Use custom_granularities in grain_to_date for a cumulative metric in semantic layer

Help
1

The problem I’m having

I am trying to create a cumulative metric with grain_to_date set as “fiscal_month” which is defined in a custom calendar, but I’m getting an error.

The context of why I’m trying to do this:

My Month to Date metric is driven based on fiscal calendar

What I’ve already tried:

Define model as follows:

models:
  - name: d_fiscal_calendar
    description: This is a fiscal year calendar that will be used as custom time spine.
    time_spine:
      standard_granularity_column: day_key
      custom_granularities:
        - name: fiscal_month
          column_name: mth_key
    columns:
      - name: day_key
        granularity: day
      - name: mth_key

And define metric as follows:

metrics:
  - name: mtd_sales
    description: 'MTD Sales Retail, based on fiscal calendar'
    type: cumulative
    label: mtd_sales
    type_params:
        measure: 
          name: sales
        grain_to_date: fiscal_month

error message i’m getting:

ValueError: Invalid enum value: `fiscal_month` in enum TimeGranularity