Bart model with onnxruntime deploying to Google Cloud Run

This repository contains the code for deploying a ONNX bart translation model to Google Cloud Run, It is based on the insights shared in the blog post 'My Journey to a serverless transformers pipeline on Google Cloud' from the Hugging Face website."

Prerequisites

• Install gcloud CLI
• Install docker

Deploying to Google Cloud Run

gcloud builds submit --tag gcr.io/<project-id>/bart-translation-zh-yue-onnx
gcloud run deploy --image gcr.io/<project-id>/bart-translation-zh-yue-onnx --platform managed \
  --command "serve" \
  --cpu 4 \
  --memory 8Gi \