f8e36141a9
Add support for mixed precision activation quantization for UNet model ( #365 )
2024-11-15 16:05:22 -08:00
6365319010
fix: Remove duplicate logger instantiation
2024-10-15 19:41:54 -07:00
8cf34376f9
SDXL and other improvements
...
Co-authored-by: Pedro Cuenca <pedro@huggingface.co >
Co-authored-by: Zach Nagengast <zacharynagengast@gmail.com >
2023-07-27 09:08:30 -10:00
7f9c58a6c1
WWDC23 update
...
Co-authored-by: davidfindlay1 <david.arthur.findlay@icloud.com >
2023-06-14 08:27:02 -07:00
7f65e1c84b
Support ControlNet ( #153 )
...
* add controlnet tentatively
* add controlnet in python code
* implement swift part
* support 8-bit quantization
* add controlnet unload when reduce memory
* remove irrelevant changes
* add more description about controlnet option in swift
* fix some for pr and update README
* pre-allocate zero shapedArray + make multi-controlnet faster
2023-04-18 16:17:59 -07:00
6cd5c7a760
Update import to nested path ( #109 )
...
* Update import to nested path
* Update python_coreml_stable_diffusion/unet.py
Co-authored-by: Pedro Cuenca <pedro@huggingface.co >
Co-authored-by: Atila Orhon <64497909+atiorh@users.noreply.github.com >
Co-authored-by: Pedro Cuenca <pedro@huggingface.co >
2023-01-26 11:12:45 -08:00
3b24fe36a7
Fix downsample channels. ( #110 )
...
Reference: https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/unet_2d_blocks.py#L749-L756
This is usually not a problem for Stable Diffusion models. However, some
architectures such as
https://huggingface.co/OFA-Sys/small-stable-diffusion-v0 wouldn't
convert.
2023-01-26 11:10:58 -08:00
583cc04ece
Initial commit
...
Co-authored-by: aseemw <aseem.elec@gmail.com >
Co-authored-by: msiracusa <msiracusa+github@gmail.com >
2022-11-30 13:41:05 -06:00
NehalBhandari