mlco2 / impact

ML has an impact on the climate. But not all models are born equal. Compute your model's emissions with our calculator and add the results to your paper with our generated latex template

This repository is passively maintained: issues will be addressed -- especially if they come with a suggested solution -- but there is no active updates to the data or code. Contributions are welcome.

Machine Learning's CO2 Impact

Checkout the online GPU emissions calculator!

By A. Lacoste, A. Luccioni, V. Schmidt

Read our paper on Quantifying Carbon Emissions of Machine Learning (NeurIPS 2019, Climate Change AI Workshop)

Use our generated latex template which automatically includes the Calculator's output for you to easily report your procedure's CO2 eq. emissions

Contributing

Setup

1. Install yarn: Node's package manager
2. Install gulp: a build tool
3. Install dependencies: from the root of this repo $ yarn install
4. Run the local server: $ gulp watch
5. Edit files! Gulp will watch for changes, build the differences and reload the browser

Content

html files are split by section in the html/ folder, and then built into the index.html file.

After editing content, if gulp watch was running, you're good ; otherwise, run gulp build to apply your changes.

Data

Anything to say or add? See data/

Acknowledgements

Special thanks to

• https://sharingbuttons.io/
• prism.js
• https://unsplash.com/