Understanding-Transformer-Attention-Weights

This repository contains code to compute and visualise the attention weights across different layers and different heads.

Name		Name	Last commit message	Last commit date
Latest commit History 5 Commits
Compute_Attention_Weights_from_Pre_trained_Projection_Matrices.ipynb		Compute_Attention_Weights_from_Pre_trained_Projection_Matrices.ipynb
LICENSE		LICENSE
Masked_Multi_head_Attention_Weights.ipynb		Masked_Multi_head_Attention_Weights.ipynb
README.md		README.md
Visualisation_Transformer_Attention_Weights.ipynb		Visualisation_Transformer_Attention_Weights.ipynb

Provide feedback