Build with OpenAI’s Whisper model in five minutes

Whisper` is OpenAI's open-source neural network for automatic speech recognition, offering a paradigm shift from "cool one day" technology to "this technology has arrived". It can capture multiple languages with stunning accuracy and is deployable as a pre-trained model or packaged and deployed using Truss, an open-source model packaging and serving library. The model invocation involves pre-processing work to turn the input into something it can use, and its deployment on Baseten takes just a few clicks. The author, a new grad software engineer, successfully deployed Whisper using Truss and has worked with the model in various situations, including a high-throughput project for Patreon.