Metatheory.jl is a general purpose metaprogramming and algebraic computation library for the Julia programming language, designed to take advantage of the powerful reflection capabilities to bridge the gap between symbolic mathema
NLP progress in Fintech. A repository to track the progress in Natural Language Processing (NLP) related to the domain of Finance, including the datasets, papers, and current state-of-the-art results for the most popular tasks.
This library contains JAX and Pytorch implementations of neural ODEs and Bayesian layers for stochastic variational inference. A rudimentary JAX implementation of differentiable SDE solvers is also provided, refer to torchsde 
Lambda Networks from the paper LambdaNetworks: Modeling Long-Range Interactions Without Attention by Bello et al. Adds support for Rank 3 and Rank 5 tensor inputs as well as initial implementation of ResNet-D with Lambda Convoluti
This repository provides a minimal implementation of adaptive gradient clipping (AGC) (as proposed in High-Performance Large-Scale Image Recognition Without Normalization1) in TensorFlow 2. The paper attributes AGC as a crucial co
Unchained is a fully type safe, compile time only units library. There is absolutely no performance loss over pure float based code (aside from insertion of possible conversion factors, but those would have to be written by hand o