Fine-Tuning LLMs: Best Practices and When to Go Small - Lecture 124

Fine-Tuning LLMs: Best Practices and When to Go Small - Lecture 124

MLOps.community via YouTube Direct link

[] LangChain Auto Evaluator evaluating LLMs

28 of 33

28 of 33

[] LangChain Auto Evaluator evaluating LLMs

Class Central Classrooms beta

YouTube videos curated by Class Central.

Classroom Contents

Fine-Tuning LLMs: Best Practices and When to Go Small - Lecture 124

Automatically move to the next video in the Classroom when playback concludes

  1. 1 [] Introduction to Mark Kim-Huang
  2. 2 [] Join the LLMs in Production Conference Part 2 on June 15-16!
  3. 3 [] Fine-Tuning LLMs: Best Practices and When to Go Small
  4. 4 [] Model approaches
  5. 5 [] You might think that you could just use OpenAI but only older base models are available
  6. 6 [] Why custom LLMs over closed source models?
  7. 7 [] Small models work well for simple tasks
  8. 8 [] Types of Fine-Tuning
  9. 9 [] Strategies for improving fine-tuning performance
  10. 10 [] Challenges
  11. 11 [] Define your task
  12. 12 [] Task framework
  13. 13 [] Defining tasks
  14. 14 [] Clustering task diversifies training data and improves out-of-domain performance
  15. 15 [] Prompt engineering
  16. 16 [] Constructing a prompt
  17. 17 [] Synthesize more data
  18. 18 [] Constructing a prompt
  19. 19 [] Increase fine-tuning efficiency with LoRa
  20. 20 [] Naive data parallelism with mixed precision is inefficient
  21. 21 [] Further reading on mixed precision
  22. 22 [] Parameter efficient fine-tuning with LoRa
  23. 23 [] LoRa Data Parallel with Mixed Precision
  24. 24 [] Summary
  25. 25 [] Q&A
  26. 26 [] Mark's journey to LLMs
  27. 27 [] Task clustering mixing with existing data sets
  28. 28 [] LangChain Auto Evaluator evaluating LLMs
  29. 29 [] Cloud platforms costs
  30. 30 [] Vector database used at Preemo
  31. 31 [] Finding a reasoning path of a model on Prompting
  32. 32 [] When to fine-tune versus prompting with a context window
  33. 33 [] Wrap up

Never Stop Learning.

Get personalized course recommendations, track subjects and courses with reminders, and more.

Someone learning on their laptop while sitting on the floor.