Penguin şablonuyla verileriniz için bir TFX ardışık düzeni oluşturun

---

Tanıtım

Bu belge, TFX Python paketi ile sağlanan penguen şablonunu kullanarak kendi veri kümeniz için bir TensorFlow Genişletilmiş (TFX) işlem hattı oluşturma talimatlarını sağlayacaktır. Oluşturulan işlem hattı başlangıçta Palmer Penguins veri kümesini kullanacak, ancak biz veri kümeniz için işlem hattını dönüştüreceğiz.

Önkoşullar

• Linux / MacOS
• Python 3.6-3.8
• Jüpiter not defteri

Adım 1. Önceden tanımlanmış şablonu proje dizininize kopyalayın.

Bu adımda, TFX'te penguen şablonundan dosyaları kopyalayarak çalışan bir boru hattı proje dizini ve dosyaları oluşturacağız. Bunu TFX boru hattı projeniz için bir iskele olarak düşünebilirsiniz.

Pip'i Güncelle

Colab'da çalışıyorsak, Pip'in en son sürümüne sahip olduğumuzdan emin olmalıyız. Yerel sistemler elbette ayrı ayrı güncellenebilir.

import sys
if 'google.colab' in sys.modules:
  !pip install --upgrade pip

Gerekli paketi yükleyin

İlk olarak, TFX ve TensorFlow Model Analizini (TFMA) kurun.

pip install -U tfx tensorflow-model-analysis

TFX'in sürümlerini kontrol edelim.

import tensorflow as tf
import tensorflow_model_analysis as tfma
import tfx

print('TF version: {}'.format(tf.__version__))
print('TFMA version: {}'.format(tfma.__version__))
print('TFX version: {}'.format(tfx.__version__))

tutucu3 l10n-yer

TF version: 2.7.1
TFMA version: 0.37.0
TFX version: 1.6.0

Bir boru hattı oluşturmaya hazırız.

PROJECT_DIR öğesini ortamınız için uygun hedefe ayarlayın. Varsayılan değer, Google Cloud AI Platform Notebook ortamı için uygun olan ~/imported/${PIPELINE_NAME} şeklindedir.

Aşağıdaki PIPELINE_NAME değiştirerek ardışık düzeninize farklı bir ad verebilirsiniz. Bu aynı zamanda dosyalarınızın yerleştirileceği proje dizininin adı olacaktır.

PIPELINE_NAME="my_pipeline"
import os
# Set this project directory to your new tfx pipeline project.
PROJECT_DIR=os.path.join(os.path.expanduser("~"), "imported", PIPELINE_NAME)

Şablon dosyalarını kopyalayın.

TFX, penguin şablonunu TFX python paketiyle birlikte içerir. penguin şablonu, bu öğreticinin amacı olan veri kümenizi boru hattına getirmek için birçok talimat içerir.

tfx template copy CLI komutu, önceden tanımlanmış şablon dosyalarını proje dizininize kopyalar.

# Set `PATH` to include user python binary directory and a directory containing `skaffold`.
PATH=%env PATH
%env PATH={PATH}:/home/jupyter/.local/bin

!tfx template copy \
  --pipeline-name={PIPELINE_NAME} \
  --destination-path={PROJECT_DIR} \
  --model=penguin

tutucu6 l10n-yer

env: PATH=/tmpfs/src/tf_docs_env/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/snap/bin:/opt/puppetlabs/bin:/opt/android-studio/current/bin:/usr/local/go/bin:/usr/local/go/packages/bin:/opt/kubernetes/client/bin/:/home/kbuilder/.local/bin:/home/jupyter/.local/bin
CLI
Copying penguin pipeline template
kubeflow_runner.py -> /home/kbuilder/imported/my_pipeline/kubeflow_runner.py
configs.py -> /home/kbuilder/imported/my_pipeline/pipeline/configs.py
pipeline.py -> /home/kbuilder/imported/my_pipeline/pipeline/pipeline.py
__init__.py -> /home/kbuilder/imported/my_pipeline/pipeline/__init__.py
model.py -> /home/kbuilder/imported/my_pipeline/models/model.py
features.py -> /home/kbuilder/imported/my_pipeline/models/features.py
features_test.py -> /home/kbuilder/imported/my_pipeline/models/features_test.py
preprocessing_test.py -> /home/kbuilder/imported/my_pipeline/models/preprocessing_test.py
preprocessing.py -> /home/kbuilder/imported/my_pipeline/models/preprocessing.py
model_test.py -> /home/kbuilder/imported/my_pipeline/models/model_test.py
__init__.py -> /home/kbuilder/imported/my_pipeline/models/__init__.py
constants.py -> /home/kbuilder/imported/my_pipeline/models/constants.py
local_runner.py -> /home/kbuilder/imported/my_pipeline/local_runner.py
__init__.py -> /home/kbuilder/imported/my_pipeline/__init__.py

Bu not defterindeki çalışma dizini bağlamını proje dizinine değiştirin.

%cd {PROJECT_DIR}

tutucu8 l10n-yer

/home/kbuilder/imported/my_pipeline

Kopyalanan kaynak dosyalarınıza göz atın

TFX şablonu, Python kaynak kodu ve örnek veriler dahil olmak üzere bir işlem hattı oluşturmak için temel iskele dosyaları sağlar. penguin şablonu, Penguin örneğiyle aynı Palmer Penguins veri kümesini ve ML modelini kullanır.

İşte Python dosyalarının her birine kısa bir giriş.

• pipeline - Bu dizin boru hattının tanımını içerir
  • configs.py — işlem hattı koşucuları için ortak sabitleri tanımlar
  • pipeline.py hattı.py — TFX bileşenlerini ve bir işlem hattını tanımlar
• models - Bu dizin, ML model tanımlarını içerir
  • feature.py , features_test.py — modelin features.py tanımlar
  • preprocessing.py , preprocessing_test.py — veriler için ön işleme rutinlerini tanımlar
  • constants.py — modelin sabitlerini tanımlar
  • model.py , model_test.py — TensorFlow gibi ML çerçevelerini kullanarak ML modelini tanımlar
• local_runner.py - yerel düzenleme motorunu kullanan yerel ortam için bir koşucu tanımlayın
• kubeflow_runner.py — Kubeflow Pipelines düzenleme motoru için bir koşucu tanımlayın

Varsayılan olarak şablon yalnızca standart TFX bileşenlerini içerir. Bazı özelleştirilmiş eylemlere ihtiyacınız varsa, işlem hattınız için özel bileşenler oluşturabilirsiniz. Ayrıntılar için lütfen TFX özel bileşen kılavuzuna bakın.

Birim test dosyaları.

_test.py olan bazı dosyalar olduğunu fark edebilirsiniz. Bunlar işlem hattının birim testleridir ve kendi işlem hatlarınızı uygularken daha fazla birim testi eklemeniz önerilir. Test dosyalarının modül adını -m bayrağı ile vererek birim testlerini çalıştırabilirsiniz. Genellikle .py uzantısını silerek ve / ile . . Örneğin:

import sys
!{sys.executable} -m models.features_test

tutucu10 l10n-yer

Running tests under Python 3.7.5: /tmpfs/src/tf_docs_env/bin/python
[ RUN      ] FeaturesTest.testLabelKey
INFO:tensorflow:time(__main__.FeaturesTest.testLabelKey): 0.0s
I0203 11:08:46.306882 140258321348416 test_util.py:2309] time(__main__.FeaturesTest.testLabelKey): 0.0s
[       OK ] FeaturesTest.testLabelKey
[ RUN      ] FeaturesTest.test_session
[  SKIPPED ] FeaturesTest.test_session
----------------------------------------------------------------------
Ran 2 tests in 0.001s

OK (skipped=1)

Yerel ortamda bir TFX işlem hattı oluşturun.

TFX, işlem hatlarını çalıştırmak için birkaç düzenleme motorunu destekler. Yerel orkestrasyon motorunu kullanacağız. Yerel düzenleme motoru, başka herhangi bir bağımlılık olmadan çalışır ve uzak bilgi işlem kümelerine bağlı olmak yerine yerel ortamda çalıştığı için geliştirme ve hata ayıklama için uygundur.

İşlem hattınızı yerel orkestratör kullanarak çalıştırmak için local_runner.py kullanacağız. Çalıştırmadan önce bir işlem hattı oluşturmanız gerekir. pipeline create .

tfx pipeline create --engine=local --pipeline_path=local_runner.py

tutucu12 l10n-yer

CLI
Creating pipeline
INFO:absl:Excluding no splits because exclude_splits is not set.
INFO:absl:Excluding no splits because exclude_splits is not set.
Pipeline "my_pipeline" created successfully.

pipeline create komutu, local_runner.py içinde tanımlanan işlem hattınızı gerçekten çalıştırmadan kaydeder.

Oluşturulan boru hattını aşağıdaki adımlarda run create komutu ile çalıştıracaksınız.

Adım 2. Verilerinizi boru hattına alın.

İlk ardışık düzen, şablona dahil edilen penguen veri kümesini alır. Verilerinizi ardışık düzene koymanız gerekir ve çoğu TFX işlem hattı, ExampleGen bileşeniyle başlar.

Bir Örnek Gen seçin

Verileriniz, işlem hattınızın erişebildiği herhangi bir yerde, yerel veya dağıtılmış bir dosya sisteminde veya sorgulanabilir bir sistemde saklanabilir. TFX, verilerinizi bir TFX ardışık düzenine getirmek için çeşitli ExampleGen bileşenleri sağlar. Aşağıdaki örnek oluşturan bileşenlerden birini seçebilirsiniz.

• CsvExampleGen: Bir dizindeki CSV dosyalarını okur. Penguen örneğinde ve Chicago taksi örneğinde kullanılmıştır.
• ImportExampleGen: TF Örnek veri formatı ile TFRecord dosyalarını alır. MNIST örneklerinde kullanılır.
• Avro veya Parke formatı için FileBasedExampleGen.
• BigQueryExampleGen : Google Cloud BigQuery'deki verileri doğrudan okur. Chicago taksi örneklerinde kullanılır.

Ayrıca kendi ExampleGen'inizi de oluşturabilirsiniz; örneğin, tfx, Presto'yu veri kaynağı olarak kullanan özel bir ExecampleGen içerir. Özel yürütücülerin nasıl kullanılacağı ve geliştirileceği hakkında daha fazla bilgi için kılavuza bakın.

Hangi ExampleGen'in kullanılacağına karar verdikten sonra, verilerinizi kullanmak için işlem hattı tanımını değiştirmeniz gerekecektir.

1. DATA_PATH içindeki local_runner.py değiştirin ve dosyalarınızın konumuna ayarlayın.

   • Yerel ortamda dosyalarınız varsa yolu belirtin. Bu, bir işlem hattı geliştirmek veya hata ayıklamak için en iyi seçenektir.
   • Dosyalar GCS'de depolanıyorsa, gs://{bucket_name}/... ile başlayan bir yol kullanabilirsiniz. Lütfen, örneğin gsutil kullanarak GCS'ye terminalinizden erişebildiğinizden emin olun. Gerekirse lütfen Google Cloud'daki yetkilendirme kılavuzunu takip edin.
   • BigQueryExampleGen gibi Sorgu tabanlı bir ÖrnekGen kullanmak istiyorsanız, veri kaynağından veri seçmek için bir Sorgu ifadesine ihtiyacınız vardır. Google Cloud BigQuery'yi veri kaynağı olarak kullanmak için ayarlamanız gereken birkaç şey daha var.
   • pipeline/configs.py :
     • GOOGLE_CLOUD_PROJECT ve GCS_BUCKET_NAME GCP projeniz ve paket adınızla değiştirin. Biz boru hattını çalıştırmadan önce kova mevcut olmalıdır.
     • BIG_QUERY_WITH_DIRECT_RUNNER_BEAM_PIPELINE_ARGS değişkeninin yorumunu kaldırın.
     • Yorumu kaldırın ve BIG_QUERY_QUERY değişkenini sorgu ifadenize ayarlayın.
   • local_runner.py :
     • pipeline.create_pipeline() yerine data_path argümanını ve uncomment query argümanını yorumlayın.
   • pipeline/pipeline.py :
     • create_pipeline() içindeki data_path argümanını ve uncomment query argümanını yorumlayın.
     • CsvExampleGen yerine BigQueryExampleGen kullanın.

2. pipeline/pipeline.py içindeki mevcut CsvExampleGen'i ExampleGen sınıfınızla değiştirin. Her ExampleGen sınıfının farklı imzası vardır. Daha fazla ayrıntı için lütfen ExampleGen bileşen kılavuzuna bakın. pipeline/pipeline.py içindeki import ifadeleri ile gerekli modülleri almayı unutmayın.

İlk işlem hattı, ExampleGen , StatisticsGen , SchemaGen ve ExampleValidator olmak üzere dört bileşenden oluşur. StatisticsGen , SchemaGen ve ExampleValidator için hiçbir şeyi değiştirmemize gerek yok. Boru hattını ilk kez çalıştıralım.

# Update and run the pipeline.
!tfx pipeline update --engine=local --pipeline_path=local_runner.py \
 && tfx run create --engine=local --pipeline_name={PIPELINE_NAME}

tutucu14 l10n-yer

CLI
Updating pipeline
INFO:absl:Excluding no splits because exclude_splits is not set.
INFO:absl:Excluding no splits because exclude_splits is not set.
Pipeline "my_pipeline" updated successfully.
CLI
Creating a run for pipeline: my_pipeline
INFO:absl:Excluding no splits because exclude_splits is not set.
INFO:absl:Excluding no splits because exclude_splits is not set.
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: "SchemaGen"
  value {
    python_class_executable_spec {
      class_path: "tfx.components.schema_gen.executor.Executor"
    }
  }
}
executor_specs {
  key: "StatisticsGen"
  value {
    beam_executable_spec {
      python_executor_spec {
        class_path: "tfx.components.statistics_gen.executor.Executor"
      }
    }
  }
}
custom_driver_specs {
  key: "CsvExampleGen"
  value {
    python_class_executable_spec {
      class_path: "tfx.components.example_gen.driver.FileBasedDriver"
    }
  }
}
metadata_connection_config {
  database_connection_config {
    sqlite {
      filename_uri: "./tfx_metadata/my_pipeline/metadata.db"
      connection_mode: READWRITE_OPENCREATE
    }
  }
}

INFO:absl:Using connection config:
 sqlite {
  filename_uri: "./tfx_metadata/my_pipeline/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: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:09:12.120566"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.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
          }
          base_type: DATASET
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "input_base"
    value {
      field_value {
        string_value: "/home/kbuilder/imported/my_pipeline/data"
      }
    }
  }
  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: "StatisticsGen"
execution_options {
  caching_options {
  }
}

INFO:absl:MetadataStore with DB connection initialized
INFO:absl:select span and version = (0, None)
INFO:absl:latest span and version = (0, None)
INFO:absl:MetadataStore with DB connection initialized
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: "./tfx_pipeline_output/my_pipeline/CsvExampleGen/examples/1"
custom_properties {
  key: "input_fingerprint"
  value {
    string_value: "split:single_split,num_files:1,total_bytes:25648,xor_checksum:1643886522,sum_checksum:1643886522"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:12.120566: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
}
base_type: DATASET
)]}), exec_properties={'input_config': '{\n  "splits": [\n    {\n      "name": "single_split",\n      "pattern": "*"\n    }\n  ]\n}', 'output_data_format': 6, '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_base': '/home/kbuilder/imported/my_pipeline/data', 'span': 0, 'version': None, 'input_fingerprint': 'split:single_split,num_files:1,total_bytes:25648,xor_checksum:1643886522,sum_checksum:1643886522'}, execution_output_uri='./tfx_pipeline_output/my_pipeline/CsvExampleGen/.system/executor_execution/1/executor_output.pb', stateful_working_dir='./tfx_pipeline_output/my_pipeline/CsvExampleGen/.system/stateful_working_dir/2022-02-03T11:09:12.120566', tmp_dir='./tfx_pipeline_output/my_pipeline/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: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:09:12.120566"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.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
          }
          base_type: DATASET
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "input_base"
    value {
      field_value {
        string_value: "/home/kbuilder/imported/my_pipeline/data"
      }
    }
  }
  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: "StatisticsGen"
execution_options {
  caching_options {
  }
}
, pipeline_info=id: "my_pipeline"
, pipeline_run_id='2022-02-03T11:09:12.120566')
INFO:absl:Generating examples.
INFO:absl:Processing input csv data /home/kbuilder/imported/my_pipeline/data/* to TFExample.
WARNING:root:Make sure that locally built Python SDK docker image has Python 3.7 interpreter.
E0203 11:09:12.848598153    5127 fork_posix.cc:70]           Fork support is only compatible with the epoll1 and poll polling strategies
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:Value type <class 'NoneType'> of key version in exec_properties is not supported, going to drop it
INFO:absl:Value type <class 'list'> of key _beam_pipeline_args in exec_properties is not supported, going to drop it
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: "./tfx_pipeline_output/my_pipeline/CsvExampleGen/examples/1"
custom_properties {
  key: "input_fingerprint"
  value {
    string_value: "split:single_split,num_files:1,total_bytes:25648,xor_checksum:1643886522,sum_checksum:1643886522"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:12.120566:CsvExampleGen:examples:0"
  }
}
custom_properties {
  key: "span"
  value {
    int_value: 0
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
, artifact_type: name: "Examples"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
properties {
  key: "version"
  value: INT
}
base_type: DATASET
)]}) for execution 1
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component CsvExampleGen is finished.
INFO:absl:Component StatisticsGen is running.
INFO:absl:Running launcher for node_info {
  type {
    name: "tfx.components.statistics_gen.component.StatisticsGen"
    base_type: PROCESS
  }
  id: "StatisticsGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:09:12.120566"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.StatisticsGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "examples"
    value {
      channels {
        producer_node_query {
          id: "CsvExampleGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:09:12.120566"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.CsvExampleGen"
            }
          }
        }
        artifact_query {
          type {
            name: "Examples"
            base_type: DATASET
          }
        }
        output_key: "examples"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "statistics"
    value {
      artifact_spec {
        type {
          name: "ExampleStatistics"
          properties {
            key: "span"
            value: INT
          }
          properties {
            key: "split_names"
            value: STRING
          }
          base_type: STATISTICS
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
}
upstream_nodes: "CsvExampleGen"
downstream_nodes: "SchemaGen"
execution_options {
  caching_options {
  }
}

INFO:absl:MetadataStore with DB connection initialized
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Going to run a new execution 2
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=2, input_dict={'examples': [Artifact(artifact: id: 1
type_id: 15
uri: "./tfx_pipeline_output/my_pipeline/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:1643886522,sum_checksum:1643886522"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:12.120566: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: "1.6.0"
  }
}
state: LIVE
create_time_since_epoch: 1643886553302
last_update_time_since_epoch: 1643886553302
, artifact_type: id: 15
name: "Examples"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
properties {
  key: "version"
  value: INT
}
base_type: DATASET
)]}, output_dict=defaultdict(<class 'list'>, {'statistics': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/2"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:12.120566:StatisticsGen:statistics:0"
  }
}
, artifact_type: name: "ExampleStatistics"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
base_type: STATISTICS
)]}), exec_properties={'exclude_splits': '[]'}, execution_output_uri='./tfx_pipeline_output/my_pipeline/StatisticsGen/.system/executor_execution/2/executor_output.pb', stateful_working_dir='./tfx_pipeline_output/my_pipeline/StatisticsGen/.system/stateful_working_dir/2022-02-03T11:09:12.120566', tmp_dir='./tfx_pipeline_output/my_pipeline/StatisticsGen/.system/executor_execution/2/.temp/', pipeline_node=node_info {
  type {
    name: "tfx.components.statistics_gen.component.StatisticsGen"
    base_type: PROCESS
  }
  id: "StatisticsGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:09:12.120566"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.StatisticsGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "examples"
    value {
      channels {
        producer_node_query {
          id: "CsvExampleGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:09:12.120566"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.CsvExampleGen"
            }
          }
        }
        artifact_query {
          type {
            name: "Examples"
            base_type: DATASET
          }
        }
        output_key: "examples"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "statistics"
    value {
      artifact_spec {
        type {
          name: "ExampleStatistics"
          properties {
            key: "span"
            value: INT
          }
          properties {
            key: "split_names"
            value: STRING
          }
          base_type: STATISTICS
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
}
upstream_nodes: "CsvExampleGen"
downstream_nodes: "SchemaGen"
execution_options {
  caching_options {
  }
}
, pipeline_info=id: "my_pipeline"
, pipeline_run_id='2022-02-03T11:09:12.120566')
INFO:absl:Generating statistics for split train.
INFO:absl:Statistics for split train written to ./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/2/Split-train.
INFO:absl:Generating statistics for split eval.
INFO:absl:Statistics for split eval written to ./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/2/Split-eval.
WARNING:root:Make sure that locally built Python SDK docker image has Python 3.7 interpreter.
INFO:absl:Cleaning up stateless execution info.
INFO:absl:Execution 2 succeeded.
INFO:absl:Cleaning up stateful execution info.
INFO:absl:Publishing output artifacts defaultdict(<class 'list'>, {'statistics': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/2"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:12.120566:StatisticsGen:statistics:0"
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
, artifact_type: name: "ExampleStatistics"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
base_type: STATISTICS
)]}) for execution 2
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component StatisticsGen is finished.
INFO:absl:Component SchemaGen is running.
INFO:absl:Running launcher for node_info {
  type {
    name: "tfx.components.schema_gen.component.SchemaGen"
    base_type: PROCESS
  }
  id: "SchemaGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:09:12.120566"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.SchemaGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "statistics"
    value {
      channels {
        producer_node_query {
          id: "StatisticsGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:09:12.120566"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.StatisticsGen"
            }
          }
        }
        artifact_query {
          type {
            name: "ExampleStatistics"
            base_type: STATISTICS
          }
        }
        output_key: "statistics"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "schema"
    value {
      artifact_spec {
        type {
          name: "Schema"
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
  parameters {
    key: "infer_feature_shape"
    value {
      field_value {
        int_value: 1
      }
    }
  }
}
upstream_nodes: "StatisticsGen"
execution_options {
  caching_options {
  }
}

INFO:absl:MetadataStore with DB connection initialized
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Going to run a new execution 3
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=3, input_dict={'statistics': [Artifact(artifact: id: 2
type_id: 17
uri: "./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/2"
properties {
  key: "split_names"
  value {
    string_value: "[\"train\", \"eval\"]"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:12.120566:StatisticsGen:statistics:0"
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
state: LIVE
create_time_since_epoch: 1643886556588
last_update_time_since_epoch: 1643886556588
, artifact_type: id: 17
name: "ExampleStatistics"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
base_type: STATISTICS
)]}, output_dict=defaultdict(<class 'list'>, {'schema': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/SchemaGen/schema/3"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:12.120566:SchemaGen:schema:0"
  }
}
, artifact_type: name: "Schema"
)]}), exec_properties={'infer_feature_shape': 1, 'exclude_splits': '[]'}, execution_output_uri='./tfx_pipeline_output/my_pipeline/SchemaGen/.system/executor_execution/3/executor_output.pb', stateful_working_dir='./tfx_pipeline_output/my_pipeline/SchemaGen/.system/stateful_working_dir/2022-02-03T11:09:12.120566', tmp_dir='./tfx_pipeline_output/my_pipeline/SchemaGen/.system/executor_execution/3/.temp/', pipeline_node=node_info {
  type {
    name: "tfx.components.schema_gen.component.SchemaGen"
    base_type: PROCESS
  }
  id: "SchemaGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:09:12.120566"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.SchemaGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "statistics"
    value {
      channels {
        producer_node_query {
          id: "StatisticsGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:09:12.120566"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.StatisticsGen"
            }
          }
        }
        artifact_query {
          type {
            name: "ExampleStatistics"
            base_type: STATISTICS
          }
        }
        output_key: "statistics"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "schema"
    value {
      artifact_spec {
        type {
          name: "Schema"
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
  parameters {
    key: "infer_feature_shape"
    value {
      field_value {
        int_value: 1
      }
    }
  }
}
upstream_nodes: "StatisticsGen"
execution_options {
  caching_options {
  }
}
, pipeline_info=id: "my_pipeline"
, pipeline_run_id='2022-02-03T11:09:12.120566')
INFO:absl:Processing schema from statistics for split train.
INFO:absl:Processing schema from statistics for split eval.
INFO:absl:Schema written to ./tfx_pipeline_output/my_pipeline/SchemaGen/schema/3/schema.pbtxt.
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'>, {'schema': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/SchemaGen/schema/3"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:12.120566:SchemaGen:schema:0"
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
, artifact_type: name: "Schema"
)]}) for execution 3
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component SchemaGen is finished.

"Component ExampleValidator tamamlandı" ifadesini görmelisiniz. boru hattı başarıyla çalıştırılırsa.

Boru hattının çıktısını inceleyin.

TFX ardışık düzen iki tür çıktı üretir, yapıtlar ve yapıtların meta verilerini ve ardışık düzen yürütmelerini içeren bir meta veri DB(MLMD) . Çıktının konumu local_runner.py içinde tanımlanır. Varsayılan olarak, tfx_pipeline_output dizini altında depolanır ve meta veriler tfx_metadata dizini altında bir sqlite veritabanı olarak depolanır.

Bu çıktıları incelemek için MLMD API'lerini kullanabilirsiniz. İlk olarak, henüz üretilmiş çıktı yapılarını aramak için bazı yardımcı fonksiyonlar tanımlayacağız.

import tensorflow as tf
import tfx
from ml_metadata import errors
from ml_metadata.proto import metadata_store_pb2
from tfx.types import artifact_utils

# TODO(b/171447278): Move these functions into TFX library.

def get_latest_executions(store, pipeline_name, component_id = None):
  """Fetch all pipeline runs."""
  if component_id is None:  # Find entire pipeline runs.
    run_contexts = [
        c for c in store.get_contexts_by_type('run')
        if c.properties['pipeline_name'].string_value == pipeline_name
    ]
  else:  # Find specific component runs.
    run_contexts = [
        c for c in store.get_contexts_by_type('component_run')
        if c.properties['pipeline_name'].string_value == pipeline_name and
           c.properties['component_id'].string_value == component_id
    ]
  if not run_contexts:
    return []
  # Pick the latest run context.
  latest_context = max(run_contexts,
                       key=lambda c: c.last_update_time_since_epoch)
  return store.get_executions_by_context(latest_context.id)

def get_latest_artifacts(store, pipeline_name, component_id = None):
  """Fetch all artifacts from latest pipeline execution."""
  executions = get_latest_executions(store, pipeline_name, component_id)

  # Fetch all artifacts produced from the given executions.
  execution_ids = [e.id for e in executions]
  events = store.get_events_by_execution_ids(execution_ids)
  artifact_ids = [
      event.artifact_id for event in events
      if event.type == metadata_store_pb2.Event.OUTPUT
  ]
  return store.get_artifacts_by_id(artifact_ids)

def find_latest_artifacts_by_type(store, artifacts, artifact_type):
  """Get the latest artifacts of a specified type."""
  # Get type information from MLMD
  try:
    artifact_type = store.get_artifact_type(artifact_type)
  except errors.NotFoundError:
    return []
  # Filter artifacts with type.
  filtered_artifacts = [aritfact for aritfact in artifacts
                        if aritfact.type_id == artifact_type.id]
  # Convert MLMD artifact data into TFX Artifact instances.
  return [artifact_utils.deserialize_artifact(artifact_type, artifact)
      for artifact in filtered_artifacts]


from tfx.orchestration.experimental.interactive import visualizations

def visualize_artifacts(artifacts):
  """Visualizes artifacts using standard visualization modules."""
  for artifact in artifacts:
    visualization = visualizations.get_registry().get_visualization(
        artifact.type_name)
    if visualization:
      visualization.display(artifact)

from tfx.orchestration.experimental.interactive import standard_visualizations
standard_visualizations.register_standard_visualizations()

import pprint

from tfx.orchestration import metadata
from tfx.types import artifact_utils
from tfx.types import standard_artifacts

def preview_examples(artifacts):
  """Preview a few records from Examples artifacts."""
  pp = pprint.PrettyPrinter()
  for artifact in artifacts:
    print("==== Examples artifact:{}({})".format(artifact.name, artifact.uri))
    for split in artifact_utils.decode_split_names(artifact.split_names):
      print("==== Reading from split:{}".format(split))
      split_uri = artifact_utils.get_split_uri([artifact], split)

      # Get the list of files in this directory (all compressed TFRecord files)
      tfrecord_filenames = [os.path.join(split_uri, name)
                            for name in os.listdir(split_uri)]
      # Create a `TFRecordDataset` to read these files
      dataset = tf.data.TFRecordDataset(tfrecord_filenames,
                                        compression_type="GZIP")
      # Iterate over the first 2 records and decode them.
      for tfrecord in dataset.take(2):
        serialized_example = tfrecord.numpy()
        example = tf.train.Example()
        example.ParseFromString(serialized_example)
        pp.pprint(example)

import local_runner

metadata_connection_config = metadata.sqlite_metadata_connection_config(
              local_runner.METADATA_PATH)

Artık çıktı yapılarının meta verilerini MLMD'den okuyabiliyoruz.

with metadata.Metadata(metadata_connection_config) as metadata_handler:
    # Search all aritfacts from the previous pipeline run.
    artifacts = get_latest_artifacts(metadata_handler.store, PIPELINE_NAME)
    # Find artifacts of Examples type.
    examples_artifacts = find_latest_artifacts_by_type(
        metadata_handler.store, artifacts,
        standard_artifacts.Examples.TYPE_NAME)
    # Find artifacts generated from StatisticsGen.
    stats_artifacts = find_latest_artifacts_by_type(
        metadata_handler.store, artifacts,
        standard_artifacts.ExampleStatistics.TYPE_NAME)
    # Find artifacts generated from SchemaGen.
    schema_artifacts = find_latest_artifacts_by_type(
        metadata_handler.store, artifacts,
        standard_artifacts.Schema.TYPE_NAME)
    # Find artifacts generated from ExampleValidator.
    anomalies_artifacts = find_latest_artifacts_by_type(
        metadata_handler.store, artifacts,
        standard_artifacts.ExampleAnomalies.TYPE_NAME)

Artık her bir bileşenin çıktılarını inceleyebiliriz. Tensorflow Data Validation (TFDV) StatisticsGen , SchemaGen ve ExampleValidator kullanılır ve TFDV bu bileşenlerden gelen çıktıları görselleştirmek için kullanılabilir.

Bu eğitimde, görselleştirmeyi göstermek için TFDV'yi dahili olarak kullanan TFX'te görselleştirme yardımcı yöntemlerini kullanacağız. Her bir bileşen hakkında daha fazla bilgi edinmek için lütfen TFX bileşenleri eğitimine bakın.

Çıktı formunu inceleyin ExampleGen

ExampleGen'in çıktısını inceleyelim. Her bölme için ilk iki örneğe bir göz atın:

preview_examples(examples_artifacts)

Varsayılan olarak, TFX ExampleGen örnekleri tren ve eval olmak üzere iki bölüme ayırır, ancak bölme yapılandırmanızı ayarlayabilirsiniz .

StatisticsGen'den çıktıyı inceleyin

visualize_artifacts(stats_artifacts)

Bu istatistikler, otomatik olarak bir veri şeması oluşturmak için SchemaGen'e sağlanır.

SchemaGen'den çıktıyı inceleyin

visualize_artifacts(schema_artifacts)

Bu şema, StatisticsGen'in çıktısından otomatik olarak çıkarılır. Bu eğitimde oluşturulan şemayı kullanacağız, ancak şemayı değiştirebilir ve özelleştirebilirsiniz .

ExampleValidator'dan çıktıyı inceleyin

visualize_artifacts(anomalies_artifacts)

Herhangi bir anormallik bulunursa, tüm örneklerin varsayımlarınızı takip ettiği verilerinizi gözden geçirebilirsiniz. StatistcsGen gibi diğer bileşenlerin çıktıları faydalı olabilir. Bulunan anormallikler, işlem hattının yürütülmesini engellemez.

Kullanılabilir özellikleri SchemaGen çıktılarından görebilirsiniz. Özellikleriniz doğrudan Trainer ML modeli oluşturmak için kullanılabilirse, bir sonraki adımı atlayabilir ve 4. Adıma geçebilirsiniz. Aksi takdirde, bir sonraki adımda bazı özellik mühendisliği çalışmaları yapabilirsiniz. Ortalamaları hesaplamak gibi tam geçişli işlemler gerektiğinde, özellikle ölçeklendirmeniz gerektiğinde, Transform bileşeni gereklidir.

Adım 3. (İsteğe bağlı) Transform bileşeniyle özellik mühendisliği.

Bu adımda, boru hattında Transform bileşeni tarafından kullanılacak çeşitli özellik mühendisliği işlerini tanımlayacaksınız. Daha fazla bilgi için Dönüştürme bileşeni kılavuzuna bakın.

Bu, yalnızca eğitim kodunuz, ExampleGen çıktısında bulunmayan ek özellik(ler) gerektiriyorsa gereklidir. Aksi takdirde, Trainer'ı kullanmanın bir sonraki adımına geçmekten çekinmeyin.

Modelin özelliklerini tanımlayın

models/features.py , özellik adları, kelime dağarcığının boyutu vb. dahil olmak üzere model için özellikleri tanımlamak için sabitler içerir. penguin modelimiz denetimli öğrenmeyi kullanarak bir sınıflandırma problemini çözdüğünden ve tüm özellikler sürekli sayısal özellikler olduğundan, varsayılan olarak penguin şablonunun iki sabit değeri vardır, FEATURE_KEYS ve LABEL_KEY . Başka bir örnek için Chicago taksi örneğindeki özellik tanımlarına bakın.

eğitim/sunma için ön işlemeyi preprocessing_fn() içinde uygulayın.

Gerçek özellik mühendisliği, models/preprocessing.py içindeki preprocessing_fn() işlevinde gerçekleşir.

preprocessing_fn , tensörlerin çıkış diktini üretmek için tensörlerin giriş diktini manipüle eden bir dizi fonksiyon tanımlayabilirsiniz. TensorFlow Transform API'sinde scale_to_0_1 ve compute_and_apply_vocabulary gibi yardımcı işlevler vardır veya normal TensorFlow işlevlerini kullanabilirsiniz. Varsayılan olarak penguin şablonu, özellik değerlerini normalleştirmek için tft.scale_to_z_score işlevinin örnek kullanımlarını içerir.

preprocessing_fn yazma hakkında daha fazla bilgi için Tensflow Transform kılavuzuna bakın.

İşlem hattına Dönüştür bileşeni ekleyin.

preprocessing_fn'niz hazırsa, ardışık düzene Transform bileşenini ekleyin.

1. pipeline/pipeline.py dosyasında, bileşeni ardışık düzene eklemek için # components.append(transform) yorumunu kaldırın.

İşlem hattını güncelleyebilir ve yeniden çalıştırabilirsiniz.

!tfx pipeline update --engine=local --pipeline_path=local_runner.py \
 && tfx run create --engine=local --pipeline_name={PIPELINE_NAME}

tutucu22 l10n-yer

CLI
Updating pipeline
INFO:absl:Excluding no splits because exclude_splits is not set.
INFO:absl:Excluding no splits because exclude_splits is not set.
Pipeline "my_pipeline" updated successfully.
CLI
Creating a run for pipeline: my_pipeline
INFO:absl:Excluding no splits because exclude_splits is not set.
INFO:absl:Excluding no splits because exclude_splits is not set.
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: "SchemaGen"
  value {
    python_class_executable_spec {
      class_path: "tfx.components.schema_gen.executor.Executor"
    }
  }
}
executor_specs {
  key: "StatisticsGen"
  value {
    beam_executable_spec {
      python_executor_spec {
        class_path: "tfx.components.statistics_gen.executor.Executor"
      }
    }
  }
}
custom_driver_specs {
  key: "CsvExampleGen"
  value {
    python_class_executable_spec {
      class_path: "tfx.components.example_gen.driver.FileBasedDriver"
    }
  }
}
metadata_connection_config {
  database_connection_config {
    sqlite {
      filename_uri: "./tfx_metadata/my_pipeline/metadata.db"
      connection_mode: READWRITE_OPENCREATE
    }
  }
}

INFO:absl:Using connection config:
 sqlite {
  filename_uri: "./tfx_metadata/my_pipeline/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: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:09:37.055994"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.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
          }
          base_type: DATASET
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "input_base"
    value {
      field_value {
        string_value: "/home/kbuilder/imported/my_pipeline/data"
      }
    }
  }
  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: "StatisticsGen"
execution_options {
  caching_options {
  }
}

INFO:absl:MetadataStore with DB connection initialized
INFO:absl:select span and version = (0, None)
INFO:absl:latest span and version = (0, None)
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={}, output_dict=defaultdict(<class 'list'>, {'examples': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/CsvExampleGen/examples/4"
custom_properties {
  key: "input_fingerprint"
  value {
    string_value: "split:single_split,num_files:1,total_bytes:25648,xor_checksum:1643886522,sum_checksum:1643886522"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:37.055994: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
}
base_type: DATASET
)]}), exec_properties={'output_file_format': 5, 'input_base': '/home/kbuilder/imported/my_pipeline/data', '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}', 'output_data_format': 6, 'input_config': '{\n  "splits": [\n    {\n      "name": "single_split",\n      "pattern": "*"\n    }\n  ]\n}', 'span': 0, 'version': None, 'input_fingerprint': 'split:single_split,num_files:1,total_bytes:25648,xor_checksum:1643886522,sum_checksum:1643886522'}, execution_output_uri='./tfx_pipeline_output/my_pipeline/CsvExampleGen/.system/executor_execution/4/executor_output.pb', stateful_working_dir='./tfx_pipeline_output/my_pipeline/CsvExampleGen/.system/stateful_working_dir/2022-02-03T11:09:37.055994', tmp_dir='./tfx_pipeline_output/my_pipeline/CsvExampleGen/.system/executor_execution/4/.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: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:09:37.055994"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.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
          }
          base_type: DATASET
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "input_base"
    value {
      field_value {
        string_value: "/home/kbuilder/imported/my_pipeline/data"
      }
    }
  }
  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: "StatisticsGen"
execution_options {
  caching_options {
  }
}
, pipeline_info=id: "my_pipeline"
, pipeline_run_id='2022-02-03T11:09:37.055994')
INFO:absl:Generating examples.
INFO:absl:Processing input csv data /home/kbuilder/imported/my_pipeline/data/* to TFExample.
WARNING:root:Make sure that locally built Python SDK docker image has Python 3.7 interpreter.
E0203 11:09:37.596944686    5287 fork_posix.cc:70]           Fork support is only compatible with the epoll1 and poll polling strategies
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:Value type <class 'NoneType'> of key version in exec_properties is not supported, going to drop it
INFO:absl:Value type <class 'list'> of key _beam_pipeline_args in exec_properties is not supported, going to drop it
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'>, {'examples': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/CsvExampleGen/examples/4"
custom_properties {
  key: "input_fingerprint"
  value {
    string_value: "split:single_split,num_files:1,total_bytes:25648,xor_checksum:1643886522,sum_checksum:1643886522"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:37.055994:CsvExampleGen:examples:0"
  }
}
custom_properties {
  key: "span"
  value {
    int_value: 0
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
, artifact_type: name: "Examples"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
properties {
  key: "version"
  value: INT
}
base_type: DATASET
)]}) for execution 4
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component CsvExampleGen is finished.
INFO:absl:Component StatisticsGen is running.
INFO:absl:Running launcher for node_info {
  type {
    name: "tfx.components.statistics_gen.component.StatisticsGen"
    base_type: PROCESS
  }
  id: "StatisticsGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:09:37.055994"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.StatisticsGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "examples"
    value {
      channels {
        producer_node_query {
          id: "CsvExampleGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:09:37.055994"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.CsvExampleGen"
            }
          }
        }
        artifact_query {
          type {
            name: "Examples"
            base_type: DATASET
          }
        }
        output_key: "examples"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "statistics"
    value {
      artifact_spec {
        type {
          name: "ExampleStatistics"
          properties {
            key: "span"
            value: INT
          }
          properties {
            key: "split_names"
            value: STRING
          }
          base_type: STATISTICS
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
}
upstream_nodes: "CsvExampleGen"
downstream_nodes: "SchemaGen"
execution_options {
  caching_options {
  }
}

INFO:absl:MetadataStore with DB connection initialized
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={'examples': [Artifact(artifact: id: 4
type_id: 15
uri: "./tfx_pipeline_output/my_pipeline/CsvExampleGen/examples/4"
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:1643886522,sum_checksum:1643886522"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:37.055994: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: "1.6.0"
  }
}
state: LIVE
create_time_since_epoch: 1643886578210
last_update_time_since_epoch: 1643886578210
, artifact_type: id: 15
name: "Examples"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
properties {
  key: "version"
  value: INT
}
base_type: DATASET
)]}, output_dict=defaultdict(<class 'list'>, {'statistics': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/5"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:37.055994:StatisticsGen:statistics:0"
  }
}
, artifact_type: name: "ExampleStatistics"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
base_type: STATISTICS
)]}), exec_properties={'exclude_splits': '[]'}, execution_output_uri='./tfx_pipeline_output/my_pipeline/StatisticsGen/.system/executor_execution/5/executor_output.pb', stateful_working_dir='./tfx_pipeline_output/my_pipeline/StatisticsGen/.system/stateful_working_dir/2022-02-03T11:09:37.055994', tmp_dir='./tfx_pipeline_output/my_pipeline/StatisticsGen/.system/executor_execution/5/.temp/', pipeline_node=node_info {
  type {
    name: "tfx.components.statistics_gen.component.StatisticsGen"
    base_type: PROCESS
  }
  id: "StatisticsGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:09:37.055994"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.StatisticsGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "examples"
    value {
      channels {
        producer_node_query {
          id: "CsvExampleGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:09:37.055994"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.CsvExampleGen"
            }
          }
        }
        artifact_query {
          type {
            name: "Examples"
            base_type: DATASET
          }
        }
        output_key: "examples"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "statistics"
    value {
      artifact_spec {
        type {
          name: "ExampleStatistics"
          properties {
            key: "span"
            value: INT
          }
          properties {
            key: "split_names"
            value: STRING
          }
          base_type: STATISTICS
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
}
upstream_nodes: "CsvExampleGen"
downstream_nodes: "SchemaGen"
execution_options {
  caching_options {
  }
}
, pipeline_info=id: "my_pipeline"
, pipeline_run_id='2022-02-03T11:09:37.055994')
INFO:absl:Generating statistics for split train.
INFO:absl:Statistics for split train written to ./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/5/Split-train.
INFO:absl:Generating statistics for split eval.
INFO:absl:Statistics for split eval written to ./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/5/Split-eval.
WARNING:root:Make sure that locally built Python SDK docker image has Python 3.7 interpreter.
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'>, {'statistics': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/5"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:37.055994:StatisticsGen:statistics:0"
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
, artifact_type: name: "ExampleStatistics"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
base_type: STATISTICS
)]}) for execution 5
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component StatisticsGen is finished.
INFO:absl:Component SchemaGen is running.
INFO:absl:Running launcher for node_info {
  type {
    name: "tfx.components.schema_gen.component.SchemaGen"
    base_type: PROCESS
  }
  id: "SchemaGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:09:37.055994"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.SchemaGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "statistics"
    value {
      channels {
        producer_node_query {
          id: "StatisticsGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:09:37.055994"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.StatisticsGen"
            }
          }
        }
        artifact_query {
          type {
            name: "ExampleStatistics"
            base_type: STATISTICS
          }
        }
        output_key: "statistics"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "schema"
    value {
      artifact_spec {
        type {
          name: "Schema"
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
  parameters {
    key: "infer_feature_shape"
    value {
      field_value {
        int_value: 1
      }
    }
  }
}
upstream_nodes: "StatisticsGen"
execution_options {
  caching_options {
  }
}

INFO:absl:MetadataStore with DB connection initialized
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Going to run a new execution 6
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=6, input_dict={'statistics': [Artifact(artifact: id: 5
type_id: 17
uri: "./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/5"
properties {
  key: "split_names"
  value {
    string_value: "[\"train\", \"eval\"]"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:37.055994:StatisticsGen:statistics:0"
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
state: LIVE
create_time_since_epoch: 1643886581527
last_update_time_since_epoch: 1643886581527
, artifact_type: id: 17
name: "ExampleStatistics"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
base_type: STATISTICS
)]}, output_dict=defaultdict(<class 'list'>, {'schema': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/SchemaGen/schema/6"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:37.055994:SchemaGen:schema:0"
  }
}
, artifact_type: name: "Schema"
)]}), exec_properties={'exclude_splits': '[]', 'infer_feature_shape': 1}, execution_output_uri='./tfx_pipeline_output/my_pipeline/SchemaGen/.system/executor_execution/6/executor_output.pb', stateful_working_dir='./tfx_pipeline_output/my_pipeline/SchemaGen/.system/stateful_working_dir/2022-02-03T11:09:37.055994', tmp_dir='./tfx_pipeline_output/my_pipeline/SchemaGen/.system/executor_execution/6/.temp/', pipeline_node=node_info {
  type {
    name: "tfx.components.schema_gen.component.SchemaGen"
    base_type: PROCESS
  }
  id: "SchemaGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:09:37.055994"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.SchemaGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "statistics"
    value {
      channels {
        producer_node_query {
          id: "StatisticsGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:09:37.055994"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.StatisticsGen"
            }
          }
        }
        artifact_query {
          type {
            name: "ExampleStatistics"
            base_type: STATISTICS
          }
        }
        output_key: "statistics"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "schema"
    value {
      artifact_spec {
        type {
          name: "Schema"
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
  parameters {
    key: "infer_feature_shape"
    value {
      field_value {
        int_value: 1
      }
    }
  }
}
upstream_nodes: "StatisticsGen"
execution_options {
  caching_options {
  }
}
, pipeline_info=id: "my_pipeline"
, pipeline_run_id='2022-02-03T11:09:37.055994')
INFO:absl:Processing schema from statistics for split train.
INFO:absl:Processing schema from statistics for split eval.
INFO:absl:Schema written to ./tfx_pipeline_output/my_pipeline/SchemaGen/schema/6/schema.pbtxt.
INFO:absl:Cleaning up stateless execution info.
INFO:absl:Execution 6 succeeded.
INFO:absl:Cleaning up stateful execution info.
INFO:absl:Publishing output artifacts defaultdict(<class 'list'>, {'schema': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/SchemaGen/schema/6"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:09:37.055994:SchemaGen:schema:0"
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
, artifact_type: name: "Schema"
)]}) for execution 6
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component SchemaGen is finished.

İşlem hattı başarılı bir şekilde çalıştıysa, "Bileşen Dönüşümü tamamlandı" ifadesini görmeniz gerekir. günlüğünde bir yerde . Transform bileşeni ve ExampleValidator bileşeni birbirine bağımlı olmadığından yürütme sırası sabit değildir. Bununla birlikte, Transform ve ExampleValidator öğelerinden biri, işlem hattı yürütmesindeki son bileşen olabilir.

Transform'dan çıktıyı inceleyin

Dönüştürme bileşeni, iki tür çıktı oluşturur, bir Tensorflow grafiği ve dönüştürülmüş örnekler. Dönüştürülen örnekler, yine ExampleGen tarafından üretilen Örnekler yapıt türüdür, ancak bu, bunun yerine dönüştürülmüş özellik değerlerini içerir.

Bunları bir önceki adımda yaptığımız gibi inceleyebilirsiniz.

with metadata.Metadata(metadata_connection_config) as metadata_handler:
    # Search all aritfacts from the previous run of Transform component.
    artifacts = get_latest_artifacts(metadata_handler.store,
                                     PIPELINE_NAME, "Transform")
    # Find artifacts of Examples type.
    transformed_examples_artifacts = find_latest_artifacts_by_type(
        metadata_handler.store, artifacts,
        standard_artifacts.Examples.TYPE_NAME)

tutucu24 l10n-yer

preview_examples(transformed_examples_artifacts)

Adım 4. Modelinizi Trainer bileşeniyle eğitin.

Trainer bileşenini kullanarak bir ML modeli oluşturacağız. Daha fazla bilgi için Eğitmen bileşeni kılavuzuna bakın. Model kodunuzu Trainer bileşenine sağlamanız gerekir.

Modelinizi tanımlayın.

Penguen şablonunda, Trainer bileşeni için models.model.run_fn argümanı olarak run_fn kullanılır. Bu, Trainer bileşeni çalıştığında models/model.py içindeki run_fn() işlevinin çağrılacağı anlamına gelir. Verilen kodda keras API kullanarak basit bir DNN modeli oluşturmak için kodu görebilirsiniz. TFX'te keras API'sini kullanma hakkında daha fazla bilgi için TFX kılavuzundaki TensorFlow 2.x'e bakın.

Bu run_fn , bir model oluşturmalı ve onu bileşen tarafından belirtilen fn_args.serving_model_dir ile gösterilen bir dizine kaydetmelisiniz. fn_args iletilen run_fn içindeki diğer bağımsız değişkenleri kullanabilirsiniz. fn_args içindeki bağımsız değişkenlerin tam listesi için ilgili kodlara bakın.

Özelliklerinizi models/features.py içinde tanımlayın ve gerektiğinde kullanın. 3. Adımda özelliklerinizi dönüştürdüyseniz, dönüştürülen özellikleri modelinize girdi olarak kullanmalısınız.

İşlem hattına Trainer bileşeni ekleyin.

run_fn'niz hazırsa, ardışık düzene Trainer bileşenini ekleyin.

1. pipeline/pipeline.py dosyasında, bileşeni ardışık düzene eklemek için # components.append(trainer) yorumunu kaldırın.

Eğitmen bileşeni için argümanlar, Dönüştürme bileşenini kullanıp kullanmadığınıza bağlı olabilir.

• Transform bileşenini KULLANMAZSANIZ , bağımsız değişkenleri değiştirmeniz gerekmez.

• Transform bileşenini kullanırsanız, bir Trainer bileşeni örneği oluştururken bağımsız değişkenleri değiştirmeniz gerekir.

  • examples bağımsız değişkenini examples=transform.outputs['transformed_examples'], . Eğitim için dönüştürülmüş örnekleri kullanmamız gerekiyor.
  • transform_graph=transform.outputs['transform_graph'], gibi transform_graph bağımsız değişkeni ekleyin. Bu grafik, dönüştürme işlemleri için TensorFlow grafiğini içerir.
  • Yukarıdaki değişikliklerden sonra, Trainer bileşeni oluşturma kodu aşağıdaki gibi görünecektir.

  # If you use a Transform component.
  trainer = Trainer(
      run_fn=run_fn,
      examples=transform.outputs['transformed_examples'],
      transform_graph=transform.outputs['transform_graph'],
      schema=schema_gen.outputs['schema'],
      ...
  

İşlem hattını güncelleyebilir ve yeniden çalıştırabilirsiniz.

!tfx pipeline update --engine=local --pipeline_path=local_runner.py \
 && tfx run create --engine=local --pipeline_name={PIPELINE_NAME}

tutucu27 l10n-yer

CLI
Updating pipeline
INFO:absl:Excluding no splits because exclude_splits is not set.
INFO:absl:Excluding no splits because exclude_splits is not set.
Pipeline "my_pipeline" updated successfully.
CLI
Creating a run for pipeline: my_pipeline
INFO:absl:Excluding no splits because exclude_splits is not set.
INFO:absl:Excluding no splits because exclude_splits is not set.
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: "SchemaGen"
  value {
    python_class_executable_spec {
      class_path: "tfx.components.schema_gen.executor.Executor"
    }
  }
}
executor_specs {
  key: "StatisticsGen"
  value {
    beam_executable_spec {
      python_executor_spec {
        class_path: "tfx.components.statistics_gen.executor.Executor"
      }
    }
  }
}
custom_driver_specs {
  key: "CsvExampleGen"
  value {
    python_class_executable_spec {
      class_path: "tfx.components.example_gen.driver.FileBasedDriver"
    }
  }
}
metadata_connection_config {
  database_connection_config {
    sqlite {
      filename_uri: "./tfx_metadata/my_pipeline/metadata.db"
      connection_mode: READWRITE_OPENCREATE
    }
  }
}

INFO:absl:Using connection config:
 sqlite {
  filename_uri: "./tfx_metadata/my_pipeline/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: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:00.469382"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.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
          }
          base_type: DATASET
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "input_base"
    value {
      field_value {
        string_value: "/home/kbuilder/imported/my_pipeline/data"
      }
    }
  }
  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: "StatisticsGen"
execution_options {
  caching_options {
  }
}

INFO:absl:MetadataStore with DB connection initialized
INFO:absl:select span and version = (0, None)
INFO:absl:latest span and version = (0, None)
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Going to run a new execution 7
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=7, input_dict={}, output_dict=defaultdict(<class 'list'>, {'examples': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/CsvExampleGen/examples/7"
custom_properties {
  key: "input_fingerprint"
  value {
    string_value: "split:single_split,num_files:1,total_bytes:25648,xor_checksum:1643886522,sum_checksum:1643886522"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:00.469382: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
}
base_type: DATASET
)]}), exec_properties={'input_base': '/home/kbuilder/imported/my_pipeline/data', '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}', 'output_data_format': 6, 'output_file_format': 5, 'input_config': '{\n  "splits": [\n    {\n      "name": "single_split",\n      "pattern": "*"\n    }\n  ]\n}', 'span': 0, 'version': None, 'input_fingerprint': 'split:single_split,num_files:1,total_bytes:25648,xor_checksum:1643886522,sum_checksum:1643886522'}, execution_output_uri='./tfx_pipeline_output/my_pipeline/CsvExampleGen/.system/executor_execution/7/executor_output.pb', stateful_working_dir='./tfx_pipeline_output/my_pipeline/CsvExampleGen/.system/stateful_working_dir/2022-02-03T11:10:00.469382', tmp_dir='./tfx_pipeline_output/my_pipeline/CsvExampleGen/.system/executor_execution/7/.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: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:00.469382"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.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
          }
          base_type: DATASET
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "input_base"
    value {
      field_value {
        string_value: "/home/kbuilder/imported/my_pipeline/data"
      }
    }
  }
  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: "StatisticsGen"
execution_options {
  caching_options {
  }
}
, pipeline_info=id: "my_pipeline"
, pipeline_run_id='2022-02-03T11:10:00.469382')
INFO:absl:Generating examples.
INFO:absl:Processing input csv data /home/kbuilder/imported/my_pipeline/data/* to TFExample.
WARNING:root:Make sure that locally built Python SDK docker image has Python 3.7 interpreter.
E0203 11:10:01.173700221    5436 fork_posix.cc:70]           Fork support is only compatible with the epoll1 and poll polling strategies
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:Value type <class 'NoneType'> of key version in exec_properties is not supported, going to drop it
INFO:absl:Value type <class 'list'> of key _beam_pipeline_args in exec_properties is not supported, going to drop it
INFO:absl:Cleaning up stateless execution info.
INFO:absl:Execution 7 succeeded.
INFO:absl:Cleaning up stateful execution info.
INFO:absl:Publishing output artifacts defaultdict(<class 'list'>, {'examples': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/CsvExampleGen/examples/7"
custom_properties {
  key: "input_fingerprint"
  value {
    string_value: "split:single_split,num_files:1,total_bytes:25648,xor_checksum:1643886522,sum_checksum:1643886522"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:00.469382:CsvExampleGen:examples:0"
  }
}
custom_properties {
  key: "span"
  value {
    int_value: 0
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
, artifact_type: name: "Examples"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
properties {
  key: "version"
  value: INT
}
base_type: DATASET
)]}) for execution 7
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component CsvExampleGen is finished.
INFO:absl:Component StatisticsGen is running.
INFO:absl:Running launcher for node_info {
  type {
    name: "tfx.components.statistics_gen.component.StatisticsGen"
    base_type: PROCESS
  }
  id: "StatisticsGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:00.469382"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.StatisticsGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "examples"
    value {
      channels {
        producer_node_query {
          id: "CsvExampleGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:10:00.469382"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.CsvExampleGen"
            }
          }
        }
        artifact_query {
          type {
            name: "Examples"
            base_type: DATASET
          }
        }
        output_key: "examples"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "statistics"
    value {
      artifact_spec {
        type {
          name: "ExampleStatistics"
          properties {
            key: "span"
            value: INT
          }
          properties {
            key: "split_names"
            value: STRING
          }
          base_type: STATISTICS
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
}
upstream_nodes: "CsvExampleGen"
downstream_nodes: "SchemaGen"
execution_options {
  caching_options {
  }
}

INFO:absl:MetadataStore with DB connection initialized
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Going to run a new execution 8
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=8, input_dict={'examples': [Artifact(artifact: id: 7
type_id: 15
uri: "./tfx_pipeline_output/my_pipeline/CsvExampleGen/examples/7"
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:1643886522,sum_checksum:1643886522"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:00.469382: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: "1.6.0"
  }
}
state: LIVE
create_time_since_epoch: 1643886601629
last_update_time_since_epoch: 1643886601629
, artifact_type: id: 15
name: "Examples"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
properties {
  key: "version"
  value: INT
}
base_type: DATASET
)]}, output_dict=defaultdict(<class 'list'>, {'statistics': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/8"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:00.469382:StatisticsGen:statistics:0"
  }
}
, artifact_type: name: "ExampleStatistics"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
base_type: STATISTICS
)]}), exec_properties={'exclude_splits': '[]'}, execution_output_uri='./tfx_pipeline_output/my_pipeline/StatisticsGen/.system/executor_execution/8/executor_output.pb', stateful_working_dir='./tfx_pipeline_output/my_pipeline/StatisticsGen/.system/stateful_working_dir/2022-02-03T11:10:00.469382', tmp_dir='./tfx_pipeline_output/my_pipeline/StatisticsGen/.system/executor_execution/8/.temp/', pipeline_node=node_info {
  type {
    name: "tfx.components.statistics_gen.component.StatisticsGen"
    base_type: PROCESS
  }
  id: "StatisticsGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:00.469382"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.StatisticsGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "examples"
    value {
      channels {
        producer_node_query {
          id: "CsvExampleGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:10:00.469382"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.CsvExampleGen"
            }
          }
        }
        artifact_query {
          type {
            name: "Examples"
            base_type: DATASET
          }
        }
        output_key: "examples"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "statistics"
    value {
      artifact_spec {
        type {
          name: "ExampleStatistics"
          properties {
            key: "span"
            value: INT
          }
          properties {
            key: "split_names"
            value: STRING
          }
          base_type: STATISTICS
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
}
upstream_nodes: "CsvExampleGen"
downstream_nodes: "SchemaGen"
execution_options {
  caching_options {
  }
}
, pipeline_info=id: "my_pipeline"
, pipeline_run_id='2022-02-03T11:10:00.469382')
INFO:absl:Generating statistics for split train.
INFO:absl:Statistics for split train written to ./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/8/Split-train.
INFO:absl:Generating statistics for split eval.
INFO:absl:Statistics for split eval written to ./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/8/Split-eval.
WARNING:root:Make sure that locally built Python SDK docker image has Python 3.7 interpreter.
INFO:absl:Cleaning up stateless execution info.
INFO:absl:Execution 8 succeeded.
INFO:absl:Cleaning up stateful execution info.
INFO:absl:Publishing output artifacts defaultdict(<class 'list'>, {'statistics': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/8"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:00.469382:StatisticsGen:statistics:0"
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
, artifact_type: name: "ExampleStatistics"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
base_type: STATISTICS
)]}) for execution 8
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component StatisticsGen is finished.
INFO:absl:Component SchemaGen is running.
INFO:absl:Running launcher for node_info {
  type {
    name: "tfx.components.schema_gen.component.SchemaGen"
    base_type: PROCESS
  }
  id: "SchemaGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:00.469382"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.SchemaGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "statistics"
    value {
      channels {
        producer_node_query {
          id: "StatisticsGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:10:00.469382"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.StatisticsGen"
            }
          }
        }
        artifact_query {
          type {
            name: "ExampleStatistics"
            base_type: STATISTICS
          }
        }
        output_key: "statistics"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "schema"
    value {
      artifact_spec {
        type {
          name: "Schema"
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
  parameters {
    key: "infer_feature_shape"
    value {
      field_value {
        int_value: 1
      }
    }
  }
}
upstream_nodes: "StatisticsGen"
execution_options {
  caching_options {
  }
}

INFO:absl:MetadataStore with DB connection initialized
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Going to run a new execution 9
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=9, input_dict={'statistics': [Artifact(artifact: id: 8
type_id: 17
uri: "./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/8"
properties {
  key: "split_names"
  value {
    string_value: "[\"train\", \"eval\"]"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:00.469382:StatisticsGen:statistics:0"
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
state: LIVE
create_time_since_epoch: 1643886605023
last_update_time_since_epoch: 1643886605023
, artifact_type: id: 17
name: "ExampleStatistics"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
base_type: STATISTICS
)]}, output_dict=defaultdict(<class 'list'>, {'schema': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/SchemaGen/schema/9"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:00.469382:SchemaGen:schema:0"
  }
}
, artifact_type: name: "Schema"
)]}), exec_properties={'exclude_splits': '[]', 'infer_feature_shape': 1}, execution_output_uri='./tfx_pipeline_output/my_pipeline/SchemaGen/.system/executor_execution/9/executor_output.pb', stateful_working_dir='./tfx_pipeline_output/my_pipeline/SchemaGen/.system/stateful_working_dir/2022-02-03T11:10:00.469382', tmp_dir='./tfx_pipeline_output/my_pipeline/SchemaGen/.system/executor_execution/9/.temp/', pipeline_node=node_info {
  type {
    name: "tfx.components.schema_gen.component.SchemaGen"
    base_type: PROCESS
  }
  id: "SchemaGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:00.469382"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.SchemaGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "statistics"
    value {
      channels {
        producer_node_query {
          id: "StatisticsGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:10:00.469382"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.StatisticsGen"
            }
          }
        }
        artifact_query {
          type {
            name: "ExampleStatistics"
            base_type: STATISTICS
          }
        }
        output_key: "statistics"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "schema"
    value {
      artifact_spec {
        type {
          name: "Schema"
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
  parameters {
    key: "infer_feature_shape"
    value {
      field_value {
        int_value: 1
      }
    }
  }
}
upstream_nodes: "StatisticsGen"
execution_options {
  caching_options {
  }
}
, pipeline_info=id: "my_pipeline"
, pipeline_run_id='2022-02-03T11:10:00.469382')
INFO:absl:Processing schema from statistics for split train.
INFO:absl:Processing schema from statistics for split eval.
INFO:absl:Schema written to ./tfx_pipeline_output/my_pipeline/SchemaGen/schema/9/schema.pbtxt.
INFO:absl:Cleaning up stateless execution info.
INFO:absl:Execution 9 succeeded.
INFO:absl:Cleaning up stateful execution info.
INFO:absl:Publishing output artifacts defaultdict(<class 'list'>, {'schema': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/SchemaGen/schema/9"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:00.469382:SchemaGen:schema:0"
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
, artifact_type: name: "Schema"
)]}) for execution 9
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component SchemaGen is finished.

Bu yürütme başarılı bir şekilde çalıştığında, artık modeliniz için ilk TFX işlem hattınızı oluşturmuş ve çalıştırmış olursunuz. Tebrikler!

Yeni modeliniz çıktı dizini altında bir yerde bulunacaktır, ancak birçok ara sonucu içeren TFX boru hattının dışında sabit bir yerde veya hizmette bir modele sahip olmak daha iyi olacaktır. Makine öğrenimi üretim sistemlerinde kritik olan yerleşik modelin sürekli değerlendirilmesiyle daha da iyi. Bir sonraki adımda TFX'te sürekli değerlendirme ve dağıtımların nasıl çalıştığını göreceğiz.

Adım 5. (Opsiyonel) Modeli Evaluator ile değerlendirin ve pusher ile yayınlayın.

Evaluator bileşeni, Trainer oluşturulan her modeli sürekli olarak değerlendirir ve Pusher , modeli dosya sisteminde önceden tanımlanmış bir konuma ve hatta Google Cloud AI Platform Modellerine kopyalar.

İşlem hattına Değerlendirici bileşeni ekler.

pipeline/pipeline.py dosyasında:

1. İşlem hattına en son model çözümleyiciyi eklemek için # components.append(model_resolver) . Evaluator, bir modeli, son işlem hattı çalıştırmasında Evaluator'ı geçen eski temel modelle karşılaştırmak için kullanılabilir. En LatestBlessedModelResolver , Evaluator'ı geçen en son modeli bulur.
2. Modeliniz için uygun tfma.MetricsSpec ayarlayın. Değerlendirme her ML modeli için farklı olabilir. Penguen şablonunda, çok kategorili bir sınıflandırma problemini çözdüğümüz için SparseCategoricalAccuracy kullanıldı. Modelinizi belirli dilimler için analiz etmek için ayrıca tfma.SliceSpec belirtmeniz gerekir. Daha fazla ayrıntı için, Değerlendirici bileşen kılavuzuna bakın.
3. Bileşeni ardışık düzene eklemek için # components.append(evaluator) yorumunu kaldırın.

İşlem hattını güncelleyebilir ve yeniden çalıştırabilirsiniz.

# Update and run the pipeline.
!tfx pipeline update --engine=local --pipeline_path=local_runner.py \
 && tfx run create --engine=local --pipeline_name={PIPELINE_NAME}

tutucu29 l10n-yer

CLI
Updating pipeline
INFO:absl:Excluding no splits because exclude_splits is not set.
INFO:absl:Excluding no splits because exclude_splits is not set.
Pipeline "my_pipeline" updated successfully.
CLI
Creating a run for pipeline: my_pipeline
INFO:absl:Excluding no splits because exclude_splits is not set.
INFO:absl:Excluding no splits because exclude_splits is not set.
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: "SchemaGen"
  value {
    python_class_executable_spec {
      class_path: "tfx.components.schema_gen.executor.Executor"
    }
  }
}
executor_specs {
  key: "StatisticsGen"
  value {
    beam_executable_spec {
      python_executor_spec {
        class_path: "tfx.components.statistics_gen.executor.Executor"
      }
    }
  }
}
custom_driver_specs {
  key: "CsvExampleGen"
  value {
    python_class_executable_spec {
      class_path: "tfx.components.example_gen.driver.FileBasedDriver"
    }
  }
}
metadata_connection_config {
  database_connection_config {
    sqlite {
      filename_uri: "./tfx_metadata/my_pipeline/metadata.db"
      connection_mode: READWRITE_OPENCREATE
    }
  }
}

INFO:absl:Using connection config:
 sqlite {
  filename_uri: "./tfx_metadata/my_pipeline/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: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:24.358660"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.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
          }
          base_type: DATASET
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "input_base"
    value {
      field_value {
        string_value: "/home/kbuilder/imported/my_pipeline/data"
      }
    }
  }
  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: "StatisticsGen"
execution_options {
  caching_options {
  }
}

INFO:absl:MetadataStore with DB connection initialized
INFO:absl:select span and version = (0, None)
INFO:absl:latest span and version = (0, None)
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Going to run a new execution 10
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=10, input_dict={}, output_dict=defaultdict(<class 'list'>, {'examples': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/CsvExampleGen/examples/10"
custom_properties {
  key: "input_fingerprint"
  value {
    string_value: "split:single_split,num_files:1,total_bytes:25648,xor_checksum:1643886522,sum_checksum:1643886522"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:24.358660: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
}
base_type: DATASET
)]}), 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}', 'output_data_format': 6, 'input_base': '/home/kbuilder/imported/my_pipeline/data', 'input_config': '{\n  "splits": [\n    {\n      "name": "single_split",\n      "pattern": "*"\n    }\n  ]\n}', 'span': 0, 'version': None, 'input_fingerprint': 'split:single_split,num_files:1,total_bytes:25648,xor_checksum:1643886522,sum_checksum:1643886522'}, execution_output_uri='./tfx_pipeline_output/my_pipeline/CsvExampleGen/.system/executor_execution/10/executor_output.pb', stateful_working_dir='./tfx_pipeline_output/my_pipeline/CsvExampleGen/.system/stateful_working_dir/2022-02-03T11:10:24.358660', tmp_dir='./tfx_pipeline_output/my_pipeline/CsvExampleGen/.system/executor_execution/10/.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: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:24.358660"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.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
          }
          base_type: DATASET
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "input_base"
    value {
      field_value {
        string_value: "/home/kbuilder/imported/my_pipeline/data"
      }
    }
  }
  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: "StatisticsGen"
execution_options {
  caching_options {
  }
}
, pipeline_info=id: "my_pipeline"
, pipeline_run_id='2022-02-03T11:10:24.358660')
INFO:absl:Generating examples.
INFO:absl:Processing input csv data /home/kbuilder/imported/my_pipeline/data/* to TFExample.
WARNING:root:Make sure that locally built Python SDK docker image has Python 3.7 interpreter.
E0203 11:10:24.894390124    5584 fork_posix.cc:70]           Fork support is only compatible with the epoll1 and poll polling strategies
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:Value type <class 'NoneType'> of key version in exec_properties is not supported, going to drop it
INFO:absl:Value type <class 'list'> of key _beam_pipeline_args in exec_properties is not supported, going to drop it
INFO:absl:Cleaning up stateless execution info.
INFO:absl:Execution 10 succeeded.
INFO:absl:Cleaning up stateful execution info.
INFO:absl:Publishing output artifacts defaultdict(<class 'list'>, {'examples': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/CsvExampleGen/examples/10"
custom_properties {
  key: "input_fingerprint"
  value {
    string_value: "split:single_split,num_files:1,total_bytes:25648,xor_checksum:1643886522,sum_checksum:1643886522"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:24.358660:CsvExampleGen:examples:0"
  }
}
custom_properties {
  key: "span"
  value {
    int_value: 0
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
, artifact_type: name: "Examples"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
properties {
  key: "version"
  value: INT
}
base_type: DATASET
)]}) for execution 10
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component CsvExampleGen is finished.
INFO:absl:Component StatisticsGen is running.
INFO:absl:Running launcher for node_info {
  type {
    name: "tfx.components.statistics_gen.component.StatisticsGen"
    base_type: PROCESS
  }
  id: "StatisticsGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:24.358660"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.StatisticsGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "examples"
    value {
      channels {
        producer_node_query {
          id: "CsvExampleGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:10:24.358660"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.CsvExampleGen"
            }
          }
        }
        artifact_query {
          type {
            name: "Examples"
            base_type: DATASET
          }
        }
        output_key: "examples"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "statistics"
    value {
      artifact_spec {
        type {
          name: "ExampleStatistics"
          properties {
            key: "span"
            value: INT
          }
          properties {
            key: "split_names"
            value: STRING
          }
          base_type: STATISTICS
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
}
upstream_nodes: "CsvExampleGen"
downstream_nodes: "SchemaGen"
execution_options {
  caching_options {
  }
}

INFO:absl:MetadataStore with DB connection initialized
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Going to run a new execution 11
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=11, input_dict={'examples': [Artifact(artifact: id: 10
type_id: 15
uri: "./tfx_pipeline_output/my_pipeline/CsvExampleGen/examples/10"
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:1643886522,sum_checksum:1643886522"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:24.358660: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: "1.6.0"
  }
}
state: LIVE
create_time_since_epoch: 1643886625515
last_update_time_since_epoch: 1643886625515
, artifact_type: id: 15
name: "Examples"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
properties {
  key: "version"
  value: INT
}
base_type: DATASET
)]}, output_dict=defaultdict(<class 'list'>, {'statistics': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/11"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:24.358660:StatisticsGen:statistics:0"
  }
}
, artifact_type: name: "ExampleStatistics"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
base_type: STATISTICS
)]}), exec_properties={'exclude_splits': '[]'}, execution_output_uri='./tfx_pipeline_output/my_pipeline/StatisticsGen/.system/executor_execution/11/executor_output.pb', stateful_working_dir='./tfx_pipeline_output/my_pipeline/StatisticsGen/.system/stateful_working_dir/2022-02-03T11:10:24.358660', tmp_dir='./tfx_pipeline_output/my_pipeline/StatisticsGen/.system/executor_execution/11/.temp/', pipeline_node=node_info {
  type {
    name: "tfx.components.statistics_gen.component.StatisticsGen"
    base_type: PROCESS
  }
  id: "StatisticsGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:24.358660"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.StatisticsGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "examples"
    value {
      channels {
        producer_node_query {
          id: "CsvExampleGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:10:24.358660"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.CsvExampleGen"
            }
          }
        }
        artifact_query {
          type {
            name: "Examples"
            base_type: DATASET
          }
        }
        output_key: "examples"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "statistics"
    value {
      artifact_spec {
        type {
          name: "ExampleStatistics"
          properties {
            key: "span"
            value: INT
          }
          properties {
            key: "split_names"
            value: STRING
          }
          base_type: STATISTICS
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
}
upstream_nodes: "CsvExampleGen"
downstream_nodes: "SchemaGen"
execution_options {
  caching_options {
  }
}
, pipeline_info=id: "my_pipeline"
, pipeline_run_id='2022-02-03T11:10:24.358660')
INFO:absl:Generating statistics for split train.
INFO:absl:Statistics for split train written to ./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/11/Split-train.
INFO:absl:Generating statistics for split eval.
INFO:absl:Statistics for split eval written to ./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/11/Split-eval.
WARNING:root:Make sure that locally built Python SDK docker image has Python 3.7 interpreter.
INFO:absl:Cleaning up stateless execution info.
INFO:absl:Execution 11 succeeded.
INFO:absl:Cleaning up stateful execution info.
INFO:absl:Publishing output artifacts defaultdict(<class 'list'>, {'statistics': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/11"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:24.358660:StatisticsGen:statistics:0"
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
, artifact_type: name: "ExampleStatistics"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
base_type: STATISTICS
)]}) for execution 11
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component StatisticsGen is finished.
INFO:absl:Component SchemaGen is running.
INFO:absl:Running launcher for node_info {
  type {
    name: "tfx.components.schema_gen.component.SchemaGen"
    base_type: PROCESS
  }
  id: "SchemaGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:24.358660"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.SchemaGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "statistics"
    value {
      channels {
        producer_node_query {
          id: "StatisticsGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:10:24.358660"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.StatisticsGen"
            }
          }
        }
        artifact_query {
          type {
            name: "ExampleStatistics"
            base_type: STATISTICS
          }
        }
        output_key: "statistics"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "schema"
    value {
      artifact_spec {
        type {
          name: "Schema"
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
  parameters {
    key: "infer_feature_shape"
    value {
      field_value {
        int_value: 1
      }
    }
  }
}
upstream_nodes: "StatisticsGen"
execution_options {
  caching_options {
  }
}

INFO:absl:MetadataStore with DB connection initialized
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Going to run a new execution 12
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=12, input_dict={'statistics': [Artifact(artifact: id: 11
type_id: 17
uri: "./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/11"
properties {
  key: "split_names"
  value {
    string_value: "[\"train\", \"eval\"]"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:24.358660:StatisticsGen:statistics:0"
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
state: LIVE
create_time_since_epoch: 1643886628941
last_update_time_since_epoch: 1643886628941
, artifact_type: id: 17
name: "ExampleStatistics"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
base_type: STATISTICS
)]}, output_dict=defaultdict(<class 'list'>, {'schema': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/SchemaGen/schema/12"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:24.358660:SchemaGen:schema:0"
  }
}
, artifact_type: name: "Schema"
)]}), exec_properties={'infer_feature_shape': 1, 'exclude_splits': '[]'}, execution_output_uri='./tfx_pipeline_output/my_pipeline/SchemaGen/.system/executor_execution/12/executor_output.pb', stateful_working_dir='./tfx_pipeline_output/my_pipeline/SchemaGen/.system/stateful_working_dir/2022-02-03T11:10:24.358660', tmp_dir='./tfx_pipeline_output/my_pipeline/SchemaGen/.system/executor_execution/12/.temp/', pipeline_node=node_info {
  type {
    name: "tfx.components.schema_gen.component.SchemaGen"
    base_type: PROCESS
  }
  id: "SchemaGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:24.358660"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.SchemaGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "statistics"
    value {
      channels {
        producer_node_query {
          id: "StatisticsGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:10:24.358660"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.StatisticsGen"
            }
          }
        }
        artifact_query {
          type {
            name: "ExampleStatistics"
            base_type: STATISTICS
          }
        }
        output_key: "statistics"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "schema"
    value {
      artifact_spec {
        type {
          name: "Schema"
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
  parameters {
    key: "infer_feature_shape"
    value {
      field_value {
        int_value: 1
      }
    }
  }
}
upstream_nodes: "StatisticsGen"
execution_options {
  caching_options {
  }
}
, pipeline_info=id: "my_pipeline"
, pipeline_run_id='2022-02-03T11:10:24.358660')
INFO:absl:Processing schema from statistics for split train.
INFO:absl:Processing schema from statistics for split eval.
INFO:absl:Schema written to ./tfx_pipeline_output/my_pipeline/SchemaGen/schema/12/schema.pbtxt.
INFO:absl:Cleaning up stateless execution info.
INFO:absl:Execution 12 succeeded.
INFO:absl:Cleaning up stateful execution info.
INFO:absl:Publishing output artifacts defaultdict(<class 'list'>, {'schema': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/SchemaGen/schema/12"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:24.358660:SchemaGen:schema:0"
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
, artifact_type: name: "Schema"
)]}) for execution 12
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component SchemaGen is finished.

Değerlendiricinin çıktısını inceleyin

Bu adım, TensorFlow Model Analysis(TFMA) Jupyter notebook uzantısını gerektirir. TFMA dizüstü bilgisayar uzantısının sürümünün, TFMA python paketinin sürümüyle aynı olması gerektiğini unutmayın.

Aşağıdaki komut, NPM kayıt defterinden TFMA notebook uzantısını yükleyecektir. Tamamlanması birkaç dakika sürebilir.

# Install TFMA notebook extension.
jupyter labextension install tensorflow_model_analysis@{tfma.__version__}

tutucu31 l10n-yer

usage: jupyter [-h] [--version] [--config-dir] [--data-dir] [--runtime-dir]
               [--paths] [--json] [--debug]
               [subcommand]

Jupyter: Interactive Computing

positional arguments:
  subcommand     the subcommand to launch

optional arguments:
  -h, --help     show this help message and exit
  --version      show the versions of core jupyter packages and exit
  --config-dir   show Jupyter config dir
  --data-dir     show Jupyter data dir
  --runtime-dir  show Jupyter runtime dir
  --paths        show all Jupyter paths. Add --json for machine-readable
                 format.
  --json         output paths as machine-readable json
  --debug        output debug information about paths

Available subcommands: bundlerextension console dejavu execute kernel
kernelspec migrate nbconvert nbextension notebook qtconsole run
serverextension troubleshoot trust

Jupyter command `jupyter-labextension` not found.

Kurulum tamamlandıysa, uzantının geçerli olması için lütfen tarayıcınızı yeniden yükleyin .

with metadata.Metadata(metadata_connection_config) as metadata_handler:
  # Search all aritfacts from the previous pipeline run.
  artifacts = get_latest_artifacts(metadata_handler.store, PIPELINE_NAME)
  model_evaluation_artifacts = find_latest_artifacts_by_type(
      metadata_handler.store, artifacts,
      standard_artifacts.ModelEvaluation.TYPE_NAME)

tutucu33 l10n-yer

if model_evaluation_artifacts:
  tfma_result = tfma.load_eval_result(model_evaluation_artifacts[0].uri)
  tfma.view.render_slicing_metrics(tfma_result)

İşlem hattına İtici bileşeni ekler.

Model umut verici görünüyorsa, modeli yayınlamamız gerekiyor. Pusher bileşeni , modeli dosya sistemindeki bir konuma veya özel bir yürütücü kullanarak GCP AI Platform Modellerine yayınlayabilir.

Evaluator bileşeni, Trainer oluşturulan her modeli sürekli olarak değerlendirir ve Pusher , modeli dosya sisteminde önceden tanımlanmış bir konuma ve hatta Google Cloud AI Platform Modellerine kopyalar.

1. local_runner.py , SERVING_MODEL_DIR öğesini yayımlanacak bir dizine ayarlayın.
2. pipeline/pipeline.py dosyasında, ardışık düzene Pusher eklemek için # components.append(pusher) yorumunu kaldırın.

İşlem hattını güncelleyebilir ve yeniden çalıştırabilirsiniz.

# Update and run the pipeline.
!tfx pipeline update --engine=local --pipeline_path=local_runner.py \
 && tfx run create --engine=local --pipeline_name={PIPELINE_NAME}

tutucu35 l10n-yer

CLI
Updating pipeline
INFO:absl:Excluding no splits because exclude_splits is not set.
INFO:absl:Excluding no splits because exclude_splits is not set.
Pipeline "my_pipeline" updated successfully.
CLI
Creating a run for pipeline: my_pipeline
INFO:absl:Excluding no splits because exclude_splits is not set.
INFO:absl:Excluding no splits because exclude_splits is not set.
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: "SchemaGen"
  value {
    python_class_executable_spec {
      class_path: "tfx.components.schema_gen.executor.Executor"
    }
  }
}
executor_specs {
  key: "StatisticsGen"
  value {
    beam_executable_spec {
      python_executor_spec {
        class_path: "tfx.components.statistics_gen.executor.Executor"
      }
    }
  }
}
custom_driver_specs {
  key: "CsvExampleGen"
  value {
    python_class_executable_spec {
      class_path: "tfx.components.example_gen.driver.FileBasedDriver"
    }
  }
}
metadata_connection_config {
  database_connection_config {
    sqlite {
      filename_uri: "./tfx_metadata/my_pipeline/metadata.db"
      connection_mode: READWRITE_OPENCREATE
    }
  }
}

INFO:absl:Using connection config:
 sqlite {
  filename_uri: "./tfx_metadata/my_pipeline/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: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:48.556314"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.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
          }
          base_type: DATASET
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "input_base"
    value {
      field_value {
        string_value: "/home/kbuilder/imported/my_pipeline/data"
      }
    }
  }
  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: "StatisticsGen"
execution_options {
  caching_options {
  }
}

INFO:absl:MetadataStore with DB connection initialized
INFO:absl:select span and version = (0, None)
INFO:absl:latest span and version = (0, None)
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Going to run a new execution 13
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=13, input_dict={}, output_dict=defaultdict(<class 'list'>, {'examples': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/CsvExampleGen/examples/13"
custom_properties {
  key: "input_fingerprint"
  value {
    string_value: "split:single_split,num_files:1,total_bytes:25648,xor_checksum:1643886522,sum_checksum:1643886522"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:48.556314: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
}
base_type: DATASET
)]}), exec_properties={'input_config': '{\n  "splits": [\n    {\n      "name": "single_split",\n      "pattern": "*"\n    }\n  ]\n}', '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}', 'output_file_format': 5, 'output_data_format': 6, 'input_base': '/home/kbuilder/imported/my_pipeline/data', 'span': 0, 'version': None, 'input_fingerprint': 'split:single_split,num_files:1,total_bytes:25648,xor_checksum:1643886522,sum_checksum:1643886522'}, execution_output_uri='./tfx_pipeline_output/my_pipeline/CsvExampleGen/.system/executor_execution/13/executor_output.pb', stateful_working_dir='./tfx_pipeline_output/my_pipeline/CsvExampleGen/.system/stateful_working_dir/2022-02-03T11:10:48.556314', tmp_dir='./tfx_pipeline_output/my_pipeline/CsvExampleGen/.system/executor_execution/13/.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: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:48.556314"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.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
          }
          base_type: DATASET
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "input_base"
    value {
      field_value {
        string_value: "/home/kbuilder/imported/my_pipeline/data"
      }
    }
  }
  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: "StatisticsGen"
execution_options {
  caching_options {
  }
}
, pipeline_info=id: "my_pipeline"
, pipeline_run_id='2022-02-03T11:10:48.556314')
INFO:absl:Generating examples.
INFO:absl:Processing input csv data /home/kbuilder/imported/my_pipeline/data/* to TFExample.
WARNING:root:Make sure that locally built Python SDK docker image has Python 3.7 interpreter.
E0203 11:10:49.163841363    5734 fork_posix.cc:70]           Fork support is only compatible with the epoll1 and poll polling strategies
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:Value type <class 'NoneType'> of key version in exec_properties is not supported, going to drop it
INFO:absl:Value type <class 'list'> of key _beam_pipeline_args in exec_properties is not supported, going to drop it
INFO:absl:Cleaning up stateless execution info.
INFO:absl:Execution 13 succeeded.
INFO:absl:Cleaning up stateful execution info.
INFO:absl:Publishing output artifacts defaultdict(<class 'list'>, {'examples': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/CsvExampleGen/examples/13"
custom_properties {
  key: "input_fingerprint"
  value {
    string_value: "split:single_split,num_files:1,total_bytes:25648,xor_checksum:1643886522,sum_checksum:1643886522"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:48.556314:CsvExampleGen:examples:0"
  }
}
custom_properties {
  key: "span"
  value {
    int_value: 0
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
, artifact_type: name: "Examples"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
properties {
  key: "version"
  value: INT
}
base_type: DATASET
)]}) for execution 13
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component CsvExampleGen is finished.
INFO:absl:Component StatisticsGen is running.
INFO:absl:Running launcher for node_info {
  type {
    name: "tfx.components.statistics_gen.component.StatisticsGen"
    base_type: PROCESS
  }
  id: "StatisticsGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:48.556314"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.StatisticsGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "examples"
    value {
      channels {
        producer_node_query {
          id: "CsvExampleGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:10:48.556314"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.CsvExampleGen"
            }
          }
        }
        artifact_query {
          type {
            name: "Examples"
            base_type: DATASET
          }
        }
        output_key: "examples"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "statistics"
    value {
      artifact_spec {
        type {
          name: "ExampleStatistics"
          properties {
            key: "span"
            value: INT
          }
          properties {
            key: "split_names"
            value: STRING
          }
          base_type: STATISTICS
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
}
upstream_nodes: "CsvExampleGen"
downstream_nodes: "SchemaGen"
execution_options {
  caching_options {
  }
}

INFO:absl:MetadataStore with DB connection initialized
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Going to run a new execution 14
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=14, input_dict={'examples': [Artifact(artifact: id: 13
type_id: 15
uri: "./tfx_pipeline_output/my_pipeline/CsvExampleGen/examples/13"
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:1643886522,sum_checksum:1643886522"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:48.556314: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: "1.6.0"
  }
}
state: LIVE
create_time_since_epoch: 1643886649739
last_update_time_since_epoch: 1643886649739
, artifact_type: id: 15
name: "Examples"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
properties {
  key: "version"
  value: INT
}
base_type: DATASET
)]}, output_dict=defaultdict(<class 'list'>, {'statistics': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/14"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:48.556314:StatisticsGen:statistics:0"
  }
}
, artifact_type: name: "ExampleStatistics"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
base_type: STATISTICS
)]}), exec_properties={'exclude_splits': '[]'}, execution_output_uri='./tfx_pipeline_output/my_pipeline/StatisticsGen/.system/executor_execution/14/executor_output.pb', stateful_working_dir='./tfx_pipeline_output/my_pipeline/StatisticsGen/.system/stateful_working_dir/2022-02-03T11:10:48.556314', tmp_dir='./tfx_pipeline_output/my_pipeline/StatisticsGen/.system/executor_execution/14/.temp/', pipeline_node=node_info {
  type {
    name: "tfx.components.statistics_gen.component.StatisticsGen"
    base_type: PROCESS
  }
  id: "StatisticsGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:48.556314"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.StatisticsGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "examples"
    value {
      channels {
        producer_node_query {
          id: "CsvExampleGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:10:48.556314"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.CsvExampleGen"
            }
          }
        }
        artifact_query {
          type {
            name: "Examples"
            base_type: DATASET
          }
        }
        output_key: "examples"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "statistics"
    value {
      artifact_spec {
        type {
          name: "ExampleStatistics"
          properties {
            key: "span"
            value: INT
          }
          properties {
            key: "split_names"
            value: STRING
          }
          base_type: STATISTICS
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
}
upstream_nodes: "CsvExampleGen"
downstream_nodes: "SchemaGen"
execution_options {
  caching_options {
  }
}
, pipeline_info=id: "my_pipeline"
, pipeline_run_id='2022-02-03T11:10:48.556314')
INFO:absl:Generating statistics for split train.
INFO:absl:Statistics for split train written to ./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/14/Split-train.
INFO:absl:Generating statistics for split eval.
INFO:absl:Statistics for split eval written to ./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/14/Split-eval.
WARNING:root:Make sure that locally built Python SDK docker image has Python 3.7 interpreter.
INFO:absl:Cleaning up stateless execution info.
INFO:absl:Execution 14 succeeded.
INFO:absl:Cleaning up stateful execution info.
INFO:absl:Publishing output artifacts defaultdict(<class 'list'>, {'statistics': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/14"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:48.556314:StatisticsGen:statistics:0"
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
, artifact_type: name: "ExampleStatistics"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
base_type: STATISTICS
)]}) for execution 14
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component StatisticsGen is finished.
INFO:absl:Component SchemaGen is running.
INFO:absl:Running launcher for node_info {
  type {
    name: "tfx.components.schema_gen.component.SchemaGen"
    base_type: PROCESS
  }
  id: "SchemaGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:48.556314"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.SchemaGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "statistics"
    value {
      channels {
        producer_node_query {
          id: "StatisticsGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:10:48.556314"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.StatisticsGen"
            }
          }
        }
        artifact_query {
          type {
            name: "ExampleStatistics"
            base_type: STATISTICS
          }
        }
        output_key: "statistics"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "schema"
    value {
      artifact_spec {
        type {
          name: "Schema"
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
  parameters {
    key: "infer_feature_shape"
    value {
      field_value {
        int_value: 1
      }
    }
  }
}
upstream_nodes: "StatisticsGen"
execution_options {
  caching_options {
  }
}

INFO:absl:MetadataStore with DB connection initialized
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Going to run a new execution 15
INFO:absl:Going to run a new execution: ExecutionInfo(execution_id=15, input_dict={'statistics': [Artifact(artifact: id: 14
type_id: 17
uri: "./tfx_pipeline_output/my_pipeline/StatisticsGen/statistics/14"
properties {
  key: "split_names"
  value {
    string_value: "[\"train\", \"eval\"]"
  }
}
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:48.556314:StatisticsGen:statistics:0"
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
state: LIVE
create_time_since_epoch: 1643886653128
last_update_time_since_epoch: 1643886653128
, artifact_type: id: 17
name: "ExampleStatistics"
properties {
  key: "span"
  value: INT
}
properties {
  key: "split_names"
  value: STRING
}
base_type: STATISTICS
)]}, output_dict=defaultdict(<class 'list'>, {'schema': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/SchemaGen/schema/15"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:48.556314:SchemaGen:schema:0"
  }
}
, artifact_type: name: "Schema"
)]}), exec_properties={'infer_feature_shape': 1, 'exclude_splits': '[]'}, execution_output_uri='./tfx_pipeline_output/my_pipeline/SchemaGen/.system/executor_execution/15/executor_output.pb', stateful_working_dir='./tfx_pipeline_output/my_pipeline/SchemaGen/.system/stateful_working_dir/2022-02-03T11:10:48.556314', tmp_dir='./tfx_pipeline_output/my_pipeline/SchemaGen/.system/executor_execution/15/.temp/', pipeline_node=node_info {
  type {
    name: "tfx.components.schema_gen.component.SchemaGen"
    base_type: PROCESS
  }
  id: "SchemaGen"
}
contexts {
  contexts {
    type {
      name: "pipeline"
    }
    name {
      field_value {
        string_value: "my_pipeline"
      }
    }
  }
  contexts {
    type {
      name: "pipeline_run"
    }
    name {
      field_value {
        string_value: "2022-02-03T11:10:48.556314"
      }
    }
  }
  contexts {
    type {
      name: "node"
    }
    name {
      field_value {
        string_value: "my_pipeline.SchemaGen"
      }
    }
  }
}
inputs {
  inputs {
    key: "statistics"
    value {
      channels {
        producer_node_query {
          id: "StatisticsGen"
        }
        context_queries {
          type {
            name: "pipeline"
          }
          name {
            field_value {
              string_value: "my_pipeline"
            }
          }
        }
        context_queries {
          type {
            name: "pipeline_run"
          }
          name {
            field_value {
              string_value: "2022-02-03T11:10:48.556314"
            }
          }
        }
        context_queries {
          type {
            name: "node"
          }
          name {
            field_value {
              string_value: "my_pipeline.StatisticsGen"
            }
          }
        }
        artifact_query {
          type {
            name: "ExampleStatistics"
            base_type: STATISTICS
          }
        }
        output_key: "statistics"
      }
      min_count: 1
    }
  }
}
outputs {
  outputs {
    key: "schema"
    value {
      artifact_spec {
        type {
          name: "Schema"
        }
      }
    }
  }
}
parameters {
  parameters {
    key: "exclude_splits"
    value {
      field_value {
        string_value: "[]"
      }
    }
  }
  parameters {
    key: "infer_feature_shape"
    value {
      field_value {
        int_value: 1
      }
    }
  }
}
upstream_nodes: "StatisticsGen"
execution_options {
  caching_options {
  }
}
, pipeline_info=id: "my_pipeline"
, pipeline_run_id='2022-02-03T11:10:48.556314')
INFO:absl:Processing schema from statistics for split train.
INFO:absl:Processing schema from statistics for split eval.
INFO:absl:Schema written to ./tfx_pipeline_output/my_pipeline/SchemaGen/schema/15/schema.pbtxt.
INFO:absl:Cleaning up stateless execution info.
INFO:absl:Execution 15 succeeded.
INFO:absl:Cleaning up stateful execution info.
INFO:absl:Publishing output artifacts defaultdict(<class 'list'>, {'schema': [Artifact(artifact: uri: "./tfx_pipeline_output/my_pipeline/SchemaGen/schema/15"
custom_properties {
  key: "name"
  value {
    string_value: "my_pipeline:2022-02-03T11:10:48.556314:SchemaGen:schema:0"
  }
}
custom_properties {
  key: "tfx_version"
  value {
    string_value: "1.6.0"
  }
}
, artifact_type: name: "Schema"
)]}) for execution 15
INFO:absl:MetadataStore with DB connection initialized
INFO:absl:Component SchemaGen is finished.

SERVING_MODEL_DIR adresinde yeni modelinizi bulabilmeniz gerekir.

Adım 6. (İsteğe bağlı) Ardışık düzeninizi GCP'de Kubeflow Pipelines'a dağıtın.

Daha önce belirtildiği gibi, local_runner.py hata ayıklama veya geliştirme amacıyla iyidir, ancak üretim iş yükleri için en iyi çözüm değildir. Bu adımda ardışık düzeni Google Cloud'daki Kubeflow Pipelines'a dağıtacağız.

Hazırlık

Bir Kubeflow Pipelines kümesine bir işlem hattı dağıtmak için kfp python paketine ve skaffold programına ihtiyacımız var.

pip install --upgrade -q kfp

# Download skaffold and set it executable.
curl -Lo skaffold https://storage.googleapis.com/skaffold/releases/latest/skaffold-linux-amd64 && chmod +x skaffold

skaffold ikilisini kabuğunuzun bulabileceği yere taşımanız gerekir. Veya --skaffold-cmd bayrağıyla tfx binary'yi çalıştırdığınızda tfx yolunu belirtebilirsiniz.

# Move skaffold binary into your path
mv skaffold /home/jupyter/.local/bin/

tutucu38 l10n-yer

mv: cannot move 'skaffold' to '/home/jupyter/.local/bin/': No such file or directory

Ayrıca işlem hattını çalıştırmak için bir Kubeflow Pipelines kümesine ihtiyacınız vardır. Lütfen TFX on Cloud AI Platform Pipelines eğitiminde Adım 1 ve 2'yi izleyin.

Kümeniz hazır olduğunda , Google bulut konsolunun Pipelines sayfasında Pipelines Dashboard'u Aç'ı tıklayarak ardışık düzen panosunu açın. Bu sayfanın URL'si, bir ardışık düzen çalıştırması istemek için ENDPOINT şeklindedir. Bitiş noktası değeri, URL'de https://'den sonra, googleusercontent.com'a kadar ve buna kadar olan her şeydir. Uç noktanızı aşağıdaki kod bloğuna yerleştirin.

ENDPOINT='' # Enter your ENDPOINT here.

Kodumuzu bir Kubeflow Pipelines kümesinde çalıştırmak için kodumuzu bir kapsayıcı görüntüsüne paketlememiz gerekiyor. Görüntü, ardışık düzenimiz dağıtılırken otomatik olarak oluşturulacaktır ve görüntünüz için yalnızca bir ad ve bir kapsayıcı kayıt defteri ayarlamanız yeterlidir. Örneğimizde, Google Container kayıt defterini kullanacağız ve adını tfx-pipeline olarak adlandıracağız.

# Read GCP project id from env.
shell_output=!gcloud config list --format 'value(core.project)' 2>/dev/null
GOOGLE_CLOUD_PROJECT=shell_output[0]

# Docker image name for the pipeline image.
CUSTOM_TFX_IMAGE='gcr.io/' + GOOGLE_CLOUD_PROJECT + '/tfx-pipeline'

Veri konumunu ayarlayın.

Verilerinize Kubeflow Pipelines kümesinden erişilebilir olmalıdır. Yerel ortamınızdaki verileri kullandıysanız, bunları Google Cloud Storage gibi uzak depolamaya yüklemeniz gerekebilir. Örneğin, aşağıdaki gibi bir Kubeflow Pipelines kümesi dağıtıldığında otomatik olarak oluşturulan varsayılan bir kovaya penguen verilerini yükleyebiliriz.

gsutil cp data/data.csv gs://{GOOGLE_CLOUD_PROJECT}-kubeflowpipelines-default/tfx-template/data/penguin/

tutucu42 l10n-yer

Copying file://data/data.csv [Content-Type=text/csv]...
NotFoundException: 404 The destination bucket gs://tf-benchmark-dashboard-kubeflowpipelines-default does not exist or the write to the destination must be restarted

DATA_PATH içinde depolanan veri konumunu DATA_PATH içinde kubeflow_runner.py .

BigQueryExampleGen kullanıyorsanız, veri dosyasını yüklemenize gerek yoktur, ancak lütfen kubeflow_runner.py dosyasının boru beam_pipeline_args pipeline.create_pipeline() işlevi için aynı query ve kubeflow_runner.py bağımsız değişkenini kullandığından emin olun.

Boru hattını dağıtın.

Her şey hazırsa, tfx pipeline create komutunu kullanarak bir işlem hattı oluşturabilirsiniz.

!tfx pipeline create  \
--engine=kubeflow \
--pipeline-path=kubeflow_runner.py \
--endpoint={ENDPOINT} \
--build-target-image={CUSTOM_TFX_IMAGE}

tutucu44 l10n-yer

CLI
[Error] --build-target-image flag was DELETED. You should specify the build target image at the `KubeflowDagRunnerConfig` class instead, and use --build-image flag without argument to build a container image when creating or updating a pipeline.

Şimdi tfx run create komutunu kullanarak yeni oluşturulan işlem hattıyla bir yürütme çalıştırması başlatın.

tfx run create --engine=kubeflow --pipeline-name={PIPELINE_NAME} --endpoint={ENDPOINT}

tutucu46 l10n-yer

CLI
Creating a run for pipeline: my_pipeline
Failed to load kube config.
Traceback (most recent call last):
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/connection.py", line 175, in _new_conn
    (self._dns_host, self.port), self.timeout, **extra_kw
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/util/connection.py", line 95, in create_connection
    raise err
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/util/connection.py", line 85, in create_connection
    sock.connect(sa)
ConnectionRefusedError: [Errno 111] Connection refused

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/connectionpool.py", line 710, in urlopen
    chunked=chunked,
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/connectionpool.py", line 398, in _make_request
    conn.request(method, url, **httplib_request_kw)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/connection.py", line 239, in request
    super(HTTPConnection, self).request(method, url, body=body, headers=headers)
  File "/usr/lib/python3.7/http/client.py", line 1256, in request
    self._send_request(method, url, body, headers, encode_chunked)
  File "/usr/lib/python3.7/http/client.py", line 1302, in _send_request
    self.endheaders(body, encode_chunked=encode_chunked)
  File "/usr/lib/python3.7/http/client.py", line 1251, in endheaders
    self._send_output(message_body, encode_chunked=encode_chunked)
  File "/usr/lib/python3.7/http/client.py", line 1030, in _send_output
    self.send(msg)
  File "/usr/lib/python3.7/http/client.py", line 970, in send
    self.connect()
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/connection.py", line 205, in connect
    conn = self._new_conn()
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/connection.py", line 187, in _new_conn
    self, "Failed to establish a new connection: %s" % e
urllib3.exceptions.NewConnectionError: <urllib3.connection.HTTPConnection object at 0x7ff729e34190>: Failed to establish a new connection: [Errno 111] Connection refused

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/tmpfs/src/tf_docs_env/bin/tfx", line 8, in <module>
    sys.exit(cli_group())
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/click/core.py", line 829, in __call__
    return self.main(*args, **kwargs)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/click/core.py", line 782, in main
    rv = self.invoke(ctx)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/click/core.py", line 1259, in invoke
    return _process_result(sub_ctx.command.invoke(sub_ctx))
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/click/core.py", line 1259, in invoke
    return _process_result(sub_ctx.command.invoke(sub_ctx))
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/click/core.py", line 1066, in invoke
    return ctx.invoke(self.callback, **ctx.params)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/click/core.py", line 610, in invoke
    return callback(*args, **kwargs)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/click/decorators.py", line 73, in new_func
    return ctx.invoke(f, obj, *args, **kwargs)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/click/core.py", line 610, in invoke
    return callback(*args, **kwargs)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/tfx/tools/cli/commands/run.py", line 94, in create_run
    handler = handler_factory.create_handler(ctx.flags_dict)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/tfx/tools/cli/handler/handler_factory.py", line 93, in create_handler
    return kubeflow_handler.KubeflowHandler(flags_dict)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/tfx/tools/cli/handler/kubeflow_handler.py", line 62, in __init__
    namespace=self.flags_dict[labels.NAMESPACE])
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/kfp/_client.py", line 197, in __init__
    if not self._context_setting['namespace'] and self.get_kfp_healthz(
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/kfp/_client.py", line 411, in get_kfp_healthz
    response = self._healthz_api.get_healthz()
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/kfp_server_api/api/healthz_service_api.py", line 63, in get_healthz
    return self.get_healthz_with_http_info(**kwargs)  # noqa: E501
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/kfp_server_api/api/healthz_service_api.py", line 148, in get_healthz_with_http_info
    collection_formats=collection_formats)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/kfp_server_api/api_client.py", line 369, in call_api
    _preload_content, _request_timeout, _host)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/kfp_server_api/api_client.py", line 185, in __call_api
    _request_timeout=_request_timeout)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/kfp_server_api/api_client.py", line 393, in request
    headers=headers)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/kfp_server_api/rest.py", line 234, in GET
    query_params=query_params)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/kfp_server_api/rest.py", line 212, in request
    headers=headers)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/request.py", line 75, in request
    method, url, fields=fields, headers=headers, **urlopen_kw
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/request.py", line 96, in request_encode_url
    return self.urlopen(method, url, **extra_kw)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/poolmanager.py", line 375, in urlopen
    response = conn.urlopen(method, u.request_uri, **kw)
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/connectionpool.py", line 826, in urlopen
    **response_kw
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/connectionpool.py", line 826, in urlopen
    **response_kw
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/connectionpool.py", line 826, in urlopen
    **response_kw
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/connectionpool.py", line 786, in urlopen
    method, url, error=e, _pool=self, _stacktrace=sys.exc_info()[2]
  File "/tmpfs/src/tf_docs_env/lib/python3.7/site-packages/urllib3/util/retry.py", line 592, in increment
    raise MaxRetryError(_pool, url, error or ResponseError(cause))
urllib3.exceptions.MaxRetryError: HTTPConnectionPool(host='localhost', port=80): Max retries exceeded with url: /apis/v1beta1/healthz (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7ff729e34190>: Failed to establish a new connection: [Errno 111] Connection refused'))

Veya işlem hattını Kubeflow Pipelines panosunda da çalıştırabilirsiniz. Yeni çalıştırma, Kubeflow Pipelines panosundaki Experiments altında listelenecektir. Denemeye tıklamak, ilerlemeyi izlemenize ve yürütme çalıştırması sırasında oluşturulan yapıları görselleştirmenize olanak tanır.

Ardışık düzeninizi Kubeflow Pipelines üzerinde çalıştırmakla ilgileniyorsanız, TFX on Cloud AI Platform Pipelines eğitiminde daha fazla talimat bulun.

Temizlemek

Bu adımda kullanılan tüm Google Cloud kaynaklarını temizlemek için eğitim için kullandığınız Google Cloud projesini silebilirsiniz .

Alternatif olarak, her bir konsolu ziyaret ederek kaynakları tek tek temizleyebilirsiniz:

• Google Bulut Depolama
• Google Kapsayıcı Kaydı
• Google Kubernetes Motoru