Runhui Huang

An open source framework that provides a simple, universal API for building distributed applications. Ray is packaged with RLlib, a scalable reinforcement learning library, and Tune, a scalable hyperparameter tuning library.

[ICML2026] Official Implementation of AlphaGRPO: Unlocking Self-Reflective Multimodal Generation in Unified Multimodal Models via Decompositional Verifiable Reward

No description.

Data and sample evaluation codes for Multimodal Rewardbench 2

No description.

用于小红书（xiaohongshu.com）的 Model Context Protocol（MCP）服务器与 CLI 工具，支持登录、发布、搜索、推荐等自动化能力

MCP for xiaohongshu.com

Collective communications library with various primitives for multi-machine training.

Example models using DeepSpeed

[CVPR 2023] Official Implementation of X-Decoder for generalized decoding for pixel, image and language

Using Low-rank adaptation to quickly fine-tune diffusion models.

Official repository of OFA (ICML 2022). Paper: OFA: Unifying Architectures, Tasks, and Modalities Through a Simple Sequence-to-Sequence Learning Framework

Repo for external large-scale work

预训练中的数据选取

PyTorch code for BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation

List of Dirty, Naughty, Obscene, and Otherwise Bad Words

🤗Transformers: State-of-the-art Natural Language Processing for Pytorch and TensorFlow 2.0.

No description.

Bazel wrapper around the pybind11 repository

Community Documents

No description.

Simple Distributed Deep Learning on TensorFlow

Fast parallel CTC.

Official Pytorch implementation of CutMix regularizer

Modular ROS visualization tool for 2D data.