view post Post 3987 Im working on a 23M ASR model, trained on 100k hours of audio See translation 4 replies · 🔥 9 9 👍 2 2 🚀 2 2 + Reply