Update DeepRVAT environment

[meyerkm]

What

This PR updates the deeprvat_env.yaml and deeprvat_env_no_gpu.yaml provided environment files. Specifically, updating the PyTorch and PyTorch-Lightning packs to 2.x versions.
Due to the PyTorch updates, subsequent updates were required in the PyTorch Lightning Trainer parameters, e.g. specified under pl_trainer in the respective config.yaml ,as well as, in the BaseModel.

This PR also addresses Issue #16 , and both Conda and Mamba can now correctly compile cuda with the provided deeprvat_env.yaml file. Note: It is still recommended by the developers to use Mamba due to speed in solving the environment over Conda, e.g. mamba env create -n deeprvat -f ~/deeprvat_env.yaml

Testing

Create the DeepRVAT environment from the .yaml file with various package managers and run the various DeepRVAT pipelines.