New posts in Caffe2

Caffe2: Load ONNX model, and inference single threaded on multi-core host / docker