Fine-Tuning Transformers with Hugging Face

About this course

Master the complete workflow for fine-tuning transformer models using the Hugging Face ecosystem. This hands-on course takes you from navigating the Hugging Face Hub to deploying production-ready models.You'll start by learning to discover, evaluate, and select models and datasets from the Hub's vast repository. Then you'll build practical skills in loading and preprocessing data, including streaming techniques for datasets too large to fit in memory.The core of the course focuses on fine-tuning transformers using the Trainer API. You'll implement custom callbacks, configure training optimizations like mixed precision, and develop comprehensive evaluation pipelines with metrics including accuracy, F1, precision, and recall.The capstone project ties everything together: you'll build an end-to-end sentiment analysis system, from data preprocessing and augmentation through training, evaluation, and publishing your model to Hugging Face Hub with professional documentation.By course end, you'll have hands-on experience with the same tools and workflows used by ML teams at leading organizations, plus a published model in your portfolio.

Course objectives

• master the end-to-end workflow of transformers
• create a sentiment analysis system
• publish a model on Hugging Face Hub