🧠 Sentiment-Driven Chatbot

This project is a sentiment-driven chatbot application powered by a PyTorch-based sentiment analysis model and the Ollama Llama model for generating empathetic responses. The chatbot runs on a Streamlit-based WhatsApp-style user interface.

🎥 Demo Video

Here is a demo video of the sentiment-driven chatbot in action: Demo Video

🔧 Project Setup

1. Prerequisites

• Install Python 3.8 or higher.
• Install the required dependencies using the command:

  pip install -r requirements.txt

2. Files in the Project

• sentiment_model.pth: A PyTorch-based classification model trained for sentiment analysis (Positive, Neutral, Negative).
• vocab.pkl: A vocabulary file mapping words to indices, used to preprocess user input for sentiment analysis.

3. Llama Model Installation

• Download and install Ollama, a locally hosted Llama model environment.
• Run the following command to download the llama3.2:1b model:

  ollama run llama3.2:1b

• Ensure the Ollama server is running on http://127.0.0.1:11436.

4. Start the Application

• To start the Streamlit application, run:

  streamlit run app.py
  

🛠 Project Details

1. Sentiment Analysis

The sentiment_model.pth is a PyTorch model that classifies user input into one of three sentiment categories:

• Positive
• Neutral
• Negative

The model is trained using a sequence of words converted to indices from the vocab.pkl file. User input is preprocessed, tokenized, and padded to a maximum length before passing through the model.

1. Llama for Response Generation

The chatbot generates empathetic responses based on user sentiment using the llama3.2:1b model. The sentiment information is appended to the user input to create a prompt that guides the Llama model.

Example Prompt: The user's sentiment is Positive. Respond empathetically to: "What a wonderful day!"

The response is fetched using the Ollama API hosted locally.

3. User Interface

The application uses a WhatsApp-style chat interface created with Streamlit. Features include:

• User messages displayed on the right in blue.
• Chatbot responses displayed on the left in green.
• Sentiment information displayed alongside user messages.

📚 Dependencies

The project uses the following Python libraries:

• Streamlit: For creating the chat interface.

• Torch: For the sentiment analysis model.

• Ollama API: To interact with the Llama model for response generation.

• Install all dependencies using:

  pip install -r requirements.txt
  

🚀 How to Use

1. Start the Ollama server with the llama3.2:1b model:

   ollama run llama3.2:1b

2. Run the chatbot application:

   streamlit run app.py

3. Interact with the chatbot in the browser at http://localhost:8501.

✨ Features

• Sentiment-based empathetic response generation.
• Intuitive WhatsApp-style chat interface.
• Locally hosted Llama model for privacy and performance.

📝 Notes

• Ensure the Ollama server is running before starting the chatbot application.
• Update the port (http://127.0.0.1:11436) in the app.py file if using a different port for Ollama.