Merge pull request #26652 from amitsrivastava78:error_5

PiperOrigin-RevId: 238305243

tensorflow/lite/experimental/examples/lstm/rnn.py

@@ -217,10 +217,10 @@ def dynamic_rnn(cell,
     parallel_iterations = parallel_iterations or 32
     if sequence_length is not None:
       sequence_length = math_ops.cast(sequence_length, dtypes.int32)
-      if sequence_length.get_shape().rank not in (None, 1):
+      if sequence_length.shape.rank not in (None, 1):
         raise ValueError(
             "sequence_length must be a vector of length batch_size, "
-            "but saw shape: %s" % sequence_length.get_shape())
+            "but saw shape: %s" % sequence_length.shape)
       sequence_length = array_ops.identity(  # Just to find it in the graph.
           sequence_length,
           name="sequence_length")

tensorflow/lite/experimental/examples/lstm/rnn_cell.py

@@ -436,9 +436,9 @@ class TFLiteLSTMCell(rnn_cell_impl.LayerRNNCell):
         aggregate="first",
         index_override=18)
 
-    input_size = inputs.get_shape().with_rank(2)[1]
+    input_size = inputs.shape.with_rank(2)[1]
     if input_size.value is None:
-      raise ValueError("Could not infer input size from inputs.get_shape()[-1]")
+      raise ValueError("Could not infer input size from inputs.shape[-1]")
 
     inputs_and_m_prev = array_ops.concat([inputs, m_prev], axis=1)
 

tensorflow/lite/experimental/microfrontend/python/ops/audio_microfrontend_op.py

@@ -96,7 +96,7 @@ def audio_microfrontend(audio,
   Raises:
     ValueError: If the audio tensor is not explicitly a vector.
   """
-  audio_shape = audio.get_shape()
+  audio_shape = audio.shape
   if audio_shape.ndims is None:
     raise ValueError("Input to `AudioMicrofrontend` should have known rank.")
   if len(audio_shape) > 1:

tensorflow/lite/python/convert.py

@@ -244,7 +244,7 @@ def build_toco_convert_protos(input_tensors,
 
   Args:
     input_tensors: List of input tensors. Type and shape are computed using
-      `foo.get_shape()` and `foo.dtype`.
+      `foo.shape` and `foo.dtype`.
     output_tensors: List of output tensors (only .name is used from this).
     inference_type: Target data type of real-number arrays in the output file.
       Must be `{tf.float32, tf.uint8}`.  (default tf.float32)
@@ -347,7 +347,7 @@ def build_toco_convert_protos(input_tensors,
                          "inference_input_type is QUANTIZED_UINT8.")
       input_array.mean_value, input_array.std_value = quantized_input_stats[idx]
     if input_shapes is None:
-      shape = input_tensor.get_shape()
+      shape = input_tensor.shape
     else:
       shape = input_shapes[idx]
     input_array.shape.dims.extend(map(int, shape))
@@ -423,7 +423,7 @@ def toco_convert_impl(input_data, input_tensors, output_tensors, *args,
   Args:
     input_data: Input data (i.e. often `sess.graph_def`),
     input_tensors: List of input tensors. Type and shape are computed using
-      `foo.get_shape()` and `foo.dtype`.
+      `foo.shape` and `foo.dtype`.
     output_tensors: List of output tensors (only .name is used from this).
     *args: See `build_toco_convert_protos`,
     **kwargs: See `build_toco_convert_protos`.
@@ -456,7 +456,7 @@ def toco_convert(input_data, input_tensors, output_tensors, *args, **kwargs):
   Args:
     input_data: Input data (i.e. often `sess.graph_def`),
     input_tensors: List of input tensors. Type and shape are computed using
-      `foo.get_shape()` and `foo.dtype`.
+      `foo.shape` and `foo.dtype`.
     output_tensors: List of output tensors (only .name is used from this).
     *args: See `build_toco_convert_protos`,
     **kwargs: See `build_toco_convert_protos`.

tensorflow/lite/python/convert_saved_model.py

@@ -215,8 +215,7 @@ def set_tensor_shapes(tensors, shapes):
           tensor.set_shape(shape)
         except ValueError as error:
           message = ("The shape of tensor '{0}' cannot be changed from {1} to "
-                     "{2}. {3}".format(name, tensor.get_shape(), shape,
-                                       str(error)))
+                     "{2}. {3}".format(name, tensor.shape, shape, str(error)))
           raise ValueError(message)
 
 

tensorflow/lite/python/lite.py

@@ -258,14 +258,14 @@ class TFLiteConverterV2(object):
     # Checks dimensions in input tensor.
     for tensor in input_tensors:
       # Note that shape_list might be empty for scalar shapes.
-      shape_list = tensor.get_shape().as_list()
+      shape_list = tensor.shape.as_list()
       if None in shape_list[1:]:
         raise ValueError(
             "None is only supported in the 1st dimension. Tensor '{0}' has "
             "invalid shape '{1}'.".format(_tensor_name(tensor), shape_list))
       elif shape_list and shape_list[0] is None:
         # Set the batch size to 1 if undefined.
-        shape = tensor.get_shape().as_list()
+        shape = tensor.shape.as_list()
         shape[0] = 1
         tensor.set_shape(shape)
 
@@ -411,7 +411,7 @@ class TFLiteConverter(object):
     Args:
       graph_def: Frozen TensorFlow GraphDef.
       input_tensors: List of input tensors. Type and shape are computed using
-        `foo.get_shape()` and `foo.dtype`.
+        `foo.shape` and `foo.dtype`.
       output_tensors: List of output tensors (only .name is used from this).
       input_arrays_with_shape: Tuple of strings representing input tensor names
         and list of integers representing input shapes
@@ -460,7 +460,7 @@ class TFLiteConverter(object):
     Args:
       sess: TensorFlow Session.
       input_tensors: List of input tensors. Type and shape are computed using
-        `foo.get_shape()` and `foo.dtype`.
+        `foo.shape` and `foo.dtype`.
       output_tensors: List of output tensors (only .name is used from this).
 
     Returns:
@@ -678,7 +678,7 @@ class TFLiteConverter(object):
     # Checks dimensions in input tensor.
     if self._has_valid_tensors():
       for tensor in self._input_tensors:
-        shape = tensor.get_shape()
+        shape = tensor.shape
         if not shape:
           raise ValueError("Provide an input shape for input array "
                            "'{0}'.".format(_tensor_name(tensor)))
@@ -808,7 +808,7 @@ class TFLiteConverter(object):
                        "use input_shapes parameter.")
 
     for tensor in self._input_tensors:
-      shape = tensor.get_shape().as_list()
+      shape = tensor.shape.as_list()
       shape[0] = batch_size
       tensor.set_shape(shape)
 

tensorflow/lite/testing/generate_examples.py

@@ -492,9 +492,8 @@ def make_zip_of_tests(zip_path,
         report["toco"] = report_lib.FAILED
         report["tf"] = report_lib.SUCCESS
         # Convert graph to toco
-        input_tensors = [(input_tensor.name.split(":")[0],
-                          input_tensor.get_shape(), input_tensor.dtype)
-                         for input_tensor in inputs]
+        input_tensors = [(input_tensor.name.split(":")[0], input_tensor.shape,
+                          input_tensor.dtype) for input_tensor in inputs]
         output_tensors = [normalize_output_name(out.name) for out in outputs]
         graph_def = freeze_graph(
             sess,

tensorflow/lite/toco/python/toco_from_protos_test.py

@@ -54,7 +54,7 @@ class TocoFromProtosTest(googletest.TestCase):
     model_flags = model_flags_pb2.ModelFlags()
     input_array = model_flags.input_arrays.add()
     input_array.name = TensorName(in_tensor)
-    input_array.shape.dims.extend(map(int, in_tensor.get_shape()))
+    input_array.shape.dims.extend(map(int, in_tensor.shape))
     model_flags.output_arrays.append(TensorName(out_tensor))
     # Shell out to run toco (in case it crashes)
     with tempfile.NamedTemporaryFile() as fp_toco, \