Crate dfdx

Ergonomics & safety focused deep learning in Rust. Main features include:

1. Tensor library, complete with const generic shapes, activation functions, and more.
2. Safe & Easy to use neural network building blocks.
3. Standard deep learning optimizers such as Sgd and Adam.
4. Reverse mode auto differentiation[1] implementation.

Modules

arrays

Collection of traits to describe Nd arrays.

devices

Provides implementations for modifying Nd arrays on the CPU.

gradients

Implementations of Gradient tapes and generic gradient containers.

losses

Standard loss functions such as mse, mae, cross entropy, and more.

nn

Neural network building blocks such as Linear and impls for tuples as feedforward networks.

numpy

Provides some generic functions to save Nd arrays in the .npy format.

optim

Optimizers such as Sgd and Adam that can optimize neural networks.

prelude

Contains all public exports.

tensor

The struct definitions for all TensorXD, Tensor trait, and more.

tensor_ops

Implementations of all operations for tensors, including activations, binary operations, and other methods.

unique_id