Stars
TorchX is a universal job launcher for PyTorch applications. TorchX is designed to have fast iteration time for training/research and support for E2E production ML pipelines when you're ready.
Open standard for mathematical programming interoperability
Itoyori: A distributed multi-threading runtime system for global-view fork-join task parallelism
Python-MIP: collection of Python tools for the modeling and solution of Mixed-Integer Linear programs
Python implementation of Regulated Evolution Strategies with Covariance Matrix Adaption for continuous "Black-Box" optimization problems.
Kubernetes Operator for MPI-based applications (distributed training, HPC, etc.)
A high-throughput and memory-efficient inference and serving engine for LLMs
Simulated Annealing for the Multiple Choice Multidimensional Knapsack Problem
A full featured, fast Command Line Argument Parser for Rust
Plays Cookie Clicker for you using Cookie Monster
Chimera: Efficiently Training Large-Scale Neural Networks with Bidirectional Pipelines.
RDMA and SHARP plugins for nccl library
MSCCL++: A GPU-driven communication stack for scalable AI applications
Communication framework for RTL simulation and emulation.
Scalable Network Stack for FPGAs (TCP/IP, RoCEv2)
oneAPI Collective Communications Library (oneCCL)
Optimized primitives for collective multi-GPU communication
TACCL: Guiding Collective Algorithm Synthesis using Communication Sketches
FlexFlow Serve: Low-Latency, High-Performance LLM Serving
Data-driven APIs for common optimization tasks
Kedro is a toolbox for production-ready data science. It uses software engineering best practices to help you create data engineering and data science pipelines that are reproducible, maintainable,…