Disco Core

Facilitates machine learning development by providing foundational components like tensors, activation functions, layers, networks, loss functions, and optimizers. Supports building and training neural networks with features such as tensor pooling, various activation functions (ReLU, Logistic, Tanh), dense and convolutional layers, sequential and residual networks, and optimizers like gradient descent and Adam. Includes trainers for sequential and computation graph-based networks, aiming to simplify the creation and training of complex machine learning models.