Arena-Hard-Auto: An automatic LLM benchmark.

Paper collections of multi-modal LLM for Math/STEM/Code.

MiMo-Audio: Audio Language Models are Few-Shot Learners

A large-scale knowledge repository for adaptive learning, learning analytics, and knowledge discovery in MOOCs, hosted by THU KEG.

End-to-End Reinforcement Learning for Multi-Turn Tool-Integrated Reasoning

Wan: Open and Advanced Large-Scale Video Generative Models

VeOmni: Scaling Any Modality Model Training with Model-Centric Distributed Recipe Zoo

SoTA open-source TTS

🤗 R1-AQA Model: mispeech/r1-aqa

Collect every awesome work about r1!

Official implementation of BLIP3o-Series

The official repository of the dots.vlm1 instruct models proposed by rednote-hilab.

gpt-oss-120b and gpt-oss-20b are two open-weight language models by OpenAI

State-of-the-art TTS model under 25MB 😻

Dream 7B, a large diffusion language model

Awesome Unified Multimodal Models

The author's implementation of FUDOKI, a multimodal large language model purely based on discrete flow matching.

My learning notes/codes for ML SYS.

Hierarchical Reasoning Model Official Release

🚀 Efficient implementations of state-of-the-art linear attention models

Text-audio foundation model from Boson AI

[ICLR 2025] MoE++: Accelerating Mixture-of-Experts Methods with Zero-Computation Experts

An Easy-to-use, Scalable and High-performance RLHF Framework based on Ray (PPO & GRPO & REINFORCE++ & vLLM & Ray & Dynamic Sampling & Async Agentic RL)

Tencent Hunyuan A13B (short as Hunyuan-A13B), an innovative and open-source LLM built on a fine-grained MoE architecture.