def reduction_cell(graph, prev, cur, out_channels):
cur = squeeze(graph, out_channels, cur)
prev = fit(graph, cur, prev)
ts = list()
outputs = list()
ts.append(seperable_conv(graph, input=prev, out_channels=out_channels,
kernels=(7,7), strides=(2,2), padding="SAME"))
ts.append(seperable_conv(graph, input=cur, out_channels=out_channels,
kernels=(5,5), strides=(2,2), padding="SAME"))
outputs.append(graph.add(ts[0], ts[1]))
ts.append(graph.maxpool2d(input=cur, kernels=(3,3), strides=(2,2), padding="SAME"))
ts.append(seperable_conv(graph, input=prev, out_channels=out_channels,
kernels=(7,7), strides=(2,2), padding="SAME"))
outputs.append(graph.add(ts[2], ts[3]))
ts.append(graph.avgpool2d(input=cur, kernels=(3,3), strides=(2,2), padding="SAME"))
ts.append(seperable_conv(graph, input=prev, out_channels=out_channels,
kernels=(5,5), strides=(2,2), padding="SAME"))
outputs.append(graph.add(ts[4], ts[5]))
ts.append(graph.maxpool2d(input=cur, kernels=(3,3), strides=(2,2), padding="SAME"))
ts.append(seperable_conv(graph, input=outputs[0], out_channels=out_channels,
kernels=(3,3), strides=(1,1), padding="SAME"))
outputs.append(graph.add(ts[6], ts[7]))
ts.append(graph.avgpool2d(input=outputs[0], kernels=(3,3), strides=(1,1), padding="SAME"))
ts.append(outputs[1])
outputs.append(graph.add(ts[8], ts[9]))
return graph.concat(1, outputs)
def normal_cell(graph, prev, cur, out_channels):
cur = squeeze(graph, out_channels, cur)
prev = fit(graph, cur, prev)
ts = list()
ts.append(seperable_conv(graph, input=cur, out_channels=out_channels,
kernels=(3,3), strides=(1,1), padding="SAME"))
ts.append(cur)
ts.append(seperable_conv(graph, input=prev, out_channels=out_channels,
kernels=(3,3), strides=(1,1), padding="SAME"))
ts.append(seperable_conv(graph, input=cur, out_channels=out_channels,
kernels=(3,3), strides=(1,1), padding="SAME"))
ts.append(graph.avgpool2d(input=cur, kernels=(3,3), strides=(1,1), padding="SAME"))
ts.append(prev)
ts.append(graph.avgpool2d(input=prev, kernels=(3,3), strides=(1,1), padding="SAME"))
ts.append(graph.avgpool2d(input=prev, kernels=(3,3), strides=(1,1), padding="SAME"))
ts.append(seperable_conv(graph, input=prev, out_channels=out_channels,
kernels=(3,3), strides=(1,1), padding="SAME"))
ts.append(seperable_conv(graph, input=prev, out_channels=out_channels,
kernels=(3,3), strides=(1,1), padding="SAME"))
assert len(ts) == 10, "Expected 10 tensors, got {}".format(len(ts))
outputs = list()
for i in range(5):
outputs.append(graph.add(ts[2*i], ts[2*i+1]))
return graph.concat(1, outputs)