TensorFlow 1.0.0 alpha

516 vues
Accéder directement au premier message non lu

Gunhan Gulsoy

non lue,
9 janv. 2017, 20:00:1809/01/2017
à dis...@tensorflow.org
Hi all,
Tensorflow 1.0.0-alpha version is now available!
This version means we are close to our 1.0 final release.
However, there might be still some breaking changes until we have the actual 1.0 final release ready. Therefore, the default documentation version under tensorflow.org and default packages provided will remain 0.12.1. You can access API documentation for 1.0 alpha version through: https://www.tensorflow.org/versions

Enjoy!


Major Features and Improvements
  • TensorFlow Debugger (tfdbg): command-line interface and API.
  • New python 3 docker images added.
  • Made pip packages pypi compliant. TensorFlow can now be installed by pip install tensorflow command.
  • Android: person detection + tracking demo implementing Scalable Object Detection using Deep Neural Networks.
  • Android: pre-built libs are now built nightly.
  • New (experimental) Java API.
Breaking Changes to the API
  • TensorFlow/models have been moved to a separate github repository.
  • Division and modulus operators (/, //, %) now match Python (flooring) semantics. This applies to tf.div and tf.mod as well. To obtain forced integer truncation based behaviors you can use tf.truncatediv and tf.truncatemod.
  • tf.divide() is now the recommended division function. tf.div() will remain, but its semantics do not respond to Python 3 or from future mechanisms.
  • tf.reverse() now takes indices of axes to be reversed. E.g. tf.reverse(a, [True, False, True]) must now be written as tf.reverse(a, [0, 2]). tf.reverse_v2() will remain until 1.0 final.
  • tf.mul, tf.sub and tf.neg are deprecated in favor of tf.multiply, tf.subtract and tf.negative.
  • tf.pack and tf.unpack are deprecated in favor of tf.stack and tf.unstack.
  • TensorArray.pack and TensorArray.unpack are getting deprecated in favor of TensorArray.stack and TensorArray.unstack.
  • The following Python functions have had their arguments changed to use axis when referring to specific dimensions. We have kept the old keyword arguments for compatibility currently, but we will be removing them well before the final 1.0.
  • tf.argmax: dimension becomes axis
  • tf.argmin: dimension becomes axis
  • tf.count_nonzero: reduction_indices becomes axis
  • tf.expand_dims: dim becomes axis
  • tf.reduce_all: reduction_indices becomes axis
  • tf.reduce_any: reduction_indices becomes axis
  • tf.reduce_join: reduction_indices becomes axis
  • tf.reduce_logsumexp: reduction_indices becomes axis
  • tf.reduce_max: reduction_indices becomes axis
  • tf.reduce_mean: reduction_indices becomes axis
  • tf.reduce_min: reduction_indices becomes axis
  • tf.reduce_prod: reduction_indices becomes axis
  • tf.reduce_sum: reduction_indices becomes axis
  • tf.reverse_sequence: batch_dim becomes batch_axis, seq_dim becomes seq_axis
  • tf.sparse_concat: concat_dim becomes axis
  • tf.sparse_reduce_sum: reduction_axes becomes axis
  • tf.sparse_reduce_sum_sparse: reduction_axes becomes axis
  • tf.sparse_split: split_dim becomes axis
  • tf.listdiff has been renamed to tf.setdiff1d to match NumPy naming.
  • tf.inv has been renamed to be tf.reciprocal (component-wise reciprocal) to avoid confusion with np.inv which is matrix inversion
  • tf.round now uses banker's rounding (round to even) semantics to match NumPy.
  • tf.split now takes arguments in a reversed order and with different keywords. In particular, we now match NumPy order as tf.split(value, num_or_size_splits, axis).
  • tf.sparse_split now takes arguments in reversed order and with different keywords. In particular we now match NumPy order as tf.sparse_split(sp_input, num_split, axis). NOTE: we have temporarily made tf.sparse_split require keyword arguments.
  • Deprecated tf.concat operator. Please switch to use tf.concat_v2 for now. In the Beta release, we will update tf.concat to match argument order of tf.concat_v2.
  • tf.image.decode_jpeg by default uses the faster DCT method, sacrificing a little fidelity for improved speed. One can revert to the old behavior by specifying the attribute dct_method='INTEGER_ACCURATE'.
  • tf.complex_abs has been removed from the Python interface. tf.abs supports complex tensors and should be used instead.
  • Template.var_scope property renamed to .variable_scope
  • SyncReplicasOptimizer is removed and SyncReplicasOptimizerV2 renamed to SyncReplicasOptimizer.
  • tf.zeros_initializer() and tf.ones_initializer() now return a callable that must be called with initializer arguments, in your code replace tf.zeros_initializer with tf.zeros_initializer().
  • SparseTensor.shape has been renamed to SparseTensor.dense_shape. Same for SparseTensorValue.shape.
  • Remove old tf summary ops, like tf.scalar_summary and tf.histogram_summary. Use tf.summary.scalar and tf.summary.histogram instead.
  • Remove tf.train.SummaryWriter and tf.train.SummaryWriterCache.
  • Removes RegisterShape from public API. Use C++ shape function registration instead.
  • Deprecated _ref dtypes from the python API.
Bug Fixes and Other Changes
  • New op: parallel_stack.
  • Introducing common tf io compression options constants for RecordReader/RecordWriter.
  • Add sparse_column_with_vocabulary_file, to specify a feature column that transform string features to IDs, where the mapping is defined by a vocabulary file.
  • Added index_to_string_table which returns a lookup table that maps indices to strings.
  • Add string_to_index_table, which returns a lookup table that matches strings to indices.
  • Add a ParallelForWithWorkerId function.
  • Add string_to_index_table, which returns a lookup table that matches strings to indices.
  • Support restore session from checkpoint files in v2 in contrib/session_bundle.
  • Added a tf.contrib.image.rotate function for arbitrary angles.
  • Added tf.contrib.framework.filter_variables as a convenience function to filter lists of variables based on regular expressions.
  • Remove old tf summary ops, like tf.scalar_summary and tf.histogram_summary. Use tf.summary.scalar and tf.summary.histogram instead.
  • make_template() takes an optional custom_getter_ param.
  • Added comment about how existing directories are handled by recursive_create_dir.
  • Added an op for QR factorizations.
  • Divides and mods in Python API now use flooring (Python) semantics.
  • Android: cmake/gradle build for TensorFlow Inference library under contrib/android/cmake
  • Android: Much more robust Session initialization code.
  • Android: TF stats now exposed directly in demo and log when debug mode is active
  • Android: new/better README.md documentation

Tyler Renelle

non lue,
23 janv. 2017, 10:29:0823/01/2017
à Discuss,gu...@google.com
Most despised question, but - any inkling on release date? As in "some months" to "a year or more"? 

Martin Wicke

non lue,
23 janv. 2017, 11:16:3723/01/2017
à Discuss,Tyler Renelle,gu...@google.com
Not that long. You should see RCs soon. 

--


You received this message because you are subscribed to the Google Groups "Discuss" group.


To unsubscribe from this group and stop receiving emails from it, send an email to discuss+u...@tensorflow.org.


To post to this group, send email to dis...@tensorflow.org.


To view this discussion on the web visit https://groups.google.com/a/tensorflow.org/d/msgid/discuss/b14cb9b9-f02f-403a-9b21-cc85b2e96330%40tensorflow.org.


Norman Heckscher

non lue,
23 janv. 2017, 19:31:2023/01/2017
à Discuss
The last available roadmap for development in the source code is June 3, 2016.


Is there a current roadmap?

Martin Wicke

non lue,
2 févr. 2017, 19:46:5002/02/2017
à Norman Heckscher,Discuss
I updated the roadmap. Sorry for the outdatedness.

To unsubscribe from this group and stop receiving emails from it, send an email to discuss+unsubscribe@tensorflow.org.

To post to this group, send email to dis...@tensorflow.org.
To view this discussion on the web visit https://groups.google.com/a/tensorflow.org/d/msgid/discuss/71481e29-2aee-4cd7-b38c-2ed34bb43788%40tensorflow.org.

Norman Heckscher

non lue,
2 févr. 2017, 23:04:1502/02/2017
à Discuss
Well done. Thank you.
Looks like rc1 has just been released. :)
Le message a été supprimé

augusti...@gmail.com

non lue,
8 mai 2017, 07:35:2208/05/2017
à Discuss,gu...@google.com
I am heavily interested in TensorFlow library and I am a Java programmer.
I came to know from your site that TensorFlow provides APIs for use in Java programs.
However, the TensorFlow Java API is not covered by the TensorFlow API stability guarantees.

I would like to know do you have any plans at least in the future to release a fully fledged Java API?

Martin Wicke

non lue,
8 mai 2017, 12:13:2608/05/2017
à augusti...@gmail.com,Discuss,Gunhan Gulsoy
We have plans, but as we are still working on the Java API, we are not yet ready to make the same guarantees we are making for the python API. If API stability is a critical concern for you, you cannot yet use the Java API.

Martin

--
You received this message because you are subscribed to the Google Groups "Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to discuss+unsubscribe@tensorflow.org.

To post to this group, send email to dis...@tensorflow.org.
To view this discussion on the web visit https://groups.google.com/a/tensorflow.org/d/msgid/discuss/b43d3bdd-e6db-42e5-8de1-9142892c8d81%40tensorflow.org.

augusti...@gmail.com

non lue,
8 mai 2017, 23:43:4608/05/2017
à Discuss,augusti...@gmail.com,gu...@google.com
Thank you very much for the reply.
Hope we will get a fully fledged Java API sometime in the future.
Répondre à tous
Répondre à l'auteur
Transférer
0 nouveau message