> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.nvidia.com/sdgm/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.nvidia.com/sdgm/_mcp/server.

# timeframe_step

#### `timeframe_step: <integer>` (Optional)

## Description

Defines the step size of generating time intervals for training table generation. This only applies to temporal queries (queries that include a temporal aggregation such as `SUM(TRANSACTIONS.AMOUNT, 0, 2, days)`). The time unit of this step is the same as the time unit in the aggregation.

* `timeframe_step` must be `< full time range - target aggregation`.

### Supported Task Types

* Temporal

### Example

For example, the value of `timeframe_step` will be in `days` for the following pQuery:

```yaml Field
PREDICT SUM(TRANSACTIONS.AMOUNT, 0, 2, days)
FOR EACH CUSTOMERS.CUSTOMER_ID
```

### Default Values

| run\_mode | Default Value |
| --------- | ------------- |
| FAST      | None          |
| NORMAL    | None          |
| BEST      | None          |