IntenseWebs/openvino
3
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

[TF FE] Add layer test for Bucketize (#16556)

Browse Source 
Signed-off-by: Kazantsev, Roman <roman.kazantsev@intel.com>
This commit is contained in:
Roman Kazantsev 2023-03-27 12:03:07 +04:00 committed by GitHub
parent a1b8a6a941
commit bb9de29062
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 24 additions and 71 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

95
tests/layer_tests/tensorflow_tests/test_tf_Bucketize.py
View File

@ -4,90 +4,43 @@
import numpy as np
import pytest
import tensorflow as tf
from common.layer_test_class import check_ir_version
from common.tf_layer_test_class import CommonTFLayerTest
from unit_tests.utils.graph import build_graph
class TestBucketize(CommonTFLayerTest):
def create_bucketize_net(self, input_shape, input_type, boundaries_size, ir_version,
use_new_frontend):
"""
Tensorflow net: IR net:
Input => Input Boundaries
| \ /
Bucketize Bucketize
{attrs: boundaries}
"""
def _prepare_input(self, inputs_info):
assert 'input' in inputs_info, "Test error: inputs_info must contain `input`"
input_shape = inputs_info['input']
input_type = self.input_type
inputs_data = {}
input_data = np.random.randint(-20, 20, input_shape).astype(input_type)
inputs_data['input'] = input_data
return inputs_data
def create_bucketize_net(self, input_shape, input_type, boundaries_size):
self.input_type = input_type
tf.compat.v1.reset_default_graph()
with tf.compat.v1.Session() as sess:
x = tf.compat.v1.placeholder(input_type, input_shape, 'Input')
constant_value = np.arange(-boundaries_size * 5, boundaries_size * 5, 10,
dtype=np.float32)
# TODO: Bucketize is not tested here. Need to re-write the test
input = tf.compat.v1.placeholder(input_type, input_shape, 'input')
# generate boundaries list
boundaries = np.sort(np.unique(np.random.randint(-200, 200, [boundaries_size]).astype(np.float32))).tolist()
tf.raw_ops.Bucketize(input=input, boundaries=boundaries)
tf.compat.v1.global_variables_initializer()
tf_net = sess.graph_def
# create reference IR net
ref_net = None
return tf_net, None
if check_ir_version(10, None, ir_version) and not use_new_frontend:
nodes_attributes = {
'input': {'kind': 'op', 'type': 'Parameter'},
'input_data': {'shape': input_shape, 'kind': 'data'},
'boundaries_input_data': {'shape': constant_value.shape, 'kind': 'data'},
'boundaries': {'type': 'Const', 'kind': 'op'},
'boundaries_data': {'kind': 'data', 'shape': constant_value.shape},
'bucketize': {'kind': 'op', 'type': 'Bucketize'},
'bucketize_data': {'shape': input_shape, 'kind': 'data'},
'result': {'kind': 'op', 'type': 'Result'}
}
test_data_basic = [
dict(input_shape=[5], input_type=np.int32, boundaries_size=1),
dict(input_shape=[3, 4], input_type=np.float32, boundaries_size=0),
dict(input_shape=[2, 3, 4], input_type=np.float32, boundaries_size=300),
]
ref_net = build_graph(nodes_attributes,
[('input', 'input_data'),
('input_data', 'bucketize', {'in': 0}),
('boundaries_input_data', 'boundaries'),
('boundaries', 'boundaries_data'),
('boundaries_data', 'bucketize', {'in': 1}),
('bucketize', 'bucketize_data'),
('bucketize_data', 'result')
])
return tf_net, ref_net
test_data_float32 = [
dict(input_shape=[5], input_type=tf.float32, boundaries_size=1),
dict(input_shape=[5], input_type=tf.float32, boundaries_size=3),
pytest.param(dict(input_shape=[4, 8], input_type=tf.float32, boundaries_size=5),
marks=pytest.mark.precommit_tf_fe),
dict(input_shape=[2, 4, 7], input_type=tf.float32, boundaries_size=10),
dict(input_shape=[2, 4, 7, 8], input_type=tf.float32, boundaries_size=12),
dict(input_shape=[2, 4, 7, 8, 10], input_type=tf.float32, boundaries_size=14)]
@pytest.mark.parametrize("params", test_data_float32)
@pytest.mark.parametrize("params", test_data_basic)
@pytest.mark.precommit_tf_fe
@pytest.mark.nightly
def test_bucketize_float32(self, params, ie_device, precision, ir_version, temp_dir,
use_new_frontend, use_old_api):
self._test(*self.create_bucketize_net(**params, ir_version=ir_version,
use_new_frontend=use_new_frontend),
ie_device, precision, ir_version, temp_dir=temp_dir,
use_new_frontend=use_new_frontend, use_old_api=use_old_api)
test_data_int32 = [
dict(input_shape=[5], input_type=tf.int32, boundaries_size=1),
dict(input_shape=[5], input_type=tf.int32, boundaries_size=3),
dict(input_shape=[4, 8], input_type=tf.int32, boundaries_size=5),
dict(input_shape=[2, 4, 7], input_type=tf.int32, boundaries_size=10),
dict(input_shape=[2, 4, 7, 8], input_type=tf.float32, boundaries_size=12),
dict(input_shape=[2, 4, 7, 8, 10], input_type=tf.float32, boundaries_size=14)]
@pytest.mark.parametrize("params", test_data_int32)
@pytest.mark.nightly
def test_bucketize_int32(self, params, ie_device, precision, ir_version, temp_dir,
def test_bucketize_basic(self, params, ie_device, precision, ir_version, temp_dir,
use_new_frontend, use_old_api):
self._test(*self.create_bucketize_net(**params, ir_version=ir_version,
use_new_frontend=use_new_frontend),
self._test(*self.create_bucketize_net(**params),
ie_device, precision, ir_version, temp_dir=temp_dir,
use_new_frontend=use_new_frontend, use_old_api=use_old_api)