THIS PROJECT IS EXPERIMENTAL. USE IT AT YOUR OWN RISK.
CLIP STUDIO PAINT filter plugin for executing Stable Diffusion txt2img, img2img processor.
This plugin filter runs Stable Diffusion in local environment.
So if you want to use this filter, greater than 12GB (allocated) of RAM and greater than 8GB (filters to 512x512) of VRAM.
Please visit INSTALLATION (Japanese).
Please visit below documents:
Basically, we believe that publishing illustrations in progress on the Internet or on services that can be censored, such as Google, would lead to the leakage of confidential information.
And they will also be at risk of account loss.
MIT by @6jz
Dependencies are not optimized. Some packages will be removed in future updates.
absl-py
aiohttp
aiosignal
async-timeout
frozenlist
google-auth
google-auth-oauthlib
grpcio
huggingface-hub
importlib-metadata
kornia
multidict
packaging
- or BSD 2 Clausepyarrow
pydeck
Pympler
python-dateutil
- or BSDpytorch-lightning
pytz-deprecation-shim
regex
requests
rsa
streamlit
tensorboard
tensorboard-data-server
tensorboard-plugin-wit
torch-fidelity
torchmetrics
tornado
transformers
tzdata
watchdog
yarl
antlr4-python3-runtime
colorama
fsspec
gitdb
GitPython
Markdown
networkx
numpy
oauthlib
omegaconf
pyasn1
Pygments
requests-oauthlib
scipy
semver
smmap
tifffile
tokenizers
toolz
torchaudio
torchvision
pybind11
albumentations
attrs
blinker
cachetools
charset-normalizer
einops
entrypoints
ftfy
future
imgaug
jsonschema
latent-diffusion
opencv-python
opencv-python-headless
pip
pubd
pyDeprecate
pyparsing
pyrsistent
pytz
PyWavelets
PyYAML
rich
setuptools
six
test-tube
toml
tqdm
- or MPL 2.0tzlocal
urllib3
validators
wcwidth
wheel
zipp
git+CLIP
git+taming-transformers