H-Huang / torch_collective_extensionLinks
A minimum demo for PyTorch distributed extension functionality for collectives.
☆12Updated 10 months ago
Alternatives and similar repositories for torch_collective_extension
Users that are interested in torch_collective_extension are comparing it to the libraries listed below
Sorting:
- ☆12Updated 2 months ago
- Hi-Speed DNN Training with Espresso: Unleashing the Full Potential of Gradient Compression with Near-Optimal Usage Strategies (EuroSys '2…☆15Updated last year
- Cupcake: A Compression Scheduler for Scalable Communication-Efficient Distributed Training (MLSys '23)☆9Updated last year
- TACCL: Guiding Collective Algorithm Synthesis using Communication Sketches☆73Updated last year
- ☆14Updated 3 years ago
- ☆23Updated 11 months ago
- PerFlow-AI is a programmable performance analysis, modeling, prediction tool for AI system.☆19Updated last month
- Managed collective communication service☆22Updated 9 months ago
- ☆23Updated 2 years ago
- ☆34Updated last year
- SHADE: Enable Fundamental Cacheability for Distributed Deep Learning Training☆35Updated 2 years ago
- ☆19Updated 2 weeks ago
- Official repository for the paper DynaPipe: Optimizing Multi-task Training through Dynamic Pipelines☆19Updated last year
- Compiler for Dynamic Neural Networks☆46Updated last year
- ☆22Updated last year
- TiledLower is a Dataflow Analysis and Codegen Framework written in Rust.☆14Updated 7 months ago
- ☆54Updated 4 years ago
- ☆25Updated last year
- ☆49Updated 6 months ago
- REEF is a GPU-accelerated DNN inference serving system that enables instant kernel preemption and biased concurrent execution in GPU sche…☆94Updated 2 years ago
- Artifacts for our ASPLOS'23 paper ElasticFlow☆52Updated last year
- Artifacts for our SIGCOMM'22 paper Muri☆42Updated last year
- ☆18Updated last year
- RPCNIC: A High-Performance and Reconfigurable PCIe-attached RPC Accelerator [HPCA2025]☆11Updated 6 months ago
- Proteus: A High-Throughput Inference-Serving System with Accuracy Scaling☆13Updated last year
- Thunder Research Group's Collective Communication Library☆37Updated last year
- Bamboo is a system for running large pipeline-parallel DNNs affordably, reliably, and efficiently using spot instances.☆50Updated 2 years ago
- ☆50Updated 2 years ago
- ☆16Updated last year
- Open-source implementation for "Helix: Serving Large Language Models over Heterogeneous GPUs and Network via Max-Flow"☆49Updated 7 months ago