/arcp_async_fastapi

Primary LanguagePythonGNU General Public License v3.0GPL-3.0

arcp_async_fastapi

  • The project is a REST API deployed using FastAPI
  • We've used transfer learning on resnet50 model to detect age of person given image
  • The API expects a base64 encoded string as input
  • I've used UTKFACE dataset for training
  • The model was trained using tensorflow 2.0
  • Then it was converted to onnx format
  • The model achieved a rmse of 5
  • You can also run the project using docker

steps to run project using docker

  • Build image docker build -t <image_name> replace <image_name> with the name of your docker image
  • Run docker docker run --name <name_of_container> <image_name> replace <name_of_container> with your container name