# `InferenceGoogleVertexAITaskSettings` [interface-InferenceGoogleVertexAITaskSettings]

| Name | Type | Description |
| - | - | - |
| `auto_truncate` | boolean | For a `text_embedding` task, truncate inputs longer than the maximum token length automatically. |
| `top_n` | [integer](./integer.md) | For a `rerank` task, the number of the top N documents that should be returned. |