Web Interface API for facebook/seamless-m4t-v2-large Model Backend Demonstration

Description:

     Demonstration of a web interface for inferring facebook/seamless-m4t-v2-large model via API calls, using Flask as the backend server.

Purposes:

     Demonstration two API calls:

• Text-to-text translation (T2TT) - t2t
• Text-to-speech translation (T2ST) - t2s
• Speech-to-text translation (S2TT) - s2t
• Speech-to-speech translation (S2ST) - s2s

File Descriptions

• seamlessm4tv2_api.py: Flask backend API provider for facebook/seamless-m4t-v2-large model.  There are currently four API calls: t2t, t2s, s2t, s2s.

• model_initializer.py: Loading model.

• index.html: A web interface to make inferences.

• seamlessm4tv2_api_s.py: Flask backend API provider for facebook/seamless-m4t-v2-large model with SSL enabled.  There are currently four API calls: t2t, t2s, s2t, s2s.

• static\index_s.html: A SSL version of web interface to make inferences.

Contributors

     ChatGPT-3.5 the coding machine!

     The voice recording portion is attributed to Bing Chatbot!

Project Attribution

Model from Hugging Face: 
- facebook/seamless-m4t-v2-large

Disclaimer

     This project is provided "as is" and without any warranty. Use it at your own risk.