Environment Setup

To set up your environment, follow the steps below:

$ python3 -m venv commonsense_tool
$ source commonsense_tool/bin/activate
(commonsense_tool) $ git clone https://github.com/reail-iitd/Robot-task-planning.git
(commonsense_tool) $ cd Robot-task-planning
(commonsense_tool) $ pip3 install -r requirements.txt

We used a system with 32GB RAM, i7 Octa-core CPU and 8GB graphics RAM to run our simulator. The models were trained using the same system. The simulator and training pipeline has been tested on Windows, MacOS and Ubuntu 16.04.

Known Errors

If you are using >=32GB RAM, graphics card with >=8GB GRAM and Windows 10 or above, or MacOS 10.15 or above or Ubuntu 16 or above you should not get any errors. However, in some cases, it is possible to run the setup even if you do not have these exact requirements.

1. MacOS 10.14 or below: You may get a multi-processing error for which you may add the following lines in app.py after the line inp = "jsons/input_home.json":

mp.set_start_method('spawn')
queue_from_webapp_to_simulator = mp.Queue()
queue_from_simulator_to_webapp = mp.Queue()
queue_for_error = mp.Queue()
queue_for_execute_to_stop = mp.Queue()
queue_for_execute_is_ongoing = mp.Queue()