Docker Hub | API documentation | Comprehensive post
ONNX model and Docker service for DeepCreamPy.
Exporting DeepCreamPy's models from Tensorflow to ONNX improves inference times when running on CPU and significantly reduces memory usage. Large parts of the model graph that are not useful for inference are eliminated through the conversion process.
This repository uses Deepshift's mirror of DeepCreamPy.
docker run --publish $PORT:$PORT --env PORT=$PORT --detach nanoskript/deepcreampy