getWeights

Retrieves weights for a dataset from the data-cluster API. For each weight it also attempts to download the weight file from S3 into a temporary local file. Returns a single weight selected by index, plus the total count of weights found.

Parameters

Param	Type	Required	Description
dataset_id	integer	Yes	Dataset ID to get weights from
weight_index	integer	No (default: 0)	Zero-based index of the weight to return. Raises an error if out of range
type	string?	No	Filter weights by type before selecting by index. Example values: voice_id, conversation, lora. When omitted, no filtering is applied

Output

Field	Type	Description
weight	object	The selected weight object
count	integer	Total number of weights found after any type filtering

weight object:

Field	Type	Description
weight_id	integer	Weight ID
external_id	string?	External ID (e.g. an ElevenLabs voice ID)
bucket_path	string	S3 bucket path to the weight file
local_path	string?	Absolute path to the downloaded temporary file. null if the download was skipped or failed
dataset_id	integer?	Source dataset ID
ai_model_id	integer?	AI model ID associated with the weight
model_metadata	object?	Arbitrary metadata stored on the weight record

note

local_path will be null if bucket_path is empty or the download fails. The node does not raise an error in that case.

Example

{
  "id": "getWeights",
  "type": "getWeights",
  "data": {
    "label": "Get Weights",
    "isExecuted": false,
    "handles": ["outputs"],
    "schema": {},
    "params": {
      "dataset_id": { "value": 42, "isExpression": false, "isAttachedToInputNode": false },
      "weight_index": { "value": 0, "isExpression": false, "isAttachedToInputNode": false },
      "type": { "value": "voice_id", "isExpression": false, "isAttachedToInputNode": false }
    },
    "inputs": [], "outputs": [], "errors": []
  },
  "position": { "x": 0, "y": 0 },
  "isSelected": false,
  "isDragging": false
}

The external_id from this node is typically passed as voice_id to the voiceGeneration node.