- 18 7月, 2018 1 次提交
-
-
由 Gabriel de Marmiesse 提交于
Removed the exceptions concerning files which were already complient with the 85 chars rule. (#10714) ### Summary ### Related Issues ### PR Overview - [ ] This PR requires new unit tests [y/n] (make sure tests are included) - [ ] This PR requires to update the documentation [y/n] (make sure the docs are up-to-date) - [x] This PR is backwards compatible [y/n] - [ ] This PR changes the current API [y/n] (all API changes need to be approved by fchollet)
-
- 17 7月, 2018 4 次提交
-
-
由 Yong Zheng Xin 提交于
### Summary Modify `docs/autogen.py` so it complies to PEP8 line length rule (line length = 85). ### Related Issues ### PR Overview - [ ] This PR requires new unit tests [y/n] (make sure tests are included) - [ ] This PR requires to update the documentation [y/n] (make sure the docs are up-to-date) - [X] This PR is backwards compatible [y/n] - [ ] This PR changes the current API [y/n] (all API changes need to be approved by fchollet)
-
由 Yong Zheng Xin 提交于
-
由 Gabriel de Marmiesse 提交于
-
由 Chanran Kim 提交于
Update `Conv2D` call to the Keras 2 API
-
- 16 7月, 2018 1 次提交
-
-
由 Gabriel de Marmiesse 提交于
### Summary I think it was meant to be passed it to the `__init__` of the parent class. ### Related Issues ### PR Overview - [ ] This PR requires new unit tests [y/n] (make sure tests are included) - [ ] This PR requires to update the documentation [y/n] (make sure the docs are up-to-date) - [ ] This PR is backwards compatible [y/n] - [x] This PR changes the current API [y/n] (all API changes need to be approved by fchollet)
-
- 15 7月, 2018 1 次提交
-
-
由 Gabriel de Marmiesse 提交于
-
- 13 7月, 2018 1 次提交
-
-
由 Masoud Kazemi 提交于
### Summary This PR fixes the redundant function module names in auto-gen docs. Currently, the module name of functions (those which are not methods of a class) is repeated two times in documentation. [For example](https://keras.io/preprocessing/sequence/#pad_sequences): > `keras.preprocessing.sequence.keras.preprocessing.sequence.pad_sequences` which should be: `keras.preprocessing.sequence.pad_sequences`. Further, this PR also removes from the docs the (wrong) module name of methods of classes like `ImageDataGenerator`. For example, [this is for `apply_transform`](https://keras.io/preprocessing/image/#apply_transform): > `keras.preprocessing.image.apply_transform` which should simply be `apply_transform`, first because it is the method of a class (i.e. `ImageDataGenerator`) and second because the referenced module name is wrong (and this [has caused confusion](https://stackoverflow.com/q/51311062/2099607)). All of this were caused by the fact that `clean_module_name()` was not called on the `function.__module__` in `render_function` (therefore this mostly affects the docs for `keras.preprocessing` and `keras.applications`). Plus, the way `autogen.py` works, the module name is repeated two times in the signature, but for the non-method functions it should be removed only once (that's why I have added a condition). ### Related Issues #10658 #10662 ### PR Overview - [n] This PR requires new unit tests [y/n] (make sure tests are included) - [n] This PR requires to update the documentation [y/n] (make sure the docs are up-to-date) - [y] This PR is backwards compatible [y/n] - [n] This PR changes the current API [y/n] (all API changes need to be approved by fchollet)
-
- 12 7月, 2018 3 次提交
-
-
由 Yu-Yang Huang 提交于
* Fix nested sequential deferred build * Add test * Unify style
-
由 fuzzythecat 提交于
-
由 Gabriel de Marmiesse 提交于
Enable automatic shape inference when using Lambda layers and CNTK. Also fixed the test_no_grad that was catching ValueErrors that had nothing to do with not having a gradient. (#10647)
-
- 11 7月, 2018 1 次提交
-
-
由 fuzzythecat 提交于
### Summary `normalize_data_format` was being called twice in `_Pooling2D` ### Related Issues ### PR Overview - [n] This PR requires new unit tests [y/n] (make sure tests are included) - [n] This PR requires to update the documentation [y/n] (make sure the docs are up-to-date) - [y] This PR is backwards compatible [y/n] - [n] This PR changes the current API [y/n] (all API changes need to be approved by fchollet)
-
- 10 7月, 2018 1 次提交
-
-
由 Masoud Kazemi 提交于
### Summary This PR adds the missing examples to the list of examples in _examples/README.md_. ### PR Overview - [n] This PR requires new unit tests [y/n] (make sure tests are included) - [n] This PR requires to update the documentation [y/n] (make sure the docs are up-to-date) - [y] This PR is backwards compatible [y/n] - [n] This PR changes the current API [y/n] (all API changes need to be approved by fchollet)
-
- 09 7月, 2018 1 次提交
-
-
由 Yanbo Liang 提交于
-
- 07 7月, 2018 3 次提交
-
-
由 Alfredo Canziani 提交于
`keras.regularizers.l1_l2()` takes two arguments. The doc was misleading.
-
由 Edgar Liberis 提交于
In light of SuperConvergence [1], the user might want to increase the learning rate during training so saying 'reducing' is no longer appropriate. [1]: https://arxiv.org/abs/1708.07120
-
由 Keith Battocchi 提交于
* Apply mean and stddev to cntk backend's random_normal * Update backend_test.py Add test case for non-standard normal distribution parameters * Make binomial, uniform, and normal random variables vary and add tests * Fix formatting issues
-
- 06 7月, 2018 2 次提交
-
-
* Reuse validation enqueuer * Fix CI * Indent val_data correctly * Typo
-
由 Fariz Rahman 提交于
-
- 05 7月, 2018 1 次提交
-
-
* Add pr template * Make uppercase to be consistent * Change according to review
-
- 04 7月, 2018 1 次提交
-
-
由 fuzzythecat 提交于
* Remove unused relu6 * Remove doc for loading relu6
-
- 02 7月, 2018 2 次提交
-
-
由 Ramon Viñas 提交于
* DOC: Fix IMDB get_word_index documentation * DOC: Fix reuters get_word_index documentation
-
* Doc * Fix indent
-
- 01 7月, 2018 3 次提交
-
-
由 Yanbo Liang 提交于
* Fix Dense layer bias_add bug when ndim > 2. * Address comment
-
由 Taehoon Lee 提交于
* Add saving tests * Add regexps to verify that the expected error is being raised
-
由 Kirill Vechera 提交于
* ignore empty KERAS_BACKEND environment variable, refs #10516 * fixed spacing (PEP8), refs #10516
-
- 26 6月, 2018 2 次提交
-
-
由 Soumya Chatterjee 提交于
-
由 Taehoon Lee 提交于
-
- 25 6月, 2018 1 次提交
-
-
由 fuzzythecat 提交于
-
- 23 6月, 2018 2 次提交
-
-
由 eTomate 提交于
Changed border_mode='same' to padding='same' in the example to match it with the parameters of Conv2D.
-
由 Kamil A. Kaczmarek 提交于
-
- 22 6月, 2018 1 次提交
-
-
由 David Silva 提交于
* Add output shape argument to transposed conv. to resolve ambiguity * Remove deconv2d legacy preprocessor * Add unit tests for output shape in 2d/3d transpose * Replace output_shape with output_padding * Update docstrings * deconv_length handles output_padding=None * Unit tests * Check if stride is greater than output padding * Pull request revision commit 65670e69dcb9249f3c7d520167d2701edfa31986 Author: David Silva <davidtvs10@gmail.com> Date: Tue Jun 19 06:00:55 2018 +0100 Replace two spaces by br tag to force new line commit d7141937317634af23b602de63363d06e96804a8 Author: David Silva <davidtvs10@gmail.com> Date: Tue Jun 19 05:26:57 2018 +0100 Add output shape equation to docstrings commit 5f06a7879bd5679eb290e379b73ff68d5cfceda1 Author: David Silva <davidtvs10@gmail.com> Date: Sat Jun 16 17:26:20 2018 +0100 Fixes following @fchollet revision commit b4530ae775ec67b79db801b1e2a514a83ba980cc Merge: 6b99d23 5fcd832b Author: David Silva <davidtvs10@gmail.com> Date: Sat Jun 16 16:17:41 2018 +0100 Merge remote-tracking branch 'upstream/master' into trans_out_pad_dev commit 5fcd832b Author: Pavithra Vijay <psv@google.com> Date: Fri Jun 15 16:29:32 2018 -0700 Fix small doc error (#10453) commit 2ec486ba Author: Francois Chollet <francois.chollet@gmail.com> Date: Fri Jun 15 14:07:28 2018 -0700 Model input validation fix. commit ffa99d69 Author: Pavithra Vijay <psv@google.com> Date: Fri Jun 15 13:52:14 2018 -0700 Fix use_bias in some convolutional layers (#10444) * Fix use_bias in some convolutional layers * Fix Conv2DTranspose test commit 6aace305 Author: ebatuhankaynak <ebatuhankaynak@users.noreply.github.com> Date: Fri Jun 15 22:00:24 2018 +0300 Update convolutional.py (#10452) Changed "# now model.output_shape == (None, 20, 16. 64)" to "# now model.output_shape == (None, 20, 16, 64)" (There is a dot instead of comma). commit 2d183db0 Author: Sachin Abeywardana <sachinruk@users.noreply.github.com> Date: Fri Jun 15 04:19:45 2018 +1000 corrected batch norm implementation in cntk (#10427) commit 52f608cb Author: Taehoon Lee <me@taehoonlee.com> Date: Thu Jun 14 05:44:22 2018 +0900 Add tests for inputs set dynamically (#10416) commit a68c516e Author: Taehoon Lee <me@taehoonlee.com> Date: Thu Jun 14 03:43:43 2018 +0900 Update a CNTK version on Travis (#10419) commit 4f90f95f Author: ebatuhankaynak <ebatuhankaynak@users.noreply.github.com> Date: Wed Jun 13 21:39:47 2018 +0300 Update faq.md (#10422) Changed "Notes that the methods...." to "Note that the methods" commit 8e5b8533 Author: Taehoon Lee <me@taehoonlee.com> Date: Thu Jun 14 03:39:13 2018 +0900 Move `get_source_inputs` (#10415) commit a40f3355 Author: Max Pumperla <max.pumperla@googlemail.com> Date: Wed Jun 13 20:36:50 2018 +0200 [docs] more details for adagrad/delta, clarifying usage of rho (#10410) * some more meat for ada docs * incorporate suggestions * line breaks commit 29a22a8d Author: fuzzythecat <fuzzy0427@gmail.com> Date: Wed Jun 13 02:57:38 2018 +0900 Add missing named arguments in ImageDataGenerator in examples (#10389) commit 0237542f Author: EyeBool <diataha.personal@gmail.com> Date: Mon Jun 11 09:44:02 2018 -0700 Fixed a type: repeated 'in common'. (#10394) commit 3eab6103 Author: Francois Chollet <francois.chollet@gmail.com> Date: Thu Jun 7 16:09:58 2018 -0700 Preserve input shape data when serializing deferred-build Sequential models. commit db849f8e Author: jlopezpena <jlopezpena@users.noreply.github.com> Date: Thu Jun 7 19:54:51 2018 +0100 Doc update (#10376) A couple of variables are "used" in two examples without being defined. For consistency with other examples where auxiliary dimensions are defined, I think it would be better to explicitly assign them a value. I just used made up values, feel free to change to whatever makes more sense! commit 67e242d9 Author: Max Pumperla <max.pumperla@googlemail.com> Date: Thu Jun 7 20:54:23 2018 +0200 Add spatial dropout and 3D global pooling to docs (#10373) * spatial dropout in docs * 3d global pooling in docs commit 4e829712 Author: Francois Chollet <francois.chollet@gmail.com> Date: Wed Jun 6 16:17:29 2018 -0700 Update docs for 2.2.0. commit 632560d9 Author: Francois Chollet <francois.chollet@gmail.com> Date: Wed Jun 6 11:53:18 2018 -0700 Style fix commit bbf42834 Author: Bohumír Zámečník <bohumir.zamecnik@gmail.com> Date: Wed Jun 6 20:52:26 2018 +0200 CuDNN RNN layers nested in TimeDistributed are not converted when loading (#10357) * Add a unit test for CuDNNGRU conversion with TimeDistributed. * Extract duplicated function convert_model() to _convert_model_weights(). * #10356 Convert weights of CuDNN/plain RNN nested in TimeDistributed. Same case as for Bidirectional, except that in TimeDistributed there's only one nested layer instead of two. commit 13548e8b Author: Taehoon Lee <me@taehoonlee.com> Date: Thu Jun 7 03:51:37 2018 +0900 Add tests for inputs set dynamically (#10367) commit 56b255cc Author: Francois Chollet <francois.chollet@gmail.com> Date: Wed Jun 6 10:51:06 2018 -0700 Update to Keras Applications 1.0.2 (fixes NASNet issue). commit 513f7b1b Author: Taehoon Lee <me@taehoonlee.com> Date: Thu Jun 7 00:47:37 2018 +0900 Fix a version number (#10361) commit 9a58f7b2 Author: Francois Chollet <francois.chollet@gmail.com> Date: Tue Jun 5 17:40:44 2018 -0700 Prepare 2.2.0 release. commit 3f4ccbb1 Author: Francois Chollet <francois.chollet@gmail.com> Date: Tue Jun 5 17:30:48 2018 -0700 Add error message when calling `summary` on unbuilt subclassed models. commit 25d0193e Author: tiferet <tgazit@gmail.com> Date: Tue Jun 5 21:23:27 2018 +0300 Supporting channels_first data format with crossentropy losses (#9715) commit 1e80c1a2 Author: KuzMenachem <35711942+KuzMenachem@users.noreply.github.com> Date: Tue Jun 5 20:10:43 2018 +0200 Fix HDF5Matrix issue when working in conjunction with TimeSeriesGenerator (#10334) * Fix issue when working in conjunction with TimeSeriesGenerator The TimeSeriesGenerator class uses xrange through six which caused an IndexError * Add test * Add corresponding test * Fix for python3 * Simplified code * Fix indent * Fix test commit ce56322a Author: David Schwertfeger <david.schwertfeger@googlemail.com> Date: Tue Jun 5 00:15:25 2018 +0200 FIX: Tensorboard callback only supports logging Embeddings layer weights (#7766) * Embed layer-outputs rather than layer-weights in TensorBoard callback * Update docstring and allow multiple inputs * Fix tests * Renaming * Set learning phase * Compute embeddings in batches * Pass embedding data explicitly * Actually process embeddings in batches * Allow multiple inputs and validate input data * Add example * Delete utils.py * Revert uncorrectly resolved merge-conflict * Minor renaming * Add comment clarifying the design choice commit b2176482 Author: Tommi Koivisto <tkoivisto@users.noreply.github.com> Date: Mon Jun 4 20:26:41 2018 +0300 Add an advanced activation layer for ReLU (#10322) The max_value argument can not be used in a layer, except custom layer or Lambda. Hence, similarly to LeakyReLU or for example Softmax, this PR adds a layer for ReLU, enabling also a capped ReLU to be used. commit 1365ed5d Author: Taehoon Lee <me@taehoonlee.com> Date: Tue Jun 5 02:07:23 2018 +0900 Reduce tests for applications (#10346) * Reduce tests for applications * Make selection over all models random commit 52e3f983 Author: Anirudh Swaminathan <aniswami97@gmail.com> Date: Sun Jun 3 01:22:09 2018 +0800 Improve the docstring of Conv3DTranspose (#10342) * Add depth as third dimension in docstring of Conv3DTranspose in convolutional.py in keras.layers commit f89b3397 Author: Taehoon Lee <me@taehoonlee.com> Date: Sun Jun 3 02:21:06 2018 +0900 Move tests for applications (#10341) commit c9cb6087 Author: François Chollet <francois.chollet@gmail.com> Date: Fri Jun 1 23:53:49 2018 -0700 Split `applications` and `preprocessing` modules. (#10339) * Split `applications` and `preprocessing` modules. * Fix dependencies. commit a3664246 Author: Zhengping Che <Peterche1990@Hotmail.com> Date: Sat Jun 2 07:13:24 2018 +0800 Handle `mask` in `TimeDistributed` wrapper. (#10242) * equip TimeDistributed with mask and unspecified input length * fix bugs in theano. add test on timedistributed + masking * skip tests on cntk with multiple unspecified time lengths. * move static shape inference to theano_backend, add docstring, etc. * fix format commit 7365a99f Author: fuzzythecat <fuzzy0427@gmail.com> Date: Fri Jun 1 04:45:26 2018 +0900 Fix doc (#10327) Fixed doc commit fe066966 Author: fuzzythecat <fuzzy0427@gmail.com> Date: Wed May 30 13:49:44 2018 +0900 Add pooling options in MobileNetV2 (#10313) * Add pooling option * Add pooling test commit 315a80ae Author: Taehoon Lee <me@taehoonlee.com> Date: Wed May 30 01:43:37 2018 +0900 Improve docstrings of applications (#10310) commit 25283eac Author: Francois Chollet <francois.chollet@gmail.com> Date: Tue May 29 09:31:14 2018 -0700 Fix typo commit 0de2adf0 Author: fuzzythecat <fuzzy0427@gmail.com> Date: Tue May 29 13:58:18 2018 +0900 Fix doc (#10308) * Fix naming convention * Add missing doc commit 9bff5b1a Author: r-kellerm <34206121+r-kellerm@users.noreply.github.com> Date: Mon May 28 03:45:35 2018 +0300 Typo fix (#10293) commit 794f8143 Author: Botty Dimanov <bottydim@users.noreply.github.com> Date: Fri May 25 22:21:41 2018 +0100 Handle capitalised extensions in list_pictures (#10220) #10219 commit 84aa7b5f Author: Wang, Zhiming <zhiming.wang@intel.com> Date: Sat May 26 05:20:57 2018 +0800 Non training Batch Norm operator has bad performance for it running into tensorflow's non fused batch norm API (#10207) * When use tensorflow as backend, let batch norm run into fused batch norm as much as possible, which has better performance. fix issue: http://github.com/keras-team/keras/issues/10058 * In Tensorflow backend, let batch norm call to FusedBatchNorm only NHWC format, also gamma and beta are not None. Test result: test env: with Tensorflow(commit a543d9471047ca3f6881c87105fcbe2cdff9207d Date: Thu May 10 17:43:30 2018, local build), python3.4, centos7.4 test cases: "pytest ./tests/keras/layers/normalization_test.py" <all passed> "pytest ./tests" <keep same result as without this commit's modification on BN> * fix code sytle. * 1. Add axis parameter in backend's batch_normalization functions. 2. Refine the batch_normalization function in tensorflow backend, Let's it call to fused batch norm as much as possible. Thanks the coments from fchollet. * Trigger * 1. add default value -1 for parameter axis in batch_normalization function in backend. 2. fix some code style. Thanks the comments from fchollet. commit e6d21795 Author: Stanley Bileschi <bileschi@google.com> Date: Fri May 25 17:19:38 2018 -0400 Adds to and alphabetizes documentation of Layer base class. (#10282) * Alphabetizes and adds to layers doc. * Responding to @cais comments * fix spacing. Remove in(out)bound_nodes commit 5eecd55a Author: Amir Alavi <s.amir.alavi@gmail.com> Date: Thu May 24 16:54:20 2018 -0400 load_weights will fail if shape mismatch (#10266) Fix for #10265 commit d476ecc7 Author: jnphilipp <mail@jnphilipp.org> Date: Thu May 24 19:52:11 2018 +0200 Use count_params function for non_trainable_count. (#10280) commit a07d9f3f Author: Stefano Cappellini <contact@stefanocappellini.com> Date: Thu May 24 19:50:21 2018 +0200 Fixes automatic doc generation problem with nested lists. Adds a new test (#10212) * Fixes automatic doc generation problem with indented lists. Adds a new test * Some style fixes on doc automatic generation files * Fixes a bad space in convolutional_recurrent.py * Changes the test_doc_auto_generation in order to include a doc string taken from the codebase. Allows text lines following nested lists commit 5a48df22 Author: Clemens Tolboom <clemens@build2be.com> Date: Thu May 24 00:00:42 2018 +0200 Make Dot documentation inline with Concatenate (#10271) Doc expects a list containing 2 tensors. commit 20644772 Author: askskro <ak@aiascience.com> Date: Wed May 23 23:59:58 2018 +0200 Typo in docstring for softplus (#10277) Softplus docstring missing a parenthesis. commit c77267af Author: Brian Nemsick <brian.nemsick@gmail.com> Date: Wed May 23 11:54:36 2018 -0700 Remove deprecated model.model from engine/saving (#10275) commit 08c87366 Author: vkk800 <vkk800@users.noreply.github.com> Date: Tue May 22 23:03:51 2018 +0200 Refactor ImageDataGenerator (#10130) * Create get_random_transform and refactor * Fix style and add tests * Add more tests * Fix documentation error * Fix documentation style issue * add apply_affine_transform * document transformation dictionary * Doc style fix commit 25a8973d Author: Taehoon Lee <me@taehoonlee.com> Date: Wed May 23 04:18:46 2018 +0900 Increase test coverages by factorizing CNTK pads (#10259) commit abd02940 Author: brge17 <33430930+brge17@users.noreply.github.com> Date: Mon May 21 18:42:07 2018 -0700 Remove Sequential.model deprecation warning (#10256) * Remove Sequential.model deprecation warning * Remove dead line of code commit 2eed2a5d Author: Taehoon Lee <me@taehoonlee.com> Date: Tue May 22 03:11:43 2018 +0900 Increase test coverages by adding invalid CNTK usecases (#10236) * Revert "Pull request revision" This reverts commit fb7c367a1560f89d9940e32962d2b3807a54ce08. * Pull request revision (v2) * Replace HTML markup by markdown
-
- 21 6月, 2018 3 次提交
-
-
由 Francois Chollet 提交于
-
由 Max Pumperla 提交于
* autogen classes and functions * remove unused, regex strings, simpler ### * parantheses, not backslash * fix signature * fix for tf backends
-
由 Anner 提交于
-
- 20 6月, 2018 1 次提交
-
-
由 Aurélien Geron 提交于
* Map TensorFlow tensor with dtype=tf.string to NumPy dtype='object', fixes #10459 * Replace _numpy_type(tensor) method with tf.as_dtype(tensor.dtype).as_numpy_dtype
-
- 19 6月, 2018 2 次提交
- 16 6月, 2018 2 次提交
-
-
由 Pavithra Vijay 提交于
-
由 Francois Chollet 提交于
-