Installation

Set up RL-CP Fusion in your environment

Follow these steps to install RL-CP Fusion and its dependencies.

System Requirements

  • Python 3.7 or higher
  • CUDA-compatible GPU (recommended)
  • 64-bit operating system

Installing Dependencies

First, install PyTorch with CUDA support:

pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

Install other required packages:

pip install gym d4rl numpy tensorboardX

Installing RL-CP Fusion

Clone the repository:

git clone https://github.com/yourusername/rl-cp-fusion.git
cd rl-cp-fusion

Install the package:

pip install -e .

Verifying Installation

Run this simple test to verify the installation:

python -c "from conformal_sac.agent_wrapper import SACAgent; print('Installation successful!')"

Common Issues

CUDA Issues

If you encounter CUDA-related errors, verify your CUDA installation:

python -c "import torch; print(torch.cuda.is_available())"

D4RL Dataset Access

If you have trouble accessing D4RL datasets, make sure you have the correct permissions:

python -c "import d4rl; import gym; env = gym.make('halfcheetah-medium-v2')"

Next Steps

After installation, you can: