Hi @<1560798754280312832:profile|AntsyPenguin90> , for Data versioning, you mean where ClearML Dataset files are uploaded? If so, you can just create an NFS mount on the training machine and point the output_uri
there
yeap, but could it support 2 location at the same time?
- output_uri_A: for data versioning
- output_uri_B: for data of experiment artifacts