Mydra logo
Artificial Intelligence
DeepLearning.AI logo

DeepLearning.AI

Pretraining LLMs

  • up to 1 hour
  • Beginner

In this course, you'll gain in-depth knowledge of pretraining large language models, from data preparation to model configuration and performance assessment. Learn innovative techniques to reduce training costs and explore various model configurations.

  • Data preparation
  • Model configuration
  • Performance evaluation
  • Text generation
  • Dataset creation

Overview

This course provides a comprehensive understanding of pretraining large language models. You'll learn to prepare data, configure models, and assess performance, with a focus on cost-effective strategies. The course covers creating high-quality datasets, using the Hugging Face library, and evaluating model performance with benchmark tasks.

  • Web Streamline Icon: https://streamlinehq.com
    Online
    course location
  • Layers 1 Streamline Icon: https://streamlinehq.com
    English
    course language
  • Self-paced
    course format
  • Live classes
    delivered online

Who is this course for?

AI Enthusiasts

Individuals interested in learning about the process of pretraining large language models.

Data Scientists

Professionals looking to expand their knowledge on model pretraining and performance evaluation.

Machine Learning Engineers

Engineers aiming to understand the complete process of pretraining LLMs for better model performance.

This course equips beginners and professionals with the skills to pretrain large language models, covering essential topics like data preparation and model configuration. Enhance your career by mastering cost-effective pretraining techniques and performance evaluation strategies.

Pre-Requisites

1 / 2

  • Basic knowledge of Python

  • Understanding of large language models

What will you learn?

Introduction
An overview of the course and its objectives.
Why Pre-training
Understanding the importance and benefits of pretraining large language models.
Data Preparation
Learn how to create and clean datasets for effective model pretraining.
Packaging Data for Pretraining
Techniques for preparing your training data for use with the Hugging Face library.
Model Initialization
Explore various options for configuring and initializing models for training.
Training in Action
Execute a training run and learn to train your own model.
Evaluation
Assess your trained model’s performance and explore common evaluation strategies.
Conclusion
Summarize the key learnings and outcomes of the course.
Quiz
Test your understanding of the course material.

Upcoming cohorts

  • Dates

    start now

Free