Getting Started with Mistral

Welcome to the repository for the course Getting Started with Mistral by DeepLearning.AI.

About the Course

In this course, you’ll access Mistral AI’s collection of open-source and commercial models, including the Mixtral 8x7B model and the latest Mixtral 8x22B. You’ll learn about selecting the right model for your use case and get hands-on with features like effective prompting techniques, function calling, JSON mode, and Retrieval Augmented Generation (RAG).

Key Topics Covered

1. Model Selection

   • Access and prompt Mistral models via API calls.
   • Decide on the complexity of tasks:
     • Simple: Classification
     • Medium: Email writing
     • Advanced: Coding
   • Consider speed requirements to choose the appropriate model.

2. Function Calling

   • Learn to use Mistral’s native function calling.
   • Enable LLMs to perform tasks better suited for traditional code, such as querying databases for numerical data.

3. Retrieval Augmented Generation (RAG)

   • Build a basic RAG system from scratch with similarity search.
   • Properly chunk data, create embeddings, and implement RAG as a function in your chat system.

4. Interactive Chat Interface

   • Build a chat interface to interact with Mistral models.
   • Ask questions about a document you upload.

Course Objectives

By the end of this course, you’ll be equipped to:

• Effectively leverage Mistral AI’s open-source and commercial models.
• Build scalable and efficient systems for various use cases.
• Implement advanced techniques like RAG and function calling.

Getting Started

1. Clone the repository:

   git clone https://github.com/RuvenGuna94/Getting-started-with-Mistral.git
   cd Getting-started-with-Mistral

2. Follow the provided notebooks to explore Mistral AI's capabilities.

Additional Resources

• Mistral AI Documentation
• DeepLearning.AI
• Hugging Face

License

This repository is provided for educational purposes.

---

Happy learning!