Advanced Prompt Engineering for Automatic Speech Recognition [P]
Reddit - Machine Learning
1 min read
About this article
I've been working on the listening part of my full-duplex speech model MichiAI and I realized that ASR prompting could be very useful. Deepgram allows for word boosting but that doesn't work that well in real word applications. Other thing that is missing is feeding a whole conversation history as context to the ASR model. This could be very useful for voice agents. TLDR, during the testing I realized the model can be fine tuned for prompting with text like: <text>Expect a license plate...
You've been blocked by network security.To continue, log in to your Reddit account or use your developer tokenIf you think you've been blocked by mistake, file a ticket below and we'll look into it.Log in File a ticket
Originally published on April 24, 2026.
Curated by AI News.
