Whisper is a versatile speech recognition model that can transcribe speech as well as perform multi-lingual translation and recognition tasks. It was trained on diverse datasets to give human-level speech recognition performance without the need for fine-tuning.
Graphcore's IPU (Intelligence Processing Unit) is a completely new kind of massively parallel processor to accelerate machine intelligence. Developers can access advanced, cost-efficient IPU compute on-demand in the cloud for building, fine-tuning and deploying AI models such as Whisper.
| Notebook | Framework | Type | Try for Free |
|---|---|---|---|
| Speech Transcription on IPUs using Whisper Tiny | Hugging Face | Inference |  |
| Multi-lingual ASR Transcription on IPUs using Whisper - Fine-tuning | Hugging Face | Fine-tuning | |
| Speech Transcription on IPUs using Whisper - Quantized Inference | Hugging Face | Inference | |
In the Speech Transcription on IPUs using Whisper Tiny | Hugging Face notebook, we demonstrate using Whisper Tiny for speech recognition and transcription on the IPU. We use Optimum Graphcore - a new open-source library and toolkit that enables developers to access IPU-optimized models certified by Hugging Face. Only a few lines of code are needed to get this state-of-the-art automated speech recognition model running on IPUs.
In the Multi-lingual ASR Transcription on IPUs using Whisper - Fine-tuning notebook, we demonstrate fine-tuning for multi-lingual speech transcription on the IPU using the Whisper implementation in the 🤗 Transformers library alongside Optimum Graphcore.
In the Speech Transcription on IPUs using Whisper - Quantized Inference notebook, we demonstrate speech transcription on the IPU using the Whisper implementation in the 🤗 Transformers library alongside Optimum Graphcore using INT4 group quantization.
To take your Whisper usage on IPUs further, or speak to an expert, please feel free to contact us.
Join our growing community and interact with AI experts, IPU developers and researchers. Hear the latest IPU news and get access to our newest models.
The contents of this repository are made available according to the terms of the Apache 2.0 license. See the included LICENSE file for details.