Unsupported language?

[dainiska]

Hello,

I have a question regarding ACE-Step 1.5 language and sacred music support.

I noticed that the language list contains:

la (Latin)
lt (Lithuanian)

but I cannot find Latvian (lv).

Is Latvian currently supported in ACE-Step 1.5 models?
If yes:

what language code should be used?
is Latvian automatically detected?

I am also curious about the inclusion of Latin (la) in the supported languages.

Since Latin is mainly used today in:

Gregorian chant
traditional church music
medieval sacred singing
Catholic liturgy

I expected the model to show some familiarity with:

Gregorian chant
church organ style
medieval sacred vocal phrasing

However, when prompting things like:

“Gregorian chant”
“church organ”
“medieval sacred music”

the model still tends to generate modern vocal and musical styles instead of traditional sacred music.

Is Gregorian chant or traditional sacred music actually part of the ACE-Step training data, or would a LoRA be needed for this kind of generation?

Also, does the selected language affect only pronunciation, or also vocal style and acoustic rendering?

Thank you.