width, height=height),
     )
+
+
+def flux_step_callback(
+    context_data: "InvocationContextData",
+    intermediate_state: PipelineIntermediateState,
+    events: "EventServiceBase",
+    is_canceled: Callable[[], bool],
+) -> None:
+    if is_canceled():
+        raise CanceledException
+    sample = intermediate_state.latents
+    latent_rgb_factors = torch.tensor(FLUX_LATENT_RGB_FACTORS, dtype=sample.dtype, device=sample.device)
+    latent_image_perm = sample.permute(1, 2, 0).to(dtype=sample.dtype, device=sample.device)
+    latent_image = latent_image_perm @ latent_rgb_factors
+    latents_ubyte = (
+        ((latent_image + 1) / 2).clamp(0, 1).mul(0xFF)  # change scale from -1..1 to 0..1  # to 0..255
+    ).to(device="cpu", dtype=torch.uint8)
+    image = Image.fromarray(latents_ubyte.cpu().numpy())
+    (width, height) = image.size
+    width *= 8
+    height *= 8
+    dataURL = image_to_dataURL(image, image_format="JPEG")
+
+    events.emit_invocation_denoise_progress(
+        context_data.queue_item,
+        context_data.invocation,
+        intermediate_state,
+        ProgressImage(dataURL=dataURL, width=width, height=height),
+    )

invokeai/backend/flux/denoise.py

@@ -0,0 +1,56 @@
+from typing import Callable
+
+import torch
+from tqdm import tqdm
+
+from invokeai.backend.flux.inpaint_extension import InpaintExtension
+from invokeai.backend.flux.model import Flux
+from invokeai.backend.stable_diffusion.diffusers_pipeline import PipelineIntermediateState
+
+
+def denoise(
+    model: Flux,
+    # model input
+    img: torch.Tensor,
+    img_ids: torch.Tensor,
+    txt: torch.Tensor,
+    txt_ids: torch.Tensor,
+    vec: torch.Tensor,
+    # sampling parameters
+    timesteps: list[float],
+    step_callback: Callable[[PipelineIntermediateState], None],
+    guidance: float,
+    inpaint_extension: InpaintExtension | None,
+):
+    step = 0
+    # guidance_vec is ignored for schnell.
+    guidance_vec = torch.full((img.shape[0],), guidance, device=img.device, dtype=img.dtype)
+    for t_curr, t_prev in tqdm(list(zip(timesteps[:-1], timesteps[1:], strict=True))):
+        t_vec = torch.full((img.shape[0],), t_curr, dtype=img.dtype, device=img.device)
+        pred = model(
+            img=img,
+            img_ids=img_ids,
+            txt=txt,
+            txt_ids=txt_ids,
+            y=vec,
+            timesteps=t_vec,
+            guidance=guidance_vec,
+        )
+        preview_img = img - t_curr * pred
+        img = img + (t_prev - t_curr) * pred
+
+        if inpaint_extension is not None:
+            img = inpaint_extension.merge_intermediate_latents_with_init_latents(img, t_prev)
+
+        step_callback(
+            PipelineIntermediateState(
+                step=step,
+                order=1,
+                total_steps=len(timesteps),
+                timestep=int(t_curr),
+                latents=preview_img,
+            ),
+        )
+        step += 1
+
+    return img

invokeai/backend/flux/inpaint_extension.py

@@ -0,0 +1,35 @@
+import torch
+
+
+class InpaintExtension:
+    """A class for managing inpainting with FLUX."""
+
+    def __init__(self, init_latents: torch.Tensor, inpaint_mask: torch.Tensor, noise: torch.Tensor):
+        """Initialize InpaintExtension.
+
+        Args:
+            init_latents (torch.Tensor): The initial latents (i.e. un-noised at timestep 0). In 'packed' format.
+            inpaint_mask (torch.Tensor): A mask specifying which elements to inpaint. Range [0, 1]. Values of 1 will be
+                re-generated. Values of 0 will remain unchanged. Values between 0 and 1 can be used to blend the
+                inpainted region with the background. In 'packed' format.
+            noise (torch.Tensor): The noise tensor used to noise the init_latents. In 'packed' format.
+        """
+        assert init_latents.shape == inpaint_mask.shape == noise.shape
+        self._init_latents = init_latents
+        self._inpaint_mask = inpaint_mask
+        self._noise = noise
+
+    def merge_intermediate_latents_with_init_latents(
+        self, intermediate_latents: torch.Tensor, timestep: float
+    ) -> torch.Tensor:
+        """Merge the intermediate latents with the initial latents for the current timestep using the inpaint mask. I.e.
+        update the intermediate latents to keep the regions that are not being inpainted on the correct noise
+        trajectory.
+
+        This function should be called after each denoising step.
+        """
+        # Noise the init latents for the current timestep.
+        noised_init_latents = self._noise * timestep + (1.0 - timestep) * self._init_latents
+
+        # Merge the intermediate latents with the noised_init_latents using the inpaint_mask.
+        return intermediate_latents * self._inpaint_mask + noised_init_latents * (1.0 - self._inpaint_mask)

invokeai/backend/flux/math.py

@@ -0,0 +1,32 @@
+# Initially pulled from https://github.com/black-forest-labs/flux
+
+import torch
+from einops import rearrange
+from torch import Tensor
+
+
+def attention(q: Tensor, k: Tensor, v: Tensor, pe: Tensor) -> Tensor:
+    q, k = apply_rope(q, k, pe)
+
+    x = torch.nn.functional.scaled_dot_product_attention(q, k, v)
+    x = rearrange(x, "B H L D -> B L (H D)")
+
+    return x
+
+
+def rope(pos: Tensor, dim: int, theta: int) -> Tensor:
+    assert dim % 2 == 0
+    scale = torch.arange(0, dim, 2, dtype=torch.float64, device=pos.device) / dim
+    omega = 1.0 / (theta**scale)
+    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()
+
+
+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_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)

invokeai/backend/flux/model.py

@@ -0,0 +1,117 @@
+# Initially pulled from https://github.com/black-forest-labs/flux
+
+from dataclasses import dataclass
+
+import torch
+from torch import Tensor, nn
+
+from invokeai.backend.flux.modules.layers import (
+    DoubleStreamBlock,
+    EmbedND,
+    LastLayer,
+    MLPEmbedder,
+    SingleStreamBlock,
+    timestep_embedding,
+)
+
+
+@dataclass
+class FluxParams:
+    in_channels: int
+    vec_in_dim: int
+    context_in_dim: int
+    hidden_size: int
+    mlp_ratio: float
+    num_heads: int
+    depth: int
+    depth_single_blocks: int
+    axes_dim: list[int]
+    theta: int
+    qkv_bias: bool
+    guidance_embed: bool
+
+
+class Flux(nn.Module):
+    """
+    Transformer model for flow matching on sequences.
+    """
+
+    def __init__(self, params: FluxParams):
+        super().__init__()
+
+        self.params = params
+        self.in_channels = params.in_channels
+        self.out_channels = 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
+        if sum(params.axes_dim) != pe_dim:
+            raise ValueError(f"Got {params.axes_dim} but expected positional dim {pe_dim}")
+        self.hidden_size = params.hidden_size
+        self.num_heads = params.num_heads
+        self.pe_embedder = EmbedND(dim=pe_dim, theta=params.theta, axes_dim=params.axes_dim)
+        self.img_in = nn.Linear(self.in_channels, self.hidden_size, bias=True)
+        self.time_in = MLPEmbedder(in_dim=256, hidden_dim=self.hidden_size)
+        self.vector_in = MLPEmbedder(params.vec_in_dim, self.hidden_size)
+        self.guidance_in = (
+            MLPEmbedder(in_dim=256, hidden_dim=self.hidden_size) if params.guidance_embed else nn.Identity()
+        )
+        self.txt_in = nn.Linear(params.context_in_dim, self.hidden_size)
+
+        self.double_blocks = nn.ModuleList(
+            [
+                DoubleStreamBlock(
+                    self.hidden_size,
+                    self.num_heads,
+                    mlp_ratio=params.mlp_ratio,
+                    qkv_bias=params.qkv_bias,
+                )
+                for _ in range(params.depth)
+            ]
+        )
+
+        self.single_blocks = nn.ModuleList(
+            [
+                SingleStreamBlock(self.hidden_size, self.num_heads, mlp_ratio=params.mlp_ratio)
+                for _ in range(params.depth_single_blocks)
+            ]
+        )
+
+        self.final_layer = LastLayer(self.hidden_size, 1, self.out_channels)
+
+    def forward(
+        self,
+        img: Tensor,
+        img_ids: Tensor,
+        txt: Tensor,
+        txt_ids: Tensor,
+        timesteps: Tensor,
+        y: Tensor,
+        guidance: Tensor | None = None,
+    ) -> Tensor:
+        if img.ndim != 3 or txt.ndim != 3:
+            raise ValueError("Input img and txt tensors must have 3 dimensions.")
+
+        # running on sequences img
+        img = self.img_in(img)
+        vec = self.time_in(timestep_embedding(timesteps, 256))
+        if self.params.guidance_embed:
+            if guidance is None:
+                raise ValueError("Didn't get guidance strength for guidance distilled model.")
+            vec = vec + self.guidance_in(timestep_embedding(guidance, 256))
+        vec = vec + self.vector_in(y)
+        txt = self.txt_in(txt)
+
+        ids = torch.cat((txt_ids, img_ids), dim=1)
+        pe = self.pe_embedder(ids)
+
+        for block in self.double_blocks:
+            img, txt = block(img=img, txt=txt, vec=vec, pe=pe)
+
+        img = torch.cat((txt, img), 1)
+        for block in self.single_blocks:
+            img = block(img, vec=vec, pe=pe)
+        img = img[:, txt.shape[1] :, ...]
+
+        img = self.final_layer(img, vec)  # (N, T, patch_size ** 2 * out_channels)
+        return img

invokeai/backend/flux/modules/autoencoder.py

@@ -0,0 +1,324 @@
+# Initially pulled from https://github.com/black-forest-labs/flux
+
+from dataclasses import dataclass
+
+import torch
+from einops import rearrange
+from torch import Tensor, nn
+
+
+@dataclass
+class AutoEncoderParams:
+    resolution: int
+    in_channels: int
+    ch: int
+    out_ch: int
+    ch_mult: list[int]
+    num_res_blocks: int
+    z_channels: int
+    scale_factor: float
+    shift_factor: float
+
+
+class AttnBlock(nn.Module):
+    def __init__(self, in_channels: int):
+        super().__init__()
+        self.in_channels = in_channels
+
+        self.norm = nn.GroupNorm(num_groups=32, num_channels=in_channels, eps=1e-6, affine=True)
+
+        self.q = nn.Conv2d(in_channels, in_channels, kernel_size=1)
+        self.k = nn.Conv2d(in_channels, in_channels, kernel_size=1)
+        self.v = nn.Conv2d(in_channels, in_channels, kernel_size=1)
+        self.proj_out = nn.Conv2d(in_channels, in_channels, kernel_size=1)
+
+    def attention(self, h_: Tensor) -> Tensor:
+        h_ = self.norm(h_)
+        q = self.q(h_)
+        k = self.k(h_)
+        v = self.v(h_)
+
+        b, c, h, w = q.shape
+        q = rearrange(q, "b c h w -> b 1 (h w) c").contiguous()
+        k = rearrange(k, "b c h w -> b 1 (h w) c").contiguous()
+        v = rearrange(v, "b c h w -> b 1 (h w) c").contiguous()
+        h_ = nn.functional.scaled_dot_product_attention(q, k, v)
+
+        return rearrange(h_, "b 1 (h w) c -> b c h w", h=h, w=w, c=c, b=b)
+
+    def forward(self, x: Tensor) -> Tensor:
+        return x + self.proj_out(self.attention(x))
+
+
+class ResnetBlock(nn.Module):
+    def __init__(self, in_channels: int, out_channels: int):
+        super().__init__()
+        self.in_channels = in_channels
+        out_channels = in_channels if out_channels is None else out_channels
+        self.out_channels = out_channels
+
+        self.norm1 = nn.GroupNorm(num_groups=32, num_channels=in_channels, eps=1e-6, affine=True)
+        self.conv1 = nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=1, padding=1)
+        self.norm2 = nn.GroupNorm(num_groups=32, num_channels=out_channels, eps=1e-6, affine=True)
+        self.conv2 = nn.Conv2d(out_channels, out_channels, kernel_size=3, stride=1, padding=1)
+        if self.in_channels != self.out_channels:
+            self.nin_shortcut = nn.Conv2d(in_channels, out_channels, kernel_size=1, stride=1, padding=0)
+
+    def forward(self, x):
+        h = x
+        h = self.norm1(h)
+        h = torch.nn.functional.silu(h)
+        h = self.conv1(h)
+
+        h = self.norm2(h)
+        h = torch.nn.functional.silu(h)
+        h = self.conv2(h)
+
+        if self.in_channels != self.out_channels:
+            x = self.nin_shortcut(x)
+
+        return x + h
+
+
+class Downsample(nn.Module):
+    def __init__(self, in_channels: int):
+        super().__init__()
+        # no asymmetric padding in torch conv, must do it ourselves
+        self.conv = nn.Conv2d(in_channels, in_channels, kernel_size=3, stride=2, padding=0)
+
+    def forward(self, x: Tensor):
+        pad = (0, 1, 0, 1)
+        x = nn.functional.pad(x, pad, mode="constant", value=0)
+        x = self.conv(x)
+        return x
+
+
+class Upsample(nn.Module):
+    def __init__(self, in_channels: int):
+        super().__init__()
+        self.conv = nn.Conv2d(in_channels, in_channels, kernel_size=3, stride=1, padding=1)
+
+    def forward(self, x: Tensor):
+        x = nn.functional.interpolate(x, scale_factor=2.0, mode="nearest")
+        x = self.conv(x)
+        return x
+
+
+class Encoder(nn.Module):
+    def __init__(
+        self,
+        resolution: int,
+        in_channels: int,
+        ch: int,
+        ch_mult: list[int],
+        num_res_blocks: int,
+        z_channels: int,
+    ):
+        super().__init__()
+        self.ch = ch
+        self.num_resolutions = len(ch_mult)
+        self.num_res_blocks = num_res_blocks
+        self.resolution = resolution
+        self.in_channels = in_channels
+        # downsampling
+        self.conv_in = nn.Conv2d(in_channels, self.ch, kernel_size=3, stride=1, padding=1)
+
+        curr_res = resolution
+        in_ch_mult = (1,) + tuple(ch_mult)
+        self.in_ch_mult = in_ch_mult
+        self.down = nn.ModuleList()
+        block_in = self.ch
+        for i_level in range(self.num_resolutions):
+            block = nn.ModuleList()
+            attn = nn.ModuleList()
+            block_in = ch * in_ch_mult[i_level]
+            block_out = ch * ch_mult[i_level]
+            for _ in range(self.num_res_blocks):
+                block.append(ResnetBlock(in_channels=block_in, out_channels=block_out))
+                block_in = block_out
+            down = nn.Module()
+            down.block = block
+            down.attn = attn
+            if i_level != self.num_resolutions - 1:
+                down.downsample = Downsample(block_in)
+                curr_res = curr_res // 2
+            self.down.append(down)
+
+        # middle
+        self.mid = nn.Module()
+        self.mid.block_1 = ResnetBlock(in_channels=block_in, out_channels=block_in)
+        self.mid.attn_1 = AttnBlock(block_in)
+        self.mid.block_2 = ResnetBlock(in_channels=block_in, out_channels=block_in)
+
+        # end
+        self.norm_out = nn.GroupNorm(num_groups=32, num_channels=block_in, eps=1e-6, affine=True)
+        self.conv_out = nn.Conv2d(block_in, 2 * z_channels, kernel_size=3, stride=1, padding=1)
+
+    def forward(self, x: Tensor) -> Tensor:
+        # downsampling
+        hs = [self.conv_in(x)]
+        for i_level in range(self.num_resolutions):
+            for i_block in range(self.num_res_blocks):
+                h = self.down[i_level].block[i_block](hs[-1])
+                if len(self.down[i_level].attn) > 0:
+                    h = self.down[i_level].attn[i_block](h)
+                hs.append(h)
+            if i_level != self.num_resolutions - 1:
+                hs.append(self.down[i_level].downsample(hs[-1]))
+
+        # middle
+        h = hs[-1]
+        h = self.mid.block_1(h)
+        h = self.mid.attn_1(h)
+        h = self.mid.block_2(h)
+        # end
+        h = self.norm_out(h)
+        h = torch.nn.functional.silu(h)
+        h = self.conv_out(h)
+        return h
+
+
+class Decoder(nn.Module):
+    def __init__(
+        self,
+        ch: int,
+        out_ch: int,
+        ch_mult: list[int],
+        num_res_blocks: int,
+        in_channels: int,
+        resolution: int,
+        z_channels: int,
+    ):
+        super().__init__()
+        self.ch = ch
+        self.num_resolutions = len(ch_mult)
+        self.num_res_blocks = num_res_blocks
+        self.resolution = resolution
+        self.in_channels = in_channels
+        self.ffactor = 2 ** (self.num_resolutions - 1)
+
+        # compute in_ch_mult, block_in and curr_res at lowest res
+        block_in = ch * ch_mult[self.num_resolutions - 1]
+        curr_res = resolution // 2 ** (self.num_resolutions - 1)
+        self.z_shape = (1, z_channels, curr_res, curr_res)
+
+        # z to block_in
+        self.conv_in = nn.Conv2d(z_channels, block_in, kernel_size=3, stride=1, padding=1)
+
+        # middle
+        self.mid = nn.Module()
+        self.mid.block_1 = ResnetBlock(in_channels=block_in, out_channels=block_in)
+        self.mid.attn_1 = AttnBlock(block_in)
+        self.mid.block_2 = ResnetBlock(in_channels=block_in, out_channels=block_in)
+
+        # upsampling
+        self.up = nn.ModuleList()
+        for i_level in reversed(range(self.num_resolutions)):
+            block = nn.ModuleList()
+            attn = nn.ModuleList()
+            block_out = ch * ch_mult[i_level]
+            for _ in range(self.num_res_blocks + 1):
+                block.append(ResnetBlock(in_channels=block_in, out_channels=block_out))
+                block_in = block_out
+            up = nn.Module()
+            up.block = block
+            up.attn = attn
+            if i_level != 0:
+                up.upsample = Upsample(block_in)
+                curr_res = curr_res * 2
+            self.up.insert(0, up)  # prepend to get consistent order
+
+        # end
+        self.norm_out = nn.GroupNorm(num_groups=32, num_channels=block_in, eps=1e-6, affine=True)
+        self.conv_out = nn.Conv2d(block_in, out_ch, kernel_size=3, stride=1, padding=1)
+
+    def forward(self, z: Tensor) -> Tensor:
+        # z to block_in
+        h = self.conv_in(z)
+
+        # middle
+        h = self.mid.block_1(h)
+        h = self.mid.attn_1(h)
+        h = self.mid.block_2(h)
+
+        # upsampling
+        for i_level in reversed(range(self.num_resolutions)):
+            for i_block in range(self.num_res_blocks + 1):
+                h = self.up[i_level].block[i_block](h)
+                if len(self.up[i_level].attn) > 0:
+                    h = self.up[i_level].attn[i_block](h)
+            if i_level != 0:
+                h = self.up[i_level].upsample(h)
+
+        # end
+        h = self.norm_out(h)
+        h = torch.nn.functional.silu(h)
+        h = self.conv_out(h)
+        return h
+
+
+class DiagonalGaussian(nn.Module):
+    def __init__(self, chunk_dim: int = 1):
+        super().__init__()
+        self.chunk_dim = chunk_dim
+
+    def forward(self, z: Tensor, sample: bool = True, generator: torch.Generator | None = None) -> Tensor:
+        mean, logvar = torch.chunk(z, 2, dim=self.chunk_dim)
+        if sample:
+            std = torch.exp(0.5 * logvar)
+            # Unfortunately, torch.randn_like(...) does not accept a generator argument at the time of writing, so we
+            # have to use torch.randn(...) instead.
+            return mean + std * torch.randn(size=mean.size(), generator=generator, dtype=mean.dtype, device=mean.device)
+        else:
+            return mean
+
+
+class AutoEncoder(nn.Module):
+    def __init__(self, params: AutoEncoderParams):
+        super().__init__()
+        self.encoder = Encoder(
+            resolution=params.resolution,
+            in_channels=params.in_channels,
+            ch=params.ch,
+            ch_mult=params.ch_mult,
+            num_res_blocks=params.num_res_blocks,
+            z_channels=params.z_channels,
+        )
+        self.decoder = Decoder(
+            resolution=params.resolution,
+            in_channels=params.in_channels,
+            ch=params.ch,
+            out_ch=params.out_ch,
+            ch_mult=params.ch_mult,
+            num_res_blocks=params.num_res_blocks,
+            z_channels=params.z_channels,
+        )
+        self.reg = DiagonalGaussian()
+
+        self.scale_factor = params.scale_factor
+        self.shift_factor = params.shift_factor
+
+    def encode(self, x: Tensor, sample: bool = True, generator: torch.Generator | None = None) -> Tensor:
+        """Run VAE encoding on input tensor x.
+
+        Args:
+            x (Tensor): Input image tensor. Shape: (batch_size, in_channels, height, width).
+            sample (bool, optional): If True, sample from the encoded distribution, else, return the distribution mean.
+                Defaults to True.
+            generator (torch.Generator | None, optional): Optional random number generator for reproducibility.
+                Defaults to None.
+
+        Returns:
+            Tensor: Encoded latent tensor. Shape: (batch_size, z_channels, latent_height, latent_width).
+        """
+
+        z = self.reg(self.encoder(x), sample=sample, generator=generator)
+        z = self.scale_factor * (z - self.shift_factor)
+        return z
+
+    def decode(self, z: Tensor) -> Tensor:
+        z = z / self.scale_factor + self.shift_factor
+        return self.decoder(z)
+
+    def forward(self, x: Tensor) -> Tensor:
+        return self.decode(self.encode(x))

invokeai/backend/flux/modules/conditioner.py

@@ -0,0 +1,33 @@
+# Initially pulled from https://github.com/black-forest-labs/flux
+
+from torch import Tensor, nn
+from transformers import PreTrainedModel, PreTrainedTokenizer
+
+
+class HFEncoder(nn.Module):
+    def __init__(self, encoder: PreTrainedModel, tokenizer: PreTrainedTokenizer, is_clip: bool, max_length: int):
+        super().__init__()
+        self.max_length = max_length
+        self.is_clip = is_clip
+        self.output_key = "pooler_output" if self.is_clip else "last_hidden_state"
+        self.tokenizer = tokenizer
+        self.hf_module = encoder
+        self.hf_module = self.hf_module.eval().requires_grad_(False)
+
+    def forward(self, text: list[str]) -> Tensor:
+        batch_encoding = self.tokenizer(
+            text,
+            truncation=True,
+            max_length=self.max_length,
+            return_length=False,
+            return_overflowing_tokens=False,
+            padding="max_length",
+            return_tensors="pt",
+        )
+
+        outputs = self.hf_module(
+            input_ids=batch_encoding["input_ids"].to(self.hf_module.device),
+            attention_mask=None,
+            output_hidden_states=False,
+        )
+        return outputs[self.output_key]

invokeai/backend/flux/modules/layers.py

@@ -0,0 +1,253 @@
+# Initially pulled from https://github.com/black-forest-labs/flux
+
+import math
+from dataclasses import dataclass
+
+import torch
+from einops import rearrange
+from torch import Tensor, nn
+
+from invokeai.backend.flux.math import attention, rope
+
+
+class EmbedND(nn.Module):
+    def __init__(self, dim: int, theta: int, axes_dim: list[int]):
+        super().__init__()
+        self.dim = dim
+        self.theta = theta
+        self.axes_dim = axes_dim
+
+    def forward(self, ids: Tensor) -> Tensor:
+        n_axes = ids.shape[-1]
+        emb = torch.cat(
+            [rope(ids[..., i], self.axes_dim[i], self.theta) for i in range(n_axes)],
+            dim=-3,
+        )
+
+        return emb.unsqueeze(1)
+
+
+def timestep_embedding(t: Tensor, dim, max_period=10000, time_factor: float = 1000.0):
+    """
+    Create sinusoidal timestep embeddings.
+    :param t: a 1-D Tensor of N indices, one per batch element.
+                      These may be fractional.
+    :param dim: the dimension of the output.
+    :param max_period: controls the minimum frequency of the embeddings.
+    :return: an (N, D) Tensor of positional embeddings.
+    """
+    t = time_factor * t
+    half = dim // 2
+    freqs = torch.exp(-math.log(max_period) * torch.arange(start=0, end=half, dtype=torch.float32) / half).to(t.device)
+
+    args = t[:, None].float() * freqs[None]
+    embedding = torch.cat([torch.cos(args), torch.sin(args)], dim=-1)
+    if dim % 2:
+        embedding = torch.cat([embedding, torch.zeros_like(embedding[:, :1])], dim=-1)
+    if torch.is_floating_point(t):
+        embedding = embedding.to(t)
+    return embedding
+
+
+class MLPEmbedder(nn.Module):
+    def __init__(self, in_dim: int, hidden_dim: int):
+        super().__init__()
+        self.in_layer = nn.Linear(in_dim, hidden_dim, bias=True)
+        self.silu = nn.SiLU()
+        self.out_layer = nn.Linear(hidden_dim, hidden_dim, bias=True)
+
+    def forward(self, x: Tensor) -> Tensor:
+        return self.out_layer(self.silu(self.in_layer(x)))
+
+
+class RMSNorm(torch.nn.Module):
+    def __init__(self, dim: int):
+        super().__init__()
+        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
+
+
+class QKNorm(torch.nn.Module):
+    def __init__(self, dim: int):
+        super().__init__()
+        self.query_norm = RMSNorm(dim)
+        self.key_norm = RMSNorm(dim)
+
+    def forward(self, q: Tensor, k: Tensor, v: Tensor) -> tuple[Tensor, Tensor]:
+        q = self.query_norm(q)
+        k = self.key_norm(k)
+        return q.to(v), k.to(v)
+
+
+class SelfAttention(nn.Module):
+    def __init__(self, dim: int, num_heads: int = 8, qkv_bias: bool = False):
+        super().__init__()
+        self.num_heads = num_heads
+        head_dim = dim // num_heads
+
+        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
+        self.norm = QKNorm(head_dim)
+        self.proj = nn.Linear(dim, dim)
+
+    def forward(self, x: Tensor, pe: Tensor) -> Tensor:
+        qkv = self.qkv(x)
+        q, k, v = rearrange(qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
+        q, k = self.norm(q, k, v)
+        x = attention(q, k, v, pe=pe)
+        x = self.proj(x)
+        return x
+
+
+@dataclass
+class ModulationOut:
+    shift: Tensor
+    scale: Tensor
+    gate: Tensor
+
+
+class Modulation(nn.Module):
+    def __init__(self, dim: int, double: bool):
+        super().__init__()
+        self.is_double = double
+        self.multiplier = 6 if double else 3
+        self.lin = nn.Linear(dim, self.multiplier * dim, bias=True)
+
+    def forward(self, vec: Tensor) -> tuple[ModulationOut, ModulationOut | None]:
+        out = self.lin(nn.functional.silu(vec))[:, None, :].chunk(self.multiplier, dim=-1)
+
+        return (
+            ModulationOut(*out[:3]),
+            ModulationOut(*out[3:]) if self.is_double else None,
+        )
+
+
+class DoubleStreamBlock(nn.Module):
+    def __init__(self, hidden_size: int, num_heads: int, mlp_ratio: float, qkv_bias: bool = False):
+        super().__init__()
+
+        mlp_hidden_dim = int(hidden_size * mlp_ratio)
+        self.num_heads = num_heads
+        self.hidden_size = hidden_size
+        self.img_mod = Modulation(hidden_size, double=True)
+        self.img_norm1 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.img_attn = SelfAttention(dim=hidden_size, num_heads=num_heads, qkv_bias=qkv_bias)
+
+        self.img_norm2 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.img_mlp = nn.Sequential(
+            nn.Linear(hidden_size, mlp_hidden_dim, bias=True),
+            nn.GELU(approximate="tanh"),
+            nn.Linear(mlp_hidden_dim, hidden_size, bias=True),
+        )
+
+        self.txt_mod = Modulation(hidden_size, double=True)
+        self.txt_norm1 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.txt_attn = SelfAttention(dim=hidden_size, num_heads=num_heads, qkv_bias=qkv_bias)
+
+        self.txt_norm2 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.txt_mlp = nn.Sequential(
+            nn.Linear(hidden_size, mlp_hidden_dim, bias=True),
+            nn.GELU(approximate="tanh"),
+            nn.Linear(mlp_hidden_dim, hidden_size, bias=True),
+        )
+
+    def forward(self, img: Tensor, txt: Tensor, vec: Tensor, pe: Tensor) -> tuple[Tensor, Tensor]:
+        img_mod1, img_mod2 = self.img_mod(vec)
+        txt_mod1, txt_mod2 = self.txt_mod(vec)
+
+        # prepare image for attention
+        img_modulated = self.img_norm1(img)
+        img_modulated = (1 + img_mod1.scale) * img_modulated + img_mod1.shift
+        img_qkv = self.img_attn.qkv(img_modulated)
+        img_q, img_k, img_v = rearrange(img_qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
+        img_q, img_k = self.img_attn.norm(img_q, img_k, img_v)
+
+        # prepare txt for attention
+        txt_modulated = self.txt_norm1(txt)
+        txt_modulated = (1 + txt_mod1.scale) * txt_modulated + txt_mod1.shift
+        txt_qkv = self.txt_attn.qkv(txt_modulated)
+        txt_q, txt_k, txt_v = rearrange(txt_qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
+        txt_q, txt_k = self.txt_attn.norm(txt_q, txt_k, txt_v)
+
+        # run actual attention
+        q = torch.cat((txt_q, img_q), dim=2)
+        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)
+        txt_attn, img_attn = attn[:, : txt.shape[1]], attn[:, txt.shape[1] :]
+
+        # calculate the img bloks
+        img = img + img_mod1.gate * self.img_attn.proj(img_attn)
+        img = img + img_mod2.gate * self.img_mlp((1 + img_mod2.scale) * self.img_norm2(img) + img_mod2.shift)
+
+        # calculate the txt bloks
+        txt = txt + txt_mod1.gate * self.txt_attn.proj(txt_attn)
+        txt = txt + txt_mod2.gate * self.txt_mlp((1 + txt_mod2.scale) * self.txt_norm2(txt) + txt_mod2.shift)
+        return img, txt
+
+
+class SingleStreamBlock(nn.Module):
+    """
+    A DiT block with parallel linear layers as described in
+    https://arxiv.org/abs/2302.05442 and adapted modulation interface.
+    """
+
+    def __init__(
+        self,
+        hidden_size: int,
+        num_heads: int,
+        mlp_ratio: float = 4.0,
+        qk_scale: float | None = None,
+    ):
+        super().__init__()
+        self.hidden_dim = hidden_size
+        self.num_heads = num_heads
+        head_dim = hidden_size // num_heads
+        self.scale = qk_scale or head_dim**-0.5
+
+        self.mlp_hidden_dim = int(hidden_size * mlp_ratio)
+        # qkv and mlp_in
+        self.linear1 = nn.Linear(hidden_size, hidden_size * 3 + self.mlp_hidden_dim)
+        # proj and mlp_out
+        self.linear2 = nn.Linear(hidden_size + self.mlp_hidden_dim, hidden_size)
+
+        self.norm = QKNorm(head_dim)
+
+        self.hidden_size = hidden_size
+        self.pre_norm = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+
+        self.mlp_act = nn.GELU(approximate="tanh")
+        self.modulation = Modulation(hidden_size, double=False)
+
+    def forward(self, x: Tensor, vec: Tensor, pe: Tensor) -> Tensor:
+        mod, _ = self.modulation(vec)
+        x_mod = (1 + mod.scale) * self.pre_norm(x) + mod.shift
+        qkv, mlp = torch.split(self.linear1(x_mod), [3 * self.hidden_size, self.mlp_hidden_dim], dim=-1)
+
+        q, k, v = rearrange(qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
+        q, k = self.norm(q, k, v)
+
+        # compute attention
+        attn = attention(q, k, v, pe=pe)
+        # compute activation in mlp stream, cat again and run second linear layer
+        output = self.linear2(torch.cat((attn, self.mlp_act(mlp)), 2))
+        return x + mod.gate * output
+
+
+class LastLayer(nn.Module):
+    def __init__(self, hidden_size: int, patch_size: int, out_channels: int):
+        super().__init__()
+        self.norm_final = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.linear = nn.Linear(hidden_size, patch_size * patch_size * out_channels, bias=True)
+        self.adaLN_modulation = nn.Sequential(nn.SiLU(), nn.Linear(hidden_size, 2 * hidden_size, bias=True))
+
+    def forward(self, x: Tensor, vec: Tensor) -> Tensor:
+        shift, scale = self.adaLN_modulation(vec).chunk(2, dim=1)
+        x = (1 + scale[:, None, :]) * self.norm_final(x) + shift[:, None, :]
+        x = self.linear(x)
+        return x

invokeai/backend/flux/sampling_utils.py

@@ -0,0 +1,135 @@
+# Initially pulled from https://github.com/black-forest-labs/flux
+
+import math
+from typing import Callable
+
+import torch
+from einops import rearrange, repeat
+
+
+def get_noise(
+    num_samples: int,
+    height: int,
+    width: int,
+    device: torch.device,
+    dtype: torch.dtype,
+    seed: int,
+):
+    # 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,
+        16,
+        # allow for packing
+        2 * math.ceil(height / 16),
+        2 * math.ceil(width / 16),
+        device=rand_device,
+        dtype=rand_dtype,
+        generator=torch.Generator(device=rand_device).manual_seed(seed),
+    ).to(device=device, dtype=dtype)
+
+
+def time_shift(mu: float, sigma: float, t: torch.Tensor) -> torch.Tensor:
+    return math.exp(mu) / (math.exp(mu) + (1 / t - 1) ** sigma)
+
+
+def get_lin_function(x1: float = 256, y1: float = 0.5, x2: float = 4096, y2: float = 1.15) -> Callable[[float], float]:
+    m = (y2 - y1) / (x2 - x1)
+    b = y1 - m * x1
+    return lambda x: m * x + b
+
+
+def get_schedule(
+    num_steps: int,
+    image_seq_len: int,
+    base_shift: float = 0.5,
+    max_shift: float = 1.15,
+    shift: bool = True,
+) -> list[float]:
+    # extra step for zero
+    timesteps = torch.linspace(1, 0, num_steps + 1)
+
+    # shifting the schedule to favor high timesteps for higher signal images
+    if shift:
+        # estimate mu based on linear estimation between two points
+        mu = get_lin_function(y1=base_shift, y2=max_shift)(image_seq_len)
+        timesteps = time_shift(mu, 1.0, timesteps)
+
+    return timesteps.tolist()
+
+
+def _find_last_index_ge_val(timesteps: list[float], val: float, eps: float = 1e-6) -> int:
+    """Find the last index in timesteps that is >= val.
+
+    We use epsilon-close equality to avoid potential floating point errors.
+    """
+    idx = len(list(filter(lambda t: t >= (val - eps), timesteps))) - 1
+    assert idx >= 0
+    return idx
+
+
+def clip_timestep_schedule(timesteps: list[float], denoising_start: float, denoising_end: float) -> list[float]:
+    """Clip the timestep schedule to the denoising range.
+
+    Args:
+        timesteps (list[float]): The original timestep schedule: [1.0, ..., 0.0].
+        denoising_start (float): A value in [0, 1] specifying the start of the denoising process. E.g. a value of 0.2
+            would mean that the denoising process start at the last timestep in the schedule >= 0.8.
+        denoising_end (float): A value in [0, 1] specifying the end of the denoising process. E.g. a value of 0.8 would
+            mean that the denoising process end at the last timestep in the schedule >= 0.2.
+
+    Returns:
+        list[float]: The clipped timestep schedule.
+    """
+    assert 0.0 <= denoising_start <= 1.0
+    assert 0.0 <= denoising_end <= 1.0
+    assert denoising_start <= denoising_end
+
+    t_start_val = 1.0 - denoising_start
+    t_end_val = 1.0 - denoising_end
+
+    t_start_idx = _find_last_index_ge_val(timesteps, t_start_val)
+    t_end_idx = _find_last_index_ge_val(timesteps, t_end_val)
+
+    clipped_timesteps = timesteps[t_start_idx : t_end_idx + 1]
+
+    return clipped_timesteps
+
+
+def unpack(x: torch.Tensor, height: int, width: int) -> torch.Tensor:
+    """Unpack flat array of patch embeddings to latent image."""
+    return rearrange(
+        x,
+        "b (h w) (c ph pw) -> b c (h ph) (w pw)",
+        h=math.ceil(height / 16),
+        w=math.ceil(width / 16),
+        ph=2,
+        pw=2,
+    )
+
+
+def pack(x: torch.Tensor) -> torch.Tensor:
+    """Pack latent image to flattented array of patch embeddings."""
+    # Pixel unshuffle with a scale of 2, and flatten the height/width dimensions to get an array of patches.
+    return rearrange(x, "b c (h ph) (w pw) -> b (h w) (c ph pw)", ph=2, pw=2)
+
+
+def generate_img_ids(h: int, w: int, batch_size: int, device: torch.device, dtype: torch.dtype) -> torch.Tensor:
+    """Generate tensor of image position ids.
+
+    Args:
+        h (int): Height of image in latent space.
+        w (int): Width of image in latent space.
+        batch_size (int): Batch size.
+        device (torch.device): Device.
+        dtype (torch.dtype): dtype.
+
+    Returns:
+        torch.Tensor: Image position ids.
+    """
+    img_ids = torch.zeros(h // 2, w // 2, 3, device=device, dtype=dtype)
+    img_ids[..., 1] = img_ids[..., 1] + torch.arange(h // 2, device=device, dtype=dtype)[:, None]
+    img_ids[..., 2] = img_ids[..., 2] + torch.arange(w // 2, device=device, dtype=dtype)[None, :]
+    img_ids = repeat(img_ids, "h w c -> b (h w) c", b=batch_size)
+    return img_ids

invokeai/backend/flux/util.py

@@ -0,0 +1,71 @@
+# Initially pulled from https://github.com/black-forest-labs/flux
+
+from dataclasses import dataclass
+from typing import Dict, Literal
+
+from invokeai.backend.flux.model import FluxParams
+from invokeai.backend.flux.modules.autoencoder import AutoEncoderParams
+
+
+@dataclass
+class ModelSpec:
+    params: FluxParams
+    ae_params: AutoEncoderParams
+    ckpt_path: str | None
+    ae_path: str | None
+    repo_id: str | None
+    repo_flow: str | None
+    repo_ae: str | None
+
+
+max_seq_lengths: Dict[str, Literal[256, 512]] = {
+    "flux-dev": 512,
+    "flux-schnell": 256,
+}
+
+
+ae_params = {
+    "flux": AutoEncoderParams(
+        resolution=256,
+        in_channels=3,
+        ch=128,
+        out_ch=3,
+        ch_mult=[1, 2, 4, 4],
+        num_res_blocks=2,
+        z_channels=16,
+        scale_factor=0.3611,
+        shift_factor=0.1159,
+    )
+}
+
+
+params = {
+    "flux-dev": FluxParams(
+        in_channels=64,
+        vec_in_dim=768,
+        context_in_dim=4096,
+        hidden_size=3072,
+        mlp_ratio=4.0,
+        num_heads=24,
+        depth=19,
+        depth_single_blocks=38,
+        axes_dim=[16, 56, 56],
+        theta=10_000,
+        qkv_bias=True,
+        guidance_embed=True,
+    ),
+    "flux-schnell": FluxParams(
+        in_channels=64,
+        vec_in_dim=768,
+        context_in_dim=4096,
+        hidden_size=3072,
+        mlp_ratio=4.0,
+        num_heads=24,
+        depth=19,
+        depth_single_blocks=38,
+        axes_dim=[16, 56, 56],
+        theta=10_000,
+        qkv_bias=True,
+        guidance_embed=False,
+    ),
+}

invokeai/backend/image_util/depth_anything/__init__.py

@@ -1,90 +0,0 @@
-from pathlib import Path
-from typing import Literal
-
-import cv2
-import numpy as np
-import torch
-import torch.nn.functional as F
-from einops import repeat
-from PIL import Image
-from torchvision.transforms import Compose
-
-from invokeai.app.services.config.config_default import get_config
-from invokeai.backend.image_util.depth_anything.model.dpt import DPT_DINOv2
-from invokeai.backend.image_util.depth_anything.utilities.util import NormalizeImage, PrepareForNet, Resize
-from invokeai.backend.util.logging import InvokeAILogger
-
-config = get_config()
-logger = InvokeAILogger.get_logger(config=config)
-
-DEPTH_ANYTHING_MODELS = {
-    "large": "https://huggingface.co/spaces/LiheYoung/Depth-Anything/resolve/main/checkpoints/depth_anything_vitl14.pth?download=true",
-    "base": "https://huggingface.co/spaces/LiheYoung/Depth-Anything/resolve/main/checkpoints/depth_anything_vitb14.pth?download=true",
-    "small": "https://huggingface.co/spaces/LiheYoung/Depth-Anything/resolve/main/checkpoints/depth_anything_vits14.pth?download=true",
-}
-
-
-transform = Compose(
-    [
-        Resize(
-            width=518,
-            height=518,
-            resize_target=False,
-            keep_aspect_ratio=True,
-            ensure_multiple_of=14,
-            resize_method="lower_bound",
-            image_interpolation_method=cv2.INTER_CUBIC,
-        ),
-        NormalizeImage(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
-        PrepareForNet(),
-    ]
-)
-
-
-class DepthAnythingDetector:
-    def __init__(self, model: DPT_DINOv2, device: torch.device) -> None:
-        self.model = model
-        self.device = device
-
-    @staticmethod
-    def load_model(
-        model_path: Path, device: torch.device, model_size: Literal["large", "base", "small"] = "small"
-    ) -> DPT_DINOv2:
-        match model_size:
-            case "small":
-                model = DPT_DINOv2(encoder="vits", features=64, out_channels=[48, 96, 192, 384])
-            case "base":
-                model = DPT_DINOv2(encoder="vitb", features=128, out_channels=[96, 192, 384, 768])
-            case "large":
-                model = DPT_DINOv2(encoder="vitl", features=256, out_channels=[256, 512, 1024, 1024])
-
-        model.load_state_dict(torch.load(model_path.as_posix(), map_location="cpu"))
-        model.eval()
-
-        model.to(device)
-        return model
-
-    def __call__(self, image: Image.Image, resolution: int = 512) -> Image.Image:
-        if not self.model:
-            logger.warn("DepthAnything model was not loaded. Returning original image")
-            return image
-
-        np_image = np.array(image, dtype=np.uint8)
-        np_image = np_image[:, :, ::-1] / 255.0
-
-        image_height, image_width = np_image.shape[:2]
-        np_image = transform({"image": np_image})["image"]
-        tensor_image = torch.from_numpy(np_image).unsqueeze(0).to(self.device)
-
-        with torch.no_grad():
-            depth = self.model(tensor_image)
-            depth = F.interpolate(depth[None], (image_height, image_width), mode="bilinear", align_corners=False)[0, 0]
-            depth = (depth - depth.min()) / (depth.max() - depth.min()) * 255.0
-
-        depth_map = repeat(depth, "h w -> h w 3").cpu().numpy().astype(np.uint8)
-        depth_map = Image.fromarray(depth_map)
-
-        new_height = int(image_height * (resolution / image_width))
-        depth_map = depth_map.resize((resolution, new_height))
-
-        return depth_map

invokeai/backend/image_util/depth_anything/depth_anything_pipeline.py

@@ -0,0 +1,31 @@
+from typing import Optional
+
+import torch
+from PIL import Image
+from transformers.pipelines import DepthEstimationPipeline
+
+from invokeai.backend.raw_model import RawModel
+
+
+class DepthAnythingPipeline(RawModel):
+    """Custom wrapper for the Depth Estimation pipeline from transformers adding compatibility
+    for Invoke's Model Management System"""
+
+    def __init__(self, pipeline: DepthEstimationPipeline) -> None:
+        self._pipeline = pipeline
+
+    def generate_depth(self, image: Image.Image) -> Image.Image:
+        depth_map = self._pipeline(image)["depth"]
+        assert isinstance(depth_map, Image.Image)
+        return depth_map
+
+    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None):
+        if device is not None and device.type not in {"cpu", "cuda"}:
+            device = None
+        self._pipeline.model.to(device=device, dtype=dtype)
+        self._pipeline.device = self._pipeline.model.device
+
+    def calc_size(self) -> int:
+        from invokeai.backend.model_manager.load.model_util import calc_module_size
+
+        return calc_module_size(self._pipeline.model)

invokeai/backend/image_util/depth_anything/model/blocks.py

@@ -1,145 +0,0 @@
-import torch.nn as nn
-
-
-def _make_scratch(in_shape, out_shape, groups=1, expand=False):
-    scratch = nn.Module()
-
-    out_shape1 = out_shape
-    out_shape2 = out_shape
-    out_shape3 = out_shape
-    if len(in_shape) >= 4:
-        out_shape4 = out_shape
-
-    if expand:
-        out_shape1 = out_shape
-        out_shape2 = out_shape * 2
-        out_shape3 = out_shape * 4
-        if len(in_shape) >= 4:
-            out_shape4 = out_shape * 8
-
-    scratch.layer1_rn = nn.Conv2d(
-        in_shape[0], out_shape1, kernel_size=3, stride=1, padding=1, bias=False, groups=groups
-    )
-    scratch.layer2_rn = nn.Conv2d(
-        in_shape[1], out_shape2, kernel_size=3, stride=1, padding=1, bias=False, groups=groups
-    )
-    scratch.layer3_rn = nn.Conv2d(
-        in_shape[2], out_shape3, kernel_size=3, stride=1, padding=1, bias=False, groups=groups
-    )
-    if len(in_shape) >= 4:
-        scratch.layer4_rn = nn.Conv2d(
-            in_shape[3], out_shape4, kernel_size=3, stride=1, padding=1, bias=False, groups=groups
-        )
-
-    return scratch
-
-
-class ResidualConvUnit(nn.Module):
-    """Residual convolution module."""
-
-    def __init__(self, features, activation, bn):
-        """Init.
-
-        Args:
-            features (int): number of features
-        """
-        super().__init__()
-
-        self.bn = bn
-
-        self.groups = 1
-
-        self.conv1 = nn.Conv2d(features, features, kernel_size=3, stride=1, padding=1, bias=True, groups=self.groups)
-
-        self.conv2 = nn.Conv2d(features, features, kernel_size=3, stride=1, padding=1, bias=True, groups=self.groups)
-
-        if self.bn:
-            self.bn1 = nn.BatchNorm2d(features)
-            self.bn2 = nn.BatchNorm2d(features)
-
-        self.activation = activation
-
-        self.skip_add = nn.quantized.FloatFunctional()
-
-    def forward(self, x):
-        """Forward pass.
-
-        Args:
-            x (tensor): input
-
-        Returns:
-            tensor: output
-        """
-
-        out = self.activation(x)
-        out = self.conv1(out)
-        if self.bn:
-            out = self.bn1(out)
-
-        out = self.activation(out)
-        out = self.conv2(out)
-        if self.bn:
-            out = self.bn2(out)
-
-        if self.groups > 1:
-            out = self.conv_merge(out)
-
-        return self.skip_add.add(out, x)
-
-
-class FeatureFusionBlock(nn.Module):
-    """Feature fusion block."""
-
-    def __init__(self, features, activation, deconv=False, bn=False, expand=False, align_corners=True, size=None):
-        """Init.
-
-        Args:
-            features (int): number of features
-        """
-        super(FeatureFusionBlock, self).__init__()
-
-        self.deconv = deconv
-        self.align_corners = align_corners
-
-        self.groups = 1
-
-        self.expand = expand
-        out_features = features
-        if self.expand:
-            out_features = features // 2
-
-        self.out_conv = nn.Conv2d(features, out_features, kernel_size=1, stride=1, padding=0, bias=True, groups=1)
-
-        self.resConfUnit1 = ResidualConvUnit(features, activation, bn)
-        self.resConfUnit2 = ResidualConvUnit(features, activation, bn)
-
-        self.skip_add = nn.quantized.FloatFunctional()
-
-        self.size = size
-
-    def forward(self, *xs, size=None):
-        """Forward pass.
-
-        Returns:
-            tensor: output
-        """
-        output = xs[0]
-
-        if len(xs) == 2:
-            res = self.resConfUnit1(xs[1])
-            output = self.skip_add.add(output, res)
-
-        output = self.resConfUnit2(output)
-
-        if (size is None) and (self.size is None):
-            modifier = {"scale_factor": 2}
-        elif size is None:
-            modifier = {"size": self.size}
-        else:
-            modifier = {"size": size}
-
-        output = nn.functional.interpolate(output, **modifier, mode="bilinear", align_corners=self.align_corners)
-
-        output = self.out_conv(output)
-
-        return output

invokeai/backend/image_util/depth_anything/model/dpt.py

@@ -1,183 +0,0 @@
-from pathlib import Path
-
-import torch
-import torch.nn as nn
-import torch.nn.functional as F
-
-from invokeai.backend.image_util.depth_anything.model.blocks import FeatureFusionBlock, _make_scratch
-
-torchhub_path = Path(__file__).parent.parent / "torchhub"
-
-
-def _make_fusion_block(features, use_bn, size=None):
-    return FeatureFusionBlock(
-        features,
-        nn.ReLU(False),
-        deconv=False,
-        bn=use_bn,
-        expand=False,
-        align_corners=True,
-        size=size,
-    )
-
-
-class DPTHead(nn.Module):
-    def __init__(self, nclass, in_channels, features, out_channels, use_bn=False, use_clstoken=False):
-        super(DPTHead, self).__init__()
-
-        self.nclass = nclass
-        self.use_clstoken = use_clstoken
-
-        self.projects = nn.ModuleList(
-            [
-                nn.Conv2d(
-                    in_channels=in_channels,
-                    out_channels=out_channel,
-                    kernel_size=1,
-                    stride=1,
-                    padding=0,
-                )
-                for out_channel in out_channels
-            ]
-        )
-
-        self.resize_layers = nn.ModuleList(
-            [
-                nn.ConvTranspose2d(
-                    in_channels=out_channels[0], out_channels=out_channels[0], kernel_size=4, stride=4, padding=0
-                ),
-                nn.ConvTranspose2d(
-                    in_channels=out_channels[1], out_channels=out_channels[1], kernel_size=2, stride=2, padding=0
-                ),
-                nn.Identity(),
-                nn.Conv2d(
-                    in_channels=out_channels[3], out_channels=out_channels[3], kernel_size=3, stride=2, padding=1
-                ),
-            ]
-        )
-
-        if use_clstoken:
-            self.readout_projects = nn.ModuleList()
-            for _ in range(len(self.projects)):
-                self.readout_projects.append(nn.Sequential(nn.Linear(2 * in_channels, in_channels), nn.GELU()))
-
-        self.scratch = _make_scratch(
-            out_channels,
-            features,
-            groups=1,
-            expand=False,
-        )
-
-        self.scratch.stem_transpose = None
-
-        self.scratch.refinenet1 = _make_fusion_block(features, use_bn)
-        self.scratch.refinenet2 = _make_fusion_block(features, use_bn)
-        self.scratch.refinenet3 = _make_fusion_block(features, use_bn)
-        self.scratch.refinenet4 = _make_fusion_block(features, use_bn)
-
-        head_features_1 = features
-        head_features_2 = 32
-
-        if nclass > 1:
-            self.scratch.output_conv = nn.Sequential(
-                nn.Conv2d(head_features_1, head_features_1, kernel_size=3, stride=1, padding=1),
-                nn.ReLU(True),
-                nn.Conv2d(head_features_1, nclass, kernel_size=1, stride=1, padding=0),
-            )
-        else:
-            self.scratch.output_conv1 = nn.Conv2d(
-                head_features_1, head_features_1 // 2, kernel_size=3, stride=1, padding=1
-            )
-
-            self.scratch.output_conv2 = nn.Sequential(
-                nn.Conv2d(head_features_1 // 2, head_features_2, kernel_size=3, stride=1, padding=1),
-                nn.ReLU(True),
-                nn.Conv2d(head_features_2, 1, kernel_size=1, stride=1, padding=0),
-                nn.ReLU(True),
-                nn.Identity(),
-            )
-
-    def forward(self, out_features, patch_h, patch_w):
-        out = []
-        for i, x in enumerate(out_features):
-            if self.use_clstoken:
-                x, cls_token = x[0], x[1]
-                readout = cls_token.unsqueeze(1).expand_as(x)
-                x = self.readout_projects[i](torch.cat((x, readout), -1))
-            else:
-                x = x[0]
-
-            x = x.permute(0, 2, 1).reshape((x.shape[0], x.shape[-1], patch_h, patch_w))
-
-            x = self.projects[i](x)
-            x = self.resize_layers[i](x)
-
-            out.append(x)
-
-        layer_1, layer_2, layer_3, layer_4 = out
-
-        layer_1_rn = self.scratch.layer1_rn(layer_1)
-        layer_2_rn = self.scratch.layer2_rn(layer_2)
-        layer_3_rn = self.scratch.layer3_rn(layer_3)
-        layer_4_rn = self.scratch.layer4_rn(layer_4)
-
-        path_4 = self.scratch.refinenet4(layer_4_rn, size=layer_3_rn.shape[2:])
-        path_3 = self.scratch.refinenet3(path_4, layer_3_rn, size=layer_2_rn.shape[2:])
-        path_2 = self.scratch.refinenet2(path_3, layer_2_rn, size=layer_1_rn.shape[2:])
-        path_1 = self.scratch.refinenet1(path_2, layer_1_rn)
-
-        out = self.scratch.output_conv1(path_1)
-        out = F.interpolate(out, (int(patch_h * 14), int(patch_w * 14)), mode="bilinear", align_corners=True)
-        out = self.scratch.output_conv2(out)
-
-        return out
-
-
-class DPT_DINOv2(nn.Module):
-    def __init__(
-        self,
-        features,
-        out_channels,
-        encoder="vitl",
-        use_bn=False,
-        use_clstoken=False,
-    ):
-        super(DPT_DINOv2, self).__init__()
-
-        assert encoder in ["vits", "vitb", "vitl"]
-
-        # # in case the Internet connection is not stable, please load the DINOv2 locally
-        # if use_local:
-        #     self.pretrained = torch.hub.load(
-        #         torchhub_path / "facebookresearch_dinov2_main",
-        #         "dinov2_{:}14".format(encoder),
-        #         source="local",
-        #         pretrained=False,
-        #     )
-        # else:
-        #     self.pretrained = torch.hub.load(
-        #         "facebookresearch/dinov2",
-        #         "dinov2_{:}14".format(encoder),
-        #     )
-
-        self.pretrained = torch.hub.load(
-            "facebookresearch/dinov2",
-            "dinov2_{:}14".format(encoder),
-        )
-
-        dim = self.pretrained.blocks[0].attn.qkv.in_features
-
-        self.depth_head = DPTHead(1, dim, features, out_channels=out_channels, use_bn=use_bn, use_clstoken=use_clstoken)
-
-    def forward(self, x):
-        h, w = x.shape[-2:]
-
-        features = self.pretrained.get_intermediate_layers(x, 4, return_class_token=True)
-
-        patch_h, patch_w = h // 14, w // 14
-
-        depth = self.depth_head(features, patch_h, patch_w)
-        depth = F.interpolate(depth, size=(h, w), mode="bilinear", align_corners=True)
-        depth = F.relu(depth)
-
-        return depth.squeeze(1)

invokeai/backend/image_util/depth_anything/utilities/util.py

@@ -1,227 +0,0 @@
-import math
-
-import cv2
-import numpy as np
-import torch
-import torch.nn.functional as F
-
-
-def apply_min_size(sample, size, image_interpolation_method=cv2.INTER_AREA):
-    """Rezise the sample to ensure the given size. Keeps aspect ratio.
-
-    Args:
-        sample (dict): sample
-        size (tuple): image size
-
-    Returns:
-        tuple: new size
-    """
-    shape = list(sample["disparity"].shape)
-
-    if shape[0] >= size[0] and shape[1] >= size[1]:
-        return sample
-
-    scale = [0, 0]
-    scale[0] = size[0] / shape[0]
-    scale[1] = size[1] / shape[1]
-
-    scale = max(scale)
-
-    shape[0] = math.ceil(scale * shape[0])
-    shape[1] = math.ceil(scale * shape[1])
-
-    # resize
-    sample["image"] = cv2.resize(sample["image"], tuple(shape[::-1]), interpolation=image_interpolation_method)
-
-    sample["disparity"] = cv2.resize(sample["disparity"], tuple(shape[::-1]), interpolation=cv2.INTER_NEAREST)
-    sample["mask"] = cv2.resize(
-        sample["mask"].astype(np.float32),
-        tuple(shape[::-1]),
-        interpolation=cv2.INTER_NEAREST,
-    )
-    sample["mask"] = sample["mask"].astype(bool)
-
-    return tuple(shape)
-
-
-class Resize(object):
-    """Resize sample to given size (width, height)."""
-
-    def __init__(
-        self,
-        width,
-        height,
-        resize_target=True,
-        keep_aspect_ratio=False,
-        ensure_multiple_of=1,
-        resize_method="lower_bound",
-        image_interpolation_method=cv2.INTER_AREA,
-    ):
-        """Init.
-
-        Args:
-            width (int): desired output width
-            height (int): desired output height
-            resize_target (bool, optional):
-                True: Resize the full sample (image, mask, target).
-                False: Resize image only.
-                Defaults to True.
-            keep_aspect_ratio (bool, optional):
-                True: Keep the aspect ratio of the input sample.
-                Output sample might not have the given width and height, and
-                resize behaviour depends on the parameter 'resize_method'.
-                Defaults to False.
-            ensure_multiple_of (int, optional):
-                Output width and height is constrained to be multiple of this parameter.
-                Defaults to 1.
-            resize_method (str, optional):
-                "lower_bound": Output will be at least as large as the given size.
-                "upper_bound": Output will be at max as large as the given size. (Output size might be smaller
-                    than given size.)
-                "minimal": Scale as least as possible.  (Output size might be smaller than given size.)
-                Defaults to "lower_bound".
-        """
-        self.__width = width
-        self.__height = height
-
-        self.__resize_target = resize_target
-        self.__keep_aspect_ratio = keep_aspect_ratio
-        self.__multiple_of = ensure_multiple_of
-        self.__resize_method = resize_method
-        self.__image_interpolation_method = image_interpolation_method
-
-    def constrain_to_multiple_of(self, x, min_val=0, max_val=None):
-        y = (np.round(x / self.__multiple_of) * self.__multiple_of).astype(int)
-
-        if max_val is not None and y > max_val:
-            y = (np.floor(x / self.__multiple_of) * self.__multiple_of).astype(int)
-
-        if y < min_val:
-            y = (np.ceil(x / self.__multiple_of) * self.__multiple_of).astype(int)
-
-        return y
-
-    def get_size(self, width, height):
-        # determine new height and width
-        scale_height = self.__height / height
-        scale_width = self.__width / width
-
-        if self.__keep_aspect_ratio:
-            if self.__resize_method == "lower_bound":
-                # scale such that output size is lower bound
-                if scale_width > scale_height:
-                    # fit width
-                    scale_height = scale_width
-                else:
-                    # fit height
-                    scale_width = scale_height
-            elif self.__resize_method == "upper_bound":
-                # scale such that output size is upper bound
-                if scale_width < scale_height:
-                    # fit width
-                    scale_height = scale_width
-                else:
-                    # fit height
-                    scale_width = scale_height
-            elif self.__resize_method == "minimal":
-                # scale as least as possbile
-                if abs(1 - scale_width) < abs(1 - scale_height):
-                    # fit width
-                    scale_height = scale_width
-                else:
-                    # fit height
-                    scale_width = scale_height
-            else:
-                raise ValueError(f"resize_method {self.__resize_method} not implemented")
-
-        if self.__resize_method == "lower_bound":
-            new_height = self.constrain_to_multiple_of(scale_height * height, min_val=self.__height)
-            new_width = self.constrain_to_multiple_of(scale_width * width, min_val=self.__width)
-        elif self.__resize_method == "upper_bound":
-            new_height = self.constrain_to_multiple_of(scale_height * height, max_val=self.__height)
-            new_width = self.constrain_to_multiple_of(scale_width * width, max_val=self.__width)
-        elif self.__resize_method == "minimal":
-            new_height = self.constrain_to_multiple_of(scale_height * height)
-            new_width = self.constrain_to_multiple_of(scale_width * width)
-        else:
-            raise ValueError(f"resize_method {self.__resize_method} not implemented")
-
-        return (new_width, new_height)
-
-    def __call__(self, sample):
-        width, height = self.get_size(sample["image"].shape[1], sample["image"].shape[0])
-
-        # resize sample
-        sample["image"] = cv2.resize(
-            sample["image"],
-            (width, height),
-            interpolation=self.__image_interpolation_method,
-        )
-
-        if self.__resize_target:
-            if "disparity" in sample:
-                sample["disparity"] = cv2.resize(
-                    sample["disparity"],
-                    (width, height),
-                    interpolation=cv2.INTER_NEAREST,
-                )
-
-            if "depth" in sample:
-                sample["depth"] = cv2.resize(sample["depth"], (width, height), interpolation=cv2.INTER_NEAREST)
-
-            if "semseg_mask" in sample:
-                # sample["semseg_mask"] = cv2.resize(
-                #     sample["semseg_mask"], (width, height), interpolation=cv2.INTER_NEAREST
-                # )
-                sample["semseg_mask"] = F.interpolate(
-                    torch.from_numpy(sample["semseg_mask"]).float()[None, None, ...], (height, width), mode="nearest"
-                ).numpy()[0, 0]
-
-            if "mask" in sample:
-                sample["mask"] = cv2.resize(
-                    sample["mask"].astype(np.float32),
-                    (width, height),
-                    interpolation=cv2.INTER_NEAREST,
-                )
-                # sample["mask"] = sample["mask"].astype(bool)
-
-        # print(sample['image'].shape, sample['depth'].shape)
-        return sample
-
-
-class NormalizeImage(object):
-    """Normlize image by given mean and std."""
-
-    def __init__(self, mean, std):
-        self.__mean = mean
-        self.__std = std
-
-    def __call__(self, sample):
-        sample["image"] = (sample["image"] - self.__mean) / self.__std
-
-        return sample
-
-
-class PrepareForNet(object):
-    """Prepare sample for usage as network input."""
-
-    def __init__(self):
-        pass
-
-    def __call__(self, sample):
-        image = np.transpose(sample["image"], (2, 0, 1))
-        sample["image"] = np.ascontiguousarray(image).astype(np.float32)
-
-        if "mask" in sample:
-            sample["mask"] = sample["mask"].astype(np.float32)
-            sample["mask"] = np.ascontiguousarray(sample["mask"])
-
-        if "depth" in sample:
-            depth = sample["depth"].astype(np.float32)
-            sample["depth"] = np.ascontiguousarray(depth)
-
-        if "semseg_mask" in sample:
-            sample["semseg_mask"] = sample["semseg_mask"].astype(np.float32)
-            sample["semseg_mask"] = np.ascontiguousarray(sample["semseg_mask"])
-
-        return sample

invokeai/backend/image_util/grounding_dino/detection_result.py

@@ -0,0 +1,22 @@
+from pydantic import BaseModel, ConfigDict
+
+
+class BoundingBox(BaseModel):
+    """Bounding box helper class."""
+
+    xmin: int
+    ymin: int
+    xmax: int
+    ymax: int
+
+
+class DetectionResult(BaseModel):
+    """Detection result from Grounding DINO."""
+
+    score: float
+    label: str
+    box: BoundingBox
+    model_config = ConfigDict(
+        # Allow arbitrary types for mask, since it will be a numpy array.
+        arbitrary_types_allowed=True
+    )

invokeai/backend/image_util/grounding_dino/grounding_dino_pipeline.py

@@ -0,0 +1,37 @@
+from typing import Optional
+
+import torch
+from PIL import Image
+from transformers.pipelines import ZeroShotObjectDetectionPipeline
+
+from invokeai.backend.image_util.grounding_dino.detection_result import DetectionResult
+from invokeai.backend.raw_model import RawModel
+
+
+class GroundingDinoPipeline(RawModel):
+    """A wrapper class for a ZeroShotObjectDetectionPipeline that makes it compatible with the model manager's memory
+    management system.
+    """
+
+    def __init__(self, pipeline: ZeroShotObjectDetectionPipeline):
+        self._pipeline = pipeline
+
+    def detect(self, image: Image.Image, candidate_labels: list[str], threshold: float = 0.1) -> list[DetectionResult]:
+        results = self._pipeline(image=image, candidate_labels=candidate_labels, threshold=threshold)
+        assert results is not None
+        results = [DetectionResult.model_validate(result) for result in results]
+        return results
+
+    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None):
+        # HACK(ryand): The GroundingDinoPipeline does not work on MPS devices. We only allow it to be moved to CPU or
+        # CUDA.
+        if device is not None and device.type not in {"cpu", "cuda"}:
+            device = None
+        self._pipeline.model.to(device=device, dtype=dtype)
+        self._pipeline.device = self._pipeline.model.device
+
+    def calc_size(self) -> int:
+        # HACK(ryand): Fix the circular import issue.
+        from invokeai.backend.model_manager.load.model_util import calc_module_size
+
+        return calc_module_size(self._pipeline.model)