Ollama Deepseek Local Chatbot 🤖

App by Build Fast with AI

A lightweight, offline-capable chatbot powered by Deepseek R1 1.5B and Ollama, running completely on your local machine. No internet required!

Features ✨

• 🏃‍♂️ Runs 100% locally - no internet or cloud services needed
• 💨 Streaming responses like ChatGPT
• 🧠 Shows AI thinking process in expandable sections
• 🎯 Uses the tiny but mighty Deepseek R1 1.5B model
• 🚀 Built with Streamlit for a clean, modern UI

Prerequisites

• Python 3.10+
• Ollama installed on your system
• Deepseek R1 1.5B model pulled in Ollama

Quick Start 🚀

1. Install Ollama and pull the Deepseek model:

2. Clone this repository:

git clone [your-repo-url]
cd [your-repo-name]

3. Install dependencies:

pip install -r requirements.txt

4. Run the app:

streamlit run deepseek-r1-streamlit.py

How It Works 🛠️

The app uses:

• Ollama for running the Deepseek R1 1.5B model locally
• LangChain for model interaction
• Streamlit for the web interface
• Streaming responses for real-time interaction
• Expandable "thinking process" sections to see the AI's reasoning

Dependencies 📦

The project uses three main Python packages:

streamlit
langchain
langchain-community

Usage Examples 💡

After starting the app:

1. Open your browser at http://localhost:8501
2. Type your question in the chat input
3. Watch the AI respond in real-time
4. Click "Show AI thinking process" to see the reasoning

Offline Usage 🔌

This chatbot is designed to work completely offline. Once you have:

1. Installed Ollama
2. Downloaded the Deepseek model
3. Installed Python dependencies

You can disconnect from the internet and continue using the chatbot!

Acknowledgments 🙏

• Ollama for the local model serving
• Deepseek for the amazing 1.5B model
• LangChain for the framework
• Streamlit for the web interface

---

Made with ❤️ for the local AI community