Clone dataset
Copy examples from a source dataset into an existing target dataset.
Each copied example gets a new ID and keeps the inputs, outputs, metadata, and source run ID of its source example. Attachments are copied when blob storage is enabled. Copying the same examples twice adds them twice. Requires the datasets:clone permission on both datasets. Returns the ID, dataset ID, creation and modification times, and source run ID of each new example. Without example IDs, a source or target that is not found can return an empty list instead of an error.
Authorizations
Body
ID of the dataset in this workspace that receives the copied examples. Its current examples are kept.
ID of the dataset in this workspace to copy examples from. It is not modified.
Only modifications made on or before this time are included. If None, the latest version of the dataset is used.
IDs of source examples to copy. Defaults to every example in the source version. When IDs are given and nothing is copied, because none match or either dataset is not found, the request fails with 404.
Split name, or list of split names, assigned to every copied example in place of its source splits. Defaults to keeping the source splits.
List of resource tag value IDs to attach to the target dataset, up to 100. Requires the datasets:tag_on_create permission.
100Response
Successful Response