Модель консерватора Tensorflow с порцией TensorFlow

Question

Я хочу экспортировать стандартную модель оценщика для обслуживания с тензорным обслуживанием, но я вижу пример, что мы должны объявить serve_input_receiver_fn .Пример следующий:

feature_spec = {'foo': tf.FixedLenFeature(...),
                'bar': tf.VarLenFeature(...)}

def serving_input_receiver_fn():
  """An input receiver that expects a serialized tf.Example."""
  serialized_tf_example = tf.placeholder(dtype=tf.string,
                                         shape=[default_batch_size],
                                         name='input_example_tensor')
  receiver_tensors = {'examples': serialized_tf_example}
  features = tf.parse_example(serialized_tf_example, feature_spec)
  return tf.estimator.export.ServingInputReceiver(features, receiver_tensors)

Я последовал этому примеру, и мой код такой:

feature_spec = {'user_id': tf.FixedLenFeature([1], tf.int64),
                    'item_id': tf.FixedLenFeature([1], tf.int64),}

def serving_input_receiver_fn():
   serialized_tf_example = tf.placeholder(dtype=tf.string, shape=[None], name='input_tensors')
   receiver_tensors = {'inputs': serialized_tf_example}

   features = tf.parse_example(serialized_tf_example, feature_spec)

   return tf.estimator.export.ServingInputReceiver(features, receiver_tensors)

, и моя модель выглядит так:

def model_fn(features, labels, mode, params):
    user_id = features["user_id"]
    item_id = features["item_id"]

ошибка в том, что:

InvalidArgumentError (see above for traceback): Restoring from checkpoint failed. This is most likely due to a mismatch between the current graph and the graph from the checkpoint. Please ensure that you have not altered the graph expected based on the checkpoint. Original error:

Assign requires shapes of both tensors to match. lhs shape= [64,128] rhs shape= [128,128]
         [[node save/Assign_5 (defined at estimator.py:142)  = Assign[T=DT_FLOAT, _class=["loc:@dense/kernel"], use_locking=true, validate_shape=true, _device="/job:localhost/replica:0/task:0/device:CPU:0"](dense/kernel, save/RestoreV2:5)]]