Bot releases are hidden (Show)
tensorflow_transform
is tested to 20.04tensorflow 2.15
.tf.saved_model.SaveOptions
to model saving functionality.apache-beam[gcp]>=2.53.0,<3
for Python 3.11 and onapache-beam[gcp]>=2.47.0,<3
for 3.9 and 3.10.protobuf>=4.25.2,<5
for Python 3.11 and on protobuf>3.20.3,<5
Published by rtg0795 about 1 year ago
reserved_tokens
parameter to vocabulary APIs, a list of tokens thatapproximate_vocabulary
now returns tokens with the same frequency intft.vocabulary
).pyarrow>=10,<11
.apache-beam>=2.47,<3
.numpy>=1.22.0
.tensorflow>=2.13.0,<3
.Published by rtg0795 over 1 year ago
RaggedTensor
s can now be automatically inferred for variable lengthrepresent_variable_length_as_ragged=true
in TFMDtft.experimental.annotate_sparse_output_shape
andtft.experimental.annotate_true_sparse_output
.DatasetKey.non_cacheable
added to allow for some datasets to not producecompute_and_apply_vocabulary
can now storestore_frequency
parameter.numpy~=1.22.0
.tensorflow>=2.12.0,<2.13
.protobuf>=3.20.3,<5
.tensorflow-metadata>=1.13.1,<1.14.0
.tfx-bsl>=1.13.0,<1.14.0
.get_vocabulary_size_by_name
to return a minimum of 1.Published by venkat2469 almost 2 years ago
tensorflow>=2.11,<2.12
tensorflow-metadata>=1.12.0,<1.13.0
.tfx-bsl>=1.12.0,<1.13.0
.Published by venkat2469 almost 2 years ago
This is the last version that supports TensorFlow 1.15.x. TF 1.15.x support
will be removed in the next version. Please check the
TF2 migration guide to migrate
to TF2.
Introduced tft.experimental.document_frequency
and tft.experimental.idf
which map each term to its document frequency and inverse document frequency
in the same order as the terms in documents.
schema_utils.schema_as_feature_spec
now supports struct features as a way
to describe tf.SequenceExample
data.
TensorRepresentations in schema used for
schema_utils.schema_as_feature_spec
can now share name with their source
features.
Introduced tft_beam.EncodeTransformedDataset
which can be used to easily
encode transformed data in preparation for materialization.
tensorflow>=1.15.5,<2
or tensorflow>=2.10,<2.11
apache-beam[gcp]>=2.41,<3
.Published by venkat2469 about 2 years ago
tfx-bsl>=1.10.1,<1.11.0
.Published by venkat2469 about 2 years ago
apache-beam[gcp]>=2.40,<3
.pyarrow>=6,<7
.tensorflow-metadata>=1.10.0,<1.11.0
.tfx-bsl>=1.10.0,<1.11.0
.Published by rtg0795 over 2 years ago
scale_by_min_max_per_key
,scale_to_0_1_per_key
and scale_to_z_score_per_key
forkey_vocabulary_filename = None
.tensorflow>=1.15.5,<2
or tensorflow>=2.9,<2.10
tensorflow-metadata>=1.9.0,<1.10.0
.tfx-bsl>=1.9.0,<1.10.0
.Published by rtg0795 over 2 years ago
tft.DatasetMetadata
and its factory method from_feature_spec
asapache-beam[gcp]>=2.38,<3
.tensorflow>=1.15.5,!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.*,!=2.4.*,!=2.5.*,!=2.6.*,!=2.7.*,<2.9
.tensorflow-metadata>=1.8.0,<1.9.0
.tfx-bsl>=1.8.0,<1.9.0
.Published by rtg0795 over 2 years ago
tensorflow>=1.15.5,!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.*,!=2.4.*,!=2.5.*,!=2.6.*,<2.9
.Published by rtg0795 over 2 years ago
tft.experimental.compute_and_apply_approximate_vocabulary
whichtft.experimental.approximate_vocabulary
with text
tft.experimental.approximate_vocabulary
tft.get_analyze_input_columns
to ensure its output includespreprocessing_fn
inputs which are not used in any TFT analyzers, but endtft.apply_buckets
.apache-beam[gcp]>=2.36,<3
.tensorflow>=1.15.5,!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.*,!=2.4.*,!=2.5.*,!=2.6.*,!=2.7.*,<2.9
.tensorflow-metadata>=1.7.0,<1.8.0
.tfx-bsl>=1.7.0,<1.8.0
.Published by rtg0795 over 2 years ago
future
package.Published by jay90099 over 2 years ago
tft.experimental.get_vocabulary_size_by_name
that can retrievetft.vocabulary
within thepreprocessing_fn
.tft.experimental.ptransform_analyzer
now supports analyzer cache using thetft.experimental.CacheablePTransformAnalyzer
container.tft.bucketize_per_key
now supports weights.numpy>=1.16,<2
.apache-beam[gcp]>=2.35,<3
.absl-py>=0.9,<2.0.0
.tensorflow-metadata>=1.6.0,<1.7.0
.tfx-bsl>=1.6.0,<1.7.0
.tensorflow>=1.15.5,!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.*,!=2.4.*,!=2.5.*,!=2.6.*,<3
.Published by jay90099 almost 3 years ago
tft.experimental.approximate_vocabulary
analyzer that is antft.vocabulary
which is more efficient with smallertop_k
threshold.tft.experimental.ptransform_analyzer
's output dtypenp.ndarray
.apache-beam[gcp]>=2.34,<3
.tensorflow>=1.15.2,!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.*,!=2.4.*,!=2.5.*,!=2.6.*,<2.8
.tensorflow-metadata>=1.5.0,<1.6.0
.tfx-bsl>=1.5.0,<1.6.0
.Published by jay90099 almost 3 years ago
tf.RaggedTensor
support to all analyzers and mappers withreduce_instance_dims=True
.tft.apply_pyfunc
) in TF2.pyarrow>=1,<6
.tensorflow-metadata>=1.4.0,<1.5.0
.tfx-bsl>=1.4.0,<1.5.0
.apache-beam[gcp]>=2.33,<3
.Published by dhruvesh09 about 3 years ago
tft.quantiles
, tft.mean
and tft.var
now ignore NaNs and infinite inputtft_beam.AnalyzeDataset
,tft_beam.AnalyzeAndTransformDataset
and tft_beam.AnalyzeDatasetWithCache
tensorflow>=1.15.2,!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.*,!=2.4.*,!=2.5.*,<2.7
.tfx-bsl>=1.3.0,<1.4.0
.tft.mean
and tft.var
caches are automatically invalidated.Published by dhruvesh09 about 3 years ago
RaggedTensor
support to output schema inference and transformedpa.RecordBatch
with TF 2.x.apache-beam[gcp]>=2.31,<3
.tensorflow-metadata>=1.2.0,<1.3.0
.tfx-bsl>=1.2.0,<1.3.0
.Published by dhruvesh09 over 3 years ago
google-cloud-bigquery>=1.28.0,<2.21
.tfx-bsl>=1.1.1,<1.2.0
.Published by dhruvesh09 over 3 years ago
tft.vocabulary
when top_k
is set by removingpreprocessing_fn
usingtft.make_and_track_object
when force_tf_compat_v1=False
with TF2six
.protobuf>=3.13,<4
.tensorflow-metadata>=1.1.0,<1.2.0
.tfx-bsl>=1.1.0,<1.2.0
.Published by dhruvesh09 over 3 years ago
apache-beam[gcp]>=2.29,<3
.tensorflow>=1.15.2,!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.*,!=2.4.*,<2.6
.tensorflow-metadata>=1.0.0,<1.1.0
.tfx-bsl>=1.0.0,<1.1.0
.tft.ptransform_analyzer
has been moved under tft.experimental
. The ordertft_beam.PTransformAnalyzer
has been moved under tft_beam.experimental
.drop_unused_features
parameter toTFTransformOutput.transform_raw_features
is now True.