mlcommons/training
Reference implementations of MLPerf® training benchmarks
This project provides standardized training benchmarks for machine learning models across various domains like language processing, image generation, and recommendation systems. It takes a specific dataset and a chosen model implementation as input, and outputs the time it takes to train that model to a target quality. It is used by deep learning engineers and researchers who want to objectively evaluate the training performance of different ML hardware and software setups.
1,748 stars. Actively maintained with 1 commit in the last 30 days.
Use this if you need a common, reproducible method to compare how quickly different machine learning systems can train a model to a specified level of accuracy.
Not ideal if you are looking for highly optimized, production-ready model implementations or if you need to perform 'real' performance measurements for production applications.
Stars
1,748
Forks
585
Language
Python
License
Apache-2.0
Category
Last pushed
Mar 12, 2026
Commits (30d)
1
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/mlcommons/training"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Compare
Related frameworks
NVIDIA/TransformerEngine
A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit and 4-bit...
mlcommons/inference
Reference implementations of MLPerf® inference benchmarks
datamade/usaddress
:us: a python library for parsing unstructured United States address strings into address components
GRAAL-Research/deepparse
Deepparse is a state-of-the-art library for parsing multinational street addresses using deep learning
CMU-SAFARI/Pythia
A customizable hardware prefetching framework using online reinforcement learning as described...