mlsys
[ICLR2025, ICML2025, NeurIPS2025 Spotlight] Quantized Attention achieves speedup of 2-5x compared to FlashAttention, without losing end-to-end metrics across language, image, and video models.
Measure and optimize the energy consumption of your AI applications!
NAACL '24 (Best Demo Paper RunnerUp) / MlSys @ NeurIPS '23 - RedCoast: A Lightweight Tool to Automate Distributed Training and Inference
Auto-graded exercises in low-level ML systems: real C++, real CUDA-C on a software GPU (no NVIDIA hardware), deterministic measured gates.
A scalable & efficient active learning/data selection system for everyone.
Machine Learning Systems: Foundations, Scaling, Agentic AI, and Physical AI (Vols I–IV) • Harvard CS249r | https://mlsysbook.ai
The RL Bridge for LLM-based Agent Applications. Made Simple & Flexible.
FedScale is a scalable and extensible open-source federated learning (FL) platform.
Audit GPU cluster communication schedules from NCCL logs. Zero dependencies. CI-ready.
A lightweight PyTorch-like library built from scratch for learning ML system design
