I am using this code to import an Inception V5 model:
with tf.gfile.FastGFile(modelFileName, 'rb') as f:
graph_def = tf.GraphDef()
graph_def.ParseFromString(f.read())
t_input = tf.placeholder(np.float32, name='input') # define the input tensor
tf.import_graph_def(graph_def, {'input':t_input})
And this works fine, however if i try the same code on an Inception V3 model then i get the error "Attempted to map inputs that were not found in graph_def: [input:0]"
I have confirmed that there is a tensor called input in my model so i dont know why its giving me this error?