永久性地将常量注入到Tensorflow图中以进行推理

Question

我用占位符训练模型is_training:

is_training_ph = tf.placeholder(tf.bool)

但是一旦完成训练和验证,我想永久地false为这个值注入常​​数,然后"重新优化"图形(即使用optimize_for_inference).有什么东西freeze_graph会这样做吗？

Answer 1

一种可能性是使用tf.import_graph_def()函数及其input_map参数来重写图中该张量的值.例如,您可以按如下方式构建程序:

with tf.Graph().as_default() as training_graph:
  # Build model.
  is_training_ph = tf.placeholder(tf.bool, name="is_training")
  # ...
training_graph_def = training_graph.as_graph_def()

with tf.Graph().as_default() as temp_graph:
  tf.import_graph_def(training_graph_def,
                      input_map={is_training_ph.name: tf.constant(False)})
temp_graph_def = temp_graph.as_graph_def()

构建之后temp_graph_def,您可以将其用作输入freeze_graph.

另一种可能freeze_graph与optimize_for_inference脚本(它们对变量名和检查点键做出假设)更兼容的替代方法是修改TensorFlow的graph_util.convert_variables_to_constants()函数,以便它转换占位符:

def convert_placeholders_to_constants(input_graph_def,
                                      placeholder_to_value_map):
  """Replaces placeholders in the given tf.GraphDef with constant values.

  Args:
    input_graph_def: GraphDef object holding the network.
    placeholder_to_value_map: A map from the names of placeholder tensors in
      `input_graph_def` to constant values.

  Returns:
    GraphDef containing a simplified version of the original.
  """

  output_graph_def = tf.GraphDef()

  for node in input_graph_def.node:
    output_node = tf.NodeDef()
    if node.op == "Placeholder" and node.name in placeholder_to_value_map:
      output_node.op = "Const"
      output_node.name = node.name
      dtype = node.attr["dtype"].type
      data = np.asarray(placeholder_to_value_map[node.name],
                        dtype=tf.as_dtype(dtype).as_numpy_dtype)
      output_node.attr["dtype"].type = dtype
      output_node.attr["value"].CopyFrom(tf.AttrValue(
          tensor=tf.contrib.util.make_tensor_proto(data,
                                                   dtype=dtype,
                                                   shape=data.shape)))
    else:
      output_node.CopyFrom(node)

    output_graph_def.node.extend([output_node])

  return output_graph_def

...然后你可以training_graph_def像上面那样构建,并写道:

temp_graph_def = convert_placeholders_to_constants(training_graph_def,
                                                   {is_training_ph.op.name: False})