-
Used to use pytorch + gym for RL.
Is there a way to make a custom neural net in pytorch and insert it into the pearl agent without using the pre-made settings for neural nets?
-
Hi,
Will you be providing any examples for real-world implementations of Pearl?
Thank you for creating this amazing project!
-
We need to go through and remove all .travis.yml files and travis badges. This would be easy to do - just need a volunteer to do it.
-
Hello
Thank you for sharing your code.
I could run two models with fixed time and pedestrian
But I had errors when running training_main.py in DEEPQLEARNING.
----- Episode 1 of 200
Traceba…
-
Hi @aqeelanwar, I'm trying to use this library to train and test the basic functionality of PEDRA but I'm running into some issues. At the moment, my config settings are like this:
```
[general_para…
-
See instructions at https://sites.google.com/view/pomdps-hackathon/pomdptools-upgrade-instructions
-
The example provided in the docstring of `KMarkovUpdater` in `src/k_previous_observations.jl` does not work for:
`pomdp = TigerPOMDP(); policy = solve(QMDPSolver(), pomdp); rng = MersenneTwister();`…
-
### Describe the issue:
I am using it to train policy gradient in RL in [colab](https://colab.research.google.com/drive/1J3VlrQyqNVRglVLa64DFJArqfe2hySzB?usp=sharing)
File "mtrand.pyx", line 928…
-
Hi,
I am facing some issues while trying to make use of PEDRA. I am currently using tensorflow 2.0, and has used the tf_upgrade_v2 script to get PEDRA to run, however, it seems that tf.contrib.lay…
-
As many of the JuliaPOMDP solver have already, I think it's time to register `RLInterface` with the General registry.
mossr updated
3 years ago