This project is a Telegram chatbot that utilizes the OpenAI GPT model for generating responses. It is developed as part of a course on using GPT in chatbots.

To run the chatbot, you need the following dependencies:

Python 3.7 or higher

• aiogram library (pip install aiogram)
• pandas library (pip install pandas)
• openai library (pip install openai)

1. Clone the repository or download the source code files.

2. Install the required dependencies mentioned in the requirements section.

3. Open the main.py file and replace YOUR_TOKEN with your Telegram bot token and OpenAI API key.

4. Make sure you have a CSV file named users.csv in the same directory. This file is used to store user data.

5. Run the main.py file to start the chatbot:

python main.py

• The /start command is used to start the conversation with the chatbot.
• If the user is registered, it sends a welcome back message.
• If the user is not registered, it registers the user and sends a welcome message.

• The /tokens command allows registered users to request token replenishment.
• Users can request tokens if the last request was made at least 3 minutes ago.
• Upon successful token replenishment, it informs the user.

• The respond function handles user messages and generates responses using the OpenAI Chat API.
• It checks if the user is registered and has enough tokens to generate a response.
• If the user is not registered, it informs the user to register first.
• If the user has run out of tokens, it informs the user to replenish their tokens.
• It manages the conversation context by storing and updating user messages.
• It sends the user's message and the context to the OpenAI Chat API for generating a response.
• The response is sent back to the user, and the context is updated with the assistant's reply.

You can customize the behavior of the chatbot by modifying the code in main.py:

• Change the welcome messages in the start_command function.
• Modify the token replenishment logic in the tokens_command function.
• Customize the system instruction and context management in the respond function.

• The chatbot relies on the OpenAI GPT-3 model for generating responses. Any limitations or biases associated with the model apply to the chatbot.
• The chatbot requires a stable internet connection to communicate with the Telegram Bot API and the OpenAI API.

This project is licensed under the MIT License. See the LICENSE file for more details.

This project was developed as part of a course on using OpenAI GPT in chatbots. Special thanks to the course instructors for their guidance and support.

Happy chatting with your GPT-powered chatbot!