wenet Docker & Self-Hosted Deployment

GitHub

Production First and Production Ready End-to-End Speech Recognition Toolkit

5,247 stars Python Python 60 files · ~51,430 tokens #asr#automatic-speech-recognition#conformer#e2e-models#production-ready#pytorch#speech-recognition#transformer
1. Pull Image (DOCKERHUB) docker pull wenet-e2e/wenet:latest
2. Docker Run (1-Click Launch) docker run -d --name wenet --restart always wenet-e2e/wenet:latest
3. Podman Run podman run -d --name wenet --restart always wenet-e2e/wenet:latest
4. Skopeo Inspect (Zero-Pull Manifest) skopeo inspect docker://docker.io/wenet-e2e/wenet:latest
5. ORAS Pull (OCI Artifact) oras pull wenet-e2e/wenet:latest
runtime/libtorch/docker/Dockerfile 
FROM ubuntu:latest
MAINTAINER <[email protected]>
ENV DEBIAN_FRONTEND=noninteractive
RUN apt-get update && apt-get install -y git cmake wget build-essential
RUN git clone https://github.com/wenet-e2e/wenet.git /home/wenet
ARG model=20210618_u2pp_conformer_libtorch.tar.gz
RUN wget -P /home https://wenet-1256283475.cos.ap-shanghai.myqcloud.com/models/aishell2/$model
RUN tar -xzf /home/$model -C /home
ARG src=/home/wenet/runtime/libtorch
RUN cmake -B $src/build -S $src -DCMAKE_BUILD_TYPE=Release -DGRAPH_TOOLS=ON && cmake --build $src/build