- 21 10月, 2016 12 次提交
-
-
由 Dan Mané 提交于
This is a fairly minimal change, some functions have been cleaned up slightly because the new summary ops dont need manual namespacing. Behavior is approximately identical. Change: 136745220
-
由 A. Unique TensorFlower 提交于
Change: 136745074
-
由 Wei Ho 提交于
Change: 136744992
-
由 A. Unique TensorFlower 提交于
Change: 136743848
-
由 A. Unique TensorFlower 提交于
Change: 136743547
-
由 Dan Mané 提交于
I originally intended to replace all the summary ops with implementations that use the TensorSummary kernel. However, I realized this would produce a lot of work without any user-visible benefit, so I decided not to do so. Although I had already created a TensorSummary wrapper for tf.summary.scalar, now that replacing all summary implementations is a non-goal, it is better to keep using ScalarSummary for consistency and backcompat. The behavior wrt. TensorBoard and the EventAccumulator is unchanged. Change: 136742625
-
由 A. Unique TensorFlower 提交于
Change: 136742352
-
由 A. Unique TensorFlower 提交于
Redefine `parameters` property to be the distribution construction arguments and add new `graph_parents` property to fulfill existing role of `parameters`. Also add validate_args to {Transformed,Quantized}Distribution ctors. Change: 136741700
-
由 A. Unique TensorFlower 提交于
Change: 136732417
-
由 A. Unique TensorFlower 提交于
separate ops for finding the min/max and doing the requantization step. Change: 136727927
-
由 A. Unique TensorFlower 提交于
not have 'hdrs'. Change: 136723996
-
由 Zongheng Yang 提交于
-
- 20 10月, 2016 28 次提交
-
-
由 A. Unique TensorFlower 提交于
Change: 136693563
-
由 A. Unique TensorFlower 提交于
Change: 136690791
-
由 A. Unique TensorFlower 提交于
Change: 136688415
-
由 A. Unique TensorFlower 提交于
Change: 136680849
-
由 A. Unique TensorFlower 提交于
Change: 136678928
-
由 A. Unique TensorFlower 提交于
Change: 136676914
-
由 A. Unique TensorFlower 提交于
Change: 136676591
-
由 Andrew Selle 提交于
This does not implement the Python API change yet for forward compatibility. e.g. eventually tf.reverse(a, [0,1,-1]) will be the same as the old API of tf.reverse(a, [True, True, False, False, False, True]) for a 6 dimensional tensor a. Change: 136675570
-
由 A. Unique TensorFlower 提交于
Change: 136672574
-
由 Jianmin Chen 提交于
Change: 136668713
-
由 A. Unique TensorFlower 提交于
Adds unit tests for handling negative arguments to tf.sqrt. Followup to cl/135111959. Change: 136665526
-
由 Vijay Vasudevan 提交于
headers from "core:framework" target like we do for framework_headers_lib, so that it is available in the pip package consistently. Tested by building the wheel with this change and validating that the files were in site-packages/tensorflow/include/tensorflow/core/framework, etc. Fixes #4996. Change: 136664737
-
由 A. Unique TensorFlower 提交于
nomsan and notsan, and concat_op_test passes in msan and tsan. Change: 136664499
-
由 A. Unique TensorFlower 提交于
Change: 136664255
-
由 Zongheng Yang 提交于
This change adapts some unit tests, that previously would fail when V2 becomes the default. They now pass with both V1 or V2 being the default. This is in preparation for the upcoming switch to V2 (~end-of-month). Change: 136663731
-
由 A. Unique TensorFlower 提交于
topological neighborhood of a device. Change: 136663586
-
由 Shanqing Cai 提交于
Change: 136663053
-
由 Vijay Vasudevan 提交于
scalar shapes during construction. Placholder/Variable currently treat [] as "unknown", because it predated scalar shapes. So we introduce new versions that are the same implementation underneath, but can differentiate between unknown and scalar. Change: 136659818
-
由 Shanqing Cai 提交于
On mac 64-bit, the library paths is actually at /usr/local/cuda/lib by default. The assumption that it is always at /usr/local/cuda/lib64 has led to build failures, e.g., http://ci.tensorflow.org/view/Nightly/job/nightly-matrix-mac-gpu/100/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=gpu-mac/console Change: 136653210
-
由 Yuan Yu 提交于
Before this CL, Device::Sync() was called only when the executor graph contains a SINK node. However the SINK node in the graph is guaranteed to be reachable from SOURCE. This CL makes sure Device::Sync() is always called before the executor exits. As a separate issue, we may consider if we want the executor graph to be "well-formed" with a single SOURCE and a single SINK. Change: 136652586
-
由 Dan Mané 提交于
Before this change, the EventAccumulator parses the new scalar_summaries as a one-element np array like this: np.array(3, dtype=np.float32). This is the same as a number wrt equality checking (i.e. expression above is equal to 3.0), but it isn't json serializable, which causes problems for TensorBoard. Change: 136647129
-
由 Mustafa Ispir 提交于
Change: 136645313
-
由 Mustafa Ispir 提交于
Change: 136644858
-
由 A. Unique TensorFlower 提交于
Change: 136644641
-
由 A. Unique TensorFlower 提交于
Change: 136643977
-
由 A. Unique TensorFlower 提交于
Change: 136641403
-
由 A. Unique TensorFlower 提交于
the second best split in finished_nodes_op. Change: 136640840
-
由 A. Unique TensorFlower 提交于
Change: 136637890
-