Session Outline
In this session at the Data Innovation Summit 2024, Thomas Capelle, Machine Learning Engineer at Weights and Biases, explores the trade-offs between using a model via an API and fine-tuning an open-source model. Using real use cases, this talk showcases how to create maximum performance while demonstrating how to achieve in an organized manner. We share some of the learnings and techniques you can apply on your own data to have a successful fine-tuning recipe.
Key Takeaways
- When and how to use API models
- Fine-tuning jobs – tricks and collaboration made easy
- What are the best models to use