Zobacz na TensorFlow.org
Uruchom w Google Colab
Wyświetl źródło na GitHub
Pobierz notatnikW tym samouczku opartym na notatniku utworzymy i uruchomimy potok TFX, który tworzy prosty model klasyfikacji i analizuje jego wydajność w wielu uruchomieniach. Notebook ten jest oparty na rurociągu TFX my zbudowany w prosty TFX Pipeline Tutorial . Jeśli jeszcze nie przeczytałeś tego samouczka, powinieneś go przeczytać przed kontynuowaniem tego notatnika.
Gdy modyfikujesz swój model lub trenujesz go za pomocą nowego zestawu danych, musisz sprawdzić, czy Twój model się poprawił, czy pogorszył. Samo sprawdzenie wskaźników najwyższego poziomu, takich jak dokładność, może nie wystarczyć. Każdy wytrenowany model powinien zostać oceniony przed wypchnięciem do produkcji.
Będziemy dodać Evaluator komponent do rurociągu utworzonej w poprzednim ćwiczeniu. Komponent oceniający przeprowadza dogłębną analizę modeli i porównuje nowy model z linią bazową, aby określić, czy są one „wystarczająco dobre”. Jest on realizowany za pomocą TensorFlow analizy modelu biblioteki.
Proszę zobaczyć Zrozumienie TFX Rurociągi , aby dowiedzieć się więcej na temat różnych koncepcji w TFX.
Ustawiać
Proces konfiguracji jest taki sam, jak w poprzednim samouczku.
Najpierw musimy zainstalować pakiet TFX Python i pobrać zestaw danych, którego użyjemy dla naszego modelu.
Ulepsz Pip
Aby uniknąć aktualizacji Pip w systemie, gdy działa on lokalnie, upewnij się, że działamy w Colab. Systemy lokalne można oczywiście aktualizować oddzielnie.
try:
import colab
!pip install --upgrade pip
except:
pass
Zainstaluj TFX
pip install -U tfx
Czy uruchomiłeś ponownie środowisko wykonawcze?
Jeśli korzystasz z Google Colab, przy pierwszym uruchomieniu powyższej komórki musisz ponownie uruchomić środowisko wykonawcze, klikając powyżej przycisk „RESTART RUNTIME” lub korzystając z menu „Runtime > Restart runtime...”. Wynika to ze sposobu, w jaki Colab ładuje paczki.
Sprawdź wersje TensorFlow i TFX.
import tensorflow as tf
print('TensorFlow version: {}'.format(tf.__version__))
from tfx import v1 as tfx
print('TFX version: {}'.format(tfx.__version__))
TensorFlow version: 2.6.2 TFX version: 1.4.0
Ustaw zmienne
Istnieje kilka zmiennych używanych do zdefiniowania potoku. Możesz dostosować te zmienne, jak chcesz. Domyślnie wszystkie dane wyjściowe z potoku będą generowane w bieżącym katalogu.
import os
PIPELINE_NAME = "penguin-tfma"
# Output directory to store artifacts generated from the pipeline.
PIPELINE_ROOT = os.path.join('pipelines', PIPELINE_NAME)
# Path to a SQLite DB file to use as an MLMD storage.
METADATA_PATH = os.path.join('metadata', PIPELINE_NAME, 'metadata.db')
# Output directory where created models from the pipeline will be exported.
SERVING_MODEL_DIR = os.path.join('serving_model', PIPELINE_NAME)
from absl import logging
logging.set_verbosity(logging.INFO) # Set default logging level.
Przygotuj przykładowe dane
Będziemy korzystać z tego samego Palmer Penguins zestaw danych .
W tym zbiorze danych znajdują się cztery cechy liczbowe, które zostały już znormalizowane tak, aby miały zakres [0,1]. Będziemy budować model klasyfikacji który przewiduje species pingwinów.
Ponieważ TFX ExampleGen odczytuje dane wejściowe z katalogu, musimy utworzyć katalog i skopiować do niego zbiór danych.
import urllib.request
import tempfile
DATA_ROOT = tempfile.mkdtemp(prefix='tfx-data') # Create a temporary directory.
_data_url = 'https://raw.githubusercontent.com/tensorflow/tfx/master/tfx/examples/penguin/data/labelled/penguins_processed.csv'
_data_filepath = os.path.join(DATA_ROOT, "data.csv")
urllib.request.urlretrieve(_data_url, _data_filepath)
('/tmp/tfx-datal5lxy_yw/data.csv', <http.client.HTTPMessage at 0x7fa18a9da150>)
Utwórz potok
Będziemy dodać Evaluator komponent do rurociągu utworzyliśmy w prosty TFX Pipeline Tutorial .
Komponent oceniającego wymaga wprowadzania danych z ExampleGen komponentu i modelu ze Trainer komponentu i tfma.EvalConfig obiektu. Opcjonalnie możemy dostarczyć model podstawowy, którego można użyć do porównania metryk z nowo wytrenowanym modelem.
Ewaluator tworzy dwa rodzaje artefaktów wyjściowych ModelEvaluation i ModelBlessing . ModelEvaluation zawiera szczegółowe wyniki oceny, które można dalej badać i wizualizować za pomocą biblioteki TFMA. ModelBlessing zawiera wynik logiczny, czy model spełnił określone kryteria i może być użyty w późniejszych komponentach, takich jak Pusher, jako sygnał.
Napisz kod szkolenia modelu
Będziemy wykorzystywać ten sam kod modelu, jak w prosty TFX Pipeline Tutorial .
_trainer_module_file = 'penguin_trainer.py'
%%writefile {_trainer_module_file}
# Copied from https://www.tensorflow.org/tfx/tutorials/tfx/penguin_simple
from typing import List
from absl import logging
import tensorflow as tf
from tensorflow import keras
from tensorflow_transform.tf_metadata import schema_utils
from tfx.components.trainer.executor import TrainerFnArgs
from tfx.components.trainer.fn_args_utils import DataAccessor
from tfx_bsl.tfxio import dataset_options
from tensorflow_metadata.proto.v0 import schema_pb2
_FEATURE_KEYS = [
'culmen_length_mm', 'culmen_depth_mm', 'flipper_length_mm', 'body_mass_g'
]
_LABEL_KEY = 'species'
_TRAIN_BATCH_SIZE = 20
_EVAL_BATCH_SIZE = 10
# Since we're not generating or creating a schema, we will instead create
# a feature spec. Since there are a fairly small number of features this is
# manageable for this dataset.
_FEATURE_SPEC = {
**{
feature: tf.io.FixedLenFeature(shape=[1], dtype=tf.float32)
for feature in _FEATURE_KEYS
},
_LABEL_KEY: tf.io.FixedLenFeature(shape=[1], dtype=tf.int64)
}
def _input_fn(file_pattern: List[str],
data_accessor: DataAccessor,
schema: schema_pb2.Schema,
batch_size: int = 200) -> tf.data.Dataset:
"""Generates features and label for training.
Args:
file_pattern: List of paths or patterns of input tfrecord files.
data_accessor: DataAccessor for converting input to RecordBatch.
schema: schema of the input data.
batch_size: representing the number of consecutive elements of returned
dataset to combine in a single batch
Returns:
A dataset that contains (features, indices) tuple where features is a
dictionary of Tensors, and indices is a single Tensor of label indices.
"""
return data_accessor.tf_dataset_factory(
file_pattern,
dataset_options.TensorFlowDatasetOptions(
batch_size=batch_size, label_key=_LABEL_KEY),
schema=schema).repeat()
def _build_keras_model() -> tf.keras.Model:
"""Creates a DNN Keras model for classifying penguin data.
Returns:
A Keras Model.
"""
# The model below is built with Functional API, please refer to
# https://www.tensorflow.org/guide/keras/overview for all API options.
inputs = [keras.layers.Input(shape=(1,), name=f) for f in _FEATURE_KEYS]
d = keras.layers.concatenate(inputs)
for _ in range(2):
d = keras.layers.Dense(8, activation='relu')(d)
outputs = keras.layers.Dense(3)(d)
model = keras.Model(inputs=inputs, outputs=outputs)
model.compile(
optimizer=keras.optimizers.Adam(1e-2),
loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
metrics=[keras.metrics.SparseCategoricalAccuracy()])
model.summary(print_fn=logging.info)
return model
# TFX Trainer will call this function.
def run_fn(fn_args: TrainerFnArgs):
"""Train the model based on given args.
Args:
fn_args: Holds args used to train the model as name/value pairs.
"""
# This schema is usually either an output of SchemaGen or a manually-curated
# version provided by pipeline author. A schema can also derived from TFT
# graph if a Transform component is used. In the case when either is missing,
# `schema_from_feature_spec` could be used to generate schema from very simple
# feature_spec, but the schema returned would be very primitive.
schema = schema_utils.schema_from_feature_spec(_FEATURE_SPEC)
train_dataset = _input_fn(
fn_args.train_files,
fn_args.data_accessor,
schema,
batch_size=_TRAIN_BATCH_SIZE)
eval_dataset = _input_fn(
fn_args.eval_files,
fn_args.data_accessor,
schema,
batch_size=_EVAL_BATCH_SIZE)
model = _build_keras_model()
model.fit(
train_dataset,
steps_per_epoch=fn_args.train_steps,
validation_data=eval_dataset,
validation_steps=fn_args.eval_steps)
# The result of the training should be saved in `fn_args.serving_model_dir`
# directory.
model.save(fn_args.serving_model_dir, save_format='tf')
Writing penguin_trainer.py
Napisz definicję potoku
Zdefiniujemy funkcję do tworzenia potoku TFX. Oprócz komponentu Evaluator wspomniano powyżej, dodamy jeszcze jeden węzeł o nazwie Resolver . Aby sprawdzić, czy nowy model staje się lepszy od poprzedniego, musimy porównać go z poprzednio opublikowanym modelem, zwanym linią bazową. ML Metadane (MLMD) śledzi wszystkie poprzednie artefakty rurociągu i Resolver można znaleźć to, czego był ostatni błogosławiony model - model przeszedł Inspekcji powodzeniem - od MLMD użyciu klasy strategię zwaną LatestBlessedModelStrategy .
import tensorflow_model_analysis as tfma
def _create_pipeline(pipeline_name: str, pipeline_root: str, data_root: str,
module_file: str, serving_model_dir: str,
metadata_path: str) -> tfx.dsl.Pipeline:
"""Creates a three component penguin pipeline with TFX."""
# Brings data into the pipeline.
example_gen = tfx.components.CsvExampleGen(input_base=data_root)
# Uses user-provided Python function that trains a model.
trainer = tfx.components.Trainer(
module_file=module_file,
examples=example_gen.outputs['examples'],
train_args=tfx.proto.TrainArgs(num_steps=100),
eval_args=tfx.proto.EvalArgs(num_steps=5))
# NEW: Get the latest blessed model for Evaluator.
model_resolver = tfx.dsl.Resolver(
strategy_class=tfx.dsl.experimental.LatestBlessedModelStrategy,
model=tfx.dsl.Channel(type=tfx.types.standard_artifacts.Model),
model_blessing=tfx.dsl.Channel(
type=tfx.types.standard_artifacts.ModelBlessing)).with_id(
'latest_blessed_model_resolver')
# NEW: Uses TFMA to compute evaluation statistics over features of a model and
# perform quality validation of a candidate model (compared to a baseline).
eval_config = tfma.EvalConfig(
model_specs=[tfma.ModelSpec(label_key='species')],
slicing_specs=[
# An empty slice spec means the overall slice, i.e. the whole dataset.
tfma.SlicingSpec(),
# Calculate metrics for each penguin species.
tfma.SlicingSpec(feature_keys=['species']),
],
metrics_specs=[
tfma.MetricsSpec(per_slice_thresholds={
'sparse_categorical_accuracy':
tfma.PerSliceMetricThresholds(thresholds=[
tfma.PerSliceMetricThreshold(
slicing_specs=[tfma.SlicingSpec()],
threshold=tfma.MetricThreshold(
value_threshold=tfma.GenericValueThreshold(
lower_bound={'value': 0.6}),
# Change threshold will be ignored if there is no
# baseline model resolved from MLMD (first run).
change_threshold=tfma.GenericChangeThreshold(
direction=tfma.MetricDirection.HIGHER_IS_BETTER,
absolute={'value': -1e-10}))
)]),
})],
)
evaluator = tfx.components.Evaluator(
examples=example_gen.outputs['examples'],
model=trainer.outputs['model'],
baseline_model=model_resolver.outputs['model'],
eval_config=eval_config)
# Checks whether the model passed the validation steps and pushes the model
# to a file destination if check passed.
pusher = tfx.components.Pusher(
model=trainer.outputs['model'],
model_blessing=evaluator.outputs['blessing'], # Pass an evaluation result.
push_destination=tfx.proto.PushDestination(
filesystem=tfx.proto.PushDestination.Filesystem(
base_directory=serving_model_dir)))
components = [
example_gen,
trainer,
# Following two components were added to the pipeline.
model_resolver,
evaluator,
pusher,
]
return tfx.dsl.Pipeline(
pipeline_name=pipeline_name,
pipeline_root=pipeline_root,
metadata_connection_config=tfx.orchestration.metadata
.sqlite_metadata_connection_config(metadata_path),
components=components)
Musimy dostarczyć następujące informacje Oceniający poprzez eval_config :
- Dodatkowe metryki do skonfigurowania (jeśli potrzebujesz więcej metryk niż zdefiniowano w modelu).
- Plasterki do skonfigurowania
- Progi walidacji modelu w celu sprawdzenia, czy należy uwzględnić walidację
Ponieważ SparseCategoricalAccuracy został już uwzględniony w model.compile() połączenia, zostanie ono automatycznie włączone do analizy. Dlatego nie dodajemy tutaj żadnych dodatkowych metryk. SparseCategoricalAccuracy zostaną wykorzystane do podjęcia decyzji, czy model jest wystarczająco dobre, zbyt.
Obliczamy metryki dla całego zbioru danych i dla każdego gatunku pingwina. SlicingSpec określa sposób możemy agregować deklarowanych wskaźników.
Istnieją dwa progi, które powinien przekroczyć nowy model, jeden to próg bezwzględny 0,6, a drugi to próg względny, który powinien być wyższy niż model bazowy. Po uruchomieniu rurociągu po raz pierwszy, change_threshold będą ignorowane i będą sprawdzane tylko value_threshold. Jeśli po uruchomieniu rurociągu więcej niż jeden raz, Resolver znajdzie model z poprzedniego uruchomienia i będzie on używany jako model odniesienia dla porównania.
Zobacz komponentu przewodnik oceniającego , aby uzyskać więcej informacji.
Uruchom rurociąg
Użyjemy LocalDagRunner jak w poprzednim ćwiczeniu.
tfx.orchestration.LocalDagRunner().run(
_create_pipeline(
pipeline_name=PIPELINE_NAME,
pipeline_root=PIPELINE_ROOT,
data_root=DATA_ROOT,
module_file=_trainer_module_file,
serving_model_dir=SERVING_MODEL_DIR,
metadata_path=METADATA_PATH))
INFO:absl:Generating ephemeral wheel package for '/tmpfs/src/temp/docs/tutorials/tfx/penguin_trainer.py' (including modules: ['penguin_trainer']).
INFO:absl:User module package has hash fingerprint version 1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703.
INFO:absl:Executing: ['/tmpfs/src/tf_docs_env/bin/python', '/tmp/tmpr3anh67s/_tfx_generated_setup.py', 'bdist_wheel', '--bdist-dir', '/tmp/tmp6s2sw4dj', '--dist-dir', '/tmp/tmp6jr76e54']
/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/setuptools/command/install.py:37: SetuptoolsDeprecationWarning: setup.py install is deprecated. Use build and pip and other standards-based tools.
setuptools.SetuptoolsDeprecationWarning,
listing git files failed - pretending there aren't any
INFO:absl:Successfully built user code wheel distribution at 'pipelines/penguin-tfma/_wheels/tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703-py3-none-any.whl'; target user module is 'penguin_trainer'.
INFO:absl:Full user module path is 'penguin_trainer@pipelines/penguin-tfma/_wheels/tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703-py3-none-any.whl'
INFO:absl:Using deployment config:
executor_specs {
key: "CsvExampleGen"
value {
beam_executable_spec {
python_executor_spec {
class_path: "tfx.components.example_gen.csv_example_gen.executor.Executor"
}
}
}
}
executor_specs {
key: "Evaluator"
value {
beam_executable_spec {
python_executor_spec {
class_path: "tfx.components.evaluator.executor.Executor"
}
}
}
}
executor_specs {
key: "Pusher"
value {
python_class_executable_spec {
class_path: "tfx.components.pusher.executor.Executor"
}
}
}
executor_specs {
key: "Trainer"
value {
python_class_executable_spec {
class_path: "tfx.components.trainer.executor.GenericExecutor"
}
}
}
custom_driver_specs {
key: "CsvExampleGen"
value {
python_class_executable_spec {
class_path: "tfx.components.example_gen.driver.FileBasedDriver"
}
}
}
metadata_connection_config {
sqlite {
filename_uri: "metadata/penguin-tfma/metadata.db"
connection_mode: READWRITE_OPENCREATE
}
}
INFO:absl:Using connection config:
sqlite {
filename_uri: "metadata/penguin-tfma/metadata.db"
connection_mode: READWRITE_OPENCREATE
}
INFO:absl:Component CsvExampleGen is running.
INFO:absl:Running launcher for node_info {
type {
name: "tfx.components.example_gen.csv_example_gen.component.CsvExampleGen"
}
id: "CsvExampleGen"
}
contexts {
contexts {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
contexts {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
contexts {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.CsvExampleGen"
}
}
}
}
outputs {
outputs {
key: "examples"
value {
artifact_spec {
type {
name: "Examples"
properties {
key: "span"
value: INT
}
properties {
key: "split_names"
value: STRING
}
properties {
key: "version"
value: INT
}
}
}
}
}
}
parameters {
parameters {
key: "input_base"
value {
field_value {
string_value: "/tmp/tfx-datal5lxy_yw"
}
}
}
parameters {
key: "input_config"
value {
field_value {
string_value: "{\n \"splits\": [\n {\n \"name\": \"single_split\",\n \"pattern\": \"*\"\n }\n ]\n}"
}
}
}
parameters {
key: "output_config"
value {
field_value {
string_value: "{\n \"split_config\": {\n \"splits\": [\n {\n \"hash_buckets\": 2,\n \"name\": \"train\"\n },\n {\n \"hash_buckets\": 1,\n \"name\": \"eval\"\n }\n ]\n }\n}"
}
}
}
parameters {
key: "output_data_format"
value {
field_value {
int_value: 6
}
}
}
parameters {
key: "output_file_format"
value {
field_value {
int_value: 5
}
}
}
}
downstream_nodes: "Evaluator"
downstream_nodes: "Trainer"
execution_options {
caching_options {
}
}
INFO:absl:MetadataStore with DB connection initialized
running bdist_wheel
running build
running build_py
creating build
creating build/lib
copying penguin_trainer.py -> build/lib
installing to /tmp/tmp6s2sw4dj
running install
running install_lib
copying build/lib/penguin_trainer.py -> /tmp/tmp6s2sw4dj
running install_egg_info
running egg_info
creating tfx_user_code_Trainer.egg-info
writing tfx_user_code_Trainer.egg-info/PKG-INFO
writing dependency_links to tfx_user_code_Trainer.egg-info/dependency_links.txt
writing top-level names to tfx_user_code_Trainer.egg-info/top_level.txt
writing manifest file 'tfx_user_code_Trainer.egg-info/SOURCES.txt'
reading manifest file 'tfx_user_code_Trainer.egg-info/SOURCES.txt'
writing manifest file 'tfx_user_code_Trainer.egg-info/SOURCES.txt'
Copying tfx_user_code_Trainer.egg-info to /tmp/tmp6s2sw4dj/tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703-py3.7.egg-info
running install_scripts
creating /tmp/tmp6s2sw4dj/tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703.dist-info/WHEEL
creating '/tmp/tmp6jr76e54/tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703-py3-none-any.whl' and adding '/tmp/tmp6s2sw4dj' to it
adding 'penguin_trainer.py'
adding 'tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703.dist-info/METADATA'
adding 'tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703.dist-info/WHEEL'
adding 'tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703.dist-info/top_level.txt'
adding 'tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703.dist-info/RECORD'
removing /tmp/tmp6s2sw4dj
WARNING: Logging before InitGoogleLogging() is written to STDERR
I1205 10:34:23.723806 28099 rdbms_metadata_access_object.cc:686] No property is defined for the Type
I1205 10:34:23.730262 28099 rdbms_metadata_access_object.cc:686] No property is defined for the Type
I1205 10:34:23.736788 28099 rdbms_metadata_access_object.cc:686] No property is defined for the Type
I1205 10:34:23.744907 28099 rdbms_metadata_access_object.cc:686] No property is defined for the Type
INFO:absl:select span and version = (0, None)
INFO:absl:latest span and version = (0, None)
INFO:absl:MetadataStore with DB connection initialized
I1205 10:34:23.758380 28099 rdbms_metadata_access_object.cc:686] No property is defined for the Type
INFO:absl:Going to run a new execution 1
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=1, input_dict={}, output_dict=defaultdict(<class 'list'>, {'examples': [Artifact(artifact: uri: "pipelines/penguin-tfma/CsvExampleGen/examples/1"
custom_properties {
key: "input_fingerprint"
value {
string_value: "split:single_split,num_files:1,total_bytes:25648,xor_checksum:1638700463,sum_checksum:1638700463"
}
}
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:CsvExampleGen:examples:0"
}
}
custom_properties {
key: "span"
value {
int_value: 0
}
}
, artifact_type: name: "Examples"
properties {
key: "span"
value: INT
}
properties {
key: "split_names"
value: STRING
}
properties {
key: "version"
value: INT
}
)]}), exec_properties={'output_file_format': 5, 'output_config': '{\n "split_config": {\n "splits": [\n {\n "hash_buckets": 2,\n "name": "train"\n },\n {\n "hash_buckets": 1,\n "name": "eval"\n }\n ]\n }\n}', 'input_config': '{\n "splits": [\n {\n "name": "single_split",\n "pattern": "*"\n }\n ]\n}', 'output_data_format': 6, 'input_base': '/tmp/tfx-datal5lxy_yw', 'span': 0, 'version': None, 'input_fingerprint': 'split:single_split,num_files:1,total_bytes:25648,xor_checksum:1638700463,sum_checksum:1638700463'}, execution_output_uri='pipelines/penguin-tfma/CsvExampleGen/.system/executor_execution/1/executor_output.pb', stateful_working_dir='pipelines/penguin-tfma/CsvExampleGen/.system/stateful_working_dir/2021-12-05T10:34:23.517028', tmp_dir='pipelines/penguin-tfma/CsvExampleGen/.system/executor_execution/1/.temp/', pipeline_node=node_info {
type {
name: "tfx.components.example_gen.csv_example_gen.component.CsvExampleGen"
}
id: "CsvExampleGen"
}
contexts {
contexts {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
contexts {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
contexts {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.CsvExampleGen"
}
}
}
}
outputs {
outputs {
key: "examples"
value {
artifact_spec {
type {
name: "Examples"
properties {
key: "span"
value: INT
}
properties {
key: "split_names"
value: STRING
}
properties {
key: "version"
value: INT
}
}
}
}
}
}
parameters {
parameters {
key: "input_base"
value {
field_value {
string_value: "/tmp/tfx-datal5lxy_yw"
}
}
}
parameters {
key: "input_config"
value {
field_value {
string_value: "{\n \"splits\": [\n {\n \"name\": \"single_split\",\n \"pattern\": \"*\"\n }\n ]\n}"
}
}
}
parameters {
key: "output_config"
value {
field_value {
string_value: "{\n \"split_config\": {\n \"splits\": [\n {\n \"hash_buckets\": 2,\n \"name\": \"train\"\n },\n {\n \"hash_buckets\": 1,\n \"name\": \"eval\"\n }\n ]\n }\n}"
}
}
}
parameters {
key: "output_data_format"
value {
field_value {
int_value: 6
}
}
}
parameters {
key: "output_file_format"
value {
field_value {
int_value: 5
}
}
}
}
downstream_nodes: "Evaluator"
downstream_nodes: "Trainer"
execution_options {
caching_options {
}
}
, pipeline_info=id: "penguin-tfma"
, pipeline_run_id='2021-12-05T10:34:23.517028')
INFO:absl:Generating examples.
WARNING:apache_beam.runners.interactive.interactive_environment:Dependencies required for Interactive Beam PCollection visualization are not available, please use: `pip install apache-beam[interactive]` to install necessary dependencies to enable all data visualization features.
INFO:absl:Processing input csv data /tmp/tfx-datal5lxy_yw/* to TFExample.
WARNING:root:Make sure that locally built Python SDK docker image has Python 3.7 interpreter.
WARNING:apache_beam.io.tfrecordio:Couldn't find python-snappy so the implementation of _TFRecordUtil._masked_crc32c is not as fast as it could be.
INFO:absl:Examples generated.
INFO:absl:Cleaning up stateless execution info.
INFO:absl:Execution 1 succeeded.
INFO:absl:Cleaning up stateful execution info.
INFO:absl:Publishing output artifacts defaultdict(<class 'list'>, {'examples': [Artifact(artifact: uri: "pipelines/penguin-tfma/CsvExampleGen/examples/1"
custom_properties {
key: "input_fingerprint"
value {
string_value: "split:single_split,num_files:1,total_bytes:25648,xor_checksum:1638700463,sum_checksum:1638700463"
}
}
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:CsvExampleGen:examples:0"
}
}
custom_properties {
key: "span"
value {
int_value: 0
}
}
custom_properties {
key: "tfx_version"
value {
string_value: "1.4.0"
}
}
, artifact_type: name: "Examples"
properties {
key: "span"
value: INT
}
properties {
key: "split_names"
value: STRING
}
properties {
key: "version"
value: INT
}
)]}) for execution 1
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component CsvExampleGen is finished.
INFO:absl:Component latest_blessed_model_resolver is running.
INFO:absl:Running launcher for node_info {
type {
name: "tfx.dsl.components.common.resolver.Resolver"
}
id: "latest_blessed_model_resolver"
}
contexts {
contexts {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
contexts {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
contexts {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.latest_blessed_model_resolver"
}
}
}
}
inputs {
inputs {
key: "model"
value {
channels {
context_queries {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
artifact_query {
type {
name: "Model"
}
}
}
}
}
inputs {
key: "model_blessing"
value {
channels {
context_queries {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
artifact_query {
type {
name: "ModelBlessing"
}
}
}
}
}
resolver_config {
resolver_steps {
class_path: "tfx.dsl.input_resolution.strategies.latest_blessed_model_strategy.LatestBlessedModelStrategy"
config_json: "{}"
input_keys: "model"
input_keys: "model_blessing"
}
}
}
downstream_nodes: "Evaluator"
execution_options {
caching_options {
}
}
INFO:absl:Running as an resolver node.
INFO:absl:MetadataStore with DB connection initialized
WARNING:absl:Artifact type Model is not found in MLMD.
WARNING:absl:Artifact type ModelBlessing is not found in MLMD.
I1205 10:34:24.899447 28099 rdbms_metadata_access_object.cc:686] No property is defined for the Type
INFO:absl:Component latest_blessed_model_resolver is finished.
INFO:absl:Component Trainer is running.
INFO:absl:Running launcher for node_info {
type {
name: "tfx.components.trainer.component.Trainer"
}
id: "Trainer"
}
contexts {
contexts {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
contexts {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
contexts {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.Trainer"
}
}
}
}
inputs {
inputs {
key: "examples"
value {
channels {
producer_node_query {
id: "CsvExampleGen"
}
context_queries {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
context_queries {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
context_queries {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.CsvExampleGen"
}
}
}
artifact_query {
type {
name: "Examples"
}
}
output_key: "examples"
}
min_count: 1
}
}
}
outputs {
outputs {
key: "model"
value {
artifact_spec {
type {
name: "Model"
}
}
}
}
outputs {
key: "model_run"
value {
artifact_spec {
type {
name: "ModelRun"
}
}
}
}
}
parameters {
parameters {
key: "custom_config"
value {
field_value {
string_value: "null"
}
}
}
parameters {
key: "eval_args"
value {
field_value {
string_value: "{\n \"num_steps\": 5\n}"
}
}
}
parameters {
key: "module_path"
value {
field_value {
string_value: "penguin_trainer@pipelines/penguin-tfma/_wheels/tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703-py3-none-any.whl"
}
}
}
parameters {
key: "train_args"
value {
field_value {
string_value: "{\n \"num_steps\": 100\n}"
}
}
}
}
upstream_nodes: "CsvExampleGen"
downstream_nodes: "Evaluator"
downstream_nodes: "Pusher"
execution_options {
caching_options {
}
}
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:MetadataStore with DB connection initialized
I1205 10:34:24.924589 28099 rdbms_metadata_access_object.cc:686] No property is defined for the Type
INFO:absl:Going to run a new execution 3
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=3, input_dict={'examples': [Artifact(artifact: id: 1
type_id: 15
uri: "pipelines/penguin-tfma/CsvExampleGen/examples/1"
properties {
key: "split_names"
value {
string_value: "[\"train\", \"eval\"]"
}
}
custom_properties {
key: "file_format"
value {
string_value: "tfrecords_gzip"
}
}
custom_properties {
key: "input_fingerprint"
value {
string_value: "split:single_split,num_files:1,total_bytes:25648,xor_checksum:1638700463,sum_checksum:1638700463"
}
}
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:CsvExampleGen:examples:0"
}
}
custom_properties {
key: "payload_format"
value {
string_value: "FORMAT_TF_EXAMPLE"
}
}
custom_properties {
key: "span"
value {
int_value: 0
}
}
custom_properties {
key: "tfx_version"
value {
string_value: "1.4.0"
}
}
state: LIVE
create_time_since_epoch: 1638700464882
last_update_time_since_epoch: 1638700464882
, artifact_type: id: 15
name: "Examples"
properties {
key: "span"
value: INT
}
properties {
key: "split_names"
value: STRING
}
properties {
key: "version"
value: INT
}
)]}, output_dict=defaultdict(<class 'list'>, {'model_run': [Artifact(artifact: uri: "pipelines/penguin-tfma/Trainer/model_run/3"
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:Trainer:model_run:0"
}
}
, artifact_type: name: "ModelRun"
)], 'model': [Artifact(artifact: uri: "pipelines/penguin-tfma/Trainer/model/3"
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:Trainer:model:0"
}
}
, artifact_type: name: "Model"
)]}), exec_properties={'train_args': '{\n "num_steps": 100\n}', 'custom_config': 'null', 'eval_args': '{\n "num_steps": 5\n}', 'module_path': 'penguin_trainer@pipelines/penguin-tfma/_wheels/tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703-py3-none-any.whl'}, execution_output_uri='pipelines/penguin-tfma/Trainer/.system/executor_execution/3/executor_output.pb', stateful_working_dir='pipelines/penguin-tfma/Trainer/.system/stateful_working_dir/2021-12-05T10:34:23.517028', tmp_dir='pipelines/penguin-tfma/Trainer/.system/executor_execution/3/.temp/', pipeline_node=node_info {
type {
name: "tfx.components.trainer.component.Trainer"
}
id: "Trainer"
}
contexts {
contexts {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
contexts {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
contexts {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.Trainer"
}
}
}
}
inputs {
inputs {
key: "examples"
value {
channels {
producer_node_query {
id: "CsvExampleGen"
}
context_queries {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
context_queries {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
context_queries {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.CsvExampleGen"
}
}
}
artifact_query {
type {
name: "Examples"
}
}
output_key: "examples"
}
min_count: 1
}
}
}
outputs {
outputs {
key: "model"
value {
artifact_spec {
type {
name: "Model"
}
}
}
}
outputs {
key: "model_run"
value {
artifact_spec {
type {
name: "ModelRun"
}
}
}
}
}
parameters {
parameters {
key: "custom_config"
value {
field_value {
string_value: "null"
}
}
}
parameters {
key: "eval_args"
value {
field_value {
string_value: "{\n \"num_steps\": 5\n}"
}
}
}
parameters {
key: "module_path"
value {
field_value {
string_value: "penguin_trainer@pipelines/penguin-tfma/_wheels/tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703-py3-none-any.whl"
}
}
}
parameters {
key: "train_args"
value {
field_value {
string_value: "{\n \"num_steps\": 100\n}"
}
}
}
}
upstream_nodes: "CsvExampleGen"
downstream_nodes: "Evaluator"
downstream_nodes: "Pusher"
execution_options {
caching_options {
}
}
, pipeline_info=id: "penguin-tfma"
, pipeline_run_id='2021-12-05T10:34:23.517028')
INFO:absl:Train on the 'train' split when train_args.splits is not set.
INFO:absl:Evaluate on the 'eval' split when eval_args.splits is not set.
INFO:absl:udf_utils.get_fn {'train_args': '{\n "num_steps": 100\n}', 'custom_config': 'null', 'eval_args': '{\n "num_steps": 5\n}', 'module_path': 'penguin_trainer@pipelines/penguin-tfma/_wheels/tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703-py3-none-any.whl'} 'run_fn'
INFO:absl:Installing 'pipelines/penguin-tfma/_wheels/tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703-py3-none-any.whl' to a temporary directory.
INFO:absl:Executing: ['/tmpfs/src/tf_docs_env/bin/python', '-m', 'pip', 'install', '--target', '/tmp/tmpc97ini82', 'pipelines/penguin-tfma/_wheels/tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703-py3-none-any.whl']
Processing ./pipelines/penguin-tfma/_wheels/tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703-py3-none-any.whl
INFO:absl:Successfully installed 'pipelines/penguin-tfma/_wheels/tfx_user_code_Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703-py3-none-any.whl'.
INFO:absl:Training model.
INFO:absl:Feature body_mass_g has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature culmen_depth_mm has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature culmen_length_mm has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature flipper_length_mm has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature species has a shape dim {
size: 1
}
. Setting to DenseTensor.
Installing collected packages: tfx-user-code-Trainer
Successfully installed tfx-user-code-Trainer-0.0+1e19049dced0ccb21e0af60dae1c6e0ef09b63d1ff0e370d7f699920c2735703
INFO:absl:Feature body_mass_g has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature culmen_depth_mm has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature culmen_length_mm has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature flipper_length_mm has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature species has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature body_mass_g has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature culmen_depth_mm has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature culmen_length_mm has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature flipper_length_mm has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature species has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature body_mass_g has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature culmen_depth_mm has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature culmen_length_mm has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature flipper_length_mm has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Feature species has a shape dim {
size: 1
}
. Setting to DenseTensor.
INFO:absl:Model: "model"
INFO:absl:__________________________________________________________________________________________________
INFO:absl:Layer (type) Output Shape Param # Connected to
INFO:absl:==================================================================================================
INFO:absl:culmen_length_mm (InputLayer) [(None, 1)] 0
INFO:absl:__________________________________________________________________________________________________
INFO:absl:culmen_depth_mm (InputLayer) [(None, 1)] 0
INFO:absl:__________________________________________________________________________________________________
INFO:absl:flipper_length_mm (InputLayer) [(None, 1)] 0
INFO:absl:__________________________________________________________________________________________________
INFO:absl:body_mass_g (InputLayer) [(None, 1)] 0
INFO:absl:__________________________________________________________________________________________________
INFO:absl:concatenate (Concatenate) (None, 4) 0 culmen_length_mm[0][0]
INFO:absl: culmen_depth_mm[0][0]
INFO:absl: flipper_length_mm[0][0]
INFO:absl: body_mass_g[0][0]
INFO:absl:__________________________________________________________________________________________________
INFO:absl:dense (Dense) (None, 8) 40 concatenate[0][0]
INFO:absl:__________________________________________________________________________________________________
INFO:absl:dense_1 (Dense) (None, 8) 72 dense[0][0]
INFO:absl:__________________________________________________________________________________________________
INFO:absl:dense_2 (Dense) (None, 3) 27 dense_1[0][0]
INFO:absl:==================================================================================================
INFO:absl:Total params: 139
INFO:absl:Trainable params: 139
INFO:absl:Non-trainable params: 0
INFO:absl:__________________________________________________________________________________________________
100/100 [==============================] - 1s 3ms/step - loss: 0.5273 - sparse_categorical_accuracy: 0.8175 - val_loss: 0.2412 - val_sparse_categorical_accuracy: 0.9600
2021-12-05 10:34:29.879208: W tensorflow/python/util/util.cc:348] Sets are not currently considered sequences, but this may change in the future, so consider avoiding using them.
INFO:tensorflow:Assets written to: pipelines/penguin-tfma/Trainer/model/3/Format-Serving/assets
INFO:tensorflow:Assets written to: pipelines/penguin-tfma/Trainer/model/3/Format-Serving/assets
INFO:absl:Training complete. Model written to pipelines/penguin-tfma/Trainer/model/3/Format-Serving. ModelRun written to pipelines/penguin-tfma/Trainer/model_run/3
INFO:absl:Cleaning up stateless execution info.
INFO:absl:Execution 3 succeeded.
INFO:absl:Cleaning up stateful execution info.
INFO:absl:Publishing output artifacts defaultdict(<class 'list'>, {'model_run': [Artifact(artifact: uri: "pipelines/penguin-tfma/Trainer/model_run/3"
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:Trainer:model_run:0"
}
}
custom_properties {
key: "tfx_version"
value {
string_value: "1.4.0"
}
}
, artifact_type: name: "ModelRun"
)], 'model': [Artifact(artifact: uri: "pipelines/penguin-tfma/Trainer/model/3"
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:Trainer:model:0"
}
}
custom_properties {
key: "tfx_version"
value {
string_value: "1.4.0"
}
}
, artifact_type: name: "Model"
)]}) for execution 3
INFO:absl:MetadataStore with DB connection initialized
I1205 10:34:30.399760 28099 rdbms_metadata_access_object.cc:686] No property is defined for the Type
I1205 10:34:30.404250 28099 rdbms_metadata_access_object.cc:686] No property is defined for the Type
INFO:absl:Component Trainer is finished.
INFO:absl:Component Evaluator is running.
INFO:absl:Running launcher for node_info {
type {
name: "tfx.components.evaluator.component.Evaluator"
}
id: "Evaluator"
}
contexts {
contexts {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
contexts {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
contexts {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.Evaluator"
}
}
}
}
inputs {
inputs {
key: "baseline_model"
value {
channels {
producer_node_query {
id: "latest_blessed_model_resolver"
}
context_queries {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
context_queries {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
context_queries {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.latest_blessed_model_resolver"
}
}
}
artifact_query {
type {
name: "Model"
}
}
output_key: "model"
}
}
}
inputs {
key: "examples"
value {
channels {
producer_node_query {
id: "CsvExampleGen"
}
context_queries {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
context_queries {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
context_queries {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.CsvExampleGen"
}
}
}
artifact_query {
type {
name: "Examples"
}
}
output_key: "examples"
}
min_count: 1
}
}
inputs {
key: "model"
value {
channels {
producer_node_query {
id: "Trainer"
}
context_queries {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
context_queries {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
context_queries {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.Trainer"
}
}
}
artifact_query {
type {
name: "Model"
}
}
output_key: "model"
}
}
}
}
outputs {
outputs {
key: "blessing"
value {
artifact_spec {
type {
name: "ModelBlessing"
}
}
}
}
outputs {
key: "evaluation"
value {
artifact_spec {
type {
name: "ModelEvaluation"
}
}
}
}
}
parameters {
parameters {
key: "eval_config"
value {
field_value {
string_value: "{\n \"metrics_specs\": [\n {\n \"per_slice_thresholds\": {\n \"sparse_categorical_accuracy\": {\n \"thresholds\": [\n {\n \"slicing_specs\": [\n {}\n ],\n \"threshold\": {\n \"change_threshold\": {\n \"absolute\": -1e-10,\n \"direction\": \"HIGHER_IS_BETTER\"\n },\n \"value_threshold\": {\n \"lower_bound\": 0.6\n }\n }\n }\n ]\n }\n }\n }\n ],\n \"model_specs\": [\n {\n \"label_key\": \"species\"\n }\n ],\n \"slicing_specs\": [\n {},\n {\n \"feature_keys\": [\n \"species\"\n ]\n }\n ]\n}"
}
}
}
parameters {
key: "example_splits"
value {
field_value {
string_value: "null"
}
}
}
parameters {
key: "fairness_indicator_thresholds"
value {
field_value {
string_value: "null"
}
}
}
}
upstream_nodes: "CsvExampleGen"
upstream_nodes: "Trainer"
upstream_nodes: "latest_blessed_model_resolver"
downstream_nodes: "Pusher"
execution_options {
caching_options {
}
}
INFO:absl:MetadataStore with DB connection initialized
I1205 10:34:30.428037 28099 rdbms_metadata_access_object.cc:686] No property is defined for the Type
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Going to run a new execution 4
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=4, input_dict={'examples': [Artifact(artifact: id: 1
type_id: 15
uri: "pipelines/penguin-tfma/CsvExampleGen/examples/1"
properties {
key: "split_names"
value {
string_value: "[\"train\", \"eval\"]"
}
}
custom_properties {
key: "file_format"
value {
string_value: "tfrecords_gzip"
}
}
custom_properties {
key: "input_fingerprint"
value {
string_value: "split:single_split,num_files:1,total_bytes:25648,xor_checksum:1638700463,sum_checksum:1638700463"
}
}
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:CsvExampleGen:examples:0"
}
}
custom_properties {
key: "payload_format"
value {
string_value: "FORMAT_TF_EXAMPLE"
}
}
custom_properties {
key: "span"
value {
int_value: 0
}
}
custom_properties {
key: "tfx_version"
value {
string_value: "1.4.0"
}
}
state: LIVE
create_time_since_epoch: 1638700464882
last_update_time_since_epoch: 1638700464882
, artifact_type: id: 15
name: "Examples"
properties {
key: "span"
value: INT
}
properties {
key: "split_names"
value: STRING
}
properties {
key: "version"
value: INT
}
)], 'model': [Artifact(artifact: id: 3
type_id: 19
uri: "pipelines/penguin-tfma/Trainer/model/3"
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:Trainer:model:0"
}
}
custom_properties {
key: "tfx_version"
value {
string_value: "1.4.0"
}
}
state: LIVE
create_time_since_epoch: 1638700470409
last_update_time_since_epoch: 1638700470409
, artifact_type: id: 19
name: "Model"
)], 'baseline_model': []}, output_dict=defaultdict(<class 'list'>, {'blessing': [Artifact(artifact: uri: "pipelines/penguin-tfma/Evaluator/blessing/4"
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:Evaluator:blessing:0"
}
}
, artifact_type: name: "ModelBlessing"
)], 'evaluation': [Artifact(artifact: uri: "pipelines/penguin-tfma/Evaluator/evaluation/4"
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:Evaluator:evaluation:0"
}
}
, artifact_type: name: "ModelEvaluation"
)]}), exec_properties={'example_splits': 'null', 'eval_config': '{\n "metrics_specs": [\n {\n "per_slice_thresholds": {\n "sparse_categorical_accuracy": {\n "thresholds": [\n {\n "slicing_specs": [\n {}\n ],\n "threshold": {\n "change_threshold": {\n "absolute": -1e-10,\n "direction": "HIGHER_IS_BETTER"\n },\n "value_threshold": {\n "lower_bound": 0.6\n }\n }\n }\n ]\n }\n }\n }\n ],\n "model_specs": [\n {\n "label_key": "species"\n }\n ],\n "slicing_specs": [\n {},\n {\n "feature_keys": [\n "species"\n ]\n }\n ]\n}', 'fairness_indicator_thresholds': 'null'}, execution_output_uri='pipelines/penguin-tfma/Evaluator/.system/executor_execution/4/executor_output.pb', stateful_working_dir='pipelines/penguin-tfma/Evaluator/.system/stateful_working_dir/2021-12-05T10:34:23.517028', tmp_dir='pipelines/penguin-tfma/Evaluator/.system/executor_execution/4/.temp/', pipeline_node=node_info {
type {
name: "tfx.components.evaluator.component.Evaluator"
}
id: "Evaluator"
}
contexts {
contexts {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
contexts {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
contexts {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.Evaluator"
}
}
}
}
inputs {
inputs {
key: "baseline_model"
value {
channels {
producer_node_query {
id: "latest_blessed_model_resolver"
}
context_queries {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
context_queries {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
context_queries {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.latest_blessed_model_resolver"
}
}
}
artifact_query {
type {
name: "Model"
}
}
output_key: "model"
}
}
}
inputs {
key: "examples"
value {
channels {
producer_node_query {
id: "CsvExampleGen"
}
context_queries {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
context_queries {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
context_queries {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.CsvExampleGen"
}
}
}
artifact_query {
type {
name: "Examples"
}
}
output_key: "examples"
}
min_count: 1
}
}
inputs {
key: "model"
value {
channels {
producer_node_query {
id: "Trainer"
}
context_queries {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
context_queries {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
context_queries {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.Trainer"
}
}
}
artifact_query {
type {
name: "Model"
}
}
output_key: "model"
}
}
}
}
outputs {
outputs {
key: "blessing"
value {
artifact_spec {
type {
name: "ModelBlessing"
}
}
}
}
outputs {
key: "evaluation"
value {
artifact_spec {
type {
name: "ModelEvaluation"
}
}
}
}
}
parameters {
parameters {
key: "eval_config"
value {
field_value {
string_value: "{\n \"metrics_specs\": [\n {\n \"per_slice_thresholds\": {\n \"sparse_categorical_accuracy\": {\n \"thresholds\": [\n {\n \"slicing_specs\": [\n {}\n ],\n \"threshold\": {\n \"change_threshold\": {\n \"absolute\": -1e-10,\n \"direction\": \"HIGHER_IS_BETTER\"\n },\n \"value_threshold\": {\n \"lower_bound\": 0.6\n }\n }\n }\n ]\n }\n }\n }\n ],\n \"model_specs\": [\n {\n \"label_key\": \"species\"\n }\n ],\n \"slicing_specs\": [\n {},\n {\n \"feature_keys\": [\n \"species\"\n ]\n }\n ]\n}"
}
}
}
parameters {
key: "example_splits"
value {
field_value {
string_value: "null"
}
}
}
parameters {
key: "fairness_indicator_thresholds"
value {
field_value {
string_value: "null"
}
}
}
}
upstream_nodes: "CsvExampleGen"
upstream_nodes: "Trainer"
upstream_nodes: "latest_blessed_model_resolver"
downstream_nodes: "Pusher"
execution_options {
caching_options {
}
}
, pipeline_info=id: "penguin-tfma"
, pipeline_run_id='2021-12-05T10:34:23.517028')
INFO:absl:udf_utils.get_fn {'example_splits': 'null', 'eval_config': '{\n "metrics_specs": [\n {\n "per_slice_thresholds": {\n "sparse_categorical_accuracy": {\n "thresholds": [\n {\n "slicing_specs": [\n {}\n ],\n "threshold": {\n "change_threshold": {\n "absolute": -1e-10,\n "direction": "HIGHER_IS_BETTER"\n },\n "value_threshold": {\n "lower_bound": 0.6\n }\n }\n }\n ]\n }\n }\n }\n ],\n "model_specs": [\n {\n "label_key": "species"\n }\n ],\n "slicing_specs": [\n {},\n {\n "feature_keys": [\n "species"\n ]\n }\n ]\n}', 'fairness_indicator_thresholds': 'null'} 'custom_eval_shared_model'
INFO:absl:Request was made to ignore the baseline ModelSpec and any change thresholds. This is likely because a baseline model was not provided: updated_config=
model_specs {
label_key: "species"
}
slicing_specs {
}
slicing_specs {
feature_keys: "species"
}
metrics_specs {
per_slice_thresholds {
key: "sparse_categorical_accuracy"
value {
thresholds {
slicing_specs {
}
threshold {
value_threshold {
lower_bound {
value: 0.6
}
}
}
}
}
}
}
INFO:absl:Using pipelines/penguin-tfma/Trainer/model/3/Format-Serving as model.
INFO:absl:The 'example_splits' parameter is not set, using 'eval' split.
INFO:absl:Evaluating model.
INFO:absl:udf_utils.get_fn {'example_splits': 'null', 'eval_config': '{\n "metrics_specs": [\n {\n "per_slice_thresholds": {\n "sparse_categorical_accuracy": {\n "thresholds": [\n {\n "slicing_specs": [\n {}\n ],\n "threshold": {\n "change_threshold": {\n "absolute": -1e-10,\n "direction": "HIGHER_IS_BETTER"\n },\n "value_threshold": {\n "lower_bound": 0.6\n }\n }\n }\n ]\n }\n }\n }\n ],\n "model_specs": [\n {\n "label_key": "species"\n }\n ],\n "slicing_specs": [\n {},\n {\n "feature_keys": [\n "species"\n ]\n }\n ]\n}', 'fairness_indicator_thresholds': 'null'} 'custom_extractors'
INFO:absl:Request was made to ignore the baseline ModelSpec and any change thresholds. This is likely because a baseline model was not provided: updated_config=
model_specs {
label_key: "species"
}
slicing_specs {
}
slicing_specs {
feature_keys: "species"
}
metrics_specs {
model_names: ""
per_slice_thresholds {
key: "sparse_categorical_accuracy"
value {
thresholds {
slicing_specs {
}
threshold {
value_threshold {
lower_bound {
value: 0.6
}
}
}
}
}
}
}
INFO:absl:Request was made to ignore the baseline ModelSpec and any change thresholds. This is likely because a baseline model was not provided: updated_config=
model_specs {
label_key: "species"
}
slicing_specs {
}
slicing_specs {
feature_keys: "species"
}
metrics_specs {
model_names: ""
per_slice_thresholds {
key: "sparse_categorical_accuracy"
value {
thresholds {
slicing_specs {
}
threshold {
value_threshold {
lower_bound {
value: 0.6
}
}
}
}
}
}
}
INFO:absl:Request was made to ignore the baseline ModelSpec and any change thresholds. This is likely because a baseline model was not provided: updated_config=
model_specs {
label_key: "species"
}
slicing_specs {
}
slicing_specs {
feature_keys: "species"
}
metrics_specs {
model_names: ""
per_slice_thresholds {
key: "sparse_categorical_accuracy"
value {
thresholds {
slicing_specs {
}
threshold {
value_threshold {
lower_bound {
value: 0.6
}
}
}
}
}
}
}
WARNING:root:Make sure that locally built Python SDK docker image has Python 3.7 interpreter.
INFO:absl:Evaluation complete. Results written to pipelines/penguin-tfma/Evaluator/evaluation/4.
INFO:absl:Checking validation results.
WARNING:tensorflow:From /tmpfs/src/tf_docs_env/lib/python3.7/site-packages/tensorflow_model_analysis/writers/metrics_plots_and_validations_writer.py:114: tf_record_iterator (from tensorflow.python.lib.io.tf_record) is deprecated and will be removed in a future version.
Instructions for updating:
Use eager execution and:
`tf.data.TFRecordDataset(path)`
WARNING:tensorflow:From /tmpfs/src/tf_docs_env/lib/python3.7/site-packages/tensorflow_model_analysis/writers/metrics_plots_and_validations_writer.py:114: tf_record_iterator (from tensorflow.python.lib.io.tf_record) is deprecated and will be removed in a future version.
Instructions for updating:
Use eager execution and:
`tf.data.TFRecordDataset(path)`
INFO:absl:Blessing result True written to pipelines/penguin-tfma/Evaluator/blessing/4.
INFO:absl:Cleaning up stateless execution info.
INFO:absl:Execution 4 succeeded.
INFO:absl:Cleaning up stateful execution info.
INFO:absl:Publishing output artifacts defaultdict(<class 'list'>, {'blessing': [Artifact(artifact: uri: "pipelines/penguin-tfma/Evaluator/blessing/4"
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:Evaluator:blessing:0"
}
}
custom_properties {
key: "tfx_version"
value {
string_value: "1.4.0"
}
}
, artifact_type: name: "ModelBlessing"
)], 'evaluation': [Artifact(artifact: uri: "pipelines/penguin-tfma/Evaluator/evaluation/4"
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:Evaluator:evaluation:0"
}
}
custom_properties {
key: "tfx_version"
value {
string_value: "1.4.0"
}
}
, artifact_type: name: "ModelEvaluation"
)]}) for execution 4
INFO:absl:MetadataStore with DB connection initialized
I1205 10:34:35.040588 28099 rdbms_metadata_access_object.cc:686] No property is defined for the Type
I1205 10:34:35.045548 28099 rdbms_metadata_access_object.cc:686] No property is defined for the Type
INFO:absl:Component Evaluator is finished.
INFO:absl:Component Pusher is running.
INFO:absl:Running launcher for node_info {
type {
name: "tfx.components.pusher.component.Pusher"
}
id: "Pusher"
}
contexts {
contexts {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
contexts {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
contexts {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.Pusher"
}
}
}
}
inputs {
inputs {
key: "model"
value {
channels {
producer_node_query {
id: "Trainer"
}
context_queries {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
context_queries {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
context_queries {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.Trainer"
}
}
}
artifact_query {
type {
name: "Model"
}
}
output_key: "model"
}
}
}
inputs {
key: "model_blessing"
value {
channels {
producer_node_query {
id: "Evaluator"
}
context_queries {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
context_queries {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
context_queries {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.Evaluator"
}
}
}
artifact_query {
type {
name: "ModelBlessing"
}
}
output_key: "blessing"
}
}
}
}
outputs {
outputs {
key: "pushed_model"
value {
artifact_spec {
type {
name: "PushedModel"
}
}
}
}
}
parameters {
parameters {
key: "custom_config"
value {
field_value {
string_value: "null"
}
}
}
parameters {
key: "push_destination"
value {
field_value {
string_value: "{\n \"filesystem\": {\n \"base_directory\": \"serving_model/penguin-tfma\"\n }\n}"
}
}
}
}
upstream_nodes: "Evaluator"
upstream_nodes: "Trainer"
execution_options {
caching_options {
}
}
INFO:absl:MetadataStore with DB connection initialized
I1205 10:34:35.068168 28099 rdbms_metadata_access_object.cc:686] No property is defined for the Type
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Going to run a new execution 5
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=5, input_dict={'model': [Artifact(artifact: id: 3
type_id: 19
uri: "pipelines/penguin-tfma/Trainer/model/3"
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:Trainer:model:0"
}
}
custom_properties {
key: "tfx_version"
value {
string_value: "1.4.0"
}
}
state: LIVE
create_time_since_epoch: 1638700470409
last_update_time_since_epoch: 1638700470409
, artifact_type: id: 19
name: "Model"
)], 'model_blessing': [Artifact(artifact: id: 4
type_id: 21
uri: "pipelines/penguin-tfma/Evaluator/blessing/4"
custom_properties {
key: "blessed"
value {
int_value: 1
}
}
custom_properties {
key: "current_model"
value {
string_value: "pipelines/penguin-tfma/Trainer/model/3"
}
}
custom_properties {
key: "current_model_id"
value {
int_value: 3
}
}
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:Evaluator:blessing:0"
}
}
custom_properties {
key: "tfx_version"
value {
string_value: "1.4.0"
}
}
state: LIVE
create_time_since_epoch: 1638700475049
last_update_time_since_epoch: 1638700475049
, artifact_type: id: 21
name: "ModelBlessing"
)]}, output_dict=defaultdict(<class 'list'>, {'pushed_model': [Artifact(artifact: uri: "pipelines/penguin-tfma/Pusher/pushed_model/5"
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:Pusher:pushed_model:0"
}
}
, artifact_type: name: "PushedModel"
)]}), exec_properties={'custom_config': 'null', 'push_destination': '{\n "filesystem": {\n "base_directory": "serving_model/penguin-tfma"\n }\n}'}, execution_output_uri='pipelines/penguin-tfma/Pusher/.system/executor_execution/5/executor_output.pb', stateful_working_dir='pipelines/penguin-tfma/Pusher/.system/stateful_working_dir/2021-12-05T10:34:23.517028', tmp_dir='pipelines/penguin-tfma/Pusher/.system/executor_execution/5/.temp/', pipeline_node=node_info {
type {
name: "tfx.components.pusher.component.Pusher"
}
id: "Pusher"
}
contexts {
contexts {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
contexts {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
contexts {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.Pusher"
}
}
}
}
inputs {
inputs {
key: "model"
value {
channels {
producer_node_query {
id: "Trainer"
}
context_queries {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
context_queries {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
context_queries {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.Trainer"
}
}
}
artifact_query {
type {
name: "Model"
}
}
output_key: "model"
}
}
}
inputs {
key: "model_blessing"
value {
channels {
producer_node_query {
id: "Evaluator"
}
context_queries {
type {
name: "pipeline"
}
name {
field_value {
string_value: "penguin-tfma"
}
}
}
context_queries {
type {
name: "pipeline_run"
}
name {
field_value {
string_value: "2021-12-05T10:34:23.517028"
}
}
}
context_queries {
type {
name: "node"
}
name {
field_value {
string_value: "penguin-tfma.Evaluator"
}
}
}
artifact_query {
type {
name: "ModelBlessing"
}
}
output_key: "blessing"
}
}
}
}
outputs {
outputs {
key: "pushed_model"
value {
artifact_spec {
type {
name: "PushedModel"
}
}
}
}
}
parameters {
parameters {
key: "custom_config"
value {
field_value {
string_value: "null"
}
}
}
parameters {
key: "push_destination"
value {
field_value {
string_value: "{\n \"filesystem\": {\n \"base_directory\": \"serving_model/penguin-tfma\"\n }\n}"
}
}
}
}
upstream_nodes: "Evaluator"
upstream_nodes: "Trainer"
execution_options {
caching_options {
}
}
, pipeline_info=id: "penguin-tfma"
, pipeline_run_id='2021-12-05T10:34:23.517028')
INFO:absl:Model version: 1638700475
INFO:absl:Model written to serving path serving_model/penguin-tfma/1638700475.
INFO:absl:Model pushed to pipelines/penguin-tfma/Pusher/pushed_model/5.
INFO:absl:Cleaning up stateless execution info.
INFO:absl:Execution 5 succeeded.
INFO:absl:Cleaning up stateful execution info.
INFO:absl:Publishing output artifacts defaultdict(<class 'list'>, {'pushed_model': [Artifact(artifact: uri: "pipelines/penguin-tfma/Pusher/pushed_model/5"
custom_properties {
key: "name"
value {
string_value: "penguin-tfma:2021-12-05T10:34:23.517028:Pusher:pushed_model:0"
}
}
custom_properties {
key: "tfx_version"
value {
string_value: "1.4.0"
}
}
, artifact_type: name: "PushedModel"
)]}) for execution 5
INFO:absl:MetadataStore with DB connection initialized
I1205 10:34:35.098553 28099 rdbms_metadata_access_object.cc:686] No property is defined for the Type
INFO:absl:Component Pusher is finished.
Po zakończeniu potoku powinieneś być w stanie zobaczyć coś takiego:
INFO:absl:Blessing result True written to pipelines/penguin-tfma/Evaluator/blessing/4.
Możesz też ręcznie sprawdzić katalog wyjściowy, w którym przechowywane są wygenerowane artefakty. Jeśli wizyta pipelines/penguin-tfma/Evaluator/blessing/ z broswer plików, można zobaczyć plik o nazwie BLESSED lub NOT_BLESSED w zależności od wyniku oceny.
Jeśli wynik jest błogosławieństwem False , Pusher odmówi pchnąć modelu do serving_model_dir , ponieważ model nie jest wystarczająco dobry, aby być wykorzystywane w produkcji.
Możesz ponownie uruchomić potok, prawdopodobnie z różnymi konfiguracjami oceny. Nawet po uruchomieniu rurociągu z dokładnie tej samej konfiguracji i zbiorze, przeszkolonych modelu mogą się nieznacznie różnić z powodu wrodzonej losowości modelu kształcenia, które mogą prowadzić do NOT_BLESSED modelu.
Zbadaj wyjścia rurociągu
Możesz użyć TFMA do zbadania i wizualizacji wyniku oceny w artefakcie ModelEvaluation.
Uzyskaj wynik analizy z artefaktów wyjściowych
Za pomocą interfejsów API MLMD można programowo lokalizować te dane wyjściowe. Najpierw zdefiniujemy kilka funkcji narzędziowych do wyszukiwania artefaktów wyjściowych, które właśnie zostały utworzone.
from ml_metadata.proto import metadata_store_pb2
# Non-public APIs, just for showcase.
from tfx.orchestration.portable.mlmd import execution_lib
# TODO(b/171447278): Move these functions into the TFX library.
def get_latest_artifacts(metadata, pipeline_name, component_id):
"""Output artifacts of the latest run of the component."""
context = metadata.store.get_context_by_type_and_name(
'node', f'{pipeline_name}.{component_id}')
executions = metadata.store.get_executions_by_context(context.id)
latest_execution = max(executions,
key=lambda e:e.last_update_time_since_epoch)
return execution_lib.get_artifacts_dict(metadata, latest_execution.id,
[metadata_store_pb2.Event.OUTPUT])
Możemy znaleźć najnowsze wykonanie Evaluator komponentu i inne artefakty wyjściowe niego.
# Non-public APIs, just for showcase.
from tfx.orchestration.metadata import Metadata
from tfx.types import standard_component_specs
metadata_connection_config = tfx.orchestration.metadata.sqlite_metadata_connection_config(
METADATA_PATH)
with Metadata(metadata_connection_config) as metadata_handler:
# Find output artifacts from MLMD.
evaluator_output = get_latest_artifacts(metadata_handler, PIPELINE_NAME,
'Evaluator')
eval_artifact = evaluator_output[standard_component_specs.EVALUATION_KEY][0]
INFO:absl:MetadataStore with DB connection initialized
Evaluator zawsze zwraca jeden artefakt oceny i możemy wizualizować go przy użyciu biblioteki TensorFlow analizy modelu. Na przykład poniższy kod wygeneruje metryki dokładności dla każdego gatunku pingwina.
import tensorflow_model_analysis as tfma
eval_result = tfma.load_eval_result(eval_artifact.uri)
tfma.view.render_slicing_metrics(eval_result, slicing_column='species')
SlicingMetricsViewer(config={'weightedExamplesColumn': 'example_count'}, data=[{'slice': 'species:0', 'metrics…
Jeśli wybierzesz „” w sparse_categorical_accuracy Show rozwijanej listy można zobaczyć wartości dokładności dla określonych gatunków. Możesz chcieć dodać więcej wycinków i sprawdzić, czy twój model jest dobry dla całej dystrybucji i czy istnieje jakikolwiek możliwy błąd.
Następne kroki
Dowiedz się więcej o analizę modelu w TensorFlow analizy modelu tutorialu biblioteki .
Można znaleźć więcej środków na https://www.tensorflow.org/tfx/tutorials
Proszę zobaczyć Zrozumienie TFX Rurociągi , aby dowiedzieć się więcej na temat różnych koncepcji w TFX.