We find the num_workers parameter for data_loader is important in some environments such as docker or limited RAM, otherwise causing failure. So we add num_workers when evaluating.
For example,
In docker environment, the shared memory is limited causing specific error:
RuntimeError: DataLoader worker (pid xxxx) is killed by signal: Bus error. It is possible that dataloader's workers are out of shared memory. Please try to raise your shared memory limit.
Hi 👋,
We find the num_workers parameter for data_loader is important in some environments such as docker or limited RAM, otherwise causing failure. So we add num_workers when evaluating.
For example, In docker environment, the shared memory is limited causing specific error:
References: https://github.com/pytorch/pytorch/issues/5040 https://github.com/ultralytics/yolov3/issues/1151