## What type of PR is this? (check all applicable)
- [X] Refactor
## Have you discussed this change with the InvokeAI team?
- [X] Extensively
- [ ] No, because:
## Have you updated all relevant documentation?
- [X] Yes
- [ ] No
## Description
As discussed with @psychedelicious and @RyanJDick, this is the first
phase of the model manager refactor. In this phase, I've added support
for storing model configuration information the `invokeai.db` SQL3
database. All the code is separate from the original model manager, so
for the time being the frontend is still using the original YAML-based
configuration, so the web app still works.
To keep things clean, I've added a new FastAPI route called
`model_records` which can add, update, retrieve and delete model
records.
The architecture is described in the first section of
`docs/contributing/MODEL_MANAGER.md`.
## QA Instructions, Screenshots, Recordings
There is a pytest for the model sql storage backend in
`tests/backend/model_manager_2/test_model_storage_sql.py`.
To populate `invokeai.db` with models from your current `models.yaml`,
do the following:
1. Stop the running server
2. Back up `invokeai.db`
3. Run `pip install -e .` to install the command used in the next step.
4. Run `invokeai-migrate-models-to-db`
This will iterate through `models.yaml` and create equivalent database
entries in the `model_config` table of `invokeai.db`. Only the models
named in the yaml file will be migrated, so anything that is autoloaded
will be ignored.
Note that in order to get the `model_records` router to be recognized by
the swagger API, I had to rebuild the frontend. Not sure why this was
necessary and would appreciate a pointer on a less radical way to do
this.
## Added/updated tests?
- [X] Yes
- [ ] No
## What type of PR is this? (check all applicable)
- [ ] Refactor
- [ ] Feature
- [X] Bug Fix
- [X] Optimization
- [ ] Documentation Update
- [ ] Community Node Submission
## Have you discussed this change with the InvokeAI team?
- [ ] Yes
- [X] No, because it's required
## Have you updated all relevant documentation?
- [ ] Yes
- [X] No, not necessary
## Description
We use Pytorch ~2.1.0 as a dependency for InvokeAI, but the installer
still installs 2.0.1 first until Invoke AIs dependencies kick in which
causes it to get deleted anyway and replaced with 2.1.0. This is
unnecessary and probably not wanted.
Fixed the dependencies for the installation script to install Pytorch
~2.1.0 to begin with.
P.s. Is there any reason why "torchmetrics==0.11.4" is pinned? What is
the reason for that? Does that change with Pytorch 2.1? It seems to work
since we use it already. It would be nice to know the reason.
Greetings
## 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
- [ ] No : _please replace this line with details on why tests
have not been included_
## [optional] Are there any post deployment tasks we need to perform?
## What type of PR is this? (check all applicable)
- [ ] Refactor
- [ ] Feature
- [ ] Bug Fix
- [x] Optimization
- [ ] Documentation Update
- [ ] Community Node Submission
## Description
Bit of a cleanup.
[chore(ui): delete unused
files](5eaea9dd64)
[feat(ui): add eslint rule
react/jsx-no-bind](3a0ec635c9)
This rule enforces no arrow functions in component props. In practice,
it means all functions passed as component props must be wrapped in
`useCallback()`.
This is a performance optimization to prevent unnecessary rerenders.
The rule is added and all violations have been fixed, whew!
[chore(ui): move useCopyImageToClipboard to
common/hooks/](f2d26a3a3c)
[chore(ui): move MM components & store to
features/](bb52861896)
Somehow they had ended up in `features/ui/tabs` which isn't right
## QA Instructions, Screenshots, Recordings
UI should still work.
It builds successfully, and I tested things out - looks good to me.
Do not use `strict=True` when scaling controlnet conditioning.
When using `guess_mode` (e.g. `more_control` or `more_prompt`), `down_block_res_samples` and `scales` are zipped.
These two objects are of different lengths, so using zip's strict mode raises an error.
In testing, `len(scales) === len(down_block_res_samples) + 1`.
It appears this behaviour is intentional, as the final "extra" item in `scales` is used immediately afterwards.
## What type of PR is this? (check all applicable)
- [ ] Refactor
- [ ] Feature
- [ ] Bug Fix
- [X] Optimization
- [ ] Documentation Update
- [ ] Community Node Submission
## Have you discussed this change with the InvokeAI team?
- [ ] Yes
- [X] No, because: This is just housekeeping
## Have you updated all relevant documentation?
- [ ] Yes
- [X] No, not needed
## Description
Update Accelerate to the most recent version. No breaking changes.
Tested for 1 week in productive use now.
## 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
- [ ] No : _please replace this line with details on why tests
have not been included_
## [optional] Are there any post deployment tasks we need to perform?
## What type of PR is this? (check all applicable)
- [ ] Refactor
- [x] Feature
- [x] Bug Fix
- [x] Optimization
- [ ] Documentation Update
- [ ] Community Node Submission
## Description
This PR introduces [`ruff`](https://github.com/astral-sh/ruff) as the
only linter and formatter needed for the project. It is really fast.
Like, alarmingly fast.
It is a drop-in replacement for flake8, isort, black, and much more.
I've configured it similarly to our existing config.
Note: we had enabled a number of flake8 plugins but didn't have the
packages themselves installed, so they did nothing. Ruff used the
existing config, and found a good number of changes needed to adhere to
those flake8 plugins. I've resolved all violations.
### Code changes
- many
[flake8-comprehensions](https://docs.astral.sh/ruff/rules/#flake8-comprehensions-c4)
violations, almost all auto-fixed
- a good handful of
[flake8-bugbear](https://docs.astral.sh/ruff/rules/#flake8-bugbear-b)
violations
- handful of
[pycodestyle](https://docs.astral.sh/ruff/rules/#pycodestyle-e-w)
violations
- some formatting
### Developer Experience
[Ruff integrates with most
editors](https://docs.astral.sh/ruff/integrations/):
- Official VSCode extension
- `ruff-lsp` python package allows it to integrate with any LSP-capable
editor (vim, emacs, etc)
- Can be configured as an external tool in PyCharm
### Github Actions
I've updated the `style-checks` action to use ruff, and deleted the
`pyflakes` action.
## 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.
-->
- Closes#5066
## QA Instructions, Screenshots, Recordings
Have a poke around, and run the app. There were some logic changes but
it was all pretty straightforward.
~~Not sure how to best test the changed github action.~~ Looks like it
just used the action from this PR, that's kinda unexpected but OK.
<!--
Please provide steps on how to test changes, any hardware or
software specifications as well as any other pertinent information.
-->
## Added/updated tests?
- [ ] Yes
- [ ] No : _please replace this line with details on why tests
have not been included_
## [optional] Are there any post deployment tasks we need to perform?
This rule enforces no arrow functions in component props. In practice, it means all functions passed as component props must be wrapped in `useCallback()`.
This is a performance optimization to prevent unnecessary rerenders.
The rule is added and all violations have been fixed, whew!
* adding VAE recall when using all parameters
* adding VAE to the RecallParameters tab in ImageMetadataActions
* checking for nil vae and casting to null if undefined
* adding default VAE to recall actions list if VAE is nullish
* fix(ui): use `lodash-es` for tree-shakeable imports
---------
Co-authored-by: psychedelicious <4822129+psychedelicious@users.noreply.github.com>
* drop-down for the color picker
* fixed the bug in alpha value
* designing done
---------
Co-authored-by: psychedelicious <4822129+psychedelicious@users.noreply.github.com>
* working
* added selector for method
* refactoring graph
* added ersgan method
* fixing yarn build
* add tooltips
* a conjuction
* rephrase
* removed manual sliders, set HRF to calculate dimensions automatically to match 512^2 pixels
* working
* working
* working
* fixed tooltip
* add hrf to use all parameters
* adding hrf method to parameters
* working on parameter recall
* working on parameter recall
* cleaning
* fix(ui): fix unnecessary casts in addHrfToGraph
* chore(ui): use camelCase in addHrfToGraph
* fix(ui): do not add HRF metadata unless HRF is added to graph
* fix(ui): remove unused imports in addHrfToGraph
* feat(ui): do not hide HRF params when disabled, only disable them
* fix(ui): remove unused vars in addHrfToGraph
* feat(ui): default HRF str to 0.35, method ESRGAN
* fix(ui): use isValidBoolean to check hrfEnabled param
* fix(nodes): update CoreMetadataInvocation fields for HRF
* feat(ui): set hrf strength default to 0.45
* fix(ui): set default hrf strength in configSlice
* feat(ui): use translations for HRF features
---------
Co-authored-by: psychedelicious <4822129+psychedelicious@users.noreply.github.com>
