cellink.tl.external.configure_livi_runner#
- cellink.tl.external.configure_livi_runner(livi_root, execution_mode='python_api', python_executable='python', device='auto')#
Configure the global LIVI runner.
Must be called once before using any other LIVI functions.
- Parameters:
livi_root (str or Path) – Path to the LIVI repository root (contains
src/andconfigs/).execution_mode ({"python_api", "subprocess"}) – How to run LIVI training.
"python_api"(recommended) imports LIVI classes directly; results are Python objects."subprocess"runs LIVI’s Hydra CLI, writing configs toconfigs/inside the LIVI repository.python_executable (str) – Python binary used in subprocess mode (default:
"python").device (str) – Compute device:
"auto"(auto-detect GPU),"cuda", or"cpu".
- Return type:
- Returns:
LIVIRunner
Examples
>>> import cellink as cl >>> cl.tl.external.configure_livi_runner("/lustre/projects/LIVI")