tensorflow latest version

v2.4.0-rc4


tensorflow v2.4.0-rc4 Release Notes

Release Date: 2020εΉ΄12月04ζ—₯ // about 5 years ago

  • πŸš€ Release 2.4.0

    Major Features and Improvements

    πŸ‘€ tf.distribute introduces experimental support for asynchronous training of Keras models via the tf.distribute.experimental.ParameterServerStrategy API. Please see below for additional details.

    πŸ“„ MultiWorkerMirroredStrategy is now a stable API and is no longer considered experimental. Some of the major improvements involve handling peer failure and many bug fixes. Please check out the detailed tutorial on Multi-worker training with Keras.

    πŸ“„ Introduces experimental support for a new module named tf.experimental.numpy which is a NumPy-compatible API for writing TF programs. See the detailed guide to learn more. Additional details below.

    βž• Adds Support for
    πŸ”Š TensorFloat-32 on Ampere based GPUs. TensorFloat-32, or TF32 for short, is a math mode for NVIDIA Ampere based GPUs and is enabled by default.

    🐎 A major refactoring of the internals of the Keras Functional API has been completed, that should improve the reliability, stability, and performance of constructing Functional models.

    Keras mixed precision API tf.keras.mixed_precision is no longer experimental and allows the use of 16-bit floating point formats during training, improving performance by up to 3x on GPUs and 60% on TPUs. Please see below for additional details.

    πŸ‘· TensorFlow Profiler now supports profiling MultiWorkerMirroredStrategy and tracing multiple workers using the sampling mode API.

    TFLite Profiler for Android is available. See the detailed guide to learn more.

    πŸ“¦ TensorFlow pip packages are now built with CUDA11 and cuDNN 8.0.2.

    πŸ’₯ Breaking Changes

    TF Core:

    • Certain float32 ops run in lower precision on Ampere based GPUs, including matmuls and convolutions, due to the use of TensorFloat-32. Specifically, inputs to such ops are rounded from 23 bits of precision to 10
      bits of precision. This is unlikely to cause issues in practice for deep learning models. In some cases, TensorFloat-32 is also used for complex64 ops.
      TensorFloat-32 can be disabled by running tf.config.experimental.enable_tensor_float_32_execution(False).
    • The byte layout for string tensors across the C-API has been updated to match TF Core/C++; i.e., a contiguous array of tensorflow::tstring/TF_TStrings.
    • C-API functions TF_StringDecode, TF_StringEncode, and TF_StringEncodedSize are no longer relevant and have been removed; see core/platform/ctstring.h for string access/modification in C.
    • tensorflow.python, tensorflow.core and tensorflow.compiler modules are now hidden. These modules are not part of TensorFlow public API.
    • tf.raw_ops.Max and tf.raw_ops.Min no longer accept inputs of type tf.complex64 or tf.complex128, because the behavior of these ops is not well defined for complex types.

    - XLA:CPU and XLA:GPU devices are no longer registered by default. Use TF_XLA_FLAGS=--tf_xla_enable_xla_devices if you really need them, but this flag will eventually be removed in subsequent releases.

    tf.keras:

    • The steps_per_execution argument in compile() is no longer experimental; if you were passing experimental_steps_per_execution, rename it to steps_per_execution in your code. This argument controls the number of batches to run during each tf.function call when calling fit(). Running multiple batches inside a single tf.function call can greatly improve performance on TPUs or small models with a large Python overhead.
    • A major refactoring of the internals of the Keras Functional API may affect code that
      is relying on certain internal details:
      • Code that uses isinstance(x, tf.Tensor) instead of tf.is_tensor when checking Keras symbolic inputs/outputs should switch to using tf.is_tensor.
      • Code that is overly dependent on the exact names attached to symbolic tensors (e.g. assumes there will be ":0" at the end of the inputs, treats names as unique identifiers instead of using tensor.ref(), etc.)
      • Code that uses get_concrete_function to trace Keras symbolic inputs directly should switch to building matching tf.TensorSpecs directly and tracing the TensorSpec objects.
      • Code that relies on the exact number and names of the op layers that TensorFlow operations were converted into may have changed.
      • Code that uses tf.map_fn/tf.cond/tf.while_loop/control flow as op layers and happens to work before TF 2.4. These will explicitly be unsupported now. Converting these ops to Functional API op layers was unreliable before TF 2.4, and prone to erroring incomprehensibly or being silently buggy.
      • Code that directly asserts on a Keras symbolic value in cases where ops like tf.rank used to return a static or symbolic value depending on if the input had a fully static shape or not. Now these ops always return symbolic values.
      • Code already susceptible to leaking tensors outside of graphs becomes slightly more likely to do so now.
      • Code that tries directly getting gradients with respect to symbolic Keras inputs/outputs. Use GradientTape on the actual Tensors passed to the already- constructed model instead.
      • Code that requires very tricky shape manipulation via converted op layers in order to work, where the Keras symbolic shape inference proves insufficient.
      • Code that tries manually walking a tf.keras.Model layer by layer and assumes layers only ever have one positional argument. This assumption doesn't hold true before TF 2.4 either, but is more likely to cause issues now.
      • Code that manually enters keras.backend.get_graph() before building a functional model is no longer needed.
      • Start enforcing input shape assumptions when calling Functional API Keras models. This may potentially break some users, in case there is a mismatch between the shape used when creating Input objects in a Functional model, and the shape of the data passed to that model. You can fix this mismatch by either calling the model with correctly-shaped data, or by relaxing Input shape assumptions (note that you can pass shapes with None entries for axes
        that are meant to be dynamic). You can also disable the input checking entirely by setting model.input_spec = None.
    • Serveral changes have been made to tf.keras.mixed_precision.experimental. Note that it is now recommended to use the non-experimental tf.keras.mixed_precision API.
      • AutoCastVariable.dtype now refers to the actual variable dtype, not the dtype it will be casted to.
      • When mixed precision is enabled, tf.keras.layers.Embedding now outputs a float16 or bfloat16 tensor instead of a float32 tensor.
      • The property tf.keras.mixed_precision.experimental.LossScaleOptimizer.loss_scale is
        ⚑️ now a tensor, not a LossScale object. This means to get a loss scale of a LossScaleOptimizer as a tensor, you must now call
        opt.loss_scale instead of opt.loss_scale().
      • The property should_cast_variables has been removed from tf.keras.mixed_precision.experimental.Policy
      • When passing a tf.mixed_precision.experimental.DynamicLossScale to
        ⚑️ tf.keras.mixed_precision.experimental.LossScaleOptimizer, the DynamicLossScale's multiplier must be 2.
      • When passing a tf.mixed_precision.experimental.DynamicLossScale to tf.keras.mixed_precision.experimental.LossScaleOptimizer,
        ⚑️ the weights of the DynanmicLossScale are copied into the LossScaleOptimizer instead of being reused. This means modifying the
        ⚑️ weights of the DynamicLossScale will no longer affect the weights of the LossScaleOptimizer, and vice versa.
      • The global policy can no longer be set to a non-floating point policy in tf.keras.mixed_precision.experimental.set_policy
      • In Layer.call, AutoCastVariables will no longer be casted within MirroredStrategy.run or ReplicaContext.merge_call. This is
        because a thread local variable is used to determine whether AutoCastVariables are casted, and those two functions run with a
        different thread. Note this only applies if one of these two functions is called within Layer.call; if one of those two functions calls Layer.call, AutoCastVariables will still be casted.

    tf.data:

    • tf.data.experimental.service.DispatchServer now takes a config tuple instead of individual arguments. Usages should be updated to tf.data.experimental.service.DispatchServer(dispatcher_config).

    - tf.data.experimental.service.WorkerServer now takes a config tuple instead of individual arguments. Usages should be updated to tf.data.experimental.service.WorkerServer(worker_config).

    tf.distribute:

    • Removes tf.distribute.Strategy.experimental_make_numpy_dataset. Please use tf.data.Dataset.from_tensor_slices instead.
    • Renames experimental_hints in tf.distribute.StrategyExtended.reduce_to, tf.distribute.StrategyExtended.batch_reduce_to, tf.distribute.ReplicaContext.all_reduce to options:
    • Renames tf.distribute.experimental.CollectiveHints to tf.distribute.experimental.CommunicationOptions.
    • Renames tf.distribute.experimental.CollectiveCommunication to tf.distribute.experimental.CommunicationImplementation.
    • Renames tf.distribute.Strategy.experimental_distribute_datasets_from_function to distribute_datasets_from_function as it is no longer experimental.

    - Removes tf.distribute.Strategy.experimental_run_v2 method, which was deprecated in TF 2.2.

    tf.lite:

    • tf.quantization.quantize_and_dequantize_v2 has been introduced, which updates the gradient definition for quantization which is outside the range
      to be 0. To simulate the V1 the behavior of tf.quantization.quantize_and_dequantize(...) use tf.grad_pass_through(tf.quantization.quantize_and_dequantize_v2)(...).

    πŸ› Bug Fixes and Other Changes

    TF Core:

    • πŸ‘ Introduces experimental support for a new module named tf.experimental.numpy, which
      is a NumPy-compatible API for writing TF programs. This module provides class ndarray, which mimics the ndarray class in NumPy, and wraps an immutable tf.Tensor under the hood. A subset of NumPy functions (e.g. numpy.add) are provided. Their inter-operation with TF facilities is seamless in most cases.
      See tensorflow/python/ops/numpy_ops/README.md
      πŸ‘ for details of what operations are supported and what are the differences from NumPy.
    • tf.types.experimental.TensorLike is a new Union type that can be used as type annotation for variables representing a Tensor or a value
      that can be converted to Tensor by tf.convert_to_tensor.
    • Calling ops with a python constants or numpy values is now consistent with tf.convert_to_tensor behavior. This avoids operations like
      tf.reshape truncating inputs such as from int64 to int32.
    • βž• Adds tf.sparse.map_values to apply a function to the .values of SparseTensor arguments.
    • The Python bitwise operators for Tensor (__and__, __or__, __xor__ and __invert__ now support non-bool arguments and apply
      πŸ‘ the corresponding bitwise ops. bool arguments continue to be supported and dispatch to logical ops. This brings them more in line with
      Python and NumPy behavior.
    • βž• Adds tf.SparseTensor.with_values. This returns a new SparseTensor with the same sparsity pattern, but with new provided values. It is
      similar to the with_values function of RaggedTensor.
    • βž• Adds StatelessCase op, and uses it if none of case branches has stateful ops.
    • Adds tf.config.experimental.get_memory_usage to return total memory usage of the device.
    • βž• Adds gradients for RaggedTensorToVariant and RaggedTensorFromVariant.
    • πŸ‘Œ Improve shape inference of nested function calls by supporting constant folding across Arg nodes which makes more static values available to shape inference functions.
    • tf.debugging:
      • tf.debugging.assert_shapes() now works on SparseTensors (Fixes #36268).
    • GPU
      • Adds Support for TensorFloat-32 on Ampere based GPUs.
        TensorFloat-32, or TF32 for short, is a math mode for NVIDIA Ampere based GPUs which causes certain float32 ops, such as matrix
        multiplications and convolutions, to run much faster on Ampere GPUs but with reduced precision. This reduced precision has not been found
        to effect convergence quality of deep learning models in practice. TensorFloat-32 is enabled by default, but can be disabled with tf.config.experimental.enable_tensor_float_32_execution.
    • tf.math:
      • Adds tf.math.erfcinv, the inverse to tf.math.erfc.
    • tf.nn:
      • tf.nn.max_pool2d now supports explicit padding.
    • tf.image:
      • Adds deterministic tf.image.stateless_random_* functions for each tf.image.random_* function. Added a new op stateless_sample_distorted_bounding_box which is a deterministic version of sample_distorted_bounding_box op. Given the same seed, these stateless functions/ops produce the same results independent of how many times the function is called, and independent of global seed settings.
      • Adds deterministic tf.image.resize backprop CUDA kernels for method=ResizeMethod.BILINEAR (the default method). Enable by setting the environment variable TF_DETERMINISTIC_OPS to "true" or "1".
    • πŸ–¨ tf.print:
      • Bug fix in tf.print() with OrderedDict where if an OrderedDict didn't have the keys sorted, the keys and values were not being printed
        in accordance with their correct mapping.
    • tf.train.Checkpoint:
      • Now accepts a root argument in the initialization, which generates a checkpoint with a root object. This allows users to create a Checkpoint object that is compatible with Keras model.save_weights() and model.load_weights. The checkpoint is also compatible with the checkpoint saved in the variables/ folder in the SavedModel.
      • When restoring, save_path can be a path to a SavedModel. The function will automatically find the checkpoint in the SavedModel.

    tf.data:

    • Adds new tf.data.experimental.service.register_dataset and tf.data.experimental.service.from_dataset_id APIs to enable one
      πŸ–¨ process to register a dataset with the tf.data service, and another process to consume data from the dataset.
    • βž• Adds support for dispatcher fault tolerance. To enable fault tolerance, configure a work_dir when running your dispatcher server and set
      dispatcher_fault_tolerance=True. The dispatcher will store its state to work_dir, so that on restart it can continue from its previous
      state after restart.
    • βž• Adds support for sharing dataset graphs via shared filesystem instead of over RPC. This reduces load on the dispatcher, improving performance
      πŸ‘· of distributing datasets. For this to work, the dispatcher's work_dir must be accessible from workers. If the worker fails to read from the
      work_dir, it falls back to using RPC for dataset graph transfer.
    • βž• Adds support for a new "distributed_epoch" processing mode. This processing mode distributes a dataset across all tf.data workers,
      πŸ“„ instead of having each worker process the full dataset. See the tf.data service docs to learn more.
    • Adds optional exclude_cols parameter to CsvDataset. This parameter is the complement of select_cols; at most one of these should be specified.
    • We have implemented an optimization which reorders data-discarding transformations such as take and shard to happen earlier in the dataset when it is safe to do so. The optimization can be disabled via the experimental_optimization.reorder_data_discarding_ops dataset option.
    • tf.data.Options were previously immutable and can now be overridden.
    • πŸ“œ tf.data.Dataset.from_generator now supports Ragged and Sparse tensors with a new output_signature argument, which allows from_generator to
      produce any type describable by a tf.TypeSpec.
    • tf.data.experimental.AUTOTUNE is now available in the core API as tf.data.AUTOTUNE.

    tf.distribute:

    • πŸ‘ Introduces experimental support for asynchronous training of Keras models via tf.distribute.experimental.ParameterServerStrategy:
      • Replaces the existing tf.distribute.experimental.ParameterServerStrategy symbol with a new class that is for parameter server training in TF2. Usage of
        the old symbol, usually with Estimator API, should be replaced with [tf.compat.v1.distribute.experimental.ParameterServerStrategy].
      • Added tf.distribute.experimental.coordinator.* namespace, including the main API ClusterCoordinator for coordinating the training cluster, the related data structure RemoteValue and PerWorkerValue.
    • βž• Adds tf.distribute.Strategy.gather and tf.distribute.ReplicaContext.all_gather APIs to support gathering dense distributed values.
    • πŸ›  Fixes various issues with saving a distributed model.

    tf.keras:

    • πŸ‘Œ Improvements from the Functional API refactoring:
      • Functional model construction does not need to maintain a global workspace graph, removing memory leaks especially when building many
        models or very large models.
      • Functional model construction should be ~8-10% faster on average.
      • Functional models can now contain non-symbolic values in their call inputs inside of the first positional argument.
      • Several classes of TF ops that were not reliably converted to Keras layers during functional API construction should now work, e.g.
        tf.image.ssim_multiscale
      • Error messages when Functional API construction goes wrong (and when ops cannot be converted to Keras layers automatically) should be
        clearer and easier to understand.
    • ⚑️ Optimizer.minimize can now accept a loss Tensor and a GradientTape as an alternative to accepting a callable loss.
    • βž• Adds beta hyperparameter to FTRL optimizer classes (Keras and others) to match FTRL paper.
    • Optimizer. __init__ now accepts a gradient_aggregator to allow for customization of how gradients are aggregated across devices, as well as
      gradients_transformers to allow for custom gradient transformations (such as gradient clipping).
    • πŸ‘Œ Improvements to Keras preprocessing layers:
      • TextVectorization can now accept a vocabulary list or file as an init arg.
      • Normalization can now accept mean and variance values as init args.
    • In Attention and AdditiveAttention layers, the call() method now accepts a return_attention_scores argument. When set to
      True, the layer returns the attention scores as an additional output argument.
    • βž• Adds tf.metrics.log_cosh and tf.metrics.logcosh API entrypoints with the same implementation as their tf.losses equivalent.
    • For Keras model, the individual call of Model.evaluate uses no cached data for evaluation, while Model.fit uses cached data when
      🐎 validation_data arg is provided for better performance.
    • Adds a save_traces argument to model.save/ tf.keras.models.save_model which determines whether the SavedModel format stores the Keras model/layer call functions. The traced functions allow Keras to revive custom models and layers without the original class definition, but if this isn't required the tracing can be disabled with the added option.
    • The tf.keras.mixed_precision API is non non-experimental. The
      non-experimental API differs from the experimental API in several ways.
      • tf.keras.mixed_precision.Policy no longer takes in a
        tf.mixed_precision.experimental.LossScale in the constructor, and no
        longer has a LossScale associated with it. Instead, Model.compile
        ⚑️ will automatically wrap the optimizer with a LossScaleOptimizer using
        dynamic loss scaling if Policy.name is "mixed_float16".
      • tf.keras.mixed_precision.LossScaleOptimizer's constructor takes in
        different arguments. In particular, it no longer takes in a LossScale,
        and there is no longer a LossScale associated with the
        ⚑️ LossScaleOptimizer. Instead, LossScaleOptimizer directly implements
        πŸ›  fixed or dynamic loss scaling. See the documentation of
        ⚑️ tf.keras.mixed_precision.experimental.LossScaleOptimizer
        for details on the differences between the experimental
        ⚑️ LossScaleOptimizer and the new non-experimental LossScaleOptimizer.
      • tf.mixed_precision.experimental.LossScale and its subclasses are
        πŸ—„ deprecated, as all of its functionality now exists within
        ⚑️ tf.keras.mixed_precision.LossScaleOptimizer

    tf.lite:

    • TFLiteConverter:
      • Support optional flags inference_input_type and inference_output_type for full integer quantized models. This allows users to modify the model input and output type to integer types (tf.int8, tf.uint8) instead of defaulting to float type (tf.float32).
    • NNAPI
      • Adds NNAPI Delegation support for requantization use cases by converting the operation into a dequantize-quantize pair.
      • Removes deprecated Interpreter.setUseNNAPI(boolean) Java API. Use Interpreter.Options.setUseNNAPI instead.
      • Deprecates Interpreter::UseNNAPI(bool) C++ API. Use NnApiDelegate() and related delegate configuration methods directly.
      • Deprecates Interpreter::SetAllowFp16PrecisionForFp32(bool) C++ API. Prefer controlling this via delegate options, e.g. tflite::StatefulNnApiDelegate::Options::allow_fp16' orTfLiteGpuDelegateOptionsV2::is_precision_loss_allowed`.
    • GPU
      • GPU acceleration now supports quantized models by default
    • DynamicBuffer::AddJoinedString() will now add a separator if the first string to be joined is empty.
    • βž• Adds support for cumulative sum (cumsum), both as builtin op and MLIR conversion.

    TensorRT

    • Issues a warning when the session_config parameter for the TF1 converter is used or the rewrite_config_template field in the TF2
      converter parameter object is used.

    TPU Enhancements:

    • βž• Adds support for the beta parameter of the FTRL optimizer for TPU embeddings. Users of other TensorFlow platforms can implement equivalent
      behavior by adjusting the l2 parameter.

    πŸ‘ XLA Support:

    • πŸ—„ xla.experimental.compile is deprecated, use tf.function(experimental_compile=True) instead.
    • Adds tf.function.experimental_get_compiler_ir which returns compiler IR (currently 'hlo' and 'optimized_hlo') for given input for given function.

    πŸ”’ Security:

    Other:

    • πŸ’… We have replaced uses of "whitelist" and "blacklist" with "allowlist" and "denylist" where possible. Please see this list for more context.
    • Adds tf.config.experimental.mlir_bridge_rollout which will help us rollout the new MLIR TPU bridge.
    • Adds tf.experimental.register_filesystem_plugin to load modular filesystem plugins from Python

    Thanks to our Contributors

    πŸš€ This release contains contributions from many people at Google and external contributors.

    8bitmp3, aaa.jq, Abhineet Choudhary, Abolfazl Shahbazi, acxz, Adam Hillier, Adrian Garcia Badaracco, Ag Ramesh, ahmedsabie, Alan Anderson, Alexander Grund, Alexandre Lissy, Alexey Ivanov, Amedeo Cavallo, anencore94, Aniket Kumar Singh, Anthony Platanios, Ashwin Phadke, Balint Cristian, Basit Ayantunde, bbbboom, Ben Barsdell, Benjamin Chetioui, Benjamin Peterson, bhack, Bhanu Prakash Bandaru Venkata, Biagio Montaruli, Brent M. Spell, bubblebooy, bzhao, cfRod, Cheng Chen, Cheng(Kit) Chen, Chris Tessum, Christian, chuanqiw, codeadmin_peritiae, COTASPAR, CuiYifeng, danielknobe, danielyou0230, dannyfriar, daria, DarrenZhang01, Denisa Roberts, dependabot[bot], Deven Desai, Dmitry Volodin, Dmitry Zakharov, drebain, Duncan Riach, Eduard Feicho, Ehsan Toosi, Elena Zhelezina, emlaprise2358, Eugene Kuznetsov, Evaderan-Lab, Evgeniy Polyakov, Fausto Morales, Felix Johnny, fo40225, Frederic Bastien, Fredrik Knutsson, fsx950223, Gaurav Singh, Gauri1 Deshpande, George Grzegorz Pawelczak, gerbauz, Gianluca Baratti, Giorgio Arena, Gmc2, Guozhong Zhuang, Hannes Achleitner, Harirai, HarisWang, Harsh188, hedgehog91, Hemal Mamtora, Hideto Ueno, Hugh Ku, Ian Beauregard, Ilya Persky, jacco, Jakub BerΓ‘nek, Jan Jongboom, Javier Montalt Tordera, Jens Elofsson, Jerry Shih, jerryyin, jgehw, Jinjing Zhou, jma, jmsmdy, Johan NordstrΓΆm, John Poole, Jonah Kohn, Jonathan Dekhtiar, jpodivin, Jung Daun, Kai Katsumata, Kaixi Hou, Kamil Rakoczy, Kaustubh Maske Patil, Kazuaki Ishizaki, Kedar Sovani, Koan-Sin Tan, Koki Ibukuro, Krzysztof Laskowski, Kushagra Sharma, Kushan Ahmadian, Lakshay Tokas, Leicong Li, levinxo, Lukas Geiger, Maderator, Mahmoud Abuzaina, Mao Yunfei, Marius Brehler, markf, Martin Hwasser, Martin Kubovčík, Matt Conley, Matthias, mazharul, mdfaijul, Michael137, MichelBr, Mikhail Startsev, Milan Straka, Ml-0, Myung-Hyun Kim, MΓ₯ns Nilsson, Nathan Luehr, ngc92, nikochiko, Niranjan Hasabnis, nyagato_00, Oceania2018, Oleg Guba, Ongun Kanat, OscarVanL, Patrik Laurell, Paul Tanger, Peter Sobot, Phil Pearl, PlusPlusUltra, Poedator, Prasad Nikam, Rahul-Kamat, Rajeshwar Reddy T, redwrasse, Rickard, Robert Szczepanski, Rohan Lekhwani, Sam Holt, Sami Kama, Samuel Holt, Sandeep Giri, sboshin, Sean Settle, settle, Sharada Shiddibhavi, Shawn Presser, ShengYang1, Shi,Guangyong, Shuxiang Gao, Sicong Li, Sidong-Wei, Srihari Humbarwadi, Srinivasan Narayanamoorthy, Steenu Johnson, Steven Clarkson, stjohnso98, Tamas Bela Feher, Tamas Nyiri, Tarandeep Singh, Teng Lu, Thibaut Goetghebuer-Planchon, Tim Bradley, Tomasz Strejczek, Tongzhou Wang, Torsten Rudolf, Trent Lo, Ty Mick, Tzu-Wei Sung, Varghese, Jojimon, Vignesh Kothapalli, Vishakha Agrawal, Vividha, Vladimir Menshakov, Vladimir Silyaev, VoVAllen, VΓ΅ VΔƒn NghΔ©a, wondertx, xiaohong1031, Xiaoming (Jason) Cui, Xinan Jiang, Yair Ehrenwald, Yasir Modak, Yasuhiro Matsumoto, Yimei Sun, Yiwen Li, Yixing, Yoav Ramon, Yong Tang, Yong Wu, yuanbopeng, Yunmo Koo, Zhangqiang, Zhou Peng, ZhuBaohe, zilinzhu, zmx


Previous changes from v2.4.0-rc3

  • πŸš€ Release 2.4.0

    Major Features and Improvements

    πŸ‘€ tf.distribute introduces experimental support for asynchronous training of Keras models via the tf.distribute.experimental.ParameterServerStrategy API. Please see below for additional details.

    πŸ“„ MultiWorkerMirroredStrategy is now a stable API and is no longer considered experimental. Some of the major improvements involve handling peer failure and many bug fixes. Please check out the detailed tutorial on Multi-worker training with Keras.

    πŸ“„ Introduces experimental support for a new module named tf.experimental.numpy which is a NumPy-compatible API for writing TF programs. See the detailed guide to learn more. Additional details below.

    βž• Adds Support for
    πŸ”Š TensorFloat-32 on Ampere based GPUs. TensorFloat-32, or TF32 for short, is a math mode for NVIDIA Ampere based GPUs and is enabled by default.

    🐎 A major refactoring of the internals of the Keras Functional API has been completed, that should improve the reliability, stability, and performance of constructing Functional models.

    Keras mixed precision API tf.keras.mixed_precision is no longer experimental and allows the use of 16-bit floating point formats during training, improving performance by up to 3x on GPUs and 60% on TPUs. Please see below for additional details.

    πŸ‘· TensorFlow Profiler now supports profiling MultiWorkerMirroredStrategy and tracing multiple workers using the sampling mode API.

    TFLite Profiler for Android is available. See the detailed guide to learn more.

    πŸ“¦ TensorFlow pip packages are now built with CUDA11 and cuDNN 8.0.2.

    πŸ’₯ Breaking Changes

    TF Core:

    • Certain float32 ops run in lower precsion on Ampere based GPUs, including matmuls and convolutions, due to the use of TensorFloat-32. Specifically, inputs to such ops are rounded from 23 bits of precision to 10
      bits of precision. This is unlikely to cause issues in practice for deep learning models. In some cases, TensorFloat-32 is also used for complex64 ops.
      TensorFloat-32 can be disabled by running tf.config.experimental.enable_tensor_float_32_execution(False).
    • The byte layout for string tensors across the C-API has been updated to match TF Core/C++; i.e., a contiguous array of tensorflow::tstring/TF_TStrings.
    • C-API functions TF_StringDecode, TF_StringEncode, and TF_StringEncodedSize are no longer relevant and have been removed; see core/platform/ctstring.h for string access/modification in C.
    • tensorflow.python, tensorflow.core and tensorflow.compiler modules are now hidden. These modules are not part of TensorFlow public API.
    • tf.raw_ops.Max and tf.raw_ops.Min no longer accept inputs of type tf.complex64 or tf.complex128, because the behavior of these ops is not well defined for complex types.

    - XLA:CPU and XLA:GPU devices are no longer registered by default. Use TF_XLA_FLAGS=--tf_xla_enable_xla_devices if you really need them, but this flag will eventually be removed in subsequent releases.

    tf.keras:

    • The steps_per_execution argument in compile() is no longer experimental; if you were passing experimental_steps_per_execution, rename it to steps_per_execution in your code. This argument controls the number of batches to run during each tf.function call when calling fit(). Running multiple batches inside a single tf.function call can greatly improve performance on TPUs or small models with a large Python overhead.
    • A major refactoring of the internals of the Keras Functional API may affect code that
      is relying on certain internal details:
      • Code that uses isinstance(x, tf.Tensor) instead of tf.is_tensor when checking Keras symbolic inputs/outputs should switch to using tf.is_tensor.
      • Code that is overly dependent on the exact names attached to symbolic tensors (e.g. assumes there will be ":0" at the end of the inputs, treats names as unique identifiers instead of using tensor.ref(), etc.)
      • Code that uses get_concrete_function to trace Keras symbolic inputs directly should switch to building matching tf.TensorSpecs directly and tracing the TensorSpec objects.
      • Code that relies on the exact number and names of the op layers that TensorFlow operations were converted into may have changed.
      • Code that uses tf.map_fn/tf.cond/tf.while_loop/control flow as op layers and happens to work before TF 2.4. These will explicitly be unsupported now. Converting these ops to Functional API op layers was unreliable before TF 2.4, and prone to erroring incomprehensibly or being silently buggy.
      • Code that directly asserts on a Keras symbolic value in cases where ops like tf.rank used to return a static or symbolic value depending on if the input had a fully static shape or not. Now these ops always return symbolic values.
      • Code already susceptible to leaking tensors outside of graphs becomes slightly more likely to do so now.
      • Code that tries directly getting gradients with respect to symbolic Keras inputs/outputs. Use GradientTape on the actual Tensors passed to the already- constructed model instead.
      • Code that requires very tricky shape manipulation via converted op layers in order to work, where the Keras symbolic shape inference proves insufficient.
      • Code that tries manually walking a tf.keras.Model layer by layer and assumes layers only ever have one positional argument. This assumption doesn't hold true before TF 2.4 either, but is more likely to cause issues now.
      • Code that manually enters keras.backend.get_graph() before building a functional model is no longer needed.
      • Start enforcing input shape assumptions when calling Functional API Keras models. This may potentially break some users, in case there is a mismatch between the shape used when creating Input objects in a Functional model, and the shape of the data passed to that model. You can fix this mismatch by either calling the model with correctly-shaped data, or by relaxing Input shape assumptions (note that you can pass shapes with None entries for axes
        that are meant to be dynamic). You can also disable the input checking entirely by setting model.input_spec = None.
    • Serveral changes have been made to tf.keras.mixed_precision.experimental. Note that it is now recommended to use the non-experimental tf.keras.mixed_precision API.
      • AutoCastVariable.dtype now refers to the actual variable dtype, not the dtype it will be casted to.
      • When mixed precision is enabled, tf.keras.layers.Embedding now outputs a float16 or bfloat16 tensor instead of a float32 tensor.
      • The property tf.keras.mixed_precision.experimental.LossScaleOptimizer.loss_scale is
        ⚑️ now a tensor, not a LossScale object. This means to get a loss scale of a LossScaleOptimizer as a tensor, you must now call
        opt.loss_scale instead of opt.loss_scale().
      • The property should_cast_variables has been removed from tf.keras.mixed_precision.experimental.Policy
      • When passing a tf.mixed_precision.experimental.DynamicLossScale to
        ⚑️ tf.keras.mixed_precision.experimental.LossScaleOptimizer, the DynamicLossScale's multiplier must be 2.
      • When passing a tf.mixed_precision.experimental.DynamicLossScale to tf.keras.mixed_precision.experimental.LossScaleOptimizer,
        ⚑️ the weights of the DynanmicLossScale are copied into the LossScaleOptimizer instead of being reused. This means modifying the
        ⚑️ weights of the DynamicLossScale will no longer affect the weights of the LossScaleOptimizer, and vice versa.
      • The global policy can no longer be set to a non-floating point policy in tf.keras.mixed_precision.experimental.set_policy
      • In Layer.call, AutoCastVariables will no longer be casted within MirroredStrategy.run or ReplicaContext.merge_call. This is
        because a thread local variable is used to determine whether AutoCastVariables are casted, and those two functions run with a
        different thread. Note this only applies if one of these two functions is called within Layer.call; if one of those two functions calls Layer.call, AutoCastVariables will still be casted.

    tf.data:

    • tf.data.experimental.service.DispatchServer now takes a config tuple instead of individual arguments. Usages should be updated to tf.data.experimental.service.DispatchServer(dispatcher_config).

    - tf.data.experimental.service.WorkerServer now takes a config tuple instead of individual arguments. Usages should be updated to tf.data.experimental.service.WorkerServer(worker_config).

    tf.distribute:

    • Removes tf.distribute.Strategy.experimental_make_numpy_dataset. Please use tf.data.Dataset.from_tensor_slices instead.
    • Renames experimental_hints in tf.distribute.StrategyExtended.reduce_to, tf.distribute.StrategyExtended.batch_reduce_to, tf.distribute.ReplicaContext.all_reduce to options:
    • Renames tf.distribute.experimental.CollectiveHints to tf.distribute.experimental.CommunicationOptions.
    • Renames tf.distribute.experimental.CollectiveCommunication to tf.distribute.experimental.CommunicationImplementation.
    • Renames tf.distribute.Strategy.experimental_distribute_datasets_from_function to distribute_datasets_from_function as it is no longer experimental.

    - Removes tf.distribute.Strategy.experimental_run_v2 method, which was deprecated in TF 2.2.

    tf.lite:

    • tf.quantization.quantize_and_dequantize_v2 has been introduced, which updates the gradient definition for quantization which is outside the range
      to be 0. To simulate the V1 the behavior of tf.quantization.quantize_and_dequantize(...) use tf.grad_pass_through(tf.quantization.quantize_and_dequantize_v2)(...).

    πŸ› Bug Fixes and Other Changes

    TF Core:

    • πŸ‘ Introduces experimental support for a new module named tf.experimental.numpy, which
      is a NumPy-compatible API for writing TF programs. This module provides class ndarray, which mimics the ndarray class in NumPy, and wraps an immutable tf.Tensor under the hood. A subset of NumPy functions (e.g. numpy.add) are provided. Their inter-operation with TF facilities is seamless in most cases.
      See tensorflow/python/ops/numpy_ops/README.md
      πŸ‘ for details of what operations are supported and what are the differences from NumPy.
    • tf.types.experimental.TensorLike is a new Union type that can be used as type annotation for variables representing a Tensor or a value
      that can be converted to Tensor by tf.convert_to_tensor.
    • Calling ops with a python constants or numpy values is now consistent with tf.convert_to_tensor behavior. This avoids operations like
      tf.reshape truncating inputs such as from int64 to int32.
    • βž• Adds tf.sparse.map_values to apply a function to the .values of SparseTensor arguments.
    • The Python bitwise operators for Tensor (__and__, __or__, __xor__ and __invert__ now support non-bool arguments and apply
      πŸ‘ the corresponding bitwise ops. bool arguments continue to be supported and dispatch to logical ops. This brings them more in line with
      Python and NumPy behavior.
    • βž• Adds tf.SparseTensor.with_values. This returns a new SparseTensor with the same sparsity pattern, but with new provided values. It is
      similar to the with_values function of RaggedTensor.
    • βž• Adds StatelessCase op, and uses it if none of case branches has stateful ops.
    • Adds tf.config.experimental.get_memory_usage to return total memory usage of the device.
    • βž• Adds gradients for RaggedTensorToVariant and RaggedTensorFromVariant.
    • πŸ‘Œ Improve shape inference of nested function calls by supporting constant folding across Arg nodes which makes more static values available to shape inference functions.
    • tf.debugging:
      • tf.debugging.assert_shapes() now works on SparseTensors (Fixes #36268).
    • GPU
      • Adds Support for TensorFloat-32 on Ampere based GPUs.
        TensorFloat-32, or TF32 for short, is a math mode for NVIDIA Ampere based GPUs which causes certain float32 ops, such as matrix
        multiplications and convolutions, to run much faster on Ampere GPUs but with reduced precision. This reduced precision has not been found
        to effect convergence quality of deep learning models in practice. TensorFloat-32 is enabled by default, but can be disabled with tf.config.experimental.enable_tensor_float_32_execution.
    • tf.math:
      • Adds tf.math.erfcinv, the inverse to tf.math.erfc.
    • tf.nn:
      • tf.nn.max_pool2d now supports explicit padding.
    • tf.image:
      • Adds deterministic tf.image.stateless_random_* functions for each tf.image.random_* function. Added a new op stateless_sample_distorted_bounding_box which is a deterministic version of sample_distorted_bounding_box op. Given the same seed, these stateless functions/ops produce the same results independent of how many times the function is called, and independent of global seed settings.
      • Adds deterministic tf.image.resize backprop CUDA kernels for method=ResizeMethod.BILINEAR (the default method). Enable by setting the environment variable TF_DETERMINISTIC_OPS to "true" or "1".
    • πŸ–¨ tf.print:
      • Bug fix in tf.print() with OrderedDict where if an OrderedDict didn't have the keys sorted, the keys and values were not being printed
        in accordance with their correct mapping.
    • tf.train.Checkpoint:
      • Now accepts a root argument in the initialization, which generates a checkpoint with a root object. This allows users to create a Checkpoint object that is compatible with Keras model.save_weights() and model.load_weights. The checkpoint is also compatible with the checkpoint saved in the variables/ folder in the SavedModel.
      • When restoring, save_path can be a path to a SavedModel. The function will automatically find the checkpoint in the SavedModel.

    tf.data:

    • Adds new tf.data.experimental.service.register_dataset and tf.data.experimental.service.from_dataset_id APIs to enable one
      πŸ–¨ process to register a dataset with the tf.data service, and another process to consume data from the dataset.
    • βž• Adds support for dispatcher fault tolerance. To enable fault tolerance, configure a work_dir when running your dispatcher server and set
      dispatcher_fault_tolerance=True. The dispatcher will store its state to work_dir, so that on restart it can continue from its previous
      state after restart.
    • βž• Adds support for sharing dataset graphs via shared filesystem instead of over RPC. This reduces load on the dispatcher, improving performance
      πŸ‘· of distributing datasets. For this to work, the dispatcher's work_dir must be accessible from workers. If the worker fails to read from the
      work_dir, it falls back to using RPC for dataset graph transfer.
    • βž• Adds support for a new "distributed_epoch" processing mode. This processing mode distributes a dataset across all tf.data workers,
      πŸ“„ instead of having each worker process the full dataset. See the tf.data service docs to learn more.
    • Adds optional exclude_cols parameter to CsvDataset. This parameter is the complement of select_cols; at most one of these should be specified.
    • We have implemented an optimization which reorders data-discarding transformations such as take and shard to happen earlier in the dataset when it is safe to do so. The optimization can be disabled via the experimental_optimization.reorder_data_discarding_ops dataset option.
    • tf.data.Options were previously immutable and can now be overridden.
    • πŸ“œ tf.data.Dataset.from_generator now supports Ragged and Sparse tensors with a new output_signature argument, which allows from_generator to
      produce any type describable by a tf.TypeSpec.
    • tf.data.experimental.AUTOTUNE is now available in the core API as tf.data.AUTOTUNE.

    tf.distribute:

    • πŸ‘ Introduces experimental support for asynchronous training of Keras models via tf.distribute.experimental.ParameterServerStrategy:
      • Replaces the existing tf.distribute.experimental.ParameterServerStrategy symbol with a new class that is for parameter server training in TF2. Usage of
        the old symbol, usually with Estimator API, should be replaced with [tf.compat.v1.distribute.experimental.ParameterServerStrategy].
      • Added tf.distribute.experimental.coordinator.* namespace, including the main API ClusterCoordinator for coordinating the training cluster, the related data structure RemoteValue and PerWorkerValue.
    • βž• Adds tf.distribute.Strategy.gather and tf.distribute.ReplicaContext.all_gather APIs to support gathering dense distributed values.
    • πŸ›  Fixes various issues with saving a distributed model.

    tf.keras:

    • πŸ‘Œ Improvements from the Functional API refactoring:
      • Functional model construction does not need to maintain a global workspace graph, removing memory leaks especially when building many
        models or very large models.
      • Functional model construction should be ~8-10% faster on average.
      • Functional models can now contain non-symbolic values in their call inputs inside of the first positional argument.
      • Several classes of TF ops that were not reliably converted to Keras layers during functional API construction should now work, e.g.
        tf.image.ssim_multiscale
      • Error messages when Functional API construction goes wrong (and when ops cannot be converted to Keras layers automatically) should be
        clearer and easier to understand.
    • ⚑️ Optimizer.minimize can now accept a loss Tensor and a GradientTape as an alternative to accepting a callable loss.
    • βž• Adds beta hyperparameter to FTRL optimizer classes (Keras and others) to match FTRL paper.
    • Optimizer. __init__ now accepts a gradient_aggregator to allow for customization of how gradients are aggregated across devices, as well as
      gradients_transformers to allow for custom gradient transformations (such as gradient clipping).
    • πŸ‘Œ Improvements to Keras preprocessing layers:
      • TextVectorization can now accept a vocabulary list or file as an init arg.
      • Normalization can now accept mean and variance values as init args.
    • In Attention and AdditiveAttention layers, the call() method now accepts a return_attention_scores argument. When set to
      True, the layer returns the attention scores as an additional output argument.
    • βž• Adds tf.metrics.log_cosh and tf.metrics.logcosh API entrypoints with the same implementation as their tf.losses equivalent.
    • For Keras model, the individual call of Model.evaluate uses no cached data for evaluation, while Model.fit uses cached data when
      🐎 validation_data arg is provided for better performance.
    • Adds a save_traces argument to model.save/ tf.keras.models.save_model which determines whether the SavedModel format stores the Keras model/layer call functions. The traced functions allow Keras to revive custom models and layers without the original class definition, but if this isn't required the tracing can be disabled with the added option.
    • The tf.keras.mixed_precision API is non non-experimental. The
      non-experimental API differs from the experimental API in several ways.
      • tf.keras.mixed_precision.Policy no longer takes in a
        tf.mixed_precision.experimental.LossScale in the constructor, and no
        longer has a LossScale associated with it. Instead, Model.compile
        ⚑️ will automatically wrap the optimizer with a LossScaleOptimizer using
        dynamic loss scaling if Policy.name is "mixed_float16".
      • tf.keras.mixed_precision.LossScaleOptimizer's constructor takes in
        different arguments. In particular, it no longer takes in a LossScale,
        and there is no longer a LossScale associated with the
        ⚑️ LossScaleOptimizer. Instead, LossScaleOptimizer directly implements
        πŸ›  fixed or dynamic loss scaling. See the documentation of
        ⚑️ tf.keras.mixed_precision.experimental.LossScaleOptimizer
        for details on the differences between the experimental
        ⚑️ LossScaleOptimizer and the new non-experimental LossScaleOptimizer.
      • tf.mixed_precision.experimental.LossScale and its subclasses are
        πŸ—„ deprecated, as all of its functionality now exists within
        ⚑️ tf.keras.mixed_precision.LossScaleOptimizer

    tf.lite:

    • TFLiteConverter:
      • Support optional flags inference_input_type and inference_output_type for full integer quantized models. This allows users to modify the model input and output type to integer types (tf.int8, tf.uint8) instead of defaulting to float type (tf.float32).
    • NNAPI
      • Adds NNAPI Delegation support for requantization use cases by converting the operation into a dequantize-quantize pair.
      • Removes deprecated Interpreter.setUseNNAPI(boolean) Java API. Use Interpreter.Options.setUseNNAPI instead.
      • Deprecates Interpreter::UseNNAPI(bool) C++ API. Use NnApiDelegate() and related delegate configuration methods directly.
      • Deprecates Interpreter::SetAllowFp16PrecisionForFp32(bool) C++ API. Prefer controlling this via delegate options, e.g. tflite::StatefulNnApiDelegate::Options::allow_fp16' orTfLiteGpuDelegateOptionsV2::is_precision_loss_allowed`.
    • GPU
      • GPU acceleration now supports quantized models by default
    • DynamicBuffer::AddJoinedString() will now add a separator if the first string to be joined is empty.
    • βž• Adds support for cumulative sum (cumsum), both as builtin op and MLIR conversion.

    TensorRT

    • Issues a warning when the session_config parameter for the TF1 converter is used or the rewrite_config_template field in the TF2
      converter parameter object is used.

    TPU Enhancements:

    • βž• Adds support for the beta parameter of the FTRL optimizer for TPU embeddings. Users of other TensorFlow platforms can implement equivalent
      behavior by adjusting the l2 parameter.

    πŸ‘ XLA Support:

    • πŸ—„ xla.experimental.compile is deprecated, use tf.function(experimental_compile=True) instead.
    • Adds tf.function.experimental_get_compiler_ir which returns compiler IR (currently 'hlo' and 'optimized_hlo') for given input for given function.

    πŸ”’ Security:

    Other:

    • πŸ’… We have replaced uses of "whitelist" and "blacklist" with "allowlist" and "denylist" where possible. Please see this list for more context.
    • Adds tf.config.experimental.mlir_bridge_rollout which will help us rollout the new MLIR TPU bridge.
    • Adds tf.experimental.register_filesystem_plugin to load modular filesystem plugins from Python

    Thanks to our Contributors

    πŸš€ This release contains contributions from many people at Google and external contributors.

    8bitmp3, aaa.jq, Abhineet Choudhary, Abolfazl Shahbazi, acxz, Adam Hillier, Adrian Garcia Badaracco, Ag Ramesh, ahmedsabie, Alan Anderson, Alexander Grund, Alexandre Lissy, Alexey Ivanov, Amedeo Cavallo, anencore94, Aniket Kumar Singh, Anthony Platanios, Ashwin Phadke, Balint Cristian, Basit Ayantunde, bbbboom, Ben Barsdell, Benjamin Chetioui, Benjamin Peterson, bhack, Bhanu Prakash Bandaru Venkata, Biagio Montaruli, Brent M. Spell, bubblebooy, bzhao, cfRod, Cheng Chen, Cheng(Kit) Chen, Chris Tessum, Christian, chuanqiw, codeadmin_peritiae, COTASPAR, CuiYifeng, danielknobe, danielyou0230, dannyfriar, daria, DarrenZhang01, Denisa Roberts, dependabot[bot], Deven Desai, Dmitry Volodin, Dmitry Zakharov, drebain, Duncan Riach, Eduard Feicho, Ehsan Toosi, Elena Zhelezina, emlaprise2358, Eugene Kuznetsov, Evaderan-Lab, Evgeniy Polyakov, Fausto Morales, Felix Johnny, fo40225, Frederic Bastien, Fredrik Knutsson, fsx950223, Gaurav Singh, Gauri1 Deshpande, George Grzegorz Pawelczak, gerbauz, Gianluca Baratti, Giorgio Arena, Gmc2, Guozhong Zhuang, Hannes Achleitner, Harirai, HarisWang, Harsh188, hedgehog91, Hemal Mamtora, Hideto Ueno, Hugh Ku, Ian Beauregard, Ilya Persky, jacco, Jakub BerΓ‘nek, Jan Jongboom, Javier Montalt Tordera, Jens Elofsson, Jerry Shih, jerryyin, jgehw, Jinjing Zhou, jma, jmsmdy, Johan NordstrΓΆm, John Poole, Jonah Kohn, Jonathan Dekhtiar, jpodivin, Jung Daun, Kai Katsumata, Kaixi Hou, Kamil Rakoczy, Kaustubh Maske Patil, Kazuaki Ishizaki, Kedar Sovani, Koan-Sin Tan, Koki Ibukuro, Krzysztof Laskowski, Kushagra Sharma, Kushan Ahmadian, Lakshay Tokas, Leicong Li, levinxo, Lukas Geiger, Maderator, Mahmoud Abuzaina, Mao Yunfei, Marius Brehler, markf, Martin Hwasser, Martin Kubovčík, Matt Conley, Matthias, mazharul, mdfaijul, Michael137, MichelBr, Mikhail Startsev, Milan Straka, Ml-0, Myung-Hyun Kim, MΓ₯ns Nilsson, Nathan Luehr, ngc92, nikochiko, Niranjan Hasabnis, nyagato_00, Oceania2018, Oleg Guba, Ongun Kanat, OscarVanL, Patrik Laurell, Paul Tanger, Peter Sobot, Phil Pearl, PlusPlusUltra, Poedator, Prasad Nikam, Rahul-Kamat, Rajeshwar Reddy T, redwrasse, Rickard, Robert Szczepanski, Rohan Lekhwani, Sam Holt, Sami Kama, Samuel Holt, Sandeep Giri, sboshin, Sean Settle, settle, Sharada Shiddibhavi, Shawn Presser, ShengYang1, Shi,Guangyong, Shuxiang Gao, Sicong Li, Sidong-Wei, Srihari Humbarwadi, Srinivasan Narayanamoorthy, Steenu Johnson, Steven Clarkson, stjohnso98, Tamas Bela Feher, Tamas Nyiri, Tarandeep Singh, Teng Lu, Thibaut Goetghebuer-Planchon, Tim Bradley, Tomasz Strejczek, Tongzhou Wang, Torsten Rudolf, Trent Lo, Ty Mick, Tzu-Wei Sung, Varghese, Jojimon, Vignesh Kothapalli, Vishakha Agrawal, Vividha, Vladimir Menshakov, Vladimir Silyaev, VoVAllen, VΓ΅ VΔƒn NghΔ©a, wondertx, xiaohong1031, Xiaoming (Jason) Cui, Xinan Jiang, Yair Ehrenwald, Yasir Modak, Yasuhiro Matsumoto, Yimei Sun, Yiwen Li, Yixing, Yoav Ramon, Yong Tang, Yong Wu, yuanbopeng, Yunmo Koo, Zhangqiang, Zhou Peng, ZhuBaohe, zilinzhu, zmx

Awesome Python is part of the LibHunt network. Terms. Privacy Policy.

(CC)
BY-SA
We recommend Spin The Wheel Of Names for a cryptographically secure random name picker.

AltStyle γ«γ‚ˆγ£γ¦ε€‰ζ›γ•γ‚ŒγŸγƒšγƒΌγ‚Έ (->γ‚ͺγƒͺγ‚ΈγƒŠγƒ«) /