aws-samples/aws-efa-nccl-baseami-pipeline

Issues

• Add more examples like nvidia-efa-ami_base/nvidia-efa-ml-ubuntu2004.yml for later version of cuda.

  #11 opened a year ago by tchaton
  0

• installing DCGM on gpu compute nodes

  #8 opened 2 years ago by rvencu
  0

• built docker image Dockerfile-cu11-pt22.03.base but nccl test fails

  #7 opened 2 years ago by rvencu
  3

• Errors validating ubuntu2004: unknown configuration key

  #6 opened 3 years ago by drscotthawley
  1

• DataLoader cash when using FI_EFA_USE_DEVICE_RDMA=1

  #4 opened 3 years ago by tohaowu
  7

• InvalidTagKey.Malformed: Tag specification must have at least one tag (error building packer image)

  #3 opened 3 years ago by brannondorsey
  2

• find_ofi_provider:543 NCCL WARN NET/OFI Couldn't find any optimal provider

  #2 opened 3 years ago by tohaowu
  0

• Can't create p4d instance with multiple EFA network interfaces.

  #1 opened 4 years ago by tohaowu
  1