## 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:
## Description
This change enhances the invocation cache logic to delete cache entries
when the resources to which they refer are deleted.
For example, a cached output may refer to "some_image.png". If that
image is deleted, and this particular cache entry is later retrieved by
a node, that node's successors will receive references to the now
non-existent "some_image.png". When they attempt to use that image, they
will fail.
To resolve this, we need to invalidate the cache when the resources to
which it refers are deleted. Two options:
- Invalidate the whole cache on every image/latents/etc delete
- Selectively invalidate cache entries when their resources are deleted
Node outputs can be any shape, with any number of resource references in
arbitrarily nested pydantic models. Traversing that structure to
identify resources is not trivial.
But invalidating the whole cache is a bit heavy-handed. It would be nice
to be more selective.
Simple solution:
- Invocation outputs' resource references are always string identifiers
- like the image's or latents' name
- Invocation outputs can be stringified, which includes said identifiers
- When the invocation is cached, we store the stringified output
alongside the "live" output classes
- When a resource is deleted, pass its identifier to the cache service,
which can then invalidate any cache entries that refer to it
The images and latents storage services have been outfitted with
`on_deleted()` callbacks, and the cache service registers itself to
handle those events. This logic was copied from `ItemStorageABC`.
`on_changed()` callback are also added to the images and latents
services, though these are not currently used. Just following the
existing pattern.
## 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
Reproduce the issue on main:
- Create a graph in workflow editor with two connected resize nodes
- Add an image to the first
- Enable cache on both
- Run the graph
- Clear Intermediates (in settings)
- Disable cache on the *second* node
- Run the graph, it should fail
Switch to the PR branch and start over, doing the exact same steps. You
shouldn't get any errors.
Example graph to start with:
## Added/updated tests?
- [~] Yes
- [ ] No : _please replace this line with details on why tests
have not been included_
* feat(ui): tweak queue UI components
* fix(ui): manually dispatch queue status query on queue item status change
RTK Query occasionally aborts the query that occurs when the tag is invalidated, especially if multples of them fire in rapid succession.
This resulted in the queue status and progress bar sometimes not reseting when the queue finishes its last item.
Manually dispatch the query now to get around this. Eventually should probably move this to a socket so we don't need to keep responding to socket with HTTP requests. Just send ti directly via socket
* chore(ui): remove errant console.logs
* fix(ui): do not accumulate node outputs in outputs area
* fix(ui): fix merge issue
---------
Co-authored-by: Kent Keirsey <31807370+hipsterusername@users.noreply.github.com>
Add `batch_id` to outbound events. This necessitates adding it to both `InvocationContext` and `InvocationQueueItem`. This allows the canvas to receive images.
When the user enqueues a batch on the canvas, it is expected that all images from that batch are directed to the canvas.
The simplest, most flexible solution is to add the `batch_id` to the invocation context-y stuff. Then everything knows what batch it came from, and we can have the canvas pick up images associated with its list of canvas `batch_id`s.
This change enhances the invocation cache logic to delete cache entries when the resources to which they refer are deleted.
For example, a cached output may refer to "some_image.png". If that image is deleted, and this particular cache entry is later retrieved by a node, that node's successors will receive references to the now non-existent "some_image.png". When they attempt to use that image, they will fail.
To resolve this, we need to invalidate the cache when the resources to which it refers are deleted. Two options:
- Invalidate the whole cache on every image/latents/etc delete
- Selectively invalidate cache entries when their resources are deleted
Node outputs can be any shape, with any number of resource references in arbitrarily nested pydantic models. Traversing that structure to identify resources is not trivial.
But invalidating the whole cache is a bit heavy-handed. It would be nice to be more selective.
Simple solution:
- Invocation outputs' resource references are always string identifiers - like the image's or latents' name
- Invocation outputs can be stringified, which includes said identifiers
- When the invocation is cached, we store the stringified output alongside the "live" output classes
- When a resource is deleted, pass its identifier to the cache service, which can then invalidate any cache entries that refer to it
The images and latents storage services have been outfitted with `on_deleted()` callbacks, and the cache service registers itself to handle those events. This logic was copied from `ItemStorageABC`.
`on_changed()` callback are also added to the images and latents services, though these are not currently used. Just following the existing pattern.
* fix(config): fix typing issues in `config/`
`config/invokeai_config.py`:
- use `Optional` for things that are optional
- fix typing of `ram_cache_size()` and `vram_cache_size()`
- remove unused and incorrectly typed method `autoconvert_path`
- fix types and logic for `parse_args()`, in which `InvokeAIAppConfig.initconf` *must* be a `DictConfig`, but function would allow it to be set as a `ListConfig`, which presumably would cause issues elsewhere
`config/base.py`:
- use `cls` for first arg of class methods
- use `Optional` for things that are optional
- fix minor type issue related to setting of `env_prefix`
- remove unused `add_subparser()` method, which calls `add_parser()` on an `ArgumentParser` (method only available on the `_SubParsersAction` object, which is returned from ArgumentParser.add_subparsers()`)
* feat: queued generation and batches
Due to a very messy branch with broad addition of `isort` on `main` alongside it, some git surgery was needed to get an agreeable git history. This commit represents all of the work on queued generation. See PR for notes.
* chore: flake8, isort, black
* fix(nodes): fix incorrect service stop() method
* fix(nodes): improve names of a few variables
* fix(tests): fix up tests after changes to batches/queue
* feat(tests): add unit tests for session queue helper functions
* feat(ui): dynamic prompts is always enabled
* feat(queue): add queue_status_changed event
* feat(ui): wip queue graphs
* feat(nodes): move cleanup til after invoker startup
* feat(nodes): add cancel_by_batch_ids
* feat(ui): wip batch graphs & UI
* fix(nodes): remove `Batch.batch_id` from required
* fix(ui): cleanup and use fixedCacheKey for all mutations
* fix(ui): remove orphaned nodes from canvas graphs
* fix(nodes): fix cancel_by_batch_ids result count
* fix(ui): only show cancel batch tooltip when batches were canceled
* chore: isort
* fix(api): return `[""]` when dynamic prompts generates no prompts
Just a simple fallback so we always have a prompt.
* feat(ui): dynamicPrompts.combinatorial is always on
There seems to be little purpose in using the combinatorial generation for dynamic prompts. I've disabled it by hiding it from the UI and defaulting combinatorial to true. If we want to enable it again in the future it's straightforward to do so.
* feat: add queue_id & support logic
* feat(ui): fix upscale button
It prepends the upscale operation to queue
* feat(nodes): return queue item when enqueuing a single graph
This facilitates one-off graph async workflows in the client.
* feat(ui): move controlnet autoprocess to queue
* fix(ui): fix non-serializable DOMRect in redux state
* feat(ui): QueueTable performance tweaks
* feat(ui): update queue list
Queue items expand to show the full queue item. Just as JSON for now.
* wip threaded session_processor
* feat(nodes,ui): fully migrate queue to session_processor
* feat(nodes,ui): add processor events
* feat(ui): ui tweaks
* feat(nodes,ui): consolidate events, reduce network requests
* feat(ui): cleanup & abstract queue hooks
* feat(nodes): optimize batch permutation
Use a generator to do only as much work as is needed.
Previously, though we only ended up creating exactly as many queue items as was needed, there was still some intermediary work that calculated *all* permutations. When that number was very high, the system had a very hard time and used a lot of memory.
The logic has been refactored to use a generator. Additionally, the batch validators are optimized to return early and use less memory.
* feat(ui): add seed behaviour parameter
This dynamic prompts parameter allows the seed to be randomized per prompt or per iteration:
- Per iteration: Use the same seed for all prompts in a single dynamic prompt expansion
- Per prompt: Use a different seed for every single prompt
"Per iteration" is appropriate for exploring a the latents space with a stable starting noise, while "Per prompt" provides more variation.
* fix(ui): remove extraneous random seed nodes from linear graphs
* fix(ui): fix controlnet autoprocess not working when queue is running
* feat(queue): add timestamps to queue status updates
Also show execution time in queue list
* feat(queue): change all execution-related events to use the `queue_id` as the room, also include `queue_item_id` in InvocationQueueItem
This allows for much simpler handling of queue items.
* feat(api): deprecate sessions router
* chore(backend): tidy logging in `dependencies.py`
* fix(backend): respect `use_memory_db`
* feat(backend): add `config.log_sql` (enables sql trace logging)
* feat: add invocation cache
Supersedes #4574
The invocation cache provides simple node memoization functionality. Nodes that use the cache are memoized and not re-executed if their inputs haven't changed. Instead, the stored output is returned.
## Results
This feature provides anywhere some significant to massive performance improvement.
The improvement is most marked on large batches of generations where you only change a couple things (e.g. different seed or prompt for each iteration) and low-VRAM systems, where skipping an extraneous model load is a big deal.
## Overview
A new `invocation_cache` service is added to handle the caching. There's not much to it.
All nodes now inherit a boolean `use_cache` field from `BaseInvocation`. This is a node field and not a class attribute, because specific instances of nodes may want to opt in or out of caching.
The recently-added `invoke_internal()` method on `BaseInvocation` is used as an entrypoint for the cache logic.
To create a cache key, the invocation is first serialized using pydantic's provided `json()` method, skipping the unique `id` field. Then python's very fast builtin `hash()` is used to create an integer key. All implementations of `InvocationCacheBase` must provide a class method `create_key()` which accepts an invocation and outputs a string or integer key.
## In-Memory Implementation
An in-memory implementation is provided. In this implementation, the node outputs are stored in memory as python classes. The in-memory cache does not persist application restarts.
Max node cache size is added as `node_cache_size` under the `Generation` config category.
It defaults to 512 - this number is up for discussion, but given that these are relatively lightweight pydantic models, I think it's safe to up this even higher.
Note that the cache isn't storing the big stuff - tensors and images are store on disk, and outputs include only references to them.
## Node Definition
The default for all nodes is to use the cache. The `@invocation` decorator now accepts an optional `use_cache: bool` argument to override the default of `True`.
Non-deterministic nodes, however, should set this to `False`. Currently, all random-stuff nodes, including `dynamic_prompt`, are set to `False`.
The field name `use_cache` is now effectively a reserved field name and possibly a breaking change if any community nodes use this as a field name. In hindsight, all our reserved field names should have been prefixed with underscores or something.
## One Gotcha
Leaf nodes probably want to opt out of the cache, because if they are not cached, their outputs are not saved again.
If you run the same graph multiple times, you only end up with a single image output, because the image storage side-effects are in the `invoke()` method, which is bypassed if we have a cache hit.
## Linear UI
The linear graphs _almost_ just work, but due to the gotcha, we need to be careful about the final image-outputting node. To resolve this, a `SaveImageInvocation` node is added and used in the linear graphs.
This node is similar to `ImagePrimitive`, except it saves a copy of its input image, and has `use_cache` set to `False` by default.
This is now the leaf node in all linear graphs, and is the only node in those graphs with `use_cache == False` _and_ the only node with `is_intermedate == False`.
## Workflow Editor
All nodes now have a footer with a new `Use Cache [ ]` checkbox. It defaults to the value set by the invocation in its python definition, but can be changed by the user.
The workflow/node validation logic has been updated to migrate old workflows to use the new default values for `use_cache`. Users may still want to review the settings that have been chosen. In the event of catastrophic failure when running this migration, the default value of `True` is applied, as this is correct for most nodes.
Users should consider saving their workflows after loading them in and having them updated.
## Future Enhancements - Callback
A future enhancement would be to provide a callback to the `use_cache` flag that would be run as the node is executed to determine, based on its own internal state, if the cache should be used or not.
This would be useful for `DynamicPromptInvocation`, where the deterministic behaviour is determined by the `combinatorial: bool` field.
## Future Enhancements - Persisted Cache
Similar to how the latents storage is backed by disk, the invocation cache could be persisted to the database or disk. We'd need to be very careful about deserializing outputs, but it's perhaps worth exploring in the future.
* fix(ui): fix queue list item width
* feat(nodes): do not send the whole node on every generator progress
* feat(ui): strip out old logic related to sessions
Things like `isProcessing` are no longer relevant with queue. Removed them all & updated everything be appropriate for queue. May be a few little quirks I've missed...
* feat(ui): fix up param collapse labels
* feat(ui): click queue count to go to queue tab
* tidy(queue): update comment, query format
* feat(ui): fix progress bar when canceling
* fix(ui): fix circular dependency
* feat(nodes): bail on node caching logic if `node_cache_size == 0`
* feat(nodes): handle KeyError on node cache pop
* feat(nodes): bypass cache codepath if caches is disabled
more better no do thing
* fix(ui): reset api cache on connect/disconnect
* feat(ui): prevent enqueue when no prompts generated
* feat(ui): add queue controls to workflow editor
* feat(ui): update floating buttons & other incidental UI tweaks
* fix(ui): fix missing/incorrect translation keys
* fix(tests): add config service to mock invocation services
invoking needs access to `node_cache_size` to occur
* optionally remove pause/resume buttons from queue UI
* option to disable prepending
* chore(ui): remove unused file
* feat(queue): remove `order_id` entirely, `item_id` is now an autoinc pk
---------
Co-authored-by: Mary Hipp <maryhipp@Marys-MacBook-Air.local>
## 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
- [ ] No
## Description (edit by @blessedcoolant , @RyanJDick )
This PR adds support for IP-Adapters (a technique for image-based
prompts) in Invoke AI. Currently only available in the Node UI.
IP-Adapter Paper: [IP-Adapter: Text Compatible Image Prompt Adapter for
Text-to-Image Diffusion Models](https://arxiv.org/abs/2308.06721)
IP-Adapter reference code: https://github.com/tencent-ailab/IP-Adapter
On order to test, install the following models via the InvokeAI UI:
Image Encoders:
[InvokeAI/ip_adapter_sd_image_encoder](https://huggingface.co/InvokeAI/ip_adapter_sd_image_encoder)
[InvokeAI/ip_adapter_sdxl_image_encoder](https://huggingface.co/InvokeAI/ip_adapter_sdxl_image_encoder)
IP-Adapters:
[InvokeAI/ip_adapter_sd15](https://huggingface.co/InvokeAI/ip_adapter_sd15)
[InvokeAI/ip_adapter_plus_sd15](https://huggingface.co/InvokeAI/ip_adapter_plus_sd15)
[InvokeAI/ip_adapter_plus_face_sd15](https://huggingface.co/InvokeAI/ip_adapter_plus_face_sd15)
[InvokeAI/ip_adapter_sdxl](https://huggingface.co/InvokeAI/ip_adapter_sdxl)
Old instructions (for reference only):
> In order to test, you need to download and place the following models
in your InvokeAI models directory.
>
> - SD 1.5 - https://huggingface.co/h94/IP-Adapter/tree/main/models -->
Download the models and the `image_encoder` folder to
`models/core/ip_adapters/sd-1`
> - SDXL - https://huggingface.co/h94/IP-Adapter/tree/main/sdxl_models
-Download the models and the `image_encoder` folder to
`models/core/ip_adapaters/sdxl`
>
> This is only temporary. This needs to be handled differently. I
outlined them here.
https://github.com/invoke-ai/InvokeAI/pull/4429#issuecomment-1705776570
## Examples using this PR
### Image variations, no text prompt
Leftmost image in each row is original image used for input to
IP-Adapter. The other rows are example outputs with different seeds,
other parameters identical.
## 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?
