-
I'm normally against implementing very recent papers before they prove to be valuable but I would like to make an exception for that one, especially because of the good results. It was recently accept…
-
@twni2016 @yufeiwang63 Thanks for your great work! I am new to reinforcement learning and have some questions about your proposed meta-SAC.
- I see that the ValueNetwork is used in `sacmeta.py`, wh…
-
Hello, @ku2482
Could you let me know what does gather method do? I would appreciate you for any example or link which help me to figure out how this function work.
https://github.com/ku2482/sac-di…
-
### 🐛 Bug
When using HER with replay_buffer that overflows at some point we get a crash with IndexError.
### To Reproduce
Run HER + SAC with small `buffer_size` and small `learning_starts`
…
-
Hi I'm trying to run SAC Discrete and I keep getting following error
```
Warning: Error detected in AddmmBackward. Traceback of forward call that caused the error:
File "results/Cart_Pole.py", …
-
**my commmand is**
python tools/train.py ./configs/detectors/detectors_cascade_rcnn_r50_1x_coco.py
**the info is**
Traceback (most recent call last):
File "tools/train.py", line 153, in
…
-
### Describe the bug
Hi,
Thank you for this amazing library!
I have a custom Gym environment that uses a discrete action space. When I try to use SAC + HER to solve it, at some point, the mod…
-
- need to actually create a tensorflow NN implementation (or two) to interact with the gym
-
Hello,
I wonder if I have to do fine-tunings to get results from Translate augmentation. It always diverges! I have tested for Cartpole, Walker, and Cheetah.
In the following figures, the diverged o…
-
**Describe the bug**
There is an issue loading trained agent files in Python 3.8.
**Code example**
I checked this on the `rl-baselines3-zoo` repository.
```bash
(base) manifold$ python en…