DLR-RM / rl-baselines3-zoo

A training framework for Stable Baselines3 reinforcement learning agents, with hyperparameter optimization and pre-trained agents included.

https://rl-baselines3-zoo.readthedocs.io

DLR-RM/rl-baselines3-zoo Issues

[Question] Results vastly different for an agent created with Stable Baselines3 using hyperparameters optimized in RL Baselines3 Zoo.
Updated 4 days ago1
[Question] Should train & eval environment seeds differ?
Closed 4 days ago4
Stuck at Local Minimum in PPO with CarRacing-v2 Environment
Closed 10 days ago
[Bug]: Docker tag is still 2.2.0a2, but latest rl-baselines3-zoo is now 2.3.0
Updated 18 days ago
[Question] How to render "info" on tensorboard
Updated 20 days ago1
[Question] The trained agent resets every 1000 episodes.
Closed a month ago3
[Question] How many startup trials in distributed optimization
Closed a month ago2
[Question] RL traning could not reach convergence for a customised environment
Closed a month ago1
[Question] Number of parallel environments with hyperparameters optimization
Closed a month ago1
[Bug]: Training suddenly stops at 25000 timesteps and Optuna optimization immediately exits in my custom environment
Closed a month ago7
[Bug]: Optimization log and optimal policy not in `--optimization-log-path` but in `--log-folder`
Updated 2 months ago1
[Bug]: Custom environment not found in gym registry, you maybe meant... error message
Updated 2 months ago1
[Question] exp_manager reward and GAE discount factors
Updated 2 months ago1
[Question] Does hyperparameter tuning support custom vectorized environments?
Closed 3 months ago6
[Bug]: TODO: add test dependencies in the `setup.py`
Closed 4 months ago1
[Bug]: video recording on Pybullet
Closed 4 months ago4
[Question] Custom Eval Callback for train/optimize
Closed 5 months ago2
[Question] RuntimeError: Unable to sample before the end of the first episode. We recommend choosing a value for learning_starts that is greater than the maximum number of timesteps in the environment.
Closed 5 months ago5
[Question] You must pass an environment when using `HerReplayBuffer`
Closed 5 months ago1
[Bug]: Custom Sub-Hyperparameters during train.py -> Optimize
Updated 5 months ago1
Training DonkeyCar with TQC algorithm with pretrained AE
Updated 6 months ago
[Error]: I got unexpected error using enjoy() with pretrain model
Closed 6 months ago4
[Question] trying to run panda environments with RL zoo
Closed 6 months ago2
[Bug]: Nan Problems for SAC, TQC, for AntBulletEnv-v0, HalfCheetahBulletEnv-v0
Updated 6 months ago9
[Feature Request] Store git hash of key repos/packages
Updated 7 months ago
[Bug]: Cannot enjoy due to error Cannot convert space of type Discrete(7). Please upgrade your code to gymnasium.
Updated 7 months ago1
[Feature Request] Call train from Python code
Closed 7 months ago2
Issue with 'feat/offline-RL' Branch for Donkey Car in rl-baselines3-zoo
Updated 7 months ago4
[Question] Training Donkey Car Without Simulator Rendering
Updated 7 months ago2
[Bug]: ppo_lstm not implemented in hyperparams_opt.py
Closed 7 months ago3
Plotting Script Improvement
Closed 8 months ago1
[Question] Support for Customized BaseFeaturesExtractor
Closed 8 months ago1
[Question] rl_zoo3 optimization pipeline for ros-based custom env
Closed 8 months ago2
[Bug]: enjoy panda policy in hugging face
Closed 8 months ago7
Error when using recommended plotting commands
Updated 8 months ago3
[Feature Request] Specify custom keyword arguments for eval environments
Closed 9 months ago2
[Feature Request] Remove gym dependency
Closed 9 months ago3
[Bug]: Kwargs in `record_video.py` not preserved by SB3 Vector Environment wrappers
Closed 9 months ago4
`optuna.trial.Trial.suggest_uniform()` deprecated
Closed 9 months ago
#enhancment
Closed 9 months ago
[Feature Request] Outdated dockerfile
Closed 9 months ago2
[Feature Request] Add the `--gym-packages` flag to `record_video`
Updated 9 months ago1
[Question] exported ONNX model does not result in same output as the original pytorch model
Updated 10 months ago5
Half-Cheetah reward function
Closed a year ago
Upgrade all MuJoco envs to v4
Updated a year ago3
[Enhancement] Deprecate --gym-packages argument
Updated a year ago5
[Question] Should "continue training" load the vecnormalize.pkl as well as the model.zip?
Closed a year ago2
[Question] Correct way of hyper parameter optimization for new algorithm?
Closed a year ago1
Add link to custom environments page in sb3 docs
Updated a year ago
[Bug]: enjoy tries to access huggingface
Closed a year ago4