Data

Tools

Indexes

Applications

Experiments

Open Source Science

Mantik

Mantik: A Workflow Platform for the Development of Artificial Intelligence on High-Performance Computing Infrastructures - Published in JOSS (2024)

https://gitlab.com/mantik-ai/mantik

Science Score: 87.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ○
    CITATION.cff file
  • ○
    codemeta.json file
  • ○
    .zenodo.json file
  • ✓
    DOI references
    Found 1 DOI reference(s) in JOSS metadata
  • ○
    Academic publication links
  • ○
    Committers with academic emails
  • ○
    Institutional organization owner
  • ✓
    JOSS paper metadata
    Published in Journal of Open Source Software

Keywords

high-performance computing python

Scientific Fields

Mathematics Computer Science - 84% confidence
Artificial Intelligence and Machine Learning Computer Science - 83% confidence
Computer Science Computer Science - 65% confidence
Last synced: 4 months ago · JSON representation

Repository

Python API and CLI for remote submission and monitoring of Python applications on HPC.

Basic Info
  • Host: gitlab.com
  • Owner: mantik-ai
  • License: mit
  • Default Branch: main
Statistics
  • Stars: 0
  • Forks: 0
  • Open Issues: 21
  • Releases: 0
Topics
high-performance computing python
Created almost 3 years ago

https://gitlab.com/mantik-ai/mantik/blob/main/ 

# Mantik - Enhancing Machine Learning Development on HPC

Mantik allows to manage the full Machine Learning workflow on HPC by
supporting MLflow and deployment of applications to HPC via the [Mantik platform](https://cloud.mantik.ai).

For a quickstart see [our documentation](https://mantik-ai.gitlab.io/mantik/getting-started/index.html)

## Extra Dependencies

By default, mantik only installs the core dependencies required to use the Mantik API.
However, additional extras are provided and can be installed:

- `mlflow`: installs the supported MLflow version (i.e. `mlflow-skinny` due do its much smaller size compared to `mlflow`)

  ```shell
  pip install "mantik[mlflow]"
  ```

- `s3`: installs the required dependencies to use the [S3 Remote File Service](https://mantik-ai.gitlab.io/mantik/remote-execution/remote-file-service.html)

  ```shell
  pip install "mantik[s3]"
  ```

- `unicore`: installs the required dependencies to use the [UNICORE Remote File Service](https://mantik-ai.gitlab.io/mantik/remote-execution/remote-file-service.html)

  ```shell
  pip install "mantik[unicore]"
  ```

- `docker`: installs the required dependencies to use the Model Docker images

  ```shell
  pip install "mantik[docker]"
  ```

## Helpdesk

For bug reports or feature requests, please email our helpdesk.
Details can be found [here](https://mantik-ai.gitlab.io/mantik/helpdesk.html).

JOSS Publication

Mantik: A Workflow Platform for the Development of Artificial Intelligence on High-Performance Computing Infrastructures
Published
June 02, 2024
DOI
Volume 9, Issue 98, Page 6136
Authors
Thomas Seidler
Ambrosys GmbH, Potsdam, Germany, Institute of Physics and Astronomy, University of Potsdam, Germany
Fabian Emmerich
Institute of Physics and Astronomy, University of Potsdam, Germany, 4Cast GmbH & Co. KG, Potsdam, Germany
Kristian Ehlert ORCID
4Cast GmbH & Co. KG, Potsdam, Germany
Rico Berner ORCID
Ambrosys GmbH, Potsdam, Germany
Oliver Nagel-Kanzler
Ambrosys GmbH, Potsdam, Germany
Norbert Schultz
Reactive Core GmbH, Berlin, Germany
Markus Quade
Ambrosys GmbH, Potsdam, Germany
Martin G. Schultz ORCID
Jülich Supercomputing Centre (JSC), Forschungszentrum Jülich GmbH, Germany
Markus Abel
Ambrosys GmbH, Potsdam, Germany
Editor
Arfon Smith ORCID
Tags
machine learning mlflow high-performance computing MLOps
View PDF Review Thread Software Archive

Committers

Last synced: 4 months ago

All Time
  • Total Commits: 1,184
  • Total Committers: 22
  • Avg Commits per committer: 53.818
  • Development Distribution Score (DDS): 0.48
Past Year
  • Commits: 29
  • Committers: 7
  • Avg Commits per committer: 4.143
  • Development Distribution Score (DDS): 0.724
Top Committers
Name Email Commits
Fabian Emmerich f****h@4****e 616
Elia Boscaini e****i@a****e 166
sara.grau s****u@4****e 101
Thomas Seidler t****r@a****e 75
Jakub Jagielski j****i@a****e 53
Rico Berner r****r@a****e 50
Thomas Seidler t****r@4****e 45
Kristian Ehlert k****t@4****e 30
Omar Ahmed o****d@a****e 15
Pawel Boening p****g@a****e 8
Jakob Fuhrmann j****n@a****e 5
Nastassia Kundas n****s@a****e 5
ankur a****a@a****e 3
Frederic Ferrera f****a@a****e 3
Amanda Maricia Raphael a****l@a****e 2
Britta Oberlack b****k@a****e 1
Cedric Lorenz c****z@a****e 1
Cedric Lorenz c****z@s****e 1
Markus Abel m****l@a****e 1
Stefan Schinkel s****l@a****e 1
Zubaer Haque z****e@4****e 1
adhish.nigam a****m@4****e 1
Committer Domains (Top 20 + Academic)
ambrosys.de: 15 4-cast.de: 6 student.hpi.de: 1

Issues and Pull Requests

Last synced: 4 months ago

Packages

  • Total packages: 1
  • Total downloads:
    • pypi 54 last-month
  • Total dependent packages: 1
  • Total dependent repositories: 4
  • Total versions: 35
  • Total maintainers: 1
pypi.org: mantik

mantik for mlflow

  • Versions: 35
  • Dependent Packages: 1
  • Dependent Repositories: 4
  • Downloads: 54 Last month
Rankings
Dependent packages count: 4.8%
Dependent repos count: 7.5%
Average: 20.5%
Downloads: 21.5%
Forks count: 29.8%
Stargazers count: 38.8%
Maintainers (1)
mantik_ai
Last synced: 4 months ago

Dependencies

demo/mlproject/Dockerfile docker
  • mquade/poetry 3.8-1.2.1 build
src/tests/resources/test-project/Dockerfile docker
  • python 3.7-slim-buster build
docs/requirements.txt pypi
  • linkify-it-py ==2.0.0
  • myst-parser ==1.0.0
  • sphinx ==5.3.0
  • sphinx-autoapi ==2.0.1
  • sphinx-autobuild ==2021.3.14
  • sphinx-click ==4.4.0
  • sphinx-copybutton ==0.5.1
  • sphinx-design ==0.3.0
  • sphinx-material ==0.0.35
  • sphinx-tippy ==0.4.1
poetry.lock pypi
  • 106 dependencies
pyproject.toml pypi
  • fastapi ^0.78.0 develop
  • freezegun ^1.2.1 develop
  • pre-commit ^2.17.0 develop
  • pytest ^7.1.2 develop
  • pytest-cov ^3.0.0 develop
  • pytest-custom-exit-code ^0.3.0 develop
  • requests-mock ^1.9.3 develop
  • Flask ^2.1.2
  • PyYAML ^6.0
  • boto3 ^1.23.6
  • click ^8.1.3
  • fastapi ^0.78.0
  • mlflow 2.2.2
  • pydantic ^1.9.0
  • python ^3.8
  • python-jose ^3.3.0
  • python-multipart ^0.0.5
  • pyunicore ^0.9.16
  • requests ^2.27.1
  • uvicorn ^0.17.6