aboutsummaryrefslogtreecommitdiff
path: root/modules
AgeCommit message (Collapse)Author
2023-01-25fix for unet hijack breaking the train tabAUTOMATIC
2023-01-25make clicking extra networks button one more time close the extra networks UIAUTOMATIC
2023-01-25Merge pull request #6510 from brkirch/unet16-upcast-precisionAUTOMATIC1111
Add upcast options, full precision sampling from float16 UNet and upcasting attention for inference using SD 2.1 models without --no-half
2023-01-25change to code for live preview fix on OSX to be bit more obviousAUTOMATIC
2023-01-25Merge pull request #7151 from brkirch/fix-approx-nnAUTOMATIC1111
Fix Approx NN previews changing first generation result
2023-01-25Add instruct-pix2pix hijackKyle
Allows loading instruct-pix2pix models via same method as inpainting models in sd_models.py and sd_hijack_ip2p.py Adds ddpm_edit.py necessary for instruct-pix2pix
2023-01-25Merge pull request #7146 from EllangoK/masterAUTOMATIC1111
Adds X/Y/Z Grid Script
2023-01-25Add UI setting for upcasting attention to float32brkirch
Adds "Upcast cross attention layer to float32" option in Stable Diffusion settings. This allows for generating images using SD 2.1 models without --no-half or xFormers. In order to make upcasting cross attention layer optimizations possible it is necessary to indent several sections of code in sd_hijack_optimizations.py so that a context manager can be used to disable autocast. Also, even though Stable Diffusion (and Diffusers) only upcast q and k, unfortunately my findings were that most of the cross attention layer optimizations could not function unless v is upcast also.
2023-01-25Add option for float32 sampling with float16 UNetbrkirch
This also handles type casting so that ROCm and MPS torch devices work correctly without --no-half. One cast is required for deepbooru in deepbooru_model.py, some explicit casting is required for img2img and inpainting. depth_model can't be converted to float16 or it won't work correctly on some systems (it's known to have issues on MPS) so in sd_models.py model.depth_model is removed for model.half().
2023-01-24remove fairscale requirement, add fake fairscale to make BLIP not complain ↵AUTOMATIC
about it mk2
2023-01-24remove fairscale requirement, add fake fairscale to make BLIP not complain ↵AUTOMATIC
about it
2023-01-24handling sub grids and merging into oneEllangoK
2023-01-24also return the removed field to sdapi/v1/upscalers because someone might ↵AUTOMATIC
have relied on it existing
2023-01-24repair sdapi/v1/upscalers returning bogus resultsAUTOMATIC
2023-01-23Fix different first gen with Approx NN previewsbrkirch
The loading of the model for approx nn live previews can change the internal state of PyTorch, resulting in a different image. This can be avoided by preloading the approx nn model in advance.
2023-01-23add image decod exception handlingVladimir Mandic
2023-01-24fix BLIP failing to import depending on configurationAUTOMATIC
2023-01-24Merge pull request #7113 from vladmandic/interrogateAUTOMATIC1111
Add selector to interrogate categories
2023-01-23add support for apostrophe in extra network namesAUTOMATIC
2023-01-23add option to skip interrogate categoriesVladimir Mandic
2023-01-23Merge branch 'AUTOMATIC1111:master' into interrogateVladimir Mandic
2023-01-23Merge pull request #7032 from gmq/extra-network-stylesAUTOMATIC1111
Extra network view style
2023-01-23api-image-formatVladimir Mandic
2023-01-23a possible fix for broken image upscalingAUTOMATIC
2023-01-23improve interrogateVladimir Mandic
2023-01-23better support for xformers flash attention on older versions of torchAUTOMATIC
2023-01-23Merge remote-tracking branch 'takuma104/xformers-flash-attention'AUTOMATIC
2023-01-23fix open directory button failingAUTOMATIC
2023-01-23Merge pull request #7031 from EllangoK/masterAUTOMATIC1111
Fixes various button overflowing UI and compact checkbox
2023-01-23Merge pull request #7093 from Shondoit/fix-dark-modeAUTOMATIC1111
Fix dark mode
2023-01-23third time's the charmAUTOMATIC
2023-01-23add missing import to previous commitAUTOMATIC
2023-01-23Fix dark modeShondoit
Fixes #7048 Co-Authored-By: J.J. Tolton <jjtolton@gmail.com>
2023-01-23rework extras tab to use script systemAUTOMATIC
2023-01-22feat(extra-networks): remove view dropdownGuillermo Moreno
2023-01-22feat(extra-networks): add default view settingGuillermo Moreno
2023-01-22feat(extra-networks): add thumbs view styleGuillermo Moreno
2023-01-22split oversize extras.py to postprocessing.pyAUTOMATIC
2023-01-22Split history extras.py to postprocessing.pyAndrey
2023-01-22Split history extras.py to postprocessing.pyAndrey
2023-01-22Split history extras.py to postprocessing.pyAndrey
2023-01-22Split history extras.py to postprocessing.pyAndrey
2023-01-22amend previous commit to work in a proper fashion when saving previewsAUTOMATIC
2023-01-22add an option to reorder tabs for extra networksAUTOMATIC
2023-01-22add option to discard weights in checkpoint merger UIAUTOMATIC
2023-01-22fix missing field for aesthetic embedding extensionAUTOMATIC
2023-01-22attention ctrl+up/down enhancementsAUTOMATIC
2023-01-21compact checkbox and fix copy image btn overflowEllangoK
also fixes type for #tab_extensions in style.css
2023-01-22fix broken textual inversion extras tabAUTOMATIC
2023-01-22add a slider for default value of added extra networksAUTOMATIC
generated by cgit v1.2.1 (git 2.18.0) at 2024-09-20 15:28:22 +0000