Visualize multiagent_traffic_light_grid
pengyuan-zhou opened this issue · 5 comments
Hi,
Sorry I'm reposing this from flow repo, kinda in a desperate rushing condition for tests.
I've tried this by just importing the MultiFeedForwardPolicy in run_eval.py, apparently doesn't work and require more work.
Any plan to deploy this soon?
Thanks.
Hey, thank you for posting this here! There were recently some incompatibility bugs in run_eval.py with Flow that I have fixed. This issue raises to light that I haven't added multi-agent policies to the visualizer though. Will update today and share with you once it is done.
Hey, thank you for posting this here! There were recently some incompatibility bugs in run_eval.py with Flow that I have fixed. This issue raises to light that I haven't added multi-agent policies to the visualizer though. Will update today and share with you once it is done.
Thank you so much. Looking forward to it 👍
Hey, thank you for posting this here! There were recently some incompatibility bugs in run_eval.py with Flow that I have fixed. This issue raises to light that I haven't added multi-agent policies to the visualizer though. Will update today and share with you once it is done.
Hi, seems the repo has developed a lot. Is there a recent plan to enable the test for traffic light grid scenarios?
hey @pengyuan-zhou I don't think i'll personally be working with grid networks in the near future. I'm adding support to any flow experiment through this repo right now though so that people can train any example provided in flow/examples (see PR #158).
I'll also be adding a description soon on how to create custom environments that can be trained in this repo. I'll be incorporating it master via PR #157
hey @pengyuan-zhou I don't think i'll personally be working with grid networks in the near future. I'm adding support to any flow experiment through this repo right now though so that people can train any example provided in flow/examples (see PR #158).
I'll also be adding a description soon on how to create custom environments that can be trained in this repo. I'll be incorporating it master via PR #157
Great, thanks for the info. Looking forward to the description. A proper TD3 or ddpg for grid light control is really necessary to tackle the space issue for singleagent learning :)