-
很奇怪的loss变化:
![image](https://user-images.githubusercontent.com/38696314/68921526-03f33680-07b4-11ea-925c-b347c9e02f03.png)
只是调了下超参数,和打印信息没有改变源代码
训练输出结果如下:
train_hyper-parameter:
map 8m
map_d…
-
when i run the command "python3 src/main.py --config=qmix_smac --env-config=sc2 with env_args.map_name=2s3z".The output comes "File "/usr/local/lib/python3.5/dist-packages/websocket/_socket.py", line…
-
I tried to implement baselines used in your paper, such as Central-V, IAC-V, under this project on 3M map, but I cannot reproduce the results reported in your paper. The following is the training cur…
-
I am getting this error when running the example (docker) command:
bash run.sh 1 python3 src/main.py --config=qmix_smac --env-config=sc2 with env_args.map_name=2s3z
Note: I replaced $GPU with 1,…
-
Hi, I want to evaluate on the 1c_3s_5z map, which is used as a benchmark for the QMIX paper, I can not find it under this directory: smac/smac/env/starcraft2/maps/SMAC_Maps/
Can you upload this map? …
-
I'm trying to add Rainbow DQN agent. Could you share the basics of your environment and point me to a code that I can take a look at to see how to use the environment?
-
[INFO 16:10:01] root Saving to FileStorageObserver in results/sacred.
[DEBUG 16:10:05] pymarl Using capture mode "fd"
[DEBUG 16:10:06] pymarl Stopping Heartbeat
[ERROR 16:10:06] pymarl Failed after…
-
### QMIX: No backprop on the (mixer) hypernetworks?
### Source code / logs
# Setup optimizer
self.params = list(self.model.parameters())
self.loss = QMixLoss(self.model…
-
Thanks for your repository.
I am trying to reproduce your results In QMIX paper. I ran the command like this:
```
python main.py --config=qmix --env-config=sc2 with env_args.map_name=3s5z
```
A…
-
### Describe the problem
It's been noted a few times that TFPolicyGraph is hard to understand for new developers. I think this is due to a few reasons:
- The policy, loss, and postprocessing are a…
ericl updated
5 years ago