About the baseline in the paper

[Jonyian]

Could you please open source the “ours” algorithm mentioned in your paper?

[BigZhaoHYZ]

同求，其次运行文件中的run_maacktr.py和run_mappo.py都会以下错误，可以提供一下解决方案吗?感谢 ERROR:root:Can not find checkpoint for ./results/Sep_19_12_22_29/models/ H:\Code-of-study\PaperCode\MARL_CAVs-main\MARL\single_agent\kfac.py:144: UserWarning: volatile was removed (Variable.volatile is always False) if input[0].volatile == False and self.steps % self.Ts == 0: H:\ProgramData\Anaconda3\envs\MARL_CAVs-main\lib\site-packages\torch\nn\modules\module.py:795: UserWarning: Using a non-full backward hook when the forward contains multiple autograd Nodes is deprecated and will be removed in future versions. This hook will be missing some grad_input. Please use register_full_backward_hook to get the documented behavior. warnings.warn("Using a non-full backward hook when the forward contains multiple autograd Nodes " H:\Code-of-study\PaperCode\MARL_CAVs-main\MARL\single_agent\Model_common.py:62: UserWarning: Implicit dimension choice for log_softmax has been deprecated. Change the call to include dim=X as an argument. act = self.actor_output_act(self.actor_linear(out)) Traceback (most recent call last): File "H:\Code-of-study\PaperCode\MARL_CAVs-main\MARL\run_maacktr.py", line 217, in train(args) File "H:\Code-of-study\PaperCode\MARL_CAVs-main\MARL\run_maacktr.py", line 128, in train maacktr.interact() File "H:\Code-of-study\PaperCode\MARL_CAVs-main\MARL\MAACKTR.py", line 96, in interact actions.append([index_to_one_hot(a, self.action_dim) for a in action]) File "H:\Code-of-study\PaperCode\MARL_CAVs-main\MARL\MAACKTR.py", line 96, in actions.append([index_to_one_hot(a, self.action_dim) for a in action]) File "H:\Code-of-study\PaperCode\MARL_CAVs-main\MARL\common\utils.py", line 37, in index_to_one_hot one_hot = np.zeros((len(index), dim)) TypeError: object of type 'numpy.int32' has no len()

[DongChen06]

@Jonyian @BigZhaoHYZ , please find the codes at: https://drive.google.com/drive/folders/1CPOOYSQzqc0_XUr8durNWDUBv66PUHaM

[DongChen06]

@BigZhaoHYZ , please look at the solution at: https://github.com/DongChen06/MARL_CAVs/issues/4. The reason may be the mismatch of the libs.

[Jonyian]

Thank you very much for your reply. I also want to ask you a question. Is the baseline of your paper the run_ma2c.py in the folder?

[DongChen06]

You can change the settings in configs.ini. For example, you can set safety_gurantee as False. ;

There are also other settings you can play with:

; concurrent
training_strategy = concurrent
actor_hidden_size = 128
critic_hidden_size = 128
shared_network = True
action_masking = True
state_split = True
; "greedy", "regionalR", "global_R"
reward_type = regionalR

[BigZhaoHYZ]

safety_guarantee = True n_step = 7 你好，运行run_ma2c.py加上configs.ini就是论文中的“基线+Tn”吗？感谢回复！ @DongChen06

[DongChen06]

@BigZhaoHYZ Yes, you are right

[Jonyian]

Thank you very much for your reply