When train on multi-gpu，encounter ‘Read -1, expected 6144, errno = 1’

TRI-ML / packnet-sfm

TRI-ML Monocular Depth Estimation Repository

https://tri-ml.github.io/packnet-sfm/

MIT License

1.23k stars 242 forks source link

When train on multi-gpu，encounter ‘Read -1, expected 6144, errno = 1’ #198

Open angiend opened 2 years ago

angiend commented 2 years ago

Sorry to bother you, when i run the python3 scripts/train.py configs/train_ddad.yaml in the docker envs, and get the error like below,but run it on single gpu is ok. would give some advice ?thanks .