Transformers from Scratch - Part 2: Building and Training a Weather Prediction Model

Transformers from Scratch - Part 2: Building and Training a Weather Prediction Model

Trelis Research via YouTube Direct link

Transformer Architecture Initialisation

8 of 15

8 of 15

Transformer Architecture Initialisation

Class Central Classrooms beta

YouTube videos curated by Class Central.

Classroom Contents

Transformers from Scratch - Part 2: Building and Training a Weather Prediction Model

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

  1. 1 Welcome and Link to Colab Notebook
  2. 2 Encoder versus Decoder Architectures
  3. 3 What is the GPT-4o architecture?
  4. 4 Recap of transformer for weather prediction
  5. 5 Pre layer norm versus post layer norm
  6. 6 RoPE vs Sinusoidal Positional Embeddings
  7. 7 Dummy Data Generation
  8. 8 Transformer Architecture Initialisation
  9. 9 Forward pass test
  10. 10 Training loop setup and test on dummy data
  11. 11 Weather data import
  12. 12 Training and Results Visualisation
  13. 13 Can the model predict the weather?
  14. 14 Is volatility in the loss graph a problem?
  15. 15 How to improve the model further?

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.