Getting Started with Mistral: An Introduction

Getting Started with Mistral: An Introduction

The open-source Mixtral 8x7B model launched by Mistral adopts a “Mixture of Experts” (MoE) architecture. Unlike traditional Transformers, the MoE model incorporates multiple expert feedforward networks (this model has 8), and during inference, a gating network is responsible for selecting two experts to work. This setup allows MoE to achieve faster inference speed while maintaining performance comparable to large models. The Mixtral 8x7B model has a total of 46.7B parameters, but only 12.9B parameters are activated during actual inference to predict the next token.

1. Mistral Website

In the “Getting Started with Mistral” course, presented by Mistral’s Sophia Yang, Ph.D., you will learn: – Explore Mistral’s open-source models (Mistral 7B, Mixtral 8x7B) and commercial models through API calls and Mistral AI’s Le Chat website. – Implement JSON formats to generate structured outputs that can be directly integrated into large software systems. – Learn how to use function calls for tool operations, such as querying tabular data with custom Python code. – Combine your large language model (LLM) responses with external knowledge sources, using RAG technology to enhance practicality. – Create a Mistral-driven chat interface that can refer to external documents. This course will help enhance your prompt engineering skills. Course link:https://deeplearning.ai/short-courses/

01.Getting Started with Mistral: Overview

getting-started-with-mistral/Getting Started with Mistral: Introduction02.Getting Started with Mistral: An Introduction

03.Getting Started with Mistral: How to Write Prompts