A safetensors extension to efficiently store sparse quantized tensors on disk

[FPGA'26 Best Paper Nomination] CXL-SpecKV: A Disaggregated FPGA Speculative KV-Cache for Datacenter LLM Serving

Fully Open Framework for Democratized Multimodal Reinforcement Learning.

Official PyTorch Implementation of "Diffusion Transformers with Representation Autoencoders"

Official repo of Promoting Efficient Reasoning with Verifiable Stepwise Reward

Fully Open Framework for Democratized Multimodal Training

Official implementation of "DPad: Efficient Diffusion Language Models with Suffix Dropout"

SDAR (Synergy of Diffusion and AutoRegression), a large diffusion language model（1.7B, 4B, 8B, 30B）

Use PEFT or Full-parameter to CPT/SFT/DPO/GRPO 600+ LLMs (Qwen3, Qwen3-MoE, DeepSeek-R1, GLM4.5, InternLM3, Llama4, ...) and 300+ MLLMs (Qwen3-VL, Qwen3-Omni, InternVL3.5, Ovis2.5, GLM4.5v, Llava, …

Easy and Efficient dLLM Fine-Tuning

Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels

Janus-Series: Unified Multimodal Understanding and Generation Models

Defeating the Training-Inference Mismatch via FP16

The open-source materials for paper "Sparsing Law: Towards Large Language Models with Greater Activation Sparsity".

Official PyTorch implementation and models for paper "Diffusion Beats Autoregressive in Data-Constrained Settings". We find diffusion models are significantly more data-efficient than standard left…

The Entropy Mechanism of Reinforcement Learning for Large Language Model Reasoning.

A Survey of Reinforcement Learning for Large Reasoning Models

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

MMaDA - Open-Sourced Multimodal Large Diffusion Language Models

Dream 7B, a large diffusion language model

ScaleRL Curve Fitting

Automatic Video Generation from Scientific Papers

Official Jax Implementation of MD4 Masked Diffusion Models

[Arxiv] Discrete Diffusion in Large Language and Multimodal Models: A Survey

Contexts Optical Compression

Falcon is a continuously-evolving, high-quality benchmark for natural-language-to-SQL (Text2SQL) systems.

A high-performance kernel library for LLM training

dInfer: An Efficient Inference Framework for Diffusion Language Models