Whisper Jax: Transcribe a 1 hour of audio in under 15 seconds

Re:Whisper Jax: Transcribe a 1 hour of audio in under 15 sec... by mysterybox on Saturday April 22, 2023 @01:54AM (#35663773)

Whisper JAX is an optimised implementation of the Whisper model by OpenAI. It runs on JAX with a TPU v4-8 in the backend. Compared to PyTorch on an A100 GPU, it is over 70x faster, making it the fastest Whisper API available.

[ Parent | on HN ]

Re:Whisper Jax: Transcribe a 1 hour of audio in under 15 sec... by __henil on Saturday April 22, 2023 @03:10AM (#35664111)

how does it compare to whisper.cpp?

[ Parent | on HN ]

Re:Whisper Jax: Transcribe a 1 hour of audio in under 15 sec... by neximo64 on Saturday April 22, 2023 @03:35AM (#35664225)

70 times faster

[ Parent | on HN ]

Re:Whisper Jax: Transcribe a 1 hour of audio in under 15 sec... by ttt3ts on Saturday April 22, 2023 @09:37AM (#35666489)

Not a good comparison. CPU vs TPU. Also, whisper.cpp perf varies a ton depending on hardware. Avx512 or neural engine?

[ Parent | on HN ]