Differences between Azure Whisper model and open-AI license on Github

Question

What are the differences between the Whisper model(s) available in Azure and those distributed under the MIT open-AI license on Github?

Answer 1

Thanks for the question, When you consume Whisper through Azure AI Speech endpoint, you can process much bigger, up to 1Gb files. More detailed info can be found here.

And if you need a quick-start guide on how to use Whisper model offline locally (e.g., "tiny" or "base" one) or access it through Azure OpenAI in a Gradio UI, feel free to re-use resources from this GitHub repo