rl-baselines3-zoo VS stable-baselines3

I'm trying to solve MountainCar-v0 enviroment from gymnasium with the A2C algorithm and the agent doesn't find a solution. I checked this so I added import stable_baselines3.common.sb2_compat.rmsprop_tf_like as RMSpropTFLike. Also checked the rl-baselines3-zoo for the hyperparameter tuning. So my code is:

RL Zoo3 (training framework): https://github.com/DLR-RM/rl-baselines3-zoo

I'm still new to the domain but wanted to shared some experimental data I've gathered from massive amount of experimentation. I don't have a strong understanding of the theory as I'm more of a software engineer than data scientist, but perhaps this will help other implementers. These notes are based on Stable Baselines 3 and RL Baselines3 Zoo with using PPO+LSTM (should apply generally to all the algos for the most part)

Try hyperparameter search. It's implemented here: https://github.com/DLR-RM/rl-baselines3-zoo for stable-baselines3. Hyperparameters make a huge difference in RL, much more than in supervised learning.

Integrating RL-baselines3-zoo

Hello, I've been running some tests of Double DQN with Stable Baselines 3 Zoo and to compare I'm using the graphs provided by Noisy Networks For Exploration.

- you can find tuned hyperparameters for DDPG, SAC, PPO in https://github.com/DLR-RM/rl-baselines3-zoo

For more complete implementation: https://github.com/DLR-RM/rl-baselines3-zoo

https://github.com/DLR-RM/rl-baselines3-zoo (PyTorch based, using https://github.com/DLR-RM/stable-baselines3)

We also release 100+ trained models in our experimental framework, the rl zoo: https://github.com/DLR-RM/rl-baselines3-zoo

The latest release (v3.0.0) of Upkie's software brings a functional sim-to-real reinforcement learning pipeline based on Stable Baselines3, with standard sim-to-real tricks. The pipeline trains on the Gymnasium environments distributed in upkie.envs (setup: pip install upkie) and is implemented in the PPO balancer. Here is a policy running on an Upkie:

PettingZoo 1.24.0 is now live! This release includes Python 3.11 support, updated Chess and Hanabi environment versions, and many bugfixes, documentation updates and testing expansions. We are also very excited to announce 3 tutorials using Stable-Baselines3, and a full training script using CleanRL with TensorBoard and WandB.

I am wondering why there is so few algorithms in Stable Baselines 3 (SB3, https://github.com/DLR-RM/stable-baselines3/tree/master)? I was expecting some algorithms like ICM, HIRO, DIAYN, ... Why there is no model-based, skill-chaining, hierarchical-RL, ... algorithms implemented there?

Discord is more focused, and they have a page for people who wants to contribute https://github.com/DLR-RM/stable-baselines3/blob/master/CONTRIBUTING.md

Therefore, I debugged this error to the ReplayBuffer that was imported from `SB3`. This is the problem function -

Have you ever wanted to use dm-control with stable-baselines3? Within Reinforcement learning (RL), a number of APIs are used to implement environments, with limited ability to convert between them. This makes training agents across different APIs highly difficult, and has resulted in a fractured ecosystem.

Changelog: https://github.com/DLR-RM/stable-baselines3/releases/tag/v1.8.0

Great project! One question though, is there any reason why you are not using existing RL models instead of creating your own, such as stable baselines?