Training library for Megatron-based models

Scalable data pre processing and curation toolkit for LLMs

Scalable toolkit for efficient model reinforcement

SEED-Voken: A Series of Powerful Visual Tokenizers

Learning to Compress Prompts with Gist Tokens - https://arxiv.org/abs/2304.08467

[Development suspended] Advanced open-source Texas Hold'em GTO solver with optimized performance (web browser version)

🌸 A command-line fuzzy finder

A high-performance Python-based I/O system for large (and small) deep learning problems, with strong support for PyTorch.

A playbook for systematically maximizing the performance of deep learning models.

📗 SheetJS Spreadsheet Data Toolkit -- New home https://git.sheetjs.com/SheetJS/sheetjs

MiniWoB++: a web interaction benchmark for reinforcement learning

Standard Go Project Layout

Figure sizes, font sizes, fonts, and more configurations at minimal overhead. Fix your journal papers, conference proceedings, and other scientific publications.

Unifying Variational Autoencoder (VAE) implementations in Pytorch (NeurIPS 2022)

A modular, primitive-first, python-first PyTorch library for Reinforcement Learning.

Parallel data preprocessing for NLP and ML.

An elegant PyTorch deep reinforcement learning library.

🚀 A simple way to launch, train, and use PyTorch models on almost any device and distributed configuration, automatic mixed precision (including fp8), and easy-to-configure FSDP and DeepSpeed support

Deep Learning project template best practices with Pytorch Lightning, Hydra, Tensorboard.

SEED RL: Scalable and Efficient Deep-RL with Accelerated Central Inference. Implements IMPALA and R2D2 algorithms in TF2 with SEED's architecture.

Python Classes Without Boilerplate

A concise but complete full-attention transformer with a set of promising experimental features from various papers

Flexible and powerful tensor operations for readable and reliable code (for pytorch, jax, TF and others)

Auriga is a minimalist LaTeX beamer presentation theme 📽

A tiny scalar-valued autograd engine and a neural net library on top of it with PyTorch-like API

Simple and easily configurable grid world environments for reinforcement learning

Shape and dimension inference (Keras-like) for PyTorch layers and neural networks

📊 Benchmark multiple object trackers (MOT) in Python

Change Python code while it's running without losing state