Releases: tensorflow/tfx
Releases · tensorflow/tfx
TFX 1.16.0 Release
Major Features and Improvements
Breaking Changes
Placeholder.__format__() is now disallowed, so you cannot use placeholders
in f-strings and str.format() calls anymore. If you get an error from this,
most likely you discovered a bug and should not use an f-string in the first
place. If it is truly your intention to print the placeholder (not its
resolved value) for debugging purposes, use repr() or !r instead.- Drop supports for the Estimator API.
For Pipeline Authors
For Component Authors
Deprecations
- KubeflowDagRunner (KFP v1 SDK) is deprecated. Use KubeflowV2DagRunner (KFP v2 pipeline spec) instead.
- Since Estimators will no longer be available in TensorFlow 2.16 and later versions, we have deprecated examples and templates that use them. We encourage you to explore Keras as a more modern and flexible high-level API for building and training models in TensorFlow.
Bug Fixes and Other Changes
Dependency Updates
| Package Name |
Version Constraints |
Previously (in v1.15.1) |
Comments |
docker |
>=7,<8 |
>=4.1,<5 |
|
Documentation Updates
TFX 1.16.0-rc0 Release
Major Features and Improvements
Breaking Changes
Placeholder.__format__() is now disallowed, so you cannot use placeholders
in f-strings and str.format() calls anymore. If you get an error from this,
most likely you discovered a bug and should not use an f-string in the first
place. If it is truly your intention to print the placeholder (not its
resolved value) for debugging purposes, use repr() or !r instead.- Drop supports for the Estimator API.
For Pipeline Authors
For Component Authors
Deprecations
- KubeflowDagRunner (KFP v1 SDK) is deprecated. Use KubeflowV2DagRunner (KFP v2 pipeline spec) instead.
- Since Estimators will no longer be available in TensorFlow 2.16 and later versions, we have deprecated examples and templates that use them. We encourage you to explore Keras as a more modern and flexible high-level API for building and training models in TensorFlow.
Bug Fixes and Other Changes
Dependency Updates
| Package Name |
Version Constraints |
Previously (in v1.15.1) |
Comments |
docker |
>=7,<8 |
>=4.1,<5 |
|
Documentation Updates
TFX 1.15.1
Version 1.15.1
Major Features and Improvements
Breaking Changes
- Support KFP pipeline spec 2.1.0 version schema and YAML files with KFP v2 DAG runner
For Pipeline Authors
For Component Authors
Deprecations
Bug Fixes and Other Changes
Dependency Updates
| Package Name |
Version Constraints |
Previously (in v1.14.0) |
Comments |
kfp-pipeline-spec |
kfp-pipeline-spec>=0.1.10,<0.2 |
>0.1.13,<0.2 |
|
Documentation Updates
TFX 1.15.0 Release
Version 1.15.0
Major Features and Improvements
- Dropped python 3.8 support.
- Extend GetPipelineRunExecutions, GetPipelineRunArtifacts APIs to support
filtering by execution create_time, type.
- ExampleValidator and DistributionValidator now support anomalies alert
generation. Users can use their own toolkits to extract and process the
alerts from the execution parameter.
- Allow DistributionValidator baseStatistics input channel artifacts to be
empty for cold start of data validation.
ph.make_proto() allows constructing proto-valued placeholders, e.g. for
larger config protos fed to a component.ph.join_path() is like os.path.join() but for placeholders.- Support passing in
experimental_debug_stripper into the Transform
pipeline runner.
Breaking Changes
Placeholder and all subclasses have been moved to other modules, their
structure has been changed and they're now immutable. Most users won't care
(the main public-facing API is unchanged and behaves the same way). If you
do special operations like isinstance() or some kind of custom
serialization on placeholders, you will have to update your code.placeholder.Placeholder.traverse() now returns more items than before,
namely also placeholder operators like _ConcatOperator (which is the
implementation of Python's + operator).- The
placeholder.RuntimeInfoKey enumeration was removed. Just hard-code the
appropriate string values in your code, and reference the new Literal type
placeholder.RuntimeInfoKeys if you want to ensure correctness.
- Arguments to
@component must now be passed as kwargs and its return type
is changed from being a Type to just being a callable that returns a new
instance (like the type's initializer). This will allow us to instead return
a factory function (which is not a Type) in future. For a given
@component def C(), this means:
- You should not use
C as a type anymore. For instance, replace
isinstance(foo, C) with something else. Depending on your use case, if
you just want to know whether it's a component, then use
isinstance(foo, tfx.types.BaseComponent) or
isinstance(foo, tfx.types.BaseFunctionalComponent).
If you want to know which component it is, check its .id instead.
Existing such checks will break type checking today and may additionally
break at runtime in future, if we migrate to a factory function.
- You can continue to use
C.test_call() like before, and it will
continue to be supported in future.
- Any type declarations using
foo: C break and must be replaced with
foo: tfx.types.BaseComponent or
foo: tfx.types.BaseFunctionalComponent.
- Any references to static class members like
C.EXECUTOR_SPEC breaks
type checking today and should be migrated away from. In particular, for
.EXECUTOR_SPEC.executor_class().Do() in unit tests, use .test_call()
instead.
- If your code previously asserted a wrong type declaration on
C, this
can now lead to (justified) type checking errors that were previously
hidden due to C being of type Any.
ph.to_list() was renamed to ph.make_list() for consistency.
For Pipeline Authors
For Component Authors
Deprecations
Bug Fixes and Other Changes
- Fixed a synchronization bug in google_cloud_ai_platform tuner.
- Print best tuning trials only from the chief worker of google_cloud_ai_platform tuner.
- Add a kpf dependency in the docker-image extra packages.
- Fix BigQueryExampleGen failure without custom_config.
Dependency Updates
| Package Name |
Version Constraints |
Previously (in v1.14.0) |
Comments |
keras-tuner |
>=1.0.4,<2,!=1.4.0,!=1.4.1 |
>=1.0.4,<2 |
|
packaging |
>=20,<21 |
>=22 |
|
attrs |
19.3.0,<22 |
19.3.0,<24 |
|
google-cloud-bigquery |
>=2.26.0,<3 |
>=3,<4 |
|
tensorflow |
>=2.15,<2.16 |
>=2.13,<2.14 |
|
tensorflow-decision-forests |
>=1.0.1,<1.9 |
>=1.0.1,<2 |
|
tensorflow-hub |
>=0.9.0,<0.14 |
>=0.15.0,<0.16 |
|
tensorflow-serving |
>=1.15,!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.*,!=2.4.*,!=2.5.*,!=2.6.*,!=2.7.*,!=2.8.*,<3 |
>=2.15,<2.16 |
|
Documentation Updates
TFX 1.15.0-rc0 Release
Major Features and Improvements
- Dropped python 3.8 support.
- Extend GetPipelineRunExecutions, GetPipelineRunArtifacts APIs to support
filtering by execution create_time, type.
- ExampleValidator and DistributionValidator now support anomalies alert
generation. Users can use their own toolkits to extract and process the
alerts from the execution parameter.
- Allow DistributionValidator baseStatistics input channel artifacts to be
empty for cold start of data validation.
ph.make_proto() allows constructing proto-valued placeholders, e.g. for
larger config protos fed to a component.ph.join_path() is like os.path.join() but for placeholders.- Support passing in
experimental_debug_stripper into the Transform
pipeline runner.
Breaking Changes
Placeholder and all subclasses have been moved to other modules, their
structure has been changed and they're now immutable. Most users won't care
(the main public-facing API is unchanged and behaves the same way). If you
do special operations like isinstance() or some kind of custom
serialization on placeholders, you will have to update your code.placeholder.Placeholder.traverse() now returns more items than before,
namely also placeholder operators like _ConcatOperator (which is the
implementation of Python's + operator).- The
placeholder.RuntimeInfoKey enumeration was removed. Just hard-code the
appropriate string values in your code, and reference the new Literal type
placeholder.RuntimeInfoKeys if you want to ensure correctness.
- Arguments to
@component must now be passed as kwargs and its return type
is changed from being a Type to just being a callable that returns a new
instance (like the type's initializer). This will allow us to instead return
a factory function (which is not a Type) in future. For a given
@component def C(), this means:
- You should not use
C as a type anymore. For instance, replace
isinstance(foo, C) with something else. Depending on your use case, if
you just want to know whether it's a component, then use
isinstance(foo, tfx.types.BaseComponent) or
isinstance(foo, tfx.types.BaseFunctionalComponent).
If you want to know which component it is, check its .id instead.
Existing such checks will break type checking today and may additionally
break at runtime in future, if we migrate to a factory function.
- You can continue to use
C.test_call() like before, and it will
continue to be supported in future.
- Any type declarations using
foo: C break and must be replaced with
foo: tfx.types.BaseComponent or
foo: tfx.types.BaseFunctionalComponent.
- Any references to static class members like
C.EXECUTOR_SPEC breaks
type checking today and should be migrated away from. In particular, for
.EXECUTOR_SPEC.executor_class().Do() in unit tests, use .test_call()
instead.
- If your code previously asserted a wrong type declaration on
C, this
can now lead to (justified) type checking errors that were previously
hidden due to C being of type Any.
ph.to_list() was renamed to ph.make_list() for consistency.
Deprecations
Bug Fixes and Other Changes
- Fixed a synchronization bug in google_cloud_ai_platform tuner.
- Print best tuning trials only from the chief worker of google_cloud_ai_platform tuner.
- Add a kpf dependency in the docker-image extra packages.
- Fix BigQueryExampleGen failure without custom_config.
Dependency Updates
| Package Name |
Version Constraints |
Previously (in v1.14.0) |
Comments |
keras-tuner |
>=1.0.4,<2,!=1.4.0,!=1.4.1 |
>=1.0.4,<2 |
|
packaging |
>=20,<21 |
>=22 |
|
attrs |
19.3.0,<22 |
19.3.0,<24 |
|
google-cloud-bigquery |
>=2.26.0,<3 |
>=3,<4 |
|
tensorflow |
>=2.15,<2.16 |
>=2.13,<2.14 |
|
tensorflow-decision-forests |
>=1.0.1,<1.9 |
>=1.0.1,<2 |
|
tensorflow-hub |
>=0.9.0,<0.14 |
>=0.15.0,<0.16 |
|
tensorflow-serving |
>=1.15,!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.*,!=2.4.*,!=2.5.*,!=2.6.*,!=2.7.*,!=2.8.*,<3 |
>=2.15,<2.16 |
|
Documentation Updates
TFX 1.14.0 Release
Major Features and Improvements
- Added python 3.10 support.
Breaking Changes
Placeholder (and _PlaceholderOperator) are no longer Jsonable.- Optimize MLMD register type to one call in most time instead of two calls.
For Pipeline Authors
For Component Authors
- Replace "tf_estimator" with "tfma_eval" as the identifier for tfma
EvalSavedModel. "tf_estimator" is now serves as the identifier for the normal
estimator model with any signature (by default 'serving').
Deprecations
Bug Fixes and Other Changes
- Apply latest TFX image vulnerability resolutions (base OS and software updates)
Dependency Updates
| Package Name |
Version Constraints |
Previously (in v1.13.0) |
Comments |
tensorflow-hub |
>=0.9.0,<0.14 |
>=0.9.0,<0.13 |
|
pyarrow |
>=10,<11 |
>=6,<7 |
|
apache-beam |
>=2.40,<3 |
>=2.47,<3 |
|
scikit-learn |
>=1.0,<2 |
>=0.23,<0.24 |
|
google-api-core |
<3 |
<1.33 |
|
google-cloud-aiplatform |
>=1.6.2,<2 |
>=1.6.2,<1.18 |
|
tflite-support |
>=0.4.3,<0.4.5 |
>=0.4.2,<0.4.3 |
|
pyyaml |
>=6,<7 |
>=3.12,<6 |
Issue with installation of PyYaml 5.4.1. (yaml/pyyaml#724) |
tensorflow |
>=2.13,<2.14 |
>=2.12,<2.13 |
|
tensorflowjs |
>=4.5,<5 |
>=3.6.0,<4 |
|
Documentation Updates
TFX 1.14.0-rc0 Release
Major Features and Improvements
- Added python 3.10 support.
Breaking Changes
Placeholder (and _PlaceholderOperator) are no longer Jsonable.- Optimize MLMD register type to one call in most time instead of two calls.
For Pipeline Authors
For Component Authors
- Replace "tf_estimator" with "tfma_eval" as the identifier for tfma
EvalSavedModel. "tf_estimator" is now serves as the identifier for the normal
estimator model with any signature (by default 'serving').
Deprecations
Bug Fixes and Other Changes
- Apply latest TFX image vulnerability resolutions (base OS and software updates)
Dependency Updates
| Package Name |
Version Constraints |
Previously (in v1.13.0) |
Comments |
tensorflow-hub |
>=0.9.0,<0.14 |
>=0.9.0,<0.13 |
|
pyarrow |
>=10,<11 |
>=6,<7 |
|
apache-beam |
>=2.40,<3 |
>=2.47,<3 |
|
scikit-learn |
>=1.0,<2 |
>=0.23,<0.24 |
|
google-api-core |
<3 |
<1.33 |
|
google-cloud-aiplatform |
>=1.6.2,<2 |
>=1.6.2,<1.18 |
|
tflite-support |
>=0.4.3,<0.4.5 |
>=0.4.2,<0.4.3 |
|
pyyaml |
>=6,<7 |
>=3.12,<6 |
Issue with installation of PyYaml 5.4.1. (yaml/pyyaml#724) |
tensorflow |
>=2.13,<2.14 |
>=2.12,<2.13 |
|
tensorflowjs |
>=4.5,<5 |
>=3.6.0,<4 |
|
Documentation Updates
TFX 1.13.0 Release
Major Features and Improvements
- Supported setting the container image at a component level for Kubeflow V2
Dag Runner.
Breaking Changes
For Pipeline Authors
-
Conditional can be used from tfx.dsl.Cond (Given from tfx import v1 as tfx).
-
Dummy channel for testing can be constructed by
tfx.testing.Channel(artifact_type).
-
placeholder.Placeholder.placeholders_involved() was replaced with
placeholder.Placeholder.traverse().
-
placeholder.Predicate.dependent_channels() was replaced with
channel_utils.get_dependent_channels(Placeholder).
-
placeholder.Predicate.encode_with_keys(...) was replaced with
channel_utils.encode_placeholder_with_channels(Placeholder, ...).
-
placeholder.Predicate.from_comparison() removed (was deprecated)
-
enable external_pipeline_artifact_query for querying artifact within one pipeline
-
Support InputArtifact[List[Artifact]] annotation in Python function custom component
For Component Authors
Deprecations
- Deprecate python 3.7 support
Bug Fixes and Other Changes
- Support to task type "workerpool1" of CLUSTER_SPEC in Vertex AI training's
service according to the changes of task type in Tuner component.
- Propagates unexpected import failures in the public v1 module.
Dependency Updates
| Package Name |
Version Constraints |
Previously (in v1.12.0) |
Comments |
click |
>=7,<9 |
>=7,<8 |
|
ml-metadata |
~=1.13.1 |
~=1.12.0 |
Synced release train |
protobuf |
>=3.13,<4 |
>=3.20.3,<5 |
To support TF 2.12 |
struct2tensor |
~=0.44.0 |
~=0.43.0 |
Synced release train |
tensorflow |
~=2.12.0 |
>=1.15.5,<2 or ~=2.11.0 |
|
tensorflow-data-validation |
~=1.13.0 |
~=1.12.0 |
Synced release train |
tensorflow-model-analysis |
~=0.44.0 |
~=0.43.0 |
Synced release train |
tensorflow-transform |
~=1.13.0 |
~=1.12.0 |
Synced release train |
tfx-bsl |
~=1.13.0 |
~=1.12.0 |
Synced release train |
Documentation Updates
- Added page for TFX-Addons
TFX 1.13.0-rc0 Release
Major Features and Improvements
- Supported setting the container image at a component level for Kubeflow V2
Dag Runner.
Breaking Changes
For Pipeline Authors
-
Conditional can be used from tfx.dsl.Cond (Given from tfx import v1 as tfx).
-
Dummy channel for testing can be constructed by
tfx.testing.Channel(artifact_type).
-
placeholder.Placeholder.placeholders_involved() was replaced with
placeholder.Placeholder.traverse().
-
placeholder.Predicate.dependent_channels() was replaced with
channel_utils.get_dependent_channels(Placeholder).
-
placeholder.Predicate.encode_with_keys(...) was replaced with
channel_utils.encode_placeholder_with_channels(Placeholder, ...).
-
placeholder.Predicate.from_comparison() removed (was deprecated)
-
enable external_pipeline_artifact_query for querying artifact within one pipeline
For Component Authors
Deprecations
- Deprecate python 3.7 support
Bug Fixes and Other Changes
- Support to task type "workerpool1" of CLUSTER_SPEC in Vertex AI training's
service according to the changes of task type in Tuner component.
- Propagates unexpected import failures in the public v1 module.
Dependency Updates
| Package Name |
Version Constraints |
Previously (in v1.12.0) |
Comments |
click |
>=7,<9 |
>=7,<8 |
|
ml-metadata |
~=1.13.1 |
~=1.12.0 |
Synced release train |
protobuf |
>=3.13,<4 |
>=3.20.3,<5 |
To support TF 2.12 |
struct2tensor |
~=0.44.0 |
~=0.43.0 |
Synced release train |
tensorflow |
~=2.12.0 |
>=1.15.5,<2 or ~=2.11.0 |
|
tensorflow-data-validation |
~=1.13.0 |
~=1.12.0 |
Synced release train |
tensorflow-model-analysis |
~=0.44.0 |
~=0.43.0 |
Synced release train |
tensorflow-transform |
~=1.13.0 |
~=1.12.0 |
Synced release train |
tfx-bsl |
~=1.13.0 |
~=1.12.0 |
Synced release train |
Documentation Updates
- Added page for TFX-Addons
TFX 1.12.0 Release
Major Features and Improvements
Breaking Changes
For Pipeline Authors
For Component Authors
Deprecations
Bug Fixes and Other Changes
- ExampleValidator and DistributionValidator now support custom validations.
Dependency Updates
| Package Name |
Version Constraints |
Previously (in v1.11.0) |
Comments |
tensorflow |
~=2.11.0 |
>=1.15.5,<2 or ~=2.10.0 |
|
tensorflow-decision-forests |
>=1.0.1,<2 |
==1.0.1 |
Make it compatible with more TF versions. |
ml-metadata |
~=1.12.0 |
~=1.11.0 |
Synced release train |
struct2tensor |
~=0.43.0 |
~=0.42.0 |
Synced release train |
tensorflow-data-validation |
~=1.12.0 |
~=1.11.0 |
Synced release train |
tensorflow-model-analysis |
~=0.43.0 |
~=0.42.0 |
Synced release train |
tensorflow-transform |
~=1.12.0 |
~=1.11.0 |
Synced release train |
tfx-bsl |
~=1.12.0 |
~=1.11.0 |
Synced release train |
Documentation Updates
