hash
stringlengths
40
40
msg
stringlengths
1
131k
author
stringlengths
1
33
email
stringlengths
0
57
date
int64
1,447B
1,698B
a97e8e878985ed255a579fb47089d5e82c3f1a97
ci_build: cleanup android sdk environment variables
Jan Prach
jendap@google.com
1,450,416,343,000
26a588ef89c0c5989d943167c2a14b0bab1df581
ci_build: do not copy .bazelrc
Jan Prach
jendap@google.com
1,450,509,770,000
2697f7f7a08ef4014c4cfb79f91403ed29ad51d2
ci_build: automatically remove transient run containers
Jan Prach
jendap@google.com
1,451,463,541,000
2e680bbcace962a7aa85baaef948d360e6a2c49b
ci_build: change android ndk root directory - bazel issue 714 workaround
Jan Prach
jendap@google.com
1,451,458,471,000
0e7fd813f32387a9055aded8c791075fdfe2dbf5
Generalize categorical processor to many categorical variables
Illia Polosukhin
ilblackdragon@gmail.com
1,451,503,347,000
cc349fccbe8de5585a62e38b2b895172e6f416c3
Merge pull request #56 from google/categorical Ref #55: Adding first implementation of categorical variables support
Illia Polosukhin
ilblackdragon@gmail.com
1,451,503,388,000
b12e497a0a0540ae9acd418c4d40ce25b13706d7
Added dask_io support to allow dask.dataframe
terrytangyuan
terrytangyuan@gmail.com
1,451,363,284,000
04816e78fd7d0d8614ee1bcb21df27e6313fa2c0
Switched to dd.Series
terrytangyuan
terrytangyuan@gmail.com
1,451,434,774,000
9ebcd2370827b6dad5a75cb87723a07066b24055
init method for DaskDataFrameFeeder
terrytangyuan
terrytangyuan@gmail.com
1,451,435,320,000
5059644b0184836d452925930928adc99d7fe11a
get_feed_dict_fn for dask data feeder
terrytangyuan
terrytangyuan@gmail.com
1,451,438,713,000
ebd0ac4dc4ab80c4e7c03424088e78f3183af920
Implementation of get_feed_dict_fn for DaskDataFeeder
terrytangyuan
terrytangyuan@gmail.com
1,451,447,964,000
9c93fc56ca9e68b8a29c3933acbcafe8f11949bb
Integrate DaskDataFeeder into base estimator
terrytangyuan
terrytangyuan@gmail.com
1,451,448,712,000
1b6ac70db4252b73b97228ddaafcba7e65a7ec9b
Lint fixes, import, dependencies
terrytangyuan
terrytangyuan@gmail.com
1,451,450,259,000
11d7768bf2d247b9c374b62040ddd6994f857d46
Fixed wrong logic for _setup_data_feeder
terrytangyuan
terrytangyuan@gmail.com
1,451,451,160,000
ed99f7a4c3821250a9e1fffc76fc1521c81fd956
Logic error and lint fixes
terrytangyuan
terrytangyuan@gmail.com
1,451,454,595,000
4a7b5a9f27259e397bd856828322e50ebc86c005
Changes according to PR comments
terrytangyuan
terrytangyuan@gmail.com
1,451,503,462,000
51df996a824ce9550e715955bbf974986a6840a1
Fixes #60: Added example of text classification wih CNNs over words.
Illia Polosukhin
ilblackdragon@gmail.com
1,451,504,043,000
b5135db52ae18ddb1dfc25224453c32ad73e31b6
Structural change of code base for data feeder
terrytangyuan
terrytangyuan@gmail.com
1,451,504,590,000
b30877a3253575e3876e5067c02a399074582d91
Fixed nosetests and pylint (+1 squashed commit) Squashed commits: [83e24b9] Fix nosetests and pylint
terrytangyuan
terrytangyuan@gmail.com
1,451,505,192,000
5826e952715f0aff05efd95630dc400ebdfdb236
Support early stopping in trainer
terrytangyuan
terrytangyuan@gmail.com
1,451,571,434,000
a1c77f5c6941c61d4a11707a8473b57544ea0335
Added early stopping to subclasses of TensorFlowEstimator
terrytangyuan
terrytangyuan@gmail.com
1,451,572,497,000
6bd1fe4252a2f53e67a4844c50bd89350ad4e2a2
Added test for early stopping
terrytangyuan
terrytangyuan@gmail.com
1,451,587,640,000
2a756b49189ea5f2a649d20eeb82046dc702bb72
Added example of using early stopping
terrytangyuan
terrytangyuan@gmail.com
1,451,587,798,000
892934535557e4d13041e421d32ee082ba53d867
Added example to README
terrytangyuan
terrytangyuan@gmail.com
1,451,587,976,000
5069e38de26f93b54dd92bf9de6661561dc4bea6
Fix typo Fix typo
Yuxin Wu
ppwwyyxxc@gmail.com
1,451,597,801,000
919270045c3d42bc3f95a0d0b7ac70e470d0625f
Merge pull request #659 from ppwwyyxx/patch-1 Fix typo
Vijay Vasudevan
vrv@google.com
1,451,598,452,000
089b0b4af40ad707d142aa340da3f7dd257a4519
Remove all {#anchor} tags in non-API Markdown docs This commit does two things: 1. Removes all '{#anchor}' syntax from Markdown header lines in the codebase 2. Replaces all links that used those #anchors with links conforming to Github/Tensorflow website auto-generated anchors The work is a continuation on #648. This commit does not change the API docs, as those Markdown files are auto-generated. Generated using Python script available here: https://github.com/samjabrahams/tensorflow_util/blob/master/py/change_header_anchor_links.py Cross-page links changed by hand using this script to find non-api anchor links: https://github.com/samjabrahams/tensorflow_util/blob/master/py/get_anchor_references.py
Sam Abrahams
sjabraha@gmail.com
1,451,590,915,000
c1bf65fc757e8d5e0e90af3b6547f62dc281c193
Merge pull request #62 from terrytangyuan/early-stopping Early stopping option (Fixed #61)
Illia Polosukhin
ilblackdragon@gmail.com
1,451,622,304,000
eaa99cf5454bf31e71b0f71365403fcd1a5c80c0
Merge pull request #59 from terrytangyuan/terry-dev DaskDataFeeder (fixes issue #53)
Illia Polosukhin
ilblackdragon@gmail.com
1,451,674,672,000
b82c94ab1917d7420addc2f9fcef8f3a288869e0
Add saver test output to gitignore
terrytangyuan
terrytangyuan@gmail.com
1,451,695,671,000
68398ab8ab5bef9e84e13eea238f9432ce1f0d96
Use pandas for examples, removed csv import
terrytangyuan
terrytangyuan@gmail.com
1,451,699,225,000
b3b7784f9ded032a1d5568d53e42819e4e02b8a6
Minor logic fix in save_restore example
terrytangyuan
terrytangyuan@gmail.com
1,451,699,831,000
0ecf7ad5bacc59236710315b431d79d99a9c798e
Merge pull request #63 from terrytangyuan/terry-dev Just some housekeeping
Illia Polosukhin
ilblackdragon@gmail.com
1,451,703,316,000
1d89c2079931f401f0831e3d27b66dd942ae3388
Fixes typos and unnecessary import in example
terrytangyuan
terrytangyuan@gmail.com
1,451,711,169,000
7f357986ddefc1492ab5e781587d77b6509af75e
+ exponential decay to trainer
terrytangyuan
terrytangyuan@gmail.com
1,451,763,862,000
2b4b216c4b0a1fbf350c840b49324d66119d5ba9
+ exponential_decay to other estimators
terrytangyuan
terrytangyuan@gmail.com
1,451,764,932,000
49aac7ecb6944c2a6ba247f6b966eb2a8e55a5fc
* Explicitly assign parameters to trainer to pass tests
terrytangyuan
terrytangyuan@gmail.com
1,451,766,235,000
b55df7854e43edb8b504aa05a2923f01bcce9c26
+ Unit test for exponential decay
terrytangyuan
terrytangyuan@gmail.com
1,451,767,050,000
948fb3b20a6f268c2445646d6fab8a01240a4725
* Locally disable pylint for redefined-variable-type
terrytangyuan
terrytangyuan@gmail.com
1,451,767,371,000
6ba647c7f2d40131b8b47d6650aa0a6307474e7b
This PR fixes a few typos in convolutional.py and os_setup.md.
Dongjoon Hyun
dongjoon@apache.org
1,451,769,059,000
d6cde414498ddf8d0fbb221aa0530d53ce51539b
Merge pull request #669 from dongjoon-hyun/fix_typos_in_convolutional_py Fix a few typos in convolutional.py and os_setup.md.
Vijay Vasudevan
vrv@google.com
1,451,769,909,000
7bf0d32f13f34fdbed0691c26b95561ba3ec3b26
Adding memoryview/buffer in BytePreprocessing to not copy memory
Illia Polosukhin
ilblackdragon@gmail.com
1,451,784,434,000
61fa3aa561a38aa4841d8ca696df8b1b6912f15f
changed the way of using memoryview/buffer into more pythonic
Illia Polosukhin
ilblackdragon@gmail.com
1,451,784,765,000
1883b00658eaa7f02ac4e07e886a2de3ef264521
unrolled prev commit
Illia Polosukhin
ilblackdragon@gmail.com
1,451,785,625,000
ad4d283fea9e97fe05c86888469afe07a012a1ae
should fix pylint errors due to memoryview/buffer
Illia Polosukhin
ilblackdragon@gmail.com
1,451,786,190,000
9aebe6cd14dc503bf06e0e9a9a7a319cbf873287
fixed pylint in py3
Illia Polosukhin
ilblackdragon@gmail.com
1,451,786,655,000
545a25a109ffe52bb6f2c7264d577055bc1f0923
Moved out all data feeder logic into data_feeder.py
Illia Polosukhin
ilblackdragon@gmail.com
1,451,793,056,000
a0a84a5277116d1449d5501dae23dc60d061e359
Added support for iterators in the predict
Illia Polosukhin
ilblackdragon@gmail.com
1,451,793,957,000
823496f5cbf850cb8b828c5903090511ea18e1c3
Make neural translation model to train. Prediction needs sampling to work.
Illia Polosukhin
ilblackdragon@gmail.com
1,451,794,152,000
8b9f49cf0016322c08fec0f72c5a868da8b50fee
Move seq2seq operations from examples to ops/seq2seq_ops
Illia Polosukhin
ilblackdragon@gmail.com
1,451,795,017,000
534c10293a9874eec99e1195afd507bf29e1077c
Making prediction work for neural translation - removing dependency on output
Illia Polosukhin
ilblackdragon@gmail.com
1,451,797,354,000
98129e01fcff01c566b31cadc382e2fba44bffa8
Added reverse to ByteProcessor to get reverse from output
Illia Polosukhin
ilblackdragon@gmail.com
1,451,805,628,000
178befcd889f9829025b10108d4063cf4b9699b6
Modified trainer and base estimator to be able to train models after restoring. Additionally added axis on which to argmax into predict
Illia Polosukhin
ilblackdragon@gmail.com
1,451,810,828,000
a55af978207e6695717e2073199cd9918ea7d99a
Switched from reduce_mean to reduce_sum in sequence loss to have a better resolution
Illia Polosukhin
ilblackdragon@gmail.com
1,451,810,861,000
524078b1a62baca11a364f13f3289b5bd9247fee
Forcing data arrays to be one byte in ByteProcessor reverse
Illia Polosukhin
ilblackdragon@gmail.com
1,451,810,891,000
cb03302064f7626b0a7ca1e602bd906e4d886f7f
Adding saving/restoring into neural translation example and outputting results
Illia Polosukhin
ilblackdragon@gmail.com
1,451,810,930,000
ccac7d50607b76b8977d707859f44ec46825badf
Fix IPython case in MNIST tutorial
Olivier Grisel
olivier.grisel@ensta.org
1,451,827,808,000
7c2bad5f7bd50e059fa7296192952a93ddc4fd74
* Allow learning_rate to accept custom decay function
terrytangyuan
terrytangyuan@gmail.com
1,451,841,662,000
8d5fe68c83549a59530037a28fd9d2cc3d6c67a2
+ Example of using customized decay function
terrytangyuan
terrytangyuan@gmail.com
1,451,842,070,000
859f9343fdd23efa9c03c692461a46986395c127
* Explicit failure message for new unit test
terrytangyuan
terrytangyuan@gmail.com
1,451,843,434,000
5e3d2e30d5276a906b3c04d7d4f194eea21f75f7
Added tests for seq2seq_ops. Moved ops tests into ops/tests folder
Illia Polosukhin
ilblackdragon@gmail.com
1,451,851,692,000
6b79f26f0bb6b0e3f11a3b2345741b0bfaf6000a
Fixing not working test in python3 for preprocessing/text
Illia Polosukhin
ilblackdragon@gmail.com
1,451,851,727,000
eb326627ff6f6025c9ad06a1b10a9158482328ef
Fixin lint errors in estimators/trainer
Illia Polosukhin
ilblackdragon@gmail.com
1,451,851,746,000
2f3477b7164d915f2fa5f040708a843055905c7b
adding coveragerc to not count tests files
Illia Polosukhin
ilblackdragon@gmail.com
1,451,851,764,000
f56d11757f406ad396cc2bc8f3d9052bf3536ddd
Added support for DataFrames in preprocessing categorical features
Illia Polosukhin
ilblackdragon@gmail.com
1,451,852,562,000
27cfb16e122f6a528a74ff2e4410b36d1ebeb913
Merge pull request #671 from ogrisel/ipython-casing Fix IPython case in MNIST tutorial
Derek Murray
derek.murray@gmail.com
1,451,852,809,000
4fe860f0519f7f1f451e37e0310f0cb47260329a
updateing coveragerc file
Illia Polosukhin
ilblackdragon@gmail.com
1,451,853,164,000
eec2cba5cec204441c3b3c97db383f2ba2a148fd
PoolAlloc: Remove div by zero, demote WARN->INFO
Vijay Vasudevan
vrv@google.com
1,451,859,749,000
2bde1e1308767ca240422596143832cbb095196b
Merge pull request #673 from tensorflow/vrv-patch-1 PoolAlloc: Remove div by zero, demote WARN->INFO
Derek Murray
derek.murray@gmail.com
1,451,860,056,000
5841c93d752da8bfc924ec8b3674e5ef339d02c9
fixed misspelling in [run] section
Illia Polosukhin
ilblackdragon@gmail.com
1,451,885,289,000
dc3b8fb08446c4a366df9bf12bd744c3a1525538
Added reference/citations for RMSprop and AdaGrad. Also updated Adam optimizer citation to all be consistent with the style for referencing papers in the docstring used in rnn_cell.
Zachary Lipton
zachary.chase@gmail.com
1,451,886,211,000
2d5f0ac61fe4e4160fbb68d8031418528111dae9
Merge pull request #670 from zackchase/master Added citations to the relevant papers for Adam, AdaGrad, and RMSprop
Vijay Vasudevan
vrv@google.com
1,451,890,808,000
8033060b347d98a7ce72948fad8ed1f43e83c2ba
Fix typo
Nathan Howell
nathan.d.howell@gmail.com
1,451,932,969,000
55392a52d75b35661d436b92961c11381bfaa4ef
Merge pull request #679 from NathanHowell/typo Fix typo
Vijay Vasudevan
vrv@google.com
1,451,935,979,000
d4f3c0a9a5d4bce0752b2883167092a8f5cfb494
Merge pull request #639 from panmari/avg_pool_cuda_helpers Using cuda_kernel_helpers in average pooling.
Vijay Vasudevan
vrv@google.com
1,451,941,924,000
2f788c4ac2a05eaf8e4f6a22b8b5d34ca0e7320b
Merge pull request #660 from samjabrahams/upstream Remove all {#anchor} tags in non-API Markdown docs
Martin Wicke
martin.wicke@gmail.com
1,451,949,384,000
ab6d3891111bd15a8c2b3db3f5e0273238937818
Fixes two minor anchor link issues.
Frank Chu
fpmc@google.com
1,451,951,224,000
972d6e7fc2fea71339cfc1fe00555bb430ed1221
Merge pull request #683 from fpmchu/master Fixes two minor anchor link issues.
Martin Wicke
martin.wicke@gmail.com
1,451,952,961,000
f7b566a7f3199faaa0b7a122a4fee5c76b9eef3c
* Lowered expected score to pass test
terrytangyuan
terrytangyuan@gmail.com
1,451,958,028,000
95b42b94e1bf5cf3a1a156e0b6135d078006fd10
+ Check for callable for learning_rate
terrytangyuan
terrytangyuan@gmail.com
1,451,961,448,000
677e64084190d5dc4e20fce200686b9b33fae0e5
Fix typos of sample python code and docs in howtos
Dongjoon Hyun
dongjoon@apache.org
1,451,972,872,000
97522f0acd53652baa57e42d06557ebb94bf0c4d
Merge pull request #689 from dongjoon-hyun/fix_typos_in_example_and_docs Fix typos of sample python code and docs in howtos
Vijay Vasudevan
vrv@google.com
1,451,973,281,000
33bc277ffc73b09bde45f6c3d6ef9c847aa224dc
* Readability and doc-string for raises
terrytangyuan
terrytangyuan@gmail.com
1,452,003,289,000
bfb30ed1430e9105391412a880a918850e73478d
Add BUILD rule to create a shared library
Jim Fleming
jimmy.fleming@gmail.com
1,452,014,775,000
295625035eefc7801cfe831bd1ba851faacb7579
Deleted the local copy of the tensor code since we're now pulling this code directly from the eigen repository
Benoit Steiner
benoit.steiner.goog@gmail.com
1,452,024,364,000
1c579361cd1e088dd5e05a394b1561a73e3667ba
Added 'logging' import to control_flow_ops which is used in the file but not imported. Change: 110842260
A. Unique TensorFlower
nobody@tensorflow.org
1,452,031,527,000
411f57e291839094108afdaa9c43094f44979eaa
Merge commit for internal changes
Vijay Vasudevan
vrv@google.com
1,452,037,575,000
3578e0ec8d70ad73112d8af7c38f7b65950d91aa
Fix file exists error
Yi-Lin Juang
frankyjuang@gmail.com
1,451,998,778,000
70eb90173f6549c8861f4ce314e62831a365e624
Merge pull request #65 from terrytangyuan/terry-dev Custom Decay Function for Learning Rate
Illia Polosukhin
ilblackdragon@gmail.com
1,452,055,565,000
b82f4b248a7b62f7b8ce82f2629a153dad6995aa
* resolve conflicts
terrytangyuan
terrytangyuan@gmail.com
1,452,059,323,000
3ccab0390db8547894baaf02043e80417701dd15
+ More explanation on saver options
terrytangyuan
terrytangyuan@gmail.com
1,452,052,218,000
b0ff98f746fd101c935ca0967c3334815d9e862f
+ saver options for other estimators
terrytangyuan
terrytangyuan@gmail.com
1,452,059,152,000
3e638c391cfe9ba22a8d50b120dd304f2bc92468
Merge branch 'master' of https://github.com/tensorflow/tensorflow
Kenton Lee
kenton.ct.lee@gmail.com
1,452,079,924,000
c42490367c403b1836338523ce81eb2f8ac05647
Removes the need for IndexedSlicesWithoutDenseShapeValue. Tests for feeding and fetching IndexedSlices without dense shapes now reuse IndexedSlicesValue. Change-Id: I84eb7289d262797275e6eaa08b5e5daf4d99fe3d
Kenton Lee
kenton.ct.lee@gmail.com
1,452,085,760,000
709c9229151cb9a06296e55e9ab71adb9827371a
Adds back the tests for fetching tuples from IndexedSlices without dense shapes. Change-Id: I95c9eb219b5c6395892f132ae5034a45460e9855
Kenton Lee
kenton.ct.lee@gmail.com
1,452,086,453,000
248db9d2935adf54c6c72e8ee2290e04394edb8f
Merge pull request #704 from jendap/ci_build ci_build updates
Martin Wicke
martin.wicke@gmail.com
1,452,100,566,000
c56bdc3302fba4b89a30f64a4ec0a18f64ccc2e4
Changing docs to match the code. Change: 111445385
A. Unique TensorFlower
nobody@tensorflow.org
1,452,113,453,000
ef1d171e53a0da06ca04a680b2d1c80eff442669
Fix a bug introduced in previous change which made colorByChanged be called only once instead of twice. It turns out calling it twice was correct in order to clear state. Otherwise switching between graphs that share the same node names causes an error. Change: 111448764
Dan Smilkov
dsmilkov@gmail.com
1,452,113,463,000
b67f8cc70d60dbef305c0b49360dcc14442937a9
Split port.h (a grab-bag) into several .h, each fairly focussed. Change: 111452493
Josh Levenberg
josh11b@tensorflow.org
1,452,113,472,000
04b52f774eab3e8176b4485183c1e5e1763e7cbd
Drop snappy.h from port.h, #include it directly from the few users. Change: 111457547
Josh Levenberg
josh11b@tensorflow.org
1,452,113,488,000