It went under radars and was barely noticed when it was released. But it's not a very good model. I'd still say it's among better local music model, especially thanks to license and multilingual capabilities. But I've used it for a bit and it could take many hours to get a desirable output and then some more time to iron out flaws like repetition or missing words.
However the model devs are currently working on Ace-Step v1.5 and it might be released by the end of January and it's leagues ahead of v1 in terms of quality, music and voices diversity, with more predictable output but should have lower hardware requirements. It's currently available to test on the official Discord server.
They said that v1 is more like a concept while v1.5 aims to compete with commercial models.