Open Internet by MindsNet

Vaibhavs10/insanely-fast-whisper

Insanely Fast Whisper is a command-line interface (CLI) tool that enables fast transcription of audio files using OpenAI's Whisper model. It leverages Transformers, Optimum, and Flash Attention to achieve blazingly fast transcription speeds. With this tool, users can transcribe 150 minutes of audio in under 98 seconds. Best for: Developers, researchers, and audio professionals seeking fast and efficient transcription solutions Use cases: Rapid transcription of large audio datasets for machine learning model training; Quickly generating subtitles for video content; Transcribing interviews or lectures for qualitative research Highlights: Transcribes 150 minutes of audio in under 98 seconds with Whisper Large v3; Optimized for speed using fp16, batching, and Flash Attention 2; Supports various models, including Whisper Large v3 and Distil-Large-V2

Computing & Technology, Computer Science, Artificial Intelligence