Skip to content

Latest commit

 

History

History
95 lines (65 loc) · 2.79 KB

README.md

File metadata and controls

95 lines (65 loc) · 2.79 KB
Raw

TutorChat README

Demo

See how the Gradio app work. Watch the demo video.

Description

A language conversation assistant is a tool that allows users to interact with a language model. Users can not only chat with the assistant but also record their voice, which will be transcribed and processed by the assistant to generate responses. Moreover, the program leverages a text-to-speech service to audibly read out the responses.

Features

  • Interaction with OpenAI's GPT language models.
  • Real-time voice recording and transcription.
  • Text-to-speech functionality to audibly present responses.
  • Gradio interface for ease of use

Prerequisites

Before you begin, ensure you have the following installed:

  • ffmpeg: This tool is crucial for audio processing.

    • Windows: You can download it from the official FFmpeg site or use a package manager like Chocolatey.

      choco install ffmpeg

    • macOS: Using Homebrew:

      brew install ffmpeg

    • Linux: Depending on your distribution, you can use apt, yum, or another package manager:

      sudo apt update
sudo apt install ffmpeg

Ensure ffmpeg is correctly installed by running:

ffmpeg -version

Requirements

  • Python 3.9 or higher.
  • Python Libraries:
    • numpy
    • gradio
    • openai
    • requests
    • pydub

Installation

  1. Clone this repository.
  2. Install the required packages with pip install -r requirements.txt

Configuration

  1. Place your OpenAI API key in a file named openaiapikey.txt.
  2. Place your Elevenlabs API key (for the text-to-speech feature) in a file named elabapikey.txt.
  3. (Optional) Modify chatbot_{language}.txt files if you wish to customize the initial behavior of the chatbot.

Usage

  1. Run the script with python main.py.
  2. Once executed, the Gradio interface will be accessible via your local browser at: http://127.0.0.1:7860/
  3. A Gradio interface will pop up, presenting a user-friendly way to interact with the chatbot.
  4. Choose your preferred language for the conversation.
  5. Select a response voice from the available options.
  6. Start your interaction by pressing the "Move to Chat button"
  7. Record your voice by using the "Record from microphone" button and send it to the chatbot with "Send Audio"
  8. The chatbot will respond both in text and audibly using the selected voice type.
  9. To end the session, simply close the Gradio interface.

Contributing

If you wish to contribute to this project, please fork the repository and submit a Pull Request.

License

This project is under the MIT License. Please refer to the LICENSE file for more details.