Deep learning library for F#. Provides tensor functionality, symbolic model differentiation, automatic differentiation and compilation to CUDA GPUs. It includes optimizers and model blocks used in deep learning.

Deep.Net is currently being ported to .NET Standard 2.0.

In the process the API is being streamlined and the documentation is being improved. The port of the numeric Tensor library is complete, but the port of the symbolic libraries is still in progress.

Go to the README of the Tensor library.