In this project, we test gemini AI with multimodal prompts. The supported model is gemini-pro-vision
.
- Complete the following configuration according to Google Cloud Doc.
- Select a proper Google Cloud project
- Enable the Vertex AI API
- Install the Google Cloud CLI
- Update and install gcloud components
-
Create a service account and generate a service account key according to this document: https://cloud.google.com/vertex-ai/docs/start/client-libraries#before_you_begin
-
Install Google CLoud Client Library
npm install @google-cloud/aiplatform
- Configure your own google cloud service account credential in backend
process.env.GOOGLE_APPLICATION_CREDENTIALS = 'Your Own Service Account Credential Path';
npm run serve
node index.js