The Future is Fine Tuned (with Dev Rishi, Predibase)
Thinking Machines: AI & Philosophy
00:00
Challenges in Developing End-to-End Language Models for Audio
The chapter explores the difficulties in creating comprehensive language models for audio, highlighting the hurdles in tokenizing and translating speech effectively. It also mentions OpenAI's Whisper as a notable step in this realm, discussing advancements in audio modeling and the surprises in model performance.
Transcript
Play full episode
