Mary Hipp
a300b6ebdd
feat(ui): graph building for FLUX in linear UI
2024-09-17 15:44:20 +10:00
Mary Hipp
00de20d102
fix(nodes): add flux generation types to choices
2024-09-17 15:44:20 +10:00
Mary Hipp
573c7d2088
feat(ui): FLUX linear - add VAE as required model field rather than allowing default
2024-09-17 15:44:20 +10:00
Mary Hipp
ffbf4aba1f
feat(ui): new fields and param layout for FLUX settings: guidance, t5 encoder, CLIP embed
2024-09-17 15:44:20 +10:00
maryhipp
8916036ed3
fix progress image for FLUX inpainting
2024-09-17 06:41:32 +03:00
psychedelicious
0fd430fc20
fix(nodes): add thresholding to lineart & lineart anime nodes
...
The lineart model often outputs a lot of almost-black noise. SD1.5 ControlNets seem to be OK with this, but SDXL ControlNets are not - they need a cleaner map. 12 was experimentally determined to be a good threshold, eliminating all the noise while keeping the actual edges. Other approaches to thresholding may be better, for example stretching the contrast or removing noise.
I tried:
- Simple thresholding (as implemented here) - works fine.
- Adaptive thresholding - doesn't work, because the thresholding is done in the context of small blocks, while we want thresholding in the context of the whole image.
- Gamma adjustment - alters the white values too much. Hard to tuen.
- Contrast stretching, with and without pre-simple-thresholding - this allows us to treshold out the noise, then stretch everything above the threshold down to almost-zero. So you have a smoother gradient of lightness near zero. It works but it also stretches contrast near white down a bit, which is probably undesired.
In the end, simple thresholding works fine and is very simple.
2024-09-17 04:04:11 +03:00
Mary Hipp
783441a89d
fix(ui): restore Settings Modal functionality to use children to render
2024-09-16 20:09:43 -04:00
psychedelicious
568cae919a
feat(ui): change upscaling icon
2024-09-16 23:58:03 +03:00
Mary Hipp
4bd7b25179
swap control layer and raster layer
2024-09-16 13:39:10 -04:00
Mary Hipp
e019df032e
change upscaling icon
2024-09-16 13:39:10 -04:00
Mary Hipp
5360740613
fix(ui): map order of layer types to the menu options
2024-09-16 13:39:10 -04:00
Ryan Dick
2934e31620
Fix bug when applying multiple LoRA models via apply_lora_sidecar_patches(), and add unit tests for the stacked LoRA case.
2024-09-16 14:48:39 +00:00
Ryan Dick
e88d3cf2f7
Assume alpha=rank for FLUX diffusers PEFT LoRA models.
2024-09-16 13:57:07 +00:00
psychedelicious
1545289a8f
fix(ui): remove accidentally added staging button
2024-09-16 15:27:55 +03:00
psychedelicious
5d3749a424
chore: v5.0.0.a7
2024-09-16 22:21:56 +10:00
psychedelicious
7fe1135ba0
feat(ui): add ref image drop targets to canvas
2024-09-16 15:10:17 +03:00
psychedelicious
41efd813e6
feat(ui): tweak drop overlay styles
2024-09-16 15:10:17 +03:00
psychedelicious
3a42285a3f
feat(ui): gallery image hover button to open in viewer
2024-09-16 15:10:17 +03:00
psychedelicious
04232876e8
feat(ui): tweak perserve mask translation
2024-09-16 15:10:17 +03:00
psychedelicious
97155bcadc
fix(ui): gallery not filling up
...
Floating point precision issue.
2024-09-16 15:10:17 +03:00
psychedelicious
0e0076d819
tidy(ui): rename canvas state "regions" -> "regionalGuidance"
2024-09-16 15:10:17 +03:00
psychedelicious
7725baf6c8
feat(ui): revised regional guidance
2024-09-16 15:10:17 +03:00
psychedelicious
4cf9933487
feat(ui): tighten up ip adapter settings layout
2024-09-16 15:10:17 +03:00
psychedelicious
26e590ed0c
feat(ui): updated add entity buttons/menu
2024-09-16 15:10:17 +03:00
psychedelicious
35012dbee8
fix(ui): pull bbox into ref image translations
2024-09-16 15:10:17 +03:00
psychedelicious
e8c218672f
tidy(ui): clean up addLayerHooks
2024-09-16 15:10:17 +03:00
psychedelicious
e967197fd9
feat(ui): close viewer button is an icon
...
Reduces the awkwardness of having two identical buttons visible at once.
2024-09-16 15:10:17 +03:00
psychedelicious
f76c998791
feat(ui): more translation updates
...
- Migrate some translations to `controlLayers` namespace
- Remove the `controlnet` namespace entirely
- Remove other unused translations
2024-09-16 15:10:17 +03:00
psychedelicious
c2fb821574
feat(ui): update translations for reference image
2024-09-16 15:10:17 +03:00
psychedelicious
0ebe070be6
feat(ui): updated canvas context menu
2024-09-16 15:10:17 +03:00
psychedelicious
26e23a43dc
tidy(ui): create/move store utils to separate file
2024-09-16 15:10:17 +03:00
psychedelicious
52a19ceac3
feat(ui): ip adapter -> reference image (internal state)
...
We are renaming "IP Adapter" to "Reference Image".
- Update all internal state
- Update all components
2024-09-16 15:10:17 +03:00
psychedelicious
2d20e9612c
feat(ui): add sections to add layer menu
2024-09-16 15:10:17 +03:00
psychedelicious
79683479fc
fix(ui): right click opens image in new tab
2024-09-16 15:10:17 +03:00
Ryan Dick
d51f2c5e00
Add bias to LoRA sidecar layer unit tests.
2024-09-15 04:39:56 +03:00
Ryan Dick
78efed4499
Revert change of make all LoRA layers torch.nn.Module's. While the code is uglier, it turns out that the Module implementation of some ops like .to(...) is noticeably slower.
2024-09-15 04:39:56 +03:00
Ryan Dick
b1cf5e989c
Replace 'torch.device("meta")' with 'accelerate.init_empty_weights()' in the hopes that the latter works properly on MacOS.
2024-09-15 04:39:56 +03:00
Ryan Dick
94668247dd
Delete duplicate file that was accidentally kept during rebase.
2024-09-15 04:39:56 +03:00
Ryan Dick
02f27c750a
Add unit tests for LoRAPatcher.apply_lora_sidecar_patches(...) and fixup dtype handling in the sidecar layers.
2024-09-15 04:39:56 +03:00
Ryan Dick
ba3ba3c23a
Add unit tests for LoRALinearSidecarLayer and ConcatenatedLoRALinearSidecarLayer.
2024-09-15 04:39:56 +03:00
Ryan Dick
61d3d566de
Minor cleanup and documentation updates.
2024-09-15 04:39:56 +03:00
Ryan Dick
ae41651346
Remove LoRA conv sidecar layers until they are needed and properly tested.
2024-09-15 04:39:56 +03:00
Ryan Dick
7ce41bf7e0
Fixup unit tests.
2024-09-15 04:39:56 +03:00
Ryan Dick
5bb0c79c14
Add links to test models for loha, lokr, ia3.
2024-09-15 04:39:56 +03:00
Ryan Dick
9438ea608c
Update all lycoris layer types to use the new torch.nn.Module base class.
2024-09-15 04:39:56 +03:00
Ryan Dick
81fbaf2b8b
Assume LoRA alpha=8 for FLUX diffusers PEFT LoRAs.
2024-09-15 04:39:56 +03:00
Ryan Dick
10c3c61cb2
Get diffusers FLUX LoRA working as sidecar patch on quantized model.
2024-09-15 04:39:56 +03:00
Ryan Dick
45bc8fcd7f
WIP - Implement sidecar LoRA layers using functional API.
2024-09-15 04:39:56 +03:00
Ryan Dick
f5f894437c
Bug fixes to get LoRA sidecar patching working for the first time.
2024-09-15 04:39:56 +03:00
Ryan Dick
3e12ac9740
WIP - LoRA sidecar layers.
2024-09-15 04:39:56 +03:00
