Shypd.aiThe largest open index of AI toolsNN-CUDA-Example provides simple examples for neural network toolkits to call custom CUDA operators. It supports PyTorch and TensorFlow, demonstrating compilation methods like JIT, setuptools, and CMake.
Trending
About
NN-CUDA-Example offers a collection of straightforward examples for integrating custom CUDA operators into popular neural network toolkits such as PyTorch and TensorFlow. It illustrates various compilation methods for CUDA kernels and their C++ wrappers, including JIT compilation, setuptools, and CMake. The repository also provides Python code examples for invoking these custom CUDA kernels, complete with functionalities for kernel time statistics and model training. This resource is particularly valuable for developers looking to optimize their neural network operations by leveraging the power of custom CUDA implementations, offering practical insights into environment setup and code structure.
Capabilities
Pricing & Plans
Open Source
Free
FAQs
Trending
