A hands-on session on deep learning, large language models, and the engineering practices that carry AI systems from training to production.
Learning Goals:
- Build and train neural networks in PyTorch, and apply data augmentation to improve how well a model generalizes
- Explain how convolutional and transformer architectures are applied to computer vision, and when transfer learning is the right choice
- Understand how large language models are built, adapted, and applied
- Weigh quantization, pruning, and distillation against latency, cost, and hardware constraints, and package a model for reliable serving
- Design a data pipeline from ingestion through feature serving, with the validation, lineage, and drift monitoring that keep it trustworthy
- Plan how an AI system is operated and evolved after deployment, using MLOps principles and maturity levels as a guide