jmservera/resnet50-embeddings repository overview

⁠Resnet50 embeddings container

Source code⁠

This is a simple container to get the embeddings of an image using a resnet50 model. The model is loaded from an onnx file from this project: https://huggingface.co/jxtc/resnet-50-embeddings⁠

⁠Local run

To run this container ensure that you publish the port. The default one is the 8000.

docker run -p 8000:8000 jmservera/resnet50-embeddings

You can change the port using the UVICORN_PORT environment variable.

Once the container is running you can test the container with something like this:

curl -F "file=@./island.jpeg" http://localhost:8000/embed

Or use some code like this:

import requests

url = "http://localhost:8000/embed"
files = {'file': open('image.jpg', 'rb')}
response = requests.post(url, files=files)
print(response.json())