fix: Fixed the issue in the `merge_chunks_with_crossfade` function handling one-dimensional audio blocks and blocks shorter than the overlap area, and removed redundant dimension expansion operations in `save_audio`.v

feat: To implement the audio chunking function with overlapping and cross-fading for processing long audio files, and to add chunk duration control to the UI.

Simplify UI - remove tabs, fix example buttons

Add progress bar, example only fills data, switch to video tab

Restore audio/video preview with tabs

Add example with Git LFS for MP4

Revert to simple clean UI

Add video preview with tabs for audio/video upload

Add MP4 and video file support

Fix: follow official HF example exactly

Follow official sam-audio example exactly

Follow sam3 pattern: global model load with spaces.GPU

Fix: lazy load model inside GPU context for ZeroGPU

chore: Update Dependencies

docs: Update the Gradio SDK version in the README.

docs: Update the Gradio SDK version to 5.0.0

docs: Update README metadata to specify the Gradio SDK and its version, and remove python_version

Configure ZeroGPU with Python 3.11

feat: Migrate the deployment to the Gradio SDK, integrate the `spaces.GPU` decorator, and remove the Dockerfile.

Switch to Docker SDK with Python 3.12

Fix python_version format without quotes

Require Python 3.11 in requirements.txt

Require Python 3.11 for perception-models compatibility

Fix import error by using sam-audio library

feat: Added the SAM Audio audio segmentation Web UI based on Gradio and its dependencies.

initial commit