Update README.md to specify Python version as 3.12 and modify requirements.txt to include a direct link for flash-attn package installation, enhancing dependency management.

Update process_input function in app.py to handle audio generation output more robustly, introducing a fallback mechanism for text generation in case of unexpected output formats. Improve error handling during audio and text generation processes. Additionally, update requirements.txt to include flash-attn for enhanced performance.

Refactor chat history processing in app.py to support both dictionary and list formats for messages, enhancing compatibility with previous data structures. Update text input handling to align with new message format, ensuring consistent user experience.

Refactor chat history management in process_input function of app.py to utilize dictionaries for user and assistant messages, improving clarity and structure. Update error handling to maintain consistent message formatting in chat history.

Refactor process_input function in app.py to improve error handling during audio and text generation. Introduce try-except blocks for better exception management and ensure audio is only saved if generated successfully. Enhance text response processing with additional error handling for decoding.

Enhance process_input function in app.py by initializing user_message_for_display at the start and improving error handling for text responses. Ensure proper formatting for multimodal inputs and update chat history management to reflect user messages accurately.

Refactor process_input and create_demo functions in app.py to enhance chat history management and improve text response handling, including the addition of user and assistant avatars. Update input clearing logic for better user experience with multimodal inputs.

Enhance process_input and create_demo functions in app.py to improve multimodal input handling, including better formatting for user messages and integration of TextStreamer for text response generation.

Refactor process_input and create_demo functions in app.py to improve chat history handling and ensure proper formatting for multimodal inputs, addressing audio processing issues.

Refactor chat history handling in process_input function to ensure it initializes correctly when not a list; improve user message formatting for multimodal inputs in app.py.

Improve chat history formatting in process_input and create_demo functions in app.py to enhance user experience with multimodal inputs, ensuring clear display of uploaded content types.

Enhance process_input function in app.py to ensure multimodal inputs are handled correctly when None; update demo creation to include hidden placeholders for image, audio, and video inputs.

Refactor process_input function in app.py to handle multimodal inputs (image, audio, video, text) and update demo creation logic accordingly.

Add accelerate dependency to requirements.txt

Update requirements.txt to add torchvision and torchaudio dependencies.

Update app.py to change torch data type to float16 and comment out attention implementation; modify requirements.txt to reorder dependencies and remove flash-attn.

Refactor app.py to import 'spaces' and add GPU decorator; update requirements.txt to remove no-build-isolation flag from flash-attn

Add application file

initial commit