NomoNotes is a tool designed to streamline your note-taking process during meetings, lectures, podcasts etc. It transcribes audio in real-time, summarizes the transcript using a language model, and saves the summary directly into Apple Notes.

• Real-time Audio Transcription: Utilizes Whisper for accurate and efficient transcription (code from https://github.com/davabase/whisper_real_time)
• Summarization: Summarizes the transcript using an LLM.
• Apple Notes Integration: Saves the summary into Apple Notes.

1. Clone the repository:

   git clone https://github.com/yourusername/nomonotes.git
   cd nomonotes
   

2. Install the required dependencies: pip install -r requirements.txt

Run the script with the following command: python transcribe.py --model medium --llm_model gpt-4o --summary_prompt meeting_prompt --notes_folder_name "NOMO Notes" To end the transcription, just keyboard interrupt the process. The script will then start generating your summary and save it to Apple Notes.

• --model: Model to use for transcription (choices: tiny, base, small, medium, large).
• --non_english: Use non-English model.
• --energy_threshold: Energy level for mic to detect.
• --llm_model: LLM model to use for summarization (this should either be an OAI model or a .gguf model you've downloaded to models/)
• --summary_prompt: Path to the summary prompt (this is the name of the file with the prompt in the prompts/ folder)
• --notes_folder_name: Name of the folder where the note will be created.

`python transcribe.py --model medium --llm_model gpt-4o --summary_prompt meeting_prompt --notes_folder_name "NOMO Notes"`

If you prefer a UI, use streamlit:

python -m streamlit run ui.py

You can use your own local models for summarization. For example, if you have a fine-tuned model that summarizes meetings and podcasts in a specific format, you can specify it using the --llm_model argument.

To use an off-the-shelf llama, you can run run_llama.sh which downloads and runs the script with an 8b llama model.