I am not usually chanced enough to listen to/watch long video/audio files :). So I built this, an open-source AI project for transcribing video and audio files, even youtube videos (from the link). It utilizes a pretrained ASR(Automatic Speech Recognition) model from the Huggingface transformers library, Streamlit for the web UI, and other python libraries for audio processing and model loading.
[27-10-24] -> I rewrote the entire thing, with extra features, and a faster model (Moonshine-base with a JAX backend)
NB: Larger files and longer videos take a longer time to transcribe...(of course)
Check it out => vidkit on HuggingFace spaces