godweiyang/NN-CUDA-Example
Examples showing how to write custom CUDA kernels and call them from PyTorch and TensorFlow for neural network operations.
Not currently ranked — collecting fresh signals.
star history
This repository provides examples of writing custom CUDA operators and integrating them with popular neural network toolkits. It demonstrates three compilation methods (JIT, setuptools, CMake) for building CUDA kernels and their C++ wrappers. The repository includes Python code for timing comparisons between custom CUDA kernels and framework operations, as well as model training examples using custom kernels. It covers both PyTorch and TensorFlow integration patterns.
Frequently asked
- What is godweiyang/NN-CUDA-Example?
- Examples showing how to write custom CUDA kernels and call them from PyTorch and TensorFlow for neural network operations.
- Is NN-CUDA-Example open source?
- Yes — godweiyang/NN-CUDA-Example is open source, released under the Apache-2.0 license.
- What language is NN-CUDA-Example written in?
- godweiyang/NN-CUDA-Example is primarily written in Python.
- How popular is NN-CUDA-Example?
- godweiyang/NN-CUDA-Example has 1.5k stars on GitHub.
- Where can I find NN-CUDA-Example?
- godweiyang/NN-CUDA-Example is on GitHub at https://github.com/godweiyang/NN-CUDA-Example.