Making OpenAPI's Open Whisper available via ReST API locally.
You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
Go to file
Drew Bednar 0ce177b290 Adding usage example 2 years ago
local_whisper #3 Adding WhisperSettings (#7) 2 years ago
pyinvoke Adding usage example 2 years ago
tests #3 Adding WhisperSettings (#7) 2 years ago
.gitignore Adding usage example 2 years ago
.pre-commit-config.yaml Working initial implementation 2 years ago
LICENSE Initial setup 2 years ago
README.md Adding usage example 2 years ago
dev_requirements.in First test 2 years ago
dev_requirements.txt First test 2 years ago
macos-py3.11-dev_requirements.txt #4 Adding macos compiled requirements (#6) 2 years ago
macos-py3.11-requirements.txt #4 Adding macos compiled requirements (#6) 2 years ago
pyproject.toml Working initial implementation 2 years ago
requirements.in Working initial implementation 2 years ago
requirements.txt Working initial implementation 2 years ago
tasks.py Initial setup 2 years ago

README.md

LocalWhisper

Making OpenAPI's Open Whisper available via ReST API locally.

Features

  • Transcription
  • Translation (Not-planned, but open to PRs)

Usage Example

Similar to the OpenAI API we can post an audio file to /audio/transcriptions as a multipart/form-data type.

curl --request POST \
  --url http://localhost:9000/audio/transcriptions \
  --header 'Content-Type: multipart/form-data' \
  --header 'Accept: application/json' \
  --form 'file=@/path/to/file/example.wav;type=audio/wav'
  --form 'model=whisper-1'

At present the model form field is not required or supported. The WHISPER_BASE_ASR_MODEL server configuration will be used to determine the Whisper model to use.