I cannot reproduce the result

[lukeprotag]

I can only obtain results in 4 robot settings, but when using weights to train 10 robots, I cannot obtain good results.

[hanruihua]

Hi

you can try to adjust the coefficients in the reward function, which is important for the training outcome

[jijiking-hh]

Hello, I also have this question. How can 10 agents set reward functions in a dynamic environment? May I ask if you can demonstrate it

[lukeprotag]

Hello, I also have this question. How can 10 agents set reward functions in a dynamic environment? May I ask if you can demonstrate it

In train_process.py, I found : par_env.add_argument('--reward_parameter', type=float, default=(3.0, 0.3, 0.0, 6.0, 0.3, 3.0, -0, 0), nargs='+')

[jijiking-hh]

Thank you，actually I have also noticed this, but the correspondence between these parameters and the reward function, their actual significance, and the basis for adjustment are still unclear