Parameter server framework for distributed machine learning
Parameter server framework for distributed machine learning
Multiverso is a parameter server based framework for training machine learning models on big data with numbers of machines. It is currently a standard C++ library and provides a series of friendly programming interfaces, and it is extended to support calling from python and Lua programs. With such easy-to-use APIs, machine learning researchers and practitioners do not need to worry about the system routine issues such as distributed model storage and operation, inter-process and inter-thread communication, multi-threading management, and so on. Instead, they are able to focus on the core machine learning logics: data, model, and training.
For more details, please view our website http://www.dmtk.io.
Linux (Tested on Ubuntu 14.04)
sudo apt-get install libopenmpi-dev openmpi-bin build-essential cmake git
git clone https://github.com/Microsoft/multiverso.git --recursive && cd multiverso
mkdir build && cd build
cmake .. && make && sudo make installWindows
Open the Multiverso.sln with Visual Studio 2013 and build.
Current distributed systems based on multiverso:
This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact [email protected] with any additional questions or comments.
No open issues yet, or sync has not completed.