Explore a comprehensive conference talk on DSPy, a cutting-edge framework from Stanford for building and optimizing language model pipelines, specifically tailored for code generation. Join Krista Opsahl-Ongy, PhD candidate at Stanford, as she delves into the intricacies of DSPy, its applications, and its potential to revolutionize complex task automation. Learn about the importance of modular pipelines in AI, challenges in manual prompt engineering, and how to build language model programs with DSPy. Gain insights from a comparative study on optimizer performance and participate in a hands-on tutorial using DSPy with the HackerCup dataset. Discover how DSPy can enhance your code generation projects and streamline AI workflows through this informative 58-minute session, complete with a Q&A segment for deeper understanding.
DSPy for Code Generation: Building and Optimizing Language Model Pipelines
Overview
Syllabus
Meet Krista: PhD Candidate at Stanford
Overview of DSPy Framework for Code Generation
The Importance of Modular Pipelines in AI
Challenges in Manual Prompt Engineering
Building Language Model Programs with DSPy
Comparative Study: Optimizer Performance
Hands-On Tutorial: Using DSPy with HackerCup Dataset
Q&A and Final Thoughts
Taught by
Weights & Biases
Related Courses
-
Applications of Stanford DSPy for Self-Improving Language Model Pipelines
-
Fundamentals of AI Agents Using RAG and LangChain
-
DSPy: Building Self-Improving Language Model Pipelines - Beyond LangChain Templates
-
Understanding DSPy - Programming AI Pipelines for Language Models
-
DSPy Assertions: Computational Constraints for Self-Refining Language Model Pipelines
-
DSPy Tutorial - Programming Language Models with 8 Key Examples
