Lincoln Stein
7159ec885f
further improvements to preload_models.py
...
- Faster startup for command line switch processing
- Specify configuration file to modify using --config option:
./scripts/preload_models.ply --config models/my-models-file.yaml
2022-10-31 11:33:05 -04:00
Lincoln Stein
b5cf734ba9
improve behavior of preload_models.py
...
- NEVER overwrite user's existing models.yaml
- Instead, merge its contents into new config file,
and rename original to models.yaml.orig (with
message)
- models.yaml has been removed from repository and renamed
models.yaml.example
2022-10-31 11:08:19 -04:00
Lincoln Stein
f7dc8eafee
restore models.yaml to virgin state
2022-10-31 10:47:35 -04:00
Lincoln Stein
89da42ad79
Merge branch 'pin-options-panel' of https://github.com/psychedelicious/stable-diffusion into psychedelicious-pin-options-panel
...
- from PR #1301
2022-10-31 09:37:13 -04:00
Lincoln Stein
e8aba99c92
Merge branch 'development' of github.com:invoke-ai/InvokeAI into development
2022-10-31 09:35:21 -04:00
Damian at mba
ced9c83e96
various prompting fixes
2022-10-31 09:34:56 -04:00
psychedelicious
247816db9a
Adds @psychedelicious to contributors list
2022-10-31 09:34:20 -04:00
Lincoln Stein
80f2cfe3e3
set default max_models to 2 internally as well as as arg
2022-10-31 09:05:38 -04:00
blessedcoolant
9a15a89e20
Tabs Styling Fix
2022-10-31 23:58:08 +11:00
blessedcoolant
c73a61b785
Add Invoke AI Logo to unpinned Options
2022-10-31 23:58:08 +11:00
psychedelicious
88203d8db2
Styling fixes
2022-10-31 23:58:08 +11:00
psychedelicious
881c69e905
Fixes invisible image uploader button
2022-10-31 23:58:08 +11:00
psychedelicious
c40278dae7
Fixes re-renders triggered by typing prompt
2022-10-31 23:58:08 +11:00
psychedelicious
7b329b7c91
Fixes: Progress bar does not activate when changing model
2022-10-31 23:58:08 +11:00
psychedelicious
c19b02ab21
Fixes disappearing cursor edge case
2022-10-31 23:58:08 +11:00
psychedelicious
6ebddf09c2
Fixes: cancel button disabled after model change
2022-10-31 23:58:08 +11:00
psychedelicious
5841e1b5be
Fixes safari process buttons style bug
2022-10-31 23:58:08 +11:00
psychedelicious
5f09ffa276
Fixes error when inpainting with show progress images enabled
2022-10-31 23:58:02 +11:00
psychedelicious
9e70c216f6
Sets defaults to show inpaint box and fill
2022-10-31 23:57:40 +11:00
psychedelicious
cbe8a9550c
Updates styling
2022-10-31 23:57:40 +11:00
psychedelicious
259ecb7b71
Fixes very slightly incorrect pixel offset
2022-10-31 23:57:40 +11:00
psychedelicious
002791ef68
Removes unused file
2022-10-31 23:57:40 +11:00
psychedelicious
21e491f878
Fixes empty canvas detection
2022-10-31 23:57:40 +11:00
psychedelicious
12c4c715aa
Demotes inpainting to img2img when mask is empty
2022-10-31 23:57:40 +11:00
psychedelicious
fe700d27df
Improves styling
2022-10-31 23:57:33 +11:00
psychedelicious
7a4ceb0f7c
Fixes empty canvas detection
2022-10-31 23:57:19 +11:00
blessedcoolant
bb5d77a9fb
Testing Build
2022-10-31 23:57:19 +11:00
blessedcoolant
3c55baf06b
Readded Bounding Box Visibility Toggle
...
Only affects preview. The backend still takes the set bounding box.
2022-10-31 23:57:19 +11:00
blessedcoolant
ca882ad5ff
Reworked Invoke Icon Layout when unpinned
2022-10-31 23:57:19 +11:00
blessedcoolant
6a7b4ef63f
Styling Fix
2022-10-31 23:57:19 +11:00
blessedcoolant
f60d22b29b
Fix an issue with the OutsideWatcher
...
The OutsideWatcher was disabling hotkeys because it was always being active -- whether the object was pinned or not. Modified the hook to now take a new optional argument called "req" which is a boolean that indicates whether to trigger it or not.
We can pass this from the component to control when the outside watcher should work and when it shouldn't.
2022-10-31 23:57:19 +11:00
blessedcoolant
6a6fbe24a3
Styling & Hotkeys Update
2022-10-31 23:57:19 +11:00
psychedelicious
5efd2ed7a8
Demotes inpainting to img2img when mask is empty
2022-10-31 23:56:49 +11:00
psychedelicious
62c346850c
Improves styling
2022-10-31 23:56:49 +11:00
psychedelicious
f6fafe3eb3
Improves bounding box fit behaviour when changing inpaint image
2022-10-31 23:56:49 +11:00
psychedelicious
6547c320a9
Adds "loopback" feature
2022-10-31 23:56:49 +11:00
psychedelicious
2d32cf4eeb
Styling improvements
2022-10-31 23:56:49 +11:00
psychedelicious
7a4e358d53
Fixes: unable to postprocess uploaded image "metadata" error
2022-10-31 23:56:49 +11:00
psychedelicious
ac1469bbd3
Fixes: inpainting bug "images do not match"
2022-10-31 23:56:49 +11:00
psychedelicious
c0c32d9daa
Fixes bug with bounding box transforming cursor
2022-10-31 23:56:49 +11:00
psychedelicious
52e74fef7c
Fixes missing gallery category on generated images
2022-10-31 23:56:49 +11:00
psychedelicious
e431d296c0
Builds fresh bundle + tidy
2022-10-31 23:56:49 +11:00
psychedelicious
1e7a5fda24
Fixes #1295
2022-10-31 23:56:49 +11:00
psychedelicious
050d72478e
Attempts to fix #1297
2022-10-31 23:56:49 +11:00
psychedelicious
d3a09f1284
Updates styles
2022-10-31 23:56:49 +11:00
psychedelicious
e096eef049
Fixes next/prev image not working if category doesn't match
2022-10-31 23:56:49 +11:00
psychedelicious
62c97dd7e6
Fixes edge cases, adds invoke button to header when options floating
2022-10-31 23:56:49 +11:00
psychedelicious
e58b7a7ef9
Adds pin feature to options panel
2022-10-31 23:56:49 +11:00
Lincoln Stein
dc556cb1a7
add max_load_models parameter for model cache control
...
- ldm.generate.Generator() now takes an argument named `max_load_models`.
This is an integer that limits the model cache size. When the cache
reaches the limit, it will start purging older models from cache.
- CLI takes an argument --max_load_models, default to 2. This will keep
one model in GPU and the other in CPU and switch back and forth
quickly.
- To not cache models at all, pass --max_load_models=1
2022-10-31 08:55:53 -04:00
Lincoln Stein
0c8f0e3386
add max_load_models parameter for model cache control
...
- ldm.generate.Generator() now takes an argument named `max_load_models`.
This is an integer that limits the model cache size. When the cache
reaches the limit, it will start purging older models from cache.
- CLI takes an argument --max_load_models, default to 2. This will keep
one model in GPU and the other in CPU and switch back and forth
quickly.
- To not cache models at all, pass --max_load_models=1
2022-10-31 08:53:16 -04:00
