A framework for efficient model inference with omni-modality models

Qwen3-VL is the multimodal large language model series developed by Qwen team, Alibaba Cloud.

Cross-platform transformer training

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

Discrete Diffusion Forcing (D2F): dLLMs Can Do Faster-Than-AR Inference

Train speculative decoding models effortlessly and port them smoothly to SGLang serving.

A Datacenter Scale Distributed Inference Serving Framework

EasyR1: An Efficient, Scalable, Multi-Modality RL Training Framework based on veRL

FlashMLA: Efficient Multi-head Latent Attention Kernels

how to optimize some algorithm in cuda.

Entropy Based Sampling and Parallel CoT Decoding

documentation for content creation

CUDA Templates and Python DSLs for High-Performance Linear Algebra

A fast multimodal LLM for real-time voice

A high-throughput and memory-efficient inference and serving engine for LLMs

Manages vllm-nccl dependency

CUDA/Metal accelerated language model inference