Confusing error with

[dentistfrankchen]

When I ran my program, I got the following error:
Could not load library libcudnn_cnn_train.so.8. Error: /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn_ops_train.so.8: symbol _Z22cudnnGenericOpTensorNdILi3EE13cudnnStatus_tP12cudnnContext16cudnnGenericOp_t21cudnnNanPropagation_tPKdPKvPK17cudnnTensorStructS8_S8_SB_S8_S8_SB_Pv, version libcudnn_ops_infer.so.8 not defined in file libcudnn_ops_infer.so.8 with link time reference

What does this mean?

[github-actions]

This issue is stale because it has been open 90 days with no activity. This issue will be closed in 30 days unless new comments are made or the stale label is removed. To skip these checks, apply the "lifecycle/frozen" label.

[ArangoGutierrez]

@dentistfrankchen Hey could you provide more context on how did you ran into that error log?
Which Version of the Toolkit are you using?
On Which System?
Did you follow the documented installation process for the toolkit? https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html#installation

[elezar]

The error is an application-level error that seems to indicate that a required runtime library cannot be found. Please repeat the experiment with an updated driver and runtime container image and reopen if you're still having problems.