autograd
Autograd automatically differentiates native Torch code, enabling automatic gradient computation for machine learning models.
What is autograd?
Autograd is an automatic differentiation library for Torch that computes gradients of native Torch code. It allows developers to define machine learning models using standard Torch operations and automatically obtain gradients for training, supporting dynamic computation graphs and higher-order derivatives.
Machine learning researchers and developers using Torch who need automatic gradient computation for training neural networks and other differentiable models, especially those working with dynamic graphs or requiring Hessian calculations.
Autograd provides a seamless and efficient way to perform automatic differentiation in Torch, with full integration into the existing ecosystem, optimized performance via code generation, and support for complex, dynamic models that change at each forward pass.
Overview
Autograd automatically differentiates native Torch code
Use Cases
Best For
- Training neural networks with dynamic computation graphs that change each iteration
Related Projects
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
