view post Post 3990 Im working on a 23M ASR model, trained on 100k hours of audio See translation 4 replies ยท ๐ฅ 9 9 ๐ 2 2 ๐ 2 2 + Reply