An example implementation of Uber's Generative Teaching Network (GTN) with Keras (tensorflow)

• GPU: docker build -f gpuDockerfile -t kerasgtn .

• CPU: docker build -f cpuDockerfile -t kerasgtn .

docker run --gpus all -u $(id -u):$(id -g) -it --rm -v $PWD:/tf kerasgtn:latest bash

Remove --gpus all if using CPU Dockerfile

docker run --gpus all -u $(id -u):$(id -g) -it --rm -v -p 8888:8888 $PWD:/tf kerasgtn:latest

Remove --gpus all if using CPU Dockerfile

from kerasgtn.gtn import GTN


class MyGTN(GTN):
    def __init__(self, **kwargs):
        super(MyGTN, self).__init__(**kwargs)
    
    def get_generator(self, input_layer):
        <implement>
    
    def get_learner(self, real_input, teacher):
        <implement>


gtn = MyGTN(input_shape=input_shape, n_classes=n_classes)
gtn.train(...)