Whisper Speech Recognition Model 147

Whisper is an open-source audio-to-text speech recognition model from OpenAI. It is trained on large datasets of diverse audio, and is also a multi-task model capable of performing multilingual speech recognition as well as speech translation and language identification.

Whisper offers five model sizes, providing a balance between speed and accuracy, with four options available for English-only models. Below are the names, approximate memory requirements, and relative speeds of the available models.