metadata

title: ML.ENERGY Leaderboard
emoji: ⚡
python_version: '3.9'
app_file: app.py
sdk: gradio
sdk_version: 3.39.0
pinned: true
tags:
  - energy
  - leaderboard

ML.ENERGY Leaderboard

How much energy do GenAI models like LLMs and Diffusion models consume?

This README focuses on explaining how to run the benchmark yourself. The actual leaderboard is here: https://ml.energy/leaderboard.

Read our paper here!

Repository Organization

leaderboard/
├── benchmark/      # Benchmark scripts & instructions
├── data/           # Benchmark results
├── deployment/     # Colosseum deployment files
├── spitfight/      # Python package for the Colosseum
├── app.py          # Leaderboard Gradio app definition
└── index.html      # Embeds the leaderboard HuggingFace Space

Colosseum

We instrumented Hugging Face TGI so that it measures and returns GPU energy consumption. Then, our controller server receives user prompts from the Gradio app, selects two models randomly, and streams model responses back with energy consumption.

Running the Benchmark

We open-sourced the entire benchmark with instructions here: ./benchmark

Citation

Please refer to our BibTeX file: citation.bib.