Build, run, manage agentic software at scale.

JHipster is a development platform to quickly generate, develop, & deploy modern web applications & microservice architectures.

All-in-one text de-duplication

An LLM playground you can run on your laptop

The SentiAspect-pt comprises 180 product reviews annotated according to implicit and explicit fine-grained opinions, which were hierarchically organized for aspect-based sentiment analysis and opin…

Plancraft is a minecraft environment and agent suite to test planning capabilities in LLMs

Python & Command-line tool to gather text and metadata on the Web: Crawling, scraping, extraction, output as CSV, JSON, HTML, MD, TXT, XML

List of resources and tools developed with focus on Portuguese.

A scalable benchmark for state representation learning in visual reinforcement learning.

Repositório de Fontes de Dados Financeiros

FactNews is the first dataset to predict sentence-level factuality of news reporting. Furthemore, we provide baseline results for sentence-level factuality and media bias predicition in Portuguese.…

0-Shot Tokenizer Transplant

Multi-Word Probabilistic based supertokenizer

Code and data to evaluate LLMs on the ENEM, the main standardized Brazilian university admission exams.

Magnificent app which corrects your previous console command.

List of resources and tools developed with focus on Portuguese.

Code for the paper "Getting the most out of your tokenizer for pre-training and domain adaptation"

A Massive Portuguese Crawling Corpus

A framework for few-shot evaluation of autoregressive language models.

A Python package to wrap a compiled soccer-twos reinforcement learning environment with gym-compatible interfaces.

Portuguese translation of the GLUE benchmark and Scitail dataset

1st Place Reinforcement Learning Competition - CEIA RL Course 2021

My solution for the MeLi Data Challenge 2021 - 3rd place

2nd place solution for the MeLi Challenge 2019