search

islambarakat99 / Multi-Robot-Formation-Control-using-Deep-Reinforcement-Learning

A leader-follower formation control using deep reinforcement learning environment, In which every agent can learn to follow the leader agent by keeping track of a certain distance to that leader, avoiding obstacles, and avoiding collision with the other agents.
MIT License
34 stars 6 forks source link

Running doubts #3

Closed syf980302 closed 1 year ago

syf980302 commented 1 year ago

Hello, I'm very interested in the formation control of multi-agent. I saw your project on github and tried to run it myself, but I have a question for your help.

image
islambarakat99 commented 1 year ago

It says that the file system can't load /home/islam/training/policy/ because it doesn't exist, as this line was hard-coded to save the data in a specific directory on my machine, which were used for other purposes that were not related to the project.

In "train.py" you can see in Line 27 

parser.add_argument("--save-dir", type=str, default="/home/islam/training/policy/", help="directory in which training state and model should be saved")

so here you have to give it your saving directory on your machine

and in lines 35 - 36:

parser.add_argument("--benchmark-dir", type=str, default="/home/islam/training/benchmark/", help="directory where benchmark data is saved")
parser.add_argument("--plots-dir", type=str, default="/home/islam/training/curves/", help="directory where plot data is saved")

In these lines you have to give it the directories for saving Benchmark Data, Plot Data too

syf980302 commented 1 year ago

ok, thank you. I see what you mean. Thank you for your help.