Best Practices for WandB Artifacts

Hi there, I am currently exploring a best practices for wandb artifacts. Can anybody let me know there best way to create artifacts rather than saving each runs as artifacts

Hi Supriya,

Runs produce artifacts, but that doesn’t mean that runs are artifacts.

Here (ttps://docs.wandb.ai/guides/artifacts/artifacts-faqs) are our documents on artifacts. When you start a run you can upload an artifact/multiple artifacts during it, in order to keep version control of that artifact. For example, you can upload your dataset to artifacts and every time your dataset changes you will be able to see its new version inside of the artifact and control that dataset’s different versions. Or you could upload your model.

Think of artifacts as a version control system something similar to github.

Cheers!
Artsiom

Hi Supriya,

We wanted to follow up with you regarding your support request as we have not heard back from you. Please let us know if we can be of further assistance or if your issue has been resolved.

Best,
Weights & Biases

Hi Supriya, since we have not heard back from you we are going to close this request. If you would like to re-open the conversation, please let us know!

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.