Transcribe
by Ben Grande · ai.useion.transcribe
Not TrustedPackage4+
Speech to text for your other apps, on your Personal Cloud's CPU: Whisper writes down what is said in a recording, in 99 languages, and gives every word the second it starts and ends, for captions, search and notes. A sound or video file in (wav, mp3, m4a, mp4, mov, webm, ogg, flac, up to 2 hours), the language and the words out, as a job with progress. A package: it has no screen.
Age rating
Rated 4+ by its publisher: suitable for ages 4 and up.
What version 0.1.0 may do
Internet
It asks to reach nothing on the internet.
While your Personal Cloud builds it, the build asks to download from:
- pypi.org
- files.pythonhosted.org
- huggingface.co
- any address under huggingface.co
- any address under hf.co
Personal Cloud tools
None. It uses nothing from your Personal Cloud beyond its own storage.
Notifications
It does not send push notifications.
Storage
Up to 4 GB on your Personal Cloud.
What your apps can use
Your other web apps on your Personal Cloud can use these once you let each one, on its own install:
- API “transcribe”: A recording in, its words out, as jobs: POST /transcribe with the sound or video file as the body (?language=en when it is known); GET /jobs/<id> for progress and the language, /jobs/<id>/transcript for the language and every word with its start and end in seconds. api:ai.useion.transcribe/transcribe
Add to Personal Cloud
Sign in to add this web app to a Personal Cloud of yours. It runs there, on your own computer, and you open it from there; nothing is installed on your phone.
Versions
0.1.0 Latest Shown above
Published · a ready-built bundle
First release: speech to text for your other apps. Whisper on your Personal Cloud's CPU writes down what is said in a sound or video file, with the language and every word's start and end, as a job with progress (api:ai.useion.transcribe/transcribe). The model is fetched when your hub builds it; it needs no internet afterwards.