Get the latest tech news

Show HN: I built a tensor library from scratch in C++/CUDA


Tensor library & inference framework for machine learning - nirw4nna/dsc

DSC is a PyTorch-compatible tensor library and inference framework for machine learning models. Programs written using DSC can seamlessly switch between backends by simply adding a dsc.set_default_device('...') instruction, no changes needed. If that is not the case you have to manually update the Makefile or set the CUDA environment variable before calling make.

Get the Android app

Or read this on Hacker News

Read more on:

Photo of Scratch

Scratch

Photo of tensor library

tensor library

Photo of C++/CUDA

C++/CUDA

Related news:

News photo

I have reimplemented Stable Diffusion 3.5 from scratch in pure PyTorch

News photo

Show HN: I wrote a BitTorrent Client from scratch

News photo

Async from scratch 3: Pinned against the wall