Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Join this 55-minute AI seminar from Neural Magic's "Random Samples" series where experts introduce SDG Hub, an open-source toolkit developed at Red Hat for customizing language models through synthetic data generation. Learn what synthetic data means for LLMs and explore SDG Hub's architecture including prompts, blocks, and flows. Discover strategies for selecting appropriate teacher models based on specific use cases like reasoning or translation. Follow along with two practical demonstrations: building a document-grounded skill using pre-built pipelines and customizing a reasoning model by creating new components. The session concludes with a demonstration of SDG Hub's new graphical interface that enables non-experts to visually construct synthetic data pipelines. Access the original InstructLab Paper at https://arxiv.org/abs/2403.01081 and explore the code repository at https://github.com/Red-Hat-AI-Innovation-Team/sdg_hub.
