- Replace legacy model manager service with the v2 manager.
- Update invocations to use new load interface.
- Fixed many but not all type checking errors in the invocations. Most
were unrelated to model manager
- Updated routes. All the new routes live under the route tag
`model_manager_v2`. To avoid confusion with the old routes,
they have the URL prefix `/api/v2/models`. The old routes
have been de-registered.
- Added a pytest for the loader.
- Updated documentation in contributing/MODEL_MANAGER.md
- Implement new model loader and modify invocations and embeddings
- Finish implementation loaders for all models currently supported by
InvokeAI.
- Move lora, textual_inversion, and model patching support into
backend/embeddings.
- Restore support for model cache statistics collection (a little ugly,
needs work).
- Fixed up invocations that load and patch models.
- Move seamless and silencewarnings utils into better location
- Cache stat collection enabled.
- Implemented ONNX loading.
- Add ability to specify the repo version variant in installer CLI.
- If caller asks for a repo version that doesn't exist, will fall back
to empty version rather than raising an error.
Unfortunately you cannot test for both a specific type of error and match its message. Splitting the error classes makes it easier to test expected error conditions.
The changes aim to deduplicate data between workflows and node templates, decoupling workflows from internal implementation details. A good amount of data that was needlessly duplicated from the node template to the workflow is removed.
These changes substantially reduce the file size of workflows (and therefore the images with embedded workflows):
- Default T2I SD1.5 workflow JSON is reduced from 23.7kb (798 lines) to 10.9kb (407 lines).
- Default tiled upscale workflow JSON is reduced from 102.7kb (3341 lines) to 51.9kb (1774 lines).
The trade-off is that we need to reference node templates to get things like the field type and other things. In practice, this is a non-issue, because we need a node template to do anything with a node anyways.
- Field types are not included in the workflow. They are always pulled from the node templates.
The field type is now properly an internal implementation detail and we can change it as needed. Previously this would require a migration for the workflow itself. With the v3 schema, the structure of a field type is an internal implementation detail that we are free to change as we see fit.
- Workflow nodes no long have an `outputs` property and there is no longer such a thing as a `FieldOutputInstance`. These are only on the templates.
These were never referenced at a time when we didn't also have the templates available, and there'd be no reason to do so.
- Node width and height are no longer stored in the node.
These weren't used. Also, per https://reactflow.dev/api-reference/types/node, we shouldn't be programmatically changing these properties. A future enhancement can properly add node resizing.
- `nodeTemplates` slice is merged back into `nodesSlice` as `nodes.templates`. Turns out it's just a hassle having these separate in separate slices.
- Workflow migration logic updated to support the new schema. V1 workflows migrate all the way to v3 now.
- Changes throughout the nodes code to accommodate the above changes.
We have two different classes named `ModelInfo` which might need to be used by API consumers. We need to export both but have to deal with this naming collision.
The `ModelInfo` I've renamed here is the one that is returned when a model is loaded. It's the object least likely to be used by API consumers.
Replace `delete_on_startup: bool` & associated logic with `ephemeral: bool` and `TemporaryDirectory`.
The temp dir is created inside of `output_dir`. For example, if `output_dir` is `invokeai/outputs/tensors/`, then the temp dir might be `invokeai/outputs/tensors/tmpvj35ht7b/`.
The temp dir is cleaned up when the service is stopped, or when it is GC'd if not properly stopped.
In the event of a catastrophic crash where the temp files are not cleaned up, the user can delete the tempdir themselves.
This situation may not occur in normal use, but if you kill the process, python cannot clean up the temp dir itself. This includes running the app in a debugger and killing the debugger process - something I do relatively often.
Tests updated.
- The default is to not delete on startup - feels safer.
- The two services using this class _do_ delete on startup.
- The class has "ephemeral" removed from its name.
- Tests & app updated for this change.
`_delete_all` logged how many items it deleted, and had to be called _after_ service start bc it needed access to logger.
Move the logger call to the startup method and return the the deleted stats from `_delete_all`. This lets `_delete_all` be called at any time.
Turns out they are just different enough in purpose that the implementations would be rather unintuitive. I've made a separate ObjectSerializer service to handle tensors and conditioning.
Refined the class a bit too.
Turns out `ItemStorageABC` was almost identical to `PickleStorageBase`. Instead of maintaining separate classes, we can use `ItemStorageABC` for both.
There's only one change needed - the `ItemStorageABC.set` method must return the newly stored item's ID. This allows us to let the service handle the responsibility of naming the item, but still create the requisite output objects during node execution.
The naming implementation is improved here. It extracts the name of the generic and appends a UUID to that string when saving items.
- New generic class `PickleStorageBase`, implements the same API as `LatentsStorageBase`, use for storing non-serializable data via pickling
- Implementation `PickleStorageTorch` uses `torch.save` and `torch.load`, same as `LatentsStorageDisk`
- Add `tensors: PickleStorageBase[torch.Tensor]` to `InvocationServices`
- Add `conditioning: PickleStorageBase[ConditioningFieldData]` to `InvocationServices`
- Remove `latents` service and all `LatentsStorage` classes
- Update `InvocationContext` and all usage of old `latents` service to use the new services/context wrapper methods
