WeDLM: The fastest diffusion language model with standard causal attention and native KV cache compatibility, delivering real speedups over vLLM-optimized baselines.

AirLLM 70B inference with single 4GB GPU

A throughput-oriented high-performance serving framework for LLMs

[ICML 2025] LaCache: Ladder-Shaped KV Caching for Efficient Long-Context Modeling of Large Language Models

ToolOrchestra is an end-to-end RL training framework for orchestrating tools and agentic workflows.

An Extensible Toolkit for Finetuning and Inference of Large Foundation Models. Large Models for All.

A CLI to estimate inference memory requirements for Hugging Face models, written in Python.

CUGA is an open-source generalist agent for the enterprise, supporting complex task execution on web and APIs, OpenAPI/MCP integrations, composable architecture, reasoning modes, and policy-aware f…

A PyTorch native platform for training generative AI models

A compact implementation of SGLang, designed to demystify the complexities of modern LLM serving systems.

[ICLR'25] BigCodeBench: Benchmarking Code Generation Towards AGI

[ICLR 2024] Efficient Streaming Language Models with Attention Sinks

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

[NeurIPS'23] H2O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Models.

[MLSys'25] QServe: W4A8KV4 Quantization and System Co-design for Efficient LLM Serving; [MLSys'25] LServe: Efficient Long-sequence LLM Serving with Unified Sparse Attention

A low-latency & high-throughput serving engine for LLMs

A framework for few-shot evaluation of language models.

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

A compilation of the best multi-agent papers

Recipe for a General, Powerful, Scalable Graph Transformer

[Preprint] Graph State Space Convolution (GSSC)

Chimera: State Space Models Beyond Sequences

Official implementation of "Hydra: Bidirectional State Space Models Through Generalized Matrix Mixers"

code for the paper "Heta: Distributed Training of Heterogeneous Graph Neural Networks"