https://github.com/bharath-tars/speech-to-speech-bot

A real-time voice bot using LLaMA 3.3-80B for intelligent responses, OpenAI Whisper V3 Turbo for speech-to-text, and gTTS for text-to-speech conversion.
https://github.com/bharath-tars/speech-to-speech-bot

gtts huggingface llama-3-70b python speech-to-speech streamlit tranformers whisper-ai

Last synced: about 2 months ago
JSON representation

A real-time voice bot using LLaMA 3.3-80B for intelligent responses, OpenAI Whisper V3 Turbo for speech-to-text, and gTTS for text-to-speech conversion.

• Host: GitHub
• URL: https://github.com/bharath-tars/speech-to-speech-bot
• Owner: Bharath-tars
• License: gpl-3.0
• Created: 2025-01-18T06:55:44.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2025-01-22T11:44:19.000Z (over 1 year ago)
• Last Synced: 2025-03-29T11:16:48.896Z (over 1 year ago)
• Topics: gtts, huggingface, llama-3-70b, python, speech-to-speech, streamlit, tranformers, whisper-ai
• Language: Python
• Homepage: https://huggingface.co/spaces/BharathTars/Nova-Voice-BOT
• Size: 19.5 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# Speech-to-Speech Conversational Bot

## Introduction

The **Speech-to-Speech Conversational Bot** is a real-time voice interaction system that combines cutting-edge technologies to enable seamless, intuitive conversations. It utilizes advanced language models and speech processing tools to provide a natural and engaging user experience. 

## Features

- **Speech Input**: Records user audio through the interface.

- **Real-time Transcription**: Converts speech to text using **OpenAI Whisper V3 Turbo** via Groq's inference client.

- **Intelligent Responses**: Generates context-aware responses using **LLaMA 3.3-80B**.

- **Speech Output**: Converts responses back to speech using **gTTS** for natural voice playback.

- **User Interface**: Built with **Streamlit** for easy interaction and audio visualization.

## Stack Design

- **Frontend**: Streamlit for audio recording and response display.

- **Speech-to-Text**: OpenAI Whisper V3 Turbo accessed through Groq's inference client.

- **Language Model**: LLaMA 3.3-80B for generating intelligent and conversational responses.

- **Text-to-Speech**: gTTS for converting text back to speech.

## Workflow

1. **Audio Recording**: The user records their query using the Streamlit interface.

2. **Speech Transcription**: The recorded audio is sent to the **Whisper V3 Turbo** model via Groq's inference client for transcription.

3. **Response Generation**: The transcribed text is passed to the **LLaMA 3.3-80B** model to generate a contextually appropriate response.

4. **Speech Synthesis**: The generated response is converted back into speech using **gTTS**.

5. **Playback**: The audio response is played back to the user through the interface.

## Installation

1. Clone the repository:

   ```bash

   git clone https://github.com/your-username/speech-to-speech-bot.git

   cd speech-to-speech-bot