SGLang is a high-performance serving framework for large language models and multimodal models.

Use PEFT or Full-parameter to CPT/SFT/DPO/GRPO 600+ LLMs (Qwen3.6, DeepSeek-V4, GLM-5.1, InternLM3, Llama4, ...) and 300+ MLLMs (Qwen3-VL, Qwen3-Omni, InternVL3.5, Ovis2.5, GLM4.5v, Gemma4, Llava, …

DeepEP: an efficient expert-parallel communication library

FlashMLA: Efficient Multi-head Latent Attention Kernels

Optimized primitives for collective multi-GPU communication

This repository contains the results and code for the MLPerf™ Training v2.0 benchmark.

Brings SQL and AI together.

a fast and user-friendly runtime for transformer inference (Bert, Albert, GPT2, Decoders, etc) on CPU and GPU.

Jittor is a high-performance deep learning framework based on JIT compiling and meta-operators.