Google, IBM & Meta Certificates – 40% Off
One plan covers every Professional Certificate on Coursera.
Unlock All Certificates
Learn how to streamline custom training and open-source generative AI model finetuning using Databricks' newly announced Serverless GPU Compute in this 29-minute conference talk. Discover best practices for leveraging serverless GPU infrastructure to eliminate the overhead of managing GPU resources while accelerating AI development workflows. Explore framework support for popular libraries including LLM Foundry, Composer, and HuggingFace, and understand how to implement distributed GPU training using PyTorch and other frameworks within the Databricks ecosystem. Master techniques for training custom deep learning models for forecasting, recommendation systems, and personalization, while gaining insights into finetuning open-source generative AI models efficiently. Examine how MLFlow and the Databricks Lakehouse platform can streamline the end-to-end development process from model training to production deployment. Understand the cost and time-saving benefits of serverless GPU compute compared to traditional GPU infrastructure management, and learn practical strategies for optimizing your AI training workflows across the entire Databricks stack.
