-
I ran with the hyperparameters given for FetchReach TD3, and it wasn't able to solve perfectly in 25k timesteps like written — can someone verify that this is the case?
-
While running ```FinRL_single_stock_trading.ipynb```, I found that the logs of Model 2 (DDPG) and Model 4 (TD3) are weird. It looks like none trading has been done, which is not same as the demo you g…
-
### The problem
this error started showing up a few versions ago and similar error appears in this case: #40166
Opening a new issue as the previous one wasn't properly resolved and the same error…
-
**Description**
**Steps to reproduce the issue:**
1. Run nightly tests
**Describe the results you received:**
Multiple cases in `drop_packets` fail because counter `RX_DRP` was not inc…
-
**High Level Description**
By checking the source code of "train.py" under the ultra directory, it seems that the training process only use 1 cpu core and does not use cuda acceleration, which makes …
-
Hello,
I have tried to use HER+DDPG to pretrain an agent based on some recorded demonstrations.
From the error I obtained i believe right now the library does not offer this feature, correct?
Whe…
-
I have a few question about your training algorithm:
1. How are shared policy parameters updated? From my understanding, it seems you are updating them once in each agent that uses the shared params.…
-
(My first issue on an open source GitHub repository)
When I am using the function _plot_param_importances(study)_ the plot is displaying different values than the ordered dict received by the functio…
-
Hello keiohta, I found that GAIL could not work in Hopper-v2 or Walker2d-v2. And SAC in this repository could not train a successful policy for Hopper-v2 too. I have checked the implementation of GAIL…
-
Hi, Merry Christmas!
Thank you for sharing the model-free-RL-library. Recently, I've been interesting for PER with continuous RL algorithms. However, I found that the performance of td3+per and sac+p…