aburkleaux-amazon / whisper-demo-sagemaker

A demo with OpenAI's Whisper model deployed to a SageMAker endpoint

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

whisper-demo-sagemaker

A demo with OpenAI's Whisper model deployed to a SageMaker endpoint

Steps to set up the demo

  1. Run the notebook in the sagemaker folder which will deploy the Whisper model to a SM endpoint
  2. Run the Cloudformation template in the ec2 folder, either via AWS CLI or in the AWS console (AWS CLI Command: aws cloudformation create-stack --stack-name whisper-demo-stack --template-body file://cf_template.yaml)
  3. Once the EC2 instance is spun up, connect to it via EC2 Instaqnce Connect and run sudo python3 whisper-demo-sagemaker/ec2/gradio-app.py. This will create a public link to the demo, which can be shared with others

About

A demo with OpenAI's Whisper model deployed to a SageMAker endpoint

License:MIT License


Languages

Language:Jupyter Notebook 71.7%Language:Python 24.2%Language:Shell 4.2%