| Age | Commit message (Collapse) | Author |
|---|
|
|
|
There is no need to re-hash the input image each iteration of the loop.
This also reverts PR #4026 as it was determined the cache hits it avoids
were actually valid.
|
|
|
|
|
|
callbacks
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
VAE Selector
|
|
multiple trivial changes for "extras" models
|
|
Fix error caused by EXIF transpose when using custom scripts
|
|
fix: should invoke callback as well in api only mode
|
|
with hires fix
|
|
* Save image/s before doing highres fix.
|
|
|
|
fix broken progress api and current image compatibility
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Release processing resources after it finishes
|
|
Add mid-kdiffusion cfgdenoiser script callback - access latents, conditionings and sigmas mid-sampling
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
prototype interrupt api
|
|
Extend extras image cache key with upscale_first arg
|
|
Added "--clip-models-path" switch
|
|
Allow TI training using 6GB VRAM when xformers is available
|
|
Add PNG info to pngs only if option is enabled.
|
|
|
|
make save dir if save dir is not exists
|
|
Correct before image saved callback
|
|
add some more classes to safety module for an extension
|
|
Unload sd_model before loading the other to solve the issue #3449
|
|
feat: add app started callback
|
