Commit Graph

6769 Commits

Author SHA1 Message Date
8b39b67ec7 Merge branch 'main' into feat/select-vram-in-config 2023-08-09 12:17:27 +10:00
a933977861 Pick correct config file for sdxl models (#4191)
## What type of PR is this? (check all applicable)

- [X] Bug Fix

## Have you discussed this change with the InvokeAI team?
- [X] Yes
- [ ] No, because:

      
## Have you updated all relevant documentation?
- [X Yes
- [ ] No


## Description

If `models.yaml` is cleared out for some reason, the model manager will
repopulate it by scanning `models`. However, this would fail with a
pydantic validation error if any SDXL checkpoint models were present
because the lack of logic to pick the correct configuration file. This
has now been added.
2023-08-09 11:16:48 +10:00
dfb41d8461 Merge branch 'main' into bugfix/autodetect-sdxl-ckpt-config 2023-08-09 03:57:44 +03:00
4d5169e16d Merge branch 'main' into feat/select-vram-in-config 2023-08-08 13:50:02 -04:00
f56f19710d allow user to interactively resize screen before UI runs 2023-08-08 12:27:25 -04:00
e77400ab62 remove deprecated options from config 2023-08-08 08:33:30 -07:00
13347f6aec blackified 2023-08-08 08:33:30 -07:00
a9bf387e5e turned on Pydantic validate_assignment 2023-08-08 08:33:30 -07:00
8258c87a9f refrain from writing deprecated legacy options to invokeai.yaml 2023-08-08 08:33:30 -07:00
1b1b399fd0 Fix crash when attempting to update a model (#4192)
## What type of PR is this? (check all applicable)

- [X] Bug Fix


## Have you discussed this change with the InvokeAI team?
- [X No, because small fix

      
## Have you updated all relevant documentation?
- [X] Yes

## Description

A logic bug was introduced in PR #4109 that caused Web-based model
updates to fail with a pydantic validation error. This corrects the
problem.

## Related Tickets & Documents

PR #4109
2023-08-08 10:54:27 -04:00
6ed7ba57dd Merge branch 'main' into bugfix/fix-model-updates 2023-08-08 09:05:25 -04:00
2b3b77a276 api(images): allow HEAD request on image/full (#4193) 2023-08-08 00:08:48 -07:00
8b8ec68b30 Merge branch 'main' into feat/image_http_head 2023-08-08 00:02:48 -07:00
e20af5aef0 feat(ui): add LoRA support to SDXL linear UI
new graph modifier `addSDXLLoRasToGraph()` handles adding LoRA to the SDXL t2i and i2i graphs.
2023-08-08 15:02:00 +10:00
57e8ec9488 chore(ui): lint/format 2023-08-08 12:53:47 +10:00
734a9e4271 invalidate board total when images deleted, only run date range logic if board has less than 20 images 2023-08-08 12:53:47 +10:00
fe924daee3 add option to disable multiselect 2023-08-08 12:53:47 +10:00
809705c30d api(images): allow HEAD request on image/full 2023-08-07 15:11:47 -07:00
4367061b19 fix(ModelManager): fix overridden VAE with relative path (#4059) 2023-08-07 12:57:32 -07:00
0fd13d3604 Merge branch 'main' into feat/select-vram-in-config 2023-08-07 15:51:59 -04:00
72a3e776b2 fix logic error introduced in PR 4109 2023-08-07 15:38:22 -04:00
af044007d5 pick correct config file for sdxl models 2023-08-07 15:19:49 -04:00
f272a44feb Merge branch 'main' into refactor/model_manager_instantiate 2023-08-07 10:59:28 -07:00
ae17d01e1d Fix hue adjustment (#4182)
* Fix hue adjustment

Hue adjustment wasn't working correctly because color channels got swapped. This has now been fixed and we're using PIL rather than cv2 to do the RGBA->HSV->RGBA conversion. The range of hue adjustment is also the more typical 0..360 degrees.
2023-08-06 23:23:51 +00:00
5bfd6cb66f Merge remote-tracking branch 'origin/main' into refactor/model_manager_instantiate
# Conflicts:
#	invokeai/backend/model_management/model_manager.py
2023-08-05 22:02:28 -07:00
d09dfc3e9b fix(api): use db_location instead of db_path_string
This may just be the SQLite memory sentinel value.
2023-08-06 14:09:04 +10:00
66f524cae7 fix(mm): fix a lot of typing issues
Most fixes are just things being typed as `str` but having default values of `None`, but there are some minor logic changes.
2023-08-06 14:09:04 +10:00
9ba50130a1 fix(api): fix db location types
The services all want strings instead of `Path`s; create variable for the string representation of the path provided by the config services.
2023-08-06 14:09:04 +10:00
d4cf2d2666 fix(api): fix ApiDependencies.invoker types
ApiDependencies.invoker` provides typing for the API's services layer. Marking it `Optional` results in all the routes seeing it as optional, which is not good.

Instead of marking it optional to satisfy the initial assignment to `None`, we can just skip the initial assignment. This preserves the IDE hinting in API layer and is types-legal.
2023-08-06 14:09:04 +10:00
b8b589c150 fix(nodes): fix hsl nodes rebase conflict 2023-08-06 09:57:49 +10:00
d93900a8de Added HSL Nodes 2023-08-06 09:57:49 +10:00
7f4c387080 test(model_management): factor out name strings 2023-08-05 15:46:46 -07:00
80876bbbd1 Merge remote-tracking branch 'origin/refactor/model_manager_instantiate' into refactor/model_manager_instantiate 2023-08-05 15:25:05 -07:00
7a4ff4c089 Merge branch 'main' into refactor/model_manager_instantiate 2023-08-05 15:23:38 -07:00
44bf308192 test(model_management): add a couple tests for _get_model_path 2023-08-05 15:22:23 -07:00
dc96a3e79d Fix random number generator
Passing in seed=0 is not equivalent to seed=None. The latter will get a new seed from entropy in the OS, and that's what we should be using.
2023-08-06 00:29:08 +10:00
c076f1397e rebuild frontend 2023-08-05 14:40:42 +10:00
2568aafc0b bump version number so that pip updates work 2023-08-05 14:40:42 +10:00
65ed224bfc Merge branch 'main' into refactor/model_manager_instantiate 2023-08-04 21:34:38 -07:00
b6e369c745 chore: black 2023-08-05 12:28:35 +10:00
ecabfc252b devices.py - Update MPS FP16 check to account for upcoming MacOS Sonoma
float16 doesn't seem to work on MacOS Sonoma due to further changes with Metal. This'll default back to float32 for Sonoma users.
2023-08-05 12:28:35 +10:00
da96a41103 Merge branch 'main' into feat/select-vram-in-config 2023-08-05 12:11:50 +10:00
d162b78767 fix broken civitai example link 2023-08-05 12:10:52 +10:00
eb6c317f04 chore: black 2023-08-05 12:05:24 +10:00
6d7223238f fix: fix typo in message 2023-08-05 12:05:24 +10:00
8607d124c5 improve message about the consequences of the --ignore_missing_core_models flag 2023-08-05 12:05:24 +10:00
23497bf759 add --ignore_missing_core_models CLI flag to bypass checking for missing core models 2023-08-05 12:05:24 +10:00
b10cf20eb1 Merge branch 'main' into refactor/model_manager_instantiate
# Conflicts:
#	invokeai/backend/model_management/model_manager.py
2023-08-04 18:28:18 -07:00
3d93851dba Installer should download fp16 models if user has specified 'auto' in config (#4129)
## What type of PR is this? (check all applicable)

- [ ] Refactor
- [ ] Feature
- [X] 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?
- [X] Yes
- [ ] No


## Description

At install time, when the user's config specified "auto" precision, the
installer was downloading the fp32 models even when an fp16 model would
be appropriate for the OS.


## 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 #4127
2023-08-05 01:56:25 +03:00
9bacd77a79 Merge branch 'main' into bugfix/fp16-models 2023-08-05 01:42:43 +03:00