This is the Caffe agent for MLModelScope, an open-source framework and hardware agnostic, extensible and customizable platform for evaluating and profiling ML models across datasets / frameworks / systems, and within AI application pipelines.

Currently it has most of the vison models from Caffe Model Zoo built in. More built-in models are comming. One can evaluate the ~40 models on any systems of insterest with either local Caffe installation or Caffe docker images.

Check out MLModelScope and welcome to contribute.

Install go if you have not done so. Please follow Go Installation.

Download and install the MLModelScope Caffe Agent:

go get -v github.com/rai-project/caffe

The agent requires The Caffe C library and other Go packages.

You can install the dependency through go get.

cd $GOPATH/src/github.com/rai-project/caffe
go get -u -v ./...

Or use Dep.

dep ensure -v

This installs the dependency in vendor/.

Note: The CGO interface passes go pointers to the C API. This is an error by the CGO runtime. Disable the error by placing

export GODEBUG=cgocheck=0

in your ~/.bashrc or ~/.zshrc file and then run either source ~/.bashrc or source ~/.zshrc

The Caffe C library is required.

If you use Caffe Docker Images (e.g. NVIDIA GPU CLOUD (NGC)), skip this step.

Refer to go-caffe for caffe installation.

Refer to External services.

Refer to Use within TensorFlow Docker Images.

Refer to Usage