Phoenix8215/BuildCudaNeuralNetworkFromScratch

Build CUDA Neural Network From Scratch

18
/ 100
Experimental

This project helps deep learning engineers understand and build neural networks that run exceptionally fast on NVIDIA GPUs. It takes raw image data, specifically handwritten digits from the MNIST dataset, and processes it through a custom-built neural network. The output is a trained network capable of classifying these digits, offering insights into high-performance computing for AI. This is ideal for deep learning researchers and practitioners.

No commits in the last 6 months.

Use this if you are a deep learning engineer or researcher looking to learn the intricacies of building high-performance neural networks from scratch using CUDA.

Not ideal if you need a high-level deep learning framework (like TensorFlow or PyTorch) for quick model development, or if you don't have experience with CUDA programming.

deep-learning-engineering GPU-acceleration neural-network-development high-performance-computing AI-research
No License Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 6 / 25
Maturity 8 / 25
Community 4 / 25

How are scores calculated?

Stars

22

Forks

1

Language

Cuda

License

Last pushed

Aug 28, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/Phoenix8215/BuildCudaNeuralNetworkFromScratch"

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