DeploymentService

Add a deployment adapter

Changed on

Attaches a LoRA adapter to a deployment. If the deployment is at adapter capacity, force can evict the oldest adapter.

post/v2/projects/{projectId}/endpoints/{endpointId}/deployments/{deploymentId}/adapters

Request

  • Base URL
    https://api.together.ai
  • URL
    https://api.together.ai/v2/projects/{projectId}/endpoints/{endpointId}/deployments/{deploymentId}/adapters
  • Auth
    HTTP bearer

Path parameters

projectIdstring required

Project identifier.

endpointIdstring required

Endpoint identifier.

deploymentIdstring required

Deployment identifier.

Request body

adapterModelIdstring required

Adapter model identifier to attach.

adapterRevisionIdstring

Optional adapter revision to pin. If omitted, the latest revision is resolved at request time.

forceboolean

Whether to evict the oldest adapter if the deployment is at adapter capacity.

Response

OK

adapterModelIdstring required

Adapter model identifier attached to the deployment.

desiredRevisionIdstring required

Adapter revision pinned on the deployment.

etagstring required

Row-level etag required for UpdateAdapter and RemoveAdapter.

adapterModelstring

Resource name of the adapter model, using projects/{projectId}/models/{adapterModelId}.

desiredRevisionstring

Resource name of the adapter model revision pinned on the deployment, using projects/{projectId}/models/{adapterModelId}/revisions/{revisionId}.

idstring required

Adapter attachment dad_ id used by get, update, and remove; stale ids never address a replacement after remove and re-add.

Changes