-
作者您好，首先感谢您优质的开源项目，代码易读性很高，且迁移到其他的环境非常方便。
我将此算法迁移到了highway-env来完成我的项目，但是我始终遇到下面图片这个问题，就是算法先收敛到一个比较好的值，但是持续一会就下降并且震荡幅度很大。
我想问一下解决这个问题是提前停止训练（在已经有比较好的结果的时候）吗？我现在在使用默认的2000000步，您在SMAC中遇到过相关的问题吗？
…
-
### What is the problem?
When trying to export models with multiple policies specified using `export_formats=` in `tune.run()`, the programs gives this error about not having a policy named `default_…
-
-
Thanks so much for your work, I find it very helpful.
I am confused with a problem, that I trained with qmix+cql,startcraft v1, 3m scneario, firstly offline and then online. I comment out the trai…
-
I took TwoStepGame for an example. My code is as follows.
I have tried different obs from TwoStepGame or TwoStepGame.with_agent_groups. It always obtained different errors.
Can any one tell me how …
-
- [ ] I have marked all applicable categories:
+ [ ] exception-raising bug
+ [ ] RL algorithm bug
+ [ ] documentation request (i.e. "X is missing from the documentation.")
+ [x] ne…
-
To expand the multi-agent RL algorithms, RLlib would greatly benefit from an implementation of the counterfactual multi-agent policy gradients (COMA) algorithm from Foerster et al. (https://arxiv.org/…
-
Hello author,
I got this error when I ran the code to replicate the results:
marl_env/lib/python3.8/site-packages/gym/logger.py:30: UserWarning: WARN: Could not seed environment
warnings.warn…
-
FileNotFoundError: [Errno 2] No such file or directory: '../exp/MMM/qmix/1/ckpt_BaseLearner_Wed_Jul_14_22_16_56_2021/iteration_9900.pth.tar'
-
作者您好！
我最近在尝试利用您编写的qmix、qtran_base和qtran_alt算法解决无人机路径规划问题。在我做实验的过程中，从实际的表现上来看，qmix算法的性能要高于qtran_base（可能是我的参数设定出现了一些问题）。所以我想请问一下您在实验中qtran算法的实际性能表现是否比qmix要好呢，如果好的话大概好多少
此外，qtran_alt是不是比qtran_base难收敛…