-
Notifications
You must be signed in to change notification settings - Fork 706
/
executor.py
220 lines (190 loc) · 9.65 KB
/
executor.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
# Lint as: python2, python3
# Copyright 2019 Google LLC. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Generic TFX model validator executor."""
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import os
from typing import Any, Dict, List, Text
import absl
import apache_beam as beam
import tensorflow_model_analysis as tfma
from tfx import types
from tfx.components.model_validator import constants
from tfx.dsl.components.base import base_executor
from tfx.dsl.io import fileio
from tfx.types import artifact_utils
from tfx.utils import io_utils
from tfx.utils import path_utils
class Executor(base_executor.BaseExecutor):
"""DEPRECATED: Please use `Evaluator` instead.
The model validator helps prevent bad models from being pushed to production.
It does this by validating exported models against known good models (e.g. the
current production model), and marking the exported model as good ("blessing
it") only if the exported model's metrics are within predefined thresholds
around the good model's metrics.
The model validator will validate tf.serving format exported models produced
by the Trainer component. The validator evaluates the models on examples
created by the ExampleGen component. The validator will also automatically
read data written by the Pusher component regarding the latest pushed models
by using ml.metadata to query the previously pushed artifacts.
To include ModelValidator in a TFX pipeline, configure your pipeline similar
to
https://github.com/tensorflow/tfx/blob/master/tfx/examples/chicago_taxi_pipeline/taxi_pipeline_simple.py#L110.
"""
# TODO(jyzhao): customized threshold support.
def _pass_threshold(self, eval_result: tfma.EvalResult) -> bool:
"""Check threshold."""
return True
# TODO(jyzhao): customized validation support.
def _compare_eval_result(self, current_model_eval_result: tfma.EvalResult,
blessed_model_eval_result: tfma.EvalResult) -> bool:
"""Compare accuracy of all metrics and return true if current is better or equal."""
for current_metric, blessed_metric in zip(
current_model_eval_result.slicing_metrics,
blessed_model_eval_result.slicing_metrics):
# slicing_metric is a tuple, index 0 is slice, index 1 is its value.
if current_metric[0] != blessed_metric[0]:
raise RuntimeError('EvalResult not match {} vs {}.'.format(
current_metric[0], blessed_metric[0]))
# TODO(b/140455644): TFMA introduced breaking change post 0.14 release.
# Remove this forward compatibility change after 0.15 release.
current_model_metrics = current_metric[1]
blessed_model_metrics = blessed_metric[1]
try:
current_model_accuracy = current_model_metrics['accuracy']
blessed_model_accuracy = blessed_model_metrics['accuracy']
except KeyError:
current_model_accuracy = current_model_metrics['']['']['accuracy']
blessed_model_accuracy = blessed_model_metrics['']['']['accuracy']
if (current_model_accuracy['doubleValue'] <
blessed_model_accuracy['doubleValue']):
absl.logging.info(
'Current model accuracy is worse than blessed model: {}'.format(
current_metric[0]))
return False
return True
def _generate_blessing_result(self, eval_examples_uri: Text,
slice_spec: List[tfma.slicer.SingleSliceSpec],
current_model_dir: Text,
blessed_model_dir: Text) -> bool:
current_model_eval_result_path = os.path.join(
self._temp_path, constants.CURRENT_MODEL_EVAL_RESULT_PATH)
blessed_model_eval_result_path = os.path.join(
self._temp_path, constants.BLESSED_MODEL_EVAL_RESULT_PATH)
with self._make_beam_pipeline() as pipeline:
eval_data = (
pipeline | 'ReadData' >> beam.io.ReadFromTFRecord(
file_pattern=io_utils.all_files_pattern(eval_examples_uri)))
current_model = tfma.default_eval_shared_model(
eval_saved_model_path=path_utils.eval_model_path(current_model_dir))
(eval_data | 'EvalCurrentModel' >> tfma.ExtractEvaluateAndWriteResults( # pylint: disable=expression-not-assigned
eval_shared_model=current_model,
slice_spec=slice_spec,
output_path=current_model_eval_result_path))
if blessed_model_dir is not None:
blessed_model = tfma.default_eval_shared_model(
eval_saved_model_path=path_utils.eval_model_path(blessed_model_dir))
(eval_data | 'EvalBlessedModel' >> tfma.ExtractEvaluateAndWriteResults( # pylint: disable=expression-not-assigned
eval_shared_model=blessed_model,
slice_spec=slice_spec,
output_path=blessed_model_eval_result_path))
absl.logging.info('all files in current_model_eval_result_path: [%s]',
str(fileio.listdir(current_model_eval_result_path)))
current_model_eval_result = tfma.load_eval_result(
output_path=current_model_eval_result_path)
if not self._pass_threshold(current_model_eval_result):
absl.logging.info('Current model does not pass threshold.')
return False
absl.logging.info('Current model passes threshold.')
if blessed_model_dir is None:
absl.logging.info('No blessed model yet.')
return True
absl.logging.info('all files in blessed_model_eval_result: [%s]',
str(fileio.listdir(blessed_model_eval_result_path)))
blessed_model_eval_result = tfma.load_eval_result(
output_path=blessed_model_eval_result_path)
if (self._compare_eval_result(current_model_eval_result,
blessed_model_eval_result)):
absl.logging.info('Current model better than blessed model.')
return True
else:
absl.logging.info('Current model worse than blessed model.')
return False
def Do(self, input_dict: Dict[Text, List[types.Artifact]],
output_dict: Dict[Text, List[types.Artifact]],
exec_properties: Dict[Text, Any]) -> None:
"""Validate current model against last blessed model.
Args:
input_dict: Input dict from input key to a list of Artifacts.
- examples: examples for eval the model.
- model: current model for validation.
output_dict: Output dict from output key to a list of Artifacts.
- blessing: model blessing result.
exec_properties: A dict of execution properties.
- blessed_model: last blessed model for validation.
- blessed_model_id: last blessed model id.
Returns:
None
"""
self._log_startup(input_dict, output_dict, exec_properties)
self._temp_path = self._get_tmp_dir()
absl.logging.info('Using temp path {} for tft.beam'.format(self._temp_path))
eval_examples_uri = artifact_utils.get_split_uri(
input_dict[constants.EXAMPLES_KEY], 'eval')
blessing = artifact_utils.get_single_instance(
output_dict[constants.BLESSING_KEY])
# Current model to be validated.
current_model = artifact_utils.get_single_instance(
input_dict[constants.MODEL_KEY])
absl.logging.info('Using {} as current model.'.format(current_model.uri))
blessing.set_string_custom_property(
constants.ARTIFACT_PROPERTY_CURRENT_MODEL_URI_KEY, current_model.uri)
blessing.set_int_custom_property(
constants.ARTIFACT_PROPERTY_CURRENT_MODEL_ID_KEY, current_model.id)
# Denote model component_name.
component_id = exec_properties['current_component_id']
blessing.set_string_custom_property('component_id', component_id)
# Previous blessed model to be validated against.
blessed_model_dir = exec_properties['blessed_model']
blessed_model_id = exec_properties['blessed_model_id']
absl.logging.info('Using {} as blessed model.'.format(blessed_model_dir))
if blessed_model_dir:
blessing.set_string_custom_property(
constants.ARTIFACT_PROPERTY_BLESSED_MODEL_URI_KEY, blessed_model_dir)
blessing.set_int_custom_property(
constants.ARTIFACT_PROPERTY_BLESSED_MODEL_ID_KEY, blessed_model_id)
absl.logging.info('Validating model.')
# TODO(b/125853306): support customized slice spec.
blessed = self._generate_blessing_result(
eval_examples_uri=eval_examples_uri,
slice_spec=[tfma.slicer.SingleSliceSpec()],
current_model_dir=current_model.uri,
blessed_model_dir=blessed_model_dir)
if blessed:
io_utils.write_string_file(
os.path.join(blessing.uri, constants.BLESSED_FILE_NAME), '')
blessing.set_int_custom_property(constants.ARTIFACT_PROPERTY_BLESSED_KEY,
constants.BLESSED_VALUE)
else:
io_utils.write_string_file(
os.path.join(blessing.uri, constants.NOT_BLESSED_FILE_NAME), '')
blessing.set_int_custom_property(constants.ARTIFACT_PROPERTY_BLESSED_KEY,
constants.NOT_BLESSED_VALUE)
absl.logging.info('Blessing result {} written to {}.'.format(
blessed, blessing.uri))
io_utils.delete_dir(self._temp_path)
absl.logging.info('Cleaned up temp path {} on executor success.'.format(
self._temp_path))