Anyone know open source models or papers for splitting audio (that contains human speech and other sounds) into two separate audio files, one containing human speech only and one containing other sounds?

Have you tried Audio Shake?

https://www.audioshake.ai

⌐◨-◨ ‘ing around. 
Nouns Fest founder.
Sundae Songs, music publishing, founder.

www.fishandfriends.fun

Never heard of it! Im looking for something that works well with regular speech vs vocals + music

Maybe worth experimenting with Audio Shake, might be able to do it…if you try keep me posted pls