Tutorial on Transformer model came out in ML from Scratch seminar series

Attention

Official Github repo in ML from Scratch seminar series. 

 

  • Lecture Slides (slides ):
    • NLP101
    • Transformer architecture
    • Training objective of language model and usage after training
    • Transformer beyond language
  • Exercise notebooks for the seminar 
    • Building your self-attention mechanism and transformer architecture (GPT2) from scratch (Open in Colab)
    • Pretrained Transformers (BERT, GPT) and training transformer on simplified language (Open in Colab
    • Notes on Einstein summation rule. (Open in Colab)