Commit Graph

10705 Commits

Author SHA1 Message Date
psychedelicious
8c15d14099 fix: use locale encoding
We have had a few bugs with v4 related to file encodings, especially on Windows.

Windows uses its own character encodings instead of `utf-8`, often `cp1252`. Some characters cannot be decoded using `utf-8`, causing `UnicodeDecodeError`.

There are a couple places where this can cause problems:
- In the installer bootstrap, we install or upgrade `pip` and decode the result, using `subprocess`.

  The input to this includes the user's home dir. In #6105, the user had one of the problematic characters in their username. `subprocess` attempts and fails to decode the username, which crashes the installer.

  To fix this, we need to use `locale.getpreferredencoding()` when executing the command.
- Similarly, in the model install service and config class, we attempt to load a yaml config file. If a problematic character is in the path to the file (which often includes the user's home dir), we can get the same error.

  One example is  #6129 in which the models.yaml migration fails.

  To fix this, we need to open the file with `locale.getpreferredencoding()`.
2024-04-04 15:30:47 +11:00
psychedelicious
38718d8c65 docs: update 020_INSTALL_MANUAL.md, remove conda section 2024-04-04 11:28:09 +11:00
psychedelicious
98ab387e2b docs: update 020_INSTALL_MANUAL.md
Redo the install the package section. It was inaccurate with respect to extra index URLs.
2024-04-04 10:54:23 +11:00
psychedelicious
a0ae2f37d7 docs: update 020_INSTALL_MANUAL.md
Tidy verbiage around the invokeai root and how it is determined
2024-04-04 10:54:23 +11:00
psychedelicious
9c51abb46e fix(config): get root from venv
This logic was a bit wonky. It only selected the `venv` parent if there was already an `invokeai.yaml` file in it. Removed this constraint.
2024-04-04 10:54:23 +11:00
psychedelicious
f887e030bb docs: update 010_INSTALL_AUTOMATED.md
Remove reference to the autodetect GPU device option.
2024-04-04 08:43:17 +11:00
psychedelicious
52b58b4a80 feat(installer): remove extra GPU options
- Remove `CUDA_AND_DML`. This was for onnx, which we have since removed.
- Remove `AUTODETECT`. This option causes problems for windows users, as it falls back on default pypi index resulting in a non-CUDA torch being installed.
- Add more explicit settings for extra index URL, based on the torch website
- Fix bug where `xformers` wasn't installed on linux and/or windows when autodetect was selected
2024-04-04 08:43:17 +11:00
psychedelicious
9fdfd4267c fix(ui): fix model name overflow
Closes #3897
2024-04-04 08:03:30 +11:00
psychedelicious
c4a6d3ddc0 docs: update FAQ for missing models solution
This will be fairly common in v4 updates. The root cause is models not being added to the `models.yaml` file in v3, so we don't correctly migrate the models to the db.

The docs describe how to use `Scan Folder` to restore missing models.
2024-04-04 07:58:11 +11:00
psychedelicious
25bbaa73b9 feat(ui): add inplace option to scan folder install ui 2024-04-04 07:58:11 +11:00
psychedelicious
2383fb93c7 fix(ui): show model install progress as 100 if finished 2024-04-04 07:58:11 +11:00
psychedelicious
63c60e6d63 feat(ui): refresh model scan results on completed model install 2024-04-04 07:58:11 +11:00
psychedelicious
3a10062b53 feat(mm): more reliable mm scan folder
Compare the installed paths to determine if the model is already installed. Fixes an issue where installed models showed up as uninstalled or vice-versa. Related to relative vs absolute path handling.
2024-04-04 07:58:11 +11:00
brandonrising
51ca59c088 Update probe to always use cpu for loading models 2024-04-04 07:34:43 +11:00
psychedelicious
216b34ac44 tests: update model rename test 2024-04-04 07:17:38 +11:00
psychedelicious
7ff2371c07 fix(mm): do not rename model file if model record is renamed
Renaming the model file to the model name introduces unnecessary contraints on model names.

For example, a model name can technically be any length, but a model _filename_ cannot be too long.

There are also constraints on valid characters for filenames which shouldn't be applied to model record names.

I believe the old behaviour is a holdover from the old system.
2024-04-04 07:17:38 +11:00
Mary Hipp
4927d1b7c9 add some test IDs for accordion targeting 2024-04-04 06:35:11 +11:00
psychedelicious
85f53f94f8 feat(mm): include needed vs free in OOM
Gives us a bit more visibility into these errors, which seem to be popping up more frequently with the new MM.
2024-04-04 06:26:15 +11:00
blessedcoolant
7da04b8333
IP-Adapter Safetensor Support (#6041)
## Summary

This PR adds support for IP Adapter safetensor files for direct usage
inside InvokeAI.

# TEST

You can download the [Composition
Adapters](https://huggingface.co/ostris/ip-composition-adapter) which
weren't previously supported in Invoke and try them out. Every other IP
Adapter model should work too.

If you pick a Safetensor IP Adapter model, you will also need to set
ViT-H or ViT-G next to it. This is a raw implementation. Can refine it
further based on feedback.

Prompt: `Spiderman holding a bunny` -- Exact same composition as the
adapter image.

![opera_UHlo1IyXPT](https://github.com/invoke-ai/InvokeAI/assets/54517381/00bf9f0b-149f-478d-87ca-3252b68d1054)
2024-04-03 22:46:45 +05:30
blessedcoolant
be574cb764 fix: incorrect suffix check in ip adapter checkpoint file 2024-04-03 22:38:28 +05:30
blessedcoolant
5f01de1993 chore: ruff and lint fixes 2024-04-03 20:41:51 +05:30
blessedcoolant
cf88bd3294 Merge branch 'main' into checkpoint-ip-adapter 2024-04-03 20:30:02 +05:30
blessedcoolant
e574815413 chore: clean up merge conflicts 2024-04-03 20:28:00 +05:30
blessedcoolant
fb293dcd84 Merge branch 'checkpoint-ip-adapter' of https://github.com/blessedcoolant/InvokeAI into checkpoint-ip-adapter 2024-04-03 20:23:07 +05:30
blessedcoolant
414851f2f0 fix: raise and present the runtime error from the exception 2024-04-03 20:21:50 +05:30
blessedcoolant
2dcbb7223b fix: use Path for ip_adapter_ckpt_path instead of str 2024-04-03 20:21:03 +05:30
psychedelicious
132aadca15 fix(ui): cancel batch status button greyed out
Closes #6110
2024-04-03 08:23:31 -04:00
blessedcoolant
14a9f74b17 cleanup: use load_file of safetensors directly for loading ip adapters 2024-04-03 12:40:13 +05:30
blessedcoolant
1372ef15b3 fix: Fail when unexpected keys are found in IP Adapter models 2024-04-03 12:40:11 +05:30
blessedcoolant
dc1681a0de fix: clip vision model auto param
Setting to 'auto' works only for InvokeAI config and auto detects the SD model but will override if user explicitly sets it. If auto used with checkpoint models, we raise an error. Checkpoints will always need to set to non-auto.
2024-04-03 12:40:11 +05:30
blessedcoolant
be1212de9a fix: Raise a better error when incorrect CLIP Vision model is used 2024-04-03 12:40:10 +05:30
blessedcoolant
a14ce0edab chore: rename IPAdapterDiffusersConfig to IPAdapterInvokeAIConfig 2024-04-03 12:40:10 +05:30
blessedcoolant
4a0dfc3b2d ui: improve the clip vision model picker layout 2024-04-03 12:40:08 +05:30
blessedcoolant
91a70c8d07 feat: Let users pick CLIP Vision model for Checkpoint IP Adapters 2024-04-03 12:40:05 +05:30
blessedcoolant
936b99bd3c chore: improve types in ip_adapter backend file 2024-04-03 12:40:02 +05:30
blessedcoolant
9ff729a7e6 fix: Update ModelView to accommodate for the new config changes to IP Adapter 2024-04-03 12:40:01 +05:30
blessedcoolant
5829b87b8d ui: update the new ip adapter configs on the frontend 2024-04-03 12:40:01 +05:30
blessedcoolant
79f7b61dfe fix: cleanup across various ip adapter files 2024-04-03 12:39:52 +05:30
blessedcoolant
b1c8266e22 feat: add base model recognition for ip adapter safetensor files 2024-04-03 12:39:52 +05:30
blessedcoolant
67afb1763e wip: Initial implementation of safetensor support for IP Adapter 2024-04-03 12:39:52 +05:30
blessedcoolant
8584171a49
docs: fix broken link (#6116)
## Summary

Fix a broken link

## Related Issues / Discussions


https://discord.com/channels/1020123559063990373/1049495067846524939/1224970148058763376

## QA Instructions

n/a

## Merge Plan

n/a

## Checklist

- [x] _The PR has a short but descriptive title, suitable for a
changelog_
- [ ] _Tests added / updated (if applicable)_ n/a
- [x] _Documentation added / updated (if applicable)_
2024-04-03 12:35:17 +05:30
psychedelicious
50951439bd docs: fix broken link 2024-04-03 17:36:15 +11:00
psychedelicious
7b93b554d7 fix(ui): add default coherence mode to generation slice migration
The valid values for this parameter changed when inpainting changed to gradient denoise. The generation slice's redux migration wasn't updated, resulting in a generation error until you change the setting or reset web UI.
2024-04-03 08:46:31 +11:00
brandonrising
21b9e96a45 Run typegen, bump version 2024-04-02 10:14:39 -04:00
psychedelicious
b6ad33ac1a perf(ui): reduce canvas max history to 100
This should further insulate canvas from excessive GCs.
2024-04-02 08:48:18 -04:00
psychedelicious
69ec14c7bb perf(ui): use rfdc for deep copying of objects
- Add and use more performant `deepClone` method for deep copying throughout the UI.

Benchmarks indicate the Really Fast Deep Clone library (`rfdc`) is the best all-around way to deep-clone large objects.

This is particularly relevant in canvas. When drawing or otherwise manipulating canvas objects, we need to do a lot of deep cloning of the canvas layer state objects.

Previously, we were using lodash's `cloneDeep`.

I did some fairly realistic benchmarks with a handful of deep-cloning algorithms/libraries (including the native `structuredClone`). I used a snapshot of the canvas state as the data to be copied:

On Chromium, `rfdc` is by far the fastest, over an order of magnitude faster than `cloneDeep`.

On FF, `fastest-json-copy` and `recursiveDeepCopy` are even faster, but are rather limited in data types. `rfdc`, while only half as fast as the former 2, is still nearly an order of magnitude faster than `cloneDeep`.

On Safari, `structuredClone` is the fastest, about 2x as fast as `cloneDeep`. `rfdc` is only 30% faster than `cloneDeep`.

`rfdc`'s peak memory usage is about 10% more than `cloneDeep` on Chrome. I couldn't get memory measurements from FF and Safari, but let's just assume the memory usage is similar relative to the other algos.

Overall, `rfdc` is the best choice for a single algo for all browsers. It's definitely the best for Chromium, by far the most popular desktop browser and thus our primary target.

A future enhancement might be to detect the browser and use that to determine which algorithm to use.
2024-04-02 08:48:18 -04:00
psychedelicious
a6c91979af fix(ui): prevent canvas history leak
There were two ways the canvas history could grow too large (past the `MAX_HISTORY` setting):
- Sometimes, when pushing to history, we didn't `shift` an item out when we exceeded the max history size.
- If the max history size was exceeded by more than one item, we still only `shift`, which removes one item.

These issue could appear after an extended canvas session, resulting in a memory leak and recurring major GCs/browser performance issues.

To fix these issues, a helper function is added for both past and future layer states, which uses slicing to ensure history never grows too large.
2024-04-02 08:48:18 -04:00
psychedelicious
e655399324 fix(config): handle windows paths in invokeai.yaml migration for legacy_conf_dir
The logic incorrectly set the `legacy_conf_dir` on windows, where the slashes go the other direction. Handle this case and update tests to catch it.
2024-04-02 08:06:59 -04:00
psychedelicious
f75de8a35c feat(db): add migration 9 - empty session queue
Empties the session queue. This is done to prevent any lingering session queue items from causing pydantic errors due to changed schemas.
2024-04-02 13:25:14 +11:00
psychedelicious
d4be945dde fix(nodes): gracefully handle custom nodes init error
Previously, exceptions raised as custom nodes are initialized were fatal errors, causing the app to exit.

With this change, any error on import is caught and the error message printed. App continues to start up without the node.

For example, a custom node that isn't updated for v4.0.0 may raise an error on import if it is attempting to import things that no longer exist.
2024-04-02 13:25:14 +11:00