No. We do not access, copy, or upload any of your train/test/dev data or models.

If you are using the community open-source version or the on-premise enterprise version, all your data, workload, models, logs, and metadata generated from your experiments stay within your secure cluster.

If you are using our fully managed version, all of your data, models, and workload is managed by your servers, we only manage the Platform itself for you.