Multigpu dense eval return errors after corpus embedding

[sirCamp]

Hi and thanks for you work. I'm trying to run the example for evaluation with multigpu (https://github.com/beir-cellar/beir/blob/main/examples/retrieval/evaluation/dense/evaluate_sbert_multi_gpu.py) but i continuosly receive this error:

File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
OSError: Expected to be able to read 2594632 bytes for message body, got 2594620
[W110 10:20:47.593865407 CudaIPCTypes.cpp:16] Producer process has been terminated before all shared CUDA tensors released. See Note [Sharing CUDA tensors]
/opt/conda/envs/campeses/lib/python3.11/multiprocessing/resource_tracker.py:254: UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown
warnings.warn('resource_tracker: There appear to be %d '

I tried to both downgrading and upgrading beir/sentence trasnformers and also datasets and pyarrow.

Did you have suggestion on how to solve this?