Release 2.8.0
Major Features and Improvements
-
tf.lite
:
- Added TFLite builtin op support for the following TF ops:
-
tf.raw_ops.Bucketize
op on CPU.
-
tf.where
op for data types tf.int32
/tf.uint32
/tf.int8
/tf.uint8
/tf.int64
.
-
tf.random.normal
op for output data type tf.float32
on CPU.
-
tf.random.uniform
op for output data type tf.float32
on CPU.
-
tf.random.categorical
op for output data type tf.int64
on CPU.
-
tensorflow.experimental.tensorrt
:
-
conversion_params
is now deprecated inside TrtGraphConverterV2
in favor of direct arguments: max_workspace_size_bytes
, precision_mode
, minimum_segment_size
, maximum_cached_engines
, use_calibration
and allow_build_at_runtime
.
- Added a new parameter called
save_gpu_specific_engines
to the .save()
function inside TrtGraphConverterV2
. When False
, the .save()
function won't save any TRT engines that have been built. When True
(default), the original behavior is preserved.
-
TrtGraphConverterV2
provides a new API called .summary()
which outputs a summary of the inference converted by TF-TRT. It namely shows each TRTEngineOp with their input(s)' and output(s)' shape and dtype. A detailed version of the summary is available which prints additionally all the TensorFlow OPs included in each of the TRTEngineOPs.
-
tf.tpu.experimental.embedding
:
-
tf.tpu.experimental.embedding.FeatureConfig
now takes an additional argument output_shape
which can specify the shape of the output activation for the feature.
-
tf.tpu.experimental.embedding.TPUEmbedding
now has the same behavior as tf.tpu.experimental.embedding.serving_embedding_lookup
which can take arbitrary rank of dense and sparse tensor. For ragged tensor, though the input tensor remains to be rank 2, the activations now can be rank 2 or above by specifying the output shape in the feature config or via the build method.
- Add
tf.config.experimental.enable_op_determinism
, which makes TensorFlow ops run deterministically at the cost of performance. Replaces the TF_DETERMINISTIC_OPS
environmental variable, which is now deprecated.
- The "Bug Fixes and Other Changes" section lists more determinism-related changes.
Bug Fixes and Other Changes
tf.data
:
- The optimization
parallel_batch
now becomes default if not disabled by users, which will parallelize copying of batch elements.
- Added the ability for
TensorSliceDataset
to identify and handle inputs that are files. This enables creating hermetic SavedModels when using datasets created from files.
tf.lite
:
- GPU
- Adds GPU Delegation support for serialization to Java API. This boosts initialization time upto 90% when OpenCL is available.
- Deprecated
Interpreter::SetNumThreads
, in favor of InterpreterBuilder::SetNumThreads
.
Adds tf.compat.v1.keras.utils.get_or_create_layer
to aid migration to TF2 by enabling tracking of nested keras models created in TF1-style, when used with the tf.compat.v1.keras.utils.track_tf1_style_variables
decorator.
tf.keras
:
- Preprocessing Layers
- Added a
tf.keras.layers.experimental.preprocessing.HashedCrossing
layer which applies the hashing trick to the concatenation of crossed scalar inputs. This provides a stateless way to try adding feature crosses of integer or string data to a model.
- Removed
keras.layers.experimental.preprocessing.CategoryCrossing
. Users should migrate to the HashedCrossing
layer or use tf.sparse.cross
/tf.ragged.cross
directly.
- Added additional
standardize
and split
modes to TextVectorization
.
standardize="lower"
will lowercase inputs.
standardize="string_punctuation"
will remove all puncuation.
split="character"
will split on every unicode character.
- Added an
output_mode
argument to the Discretization
and Hashing
layers with the same semantics as other preprocessing layers. All categorical preprocessing layers now support output_mode
.
- All preprocessing layer output will follow the compute dtype of a
tf.keras.mixed_precision.Policy
, unless constructed with output_mode="int"
in which case output will be tf.int64
. The output type of any preprocessing layer can be controlled individually by passing a dtype
argument to the layer.
tf.random.Generator
for keras initializers and all RNG code.
- Added 3 new APIs for enable/disable/check the usage of
tf.random.Generator
in keras backend, which will be the new backend for all the RNG in Keras. We plan to switch on the new code path by default in TF 2.8, and the behavior change will likely to cause some breakage on user side (eg. if the test is checking against some golden number). These 3 APIs will allow user to disable and switch back to legacy behavior if they prefer. In future (eg tf 2.10), we expect to totally remove the legacy code path (stateful random Ops), and these 3 APIs will be removed as well.
tf.keras.callbacks.experimental.BackupAndRestore
is now available as tf.keras.callbacks.BackupAndRestore
. The experimental endpoint is deprecated and will be removed in a future release.
tf.keras.experimental.SidecarEvaluator
is now available as tf.keras.utils.SidecarEvaluator
. The experimental endpoint is deprecated and will be removed in a future release.
- Metrics update and collection logic in default
Model.train_step()
is now customizable via overriding Model.compute_metrics()
.
- Losses computation logic in default
Model.train_step()
is now customizable via overriding Model.compute_loss()
.
jit_compile
added to Model.compile()
on an opt-in basis to compile the model's training step with XLA. Note that jit_compile=True
may not necessarily work for all models.
Deterministic Op Functionality
- Add determinsitic GPU implementations of:
tf.function(jit_compile=True)
's that use Scatter
.
- (since v2.7) Stateful ops used in
tf.data.Dataset
- (since v2.7)
tf.convert_to_tensor
when fed with (sparse) tf.IndexedSlices
(because it uses tf.math.unsorted_segment_sum
)
- (since v2.7)
tf.gather
backprop (because tf.convert_to_tensor
reduces tf.gather
's (sparse) tf.IndexedSlices
gradients into its dense params
input)
- (since v2.7)
tf.math.segment_mean
- (since v2.7)
tf.math.segment_prod
- (since v2.7)
tf.math.segment_sum
- (since v2.7)
tf.math.unsorted_segment_mean
- (since v2.7)
tf.math.unsorted_segment_prod
- (since v2.7)
tf.math.unsorted_segment_sum
- (since v2.7)
tf.math.unsorted_segment_sqrt
- (since v2.7)
tf.nn.ctc_loss
(resolved, possibly in prior release, and confirmed with tests)
- (since v2.7)
tf.nn.sparse_softmax_crossentropy_with_logits
- (since v2.7) Run the following ops on CPU (with significant performance penalty):
tf.scatter_nd
and other related scatter functions, such as tf.tensor_scatter_nd_update
- Add determinism-unimplemented exception-throwing to the following ops. When op-determinism is expected (i.e. after
tf.config.experimental.enable_op_determinism
has been called), an attempt to use the specified paths through the following ops on a GPU will cause tf.errors.UnimplementedError
(with an understandable message), unless otherwise specified, to be thrown.
FakeQuantWithMinMaxVarsGradient
and FakeQuantWithMinMaxVarsPerChannelGradient
- (since v2.7)
tf.compat.v1.get_seed
if the global random seed has not yet been set (via tf.random.set_seed
). Throws RuntimeError
from Python or InvalidArgument
from C++
- (since v2.7)
tf.compat.v1.nn.fused_batch_norm
backprop to offset
when is_training=False
- (since v2.7)
tf.image.adjust_contrast
forward
- (since v2.7)
tf.image.resize
with method=ResizeMethod.NEAREST
backprop
- (since v2.7)
tf.linalg.svd
- (since v2.7)
tf.math.bincount
- (since v2.7)
tf.nn.depthwise_conv2d
backprop to filter
when not using cuDNN convolution
- (since v2.7)
tf.nn.dilation2d
gradient
- (since v2.7)
tf.nn.max_pool_with_argmax
gradient
- (since v2.7)
tf.raw_ops.DebugNumericSummary
and tf.raw_ops.DebugNumericSummaryV2
- (since v2.7)
tf.timestamp
. Throws FailedPrecondition
- (since v2.7)
tf.Variable.scatter_add
(and other scatter methods, both on ref and resource variables)
- (since v2.7) The random-number-generating ops in the
tf.random
module when the global random seed has not yet been set (via tf.random.set_seed
). Throws RuntimeError
from Python or InvalidArgument
from C++
Add tf.config.experimental.enable_op_determinism
, which makes TensorFlow ops run deterministically at the cost of performance. This is equivalent to setting the previously-existing TF_DETERMINISTIC_OPS
environmental variable to 1
. The environmental variable is now deprecated, so the enable_op_determinism
function should be used instead.
Thanks to our Contributors
This release contains contributions from many people at Google, as well as:
8bitmp3, Adam Lanicek, ag.ramesh, alesapin, Andrew Goodbody, annasuheyla, Ariel Elkin, Arnab Dutta, Ben Barsdell, bhack, cfRod, Chengji Yao, Christopher Bate, dan, Dan F-M, David Korczynski, DEKHTIARJonathan, dengzhiyuan, Deven Desai, Duncan Riach, Eli Osherovich, Ewout Ter Hoeven, ez2take, Faijul Amin, fo40225, Frederic Bastien, gadagashwini, Gauri1 Deshpande, Georgiy Manuilov, Guilherme De Lázari, Guozhong Zhuang, H1Gdev, homuler, Hongxu Jia, Jacky_Yin, jayfurmanek, jgehw, Jhalak Patel, Jinzhe Zeng, Johan Gunnarsson, Jonathan Dekhtiar, Kaixi Hou, Kanvi Khanna, Kevin Cheng, Koan-Sin Tan, Kruglov-Dmitry, Kun Lu, Lemo, Lequn Chen, long.chen, Louis Sugy, Mahmoud Abuzaina, Mao, Marius Brehler, Mark Harfouche, Martin Patz, Maxiwell S. Garcia, Meenakshi Venkataraman, Michael Melesse, Mrinal Tyagi, Måns Nilsson, Nathan John Sircombe, Nathan Luehr, Nilesh Agarwalla, Oktay Ozturk, Patrice Vignola, Pawel-Polyai, Rama Ketineni, Ramesh Sampath, Reza Rahimi, Rob Suderman, Robert Kalmar, Rohit Santhanam, Sachin Muradi, Saduf2019, Samuel Marks, Shi,Guangyong, Sidong-Wei, Srinivasan Narayanamoorthy, Srishti Srivastava, Steven I Reeves, stevenireeves, Supernovae, Tamas Bela Feher, Tao Xu, Thibaut Goetghebuer-Planchon, Thomas Schmeyer, tilakrayal, Valery Mironov, Victor Guo, Vignesh Kothapalli, Vishnuvardhan Janapati, wamuir, Wang,Quintin, William Muir, William Raveane, Yash Goel, Yimei Sun, Yong Tang, Yuduo Wu