Update communityNodes.md (#3873)

Added the Ideal Size node

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

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


## Have you discussed this change with the InvokeAI team?
- [ ] Yes
- [X] No, because: It's a community node addition

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


## Description

Added a reference to my community node that calculates the ideal size
for initial latent generation that avoids duplication. This is the logic
that was present in 2.3.5's first pass of high-res optimization.

## Related Tickets & Documents

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

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

- Related Issue #
- Closes #

## QA Instructions, Screenshots, Recordings

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

## Added/updated tests?

- [ ] Yes
- [X] No : This is a documentation change that references my community
node.

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

.github/CODEOWNERS

@@ -6,7 +6,7 @@
 /mkdocs.yml @lstein  @blessedcoolant
 
 # nodes
-/invokeai/app/ @Kyle0654 @blessedcoolant
+/invokeai/app/ @Kyle0654 @blessedcoolant @psychedelicious @brandonrising 
 
 # installation and configuration
 /pyproject.toml  @lstein @blessedcoolant
@@ -22,7 +22,7 @@
 /invokeai/backend @blessedcoolant @psychedelicious @lstein @maryhipp
 
 # generation, model management, postprocessing
-/invokeai/backend  @damian0815 @lstein @blessedcoolant @jpphoto @gregghelt2 @StAlKeR7779
+/invokeai/backend  @damian0815 @lstein @blessedcoolant @gregghelt2 @StAlKeR7779 @brandonrising
 
 # front ends
 /invokeai/frontend/CLI @lstein

.github/workflows/mkdocs-material.yml

@@ -2,7 +2,7 @@ name: mkdocs-material
 on:
   push:
     branches:
-      - 'refs/heads/v2.3'
+      - 'refs/heads/main'
 
 permissions:
     contents: write
@@ -43,7 +43,7 @@ jobs:
             --verbose
 
       - name: deploy to gh-pages
-        if: ${{ github.ref == 'refs/heads/v2.3' }}
+        if: ${{ github.ref == 'refs/heads/main' }}
         run: |
           python -m \
             mkdocs gh-deploy \

README.md

@@ -36,15 +36,6 @@
 
 </div>
 
-_**Note: This is an alpha release. Bugs are expected and not all
-features are fully implemented. Please use the GitHub [Issues
-pages](https://github.com/invoke-ai/InvokeAI/issues?q=is%3Aissue+is%3Aopen)
-to report unexpected problems. Also note that InvokeAI root directory
-which contains models, outputs and configuration files, has changed
-between the 2.x and 3.x release. If you wish to use your v2.3 root
-directory with v3.0, please follow the directions in [Migrating a 2.3
-root directory to 3.0](#migrating-to-3).**_
-
 InvokeAI is a leading creative engine built to empower professionals
 and enthusiasts alike. Generate and create stunning visual media using
 the latest AI-driven technologies. InvokeAI offers an industry leading
@@ -132,8 +123,10 @@ and go to http://localhost:9090.
 
 ### Command-Line Installation (for developers and users familiar with Terminals)
 
-You must have Python 3.9 or 3.10 installed on your machine. Earlier or later versions are
-not supported.
+You must have Python 3.9 or 3.10 installed on your machine. Earlier or
+later versions are not supported.
+Node.js also needs to be installed along with yarn (can be installed with
+the command `npm install -g yarn` if needed)
 
 1. Open a command-line window on your machine. The PowerShell is recommended for Windows.
 2. Create a directory to install InvokeAI into. You'll need at least 15 GB of free space:
@@ -197,11 +190,18 @@ not supported.
 7. Launch the web server (do it every time you run InvokeAI):
 
     ```terminal
-    invokeai --web
+    invokeai-web
     ```
 
-8. Point your browser to http://localhost:9090 to bring up the web interface.
-9. Type `banana sushi` in the box on the top left and click `Invoke`.
+8. Build Node.js assets
+
+  ```terminal
+  cd invokeai/frontend/web/
+  yarn vite build
+  ```
+
+9. Point your browser to http://localhost:9090 to bring up the web interface.
+10. Type `banana sushi` in the box on the top left and click `Invoke`.
 
 Be sure to activate the virtual environment each time before re-launching InvokeAI,
 using `source .venv/bin/activate` or `.venv\Scripts\activate`.
@@ -255,19 +255,24 @@ old models directory (which contains the models selected at install
 time) will be renamed `models.orig` and can be deleted once you have
 confirmed that the migration was successful.
 
+ If you wish, you can pass the 2.3 root directory to both `--from` and
+`--to` in order to update in place. Warning: this directory will no
+longer be usable with InvokeAI 2.3.
+
 #### Migrating in place
 
 For the adventurous, you may do an in-place upgrade from 2.3 to 3.0
-without touching the command line. The recipe is as follows>
+without touching the command line. ***This recipe does not work on
+Windows platforms due to a bug in the Windows version of the 2.3
+upgrade script.** See the next section for a Windows recipe.
+
+##### For Mac and Linux Users:
 
 1. Launch the InvokeAI launcher script in your current v2.3 root directory.
 
 2. Select option [9] "Update InvokeAI" to bring up the updater dialog.
 
-3a. During the alpha release phase, select option [3] and manually
-enter the tag name `v3.0.0+a2`.
-
-3b. Once 3.0 is released, select option [1] to upgrade to the latest release.
+3. Select option [1] to upgrade to the latest release.
 
 4. Once the upgrade is finished you will be returned to the launcher
 menu. Select option [7] "Re-run the configure script to fix a broken
@@ -286,14 +291,33 @@ worked, you can safely remove these files. Alternatively you can
 restore a working v2.3 directory by removing the new files and
 restoring the ".orig" files' original names.
 
+##### For Windows Users:
+
+Windows Users can upgrade with the
+
+1. Enter the 2.3 root directory you wish to upgrade
+2. Launch `invoke.sh` or `invoke.bat`
+3. Select the "Developer's console" option [8]
+4. Type the following commands
+
+```
+pip install "invokeai @ https://github.com/invoke-ai/InvokeAI/archive/refs/tags/v3.0.0" --use-pep517 --upgrade
+invokeai-configure --root .
+```
+(Replace `v3.0.0` with the current release number if this document is out of date).
+
+The first command will install and upgrade new software to run
+InvokeAI. The second will prepare the 2.3 directory for use with 3.0.
+You may now launch the WebUI in the usual way, by selecting option [1]
+from the launcher script
+
 #### Migration Caveats
 
 The migration script will migrate your invokeai settings and models,
 including textual inversion models, LoRAs and merges that you may have
 installed previously. However it does **not** migrate the generated
-images stored in your 2.3-format outputs directory. The released
-version of 3.0 is expected to have an interface for importing an
-entire directory of image files as a batch.
+images stored in your 2.3-format outputs directory. You will need to
+manually import selected images into the 3.0 gallery via drag-and-drop.
 
 ## Hardware Requirements
 
@@ -305,9 +329,12 @@ AMD card (using the ROCm driver).
 
 You will need one of the following:
 
-- An NVIDIA-based graphics card with 4 GB or more VRAM memory.
+- An NVIDIA-based graphics card with 4 GB or more VRAM memory. 6-8 GB
+  of VRAM is highly recommended for rendering using the Stable
+  Diffusion XL models
 - An Apple computer with an M1 chip.
-- An AMD-based graphics card with 4GB or more VRAM memory. (Linux only)
+- An AMD-based graphics card with 4GB or more VRAM memory (Linux
+  only), 6-8 GB for XL rendering.
 
 We do not recommend the GTX 1650 or 1660 series video cards. They are
 unable to run in half-precision mode and do not have sufficient VRAM
@@ -340,13 +367,12 @@ Invoke AI provides an organized gallery system for easily storing, accessing, an
 ### Other features
 
 - *Support for both ckpt and diffusers models*
-- *SD 2.0, 2.1 support*
+- *SD 2.0, 2.1, XL support*
 - *Upscaling Tools*
 - *Embedding Manager & Support*
 - *Model Manager & Support*
 - *Node-Based Architecture*
 - *Node-Based Plug-&-Play UI (Beta)*
-- *SDXL Support* (Coming soon)
 
 ### Latest Changes
 

docs/CHANGELOG.md

@@ -617,8 +617,6 @@ sections describe what's new for InvokeAI.
 - `dream.py` script renamed `invoke.py`. A `dream.py` script wrapper remains for
   backward compatibility.
 - Completely new WebGUI - launch with `python3 scripts/invoke.py --web`
-- Support for [inpainting](deprecated/INPAINTING.md) and
-  [outpainting](features/OUTPAINTING.md)
 - img2img runs on all k\* samplers
 - Support for
   [negative prompts](features/PROMPTS.md#negative-and-unconditioned-prompts)

docs/contributing/CONTRIBUTING.md

@@ -1,42 +1,38 @@
+# How to Contribute
+
 ## Welcome to Invoke AI
-
-We're thrilled to have you here and we're excited for you to contribute. 
-
 Invoke AI originated as a project built by the community, and that vision carries forward today as we aim to build the best pro-grade tools available. We work together to incorporate the latest in AI/ML research, making these tools available in over 20 languages to artists and creatives around the world as part of our fully permissive OSS project designed for individual users to self-host and use.
 
-Here are some guidelines to help you get started:
 
-### Technical Prerequisites
+## Contributing to Invoke AI
+Anyone who wishes to contribute to InvokeAI, whether features, bug fixes, code cleanup, testing, code reviews, documentation or translation is very much encouraged to do so.
 
-Front-end: You'll need a working knowledge of React and TypeScript.
+To join, just raise your hand on the InvokeAI Discord server (#dev-chat) or the GitHub discussion board.
 
-Back-end: Depending on the scope of your contribution, you may need to know SQLite, FastAPI, Python, and Socketio. Also, a good majority of the backend logic involved in processing images is built in a modular way using a concept called "Nodes", which are isolated functions that carry out individual, discrete operations. This design allows for easy contributions of novel pipelines and capabilities.
+### Areas of contribution: 
 
-### How to Submit Contributions
+#### Development
+If you’d like to help with development, please see our [development guide](contribution_guides/development.md). If you’re unfamiliar with contributing to open source projects, there is a tutorial contained within the development guide.
 
-To start contributing, please follow these steps:
+#### Documentation
+If you’d like to help with documentation, please see our [documentation guide](contribution_guides/documenation.md).
 
-1. Familiarize yourself with our roadmap and open projects to see where your skills and interests align. These documents can serve as a source of inspiration.
-2. Open a Pull Request (PR) with a clear description of the feature you're adding or the problem you're solving. Make sure your contribution aligns with the project's vision.
-3. Adhere to general best practices. This includes assuming interoperability with other nodes, keeping the scope of your functions as small as possible, and organizing your code according to our architecture documents.
+#### Translation
+If you'd like to help with translation, please see our [translation guide](docs/contributing/.contribution_guides/translation.md).
 
-### Types of Contributions We're Looking For
+#### Tutorials 
+Please reach out to @imic or @hipsterusername on [Discord](https://discord.gg/ZmtBAhwWhy) to help create tutorials for InvokeAI.
 
-We welcome all contributions that improve the project. Right now, we're especially looking for:
+We hope you enjoy using our software as much as we enjoy creating it, and we hope that some of those of you who are reading this will elect to become part of our contributor community.
 
-1. Quality of life (QOL) enhancements on the front-end.
-2. New backend capabilities added through nodes.
-3. Incorporating additional optimizations from the broader open-source software community.
 
-### Communication and Decision-making Process
+### Contributors
 
-Project maintainers and code owners review PRs to ensure they align with the project's goals. They may provide design or architectural guidance, suggestions on user experience, or provide more significant feedback on the contribution itself. Expect to receive feedback on your submissions, and don't hesitate to ask questions or propose changes.
+This project is a combined effort of dedicated people from across the world. [Check out the list of all these amazing people](https://invoke-ai.github.io/InvokeAI/other/CONTRIBUTORS/). We thank them for their time, hard work and effort.
 
-For more robust discussions, or if you're planning to add capabilities not currently listed on our roadmap, please reach out to us on our Discord server. That way, we can ensure your proposed contribution aligns with the project's direction before you start writing code.
+### Code of Conduct
 
-### Code of Conduct and Contribution Expectations
-
-We want everyone in our community to have a positive experience. To facilitate this, we've established a code of conduct and a statement of values that we expect all contributors to adhere to. Please take a moment to review these documents—they're essential to maintaining a respectful and inclusive environment.
+The InvokeAI community is a welcoming place, and we want your help in maintaining that. Please review our [Code of Conduct](https://github.com/invoke-ai/InvokeAI/blob/main/CODE_OF_CONDUCT.md) to learn more - it's essential to maintaining a respectful and inclusive environment.
 
 By making a contribution to this project, you certify that:
 
@@ -49,6 +45,12 @@ This disclaimer is not a license and does not grant any rights or permissions. Y
 
 This disclaimer is provided "as is" without warranty of any kind, whether expressed or implied, including but not limited to the warranties of merchantability, fitness for a particular purpose, or non-infringement. In no event shall the authors or copyright holders be liable for any claim, damages, or other liability, whether in an action of contract, tort, or otherwise, arising from, out of, or in connection with the contribution or the use or other dealings in the contribution.
 
+### Support
+
+For support, please use this repository's [GitHub Issues](https://github.com/invoke-ai/InvokeAI/issues), or join the [Discord](https://discord.gg/ZmtBAhwWhy).
+
+Original portions of the software are Copyright (c) 2023 by respective contributors.
+
 ---
 
 Remember, your contributions help make this project great. We're excited to see what you'll bring to our community!

docs/contributing/LOCAL_DEVELOPMENT.md

@@ -81,3 +81,193 @@ pytest --cov; open ./coverage/html/index.html
 <!--#TODO: get input from blessedcoolant here, for the moment inserted the frontend README via snippets extension.-->
 
 --8<-- "invokeai/frontend/web/README.md"
+
+## Developing InvokeAI in VSCode
+
+VSCode offers some nice tools:
+
+- python debugger
+- automatic `venv` activation
+- remote dev (e.g. run InvokeAI on a beefy linux desktop while you type in
+  comfort on your macbook)
+
+### Setup
+
+You'll need the
+[Python](https://marketplace.visualstudio.com/items?itemName=ms-python.python)
+and
+[Pylance](https://marketplace.visualstudio.com/items?itemName=ms-python.vscode-pylance)
+extensions installed first.
+
+It's also really handy to install the `Jupyter` extensions:
+
+- [Jupyter](https://marketplace.visualstudio.com/items?itemName=ms-toolsai.jupyter)
+- [Jupyter Cell Tags](https://marketplace.visualstudio.com/items?itemName=ms-toolsai.vscode-jupyter-cell-tags)
+- [Jupyter Notebook Renderers](https://marketplace.visualstudio.com/items?itemName=ms-toolsai.jupyter-renderers)
+- [Jupyter Slide Show](https://marketplace.visualstudio.com/items?itemName=ms-toolsai.vscode-jupyter-slideshow)
+
+#### InvokeAI workspace
+
+Creating a VSCode workspace for working on InvokeAI is highly recommended. It
+can hold InvokeAI-specific settings and configs.
+
+To make a workspace:
+
+- Open the InvokeAI repo dir in VSCode
+- `File` > `Save Workspace As` > save it _outside_ the repo
+
+#### Default python interpreter (i.e. automatic virtual environment activation)
+
+- Use command palette to run command
+  `Preferences: Open Workspace Settings (JSON)`
+- Add `python.defaultInterpreterPath` to `settings`, pointing to your `venv`'s
+  python
+
+Should look something like this:
+
+```jsonc
+{
+  // I like to have all InvokeAI-related folders in my workspace
+  "folders": [
+    {
+      // repo root
+      "path": "InvokeAI"
+    },
+    {
+      // InvokeAI root dir, where `invokeai.yaml` lives
+      "path": "/path/to/invokeai_root"
+    }
+  ],
+  "settings": {
+    // Where your InvokeAI `venv`'s python executable lives
+    "python.defaultInterpreterPath": "/path/to/invokeai_root/.venv/bin/python"
+  }
+}
+```
+
+Now when you open the VSCode integrated terminal, or do anything that needs to
+run python, it will automatically be in your InvokeAI virtual environment.
+
+Bonus: When you create a Jupyter notebook, when you run it, you'll be prompted
+for the python interpreter to run in. This will default to your `venv` python,
+and so you'll have access to the same python environment as the InvokeAI app.
+
+This is _super_ handy.
+
+#### Debugging configs with `launch.json`
+
+Debugging configs are managed in a `launch.json` file. Like most VSCode configs,
+these can be scoped to a workspace or folder.
+
+Follow the [official guide](https://code.visualstudio.com/docs/python/debugging)
+to set up your `launch.json` and try it out.
+
+Now we can create the InvokeAI debugging configs:
+
+```jsonc
+{
+  // Use IntelliSense to learn about possible attributes.
+  // Hover to view descriptions of existing attributes.
+  // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
+  "version": "0.2.0",
+  "configurations": [
+    {
+      // Run the InvokeAI backend & serve the pre-built UI
+      "name": "InvokeAI Web",
+      "type": "python",
+      "request": "launch",
+      "program": "scripts/invokeai-web.py",
+      "args": [
+        // Your InvokeAI root dir (where `invokeai.yaml` lives)
+        "--root",
+        "/path/to/invokeai_root",
+        // Access the app from anywhere on your local network
+        "--host",
+        "0.0.0.0"
+      ],
+      "justMyCode": true
+    },
+    {
+      // Run the nodes-based CLI
+      "name": "InvokeAI CLI",
+      "type": "python",
+      "request": "launch",
+      "program": "scripts/invokeai-cli.py",
+      "justMyCode": true
+    },
+    {
+      // Run tests
+      "name": "InvokeAI Test",
+      "type": "python",
+      "request": "launch",
+      "module": "pytest",
+      "args": ["--capture=no"],
+      "justMyCode": true
+    },
+    {
+      // Run a single test
+      "name": "InvokeAI Single Test",
+      "type": "python",
+      "request": "launch",
+      "module": "pytest",
+      "args": [
+        // Change this to point to the specific test you are working on
+        "tests/nodes/test_invoker.py"
+      ],
+      "justMyCode": true
+    },
+    {
+      // This is the default, useful to just run a single file
+      "name": "Python: File",
+      "type": "python",
+      "request": "launch",
+      "program": "${file}",
+      "justMyCode": true
+    }
+  ]
+}
+```
+
+You'll see these configs in the debugging configs drop down. Running them will
+start InvokeAI with attached debugger, in the correct environment, and work just
+like the normal app.
+
+Enjoy debugging InvokeAI with ease (not that we have any bugs of course).
+
+#### Remote dev
+
+This is very easy to set up and provides the same very smooth experience as
+local development. Environments and debugging, as set up above, just work,
+though you'd need to recreate the workspace and debugging configs on the remote.
+
+Consult the
+[official guide](https://code.visualstudio.com/docs/remote/remote-overview) to
+get it set up.
+
+Suggest using VSCode's included settings sync so that your remote dev host has
+all the same app settings and extensions automagically.
+
+##### One remote dev gotcha
+
+I've found the automatic port forwarding to be very flakey. You can disable it
+in `Preferences: Open Remote Settings (ssh: hostname)`. Search for
+`remote.autoForwardPorts` and untick the box.
+
+To forward ports very reliably, use SSH on the remote dev client (e.g. your
+macbook). Here's how to forward both backend API port (`9090`) and the frontend
+live dev server port (`5173`):
+
+```bash
+ssh \
+    -L 9090:localhost:9090 \
+    -L 5173:localhost:5173 \
+    user@remote-dev-host
+```
+
+The forwarding stops when you close the terminal window, so suggest to do this
+_outside_ the VSCode integrated terminal in case you need to restart VSCode for
+an extension update or something
+
+Now, on your remote dev client, you can open `localhost:9090` and access the UI,
+now served from the remote dev host, just the same as if it was running on the
+client.

docs/contributing/contribution_guides/development.md

@@ -0,0 +1,91 @@
+# Development
+
+## **What do I need to know to help?**
+
+If you are looking to help to with a code contribution, InvokeAI uses several different technologies under the hood: Python (Pydantic, FastAPI, diffusers) and Typescript (React, Redux Toolkit, ChakraUI, Mantine, Konva). Familiarity with StableDiffusion and image generation concepts is helpful, but not essential. 
+
+For more information, please review our area specific documentation:
+
+* #### [InvokeAI Architecure](../ARCHITECTURE.md)
+* #### [Frontend Documentation](development_guides/contributingToFrontend.md)
+* #### [Node Documentation](../INVOCATIONS.md)
+* #### [Local Development](../LOCAL_DEVELOPMENT.md)
+
+If you don't feel ready to make a code contribution yet, no problem! You can also help out in other ways, such as [documentation](documentation.md) or [translation](translation.md).
+
+There are two paths to making a development contribution: 
+
+1. Choosing an open issue to address. Open issues can be found in the [Issues](https://github.com/invoke-ai/InvokeAI/issues?q=is%3Aissue+is%3Aopen) section of the InvokeAI repository. These are tagged by the issue type (bug, enhancement, etc.) along with the “good first issues” tag denoting if they are suitable for first time contributors.
+    1. Additional items can be found on our roadmap <******************************link to roadmap>******************************. The roadmap is organized in terms of priority, and contains features of varying size and complexity. If there is an inflight item you’d like to help with, reach out to the contributor assigned to the item to see how you can help. 
+2. Opening a new issue or feature to add. **Please make sure you have searched through existing issues before creating new ones.**
+
+*Regardless of what you choose, please post in the  [#dev-chat](https://discord.com/channels/1020123559063990373/1049495067846524939) channel of the Discord before you start development in order to confirm that the issue or feature is aligned with the current direction of the project. We value our contributors time and effort and want to ensure that no one’s time is being misspent.*
+
+## Best Practices: 
+* Keep your pull requests small. Smaller pull requests are more likely to be accepted and merged
+* Comments! Commenting your code helps reviwers easily understand your contribution
+* Use Python and Typescript’s typing systems, and consider using an editor with [LSP](https://microsoft.github.io/language-server-protocol/) support to streamline development
+* Make all communications public. This ensure knowledge is shared with the whole community
+
+## **How do I make a contribution?**
+
+Never made an open source contribution before? Wondering how contributions work in our project? Here's a quick rundown!
+
+Before starting these steps, ensure you have your local environment [configured for development](../LOCAL_DEVELOPMENT.md).
+
+1.  Find a [good first issue](https://github.com/invoke-ai/InvokeAI/contribute) that you are interested in addressing or a feature that you would like to add. Then, reach out to our team in the [#dev-chat](https://discord.com/channels/1020123559063990373/1049495067846524939) channel of the Discord to ensure you are  setup for success. 
+2. Fork the [InvokeAI](https://github.com/invoke-ai/InvokeAI) repository to your GitHub profile. This means that you will have a copy of the repository under **your-GitHub-username/InvokeAI**.
+3. Clone the repository to your local machine using:
+
+```bash
+git clone https://github.com/your-GitHub-username/InvokeAI.git
+```
+
+If you're unfamiliar with using Git through the commandline, [GitHub Desktop](https://desktop.github.com) is a easy-to-use alternative with a UI. You can do all the same steps listed here, but through the interface. 
+
+4. Create a new branch for your fix using:
+
+```bash
+git checkout -b branch-name-here
+```
+
+5. Make the appropriate changes for the issue you are trying to address or the feature that you want to add.
+6. Add the file contents of the changed files to the "snapshot" git uses to manage the state of the project, also known as the index:
+
+```bash
+git add insert-paths-of-changed-files-here
+```
+
+7. Store the contents of the index with a descriptive message.
+
+```bash
+git commit -m "Insert a short message of the changes made here"
+```
+
+8. Push the changes to the remote repository using
+
+```markdown
+git push origin branch-name-here
+```
+
+9. Submit a pull request to the **main** branch of the InvokeAI repository.
+10. Title the pull request with a short description of the changes made and the issue or bug number associated with your change. For example, you can title an issue like so "Added more log outputting to resolve #1234".
+11. In the description of the pull request, explain the changes that you made, any issues you think exist with the pull request you made, and any questions you have for the maintainer. It's OK if your pull request is not perfect (no pull request is), the reviewer will be able to help you fix any problems and improve it!
+12. Wait for the pull request to be reviewed by other collaborators.
+13. Make changes to the pull request if the reviewer(s) recommend them.
+14. Celebrate your success after your pull request is merged!
+
+If you’d like to learn more about contributing to Open Source projects, here is a [Getting Started Guide](https://opensource.com/article/19/7/create-pull-request-github). 
+
+## **Where can I go for help?**
+
+If you need help, you can ask questions in the [#dev-chat](https://discord.com/channels/1020123559063990373/1049495067846524939) channel of the Discord.
+
+For frontend related work, **@pyschedelicious** is the best person to reach out to. 
+
+For backend related work, please reach out to **@blessedcoolant**, **@lstein**, **@StAlKeR7779** or **@pyschedelicious**.
+
+## **What does the Code of Conduct mean for me?**
+
+Our [Code of Conduct](CODE_OF_CONDUCT.md)  means that you are responsible for treating everyone on the project with respect and courtesy regardless of their identity. If you are the victim of any inappropriate behavior or comments as described in our Code of Conduct, we are here for you and will do the best to ensure that the abuser is reprimanded appropriately, per our code.
+

docs/contributing/contribution_guides/development_guides/contributingToFrontend.md

@@ -0,0 +1,75 @@
+# Contributing to the Frontend
+
+# InvokeAI Web UI
+
+- [InvokeAI Web UI](https://github.com/invoke-ai/InvokeAI/tree/main/invokeai/frontend/web/docs#invokeai-web-ui)
+    - [Stack](https://github.com/invoke-ai/InvokeAI/tree/main/invokeai/frontend/web/docs#stack)
+    - [Contributing](https://github.com/invoke-ai/InvokeAI/tree/main/invokeai/frontend/web/docs#contributing)
+        - [Dev Environment](https://github.com/invoke-ai/InvokeAI/tree/main/invokeai/frontend/web/docs#dev-environment)
+        - [Production builds](https://github.com/invoke-ai/InvokeAI/tree/main/invokeai/frontend/web/docs#production-builds)
+
+The UI is a fairly straightforward Typescript React app, with the Unified Canvas being more complex.
+
+Code is located in `invokeai/frontend/web/` for review.
+
+## Stack
+
+State management is Redux via [Redux Toolkit](https://github.com/reduxjs/redux-toolkit). We lean heavily on RTK:
+
+- `createAsyncThunk` for HTTP requests
+- `createEntityAdapter` for fetching images and models
+- `createListenerMiddleware` for workflows
+
+The API client and associated types are generated from the OpenAPI schema. See API_CLIENT.md.
+
+Communication with server is a mix of HTTP and [socket.io](https://github.com/socketio/socket.io-client) (with a simple socket.io redux middleware to help).
+
+[Chakra-UI](https://github.com/chakra-ui/chakra-ui) & [Mantine](https://github.com/mantinedev/mantine) for components and styling.
+
+[Konva](https://github.com/konvajs/react-konva) for the canvas, but we are pushing the limits of what is feasible with it (and HTML canvas in general). We plan to rebuild it with [PixiJS](https://github.com/pixijs/pixijs) to take advantage of WebGL's improved raster handling.
+
+[Vite](https://vitejs.dev/) for bundling.
+
+Localisation is via [i18next](https://github.com/i18next/react-i18next), but translation happens on our [Weblate](https://hosted.weblate.org/engage/invokeai/) project. Only the English source strings should be changed on this repo.
+
+## Contributing
+
+Thanks for your interest in contributing to the InvokeAI Web UI!
+
+We encourage you to ping @psychedelicious and @blessedcoolant on [Discord](https://discord.gg/ZmtBAhwWhy) if you want to contribute, just to touch base and ensure your work doesn't conflict with anything else going on. The project is very active.
+
+### Dev Environment
+
+**Setup** 
+
+1. Install [node](https://nodejs.org/en/download/). You can confirm node is installed with:
+```bash
+node --version
+```
+2. Install [yarn classic](https://classic.yarnpkg.com/lang/en/) and confirm it is installed by running this:
+```bash
+npm install --global yarn
+yarn --version
+```
+
+From `invokeai/frontend/web/` run `yarn install` to get everything set up.
+
+Start everything in dev mode:
+1. Ensure your virtual environment is running
+2. Start the dev server: `yarn dev`
+3. Start the InvokeAI Nodes backend: `python scripts/invokeai-web.py # run from the repo root`
+4. Point your browser to the dev server address e.g. [http://localhost:5173/](http://localhost:5173/)
+
+### VSCode Remote Dev
+
+We've noticed an intermittent issue with the VSCode Remote Dev port forwarding. If you use this feature of VSCode, you may intermittently click the Invoke button and then get nothing until the request times out. Suggest disabling the IDE's port forwarding feature and doing it manually via SSH:
+
+`ssh -L 9090:localhost:9090 -L 5173:localhost:5173 user@host`
+
+### Production builds
+
+For a number of technical and logistical reasons, we need to commit UI build artefacts to the repo.
+
+If you submit a PR, there is a good chance we will ask you to include a separate commit with a build of the app.
+
+To build for production, run `yarn build`.

docs/contributing/contribution_guides/documentation.md

@@ -0,0 +1,13 @@
+# Documentation
+
+Documentation is an important part of any open source project. It provides a clear and concise way to communicate how the software works, how to use it, and how to troubleshoot issues. Without proper documentation, it can be difficult for users to understand the purpose and functionality of the project. 
+
+## Contributing
+
+All documentation is maintained in the InvokeAI GitHub repository. If you come across documentation that is out of date or incorrect, please submit a pull request with the necessary changes. 
+
+When updating or creating documentation, please keep in mind InvokeAI is a tool for everyone, not just those who have familiarity with generative art. 
+
+## Help & Questions
+
+Please ping @imic1 or @hipsterusername in the [Discord](https://discord.com/channels/1020123559063990373/1049495067846524939) if you have any questions.

docs/contributing/contribution_guides/translation.md

@@ -0,0 +1,19 @@
+# Translation
+
+InvokeAI uses [Weblate](https://weblate.org/) for translation. Weblate is a FOSS project providing a scalable translation service. Weblate automates the tedious parts of managing translation of a growing project, and the service is generously provided at no cost to FOSS projects like InvokeAI.
+
+## Contributing
+
+If you'd like to contribute by adding or updating a translation, please visit our [Weblate project](https://hosted.weblate.org/engage/invokeai/). You'll need to sign in with your GitHub account (a number of other accounts are supported, including Google).
+
+Once signed in, select a language and then the Web UI component. From here you can Browse and Translate strings from English to your chosen language. Zen mode offers a simpler translation experience.
+
+Your changes will be attributed to you in the automated PR process; you don't need to do anything else.
+
+## Help & Questions
+
+Please check Weblate's [documentation](https://docs.weblate.org/en/latest/index.html) or ping @Harvestor on [Discord](https://discord.com/channels/1020123559063990373/1049495067846524939) if you have any questions.
+
+## Thanks
+
+Thanks to the InvokeAI community for their efforts to translate the project!

docs/contributing/contribution_guides/tutorials.md

@@ -0,0 +1,11 @@
+# Tutorials
+
+Tutorials help new & existing users expand their abilty to use InvokeAI to the full extent of our features and services.  
+
+Currently, we have a set of tutorials available on our [YouTube channel](https://www.youtube.com/@invokeai), but as InvokeAI continues to evolve with new updates, we want to ensure that we are giving our users the resources they need to succeed. 
+
+Tutorials can be in the form of videos or article walkthroughs on a subject of your choice. We recommend focusing tutorials on the key image generation methods, or on a specific component within one of the image generation methods.
+
+## Contributing
+
+Please reach out to @imic or @hipsterusername on [Discord](https://discord.gg/ZmtBAhwWhy) to help create tutorials for InvokeAI.

docs/features/CONCEPTS.md

@@ -1,8 +1,8 @@
 ---
-title: Concepts 
+title: Textual Inversion Embeddings and LoRAs
 ---
 
-# :material-library-shelves: The Hugging Face Concepts Library and Importing Textual Inversion files
+# :material-library-shelves: Textual Inversions and LoRAs
 
 With the advances in research, many new capabilities are available to customize the knowledge and understanding of novel concepts not originally contained in the base model. 
 
@@ -64,21 +64,25 @@ select the embedding you'd like to use. This UI has type-ahead support, so you c
 
 ## Using LoRAs
 
-LoRA files are models that customize the output of Stable Diffusion image generation.
-Larger than embeddings, but much smaller than full models, they augment SD with improved
-understanding of subjects and artistic styles.
+LoRA files are models that customize the output of Stable Diffusion
+image generation.  Larger than embeddings, but much smaller than full
+models, they augment SD with improved understanding of subjects and
+artistic styles.
 
-Unlike TI files, LoRAs do not introduce novel vocabulary into the model's known tokens. Instead,
-LoRAs augment the model's weights that are applied to generate imagery. LoRAs may be supplied
-with a "trigger" word that they have been explicitly trained on, or may simply apply their 
-effect without being triggered.
+Unlike TI files, LoRAs do not introduce novel vocabulary into the
+model's known tokens. Instead, LoRAs augment the model's weights that
+are applied to generate imagery. LoRAs may be supplied with a
+"trigger" word that they have been explicitly trained on, or may
+simply apply their effect without being triggered.
 
-LoRAs are typically stored in .safetensors files, which are the most secure way to store and transmit
-these types of weights. You may install any number of `.safetensors` LoRA files simply by copying them into
-the `lora` directory of the corresponding InvokeAI models directory (usually `invokeai`
-in your home directory). For example, you can simply move a Stable Diffusion 1.5 LoRA file to
-the `sd-1/lora` folder.
+LoRAs are typically stored in .safetensors files, which are the most
+secure way to store and transmit these types of weights. You may
+install any number of `.safetensors` LoRA files simply by copying them
+into the `autoimport/lora` directory of the corresponding InvokeAI models
+directory (usually `invokeai` in your home directory).
 
-To use these when generating, open the LoRA menu item in the options panel, select the LoRAs you want to apply
-and ensure that they have the appropriate weight recommended by the model provider. Typically, most LoRAs perform best at a weight of .75-1. 
+To use these when generating, open the LoRA menu item in the options
+panel, select the LoRAs you want to apply and ensure that they have
+the appropriate weight recommended by the model provider. Typically,
+most LoRAs perform best at a weight of .75-1.
 

docs/features/CONFIGURATION.md

@@ -0,0 +1,287 @@
+---
+title: Configuration
+---
+
+# :material-tune-variant: InvokeAI Configuration
+
+## Intro
+
+InvokeAI has numerous runtime settings which can be used to adjust
+many aspects of its operations, including the location of files and
+directories, memory usage, and performance. These settings can be
+viewed and customized in several ways:
+
+1. By editing settings in the `invokeai.yaml` file.
+2. By setting environment variables.
+3. On the command-line, when InvokeAI is launched.
+
+In addition, the most commonly changed settings are accessible
+graphically via the `invokeai-configure` script.
+
+### How the Configuration System Works
+
+When InvokeAI is launched, the very first thing it needs to do is to
+find its "root" directory, which contains its configuration files,
+installed models, its database of images, and the folder(s) of
+generated images themselves. In this document, the root directory will
+be referred to as ROOT.
+
+#### Finding the Root Directory
+
+To find its root directory, InvokeAI uses the following recipe:
+
+1. It first looks for the argument `--root <path>` on the command line
+it was launched from, and uses the indicated path if present.
+
+2. Next it looks for the environment variable INVOKEAI_ROOT, and uses
+the directory path found there if present.
+
+3. If neither of these are present, then InvokeAI looks for the
+folder containing the `.venv` Python virtual environment directory for
+the currently active environment. This directory is checked for files
+expected inside the InvokeAI root before it is used.
+
+4. Finally, InvokeAI looks for a directory in the current user's home
+directory named `invokeai`.
+
+#### Reading the InvokeAI Configuration File
+
+Once the root directory has been located, InvokeAI looks for a file
+named `ROOT/invokeai.yaml`, and if present reads configuration values
+from it. The top of this file looks like this:
+
+```
+InvokeAI:
+  Web Server:
+    host: localhost
+    port: 9090
+    allow_origins: []
+    allow_credentials: true
+    allow_methods:
+    - '*'
+    allow_headers:
+    - '*'
+  Features:
+    esrgan: true
+    internet_available: true
+    log_tokenization: false
+    nsfw_checker: false
+    patchmatch: true
+    restore: true
+...
+```
+
+This lines in this file are used to establish default values for
+Invoke's settings. In the above fragment, the Web Server's listening
+port is set to 9090 by the `port` setting.
+
+You can edit this file with a text editor such as "Notepad" (do not
+use Word or any other word processor). When editing, be careful to
+maintain the indentation, and do not add extraneous text, as syntax
+errors will prevent InvokeAI from launching. A basic guide to the
+format of YAML files can be found
+[here](https://circleci.com/blog/what-is-yaml-a-beginner-s-guide/).
+
+You can fix a broken `invokeai.yaml` by deleting it and running the
+configuration script again -- option [7] in the launcher, "Re-run the
+configure script".
+
+#### Reading Environment Variables
+
+Next InvokeAI looks for defined environment variables in the format
+`INVOKEAI_<setting_name>`, for example `INVOKEAI_port`. Environment
+variable values take precedence over configuration file variables. On
+a Macintosh system, for example, you could change the port that the
+web server listens on by setting the environment variable this way:
+
+```
+export INVOKEAI_port=8000
+invokeai-web
+```
+
+Please check out these
+[Macintosh](https://phoenixnap.com/kb/set-environment-variable-mac)
+and
+[Windows](https://phoenixnap.com/kb/windows-set-environment-variable)
+guides for setting temporary and permanent environment variables.
+
+#### Reading the Command Line
+
+Lastly, InvokeAI takes settings from the command line, which override
+everything else. The command-line settings have the same name as the
+corresponding configuration file settings, preceded by a `--`, for
+example `--port 8000`.
+
+If you are using the launcher (`invoke.sh` or `invoke.bat`) to launch
+InvokeAI, then just pass the command-line arguments to the launcher:
+
+```
+invoke.bat --port 8000 --host 0.0.0.0
+```
+
+The arguments will be applied when you select the web server option
+(and the other options as well).
+
+If, on the other hand, you prefer to launch InvokeAI directly from the
+command line, you would first activate the virtual environment (known
+as the "developer's console" in the launcher), and run `invokeai-web`:
+
+```
+> C:\Users\Fred\invokeai\.venv\scripts\activate
+(.venv) > invokeai-web --port 8000 --host 0.0.0.0
+```
+
+You can get a listing and brief instructions for each of the
+command-line options by giving the `--help` argument:
+
+```
+(.venv) > invokeai-web --help
+usage: InvokeAI [-h] [--host HOST] [--port PORT] [--allow_origins [ALLOW_ORIGINS ...]] [--allow_credentials | --no-allow_credentials]
+                [--allow_methods [ALLOW_METHODS ...]] [--allow_headers [ALLOW_HEADERS ...]] [--esrgan | --no-esrgan]
+                [--internet_available | --no-internet_available] [--log_tokenization | --no-log_tokenization]
+                [--nsfw_checker | --no-nsfw_checker] [--patchmatch | --no-patchmatch] [--restore | --no-restore]
+                [--always_use_cpu | --no-always_use_cpu] [--free_gpu_mem | --no-free_gpu_mem] [--max_cache_size MAX_CACHE_SIZE]
+                [--max_vram_cache_size MAX_VRAM_CACHE_SIZE] [--precision {auto,float16,float32,autocast}]
+                [--sequential_guidance | --no-sequential_guidance] [--xformers_enabled | --no-xformers_enabled]
+                [--tiled_decode | --no-tiled_decode] [--root ROOT] [--autoimport_dir AUTOIMPORT_DIR] [--lora_dir LORA_DIR]
+                [--embedding_dir EMBEDDING_DIR] [--controlnet_dir CONTROLNET_DIR] [--conf_path CONF_PATH] [--models_dir MODELS_DIR]
+                [--legacy_conf_dir LEGACY_CONF_DIR] [--db_dir DB_DIR] [--outdir OUTDIR] [--from_file FROM_FILE]
+                [--use_memory_db | --no-use_memory_db] [--model MODEL] [--log_handlers [LOG_HANDLERS ...]]
+                [--log_format {plain,color,syslog,legacy}] [--log_level {debug,info,warning,error,critical}]
+...
+```
+
+## The Configuration Settings
+
+The configuration settings are divided into several distinct
+groups in `invokeia.yaml`:
+
+### Web Server
+
+| Setting  | Default Value  |  Description |
+|----------|----------------|--------------|
+| `host`     | `localhost`      | Name or IP address of the network interface that the web server will listen on  |
+| `port`     | `9090`           | Network port number that the web server will listen on  |
+| `allow_origins`  | `[]`       | A list of host names or IP addresses that are allowed to connect to the InvokeAI API in the format `['host1','host2',...]` |
+| `allow_credentials | `true`   | Require credentials for a foreign host to access the InvokeAI API (don't change this) |
+| `allow_methods` | `*`         | List of HTTP methods ("GET", "POST") that the web server is allowed to use when accessing the API  |
+| `allow_headers` | `*`         | List of HTTP headers that the web server will accept when accessing the API  |
+
+The documentation for InvokeAI's API can be accessed by browsing to the following URL: [http://localhost:9090/docs].
+
+### Features
+
+These configuration settings allow you to enable and disable various InvokeAI features:
+
+| Setting  | Default Value  |  Description |
+|----------|----------------|--------------|
+| `esrgan`     | `true`      | Activate the ESRGAN upscaling options|
+| `internet_available` | `true`     | When a resource is not available locally, try to fetch it via the internet |
+| `log_tokenization` | `false`      | Before each text2image generation, print a color-coded representation of the prompt to the console; this can help understand why a prompt is not working as expected |
+| `nsfw_checker` | `true`     | Activate the NSFW checker to blur out risque images |
+| `patchmatch` | `true`     | Activate the "patchmatch" algorithm for improved inpainting |
+| `restore`    | `true`     | Activate the facial restoration features (DEPRECATED; restoration features will be removed in 3.0.0) |
+
+### Memory/Performance
+
+These options tune InvokeAI's memory and performance characteristics.
+
+| Setting  | Default Value  |  Description |
+|----------|----------------|--------------|
+| `always_use_cpu`     | `false`      | Use the CPU to generate images, even if a GPU is available |
+| `free_gpu_mem`       | `false`      | Aggressively free up GPU memory after each operation; this will allow you to run in low-VRAM environments with some performance penalties |
+| `max_cache_size`       | `6`      | Amount of CPU RAM (in GB) to reserve for caching models in memory; more cache allows you to keep models in memory and switch among them quickly |
+| `max_vram_cache_size`  | `2.75`   | Amount of GPU VRAM (in GB) to reserve for caching models in VRAM; more cache speeds up generation but reduces the size of the images that can be generated. This can be set to zero to maximize the amount of memory available for generation. |
+| `precision`       | `auto`      | Floating point precision. One of `auto`, `float16` or `float32`. `float16` will consume half the memory of `float32` but produce slightly lower-quality images. The `auto` setting will guess the proper precision based on your video card and operating system |
+| `sequential_guidance`     | `false`      | Calculate guidance in serial rather than in parallel, lowering memory requirements at the cost of some performance loss |
+| `xformers_enabled`        | `true`      | If the x-formers memory-efficient attention module is installed, activate it for better memory usage and generation speed|
+| `tiled_decode`            | `false`     | If true, then during the VAE decoding phase the image will be decoded a section at a time, reducing memory consumption at the cost of a performance hit |
+
+### Paths
+
+These options set the paths of various directories and files used by
+InvokeAI. Relative paths are interpreted relative to INVOKEAI_ROOT, so
+if INVOKEAI_ROOT is `/home/fred/invokeai` and the path is
+`autoimport/main`, then the corresponding directory will be located at
+`/home/fred/invokeai/autoimport/main`.
+
+| Setting  | Default Value  |  Description |
+|----------|----------------|--------------|
+| `autoimport_dir` | `autoimport/main`     | At startup time, read and import any main model files found in this directory |
+| `lora_dir` | `autoimport/lora`     | At startup time, read and import any LoRA/LyCORIS models found in this directory |
+| `embedding_dir` | `autoimport/embedding`  | At startup time, read and import any textual inversion (embedding) models found in this directory |
+| `controlnet_dir` | `autoimport/controlnet`  | At startup time, read and import any ControlNet models found in this directory |
+| `conf_path` | `configs/models.yaml`  | Location of the `models.yaml` model configuration file |
+| `models_dir` | `models`  | Location of the directory containing models installed by InvokeAI's model manager |
+| `legacy_conf_dir` | `configs/stable-diffusion`  | Location of the directory containing the .yaml configuration files for legacy checkpoint models |
+| `db_dir` | `databases`  | Location of the directory containing InvokeAI's image, schema and session database |
+| `outdir` | `outputs`  | Location of the directory in which the gallery of generated and uploaded images will be stored |
+| `use_memory_db` | `false`  | Keep database information in memory rather than on disk; this will not preserve image gallery information across restarts |
+
+Note that the autoimport directories will be searched recursively,
+allowing you to organize the models into folders and subfolders in any
+way you wish. In addition, while we have split up autoimport
+directories by the type of model they contain, this isn't
+necessary. You can combine different model types in the same folder
+and InvokeAI will figure out what they are. So you can easily use just
+one autoimport directory by commenting out the unneeded paths:
+
+```
+Paths:
+  autoimport_dir: autoimport
+#  lora_dir: null
+#  embedding_dir: null
+#  controlnet_dir: null
+```
+
+### Logging
+
+These settings control the information, warning, and debugging
+messages printed to the console log while InvokeAI is running:
+
+| Setting  | Default Value  |  Description |
+|----------|----------------|--------------|
+| `log_handlers` | `console` | This controls where log messages are sent, and can be a list of one or more destinations. Values include `console`, `file`, `syslog` and `http`. These are described in more detail below |
+| `log_format` | `color` | This controls the formatting of the log messages. Values are `plain`, `color`, `legacy` and `syslog` |
+| `log_level`  | `debug` | This filters messages according to the level of severity and can be one of `debug`, `info`, `warning`, `error` and `critical`. For example, setting to `warning` will display all messages at the warning level or higher, but won't display "debug" or "info" messages |
+
+Several different log handler destinations are available, and multiple destinations are supported by providing a list:
+
+```
+  log_handlers:
+     - console
+     - syslog=localhost
+     - file=/var/log/invokeai.log
+```
+
+* `console` is the default. It prints log messages to the command-line window from which InvokeAI was launched.
+
+* `syslog` is only available on Linux and Macintosh systems. It uses
+  the operating system's "syslog" facility to write log file entries
+  locally or to a remote logging machine. `syslog` offers a variety
+  of configuration options:
+
+```
+  syslog=/dev/log`      - log to the /dev/log device
+  syslog=localhost`     - log to the network logger running on the local machine
+  syslog=localhost:512` - same as above, but using a non-standard port
+  syslog=fredserver,facility=LOG_USER,socktype=SOCK_DRAM`
+                        - Log to LAN-connected server "fredserver" using the facility LOG_USER and datagram packets.
+```
+
+* `http` can be used to log to a remote web server. The server must be
+  properly configured to receive and act on log messages. The option
+  accepts the URL to the web server, and a `method` argument
+  indicating whether the message should be submitted using the GET or
+  POST method.
+
+```
+ http=http://my.server/path/to/logger,method=POST
+```
+
+The `log_format` option provides several alternative formats:
+
+* `color`    - default format providing time, date and a message, using text colors to distinguish different log severities
+* `plain`    - same as above, but monochrome text only
+* `syslog`   - the log level and error message only, allowing the syslog system to attach the time and date
+* `legacy`   - a format similar to the one used by the legacy 2.3 InvokeAI releases.

docs/features/CONTROLNET.md

@@ -8,20 +8,64 @@ title: ControlNet
 
 ControlNet
 
-ControlNet is a powerful set of features developed by the open-source community (notably, Stanford researcher [**@ilyasviel**](https://github.com/lllyasviel)) that allows you to apply a secondary neural network model to your image generation process in Invoke.
+ControlNet is a powerful set of features developed by the open-source
+community (notably, Stanford researcher
+[**@ilyasviel**](https://github.com/lllyasviel)) that allows you to
+apply a secondary neural network model to your image generation
+process in Invoke.
 
-With ControlNet, you can get more control over the output of your image generation, providing you with a way to direct the network towards generating images that better fit your desired style or outcome.
+With ControlNet, you can get more control over the output of your
+image generation, providing you with a way to direct the network
+towards generating images that better fit your desired style or
+outcome.
 
 
 ### How it works
 
-ControlNet works by analyzing an input image, pre-processing that image to identify relevant information that can be interpreted by each specific ControlNet model, and then inserting that control information into the generation process. This can be used to adjust the style, composition, or other aspects of the image to better achieve a specific result.
+ControlNet works by analyzing an input image, pre-processing that
+image to identify relevant information that can be interpreted by each
+specific ControlNet model, and then inserting that control information
+into the generation process. This can be used to adjust the style,
+composition, or other aspects of the image to better achieve a
+specific result.
 
 
 ### Models
 
-As part of the model installation, ControlNet models can be selected including a variety of pre-trained models that have been added to achieve different effects or styles in your generated images. Further ControlNet models may require additional code functionality to also be incorporated into Invoke's Invocations folder. You should expect to follow any installation instructions for ControlNet models loaded  outside the default models provided by Invoke. The default models include:
+InvokeAI provides access to a series of ControlNet models that provide
+different effects or styles in your generated images.  Currently
+InvokeAI only supports "diffuser" style ControlNet models. These are
+folders that contain the files `config.json` and/or
+`diffusion_pytorch_model.safetensors` and
+`diffusion_pytorch_model.fp16.safetensors`. The name of the folder is
+the name of the model.
 
+***InvokeAI does not currently support checkpoint-format
+ControlNets. These come in the form of a single file with the
+extension `.safetensors`.***
+
+Diffuser-style ControlNet models are available at HuggingFace
+(http://huggingface.co) and accessed via their repo IDs (identifiers
+in the format "author/modelname"). The easiest way to install them is
+to use the InvokeAI model installer application. Use the
+`invoke.sh`/`invoke.bat` launcher to select item [5] and then navigate
+to the CONTROLNETS section. Select the models you wish to install and
+press "APPLY CHANGES". You may also enter additional HuggingFace
+repo_ids in the "Additional models" textbox:
+
+![Model Installer -
+Controlnetl](../assets/installing-models/model-installer-controlnet.png){:width="640px"}
+
+Command-line users can launch the model installer using the command
+`invokeai-model-install`.
+
+_Be aware that some ControlNet models require additional code
+functionality in order to work properly, so just installing a
+third-party ControlNet model may not have the desired effect._ Please
+read and follow the documentation for installing a third party model
+not currently included among InvokeAI's default list.
+
+The models currently supported include:
 
 **Canny**:
 

docs/features/NODES.md

@@ -0,0 +1,206 @@
+# Nodes Editor (Experimental)
+
+🚨
+*The node editor is experimental. We've made it accessible because we use it to develop the application, but we have not addressed the many known rough edges. It's very easy to shoot yourself in the foot, and we cannot offer support for it until it sees full release (ETA v3.1). Everything is subject to change without warning.* 
+🚨
+
+The nodes editor is a blank canvas allowing for the use of individual functions and image transformations to control the image generation workflow. The node processing flow is usually done from left (inputs) to right (outputs), though linearity can become abstracted the more complex the node graph becomes. Nodes inputs and outputs are connected by dragging connectors from node to node.
+
+To better understand how nodes are used, think of how an electric power bar works. It takes in one input (electricity from a wall outlet) and passes it to multiple devices through multiple outputs. Similarly, a node could have multiple inputs and outputs functioning at the same (or different) time, but all node outputs pass information onward like a power bar passes electricity. Not all outputs are compatible with all inputs, however - Each node has different constraints on how it is expecting to input/output information. In general, node outputs are colour-coded to match compatible inputs of other nodes.
+
+## Anatomy of a Node
+
+Individual nodes are made up of the following:
+
+- Inputs: Edge points on the left side of the node window where you connect outputs from other nodes.
+- Outputs: Edge points on the right side of the node window where you connect to inputs on other nodes.
+- Options: Various options which are either manually configured, or overridden by connecting an output from another node to the input.
+
+## Diffusion Overview
+
+Taking the time to understand the diffusion process will help you to understand how to set up your nodes in the nodes editor. 
+
+There are two main spaces Stable Diffusion works in: image space and latent space.
+
+Image space represents images in pixel form that you look at. Latent space represents compressed inputs. It’s in latent space that Stable Diffusion processes images. A VAE (Variational Auto Encoder) is responsible for compressing and encoding inputs into latent space, as well as decoding outputs back into image space.
+
+When you generate an image using text-to-image, multiple steps occur in latent space:
+1. Random noise is generated at the chosen height and width. The noise’s characteristics are dictated by the chosen (or not chosen) seed. This noise tensor is passed into latent space. We’ll call this noise A.
+1. Using a model’s U-Net, a noise predictor examines noise A, and the words tokenized by CLIP from your prompt (conditioning). It generates its own noise tensor to predict what the final image might look like in latent space. We’ll call this noise B.
+1. Noise B is subtracted from noise A in an attempt to create a final latent image indicative of the inputs. This step is repeated for the number of sampler steps chosen.
+1. The VAE decodes the final latent image from latent space into image space.
+
+image-to-image is a similar process, with only step 1 being different:
+1. The input image is decoded from image space into latent space by the VAE. Noise is then added to the input latent image. Denoising Strength dictates how much noise is added, 0 being none, and 1 being all-encompassing. We’ll call this noise A. The process is then the same as steps 2-4 in the text-to-image explanation above. 
+
+Furthermore, a model provides the CLIP prompt tokenizer, the VAE, and a U-Net (where noise prediction occurs given a prompt and initial noise tensor).
+
+A noise scheduler (eg. DPM++ 2M Karras) schedules the subtraction of noise from the latent image across the sampler steps chosen (step 3 above). Less noise is usually subtracted at higher sampler steps. 
+
+## Node Types (Base Nodes)
+
+| Node <img width=160 align="right"> | Function                                                                              |
+| ---------------------------------- | --------------------------------------------------------------------------------------|
+| Add                                | Adds two numbers |
+| CannyImageProcessor                | Canny edge detection for ControlNet |
+| ClipSkip                           | Skip layers in clip text_encoder model |
+| Collect                            | Collects values into a collection |
+| Prompt (Compel)                    | Parse prompt using compel package to conditioning |
+| ContentShuffleImageProcessor       | Applies content shuffle processing to image |
+| ControlNet                         | Collects ControlNet info to pass to other nodes |
+| CvInpaint                          | Simple inpaint using opencv |
+| Divide                             | Divides two numbers |
+| DynamicPrompt                      | Parses a prompt using adieyal/dynamic prompt's random or combinatorial generator |
+| FloatLinearRange                   | Creates a range |
+| HedImageProcessor                  | Applies HED edge detection to image |
+| ImageBlur                          | Blurs an image |
+| ImageChannel                       | Gets a channel from an image |
+| ImageCollection                    | Load a collection of images and provide it as output |
+| ImageConvert                       | Converts an image to a different mode |
+| ImageCrop                          | Crops an image to a specified box. The box can be outside of the image. |
+| ImageInverseLerp                   | Inverse linear interpolation of all pixels of an image |
+| ImageLerp                          | Linear interpolation of all pixels of an image |
+| ImageMultiply                      | Multiplies two images together using `PIL.ImageChops.Multiply()` |
+| ImagePaste                         | Pastes an image into another image |
+| ImageProcessor                     | Base class for invocations that reprocess images for ControlNet |
+| ImageResize                        | Resizes an image to specific dimensions |
+| ImageScale                         | Scales an image by a factor |
+| ImageToLatents                     | Scales latents by a given factor |
+| InfillColor                        | Infills transparent areas of an image with a solid color |
+| InfillPatchMatch                   | Infills transparent areas of an image using the PatchMatch algorithm |
+| InfillTile                         | Infills transparent areas of an image with tiles of the image |
+| Inpaint                            | Generates an image using inpaint |
+| Iterate                            | Iterates over a list of items |
+| LatentsToImage                     | Generates an image from latents |
+| LatentsToLatents                   | Generates latents using latents as base image |
+| LeresImageProcessor                | Applies leres processing to image |
+| LineartAnimeImageProcessor         | Applies line art anime processing to image |
+| LineartImageProcessor              | Applies line art processing to image |
+| LoadImage                          | Load an image and provide it as output |
+| Lora Loader                        | Apply selected lora to unet and text_encoder |
+| Model Loader                       | Loads a main model, outputting its submodels |
+| MaskFromAlpha                      | Extracts the alpha channel of an image as a mask |
+| MediapipeFaceProcessor             | Applies mediapipe face processing to image |
+| MidasDepthImageProcessor           | Applies Midas depth processing to image |
+| MlsdImageProcessor                 | Applied MLSD processing to image |
+| Multiply                           | Multiplies two numbers |
+| Noise                              | Generates latent noise |
+| NormalbaeImageProcessor            | Applies NormalBAE processing to image |
+| OpenposeImageProcessor             | Applies Openpose processing to image |
+| ParamFloat                         | A float parameter |
+| ParamInt                           | An integer parameter |
+| PidiImageProcessor                 | Applies PIDI processing to an image |
+| Progress Image                     | Displays the progress image in the Node Editor |
+| RandomInit                         | Outputs a single random integer |
+| RandomRange                        | Creates a collection of random numbers |
+| Range                              | Creates a range of numbers from start to stop with step |
+| RangeOfSize                        | Creates a range from start to start + size with step |
+| ResizeLatents                      | Resizes latents to explicit width/height (in pixels). Provided dimensions are floor-divided by 8. |
+| RestoreFace                        | Restores faces in the image |
+| ScaleLatents                       | Scales latents by a given factor |
+| SegmentAnythingProcessor           | Applies segment anything processing to image |
+| ShowImage                          | Displays a provided image, and passes it forward in the pipeline |
+| StepParamEasing                    | Experimental per-step parameter for easing for denoising steps |
+| Subtract                           | Subtracts two numbers |
+| TextToLatents                      | Generates latents from conditionings |
+| TileResampleProcessor              | Bass class for invocations that preprocess images for ControlNet |
+| Upscale                            | Upscales an image |
+| VAE Loader                         | Loads a VAE model, outputting a VaeLoaderOutput |
+| ZoeDepthImageProcessor             | Applies Zoe depth processing to image |
+
+## Node Grouping Concepts
+
+There are several node grouping concepts that can be examined with a narrow focus. These (and other) groupings can be pieced together to make up functional graph setups, and are important to understanding how groups of nodes work together as part of a whole. Note that the screenshots below aren't examples of complete functioning node graphs (see Examples).
+
+### Noise
+
+As described, an initial noise tensor is necessary for the latent diffusion process. As a result, all non-image *ToLatents nodes require a noise node input.  
+
+<img width="654" alt="groupsnoise" src="https://github.com/ymgenesis/InvokeAI/assets/25252829/2e8d297e-ad55-4d27-bc93-c119dad2a2c5">
+
+### Conditioning
+
+As described, conditioning is necessary for the latent diffusion process, whether empty or not. As a result, all non-image *ToLatents nodes require positive and negative conditioning inputs. Conditioning is reliant on a CLIP tokenizer provided by the Model Loader node.
+
+<img width="1024" alt="groupsconditioning" src="https://github.com/ymgenesis/InvokeAI/assets/25252829/f8f7ad8a-8d9c-418e-b5ad-1437b774b27e">
+
+### Image Space & VAE
+
+The ImageToLatents node doesn't require a noise node input, but requires a VAE input to convert the image from image space into latent space. In reverse, the LatentsToImage node requires a VAE input to convert from latent space back into image space.
+
+<img width="637" alt="groupsimgvae" src="https://github.com/ymgenesis/InvokeAI/assets/25252829/dd99969c-e0a8-4f78-9b17-3ffe179cef9a">
+
+### Defined & Random Seeds
+
+It is common to want to use both the same seed (for continuity) and random seeds (for variance). To define a seed, simply enter it into the 'Seed' field on a noise node. Conversely, the RandomInt node generates a random integer between 'Low' and 'High', and can be used as input to the 'Seed' edge point on a noise node to randomize your seed.
+
+<img width="922" alt="groupsrandseed" src="https://github.com/ymgenesis/InvokeAI/assets/25252829/af55bc20-60f6-438e-aba5-3ec871443710">
+
+### Control
+
+Control means to guide the diffusion process to adhere to a defined input or structure. Control can be provided as input to non-image *ToLatents nodes from ControlNet nodes. ControlNet nodes usually require an image processor which converts an input image for use with ControlNet.
+
+<img width="805" alt="groupscontrol" src="https://github.com/ymgenesis/InvokeAI/assets/25252829/cc9c5de7-23a7-46c8-bbad-1f3609d999a6">
+
+### LoRA
+
+The Lora Loader node lets you load a LoRA (say that ten times fast) and pass it as output to both the Prompt (Compel) and non-image *ToLatents nodes. A model's CLIP tokenizer is passed through the LoRA into Prompt (Compel), where it affects conditioning. A model's U-Net is also passed through the LoRA into a non-image *ToLatents node, where it affects noise prediction.
+
+<img width="993" alt="groupslora" src="https://github.com/ymgenesis/InvokeAI/assets/25252829/630962b0-d914-4505-b3ea-ccae9b0269da">
+
+### Scaling
+
+Use the ImageScale, ScaleLatents, and Upscale nodes to upscale images and/or latent images. The chosen method differs across contexts. However, be aware that latents are already noisy and compressed at their original resolution; scaling an image could produce more detailed results.
+
+<img width="644" alt="groupsallscale" src="https://github.com/ymgenesis/InvokeAI/assets/25252829/99314f05-dd9f-4b6d-b378-31de55346a13">
+
+### Iteration + Multiple Images as Input
+
+Iteration is a common concept in any processing, and means to repeat a process with given input. In nodes, you're able to use the Iterate node to iterate through collections usually gathered by the Collect node. The Iterate node has many potential uses, from processing a collection of images one after another, to varying seeds across multiple image generations and more. This screenshot demonstrates how to collect several images and pass them out one at a time.
+
+<img width="788" alt="groupsiterate" src="https://github.com/ymgenesis/InvokeAI/assets/25252829/4af5ca27-82c9-4018-8c5b-024d3ee0a121">
+
+### Multiple Image Generation + Random Seeds
+
+Multiple image generation in the node editor is done using the RandomRange node. In this case, the 'Size' field represents the number of images to generate. As RandomRange produces a collection of integers, we need to add the Iterate node to iterate through the collection. 
+
+To control seeds across generations takes some care. The first row in the screenshot will generate multiple images with different seeds, but using the same RandomRange parameters across invocations will result in the same group of random seeds being used across the images, producing repeatable results. In the second row, adding the RandomInt node as input to RandomRange's 'Seed' edge point will ensure that seeds are varied across all images across invocations, producing varied results.
+
+<img width="1027" alt="groupsmultigenseeding" src="https://github.com/ymgenesis/InvokeAI/assets/25252829/518d1b2b-fed1-416b-a052-ab06552521b3">
+
+## Examples
+
+With our knowledge of node grouping and the diffusion process, let’s break down some basic graphs in the nodes editor. Note that a node's options can be overridden by inputs from other nodes. These examples aren't strict rules to follow and only demonstrate some basic configurations.
+
+### Basic text-to-image Node Graph
+
+<img width="875" alt="nodest2i" src="https://github.com/ymgenesis/InvokeAI/assets/25252829/17c67720-c376-4db8-94f0-5e00381a61ee">
+
+- Model Loader: A necessity to generating images (as we’ve read above). We choose our model from the dropdown. It outputs a U-Net, CLIP tokenizer, and VAE.
+- Prompt (Compel): Another necessity. Two prompt nodes are created. One will output positive conditioning (what you want, ‘dog’), one will output negative (what you don’t want, ‘cat’). They both input the CLIP tokenizer that the Model Loader node outputs.
+- Noise: Consider this noise A from step one of the text-to-image explanation above. Choose a seed number, width, and height.
+- TextToLatents: This node takes many inputs for converting and processing text & noise from image space into latent space, hence the name TextTo**Latents**. In this setup, it inputs positive and negative conditioning from the prompt nodes for processing (step 2 above). It inputs noise from the noise node for processing (steps 2 & 3 above). Lastly, it inputs a U-Net from the Model Loader node for processing (step 2 above). It outputs latents for use in the next LatentsToImage node. Choose number of sampler steps, CFG scale, and scheduler.
+- LatentsToImage: This node takes in processed latents from the TextToLatents node, and the model’s VAE from the Model Loader node which is responsible for decoding latents back into the image space, hence the name LatentsTo**Image**. This node is the last stop, and once the image is decoded, it is saved to the gallery.
+
+### Basic image-to-image Node Graph
+
+<img width="998" alt="nodesi2i" src="https://github.com/ymgenesis/InvokeAI/assets/25252829/3f2c95d5-cee7-4415-9b79-b46ee60a92fe">
+
+- Model Loader: Choose a model from the dropdown.
+- Prompt (Compel): Two prompt nodes. One positive (dog), one negative (dog). Same CLIP inputs from the Model Loader node as before.
+- ImageToLatents: Upload a source image directly in the node window, via drag'n'drop from the gallery, or passed in as input. The ImageToLatents node inputs the VAE from the Model Loader node to decode the chosen image from image space into latent space, hence the name ImageTo**Latents**. It outputs latents for use in the next LatentsToLatents node. It also outputs the source image's width and height for use in the next Noise node if the final image is to be the same dimensions as the source image.
+- Noise: A noise tensor is created with the width and height of the source image, and connected to the next LatentsToLatents node. Notice the width and height fields are overridden by the input from the ImageToLatents width and height outputs.
+- LatentsToLatents: The inputs and options are nearly identical to TextToLatents, except that LatentsToLatents also takes latents as an input. Considering our source image is already converted to latents in the last ImageToLatents node, and text + noise are no longer the only inputs to process, we use the LatentsToLatents node. 
+- LatentsToImage: Like previously, the LatentsToImage node will use the VAE from the Model Loader as input to decode the latents from LatentsToLatents into image space, and save it to the gallery.
+
+### Basic ControlNet Node Graph
+
+<img width="703" alt="nodescontrol" src="https://github.com/ymgenesis/InvokeAI/assets/25252829/b02ded86-ceb4-44a2-9910-e19ad184d471">
+
+- Model Loader
+- Prompt (Compel)
+- Noise: Width and height of the CannyImageProcessor ControlNet image is passed in to set the dimensions of the noise passed to TextToLatents.
+- CannyImageProcessor: The CannyImageProcessor node is used to process the source image being used as a ControlNet. Each ControlNet processor node applies control in different ways, and has some different options to configure. Width and height are passed to noise, as mentioned. The processed ControlNet image is output to the ControlNet node.
+- ControlNet: Select the type of control model. In this case, canny is chosen as the CannyImageProcessor was used to generate the ControlNet image. Configure the control node options, and pass the control output to TextToLatents.
+- TextToLatents: Similar to the basic text-to-image example, except ControlNet is passed to the control input edge point.
+- LatentsToImage

docs/features/WEB.md

@@ -4,15 +4,19 @@ title: InvokeAI Web Server
 
 # :material-web: InvokeAI Web Server
 
-As of version 2.0.0, this distribution comes with a full-featured web server
-(see screenshot).
+## Quick guided walkthrough of the WebUI's features
 
-To use it, launch the `invoke.sh`/`invoke.bat` script and select
-option (2). Alternatively, with the InvokeAI environment active, run
-the `invokeai` script by adding the `--web` option:
+While most of the WebUI's features are intuitive, here is a guided walkthrough
+through its various components.
+
+### Launching the WebUI
+
+To run the InvokeAI web server, start the `invoke.sh`/`invoke.bat`
+script and select option (1). Alternatively, with the InvokeAI
+environment active, run `invokeai-web`:
 
 ```bash
-invokeai --web
+invokeai-web
 ```
 
 You can then connect to the server by pointing your web browser at
@@ -28,33 +32,32 @@ invoke.sh --host 0.0.0.0
 or
 
 ```bash
-invokeai --web --host 0.0.0.0
+invokeai-web --host 0.0.0.0
 ```
 
-## Quick guided walkthrough of the WebUI's features
-
-While most of the WebUI's features are intuitive, here is a guided walkthrough
-through its various components.
+### The InvokeAI Web Interface
 
 ![Invoke Web Server - Major Components](../assets/invoke-web-server-1.png){:width="640px"}
 
 The screenshot above shows the Text to Image tab of the WebUI. There are three
 main sections:
 
-1. A **control panel** on the left, which contains various settings for text to
-   image generation. The most important part is the text field (currently
-   showing `strawberry sushi`) for entering the text prompt, and the camera icon
-   directly underneath that will render the image. We'll call this the _Invoke_
-   button from now on.
+1. A **control panel** on the left, which contains various settings
+   for text to image generation. The most important part is the text
+   field (currently showing `fantasy painting, horned demon`) for
+   entering the positive text prompt, another text field right below it for an
+   optional negative text prompt (concepts to exclude), and a _Invoke_ button 
+   to begin the image rendering process.
 
-2. The **current image** section in the middle, which shows a large format
-   version of the image you are currently working on. A series of buttons at the
-   top ("image to image", "Use All", "Use Seed", etc) lets you modify the image
-   in various ways.
+2. The **current image** section in the middle, which shows a large
+   format version of the image you are currently working on. A series
+   of buttons at the top lets you modify and manipulate the image in
+   various ways.
 
-3. A \*_gallery_ section on the left that contains a history of the images you
+3. A **gallery** section on the left that contains a history of the images you
    have generated. These images are read and written to the directory specified
-   at launch time in `--outdir`.
+   in the `INVOKEAIROOT/invokeai.yaml` initialization file, usually a directory
+   named `outputs` in `INVOKEAIROOT`.
 
 In addition to these three elements, there are a series of icons for changing
 global settings, reporting bugs, and changing the theme on the upper right.
@@ -76,14 +79,10 @@ From top to bottom, these are:
    with outpainting,and modify interior portions of the image with
    inpainting, erase portions of a starting image and have the AI fill in
    the erased region from a text prompt.
-4. Workflow Management (not yet implemented) - this panel will allow you to create
+4. Node Editor - (experimental) this panel allows you to create
    pipelines of common operations and combine them into workflows.
-5. Training (not yet implemented) - this panel will provide an interface to [textual
-   inversion training](TEXTUAL_INVERSION.md) and fine tuning.
-
-The inpainting, outpainting and postprocessing tabs are currently in
-development. However, limited versions of their features can already be accessed
-through the Text to Image and Image to Image tabs.
+5. Model Manager - this panel allows you to import and configure new
+   models using URLs, local paths, or HuggingFace diffusers repo_ids.
 
 ## Walkthrough
 
@@ -92,43 +91,54 @@ feature set.
 
 ### Text to Image
 
-1. Launch the WebUI using `python scripts/invoke.py --web` and connect to it
-   with your browser by accessing `http://localhost:9090`. If the browser and
-   server are running on different machines on your LAN, add the option
-   `--host 0.0.0.0` to the launch command line and connect to the machine
-   hosting the web server using its IP address or domain name.
+1. Launch the WebUI using launcher option [1] and connect to it with
+   your browser by accessing `http://localhost:9090`. If the browser
+   and server are running on different machines on your LAN, add the
+   option `--host 0.0.0.0` to the `invoke.sh` launch command line and connect to
+   the machine hosting the web server using its IP address or domain
+   name.
 
-2. If all goes well, the WebUI should come up and you'll see a green
-   `connected` message on the upper right.
+2. If all goes well, the WebUI should come up and you'll see a green dot
+   meaning `connected`  on the upper right.
+
+![Invoke Web Server - Control Panel](../assets/invoke-control-panel-1.png){ align=right width=300px }
 
 #### Basics
 
-1.  Generate an image by typing _strawberry sushi_ into the large prompt field
-    on the upper left and then clicking on the Invoke button (the one with the
-    Camera icon). After a short wait, you'll see a large image of sushi in the
+1.  Generate an image by typing _bluebird_ into the large prompt field
+    on the upper left and then clicking on the Invoke button or pressing
+	the return button.
+	After a short wait, you'll see a large image of a bluebird in the
     image panel, and a new thumbnail in the gallery on the right.
 
-    If you need more room on the screen, you can turn the gallery off by
-    clicking on the **x** to the right of "Your Invocations". You can turn it
-    back on later by clicking the image icon that appears in the gallery's
-    place.
+    If you need more room on the screen, you can turn the gallery off
+    by typing the **g** hotkey. You can turn it back on later by clicking the
+    image icon that appears in the gallery's place. The list of hotkeys can
+	be found by clicking on the keyboard icon above the image gallery.
 
-    The images are written into the directory indicated by the `--outdir` option
-    provided at script launch time. By default, this is `outputs/img-samples`
-    under the InvokeAI directory.
-
-2.  Generate a bunch of strawberry sushi images by increasing the number of
-    requested images by adjusting the Images counter just below the Camera
+2.  Generate a bunch of bluebird images by increasing the number of
+    requested images by adjusting the Images counter just below the Invoke
     button. As each is generated, it will be added to the gallery. You can
     switch the active image by clicking on the gallery thumbnails.
+	
+	If you'd like to watch the image generation progress, click the hourglass
+	icon above the main image area. As generation progresses, you'll see
+	increasingly detailed versions of the ultimate image.
 
-3.  Try playing with different settings, including image width and height, the
-    Sampler, the Steps and the CFG scale.
+3.  Try playing with different settings, including changing the main
+    model, the image width and height, the Scheduler, the Steps and
+    the CFG scale.
+	
+	The _Model_ changes the main model. Thousands of custom models are
+	now available, which generate a variety of image styles and
+	subjects. While InvokeAI comes with a few starter models, it is
+	easy to import new models into the application. See [Installing
+	Models](../installation/050_INSTALLING_MODELS.md) for more details.
 
     Image _Width_ and _Height_ do what you'd expect. However, be aware that
     larger images consume more VRAM memory and take longer to generate.
 
-    The _Sampler_ controls how the AI selects the image to display. Some
+    The _Scheduler_ controls how the AI selects the image to display. Some
     samplers are more "creative" than others and will produce a wider range of
     variations (see next section). Some samplers run faster than others.
 
@@ -142,17 +152,27 @@ feature set.
     to the input prompt. You can go as high or low as you like, but generally
     values greater than 20 won't improve things much, and values lower than 5
     will produce unexpected images. There are complex interactions between
-    _Steps_, _CFG Scale_ and the _Sampler_, so experiment to find out what works
+    _Steps_, _CFG Scale_ and the _Scheduler_, so experiment to find out what works
     for you.
+	
+	The _Seed_ controls the series of values returned by InvokeAI's
+    random number generator. Each unique seed value will generate a different
+	image. To regenerate a previous image, simply use the original image's
+	seed value. A slider to the right of the _Seed_ field will change the
+	seed each time an image is generated.
 
-4.  To regenerate a previously-generated image, select the image you want and
-    click _Use All_. This loads the text prompt and other original settings into
-    the control panel. If you then press _Invoke_ it will regenerate the image
-    exactly. You can also selectively modify the prompt or other settings to
-    tweak the image.
+![Invoke Web Server - Control Panel 2](../assets/control-panel-2.png){ align=right width=400px }
 
-    Alternatively, you may click on _Use Seed_ to load just the image's seed,
-    and leave other settings unchanged.
+4.  To regenerate a previously-generated image, select the image you
+    want and click the asterisk ("*") button at the top of the
+    image. This loads the text prompt and other original settings into
+    the control panel. If you then press _Invoke_ it will regenerate
+    the image exactly. You can also selectively modify the prompt or
+    other settings to tweak the image.
+
+    Alternatively, you may click on the "sprouting plant icon" to load
+    just the image's seed, and leave other settings unchanged or the
+    quote icon to load just the positive and negative prompts.
 
 5.  To regenerate a Stable Diffusion image that was generated by another SD
     package, you need to know its text prompt and its _Seed_. Copy-paste the
@@ -161,62 +181,22 @@ feature set.
     you Invoke, you will get something similar to the original image. It will
     not be exact unless you also set the correct values for the original
     sampler, CFG, steps and dimensions, but it will (usually) be close.
+	
+6.  To save an image, right click on it to bring up a menu that will
+	let you download the image, save it to a named image gallery, and
+	copy it to the clipboard, among other things.
 
-#### Variations on a theme
+#### Upscaling
 
-1.  Let's try generating some variations. Select your favorite sushi image from
-    the gallery to load it. Then select "Use All" from the list of buttons
-    above. This will load up all the settings used to generate this image,
-    including its unique seed.
+![Invoke Web Server - Upscaling](../assets/upscaling.png){ align=right width=400px }
 
-    Go down to the Variations section of the Control Panel and set the button to
-    On. Set Variation Amount to 0.2 to generate a modest number of variations on
-    the image, and also set the Image counter to `4`. Press the `invoke` button.
-    This will generate a series of related images. To obtain smaller variations,
-    just lower the Variation Amount. You may also experiment with changing the
-    Sampler. Some samplers generate more variability than others. _k_euler_a_ is
-    particularly creative, while _ddim_ is pretty conservative.
-
-2.  For even more variations, experiment with increasing the setting for
-    _Perlin_. This adds a bit of noise to the image generation process. Note
-    that values of Perlin noise greater than 0.15 produce poor images for
-    several of the samplers.
-
-#### Facial reconstruction and upscaling
-
-Stable Diffusion frequently produces mangled faces, particularly when there are
-multiple figures in the same scene. Stable Diffusion has particular issues with
-generating reallistic eyes. InvokeAI provides the ability to reconstruct faces
-using either the GFPGAN or CodeFormer libraries. For more information see
-[POSTPROCESS](POSTPROCESS.md).
-
-1.  Invoke a prompt that generates a mangled face. A prompt that often gives
-    this is "portrait of a lawyer, 3/4 shot" (this is not intended as a slur
-    against lawyers!) Once you have an image that needs some touching up, load
-    it into the Image panel, and press the button with the face icon
-    (highlighted in the first screenshot below). A dialog box will appear. Leave
-    _Strength_ at 0.8 and press \*Restore Faces". If all goes well, the eyes and
-    other aspects of the face will be improved (see the second screenshot)
-
-    ![Invoke Web Server - Original Image](../assets/invoke-web-server-3.png)
-
-    ![Invoke Web Server - Retouched Image](../assets/invoke-web-server-4.png)
-
-    The facial reconstruction _Strength_ field adjusts how aggressively the face
-    library will try to alter the face. It can be as high as 1.0, but be aware
-    that this often softens the face airbrush style, losing some details. The
-    default 0.8 is usually sufficient.
-
-2.  "Upscaling" is the process of increasing the size of an image while
-    retaining the sharpness. InvokeAI uses an external library called "ESRGAN"
-    to do this. To invoke upscaling, simply select an image and press the _HD_
-    button above it. You can select between 2X and 4X upscaling, and adjust the
-    upscaling strength, which has much the same meaning as in facial
-    reconstruction. Try running this on one of your previously-generated images.
-
-3.  Finally, you can run facial reconstruction and/or upscaling automatically
-    after each Invocation. Go to the Advanced Options section of the Control
-    Panel and turn on _Restore Face_ and/or _Upscale_.
+"Upscaling" is the process of increasing the size of an image while
+    retaining the sharpness. InvokeAI uses an external library called
+    "ESRGAN" to do this. To invoke upscaling, simply select an image
+    and press the "expanding arrows" button above it. You can select
+    between 2X and 4X upscaling, and adjust the upscaling strength,
+    which has much the same meaning as in facial reconstruction. Try
+    running this on one of your previously-generated images.
 
 ### Image to Image
 
@@ -224,24 +204,14 @@ InvokeAI lets you take an existing image and use it as the basis for a new
 creation. You can use any sort of image, including a photograph, a scanned
 sketch, or a digital drawing, as long as it is in PNG or JPEG format.
 
-For this tutorial, we'll use files named
-[Lincoln-and-Parrot-512.png](../assets/Lincoln-and-Parrot-512.png), and
-[Lincoln-and-Parrot-512-transparent.png](../assets/Lincoln-and-Parrot-512-transparent.png).
-Download these images to your local machine now to continue with the
-walkthrough.
+For this tutorial, we'll use the file named
+[Lincoln-and-Parrot-512.png](../assets/Lincoln-and-Parrot-512.png).
 
-1.  Click on the _Image to Image_ tab icon, which is the second icon from the
-    top on the left-hand side of the screen:
+1.  Click on the _Image to Image_ tab icon, which is the second icon
+    from the top on the left-hand side of the screen. This will bring
+    you to a screen similar to the one shown here:
 
-    <figure markdown>
-    ![Invoke Web Server - Image to Image Icon](../assets/invoke-web-server-5.png)
-    </figure>
-
-    This will bring you to a screen similar to the one shown here:
-
-    <figure markdown>
-    ![Invoke Web Server - Image to Image Tab](../assets/invoke-web-server-6.png){:width="640px"}
-    </figure>
+    ![Invoke Web Server - Image to Image Tab](../assets/invoke-web-server-6.png){ width="640px" }
 
 2.  Drag-and-drop the Lincoln-and-Parrot image into the Image panel, or click
     the blank area to get an upload dialog. The image will load into an area
@@ -255,120 +225,99 @@ walkthrough.
     ![Invoke Web Server - Image to Image example](../assets/invoke-web-server-7.png){:width="640px"}
 
 4.  Experiment with the different settings. The most influential one in Image to
-    Image is _Image to Image Strength_ located about midway down the control
+    Image is _Denoising Strength_ located about midway down the control
     panel. By default it is set to 0.75, but can range from 0.0 to 0.99. The
     higher the value, the more of the original image the AI will replace. A
     value of 0 will leave the initial image completely unchanged, while 0.99
-    will replace it completely. However, the Sampler and CFG Scale also
+    will replace it completely. However, the _Scheduler_ and _CFG Scale_ also
     influence the final result. You can also generate variations in the same way
     as described in Text to Image.
 
-5.  What if we only want to change certain part(s) of the image and leave the
-    rest intact? This is called Inpainting, and a future version of the InvokeAI
-    web server will provide an interactive painting canvas on which you can
-    directly draw the areas you wish to Inpaint into. For now, you can achieve
-    this effect by using an external photoeditor tool to make one or more
-    regions of the image transparent as described in [INPAINTING.md] and
-    uploading that.
-
-    The file
-    [Lincoln-and-Parrot-512-transparent.png](../assets/Lincoln-and-Parrot-512-transparent.png)
-    is a version of the earlier image in which the area around the parrot has
-    been replaced with transparency. Click on the "x" in the upper right of the
-    Initial Image and upload the transparent version. Using the same prompt "old
-    sea captain with raven on shoulder" try Invoking an image. This time, only
-    the parrot will be replaced, leaving the rest of the original image intact:
-
-    <figure markdown>
-    ![Invoke Web Server - Inpainting](../assets/invoke-web-server-8.png){:width="640px"}
-    </figure>
+5.  What if we only want to change certain part(s) of the image and
+    leave the rest intact? This is called Inpainting, and you can do
+    it in the [Unified Canvas](UNIFIED_CANVAS.md). The Unified Canvas
+    also allows you to extend borders of the image and fill in the
+    blank areas, a process called outpainting.
 
 6.  Would you like to modify a previously-generated image using the Image to
-    Image facility? Easy! While in the Image to Image panel, hover over any of
-    the gallery images to see a little menu of icons pop up. Click the picture
-    icon to instantly send the selected image to Image to Image as the initial
-    image.
+    Image facility? Easy! While in the Image to Image panel, drag and drop any
+	image in the gallery into the Initial Image area, and it will be ready for
+	use. You can do the same thing with the main image display. Click on the
+	_Send to_ icon to get a menu of
+	commands and choose "Send to Image to Image".
+	
+	![Send To Icon](../assets/send-to-icon.png) 
 
-You can do the same from the Text to Image tab by clicking on the picture icon
-above the central image panel. The screenshot below shows where the "use as
-initial image" icons are located.
+### Textual Inversion, LoRA and ControlNet
 
-![Invoke Web Server - Use as Image Links](../assets/invoke-web-server-9.png){:width="640px"}
+InvokeAI supports several different types of model files that
+extending the capabilities of the main model by adding artistic
+styles, special effects, or subjects. By mixing and matching textual
+inversion, LoRA and ControlNet models, you can achieve many
+interesting and beautiful effects.
 
-### Unified Canvas
+We will give an example using a LoRA model named "Ink Scenery". This
+LoRA, which can be downloaded from Civitai (civitai.com), is
+specialized to paint landscapes that look like they were made with
+dripping india ink. To install this LoRA, we first download it and 
+put it into the `autoimport/lora` folder located inside the
+`invokeai` root directory. After restarting the web server, the
+LoRA will now become available for use.
 
-See the [Unified Canvas Guide](UNIFIED_CANVAS.md)
+To see this LoRA at work, we'll first generate an image without it
+using the standard `stable-diffusion-v1-5` model. Choose this
+model and enter the prompt "mountains, ink". Here is a typical
+generated image, a mountain range rendered in ink and watercolor
+wash:
 
-## Reference
+![Ink Scenery without LoRA](../assets/lora-example-0.png){ width=512px }
 
-### Additional Options
+Now let's install and activate the Ink Scenery LoRA. Go to
+https://civitai.com/models/78605/ink-scenery-or and download the LoRA
+model file to `invokeai/autoimport/lora` and restart the web
+server. (Alternatively, you can use [InvokeAI's Web Model
+Manager](../installation/050_INSTALLING_MODELS.md) to download and
+install the LoRA directly by typing its URL into the _Import
+Models_->_Location_ field).
 
-| parameter <img width=160 align="right"> | effect                                                                                                                                     |
-| --------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------ |
-| `--web_develop`                         | Starts the web server in development mode.                                                                                                 |
-| `--web_verbose`                         | Enables verbose logging                                                                                                                    |
-| `--cors [CORS ...]`                     | Additional allowed origins, comma-separated                                                                                                |
-| `--host HOST`                           | Web server: Host or IP to listen on. Set to 0.0.0.0 to accept traffic from other devices on your network.                                  |
-| `--port PORT`                           | Web server: Port to listen on                                                                                                              |
-| `--certfile CERTFILE`                   | Web server: Path to certificate file to use for SSL. Use together with --keyfile                                                           |
-| `--keyfile KEYFILE`                     | Web server: Path to private key file to use for SSL. Use together with --certfile'                                                         |
-| `--gui`                                 | Start InvokeAI GUI - This is the "desktop mode" version of the web app. It uses Flask to create a desktop app experience of the webserver. |
+Scroll down the control panel until you get to the LoRA accordion
+section, and open it:
 
-### Web Specific Features
+![LoRA Section](../assets/lora-example-1.png){ width=512px }
 
-The web experience offers an incredibly easy-to-use experience for interacting
-with the InvokeAI toolkit. For detailed guidance on individual features, see the
-Feature-specific help documents available in this directory. Note that the
-latest functionality available in the CLI may not always be available in the Web
-interface.
+Click the popup menu and select "Ink scenery". (If it isn't there, then
+the model wasn't installed to the right place, or perhaps you forgot
+to restart the web server.) The LoRA section will change to look like this:
 
-#### Dark Mode & Light Mode
+![LoRA Section Loaded](../assets/lora-example-2.png){ width=512px }
 
-The InvokeAI interface is available in a nano-carbon black & purple Dark Mode,
-and a "burn your eyes out Nosferatu" Light Mode. These can be toggled by
-clicking the Sun/Moon icons at the top right of the interface.
+Note that there is now a slider control for _Ink scenery_. The slider
+controls how much influence the LoRA model will have on the generated
+image.
 
-![InvokeAI Web Server - Dark Mode](../assets/invoke_web_dark.png)
+Run the "mountains, ink" prompt again and observe the change in style:
 
-![InvokeAI Web Server - Light Mode](../assets/invoke_web_light.png)
+![Ink Scenery](../assets/lora-example-3.png){ width=512px }
 
-#### Invocation Toolbar
+Try adjusting the weight slider for larger and smaller weights and
+generate the image after each adjustment. The higher the weight, the
+more influence the LoRA will have.
 
-The left side of the InvokeAI interface is available for customizing the prompt
-and the settings used for invoking your new image. Typing your prompt into the
-open text field and clicking the Invoke button will produce the image based on
-the settings configured in the toolbar.
+To remove the LoRA completely, just click on its trash can icon.
 
-See below for additional documentation related to each feature:
+Multiple LoRAs can be added simultaneously and combined with textual
+inversions and ControlNet models. Please see [Textual Inversions and
+LoRAs](CONCEPTS.md) and [Using ControlNet](CONTROLNET.md) for details.
 
-- [Variations](./VARIATIONS.md)
-- [Upscaling](./POSTPROCESS.md#upscaling)
-- [Image to Image](./IMG2IMG.md)
-- [Other](./OTHER.md)
+## Summary
 
-#### Invocation Gallery
-
-The currently selected --outdir (or the default outputs folder) will display all
-previously generated files on load. As new invocations are generated, these will
-be dynamically added to the gallery, and can be previewed by selecting them.
-Each image also has a simple set of actions (e.g., Delete, Use Seed, Use All
-Parameters, etc.) that can be accessed by hovering over the image.
-
-#### Image Workspace
-
-When an image from the Invocation Gallery is selected, or is generated, the
-image will be displayed within the center of the interface. A quickbar of common
-image interactions are displayed along the top of the image, including:
-
-- Use image in the `Image to Image` workflow
-- Initialize Face Restoration on the selected file
-- Initialize Upscaling on the selected file
-- View File metadata and details
-- Delete the file
+This walkthrough just skims the surface of the many things InvokeAI
+can do. Please see [Features](index.md) for more detailed reference
+guides.
 
 ## Acknowledgements
 
-A huge shout-out to the core team working to make this vision a reality,
+A huge shout-out to the core team working to make the Web GUI a reality,
 including [psychedelicious](https://github.com/psychedelicious),
 [Kyle0654](https://github.com/Kyle0654) and
 [blessedcoolant](https://github.com/blessedcoolant).

docs/features/index.md

@@ -17,8 +17,12 @@ a single convenient digital artist-optimized user interface.
 ### * [Prompt Engineering](PROMPTS.md)
 Get the images you want with the InvokeAI  prompt engineering language.
 
-## * The [Concepts Library](CONCEPTS.md)
-Add custom subjects and styles using HuggingFace's repository of embeddings.
+### * The [LoRA, LyCORIS and Textual Inversion Models](CONCEPTS.md)
+Add custom subjects and styles using a variety of fine-tuned models.
+
+### * [ControlNet](CONTROLNET.md)
+Learn how to install and use ControlNet models for fine control over
+image output.
 
 ### * [Image-to-Image Guide](IMG2IMG.md)
 Use a seed image to build new creations in the CLI.
@@ -29,26 +33,28 @@ are the ticket.
 
 ## Model Management
 
-## * [Model Installation](../installation/050_INSTALLING_MODELS.md)
+### * [Model Installation](../installation/050_INSTALLING_MODELS.md)
 Learn how to import third-party models and switch among them. This
 guide also covers optimizing models to load quickly.
 
-## * [Merging Models](MODEL_MERGING.md)
+### * [Merging Models](MODEL_MERGING.md)
 Teach an old model new tricks. Merge 2-3 models together to create a
 new model that combines characteristics of the originals.
 
-## * [Textual Inversion](TEXTUAL_INVERSION.md)
+### * [Textual Inversion](TRAINING.md)
 Personalize models by adding your own style or subjects.
 
-# Other Features
+## Other Features
 
-## * [The NSFW Checker](NSFW.md)
+### * [The NSFW Checker](NSFW.md)
 Prevent InvokeAI from displaying unwanted racy images.
 
-## * [Controlling Logging](LOGGING.md)
+### * [Controlling Logging](LOGGING.md)
 Control how InvokeAI logs status messages.
 
-## * [Miscellaneous](OTHER.md)
+<!-- OUT OF DATE
+### * [Miscellaneous](OTHER.md)
 Run InvokeAI on Google Colab, generate images with repeating patterns,
 batch process a file of prompts, increase the "creativity" of image
 generation by adding initial noise, and more!
+-->

docs/index.md

@@ -24,7 +24,7 @@ title: Home
 
 [![CI checks on main badge]][ci checks on main link]
 [![CI checks on dev badge]][ci checks on dev link]
-[![latest commit to dev badge]][latest commit to dev link]
+<!-- [![latest commit to dev badge]][latest commit to dev link] -->
 
 [![github open issues badge]][github open issues link]
 [![github open prs badge]][github open prs link]
@@ -54,10 +54,10 @@ title: Home
 [github stars badge]:
   https://flat.badgen.net/github/stars/invoke-ai/InvokeAI?icon=github
 [github stars link]: https://github.com/invoke-ai/InvokeAI/stargazers
-[latest commit to dev badge]:
+<!-- [latest commit to dev badge]:
   https://flat.badgen.net/github/last-commit/invoke-ai/InvokeAI/development?icon=github&color=yellow&label=last%20dev%20commit&cache=900
 [latest commit to dev link]:
-  https://github.com/invoke-ai/InvokeAI/commits/development
+  https://github.com/invoke-ai/InvokeAI/commits/main -->
 [latest release badge]:
   https://flat.badgen.net/github/release/invoke-ai/InvokeAI/development?icon=github
 [latest release link]: https://github.com/invoke-ai/InvokeAI/releases
@@ -82,6 +82,25 @@ Q&A</a>]
 
     This fork is rapidly evolving. Please use the [Issues tab](https://github.com/invoke-ai/InvokeAI/issues) to report bugs and make feature requests. Be sure to use the provided templates. They will help aid diagnose issues faster.
 
+## :octicons-package-dependencies-24: Installation
+
+This fork is supported across Linux, Windows and Macintosh. Linux users can use
+either an Nvidia-based card (with CUDA support) or an AMD card (using the ROCm
+driver).
+
+### [Installation Getting Started Guide](installation)
+#### **[Automated Installer](installation/010_INSTALL_AUTOMATED.md)**
+✅ This is the recommended installation method for first-time users. 
+#### [Manual Installation](installation/020_INSTALL_MANUAL.md)
+This method is recommended for experienced users and developers
+#### [Docker Installation](installation/040_INSTALL_DOCKER.md)
+This method is recommended for those familiar with running Docker containers
+### Other Installation Guides
+  - [PyPatchMatch](installation/060_INSTALL_PATCHMATCH.md)
+  - [XFormers](installation/070_INSTALL_XFORMERS.md)
+  - [CUDA and ROCm Drivers](installation/030_INSTALL_CUDA_AND_ROCM.md)
+  - [Installing New Models](installation/050_INSTALLING_MODELS.md)
+
 ## :fontawesome-solid-computer: Hardware Requirements
 
 ### :octicons-cpu-24: System
@@ -107,24 +126,6 @@ images in full-precision mode:
 - At least 18 GB of free disk space for the machine learning model, Python, and
   all its dependencies.
 
-## :octicons-package-dependencies-24: Installation
-
-This fork is supported across Linux, Windows and Macintosh. Linux users can use
-either an Nvidia-based card (with CUDA support) or an AMD card (using the ROCm
-driver).
-
-### [Installation Getting Started Guide](installation)
-#### [Automated Installer](installation/010_INSTALL_AUTOMATED.md)
-This method is recommended for 1st time users
-#### [Manual Installation](installation/020_INSTALL_MANUAL.md)
-This method is recommended for experienced users and developers
-#### [Docker Installation](installation/040_INSTALL_DOCKER.md)
-This method is recommended for those familiar with running Docker containers
-### Other Installation Guides
-  - [PyPatchMatch](installation/060_INSTALL_PATCHMATCH.md)
-  - [XFormers](installation/070_INSTALL_XFORMERS.md)
-  - [CUDA and ROCm Drivers](installation/030_INSTALL_CUDA_AND_ROCM.md)
-  - [Installing New Models](installation/050_INSTALLING_MODELS.md)
 
 ## :octicons-gift-24: InvokeAI Features
 
@@ -145,14 +146,17 @@ This method is recommended for those familiar with running Docker containers
 ### Model Management
 - [Installing](installation/050_INSTALLING_MODELS.md)
 - [Model Merging](features/MODEL_MERGING.md)
+- [ControlNet Models](features/CONTROLNET.md)
 - [Style/Subject Concepts and Embeddings](features/CONCEPTS.md)
-- [Textual Inversion](features/TEXTUAL_INVERSION.md)
 - [Not Safe for Work (NSFW) Checker](features/NSFW.md)
 <!-- seperator -->
 ### Prompt Engineering
 - [Prompt Syntax](features/PROMPTS.md)
 - [Generating Variations](features/VARIATIONS.md)
 
+### InvokeAI Configuration
+- [Guide to InvokeAI Runtime Settings](features/CONFIGURATION.md)
+
 ## :octicons-log-16: Important Changes Since Version 2.3
 
 ### Nodes
@@ -219,14 +223,10 @@ get solutions for common installation problems and other issues.
 
 Anyone who wishes to contribute to this project, whether documentation,
 features, bug fixes, code cleanup, testing, or code reviews, is very much
-encouraged to do so. If you are unfamiliar with how to contribute to GitHub
-projects, here is a
-[Getting Started Guide](https://opensource.com/article/19/7/create-pull-request-github).
+encouraged to do so. 
 
-A full set of contribution guidelines, along with templates, are in progress,
-but for now the most important thing is to **make your pull request against the
-"development" branch**, and not against "main". This will help keep public
-breakage to a minimum and will allow you to propose more radical changes.
+[Please take a look at our Contribution documentation to learn more about contributing to InvokeAI. 
+](contributing/CONTRIBUTING.md)
 
 ## :octicons-person-24: Contributors
 

docs/installation/010_INSTALL_AUTOMATED.md

@@ -124,9 +124,9 @@ experimental versions later.
     [latest release](https://github.com/invoke-ai/InvokeAI/releases/latest),
     and look for a file named:
 
-    - InvokeAI-installer-v2.X.X.zip
+    - InvokeAI-installer-v3.X.X.zip
 
-    where "2.X.X" is the latest released version. The file is located
+    where "3.X.X" is the latest released version. The file is located
     at the very bottom of the release page, under **Assets**.
 
 4.  **Unpack the installer**: Unpack the zip file into a convenient directory. This will create a new
@@ -354,8 +354,8 @@ experimental versions later.
 
 12. **InvokeAI Options**: You can launch InvokeAI with several different command-line arguments that
     customize its behavior. For example, you can change the location of the
-    image output directory, or select your favorite sampler. See the
-    [Command-Line Interface](../features/CLI.md) for a full list of the options.
+    image output directory or balance memory usage vs performance. See
+    [Configuration](../features/CONFIGURATION.md) for a full list of the options.
 
     - To set defaults that will take effect every time you launch InvokeAI,
       use a text editor (e.g. Notepad) to exit the file

docs/installation/020_INSTALL_MANUAL.md

@@ -256,7 +256,7 @@ manager, please follow these steps:
 
 10.  Render away!
 
-    Browse the [features](../features/CLI.md) section to learn about all the
+    Browse the [features](../features/index.md) section to learn about all the
     things you can do with InvokeAI.
 
 
@@ -270,7 +270,7 @@ manager, please follow these steps:
 
 12. Other scripts
 
-    The [Textual Inversion](../features/TEXTUAL_INVERSION.md) script can be launched with the command:
+    The [Textual Inversion](../features/TRAINING.md) script can be launched with the command:
 
     ```bash
     invokeai-ti --gui

docs/installation/050_INSTALLING_MODELS.md

@@ -43,24 +43,7 @@ InvokeAI comes with support for a good set of starter models. You'll
 find them listed in the master models file
 `configs/INITIAL_MODELS.yaml` in the InvokeAI root directory. The
 subset that are currently installed are found in
-`configs/models.yaml`. As of v2.3.1, the list of starter models is:
-
-|Model Name | HuggingFace Repo ID | Description | URL |
-|---------- | ---------- | ----------- | --- |
-|stable-diffusion-1.5|runwayml/stable-diffusion-v1-5|Stable Diffusion version 1.5 diffusers model (4.27 GB)|https://huggingface.co/runwayml/stable-diffusion-v1-5 |
-|sd-inpainting-1.5|runwayml/stable-diffusion-inpainting|RunwayML SD 1.5 model optimized for inpainting, diffusers version (4.27 GB)|https://huggingface.co/runwayml/stable-diffusion-inpainting |
-|stable-diffusion-2.1|stabilityai/stable-diffusion-2-1|Stable Diffusion version 2.1 diffusers model, trained on 768 pixel images (5.21 GB)|https://huggingface.co/stabilityai/stable-diffusion-2-1 |
-|sd-inpainting-2.0|stabilityai/stable-diffusion-2-inpainting|Stable Diffusion version 2.0 inpainting model (5.21 GB)|https://huggingface.co/stabilityai/stable-diffusion-2-inpainting |
-|analog-diffusion-1.0|wavymulder/Analog-Diffusion|An SD-1.5 model trained on diverse analog photographs (2.13 GB)|https://huggingface.co/wavymulder/Analog-Diffusion |
-|deliberate-1.0|XpucT/Deliberate|Versatile model that produces detailed images up to 768px (4.27 GB)|https://huggingface.co/XpucT/Deliberate |
-|d&d-diffusion-1.0|0xJustin/Dungeons-and-Diffusion|Dungeons & Dragons characters (2.13 GB)|https://huggingface.co/0xJustin/Dungeons-and-Diffusion |
-|dreamlike-photoreal-2.0|dreamlike-art/dreamlike-photoreal-2.0|A photorealistic model trained on 768 pixel images based on SD 1.5 (2.13 GB)|https://huggingface.co/dreamlike-art/dreamlike-photoreal-2.0 |
-|inkpunk-1.0|Envvi/Inkpunk-Diffusion|Stylized illustrations inspired by Gorillaz, FLCL and Shinkawa; prompt with "nvinkpunk" (4.27 GB)|https://huggingface.co/Envvi/Inkpunk-Diffusion |
-|openjourney-4.0|prompthero/openjourney|An SD 1.5 model fine tuned on Midjourney; prompt with "mdjrny-v4 style" (2.13 GB)|https://huggingface.co/prompthero/openjourney |
-|portrait-plus-1.0|wavymulder/portraitplus|An SD-1.5 model trained on close range portraits of people; prompt with "portrait+" (2.13 GB)|https://huggingface.co/wavymulder/portraitplus |
-|seek-art-mega-1.0|coreco/seek.art_MEGA|A general use SD-1.5 "anything" model that supports multiple styles (2.1 GB)|https://huggingface.co/coreco/seek.art_MEGA |
-|trinart-2.0|naclbit/trinart_stable_diffusion_v2|An SD-1.5 model finetuned with ~40K assorted high resolution manga/anime-style images (2.13 GB)|https://huggingface.co/naclbit/trinart_stable_diffusion_v2 |
-|waifu-diffusion-1.4|hakurei/waifu-diffusion|An SD-1.5 model trained on 680k anime/manga-style images (2.13 GB)|https://huggingface.co/hakurei/waifu-diffusion |
+`configs/models.yaml`.
 
 Note that these files are covered by an "Ethical AI" license which
 forbids certain uses. When you initially download them, you are asked
@@ -71,8 +54,7 @@ with the model terms by visiting the URLs in the table above.
 
 ## Community-Contributed Models
 
-There are too many to list here and more are being contributed every
-day.  [HuggingFace](https://huggingface.co/models?library=diffusers)
+[HuggingFace](https://huggingface.co/models?library=diffusers)
 is a great resource for diffusers models, and is also the home of a
 [fast-growing repository](https://huggingface.co/sd-concepts-library)
 of embedding (".bin") models that add subjects and/or styles to your
@@ -86,310 +68,106 @@ only `.safetensors` and `.ckpt` models, but they can be easily loaded
 into InvokeAI and/or converted into optimized `diffusers` models. Be
 aware that CIVITAI hosts many models that generate NSFW content.
 
-!!! note
-
-    InvokeAI 2.3.x does not support directly importing and
-    running Stable Diffusion version 2 checkpoint models. You may instead
-    convert them into `diffusers` models using the conversion methods
-    described below.
-
 ## Installation
 
-There are multiple ways to install and manage models:
+There are two ways to install and manage models:
 
-1. The `invokeai-configure` script which will download and install them for you.
+1. The `invokeai-model-install` script which will download and install
+them for you.  In addition to supporting main models, you can install
+ControlNet, LoRA and Textual Inversion models.
 
-2. The command-line tool (CLI) has commands that allows you to import, configure and modify
-   models files.
-
-3. The web interface (WebUI) has a GUI for importing and managing
+2. The web interface (WebUI) has a GUI for importing and managing
    models.
 
-### Installation via `invokeai-configure`
+3. By placing models (or symbolic links to models) inside one of the
+InvokeAI root directory's `autoimport` folder.
 
-From the `invoke` launcher, choose option (6) "re-run the configure
-script to download new models." This will launch the same script that
-prompted you to select models at install time. You can use this to add
-models that you skipped the first time around. It is all right to
-specify a model that was previously downloaded; the script will just
-confirm that the files are complete.
+### Installation via `invokeai-model-install`
 
-### Installation via the CLI
+From the `invoke` launcher, choose option [5] "Download and install
+models." This will launch the same script that prompted you to select
+models at install time. You can use this to add models that you
+skipped the first time around. It is all right to specify a model that
+was previously downloaded; the script will just confirm that the files
+are complete.
 
-You can install a new model, including any of the community-supported ones, via
-the command-line client's `!import_model` command.
+The installer has different panels for installing main models from
+HuggingFace, models from Civitai and other arbitrary web sites,
+ControlNet models, LoRA/LyCORIS models, and Textual Inversion
+embeddings. Each section has a text box in which you can enter a new
+model to install. You can refer to a model using its:
 
-#### Installing individual `.ckpt` and `.safetensors` models
+1. Local path to the .ckpt, .safetensors or diffusers folder on your local machine
+2. A directory on your machine that contains multiple models
+3. A URL that points to a downloadable model
+4. A HuggingFace repo id
 
-If the model is already downloaded to your local disk, use
-`!import_model /path/to/file.ckpt` to load it. For example:
+Previously-installed models are shown with checkboxes. Uncheck a box
+to unregister the model from InvokeAI. Models that are physically
+installed inside the InvokeAI root directory will be deleted and
+purged (after a confirmation warning). Models that are located outside
+the InvokeAI root directory will be unregistered but not deleted.
 
-```bash
-invoke> !import_model C:/Users/fred/Downloads/martians.safetensors
+Note: The installer script uses a console-based text interface that requires
+significant amounts of horizontal and vertical space. If the display
+looks messed up, just enlarge the terminal window and/or relaunch the
+script.
+
+If you wish you can script model addition and deletion, as well as
+listing installed models. Start the "developer's console" and give the
+command `invokeai-model-install --help`. This will give you a series
+of command-line parameters that will let you control model
+installation. Examples:
+
+```
+# (list all controlnet models)
+invokeai-model-install --list controlnet
+
+# (install the model at the indicated URL)
+invokeai-model-install --add http://civitai.com/2860
+
+# (delete the named model)
+invokeai-model-install --delete sd-1/main/analog-diffusion
 ```
 
-!!! tip "Forward Slashes"
-    On Windows systems, use forward slashes rather than backslashes
-    in your file paths.
-    If you do use backslashes,
-    you must double them like this:
-    `C:\\Users\\fred\\Downloads\\martians.safetensors`
+### Installation via the Web GUI
 
-Alternatively you can directly import the file using its URL:
+To install a new model using the Web GUI, do the following:
 
-```bash
-invoke> !import_model https://example.org/sd_models/martians.safetensors
-```
+1. Open the InvokeAI Model Manager (cube at the bottom of the
+left-hand panel) and navigate to *Import Models*
 
-For this to work, the URL must not be password-protected. Otherwise
-you will receive a 404 error.
+2. In the field labeled *Location* type in the path to the model you
+wish to install. You may use a URL, HuggingFace repo id, or a path on
+your local disk.
 
-When you import a legacy model, the CLI will first ask you what type
-of model this is. You can indicate whether it is a model based on
-Stable Diffusion 1.x (1.4 or 1.5), one based on Stable Diffusion 2.x,
-or a 1.x inpainting model. Be careful to indicate the correct model
-type, or it will not load correctly. You can correct the model type
-after the fact using the `!edit_model` command.
+3. Alternatively, the *Scan for Models* button allows you to paste in
+the path to a folder somewhere on your machine. It will be scanned for
+importable models and prompt you to add the ones of your choice.
 
-The system will then ask you a few other questions about the model,
-including what size image it was trained on (usually 512x512), what
-name and description you wish to use for it, and whether you would
-like to install a custom VAE (variable autoencoder) file for the
-model. For recent models, the answer to the VAE question is usually
-"no," but it won't hurt to answer "yes".
+4. Press *Add Model* and wait for confirmation that the model
+was added.
 
-After importing, the model will load. If this is successful, you will
-be asked if you want to keep the model loaded in memory to start
-generating immediately. You'll also be asked if you wish to make this
-the default model on startup. You can change this later using
-`!edit_model`.
+To delete a model, Select *Model Manager* to list all the currently
+installed models. Press the trash can icons to delete any models you
+wish to get rid of. Models whose weights are located inside the
+InvokeAI `models` directory will be purged from disk, while those
+located outside will be unregistered from InvokeAI, but not deleted.
 
-#### Importing a batch of `.ckpt` and `.safetensors` models from a directory
+You can see where model weights are located by clicking on the model name.
+This will bring up an editable info panel showing the model's characteristics,
+including the `Model Location` of its files.
 
-You may also point `!import_model` to a directory containing a set of
-`.ckpt` or `.safetensors` files. They will be imported _en masse_.
+### Installation via the `autoimport` function
 
-!!! example
+In the InvokeAI root directory you will find a series of folders under
+`autoimport`, one each for main models, controlnets, embeddings and
+Loras.  Any models that you add to these directories will be scanned
+at startup time and registered automatically.
 
-    ```console
-    invoke> !import_model C:/Users/fred/Downloads/civitai_models/
-    ```
+You may create symbolic links from these folders to models located
+elsewhere on disk and they will be autoimported. You can also create
+subfolders and organize them as you wish.
 
-You will be given the option to import all models found in the
-directory, or select which ones to import. If there are subfolders
-within the directory, they will be searched for models to import.
-
-#### Installing `diffusers` models
-
-You can install a `diffusers` model from the HuggingFace site using
-`!import_model` and the HuggingFace repo_id for the model:
-
-```bash
-invoke> !import_model andite/anything-v4.0
-```
-
-Alternatively, you can download the model to disk and import it from
-there. The model may be distributed as a ZIP file, or as a Git
-repository:
-
-```bash
-invoke> !import_model C:/Users/fred/Downloads/andite--anything-v4.0
-```
-
-!!! tip "The CLI supports file path autocompletion"
-         Type a bit of the path name and hit ++tab++ in order to get a choice of
-         possible completions.
-
-!!! tip "On Windows, you can drag model files onto the command-line"
-         Once you have typed in `!import_model `, you can drag the
-         model  file or directory onto the command-line to insert the model path. This way, you don't need to
-         type it or copy/paste. However, you will need to reverse or
-         double backslashes as noted above.
-
-Before installing, the CLI will ask you for a short name and
-description for the model, whether to make this the default model that
-is loaded at InvokeAI startup time, and whether to replace its
-VAE. Generally the answer to the latter question is "no".
-
-### Converting legacy models into `diffusers`
-
-The CLI `!convert_model` will convert a `.safetensors` or `.ckpt`
-models file into `diffusers` and install it.This will enable the model
-to load and run faster without loss of image quality.
-
-The usage is identical to `!import_model`. You may point the command
-to either a downloaded model file on disk, or to a (non-password
-protected) URL:
-
-```bash
-invoke> !convert_model C:/Users/fred/Downloads/martians.safetensors
-```
-
-After a successful conversion, the CLI will offer you the option of
-deleting the original `.ckpt` or `.safetensors` file.
-
-### Optimizing a previously-installed model
-
-Lastly, if you have previously installed a `.ckpt` or `.safetensors`
-file and wish to convert it into a `diffusers` model, you can do this
-without re-downloading and converting the original file using the
-`!optimize_model` command. Simply pass the short name of an existing
-installed model:
-
-```bash
-invoke> !optimize_model martians-v1.0
-```
-
-The model will be converted into `diffusers` format and replace the
-previously installed version. You will again be offered the
-opportunity to delete the original `.ckpt` or `.safetensors` file.
-
-### Related CLI Commands
-
-There are a whole series of additional model management commands in
-the CLI that you can read about in [Command-Line
-Interface](../features/CLI.md). These include:
-
-* `!models` - List all installed models
-* `!switch <model name>` - Switch to the indicated model
-* `!edit_model <model name>` - Edit the indicated model to change its name, description or other properties
-* `!del_model <model name>` - Delete the indicated model
-
-### Manually editing `configs/models.yaml`
-
-
-If you are comfortable with a text editor then you may simply edit `models.yaml`
-directly.
-
-You will need to download the desired `.ckpt/.safetensors` file and
-place it somewhere on your machine's filesystem. Alternatively, for a
-`diffusers` model, record the repo_id or download the whole model
-directory. Then using a **text** editor (e.g. the Windows Notepad
-application), open the file `configs/models.yaml`, and add a new
-stanza that follows this model:
-
-#### A legacy model
-
-A legacy `.ckpt` or `.safetensors` entry will look like this:
-
-```yaml
-arabian-nights-1.0:
-  description: A great fine-tune in Arabian Nights style
-  weights: ./path/to/arabian-nights-1.0.ckpt
-  config: ./configs/stable-diffusion/v1-inference.yaml
-  format: ckpt
-  width: 512
-  height: 512
-  default: false
-```
-
-Note that `format` is `ckpt` for both `.ckpt` and `.safetensors` files.
-
-#### A diffusers model
-
-A stanza for a `diffusers` model will look like this for a HuggingFace
-model with a repository ID:
-
-```yaml
-arabian-nights-1.1:
-  description: An even better fine-tune of the Arabian Nights
-  repo_id: captahab/arabian-nights-1.1
-  format: diffusers
-  default: true
-```
-
-And for a downloaded directory:
-
-```yaml
-arabian-nights-1.1:
-  description: An even better fine-tune of the Arabian Nights
-  path: /path/to/captahab-arabian-nights-1.1
-  format: diffusers
-  default: true
-```
-
-There is additional syntax for indicating an external VAE to use with
-this model. See `INITIAL_MODELS.yaml` and `models.yaml` for examples.
-
-After you save the modified `models.yaml` file relaunch
-`invokeai`. The new model will now be available for your use.
-
-### Installation via the WebUI
-
-To access the WebUI Model Manager, click on the button that looks like
-a cube in the upper right side of the browser screen. This will bring
-up a dialogue that lists the models you have already installed, and
-allows you to load, delete or edit them:
-
-<figure markdown>
-
-![model-manager](../assets/installing-models/webui-models-1.png)
-
-</figure>
-
-To add a new model, click on **+ Add New** and select to either a
-checkpoint/safetensors model, or a diffusers model:
-
-<figure markdown>
-
-![model-manager-add-new](../assets/installing-models/webui-models-2.png)
-
-</figure>
-
-In this example, we chose **Add Diffusers**. As shown in the figure
-below, a new dialogue prompts you to enter the name to use for the
-model, its description, and either the location of the `diffusers`
-model on disk, or its Repo ID on the HuggingFace web site. If you
-choose to enter a path to disk, the system will autocomplete for you
-as you type:
-
-<figure markdown>
-
-![model-manager-add-diffusers](../assets/installing-models/webui-models-3.png)
-
-</figure>
-
-Press **Add Model** at the bottom of the dialogue (scrolled out of
-site in the figure), and the model will be downloaded, imported, and
-registered in `models.yaml`.
-
-The **Add Checkpoint/Safetensor Model** option is similar, except that
-in this case you can choose to scan an entire folder for
-checkpoint/safetensors files to import. Simply type in the path of the
-directory and press the "Search" icon. This will display the
-`.ckpt` and `.safetensors` found inside the directory and its
-subfolders, and allow you to choose which ones to import:
-
-<figure markdown>
-
-![model-manager-add-checkpoint](../assets/installing-models/webui-models-4.png)
-
-</figure>
-
-## Model Management Startup Options
-
-The `invoke` launcher and the `invokeai` script accept a series of
-command-line arguments that modify InvokeAI's behavior when loading
-models. These can be provided on the command line, or added to the
-InvokeAI root directory's `invokeai.init` initialization file.
-
-The arguments are:
-
-* `--model <model name>` -- Start up with the indicated model loaded
-* `--ckpt_convert` -- When a checkpoint/safetensors model is loaded, convert it into a `diffusers` model in memory. This does not permanently save the converted model to disk.
-* `--autoconvert <path/to/directory>` -- Scan the indicated directory path for new checkpoint/safetensors files, convert them into `diffusers` models, and import them into InvokeAI.
-
-Here is an example of providing an argument on the command line using
-the `invoke.sh` launch script:
-
-```bash
-invoke.sh --autoconvert /home/fred/stable-diffusion-checkpoints
-```
-
-And here is what the same argument looks like in `invokeai.init`:
-
-```bash
---outdir="/home/fred/invokeai/outputs
---no-nsfw_checker
---autoconvert /home/fred/stable-diffusion-checkpoints
-```
+The location of the autoimport directories are controlled by settings
+in `invokeai.yaml`. See [Configuration](../features/CONFIGURATION.md).

docs/installation/index.md

@@ -15,7 +15,7 @@ See the [troubleshooting
 section](010_INSTALL_AUTOMATED.md#troubleshooting) of the automated
 install guide for frequently-encountered installation issues.
 
-## Main Application
+## Installation options
 
 1. [Automated Installer](010_INSTALL_AUTOMATED.md)
 
@@ -24,6 +24,9 @@ install guide for frequently-encountered installation issues.
     "developer console" which will help us debug problems with you and
     give you to access experimental features.
 
+
+    ✅ This is the recommended option for first time users. 
+
 2. [Manual Installation](020_INSTALL_MANUAL.md)
 
     In this method you will manually run the commands needed to install

docs/nodes/communityNodes.md

@@ -0,0 +1,52 @@
+# Community Nodes
+
+These are nodes that have been developed by the community, for the community. If you're not sure what a node is, you can learn more about nodes [here](overview.md).
+
+If you'd like to submit a node for the community, please refer to the [node creation overview](./overview.md#contributing-nodes).
+
+To download a node, simply download the `.py` node file from the link and add it to the `invokeai/app/invocations/` folder in your Invoke AI install location. Along with the node, an example node graph should be provided to help you get started with the node. 
+
+To use a community node graph, download the the `.json` node graph file and load it into Invoke AI via the **Load Nodes** button on the Node Editor. 
+
+## Disclaimer
+
+The nodes linked below have been developed and contributed by members of the Invoke AI community. While we strive to ensure the quality and safety of these contributions, we do not guarantee the reliability or security of the nodes. If you have issues or concerns with any of the nodes below, please raise it on GitHub or in the Discord.
+
+## List of Nodes
+
+### Face Mask
+
+**Description:** This node autodetects a face in the image using MediaPipe and masks it by making it transparent. Via outpainting you can swap faces with other faces, or invert the mask and swap things around the face with other things. Additionally, you can supply X and Y offset values to scale/change the shape of the mask for finer control. The node also outputs an all-white mask in the same dimensions as the input image. This is needed by the inpaint node (and unified canvas) for outpainting.
+
+**Node Link:** https://github.com/ymgenesis/InvokeAI/blob/facemaskmediapipe/invokeai/app/invocations/facemask.py
+
+**Example Node Graph:** https://www.mediafire.com/file/gohn5sb1bfp8use/21-July_2023-FaceMask.json/file
+
+**Output Examples** 
+
+![2e3168cb-af6a-475d-bfac-c7b7fd67b4c2](https://github.com/ymgenesis/InvokeAI/assets/25252829/a5ad7d44-2ada-4b3c-a56e-a21f8244a1ac)
+![2_annotated](https://github.com/ymgenesis/InvokeAI/assets/25252829/53416c8a-a23b-4d76-bb6d-3cfd776e0096)
+![2fe2150c-fd08-4e26-8c36-f0610bf441bb](https://github.com/ymgenesis/InvokeAI/assets/25252829/b0f7ecfe-f093-4147-a904-b9f131b41dc9)
+![831b6b98-4f0f-4360-93c8-69a9c1338cbe](https://github.com/ymgenesis/InvokeAI/assets/25252829/fc7b0622-e361-4155-8a76-082894d084f0)
+
+--------------------------------
+### Super Cool Node Template
+
+**Description:** This node allows you to do super cool things with InvokeAI.
+
+**Node Link:** https://github.com/invoke-ai/InvokeAI/fake_node.py
+
+**Example Node Graph:**  https://github.com/invoke-ai/InvokeAI/fake_node_graph.json
+
+**Output Examples** 
+
+![Invoke AI](https://invoke-ai.github.io/InvokeAI/assets/invoke_ai_banner.png)
+
+### Ideal Size
+
+**Description:** This node calculates an ideal image size for a first pass of a multi-pass upscaling. The aim is to avoid duplication that results from choosing a size larger than the model is capable of.
+
+**Node Link:** https://github.com/JPPhoto/ideal-size-node
+
+## Help
+If you run into any issues with a node, please post in the [InvokeAI Discord](https://discord.gg/ZmtBAhwWhy). 

docs/nodes/overview.md

@@ -0,0 +1,42 @@
+# Nodes
+
+## What are Nodes?
+An Node is simply a single operation that takes in some inputs and gives
+out some outputs. We can then chain multiple nodes together to create more
+complex functionality. All InvokeAI features are added through nodes.
+
+This means nodes can be used to easily extend the image generation capabilities of InvokeAI, and allow you build workflows to suit your needs. 
+
+You can read more about nodes and the node editor [here](../features/NODES.md). 
+
+
+## Downloading Nodes
+To download a new node, visit our list of [Community Nodes](communityNodes.md). These are nodes that have been created by the community, for the community. 
+
+
+## Contributing Nodes
+
+To learn about creating a new node, please visit our [Node creation documenation](../contributing/INVOCATIONS.md). 
+
+Once you’ve created a node and confirmed that it behaves as expected locally, follow these steps: 
+* Make sure the node is contained in a new Python (.py) file 
+* Submit a pull request with a link to your node in GitHub against the `nodes` branch to add the node to the [Community Nodes](Community Nodes) list
+    * Make sure you are following the template below and have provided all relevant details about the node and what it does.
+* A maintainer will review the pull request and node. If the node is aligned with the direction of the project, you might be asked for permission to include it in the core project.
+
+### Community Node Template
+
+```markdown
+--------------------------------
+### Super Cool Node Template
+
+**Description:** This node allows you to do super cool things with InvokeAI.
+
+**Node Link:** https://github.com/invoke-ai/InvokeAI/fake_node.py
+
+**Example Node Graph:**  https://github.com/invoke-ai/InvokeAI/fake_node_graph.json
+
+**Output Examples** 
+
+![InvokeAI](https://invoke-ai.github.io/InvokeAI/assets/invoke_ai_banner.png)
+```

docs/other/CONTRIBUTORS.md

@@ -17,67 +17,267 @@ We thank them for all of their time and hard work.
 
 * @lstein (Lincoln Stein) - Co-maintainer
 * @blessedcoolant - Co-maintainer
-* @hipsterusername (Kent Keirsey) - Product Manager
-* @psychedelicious - Web Team Leader
+* @hipsterusername (Kent Keirsey) - Co-maintainer, CEO, Positive Vibes
+* @psychedelicious (Spencer Mabrito) - Web Team Leader
 * @Kyle0654 (Kyle Schouviller) - Node Architect and General Backend Wizard
-* @damian0815 - Attention Systems and Gameplay Engineer
-* @mauwii (Matthias Wild) - Continuous integration and product maintenance engineer
-* @Netsvetaev (Artur Netsvetaev) - UI/UX Developer
-* @tildebyte - General gadfly and resident (self-appointed) know-it-all
-* @keturn - Lead for Diffusers port
+* @damian0815 - Attention Systems and Compel Maintainer
 * @ebr (Eugene Brodsky) - Cloud/DevOps/Sofware engineer; your friendly neighbourhood cluster-autoscaler
-* @jpphoto (Jonathan Pollack) - Inference and rendering engine optimization
-* @genomancer (Gregg Helt) - Model training and merging
+* @genomancer (Gregg Helt) - Controlnet support
+* @StAlKeR7779 (Sergey Borisov) - Torch stack, ONNX, model management, optimization
+* @cheerio (Mary Rogers) - Lead Engineer & Web App Development
+* @brandon (Brandon Rising) - Platform, Infrastructure, Backend Systems
+* @ryanjdick (Ryan Dick) - Machine Learning & Training
+* @millu (Millun Atluri) - Community Manager, Documentation, Node-wrangler
+* @chainchompa (Jennifer Player) - Web Development & Chain-Chomping
+* @keturn (Kevin Turner) - Diffusers
+* @gogurt enjoyer - Discord moderator and end user support
+* @whosawhatsis - Discord moderator and end user support
+* @dwinrger - Discord moderator and end user support
+* @526christian - Discord moderator and end user support
 
-## **Contributions by**
+## **Full List of Contributors by Commit Name**
 
-- [Sean McLellan](https://github.com/Oceanswave)
-- [Kevin Gibbons](https://github.com/bakkot)
-- [Tesseract Cat](https://github.com/TesseractCat)
-- [blessedcoolant](https://github.com/blessedcoolant)
-- [David Ford](https://github.com/david-ford)
-- [yunsaki](https://github.com/yunsaki)
-- [James Reynolds](https://github.com/magnusviri)
-- [David Wager](https://github.com/maddavid123)
-- [Jason Toffaletti](https://github.com/toffaletti)
-- [tildebyte](https://github.com/tildebyte)
-- [Cragin Godley](https://github.com/cgodley)
-- [BlueAmulet](https://github.com/BlueAmulet)
-- [Benjamin Warner](https://github.com/warner-benjamin)
-- [Cora Johnson-Roberson](https://github.com/corajr)
-- [veprogames](https://github.com/veprogames)
-- [JigenD](https://github.com/JigenD)
-- [Niek van der Maas](https://github.com/Niek)
-- [Henry van Megen](https://github.com/hvanmegen)
-- [Håvard Gulldahl](https://github.com/havardgulldahl)
-- [greentext2](https://github.com/greentext2)
-- [Simon Vans-Colina](https://github.com/simonvc)
-- [Gabriel Rotbart](https://github.com/gabrielrotbart)
-- [Eric Khun](https://github.com/erickhun)
-- [Brent Ozar](https://github.com/BrentOzar)
-- [nderscore](https://github.com/nderscore)
-- [Mikhail Tishin](https://github.com/tishin)
-- [Tom Elovi Spruce](https://github.com/ilovecomputers)
-- [spezialspezial](https://github.com/spezialspezial)
-- [Yosuke Shinya](https://github.com/shinya7y)
-- [Andy Pilate](https://github.com/Cubox)
-- [Muhammad Usama](https://github.com/SMUsamaShah)
-- [Arturo Mendivil](https://github.com/artmen1516)
-- [Paul Sajna](https://github.com/sajattack)
-- [Samuel Husso](https://github.com/shusso)
-- [nicolai256](https://github.com/nicolai256)
-- [Mihai](https://github.com/mh-dm)
-- [Any Winter](https://github.com/any-winter-4079)
-- [Doggettx](https://github.com/doggettx)
-- [Matthias Wild](https://github.com/mauwii)
-- [Kyle Schouviller](https://github.com/kyle0654)
-- [rabidcopy](https://github.com/rabidcopy)
-- [Dominic Letz](https://github.com/dominicletz)
-- [Dmitry T.](https://github.com/ArDiouscuros)
-- [Kent Keirsey](https://github.com/hipsterusername)
-- [psychedelicious](https://github.com/psychedelicious)
-- [damian0815](https://github.com/damian0815)
-- [Eugene Brodsky](https://github.com/ebr)
+- AbdBarho
+- ablattmann
+- AdamOStark
+- Adam Rice
+- Airton Silva
+- Alexander Eichhorn
+- Alexandre D. Roberge
+- Andreas Rozek
+- Andre LaBranche
+- Andy Bearman
+- Andy Luhrs
+- Andy Pilate
+- Any-Winter-4079
+- apolinario
+- ArDiouscuros
+- Armando C. Santisbon
+- Arthur Holstvoogd
+- artmen1516
+- Artur
+- Arturo Mendivil
+- Ben Alkov
+- Benjamin Warner
+- Bernard Maltais
+- blessedcoolant
+- blhook
+- BlueAmulet
+- Bouncyknighter
+- Brandon Rising
+- Brent Ozar
+- Brian Racer
+- bsilvereagle
+- c67e708d
+- CapableWeb
+- Carson Katri
+- Chloe
+- Chris Dawson
+- Chris Hayes
+- Chris Jones
+- chromaticist
+- Claus F. Strasburger
+- cmdr2
+- cody
+- Conor Reid
+- Cora Johnson-Roberson
+- coreco
+- cosmii02
+- cpacker
+- Cragin Godley
+- creachec
+- Damian Stewart
+- Daniel Manzke
+- Danny Beer
+- Dan Sully
+- David Burnett
+- David Ford
+- David Regla
+- David Wager
+- Daya Adianto
+- db3000
+- Denis Olshin
+- Dennis
+- Dominic Letz
+- DrGunnarMallon
+- Edward Johan
+- elliotsayes
+- Elrik
+- ElrikUnderlake
+- Eric Khun
+- Eric Wolf
+- Eugene Brodsky
+- ExperimentalCyborg
+- Fabian Bahl
+- Fabio 'MrWHO' Torchetti
+- fattire
+- Felipe Nogueira
+- Félix Sanz
+- figgefigge
+- Gabriel Mackievicz Telles
+- gabrielrotbart
+- gallegonovato
+- Gérald LONLAS
+- GitHub Actions Bot
+- gogurtenjoyer
+- greentext2
+- Gregg Helt
+- H4rk
+- Håvard Gulldahl
+- henry
+- Henry van Megen
+- hipsterusername
+- hj
+- Hosted Weblate
+- Iman Karim
+- ismail ihsan bülbül
+- Ivan Efimov
+- jakehl
+- Jakub Kolčář
+- JamDon2
+- James Reynolds
+- Jan Skurovec
+- Jari Vetoniemi
+- Jason Toffaletti
+- Jaulustus
+- Jeff Mahoney
+- jeremy
+- Jeremy Clark
+- JigenD
+- Jim Hays
+- Johan Roxendal
+- Johnathon Selstad
+- Jonathan
+- Joseph Dries III
+- JPPhoto
+- jspraul
+- Justin Wong
+- Juuso V
+- Kaspar Emanuel
+- Katsuyuki-Karasawa
+- Kent Keirsey
+- Kevin Coakley
+- Kevin Gibbons
+- Kevin Schaul
+- Kevin Turner
+- krummrey
+- Kyle Lacy
+- Kyle Schouviller
+- Lawrence Norton
+- LemonDouble
+- Leo Pasanen
+- Lincoln Stein
+- LoganPederson
+- Lynne Whitehorn
+- majick
+- Marco Labarile
+- Martin Kristiansen
+- Mary Hipp Rogers
+- mastercaster9000
+- Matthias Wild
+- michaelk71
+- mickr777
+- Mihai
+- Mihail Dumitrescu
+- Mikhail Tishin
+- Millun Atluri
+- Minjune Song
+- mitien
+- mofuzz
+- Muhammad Usama
+- Name
+- _nderscore
+- Netzer R
+- Nicholas Koh
+- Nicholas Körfer
+- nicolai256
+- Niek van der Maas
+- noodlebox
+- Nuno Coração
+- ofirkris
+- Olivier Louvignes
+- owenvincent
+- Patrick Esser
+- Patrick Tien
+- Patrick von Platen
+- Paul Sajna
+- pejotr
+- Peter Baylies
+- Peter Lin
+- plucked
+- prixt
+- psychedelicious
+- Rainer Bernhardt
+- Riccardo Giovanetti
+- Rich Jones
+- rmagur1203
+- Rob Baines
+- Robert Bolender
+- Robin Rombach
+- Rohan Barar
+- rpagliuca
+- rromb
+- Rupesh Sreeraman
+- Ryan Cao
+- Saifeddine
+- Saifeddine ALOUI
+- SammCheese
+- Sammy
+- sammyf
+- Samuel Husso
+- Scott Lahteine
+- Sean McLellan
+- Sebastian Aigner
+- Sergey Borisov
+- Sergey Krashevich
+- Shapor Naghibzadeh
+- Shawn Zhong
+- Simon Vans-Colina
+- skunkworxdark
+- slashtechno
+- spezialspezial
+- ssantos
+- StAlKeR7779
+- Stephan Koglin-Fischer
+- SteveCaruso
+- Steve Martinelli
+- Steven Frank
+- System X - Files
+- Taylor Kems
+- techicode
+- techybrain-dev
+- tesseractcat
+- thealanle
+- Thomas
+- tildebyte
+- Tim Cabbage
+- Tom
+- Tom Elovi Spruce
+- Tom Gouville
+- tomosuto
+- Travco
+- Travis Palmer
+- tyler
+- unknown
+- user1
+- Vedant Madane
+- veprogames
+- wa.code
+- wfng92
+- whosawhatsis
+- Will
+- William Becher
+- William Chong
+- xra
+- Yeung Yiu Hung
+- ymgenesis
+- Yorzaren
+- Yosuke Shinya
+- yun saki
+- Zadagu
+- zeptofine
+- 冯不游
+- 唐澤 克幸
 
 ## **Original CompVis Authors**
 

installer/create_installer.sh

@@ -24,7 +24,8 @@ read -e -p "Tag this repo with '${VERSION}' and '${LATEST_TAG}'? [n]: " input
 RESPONSE=${input:='n'}
 if [ "$RESPONSE" == 'y' ]; then
 
-    if ! git tag $VERSION ; then
+    git push origin :refs/tags/$VERSION
+    if ! git tag -fa $VERSION ; then
 	    echo "Existing/invalid tag"
 	    exit -1
     fi

installer/install.bat.in

@@ -38,7 +38,7 @@ echo    https://learn.microsoft.com/en-US/cpp/windows/latest-supported-vc-redist
 echo.
 echo See %INSTRUCTIONS% for more details.
 echo.
-echo "For the best user experience we suggest enlarging or maximizing this window now."
+echo FOR THE BEST USER EXPERIENCE WE SUGGEST MAXIMIZING THIS WINDOW NOW.
 pause
 
 @rem ---------------------------- check Python version ---------------

installer/templates/invoke.bat.in

@@ -19,7 +19,7 @@ echo 8. Open the developer console
 echo 9. Update InvokeAI
 echo 10. Command-line help
 echo Q - Quit
-set /P choice="Please enter 1-10, Q: [2] "
+set /P choice="Please enter 1-10, Q: [1] "
 if not defined choice set choice=1
 IF /I "%choice%" == "1" (
     echo Starting the InvokeAI browser-based UI..

invokeai/app/api/dependencies.py

@@ -11,9 +11,9 @@ from invokeai.app.services.board_images import (
 )
 from invokeai.app.services.board_record_storage import SqliteBoardRecordStorage
 from invokeai.app.services.boards import BoardService, BoardServiceDependencies
+from invokeai.app.services.config import InvokeAIAppConfig
 from invokeai.app.services.image_record_storage import SqliteImageRecordStorage
 from invokeai.app.services.images import ImageService, ImageServiceDependencies
-from invokeai.app.services.metadata import CoreMetadataService
 from invokeai.app.services.resource_name import SimpleNameService
 from invokeai.app.services.urls import LocalUrlService
 from invokeai.backend.util.logging import InvokeAILogger
@@ -21,7 +21,6 @@ from invokeai.version.invokeai_version import __version__
 
 from ..services.default_graphs import create_system_graphs
 from ..services.latent_storage import DiskLatentsStorage, ForwardCacheLatentsStorage
-from ..services.restoration_services import RestorationServices
 from ..services.graph import GraphExecutionState, LibraryGraph
 from ..services.image_file_storage import DiskImageFileStorage
 from ..services.invocation_queue import MemoryInvocationQueue
@@ -58,8 +57,9 @@ class ApiDependencies:
     invoker: Invoker = None
 
     @staticmethod
-    def initialize(config, event_handler_id: int, logger: Logger = logger):
-        logger.debug(f'InvokeAI version {__version__}')
+    def initialize(config: InvokeAIAppConfig, event_handler_id: int, logger: Logger = logger):
+        logger.info(f"InvokeAI version {__version__}")
+        logger.info(f"Root directory = {str(config.root_path)}")
         logger.debug(f"Internet connectivity is {config.internet_available}")
 
         events = FastAPIEventService(event_handler_id)
@@ -75,7 +75,6 @@ class ApiDependencies:
         )
 
         urls = LocalUrlService()
-        metadata = CoreMetadataService()
         image_record_storage = SqliteImageRecordStorage(db_location)
         image_file_storage = DiskImageFileStorage(f"{output_folder}/images")
         names = SimpleNameService()
@@ -111,7 +110,6 @@ class ApiDependencies:
                 board_image_record_storage=board_image_record_storage,
                 image_record_storage=image_record_storage,
                 image_file_storage=image_file_storage,
-                metadata=metadata,
                 url=urls,
                 logger=logger,
                 names=names,
@@ -120,7 +118,7 @@ class ApiDependencies:
         )
 
         services = InvocationServices(
-            model_manager=ModelManagerService(config,logger),
+            model_manager=ModelManagerService(config, logger),
             events=events,
             latents=latents,
             images=images,
@@ -132,7 +130,6 @@ class ApiDependencies:
             ),
             graph_execution_manager=graph_execution_manager,
             processor=DefaultInvocationProcessor(),
-            restoration=RestorationServices(config, logger),
             configuration=config,
             logger=logger,
         )

invokeai/app/api/routers/app_info.py

@@ -1,18 +1,73 @@
+from enum import Enum
+from fastapi import Body
 from fastapi.routing import APIRouter
-from pydantic import BaseModel
+from pydantic import BaseModel, Field
 
+from invokeai.backend.image_util.patchmatch import PatchMatch
 from invokeai.version import __version__
 
-app_router = APIRouter(prefix="/v1/app", tags=['app'])
+from ..dependencies import ApiDependencies
+from invokeai.backend.util.logging import logging
+
+class LogLevel(int, Enum):
+    NotSet = logging.NOTSET
+    Debug = logging.DEBUG
+    Info = logging.INFO
+    Warning = logging.WARNING
+    Error = logging.ERROR
+    Critical = logging.CRITICAL
+    
+app_router = APIRouter(prefix="/v1/app", tags=["app"])
 
 
 class AppVersion(BaseModel):
     """App Version Response"""
-    version: str
+
+    version: str = Field(description="App version")
 
 
-@app_router.get('/version', operation_id="app_version",
-                status_code=200,
-                response_model=AppVersion)
+class AppConfig(BaseModel):
+    """App Config Response"""
+
+    infill_methods: list[str] = Field(description="List of available infill methods")
+
+
+@app_router.get(
+    "/version", operation_id="app_version", status_code=200, response_model=AppVersion
+)
 async def get_version() -> AppVersion:
     return AppVersion(version=__version__)
+
+
+@app_router.get(
+    "/config", operation_id="get_config", status_code=200, response_model=AppConfig
+)
+async def get_config() -> AppConfig:
+    infill_methods = ['tile']
+    if PatchMatch.patchmatch_available():
+        infill_methods.append('patchmatch')
+    return AppConfig(infill_methods=infill_methods)
+
+@app_router.get(
+    "/logging",
+    operation_id="get_log_level",
+    responses={200: {"description" : "The operation was successful"}},
+    response_model = LogLevel,
+)
+async def get_log_level(
+) -> LogLevel:
+    """Returns the log level"""
+    return LogLevel(ApiDependencies.invoker.services.logger.level)
+
+@app_router.post(
+    "/logging",
+    operation_id="set_log_level",
+    responses={200: {"description" : "The operation was successful"}},
+    response_model = LogLevel,
+)
+async def set_log_level(
+        level: LogLevel = Body(description="New log verbosity level"),
+) -> LogLevel:
+    """Sets the log verbosity level"""
+    ApiDependencies.invoker.services.logger.setLevel(level)
+    return LogLevel(ApiDependencies.invoker.services.logger.level)

invokeai/app/api/routers/board_images.py

@@ -24,11 +24,14 @@ async def create_board_image(
 ):
     """Creates a board_image"""
     try:
-        result = ApiDependencies.invoker.services.board_images.add_image_to_board(board_id=board_id, image_name=image_name)
+        result = ApiDependencies.invoker.services.board_images.add_image_to_board(
+            board_id=board_id, image_name=image_name
+        )
         return result
     except Exception as e:
         raise HTTPException(status_code=500, detail="Failed to add to board")
-    
+
+
 @board_images_router.delete(
     "/",
     operation_id="remove_board_image",
@@ -43,27 +46,10 @@ async def remove_board_image(
 ):
     """Deletes a board_image"""
     try:
-        result = ApiDependencies.invoker.services.board_images.remove_image_from_board(board_id=board_id, image_name=image_name)
+        result = ApiDependencies.invoker.services.board_images.remove_image_from_board(
+            board_id=board_id, image_name=image_name
+        )
         return result
     except Exception as e:
         raise HTTPException(status_code=500, detail="Failed to update board")
 
-
-
-@board_images_router.get(
-    "/{board_id}",
-    operation_id="list_board_images",
-    response_model=OffsetPaginatedResults[ImageDTO],
-)
-async def list_board_images(
-    board_id: str = Path(description="The id of the board"),
-    offset: int = Query(default=0, description="The page offset"),
-    limit: int = Query(default=10, description="The number of boards per page"),
-) -> OffsetPaginatedResults[ImageDTO]:
-    """Gets a list of images for a board"""
-
-    results = ApiDependencies.invoker.services.board_images.get_images_for_board(
-        board_id,
-    )
-    return results
-

invokeai/app/api/routers/boards.py

@@ -1,16 +1,28 @@
 from typing import Optional, Union
+
 from fastapi import Body, HTTPException, Path, Query
 from fastapi.routing import APIRouter
+from pydantic import BaseModel, Field
+
 from invokeai.app.services.board_record_storage import BoardChanges
 from invokeai.app.services.image_record_storage import OffsetPaginatedResults
 from invokeai.app.services.models.board_record import BoardDTO
 
-
 from ..dependencies import ApiDependencies
 
 boards_router = APIRouter(prefix="/v1/boards", tags=["boards"])
 
 
+class DeleteBoardResult(BaseModel):
+    board_id: str = Field(description="The id of the board that was deleted.")
+    deleted_board_images: list[str] = Field(
+        description="The image names of the board-images relationships that were deleted."
+    )
+    deleted_images: list[str] = Field(
+        description="The names of the images that were deleted."
+    )
+
+
 @boards_router.post(
     "/",
     operation_id="create_board",
@@ -69,25 +81,42 @@ async def update_board(
         raise HTTPException(status_code=500, detail="Failed to update board")
 
 
-@boards_router.delete("/{board_id}", operation_id="delete_board")
+@boards_router.delete(
+    "/{board_id}", operation_id="delete_board", response_model=DeleteBoardResult
+)
 async def delete_board(
     board_id: str = Path(description="The id of board to delete"),
     include_images: Optional[bool] = Query(
         description="Permanently delete all images on the board", default=False
     ),
-) -> None:
+) -> DeleteBoardResult:
     """Deletes a board"""
     try:
         if include_images is True:
+            deleted_images = ApiDependencies.invoker.services.board_images.get_all_board_image_names_for_board(
+                board_id=board_id
+            )
             ApiDependencies.invoker.services.images.delete_images_on_board(
                 board_id=board_id
             )
             ApiDependencies.invoker.services.boards.delete(board_id=board_id)
+            return DeleteBoardResult(
+                board_id=board_id,
+                deleted_board_images=[],
+                deleted_images=deleted_images,
+            )
         else:
+            deleted_board_images = ApiDependencies.invoker.services.board_images.get_all_board_image_names_for_board(
+                board_id=board_id
+            )
             ApiDependencies.invoker.services.boards.delete(board_id=board_id)
+            return DeleteBoardResult(
+                board_id=board_id,
+                deleted_board_images=deleted_board_images,
+                deleted_images=[],
+            )
     except Exception as e:
-        # TODO: Does this need any exception handling at all?
-        pass
+        raise HTTPException(status_code=500, detail="Failed to delete board")
 
 
 @boards_router.get(
@@ -115,3 +144,19 @@ async def list_boards(
             status_code=400,
             detail="Invalid request: Must provide either 'all' or both 'offset' and 'limit'",
         )
+
+
+@boards_router.get(
+    "/{board_id}/image_names",
+    operation_id="list_all_board_image_names",
+    response_model=list[str],
+)
+async def list_all_board_image_names(
+    board_id: str = Path(description="The id of the board"),
+) -> list[str]:
+    """Gets a list of images for a board"""
+
+    image_names = ApiDependencies.invoker.services.board_images.get_all_board_image_names_for_board(
+        board_id,
+    )
+    return image_names

invokeai/app/api/routers/images.py

@@ -1,25 +1,28 @@
 import io
 from typing import Optional
+
 from fastapi import Body, HTTPException, Path, Query, Request, Response, UploadFile
-from fastapi.routing import APIRouter
 from fastapi.responses import FileResponse
+from fastapi.routing import APIRouter
 from PIL import Image
-from invokeai.app.models.image import (
-    ImageCategory,
-    ResourceOrigin,
-)
+
+from invokeai.app.invocations.metadata import ImageMetadata
+from invokeai.app.models.image import ImageCategory, ResourceOrigin
 from invokeai.app.services.image_record_storage import OffsetPaginatedResults
+from invokeai.app.services.item_storage import PaginatedResults
 from invokeai.app.services.models.image_record import (
     ImageDTO,
     ImageRecordChanges,
     ImageUrlsDTO,
 )
-from invokeai.app.services.item_storage import PaginatedResults
 
 from ..dependencies import ApiDependencies
 
 images_router = APIRouter(prefix="/v1/images", tags=["images"])
 
+# images are immutable; set a high max-age
+IMAGE_MAX_AGE = 31536000
+
 
 @images_router.post(
     "/",
@@ -37,9 +40,15 @@ async def upload_image(
     response: Response,
     image_category: ImageCategory = Query(description="The category of the image"),
     is_intermediate: bool = Query(description="Whether this is an intermediate image"),
+    board_id: Optional[str] = Query(
+        default=None, description="The board to add this image to, if any"
+    ),
     session_id: Optional[str] = Query(
         default=None, description="The session ID associated with this upload, if any"
     ),
+    crop_visible: Optional[bool] = Query(
+        default=False, description="Whether to crop the image"
+    ),
 ) -> ImageDTO:
     """Uploads an image"""
     if not file.content_type.startswith("image"):
@@ -49,6 +58,9 @@ async def upload_image(
 
     try:
         pil_image = Image.open(io.BytesIO(contents))
+        if crop_visible:
+            bbox = pil_image.getbbox()
+            pil_image = pil_image.crop(bbox)
     except:
         # Error opening the image
         raise HTTPException(status_code=415, detail="Failed to read image")
@@ -59,6 +71,7 @@ async def upload_image(
             image_origin=ResourceOrigin.EXTERNAL,
             image_category=image_category,
             session_id=session_id,
+            board_id=board_id,
             is_intermediate=is_intermediate,
         )
 
@@ -83,6 +96,18 @@ async def delete_image(
         pass
 
 
+@images_router.post("/clear-intermediates", operation_id="clear_intermediates")
+async def clear_intermediates() -> int:
+    """Clears all intermediates"""
+
+    try:
+        count_deleted = ApiDependencies.invoker.services.images.delete_intermediates()
+        return count_deleted
+    except Exception as e:
+        raise HTTPException(status_code=500, detail="Failed to clear intermediates")
+        pass
+
+
 @images_router.patch(
     "/{image_name}",
     operation_id="update_image",
@@ -103,14 +128,14 @@ async def update_image(
 
 
 @images_router.get(
-    "/{image_name}/metadata",
-    operation_id="get_image_metadata",
+    "/{image_name}",
+    operation_id="get_image_dto",
     response_model=ImageDTO,
 )
-async def get_image_metadata(
+async def get_image_dto(
     image_name: str = Path(description="The name of image to get"),
 ) -> ImageDTO:
-    """Gets an image's metadata"""
+    """Gets an image's DTO"""
 
     try:
         return ApiDependencies.invoker.services.images.get_dto(image_name)
@@ -119,7 +144,23 @@ async def get_image_metadata(
 
 
 @images_router.get(
-    "/{image_name}",
+    "/{image_name}/metadata",
+    operation_id="get_image_metadata",
+    response_model=ImageMetadata,
+)
+async def get_image_metadata(
+    image_name: str = Path(description="The name of image to get"),
+) -> ImageMetadata:
+    """Gets an image's metadata"""
+
+    try:
+        return ApiDependencies.invoker.services.images.get_metadata(image_name)
+    except Exception as e:
+        raise HTTPException(status_code=404)
+
+
+@images_router.get(
+    "/{image_name}/full",
     operation_id="get_image_full",
     response_class=Response,
     responses={
@@ -141,12 +182,14 @@ async def get_image_full(
         if not ApiDependencies.invoker.services.images.validate_path(path):
             raise HTTPException(status_code=404)
 
-        return FileResponse(
+        response = FileResponse(
             path,
             media_type="image/png",
             filename=image_name,
             content_disposition_type="inline",
         )
+        response.headers["Cache-Control"] = f"max-age={IMAGE_MAX_AGE}"
+        return response
     except Exception as e:
         raise HTTPException(status_code=404)
 
@@ -175,9 +218,11 @@ async def get_image_thumbnail(
         if not ApiDependencies.invoker.services.images.validate_path(path):
             raise HTTPException(status_code=404)
 
-        return FileResponse(
+        response = FileResponse(
             path, media_type="image/webp", content_disposition_type="inline"
         )
+        response.headers["Cache-Control"] = f"max-age={IMAGE_MAX_AGE}"
+        return response
     except Exception as e:
         raise HTTPException(status_code=404)
 
@@ -208,26 +253,27 @@ async def get_image_urls(
 
 @images_router.get(
     "/",
-    operation_id="list_images_with_metadata",
+    operation_id="list_image_dtos",
     response_model=OffsetPaginatedResults[ImageDTO],
 )
-async def list_images_with_metadata(
+async def list_image_dtos(
     image_origin: Optional[ResourceOrigin] = Query(
-        default=None, description="The origin of images to list"
+        default=None, description="The origin of images to list."
     ),
     categories: Optional[list[ImageCategory]] = Query(
-        default=None, description="The categories of image to include"
+        default=None, description="The categories of image to include."
     ),
     is_intermediate: Optional[bool] = Query(
-        default=None, description="Whether to list intermediate images"
+        default=None, description="Whether to list intermediate images."
     ),
     board_id: Optional[str] = Query(
-        default=None, description="The board id to filter by"
+        default=None,
+        description="The board id to filter by. Use 'none' to find images without a board.",
     ),
     offset: int = Query(default=0, description="The page offset"),
     limit: int = Query(default=10, description="The number of images per page"),
 ) -> OffsetPaginatedResults[ImageDTO]:
-    """Gets a list of images"""
+    """Gets a list of image DTOs"""
 
     image_dtos = ApiDependencies.invoker.services.images.get_many(
         offset,

invokeai/app/api/routers/models.py

@@ -1,6 +1,7 @@
-# Copyright (c) 2023 Kyle Schouviller (https://github.com/kyle0654), 2023 Kent Keirsey (https://github.com/hipsterusername), 2024 Lincoln Stein
+# Copyright (c) 2023 Kyle Schouviller (https://github.com/kyle0654), 2023 Kent Keirsey (https://github.com/hipsterusername), 2023 Lincoln D. Stein
 
 
+import pathlib
 from typing import Literal, List, Optional, Union
 
 from fastapi import Body, Path, Query, Response
@@ -12,8 +13,11 @@ from invokeai.backend import BaseModelType, ModelType
 from invokeai.backend.model_management.models import (
     OPENAPI_MODEL_CONFIGS,
     SchedulerPredictionType,
+    ModelNotFoundException,
+    InvalidModelException,
 )
 from invokeai.backend.model_management import MergeInterpolationMethod
+
 from ..dependencies import ApiDependencies
 
 models_router = APIRouter(prefix="/v1/models", tags=["models"])
@@ -22,6 +26,7 @@ UpdateModelResponse = Union[tuple(OPENAPI_MODEL_CONFIGS)]
 ImportModelResponse = Union[tuple(OPENAPI_MODEL_CONFIGS)]
 ConvertModelResponse = Union[tuple(OPENAPI_MODEL_CONFIGS)]
 MergeModelResponse = Union[tuple(OPENAPI_MODEL_CONFIGS)]
+ImportModelAttributes = Union[tuple(OPENAPI_MODEL_CONFIGS)]
 
 class ModelsList(BaseModel):
     models: list[Union[tuple(OPENAPI_MODEL_CONFIGS)]]
@@ -32,11 +37,16 @@ class ModelsList(BaseModel):
     responses={200: {"model": ModelsList }},
 )
 async def list_models(
-    base_model: Optional[BaseModelType] = Query(default=None, description="Base model"),
+    base_models: Optional[List[BaseModelType]] = Query(default=None, description="Base models to include"),
     model_type: Optional[ModelType] = Query(default=None, description="The type of model to get"),
 ) -> ModelsList:
     """Gets a list of models"""
-    models_raw = ApiDependencies.invoker.services.model_manager.list_models(base_model, model_type)
+    if base_models and len(base_models)>0:
+        models_raw = list()
+        for base_model in base_models:
+            models_raw.extend(ApiDependencies.invoker.services.model_manager.list_models(base_model, model_type))
+    else:
+        models_raw = ApiDependencies.invoker.services.model_manager.list_models(None, model_type)
     models = parse_obj_as(ModelsList, { "models": models_raw })
     return models
 
@@ -44,8 +54,9 @@ async def list_models(
     "/{base_model}/{model_type}/{model_name}",
     operation_id="update_model",
     responses={200: {"description" : "The model was updated successfully"},
+               400: {"description" : "Bad request"},
                404: {"description" : "The model could not be found"},
-               400: {"description" : "Bad request"}
+               409: {"description" : "There is already a model corresponding to the new name"},
                },
     status_code = 200,
     response_model = UpdateModelResponse,
@@ -56,33 +67,69 @@ async def update_model(
         model_name: str = Path(description="model name"),
         info: Union[tuple(OPENAPI_MODEL_CONFIGS)] = Body(description="Model configuration"),
 ) -> UpdateModelResponse:
-    """ Add Model """
+    """ Update model contents with a new config. If the model name or base fields are changed, then the model is renamed. """
+    logger = ApiDependencies.invoker.services.logger
+
+    
     try:
+        previous_info = ApiDependencies.invoker.services.model_manager.list_model(
+            model_name=model_name,
+            base_model=base_model,
+            model_type=model_type,
+        )
+
+        # rename operation requested
+        if info.model_name != model_name or info.base_model != base_model:
+            ApiDependencies.invoker.services.model_manager.rename_model(
+                base_model = base_model,
+                model_type = model_type,
+                model_name = model_name,
+                new_name = info.model_name,
+                new_base = info.base_model,
+            )
+            logger.info(f'Successfully renamed {base_model}/{model_name}=>{info.base_model}/{info.model_name}')
+            # update information to support an update of attributes
+            model_name = info.model_name
+            base_model = info.base_model
+            new_info = ApiDependencies.invoker.services.model_manager.list_model(
+                model_name=model_name,
+                base_model=base_model,
+                model_type=model_type,
+            )
+            if new_info.get('path') != previous_info.get('path'):  # model manager moved model path during rename - don't overwrite it
+                info.path = new_info.get('path')
+            
         ApiDependencies.invoker.services.model_manager.update_model(
             model_name=model_name,
             base_model=base_model,
             model_type=model_type,
             model_attributes=info.dict()
         )
+            
         model_raw = ApiDependencies.invoker.services.model_manager.list_model(
             model_name=model_name,
             base_model=base_model,
             model_type=model_type,
         )
         model_response = parse_obj_as(UpdateModelResponse, model_raw)
-    except KeyError as e:
+    except ModelNotFoundException as e:
         raise HTTPException(status_code=404, detail=str(e))
     except ValueError as e:
+        logger.error(str(e))
+        raise HTTPException(status_code=409, detail=str(e))
+    except Exception as e:
+        logger.error(str(e))
         raise HTTPException(status_code=400, detail=str(e))
 
     return model_response
 
 @models_router.post(
-    "/",
+    "/import",
     operation_id="import_model",
     responses= {
         201: {"description" : "The model imported successfully"},
         404: {"description" : "The model could not be found"},
+        415: {"description" : "Unrecognized file/folder format"},
         424: {"description" : "The model appeared to import successfully, but could not be found in the model manager"},
         409: {"description" : "There is already a model corresponding to this path or repo_id"},
     },
@@ -94,7 +141,7 @@ async def import_model(
         prediction_type: Optional[Literal['v_prediction','epsilon','sample']] = \
                 Body(description='Prediction type for SDv2 checkpoint files', default="v_prediction"),
 ) -> ImportModelResponse:
-    """ Add a model using its local path, repo_id, or remote URL """
+    """ Add a model using its local path, repo_id, or remote URL. Model characteristics will be probed and configured automatically """
     
     items_to_import = {location}
     prediction_types = { x.value: x for x in SchedulerPredictionType }
@@ -109,7 +156,7 @@ async def import_model(
 
         if not info:
             logger.error("Import failed")
-            raise HTTPException(status_code=424)
+            raise HTTPException(status_code=415)
         
         logger.info(f'Successfully imported {location}, got {info}')
         model_raw = ApiDependencies.invoker.services.model_manager.list_model(
@@ -119,25 +166,66 @@ async def import_model(
         )
         return parse_obj_as(ImportModelResponse, model_raw)
     
-    except KeyError as e:
+    except ModelNotFoundException as e:
+        logger.error(str(e))
+        raise HTTPException(status_code=404, detail=str(e))
+    except InvalidModelException as e:
+        logger.error(str(e))
+        raise HTTPException(status_code=415)
+    except ValueError as e:
+        logger.error(str(e))
+        raise HTTPException(status_code=409, detail=str(e))
+        
+@models_router.post(
+    "/add",
+    operation_id="add_model",
+    responses= {
+        201: {"description" : "The model added successfully"},
+        404: {"description" : "The model could not be found"},
+        424: {"description" : "The model appeared to add successfully, but could not be found in the model manager"},
+        409: {"description" : "There is already a model corresponding to this path or repo_id"},
+    },
+    status_code=201,
+    response_model=ImportModelResponse
+)
+async def add_model(
+        info: Union[tuple(OPENAPI_MODEL_CONFIGS)] = Body(description="Model configuration"),
+) -> ImportModelResponse:
+    """ Add a model using the configuration information appropriate for its type. Only local models can be added by path"""
+    
+    logger = ApiDependencies.invoker.services.logger
+
+    try:
+        ApiDependencies.invoker.services.model_manager.add_model(
+            info.model_name,
+            info.base_model,
+            info.model_type,
+            model_attributes = info.dict()
+        )
+        logger.info(f'Successfully added {info.model_name}')
+        model_raw = ApiDependencies.invoker.services.model_manager.list_model(
+            model_name=info.model_name,
+            base_model=info.base_model,
+            model_type=info.model_type
+        )
+        return parse_obj_as(ImportModelResponse, model_raw)
+    except ModelNotFoundException as e:
         logger.error(str(e))
         raise HTTPException(status_code=404, detail=str(e))
     except ValueError as e:
         logger.error(str(e))
         raise HTTPException(status_code=409, detail=str(e))
-        
 
+    
 @models_router.delete(
     "/{base_model}/{model_type}/{model_name}",
     operation_id="del_model",
     responses={
-        204: {
-        "description": "Model deleted successfully"
-        }, 
-        404: {
-        "description": "Model not found"
-        }
+        204: { "description": "Model deleted successfully" }, 
+        404: { "description": "Model not found" }
     },
+    status_code = 204,
+    response_model = None,
 )
 async def delete_model(
         base_model: BaseModelType = Path(description="Base model"),
@@ -154,9 +242,9 @@ async def delete_model(
                                                                  )
         logger.info(f"Deleted model: {model_name}")
         return Response(status_code=204)
-    except KeyError:
-        logger.error(f"Model not found: {model_name}")
-        raise HTTPException(status_code=404, detail=f"Model '{model_name}' not found")
+    except ModelNotFoundException as e:
+        logger.error(str(e))
+        raise HTTPException(status_code=404, detail=str(e))
 
 @models_router.put(
     "/convert/{base_model}/{model_type}/{model_name}",
@@ -173,24 +261,75 @@ async def convert_model(
         base_model: BaseModelType = Path(description="Base model"),
         model_type: ModelType = Path(description="The type of model"),
         model_name: str = Path(description="model name"),
+        convert_dest_directory: Optional[str] = Query(default=None, description="Save the converted model to the designated directory"),
 ) -> ConvertModelResponse:
-    """Convert a checkpoint model into a diffusers model"""
+    """Convert a checkpoint model into a diffusers model, optionally saving to the indicated destination directory, or `models` if none."""
     logger = ApiDependencies.invoker.services.logger
     try:
         logger.info(f"Converting model: {model_name}")
+        dest = pathlib.Path(convert_dest_directory) if convert_dest_directory else None
         ApiDependencies.invoker.services.model_manager.convert_model(model_name,
                                                                      base_model = base_model,
-                                                                     model_type = model_type
+                                                                     model_type = model_type,
+                                                                     convert_dest_directory = dest,
                                                                      )
         model_raw = ApiDependencies.invoker.services.model_manager.list_model(model_name,
                                                                               base_model = base_model,
                                                                               model_type = model_type)
         response = parse_obj_as(ConvertModelResponse, model_raw)
-    except KeyError:
-        raise HTTPException(status_code=404, detail=f"Model '{model_name}' not found")
+    except ModelNotFoundException as e:
+        raise HTTPException(status_code=404, detail=f"Model '{model_name}' not found: {str(e)}")
     except ValueError as e:
         raise HTTPException(status_code=400, detail=str(e))
     return response
+
+@models_router.get(
+    "/search",
+    operation_id="search_for_models",
+    responses={
+        200: { "description": "Directory searched successfully" },
+        404: { "description": "Invalid directory path"  },
+    },
+    status_code = 200,
+    response_model = List[pathlib.Path]
+)
+async def search_for_models(
+        search_path: pathlib.Path = Query(description="Directory path to search for models")
+)->List[pathlib.Path]:
+    if not search_path.is_dir():
+        raise HTTPException(status_code=404, detail=f"The search path '{search_path}' does not exist or is not directory")
+    return ApiDependencies.invoker.services.model_manager.search_for_models([search_path])
+
+@models_router.get(
+    "/ckpt_confs",
+    operation_id="list_ckpt_configs",
+    responses={
+        200: { "description" : "paths retrieved successfully" },
+    },
+    status_code = 200,
+    response_model = List[pathlib.Path]
+)
+async def list_ckpt_configs(
+)->List[pathlib.Path]:
+    """Return a list of the legacy checkpoint configuration files stored in `ROOT/configs/stable-diffusion`, relative to ROOT."""
+    return ApiDependencies.invoker.services.model_manager.list_checkpoint_configs()
+    
+        
+@models_router.post(
+    "/sync",
+    operation_id="sync_to_config",
+    responses={
+        201: { "description": "synchronization successful" },
+    },
+    status_code = 201,
+    response_model = bool
+)
+async def sync_to_config(
+)->bool:
+    """Call after making changes to models.yaml, autoimport directories or models directory to synchronize
+    in-memory data structures with disk data structures."""
+    ApiDependencies.invoker.services.model_manager.sync_to_config()
+    return True
         
 @models_router.put(
     "/merge/{base_model}",
@@ -210,23 +349,27 @@ async def merge_models(
         alpha: Optional[float]                     = Body(description="Alpha weighting strength to apply to 2d and 3d models", default=0.5),
         interp: Optional[MergeInterpolationMethod] = Body(description="Interpolation method"),
         force: Optional[bool]                      = Body(description="Force merging of models created with different versions of diffusers", default=False),
+        merge_dest_directory: Optional[str]       = Body(description="Save the merged model to the designated directory (with 'merged_model_name' appended)", default=None)
 ) -> MergeModelResponse:
     """Convert a checkpoint model into a diffusers model"""
     logger = ApiDependencies.invoker.services.logger
     try:
-        logger.info(f"Merging models: {model_names}")
+        logger.info(f"Merging models: {model_names} into {merge_dest_directory or '<MODELS>'}/{merged_model_name}")
+        dest = pathlib.Path(merge_dest_directory) if merge_dest_directory else None
         result = ApiDependencies.invoker.services.model_manager.merge_models(model_names,
                                                                              base_model,
-                                                                             merged_model_name or "+".join(model_names),
-                                                                             alpha,
-                                                                             interp,
-                                                                             force)
+                                                                             merged_model_name=merged_model_name or "+".join(model_names),
+                                                                             alpha=alpha,
+                                                                             interp=interp,
+                                                                             force=force,
+                                                                             merge_dest_directory = dest
+                                                                             )
         model_raw = ApiDependencies.invoker.services.model_manager.list_model(result.name,
                                                                               base_model = base_model,
                                                                               model_type = ModelType.Main,
                                                                               )
         response = parse_obj_as(ConvertModelResponse, model_raw)
-    except KeyError:
+    except ModelNotFoundException:
         raise HTTPException(status_code=404, detail=f"One or more of the models '{model_names}' not found")
     except ValueError as e:
         raise HTTPException(status_code=400, detail=str(e))

invokeai/app/api_app.py

@@ -4,6 +4,7 @@ import sys
 from inspect import signature
 
 import uvicorn
+import socket
 
 from fastapi import FastAPI
 from fastapi.middleware.cors import CORSMiddleware
@@ -39,6 +40,7 @@ from .invocations.baseinvocation import BaseInvocation
     
 
 import torch
+import invokeai.backend.util.hotfixes
 if torch.backends.mps.is_available():
     import invokeai.backend.util.mps_fixes
 
@@ -192,9 +194,22 @@ app.mount("/",
          )
 
 def invoke_api():
+    def find_port(port: int):
+        """Find a port not in use starting at given port"""
+        # Taken from https://waylonwalker.com/python-find-available-port/, thanks Waylon!
+        # https://github.com/WaylonWalker
+        with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as s:
+            if s.connect_ex(("localhost", port)) == 0:
+                return find_port(port=port + 1)
+            else:
+                return port
+
+    port = find_port(app_config.port)
+    if port != app_config.port:
+        logger.warn(f"Port {app_config.port} in use, using port {port}")
     # Start our own event loop for eventing usage
     loop = asyncio.new_event_loop()
-    config = uvicorn.Config(app=app, host=app_config.host, port=app_config.port, loop=loop)
+    config = uvicorn.Config(app=app, host=app_config.host, port=port, loop=loop)
     # Use access_log to turn off logging
     server = uvicorn.Server(config)
     loop.run_until_complete(server.serve())

invokeai/app/cli_app.py

@@ -34,7 +34,6 @@ from invokeai.app.services.board_record_storage import SqliteBoardRecordStorage
 from invokeai.app.services.boards import BoardService, BoardServiceDependencies
 from invokeai.app.services.image_record_storage import SqliteImageRecordStorage
 from invokeai.app.services.images import ImageService, ImageServiceDependencies
-from invokeai.app.services.metadata import CoreMetadataService
 from invokeai.app.services.resource_name import SimpleNameService
 from invokeai.app.services.urls import LocalUrlService
 from .services.default_graphs import (default_text_to_image_graph_id,
@@ -55,10 +54,10 @@ from .services.invocation_services import InvocationServices
 from .services.invoker import Invoker
 from .services.model_manager_service import ModelManagerService
 from .services.processor import DefaultInvocationProcessor
-from .services.restoration_services import RestorationServices
 from .services.sqlite import SqliteItemStorage
 
 import torch
+import invokeai.backend.util.hotfixes
 if torch.backends.mps.is_available():
     import invokeai.backend.util.mps_fixes
 
@@ -244,7 +243,6 @@ def invoke_cli():
         )
 
     urls = LocalUrlService()
-    metadata = CoreMetadataService()
     image_record_storage = SqliteImageRecordStorage(db_location)
     image_file_storage = DiskImageFileStorage(f"{output_folder}/images")
     names = SimpleNameService()
@@ -277,7 +275,6 @@ def invoke_cli():
             board_image_record_storage=board_image_record_storage,
             image_record_storage=image_record_storage,
             image_file_storage=image_file_storage,
-            metadata=metadata,
             url=urls,
             logger=logger,
             names=names,
@@ -298,7 +295,6 @@ def invoke_cli():
         ),
         graph_execution_manager=graph_execution_manager,
         processor=DefaultInvocationProcessor(),
-        restoration=RestorationServices(config,logger=logger),
         logger=logger,
         configuration=config,
     )

invokeai/app/invocations/collections.py

@@ -4,17 +4,12 @@ from typing import Literal
 
 import numpy as np
 from pydantic import Field, validator
-from invokeai.app.models.image import ImageField
 
+from invokeai.app.models.image import ImageField
 from invokeai.app.util.misc import SEED_MAX, get_random_seed
 
-from .baseinvocation import (
-    BaseInvocation,
-    InvocationConfig,
-    InvocationContext,
-    BaseInvocationOutput,
-    UIConfig,
-)
+from .baseinvocation import (BaseInvocation, BaseInvocationOutput,
+                             InvocationConfig, InvocationContext, UIConfig)
 
 
 class IntCollectionOutput(BaseInvocationOutput):
@@ -32,7 +27,8 @@ class FloatCollectionOutput(BaseInvocationOutput):
     type: Literal["float_collection"] = "float_collection"
 
     # Outputs
-    collection: list[float] = Field(default=[], description="The float collection")
+    collection: list[float] = Field(
+        default=[], description="The float collection")
 
 
 class ImageCollectionOutput(BaseInvocationOutput):
@@ -41,7 +37,8 @@ class ImageCollectionOutput(BaseInvocationOutput):
     type: Literal["image_collection"] = "image_collection"
 
     # Outputs
-    collection: list[ImageField] = Field(default=[], description="The output images")
+    collection: list[ImageField] = Field(
+        default=[], description="The output images")
 
     class Config:
         schema_extra = {"required": ["type", "collection"]}
@@ -57,6 +54,14 @@ class RangeInvocation(BaseInvocation):
     stop: int = Field(default=10, description="The stop of the range")
     step: int = Field(default=1, description="The step of the range")
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Range",
+                "tags": ["range", "integer", "collection"]
+            },
+        }
+
     @validator("stop")
     def stop_gt_start(cls, v, values):
         if "start" in values and v <= values["start"]:
@@ -79,10 +84,20 @@ class RangeOfSizeInvocation(BaseInvocation):
     size: int = Field(default=1, description="The number of values")
     step: int = Field(default=1, description="The step of the range")
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Sized Range",
+                "tags": ["range", "integer", "size", "collection"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> IntCollectionOutput:
         return IntCollectionOutput(
-            collection=list(range(self.start, self.start + self.size, self.step))
-        )
+            collection=list(
+                range(
+                    self.start, self.start + self.size,
+                    self.step)))
 
 
 class RandomRangeInvocation(BaseInvocation):
@@ -103,11 +118,21 @@ class RandomRangeInvocation(BaseInvocation):
         default_factory=get_random_seed,
     )
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Random Range",
+                "tags": ["range", "integer", "random", "collection"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> IntCollectionOutput:
         rng = np.random.default_rng(self.seed)
         return IntCollectionOutput(
-            collection=list(rng.integers(low=self.low, high=self.high, size=self.size))
-        )
+            collection=list(
+                rng.integers(
+                    low=self.low, high=self.high,
+                    size=self.size)))
 
 
 class ImageCollectionInvocation(BaseInvocation):
@@ -121,6 +146,7 @@ class ImageCollectionInvocation(BaseInvocation):
         default=[], description="The image collection to load"
     )
     # fmt: on
+
     def invoke(self, context: InvocationContext) -> ImageCollectionOutput:
         return ImageCollectionOutput(collection=self.images)
 
@@ -128,6 +154,7 @@ class ImageCollectionInvocation(BaseInvocation):
         schema_extra = {
             "ui": {
                 "type_hints": {
+                    "title": "Image Collection",
                     "images": "image_collection",
                 }
             },

invokeai/app/invocations/compel.py

@@ -1,8 +1,8 @@
-from typing import Literal, Optional, Union, List
+from typing import Literal, Optional, Union, List, Annotated
 from pydantic import BaseModel, Field
 import re
 import torch
-from compel import Compel
+from compel import Compel, ReturnedEmbeddingsType
 from compel.prompt_parser import (Blend, Conjunction,
                                   CrossAttentionControlSubstitute,
                                   FlattenedPrompt, Fragment)
@@ -14,6 +14,7 @@ from ...backend.stable_diffusion.diffusion import InvokeAIDiffuserComponent
 from .baseinvocation import (BaseInvocation, BaseInvocationOutput,
                              InvocationConfig, InvocationContext)
 from .model import ClipField
+from dataclasses import dataclass
 
 
 class ConditioningField(BaseModel):
@@ -23,6 +24,34 @@ class ConditioningField(BaseModel):
     class Config:
         schema_extra = {"required": ["conditioning_name"]}
 
+@dataclass
+class BasicConditioningInfo:
+    #type: Literal["basic_conditioning"] = "basic_conditioning"
+    embeds: torch.Tensor
+    extra_conditioning: Optional[InvokeAIDiffuserComponent.ExtraConditioningInfo]
+    # weight: float
+    # mode: ConditioningAlgo
+
+@dataclass
+class SDXLConditioningInfo(BasicConditioningInfo):
+    #type: Literal["sdxl_conditioning"] = "sdxl_conditioning"
+    pooled_embeds: torch.Tensor
+    add_time_ids: torch.Tensor
+
+ConditioningInfoType = Annotated[
+    Union[BasicConditioningInfo, SDXLConditioningInfo],
+    Field(discriminator="type")
+]
+
+@dataclass
+class ConditioningFieldData:
+    conditionings: List[Union[BasicConditioningInfo, SDXLConditioningInfo]]
+    #unconditioned: Optional[torch.Tensor]
+
+#class ConditioningAlgo(str, Enum):
+#    Compose = "compose"
+#    ComposeEx = "compose_ex"
+#    PerpNeg = "perp_neg"
 
 class CompelOutput(BaseInvocationOutput):
     """Compel parser output"""
@@ -57,10 +86,10 @@ class CompelInvocation(BaseInvocation):
     @torch.no_grad()
     def invoke(self, context: InvocationContext) -> CompelOutput:
         tokenizer_info = context.services.model_manager.get_model(
-            **self.clip.tokenizer.dict(),
+            **self.clip.tokenizer.dict(), context=context,
         )
         text_encoder_info = context.services.model_manager.get_model(
-            **self.clip.text_encoder.dict(),
+            **self.clip.text_encoder.dict(), context=context,
         )
 
         def _lora_loader():
@@ -82,6 +111,7 @@ class CompelInvocation(BaseInvocation):
                         model_name=name,
                         base_model=self.clip.text_encoder.base_model,
                         model_type=ModelType.TextualInversion,
+                        context=context,
                     ).context.model
                 )
             except ModelNotFoundException:
@@ -100,7 +130,7 @@ class CompelInvocation(BaseInvocation):
                 text_encoder=text_encoder,
                 textual_inversion_manager=ti_manager,
                 dtype_for_device_getter=torch_dtype,
-                truncate_long_prompts=True,  # TODO:
+                truncate_long_prompts=True,
             )
 
             conjunction = Compel.parse_prompt_string(self.prompt)
@@ -112,19 +142,25 @@ class CompelInvocation(BaseInvocation):
             c, options = compel.build_conditioning_tensor_for_prompt_object(
                 prompt)
 
-            # TODO: long prompt support
-            # if not self.truncate_long_prompts:
-            #    [c, uc] = compel.pad_conditioning_tensors_to_same_length([c, uc])
             ec = InvokeAIDiffuserComponent.ExtraConditioningInfo(
                 tokens_count_including_eos_bos=get_max_token_count(
                     tokenizer, conjunction),
                 cross_attention_control_args=options.get(
                     "cross_attention_control", None),)
 
-        conditioning_name = f"{context.graph_execution_state_id}_{self.id}_conditioning"
+        c = c.detach().to("cpu")
 
-        # TODO: hacky but works ;D maybe rename latents somehow?
-        context.services.latents.save(conditioning_name, (c, ec))
+        conditioning_data = ConditioningFieldData(
+            conditionings=[
+                BasicConditioningInfo(
+                    embeds=c,
+                    extra_conditioning=ec,
+                )
+            ]
+        )
+
+        conditioning_name = f"{context.graph_execution_state_id}_{self.id}_conditioning"
+        context.services.latents.save(conditioning_name, conditioning_data)
 
         return CompelOutput(
             conditioning=ConditioningField(
@@ -132,6 +168,397 @@ class CompelInvocation(BaseInvocation):
             ),
         )
 
+class SDXLPromptInvocationBase:
+    def run_clip_raw(self, context, clip_field, prompt, get_pooled):
+        tokenizer_info = context.services.model_manager.get_model(
+            **clip_field.tokenizer.dict(),
+        )
+        text_encoder_info = context.services.model_manager.get_model(
+            **clip_field.text_encoder.dict(),
+        )
+
+        def _lora_loader():
+            for lora in clip_field.loras:
+                lora_info = context.services.model_manager.get_model(
+                    **lora.dict(exclude={"weight"}))
+                yield (lora_info.context.model, lora.weight)
+                del lora_info
+            return
+
+        #loras = [(context.services.model_manager.get_model(**lora.dict(exclude={"weight"})).context.model, lora.weight) for lora in self.clip.loras]
+
+        ti_list = []
+        for trigger in re.findall(r"<[a-zA-Z0-9., _-]+>", prompt):
+            name = trigger[1:-1]
+            try:
+                ti_list.append(
+                    context.services.model_manager.get_model(
+                        model_name=name,
+                        base_model=clip_field.text_encoder.base_model,
+                        model_type=ModelType.TextualInversion,
+                    ).context.model
+                )
+            except ModelNotFoundException:
+                # print(e)
+                #import traceback
+                #print(traceback.format_exc())
+                print(f"Warn: trigger: \"{trigger}\" not found")
+
+        with ModelPatcher.apply_lora_text_encoder(text_encoder_info.context.model, _lora_loader()),\
+                ModelPatcher.apply_ti(tokenizer_info.context.model, text_encoder_info.context.model, ti_list) as (tokenizer, ti_manager),\
+                ModelPatcher.apply_clip_skip(text_encoder_info.context.model, clip_field.skipped_layers),\
+                text_encoder_info as text_encoder:
+
+            text_inputs = tokenizer(
+                prompt,
+                padding="max_length",
+                max_length=tokenizer.model_max_length,
+                truncation=True,
+                return_tensors="pt",
+            )
+            text_input_ids = text_inputs.input_ids
+            prompt_embeds = text_encoder(
+                text_input_ids.to(text_encoder.device),
+                output_hidden_states=True,
+            )
+            if get_pooled:
+                c_pooled = prompt_embeds[0]
+            else:
+                c_pooled = None
+            c = prompt_embeds.hidden_states[-2]
+
+        del tokenizer
+        del text_encoder
+        del tokenizer_info
+        del text_encoder_info
+
+        c = c.detach().to("cpu")
+        if c_pooled is not None:
+            c_pooled = c_pooled.detach().to("cpu")
+
+        return c, c_pooled, None
+
+    def run_clip_compel(self, context, clip_field, prompt, get_pooled):
+        tokenizer_info = context.services.model_manager.get_model(
+            **clip_field.tokenizer.dict(),
+        )
+        text_encoder_info = context.services.model_manager.get_model(
+            **clip_field.text_encoder.dict(),
+        )
+
+        def _lora_loader():
+            for lora in clip_field.loras:
+                lora_info = context.services.model_manager.get_model(
+                    **lora.dict(exclude={"weight"}))
+                yield (lora_info.context.model, lora.weight)
+                del lora_info
+            return
+
+        #loras = [(context.services.model_manager.get_model(**lora.dict(exclude={"weight"})).context.model, lora.weight) for lora in self.clip.loras]
+
+        ti_list = []
+        for trigger in re.findall(r"<[a-zA-Z0-9., _-]+>", prompt):
+            name = trigger[1:-1]
+            try:
+                ti_list.append(
+                    context.services.model_manager.get_model(
+                        model_name=name,
+                        base_model=clip_field.text_encoder.base_model,
+                        model_type=ModelType.TextualInversion,
+                    ).context.model
+                )
+            except ModelNotFoundException:
+                # print(e)
+                #import traceback
+                #print(traceback.format_exc())
+                print(f"Warn: trigger: \"{trigger}\" not found")
+
+        with ModelPatcher.apply_lora_text_encoder(text_encoder_info.context.model, _lora_loader()),\
+                ModelPatcher.apply_ti(tokenizer_info.context.model, text_encoder_info.context.model, ti_list) as (tokenizer, ti_manager),\
+                ModelPatcher.apply_clip_skip(text_encoder_info.context.model, clip_field.skipped_layers),\
+                text_encoder_info as text_encoder:
+
+            compel = Compel(
+                tokenizer=tokenizer,
+                text_encoder=text_encoder,
+                textual_inversion_manager=ti_manager,
+                dtype_for_device_getter=torch_dtype,
+                truncate_long_prompts=True,  # TODO:
+                returned_embeddings_type=ReturnedEmbeddingsType.PENULTIMATE_HIDDEN_STATES_NON_NORMALIZED, # TODO: clip skip
+                requires_pooled=True,
+            )
+
+            conjunction = Compel.parse_prompt_string(prompt)
+
+            if context.services.configuration.log_tokenization:
+                # TODO: better logging for and syntax
+                for prompt_obj in conjunction.prompts:
+                    log_tokenization_for_prompt_object(prompt_obj, tokenizer)
+
+            # TODO: ask for optimizations? to not run text_encoder twice
+            c, options = compel.build_conditioning_tensor_for_conjunction(conjunction)
+            if get_pooled:
+                c_pooled = compel.conditioning_provider.get_pooled_embeddings([prompt])
+            else:
+                c_pooled = None
+
+            ec = InvokeAIDiffuserComponent.ExtraConditioningInfo(
+                tokens_count_including_eos_bos=get_max_token_count(tokenizer, conjunction),
+                cross_attention_control_args=options.get("cross_attention_control", None),
+            )
+
+        del tokenizer
+        del text_encoder
+        del tokenizer_info
+        del text_encoder_info
+
+        c = c.detach().to("cpu")
+        if c_pooled is not None:
+            c_pooled = c_pooled.detach().to("cpu")
+
+        return c, c_pooled, ec
+
+class SDXLCompelPromptInvocation(BaseInvocation, SDXLPromptInvocationBase):
+    """Parse prompt using compel package to conditioning."""
+
+    type: Literal["sdxl_compel_prompt"] = "sdxl_compel_prompt"
+
+    prompt: str = Field(default="", description="Prompt")
+    style: str = Field(default="", description="Style prompt")
+    original_width: int = Field(1024, description="")
+    original_height: int = Field(1024, description="")
+    crop_top: int = Field(0, description="")
+    crop_left: int = Field(0, description="")
+    target_width: int = Field(1024, description="")
+    target_height: int = Field(1024, description="")
+    clip: ClipField = Field(None, description="Clip to use")
+    clip2: ClipField = Field(None, description="Clip2 to use")
+
+    # Schema customisation
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "SDXL Prompt (Compel)",
+                "tags": ["prompt", "compel"],
+                "type_hints": {
+                    "model": "model"
+                }
+            },
+        }
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> CompelOutput:
+        c1, c1_pooled, ec1 = self.run_clip_compel(context, self.clip, self.prompt, False)
+        if self.style.strip() == "":
+            c2, c2_pooled, ec2 = self.run_clip_compel(context, self.clip2, self.prompt, True)
+        else:
+            c2, c2_pooled, ec2 = self.run_clip_compel(context, self.clip2, self.style, True)
+
+        original_size = (self.original_height, self.original_width)
+        crop_coords = (self.crop_top, self.crop_left)
+        target_size = (self.target_height, self.target_width)
+
+        add_time_ids = torch.tensor([
+            original_size + crop_coords + target_size
+        ])
+
+        conditioning_data = ConditioningFieldData(
+            conditionings=[
+                SDXLConditioningInfo(
+                    embeds=torch.cat([c1, c2], dim=-1),
+                    pooled_embeds=c2_pooled,
+                    add_time_ids=add_time_ids,
+                    extra_conditioning=ec1,
+                )
+            ]
+        )
+
+        conditioning_name = f"{context.graph_execution_state_id}_{self.id}_conditioning"
+        context.services.latents.save(conditioning_name, conditioning_data)
+
+        return CompelOutput(
+            conditioning=ConditioningField(
+                conditioning_name=conditioning_name,
+            ),
+        )
+
+class SDXLRefinerCompelPromptInvocation(BaseInvocation, SDXLPromptInvocationBase):
+    """Parse prompt using compel package to conditioning."""
+
+    type: Literal["sdxl_refiner_compel_prompt"] = "sdxl_refiner_compel_prompt"
+
+    style: str = Field(default="", description="Style prompt") # TODO: ?
+    original_width: int = Field(1024, description="")
+    original_height: int = Field(1024, description="")
+    crop_top: int = Field(0, description="")
+    crop_left: int = Field(0, description="")
+    aesthetic_score: float = Field(6.0, description="")
+    clip2: ClipField = Field(None, description="Clip to use")
+
+    # Schema customisation
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "SDXL Refiner Prompt (Compel)",
+                "tags": ["prompt", "compel"],
+                "type_hints": {
+                    "model": "model"
+                }
+            },
+        }
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> CompelOutput:
+        c2, c2_pooled, ec2 = self.run_clip_compel(context, self.clip2, self.style, True)
+
+        original_size = (self.original_height, self.original_width)
+        crop_coords = (self.crop_top, self.crop_left)
+
+        add_time_ids = torch.tensor([
+            original_size + crop_coords + (self.aesthetic_score,)
+        ])
+
+        conditioning_data = ConditioningFieldData(
+            conditionings=[
+                SDXLConditioningInfo(
+                    embeds=c2,
+                    pooled_embeds=c2_pooled,
+                    add_time_ids=add_time_ids,
+                    extra_conditioning=ec2, # or None
+                )
+            ]
+        )
+
+        conditioning_name = f"{context.graph_execution_state_id}_{self.id}_conditioning"
+        context.services.latents.save(conditioning_name, conditioning_data)
+
+        return CompelOutput(
+            conditioning=ConditioningField(
+                conditioning_name=conditioning_name,
+            ),
+        )
+
+class SDXLRawPromptInvocation(BaseInvocation, SDXLPromptInvocationBase):
+    """Pass unmodified prompt to conditioning without compel processing."""
+
+    type: Literal["sdxl_raw_prompt"] = "sdxl_raw_prompt"
+
+    prompt: str = Field(default="", description="Prompt")
+    style: str = Field(default="", description="Style prompt")
+    original_width: int = Field(1024, description="")
+    original_height: int = Field(1024, description="")
+    crop_top: int = Field(0, description="")
+    crop_left: int = Field(0, description="")
+    target_width: int = Field(1024, description="")
+    target_height: int = Field(1024, description="")
+    clip: ClipField = Field(None, description="Clip to use")
+    clip2: ClipField = Field(None, description="Clip2 to use")
+
+    # Schema customisation
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "SDXL Prompt (Raw)",
+                "tags": ["prompt", "compel"],
+                "type_hints": {
+                    "model": "model"
+                }
+            },
+        }
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> CompelOutput:
+        c1, c1_pooled, ec1 = self.run_clip_raw(context, self.clip, self.prompt, False)
+        if self.style.strip() == "":
+            c2, c2_pooled, ec2 = self.run_clip_raw(context, self.clip2, self.prompt, True)
+        else:
+            c2, c2_pooled, ec2 = self.run_clip_raw(context, self.clip2, self.style, True)
+
+        original_size = (self.original_height, self.original_width)
+        crop_coords = (self.crop_top, self.crop_left)
+        target_size = (self.target_height, self.target_width)
+
+        add_time_ids = torch.tensor([
+            original_size + crop_coords + target_size
+        ])
+
+        conditioning_data = ConditioningFieldData(
+            conditionings=[
+                SDXLConditioningInfo(
+                    embeds=torch.cat([c1, c2], dim=-1),
+                    pooled_embeds=c2_pooled,
+                    add_time_ids=add_time_ids,
+                    extra_conditioning=ec1,
+                )
+            ]
+        )
+
+        conditioning_name = f"{context.graph_execution_state_id}_{self.id}_conditioning"
+        context.services.latents.save(conditioning_name, conditioning_data)
+
+        return CompelOutput(
+            conditioning=ConditioningField(
+                conditioning_name=conditioning_name,
+            ),
+        )
+
+class SDXLRefinerRawPromptInvocation(BaseInvocation, SDXLPromptInvocationBase):
+    """Parse prompt using compel package to conditioning."""
+
+    type: Literal["sdxl_refiner_raw_prompt"] = "sdxl_refiner_raw_prompt"
+
+    style: str = Field(default="", description="Style prompt") # TODO: ?
+    original_width: int = Field(1024, description="")
+    original_height: int = Field(1024, description="")
+    crop_top: int = Field(0, description="")
+    crop_left: int = Field(0, description="")
+    aesthetic_score: float = Field(6.0, description="")
+    clip2: ClipField = Field(None, description="Clip to use")
+
+    # Schema customisation
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "SDXL Refiner Prompt (Raw)",
+                "tags": ["prompt", "compel"],
+                "type_hints": {
+                    "model": "model"
+                }
+            },
+        }
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> CompelOutput:
+        c2, c2_pooled, ec2 = self.run_clip_raw(context, self.clip2, self.style, True)
+
+        original_size = (self.original_height, self.original_width)
+        crop_coords = (self.crop_top, self.crop_left)
+
+        add_time_ids = torch.tensor([
+            original_size + crop_coords + (self.aesthetic_score,)
+        ])
+
+        conditioning_data = ConditioningFieldData(
+            conditionings=[
+                SDXLConditioningInfo(
+                    embeds=c2,
+                    pooled_embeds=c2_pooled,
+                    add_time_ids=add_time_ids,
+                    extra_conditioning=ec2, # or None
+                )
+            ]
+        )
+
+        conditioning_name = f"{context.graph_execution_state_id}_{self.id}_conditioning"
+        context.services.latents.save(conditioning_name, conditioning_data)
+
+        return CompelOutput(
+            conditioning=ConditioningField(
+                conditioning_name=conditioning_name,
+            ),
+        )
+
+
 class ClipSkipInvocationOutput(BaseInvocationOutput):
     """Clip skip node output"""
     type: Literal["clip_skip_output"] = "clip_skip_output"
@@ -144,6 +571,14 @@ class ClipSkipInvocation(BaseInvocation):
     clip: ClipField = Field(None, description="Clip to use")
     skipped_layers: int = Field(0, description="Number of layers to skip in text_encoder")
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "CLIP Skip",
+                "tags": ["clip", "skip"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ClipSkipInvocationOutput:
         self.clip.skipped_layers += self.skipped_layers
         return ClipSkipInvocationOutput(

invokeai/app/invocations/controlnet_image_processors.py

@@ -1,42 +1,25 @@
 # Invocations for ControlNet image preprocessors
 # initial implementation by Gregg Helt, 2023
 # heavily leverages controlnet_aux package: https://github.com/patrickvonplaten/controlnet_aux
-from builtins import float, bool
+from builtins import bool, float
+from typing import Dict, List, Literal, Optional, Union
 
 import cv2
 import numpy as np
-from typing import Literal, Optional, Union, List, Dict
+from controlnet_aux import (CannyDetector, ContentShuffleDetector, HEDdetector,
+                            LeresDetector, LineartAnimeDetector,
+                            LineartDetector, MediapipeFaceDetector,
+                            MidasDetector, MLSDdetector, NormalBaeDetector,
+                            OpenposeDetector, PidiNetDetector, SamDetector,
+                            ZoeDetector)
+from controlnet_aux.util import HWC3, ade_palette
 from PIL import Image
 from pydantic import BaseModel, Field, validator
 
-from ..models.image import ImageField, ImageCategory, ResourceOrigin
-from .baseinvocation import (
-    BaseInvocation,
-    BaseInvocationOutput,
-    InvocationContext,
-    InvocationConfig,
-)
-
-from controlnet_aux import (
-    CannyDetector,
-    HEDdetector,
-    LineartDetector,
-    LineartAnimeDetector,
-    MidasDetector,
-    MLSDdetector,
-    NormalBaeDetector,
-    OpenposeDetector,
-    PidiNetDetector,
-    ContentShuffleDetector,
-    ZoeDetector,
-    MediapipeFaceDetector,
-    SamDetector,
-    LeresDetector,
-)
-
-from controlnet_aux.util import HWC3, ade_palette
-
-
+from ...backend.model_management import BaseModelType, ModelType
+from ..models.image import ImageCategory, ImageField, ResourceOrigin
+from .baseinvocation import (BaseInvocation, BaseInvocationOutput,
+                             InvocationConfig, InvocationContext)
 from .image import ImageOutput, PILInvocationConfig
 
 CONTROLNET_DEFAULT_MODELS = [
@@ -74,66 +57,83 @@ CONTROLNET_DEFAULT_MODELS = [
     "lllyasviel/control_v11e_sd15_ip2p",
     "lllyasviel/control_v11f1e_sd15_tile",
 
-     #################################################
-     #  thibaud sd v2.1 models (ControlNet v1.0? or v1.1?
-     ##################################################
-     "thibaud/controlnet-sd21-openpose-diffusers",
-     "thibaud/controlnet-sd21-canny-diffusers",
-     "thibaud/controlnet-sd21-depth-diffusers",
-     "thibaud/controlnet-sd21-scribble-diffusers",
-     "thibaud/controlnet-sd21-hed-diffusers",
-     "thibaud/controlnet-sd21-zoedepth-diffusers",
-     "thibaud/controlnet-sd21-color-diffusers",
-     "thibaud/controlnet-sd21-openposev2-diffusers",
-     "thibaud/controlnet-sd21-lineart-diffusers",
-     "thibaud/controlnet-sd21-normalbae-diffusers",
-     "thibaud/controlnet-sd21-ade20k-diffusers",
+    #################################################
+    #  thibaud sd v2.1 models (ControlNet v1.0? or v1.1?
+    ##################################################
+    "thibaud/controlnet-sd21-openpose-diffusers",
+    "thibaud/controlnet-sd21-canny-diffusers",
+    "thibaud/controlnet-sd21-depth-diffusers",
+    "thibaud/controlnet-sd21-scribble-diffusers",
+    "thibaud/controlnet-sd21-hed-diffusers",
+    "thibaud/controlnet-sd21-zoedepth-diffusers",
+    "thibaud/controlnet-sd21-color-diffusers",
+    "thibaud/controlnet-sd21-openposev2-diffusers",
+    "thibaud/controlnet-sd21-lineart-diffusers",
+    "thibaud/controlnet-sd21-normalbae-diffusers",
+    "thibaud/controlnet-sd21-ade20k-diffusers",
 
-     ##############################################
-     #  ControlNetMediaPipeface, ControlNet v1.1
-     ##############################################
-     # ["CrucibleAI/ControlNetMediaPipeFace", "diffusion_sd15"],  # SD 1.5
-     #    diffusion_sd15 needs to be passed to from_pretrained() as subfolder arg
-     #    hacked t2l to split to model & subfolder if format is "model,subfolder"
-     "CrucibleAI/ControlNetMediaPipeFace,diffusion_sd15",  # SD 1.5
-     "CrucibleAI/ControlNetMediaPipeFace",  # SD 2.1?
+    ##############################################
+    #  ControlNetMediaPipeface, ControlNet v1.1
+    ##############################################
+    # ["CrucibleAI/ControlNetMediaPipeFace", "diffusion_sd15"],  # SD 1.5
+    #    diffusion_sd15 needs to be passed to from_pretrained() as subfolder arg
+    #    hacked t2l to split to model & subfolder if format is "model,subfolder"
+    "CrucibleAI/ControlNetMediaPipeFace,diffusion_sd15",  # SD 1.5
+    "CrucibleAI/ControlNetMediaPipeFace",  # SD 2.1?
 ]
 
 CONTROLNET_NAME_VALUES = Literal[tuple(CONTROLNET_DEFAULT_MODELS)]
-CONTROLNET_MODE_VALUES = Literal[tuple(["balanced", "more_prompt", "more_control", "unbalanced"])]
-# crop and fill options not ready yet
-# CONTROLNET_RESIZE_VALUES = Literal[tuple(["just_resize", "crop_resize", "fill_resize"])]
+CONTROLNET_MODE_VALUES = Literal[tuple(
+    ["balanced", "more_prompt", "more_control", "unbalanced"])]
+CONTROLNET_RESIZE_VALUES = Literal[tuple(
+    ["just_resize", "crop_resize", "fill_resize", "just_resize_simple",])]
+
+
+class ControlNetModelField(BaseModel):
+    """ControlNet model field"""
+
+    model_name: str = Field(description="Name of the ControlNet model")
+    base_model: BaseModelType = Field(description="Base model")
 
 
 class ControlField(BaseModel):
     image: ImageField = Field(default=None, description="The control image")
-    control_model: Optional[str] = Field(default=None, description="The ControlNet model to use")
+    control_model: Optional[ControlNetModelField] = Field(
+        default=None, description="The ControlNet model to use")
     # control_weight: Optional[float] = Field(default=1, description="weight given to controlnet")
-    control_weight: Union[float, List[float]] = Field(default=1, description="The weight given to the ControlNet")
-    begin_step_percent: float = Field(default=0, ge=0, le=1,
-                                      description="When the ControlNet is first applied (% of total steps)")
-    end_step_percent: float = Field(default=1, ge=0, le=1,
-                                    description="When the ControlNet is last applied (% of total steps)")
-    control_mode: CONTROLNET_MODE_VALUES = Field(default="balanced", description="The control mode to use")
-    # resize_mode: CONTROLNET_RESIZE_VALUES = Field(default="just_resize", description="The resize mode to use")
+    control_weight: Union[float, List[float]] = Field(
+        default=1, description="The weight given to the ControlNet")
+    begin_step_percent: float = Field(
+        default=0, ge=0, le=1,
+        description="When the ControlNet is first applied (% of total steps)")
+    end_step_percent: float = Field(
+        default=1, ge=0, le=1,
+        description="When the ControlNet is last applied (% of total steps)")
+    control_mode: CONTROLNET_MODE_VALUES = Field(
+        default="balanced", description="The control mode to use")
+    resize_mode: CONTROLNET_RESIZE_VALUES = Field(
+        default="just_resize", description="The resize mode to use")
 
     @validator("control_weight")
-    def abs_le_one(cls, v):
-        """validate that all abs(values) are <=1"""
+    def validate_control_weight(cls, v):
+        """Validate that all control weights in the valid range"""
         if isinstance(v, list):
             for i in v:
-                if abs(i) > 1:
-                    raise ValueError('all abs(control_weight) must be <= 1')
+                if i < -1 or i > 2:
+                    raise ValueError(
+                        'Control weights must be within -1 to 2 range')
         else:
-            if abs(v) > 1:
-                raise ValueError('abs(control_weight) must be <= 1')
+            if v < -1 or v > 2:
+                raise ValueError('Control weights must be within -1 to 2 range')
         return v
+
     class Config:
         schema_extra = {
             "required": ["image", "control_model", "control_weight", "begin_step_percent", "end_step_percent"],
             "ui": {
                 "type_hints": {
                     "control_weight": "float",
+                    "control_model": "controlnet_model",
                     # "control_weight": "number",
                 }
             }
@@ -154,26 +154,28 @@ class ControlNetInvocation(BaseInvocation):
     type: Literal["controlnet"] = "controlnet"
     # Inputs
     image: ImageField = Field(default=None, description="The control image")
-    control_model: CONTROLNET_NAME_VALUES = Field(default="lllyasviel/sd-controlnet-canny",
+    control_model: ControlNetModelField = Field(default="lllyasviel/sd-controlnet-canny",
                                                   description="control model used")
     control_weight: Union[float, List[float]] = Field(default=1.0, description="The weight given to the ControlNet")
-    begin_step_percent: float = Field(default=0, ge=0, le=1,
+    begin_step_percent: float = Field(default=0, ge=-1, le=2,
                                       description="When the ControlNet is first applied (% of total steps)")
     end_step_percent: float = Field(default=1, ge=0, le=1,
                                     description="When the ControlNet is last applied (% of total steps)")
     control_mode: CONTROLNET_MODE_VALUES = Field(default="balanced", description="The control mode used")
+    resize_mode: CONTROLNET_RESIZE_VALUES = Field(default="just_resize", description="The resize mode used")
     # fmt: on
 
     class Config(InvocationConfig):
         schema_extra = {
             "ui": {
-                "tags": ["latents"],
+                "title": "ControlNet",
+                "tags": ["controlnet", "latents"],
                 "type_hints": {
-                  "model": "model",
-                  "control": "control",
-                  # "cfg_scale": "float",
-                  "cfg_scale": "number",
-                  "control_weight": "float",
+                    "model": "model",
+                    "control": "control",
+                    # "cfg_scale": "float",
+                    "cfg_scale": "number",
+                    "control_weight": "float",
                 }
             },
         }
@@ -187,6 +189,7 @@ class ControlNetInvocation(BaseInvocation):
                 begin_step_percent=self.begin_step_percent,
                 end_step_percent=self.end_step_percent,
                 control_mode=self.control_mode,
+                resize_mode=self.resize_mode,
             ),
         )
 
@@ -200,6 +203,13 @@ class ImageProcessorInvocation(BaseInvocation, PILInvocationConfig):
     image: ImageField = Field(default=None, description="The image to process")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Image Processor",
+                "tags": ["image", "processor"]
+            },
+        }
 
     def run_processor(self, image):
         # superclass just passes through image without processing
@@ -231,14 +241,15 @@ class ImageProcessorInvocation(BaseInvocation, PILInvocationConfig):
         return ImageOutput(
             image=processed_image_field,
             # width=processed_image.width,
-            width = image_dto.width,
+            width=image_dto.width,
             # height=processed_image.height,
-            height = image_dto.height,
+            height=image_dto.height,
             # mode=processed_image.mode,
         )
 
 
-class CannyImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig):
+class CannyImageProcessorInvocation(
+        ImageProcessorInvocation, PILInvocationConfig):
     """Canny edge detection for ControlNet"""
     # fmt: off
     type: Literal["canny_image_processor"] = "canny_image_processor"
@@ -247,13 +258,23 @@ class CannyImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfi
     high_threshold: int = Field(default=200, ge=0, le=255, description="The high threshold of the Canny pixel gradient (0-255)")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Canny Processor",
+                "tags": ["controlnet", "canny", "image", "processor"]
+            },
+        }
+
     def run_processor(self, image):
         canny_processor = CannyDetector()
-        processed_image = canny_processor(image, self.low_threshold, self.high_threshold)
+        processed_image = canny_processor(
+            image, self.low_threshold, self.high_threshold)
         return processed_image
 
 
-class HedImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig):
+class HedImageProcessorInvocation(
+        ImageProcessorInvocation, PILInvocationConfig):
     """Applies HED edge detection to image"""
     # fmt: off
     type: Literal["hed_image_processor"] = "hed_image_processor"
@@ -265,6 +286,14 @@ class HedImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig)
     scribble: bool = Field(default=False, description="Whether to use scribble mode")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Softedge(HED) Processor",
+                "tags": ["controlnet", "softedge", "hed", "image", "processor"]
+            },
+        }
+
     def run_processor(self, image):
         hed_processor = HEDdetector.from_pretrained("lllyasviel/Annotators")
         processed_image = hed_processor(image,
@@ -277,7 +306,8 @@ class HedImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig)
         return processed_image
 
 
-class LineartImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig):
+class LineartImageProcessorInvocation(
+        ImageProcessorInvocation, PILInvocationConfig):
     """Applies line art processing to image"""
     # fmt: off
     type: Literal["lineart_image_processor"] = "lineart_image_processor"
@@ -287,16 +317,25 @@ class LineartImageProcessorInvocation(ImageProcessorInvocation, PILInvocationCon
     coarse: bool = Field(default=False, description="Whether to use coarse mode")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Lineart Processor",
+                "tags": ["controlnet", "lineart", "image", "processor"]
+            },
+        }
+
     def run_processor(self, image):
-        lineart_processor = LineartDetector.from_pretrained("lllyasviel/Annotators")
-        processed_image = lineart_processor(image,
-                                            detect_resolution=self.detect_resolution,
-                                            image_resolution=self.image_resolution,
-                                            coarse=self.coarse)
+        lineart_processor = LineartDetector.from_pretrained(
+            "lllyasviel/Annotators")
+        processed_image = lineart_processor(
+            image, detect_resolution=self.detect_resolution,
+            image_resolution=self.image_resolution, coarse=self.coarse)
         return processed_image
 
 
-class LineartAnimeImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig):
+class LineartAnimeImageProcessorInvocation(
+        ImageProcessorInvocation, PILInvocationConfig):
     """Applies line art anime processing to image"""
     # fmt: off
     type: Literal["lineart_anime_image_processor"] = "lineart_anime_image_processor"
@@ -305,8 +344,17 @@ class LineartAnimeImageProcessorInvocation(ImageProcessorInvocation, PILInvocati
     image_resolution: int = Field(default=512, ge=0, description="The pixel resolution for the output image")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Lineart Anime Processor",
+                "tags": ["controlnet", "lineart", "anime", "image", "processor"]
+            },
+        }
+
     def run_processor(self, image):
-        processor = LineartAnimeDetector.from_pretrained("lllyasviel/Annotators")
+        processor = LineartAnimeDetector.from_pretrained(
+            "lllyasviel/Annotators")
         processed_image = processor(image,
                                     detect_resolution=self.detect_resolution,
                                     image_resolution=self.image_resolution,
@@ -314,7 +362,8 @@ class LineartAnimeImageProcessorInvocation(ImageProcessorInvocation, PILInvocati
         return processed_image
 
 
-class OpenposeImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig):
+class OpenposeImageProcessorInvocation(
+        ImageProcessorInvocation, PILInvocationConfig):
     """Applies Openpose processing to image"""
     # fmt: off
     type: Literal["openpose_image_processor"] = "openpose_image_processor"
@@ -324,17 +373,26 @@ class OpenposeImageProcessorInvocation(ImageProcessorInvocation, PILInvocationCo
     image_resolution: int = Field(default=512, ge=0, description="The pixel resolution for the output image")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Openpose Processor",
+                "tags": ["controlnet", "openpose", "image", "processor"]
+            },
+        }
+
     def run_processor(self, image):
-        openpose_processor = OpenposeDetector.from_pretrained("lllyasviel/Annotators")
-        processed_image = openpose_processor(image,
-                                             detect_resolution=self.detect_resolution,
-                                             image_resolution=self.image_resolution,
-                                             hand_and_face=self.hand_and_face,
-                                             )
+        openpose_processor = OpenposeDetector.from_pretrained(
+            "lllyasviel/Annotators")
+        processed_image = openpose_processor(
+            image, detect_resolution=self.detect_resolution,
+            image_resolution=self.image_resolution,
+            hand_and_face=self.hand_and_face,)
         return processed_image
 
 
-class MidasDepthImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig):
+class MidasDepthImageProcessorInvocation(
+        ImageProcessorInvocation, PILInvocationConfig):
     """Applies Midas depth processing to image"""
     # fmt: off
     type: Literal["midas_depth_image_processor"] = "midas_depth_image_processor"
@@ -345,6 +403,14 @@ class MidasDepthImageProcessorInvocation(ImageProcessorInvocation, PILInvocation
     # depth_and_normal: bool = Field(default=False, description="whether to use depth and normal mode")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Midas (Depth) Processor",
+                "tags": ["controlnet", "midas", "depth", "image", "processor"]
+            },
+        }
+
     def run_processor(self, image):
         midas_processor = MidasDetector.from_pretrained("lllyasviel/Annotators")
         processed_image = midas_processor(image,
@@ -356,7 +422,8 @@ class MidasDepthImageProcessorInvocation(ImageProcessorInvocation, PILInvocation
         return processed_image
 
 
-class NormalbaeImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig):
+class NormalbaeImageProcessorInvocation(
+        ImageProcessorInvocation, PILInvocationConfig):
     """Applies NormalBae processing to image"""
     # fmt: off
     type: Literal["normalbae_image_processor"] = "normalbae_image_processor"
@@ -365,15 +432,25 @@ class NormalbaeImageProcessorInvocation(ImageProcessorInvocation, PILInvocationC
     image_resolution: int = Field(default=512, ge=0, description="The pixel resolution for the output image")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Normal BAE Processor",
+                "tags": ["controlnet", "normal", "bae", "image", "processor"]
+            },
+        }
+
     def run_processor(self, image):
-        normalbae_processor = NormalBaeDetector.from_pretrained("lllyasviel/Annotators")
-        processed_image = normalbae_processor(image,
-                                              detect_resolution=self.detect_resolution,
-                                              image_resolution=self.image_resolution)
+        normalbae_processor = NormalBaeDetector.from_pretrained(
+            "lllyasviel/Annotators")
+        processed_image = normalbae_processor(
+            image, detect_resolution=self.detect_resolution,
+            image_resolution=self.image_resolution)
         return processed_image
 
 
-class MlsdImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig):
+class MlsdImageProcessorInvocation(
+        ImageProcessorInvocation, PILInvocationConfig):
     """Applies MLSD processing to image"""
     # fmt: off
     type: Literal["mlsd_image_processor"] = "mlsd_image_processor"
@@ -384,17 +461,25 @@ class MlsdImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig
     thr_d: float = Field(default=0.1, ge=0, description="MLSD parameter `thr_d`")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "MLSD Processor",
+                "tags": ["controlnet", "mlsd", "image", "processor"]
+            },
+        }
+
     def run_processor(self, image):
         mlsd_processor = MLSDdetector.from_pretrained("lllyasviel/Annotators")
-        processed_image = mlsd_processor(image,
-                                         detect_resolution=self.detect_resolution,
-                                         image_resolution=self.image_resolution,
-                                         thr_v=self.thr_v,
-                                         thr_d=self.thr_d)
+        processed_image = mlsd_processor(
+            image, detect_resolution=self.detect_resolution,
+            image_resolution=self.image_resolution, thr_v=self.thr_v,
+            thr_d=self.thr_d)
         return processed_image
 
 
-class PidiImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig):
+class PidiImageProcessorInvocation(
+        ImageProcessorInvocation, PILInvocationConfig):
     """Applies PIDI processing to image"""
     # fmt: off
     type: Literal["pidi_image_processor"] = "pidi_image_processor"
@@ -405,17 +490,26 @@ class PidiImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig
     scribble: bool = Field(default=False, description="Whether to use scribble mode")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "PIDI Processor",
+                "tags": ["controlnet", "pidi", "image", "processor"]
+            },
+        }
+
     def run_processor(self, image):
-        pidi_processor = PidiNetDetector.from_pretrained("lllyasviel/Annotators")
-        processed_image = pidi_processor(image,
-                                         detect_resolution=self.detect_resolution,
-                                         image_resolution=self.image_resolution,
-                                         safe=self.safe,
-                                         scribble=self.scribble)
+        pidi_processor = PidiNetDetector.from_pretrained(
+            "lllyasviel/Annotators")
+        processed_image = pidi_processor(
+            image, detect_resolution=self.detect_resolution,
+            image_resolution=self.image_resolution, safe=self.safe,
+            scribble=self.scribble)
         return processed_image
 
 
-class ContentShuffleImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig):
+class ContentShuffleImageProcessorInvocation(
+        ImageProcessorInvocation, PILInvocationConfig):
     """Applies content shuffle processing to image"""
     # fmt: off
     type: Literal["content_shuffle_image_processor"] = "content_shuffle_image_processor"
@@ -427,6 +521,14 @@ class ContentShuffleImageProcessorInvocation(ImageProcessorInvocation, PILInvoca
     f: Optional[int] = Field(default=256, ge=0, description="Content shuffle `f` parameter")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Content Shuffle Processor",
+                "tags": ["controlnet", "contentshuffle", "image", "processor"]
+            },
+        }
+
     def run_processor(self, image):
         content_shuffle_processor = ContentShuffleDetector()
         processed_image = content_shuffle_processor(image,
@@ -440,19 +542,30 @@ class ContentShuffleImageProcessorInvocation(ImageProcessorInvocation, PILInvoca
 
 
 # should work with controlnet_aux >= 0.0.4 and timm <= 0.6.13
-class ZoeDepthImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig):
+class ZoeDepthImageProcessorInvocation(
+        ImageProcessorInvocation, PILInvocationConfig):
     """Applies Zoe depth processing to image"""
     # fmt: off
     type: Literal["zoe_depth_image_processor"] = "zoe_depth_image_processor"
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Zoe (Depth) Processor",
+                "tags": ["controlnet", "zoe", "depth", "image", "processor"]
+            },
+        }
+
     def run_processor(self, image):
-        zoe_depth_processor = ZoeDetector.from_pretrained("lllyasviel/Annotators")
+        zoe_depth_processor = ZoeDetector.from_pretrained(
+            "lllyasviel/Annotators")
         processed_image = zoe_depth_processor(image)
         return processed_image
 
 
-class MediapipeFaceProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig):
+class MediapipeFaceProcessorInvocation(
+        ImageProcessorInvocation, PILInvocationConfig):
     """Applies mediapipe face processing to image"""
     # fmt: off
     type: Literal["mediapipe_face_processor"] = "mediapipe_face_processor"
@@ -461,16 +574,27 @@ class MediapipeFaceProcessorInvocation(ImageProcessorInvocation, PILInvocationCo
     min_confidence: float = Field(default=0.5, ge=0, le=1, description="Minimum confidence for face detection")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Mediapipe Processor",
+                "tags": ["controlnet", "mediapipe", "image", "processor"]
+            },
+        }
+
     def run_processor(self, image):
         # MediaPipeFaceDetector throws an error if image has alpha channel
         #     so convert to RGB if needed
         if image.mode == 'RGBA':
             image = image.convert('RGB')
         mediapipe_face_processor = MediapipeFaceDetector()
-        processed_image = mediapipe_face_processor(image, max_faces=self.max_faces, min_confidence=self.min_confidence)
+        processed_image = mediapipe_face_processor(
+            image, max_faces=self.max_faces, min_confidence=self.min_confidence)
         return processed_image
 
-class LeresImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig):
+
+class LeresImageProcessorInvocation(
+        ImageProcessorInvocation, PILInvocationConfig):
     """Applies leres processing to image"""
     # fmt: off
     type: Literal["leres_image_processor"] = "leres_image_processor"
@@ -482,18 +606,25 @@ class LeresImageProcessorInvocation(ImageProcessorInvocation, PILInvocationConfi
     image_resolution: int = Field(default=512, ge=0, description="The pixel resolution for the output image")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Leres (Depth) Processor",
+                "tags": ["controlnet", "leres", "depth", "image", "processor"]
+            },
+        }
+
     def run_processor(self, image):
         leres_processor = LeresDetector.from_pretrained("lllyasviel/Annotators")
-        processed_image = leres_processor(image,
-                                          thr_a=self.thr_a,
-                                          thr_b=self.thr_b,
-                                          boost=self.boost,
-                                          detect_resolution=self.detect_resolution,
-                                          image_resolution=self.image_resolution)
+        processed_image = leres_processor(
+            image, thr_a=self.thr_a, thr_b=self.thr_b, boost=self.boost,
+            detect_resolution=self.detect_resolution,
+            image_resolution=self.image_resolution)
         return processed_image
 
 
-class TileResamplerProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig):
+class TileResamplerProcessorInvocation(
+        ImageProcessorInvocation, PILInvocationConfig):
 
     # fmt: off
     type: Literal["tile_image_processor"] = "tile_image_processor"
@@ -502,6 +633,14 @@ class TileResamplerProcessorInvocation(ImageProcessorInvocation, PILInvocationCo
     down_sampling_rate: float = Field(default=1.0, ge=1.0, le=8.0, description="Down sampling rate")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Tile Resample Processor",
+                "tags": ["controlnet", "tile", "resample", "image", "processor"]
+            },
+        }
+
     # tile_resample copied from sd-webui-controlnet/scripts/processor.py
     def tile_resample(self,
                       np_img: np.ndarray,
@@ -520,28 +659,33 @@ class TileResamplerProcessorInvocation(ImageProcessorInvocation, PILInvocationCo
     def run_processor(self, img):
         np_img = np.array(img, dtype=np.uint8)
         processed_np_image = self.tile_resample(np_img,
-                                                #res=self.tile_size,
+                                                # res=self.tile_size,
                                                 down_sampling_rate=self.down_sampling_rate
                                                 )
         processed_image = Image.fromarray(processed_np_image)
         return processed_image
 
 
-
-
-class SegmentAnythingProcessorInvocation(ImageProcessorInvocation, PILInvocationConfig):
+class SegmentAnythingProcessorInvocation(
+        ImageProcessorInvocation, PILInvocationConfig):
     """Applies segment anything processing to image"""
     # fmt: off
     type: Literal["segment_anything_processor"] = "segment_anything_processor"
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {"ui": {"title": "Segment Anything Processor", "tags": [
+            "controlnet", "segment", "anything", "sam", "image", "processor"]}, }
+
     def run_processor(self, image):
         # segment_anything_processor = SamDetector.from_pretrained("ybelkada/segment-anything", subfolder="checkpoints")
-        segment_anything_processor = SamDetectorReproducibleColors.from_pretrained("ybelkada/segment-anything", subfolder="checkpoints")
+        segment_anything_processor = SamDetectorReproducibleColors.from_pretrained(
+            "ybelkada/segment-anything", subfolder="checkpoints")
         np_img = np.array(image, dtype=np.uint8)
         processed_image = segment_anything_processor(np_img)
         return processed_image
 
+
 class SamDetectorReproducibleColors(SamDetector):
 
     # overriding SamDetector.show_anns() method to use reproducible colors for segmentation image
@@ -553,7 +697,8 @@ class SamDetectorReproducibleColors(SamDetector):
             return
         sorted_anns = sorted(anns, key=(lambda x: x['area']), reverse=True)
         h, w = anns[0]['segmentation'].shape
-        final_img = Image.fromarray(np.zeros((h, w, 3), dtype=np.uint8), mode="RGB")
+        final_img = Image.fromarray(
+            np.zeros((h, w, 3), dtype=np.uint8), mode="RGB")
         palette = ade_palette()
         for i, ann in enumerate(sorted_anns):
             m = ann['segmentation']
@@ -561,5 +706,8 @@ class SamDetectorReproducibleColors(SamDetector):
             # doing modulo just in case number of annotated regions exceeds number of colors in palette
             ann_color = palette[i % len(palette)]
             img[:, :] = ann_color
-            final_img.paste(Image.fromarray(img, mode="RGB"), (0, 0), Image.fromarray(np.uint8(m * 255)))
+            final_img.paste(
+                Image.fromarray(img, mode="RGB"),
+                (0, 0),
+                Image.fromarray(np.uint8(m * 255)))
         return np.array(final_img, dtype=np.uint8)

invokeai/app/invocations/cv.py

@@ -35,6 +35,14 @@ class CvInpaintInvocation(BaseInvocation, CvInvocationConfig):
     mask: ImageField = Field(default=None, description="The mask to use when inpainting")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "OpenCV Inpaint",
+                "tags": ["opencv", "inpaint"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
         mask = context.services.images.get_pil_image(self.mask.image_name)

invokeai/app/invocations/generate.py

@@ -130,6 +130,7 @@ class InpaintInvocation(BaseInvocation):
         schema_extra = {
             "ui": {
                 "tags": ["stable-diffusion", "image"],
+                "title": "Inpaint"
             },
         }
 
@@ -146,48 +147,54 @@ class InpaintInvocation(BaseInvocation):
             source_node_id=source_node_id,
         )
 
-    def get_conditioning(self, context):
-        c, extra_conditioning_info = context.services.latents.get(self.positive_conditioning.conditioning_name)
-        uc, _ = context.services.latents.get(self.negative_conditioning.conditioning_name)
+    def get_conditioning(self, context, unet):
+        positive_cond_data = context.services.latents.get(self.positive_conditioning.conditioning_name)
+        c = positive_cond_data.conditionings[0].embeds.to(device=unet.device, dtype=unet.dtype)
+        extra_conditioning_info = positive_cond_data.conditionings[0].extra_conditioning
+
+        negative_cond_data = context.services.latents.get(self.negative_conditioning.conditioning_name)
+        uc = negative_cond_data.conditionings[0].embeds.to(device=unet.device, dtype=unet.dtype)
 
         return (uc, c, extra_conditioning_info)
 
     @contextmanager
     def load_model_old_way(self, context, scheduler):
-        unet_info = context.services.model_manager.get_model(**self.unet.unet.dict())
-        vae_info = context.services.model_manager.get_model(**self.vae.vae.dict())
+        def _lora_loader():
+            for lora in self.unet.loras:
+                lora_info = context.services.model_manager.get_model(
+                    **lora.dict(exclude={"weight"}), context=context,)
+                yield (lora_info.context.model, lora.weight)
+                del lora_info
+            return
+        
+        unet_info = context.services.model_manager.get_model(**self.unet.unet.dict(), context=context,)
+        vae_info = context.services.model_manager.get_model(**self.vae.vae.dict(), context=context,)
 
-        #unet = unet_info.context.model
-        #vae = vae_info.context.model
+        with vae_info as vae,\
+                ModelPatcher.apply_lora_unet(unet_info.context.model, _lora_loader()),\
+                unet_info as unet:
 
-        with ExitStack() as stack:
-            loras = [(stack.enter_context(context.services.model_manager.get_model(**lora.dict(exclude={"weight"}))), lora.weight) for lora in self.unet.loras]
+            device = context.services.model_manager.mgr.cache.execution_device
+            dtype = context.services.model_manager.mgr.cache.precision
 
-            with vae_info as vae,\
-                 unet_info as unet,\
-                 ModelPatcher.apply_lora_unet(unet, loras):
+            pipeline = StableDiffusionGeneratorPipeline(
+                vae=vae,
+                text_encoder=None,
+                tokenizer=None,
+                unet=unet,
+                scheduler=scheduler,
+                safety_checker=None,
+                feature_extractor=None,
+                requires_safety_checker=False,
+                precision="float16" if dtype == torch.float16 else "float32",
+                execution_device=device,
+            )
 
-                device = context.services.model_manager.mgr.cache.execution_device
-                dtype = context.services.model_manager.mgr.cache.precision
-
-                pipeline = StableDiffusionGeneratorPipeline(
-                    vae=vae,
-                    text_encoder=None,
-                    tokenizer=None,
-                    unet=unet,
-                    scheduler=scheduler,
-                    safety_checker=None,
-                    feature_extractor=None,
-                    requires_safety_checker=False,
-                    precision="float16" if dtype == torch.float16 else "float32",
-                    execution_device=device,
-                )
-
-                yield OldModelInfo(
-                    name=self.unet.unet.model_name,
-                    hash="<NO-HASH>",
-                    model=pipeline,
-                )
+            yield OldModelInfo(
+                name=self.unet.unet.model_name,
+                hash="<NO-HASH>",
+                model=pipeline,
+            )
 
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = (
@@ -207,7 +214,6 @@ class InpaintInvocation(BaseInvocation):
         )
         source_node_id = graph_execution_state.prepared_source_mapping[self.id]
 
-        conditioning = self.get_conditioning(context)
         scheduler = get_scheduler(
             context=context,
             scheduler_info=self.unet.scheduler,
@@ -215,6 +221,8 @@ class InpaintInvocation(BaseInvocation):
         )
 
         with self.load_model_old_way(context, scheduler) as model:
+            conditioning = self.get_conditioning(context, model.context.model.unet)
+
             outputs = Inpaint(model).generate(
                 conditioning=conditioning,
                 scheduler=scheduler,
@@ -226,21 +234,21 @@ class InpaintInvocation(BaseInvocation):
                 ),  # Shorthand for passing all of the parameters above manually
             )
 
-        # Outputs is an infinite iterator that will return a new InvokeAIGeneratorOutput object
-        # each time it is called. We only need the first one.
-        generator_output = next(outputs)
+            # Outputs is an infinite iterator that will return a new InvokeAIGeneratorOutput object
+            # each time it is called. We only need the first one.
+            generator_output = next(outputs)
 
-        image_dto = context.services.images.create(
-            image=generator_output.image,
-            image_origin=ResourceOrigin.INTERNAL,
-            image_category=ImageCategory.GENERAL,
-            session_id=context.graph_execution_state_id,
-            node_id=self.id,
-            is_intermediate=self.is_intermediate,
-        )
+            image_dto = context.services.images.create(
+                image=generator_output.image,
+                image_origin=ResourceOrigin.INTERNAL,
+                image_category=ImageCategory.GENERAL,
+                session_id=context.graph_execution_state_id,
+                node_id=self.id,
+                is_intermediate=self.is_intermediate,
+            )
 
-        return ImageOutput(
-            image=ImageField(image_name=image_dto.image_name),
-            width=image_dto.width,
-            height=image_dto.height,
-        )
+            return ImageOutput(
+                image=ImageField(image_name=image_dto.image_name),
+                width=image_dto.width,
+                height=image_dto.height,
+            )

invokeai/app/invocations/image.py

@@ -5,6 +5,7 @@ from typing import Literal, Optional
 import numpy
 from PIL import Image, ImageFilter, ImageOps, ImageChops
 from pydantic import BaseModel, Field
+from typing import Union
 
 from ..models.image import ImageCategory, ImageField, ResourceOrigin
 from .baseinvocation import (
@@ -70,6 +71,15 @@ class LoadImageInvocation(BaseInvocation):
         default=None, description="The image to load"
     )
     # fmt: on
+
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Load Image",
+                "tags": ["image", "load"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
 
@@ -90,6 +100,14 @@ class ShowImageInvocation(BaseInvocation):
         default=None, description="The image to show"
     )
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Show Image",
+                "tags": ["image", "show"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
         if image:
@@ -118,6 +136,14 @@ class ImageCropInvocation(BaseInvocation, PILInvocationConfig):
     height: int = Field(default=512, gt=0, description="The height of the crop rectangle")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Crop Image",
+                "tags": ["image", "crop"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
 
@@ -156,6 +182,14 @@ class ImagePasteInvocation(BaseInvocation, PILInvocationConfig):
     y:                     int = Field(default=0, description="The top y coordinate at which to paste the image")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Paste Image",
+                "tags": ["image", "paste"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         base_image = context.services.images.get_pil_image(self.base_image.image_name)
         image = context.services.images.get_pil_image(self.image.image_name)
@@ -206,6 +240,14 @@ class MaskFromAlphaInvocation(BaseInvocation, PILInvocationConfig):
     invert:      bool = Field(default=False, description="Whether or not to invert the mask")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Mask From Alpha",
+                "tags": ["image", "mask", "alpha"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> MaskOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
 
@@ -240,6 +282,14 @@ class ImageMultiplyInvocation(BaseInvocation, PILInvocationConfig):
     image2: Optional[ImageField]  = Field(default=None, description="The second image to multiply")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Multiply Images",
+                "tags": ["image", "multiply"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image1 = context.services.images.get_pil_image(self.image1.image_name)
         image2 = context.services.images.get_pil_image(self.image2.image_name)
@@ -276,6 +326,14 @@ class ImageChannelInvocation(BaseInvocation, PILInvocationConfig):
     channel: IMAGE_CHANNELS  = Field(default="A", description="The channel to get")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Image Channel",
+                "tags": ["image", "channel"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
 
@@ -311,6 +369,14 @@ class ImageConvertInvocation(BaseInvocation, PILInvocationConfig):
     mode: IMAGE_MODES  = Field(default="L", description="The mode to convert to")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Convert Image",
+                "tags": ["image", "convert"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
 
@@ -344,6 +410,14 @@ class ImageBlurInvocation(BaseInvocation, PILInvocationConfig):
     blur_type: Literal["gaussian", "box"] = Field(default="gaussian", description="The type of blur")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Blur Image",
+                "tags": ["image", "blur"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
 
@@ -398,11 +472,19 @@ class ImageResizeInvocation(BaseInvocation, PILInvocationConfig):
 
     # Inputs
     image: Optional[ImageField]  = Field(default=None, description="The image to resize")
-    width:                         int = Field(ge=64, multiple_of=8, description="The width to resize to (px)")
-    height:                        int = Field(ge=64, multiple_of=8, description="The height to resize to (px)")
+    width:                         Union[int, None] = Field(ge=64, multiple_of=8, description="The width to resize to (px)")
+    height:                        Union[int, None] = Field(ge=64, multiple_of=8, description="The height to resize to (px)")
     resample_mode:  PIL_RESAMPLING_MODES = Field(default="bicubic", description="The resampling mode")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Resize Image",
+                "tags": ["image", "resize"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
 
@@ -436,11 +518,19 @@ class ImageScaleInvocation(BaseInvocation, PILInvocationConfig):
     type: Literal["img_scale"] = "img_scale"
 
     # Inputs
-    image:       Optional[ImageField] = Field(default=None, description="The image to scale")
-    scale_factor:                  float = Field(gt=0, description="The factor by which to scale the image")
+    image:          Optional[ImageField] = Field(default=None, description="The image to scale")
+    scale_factor:        Optional[float] = Field(default=2.0, gt=0, description="The factor by which to scale the image")
     resample_mode:  PIL_RESAMPLING_MODES = Field(default="bicubic", description="The resampling mode")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Scale Image",
+                "tags": ["image", "scale"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
 
@@ -481,6 +571,14 @@ class ImageLerpInvocation(BaseInvocation, PILInvocationConfig):
     max: int = Field(default=255, ge=0, le=255, description="The maximum output value")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Image Linear Interpolation",
+                "tags": ["image", "linear", "interpolation", "lerp"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
 
@@ -517,6 +615,14 @@ class ImageInverseLerpInvocation(BaseInvocation, PILInvocationConfig):
     max: int = Field(default=255, ge=0, le=255, description="The maximum input value")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Image Inverse Linear Interpolation",
+                "tags": ["image", "linear", "interpolation", "inverse"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
 

invokeai/app/invocations/infill.py

@@ -14,6 +14,7 @@ from invokeai.backend.image_util.patchmatch import PatchMatch
 from ..models.image import ColorField, ImageCategory, ImageField, ResourceOrigin
 from .baseinvocation import (
     BaseInvocation,
+    InvocationConfig,
     InvocationContext,
 )
 
@@ -133,6 +134,14 @@ class InfillColorInvocation(BaseInvocation):
         description="The color to use to infill",
     )
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Color Infill",
+                "tags": ["image", "inpaint", "color", "infill"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
 
@@ -173,6 +182,14 @@ class InfillTileInvocation(BaseInvocation):
         default_factory=get_random_seed,
     )
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Tile Infill",
+                "tags": ["image", "inpaint", "tile", "infill"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
 
@@ -206,6 +223,14 @@ class InfillPatchMatchInvocation(BaseInvocation):
         default=None, description="The image to infill"
     )
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Patch Match Infill",
+                "tags": ["image", "inpaint", "patchmatch", "infill"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
 

invokeai/app/invocations/latent.py

@@ -1,5 +1,6 @@
 # Copyright (c) 2023 Kyle Schouviller (https://github.com/kyle0654)
 
+from contextlib import ExitStack
 from typing import List, Literal, Optional, Union
 
 import einops
@@ -9,9 +10,10 @@ from diffusers.image_processor import VaeImageProcessor
 from diffusers.schedulers import SchedulerMixin as Scheduler
 from pydantic import BaseModel, Field, validator
 
+from invokeai.app.invocations.metadata import CoreMetadata
 from invokeai.app.util.step_callback import stable_diffusion_step_callback
+from invokeai.backend.model_management.models.base import ModelType
 
-from ..models.image import ImageCategory, ImageField, ResourceOrigin
 from ...backend.model_management.lora import ModelPatcher
 from ...backend.stable_diffusion import PipelineIntermediateState
 from ...backend.stable_diffusion.diffusers_pipeline import (
@@ -20,13 +22,25 @@ from ...backend.stable_diffusion.diffusers_pipeline import (
 from ...backend.stable_diffusion.diffusion.shared_invokeai_diffusion import \
     PostprocessingSettings
 from ...backend.stable_diffusion.schedulers import SCHEDULER_MAP
-from ...backend.util.devices import torch_dtype
+from ...backend.util.devices import choose_torch_device, torch_dtype, choose_precision
+from ..models.image import ImageCategory, ImageField, ResourceOrigin
 from .baseinvocation import (BaseInvocation, BaseInvocationOutput,
                              InvocationConfig, InvocationContext)
 from .compel import ConditioningField
 from .controlnet_image_processors import ControlField
 from .image import ImageOutput
 from .model import ModelInfo, UNetField, VaeField
+from invokeai.app.util.controlnet_utils import prepare_control_image
+
+from diffusers.models.attention_processor import (
+    AttnProcessor2_0,
+    LoRAAttnProcessor2_0,
+    LoRAXFormersAttnProcessor,
+    XFormersAttnProcessor,
+)
+
+
+DEFAULT_PRECISION = choose_precision(choose_torch_device())
 
 
 class LatentsField(BaseModel):
@@ -70,16 +84,21 @@ def get_scheduler(
     scheduler_name: str,
 ) -> Scheduler:
     scheduler_class, scheduler_extra_config = SCHEDULER_MAP.get(
-        scheduler_name, SCHEDULER_MAP['ddim'])
+        scheduler_name, SCHEDULER_MAP['ddim']
+    )
     orig_scheduler_info = context.services.model_manager.get_model(
-        **scheduler_info.dict())
+        **scheduler_info.dict(), context=context,
+    )
     with orig_scheduler_info as orig_scheduler:
         scheduler_config = orig_scheduler.config
 
     if "_backup" in scheduler_config:
         scheduler_config = scheduler_config["_backup"]
-    scheduler_config = {**scheduler_config, **
-                        scheduler_extra_config, "_backup": scheduler_config}
+    scheduler_config = {
+        **scheduler_config,
+        **scheduler_extra_config,
+        "_backup": scheduler_config,
+    }
     scheduler = scheduler_class.from_config(scheduler_config)
 
     # hack copied over from generate.py
@@ -124,6 +143,7 @@ class TextToLatentsInvocation(BaseInvocation):
     class Config(InvocationConfig):
         schema_extra = {
             "ui": {
+                "title": "Text To Latents",
                 "tags": ["latents"],
                 "type_hints": {
                     "model": "model",
@@ -136,8 +156,11 @@ class TextToLatentsInvocation(BaseInvocation):
 
     # TODO: pass this an emitter method or something? or a session for dispatching?
     def dispatch_progress(
-            self, context: InvocationContext, source_node_id: str,
-            intermediate_state: PipelineIntermediateState) -> None:
+        self,
+        context: InvocationContext,
+        source_node_id: str,
+        intermediate_state: PipelineIntermediateState,
+    ) -> None:
         stable_diffusion_step_callback(
             context=context,
             intermediate_state=intermediate_state,
@@ -146,11 +169,17 @@ class TextToLatentsInvocation(BaseInvocation):
         )
 
     def get_conditioning_data(
-            self, context: InvocationContext, scheduler) -> ConditioningData:
-        c, extra_conditioning_info = context.services.latents.get(
-            self.positive_conditioning.conditioning_name)
-        uc, _ = context.services.latents.get(
-            self.negative_conditioning.conditioning_name)
+        self,
+        context: InvocationContext,
+        scheduler,
+        unet,
+    ) -> ConditioningData:
+        positive_cond_data = context.services.latents.get(self.positive_conditioning.conditioning_name)
+        c = positive_cond_data.conditionings[0].embeds.to(device=unet.device, dtype=unet.dtype)
+        extra_conditioning_info = positive_cond_data.conditionings[0].extra_conditioning
+
+        negative_cond_data = context.services.latents.get(self.negative_conditioning.conditioning_name)
+        uc = negative_cond_data.conditionings[0].embeds.to(device=unet.device, dtype=unet.dtype)
 
         conditioning_data = ConditioningData(
             unconditioned_embeddings=uc,
@@ -172,12 +201,15 @@ class TextToLatentsInvocation(BaseInvocation):
             eta=0.0,  # ddim_eta
 
             # for ancestral and sde schedulers
-            generator=torch.Generator(device=uc.device).manual_seed(0),
+            generator=torch.Generator(device=unet.device).manual_seed(0),
         )
         return conditioning_data
 
     def create_pipeline(
-            self, unet, scheduler) -> StableDiffusionGeneratorPipeline:
+        self,
+        unet,
+        scheduler,
+    ) -> StableDiffusionGeneratorPipeline:
         # TODO:
         # configure_model_padding(
         #    unet,
@@ -212,6 +244,7 @@ class TextToLatentsInvocation(BaseInvocation):
         model: StableDiffusionGeneratorPipeline,
         control_input: List[ControlField],
         latents_shape: List[int],
+        exit_stack: ExitStack,
         do_classifier_free_guidance: bool = True,
     ) -> List[ControlNetData]:
 
@@ -237,31 +270,26 @@ class TextToLatentsInvocation(BaseInvocation):
             control_data = []
             control_models = []
             for control_info in control_list:
-                # handle control models
-                if ("," in control_info.control_model):
-                    control_model_split = control_info.control_model.split(",")
-                    control_name = control_model_split[0]
-                    control_subfolder = control_model_split[1]
-                    print("Using HF model subfolders")
-                    print("    control_name: ", control_name)
-                    print("    control_subfolder: ", control_subfolder)
-                    control_model = ControlNetModel.from_pretrained(
-                        control_name, subfolder=control_subfolder,
-                        torch_dtype=model.unet.dtype).to(
-                        model.device)
-                else:
-                    control_model = ControlNetModel.from_pretrained(
-                        control_info.control_model, torch_dtype=model.unet.dtype).to(model.device)
+                control_model = exit_stack.enter_context(
+                    context.services.model_manager.get_model(
+                        model_name=control_info.control_model.model_name,
+                        model_type=ModelType.ControlNet,
+                        base_model=control_info.control_model.base_model,
+                        context=context,
+                    )
+                )
+
                 control_models.append(control_model)
                 control_image_field = control_info.image
                 input_image = context.services.images.get_pil_image(
-                    control_image_field.image_name)
+                    control_image_field.image_name
+                )
                 # self.image.image_type, self.image.image_name
                 # FIXME: still need to test with different widths, heights, devices, dtypes
                 #        and add in batch_size, num_images_per_prompt?
                 #        and do real check for classifier_free_guidance?
                 # prepare_control_image should return torch.Tensor of shape(batch_size, 3, height, width)
-                control_image = model.prepare_control_image(
+                control_image = prepare_control_image(
                     image=input_image,
                     do_classifier_free_guidance=do_classifier_free_guidance,
                     width=control_width_resize,
@@ -271,13 +299,19 @@ class TextToLatentsInvocation(BaseInvocation):
                     device=control_model.device,
                     dtype=control_model.dtype,
                     control_mode=control_info.control_mode,
+                    resize_mode=control_info.resize_mode,
                 )
                 control_item = ControlNetData(
-                    model=control_model, image_tensor=control_image,
+                    model=control_model,
+                    image_tensor=control_image,
                     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,)
+                    control_mode=control_info.control_mode,
+                    # any resizing needed should currently be happening in prepare_control_image(),
+                    #    but adding resize_mode to ControlNetData in case needed in the future
+                    resize_mode=control_info.resize_mode,
+                )
                 control_data.append(control_item)
                 # MultiControlNetModel has been refactored out, just need list[ControlNetData]
         return control_data
@@ -288,7 +322,8 @@ class TextToLatentsInvocation(BaseInvocation):
 
         # Get the source node id (we are invoking the prepared node)
         graph_execution_state = context.services.graph_execution_manager.get(
-            context.graph_execution_state_id)
+            context.graph_execution_state_id
+        )
         source_node_id = graph_execution_state.prepared_source_mapping[self.id]
 
         def step_callback(state: PipelineIntermediateState):
@@ -297,16 +332,21 @@ class TextToLatentsInvocation(BaseInvocation):
         def _lora_loader():
             for lora in self.unet.loras:
                 lora_info = context.services.model_manager.get_model(
-                    **lora.dict(exclude={"weight"}))
+                    **lora.dict(exclude={"weight"}), context=context,
+                )
                 yield (lora_info.context.model, lora.weight)
                 del lora_info
             return
 
         unet_info = context.services.model_manager.get_model(
-            **self.unet.unet.dict())
-        with ModelPatcher.apply_lora_unet(unet_info.context.model, _lora_loader()),\
+            **self.unet.unet.dict(), context=context,
+        )
+        with ExitStack() as exit_stack,\
+                ModelPatcher.apply_lora_unet(unet_info.context.model, _lora_loader()),\
                 unet_info as unet:
 
+            noise = noise.to(device=unet.device, dtype=unet.dtype)
+
             scheduler = get_scheduler(
                 context=context,
                 scheduler_info=self.unet.scheduler,
@@ -314,13 +354,14 @@ class TextToLatentsInvocation(BaseInvocation):
             )
 
             pipeline = self.create_pipeline(unet, scheduler)
-            conditioning_data = self.get_conditioning_data(context, scheduler)
+            conditioning_data = self.get_conditioning_data(context, scheduler, unet)
 
             control_data = self.prep_control_data(
                 model=pipeline, context=context, control_input=self.control,
                 latents_shape=noise.shape,
                 # do_classifier_free_guidance=(self.cfg_scale >= 1.0))
                 do_classifier_free_guidance=True,
+                exit_stack=exit_stack,
             )
 
             # TODO: Verify the noise is the right size
@@ -334,6 +375,7 @@ class TextToLatentsInvocation(BaseInvocation):
             )
 
         # https://discuss.huggingface.co/t/memory-usage-by-later-pipeline-stages/23699
+        result_latents = result_latents.to("cpu")
         torch.cuda.empty_cache()
 
         name = f'{context.graph_execution_state_id}__{self.id}'
@@ -357,6 +399,7 @@ class LatentsToLatentsInvocation(TextToLatentsInvocation):
     class Config(InvocationConfig):
         schema_extra = {
             "ui": {
+                "title": "Latent To Latents",
                 "tags": ["latents"],
                 "type_hints": {
                     "model": "model",
@@ -373,7 +416,8 @@ class LatentsToLatentsInvocation(TextToLatentsInvocation):
 
         # Get the source node id (we are invoking the prepared node)
         graph_execution_state = context.services.graph_execution_manager.get(
-            context.graph_execution_state_id)
+            context.graph_execution_state_id
+        )
         source_node_id = graph_execution_state.prepared_source_mapping[self.id]
 
         def step_callback(state: PipelineIntermediateState):
@@ -382,16 +426,22 @@ class LatentsToLatentsInvocation(TextToLatentsInvocation):
         def _lora_loader():
             for lora in self.unet.loras:
                 lora_info = context.services.model_manager.get_model(
-                    **lora.dict(exclude={"weight"}))
+                    **lora.dict(exclude={"weight"}), context=context,
+                )
                 yield (lora_info.context.model, lora.weight)
                 del lora_info
             return
 
         unet_info = context.services.model_manager.get_model(
-            **self.unet.unet.dict())
-        with ModelPatcher.apply_lora_unet(unet_info.context.model, _lora_loader()),\
+            **self.unet.unet.dict(), context=context,
+        )
+        with ExitStack() as exit_stack,\
+                ModelPatcher.apply_lora_unet(unet_info.context.model, _lora_loader()),\
                 unet_info as unet:
 
+            noise = noise.to(device=unet.device, dtype=unet.dtype)
+            latent = latent.to(device=unet.device, dtype=unet.dtype)
+
             scheduler = get_scheduler(
                 context=context,
                 scheduler_info=self.unet.scheduler,
@@ -399,18 +449,20 @@ class LatentsToLatentsInvocation(TextToLatentsInvocation):
             )
 
             pipeline = self.create_pipeline(unet, scheduler)
-            conditioning_data = self.get_conditioning_data(context, scheduler)
+            conditioning_data = self.get_conditioning_data(context, scheduler, unet)
 
             control_data = self.prep_control_data(
                 model=pipeline, context=context, control_input=self.control,
                 latents_shape=noise.shape,
                 # do_classifier_free_guidance=(self.cfg_scale >= 1.0))
                 do_classifier_free_guidance=True,
+                exit_stack=exit_stack,
             )
 
             # TODO: Verify the noise is the right size
             initial_latents = latent if self.strength < 1.0 else torch.zeros_like(
-                latent, device=unet.device, dtype=latent.dtype)
+                latent, device=unet.device, dtype=latent.dtype
+            )
 
             timesteps, _ = pipeline.get_img2img_timesteps(
                 self.steps,
@@ -429,6 +481,7 @@ class LatentsToLatentsInvocation(TextToLatentsInvocation):
             )
 
         # https://discuss.huggingface.co/t/memory-usage-by-later-pipeline-stages/23699
+        result_latents = result_latents.to("cpu")
         torch.cuda.empty_cache()
 
         name = f'{context.graph_execution_state_id}__{self.id}'
@@ -449,11 +502,14 @@ class LatentsToImageInvocation(BaseInvocation):
     tiled: bool = Field(
         default=False,
         description="Decode latents by overlaping tiles(less memory consumption)")
+    fp32: bool = Field(DEFAULT_PRECISION=='float32', description="Decode in full precision")
+    metadata: Optional[CoreMetadata] = Field(default=None, description="Optional core metadata to be written to the image")
 
     # Schema customisation
     class Config(InvocationConfig):
         schema_extra = {
             "ui": {
+                "title": "Latents To Image",
                 "tags": ["latents", "image"],
             },
         }
@@ -463,10 +519,36 @@ class LatentsToImageInvocation(BaseInvocation):
         latents = context.services.latents.get(self.latents.latents_name)
 
         vae_info = context.services.model_manager.get_model(
-            **self.vae.vae.dict(),
+            **self.vae.vae.dict(), context=context,
         )
 
         with vae_info as vae:
+            latents = latents.to(vae.device)
+            if self.fp32:
+                vae.to(dtype=torch.float32)
+
+                use_torch_2_0_or_xformers = isinstance(
+                    vae.decoder.mid_block.attentions[0].processor,
+                    (
+                        AttnProcessor2_0,
+                        XFormersAttnProcessor,
+                        LoRAXFormersAttnProcessor,
+                        LoRAAttnProcessor2_0,
+                    ),
+                )
+                # if xformers or torch_2_0 is used attention block does not need
+                # to be in float32 which can save lots of memory
+                if use_torch_2_0_or_xformers:
+                    vae.post_quant_conv.to(latents.dtype)
+                    vae.decoder.conv_in.to(latents.dtype)
+                    vae.decoder.mid_block.to(latents.dtype)
+                else:
+                    latents = latents.float()
+
+            else:
+                vae.to(dtype=torch.float16)
+                latents = latents.half()
+
             if self.tiled or context.services.configuration.tiled_decode:
                 vae.enable_tiling()
             else:
@@ -493,7 +575,8 @@ class LatentsToImageInvocation(BaseInvocation):
             image_category=ImageCategory.GENERAL,
             node_id=self.id,
             session_id=context.graph_execution_state_id,
-            is_intermediate=self.is_intermediate
+            is_intermediate=self.is_intermediate,
+            metadata=self.metadata.dict() if self.metadata else None,
         )
 
         return ImageOutput(
@@ -515,9 +598,9 @@ class ResizeLatentsInvocation(BaseInvocation):
     # Inputs
     latents: Optional[LatentsField] = Field(
         description="The latents to resize")
-    width:                         int = Field(
+    width:                         Union[int, None] = Field(default=512,
         ge=64, multiple_of=8, description="The width to resize to (px)")
-    height:                        int = Field(
+    height:                        Union[int, None] = Field(default=512,
         ge=64, multiple_of=8, description="The height to resize to (px)")
     mode:   LATENTS_INTERPOLATION_MODE = Field(
         default="bilinear", description="The interpolation mode")
@@ -525,15 +608,28 @@ class ResizeLatentsInvocation(BaseInvocation):
         default=False,
         description="Whether or not to antialias (applied in bilinear and bicubic modes only)")
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Resize Latents",
+                "tags": ["latents", "resize"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> LatentsOutput:
         latents = context.services.latents.get(self.latents.latents_name)
 
+        # TODO:
+        device=choose_torch_device()
+
         resized_latents = torch.nn.functional.interpolate(
-            latents, size=(self.height // 8, self.width // 8),
+            latents.to(device), size=(self.height // 8, self.width // 8),
             mode=self.mode, antialias=self.antialias
-            if self.mode in ["bilinear", "bicubic"] else False,)
+            if self.mode in ["bilinear", "bicubic"] else False,
+        )
 
         # https://discuss.huggingface.co/t/memory-usage-by-later-pipeline-stages/23699
+        resized_latents = resized_latents.to("cpu")
         torch.cuda.empty_cache()
 
         name = f"{context.graph_execution_state_id}__{self.id}"
@@ -558,16 +654,29 @@ class ScaleLatentsInvocation(BaseInvocation):
         default=False,
         description="Whether or not to antialias (applied in bilinear and bicubic modes only)")
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Scale Latents",
+                "tags": ["latents", "scale"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> LatentsOutput:
         latents = context.services.latents.get(self.latents.latents_name)
 
+        # TODO:
+        device=choose_torch_device()
+
         # resizing
         resized_latents = torch.nn.functional.interpolate(
-            latents, scale_factor=self.scale_factor, mode=self.mode,
+            latents.to(device), scale_factor=self.scale_factor, mode=self.mode,
             antialias=self.antialias
-            if self.mode in ["bilinear", "bicubic"] else False,)
+            if self.mode in ["bilinear", "bicubic"] else False,
+        )
 
         # https://discuss.huggingface.co/t/memory-usage-by-later-pipeline-stages/23699
+        resized_latents = resized_latents.to("cpu")
         torch.cuda.empty_cache()
 
         name = f"{context.graph_execution_state_id}__{self.id}"
@@ -587,12 +696,15 @@ class ImageToLatentsInvocation(BaseInvocation):
     tiled: bool = Field(
         default=False,
         description="Encode latents by overlaping tiles(less memory consumption)")
+    fp32: bool = Field(DEFAULT_PRECISION=='float32', description="Decode in full precision")
+
 
     # Schema customisation
     class Config(InvocationConfig):
         schema_extra = {
             "ui": {
-                "tags": ["latents", "image"],
+                "title": "Image To Latents",
+                "tags": ["latents", "image"]
             },
         }
 
@@ -605,7 +717,7 @@ class ImageToLatentsInvocation(BaseInvocation):
 
         #vae_info = context.services.model_manager.get_model(**self.vae.vae.dict())
         vae_info = context.services.model_manager.get_model(
-            **self.vae.vae.dict(),
+            **self.vae.vae.dict(), context=context,
         )
 
         image_tensor = image_resized_to_grid_as_tensor(image.convert("RGB"))
@@ -613,6 +725,32 @@ class ImageToLatentsInvocation(BaseInvocation):
             image_tensor = einops.rearrange(image_tensor, "c h w -> 1 c h w")
 
         with vae_info as vae:
+            orig_dtype = vae.dtype
+            if self.fp32:
+                vae.to(dtype=torch.float32)
+
+                use_torch_2_0_or_xformers = isinstance(
+                    vae.decoder.mid_block.attentions[0].processor,
+                    (
+                        AttnProcessor2_0,
+                        XFormersAttnProcessor,
+                        LoRAXFormersAttnProcessor,
+                        LoRAAttnProcessor2_0,
+                    ),
+                )
+                # if xformers or torch_2_0 is used attention block does not need
+                # to be in float32 which can save lots of memory
+                if use_torch_2_0_or_xformers:
+                    vae.post_quant_conv.to(orig_dtype)
+                    vae.decoder.conv_in.to(orig_dtype)
+                    vae.decoder.mid_block.to(orig_dtype)
+                #else:
+                #    latents = latents.float()
+
+            else:
+                vae.to(dtype=torch.float16)
+                #latents = latents.half()
+
             if self.tiled:
                 vae.enable_tiling()
             else:
@@ -626,9 +764,10 @@ class ImageToLatentsInvocation(BaseInvocation):
                     dtype=vae.dtype
                 )  # FIXME: uses torch.randn. make reproducible!
 
-            latents = 0.18215 * latents
+            latents = vae.config.scaling_factor * latents
+            latents = latents.to(dtype=orig_dtype)
 
         name = f"{context.graph_execution_state_id}__{self.id}"
-        # context.services.latents.set(name, latents)
+        latents = latents.to("cpu")
         context.services.latents.save(name, latents)
         return build_latents_output(latents_name=name, latents=latents)

invokeai/app/invocations/math.py

@@ -52,6 +52,14 @@ class AddInvocation(BaseInvocation, MathInvocationConfig):
     b: int = Field(default=0, description="The second number")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Add",
+                "tags": ["math", "add"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> IntOutput:
         return IntOutput(a=self.a + self.b)
 
@@ -65,6 +73,14 @@ class SubtractInvocation(BaseInvocation, MathInvocationConfig):
     b: int = Field(default=0, description="The second number")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Subtract",
+                "tags": ["math", "subtract"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> IntOutput:
         return IntOutput(a=self.a - self.b)
 
@@ -78,6 +94,14 @@ class MultiplyInvocation(BaseInvocation, MathInvocationConfig):
     b: int = Field(default=0, description="The second number")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Multiply",
+                "tags": ["math", "multiply"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> IntOutput:
         return IntOutput(a=self.a * self.b)
 
@@ -91,6 +115,14 @@ class DivideInvocation(BaseInvocation, MathInvocationConfig):
     b: int = Field(default=0, description="The second number")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Divide",
+                "tags": ["math", "divide"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> IntOutput:
         return IntOutput(a=int(self.a / self.b))
 
@@ -105,5 +137,14 @@ class RandomIntInvocation(BaseInvocation):
         default=np.iinfo(np.int32).max, description="The exclusive high value"
     )
     # fmt: on
+
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Random Integer",
+                "tags": ["math", "random", "integer"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> IntOutput:
         return IntOutput(a=np.random.randint(self.low, self.high))

invokeai/app/invocations/metadata.py

@@ -0,0 +1,132 @@
+from typing import Literal, Optional, Union
+
+from pydantic import BaseModel, Field
+
+from invokeai.app.invocations.baseinvocation import (BaseInvocation,
+                                                     BaseInvocationOutput, InvocationConfig,
+                                                     InvocationContext)
+from invokeai.app.invocations.controlnet_image_processors import ControlField
+from invokeai.app.invocations.model import (LoRAModelField, MainModelField,
+                                            VAEModelField)
+
+
+class LoRAMetadataField(BaseModel):
+    """LoRA metadata for an image generated in InvokeAI."""
+    lora: LoRAModelField = Field(description="The LoRA model")
+    weight: float = Field(description="The weight of the LoRA model")
+
+
+class CoreMetadata(BaseModel):
+    """Core generation metadata for an image generated in InvokeAI."""
+
+    generation_mode: str = Field(description="The generation mode that output this image",)
+    positive_prompt: str = Field(description="The positive prompt parameter")
+    negative_prompt: str = Field(description="The negative prompt parameter")
+    width: int = Field(description="The width parameter")
+    height: int = Field(description="The height parameter")
+    seed: int = Field(description="The seed used for noise generation")
+    rand_device: str = Field(description="The device used for random number generation")
+    cfg_scale: float = Field(description="The classifier-free guidance scale parameter")
+    steps: int = Field(description="The number of steps used for inference")
+    scheduler: str = Field(description="The scheduler used for inference")
+    clip_skip: int = Field(description="The number of skipped CLIP layers",)
+    model: MainModelField = Field(description="The main model used for inference")
+    controlnets: list[ControlField]= Field(description="The ControlNets used for inference")
+    loras: list[LoRAMetadataField] = Field(description="The LoRAs used for inference")
+    strength: Union[float, None] = Field(
+        default=None,
+        description="The strength used for latents-to-latents",
+    )
+    init_image: Union[str, None] = Field(
+        default=None, description="The name of the initial image"
+    )
+    vae: Union[VAEModelField, None] = Field(
+        default=None,
+        description="The VAE used for decoding, if the main model's default was not used",
+    )
+
+
+class ImageMetadata(BaseModel):
+    """An image's generation metadata"""
+
+    metadata: Optional[dict] = Field(
+        default=None,
+        description="The image's core metadata, if it was created in the Linear or Canvas UI",
+    )
+    graph: Optional[dict] = Field(
+        default=None, description="The graph that created the image"
+    )
+
+
+class MetadataAccumulatorOutput(BaseInvocationOutput):
+    """The output of the MetadataAccumulator node"""
+
+    type: Literal["metadata_accumulator_output"] = "metadata_accumulator_output"
+
+    metadata: CoreMetadata = Field(description="The core metadata for the image")
+
+
+class MetadataAccumulatorInvocation(BaseInvocation):
+    """Outputs a Core Metadata Object"""
+
+    type: Literal["metadata_accumulator"] = "metadata_accumulator"
+
+    generation_mode: str = Field(description="The generation mode that output this image",)
+    positive_prompt: str = Field(description="The positive prompt parameter")
+    negative_prompt: str = Field(description="The negative prompt parameter")
+    width: int = Field(description="The width parameter")
+    height: int = Field(description="The height parameter")
+    seed: int = Field(description="The seed used for noise generation")
+    rand_device: str = Field(description="The device used for random number generation")
+    cfg_scale: float = Field(description="The classifier-free guidance scale parameter")
+    steps: int = Field(description="The number of steps used for inference")
+    scheduler: str = Field(description="The scheduler used for inference")
+    clip_skip: int = Field(description="The number of skipped CLIP layers",)
+    model: MainModelField = Field(description="The main model used for inference")
+    controlnets: list[ControlField]= Field(description="The ControlNets used for inference")
+    loras: list[LoRAMetadataField] = Field(description="The LoRAs used for inference")
+    strength: Union[float, None] = Field(
+        default=None,
+        description="The strength used for latents-to-latents",
+    )
+    init_image: Union[str, None] = Field(
+        default=None, description="The name of the initial image"
+    )
+    vae: Union[VAEModelField, None] = Field(
+        default=None,
+        description="The VAE used for decoding, if the main model's default was not used",
+    )
+
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Metadata Accumulator",
+                "tags": ["image", "metadata", "generation"]
+            },
+        }
+
+
+    def invoke(self, context: InvocationContext) -> MetadataAccumulatorOutput:
+        """Collects and outputs a CoreMetadata object"""
+
+        return MetadataAccumulatorOutput(
+            metadata=CoreMetadata(
+                generation_mode=self.generation_mode,
+                positive_prompt=self.positive_prompt,
+                negative_prompt=self.negative_prompt,
+                width=self.width,
+                height=self.height,
+                seed=self.seed,
+                rand_device=self.rand_device,
+                cfg_scale=self.cfg_scale,
+                steps=self.steps,
+                scheduler=self.scheduler,
+                model=self.model,
+                strength=self.strength,
+                init_image=self.init_image,
+                vae=self.vae,
+                controlnets=self.controlnets,
+                loras=self.loras,
+                clip_skip=self.clip_skip,
+            )
+        )

invokeai/app/invocations/model.py

@@ -33,7 +33,6 @@ class ClipField(BaseModel):
     skipped_layers: int = Field(description="Number of skipped layers in text_encoder")
     loras: List[LoraInfo] = Field(description="Loras to apply on model loading")
 
-
 class VaeField(BaseModel):
     # TODO: better naming?
     vae: ModelInfo = Field(description="Info to load vae submodel")
@@ -50,7 +49,6 @@ class ModelLoaderOutput(BaseInvocationOutput):
     vae: VaeField = Field(default=None, description="Vae submodel")
     # fmt: on
 
-
 class MainModelField(BaseModel):
     """Main model field"""
 
@@ -64,7 +62,6 @@ class LoRAModelField(BaseModel):
     model_name: str = Field(description="Name of the LoRA model")
     base_model: BaseModelType = Field(description="Base model")
 
-
 class MainModelLoaderInvocation(BaseInvocation):
     """Loads a main model, outputting its submodels."""
 
@@ -157,6 +154,22 @@ class MainModelLoaderInvocation(BaseInvocation):
                 loras=[],
                 skipped_layers=0,
             ),
+            clip2=ClipField(
+                tokenizer=ModelInfo(
+                    model_name=model_name,
+                    base_model=base_model,
+                    model_type=model_type,
+                    submodel=SubModelType.Tokenizer2,
+                ),
+                text_encoder=ModelInfo(
+                    model_name=model_name,
+                    base_model=base_model,
+                    model_type=model_type,
+                    submodel=SubModelType.TextEncoder2,
+                ),
+                loras=[],
+                skipped_layers=0,
+            ),
             vae=VaeField(
                 vae=ModelInfo(
                     model_name=model_name,
@@ -167,7 +180,7 @@ class MainModelLoaderInvocation(BaseInvocation):
             ),
         )
 
-
+    
 class LoraLoaderOutput(BaseInvocationOutput):
     """Model loader output"""
 

invokeai/app/invocations/noise.py

@@ -48,7 +48,7 @@ def get_noise(
         dtype=torch_dtype(device),
         device=noise_device_type,
         generator=generator,
-    ).to(device)
+    ).to("cpu")
 
     return noise_tensor
 
@@ -112,6 +112,7 @@ class NoiseInvocation(BaseInvocation):
     class Config(InvocationConfig):
         schema_extra = {
             "ui": {
+                "title": "Noise",
                 "tags": ["latents", "noise"],
             },
         }

invokeai/app/invocations/param_easing.py

@@ -43,6 +43,14 @@ class FloatLinearRangeInvocation(BaseInvocation):
     stop: float = Field(default=10, description="The last value of the range")
     steps: int = Field(default=30, description="number of values to interpolate over (including start and stop)")
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Linear Range (Float)",
+                "tags": ["math", "float", "linear", "range"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> FloatCollectionOutput:
         param_list = list(np.linspace(self.start, self.stop, self.steps))
         return FloatCollectionOutput(
@@ -113,6 +121,14 @@ class StepParamEasingInvocation(BaseInvocation):
     show_easing_plot: bool = Field(default=False, description="show easing plot")
     # fmt: on
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Param Easing By Step",
+                "tags": ["param", "step", "easing"]
+            },
+        }
+
 
     def invoke(self, context: InvocationContext) -> FloatCollectionOutput:
         log_diagnostics = False

invokeai/app/invocations/params.py

@@ -1,9 +1,12 @@
 # Copyright (c) 2023 Kyle Schouviller (https://github.com/kyle0654)
 
 from typing import Literal
+
 from pydantic import Field
-from .baseinvocation import BaseInvocation, BaseInvocationOutput, InvocationContext
-from .math import IntOutput, FloatOutput
+
+from .baseinvocation import (BaseInvocation, BaseInvocationOutput,
+                             InvocationConfig, InvocationContext)
+from .math import FloatOutput, IntOutput
 
 # Pass-through parameter nodes - used by subgraphs
 
@@ -14,6 +17,14 @@ class ParamIntInvocation(BaseInvocation):
     a: int = Field(default=0, description="The integer value")
     #fmt: on
 
+    class Config(InvocationConfig):
+      schema_extra = {
+          "ui": {
+              "tags": ["param", "integer"],
+              "title": "Integer Parameter"
+          },
+      }
+
     def invoke(self, context: InvocationContext) -> IntOutput:
         return IntOutput(a=self.a)
 
@@ -24,5 +35,36 @@ class ParamFloatInvocation(BaseInvocation):
     param: float = Field(default=0.0, description="The float value")
     #fmt: on
 
+    class Config(InvocationConfig):
+      schema_extra = {
+          "ui": {
+              "tags": ["param", "float"],
+              "title": "Float Parameter"
+          },
+      }
+
     def invoke(self, context: InvocationContext) -> FloatOutput:
         return FloatOutput(param=self.param)
+
+class StringOutput(BaseInvocationOutput):
+    """A string output"""
+    type: Literal["string_output"] = "string_output"
+    text: str = Field(default=None, description="The output string")
+
+
+class ParamStringInvocation(BaseInvocation):
+    """A string parameter"""
+    type: Literal['param_string'] = 'param_string'
+    text: str = Field(default='', description='The string value')
+
+    class Config(InvocationConfig):
+      schema_extra = {
+          "ui": {
+              "tags": ["param", "string"],
+              "title": "String Parameter"
+          },
+      }
+
+    def invoke(self, context: InvocationContext) -> StringOutput:
+        return StringOutput(text=self.text)
+    

invokeai/app/invocations/prompt.py

@@ -1,8 +1,10 @@
-from typing import Literal
+from os.path import exists
+from typing import Literal, Optional
 
-from pydantic.fields import Field
+import numpy as np
+from pydantic import Field, validator
 
-from .baseinvocation import BaseInvocation, BaseInvocationOutput, InvocationContext
+from .baseinvocation import BaseInvocation, BaseInvocationOutput, InvocationConfig, InvocationContext
 from dynamicprompts.generators import RandomPromptGenerator, CombinatorialPromptGenerator
 
 class PromptOutput(BaseInvocationOutput):
@@ -46,6 +48,14 @@ class DynamicPromptInvocation(BaseInvocation):
         default=False, description="Whether to use the combinatorial generator"
     )
 
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Dynamic Prompt",
+                "tags": ["prompt", "dynamic"]
+            },
+        }
+
     def invoke(self, context: InvocationContext) -> PromptCollectionOutput:
         if self.combinatorial:
             generator = CombinatorialPromptGenerator()
@@ -55,3 +65,49 @@ class DynamicPromptInvocation(BaseInvocation):
             prompts = generator.generate(self.prompt, num_images=self.max_prompts)
 
         return PromptCollectionOutput(prompt_collection=prompts, count=len(prompts))
+    
+
+class PromptsFromFileInvocation(BaseInvocation):
+    '''Loads prompts from a text file'''
+    # fmt: off
+    type: Literal['prompt_from_file'] = 'prompt_from_file'
+
+    # Inputs
+    file_path: str = Field(description="Path to prompt text file")
+    pre_prompt: Optional[str] = Field(description="String to prepend to each prompt")
+    post_prompt: Optional[str] = Field(description="String to append to each prompt")
+    start_line: int = Field(default=1, ge=1, description="Line in the file to start start from")
+    max_prompts: int = Field(default=1, ge=0, description="Max lines to read from file (0=all)")
+    #fmt: on
+
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "Prompts From File",
+                "tags": ["prompt", "file"]
+            },
+        }
+
+    @validator("file_path")
+    def file_path_exists(cls, v):
+        if not exists(v):
+            raise ValueError(FileNotFoundError)
+        return v
+
+    def promptsFromFile(self, file_path: str, pre_prompt: str, post_prompt: str, start_line: int, max_prompts: int):
+        prompts = []
+        start_line -= 1
+        end_line = start_line + max_prompts
+        if max_prompts <= 0:
+            end_line = np.iinfo(np.int32).max
+        with open(file_path) as f:
+            for i, line in enumerate(f):
+                if i >= start_line and i < end_line:
+                    prompts.append((pre_prompt or '') + line.strip() + (post_prompt or ''))
+                if i >= end_line:
+                    break
+        return prompts
+
+    def invoke(self, context: InvocationContext) -> PromptCollectionOutput:
+        prompts = self.promptsFromFile(self.file_path, self.pre_prompt, self.post_prompt, self.start_line, self.max_prompts)
+        return PromptCollectionOutput(prompt_collection=prompts, count=len(prompts))

invokeai/app/invocations/reconstruct.py

@@ -1,55 +0,0 @@
-from typing import Literal, Optional
-
-from pydantic import Field
-
-from invokeai.app.models.image import ImageCategory, ImageField, ResourceOrigin
-
-from .baseinvocation import BaseInvocation, InvocationContext, InvocationConfig
-from .image import ImageOutput
-
-
-class RestoreFaceInvocation(BaseInvocation):
-    """Restores faces in an image."""
-
-    # fmt: off
-    type:  Literal["restore_face"] = "restore_face"
-
-    # Inputs
-    image: Optional[ImageField] = Field(description="The input image")
-    strength:                float = Field(default=0.75, gt=0, le=1, description="The strength of the restoration"  )
-    # fmt: on
-
-    # Schema customisation
-    class Config(InvocationConfig):
-        schema_extra = {
-            "ui": {
-                "tags": ["restoration", "image"],
-            },
-        }
-
-    def invoke(self, context: InvocationContext) -> ImageOutput:
-        image = context.services.images.get_pil_image(self.image.image_name)
-        results = context.services.restoration.upscale_and_reconstruct(
-            image_list=[[image, 0]],
-            upscale=None,
-            strength=self.strength,  # GFPGAN strength
-            save_original=False,
-            image_callback=None,
-        )
-
-        # Results are image and seed, unwrap for now
-        # TODO: can this return multiple results?
-        image_dto = context.services.images.create(
-            image=results[0][0],
-            image_origin=ResourceOrigin.INTERNAL,
-            image_category=ImageCategory.GENERAL,
-            node_id=self.id,
-            session_id=context.graph_execution_state_id,
-            is_intermediate=self.is_intermediate,
-        )
-
-        return ImageOutput(
-            image=ImageField(image_name=image_dto.image_name),
-            width=image_dto.width,
-            height=image_dto.height,
-        )

invokeai/app/invocations/sdxl.py

@@ -0,0 +1,709 @@
+import torch
+import inspect
+from tqdm import tqdm
+from typing import List, Literal, Optional, Union
+
+from pydantic import Field, validator
+
+from ...backend.model_management import ModelType, SubModelType
+from invokeai.app.util.step_callback import stable_diffusion_xl_step_callback
+from .baseinvocation import (BaseInvocation, BaseInvocationOutput,
+                             InvocationConfig, InvocationContext)
+
+from .model import UNetField, ClipField, VaeField, MainModelField, ModelInfo
+from .compel import ConditioningField
+from .latent import LatentsField, SAMPLER_NAME_VALUES, LatentsOutput, get_scheduler, build_latents_output
+
+class SDXLModelLoaderOutput(BaseInvocationOutput):
+    """SDXL base model loader output"""
+
+    # fmt: off
+    type: Literal["sdxl_model_loader_output"] = "sdxl_model_loader_output"
+
+    unet: UNetField = Field(default=None, description="UNet submodel")
+    clip: ClipField = Field(default=None, description="Tokenizer and text_encoder submodels")
+    clip2: ClipField = Field(default=None, description="Tokenizer and text_encoder submodels")
+    vae: VaeField = Field(default=None, description="Vae submodel")
+    # fmt: on
+
+class SDXLRefinerModelLoaderOutput(BaseInvocationOutput):
+    """SDXL refiner model loader output"""
+    # fmt: off
+    type: Literal["sdxl_refiner_model_loader_output"] = "sdxl_refiner_model_loader_output"
+    unet: UNetField = Field(default=None, description="UNet submodel")
+    clip2: ClipField = Field(default=None, description="Tokenizer and text_encoder submodels")
+    vae: VaeField = Field(default=None, description="Vae submodel")
+    # fmt: on
+    #fmt: on
+    
+class SDXLModelLoaderInvocation(BaseInvocation):
+    """Loads an sdxl base model, outputting its submodels."""
+
+    type: Literal["sdxl_model_loader"] = "sdxl_model_loader"
+
+    model: MainModelField = Field(description="The model to load")
+    # TODO: precision?
+
+    # Schema customisation
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "SDXL Model Loader",
+                "tags": ["model", "loader", "sdxl"],
+                "type_hints": {"model": "model"},
+            },
+        }
+
+    def invoke(self, context: InvocationContext) -> SDXLModelLoaderOutput:
+        base_model = self.model.base_model
+        model_name = self.model.model_name
+        model_type = ModelType.Main
+
+        # TODO: not found exceptions
+        if not context.services.model_manager.model_exists(
+            model_name=model_name,
+            base_model=base_model,
+            model_type=model_type,
+        ):
+            raise Exception(f"Unknown {base_model} {model_type} model: {model_name}")
+
+        return SDXLModelLoaderOutput(
+            unet=UNetField(
+                unet=ModelInfo(
+                    model_name=model_name,
+                    base_model=base_model,
+                    model_type=model_type,
+                    submodel=SubModelType.UNet,
+                ),
+                scheduler=ModelInfo(
+                    model_name=model_name,
+                    base_model=base_model,
+                    model_type=model_type,
+                    submodel=SubModelType.Scheduler,
+                ),
+                loras=[],
+            ),
+            clip=ClipField(
+                tokenizer=ModelInfo(
+                    model_name=model_name,
+                    base_model=base_model,
+                    model_type=model_type,
+                    submodel=SubModelType.Tokenizer,
+                ),
+                text_encoder=ModelInfo(
+                    model_name=model_name,
+                    base_model=base_model,
+                    model_type=model_type,
+                    submodel=SubModelType.TextEncoder,
+                ),
+                loras=[],
+                skipped_layers=0,
+            ),
+            clip2=ClipField(
+                tokenizer=ModelInfo(
+                    model_name=model_name,
+                    base_model=base_model,
+                    model_type=model_type,
+                    submodel=SubModelType.Tokenizer2,
+                ),
+                text_encoder=ModelInfo(
+                    model_name=model_name,
+                    base_model=base_model,
+                    model_type=model_type,
+                    submodel=SubModelType.TextEncoder2,
+                ),
+                loras=[],
+                skipped_layers=0,
+            ),
+            vae=VaeField(
+                vae=ModelInfo(
+                    model_name=model_name,
+                    base_model=base_model,
+                    model_type=model_type,
+                    submodel=SubModelType.Vae,
+                ),
+            ),
+        )
+
+class SDXLRefinerModelLoaderInvocation(BaseInvocation):
+    """Loads an sdxl refiner model, outputting its submodels."""
+    type: Literal["sdxl_refiner_model_loader"] = "sdxl_refiner_model_loader"
+
+    model: MainModelField = Field(description="The model to load")
+    # TODO: precision?
+
+    # Schema customisation
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "SDXL Refiner Model Loader",
+                "tags": ["model", "loader", "sdxl_refiner"],
+                "type_hints": {"model": "model"},
+            },
+        }
+
+    def invoke(self, context: InvocationContext) -> SDXLRefinerModelLoaderOutput:
+        base_model = self.model.base_model
+        model_name = self.model.model_name
+        model_type = ModelType.Main
+
+        # TODO: not found exceptions
+        if not context.services.model_manager.model_exists(
+            model_name=model_name,
+            base_model=base_model,
+            model_type=model_type,
+        ):
+            raise Exception(f"Unknown {base_model} {model_type} model: {model_name}")
+
+        return SDXLRefinerModelLoaderOutput(
+            unet=UNetField(
+                unet=ModelInfo(
+                    model_name=model_name,
+                    base_model=base_model,
+                    model_type=model_type,
+                    submodel=SubModelType.UNet,
+                ),
+                scheduler=ModelInfo(
+                    model_name=model_name,
+                    base_model=base_model,
+                    model_type=model_type,
+                    submodel=SubModelType.Scheduler,
+                ),
+                loras=[],
+            ),
+            clip2=ClipField(
+                tokenizer=ModelInfo(
+                    model_name=model_name,
+                    base_model=base_model,
+                    model_type=model_type,
+                    submodel=SubModelType.Tokenizer2,
+                ),
+                text_encoder=ModelInfo(
+                    model_name=model_name,
+                    base_model=base_model,
+                    model_type=model_type,
+                    submodel=SubModelType.TextEncoder2,
+                ),
+                loras=[],
+                skipped_layers=0,
+            ),
+            vae=VaeField(
+                vae=ModelInfo(
+                    model_name=model_name,
+                    base_model=base_model,
+                    model_type=model_type,
+                    submodel=SubModelType.Vae,
+                ),
+            ),
+        )
+    
+# Text to image
+class SDXLTextToLatentsInvocation(BaseInvocation):
+    """Generates latents from conditionings."""
+
+    type: Literal["t2l_sdxl"] = "t2l_sdxl"
+
+    # Inputs
+    # fmt: off
+    positive_conditioning: Optional[ConditioningField] = Field(description="Positive conditioning for generation")
+    negative_conditioning: Optional[ConditioningField] = Field(description="Negative conditioning for generation")
+    noise: Optional[LatentsField] = Field(description="The noise to use")
+    steps:       int = Field(default=10, gt=0, description="The number of steps to use to generate the image")
+    cfg_scale: Union[float, List[float]] = Field(default=7.5, ge=1, description="The Classifier-Free Guidance, higher values may result in a result closer to the prompt", )
+    scheduler: SAMPLER_NAME_VALUES = Field(default="euler", description="The scheduler to use" )
+    unet: UNetField = Field(default=None, description="UNet submodel")
+    denoising_end: float = Field(default=1.0, gt=0, le=1, description="")
+    #control: Union[ControlField, list[ControlField]] = Field(default=None, description="The control to use")
+    #seamless:   bool = Field(default=False, description="Whether or not to generate an image that can tile without seams", )
+    #seamless_axes: str = Field(default="", description="The axes to tile the image on, 'x' and/or 'y'")
+    # fmt: on
+
+    @validator("cfg_scale")
+    def ge_one(cls, v):
+        """validate that all cfg_scale values are >= 1"""
+        if isinstance(v, list):
+            for i in v:
+                if i < 1:
+                    raise ValueError('cfg_scale must be greater than 1')
+        else:
+            if v < 1:
+                raise ValueError('cfg_scale must be greater than 1')
+        return v
+
+    # Schema customisation
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "SDXL Text To Latents",
+                "tags": ["latents"],
+                "type_hints": {
+                  "model": "model",
+                  # "cfg_scale": "float",
+                  "cfg_scale": "number"
+                }
+            },
+        }
+
+    def dispatch_progress(
+        self,
+        context: InvocationContext,
+        source_node_id: str,
+        sample,
+        step,
+        total_steps,
+    ) -> None:
+        stable_diffusion_xl_step_callback(
+            context=context,
+            node=self.dict(),
+            source_node_id=source_node_id,
+            sample=sample,
+            step=step,
+            total_steps=total_steps,
+        )
+
+    # based on
+    # https://github.com/huggingface/diffusers/blob/3ebbaf7c96801271f9e6c21400033b6aa5ffcf29/src/diffusers/pipelines/stable_diffusion/pipeline_onnx_stable_diffusion.py#L375
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> LatentsOutput:
+        graph_execution_state = context.services.graph_execution_manager.get(
+            context.graph_execution_state_id
+        )
+        source_node_id = graph_execution_state.prepared_source_mapping[self.id]
+        latents = context.services.latents.get(self.noise.latents_name)
+
+        positive_cond_data = context.services.latents.get(self.positive_conditioning.conditioning_name)
+        prompt_embeds = positive_cond_data.conditionings[0].embeds
+        pooled_prompt_embeds = positive_cond_data.conditionings[0].pooled_embeds
+        add_time_ids = positive_cond_data.conditionings[0].add_time_ids
+
+        negative_cond_data = context.services.latents.get(self.negative_conditioning.conditioning_name)
+        negative_prompt_embeds = negative_cond_data.conditionings[0].embeds
+        negative_pooled_prompt_embeds = negative_cond_data.conditionings[0].pooled_embeds
+        add_neg_time_ids = negative_cond_data.conditionings[0].add_time_ids
+
+        scheduler = get_scheduler(
+            context=context,
+            scheduler_info=self.unet.scheduler,
+            scheduler_name=self.scheduler,
+        )
+
+        num_inference_steps = self.steps
+        scheduler.set_timesteps(num_inference_steps)
+        timesteps = scheduler.timesteps
+
+        latents = latents * scheduler.init_noise_sigma
+
+
+        unet_info = context.services.model_manager.get_model(
+            **self.unet.unet.dict()
+        )
+        do_classifier_free_guidance = True
+        cross_attention_kwargs = None
+        with unet_info as unet:
+
+            extra_step_kwargs = dict()
+            if "eta" in set(inspect.signature(scheduler.step).parameters.keys()):
+                extra_step_kwargs.update(
+                    eta=0.0,
+                )
+            if "generator" in set(inspect.signature(scheduler.step).parameters.keys()):
+                extra_step_kwargs.update(
+                    generator=torch.Generator(device=unet.device).manual_seed(0),
+                )
+
+            num_warmup_steps = len(timesteps) - self.steps * scheduler.order
+
+            # apply denoising_end
+            skipped_final_steps = int(round((1 - self.denoising_end) * self.steps))
+            num_inference_steps = num_inference_steps - skipped_final_steps
+            timesteps = timesteps[: num_warmup_steps + scheduler.order * num_inference_steps]
+
+            if not context.services.configuration.sequential_guidance:
+                prompt_embeds = torch.cat([negative_prompt_embeds, prompt_embeds], dim=0)
+                add_text_embeds = torch.cat([negative_pooled_prompt_embeds, pooled_prompt_embeds], dim=0)
+                add_time_ids = torch.cat([add_neg_time_ids, add_time_ids], dim=0)
+
+                prompt_embeds = prompt_embeds.to(device=unet.device, dtype=unet.dtype)
+                add_text_embeds = add_text_embeds.to(device=unet.device, dtype=unet.dtype)
+                add_time_ids = add_time_ids.to(device=unet.device, dtype=unet.dtype)
+                latents = latents.to(device=unet.device, dtype=unet.dtype)
+
+                with tqdm(total=num_inference_steps) as progress_bar:
+                    for i, t in enumerate(timesteps):
+                        # expand the latents if we are doing classifier free guidance
+                        latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents
+
+                        latent_model_input = scheduler.scale_model_input(latent_model_input, t)
+
+                        # predict the noise residual
+                        added_cond_kwargs = {"text_embeds": add_text_embeds, "time_ids": add_time_ids}
+                        noise_pred = unet(
+                            latent_model_input,
+                            t,
+                            encoder_hidden_states=prompt_embeds,
+                            cross_attention_kwargs=cross_attention_kwargs,
+                            added_cond_kwargs=added_cond_kwargs,
+                            return_dict=False,
+                        )[0]
+
+                        # perform guidance
+                        if do_classifier_free_guidance:
+                            noise_pred_uncond, noise_pred_text = noise_pred.chunk(2)
+                            noise_pred = noise_pred_uncond + self.cfg_scale * (noise_pred_text - noise_pred_uncond)
+                            #del noise_pred_uncond
+                            #del noise_pred_text
+
+                        #if do_classifier_free_guidance and guidance_rescale > 0.0:
+                        #    # Based on 3.4. in https://arxiv.org/pdf/2305.08891.pdf
+                        #    noise_pred = rescale_noise_cfg(noise_pred, noise_pred_text, guidance_rescale=guidance_rescale)
+
+                        # compute the previous noisy sample x_t -> x_t-1
+                        latents = scheduler.step(noise_pred, t, latents, **extra_step_kwargs, return_dict=False)[0]
+
+                        # call the callback, if provided
+                        if i == len(timesteps) - 1 or ((i + 1) > num_warmup_steps and (i + 1) % scheduler.order == 0):
+                            progress_bar.update()
+                            self.dispatch_progress(context, source_node_id, latents, i, num_inference_steps)
+                            #if callback is not None and i % callback_steps == 0:
+                            #    callback(i, t, latents)
+            else:
+                negative_pooled_prompt_embeds = negative_pooled_prompt_embeds.to(device=unet.device, dtype=unet.dtype)
+                negative_prompt_embeds = negative_prompt_embeds.to(device=unet.device, dtype=unet.dtype)
+                add_neg_time_ids = add_neg_time_ids.to(device=unet.device, dtype=unet.dtype)
+                pooled_prompt_embeds = pooled_prompt_embeds.to(device=unet.device, dtype=unet.dtype)
+                prompt_embeds = prompt_embeds.to(device=unet.device, dtype=unet.dtype)
+                add_time_ids = add_time_ids.to(device=unet.device, dtype=unet.dtype)
+                latents = latents.to(device=unet.device, dtype=unet.dtype)
+
+                with tqdm(total=num_inference_steps) as progress_bar:
+                    for i, t in enumerate(timesteps):
+                        # expand the latents if we are doing classifier free guidance
+                        #latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents
+
+                        latent_model_input = scheduler.scale_model_input(latents, t)
+
+                        #import gc
+                        #gc.collect()
+                        #torch.cuda.empty_cache()
+
+                        # predict the noise residual
+
+                        added_cond_kwargs = {"text_embeds": negative_pooled_prompt_embeds, "time_ids": add_neg_time_ids}
+                        noise_pred_uncond = unet(
+                            latent_model_input,
+                            t,
+                            encoder_hidden_states=negative_prompt_embeds,
+                            cross_attention_kwargs=cross_attention_kwargs,
+                            added_cond_kwargs=added_cond_kwargs,
+                            return_dict=False,
+                        )[0]
+
+                        added_cond_kwargs = {"text_embeds": pooled_prompt_embeds, "time_ids": add_time_ids}
+                        noise_pred_text = unet(
+                            latent_model_input,
+                            t,
+                            encoder_hidden_states=prompt_embeds,
+                            cross_attention_kwargs=cross_attention_kwargs,
+                            added_cond_kwargs=added_cond_kwargs,
+                            return_dict=False,
+                        )[0]
+
+                        # perform guidance
+                        noise_pred = noise_pred_uncond + self.cfg_scale * (noise_pred_text - noise_pred_uncond)
+
+                        #del noise_pred_text
+                        #del noise_pred_uncond
+                        #import gc
+                        #gc.collect()
+                        #torch.cuda.empty_cache()
+
+                        #if do_classifier_free_guidance and guidance_rescale > 0.0:
+                        #    # Based on 3.4. in https://arxiv.org/pdf/2305.08891.pdf
+                        #    noise_pred = rescale_noise_cfg(noise_pred, noise_pred_text, guidance_rescale=guidance_rescale)
+
+                        # compute the previous noisy sample x_t -> x_t-1
+                        latents = scheduler.step(noise_pred, t, latents, **extra_step_kwargs, return_dict=False)[0]
+
+                        #del noise_pred
+                        #import gc
+                        #gc.collect()
+                        #torch.cuda.empty_cache()
+
+                        # call the callback, if provided
+                        if i == len(timesteps) - 1 or ((i + 1) > num_warmup_steps and (i + 1) % scheduler.order == 0):
+                            progress_bar.update()
+                            self.dispatch_progress(context, source_node_id, latents, i, num_inference_steps)
+                            #if callback is not None and i % callback_steps == 0:
+                            #    callback(i, t, latents)
+
+
+
+        #################
+
+        latents = latents.to("cpu")
+        torch.cuda.empty_cache()
+
+        name = f'{context.graph_execution_state_id}__{self.id}'
+        context.services.latents.save(name, latents)
+        return build_latents_output(latents_name=name, latents=latents)
+
+class SDXLLatentsToLatentsInvocation(BaseInvocation):
+    """Generates latents from conditionings."""
+
+    type: Literal["l2l_sdxl"] = "l2l_sdxl"
+
+    # Inputs
+    # fmt: off
+    positive_conditioning: Optional[ConditioningField] = Field(description="Positive conditioning for generation")
+    negative_conditioning: Optional[ConditioningField] = Field(description="Negative conditioning for generation")
+    noise: Optional[LatentsField] = Field(description="The noise to use")
+    steps:       int = Field(default=10, gt=0, description="The number of steps to use to generate the image")
+    cfg_scale: Union[float, List[float]] = Field(default=7.5, ge=1, description="The Classifier-Free Guidance, higher values may result in a result closer to the prompt", )
+    scheduler: SAMPLER_NAME_VALUES = Field(default="euler", description="The scheduler to use" )
+    unet: UNetField = Field(default=None, description="UNet submodel")
+    latents: Optional[LatentsField] = Field(description="Initial latents")
+
+    denoising_start: float = Field(default=0.0, ge=0, lt=1, description="")
+    denoising_end: float = Field(default=1.0, gt=0, le=1, description="")
+
+    #control: Union[ControlField, list[ControlField]] = Field(default=None, description="The control to use")
+    #seamless:   bool = Field(default=False, description="Whether or not to generate an image that can tile without seams", )
+    #seamless_axes: str = Field(default="", description="The axes to tile the image on, 'x' and/or 'y'")
+    # fmt: on
+
+    @validator("cfg_scale")
+    def ge_one(cls, v):
+        """validate that all cfg_scale values are >= 1"""
+        if isinstance(v, list):
+            for i in v:
+                if i < 1:
+                    raise ValueError('cfg_scale must be greater than 1')
+        else:
+            if v < 1:
+                raise ValueError('cfg_scale must be greater than 1')
+        return v
+
+    # Schema customisation
+    class Config(InvocationConfig):
+        schema_extra = {
+            "ui": {
+                "title": "SDXL Latents to Latents",
+                "tags": ["latents"],
+                "type_hints": {
+                  "model": "model",
+                  # "cfg_scale": "float",
+                  "cfg_scale": "number"
+                }
+            },
+        }
+
+    def dispatch_progress(
+        self,
+        context: InvocationContext,
+        source_node_id: str,
+        sample,
+        step,
+        total_steps,
+    ) -> None:
+        stable_diffusion_xl_step_callback(
+            context=context,
+            node=self.dict(),
+            source_node_id=source_node_id,
+            sample=sample,
+            step=step,
+            total_steps=total_steps,
+        )
+
+    # based on
+    # https://github.com/huggingface/diffusers/blob/3ebbaf7c96801271f9e6c21400033b6aa5ffcf29/src/diffusers/pipelines/stable_diffusion/pipeline_onnx_stable_diffusion.py#L375
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> LatentsOutput:
+        graph_execution_state = context.services.graph_execution_manager.get(
+            context.graph_execution_state_id
+        )
+        source_node_id = graph_execution_state.prepared_source_mapping[self.id]
+        latents = context.services.latents.get(self.latents.latents_name)
+
+        positive_cond_data = context.services.latents.get(self.positive_conditioning.conditioning_name)
+        prompt_embeds = positive_cond_data.conditionings[0].embeds
+        pooled_prompt_embeds = positive_cond_data.conditionings[0].pooled_embeds
+        add_time_ids = positive_cond_data.conditionings[0].add_time_ids
+
+        negative_cond_data = context.services.latents.get(self.negative_conditioning.conditioning_name)
+        negative_prompt_embeds = negative_cond_data.conditionings[0].embeds
+        negative_pooled_prompt_embeds = negative_cond_data.conditionings[0].pooled_embeds
+        add_neg_time_ids = negative_cond_data.conditionings[0].add_time_ids
+
+        scheduler = get_scheduler(
+            context=context,
+            scheduler_info=self.unet.scheduler,
+            scheduler_name=self.scheduler,
+        )
+
+        # apply denoising_start
+        num_inference_steps = self.steps
+        scheduler.set_timesteps(num_inference_steps)
+
+        t_start = int(round(self.denoising_start * num_inference_steps))
+        timesteps = scheduler.timesteps[t_start * scheduler.order:]
+        num_inference_steps = num_inference_steps - t_start
+
+        # apply noise(if provided)
+        if self.noise is not None:
+            noise = context.services.latents.get(self.noise.latents_name)
+            latents = scheduler.add_noise(latents, noise, timesteps[:1])
+            del noise
+
+        unet_info = context.services.model_manager.get_model(
+            **self.unet.unet.dict()
+        )
+        do_classifier_free_guidance = True
+        cross_attention_kwargs = None
+        with unet_info as unet:
+
+            # apply scheduler extra args
+            extra_step_kwargs = dict()
+            if "eta" in set(inspect.signature(scheduler.step).parameters.keys()):
+                extra_step_kwargs.update(
+                    eta=0.0,
+                )
+            if "generator" in set(inspect.signature(scheduler.step).parameters.keys()):
+                extra_step_kwargs.update(
+                    generator=torch.Generator(device=unet.device).manual_seed(0),
+                )
+
+            num_warmup_steps = max(len(timesteps) - num_inference_steps * scheduler.order, 0)
+
+            # apply denoising_end
+            skipped_final_steps = int(round((1 - self.denoising_end) * self.steps))
+            num_inference_steps = num_inference_steps - skipped_final_steps
+            timesteps = timesteps[: num_warmup_steps + scheduler.order * num_inference_steps]
+
+            if not context.services.configuration.sequential_guidance:
+                prompt_embeds = torch.cat([negative_prompt_embeds, prompt_embeds], dim=0)
+                add_text_embeds = torch.cat([negative_pooled_prompt_embeds, pooled_prompt_embeds], dim=0)
+                add_time_ids = torch.cat([add_neg_time_ids, add_time_ids], dim=0)
+
+                prompt_embeds = prompt_embeds.to(device=unet.device, dtype=unet.dtype)
+                add_text_embeds = add_text_embeds.to(device=unet.device, dtype=unet.dtype)
+                add_time_ids = add_time_ids.to(device=unet.device, dtype=unet.dtype)
+                latents = latents.to(device=unet.device, dtype=unet.dtype)
+
+                with tqdm(total=num_inference_steps) as progress_bar:
+                    for i, t in enumerate(timesteps):
+                        # expand the latents if we are doing classifier free guidance
+                        latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents
+
+                        latent_model_input = scheduler.scale_model_input(latent_model_input, t)
+
+                        # predict the noise residual
+                        added_cond_kwargs = {"text_embeds": add_text_embeds, "time_ids": add_time_ids}
+                        noise_pred = unet(
+                            latent_model_input,
+                            t,
+                            encoder_hidden_states=prompt_embeds,
+                            cross_attention_kwargs=cross_attention_kwargs,
+                            added_cond_kwargs=added_cond_kwargs,
+                            return_dict=False,
+                        )[0]
+
+                        # perform guidance
+                        if do_classifier_free_guidance:
+                            noise_pred_uncond, noise_pred_text = noise_pred.chunk(2)
+                            noise_pred = noise_pred_uncond + self.cfg_scale * (noise_pred_text - noise_pred_uncond)
+                            #del noise_pred_uncond
+                            #del noise_pred_text
+
+                        #if do_classifier_free_guidance and guidance_rescale > 0.0:
+                        #    # Based on 3.4. in https://arxiv.org/pdf/2305.08891.pdf
+                        #    noise_pred = rescale_noise_cfg(noise_pred, noise_pred_text, guidance_rescale=guidance_rescale)
+
+                        # compute the previous noisy sample x_t -> x_t-1
+                        latents = scheduler.step(noise_pred, t, latents, **extra_step_kwargs, return_dict=False)[0]
+
+                        # call the callback, if provided
+                        if i == len(timesteps) - 1 or ((i + 1) > num_warmup_steps and (i + 1) % scheduler.order == 0):
+                            progress_bar.update()
+                            self.dispatch_progress(context, source_node_id, latents, i, num_inference_steps)
+                            #if callback is not None and i % callback_steps == 0:
+                            #    callback(i, t, latents)
+            else:
+                negative_pooled_prompt_embeds = negative_pooled_prompt_embeds.to(device=unet.device, dtype=unet.dtype)
+                negative_prompt_embeds = negative_prompt_embeds.to(device=unet.device, dtype=unet.dtype)
+                add_neg_time_ids = add_neg_time_ids.to(device=unet.device, dtype=unet.dtype)
+                pooled_prompt_embeds = pooled_prompt_embeds.to(device=unet.device, dtype=unet.dtype)
+                prompt_embeds = prompt_embeds.to(device=unet.device, dtype=unet.dtype)
+                add_time_ids = add_time_ids.to(device=unet.device, dtype=unet.dtype)
+                latents = latents.to(device=unet.device, dtype=unet.dtype)
+
+                with tqdm(total=num_inference_steps) as progress_bar:
+                    for i, t in enumerate(timesteps):
+                        # expand the latents if we are doing classifier free guidance
+                        #latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents
+
+                        latent_model_input = scheduler.scale_model_input(latents, t)
+
+                        #import gc
+                        #gc.collect()
+                        #torch.cuda.empty_cache()
+
+                        # predict the noise residual
+
+                        added_cond_kwargs = {"text_embeds": negative_pooled_prompt_embeds, "time_ids": add_time_ids}
+                        noise_pred_uncond = unet(
+                            latent_model_input,
+                            t,
+                            encoder_hidden_states=negative_prompt_embeds,
+                            cross_attention_kwargs=cross_attention_kwargs,
+                            added_cond_kwargs=added_cond_kwargs,
+                            return_dict=False,
+                        )[0]
+
+                        added_cond_kwargs = {"text_embeds": pooled_prompt_embeds, "time_ids": add_time_ids}
+                        noise_pred_text = unet(
+                            latent_model_input,
+                            t,
+                            encoder_hidden_states=prompt_embeds,
+                            cross_attention_kwargs=cross_attention_kwargs,
+                            added_cond_kwargs=added_cond_kwargs,
+                            return_dict=False,
+                        )[0]
+
+                        # perform guidance
+                        noise_pred = noise_pred_uncond + self.cfg_scale * (noise_pred_text - noise_pred_uncond)
+
+                        #del noise_pred_text
+                        #del noise_pred_uncond
+                        #import gc
+                        #gc.collect()
+                        #torch.cuda.empty_cache()
+
+                        #if do_classifier_free_guidance and guidance_rescale > 0.0:
+                        #    # Based on 3.4. in https://arxiv.org/pdf/2305.08891.pdf
+                        #    noise_pred = rescale_noise_cfg(noise_pred, noise_pred_text, guidance_rescale=guidance_rescale)
+
+                        # compute the previous noisy sample x_t -> x_t-1
+                        latents = scheduler.step(noise_pred, t, latents, **extra_step_kwargs, return_dict=False)[0]
+
+                        #del noise_pred
+                        #import gc
+                        #gc.collect()
+                        #torch.cuda.empty_cache()
+
+                        # call the callback, if provided
+                        if i == len(timesteps) - 1 or ((i + 1) > num_warmup_steps and (i + 1) % scheduler.order == 0):
+                            progress_bar.update()
+                            self.dispatch_progress(context, source_node_id, latents, i, num_inference_steps)
+                            #if callback is not None and i % callback_steps == 0:
+                            #    callback(i, t, latents)
+
+
+
+        #################
+
+        latents = latents.to("cpu")
+        torch.cuda.empty_cache()
+
+        name = f'{context.graph_execution_state_id}__{self.id}'
+        context.services.latents.save(name, latents)
+        return build_latents_output(latents_name=name, latents=latents)

invokeai/app/invocations/upscale.py

@@ -1,48 +1,119 @@
-# Copyright (c) 2022 Kyle Schouviller (https://github.com/kyle0654)
-
-from typing import Literal, Optional
+# Copyright (c) 2022 Kyle Schouviller (https://github.com/kyle0654) & the InvokeAI Team
+from pathlib import Path
+from typing import Literal, Union
 
+import cv2 as cv
+import numpy as np
+from basicsr.archs.rrdbnet_arch import RRDBNet
+from PIL import Image
 from pydantic import Field
+from realesrgan import RealESRGANer
 
 from invokeai.app.models.image import ImageCategory, ImageField, ResourceOrigin
-from .baseinvocation import BaseInvocation, InvocationContext, InvocationConfig
+
+from .baseinvocation import BaseInvocation, InvocationConfig, InvocationContext
 from .image import ImageOutput
 
+# TODO: Populate this from disk?
+# TODO: Use model manager to load?
+ESRGAN_MODELS = Literal[
+    "RealESRGAN_x4plus.pth",
+    "RealESRGAN_x4plus_anime_6B.pth",
+    "ESRGAN_SRx4_DF2KOST_official-ff704c30.pth",
+    "RealESRGAN_x2plus.pth",
+]
 
-class UpscaleInvocation(BaseInvocation):
-    """Upscales an image."""
 
-    # fmt: off
-    type: Literal["upscale"] = "upscale"
+class ESRGANInvocation(BaseInvocation):
+    """Upscales an image using RealESRGAN."""
 
-    # Inputs
-    image: Optional[ImageField] = Field(description="The input image", default=None)
-    strength: float = Field(default=0.75, gt=0, le=1, description="The strength")
-    level: Literal[2, 4] = Field(default=2, description="The upscale level")
-    # fmt: on
+    type: Literal["esrgan"] = "esrgan"
+    image: Union[ImageField, None] = Field(default=None, description="The input image")
+    model_name: ESRGAN_MODELS = Field(
+        default="RealESRGAN_x4plus.pth", description="The Real-ESRGAN model to use"
+    )
 
-    # Schema customisation
     class Config(InvocationConfig):
         schema_extra = {
             "ui": {
-                "tags": ["upscaling", "image"],
+                "title": "Upscale (RealESRGAN)",
+                "tags": ["image", "upscale", "realesrgan"]
             },
         }
 
     def invoke(self, context: InvocationContext) -> ImageOutput:
         image = context.services.images.get_pil_image(self.image.image_name)
-        results = context.services.restoration.upscale_and_reconstruct(
-            image_list=[[image, 0]],
-            upscale=(self.level, self.strength),
-            strength=0.0,  # GFPGAN strength
-            save_original=False,
-            image_callback=None,
+        models_path = context.services.configuration.models_path
+
+        rrdbnet_model = None
+        netscale = None
+        esrgan_model_path = None
+
+        if self.model_name in [
+            "RealESRGAN_x4plus.pth",
+            "ESRGAN_SRx4_DF2KOST_official-ff704c30.pth",
+        ]:
+            # x4 RRDBNet model
+            rrdbnet_model = RRDBNet(
+                num_in_ch=3,
+                num_out_ch=3,
+                num_feat=64,
+                num_block=23,
+                num_grow_ch=32,
+                scale=4,
+            )
+            netscale = 4
+        elif self.model_name in ["RealESRGAN_x4plus_anime_6B.pth"]:
+            # x4 RRDBNet model, 6 blocks
+            rrdbnet_model = RRDBNet(
+                num_in_ch=3,
+                num_out_ch=3,
+                num_feat=64,
+                num_block=6,  # 6 blocks
+                num_grow_ch=32,
+                scale=4,
+            )
+            netscale = 4
+        elif self.model_name in ["RealESRGAN_x2plus.pth"]:
+            # x2 RRDBNet model
+            rrdbnet_model = RRDBNet(
+                num_in_ch=3,
+                num_out_ch=3,
+                num_feat=64,
+                num_block=23,
+                num_grow_ch=32,
+                scale=2,
+            )
+            netscale = 2
+        else:
+            msg = f"Invalid RealESRGAN model: {self.model_name}"
+            context.services.logger.error(msg)
+            raise ValueError(msg)
+
+        esrgan_model_path = Path(f"core/upscaling/realesrgan/{self.model_name}")
+
+        upsampler = RealESRGANer(
+            scale=netscale,
+            model_path=str(models_path / esrgan_model_path),
+            model=rrdbnet_model,
+            half=False,
         )
 
-        # Results are image and seed, unwrap for now
-        # TODO: can this return multiple results?
+        # prepare image - Real-ESRGAN uses cv2 internally, and cv2 uses BGR vs RGB for PIL
+        cv_image = cv.cvtColor(np.array(image.convert("RGB")), cv.COLOR_RGB2BGR)
+
+        # We can pass an `outscale` value here, but it just resizes the image by that factor after
+        # upscaling, so it's kinda pointless for our purposes. If you want something other than 4x
+        # upscaling, you'll need to add a resize node after this one.
+        upscaled_image, img_mode = upsampler.enhance(cv_image)
+
+        # back to PIL
+        pil_image = Image.fromarray(
+            cv.cvtColor(upscaled_image, cv.COLOR_BGR2RGB)
+        ).convert("RGBA")
+
         image_dto = context.services.images.create(
-            image=results[0][0],
+            image=pil_image,
             image_origin=ResourceOrigin.INTERNAL,
             image_category=ImageCategory.GENERAL,
             node_id=self.id,

invokeai/app/models/metadata.py

@@ -1,93 +0,0 @@
-from typing import Optional, Union, List
-from pydantic import BaseModel, Extra, Field, StrictFloat, StrictInt, StrictStr
-
-
-class ImageMetadata(BaseModel):
-    """
-    Core generation metadata for an image/tensor generated in InvokeAI.
-
-    Also includes any metadata from the image's PNG tEXt chunks.
-
-    Generated by traversing the execution graph, collecting the parameters of the nearest ancestors
-    of a given node.
-
-    Full metadata may be accessed by querying for the session in the `graph_executions` table.
-    """
-
-    class Config:
-        extra = Extra.allow
-        """
-        This lets the ImageMetadata class accept arbitrary additional fields. The CoreMetadataService
-        won't add any fields that are not already defined, but other a different metadata service
-        implementation might.
-        """
-
-    type: Optional[StrictStr] = Field(
-        default=None,
-        description="The type of the ancestor node of the image output node.",
-    )
-    """The type of the ancestor node of the image output node."""
-    positive_conditioning: Optional[StrictStr] = Field(
-        default=None, description="The positive conditioning."
-    )
-    """The positive conditioning"""
-    negative_conditioning: Optional[StrictStr] = Field(
-        default=None, description="The negative conditioning."
-    )
-    """The negative conditioning"""
-    width: Optional[StrictInt] = Field(
-        default=None, description="Width of the image/latents in pixels."
-    )
-    """Width of the image/latents in pixels"""
-    height: Optional[StrictInt] = Field(
-        default=None, description="Height of the image/latents in pixels."
-    )
-    """Height of the image/latents in pixels"""
-    seed: Optional[StrictInt] = Field(
-        default=None, description="The seed used for noise generation."
-    )
-    """The seed used for noise generation"""
-    # cfg_scale: Optional[StrictFloat] = Field(
-    # cfg_scale: Union[float, list[float]] = Field(
-    cfg_scale: Union[StrictFloat, List[StrictFloat]] = Field(
-        default=None, description="The classifier-free guidance scale."
-    )
-    """The classifier-free guidance scale"""
-    steps: Optional[StrictInt] = Field(
-        default=None, description="The number of steps used for inference."
-    )
-    """The number of steps used for inference"""
-    scheduler: Optional[StrictStr] = Field(
-        default=None, description="The scheduler used for inference."
-    )
-    """The scheduler used for inference"""
-    model: Optional[StrictStr] = Field(
-        default=None, description="The model used for inference."
-    )
-    """The model used for inference"""
-    strength: Optional[StrictFloat] = Field(
-        default=None,
-        description="The strength used for image-to-image/latents-to-latents.",
-    )
-    """The strength used for image-to-image/latents-to-latents."""
-    latents: Optional[StrictStr] = Field(
-        default=None, description="The ID of the initial latents."
-    )
-    """The ID of the initial latents"""
-    vae: Optional[StrictStr] = Field(
-        default=None, description="The VAE used for decoding."
-    )
-    """The VAE used for decoding"""
-    unet: Optional[StrictStr] = Field(
-        default=None, description="The UNet used dor inference."
-    )
-    """The UNet used dor inference"""
-    clip: Optional[StrictStr] = Field(
-        default=None, description="The CLIP Encoder used for conditioning."
-    )
-    """The CLIP Encoder used for conditioning"""
-    extra: Optional[StrictStr] = Field(
-        default=None,
-        description="Uploaded image metadata, extracted from the PNG tEXt chunk.",
-    )
-    """Uploaded image metadata, extracted from the PNG tEXt chunk."""

invokeai/app/services/board_image_record_storage.py

@@ -32,11 +32,11 @@ class BoardImageRecordStorageBase(ABC):
         pass
 
     @abstractmethod
-    def get_images_for_board(
+    def get_all_board_image_names_for_board(
         self,
         board_id: str,
-    ) -> OffsetPaginatedResults[ImageRecord]:
-        """Gets images for a board."""
+    ) -> list[str]:
+        """Gets all board images for a board, as a list of the image names."""
         pass
 
     @abstractmethod
@@ -211,6 +211,26 @@ class SqliteBoardImageRecordStorage(BoardImageRecordStorageBase):
             items=images, offset=offset, limit=limit, total=count
         )
 
+    def get_all_board_image_names_for_board(self, board_id: str) -> list[str]:
+        try:
+            self._lock.acquire()
+            self._cursor.execute(
+                """--sql
+                SELECT image_name
+                FROM board_images
+                WHERE board_id = ?;
+                """,
+                (board_id,),
+            )
+            result = cast(list[sqlite3.Row], self._cursor.fetchall())
+            image_names = list(map(lambda r: r[0], result))
+            return image_names
+        except sqlite3.Error as e:
+            self._conn.rollback()
+            raise e
+        finally:
+            self._lock.release()
+
     def get_board_for_image(
         self,
         image_name: str,

invokeai/app/services/board_images.py

@@ -38,11 +38,11 @@ class BoardImagesServiceABC(ABC):
         pass
 
     @abstractmethod
-    def get_images_for_board(
+    def get_all_board_image_names_for_board(
         self,
         board_id: str,
-    ) -> OffsetPaginatedResults[ImageDTO]:
-        """Gets images for a board."""
+    ) -> list[str]:
+        """Gets all board images for a board, as a list of the image names."""
         pass
 
     @abstractmethod
@@ -98,30 +98,13 @@ class BoardImagesService(BoardImagesServiceABC):
     ) -> None:
         self._services.board_image_records.remove_image_from_board(board_id, image_name)
 
-    def get_images_for_board(
+    def get_all_board_image_names_for_board(
         self,
         board_id: str,
-    ) -> OffsetPaginatedResults[ImageDTO]:
-        image_records = self._services.board_image_records.get_images_for_board(
+    ) -> list[str]:
+        return self._services.board_image_records.get_all_board_image_names_for_board(
             board_id
         )
-        image_dtos = list(
-            map(
-                lambda r: image_record_to_dto(
-                    r,
-                    self._services.urls.get_image_url(r.image_name),
-                    self._services.urls.get_image_url(r.image_name, True),
-                    board_id,
-                ),
-                image_records.items,
-            )
-        )
-        return OffsetPaginatedResults[ImageDTO](
-            items=image_dtos,
-            offset=image_records.offset,
-            limit=image_records.limit,
-            total=image_records.total,
-        )
 
     def get_board_for_image(
         self,
@@ -136,7 +119,7 @@ def board_record_to_dto(
 ) -> BoardDTO:
     """Converts a board record to a board DTO."""
     return BoardDTO(
-        **board_record.dict(exclude={'cover_image_name'}),
+        **board_record.dict(exclude={"cover_image_name"}),
         cover_image_name=cover_image_name,
         image_count=image_count,
     )

invokeai/app/services/config.py

@@ -200,7 +200,7 @@ class InvokeAISettings(BaseSettings):
         type = get_args(get_type_hints(cls)['type'])[0]
         field_dict = dict({type:dict()})
         for name,field in self.__fields__.items():
-            if name in cls._excluded():
+            if name in cls._excluded_from_yaml():
                 continue
             category = field.field_info.extra.get("category") or "Uncategorized"
             value = getattr(self,name)
@@ -271,8 +271,13 @@ class InvokeAISettings(BaseSettings):
 
     @classmethod
     def _excluded(self)->List[str]:
-        # combination of deprecated parameters and internal ones
-        return ['type','initconf', 'gpu_mem_reserved', 'max_loaded_models', 'version']
+        # internal fields that shouldn't be exposed as command line options
+        return ['type','initconf']
+    
+    @classmethod
+    def _excluded_from_yaml(self)->List[str]:
+        # combination of deprecated parameters and internal ones that shouldn't be exposed as invokeai.yaml options
+        return ['type','initconf', 'gpu_mem_reserved', 'max_loaded_models', 'version', 'from_file', 'model', 'restore', 'root']
 
     class Config:
         env_file_encoding = 'utf-8'
@@ -361,7 +366,7 @@ setting environment variables INVOKEAI_<setting>.
     log_tokenization    : bool = Field(default=False, description="Enable logging of parsed prompt tokens.", category='Features')
     nsfw_checker        : bool = Field(default=True, description="Enable/disable the NSFW checker", category='Features')
     patchmatch          : bool = Field(default=True, description="Enable/disable patchmatch inpaint code", category='Features')
-    restore             : bool = Field(default=True, description="Enable/disable face restoration code", category='Features')
+    restore             : bool = Field(default=True, description="Enable/disable face restoration code (DEPRECATED)", category='DEPRECATED')
 
     always_use_cpu      : bool = Field(default=False, description="If true, use the CPU for rendering even if a GPU is available.", category='Memory/Performance')
     free_gpu_mem        : bool = Field(default=False, description="If true, purge model from GPU after each generation.", category='Memory/Performance')
@@ -369,16 +374,16 @@ setting environment variables INVOKEAI_<setting>.
     max_cache_size      : float = Field(default=6.0, gt=0, description="Maximum memory amount used by model cache for rapid switching", category='Memory/Performance')
     max_vram_cache_size : float = Field(default=2.75, ge=0, description="Amount of VRAM reserved for model storage", category='Memory/Performance')
     gpu_mem_reserved    : float = Field(default=2.75, ge=0, description="DEPRECATED: use max_vram_cache_size. Amount of VRAM reserved for model storage", category='DEPRECATED')
-    precision           : Literal[tuple(['auto','float16','float32','autocast'])] = Field(default='float16',description='Floating point precision', category='Memory/Performance')
+    precision           : Literal[tuple(['auto','float16','float32','autocast'])] = Field(default='auto',description='Floating point precision', category='Memory/Performance')
     sequential_guidance : bool = Field(default=False, description="Whether to calculate guidance in serial instead of in parallel, lowering memory requirements", category='Memory/Performance')
     xformers_enabled    : bool = Field(default=True, description="Enable/disable memory-efficient attention", category='Memory/Performance')
     tiled_decode        : bool = Field(default=False, description="Whether to enable tiled VAE decode (reduces memory consumption with some performance penalty)", category='Memory/Performance')
 
     root                : Path = Field(default=_find_root(), description='InvokeAI runtime root directory', category='Paths')
-    autoimport_dir      : Path = Field(default='autoimport/main', description='Path to a directory of models files to be imported on startup.', category='Paths')
-    lora_dir            : Path = Field(default='autoimport/lora', description='Path to a directory of LoRA/LyCORIS models to be imported on startup.', category='Paths')
-    embedding_dir       : Path = Field(default='autoimport/embedding', description='Path to a directory of Textual Inversion embeddings to be imported on startup.', category='Paths')
-    controlnet_dir      : Path = Field(default='autoimport/controlnet', description='Path to a directory of ControlNet embeddings to be imported on startup.', category='Paths')
+    autoimport_dir      : Path = Field(default='autoimport', description='Path to a directory of models files to be imported on startup.', category='Paths')
+    lora_dir            : Path = Field(default=None, description='Path to a directory of LoRA/LyCORIS models to be imported on startup.', category='Paths')
+    embedding_dir       : Path = Field(default=None, description='Path to a directory of Textual Inversion embeddings to be imported on startup.', category='Paths')
+    controlnet_dir      : Path = Field(default=None, description='Path to a directory of ControlNet embeddings to be imported on startup.', category='Paths')
     conf_path           : Path = Field(default='configs/models.yaml', description='Path to models definition file', category='Paths')
     models_dir          : Path = Field(default='models', description='Path to the models directory', category='Paths')
     legacy_conf_dir     : Path = Field(default='configs/stable-diffusion', description='Path to directory of legacy checkpoint config files', category='Paths')
@@ -392,7 +397,7 @@ setting environment variables INVOKEAI_<setting>.
     log_handlers        : List[str] = Field(default=["console"], description='Log handler. Valid options are "console", "file=<path>", "syslog=path|address:host:port", "http=<url>"', category="Logging")
     # note - would be better to read the log_format values from logging.py, but this creates circular dependencies issues
     log_format          : Literal[tuple(['plain','color','syslog','legacy'])] = Field(default="color", description='Log format. Use "plain" for text-only, "color" for colorized output, "legacy" for 2.3-style logging and "syslog" for syslog-style', category="Logging")
-    log_level           : Literal[tuple(["debug","info","warning","error","critical"])] = Field(default="debug", description="Emit logging messages at this level or  higher", category="Logging")
+    log_level           : Literal[tuple(["debug","info","warning","error","critical"])] = Field(default="info", description="Emit logging messages at this level or  higher", category="Logging")
 
     version             : bool = Field(default=False, description="Show InvokeAI version and exit", category="Other")
     #fmt: on
@@ -441,7 +446,7 @@ setting environment variables INVOKEAI_<setting>.
         Path to the runtime root directory
         '''
         if self.root:
-            return Path(self.root).expanduser()
+            return Path(self.root).expanduser().absolute()
         else:
             return self.find_root()
 

invokeai/app/services/events.py

@@ -105,8 +105,6 @@ class EventServiceBase:
     def emit_model_load_started (
             self,
             graph_execution_state_id: str,
-            node: dict,
-            source_node_id: str,
             model_name: str,
             base_model: BaseModelType,
             model_type: ModelType,
@@ -117,8 +115,6 @@ class EventServiceBase:
             event_name="model_load_started",
             payload=dict(
                 graph_execution_state_id=graph_execution_state_id,
-                node=node,
-                source_node_id=source_node_id,
                 model_name=model_name,
                 base_model=base_model,
                 model_type=model_type,
@@ -129,8 +125,6 @@ class EventServiceBase:
     def emit_model_load_completed(
             self,
             graph_execution_state_id: str,
-            node: dict,
-            source_node_id: str,
             model_name: str,
             base_model: BaseModelType,
             model_type: ModelType,
@@ -142,12 +136,12 @@ class EventServiceBase:
             event_name="model_load_completed",
             payload=dict(
                 graph_execution_state_id=graph_execution_state_id,
-                node=node,
-                source_node_id=source_node_id,
                 model_name=model_name,
                 base_model=base_model,
                 model_type=model_type,
                 submodel=submodel,
-                model_info=model_info,
+                hash=model_info.hash,
+                location=str(model_info.location),
+                precision=str(model_info.precision),
             ),
         )

invokeai/app/services/image_file_storage.py

@@ -1,14 +1,14 @@
 # Copyright (c) 2022 Kyle Schouviller (https://github.com/kyle0654) and the InvokeAI Team
+import json
 from abc import ABC, abstractmethod
 from pathlib import Path
 from queue import Queue
 from typing import Dict, Optional, Union
 
-from PIL.Image import Image as PILImageType
 from PIL import Image, PngImagePlugin
+from PIL.Image import Image as PILImageType
 from send2trash import send2trash
 
-from invokeai.app.models.metadata import ImageMetadata
 from invokeai.app.util.thumbnails import get_thumbnail_name, make_thumbnail
 
 
@@ -59,7 +59,8 @@ class ImageFileStorageBase(ABC):
         self,
         image: PILImageType,
         image_name: str,
-        metadata: Optional[ImageMetadata] = None,
+        metadata: Optional[dict] = None,
+        graph: Optional[dict] = None,
         thumbnail_size: int = 256,
     ) -> None:
         """Saves an image and a 256x256 WEBP thumbnail. Returns a tuple of the image name, thumbnail name, and created timestamp."""
@@ -110,20 +111,22 @@ class DiskImageFileStorage(ImageFileStorageBase):
         self,
         image: PILImageType,
         image_name: str,
-        metadata: Optional[ImageMetadata] = None,
+        metadata: Optional[dict] = None,
+        graph: Optional[dict] = None,
         thumbnail_size: int = 256,
     ) -> None:
         try:
             self.__validate_storage_folders()
             image_path = self.get_path(image_name)
 
+            pnginfo = PngImagePlugin.PngInfo()
+            
             if metadata is not None:
-                pnginfo = PngImagePlugin.PngInfo()
-                pnginfo.add_text("invokeai", metadata.json())
-                image.save(image_path, "PNG", pnginfo=pnginfo)
-            else:
-                image.save(image_path, "PNG")
+                pnginfo.add_text("invokeai_metadata", json.dumps(metadata))
+            if graph is not None:
+                pnginfo.add_text("invokeai_graph", json.dumps(graph))
 
+            image.save(image_path, "PNG", pnginfo=pnginfo)
             thumbnail_name = get_thumbnail_name(image_name)
             thumbnail_path = self.get_path(thumbnail_name, thumbnail=True)
             thumbnail_image = make_thumbnail(image, thumbnail_size)

invokeai/app/services/image_record_storage.py

@@ -1,3 +1,4 @@
+import json
 import sqlite3
 import threading
 from abc import ABC, abstractmethod
@@ -8,9 +9,11 @@ from pydantic import BaseModel, Field
 from pydantic.generics import GenericModel
 
 from invokeai.app.models.image import ImageCategory, ResourceOrigin
-from invokeai.app.models.metadata import ImageMetadata
 from invokeai.app.services.models.image_record import (
-    ImageRecord, ImageRecordChanges, deserialize_image_record)
+    ImageRecord,
+    ImageRecordChanges,
+    deserialize_image_record,
+)
 
 T = TypeVar("T", bound=BaseModel)
 
@@ -48,6 +51,28 @@ class ImageRecordDeleteException(Exception):
         super().__init__(message)
 
 
+IMAGE_DTO_COLS = ", ".join(
+    list(
+        map(
+            lambda c: "images." + c,
+            [
+                "image_name",
+                "image_origin",
+                "image_category",
+                "width",
+                "height",
+                "session_id",
+                "node_id",
+                "is_intermediate",
+                "created_at",
+                "updated_at",
+                "deleted_at",
+            ],
+        )
+    )
+)
+
+
 class ImageRecordStorageBase(ABC):
     """Low-level service responsible for interfacing with the image record store."""
 
@@ -58,6 +83,11 @@ class ImageRecordStorageBase(ABC):
         """Gets an image record."""
         pass
 
+    @abstractmethod
+    def get_metadata(self, image_name: str) -> Optional[dict]:
+        """Gets an image's metadata'."""
+        pass
+
     @abstractmethod
     def update(
         self,
@@ -70,8 +100,8 @@ class ImageRecordStorageBase(ABC):
     @abstractmethod
     def get_many(
         self,
-        offset: int = 0,
-        limit: int = 10,
+        offset: Optional[int] = None,
+        limit: Optional[int] = None,
         image_origin: Optional[ResourceOrigin] = None,
         categories: Optional[list[ImageCategory]] = None,
         is_intermediate: Optional[bool] = None,
@@ -92,6 +122,11 @@ class ImageRecordStorageBase(ABC):
         """Deletes many image records."""
         pass
 
+    @abstractmethod
+    def delete_intermediates(self) -> list[str]:
+        """Deletes all intermediate image records, returning a list of deleted image names."""
+        pass
+
     @abstractmethod
     def save(
         self,
@@ -102,7 +137,7 @@ class ImageRecordStorageBase(ABC):
         height: int,
         session_id: Optional[str],
         node_id: Optional[str],
-        metadata: Optional[ImageMetadata],
+        metadata: Optional[dict],
         is_intermediate: bool = False,
     ) -> datetime:
         """Saves an image record."""
@@ -206,7 +241,7 @@ class SqliteImageRecordStorage(ImageRecordStorageBase):
 
             self._cursor.execute(
                 f"""--sql
-                SELECT * FROM images
+                SELECT {IMAGE_DTO_COLS} FROM images
                 WHERE image_name = ?;
                 """,
                 (image_name,),
@@ -224,6 +259,28 @@ class SqliteImageRecordStorage(ImageRecordStorageBase):
 
         return deserialize_image_record(dict(result))
 
+    def get_metadata(self, image_name: str) -> Optional[dict]:
+        try:
+            self._lock.acquire()
+
+            self._cursor.execute(
+                f"""--sql
+                SELECT images.metadata FROM images
+                WHERE image_name = ?;
+                """,
+                (image_name,),
+            )
+
+            result = cast(Optional[sqlite3.Row], self._cursor.fetchone())
+            if not result or not result[0]:
+                return None
+            return json.loads(result[0])
+        except sqlite3.Error as e:
+            self._conn.rollback()
+            raise ImageRecordNotFoundException from e
+        finally:
+            self._lock.release()
+
     def update(
         self,
         image_name: str,
@@ -273,8 +330,8 @@ class SqliteImageRecordStorage(ImageRecordStorageBase):
 
     def get_many(
         self,
-        offset: int = 0,
-        limit: int = 10,
+        offset: Optional[int] = None,
+        limit: Optional[int] = None,
         image_origin: Optional[ResourceOrigin] = None,
         categories: Optional[list[ImageCategory]] = None,
         is_intermediate: Optional[bool] = None,
@@ -291,8 +348,8 @@ class SqliteImageRecordStorage(ImageRecordStorageBase):
             WHERE 1=1
             """
 
-            images_query = """--sql
-            SELECT images.*
+            images_query = f"""--sql
+            SELECT {IMAGE_DTO_COLS}
             FROM images
             LEFT JOIN board_images ON board_images.image_name = images.image_name
             WHERE 1=1
@@ -328,11 +385,15 @@ class SqliteImageRecordStorage(ImageRecordStorageBase):
 
                 query_params.append(is_intermediate)
 
-            if board_id is not None:
+            # board_id of "none" is reserved for images without a board
+            if board_id == "none":
+                query_conditions += """--sql
+                AND board_images.board_id IS NULL
+                """
+            elif board_id is not None:
                 query_conditions += """--sql
                 AND board_images.board_id = ?
                 """
-
                 query_params.append(board_id)
 
             query_pagination = """--sql
@@ -343,8 +404,12 @@ class SqliteImageRecordStorage(ImageRecordStorageBase):
             images_query += query_conditions + query_pagination + ";"
             # Add all the parameters
             images_params = query_params.copy()
-            images_params.append(limit)
-            images_params.append(offset)
+
+            if limit is not None:
+                images_params.append(limit)
+            if offset is not None:
+                images_params.append(offset)
+
             # Build the list of images, deserializing each row
             self._cursor.execute(images_query, images_params)
             result = cast(list[sqlite3.Row], self._cursor.fetchall())
@@ -401,6 +466,32 @@ class SqliteImageRecordStorage(ImageRecordStorageBase):
         finally:
             self._lock.release()
 
+
+    def delete_intermediates(self) -> list[str]:
+        try:
+            self._lock.acquire()
+            self._cursor.execute(
+                """--sql
+                SELECT image_name FROM images
+                WHERE is_intermediate = TRUE;
+                """
+            )
+            result = cast(list[sqlite3.Row], self._cursor.fetchall())
+            image_names = list(map(lambda r: r[0], result))
+            self._cursor.execute(
+                """--sql
+                DELETE FROM images
+                WHERE is_intermediate = TRUE;
+                """
+            )
+            self._conn.commit()
+            return image_names
+        except sqlite3.Error as e:
+            self._conn.rollback()
+            raise ImageRecordDeleteException from e
+        finally:
+            self._lock.release()
+
     def save(
         self,
         image_name: str,
@@ -410,12 +501,12 @@ class SqliteImageRecordStorage(ImageRecordStorageBase):
         width: int,
         height: int,
         node_id: Optional[str],
-        metadata: Optional[ImageMetadata],
+        metadata: Optional[dict],
         is_intermediate: bool = False,
     ) -> datetime:
         try:
             metadata_json = (
-                None if metadata is None else metadata.json(exclude_none=True)
+                None if metadata is None else json.dumps(metadata)
             )
             self._lock.acquire()
             self._cursor.execute(
@@ -465,9 +556,7 @@ class SqliteImageRecordStorage(ImageRecordStorageBase):
         finally:
             self._lock.release()
 
-    def get_most_recent_image_for_board(
-        self, board_id: str
-    ) -> Optional[ImageRecord]:
+    def get_most_recent_image_for_board(self, board_id: str) -> Optional[ImageRecord]:
         try:
             self._lock.acquire()
             self._cursor.execute(

invokeai/app/services/images.py

@@ -1,16 +1,24 @@
+import json
 from abc import ABC, abstractmethod
 from logging import Logger
-from typing import Optional, TYPE_CHECKING, Union
+from typing import TYPE_CHECKING, Optional
+
 from PIL.Image import Image as PILImageType
 
+from invokeai.app.invocations.metadata import ImageMetadata
 from invokeai.app.models.image import (
     ImageCategory,
-    ResourceOrigin,
     InvalidImageCategoryException,
     InvalidOriginException,
+    ResourceOrigin,
 )
-from invokeai.app.models.metadata import ImageMetadata
 from invokeai.app.services.board_image_record_storage import BoardImageRecordStorageBase
+from invokeai.app.services.image_file_storage import (
+    ImageFileDeleteException,
+    ImageFileNotFoundException,
+    ImageFileSaveException,
+    ImageFileStorageBase,
+)
 from invokeai.app.services.image_record_storage import (
     ImageRecordDeleteException,
     ImageRecordNotFoundException,
@@ -18,22 +26,16 @@ from invokeai.app.services.image_record_storage import (
     ImageRecordStorageBase,
     OffsetPaginatedResults,
 )
+from invokeai.app.services.item_storage import ItemStorageABC
 from invokeai.app.services.models.image_record import (
-    ImageRecord,
     ImageDTO,
+    ImageRecord,
     ImageRecordChanges,
     image_record_to_dto,
 )
-from invokeai.app.services.image_file_storage import (
-    ImageFileDeleteException,
-    ImageFileNotFoundException,
-    ImageFileSaveException,
-    ImageFileStorageBase,
-)
-from invokeai.app.services.item_storage import ItemStorageABC, PaginatedResults
-from invokeai.app.services.metadata import MetadataServiceBase
 from invokeai.app.services.resource_name import NameServiceBase
 from invokeai.app.services.urls import UrlServiceBase
+from invokeai.app.util.metadata import get_metadata_graph_from_raw_session
 
 if TYPE_CHECKING:
     from invokeai.app.services.graph import GraphExecutionState
@@ -50,7 +52,9 @@ class ImageServiceABC(ABC):
         image_category: ImageCategory,
         node_id: Optional[str] = None,
         session_id: Optional[str] = None,
+        board_id: Optional[str] = None,
         is_intermediate: bool = False,
+        metadata: Optional[dict] = None,
     ) -> ImageDTO:
         """Creates an image, storing the file and its metadata."""
         pass
@@ -79,6 +83,11 @@ class ImageServiceABC(ABC):
         """Gets an image DTO."""
         pass
 
+    @abstractmethod
+    def get_metadata(self, image_name: str) -> ImageMetadata:
+        """Gets an image's metadata."""
+        pass
+
     @abstractmethod
     def get_path(self, image_name: str, thumbnail: bool = False) -> str:
         """Gets an image's path."""
@@ -112,6 +121,11 @@ class ImageServiceABC(ABC):
         """Deletes an image."""
         pass
 
+    @abstractmethod
+    def delete_intermediates(self) -> int:
+        """Deletes all intermediate images."""
+        pass
+
     @abstractmethod
     def delete_images_on_board(self, board_id: str):
         """Deletes all images on a board."""
@@ -124,7 +138,6 @@ class ImageServiceDependencies:
     image_records: ImageRecordStorageBase
     image_files: ImageFileStorageBase
     board_image_records: BoardImageRecordStorageBase
-    metadata: MetadataServiceBase
     urls: UrlServiceBase
     logger: Logger
     names: NameServiceBase
@@ -135,7 +148,6 @@ class ImageServiceDependencies:
         image_record_storage: ImageRecordStorageBase,
         image_file_storage: ImageFileStorageBase,
         board_image_record_storage: BoardImageRecordStorageBase,
-        metadata: MetadataServiceBase,
         url: UrlServiceBase,
         logger: Logger,
         names: NameServiceBase,
@@ -144,7 +156,6 @@ class ImageServiceDependencies:
         self.image_records = image_record_storage
         self.image_files = image_file_storage
         self.board_image_records = board_image_record_storage
-        self.metadata = metadata
         self.urls = url
         self.logger = logger
         self.names = names
@@ -164,7 +175,9 @@ class ImageService(ImageServiceABC):
         image_category: ImageCategory,
         node_id: Optional[str] = None,
         session_id: Optional[str] = None,
+        board_id: Optional[str] = None,
         is_intermediate: bool = False,
+        metadata: Optional[dict] = None,
     ) -> ImageDTO:
         if image_origin not in ResourceOrigin:
             raise InvalidOriginException
@@ -174,7 +187,16 @@ class ImageService(ImageServiceABC):
 
         image_name = self._services.names.create_image_name()
 
-        metadata = self._get_metadata(session_id, node_id)
+        graph = None
+
+        if session_id is not None:
+            session_raw = self._services.graph_execution_manager.get_raw(session_id)
+            if session_raw is not None:
+                try:
+                    graph = get_metadata_graph_from_raw_session(session_raw)
+                except Exception as e:
+                    self._services.logger.warn(f"Failed to parse session graph: {e}")
+                    graph = None
 
         (width, height) = image.size
 
@@ -191,14 +213,17 @@ class ImageService(ImageServiceABC):
                 is_intermediate=is_intermediate,
                 # Nullable fields
                 node_id=node_id,
-                session_id=session_id,
                 metadata=metadata,
+                session_id=session_id,
             )
 
+            if board_id is not None:
+                self._services.board_image_records.add_image_to_board(
+                    board_id=board_id, image_name=image_name
+                )
+
             self._services.image_files.save(
-                image_name=image_name,
-                image=image,
-                metadata=metadata,
+                image_name=image_name, image=image, metadata=metadata, graph=graph
             )
 
             image_dto = self.get_dto(image_name)
@@ -268,6 +293,34 @@ class ImageService(ImageServiceABC):
             self._services.logger.error("Problem getting image DTO")
             raise e
 
+    def get_metadata(self, image_name: str) -> Optional[ImageMetadata]:
+        try:
+            image_record = self._services.image_records.get(image_name)
+
+            if not image_record.session_id:
+                return ImageMetadata()
+
+            session_raw = self._services.graph_execution_manager.get_raw(
+                image_record.session_id
+            )
+            graph = None
+
+            if session_raw:
+                try:
+                    graph = get_metadata_graph_from_raw_session(session_raw)
+                except Exception as e:
+                    self._services.logger.warn(f"Failed to parse session graph: {e}")
+                    graph = None
+
+            metadata = self._services.image_records.get_metadata(image_name)
+            return ImageMetadata(graph=graph, metadata=metadata)
+        except ImageRecordNotFoundException:
+            self._services.logger.error("Image record not found")
+            raise
+        except Exception as e:
+            self._services.logger.error("Problem getting image DTO")
+            raise e
+
     def get_path(self, image_name: str, thumbnail: bool = False) -> str:
         try:
             return self._services.image_files.get_path(image_name, thumbnail)
@@ -348,16 +401,14 @@ class ImageService(ImageServiceABC):
 
     def delete_images_on_board(self, board_id: str):
         try:
-            images = self._services.board_image_records.get_images_for_board(board_id)
-            image_name_list = list(
-                map(
-                    lambda r: r.image_name,
-                    images.items,
+            image_names = (
+                self._services.board_image_records.get_all_board_image_names_for_board(
+                    board_id
                 )
             )
-            for image_name in image_name_list:
+            for image_name in image_names:
                 self._services.image_files.delete(image_name)
-            self._services.image_records.delete_many(image_name_list)
+            self._services.image_records.delete_many(image_names)
         except ImageRecordDeleteException:
             self._services.logger.error(f"Failed to delete image records")
             raise
@@ -368,14 +419,19 @@ class ImageService(ImageServiceABC):
             self._services.logger.error("Problem deleting image records and files")
             raise e
 
-    def _get_metadata(
-        self, session_id: Optional[str] = None, node_id: Optional[str] = None
-    ) -> Optional[ImageMetadata]:
-        """Get the metadata for a node."""
-        metadata = None
-
-        if node_id is not None and session_id is not None:
-            session = self._services.graph_execution_manager.get(session_id)
-            metadata = self._services.metadata.create_image_metadata(session, node_id)
-
-        return metadata
+    def delete_intermediates(self) -> int:
+        try:
+            image_names = self._services.image_records.delete_intermediates()
+            count = len(image_names)
+            for image_name in image_names:
+                self._services.image_files.delete(image_name)
+            return count
+        except ImageRecordDeleteException:
+            self._services.logger.error(f"Failed to delete image records")
+            raise
+        except ImageFileDeleteException:
+            self._services.logger.error(f"Failed to delete image files")
+            raise
+        except Exception as e:
+            self._services.logger.error("Problem deleting image records and files")
+            raise e

invokeai/app/services/invocation_services.py

@@ -10,10 +10,9 @@ if TYPE_CHECKING:
     from invokeai.app.services.model_manager_service import ModelManagerServiceBase
     from invokeai.app.services.events import EventServiceBase
     from invokeai.app.services.latent_storage import LatentsStorageBase
-    from invokeai.app.services.restoration_services import RestorationServices
     from invokeai.app.services.invocation_queue import InvocationQueueABC
     from invokeai.app.services.item_storage import ItemStorageABC
-    from invokeai.app.services.config import InvokeAISettings
+    from invokeai.app.services.config import InvokeAIAppConfig
     from invokeai.app.services.graph import GraphExecutionState, LibraryGraph
     from invokeai.app.services.invoker import InvocationProcessorABC
 
@@ -24,7 +23,7 @@ class InvocationServices:
     # TODO: Just forward-declared everything due to circular dependencies. Fix structure.
     board_images: "BoardImagesServiceABC"
     boards: "BoardServiceABC"
-    configuration: "InvokeAISettings"
+    configuration: "InvokeAIAppConfig"
     events: "EventServiceBase"
     graph_execution_manager: "ItemStorageABC"["GraphExecutionState"]
     graph_library: "ItemStorageABC"["LibraryGraph"]
@@ -34,13 +33,12 @@ class InvocationServices:
     model_manager: "ModelManagerServiceBase"
     processor: "InvocationProcessorABC"
     queue: "InvocationQueueABC"
-    restoration: "RestorationServices"
 
     def __init__(
         self,
         board_images: "BoardImagesServiceABC",
         boards: "BoardServiceABC",
-        configuration: "InvokeAISettings",
+        configuration: "InvokeAIAppConfig",
         events: "EventServiceBase",
         graph_execution_manager: "ItemStorageABC"["GraphExecutionState"],
         graph_library: "ItemStorageABC"["LibraryGraph"],
@@ -50,7 +48,6 @@ class InvocationServices:
         model_manager: "ModelManagerServiceBase",
         processor: "InvocationProcessorABC",
         queue: "InvocationQueueABC",
-        restoration: "RestorationServices",
     ):
         self.board_images = board_images
         self.boards = boards
@@ -65,4 +62,3 @@ class InvocationServices:
         self.model_manager = model_manager
         self.processor = processor
         self.queue = queue
-        self.restoration = restoration

invokeai/app/services/item_storage.py

@@ -1,5 +1,5 @@
 from abc import ABC, abstractmethod
-from typing import Callable, Generic, TypeVar
+from typing import Callable, Generic, Optional, TypeVar
 
 from pydantic import BaseModel, Field
 from pydantic.generics import GenericModel
@@ -29,14 +29,22 @@ class ItemStorageABC(ABC, Generic[T]):
 
     @abstractmethod
     def get(self, item_id: str) -> T:
+        """Gets the item, parsing it into a Pydantic model"""
+        pass
+
+    @abstractmethod
+    def get_raw(self, item_id: str) -> Optional[str]:
+        """Gets the raw item as a string, skipping Pydantic parsing"""
         pass
 
     @abstractmethod
     def set(self, item: T) -> None:
+        """Sets the item"""
         pass
 
     @abstractmethod
     def list(self, page: int = 0, per_page: int = 10) -> PaginatedResults[T]:
+        """Gets a paginated list of items"""
         pass
 
     @abstractmethod

invokeai/app/services/metadata.py

@@ -1,142 +0,0 @@
-from abc import ABC, abstractmethod
-from typing import Any, Optional
-import networkx as nx
-
-from invokeai.app.models.metadata import ImageMetadata
-from invokeai.app.services.graph import Graph, GraphExecutionState
-
-
-class MetadataServiceBase(ABC):
-    """Handles building metadata for nodes, images, and outputs."""
-
-    @abstractmethod
-    def create_image_metadata(
-        self, session: GraphExecutionState, node_id: str
-    ) -> ImageMetadata:
-        """Builds an ImageMetadata object for a node."""
-        pass
-
-
-class CoreMetadataService(MetadataServiceBase):
-    _ANCESTOR_TYPES = ["t2l", "l2l"]
-    """The ancestor types that contain the core metadata"""
-
-    _ANCESTOR_PARAMS = ["type", "steps", "model", "cfg_scale", "scheduler", "strength"]
-    """The core metadata parameters in the ancestor types"""
-
-    _NOISE_FIELDS = ["seed", "width", "height"]
-    """The core metadata parameters in the noise node"""
-
-    def create_image_metadata(
-        self, session: GraphExecutionState, node_id: str
-    ) -> ImageMetadata:
-        metadata = self._build_metadata_from_graph(session, node_id)
-
-        return metadata
-
-    def _find_nearest_ancestor(self, G: nx.DiGraph, node_id: str) -> Optional[str]:
-        """
-        Finds the id of the nearest ancestor (of a valid type) of a given node.
-
-        Parameters:
-        G (nx.DiGraph): The execution graph, converted in to a networkx DiGraph. Its nodes must
-        have the same data as the execution graph.
-        node_id (str): The ID of the node.
-
-        Returns:
-        str | None: The ID of the nearest ancestor, or None if there are no valid ancestors.
-        """
-
-        # Retrieve the node from the graph
-        node = G.nodes[node_id]
-
-        # If the node type is one of the core metadata node types, return its id
-        if node.get("type") in self._ANCESTOR_TYPES:
-            return node.get("id")
-
-        # Else, look for the ancestor in the predecessor nodes
-        for predecessor in G.predecessors(node_id):
-            result = self._find_nearest_ancestor(G, predecessor)
-            if result:
-                return result
-
-        # If there are no valid ancestors, return None
-        return None
-
-    def _get_additional_metadata(
-        self, graph: Graph, node_id: str
-    ) -> Optional[dict[str, Any]]:
-        """
-        Returns additional metadata for a given node.
-
-        Parameters:
-        graph (Graph): The execution graph.
-        node_id (str): The ID of the node.
-
-        Returns:
-        dict[str, Any] | None: A dictionary of additional metadata.
-        """
-
-        metadata = {}
-
-        # Iterate over all edges in the graph
-        for edge in graph.edges:
-            dest_node_id = edge.destination.node_id
-            dest_field = edge.destination.field
-            source_node_dict = graph.nodes[edge.source.node_id].dict()
-
-            # If the destination node ID matches the given node ID, gather necessary metadata
-            if dest_node_id == node_id:
-                # Prompt
-                if dest_field == "positive_conditioning":
-                    metadata["positive_conditioning"] = source_node_dict.get("prompt")
-                # Negative prompt
-                if dest_field == "negative_conditioning":
-                    metadata["negative_conditioning"] = source_node_dict.get("prompt")
-                # Seed, width and height
-                if dest_field == "noise":
-                    for field in self._NOISE_FIELDS:
-                        metadata[field] = source_node_dict.get(field)
-        return metadata
-
-    def _build_metadata_from_graph(
-        self, session: GraphExecutionState, node_id: str
-    ) -> ImageMetadata:
-        """
-        Builds an ImageMetadata object for a node.
-
-        Parameters:
-        session (GraphExecutionState): The session.
-        node_id (str): The ID of the node.
-
-        Returns:
-        ImageMetadata: The metadata for the node.
-        """
-
-        # We need to do all the traversal on the execution graph
-        graph = session.execution_graph
-
-        # Find the nearest `t2l`/`l2l` ancestor of the given node
-        ancestor_id = self._find_nearest_ancestor(graph.nx_graph_with_data(), node_id)
-
-        # If no ancestor was found, return an empty ImageMetadata object
-        if ancestor_id is None:
-            return ImageMetadata()
-
-        ancestor_node = graph.get_node(ancestor_id)
-
-        # Grab all the core metadata from the ancestor node
-        ancestor_metadata = {
-            param: val
-            for param, val in ancestor_node.dict().items()
-            if param in self._ANCESTOR_PARAMS
-        }
-
-        # Get this image's prompts and noise parameters
-        addl_metadata = self._get_additional_metadata(graph, ancestor_id)
-
-        # If additional metadata was found, add it to the main metadata
-        if addl_metadata is not None:
-            ancestor_metadata.update(addl_metadata)
-
-        return ImageMetadata(**ancestor_metadata)

invokeai/app/services/model_manager_service.py

@@ -18,8 +18,9 @@ from invokeai.backend.model_management import (
     SchedulerPredictionType,
     ModelMerger,
     MergeInterpolationMethod,
+    ModelNotFoundException,
 )
-
+from invokeai.backend.model_management.model_search import FindModels
 
 import torch
 from invokeai.app.models.exceptions import CanceledException
@@ -145,7 +146,7 @@ class ModelManagerServiceBase(ABC):
     ) -> AddModelResult:
         """
         Update the named model with a dictionary of attributes. Will fail with a
-        KeyErrorException if the name does not already exist.
+        ModelNotFoundException if the name does not already exist.
 
         On a successful update, the config will be changed in memory. Will fail 
         with an assertion error if provided attributes are incorrect or 
@@ -167,6 +168,27 @@ class ModelManagerServiceBase(ABC):
         """
         pass
 
+    @abstractmethod
+    def rename_model(self,
+                     model_name: str,
+                     base_model: BaseModelType,
+                     model_type: ModelType,
+                     new_name: str,
+                     ):
+        """
+        Rename the indicated model.
+        """
+        pass
+
+    @abstractmethod
+    def list_checkpoint_configs(
+        self
+    )->List[Path]:
+        """
+        List the checkpoint config paths from ROOT/configs/stable-diffusion.
+        """
+        pass
+
     @abstractmethod
     def convert_model(
         self,
@@ -220,6 +242,7 @@ class ModelManagerServiceBase(ABC):
             alpha: Optional[float] = 0.5,
             interp: Optional[MergeInterpolationMethod] = None,
             force: Optional[bool] = False,
+            merge_dest_directory: Optional[Path] = None
     ) -> AddModelResult:
         """
         Merge two to three diffusrs pipeline models and save as a new model.
@@ -228,9 +251,26 @@ class ModelManagerServiceBase(ABC):
         :param merged_model_name: Name of destination merged model
         :param alpha: Alpha strength to apply to 2d and 3d model
         :param interp: Interpolation method. None (default) 
+        :param merge_dest_directory: Save the merged model to the designated directory (with 'merged_model_name' appended)
         """
         pass
-    
+
+    @abstractmethod
+    def search_for_models(self, directory: Path)->List[Path]:
+        """
+        Return list of all models found in the designated directory.
+        """
+        pass
+        
+    @abstractmethod
+    def sync_to_config(self):
+        """
+        Re-read models.yaml, rescan the models directory, and reimport models 
+        in the autoimport directories. Call after making changes outside the
+        model manager API.
+        """
+        pass
+        
     @abstractmethod
     def commit(self, conf_file: Optional[Path] = None) -> None:
         """
@@ -259,10 +299,11 @@ class ModelManagerService(ModelManagerServiceBase):
         else:
             config_file = config.root_dir / "configs/models.yaml"
             
-        logger.debug(f'config file={config_file}')
+        logger.debug(f'Config file={config_file}')
 
         device = torch.device(choose_torch_device())
-        logger.debug(f'GPU device = {device}')
+        device_name =  torch.cuda.get_device_name() if device==torch.device('cuda') else ''
+        logger.info(f'GPU device = {device} {device_name}')
 
         precision = config.precision
         if precision == "auto":
@@ -299,7 +340,6 @@ class ModelManagerService(ModelManagerServiceBase):
         base_model: BaseModelType,
         model_type: ModelType,
         submodel: Optional[SubModelType] = None,
-        node: Optional[BaseInvocation] = None,
         context: Optional[InvocationContext] = None,
     ) -> ModelInfo:
         """
@@ -307,11 +347,9 @@ class ModelManagerService(ModelManagerServiceBase):
         part (such as the vae) of a diffusers mode.
         """
 
-        # if we are called from within a node, then we get to emit
-        # load start and complete events
-        if node and context:
+        # we can emit model loading events if we are executing with access to the invocation context
+        if context:
             self._emit_load_event(
-                node=node,
                 context=context,
                 model_name=model_name,
                 base_model=base_model,
@@ -326,9 +364,8 @@ class ModelManagerService(ModelManagerServiceBase):
             submodel,
         )
 
-        if node and context:
+        if context:
             self._emit_load_event(
-                node=node,
                 context=context,
                 model_name=model_name,
                 base_model=base_model,
@@ -412,14 +449,14 @@ class ModelManagerService(ModelManagerServiceBase):
     ) -> AddModelResult:
         """
         Update the named model with a dictionary of attributes. Will fail with a
-        KeyError exception if the name does not already exist.
+        ModelNotFoundException exception if the name does not already exist.
         On a successful update, the config will be changed in memory. Will fail 
         with an assertion error if provided attributes are incorrect or 
         the model name is missing. Call commit() to write changes to disk.
         """
         self.logger.debug(f'update model {model_name}')
         if not self.model_exists(model_name, base_model, model_type):
-            raise KeyError(f"Unknown model {model_name}")
+            raise ModelNotFoundException(f"Unknown model {model_name}")
         return self.add_model(model_name, base_model, model_type, model_attributes, clobber=True)
     
     def del_model(
@@ -431,16 +468,18 @@ class ModelManagerService(ModelManagerServiceBase):
         """
         Delete the named model from configuration. If delete_files is true,
         then the underlying weight file or diffusers directory will be deleted
-        as well. Call commit() to write to disk.
+        as well.
         """
         self.logger.debug(f'delete model {model_name}')
         self.mgr.del_model(model_name, base_model, model_type)
+        self.mgr.commit()
 
     def convert_model(
         self,
         model_name: str,
         base_model: BaseModelType,
         model_type: Union[ModelType.Main,ModelType.Vae],
+        convert_dest_directory: Optional[Path] = Field(default=None, description="Optional directory location for merged model"),        
     ) -> AddModelResult:
         """
         Convert a checkpoint file into a diffusers folder, deleting the cached
@@ -449,13 +488,14 @@ class ModelManagerService(ModelManagerServiceBase):
         :param model_name: Name of the model to convert
         :param base_model: Base model type
         :param model_type: Type of model ['vae' or 'main']
+        :param convert_dest_directory: Save the converted model to the designated directory (`models/etc/etc` by default)
 
         This will raise a ValueError unless the model is not a checkpoint. It will
         also raise a ValueError in the event that there is a similarly-named diffusers
         directory already in place.
         """
         self.logger.debug(f'convert model {model_name}')        
-        return self.mgr.convert_model(model_name, base_model, model_type)
+        return self.mgr.convert_model(model_name, base_model, model_type, convert_dest_directory)
 
     def commit(self, conf_file: Optional[Path]=None):
         """
@@ -467,23 +507,19 @@ class ModelManagerService(ModelManagerServiceBase):
 
     def _emit_load_event(
         self,
-        node,
         context,
         model_name: str,
         base_model: BaseModelType,
         model_type: ModelType,
-        submodel: SubModelType,
+        submodel: Optional[SubModelType] = None,
         model_info: Optional[ModelInfo] = None,
     ):
         if context.services.queue.is_canceled(context.graph_execution_state_id):
             raise CanceledException()
-        graph_execution_state = context.services.graph_execution_manager.get(context.graph_execution_state_id)
-        source_node_id = graph_execution_state.prepared_source_mapping[node.id]
+
         if model_info:
             context.services.events.emit_model_load_completed(
                 graph_execution_state_id=context.graph_execution_state_id,
-                node=node.dict(),
-                source_node_id=source_node_id,
                 model_name=model_name,
                 base_model=base_model,
                 model_type=model_type,
@@ -493,8 +529,6 @@ class ModelManagerService(ModelManagerServiceBase):
         else:
             context.services.events.emit_model_load_started(
                 graph_execution_state_id=context.graph_execution_state_id,
-                node=node.dict(),
-                source_node_id=source_node_id,
                 model_name=model_name,
                 base_model=base_model,
                 model_type=model_type,
@@ -536,6 +570,7 @@ class ModelManagerService(ModelManagerServiceBase):
             alpha: Optional[float] = 0.5,
             interp: Optional[MergeInterpolationMethod] = None,
             force: Optional[bool] = False,
+            merge_dest_directory: Optional[Path] = Field(default=None, description="Optional directory location for merged model"),
     ) -> AddModelResult:
         """
         Merge two to three diffusrs pipeline models and save as a new model.
@@ -544,6 +579,7 @@ class ModelManagerService(ModelManagerServiceBase):
         :param merged_model_name: Name of destination merged model
         :param alpha: Alpha strength to apply to 2d and 3d model
         :param interp: Interpolation method. None (default) 
+        :param merge_dest_directory: Save the merged model to the designated directory (with 'merged_model_name' appended)
         """
         merger = ModelMerger(self.mgr)
         try:
@@ -554,7 +590,55 @@ class ModelManagerService(ModelManagerServiceBase):
                 alpha = alpha,
                 interp = interp,
                 force = force,
+                merge_dest_directory=merge_dest_directory,
             )
         except AssertionError as e:
             raise ValueError(e)
         return result
+
+    def search_for_models(self, directory: Path)->List[Path]:
+        """
+        Return list of all models found in the designated directory.
+        """
+        search = FindModels(directory,self.logger)
+        return search.list_models()
+
+    def sync_to_config(self):
+        """
+        Re-read models.yaml, rescan the models directory, and reimport models 
+        in the autoimport directories. Call after making changes outside the
+        model manager API.
+        """
+        return self.mgr.sync_to_config()
+
+    def list_checkpoint_configs(self)->List[Path]:
+        """
+        List the checkpoint config paths from ROOT/configs/stable-diffusion.
+        """
+        config = self.mgr.app_config
+        conf_path = config.legacy_conf_path
+        root_path = config.root_path
+        return [(conf_path / x).relative_to(root_path) for x in conf_path.glob('**/*.yaml')]
+
+    def rename_model(self,
+                     model_name: str,
+                     base_model: BaseModelType,
+                     model_type: ModelType,
+                     new_name: str = None,
+                     new_base: BaseModelType = None,
+                     ):
+        """
+        Rename the indicated model. Can provide a new name and/or a new base.
+        :param model_name: Current name of the model
+        :param base_model: Current base of the model
+        :param model_type: Model type (can't be changed)
+        :param new_name: New name for the model
+        :param new_base: New base for the model
+        """
+        self.mgr.rename_model(base_model = base_model,
+                              model_type = model_type,
+                              model_name = model_name,
+                              new_name = new_name,
+                              new_base = new_base,
+                              )
+        

invokeai/app/services/models/image_record.py

@@ -1,13 +1,14 @@
 import datetime
 from typing import Optional, Union
+
 from pydantic import BaseModel, Extra, Field, StrictBool, StrictStr
+
 from invokeai.app.models.image import ImageCategory, ResourceOrigin
-from invokeai.app.models.metadata import ImageMetadata
 from invokeai.app.util.misc import get_iso_timestamp
 
 
 class ImageRecord(BaseModel):
-    """Deserialized image record."""
+    """Deserialized image record without metadata."""
 
     image_name: str = Field(description="The unique name of the image.")
     """The unique name of the image."""
@@ -43,11 +44,6 @@ class ImageRecord(BaseModel):
         description="The node ID that generated this image, if it is a generated image.",
     )
     """The node ID that generated this image, if it is a generated image."""
-    metadata: Optional[ImageMetadata] = Field(
-        default=None,
-        description="A limited subset of the image's generation metadata. Retrieve the image's session for full metadata.",
-    )
-    """A limited subset of the image's generation metadata. Retrieve the image's session for full metadata."""
 
 
 class ImageRecordChanges(BaseModel, extra=Extra.forbid):
@@ -112,6 +108,7 @@ def deserialize_image_record(image_dict: dict) -> ImageRecord:
 
     # Retrieve all the values, setting "reasonable" defaults if they are not present.
 
+    # TODO: do we really need to handle default values here? ideally the data is the correct shape...
     image_name = image_dict.get("image_name", "unknown")
     image_origin = ResourceOrigin(
         image_dict.get("image_origin", ResourceOrigin.INTERNAL.value)
@@ -128,13 +125,6 @@ def deserialize_image_record(image_dict: dict) -> ImageRecord:
     deleted_at = image_dict.get("deleted_at", get_iso_timestamp())
     is_intermediate = image_dict.get("is_intermediate", False)
 
-    raw_metadata = image_dict.get("metadata")
-
-    if raw_metadata is not None:
-        metadata = ImageMetadata.parse_raw(raw_metadata)
-    else:
-        metadata = None
-
     return ImageRecord(
         image_name=image_name,
         image_origin=image_origin,
@@ -143,7 +133,6 @@ def deserialize_image_record(image_dict: dict) -> ImageRecord:
         height=height,
         session_id=session_id,
         node_id=node_id,
-        metadata=metadata,
         created_at=created_at,
         updated_at=updated_at,
         deleted_at=deleted_at,

invokeai/app/services/restoration_services.py

@@ -1,113 +0,0 @@
-import sys
-import traceback
-import torch
-from typing import types
-from ...backend.restoration import Restoration
-from ...backend.util import choose_torch_device, CPU_DEVICE, MPS_DEVICE
-
-# This should be a real base class for postprocessing functions,
-# but right now we just instantiate the existing gfpgan, esrgan
-# and codeformer functions.
-class RestorationServices:
-    '''Face restoration and upscaling'''
-    
-    def __init__(self,args,logger:types.ModuleType):
-        try:
-            gfpgan, codeformer, esrgan = None, None, None
-            if args.restore or args.esrgan:
-                restoration = Restoration()
-                # TODO: redo for new model structure
-                if False and args.restore:
-                    gfpgan, codeformer = restoration.load_face_restore_models(
-                        args.gfpgan_model_path
-                    )
-                else:
-                    logger.info("Face restoration disabled")
-                    if False and args.esrgan:
-                        esrgan = restoration.load_esrgan(args.esrgan_bg_tile)
-                    else:
-                        logger.info("Upscaling disabled")
-            else:
-                logger.info("Face restoration and upscaling disabled")
-        except (ModuleNotFoundError, ImportError):
-            print(traceback.format_exc(), file=sys.stderr)
-            logger.info("You may need to install the ESRGAN and/or GFPGAN modules")
-        self.device = torch.device(choose_torch_device())
-        self.gfpgan = gfpgan
-        self.codeformer = codeformer
-        self.esrgan = esrgan
-        self.logger = logger
-        self.logger.info('Face restoration initialized')
-
-    # note that this one method does gfpgan and codepath reconstruction, as well as
-    # esrgan upscaling
-    # TO DO: refactor into separate methods
-    def upscale_and_reconstruct(
-        self,
-        image_list,
-        facetool="gfpgan",
-        upscale=None,
-        upscale_denoise_str=0.75,
-        strength=0.0,
-        codeformer_fidelity=0.75,
-        save_original=False,
-        image_callback=None,
-        prefix=None,
-    ):
-        results = []
-        for r in image_list:
-            image, seed = r
-            try:
-                if strength > 0:
-                    if self.gfpgan is not None or self.codeformer is not None:
-                        if facetool == "gfpgan":
-                            if self.gfpgan is None:
-                                self.logger.info(
-                                    "GFPGAN not found. Face restoration is disabled."
-                                )
-                            else:
-                                image = self.gfpgan.process(image, strength, seed)
-                        if facetool == "codeformer":
-                            if self.codeformer is None:
-                                self.logger.info(
-                                    "CodeFormer not found. Face restoration is disabled."
-                                )
-                            else:
-                                cf_device = (
-                                    CPU_DEVICE if self.device == MPS_DEVICE else self.device
-                                )
-                                image = self.codeformer.process(
-                                    image=image,
-                                    strength=strength,
-                                    device=cf_device,
-                                    seed=seed,
-                                    fidelity=codeformer_fidelity,
-                                )
-                    else:
-                        self.logger.info("Face Restoration is disabled.")
-                if upscale is not None:
-                    if self.esrgan is not None:
-                        if len(upscale) < 2:
-                            upscale.append(0.75)
-                        image = self.esrgan.process(
-                            image,
-                            upscale[1],
-                            seed,
-                            int(upscale[0]),
-                            denoise_str=upscale_denoise_str,
-                        )
-                    else:
-                        self.logger.info("ESRGAN is disabled. Image not upscaled.")
-            except Exception as e:
-                self.logger.info(
-                    f"Error running RealESRGAN or GFPGAN. Your image was not upscaled.\n{e}"
-                )
-
-            if image_callback is not None:
-                image_callback(image, seed, upscaled=True, use_prefix=prefix)
-            else:
-                r[0] = image
-
-            results.append([image, seed])
-
-        return results

invokeai/app/services/sqlite.py

@@ -1,6 +1,6 @@
 import sqlite3
 from threading import Lock
-from typing import Generic, TypeVar, Optional, Union, get_args
+from typing import Generic, Optional, TypeVar, get_args
 
 from pydantic import BaseModel, parse_raw_as
 
@@ -78,6 +78,21 @@ class SqliteItemStorage(ItemStorageABC, Generic[T]):
 
         return self._parse_item(result[0])
 
+    def get_raw(self, id: str) -> Optional[str]:
+        try:
+            self._lock.acquire()
+            self._cursor.execute(
+                f"""SELECT item FROM {self._table_name} WHERE id = ?;""", (str(id),)
+            )
+            result = self._cursor.fetchone()
+        finally:
+            self._lock.release()
+
+        if not result:
+            return None
+
+        return result[0]
+
     def delete(self, id: str):
         try:
             self._lock.acquire()

invokeai/app/services/urls.py

@@ -22,4 +22,4 @@ class LocalUrlService(UrlServiceBase):
         if thumbnail:
             return f"{self._base_url}/images/{image_basename}/thumbnail"
 
-        return f"{self._base_url}/images/{image_basename}"
+        return f"{self._base_url}/images/{image_basename}/full"

invokeai/app/util/controlnet_utils.py

@@ -0,0 +1,342 @@
+import torch
+import numpy as np
+import cv2
+from PIL import Image
+from diffusers.utils import PIL_INTERPOLATION
+
+from einops import rearrange
+from controlnet_aux.util import HWC3, resize_image
+
+###################################################################
+# Copy of scripts/lvminthin.py from Mikubill/sd-webui-controlnet
+###################################################################
+# High Quality Edge Thinning using Pure Python
+# Written by Lvmin Zhangu
+# 2023 April
+# Stanford University
+# If you use this, please Cite "High Quality Edge Thinning using Pure Python", Lvmin Zhang, In Mikubill/sd-webui-controlnet.
+
+lvmin_kernels_raw = [
+    np.array([
+        [-1, -1, -1],
+        [0, 1, 0],
+        [1, 1, 1]
+    ], dtype=np.int32),
+    np.array([
+        [0, -1, -1],
+        [1, 1, -1],
+        [0, 1, 0]
+    ], dtype=np.int32)
+]
+
+lvmin_kernels = []
+lvmin_kernels += [np.rot90(x, k=0, axes=(0, 1)) for x in lvmin_kernels_raw]
+lvmin_kernels += [np.rot90(x, k=1, axes=(0, 1)) for x in lvmin_kernels_raw]
+lvmin_kernels += [np.rot90(x, k=2, axes=(0, 1)) for x in lvmin_kernels_raw]
+lvmin_kernels += [np.rot90(x, k=3, axes=(0, 1)) for x in lvmin_kernels_raw]
+
+lvmin_prunings_raw = [
+    np.array([
+        [-1, -1, -1],
+        [-1, 1, -1],
+        [0, 0, -1]
+    ], dtype=np.int32),
+    np.array([
+        [-1, -1, -1],
+        [-1, 1, -1],
+        [-1, 0, 0]
+    ], dtype=np.int32)
+]
+
+lvmin_prunings = []
+lvmin_prunings += [np.rot90(x, k=0, axes=(0, 1)) for x in lvmin_prunings_raw]
+lvmin_prunings += [np.rot90(x, k=1, axes=(0, 1)) for x in lvmin_prunings_raw]
+lvmin_prunings += [np.rot90(x, k=2, axes=(0, 1)) for x in lvmin_prunings_raw]
+lvmin_prunings += [np.rot90(x, k=3, axes=(0, 1)) for x in lvmin_prunings_raw]
+
+
+def remove_pattern(x, kernel):
+    objects = cv2.morphologyEx(x, cv2.MORPH_HITMISS, kernel)
+    objects = np.where(objects > 127)
+    x[objects] = 0
+    return x, objects[0].shape[0] > 0
+
+
+def thin_one_time(x, kernels):
+    y = x
+    is_done = True
+    for k in kernels:
+        y, has_update = remove_pattern(y, k)
+        if has_update:
+            is_done = False
+    return y, is_done
+
+
+def lvmin_thin(x, prunings=True):
+    y = x
+    for i in range(32):
+        y, is_done = thin_one_time(y, lvmin_kernels)
+        if is_done:
+            break
+    if prunings:
+        y, _ = thin_one_time(y, lvmin_prunings)
+    return y
+
+
+def nake_nms(x):
+    f1 = np.array([[0, 0, 0], [1, 1, 1], [0, 0, 0]], dtype=np.uint8)
+    f2 = np.array([[0, 1, 0], [0, 1, 0], [0, 1, 0]], dtype=np.uint8)
+    f3 = np.array([[1, 0, 0], [0, 1, 0], [0, 0, 1]], dtype=np.uint8)
+    f4 = np.array([[0, 0, 1], [0, 1, 0], [1, 0, 0]], dtype=np.uint8)
+    y = np.zeros_like(x)
+    for f in [f1, f2, f3, f4]:
+        np.putmask(y, cv2.dilate(x, kernel=f) == x, x)
+    return y
+
+
+################################################################################
+# copied from Mikubill/sd-webui-controlnet external_code.py and modified for InvokeAI
+################################################################################
+# FIXME: not using yet, if used in the future will most likely require modification of preprocessors
+def pixel_perfect_resolution(
+        image: np.ndarray,
+        target_H: int,
+        target_W: int,
+        resize_mode: str,
+) -> int:
+    """
+    Calculate the estimated resolution for resizing an image while preserving aspect ratio.
+
+    The function first calculates scaling factors for height and width of the image based on the target
+    height and width. Then, based on the chosen resize mode, it either takes the smaller or the larger
+    scaling factor to estimate the new resolution.
+
+    If the resize mode is OUTER_FIT, the function uses the smaller scaling factor, ensuring the whole image
+    fits within the target dimensions, potentially leaving some empty space.
+
+    If the resize mode is not OUTER_FIT, the function uses the larger scaling factor, ensuring the target
+    dimensions are fully filled, potentially cropping the image.
+
+    After calculating the estimated resolution, the function prints some debugging information.
+
+    Args:
+        image (np.ndarray): A 3D numpy array representing an image. The dimensions represent [height, width, channels].
+        target_H (int): The target height for the image.
+        target_W (int): The target width for the image.
+        resize_mode (ResizeMode): The mode for resizing.
+
+    Returns:
+        int: The estimated resolution after resizing.
+    """
+    raw_H, raw_W, _ = image.shape
+
+    k0 = float(target_H) / float(raw_H)
+    k1 = float(target_W) / float(raw_W)
+
+    if resize_mode == "fill_resize":
+        estimation = min(k0, k1) * float(min(raw_H, raw_W))
+    else: # "crop_resize" or "just_resize" (or possibly "just_resize_simple"?)
+        estimation = max(k0, k1) * float(min(raw_H, raw_W))
+
+    # print(f"Pixel Perfect Computation:")
+    # print(f"resize_mode = {resize_mode}")
+    # print(f"raw_H = {raw_H}")
+    # print(f"raw_W = {raw_W}")
+    # print(f"target_H = {target_H}")
+    # print(f"target_W = {target_W}")
+    # print(f"estimation = {estimation}")
+
+    return int(np.round(estimation))
+
+
+###########################################################################
+# Copied from detectmap_proc method in scripts/detectmap_proc.py in Mikubill/sd-webui-controlnet
+#    modified for InvokeAI
+###########################################################################
+# def detectmap_proc(detected_map, module, resize_mode, h, w):
+def np_img_resize(
+    np_img: np.ndarray,
+    resize_mode: str,
+    h: int,
+    w: int,
+    device: torch.device = torch.device('cpu')
+):
+    # if 'inpaint' in module:
+    #     np_img = np_img.astype(np.float32)
+    # else:
+    #     np_img = HWC3(np_img)
+    np_img = HWC3(np_img)
+
+    def safe_numpy(x):
+        # A very safe method to make sure that Apple/Mac works
+        y = x
+
+        # below is very boring but do not change these. If you change these Apple or Mac may fail.
+        y = y.copy()
+        y = np.ascontiguousarray(y)
+        y = y.copy()
+        return y
+
+    def get_pytorch_control(x):
+        # A very safe method to make sure that Apple/Mac works
+        y = x
+
+        # below is very boring but do not change these. If you change these Apple or Mac may fail.
+        y = torch.from_numpy(y)
+        y = y.float() / 255.0
+        y = rearrange(y, 'h w c -> 1 c h w')
+        y = y.clone()
+        # y = y.to(devices.get_device_for("controlnet"))
+        y = y.to(device)
+        y = y.clone()
+        return y
+
+    def high_quality_resize(x: np.ndarray,
+                            size):
+        # Written by lvmin
+        # Super high-quality control map up-scaling, considering binary, seg, and one-pixel edges
+        inpaint_mask = None
+        if x.ndim == 3 and x.shape[2] == 4:
+            inpaint_mask = x[:, :, 3]
+            x = x[:, :, 0:3]
+
+        new_size_is_smaller = (size[0] * size[1]) < (x.shape[0] * x.shape[1])
+        new_size_is_bigger = (size[0] * size[1]) > (x.shape[0] * x.shape[1])
+        unique_color_count = np.unique(x.reshape(-1, x.shape[2]), axis=0).shape[0]
+        is_one_pixel_edge = False
+        is_binary = False
+        if unique_color_count == 2:
+            is_binary = np.min(x) < 16 and np.max(x) > 240
+            if is_binary:
+                xc = x
+                xc = cv2.erode(xc, np.ones(shape=(3, 3), dtype=np.uint8), iterations=1)
+                xc = cv2.dilate(xc, np.ones(shape=(3, 3), dtype=np.uint8), iterations=1)
+                one_pixel_edge_count = np.where(xc < x)[0].shape[0]
+                all_edge_count = np.where(x > 127)[0].shape[0]
+                is_one_pixel_edge = one_pixel_edge_count * 2 > all_edge_count
+
+        if 2 < unique_color_count < 200:
+            interpolation = cv2.INTER_NEAREST
+        elif new_size_is_smaller:
+            interpolation = cv2.INTER_AREA
+        else:
+            interpolation = cv2.INTER_CUBIC  # Must be CUBIC because we now use nms. NEVER CHANGE THIS
+
+        y = cv2.resize(x, size, interpolation=interpolation)
+        if inpaint_mask is not None:
+            inpaint_mask = cv2.resize(inpaint_mask, size, interpolation=interpolation)
+
+        if is_binary:
+            y = np.mean(y.astype(np.float32), axis=2).clip(0, 255).astype(np.uint8)
+            if is_one_pixel_edge:
+                y = nake_nms(y)
+                _, y = cv2.threshold(y, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
+                y = lvmin_thin(y, prunings=new_size_is_bigger)
+            else:
+                _, y = cv2.threshold(y, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
+            y = np.stack([y] * 3, axis=2)
+
+        if inpaint_mask is not None:
+            inpaint_mask = (inpaint_mask > 127).astype(np.float32) * 255.0
+            inpaint_mask = inpaint_mask[:, :, None].clip(0, 255).astype(np.uint8)
+            y = np.concatenate([y, inpaint_mask], axis=2)
+
+        return y
+
+    # if resize_mode == external_code.ResizeMode.RESIZE:
+    if resize_mode == "just_resize": # RESIZE
+        np_img = high_quality_resize(np_img, (w, h))
+        np_img = safe_numpy(np_img)
+        return get_pytorch_control(np_img), np_img
+
+    old_h, old_w, _ = np_img.shape
+    old_w = float(old_w)
+    old_h = float(old_h)
+    k0 = float(h) / old_h
+    k1 = float(w) / old_w
+
+    safeint = lambda x: int(np.round(x))
+
+    # if resize_mode == external_code.ResizeMode.OUTER_FIT:
+    if resize_mode == "fill_resize":  # OUTER_FIT
+        k = min(k0, k1)
+        borders = np.concatenate([np_img[0, :, :], np_img[-1, :, :], np_img[:, 0, :], np_img[:, -1, :]], axis=0)
+        high_quality_border_color = np.median(borders, axis=0).astype(np_img.dtype)
+        if len(high_quality_border_color) == 4:
+            # Inpaint hijack
+            high_quality_border_color[3] = 255
+        high_quality_background = np.tile(high_quality_border_color[None, None], [h, w, 1])
+        np_img = high_quality_resize(np_img, (safeint(old_w * k), safeint(old_h * k)))
+        new_h, new_w, _ = np_img.shape
+        pad_h = max(0, (h - new_h) // 2)
+        pad_w = max(0, (w - new_w) // 2)
+        high_quality_background[pad_h:pad_h + new_h, pad_w:pad_w + new_w] = np_img
+        np_img = high_quality_background
+        np_img = safe_numpy(np_img)
+        return get_pytorch_control(np_img), np_img
+    else: # resize_mode == "crop_resize"  (INNER_FIT)
+        k = max(k0, k1)
+        np_img = high_quality_resize(np_img, (safeint(old_w * k), safeint(old_h * k)))
+        new_h, new_w, _ = np_img.shape
+        pad_h = max(0, (new_h - h) // 2)
+        pad_w = max(0, (new_w - w) // 2)
+        np_img = np_img[pad_h:pad_h + h, pad_w:pad_w + w]
+        np_img = safe_numpy(np_img)
+        return get_pytorch_control(np_img), np_img
+
+def prepare_control_image(
+    # image used to be Union[PIL.Image.Image, List[PIL.Image.Image], torch.Tensor, List[torch.Tensor]]
+    # but now should be able to assume that image is a single PIL.Image, which simplifies things
+    image: Image,
+    # FIXME: need to fix hardwiring of width and height, change to basing on latents dimensions?
+    # latents_to_match_resolution, # TorchTensor of shape (batch_size, 3, height, width)
+    width=512,  # should be 8 * latent.shape[3]
+    height=512,  # should be 8 * latent height[2]
+    # batch_size=1, # currently no batching
+    # num_images_per_prompt=1, # currently only single image
+    device="cuda",
+    dtype=torch.float16,
+    do_classifier_free_guidance=True,
+    control_mode="balanced",
+    resize_mode="just_resize_simple",
+):
+    # FIXME: implement "crop_resize_simple" and "fill_resize_simple", or pull them out
+    if (resize_mode == "just_resize_simple" or
+        resize_mode == "crop_resize_simple" or
+        resize_mode == "fill_resize_simple"):
+        image = image.convert("RGB")
+        if (resize_mode == "just_resize_simple"):
+            image = image.resize((width, height), resample=PIL_INTERPOLATION["lanczos"])
+        elif (resize_mode == "crop_resize_simple"): # not yet implemented
+            pass
+        elif (resize_mode == "fill_resize_simple"): # not yet implemented
+            pass
+        nimage = np.array(image)
+        nimage = nimage[None, :]
+        nimage = np.concatenate([nimage], axis=0)
+        # normalizing RGB values to [0,1] range (in PIL.Image they are [0-255])
+        nimage = np.array(nimage).astype(np.float32) / 255.0
+        nimage = nimage.transpose(0, 3, 1, 2)
+        timage = torch.from_numpy(nimage)
+
+    # use fancy lvmin controlnet resizing
+    elif (resize_mode == "just_resize" or resize_mode == "crop_resize" or resize_mode == "fill_resize"):
+        nimage = np.array(image)
+        timage, nimage = np_img_resize(
+            np_img=nimage,
+            resize_mode=resize_mode,
+            h=height,
+            w=width,
+            # device=torch.device('cpu')
+            device=device,
+        )
+    else:
+        pass
+        print("ERROR: invalid resize_mode ==> ", resize_mode)
+        exit(1)
+
+    timage = timage.to(device=device, dtype=dtype)
+    cfg_injection = (control_mode == "more_control" or control_mode == "unbalanced")
+    if do_classifier_free_guidance and not cfg_injection:
+        timage = torch.cat([timage] * 2)
+    return timage

invokeai/app/util/metadata.py

@@ -0,0 +1,55 @@
+import json
+from typing import Optional
+
+from pydantic import ValidationError
+
+from invokeai.app.services.graph import Edge
+
+
+def get_metadata_graph_from_raw_session(session_raw: str) -> Optional[dict]:
+    """
+    Parses raw session string, returning a dict of the graph.
+
+    Only the general graph shape is validated; none of the fields are validated.
+
+    Any `metadata_accumulator` nodes and edges are removed.
+
+    Any validation failure will return None.
+    """
+
+    graph = json.loads(session_raw).get("graph", None)
+
+    # sanity check make sure the graph is at least reasonably shaped
+    if (
+        type(graph) is not dict
+        or "nodes" not in graph
+        or type(graph["nodes"]) is not dict
+        or "edges" not in graph
+        or type(graph["edges"]) is not list
+    ):
+        # something has gone terribly awry, return an empty dict
+        return None
+
+    try:
+        # delete the `metadata_accumulator` node
+        del graph["nodes"]["metadata_accumulator"]
+    except KeyError:
+        # no accumulator node, all good
+        pass
+
+    # delete any edges to or from it
+    for i, edge in enumerate(graph["edges"]):
+        try:
+            # try to parse the edge
+            Edge(**edge)
+        except ValidationError:
+            # something has gone terribly awry, return an empty dict
+            return None
+
+        if (
+            edge["source"]["node_id"] == "metadata_accumulator"
+            or edge["destination"]["node_id"] == "metadata_accumulator"
+        ):
+            del graph["edges"][i]
+
+    return graph

invokeai/app/util/step_callback.py

@@ -1,9 +1,30 @@
+import torch
+from PIL import Image
 from invokeai.app.models.exceptions import CanceledException
 from invokeai.app.models.image import ProgressImage
 from ..invocations.baseinvocation import InvocationContext
 from ...backend.util.util import image_to_dataURL
 from ...backend.generator.base import Generator
 from ...backend.stable_diffusion import PipelineIntermediateState
+from invokeai.app.services.config import InvokeAIAppConfig
+
+
+def sample_to_lowres_estimated_image(samples, latent_rgb_factors, smooth_matrix = None):
+    latent_image = samples[0].permute(1, 2, 0) @ latent_rgb_factors
+
+    if smooth_matrix is not None:
+        latent_image = latent_image.unsqueeze(0).permute(3, 0, 1, 2)
+        latent_image = torch.nn.functional.conv2d(latent_image, smooth_matrix.reshape((1,1,3,3)), padding=1)
+        latent_image = latent_image.permute(1, 2, 3, 0).squeeze(0)
+
+    latents_ubyte = (
+        ((latent_image + 1) / 2)
+        .clamp(0, 1)  # change scale from -1..1 to 0..1
+        .mul(0xFF)  # to 0..255
+        .byte()
+    ).cpu()
+
+    return Image.fromarray(latents_ubyte.numpy())
 
 
 def stable_diffusion_step_callback(
@@ -37,7 +58,24 @@ def stable_diffusion_step_callback(
     #     step = intermediate_state.step
 
     # TODO: only output a preview image when requested
-    image = Generator.sample_to_lowres_estimated_image(sample)
+
+    # origingally adapted from code by @erucipe and @keturn here:
+    # https://discuss.huggingface.co/t/decoding-latents-to-rgb-without-upscaling/23204/7
+
+    # these updated numbers for v1.5 are from @torridgristle
+    v1_5_latent_rgb_factors = torch.tensor(
+        [
+            #    R        G        B
+            [0.3444, 0.1385, 0.0670],  # L1
+            [0.1247, 0.4027, 0.1494],  # L2
+            [-0.3192, 0.2513, 0.2103],  # L3
+            [-0.1307, -0.1874, -0.7445],  # L4
+        ],
+        dtype=sample.dtype,
+        device=sample.device,
+    )
+
+    image = sample_to_lowres_estimated_image(sample, v1_5_latent_rgb_factors)
 
     (width, height) = image.size
     width *= 8
@@ -53,3 +91,56 @@ def stable_diffusion_step_callback(
         step=intermediate_state.step,
         total_steps=node["steps"],
     )
+
+def stable_diffusion_xl_step_callback(
+    context: InvocationContext,
+    node: dict,
+    source_node_id: str,
+    sample,
+    step,
+    total_steps,
+):
+    if context.services.queue.is_canceled(context.graph_execution_state_id):
+        raise CanceledException
+
+    sdxl_latent_rgb_factors = torch.tensor(
+        [
+            #   R        G        B
+            [ 0.3816,  0.4930,  0.5320],
+            [-0.3753,  0.1631,  0.1739],
+            [ 0.1770,  0.3588, -0.2048],
+            [-0.4350, -0.2644, -0.4289],
+        ],
+        dtype=sample.dtype,
+        device=sample.device,
+    )
+
+    sdxl_smooth_matrix = torch.tensor(
+        [
+            #[ 0.0478,  0.1285,  0.0478],
+            #[ 0.1285,  0.2948,  0.1285],
+            #[ 0.0478,  0.1285,  0.0478],
+            [0.0358, 0.0964, 0.0358],
+            [0.0964, 0.4711, 0.0964],
+            [0.0358, 0.0964, 0.0358],
+        ],
+        dtype=sample.dtype,
+        device=sample.device,
+    )
+
+    image = sample_to_lowres_estimated_image(sample, sdxl_latent_rgb_factors, sdxl_smooth_matrix)
+
+    (width, height) = image.size
+    width *= 8
+    height *= 8
+
+    dataURL = image_to_dataURL(image, image_format="JPEG")
+
+    context.services.events.emit_generator_progress(
+        graph_execution_state_id=context.graph_execution_state_id,
+        node=node,
+        source_node_id=source_node_id,
+        progress_image=ProgressImage(width=width, height=height, dataURL=dataURL),
+        step=step,
+        total_steps=total_steps,
+    )

invokeai/backend/install/invokeai_configure.py

@@ -23,6 +23,7 @@ from urllib import request
 
 import npyscreen
 import transformers
+import omegaconf
 from diffusers import AutoencoderKL
 from diffusers.pipelines.stable_diffusion.safety_checker import StableDiffusionSafetyChecker
 from huggingface_hub import HfFolder
@@ -30,8 +31,6 @@ from huggingface_hub import login as hf_hub_login
 from omegaconf import OmegaConf
 from tqdm import tqdm
 from transformers import (
-    AutoProcessor,
-    CLIPSegForImageSegmentation,
     CLIPTextModel,
     CLIPTokenizer,
     AutoFeatureExtractor,
@@ -45,8 +44,8 @@ from invokeai.app.services.config import (
 from invokeai.backend.util.logging import InvokeAILogger
 from invokeai.frontend.install.model_install import addModelsForm, process_and_execute
 from invokeai.frontend.install.widgets import (
-    SingleSelectColumns,
     CenteredButtonPress,
+    FileBox,
     IntTitleSlider,
     set_min_terminal_size,
     CyclingForm,
@@ -72,7 +71,6 @@ transformers.logging.set_verbosity_error()
 config = InvokeAIAppConfig.get_config()
 
 Model_dir = "models"
-Weights_dir = "ldm/stable-diffusion-v1/"
 
 Default_config_file = config.model_conf_path
 SD_Configs = config.legacy_conf_path
@@ -226,64 +224,35 @@ def download_conversion_models():
 
 # ---------------------------------------------
 def download_realesrgan():
-    logger.info("Installing models from RealESRGAN...")
-    model_url = "https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.5.0/realesr-general-x4v3.pth"
-    wdn_model_url = "https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.5.0/realesr-general-wdn-x4v3.pth"
-
-    model_dest = config.root_path / "models/core/upscaling/realesrgan/realesr-general-x4v3.pth"
-    wdn_model_dest = config.root_path / "models/core/upscaling/realesrgan/realesr-general-wdn-x4v3.pth"
-
-    download_with_progress_bar(model_url, str(model_dest), "RealESRGAN")
-    download_with_progress_bar(wdn_model_url, str(wdn_model_dest), "RealESRGANwdn")
-
-
-def download_gfpgan():
-    logger.info("Installing GFPGAN models...")
-    for model in (
-        [
-            "https://github.com/TencentARC/GFPGAN/releases/download/v1.3.0/GFPGANv1.4.pth",
-            "./models/core/face_restoration/gfpgan/GFPGANv1.4.pth",
-        ],
-        [
-            "https://github.com/xinntao/facexlib/releases/download/v0.1.0/detection_Resnet50_Final.pth",
-            "./models/core/face_restoration/gfpgan/weights/detection_Resnet50_Final.pth",
-        ],
-        [
-            "https://github.com/xinntao/facexlib/releases/download/v0.2.2/parsing_parsenet.pth",
-            "./models/core/face_restoration/gfpgan/weights/parsing_parsenet.pth",
-        ],
-    ):
-        model_url, model_dest = model[0], config.root_path / model[1]
-        download_with_progress_bar(model_url, str(model_dest), "GFPGAN weights")
-
+    logger.info("Installing ESRGAN Upscaling models...")
+    URLs = [
+        dict(
+            url = "https://github.com/xinntao/Real-ESRGAN/releases/download/v0.1.0/RealESRGAN_x4plus.pth",
+            dest = "core/upscaling/realesrgan/RealESRGAN_x4plus.pth",
+            description = "RealESRGAN_x4plus.pth",
+        ),
+        dict(
+            url = "https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.2.4/RealESRGAN_x4plus_anime_6B.pth",
+            dest = "core/upscaling/realesrgan/RealESRGAN_x4plus_anime_6B.pth",
+            description = "RealESRGAN_x4plus_anime_6B.pth",
+        ),
+        dict(
+            url= "https://github.com/xinntao/Real-ESRGAN/releases/download/v0.1.1/ESRGAN_SRx4_DF2KOST_official-ff704c30.pth",
+            dest= "core/upscaling/realesrgan/ESRGAN_SRx4_DF2KOST_official-ff704c30.pth",
+            description = "ESRGAN_SRx4_DF2KOST_official.pth",
+        ),
+        dict(
+            url= "https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.1/RealESRGAN_x2plus.pth",
+            dest= "core/upscaling/realesrgan/RealESRGAN_x2plus.pth",
+            description = "RealESRGAN_x2plus.pth",
+        ),
+    ]
+    for model in URLs:
+        download_with_progress_bar(model['url'], config.models_path / model['dest'], model['description'])
 
 # ---------------------------------------------
-def download_codeformer():
-    logger.info("Installing CodeFormer model file...")
-    model_url = (
-        "https://github.com/sczhou/CodeFormer/releases/download/v0.1.0/codeformer.pth"
-    )
-    model_dest = config.root_path / "models/core/face_restoration/codeformer/codeformer.pth"
-    download_with_progress_bar(model_url, str(model_dest), "CodeFormer")
-
-
-# ---------------------------------------------
-def download_clipseg():
-    logger.info("Installing clipseg model for text-based masking...")
-    CLIPSEG_MODEL = "CIDAS/clipseg-rd64-refined"
-    try:
-        hf_download_from_pretrained(AutoProcessor, CLIPSEG_MODEL, config.root_path / 'models/core/misc/clipseg')
-        hf_download_from_pretrained(CLIPSegForImageSegmentation, CLIPSEG_MODEL, config.root_path / 'models/core/misc/clipseg')
-    except Exception:
-        logger.info("Error installing clipseg model:")
-        logger.info(traceback.format_exc())
-
-
 def download_support_models():
     download_realesrgan()
-    download_gfpgan()
-    download_codeformer()
-    download_clipseg()
     download_conversion_models()
 
 # -------------------------------------
@@ -442,21 +411,21 @@ Use cursor arrows to make a checkbox selection, and space to toggle.
         self.nextrely += 1
         self.add_widget_intelligent(
             npyscreen.FixedText,
-            value="Directories containing textual inversion, controlnet and LoRA models (<tab> autocompletes, ctrl-N advances):",
+            value="Folder to recursively scan for new checkpoints, ControlNets, LoRAs and TI models (<tab> autocompletes, ctrl-N advances):",
             editable=False,
             color="CONTROL",
         )
         self.autoimport_dirs = {}
-        for description, config_name, path in autoimport_paths(old_opts):
-            self.autoimport_dirs[config_name] = self.add_widget_intelligent(
-                npyscreen.TitleFilename,
-                name=description+':',
-                value=str(path),
+        self.autoimport_dirs['autoimport_dir'] = self.add_widget_intelligent(
+                FileBox,
+                name=f'Autoimport Folder',
+                value=str(config.root_path / config.autoimport_dir),
                 select_dir=True,
                 must_exist=False,
                 use_two_lines=False,
                 labelColor="GOOD",
                 begin_entry_at=32,
+                max_height = 3,
                 scroll_exit=True
             )
         self.nextrely += 1
@@ -593,7 +562,6 @@ def edit_opts(program_opts: Namespace, invokeai_opts: Namespace) -> argparse.Nam
     editApp.run()
     return editApp.new_opts()
 
-
 def default_startup_options(init_file: Path) -> Namespace:
     opts = InvokeAIAppConfig.get_config()
     if not init_file.exists():
@@ -601,7 +569,14 @@ def default_startup_options(init_file: Path) -> Namespace:
     return opts
 
 def default_user_selections(program_opts: Namespace) -> InstallSelections:
-    installer = ModelInstall(config)
+    
+    try:
+        installer = ModelInstall(config)
+    except omegaconf.errors.ConfigKeyError:
+        logger.warning('Your models.yaml file is corrupt or out of date. Reinitializing')
+        initialize_rootdir(config.root_path, True)
+        installer = ModelInstall(config)
+        
     models = installer.all_models()
     return InstallSelections(
         install_models=[models[installer.default_model()].path or models[installer.default_model()].repo_id]
@@ -609,19 +584,8 @@ def default_user_selections(program_opts: Namespace) -> InstallSelections:
         else [models[x].path or models[x].repo_id for x in installer.recommended_models()]
         if program_opts.yes_to_all
         else list(),
-#        scan_directory=None,
-#        autoscan_on_startup=None,
     )
 
-# -------------------------------------
-def autoimport_paths(config: InvokeAIAppConfig):
-    return [
-        ('Checkpoints & diffusers models', 'autoimport_dir', config.root_path / config.autoimport_dir),
-        ('LoRA/LyCORIS models',            'lora_dir',       config.root_path / config.lora_dir),
-        ('Controlnet models',              'controlnet_dir', config.root_path / config.controlnet_dir),
-        ('Textual Inversion Embeddings',   'embedding_dir',  config.root_path / config.embedding_dir),
-    ]
-    
 # -------------------------------------
 def initialize_rootdir(root: Path, yes_to_all: bool = False):
     logger.info("** INITIALIZING INVOKEAI RUNTIME DIRECTORY **")
@@ -666,7 +630,7 @@ def run_console_ui(
 
     # The third argument is needed in the Windows 11 environment to
     # launch a console window running this program.
-    set_min_terminal_size(MIN_COLS, MIN_LINES,'invokeai-configure')
+    set_min_terminal_size(MIN_COLS, MIN_LINES)
 
     # the install-models application spawns a subprocess to install
     # models, and will crash unless this is set before running.
@@ -697,6 +661,9 @@ def write_opts(opts: Namespace, init_file: Path):
     with open(init_file,'w', encoding='utf-8') as file:
         file.write(new_config.to_yaml())
 
+    if hasattr(opts,'hf_token') and opts.hf_token:
+        HfLogin(opts.hf_token)
+
 # -------------------------------------
 def default_output_dir() -> Path:
     return config.root_path / "outputs"
@@ -743,7 +710,7 @@ def migrate_if_needed(opt: Namespace, root: Path)->bool:
     old_init_file = root / 'invokeai.init'
     new_init_file = root / 'invokeai.yaml'
     old_hub = root / 'models/hub'
-    migration_needed =  old_init_file.exists() and not new_init_file.exists() or old_hub.exists()
+    migration_needed =  (old_init_file.exists() and not new_init_file.exists()) and old_hub.exists()
     
     if migration_needed:
         if opt.yes_to_all or \
@@ -858,9 +825,9 @@ def main():
             download_support_models()
 
         if opt.skip_sd_weights:
-            logger.info("\n** SKIPPING DIFFUSION WEIGHTS DOWNLOAD PER USER REQUEST **")
+            logger.warning("SKIPPING DIFFUSION WEIGHTS DOWNLOAD PER USER REQUEST")
         elif models_to_download:
-            logger.info("\n** DOWNLOADING DIFFUSION WEIGHTS **")
+            logger.info("DOWNLOADING DIFFUSION WEIGHTS")
             process_and_execute(opt, models_to_download)
 
         postscript(errors=errors)

invokeai/backend/install/migrate_to_3.py

@@ -593,9 +593,12 @@ script, which will perform a full upgrade in place."""
     config = InvokeAIAppConfig.get_config()
     config.parse_args(['--root',str(dest_root)])
 
-    # TODO: revisit
-    # assert (dest_root / 'models').is_dir(), f"{dest_root} does not contain a 'models' subdirectory"
-    # assert (dest_root / 'invokeai.yaml').exists(), f"{dest_root} does not contain an InvokeAI init file."
+    # TODO: revisit - don't rely on invokeai.yaml to exist yet!
+    dest_is_setup = (dest_root / 'models/core').exists() and (dest_root / 'databases').exists()
+    if not dest_is_setup:
+        import invokeai.frontend.install.invokeai_configure
+        from invokeai.backend.install.invokeai_configure import initialize_rootdir
+        initialize_rootdir(dest_root, True)
 
     do_migrate(src_root,dest_root)
 

invokeai/backend/install/model_install_backend.py

@@ -10,7 +10,7 @@ from tempfile import TemporaryDirectory
 from typing import List, Dict, Callable, Union, Set
 
 import requests
-from diffusers import StableDiffusionPipeline
+from diffusers import DiffusionPipeline
 from diffusers import logging as dlogging
 from huggingface_hub import hf_hub_url, HfFolder, HfApi
 from omegaconf import OmegaConf
@@ -71,8 +71,6 @@ class ModelInstallList:
 class InstallSelections():
     install_models: List[str]= field(default_factory=list)
     remove_models: List[str]=field(default_factory=list)
-#    scan_directory: Path = None
-#    autoscan_on_startup: bool=False
 
 @dataclass
 class ModelLoadInfo():
@@ -119,10 +117,11 @@ class ModelInstall(object):
 
         # supplement with entries in models.yaml
         installed_models = self.mgr.list_models()
+        
         for md in installed_models:
             base = md['base_model']
-            model_type = md['type']
-            name = md['name']
+            model_type = md['model_type']
+            name = md['model_name']
             key = ModelManager.create_key(name, base, model_type)
             if key in model_dict:
                 model_dict[key].installed = True
@@ -136,6 +135,12 @@ class ModelInstall(object):
                 )
         return {x : model_dict[x] for x in sorted(model_dict.keys(),key=lambda y: model_dict[y].name.lower())}
 
+    def list_models(self, model_type):
+        installed = self.mgr.list_models(model_type=model_type)
+        print(f'Installed models of type `{model_type}`:')
+        for i in installed:
+            print(f"{i['model_name']}\t{i['base_model']}\t{i['path']}")
+
     def starter_models(self)->Set[str]:
         models = set()
         for key, value in self.datasets.items():
@@ -207,7 +212,7 @@ class ModelInstall(object):
                                     {'config.json','model_index.json','learned_embeds.bin','pytorch_lora_weights.bin'}
                                     ]
                                    ):
-            models_installed.update(self._install_path(path))
+            models_installed.update({str(model_path_id_or_url): self._install_path(path)})
 
         # recursive scan
         elif path.is_dir():
@@ -305,6 +310,8 @@ class ModelInstall(object):
         if key := self.reverse_paths.get(path_name):
             (name, base, mtype) = ModelManager.parse_key(key)
             return name
+        elif location.is_dir():
+            return location.name
         else:
             return location.stem
 
@@ -360,7 +367,7 @@ class ModelInstall(object):
         model = None
         for revision in revisions:
             try:
-                model = StableDiffusionPipeline.from_pretrained(repo_id,revision=revision,safety_checker=None)
+                model = DiffusionPipeline.from_pretrained(repo_id,revision=revision,safety_checker=None)
             except:  # most errors are due to fp16 not being present. Fix this to catch other errors
                 pass
             if model:

invokeai/backend/model_management/__init__.py

@@ -3,6 +3,6 @@ Initialization file for invokeai.backend.model_management
 """
 from .model_manager import ModelManager, ModelInfo, AddModelResult, SchedulerPredictionType
 from .model_cache import ModelCache
-from .models import BaseModelType, ModelType, SubModelType, ModelVariantType
+from .models import BaseModelType, ModelType, SubModelType, ModelVariantType, ModelNotFoundException, DuplicateModelException
 from .model_merge import ModelMerger, MergeInterpolationMethod
 

invokeai/backend/model_management/convert_ckpt_to_diffusers.py

@@ -21,6 +21,7 @@ import re
 import warnings
 from pathlib import Path
 from typing import Union
+from packaging import version
 
 import torch
 from safetensors.torch import load_file
@@ -63,6 +64,7 @@ from diffusers.pipelines.stable_diffusion.safety_checker import (
     StableDiffusionSafetyChecker,
 )
 from diffusers.utils import is_safetensors_available
+import transformers
 from transformers import (
     AutoFeatureExtractor,
     BertTokenizerFast,
@@ -841,7 +843,16 @@ def convert_ldm_clip_checkpoint(checkpoint):
                 key
             ]
 
-    text_model.load_state_dict(text_model_dict)
+    # transformers 4.31.0 and higher - this key no longer in state dict
+    if version.parse(transformers.__version__) >= version.parse("4.31.0"):
+        position_ids = text_model_dict.pop("text_model.embeddings.position_ids", None)
+        text_model.load_state_dict(text_model_dict)
+        if position_ids is not None:
+            text_model.text_model.embeddings.position_ids.copy_(position_ids)
+
+    # transformers 4.30.2 and lower - position_ids is part of state_dict
+    else:
+        text_model.load_state_dict(text_model_dict)        
 
     return text_model
 
@@ -947,7 +958,16 @@ def convert_open_clip_checkpoint(checkpoint):
 
                 text_model_dict[new_key] = checkpoint[key]
 
-    text_model.load_state_dict(text_model_dict)
+    # transformers 4.31.0 and higher - this key no longer in state dict
+    if version.parse(transformers.__version__) >= version.parse("4.31.0"):
+        position_ids = text_model_dict.pop("text_model.embeddings.position_ids", None)
+        text_model.load_state_dict(text_model_dict)
+        if position_ids is not None:
+            text_model.text_model.embeddings.position_ids.copy_(position_ids)
+
+    # transformers 4.30.2 and lower - position_ids is part of state_dict
+    else:
+        text_model.load_state_dict(text_model_dict)  
 
     return text_model
 

invokeai/backend/model_management/model_cache.py

@@ -104,7 +104,8 @@ class ModelCache(object):
         :param sha_chunksize: Chunksize to use when calculating sha256 model hash
         '''
         self.model_infos: Dict[str, ModelBase] = dict()
-        self.lazy_offloading = lazy_offloading
+        # allow lazy offloading only when vram cache enabled
+        self.lazy_offloading = lazy_offloading and max_vram_cache_size > 0
         self.precision: torch.dtype=precision
         self.max_cache_size: float=max_cache_size
         self.max_vram_cache_size: float=max_vram_cache_size
@@ -327,6 +328,25 @@ class ModelCache(object):
 
             refs = sys.getrefcount(cache_entry.model)
 
+            # manualy clear local variable references of just finished function calls
+            # for some reason python don't want to collect it even by gc.collect() immidiately
+            if refs > 2:
+                while True:
+                    cleared = False
+                    for referrer in gc.get_referrers(cache_entry.model):
+                        if type(referrer).__name__ == "frame":
+                            # RuntimeError: cannot clear an executing frame
+                            with suppress(RuntimeError):
+                                referrer.clear()
+                                cleared = True
+                                #break
+
+                    # repeat if referrers changes(due to frame clear), else exit loop
+                    if cleared:
+                        gc.collect()
+                    else:
+                        break
+
             device = cache_entry.model.device if hasattr(cache_entry.model, "device") else None
             self.logger.debug(f"Model: {model_key}, locks: {cache_entry._locks}, device: {device}, loaded: {cache_entry.loaded}, refs: {refs}")
 
@@ -362,6 +382,9 @@ class ModelCache(object):
                 self.logger.debug(f'GPU VRAM freed: {(mem.vram_used/GIG):.2f} GB')
                 vram_in_use += mem.vram_used  # note vram_used is negative
                 self.logger.debug(f'{(vram_in_use/GIG):.2f}GB VRAM used for models; max allowed={(reserved/GIG):.2f}GB')
+
+        gc.collect()
+        torch.cuda.empty_cache()
         
     def _local_model_hash(self, model_path: Union[str, Path]) -> str:
         sha = hashlib.sha256()

invokeai/backend/model_management/model_manager.py

@@ -106,16 +106,16 @@ providing information about a model defined in models.yaml. For example:
 
    >>> models = mgr.list_models()
    >>> json.dumps(models[0])
-  {"path": "/home/lstein/invokeai-main/models/sd-1/controlnet/canny", 
-    "model_format": "diffusers", 
-    "name": "canny", 
-    "base_model": "sd-1", 
+  {"path": "/home/lstein/invokeai-main/models/sd-1/controlnet/canny",
+    "model_format": "diffusers",
+    "name": "canny",
+    "base_model": "sd-1",
     "type": "controlnet"
    }
 
 You can filter by model type and base model as shown here:
 
-  
+
   controlnets = mgr.list_models(model_type=ModelType.ControlNet,
                                 base_model=BaseModelType.StableDiffusion1)
   for c in controlnets:
@@ -140,14 +140,14 @@ Layout of the `models` directory:
 
  models
  ├── sd-1
- │   ├── controlnet
- │   ├── lora
- │   ├── main
- │   └── embedding
+ │   ├── controlnet
+ │   ├── lora
+ │   ├── main
+ │   └── embedding
  ├── sd-2
- │   ├── controlnet
- │   ├── lora
- │   ├── main
+ │   ├── controlnet
+ │   ├── lora
+ │   ├── main
  │   └── embedding
  └── core
      ├── face_reconstruction
@@ -195,7 +195,7 @@ name, base model, type and a dict of model attributes. See
 `invokeai/backend/model_management/models` for the attributes required
 by each model type.
 
-A model can be deleted using `del_model()`, providing the same 
+A model can be deleted using `del_model()`, providing the same
 identifying information as `get_model()`
 
 The `heuristic_import()` method will take a set of strings
@@ -247,10 +247,13 @@ import invokeai.backend.util.logging as logger
 from invokeai.app.services.config import InvokeAIAppConfig
 from invokeai.backend.util import CUDA_DEVICE, Chdir
 from .model_cache import ModelCache, ModelLocker
+from .model_search import ModelSearch
 from .models import (
     BaseModelType, ModelType, SubModelType,
     ModelError, SchedulerPredictionType, MODEL_CLASSES,
-    ModelConfigBase, ModelNotFoundException, InvalidModelException,
+    ModelConfigBase,
+    ModelNotFoundException, InvalidModelException,
+    DuplicateModelException,
 )
 
 # We are only starting to number the config file with release 3.
@@ -303,7 +306,7 @@ class ModelManager(object):
         logger: types.ModuleType = logger,
     ):
         """
-        Initialize with the path to the models.yaml config file. 
+        Initialize with the path to the models.yaml config file.
         Optional parameters are the torch device type, precision, max_models,
         and sequential_offload boolean. Note that the default device
         type and precision are set up for a CUDA system running at half precision.
@@ -323,15 +326,6 @@ class ModelManager(object):
         # TODO: metadata not found
         # TODO: version check
 
-        self.models = dict()
-        for model_key, model_config in config.items():
-            model_name, base_model, model_type = self.parse_key(model_key)
-            model_class = MODEL_CLASSES[base_model][model_type]
-            # alias for config file
-            model_config["model_format"] = model_config.pop("format")
-            self.models[model_key] = model_class.create_config(**model_config)
-
-        # check config version number and update on disk/RAM if necessary
         self.app_config = InvokeAIAppConfig.get_config()
         self.logger = logger
         self.cache = ModelCache(
@@ -342,11 +336,41 @@ class ModelManager(object):
             sequential_offload = sequential_offload,
             logger = logger,
         )
+
+        self._read_models(config)
+
+    def _read_models(self, config: Optional[DictConfig] = None):
+        if not config:
+            if self.config_path:
+                config = OmegaConf.load(self.config_path)
+            else:
+                return
+
+        self.models = dict()
+        for model_key, model_config in config.items():
+            if model_key.startswith('_'):
+                continue
+            model_name, base_model, model_type = self.parse_key(model_key)
+            model_class = MODEL_CLASSES[base_model][model_type]
+            # alias for config file
+            model_config["model_format"] = model_config.pop("format")
+            self.models[model_key] = model_class.create_config(**model_config)
+
+        # check config version number and update on disk/RAM if necessary
         self.cache_keys = dict()
 
         # add controlnet, lora and textual_inversion models from disk
         self.scan_models_directory()
 
+    def sync_to_config(self):
+        """
+        Call this when `models.yaml` has been changed externally.
+        This will reinitialize internal data structures
+        """
+        # Reread models directory; note that this will reinitialize the cache,
+        # causing otherwise unreferenced models to be removed from memory
+        self._read_models()
+
     def model_exists(
         self,
         model_name: str,
@@ -409,7 +433,7 @@ class ModelManager(object):
         :param model_name: symbolic name of the model in models.yaml
         :param model_type: ModelType enum indicating the type of model to return
         :param base_model: BaseModelType enum indicating the base model used by this model
-        :param submode_typel: an ModelType enum indicating the portion of 
+        :param submode_typel: an ModelType enum indicating the portion of
                the model to retrieve (e.g. ModelType.Vae)
         """
         model_class = MODEL_CLASSES[base_model][model_type]
@@ -434,7 +458,7 @@ class ModelManager(object):
                 raise ModelNotFoundException(f"Model not found - {model_key}")
 
         # vae/movq override
-        # TODO: 
+        # TODO:
         if submodel_type is not None and hasattr(model_config, submodel_type):
             override_path = getattr(model_config, submodel_type)
             if override_path:
@@ -467,7 +491,7 @@ class ModelManager(object):
         self.cache_keys[model_key].add(model_context.key)
 
         model_hash = "<NO_HASH>" # TODO:
-            
+
         return ModelInfo(
             context = model_context,
             name = model_name,
@@ -496,7 +520,7 @@ class ModelManager(object):
 
     def model_names(self) -> List[Tuple[str, BaseModelType, ModelType]]:
         """
-        Return a list of (str, BaseModelType, ModelType) corresponding to all models 
+        Return a list of (str, BaseModelType, ModelType) corresponding to all models
         known to the configuration.
         """
         return [(self.parse_key(x)) for x in self.models.keys()]
@@ -527,7 +551,10 @@ class ModelManager(object):
         model_keys = [self.create_key(model_name, base_model, model_type)] if model_name else sorted(self.models, key=str.casefold)
         models = []
         for model_key in model_keys:
-            model_config = self.models[model_key]
+            model_config = self.models.get(model_key)
+            if not model_config:
+                self.logger.error(f'Unknown model {model_name}')
+                raise ModelNotFoundException(f'Unknown model {model_name}')
 
             cur_model_name, cur_base_model, cur_model_type = self.parse_key(model_key)
             if base_model is not None and cur_base_model != base_model:
@@ -538,11 +565,14 @@ class ModelManager(object):
             model_dict = dict(
                 **model_config.dict(exclude_defaults=True),
                 # OpenAPIModelInfoBase
-                name=cur_model_name,
+                model_name=cur_model_name,
                 base_model=cur_base_model,
-                type=cur_model_type,
+                model_type=cur_model_type,
             )
 
+            # expose paths as absolute to help web UI
+            if path := model_dict.get('path'):
+                model_dict['path'] = str(self.app_config.root_path / path)
             models.append(model_dict)
 
         return models
@@ -571,7 +601,7 @@ class ModelManager(object):
         model_cfg = self.models.pop(model_key, None)
 
         if model_cfg is None:
-            raise KeyError(f"Unknown model {model_key}")
+            raise ModelNotFoundException(f"Unknown model {model_key}")
 
         # note: it not garantie to release memory(model can has other references)
         cache_ids = self.cache_keys.pop(model_key, [])
@@ -589,6 +619,7 @@ class ModelManager(object):
                 rmtree(str(model_path))
             else:
                 model_path.unlink()
+        self.commit()
 
     # LS: tested
     def add_model(
@@ -609,6 +640,10 @@ class ModelManager(object):
         The returned dict has the same format as the dict returned by
         model_info().
         """
+        # relativize paths as they go in - this makes it easier to move the root directory around
+        if path := model_attributes.get('path'):
+            if Path(path).is_relative_to(self.app_config.root_path):
+                model_attributes['path'] = str(Path(path).relative_to(self.app_config.root_path))
 
         model_class = MODEL_CLASSES[base_model][model_type]
         model_config = model_class.create_config(**model_attributes)
@@ -645,11 +680,61 @@ class ModelManager(object):
             config = model_config,
         )
 
+    def rename_model(
+            self,
+            model_name: str,
+            base_model: BaseModelType,
+            model_type: ModelType,
+            new_name: str = None,
+            new_base: BaseModelType = None,
+    ):
+        '''
+        Rename or rebase a model.
+        '''
+        if new_name is None and new_base is None:
+            self.logger.error("rename_model() called with neither a new_name nor a new_base. {model_name} unchanged.")
+            return
+
+        model_key = self.create_key(model_name, base_model, model_type)
+        model_cfg = self.models.get(model_key, None)
+        if not model_cfg:
+            raise ModelNotFoundException(f"Unknown model: {model_key}")
+
+        old_path = self.app_config.root_path / model_cfg.path
+        new_name = new_name or model_name
+        new_base = new_base or base_model
+        new_key = self.create_key(new_name, new_base, model_type)
+        if new_key in self.models:
+            raise ValueError(f'Attempt to overwrite existing model definition "{new_key}"')
+
+        # if this is a model file/directory that we manage ourselves, we need to move it
+        if old_path.is_relative_to(self.app_config.models_path):
+            new_path = self.app_config.root_path / 'models' / BaseModelType(new_base).value / ModelType(model_type).value / new_name
+            move(old_path, new_path)
+            model_cfg.path = str(new_path.relative_to(self.app_config.root_path))
+
+        # clean up caches
+        old_model_cache = self._get_model_cache_path(old_path)
+        if old_model_cache.exists():
+            if old_model_cache.is_dir():
+                rmtree(str(old_model_cache))
+            else:
+                old_model_cache.unlink()
+
+        cache_ids = self.cache_keys.pop(model_key, [])
+        for cache_id in cache_ids:
+            self.cache.uncache_model(cache_id)
+
+        self.models.pop(model_key, None) # delete
+        self.models[new_key] = model_cfg
+        self.commit()
+
     def convert_model (
             self,
             model_name: str,
             base_model: BaseModelType,
             model_type: Union[ModelType.Main,ModelType.Vae],
+            dest_directory: Optional[Path]=None,
     ) -> AddModelResult:
         '''
         Convert a checkpoint file into a diffusers folder, deleting the cached
@@ -676,14 +761,14 @@ class ModelManager(object):
                                )
         checkpoint_path = self.app_config.root_path / info["path"]
         old_diffusers_path = self.app_config.models_path / model.location
-        new_diffusers_path = self.app_config.models_path / base_model.value / model_type.value / model_name
+        new_diffusers_path = (dest_directory or self.app_config.models_path / base_model.value / model_type.value) / model_name
         if new_diffusers_path.exists():
             raise ValueError(f"A diffusers model already exists at {new_diffusers_path}")
 
         try:
             move(old_diffusers_path,new_diffusers_path)
             info["model_format"] = "diffusers"
-            info["path"] = str(new_diffusers_path.relative_to(self.app_config.root_path))
+            info["path"] = str(new_diffusers_path) if dest_directory else str(new_diffusers_path.relative_to(self.app_config.root_path))
             info.pop('config')
 
             result = self.add_model(model_name, base_model, model_type,
@@ -693,12 +778,12 @@ class ModelManager(object):
             # something went wrong, so don't leave dangling diffusers model in directory or it will cause a duplicate model error!
             rmtree(new_diffusers_path)
             raise
-        
+
         if checkpoint_path.exists() and checkpoint_path.is_relative_to(self.app_config.models_path):
             checkpoint_path.unlink()
-        
+
         return result
-    
+
     def search_models(self, search_folder):
         self.logger.info(f"Finding Models In: {search_folder}")
         models_folder_ckpt = Path(search_folder).glob("**/*.ckpt")
@@ -741,10 +826,14 @@ class ModelManager(object):
         assert config_file_path is not None,'no config file path to write to'
         config_file_path = self.app_config.root_path / config_file_path
         tmpfile = os.path.join(os.path.dirname(config_file_path), "new_config.tmp")
-        with open(tmpfile, "w", encoding="utf-8") as outfile:
-            outfile.write(self.preamble())
-            outfile.write(yaml_str)
-        os.replace(tmpfile, config_file_path)
+        try:
+            with open(tmpfile, "w", encoding="utf-8") as outfile:
+                outfile.write(self.preamble())
+                outfile.write(yaml_str)
+            os.replace(tmpfile, config_file_path)
+        except OSError as err:
+            self.logger.warning(f"Could not modify the config file at {config_file_path}")
+            self.logger.warning(err)
 
     def preamble(self) -> str:
         """
@@ -771,7 +860,7 @@ class ModelManager(object):
         loaded_files = set()
         new_models_found = False
 
-        self.logger.info(f'scanning {self.app_config.models_path} for new models')
+        self.logger.info(f'Scanning {self.app_config.models_path} for new models')
         with Chdir(self.app_config.root_path):
             for model_key, model_config in list(self.models.items()):
                 model_name, cur_base_model, cur_model_type = self.parse_key(model_key)
@@ -804,15 +893,18 @@ class ModelManager(object):
                             model_name = model_path.name if model_path.is_dir() else model_path.stem
                             model_key = self.create_key(model_name, cur_base_model, cur_model_type)
 
-                            if model_key in self.models:
-                                raise Exception(f"Model with key {model_key} added twice")
-
-                            if model_path.is_relative_to(self.app_config.root_path):
-                                model_path = model_path.relative_to(self.app_config.root_path)
                             try:
+                                if model_key in self.models:
+                                    raise DuplicateModelException(f"Model with key {model_key} added twice")
+
+                                if model_path.is_relative_to(self.app_config.root_path):
+                                    model_path = model_path.relative_to(self.app_config.root_path)
+                                    
                                 model_config: ModelConfigBase = model_class.probe_config(str(model_path))
                                 self.models[model_key] = model_config
                                 new_models_found = True
+                            except DuplicateModelException as e:
+                                self.logger.warning(e)
                             except InvalidModelException:
                                 self.logger.warning(f"Not a valid model: {model_path}")
                             except NotImplementedError as e:
@@ -823,6 +915,7 @@ class ModelManager(object):
         if (new_models_found or imported_models) and self.config_path:
             self.commit()
 
+
     def autoimport(self)->Dict[str, AddModelResult]:
         '''
         Scan the autoimport directory (if defined) and import new models, delete defunct models.
@@ -830,63 +923,50 @@ class ModelManager(object):
         # avoid circular import
         from invokeai.backend.install.model_install_backend import ModelInstall
         from invokeai.frontend.install.model_install import ask_user_for_prediction_type
-        
+
+        class ScanAndImport(ModelSearch):
+            def __init__(self, directories, logger, ignore: Set[Path], installer: ModelInstall):
+                super().__init__(directories, logger)
+                self.installer = installer
+                self.ignore = ignore
+
+            def on_search_started(self):
+                self.new_models_found = dict()
+
+            def on_model_found(self, model: Path):
+                if model not in self.ignore:
+                    self.new_models_found.update(self.installer.heuristic_import(model))
+
+            def on_search_completed(self):
+                self.logger.info(f'Scanned {self._items_scanned} files and directories, imported {len(self.new_models_found)} models')
+
+            def models_found(self):
+                return self.new_models_found
+
+        config = self.app_config
+
+        # LS: hacky
+        # Patch in the SD VAE from core so that it is available for use by the UI
+        try:
+            self.heuristic_import({config.root_path / 'models/core/convert/sd-vae-ft-mse'})
+        except:
+            pass
+
         installer = ModelInstall(config = self.app_config,
                                  model_manager = self,
                                  prediction_type_helper = ask_user_for_prediction_type,
                                  )
+        known_paths = {config.root_path / x['path'] for x in self.list_models()}
+        directories = {config.root_path / x for x in [config.autoimport_dir,
+                                                      config.lora_dir,
+                                                      config.embedding_dir,
+                                                      config.controlnet_dir,
+                                                      ] if x
+                       }
+        scanner = ScanAndImport(directories, self.logger, ignore=known_paths, installer=installer)
+        scanner.search()
         
-        scanned_dirs = set()
-        
-        config = self.app_config
-        known_paths = {(self.app_config.root_path / x['path']) for x in self.list_models()}
-
-        for autodir in [config.autoimport_dir,
-                        config.lora_dir,
-                        config.embedding_dir,
-                        config.controlnet_dir]:
-            if autodir is None:
-                continue
-
-            self.logger.info(f'Scanning {autodir} for models to import')
-            installed = dict()
-        
-            autodir = self.app_config.root_path / autodir
-            if not autodir.exists():
-                continue
-
-            items_scanned = 0
-            new_models_found = dict()
-            
-            for root, dirs, files in os.walk(autodir):
-                items_scanned += len(dirs) + len(files)
-                for d in dirs:
-                    path = Path(root) / d
-                    if path in known_paths or path.parent in scanned_dirs:
-                        scanned_dirs.add(path)
-                        continue
-                    if any([(path/x).exists() for x in {'config.json','model_index.json','learned_embeds.bin','pytorch_lora_weights.bin'}]):
-                        try:
-                            new_models_found.update(installer.heuristic_import(path))
-                            scanned_dirs.add(path)
-                        except ValueError as e:
-                            self.logger.warning(str(e))
-
-                for f in files:
-                    path = Path(root) / f
-                    if path in known_paths or path.parent in scanned_dirs:
-                        continue
-                    if path.suffix in {'.ckpt','.bin','.pth','.safetensors','.pt'}:
-                        try:
-                            import_result = installer.heuristic_import(path)
-                            new_models_found.update(import_result)
-                        except ValueError as e:
-                            self.logger.warning(str(e))
-
-            self.logger.info(f'Scanned {items_scanned} files and directories, imported {len(new_models_found)} models')
-            installed.update(new_models_found)
-
-        return installed
+        return scanner.models_found()
 
     def heuristic_import(self,
                          items_to_import: Set[str],
@@ -909,18 +989,18 @@ class ModelManager(object):
         that model.
 
         May return the following exceptions:
-        - KeyError   - one or more of the items to import is not a valid path, repo_id or URL
+        - ModelNotFoundException   - one or more of the items to import is not a valid path, repo_id or URL
         - ValueError - a corresponding model already exists
         '''
         # avoid circular import here
         from invokeai.backend.install.model_install_backend import ModelInstall
         successfully_installed = dict()
-        
+
         installer = ModelInstall(config = self.app_config,
                                  prediction_type_helper = prediction_type_helper,
                                  model_manager = self)
         for thing in items_to_import:
             installed = installer.heuristic_import(thing)
             successfully_installed.update(installed)
-        self.commit()                
+        self.commit()
         return successfully_installed

invokeai/backend/model_management/model_merge.py

@@ -11,7 +11,7 @@ from enum import Enum
 from pathlib import Path
 from diffusers import DiffusionPipeline
 from diffusers import logging as dlogging
-from typing import List, Union
+from typing import List, Union, Optional
 
 import invokeai.backend.util.logging as logger
 
@@ -74,6 +74,7 @@ class ModelMerger(object):
             alpha: float = 0.5,
             interp: MergeInterpolationMethod = None,
             force: bool = False,
+            merge_dest_directory: Optional[Path] = None,
             **kwargs,
     ) -> AddModelResult:
         """
@@ -85,7 +86,7 @@ class ModelMerger(object):
         :param interp: The interpolation method to use for the merging. Supports "weighted_average", "sigmoid", "inv_sigmoid", "add_difference" and None.
                    Passing None uses the default interpolation which is weighted sum interpolation. For merging three checkpoints, only "add_difference" is supported. Add_difference is A+(B-C).
         :param force:  Whether to ignore mismatch in model_config.json for the current models. Defaults to False.
-
+        :param merge_dest_directory: Save the merged model to the designated directory (with 'merged_model_name' appended)
         **kwargs - the default DiffusionPipeline.get_config_dict kwargs:
              cache_dir, resume_download, force_download, proxies, local_files_only, use_auth_token, revision, torch_dtype, device_map
         """
@@ -111,7 +112,7 @@ class ModelMerger(object):
         merged_pipe = self.merge_diffusion_models(
            model_paths, alpha, merge_method, force, **kwargs
         )
-        dump_path = config.models_path / base_model.value / ModelType.Main.value
+        dump_path = Path(merge_dest_directory) if merge_dest_directory else config.models_path / base_model.value / ModelType.Main.value
         dump_path.mkdir(parents=True, exist_ok=True)
         dump_path = dump_path / merged_model_name
 

invokeai/backend/model_management/model_probe.py

@@ -12,6 +12,7 @@ from picklescan.scanner import scan_file_path
 from .models import (
     BaseModelType, ModelType, ModelVariantType,
     SchedulerPredictionType, SilenceWarnings,
+    InvalidModelException
 )
 from .models.base import read_checkpoint_meta
 
@@ -38,6 +39,9 @@ class ModelProbe(object):
 
     CLASS2TYPE = {
         'StableDiffusionPipeline' : ModelType.Main,
+        'StableDiffusionInpaintPipeline' : ModelType.Main,
+        'StableDiffusionXLPipeline' : ModelType.Main,
+        'StableDiffusionXLImg2ImgPipeline' : ModelType.Main,
         'AutoencoderKL' : ModelType.Vae,
         'ControlNetModel' : ModelType.ControlNet,
     }
@@ -59,7 +63,7 @@ class ModelProbe(object):
         elif isinstance(model,(dict,ModelMixin,ConfigMixin)):
             return cls.probe(model_path=None, model=model, prediction_type_helper=prediction_type_helper)
         else:
-            raise ValueError("model parameter {model} is neither a Path, nor a model")
+            raise InvalidModelException("model parameter {model} is neither a Path, nor a model")
 
     @classmethod
     def probe(cls,
@@ -99,9 +103,10 @@ class ModelProbe(object):
                 upcast_attention = (base_type==BaseModelType.StableDiffusion2 \
                                      and prediction_type==SchedulerPredictionType.VPrediction),
                 format = format,
-                image_size = 768 if (base_type==BaseModelType.StableDiffusion2 \
-                                     and prediction_type==SchedulerPredictionType.VPrediction \
-                                     ) else 512,
+                image_size = 1024 if (base_type in {BaseModelType.StableDiffusionXL,BaseModelType.StableDiffusionXLRefiner}) else \
+                              768 if (base_type==BaseModelType.StableDiffusion2 \
+                                     and prediction_type==SchedulerPredictionType.VPrediction ) else \
+                              512
             )
         except Exception:
             raise
@@ -138,7 +143,7 @@ class ModelProbe(object):
             if len(ckpt) < 10 and all(isinstance(v, torch.Tensor) for v in ckpt.values()):
                 return ModelType.TextualInversion
         
-        raise ValueError(f"Unable to determine model type for {model_path}")
+        raise InvalidModelException(f"Unable to determine model type for {model_path}")
 
     @classmethod
     def get_model_type_from_folder(cls, folder_path: Path, model: ModelMixin)->ModelType:
@@ -168,7 +173,7 @@ class ModelProbe(object):
             return type
 
         # give up
-        raise ValueError(f"Unable to determine model type for {folder_path}")
+        raise InvalidModelException(f"Unable to determine model type for {folder_path}")
 
     @classmethod
     def _scan_and_load_checkpoint(cls,model_path: Path)->dict:
@@ -237,7 +242,7 @@ class CheckpointProbeBase(ProbeBase):
         elif in_channels == 4:
             return ModelVariantType.Normal
         else:
-            raise ValueError(f"Cannot determine variant type (in_channels={in_channels}) at {self.checkpoint_path}")
+            raise InvalidModelException(f"Cannot determine variant type (in_channels={in_channels}) at {self.checkpoint_path}")
 
 class PipelineCheckpointProbe(CheckpointProbeBase):
     def get_base_type(self)->BaseModelType:
@@ -248,7 +253,10 @@ class PipelineCheckpointProbe(CheckpointProbeBase):
             return BaseModelType.StableDiffusion1
         if key_name in state_dict and state_dict[key_name].shape[-1] == 1024:
             return BaseModelType.StableDiffusion2
-        raise ValueError("Cannot determine base type")
+        # TODO: Verify that this is correct! Need an XL checkpoint file for this.
+        if key_name in state_dict and state_dict[key_name].shape[-1] == 2048:
+            return BaseModelType.StableDiffusionXL
+        raise InvalidModelException("Cannot determine base type")
 
     def get_scheduler_prediction_type(self)->SchedulerPredictionType:
         type = self.get_base_type()
@@ -329,7 +337,7 @@ class ControlNetCheckpointProbe(CheckpointProbeBase):
                 return BaseModelType.StableDiffusion2
             elif self.checkpoint_path and self.helper:
                 return self.helper(self.checkpoint_path)
-        raise ValueError("Unable to determine base type for {self.checkpoint_path}")
+        raise InvalidModelException("Unable to determine base type for {self.checkpoint_path}")
 
 ########################################################
 # classes for probing folders
@@ -360,8 +368,12 @@ class PipelineFolderProbe(FolderProbeBase):
             return BaseModelType.StableDiffusion1  
         elif unet_conf['cross_attention_dim'] == 1024:
             return BaseModelType.StableDiffusion2
+        elif unet_conf['cross_attention_dim'] == 1280:
+            return BaseModelType.StableDiffusionXLRefiner
+        elif unet_conf['cross_attention_dim'] == 2048:
+            return BaseModelType.StableDiffusionXL
         else:
-            raise ValueError(f'Unknown base model for {self.folder_path}')
+            raise InvalidModelException(f'Unknown base model for {self.folder_path}')
 
     def get_scheduler_prediction_type(self)->SchedulerPredictionType:
         if self.model:
@@ -390,7 +402,7 @@ class PipelineFolderProbe(FolderProbeBase):
                 
             in_channels = conf['in_channels']
             if in_channels == 9:
-                return ModelVariantType.Inpainting
+                return ModelVariantType.Inpaint
             elif in_channels == 5:
                 return ModelVariantType.Depth
             elif in_channels == 4:
@@ -418,7 +430,7 @@ class ControlNetFolderProbe(FolderProbeBase):
     def get_base_type(self)->BaseModelType:
         config_file = self.folder_path / 'config.json'
         if not config_file.exists():
-            raise ValueError(f"Cannot determine base type for {self.folder_path}")
+            raise InvalidModelException(f"Cannot determine base type for {self.folder_path}")
         with open(config_file,'r') as file:
             config = json.load(file)
         # no obvious way to distinguish between sd2-base and sd2-768
@@ -435,7 +447,7 @@ class LoRAFolderProbe(FolderProbeBase):
                 model_file = base_file
                 break
         if not model_file:
-            raise ValueError('Unknown LoRA format encountered')
+            raise InvalidModelException('Unknown LoRA format encountered')
         return LoRACheckpointProbe(model_file,None).get_base_type()
 
 ############## register probe classes ######

invokeai/backend/model_management/model_search.py

@@ -0,0 +1,103 @@
+# Copyright 2023, Lincoln D. Stein and the InvokeAI Team
+"""
+Abstract base class for recursive directory search for models.
+"""
+
+import os
+from abc import ABC, abstractmethod
+from typing import List, Set, types
+from pathlib import Path
+
+import invokeai.backend.util.logging as logger
+
+class ModelSearch(ABC):
+    def __init__(self, directories: List[Path], logger: types.ModuleType=logger):
+        """
+        Initialize a recursive model directory search.
+        :param directories: List of directory Paths to recurse through
+        :param logger: Logger to use
+        """
+        self.directories = directories
+        self.logger = logger
+        self._items_scanned = 0
+        self._models_found = 0
+        self._scanned_dirs = set()
+        self._scanned_paths = set()
+        self._pruned_paths = set()
+
+    @abstractmethod
+    def on_search_started(self):
+        """
+        Called before the scan starts.
+        """
+        pass
+
+    @abstractmethod
+    def on_model_found(self, model: Path):
+        """
+        Process a found model. Raise an exception if something goes wrong.
+        :param model: Model to process - could be a directory or checkpoint.
+        """
+        pass
+
+    @abstractmethod
+    def on_search_completed(self):
+        """
+        Perform some activity when the scan is completed. May use instance
+        variables, items_scanned and models_found
+        """
+        pass
+
+    def search(self):
+        self.on_search_started()
+        for dir in self.directories:
+            self.walk_directory(dir)
+        self.on_search_completed()
+
+    def walk_directory(self, path: Path):
+        for root, dirs, files in os.walk(path):
+            if str(Path(root).name).startswith('.'):
+                self._pruned_paths.add(root)
+            if any([Path(root).is_relative_to(x) for x in self._pruned_paths]):
+                continue
+            
+            self._items_scanned += len(dirs) + len(files)
+            for d in dirs:
+                path = Path(root) / d
+                if path in self._scanned_paths or path.parent in self._scanned_dirs:
+                    self._scanned_dirs.add(path)
+                    continue
+                if any([(path/x).exists() for x in {'config.json','model_index.json','learned_embeds.bin','pytorch_lora_weights.bin'}]):
+                    try:
+                        self.on_model_found(path)
+                        self._models_found += 1
+                        self._scanned_dirs.add(path)
+                    except Exception as e:
+                        self.logger.warning(str(e))
+
+            for f in files:
+                path = Path(root) / f
+                if path.parent in self._scanned_dirs:
+                    continue
+                if path.suffix in {'.ckpt','.bin','.pth','.safetensors','.pt'}:
+                    try:
+                        self.on_model_found(path)
+                        self._models_found += 1
+                    except Exception as e:
+                        self.logger.warning(str(e))
+
+class FindModels(ModelSearch):
+    def on_search_started(self):
+        self.models_found: Set[Path] = set()
+
+    def on_model_found(self,model: Path):
+        self.models_found.add(model)
+
+    def on_search_completed(self):
+        pass
+
+    def list_models(self) -> List[Path]:
+        self.search()
+        return self.models_found
+
+    

invokeai/backend/model_management/models/__init__.py

@@ -2,8 +2,13 @@ import inspect
 from enum import Enum
 from pydantic import BaseModel
 from typing import Literal, get_origin
-from .base import BaseModelType, ModelType, SubModelType, ModelBase, ModelConfigBase, ModelVariantType, SchedulerPredictionType, ModelError, SilenceWarnings, ModelNotFoundException, InvalidModelException
+from .base import (
+    BaseModelType, ModelType, SubModelType, ModelBase, ModelConfigBase,
+    ModelVariantType, SchedulerPredictionType, ModelError, SilenceWarnings,
+    ModelNotFoundException, InvalidModelException, DuplicateModelException
+    )
 from .stable_diffusion import StableDiffusion1Model, StableDiffusion2Model
+from .sdxl import StableDiffusionXLModel
 from .vae import VaeModel
 from .lora import LoRAModel
 from .controlnet import ControlNetModel # TODO:
@@ -24,6 +29,22 @@ MODEL_CLASSES = {
         ModelType.ControlNet: ControlNetModel,
         ModelType.TextualInversion: TextualInversionModel,
     },
+    BaseModelType.StableDiffusionXL: {
+        ModelType.Main: StableDiffusionXLModel,
+        ModelType.Vae: VaeModel,
+        # will not work until support written
+        ModelType.Lora: LoRAModel,
+        ModelType.ControlNet: ControlNetModel,
+        ModelType.TextualInversion: TextualInversionModel,
+    },
+    BaseModelType.StableDiffusionXLRefiner: {
+        ModelType.Main: StableDiffusionXLModel,
+        ModelType.Vae: VaeModel,
+        # will not work until support written
+        ModelType.Lora: LoRAModel,
+        ModelType.ControlNet: ControlNetModel,
+        ModelType.TextualInversion: TextualInversionModel,
+    },
     #BaseModelType.Kandinsky2_1: {
     #    ModelType.Main: Kandinsky2_1Model,
     #    ModelType.MoVQ: MoVQModel,
@@ -37,9 +58,9 @@ MODEL_CONFIGS = list()
 OPENAPI_MODEL_CONFIGS = list()
 
 class OpenAPIModelInfoBase(BaseModel):
-    name: str
+    model_name: str
     base_model: BaseModelType
-    type: ModelType
+    model_type: ModelType
 
 
 for base_model, models in MODEL_CLASSES.items():
@@ -48,7 +69,9 @@ for base_model, models in MODEL_CLASSES.items():
         model_configs.discard(None)
         MODEL_CONFIGS.extend(model_configs)
 
-        for cfg in model_configs:
+        # LS: sort to get the checkpoint configs first, which makes
+        # for a better template in the Swagger docs
+        for cfg in sorted(model_configs, key=lambda x: str(x)):
             model_name, cfg_name = cfg.__qualname__.split('.')[-2:]
             openapi_cfg_name = model_name + cfg_name
             if openapi_cfg_name in vars():
@@ -56,7 +79,7 @@ for base_model, models in MODEL_CLASSES.items():
 
             api_wrapper = type(openapi_cfg_name, (cfg, OpenAPIModelInfoBase), dict(
                 __annotations__ = dict(
-                    type=Literal[model_type.value],
+                    model_type=Literal[model_type.value],
                 ),
             ))