maro/maro
Jinyu-W 297709736c
Add DDQN (#598)
* Refine explore strategy, add prioritized sampling support; add DDQN example; add DQN test (#590)

* Runnable. Should setup a benchmark and test performance.

* Refine logic

* Test DQN on GYM passed

* Refine explore strategy

* Minor

* Minor

* Add Dueling DQN in CIM scenario

* Resolve PR comments

* Add one more explanation

* fix env_sampler eval info list issue

* update version to 0.3.2a4

---------

Co-authored-by: Huoran Li <huoranli@microsoft.com>
Co-authored-by: Jinyu Wang <Wang.Jinyu@microsoft.com>
2023-10-27 14:12:46 +08:00
..
backends Format code on `master` (black + isort) (#538) 2022-06-14 07:37:59 +08:00
cli V0.3: Upgrade RL Workflow; Add RL Benchmarks; Update Package Version (#588) 2023-03-30 09:57:25 +08:00
communication Format code on `master` (black + isort) (#538) 2022-06-14 07:37:59 +08:00
data_lib Add local file config in isort & Add missing copyright info (#547) 2022-06-14 14:47:47 +08:00
event_buffer Merge V0.3 into master: update decision event logic & rl component bundle (#569) 2022-12-27 17:15:46 +08:00
forecasting Add local file config in isort & Add missing copyright info (#547) 2022-06-14 14:47:47 +08:00
rl Add DDQN (#598) 2023-10-27 14:12:46 +08:00
simulator Merge V0.3 into master: update decision event logic & rl component bundle (#569) 2022-12-27 17:15:46 +08:00
streamit V0.3: Upgrade RL Workflow; Add RL Benchmarks; Update Package Version (#588) 2023-03-30 09:57:25 +08:00
utils Add local file config in isort & Add missing copyright info (#547) 2022-06-14 14:47:47 +08:00
vector_env Format code on `master` (black + isort) (#538) 2022-06-14 07:37:59 +08:00
README.rst MARO v0.3: a new design of RL Toolkit, CLI refactorization, and corresponding updates. (#539) 2022-06-01 15:05:22 +08:00
__init__.py Format code on `master` (black + isort) (#538) 2022-06-14 07:37:59 +08:00
__misc__.py Add DDQN (#598) 2023-10-27 14:12:46 +08:00
common.py Merge V0.3 into master: update decision event logic & rl component bundle (#569) 2022-12-27 17:15:46 +08:00
requirements.build.txt Merge V0.3 into master: update decision event logic & rl component bundle (#569) 2022-12-27 17:15:46 +08:00

README.rst


.. image:: https://img.shields.io/pypi/l/pymaro
   :target: https://github.com/microsoft/maro/blob/master/LICENSE
   :alt: License


.. image:: https://raw.githubusercontent.com/microsoft/maro/master/docs/source/images/badges/platform.svg
   :target: https://pypi.org/project/pymaro/
   :alt: Platform


.. image:: https://img.shields.io/pypi/pyversions/pymaro.svg?logo=python&logoColor=white
   :target: https://pypi.org/project/pymaro/#files
   :alt: Python Versions


.. image:: https://img.shields.io/github/languages/code-size/microsoft/maro
   :target: https://github.com/microsoft/maro
   :alt: Code Size


.. image:: https://img.shields.io/docker/image-size/maro2020/maro
   :target: https://hub.docker.com/repository/docker/maro2020/maro/tags?page=1
   :alt: Docker Size


.. image:: https://img.shields.io/github/issues/microsoft/maro
   :target: https://github.com/microsoft/maro/issues
   :alt: Issues


.. image:: https://img.shields.io/github/issues-pr/microsoft/maro
   :target: https://github.com/microsoft/maro/pulls
   :alt: Pull Requests


.. image:: https://img.shields.io/librariesio/github/microsoft/maro
   :target: https://libraries.io/pypi/pymaro
   :alt: Dependencies


.. image:: https://github.com/microsoft/maro/workflows/test/badge.svg
  :target: https://github.com/microsoft/maro/actions?query=workflow%3Atest
  :alt: test


.. image:: https://github.com/microsoft/maro/workflows/build/badge.svg
  :target: https://github.com/microsoft/maro/actions?query=workflow%3Abuild
  :alt: build


.. image:: https://github.com/microsoft/maro/workflows/docker/badge.svg
   :target: https://hub.docker.com/repository/docker/maro2020/maro
   :alt: docker


.. image:: https://readthedocs.org/projects/maro/badge/?version=latest
  :target: https://maro.readthedocs.io/
  :alt: docs


.. image:: https://img.shields.io/pypi/v/pymaro
   :target: https://pypi.org/project/pymaro/#files
   :alt: PypI Versions


.. image:: https://img.shields.io/pypi/wheel/pymaro
   :target: https://pypi.org/project/pymaro/#files
   :alt: Wheel


.. image:: https://raw.githubusercontent.com/microsoft/maro/master/docs/source/images/badges/citi_bike.svg
   :target: https://maro.readthedocs.io/en/latest/scenarios/citi_bike.html
   :alt: Citi Bike


.. image:: https://raw.githubusercontent.com/microsoft/maro/master/docs/source/images/badges/cim.svg
   :target: https://maro.readthedocs.io/en/latest/scenarios/container_inventory_management.html
   :alt: CIM


.. image:: https://raw.githubusercontent.com/microsoft/maro/master/docs/source/images/badges/vm_scheduling.svg
   :target: https://maro.readthedocs.io/en/latest/scenarios/vm_scheduling.html
   :alt: VM Scheduling


.. image:: https://img.shields.io/gitter/room/microsoft/maro
   :target: https://gitter.im/Microsoft/MARO#
   :alt: Gitter


.. image:: https://raw.githubusercontent.com/microsoft/maro/master/docs/source/images/badges/stack_overflow.svg
   :target: https://stackoverflow.com/questions/ask?tags=maro
   :alt: Stack Overflow


.. image:: https://img.shields.io/github/release-date-pre/microsoft/maro
   :target: https://github.com/microsoft/maro/releases
   :alt: Releases


.. image:: https://img.shields.io/github/commits-since/microsoft/maro/latest/master
   :target: https://github.com/microsoft/maro/commits/master
   :alt: Commits


.. image:: https://github.com/microsoft/maro/workflows/vulnerability%20scan/badge.svg
   :target: https://github.com/microsoft/maro/actions?query=workflow%3A%22vulnerability+scan%22
   :alt: Vulnerability Scan


.. image:: https://github.com/microsoft/maro/workflows/lint/badge.svg
   :target: https://github.com/microsoft/maro/actions?query=workflow%3Alint
   :alt: Lint


.. image:: https://img.shields.io/codecov/c/github/microsoft/maro
   :target: https://codecov.io/gh/microsoft/maro
   :alt: Coverage


.. image:: https://img.shields.io/pypi/dm/pymaro
   :target: https://pypi.org/project/pymaro/#files
   :alt: Downloads


.. image:: https://img.shields.io/docker/pulls/maro2020/maro
   :target: https://hub.docker.com/repository/docker/maro2020/maro
   :alt: Docker Pulls


.. image:: https://raw.githubusercontent.com/microsoft/maro/master/docs/source/images/badges/play_with_maro.svg
   :target: https://hub.docker.com/r/maro2020/maro
   :alt: Play with MARO



.. image:: https://github.com/microsoft/maro/blob/master/docs/source/images/logo.svg
   :target: https://maro.readthedocs.io/en/latest/
   :alt: MARO LOGO

================================================================================================================


.. image:: https://raw.githubusercontent.com/microsoft/maro/master/docs/source/images/badges/vm_scheduling.svg
  :target: https://maro.readthedocs.io/en/latest/scenarios/vm_scheduling.html
  :alt: VM Scheduling


.. image:: https://img.shields.io/gitter/room/microsoft/maro
  :target: https://gitter.im/Microsoft/MARO#
  :alt: Gitter


.. image:: https://raw.githubusercontent.com/microsoft/maro/master/docs/source/images/badges/stack_overflow.svg
  :target: https://stackoverflow.com/questions/ask?tags=maro
  :alt: Stack Overflow


.. image:: https://img.shields.io/github/release-date-pre/microsoft/maro
  :target: https://github.com/microsoft/maro/releases
  :alt: Releases


.. image:: https://img.shields.io/github/commits-since/microsoft/maro/latest/master
  :target: https://github.com/microsoft/maro/commits/master
  :alt: Commits


.. image:: https://github.com/microsoft/maro/workflows/vulnerability%20scan/badge.svg
  :target: https://github.com/microsoft/maro/actions?query=workflow%3A%22vulnerability+scan%22
  :alt: Vulnerability Scan


.. image:: https://github.com/microsoft/maro/workflows/lint/badge.svg
  :target: https://github.com/microsoft/maro/actions?query=workflow%3Alint
  :alt: Lint


.. image:: https://img.shields.io/codecov/c/github/microsoft/maro
  :target: https://codecov.io/gh/microsoft/maro
  :alt: Coverage


.. image:: https://img.shields.io/pypi/dm/pymaro
  :target: https://pypi.org/project/pymaro/#files
  :alt: Downloads


.. image:: https://img.shields.io/docker/pulls/maro2020/maro
  :target: https://hub.docker.com/repository/docker/maro2020/maro
  :alt: Docker Pulls


.. image:: https://raw.githubusercontent.com/microsoft/maro/master/docs/source/images/badges/play_with_maro.svg
  :target: https://hub.docker.com/r/maro2020/maro
  :alt: Play with MARO



.. image:: https://github.com/microsoft/maro/blob/master/docs/source/images/logo.svg
  :target: https://maro.readthedocs.io/en/latest/
  :alt: MARO LOGO

================================================================================================================

Multi-Agent Resource Optimization (MARO) platform is an instance of Reinforcement
learning as a Service (RaaS) for real-world resource optimization. It can be
applied to many important industrial domains, such as `container inventory
management <https://maro.readthedocs.io/en/latest/scenarios/container_inventory_management.html>`_
in logistics, `bike repositioning <https://maro.readthedocs.io/en/latest/scenarios/citi_bike.html>`_
in transportation, `virtual machine <https://maro.readthedocs.io/en/latest/scenarios/vm_scheduling.html>`_ provisioning in data centers, and asset management in finance. Besides
`Reinforcement Learning <https://www.andrew.cmu.edu/course/10-703/textbook/BartoSutton.pdf>`_ (RL),
it also supports other planning/decision mechanisms, such as
`Operations Research <https://en.wikipedia.org/wiki/Operations_research>`_.

Key Components of MARO:


* Simulation toolkit: it provides some predefined scenarios, and the reusable
  wheels for building new scenarios.
* RL toolkit: it provides a full-stack abstraction for RL, such as agent manager,
  agent, RL algorithms, learner, actor, and various shapers.
* Distributed toolkit: it provides distributed communication components, interface
  of user-defined functions for message auto-handling, cluster provision, and job orchestration.


.. image:: https://github.com/microsoft/maro/blob/master/docs/source/images/maro_overview.svg
   :target: https://github.com/microsoft/maro/blob/master/docs/source/images/maro_overview.svg
   :alt: MARO Key Components


Contents
--------

.. list-table::
  :header-rows: 1

  * - File/folder
    - Description
  * - ``maro``
    - MARO source code.
  * - ``docs``
    - MARO docs, it is host on `readthedocs <https://maro.readthedocs.io/en/latest/>`_.
  * - ``examples``
    - Showcase of MARO.
  * - ``notebooks``
    - MARO quick-start notebooks.


*Try `MARO playground <#run-playground>`_ to have a quick experience.*

Install MARO from `PyPI <https://pypi.org/project/pymaro/#files>`_
----------------------------------------------------------------------

*Notes: The CLI commands (including the visualization tool) are not included in pymaro package. To enable these support, you need to install from source.*


*
  Mac OS / Linux

  .. code-block:: sh

    pip install pymaro

*
  Windows

  .. code-block:: powershell

    # Install torch first, if you don't have one.
    pip install torch===1.6.0 torchvision===0.7.0 -f https://download.pytorch.org/whl/torch_stable.html

    pip install pymaro

Install MARO from Source
------------------------

*Notes: Install from source if you want to use the CLI commands (including the visualization tool).*


*
  Prerequisites


  * C++ Compiler

    * Linux or Mac OS X: ``gcc``
    * Windows: `Build Tools for Visual Studio 2017 <https://visualstudio.microsoft.com/thank-you-downloading-visual-studio/?sku=BuildTools&rel=15>`_

*
  Enable Virtual Environment


  *
    Mac OS / Linux

    .. code-block:: sh

      # If your environment is not clean, create a virtual environment firstly.
      python -m venv maro_venv
      source ./maro_venv/bin/activate

  *
    Windows

    .. code-block:: powershell

       # If your environment is not clean, create a virtual environment firstly.
       python -m venv maro_venv

       # You may need this for SecurityError in PowerShell.
       Set-ExecutionPolicy -Scope CurrentUser -ExecutionPolicy Unrestricted

       # Activate the virtual environment.
       .\maro_venv\Scripts\activate

*
  Install MARO

  .. code-block:: sh

     # Git Clone the whole source code.
     git clone https://github.com/microsoft/maro.git


  *
    Mac OS / Linux

    .. code-block:: sh

      # Install MARO from source.
      bash scripts/install_maro.sh

  *
    Windows

    .. code-block:: powershell

      # Install MARO from source.
      .\scripts\install_maro.bat

*
  *Notes: If your package is not found, remember to set your PYTHONPATH*


  * Mac OS / Linux

  .. code-block:: sh

     export PYTHONPATH=PATH-TO-MARO


  * Windows

  .. code-block:: powershell

     $Env:PYTHONPATH=PATH-TO-MARO

Quick Example
-------------

.. code-block:: python

  from maro.simulator import Env

  env = Env(scenario="cim", topology="toy.5p_ssddd_l0.0", start_tick=0, durations=100)

  metrics, decision_event, is_done = env.step(None)

  while not is_done:
      metrics, decision_event, is_done = env.step(None)

  print(f"environment metrics: {env.metrics}")

`Environment Visualization <https://maro.readthedocs.io/en/latest/>`_
-------------------------------------------------------------------------

.. code-block:: sh

   # Enable environment dump feature, when initializing the environment instance
   env = Env(scenario="cim",
             topology="toy.5p_ssddd_l0.0",
             start_tick=0,
             durations=100,
             options={"enable-dump-snapshot": "./dump_data"})

   # Inspect environment with the dump data
   maro inspector dashboard --source_path ./dump_data/YOUR_SNAPSHOT_DUMP_FOLDER

Show Cases
^^^^^^^^^^


*
  Case I - Container Inventory Management

  .. image:: https://github.com/microsoft/maro/blob/master/docs/source/images/visualization/dashboard/cim_inter_epoch.gif
     :target: https://github.com/microsoft/maro/blob/master/docs/source/images/visualization/dashboard/cim_inter_epoch.gif
     :alt: CIM Inter Epoch


  .. image:: https://github.com/microsoft/maro/blob/master/docs/source/images/visualization/dashboard/cim_intra_epoch_by_ports.gif
     :target: https://github.com/microsoft/maro/blob/master/docs/source/images/visualization/dashboard/cim_intra_epoch_by_ports.gif
     :alt: CIM Intra Epoch


*
  Case II - Citi Bike

  .. image:: https://github.com/microsoft/maro/blob/master/docs/source/images/visualization/dashboard/citi_bike_inter_epoch.gif
     :target: https://github.com/microsoft/maro/blob/master/docs/source/images/visualization/dashboard/citi_bike_inter_epoch.gif
     :alt: Citi Bike Inter Epoch


  .. image:: https://github.com/microsoft/maro/blob/master/docs/source/images/visualization/dashboard/citi_bike_intra_epoch_by_station.gif
     :target: https://github.com/microsoft/maro/blob/master/docs/source/images/visualization/dashboard/citi_bike_intra_epoch_by_station.gif
     :alt: Citi Bike Intra Epoch


Run Playground
--------------


*
  Pull from `Docker Hub <https://hub.docker.com/r/maro2020/playground>`_

  .. code-block:: sh

     # Pull the docker image from docker hub
     docker pull maro2020/playground

     # Run playground container.
     # Redis commander (GUI for redis) -> http://127.0.0.1:40009
     # Jupyter lab with maro -> http://127.0.0.1:40010
     docker run -p 40009:40009 -p 40010:40010 maro2020/playground

*
  Build from source


  *
    Mac OS / Linux

    .. code-block:: sh

      # Build playground image.
      bash ./scripts/build_playground.sh

       # Run playground container.
       # Redis commander (GUI for redis) -> http://127.0.0.1:40009
       # Jupyter lab with maro -> http://127.0.0.1:40010
       docker run -p 40009:40009 -p 40010:40010 maro2020/playground

  *
    Windows

    .. code-block:: powershell

      # Build playground image.
      .\scripts\build_playground.bat

       # Run playground container.
       # Redis commander (GUI for redis) -> http://127.0.0.1:40009
       # Jupyter lab with maro -> http://127.0.0.1:40010
       docker run -p 40009:40009 -p 40010:40010 maro2020/playground

Contributing
------------

This project welcomes contributions and suggestions. Most contributions require
you to agree to a Contributor License Agreement (CLA) declaring that you have
the right to, and actually do, grant us the rights to use your contribution. For
details, visit https://cla.opensource.microsoft.com.

When you submit a pull request, a CLA bot will automatically determine whether
you need to provide a CLA and decorate the PR appropriately (e.g., status check,
comment). Simply follow the instructions provided by the bot. You will only need
to do this once across all repos using our CLA.

This project has adopted the
`Microsoft Open Source Code of Conduct <https://opensource.microsoft.com/codeofconduct/>`_.
For more information see the
`Code of Conduct FAQ <https://opensource.microsoft.com/codeofconduct/faq/>`_
or contact `opencode@microsoft.com <mailto:opencode@microsoft.com>`_
with any additional questions or comments.

Related Papers
--------------

`Container Inventory Management <https://maro.readthedocs.io/en/latest/scenarios/container_inventory_management.html>`_
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^


.. image:: https://github.com/microsoft/maro/blob/master/docs/source/images/scenario/cim_vis.gif
   :target: https://github.com/microsoft/maro/blob/master/docs/source/images/scenario/cim_vis.gif
   :alt: CIM Vis


Wenlei Shi, Xinran Wei, Jia Zhang, Xiaoyuan Ni, Arthur Jiang, Jiang Bian, Tie-Yan Liu. "\ `Cooperative Policy Learning with Pre-trained Heterogeneous Observation Representations <https://arxiv.org/abs/2012.13099>`_\ ". AAMAS 2021

Xihan Li, Jia Zhang, Jiang Bian, Yunhai Tong, Tie-Yan Liu. "\ `A Cooperative Multi-Agent Reinforcement Learning Framework for Resource Balancing in Complex Logistics Network <https://arxiv.org/abs/1903.00714>`_\ ". AAMAS 2019

Related News
------------

`MSRA Top-10 Hack-Techs in 2021 <https://mp.weixin.qq.com/s/Y4kjQq8gKcsEsVadjdwnEQ>`_

`Open Source Platform MARO: Anywhere Door for Resource Optimization <https://mp.weixin.qq.com/s/zXIpgzomLhDWS_YUFmRlEQ>`_

`AI from "Point" to "Surface" <https://mp.weixin.qq.com/s/PggO49wwmS7grTO0nEMgVQ>`_

`Cite Us <https://github.com/microsoft/maro/blob/master/CITATION>`_
---------------------------------------------------------------------

License
-------

Copyright (c) Microsoft Corporation. All rights reserved.

Licensed under the `MIT <https://github.com/microsoft/maro/blob/master/LICENSE>`_ License.