srush/Triton-Puzzles

Puzzles for learning Triton

45
/ 100
Emerging

This project helps AI engineers and researchers learn how to program GPUs efficiently using Triton, an open-source language. It provides interactive coding puzzles that start with basic concepts and progress to complex algorithms like Flash Attention and quantized neural networks. Learners input their Triton code solutions and receive feedback, helping them master low-level details crucial for high-performance AI systems.

2,332 stars. No commits in the last 6 months.

Use this if you are an AI developer looking to master Triton for high-performance GPU programming without needing immediate access to a GPU.

Not ideal if you are a data scientist or machine learning practitioner who primarily uses high-level frameworks and does not need to delve into low-level GPU programming.

GPU programming AI development Deep learning optimization Accelerator programming Performance engineering
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 19 / 25

How are scores calculated?

Stars

2,332

Forks

206

Language

Jupyter Notebook

License

Apache-2.0

Last pushed

Nov 18, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/srush/Triton-Puzzles"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.