Commits

Breadcrumbs

History for

onmaster

User selector

All users

All time

Commit History

Commits on Nov 30, 2016

• Added AVX-512 support to Eigen and TensorFlow.

  Show description for 35b6050
  tensorflower-gardener

  committed
  35b6050

  Copy full SHA for 35b6050

Commits on Oct 28, 2016

• Merge changes from github.

  Show description for e2d51a8
  

  zheng-xq

  authored and

  tensorflower-gardener

  committed
  e2d51a8

  Copy full SHA for e2d51a8

Commits on Oct 23, 2016

• Fix various warnings exposed by current versions of Clang.

  Show description for a320ea4
  tensorflower-gardener

  committed
  a320ea4

  Copy full SHA for a320ea4

Commits on Oct 14, 2016

• Adds a resource reference dtype to TF.

  Show description for 27b7506
  tensorflower-gardener

  committed
  27b7506

  Copy full SHA for 27b7506

Commits on Sep 30, 2016

• Add a core:framework_minimal library that includes minimal headers for types,

  Show description for 05ce371
  tensorflower-gardener

  committed
  05ce371

  Copy full SHA for 05ce371

Commits on Sep 26, 2016

• allocator.h: Don't include what you don't use.

  Show description for 7d7e082
  

  mrry

  authored and

  tensorflower-gardener

  committed
  7d7e082

  Copy full SHA for 7d7e082

Commits on Jun 2, 2016

• Update copyright for 3p/tf/core.

  Show description for c8b59c0
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  c8b59c0

  Copy full SHA for c8b59c0

Commits on May 19, 2016

• Add the ability to return the cost model to the client as part of run metadata.

  Show description for be8a3e2
  

  yuanbyu

  authored and

  tensorflower-gardener

  committed
  be8a3e2

  Copy full SHA for be8a3e2

Commits on Apr 18, 2016

• Added a format for saving an inference graph that can be memmapped and an utility to convert a freezed graph into this format.

  Show description for 3c280f6
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  3c280f6

  Copy full SHA for 3c280f6

Commits on Mar 30, 2016

• Added a config to enable the use of a tracking allocator on CPU

  Show description for 7b7b7fb
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  7b7b7fb

  Copy full SHA for 7b7b7fb

Commits on Mar 22, 2016

• Add a half type to TensorFlow core, based on Eigen::half.

  Show description for 9d03824
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  9d03824

  Copy full SHA for 9d03824

Commits on Mar 18, 2016

• Change Cuda pinned memory allocation to BFC allocator.

  Show description for 2b7ca5c
  

  zheng-xq

  authored and

  tensorflower-gardener

  committed
  2b7ca5c

  Copy full SHA for 2b7ca5c

Commits on Mar 16, 2016

• Rollforward of "Merge changes from github."

  Show description for 4c85a08
  

  girving

  authored and

  tensorflower-gardener

  committed
  4c85a08

  Copy full SHA for 4c85a08

• Rollback of: "Merge changes from github."

  Show description for 9a4878c
  

  Vijay Vasudevan

  authored and

  tensorflower-gardener

  committed
  9a4878c

  Copy full SHA for 9a4878c

• Merge changes from github.

  Show description for 3ae663c
  

  martinwicke

  authored and

  tensorflower-gardener

  committed
  3ae663c

  Copy full SHA for 3ae663c

Commits on Mar 2, 2016

• Add optional comprehensive logging of memory allocation/deallocation events. When enabled, the following events are recorded:

  Show description for ec1403e
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  ec1403e

  Copy full SHA for ec1403e

Commits on Feb 22, 2016

• Added the ability to track allocation sizes in the tracking allocator itself if the underlying

  Show description for 8f8f7f1
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  8f8f7f1

  Copy full SHA for 8f8f7f1

Commits on Feb 20, 2016

• Exposes the memory limit in the allocator's stats.

  Show description for b024048
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  b024048

  Copy full SHA for b024048

Commits on Feb 18, 2016

• Tracks some basic statistics in cpu and gpu allocator.

  Show description for 648e265
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  648e265

  Copy full SHA for 648e265

Commits on Feb 17, 2016

• Merge changes from github.

  Show description for fe056f0
  

  Vijay Vasudevan

  authored and

  tensorflower-gardener

  committed
  fe056f0

  Copy full SHA for fe056f0

• Restore AllocatorAttributes as a 32-bit value, add a test to make sure the

  Show description for 69a7321
  

  Vijay Vasudevan

  authored and

  tensorflower-gardener

  committed
  69a7321

  Copy full SHA for 69a7321

Commits on Feb 11, 2016

• Expose the ability to store and retrieve a datatype from the TensorShape

  Show description for 9934b9c
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  9934b9c

  Copy full SHA for 9934b9c

Commits on Feb 6, 2016

• Add support for allocators that require tensors with zero

  Show description for 8f80ef1
  

  A. Unique TensorFlower

  authored and

  Vijay Vasudevan

  committed
  8f80ef1

  Copy full SHA for 8f80ef1

Commits on Jan 26, 2016

• Global search & replace to move to the new location for

  Show description for 64aa098
  

  josh11b

  authored and

  Vijay Vasudevan

  committed
  64aa098

  Copy full SHA for 64aa098

Commits on Jan 18, 2016

• Adds enough auditing to make it possible to track tensor buffers throughout an execution, and build a cost model of memory usage.

  Show description for 6f62e43
  

  A. Unique TensorFlower

  authored and

  keveman

  committed
  6f62e43

  Copy full SHA for 6f62e43

Commits on Jan 14, 2016

• Move responsibility for constructing/destructing complex objects from

  Show description for 861f8f0
  

  A. Unique TensorFlower

  authored and

  Vijay Vasudevan

  committed
  861f8f0

  Copy full SHA for 861f8f0

Commits on Jan 7, 2016

• The default memory manager would delay 10s for any memory failure. This is not good for Cudnn scratch space allocation, which is only optional with performance gains.

  Show description for f13442a
  

  A. Unique TensorFlower

  authored and

  Vijay Vasudevan

  committed
  f13442a

  Copy full SHA for f13442a

Commits on Nov 20, 2015

• TensorFlow: Improve performance of Alexnet

  Show description for 9c3043f
  keveman

  committed
  9c3043f

  Copy full SHA for 9c3043f

Commits on Nov 16, 2015

• TensorFlow: Doc and linter fixes, some additional tests and

  Show description for 56313de
  

  Vijay Vasudevan

  committed
  56313de

  Copy full SHA for 56313de

Commits on Nov 7, 2015

• TensorFlow: Initial commit of TensorFlow library.

  Show description for f41959c
  keveman

  committed
  f41959c

  Copy full SHA for f41959c

Loading