Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

awesome-Auto-Parallelism

A baseline repository of Auto-Parallelism in Training Neural Networks
https://github.com/ConnollyLeon/awesome-Auto-Parallelism

Last synced: 6 days ago
JSON representation

  • Pipeline Parallelism or Inter-layer Model Parallelism only:

  • Data Parallelism + Pipeline Parallelism (or Inter-layer Model Parallelism):

    • mlr.press
    • nips
    • REGAL
    • arxiv
    • usenix
    • mdpi
    • RaNNC - scale neural networks. | DIRECT and University of Tokyo | [arxiv](http://arxiv.org/abs/2103.16063) | PyTorch | IPDPS 2021 | dynamic programming
    • HeterPS
    • PipeDream - 2BW | Microsoft Fiddle| [arxiv](https://arxiv.org/pdf/1806.03377.pdf), | PyTorch | 2018 on arxiv, SOSP 2019 | Dynamic Programming with Profile
    • DNN-partitioning - of-concept implementation | NIPS 2020 |Dynamic Programming and Integer Programming
    • DAPPLE
    • PipeTransformer
    • Chimera - scale neural networks with bidirectional pipelines | Department of Computer Science, ETH Zurich Switzerland | [dl.acm](https://dl.acm.org/doi/pdf/10.1145/3458817.3476145) | PyTorch | SC 2021 | Performance model with brute force
    • FTPipe - GPU one. | Technion-Israel Institute of Technology | [usenix](https://usenix.org/system/files/atc21-eliad.pdf) | PyTorch | 2021 | multiprocessor scheduling problem with profiling.
    • arxiv

  • Data Parallelism + Intra-layer Model Parallelism (or Tensor Parallelism):

    • TensorOpt - Parallelism | CUHK & Huawei | [arxiv](https://arxiv.org/pdf/2004.10856.pdf) | MindSpore | 2020 on arxiv | Dynamic Programming based graph search algorithm
    • Double Recursive - 3-030-85665-6_13) | MindSpore | Euro-Par 2021 | Double Recursive
    • arxiv
    • arxiv
    • AccPar
    • ROC - Paper.pdf) | On top of Flexflow | MLSys 2020 | uses a novel online linear regression model to achieve efficient graph partitioning, and introduces a dynamic programming algorithm to minimize data transfer cost.
    • PaSE - research/PaSE/raw/master/docs/PaSE_ipdps2021.pdf) | prototype | IPDPS 2021 | Dynamic Programming
    • FlexFlow
    • arxiv
    • arxiv

  • Data Parallelism + Model Parallelism (or Tensor Parallelism) + Pipeline Parallelism:

    • arxiv
    • GSPMD
    • arxiv - Karp Algorithm
    • arxiv - given.
    • Piper - of-concept implementation) and input files (profiled DNN models / workloads) from the paper "Piper: Multidimensional Planner for DNN Parallelization" published at NeurIPS 2021. An extension of DNN partitioning| Microsoft Fiddle| [link](https://www.microsoft.com/en-us/research/publication/piper-multidimensional-planner-for-dnn-parallelization/) | proof-of-concept implementation | NIPS 2021 | two-level dynamic programming
    • DistIR - Search Simulator
    • Alpa - and Intra-Operator Parallelism for Distributed Deep Learning | UC Berkley, Google, etc. | [arxiv](https://arxiv.org/pdf/2201.12023.pdf) | Jax, XLA | 2022 | Integer Linear for Intra, Dynamic programming for inter
    • arxiv
    • arxiv - given.

  • Other Interesting automatic work

Programming Languages
Python 10 C++ 6
Categories
Data Parallelism + Pipeline Parallelism (or Inter-layer Model Parallelism): 15 Data Parallelism + Intra-layer Model Parallelism (or Tensor Parallelism): 10 Data Parallelism + Model Parallelism (or Tensor Parallelism) + Pipeline Parallelism: 9 Pipeline Parallelism or Inter-layer Model Parallelism only: 7 Other Interesting automatic work 1
Sub Categories
Keywords
pipeline-parallelism 4 deep-learning 3 deep-neural-networks 2 distributed-training 2 checkpointing 1 gpipe 1 model-parallelism 1 parallelism 1 pytorch 1 distribution-strategy-planner 1 hybrid-parallelism 1 distributed-deep-learning 1 transformers 1 fine-tuning 1 nlp 1 t5 1 alpa 1 auto-parallelization 1 compiler 1 distributed-computing 1 high-performance-computing 1 jax 1 llm 1 machine-learning 1 inference-optimization 1