Dataset Collection Tags

Collection tags are a commercial-edition feature in FastGPT. They let you tag and categorize data collections within a Dataset for more efficient data management.

You can also use tags as collection filters during Dataset searches for more precise results.

Basic Tag Operations

On the Dataset detail page, you can manage tags with the following operations:

• Create a tag
• Rename a tag
• Delete a tag
• Assign a tag to multiple collections
• Add multiple tags to a single collection

You can also filter collections by tags.

Collection Filtering in Dataset search

Tags can be used to filter collections during Dataset searches by filling in the "Collection Filter" field. Here's an example:

{
  "tags": {
    "$and": ["Tag 1", "Tag 2"],
    "$or": ["When $and tags are present, $and takes effect and $or is ignored"]
  },
  "createTime": {
    "$gte": "YYYY-MM-DD HH:mm format, matches collections created after this time",
    "$lte": "YYYY-MM-DD HH:mm format, matches collections created before this time. Can be used together with $gte"
  }
}

Two important notes:

• Tag values can be a string tag name or null, where null represents collections with no tags assigned
• There are two filter condition types: $and and $or. When both are set, only $and takes effect