title: Hexviz | |
emoji: 👁️🧬 | |
colorFrom: green | |
colorTo: purple | |
sdk: streamlit | |
sdk_version: 1.17.0 | |
python_version: 3.10.5 | |
app_file: ./hexviz/🧬Attention_Visualization.py | |
pinned: true | |
tags: | |
- protein language models | |
- attention analysis | |
- protein structure | |
- biology | |
# hexviz | |
Visualize attention pattern on 3D protein structures | |
## Install and run | |
```shell | |
poetry install | |
poetry run streamlit run hexviz/streamlit/Attention_On_Structure.py | |
``` | |
## Export dependecies from poetry | |
Spaces [require](https://huggingface.co/docs/hub/spaces-dependencies#adding-your-own-dependencies) dependencies in a `requirements.txt` file. Export depencies from poetry's `pyproject.toml` file with: | |
```shell | |
poetry export -f requirements.txt --output requirements.txt --without-hashes | |
``` | |
## Acknowledgements | |
This project builds on the attention visualization introduced and developed in | |
https://github.com/salesforce/provis#provis-attention-visualizer | |