Issue: Conda Environment Creation Taking Too Long

[Jeevan-Neupane]

Description

When trying to create a Conda environment using the benchmark-requirements-cpu.yaml file, the process gets stuck at "Solving environment" for an unusually long time and it doesn't get resolved.

Steps to Reproduce

Run the following command

conda env create --name xrayto3d --file benchmark-requirements-cpu.yaml

The process halts at:

Solving environment: \

It takes an excessive amount of time without proceeding and doesn't get solved.

[Jeevan-Neupane]

It looks like the issue was caused by unnecessary GPU-related dependencies (pytorch-gpu, cudatoolkit=10.1, pytorch-cuda). Since this setup is running on CPU-only, Conda was struggling to resolve GPU-related packages that aren’t needed.

Updated benchmark-requirements-cpu.yaml (CPU-Only)

name: xrayto3d
channels:
  - conda-forge
  - defaults
dependencies:
  - python=3.8
  - numpy
  - monai
  - torchvision
  - nibabel
  - itk
  - simpleitk
  - scikit-image
  - pillow
  - pandas
  - matplotlib
  - pytorch  # CPU version (no GPU)
  - wandb
  - tqdm
  - einops
  - pytorch-lightning

Previous benchmark-requirements-cpu.yaml (With GPU Dependencies)

channels:
  - defaults
  - conda-forge
  - anaconda
dependencies:
  - python=3.8
  - numpy
  - monai
  - torchvision
  - nibabel
  - itk
  - simpleitk=2.0
  - scikit-image
  - pillow
  - pandas
  - matplotlib
  - pytorch-gpu  #  Unnecessary for CPU
  - cudatoolkit=10.1  # Causes dependency resolution issues
  - wandb
  - tqdm
  - einops
  - pytorch-lightning

[msrepo]

@Jeevan-Neupane Lets see if using mamba instead of conda speeds up conda env creation especially pulling the required dependencies. mamba is a like for like replacement of conda commands.