Streamlining reinforcement learning with RLOps. State-of-the-art RL algorithms and tools.

🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.

(Linear-chain) Conditional random field in PyTorch.

Minimal Interactive Attention Visualization

Open-Source Machine Translation Quality Estimation in PyTorch

Alternating Directions Dual Decomposition

Reformer, the efficient Transformer, in Pytorch

The entmax mapping and its loss, a family of sparse softmax alternatives.

Code for the paper "Language Models are Unsupervised Multitask Learners"

We introduce the Fake.Br Corpus, which is composed of aligned true and fake news written in Brazilian Portuguese.

Lisbon Machine Learning Summer School Lab Guide

A multilingual dependency parser based on linear programming relaxations.

COMBO is jointly trained tagger, lemmatizer and dependency parser.

Universal segmenter based on the Universal Dependency framework, written by Y. Shao, Uppsala University

Tree edit distance using the Zhang Shasha algorithm

CoNLL-U Parser parses a CoNLL-U formatted string into a nested python dictionary.

BiMPM: Bilateral Multi-Perspective Matching for Natural Language Sentences