facebookresearch / CollaQ

A code implementation for our arXiv paper "Multi-agent Adhoc Team Play using Decompositional Q function"
Other
127 stars 24 forks source link

4 questions about the experiment mentioned in the paper #16

Open yyf17 opened 3 years ago

yyf17 commented 3 years ago

Dear author, I have 4 questions for you about the experiment mentioned in the paper as followings:

1 Question :What is the map used in Table 4?

image

2 Question:Which parameter is used for "without random agent IDs" in Figure 10?

image

3 Question:How to set the configuration file of "Random Action" algorithm in Figure 9(In the Ad hoc Resource Collection Env)?

image

4 Question:Some experimental maps mentioned in the article were not found. Can the designed maps 3s1z_vs_16zg, 1s3z_vs_16zg and 2s2z_vs_16zg be shared?

Can the modified maps 28m_vs_30m and 29m_vs_30m be shared?