NVlabs/tiny-cuda-nn
A high-performance C++/CUDA framework for training and querying neural networks with GPU acceleration.
Not currently ranked — collecting fresh signals.
star history
Tiny CUDA Neural Networks is a lightweight framework for training and querying neural networks on GPUs. It provides fully-fused multi-layer perceptrons and multiresolution hash encodings optimized for speed, supporting various input encodings, loss functions like L2, and optimizers like Adam. The framework exposes a C++/CUDA API and offers optional JIT fusion for additional performance gains.
Frequently asked
- What is NVlabs/tiny-cuda-nn?
- A high-performance C++/CUDA framework for training and querying neural networks with GPU acceleration.
- Is tiny-cuda-nn open source?
- Yes — NVlabs/tiny-cuda-nn is an open-source project tracked on heatdrop.
- What language is tiny-cuda-nn written in?
- NVlabs/tiny-cuda-nn is primarily written in C++.
- How popular is tiny-cuda-nn?
- NVlabs/tiny-cuda-nn has 4.5k stars on GitHub.
- Where can I find tiny-cuda-nn?
- NVlabs/tiny-cuda-nn is on GitHub at https://github.com/NVlabs/tiny-cuda-nn.