1# Copyright 2019 The TensorFlow Authors. All Rights Reserved. 2# 3# Licensed under the Apache License, Version 2.0 (the "License"); 4# you may not use this file except in compliance with the License. 5# You may obtain a copy of the License at 6# 7# http://www.apache.org/licenses/LICENSE-2.0 8# 9# Unless required by applicable law or agreed to in writing, software 10# distributed under the License is distributed on an "AS IS" BASIS, 11# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 12# See the License for the specific language governing permissions and 13# limitations under the License. 14# ============================================================================== 15"""Test configs for cond.""" 16from __future__ import absolute_import 17from __future__ import division 18from __future__ import print_function 19 20import numpy as np 21import tensorflow.compat.v1 as tf 22from tensorflow.lite.testing.zip_test_utils import create_tensor_data 23from tensorflow.lite.testing.zip_test_utils import make_zip_of_tests 24from tensorflow.lite.testing.zip_test_utils import register_make_test_function 25from tensorflow.python.framework import test_util 26 27 28@register_make_test_function("make_cond_tests") 29@test_util.enable_control_flow_v2 30def make_cond_tests(options): 31 """Make a set of tests to do relu1.""" 32 33 # Chose a set of parameters 34 test_parameters = [{ 35 # Note: The `tf.string` test case also serves as a regression test to 36 # ensure that branch subgraph with dynamically allocated inputs/outputs 37 # are handled correctly. 38 "dtype": [tf.float32, tf.string], 39 "pred": [False, True], 40 }] 41 42 def build_graph(parameters): 43 """Build the graph for cond tests.""" 44 input1 = tf.placeholder(dtype=parameters["dtype"], shape=(1,)) 45 input2 = tf.placeholder(dtype=parameters["dtype"], shape=(1,)) 46 # MLIR TFLite converter can't handle scalar inputs. This is a workaround 47 # to input (1,) tensors and then reshape to scalar. 48 # TODO(b/129003347): Remove the workaround after scalar inputs are 49 # supported. 50 pred = tf.placeholder(dtype=tf.bool, shape=(1,)) 51 pred_scalar = tf.reshape(pred, ()) 52 53 out = tf.cond(pred_scalar, lambda: input1, lambda: input2) 54 return [input1, input2, pred], [out] 55 56 def build_inputs(parameters, sess, inputs, outputs): 57 input_values = [ 58 create_tensor_data(parameters["dtype"], (1,)), 59 create_tensor_data(parameters["dtype"], (1,)), 60 np.array([parameters["pred"]], dtype=np.bool_), 61 ] 62 return input_values, sess.run( 63 outputs, feed_dict=dict(zip(inputs, input_values))) 64 65 make_zip_of_tests(options, test_parameters, build_graph, build_inputs) 66