Spaces:
Runtime error
Runtime error
Commit History
-
[Admin maintenance] Support new ZeroGPU hardware (#7)
- [Admin maintenance] Support new ZeroGPU hardware (949cc4d23677acef292f54d41a455aaac8122c98) Co-authored-by: Apolinário from multimodal AI art <multimodalart@users.noreply.huggingface.co>24a6952
-
feat: Add width and height parameters to submit_function_flux for enhanced image processing
- Introduced `width` and `height` parameters in the `submit_function_flux` function, allowing users to specify the dimensions of the output images. This change improves flexibility in image processing and enhances user control over the generated results.c35ec9e -
refactor: Simplify output handling in submit_function_flux by removing return_dict
- Removed the `return_dict` parameter from the `submit_function_flux` function, streamlining the output to directly return the first image from the results. This change enhances code clarity and reduces complexity in the function's return value management.f97271b -
feat: Add return_dict parameter to submit_function_flux for enhanced output control
- Introduced the `return_dict` parameter in the `submit_function_flux` function, allowing users to specify the output format of the function's results. This change improves flexibility in handling the function's return values.f86262e
-
refactor: Rename mask parameter to mask_image in submit_function_flux
- Updated the parameter name from `mask` to `mask_image` in the `submit_function_flux` function for improved clarity and consistency in the codebase.d63caf4 -
refactor: Remove resolution parameters and update guidance scale in app.py
- Eliminated the resolution parameter from the submit_function_flux, simplifying the function signature. - Adjusted the guidance scale slider's maximum value from 7.5 to 50, enhancing user control over the model's behavior. - Updated the UI layout for result image display, ensuring consistency in the application interface.7f24de3 -
feat: Implement mask-based and Flux.1 fill development tab in app.py
- Added a new tab for mask-based and Flux.1 fill development, allowing users to upload person and condition images for virtual try-on. - Introduced advanced options including inference steps, guidance scale, and random seed settings. - Enhanced user experience with example images for both person and condition inputs. - Updated the UI layout to improve accessibility and interaction for users.0ed20b9 -
fix: Update attention checkpoint version in CatVTON pipeline
- Changed the attention checkpoint version from "mix" to "mix-48k-1024" in app.py to improve model performance and ensure compatibility with the latest updates.6025240 -
feat: Add authentication token support for Flux and CatVTON pipelines in app.py
- Integrated the use of an authentication token for accessing Hugging Face models in the FluxTryOnPipeline and CatVTONPix2PixPipeline. - Updated the model loading functions to include `use_auth_token` parameter, enhancing security and access control for model repositories.bde7aa3 -
fix: Update attention checkpoint path in app.py for CatVTON pipeline
- Changed the attention checkpoint path from the previous repository to the MaskFree version for improved model performance. - Updated the attention checkpoint version to a more general reference, enhancing compatibility with future updates.673064b -
feat: Add new argument for p2p base model path in app.py
- Introduced a new command-line argument `--p2p_base_model_path` to specify the path to the base model for evaluation, defaulting to "timbrooks/instruct-pix2pix". This enhancement allows for greater flexibility in model selection during runtime.26c80cc -
feat: Enhance CatVTON functionality with new pipelines and UI improvements
- Added CatVTONPix2PixPipeline and FluxTryOnPipeline to support additional virtual try-on methods. - Implemented new submit functions for mask-free and Flux-based try-on. - Updated UI to include separate tabs for mask-based and mask-free options, enhancing user experience. - Modified requirements.txt to include new dependencies and updated existing ones. - Improved error handling and image processing in the submission functions.905c952
-
chore: Update default base model path in app.py
3e791eb -
chore: Update default base model path in app.py
1519098
-
chore: Update app.py to improve demo instructions and links
32b0bc1 -
chore: Update app.py to improve demo instructions and links
b339a83 -
feat: Add GPU duration parameter to submit_function
640d658 -
chore: Update app.py to fix torch.jit.script assignment
6927e60
-
Update app.py
Verified signature7b80416
-
chore: Set CUDA environment variables in app.py
fe2cfb5 -
chore: Update dependencies and code structure
6eb1d7d -
chore: Update segformer_ckpt in app.py to use mattmdjaga/segformer_b2_clothes
19a1fa7 -
chore: Remove unused files and dependencies
21c6c10
-
chore: Update app.py to set CUDA_HOME environment variable and add CUDA bin directory to PATH
ZhengChong864fc50 -
chore: Set CUDA_HOME environment variable for GPU support
ZhengChonge730f66 -
Add spaces library for GPU support
ZhengChong8d0429d -
chore: Add SCHP model and detectron2 dependencies
ZhengChong6a6227f