caf84e8233
Expose inpainting mask and composite
...
For inpainting, this exposes the mask and masked composite and gives
the user the ability to display these in the web UI,
save to disk, or both.
3 years ago
4d26c7da57
initialize extra_network_data before use
3 years ago
da3f942ab2
Merge pull request #8017 from space-nuko/before-process-batch
...
Add `before_process_batch` script callback
3 years ago
ac38ad7e60
Merge remote-tracking branch 'origin/master' into unipc
3 years ago
a2d635ad13
Add before_process_batch script callback
3 years ago
15f4b217b1
fix the a merge conflict resolve i did that entirely breaks image generation
3 years ago
6911deb242
Merge branch 'master' into fix-save-params-2
3 years ago
b20737815a
Fix params.txt saving for infotexts modified by process_batch
3 years ago
0a4917ac40
Apply extra networks per-batch instead of per-session (fixes wildcards)
3 years ago
fb274229b2
bug fix
3 years ago
c88dcc20d4
UniPC does not support img2img (for now)
3 years ago
1253199889
Working UniPC (for batch size 1)
3 years ago
c27c0de0f7
txt2img Hires Fix
3 years ago
6c6c6636bb
Image CFG Added (Full Implementation)
...
Uses separate denoiser for edit (instruct-pix2pix) models
No impact to txt2img or regular img2img
"Image CFG Scale" will only apply to instruct-pix2pix models and metadata will only be added if using such model
3 years ago
3b2ad20ac1
Processing only, no CFGDenoiser change
...
Allows instruct-pix2pix
3 years ago
cf0cfefe91
Revert "instruct-pix2pix support"
...
This reverts commit 269833067d
3 years ago
269833067d
instruct-pix2pix support
3 years ago
040ec7a80e
make the program read Eta and Eta DDIM from generation parameters
3 years ago
00dab8f10d
remove Batch size and Batch pos from textinfo (goodbye)
3 years ago
fecb990deb
Merge pull request #7309 from brkirch/fix-embeddings
...
Fix embeddings, upscalers, and refactor `--upcast-sampling`
3 years ago
ada17dbd7c
Refactor conditional casting, fix upscalers
3 years ago
5eee2ac398
add data-dir flag and set all user data directories based on it
3 years ago
7a14c8ab45
add an option to enable sections from extras tab in txt2img/img2img
...
fix some style inconsistenices
3 years ago
10421f93c3
Fix full previews, --no-half-vae
3 years ago
d1d6ce2983
add edit_image_conditioning from my earlier edits in case there's an attempt to inegrate pix2pix properly
...
this allows to use pix2pix model in img2img though it won't work well this way
3 years ago
1574e96729
Merge pull request #6510 from brkirch/unet16-upcast-precision
...
Add upcast options, full precision sampling from float16 UNet and upcasting attention for inference using SD 2.1 models without --no-half
3 years ago
57c1baa774
change to code for live preview fix on OSX to be bit more obvious
3 years ago
e3b53fd295
Add UI setting for upcasting attention to float32
...
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.
3 years ago
84d9ce30cb
Add option for float32 sampling with float16 UNet
...
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().
3 years ago
f64af77adc
Fix different first gen with Approx NN previews
...
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.
3 years ago
78f59a4e01
enable compact view for train tab
...
prevent previews from ruining hypernetwork training
3 years ago
3deea34135
extract extra network data from prompt earlier
3 years ago
92fb1096db
make it so that extra networks are not removed from infotext
3 years ago
40ff6db532
extra networks UI
...
rework of hypernets: rather than via settings, hypernets are added directly to prompt as <hypernet:name:weight>
3 years ago
a8322ad75b
Merge pull request #6854 from EllangoK/master
...
Saves Extra Generation Parameters to params.txt
3 years ago
b186d44dcd
use DDIM in hires fix is the sampler is PLMS
3 years ago
5e15a0b422
Changed params.txt save to after manual init call
3 years ago
e0e8005009
make StableDiffusionProcessing class not hold a reference to shared.sd_model object
3 years ago
9991967f40
Add a check and explanation for tensor with all NaNs.
3 years ago
f9ac3352cb
change hypernets to use sha256 hashes
3 years ago
88416ab5ff
Fix extension parameters not being saved to last used parameters
3 years ago
d4fd2418ef
add an option to use old hiresfix width/height behavior
...
add a visual effect to inactive hires fix elements
3 years ago
50e2536279
Merge branch 'AUTOMATIC1111:master' into img2img-api-scripts
3 years ago
1a5b86ad65
rework hires fix preview for #6437 : movie it to where it takes less place, make it actually account for all relevant sliders and calculate dimensions correctly
3 years ago
b5253f0dab
allow img2img api to run scripts
3 years ago
847f869c67
experimental optimization
3 years ago
2e30997450
move sd_model assignment to the place where we change the sd_model
3 years ago
83ca8dd0c9
Merge branch 'AUTOMATIC1111:master' into fix-sd-arch-switch-in-override-settings
3 years ago
99b67cff0b
make hires fix not do anything if the user chooses the second pass resolution to be the same as first pass resolution
3 years ago
bc43293c64
fix incorrect display/calculation for number of steps for hires fix in progress bars
3 years ago
catboxanon
high_byte
AUTOMATIC1111
space-nuko
Kyle
brkirch
Max Audron
EllangoK
noodleanon
Philpax