Completed
- Setting up evaluation criteria and format requirements
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
Improving LLM Performance Through Evaluation and Few-Shot Examples - Part 2
Automatically move to the next video in the Classroom when playback concludes
- 1 - Introduction to LLM evals and their importance
- 2 - Overview of creating high-quality prompts from scratch
- 3 - Emphasis on importance of high-quality examples
- 4 - Introduction to systematic approach for creating examples
- 5 - Overview of demonstration using touch rugby example
- 6 - Introduction to LLM eval repo and UI demonstration
- 7 - Start of UI demonstration with pipeline creation
- 8 - Creating initial pipeline with Claude Sonnet
- 9 - Creating first dataset for touch rugby Q&A
- 10 - Setting up evaluation criteria and format requirements
- 11 - Demonstration of generating ground truth answers
- 12 - Creating second evaluation task
- 13 - Introduction to creating few-shot examples
- 14 - Setting up pipeline with few-shot examples
- 15 - Creating training examples for few-shot learning
- 16 - Demonstration of improved performance with few-shot examples
- 17 - Discussion of pipeline customization options
- 18 - Final tips on judges and evaluation
- 19 - Recommendations for managing examples
- 20 - Discussion of OpenAI's o1 model considerations
- 21 - Conclusion and future topics