Tensorflow：从图中删除节点

Question

I'm trying to delete some nodes from graph and save it in .pb

Only needed nodes can be added to new mod_graph_def graph, but the problem that graph still have some references to deleted node in other nodes inputs, but I can't modify inputs of node:

def delete_ops_from_graph():
    with open(input_model_filepath, 'rb') as f:
        graph_def = tf.GraphDef()
        graph_def.ParseFromString(f.read())

    nodes = []
    for node in graph_def.node:
        if 'Neg' in node.name:
            print('Drop', node.name)
        else:
            nodes.append(node)

    mod_graph_def = tf.GraphDef()
    mod_graph_def.node.extend(nodes)

    # The problem that graph still have some references to deleted node in other nodes inputs
    for node in mod_graph_def.node:
        inp_names = []
        for inp in node.input:
            if 'Neg' in inp:
                pass
            else:
                inp_names.append(inp)

        node.input = inp_names # TypeError: Can't set composite field

    with open(output_model_filepath, 'wb') as f:
        f.write(mod_graph_def.SerializeToString())

Answer 1

def delete_ops_from_graph():
    with open(input_model_filepath, 'rb') as f:
        graph_def = tf.GraphDef()
        graph_def.ParseFromString(f.read())

    # Delete nodes
    nodes = []
    for node in graph_def.node:
        if 'Neg' in node.name:
            print('Drop', node.name)
        else:
            nodes.append(node)

    mod_graph_def = tf.GraphDef()
    mod_graph_def.node.extend(nodes)

    # Delete references to deleted nodes
    for node in mod_graph_def.node:
        inp_names = []
        for inp in node.input:
            if 'Neg' in inp:
                pass
            else:
                inp_names.append(inp)

        del node.input[:]
        node.input.extend(inp_names)

    with open(output_model_filepath, 'wb') as f:
        f.write(mod_graph_def.SerializeToString())

Answer 2

Previous answer is good but I would suggest to tie removed node input to the next node input. Like if we have a chain A-input b->B-input c->C-input d->D and going to remove say node B then we should not just remove input c but replace it with input b . Look at code below:

#  remove node and connect its input to follower
def remove_node(graph_def, node_name, input_name):
    nodes = []
    for node in graph_def.node:
        if node.name == node_name:
            assert(input_name in node.input or len(node.input) == 0),\
                "Node input to use is not among inputs of node to remove"
            input_of_removed_node = input_name if len(node.input) else ''
            print("Removing {} and using its input {}".format(node.name, 
                   input_of_removed_node))
            continue
        nodes.append(node)
    
    # modify inputs where required
    # removed name must be replaced with input from removed node
    for node in nodes:
        inp_names = []
        replace = False
        for inp in node.input:
            if inp == node_name:
                inp_names.append(input_of_removed_node)
                print("For node {} replacing input {} 
                       with {}".format(node.name, inp, input_of_removed_node))
                replace = True
            else:
                inp_names.append(inp)
        if replace:
            del node.input[:]
            node.input.extend(inp_names)
    mod_graph_def = tf.GraphDef()
    mod_graph_def.node.extend(nodes)
    return mod_graph_def