This tool will achieve the following tasks: transcribe an audio file to text, create sentence segments from the transcription, identify the interview questions and corresponding answers, and identify key elements of the answers.
The following downloads are required: ffmpeg
Additional steps are needed to integrate Punctuator:
- Download the model from https://drive.google.com/drive/u/1/folders/0B7BsN5f2F1fZQnFsbzJ3TWxxMms?resourcekey=0-6yhuY9FOeITBBWWNdyG2aw
- I used the Demo-Europarl-en.pcl
- in the punctuator folder, edit punc.py line 171 to include the file path for the model
To activate the server:
- Create venv using 'python3 -m venv venv' from the root directory.
- Activate virtual environment:
- For Mac/Linux: 'source venv/bin/activate'
- For Windows: 'venv\Scripts\activate.bat'
- Install dependencies using 'pip install -r requirements.txt'.
- Start server by running 'python3 server.py' from the root directory.