This README provides an abbreviated documentation of the DLIO code. Please refer to https://argonne-lcf.github.io/dlio_benchmark/ for full user documentation.

DLIO is an I/O benchmark for Deep Learning. DLIO is aimed at emulating the I/O behavior of various deep learning applications. The benchmark is delivered as an executable that can be configured for various I/O patterns. It uses a modular design to incorporate more data loaders, data formats, datasets, and configuration parameters. It emulates modern deep learning applications using Benchmark Runner, Data Generator, Format Handler, and I/O Profiler modules.

git clone https://github.com/argonne-lcf/dlio_benchmark
cd dlio_benchmark/
pip install -r requirements.txt
export PYTHONPATH=$PWD/:$PYTHONPATH
python ./src/dlio_benchmark.py ++workload.workflow.generate_data=True

Additionally, to generate the report iostat is needed and can be installed from the sysstat package using your package manager.

git clone https://github.com/argonne-lcf/dlio_benchmark
cd dlio_benchmark/
docker build -t dlio .
docker run -t dlio python ./src/dlio_benchmark.py ++workload.workflow.generate_data=True

You can also pull rebuilt container from docker hub (might not reflect the most recent change of the code):

docker docker.io/zhenghh04/dlio:latest
docker run -t docker.io/zhenghh04/dlio:latest python ./src/dlio_benchmark.py ++workload.workflow.generate_data=True

One can also run interactively inside the container

docker run -t docker.io/zhenghh04/dlio:latest /bin/bash
root@30358dd47935:/workspace/dlio$ python ./src/dlio_benchmark.py ++workload.workflow.generate_data=True

PowerPC requires installation through anaconda.

# Setup required channels
conda config --prepend channels conda-forge
conda config --prepend channels https://public.dhe.ibm.com/ibmdl/export/pub/software/server/ibm-ai/conda/

# create and activate environment
conda env create --prefix ./dlio_env_ppc --file environment-ppc.yaml --force
conda activate ./dlio_env_ppc

# install hydra dependency by source
mkdir external
cd external
git clone [email protected]:facebookresearch/hydra.git
cd hydra
python -m pip install .

A DLIO run is split in 3 phases:

• Generate synthetic data that DLIO will use
• Run the benchmark using the previously generated data
• Post-process the results to generate a report

The configurations of a workload can be specified through a yaml file. Examples of yaml files can be found in ./configs/workload/.

One can specify the workload through the workload= option on the command line. Specific configuration fields can then be overridden following the hydra framework convention (e.g. ++workload.framework=tensorflow).

First, generate the data

mpirun -np 8 python3 src/dlio_benchmark.py workload=unet3d ++workload.workflow.generate_data=True ++workload.workflow.train=False

If possible, one can flush the filesystem caches in order to properly capture device I/O

sudo sync && echo 3 | sudo tee /proc/sys/vm/drop_caches

Finally, run the benchmark with iostat profiling, listing the io devices you would like to trace.

mpirun -np 8 python3 src/dlio_benchmark.py workload=unet3d ++workload.workflow.profiling=True ++workload.profiling.profiler=iostat ++workload.profiling.iostat_devices=[sda,sdb]

All the outputs will be stored in hydra_log/unet3d/$DATE-$TIME folder. To post process the data, one can do

python3 src/dlio_postprocessor.py --output-folder hydra_log/unet3d/$DATE-$TIME

This will generate DLIO_$model_report.txt in the output folder.

Workload characteristics are specified by a YAML configuration file. Below is an example of a YAML file for the UNet3D workload which is used for 3D image segmentation.

# contents of unet3d.yaml
model: unet3d

framework: pytorch

workflow:
  generate_data: False
  train: True
  checkpoint: True

dataset: 
  data_folder: data/unet3d/
  format: npz
  num_files_train: 168
  num_samples_per_file: 1
  record_length: 234560851
  record_length_stdev: 109346892
  record_length_resize: 2097152
  
reader: 
  data_loader: pytorch
  batch_size: 4
  read_threads: 4
  file_shuffle: seed
  sample_shuffle: seed

train:
  epochs: 10
  computation_time: 1.3604

checkpoint:
  checkpoint_folder: checkpoints/unet3d
  checkpoint_after_epoch: 5
  epochs_between_checkpoints: 2
  model_size: 499153191

The full list of configurations can be found in: https://argonne-lcf.github.io/dlio_benchmark/config.html

The YAML file is loaded through hydra (https://hydra.cc/). The default setting are overridden by the configurations loaded from the YAML file. One can override the configuration through command line (https://hydra.cc/docs/advanced/override_grammar/basic/).

• DLIO currently assumes the samples to always be 2D images, even though one can set the size of each sample through --record_length. We expect the shape of the sample to have minimal impact to the I/O itself. This yet to be validated for case by case perspective. We plan to add option to allow specifying the shape of the sample.

• We assume the data/label pairs are stored in the same file. Storing data and labels in separate files will be supported in future.

• File format support: we only support tfrecord, hdf5, npz, csv, jpg, jpeg formats. Other data formats can be extended.

• Data Loader support: we support reading datasets using TensorFlow tf.data data loader, PyTorch DataLoader, and a set of custom data readers implemented in ./reader. For TensorFlow tf.data data loader, PyTorch DataLoader

  • We have complete support for tfrecord format in TensorFlow data loader.
  • For npz, jpg, jpeg, hdf5, we currently only support one sample per file case. In other words, each sample is stored in an independent file. Multiple samples per file case will be supported in future.

We welcome contributions from the community to the benchmark code. Specifically, we welcome contribution in the following aspects: General new features needed including:

• support for new workloads: if you think that your workload(s) would be interested to the public, and would like to provide the yaml file to be included in the repo, please submit an issue.
• support for new data loaders, such as DALI loader, MxNet loader, etc
• support for new frameworks, such as MxNet
• support for noval file systems or storage, such as AWS S3.
• support for loading new data formats.

If you would like to contribute, please submit an issue to https://github.com/argonne-lcf/dlio_benchmark/issues, and contact ALCF DLIO team, Huihuo Zheng at [email protected]

The original CCGrid'21 paper describes the design and implementation of DLIO code. Please cite this paper if you use DLIO for your research.

@article{devarajan2021dlio,
  title={DLIO: A Data-Centric Benchmark for Scientific Deep Learning Applications},
  author={H. Devarajan and H. Zheng and A. Kougkas and X.-H. Sun and V. Vishwanath},
  booktitle={IEEE/ACM International Symposium in Cluster, Cloud, and Internet Computing (CCGrid'21)},
  year={2021},
  volume={},
  number={81--91},
  pages={},
  publisher={IEEE/ACM}
}

We also encourage people to take a look at a relevant work from MLPerf Storage working group.

@article{balmau2022mlperfstorage,
  title={Characterizing I/O in Machine Learning with MLPerf Storage},
  author={O. Balmau},
  booktitle={SIGMOD Record DBrainstorming},
  year={2022},
  volume={51},
  number={3},
  publisher={ACM}
}

This work used resources of the Argonne Leadership Computing Facility, which is a DOE Office of Science User Facility under Contract DE-AC02-06CH11357 and is supported in part by National Science Foundation under NSF, OCI-1835764 and NSF, CSR-1814872.

Apache 2.0 LICENSE

Copyright (c) 2022, UChicago Argonne, LLC All Rights Reserved

If you have questions about your rights to use or distribute this software, please contact Argonne Intellectual Property Office at [email protected]

NOTICE. This Software was developed under funding from the U.S. Department of Energy and the U.S. Government consequently retains certain rights. As such, the U.S. Government has been granted for itself and others acting on its behalf a paid-up, nonexclusive, irrevocable, worldwide license in the Software to reproduce, distribute copies to the public, prepare derivative works, and perform publicly and display publicly, and to permit others to do so.