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 transpose.""" 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 25 26 27@register_make_test_function() 28def make_transpose_tests(options): 29 """Make a set of tests to do transpose.""" 30 31 # TODO(nupurgarg): Add test for uint8. 32 test_parameters = [{ 33 "dtype": [tf.int32, tf.int64, tf.float32], 34 "input_shape": [[2, 2, 3]], 35 "perm": [[0, 1, 2], [0, 2, 1]], 36 "constant_perm": [True, False], 37 "fully_quantize": [False], 38 }, { 39 "dtype": [tf.float32], 40 "input_shape": [[1, 2, 3, 4]], 41 "perm": [[0, 1, 2, 3], [3, 0, 1, 2]], 42 "constant_perm": [True, False], 43 "fully_quantize": [False], 44 }, { 45 "dtype": [tf.float32], 46 "input_shape": [[1, 2, 3, 4, 5]], 47 "perm": [[4, 3, 2, 1, 0]], 48 "constant_perm": [True, False], 49 "fully_quantize": [False], 50 }, { 51 "dtype": [tf.float32], 52 "input_shape": [[2, 2, 3]], 53 "perm": [[0, 1, 2], [0, 2, 1]], 54 "constant_perm": [True], 55 "fully_quantize": [True], 56 }, { 57 "dtype": [tf.float32], 58 "input_shape": [[1, 2, 3, 4]], 59 "perm": [[0, 1, 2, 3], [3, 0, 1, 2]], 60 "constant_perm": [True], 61 "fully_quantize": [True], 62 }, { 63 "dtype": [tf.float32], 64 "input_shape": [[1, 2, 3, 4, 5]], 65 "perm": [[0, 1, 2, 3, 4], [3, 4, 0, 1, 2]], 66 "constant_perm": [True], 67 "fully_quantize": [True, False], 68 }] 69 70 def build_graph(parameters): 71 """Build a transpose graph given `parameters`.""" 72 input_tensor = tf.compat.v1.placeholder( 73 dtype=parameters["dtype"], 74 name="input", 75 shape=parameters["input_shape"]) 76 77 if parameters["constant_perm"]: 78 perm = parameters["perm"] 79 input_tensors = [input_tensor] 80 else: 81 shape = [len(parameters["perm"]), 2] 82 perm = tf.compat.v1.placeholder(dtype=tf.int32, name="perm", shape=shape) 83 input_tensors = [input_tensor, perm] 84 85 out = tf.transpose(input_tensor, perm=perm) 86 return input_tensors, [out] 87 88 def build_inputs(parameters, sess, inputs, outputs): 89 values = [ 90 create_tensor_data(parameters["dtype"], parameters["input_shape"], 91 min_value=-1, max_value=1) 92 ] 93 if not parameters["constant_perm"]: 94 values.append(np.array(parameters["perm"])) 95 return values, sess.run(outputs, feed_dict=dict(zip(inputs, values))) 96 97 make_zip_of_tests( 98 options, 99 test_parameters, 100 build_graph, 101 build_inputs, 102 expected_tf_failures=9) 103