Skip to content

Latest commit

 

History

History
59 lines (47 loc) · 2.41 KB

chatbots.md

File metadata and controls

59 lines (47 loc) · 2.41 KB
Raw

Creating a Basic Chatbot with ASR & TTS Using Streamlit

Introduction

You now have the opportunity to put your knowledge into action by building a simple chatbot. Keep in mind that these concepts can be applied to a wide range of use-cases. This demonstration will also illustrate how to incorporate Automatic Speech Recognition (ASR) and Text-to-Speech (TTS) capabilities (PLUS GPT-4 VISION!). You can interact with the chatbot through speech input and receive responses in speech format.

How to Get Started

  1. Make sure you have your credentials correctly saved in the .env file. THE FILE MUST BE ".env" ONLY
  2. Make sure you're running within your venv environment. Use the following commands:

Windows Run the below under powershell or terminal as admin:

Set-ExecutionPolicy RemoteSigned

Once that's done you can run the following in VS Code:

./venv/scripts/activate.ps1

in Cmd you can run

c:\pathtovenv\scripts\activate.bat

On *nix/mac you can run this command

source ./venv/bin/activate

  1. Once in the environment, run the following command to install all the required modules for this application:

    pip install -r requirements.txt

To launch the basic chatbot

The basic chatbot just provides chat functionality with ASR, TTS & Vision. You can run it with the following command:

streamlit run 'basic_chatbot.py'

To launch the advanced autonomous agent chatbot

The advanced chatbot includes all basic functions (minus vision as it doesnt support functions) in addition to automated function calling of internet and kb skills.

You can run it with the following command:

streamlit run 'advanced_chatbot.py'

To launch the OpenAI Assistants & Threads chatbot

The assistant chatbot uses the new OpenAI Assistants and Threads functions.

You can run it with the following command:

streamlit run 'chatbot_assistants.py'

Please make sure to complete the course before going into this file as i will slowly explain everything

Feel free to engage in a conversation with the chatbot. You can also customize and enhance the system prompts as needed. This is all about effectively providing input and using the output to your advantage. While we are currently focused on speech, it won't be long before we gain access to using photos and videos as inputs and outputs as well! Have fun, and best of luck with your future endeavors.

Wanis Elabbar