Gráfico de Commits

Selecionar branches

Esconder Pull Requests

api-consistency

keras-1

keras-2

master

revert-6817-getfullargspec

tf-keras

0.1.0

0.1.1

0.1.2

0.1.3

0.2.0

0.3.0

0.3.1

0.3.2

0.3.3

1.0.0

1.0.1

1.0.2

1.0.3

1.0.4

1.0.5

1.0.6

1.0.7

1.0.8

1.1.0

1.1.1

1.1.2

1.2.0

1.2.1

1.2.2

2.0.0

2.0.1

2.0.2

2.0.3

2.0.4

2.0.5

Monocromático Colorido

• 205c61178d Update backend strings (#7214) master Taehoon Lee 2017-07-06 20:27:07 +09:00
• 3fe7837691 Update the old interfaces (#7252) Taehoon Lee 2017-07-06 20:26:47 +09:00
• bcc68e9040 Improve docstring on order of transformations (#7248) Dirk Gorissen 2017-07-06 12:23:33 +01:00
• 985c441014 Improved error message for missing class_weights. (#7238) jorgecarleitao 2017-07-05 20:20:12 +02:00
• 59cd1c3994 Fix typos (#7213) Taehoon Lee 2017-07-03 11:23:33 +09:00
• d135eda40e Update backend docs. api-consistency Francois Chollet 2017-06-30 16:38:16 -07:00
• 4de120f051 Fix computation of receptive field size for 'channels_last' (#7200) Larissa Heinrich 2017-06-30 19:36:22 -04:00
• 49f7649036 Fix backend tests Francois Chollet 2017-06-30 15:40:14 -07:00
• 6b16f3c135 Style cleanup, in particular crossentropy backend API Francois Chollet 2017-06-30 15:07:09 -07:00
• d9255f15a4 Redirect mobilenet weights path. Francois Chollet 2017-06-30 13:59:04 -07:00
• b42f6760bc Fix mobilenet weight preprocessing. Francois Chollet 2017-06-30 13:52:14 -07:00
• 6b51d149ca Merge branch 'master' of github.com:fchollet/keras Francois Chollet 2017-06-30 12:24:54 -07:00
• 8040ad72dd Fix mobilenet bugs. Francois Chollet 2017-06-30 12:24:40 -07:00
• ec048edffc Fix typo (#7191) nzw 2017-07-01 03:05:59 +09:00
• 23269507fd CI now fails for test coverage below 80%. Francois Chollet 2017-06-29 12:14:04 -07:00
• eafdffff75 Fix NotImplementedError in data_utils.py Francois Chollet 2017-06-29 12:13:34 -07:00
• 5834747dc7 Make load_model's convert_custom_objects recursive jorgecarleitao 2017-06-29 20:51:56 +02:00
• 6c2dea64fc conv3d_transpose in tf, th, and cntk (#7161) NC Cullen 2017-06-29 12:09:12 -04:00
• ddcf66fbe8 Update README.md (#7163) Alan Yee 2017-06-29 09:05:17 -07:00
• 5edd96a9f8 Don't warn of repeated data with one worker (#7175) Dapid 2017-06-29 18:03:24 +02:00
• c6ec39258b Fix styles (#7154) Taehoon Lee 2017-06-28 22:33:21 +09:00
• 7866fbaa1a Fix some typos. (#7144) Kongsea 2017-06-27 20:00:58 -05:00
• 94397e08ae Further style fixes. Francois Chollet 2017-06-27 16:40:12 -07:00
• 98db0285ee Style fix (avoid use of reserved variable name). Francois Chollet 2017-06-27 16:08:10 -07:00
• 7413956e7e Merge branch 'master' of github.com:fchollet/keras Francois Chollet 2017-06-27 15:46:05 -07:00
• 9a4598da50 Further style fixes in mobilenet Francois Chollet 2017-06-27 15:45:51 -07:00
• a09c9f6c2d Fix test in training (#7149) Frédéric Branchaud-Charron 2017-06-27 16:19:54 -04:00
• e4ab777d07 Improve documentation (#7150) nzw 2017-06-28 05:19:24 +09:00
• 4fab0bf9a8 Style fixes. Francois Chollet 2017-06-27 11:18:57 -07:00
• 49a7c7376d Addition of MobileNet to application (#7009) Somshubra Majumdar 2017-06-27 13:11:48 -05:00
• fb9dbdb10c Remove name-based trainable weight sorting, as it was a source of bugs. Francois Chollet 2017-06-26 19:52:14 -07:00
• a85263fb3e unify coding style (#7129) Hideaki Kanehara 2017-06-27 03:16:03 +09:00
• f6c1730cf3 better error message when ReduceLROnPlateau conditioned on non-existing metric (#7134) Gleb Sidora 2017-06-26 21:15:22 +03:00
• 598954d2c8 * Replace tensorflow deprecated attribute : reduction_indices -> axis (#7126) vfdev 2017-06-26 00:17:56 +02:00
• 21c5e5479b Add Sequence section in utils page (#7123) nzw 2017-06-26 02:33:25 +09:00
• d1ad183770 Merge branch 'master' of github.com:fchollet/keras Francois Chollet 2017-06-23 13:57:34 -07:00
• 9dca90e705 Fix PEP8 Francois Chollet 2017-06-23 13:56:54 -07:00
• 5a7f6b0e74 fix dtype handling in 2 optimizers and 1 layer (#7088) Ben 2017-06-23 16:41:11 -04:00
• 7dcd2982b2 Allow custom print functions for summary Francois Chollet 2017-06-23 13:24:20 -07:00
• 3462835597 Improve formatting of CONTRIBUTING.md. Francois Chollet 2017-06-22 16:04:52 -07:00
• 585f33f6b7 Add API design review process to CONTRIBUTING.md Francois Chollet 2017-06-22 16:03:51 -07:00
• 1aa9e9199b Add more meaningful message when cnkt can't handle variable length input (#7094) Tang, Cheng 2017-06-22 13:11:45 -07:00
• fb97b6e0fa _check_array_lengths properly handles corner cases with None (#7063) Andrew Hundt 2017-06-22 00:29:33 -04:00
• 60cf7ca6b2 Fix typos (#7087) Taehoon Lee 2017-06-22 13:28:57 +09:00
• 1b539993aa training.py _slice_arrays() fix crash when arrays are None (#7069) Andrew Hundt 2017-06-21 21:01:30 -04:00
• de73eda89a .gitignore visual studio code IDE excluded (#7070) Andrew Hundt 2017-06-21 16:47:52 -04:00
• 219d6ee5be training.py _weighted_masked_objective fix crash when weights is None (#7068) Andrew Hundt 2017-06-21 16:46:55 -04:00
• f430de10fb Style fixes (#7073) Taehoon Lee 2017-06-22 02:58:38 +09:00
• a2f6ae2c66 Style fix (#7079) nzw 2017-06-22 02:53:59 +09:00
• b713122e77 wrappers_test.py quick fix for flaky TimeDistributed test (#7062) Andrew Hundt 2017-06-21 13:50:58 -04:00
• 75470e380f join() OrderedEnqueuer.executor to prevent zombies (#7059) Gökçen Eraslan 2017-06-21 19:50:10 +02:00
• b5ad5334fc Add more numpy-style attributes to HDF5Matrix (#6982) moi90 2017-06-21 18:59:29 +02:00
• 04a20177cf Avoid DeprecationWarning from inspect.getargspec (rebased) (#7035) Chris 2017-06-21 18:58:38 +02:00
• 58d1d0678f Fix test in multiprocessing (#7058) Frédéric Branchaud-Charron 2017-06-20 18:56:45 -04:00
• abf8691ade Temp fix to multiprocessing tests Francois Chollet 2017-06-20 12:48:18 -07:00
• 7425e68cd6 Style fixes Francois Chollet 2017-06-20 12:47:52 -07:00
• ab6b82c2db Fix the ordering bugs when using pickle_safe=True (#6891) Frédéric Branchaud-Charron 2017-06-20 13:55:46 -04:00
• 6814506528 Renamed one_hot function to hashing_trick, made hashing stable (#6887) Andrea Esuli 2017-06-20 19:22:02 +02:00
• 1ddf23528e Fix bug in preprocess_weights_for_loading (#6960) Shaofan Lai 2017-06-21 01:21:14 +08:00
• 86c8d1dd45 Fixed default header in RemoteMonitor callback. (#7040) Sebastian Brandes 2017-06-20 19:20:28 +02:00
• 929ae992c2 test_multiprocessing.py fix incorrect use of numpy.random.randint (#7030) Andrew Hundt 2017-06-20 13:18:28 -04:00
• 3d9428d344 Increase test coverage (#7037) Taehoon Lee 2017-06-20 04:12:50 +09:00
• e0543fbfc8 throw an error when receiving empty batch, instead of returning nans at the end (#7038) Rik Nijessen 2017-06-19 20:23:41 +02:00
• 767846e642 fix url to LSTM paper (#7025) Stefan Schweter 2017-06-19 00:59:05 +02:00
• d852c2d772 Replace plot_loss_callback example with json_logging_callback (#4116) (#6941) Andrey M 2017-06-18 23:21:39 +02:00
• ff45159b69 Rename tensorboard histogram names to silence warnings (#7017) Elliot Saba 2017-06-17 18:22:58 -07:00
• 7766ab341f Speed up Theano tests (#7011) Taehoon Lee 2017-06-18 08:26:07 +09:00
• 4135aeebc4 Revert "Avoid DeprecationWarning from inspect.getargspec (#6817)" (#7018) François Chollet 2017-06-16 20:47:14 -07:00
• a748529d1f Revert "Avoid DeprecationWarning from inspect.getargspec (#6817)" revert-6817-getfullargspec François Chollet 2017-06-16 20:46:58 -07:00
• ced84c4b42 Avoid DeprecationWarning from inspect.getargspec (#6817) Chris 2017-06-17 00:09:25 +02:00
• 8d5b2ce60c Add example to compare RELU with SELU (#6990) Zafarali Ahmed 2017-06-16 18:00:16 -04:00
• c0f0b660a6 mean_square_error => mean_squared_error (#7015) Andrew Hundt 2017-06-16 16:01:31 -04:00
• d3c33613a1 Keras 2 _*generator displaying warning always about semantic changes from Keras 1 (#7001) Olli Huotari 2017-06-16 22:48:44 +03:00
• 5ca5699b00 Fixed some descriptions in backend (#6778) Mako 2017-06-17 03:12:14 +09:00
• be6503a8a8 Add space in error message Daniel Høyer Iversen 2017-06-16 19:47:40 +02:00
• c73ba916f6 Fix model loading for LeakyReLU layer (#7010) Yu-Yang 2017-06-17 00:12:18 +08:00
• e1c3988198 "here" links are difficult for individuals that need a screen reader for accessibility. (#6976) Andrew Hundt 2017-06-16 00:30:39 -04:00
• f65a56fb65 Exclude CNTK from TimeDistributed learning phase test Francois Chollet 2017-06-14 16:09:52 -07:00
• 00a2724260 Merge branch 'EntilZha-master' Francois Chollet 2017-06-14 14:51:49 -07:00
• a625fcde5c Add learning phase support for TimeDistributed Francois Chollet 2017-06-14 14:51:06 -07:00
• 73f374ec67 Merge branch 'master' of https://github.com/EntilZha/keras into EntilZha-master Francois Chollet 2017-06-14 14:39:26 -07:00
• 21cf50734a add Scaled Exponential Linear Unit activation (#6924) Rizky Luthfianto 2017-06-15 01:38:09 +07:00
• 4a6f06f06d Increase test coverage (#6765) Taehoon Lee 2017-06-15 01:34:15 +09:00
• 295e4f8064 Update lstm_benchmark.py (#6966) Simon Brugman 2017-06-13 20:13:21 +02:00
• f4cb890024 Minor typo correction in faq.md (#6964) td2014 2017-06-12 17:19:58 -07:00
• cd943231d1 Generator should use Process Lock when pickle safe instead of Threading Lock (#6911) Leoyzen 2017-06-13 04:59:45 +08:00
• f7b925a893 Use the pytest tmpdir fixture (#6901) Chris 2017-06-12 22:05:40 +02:00
• 5d63ab4251 More descriptive message when user specifies non-existing metric for early stopping (#6954) Gleb Sidora 2017-06-12 23:04:44 +03:00
• d4b618bf23 Prepare new PyPI release. 2.0.5 Francois Chollet 2017-06-12 11:46:21 -07:00
• 5012678e17 Fix typos (#6949) Taehoon Lee 2017-06-12 09:22:16 +09:00
• 11d9c995cc fixes #3859 clipnorm tensorflow (#6859) Chen 2017-06-10 22:08:04 +03:00
• d92fab69a2 Parallel directory iterator initialization (#6890) GPhilo 2017-06-10 04:41:44 +02:00
• 19463a19b8 Fine-tuning InceptionV3: Correct number of layers for two last inception blocks. (#6918) Martin Hallén 2017-06-10 02:54:06 +02:00
• ca1122fe80 update with more meaningful error message for CNTK backend (#6915) Tang, Cheng 2017-06-09 13:05:57 -07:00
• 7fc707e13e Fix comments for binary_crossentropy and sparse_crossentropy. (#6919) Jan Zikes 2017-06-09 19:55:40 +02:00
• 846d25ab97 fix a bug when model.fit assumed that x[0] has len, not true for sparse matrices (#6916) Maxim Grechkin 2017-06-08 21:20:00 -07:00
• 8c0a8b4b04 Use batch_input_shape in input layer even if input_tensor is set (#6883) Erik Smistad 2017-06-07 21:35:35 +02:00
• 1b1e09a366 Remove duplicated batch_set_value in cntk_backend (#6878) Daniel Høyer Iversen 2017-06-07 09:00:25 +02:00
• fd427b8cdb Fix typos (#6879) Taehoon Lee 2017-06-07 16:00:11 +09:00
• 53303fdb10 Improvements to style transfer as discussed in https://github.com/fchollet/keras/pull/6872 (#6877) Somshubra Majumdar 2017-06-07 01:40:32 -05:00