Skip to content

Automatic Speech Recognition server on Docker using OpenAI Whisper and Gradio

Notifications You must be signed in to change notification settings

alessiodevoto/whisper-server

Repository files navigation

Dockerized Whisper Automatic Speech Recognition

Deploy a speech to text demo with OpenAI Whisper Automatic Speech Recognition via Docker.

Setup (in 5 minutes)

First of all, make sure you are in a directoy containing both this repository and model weights. Then, build the Docker image:

docker build -t repo/whisper-server:0.0.1 -f whisper-server/Dockerfile .

When building the image we copy app files and model weights from local file system. In case the directories containing app files and models are at a different location, you should edit the dockerfile accordingly.

You can then run a container using the run_service.sh script. This will create a new container and start up a server inside it.

bash run_service.sh --port <your_port> --gpu_index <cuda_device_index> --model_size <model_size> --image_name <image_name>

Beware that:

  • if you provide a port which is already allocated, the container will stop immediately
  • <image_name> is the name you gave to your image when your built it.
  • <cuda_device_index> is the index of the gpu you want to use for inferece. Defaults to 0.
  • <model_size> is the size of the whisper model you want to use. You can choose between tiny, medium and large-v3 (recommended)

About

Automatic Speech Recognition server on Docker using OpenAI Whisper and Gradio

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published