- A python chatbot using python's sentence-transformers with a React frontend.
- User inputs are embedded numerically to create a vector representation for a more accurate, quicker response.
- This version takes my resume to answer job interview questions.
- Can easily be adjusted to use on your e-commerce, personal, and professional websites.
-
clone
-
venv
-
pip install -r requirements.txt
-
make your json file following this structure:
{ "questions": [ { "patterns": ["Hello", "Hi", "Hey", "Good day", "Greetings"], "response": "Hello! How can I assist you today?" }, { "patterns": [ "What is your name?", "Who are you?", "Can you tell me about yourself?" ], "response": "I am a chatbot created to share information about _." } ] }
(ChatGPT can create this file for you, given the structure and the document you want to create questions from. The more patterns you provide, the more accurate the bot becomes.)
-
Adjust line 12 to the correct json file name.
-
client, npm install, run React, run flask
- Connect to database to hold user inputs to generate more responses.