refactor: move inline text styling to a dedicated CSS class

docs: Update UI text to include specific button click instructions for generating transitions.

style: Apply 'Generate transition artifacts' button styling using element ID and remove background image.

style: Apply custom styling to the 'Generate transition artifacts' button.

fix: Update guitar-style LoRA repository path.

chore: Modify default demo repository to 'yng314/audio-demo-private'.

chore: reduce GPU duration for `_run_transition` function from 240 to 120 seconds

feat: Add descriptive `info` text to various Gradio UI components.

docs: Update UI description to recommend using default demo songs and parameters for best results.

feat: Update default audio generation parameter values and ignore .DS_Store.

feat: enhance UI clarity by updating descriptions, adding LoRA adapter details, and enabling multi-line audio output labels.

feat: Set default LoRA adapter dropdown value to 'Our Trained Guitar-Style LoRA'.

feat: Force dark theme and add API and Gradio links to the application footer.

chore: Increase GPU duration for `_run_transition` to 240 seconds.

config: Reduce GPU duration for `_run_transition` function from 240 to 120 seconds.

chore: Reduce GPU duration for the `_run_transition` function.

refactor: Update custom LoRA option label to 'Guitar-Style LoRA'.

feat: Add custom LoRA option to dropdown and map its repository.

feat: Implement LoRA comparison by generating and displaying both baseline and LoRA-enabled audio transition artifacts, increasing GPU duration for the process.

feat: Implement default demo song loading from Hugging Face Hub.

refactor: Standardize audio preview generation for hard splice, rough, and final transitions using a shared comparison window and remove `_assemble_substitute_mix`.

feat: Generate a short, faded, seam-focused preview for hard splice audio to improve UI clarity.

feat: Implement robust audio file loading with `torchaudio` and `soundfile` fallback.

feat: Implement Demucs model prefetching at startup and configure GPU decorator duration to 240 seconds.

chore: manage project dependencies.

feat: Implement ACE-Step V1.5 audio generation system with FastAPI server and Gradio UI.

feat: Add Docker support for the audio generation application by creating Dockerfile, .dockerignore, and updating application launch parameters.

docs: Update Gradio SDK version from 6.2.0 to 6.7.0 in README.

Adjust project dependencies.

refactor: dynamically filter `initialize_service` arguments using introspection to improve robustness.

feat: Conditionally enable MLX DIT based on system architecture or environment variable and update Python version to 3.11.

Downgraded `sdk_version` in `README.md` and integrated `spaces.GPU` decorator into `app.py`.

feat: implement audio transition generation pipeline with modules for transition generation, cue point selection, and audio utilities.

initial commit