## 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?
- [X ] Yes
- [ ] No, because:
## Have you updated all relevant documentation?
- [ ] Yes
- [ X] No - not yet WIP
## Description
This PR adds support for loading and converting checkpoint-format
ControlNet and SDXL models. The SDXL and SDXL-refiner model conversions
are working; however saving the unet in safetensors format leads to
corrupted model files, so currently is saving in .bin format (after
scanning the input model).
ControlNet conversion seems to be working but needs further testing.
To use this PR, you will need to copy the files
`invokeai/configs/stable-diffusion/sd_xl_base.yaml` and
`invokeai/configs/stable-diffusion/sd_xl_refiner.yaml` into
`INVOKEAI/configs/stable-diffusion`. You will also need to run
`invokeai-configure --yes --skip-sd` in order to install additional core
model files needed by the converter.
## What type of PR is this? (check all applicable)
- [x] Feature
## Have you discussed this change with the InvokeAI team?
- [x] Yes
## Description
- Update the Aspect Ratio tags to show the aspect ratio values rather
than Wide / Square and etc.
- Updated Lora Input to take values between -50 and 50 coz I found some
LoRA that are actually trained to work until -25 and +15 too. So these
input caps should mostly suffice. If there's ever a LoRA that goes
bonkers on that, we can change it.
- Fixed LoRA's being sorted the wrong way in Lora Select.
- Fixed Embeddings being sorted the wrong way in Embedding Select.
## 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?
- add `addNSFWCheckerToGraph` and `addWatermarkerToGraph` functions
- use them in all linear graph creation
- add state & toggles to settings modal to enable these
- trigger queries for app config on socket connect
- disable the nsfw/watermark booleans if we get the app config and they are not available
## What type of PR is this? (check all applicable)
- [x] Feature
## Have you discussed this change with the InvokeAI team?
- [x] Yes
## Description
This PR adds support for SDXL Models in the Linear UI
### DONE
- SDXL Base Text To Image Support
- SDXL Base Image To Image Support
- SDXL Refiner Support
- SDXL Relevant UI
## [optional] Are there any post deployment tasks we need to perform?
Double check to ensure nothing major changed with 1.0 -- In any case
those changes would be backend related mostly. If Refiner is scrapped
for 1.0 models, then we simply disable the Refiner Graph.
Rolled back the earlier split of the refiner model query.
Now, when you use `useGetMainModelsQuery()`, you must provide it an array of base model types.
They are provided as constants for simplicity:
- ALL_BASE_MODELS
- NON_REFINER_BASE_MODELS
- REFINER_BASE_MODELS
Opted to just use args for the hook instead of wrapping the hook in another hook, we can tidy this up later if desired.
We can derive `isRefinerAvailable` from the query result (eg are there any refiner models installed). This is a piece of server state, so by using the list models response directly, we can avoid needing to manually keep the client in sync with the server.
Created a `useIsRefinerAvailable()` hook to return this boolean wherever it is needed.
Also updated the main models & refiner models endpoints to only return the appropriate models. Now we don't need to filter the data on these endpoints.
