rl
rl copied to clipboard
pytorch
A modular, primitive-first, python-first PyTorch library for Reinforcement Learning.
Is there a tutorial on how to implement an RL algorithm like PPO in a PettingZoo env using TorchRL. I used the PettingZooWrapper() function on the already created environment but...
## Description This PR offers a convenient `lightning.pytorch.LightningModule` base class, from which one can inherit to be able to train a `torchrl` model using `lightning`. ## Motivation and Context This...
## Description Adding [CrossQ](https://openreview.net/pdf?id=PczQtTsTIX) ## Motivation and Context Why is this change required? What problem does it solve? If it fixes an open issue, please link to the issue here....
## Motivation Simulated environments are capable of rendering depth images which can be very useful in downstream training as they enable 3D understanding of the task being performed. RoboHive environments,...
## Description Per #2003 adds multi-agent GRU and LSTMs to torchRL's multiagent modules. Modifies the MultiAgentNetBase class to take in multiple input tensors and output tensors, which allows these recurrent...
TODO: - [x] Doc
## Describe the bug The VecNorm transform produces a bug when calling the to_observation_norm method with multiple keys ## To Reproduce ``` gym_env = GymEnv("MountainCarContinuous-v0", device='cpu') transformed_env = TransformedEnv(gym_env, VecNorm(in_keys=["observation",...
Hello, I have a question regarding the construction of replay buffers in distributed training (DDP). Across multiple workers, I would like to use a single, large prioritized replay buffer. With...
## Describe the bug When the batched env device is `cuda` the step count on the batched env seems completely off from what it should be. When the batches env...
