Define protocols and networks, run the protocols and see the results!
git clone --recursive git@github.com:FaridZandi/psim.git
# install gcc-9.4, python3.8
sudo apt-get install cmake libboost-all-dev -y
python3 -m pip install matplotlib numpy pandas networkx seaborn scipy
build is done with cmake.
mkdir build
cd build
cmake ..
make -j
This simulator was used to evaluate the paper "Foresight: Joint Time and Space Scheduling for Efficient Distributed ML Training" published in the proceedings of the IFIP Networking 2025 Conference.
To Run the experiments, cd into the run directory, and run any of the following:
# figure 5
python sweep-components-jobsizes.py
python sweep-components-oversub.py
# Figure 6
python sweep-placement.py
# Figure 7
python sweep-intensity.py
python sweep-topology.py
The results will be generated under the results/exps/.