JAX implementation of OpenAI's Whisper model for up to 70x speed-up on TPU.
updated at May 12, 2024, 9:30 a.m.
40 +0
4,112 +17
349 +4