Large Language Models - Will They Keep Getting Bigger?

Large Language Models - Will They Keep Getting Bigger?

MIT Embodied Intelligence via YouTube Direct link

compute vs data

50 of 50

50 of 50

compute vs data

Class Central Classrooms beta

YouTube videos curated by Class Central.

Classroom Contents

Large Language Models - Will They Keep Getting Bigger?

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

  1. 1 Introduction
  2. 2 What are language models
  3. 3 Modern NLP
  4. 4 Scaling
  5. 5 sparse models
  6. 6 Gshard
  7. 7 Base Layers
  8. 8 Formal Optimization
  9. 9 Algorithmic Optimization
  10. 10 Experiments
  11. 11 Comparison
  12. 12 Benefits
  13. 13 Dmxlayers
  14. 14 Representations
  15. 15 Simple routing
  16. 16 Training time
  17. 17 Parallel training
  18. 18 Data curation
  19. 19 Unrealistic setting
  20. 20 Domain structure
  21. 21 Inference procedure
  22. 22 Perplexity numbers
  23. 23 Modularity
  24. 24 Remove experts
  25. 25 Summary
  26. 26 Generic language models
  27. 27 Hot dog example
  28. 28 Hot pan example
  29. 29 Common sense example
  30. 30 Large language models
  31. 31 The fundamental challenge
  32. 32 Surface form competition
  33. 33 Flip the reasoning
  34. 34 Key intuition
  35. 35 Noisey channel models
  36. 36 Finetuning
  37. 37 Scoring Strings
  38. 38 Web Crawls
  39. 39 Example Output
  40. 40 Structure Data
  41. 41 Efficiency
  42. 42 Questions
  43. 43 Density estimation
  44. 44 Better training objectives
  45. 45 Optimization
  46. 46 Probability
  47. 47 Induction
  48. 48 multimodality
  49. 49 outliers
  50. 50 compute vs data

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.