Explore managed training with Amazon SageMaker and Hugging Face Transformers in this 25-minute video featuring Philipp Schmid, Machine Learning Engineer and Tech Lead at Hugging Face. Learn about SageMaker integration, preprocessing, S3 integration, and the Hugging Face Hub. Discover how to start training, access models, and leverage existing training scripts and the Transformer Trainer. Gain insights into hyperparameter search, GPU instances, and the challenges of model optimization. Hear Philipp's perspective on exciting developments in the field and learn about Optimum, a tool for model optimization. Get a glimpse into a day in the life of a Machine Learning Engineer and the future of NLP model democratization and productionization.
Overview
Syllabus
Introduction
What is SageMaker
Getting started
Preprocessing
S3 Integration
Hiking Phase Hub
Starting the Training
Accessing the Model
How SageMaker Works
Using Existing Training Scripts
Leveraging the Trainer from Transformers
Hyperparameter Search with SageMaker
What does a day in the life of Philip look like
Where do you see the most exciting developments in this space
What are the sort of challenges that happen when you want to make these models smaller
What is Optimum
GPU Instances
Taught by
Hugging Face
Related Courses
-
Working with Hugging Face
-
Train Your First NLP Transformer Model with Amazon SageMaker
-
Accelerating Transformers with Hugging Face Optimum and Infinity
-
Vision Transformers Explained + Fine-Tuning in Python
-
Building and Deploying Machine Learning Models with Amazon SageMaker
-
Simple Training with the Transformers Trainer
