Gets all validation sets available to the user that a rapid with the provided parameters can be added to.
Changed onMatching is strict: a validation set is only returned when its asset type, modality, and prompt type equal the provided parameters exactly — the same constraint enforced when actually adding a rapid. Uninitialized validation sets (no modality yet — they adopt the classifiers of the first rapid added to them) are compatible with any rapid and are always returned, even while still empty. Results cover the caller's own and public validation sets; callers allowed to query across organizations see compatible sets from every organization. This differs from /validation-set/recommended, which uses loose overlap matching and may return sets that then reject the rapid. Returns an empty array when no compatible set exists.
Request
- Base URLhttps://api.rapidata.ai
- URLhttps://api.rapidata.ai/validation-sets/compatible
- AuthOpenID Connect (https://auth.rapidata.ai/.well-known/openid-configuration, scopes: openid, profile, email, offline_access, api)
Query parameters
The asset type that the validation set must contain. An asset type can be an image, video, audio, text, or any combination of these.
The rapid modality that the validation set must contain. The modality is the type of rapid such as classify, compare, locate, etc.
The prompt type that the validation set must contain. A prompt type is the additional information that is presented to the user when solving a rapid. For example, a prompt type can be either text or an asset if the context is an image or video.
An instruction used to find validation sets that have the same instruction. An instruction is a text that is presented to the user when solving a rapid that explains how to solve the rapid.
Response
OK
Changes
- info
the security scope
apiwas added to the endpoint's security schemeOpenIdConnect
- info