How to compile a model trianed with tensorflow estimator

[idata]

I have trained a tensorflow model using the code for tensorflow triplet loss. The model was trained using a Tensorflow estimator. I tried to freeze the graph and run mvNCCompile but I'm getting errors. The list of operations is the following:

global_step/Initializer/zeros/shape_as_tensor

global_step/Initializer/zeros/Constglobal_step/Initializer/zerosglobal_stepglobal_step/Assignglobal_step/readglobal_step/IsVariableInitializedglobal_step/cond/Switchglobal_step/cond/switch_tglobal_step/cond/switch_fglobal_step/cond/pred_idglobal_step/cond/readglobal_step/cond/read/Switchglobal_step/cond/Switch_1global_step/cond/Mergeglobal_step/add/yglobal_step/addfilenamescompression_typebuffer_sizebuffer_size_1seedseed2countbatch_sizebuffer_size_2IteratorTFRecordDatasetMapDatasetShuffleDatasetRepeatDatasetBatchDatasetPrefetchDatasetMakeIteratorIteratorToStringHandleIteratorGetNextReshape/shapeReshapemodel/block_1/conv2d/kernel/Initializer/random_uniform/shapemodel/block_1/conv2d/kernel/Initializer/random_uniform/minmodel/block_1/conv2d/kernel/Initializer/random_uniform/maxmodel/block_1/conv2d/kernel/Initializer/random_uniform/RandomUniformmodel/block_1/conv2d/kernel/Initializer/random_uniform/submodel/block_1/conv2d/kernel/Initializer/random_uniform/mulmodel/block_1/conv2d/kernel/Initializer/random_uniformmodel/block_1/conv2d/kernelmodel/block_1/conv2d/kernel/Assignmodel/block_1/conv2d/kernel/readmodel/block_1/conv2d/bias/Initializer/zeros/shape_as_tensormodel/block_1/conv2d/bias/Initializer/zeros/Constmodel/block_1/conv2d/bias/Initializer/zerosmodel/block_1/conv2d/biasmodel/block_1/conv2d/bias/Assignmodel/block_1/conv2d/bias/readmodel/block_1/conv2d/dilation_ratemodel/block_1/conv2d/Conv2Dmodel/block_1/conv2d/BiasAddmodel/block_1/Relumodel/block_1/max_pooling2d/MaxPool

[other convolutions…]

model/block_4/max_pooling2d/MaxPool

model/Flatten/flatten/Shapemodel/Flatten/flatten/strided_slice/stackmodel/Flatten/flatten/strided_slice/stack_1model/Flatten/flatten/strided_slice/stack_2model/Flatten/flatten/strided_slicemodel/Flatten/flatten/Reshape/shape/1model/Flatten/flatten/Reshape/shapemodel/Flatten/flatten/Reshapemodel/fc_1/dense/kernel/Initializer/random_uniform/shape

[fully connected operations…]

model/fc_2/dense/bias/read

model/fc_2/dense/MatMulmodel/fc_2/dense/BiasAdd

[other operations]

I am confused about what the input node to be passed to mvNCCompile should be, since there is not a "input" node.