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

Code and Data for Tau-Bench

🌎💪 BrowserGym, a Gym environment for web task automation

Open sourced predictions, execution logs, trajectories, and results from model inference + evaluation runs on the SWE-bench task.

A curated list of of awesome UI agents resources, encompassing Web, App, OS, and beyond (continually updated)

A bug repository that keeps growing

aider is AI pair programming in your terminal

Estimate MFU for DeepSeekV3

Next Token Prediction Towards Multimodal Intelligence: A Comprehensive Survey

DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding

Maya: An Instruction Finetuned Multilingual Multimodal Model using Aya

A PyTorch library for implementing flow matching algorithms, featuring continuous and discrete flow matching implementations. It includes practical examples for both text and image modalities.

HunyuanVideo: A Systematic Framework For Large Video Generation Model

The simplest implementation of recent Sparse Attention patterns for efficient LLM inference.

The Open Cookbook for Top-Tier Code Large Language Model

The paper collections for the autoregressive models in vision.

This is the repository that contains the source code for the Self-Evaluation Guided MCTS for online DPO.

Replicating O1 inference-time scaling laws

The official CLIP training codebase of Inf-CL: "Breaking the Memory Barrier: Near Infinite Batch Size Scaling for Contrastive Loss". A super memory-efficiency CLIP training scheme.

Meta Lingua: a lean, efficient, and easy-to-hack codebase to research LLMs.

Official inference framework for 1-bit LLMs

Janus-Series: Unified Multimodal Understanding and Generation Models

Movie Gen Bench - two media generation evaluation benchmarks released with Meta Movie Gen

Scalable data pre processing and curation toolkit for LLMs