lm-evaluation-harness/lm_eval/__init__.py

from .evaluator import evaluate, simple_evaluate
import habana_frameworks.torch.gpu_migration
import habana_frameworks.torch.core as htcore