机器学习库 TensorFlow 1.9.0 发布,常规更新版本

机器学习库 TensorFlow 1.9.0 发布,常规更新版本

达尔文
达尔文
发布于2018年07月12日
收藏 2

机器学习库 TensorFlow 1.9.0 已发布,该版本没有重大更新,只是一个常规更新版本。

主要功能和改进

  • Updated docs for tf.keras: New Keras-based get started and programmers guide page.

  • Update tf.keras to the Keras 2.1.6 API.

  • Added tf.keras.layers.CuDNNGRU and tf.keras.layers.CuDNNLSTM layers. Try it.

  • Adding support of core feature columns and losses to gradient boosted trees estimators.

  • The python interface for the TFLite Optimizing Converter has been expanded, and the command line interface (AKA: toco, tflite_convert) is once again included in the standard pip installation.

  • Improved data-loading and text processing with:

  • tf.decode_compressed

  • tf.string_strip

  • tf.strings.regex_full_match

  • Added experimental support for new pre-made Estimators:

  • tf.contrib.estimator.BaselineEstimator

  • tf.contrib.estimator.RNNClassifier

  • tf.contrib.estimator.RNNEstimator

  • The distributions.Bijector API supports broadcasting for Bijectors with new API changes.

可以看到,主要是升级和改进了一些 API,以及改进了文档。

破坏性变更

  • If you’re opening empty variable scopes; replace variable_scope(”, …) by variable_scope(tf.get_variable_scope(), …).

  • Headers used for building custom ops have been moved from site-packages/external into site-packages/tensorflow/include/external.

详情请查看 https://github.com/tensorflow/tensorflow/releases/tag/v1.9.0