Update the behavior on custom op. (#459)

keras2onnx/_parser_1x.py

@@ -7,7 +7,7 @@
 
 from .common import k2o_logger
 from .funcbook import get_converter
-from ._parse_tf import infer_variable_type, tsname_to_node, adjust_input_batch_size
+from ._parser_tf import infer_variable_type, tsname_to_node, adjust_input_batch_size
 
 
 def extract_inbound_nodes(layer):

keras2onnx/_tf_ops.py

@@ -36,11 +36,16 @@ def convert_tf_random_standard_normal(scope, operator, container):
 
 
 def pass_thru_converter(scope, operator, container):
+ """
+ This converter is to copy the original graph node with its def into a ONNX node format.
+ """
  tf_op = operator.raw_operator
  attrs = _to_onnx_attrs(tf_op)
 
  container.add_node(operator.type,
  operator.input_full_names,
  operator.output_full_names,
  name=operator.full_name,
+ op_domain='ai.onnx.contrib',
+ op_version=1,
  **attrs)

keras2onnx/main.py

@@ -15,9 +15,10 @@
 from .topology import Topology
 from .common.utils import set_logger_level, k2o_logger
 from .funcbook import set_converter
+from ._tf_utils import tsname_to_node
 from ._builtin import register_direct_tf_ops
 from ._parser_1x import build_opdict_from_keras
-from ._parse_tf import tsname_to_node, build_layer_output_from_model
+from ._parser_tf import build_layer_output_from_model
 
 
 def convert_keras(model, name=None, doc_string='', target_opset=None,

keras2onnx/parser.py

@@ -14,8 +14,8 @@
 from .funcbook import get_converter, set_converter
 from ._consts import TYPES
 from ._tf_ops import pass_thru_converter
-from ._parse_tf import (infer_variable_type, LayerInfo, is_placeholder_node,
- tsname_to_node, on_parsing_keras_layer_v2, adjust_input_batch_size as _adjust_input_batch_size)
+from ._parser_tf import (infer_variable_type, LayerInfo, is_placeholder_node,
+  tsname_to_node, on_parsing_keras_layer_v2, adjust_input_batch_size as _adjust_input_batch_size)
 from ._parser_1x import (extract_inbound_nodes,
  list_input_tensors, list_input_mask, list_output_mask,
  list_output_tensors, list_input_shapes, list_output_shapes, on_parsing_keras_layer)
@@ -294,15 +294,15 @@ def _check_tfnode_converter_availability(graph, node):
 
 
 def _check_tfnodes_converter_availability(graph, nodelist, debug_mode):
+ status = True
  for n_ in nodelist:
  if not _check_tfnode_converter_availability(graph, n_):
  k2o_logger().warning(
- "The tf.op node {} of type {} cannot be converted".format(n_.name, n_.type))
- if debug_mode:
- continue
- return False
+ "WARN: No corresponding ONNX op matches the tf.op node {} of type {}".format(n_.name, n_.type) +
+ "\n The generated ONNX model needs run with the custom op supports.")
+ status = False
 
- return True
+ return status
 
 
 def _on_parsing_tf_nodes(graph, nodelist, varset, debug_mode):