Resources for those studying the fast-moving pace of AI policy

Markdown memory system for you and your AI agent

Machine Learning Systems

A curated list of LLM datasets for African languages.

Library for fast text representation and classification. Fix compatibility with numpy 2

OSS Investor relationship hub for founders

Adaptive Softmax implementation for PyTorch

Master's Thesis. Open source and LaTeX.

LLM Eval leaderboard for African Languages

A scalable generative AI framework built for researchers and developers working on Large Language Models, Multimodal, and Speech AI (Automatic Speech Recognition and Text-to-Speech)

Bringing BERT into modernity via both architecture changes and scaling

Minimalistic 4D-parallelism distributed training framework for education purpose

Geographically-informed language identification

Whisperer provides an unopinionated approach for running multiple agents in Elixir

Benchmarking Large Language Models for FHIR

Direct Preference Optimization from scratch in PyTorch

Tools for merging pretrained large language models.

evolve llm training instruction, from english instruction to any language.

Crowd-sourced lists of urls to help Common Crawl crawl under-resourced languages. See https://github.com/commoncrawl/web-languages-code/ for the code

Model Merging in LLMs, MLLMs, and Beyond: Methods, Theories, Applications and Opportunities. ACM Computing Surveys, 2026.

PyTorch version of Stable Baselines, reliable implementations of reinforcement learning algorithms.

IA's public Wayback Machine (moved from SourceForge)

MedEmbed is a collection of embedding models fine-tuned specifically for medical and clinical data.

stoplists for African languages generated from the ASP corpus

Source stories from the African Storybook Project in Markdown format

maximal update parametrization (µP)

A Production-ready Reinforcement Learning AI Agent Library brought by the Applied Reinforcement Learning team at Meta.