-
When I am conducting experiments and find that some approach works well, I want to compare my results to an established baseline. OpenAI Baselines is one, but (for obvious reasons listed in the readme…
dniku updated
5 years ago
-
1)Code Patch: (examples/actor_critic.py)
# driver = PyDriver(
# tf_env,
# policy,
# observers=train_metrics,
# max_steps=num_steps_per_…
-
Hi! Thank you for sharing your work! In the paper you are mentioning that you tested the HAC in a discrete gridworld environment. I assume that you do not need Mujoco for this. Is the code you are sha…
-
There are many algorithms that import Mujoco environments because they are not separated. In my case I dont care about Mujoco, in fact I had to get a trial license just to avoid having to remove code …
-
**TF Version: 2.0.0-dev20190214
Windows 10
Anaconda Python 3.6.5
GPU: GeForce GTX 1070 Max-Q Design
[Tensorflow 2.0 (gpu) nightly](https://pypi.org/project/tf-nightly-gpu-2.0-preview/) installed v…
-
in the manual of HFO, it writes:
![Screenshot from 2019-04-18 09-49-50](https://user-images.githubusercontent.com/35871168/56331560-6739cd80-61bf-11e9-8c1f-2dc5df14bdec.png)
so the power's range is …
-
Seems a good candidate for inclusion:
https://arxiv.org/abs/1801.01290
Applications:
https://arxiv.org/abs/1812.05905
-
### System information
- **OS Platform and Distribution (e.g., Linux Ubuntu 16.04)**: macOS Mojave
- **Ray installed from (source or binary)**: binary
- **Ray version**: 0.8.0.dev1
- **Python ve…
-
/SLM-Lab/slm_lab/lib/util.py", line 5, in
import cv2
ModuleNotFoundError: No module named 'cv2'
**To Reproduce**
1. OS used: Ubuntu 18 LTS
2. SLM-Lab git: git cloned
3. demo.json not wor…
-
Hi,
I already understand the keras functional model, my question is more specific to Keras RL and the structure of data in the Environment objects for models with multiple inputs.
how can I set …