search

ahmetoner / whisper-asr-webservice

OpenAI Whisper ASR Webservice API
https://ahmetoner.github.io/whisper-asr-webservice
MIT License
2.14k stars 380 forks source link
asr automatic-speech-recognition docker openai-whisper speech speech-recognition speech-to-text

readme

Release Docker Pulls Build Licence

Whisper ASR Webservice

Whisper is a general-purpose speech recognition model. It is trained on a large dataset of diverse audio and is also a multitask model that can perform multilingual speech recognition as well as speech translation and language identification. For more details: github.com/openai/whisper

Features

Current release (v1.6.0) supports following whisper models:

Quick Usage

CPU

docker run -d -p 9000:9000 -e ASR_MODEL=base -e ASR_ENGINE=openai_whisper onerahmet/openai-whisper-asr-webservice:latest

GPU

docker run -d --gpus all -p 9000:9000 -e ASR_MODEL=base -e ASR_ENGINE=openai_whisper onerahmet/openai-whisper-asr-webservice:latest-gpu

for more information:

Documentation

Explore the documentation by clicking here.

Credits