Commits

Breadcrumbs

History for

onr0.9

User selector

All users

All time

Commit History

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 17, 2016

• Avoid calling LogMemory::IsEnabled in allocate_tensor. This saves a couple

  Show description for 20c7eca
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  20c7eca

  Copy full SHA for 20c7eca

Commits on May 4, 2016

• Only construct UniqueTensorReferences if they are needed.

  Show description for fc99eea
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  fc99eea

  Copy full SHA for fc99eea

Commits on Apr 21, 2016

• Check for op deprecation in C++ API and graph validation

  Show description for 18429eb
  

  girving

  authored and

  tensorflower-gardener

  committed
  18429eb

  Copy full SHA for 18429eb

• Change tensorflow code to use proto_text for DebugString and ShortDebugString

  Show description for 3883578
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  3883578

  Copy full SHA for 3883578

Commits on Apr 20, 2016

• Add a .Deprecated method to REGISTER_OP

  Show description for 30334d2
  

  girving

  authored and

  tensorflower-gardener

  committed
  30334d2

  Copy full SHA for 30334d2

• Rollback of "Change tensorflow code to use proto_text for DebugString andS hortDebugString"

  Show description for 43fc42b
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  43fc42b

  Copy full SHA for 43fc42b

Commits on Apr 19, 2016

• Change tensorflow code to use proto_text for DebugString and ShortDebugString

  Show description for 50d0241
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  50d0241

  Copy full SHA for 50d0241

Commits on Mar 29, 2016

• Fixing some implicit int64->32 downcast errors. (In most cases, by

  Show description for dff1e63
  

  David G. Andersen

  authored and

  tensorflower-gardener

  committed
  dff1e63

  Copy full SHA for dff1e63

Commits on Mar 28, 2016

• Change selective registration of op kernels to work off of the class name

  Show description for e30d80d
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  e30d80d

  Copy full SHA for e30d80d

Commits on Mar 24, 2016

• Several changes to reduce TensorFlow code size (especially important

  Show description for 821920c
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  821920c

  Copy full SHA for 821920c

Commits on Mar 12, 2016

• Add optional cc file whitelisting for operator kernel registration.

  Show description for eb56025
  

  andrewharp

  authored and

  tensorflower-gardener

  committed
  eb56025

  Copy full SHA for eb56025

Commits on Mar 11, 2016

• Changes to make custom user ops not statically include tensorflow/core:framework.

  Show description for e774fa8
  

  keveman

  authored and

  tensorflower-gardener

  committed
  e774fa8

  Copy full SHA for e774fa8

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

• Changes to reduce code size (especially useful for mobile):

  Show description for b76812c
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  b76812c

  Copy full SHA for b76812c

Commits on Feb 25, 2016

• Avoid some over-inlined routines. Reduces code size of TensorFlow binaries

  Show description for 9ccc4b6
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  9ccc4b6

  Copy full SHA for 9ccc4b6

Commits on Feb 16, 2016

• Several changes to reduce the overhead and the cache footprint of dispatching

  Show description for cf66101
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  cf66101

  Copy full SHA for cf66101

Commits on Feb 12, 2016

• Hardening the error checking related to memory types.

  Show description for 0e28e70
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  0e28e70

  Copy full SHA for 0e28e70

Commits on Feb 11, 2016

• Moves MemoryType inference code out of OpKernel so that it can reused.

  Show description for 7415dfa
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  7415dfa

  Copy full SHA for 7415dfa

Commits on Feb 5, 2016

• Allows int32 on host memory work when a function passes int32 in/out.

  Show description for 0c6c4b9
  

  A. Unique TensorFlower

  authored and

  keveman

  committed
  0c6c4b9

  Copy full SHA for 0c6c4b9

Commits on Feb 3, 2016

• Add RemoveNewDefaultAttrsFromGraphDef() for preparing a GraphDef for

  Show description for 7b4a463
  

  josh11b

  authored and

  Vijay Vasudevan

  committed
  7b4a463

  Copy full SHA for 7b4a463

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 25, 2016

• Eliminate per-op allocation of gpu device wrapper. The PerOpGpuDevice is allocated once in the OpKernelContext::Params struct, then re-used every time a new OpKernelContext uses the Params. Thus in…

  Show description for ff8522d
  

  A. Unique TensorFlower

  authored and

  Vijay Vasudevan

  committed
  ff8522d

  Copy full SHA for ff8522d

Commits on Jan 22, 2016

• Move #include <vector> out of port.h to users of std::vector<>.

  Show description for b481783
  

  josh11b

  authored and

  Vijay Vasudevan

  committed
  b481783

  Copy full SHA for b481783

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 13, 2016

• Added OpKernel::is_internal and used it to avoid a memcmp against the

  Show description for 94538a9
  

  A. Unique TensorFlower

  authored and

  Vijay Vasudevan

  committed
  94538a9

  Copy full SHA for 94538a9

Commits on Jan 12, 2016

• Add GraphDef version 6 implementing scalar strictness, but don't use it

  Show description for cbff45c
  

  girving

  authored and

  Vijay Vasudevan

  committed
  cbff45c

  Copy full SHA for cbff45c

Commits on Jan 7, 2016

• #include "tensorflow/core/platform/mutex.h"

  Show description for b1d45c9
  

  josh11b

  authored and

  Vijay Vasudevan

  committed
  b1d45c9

  Copy full SHA for b1d45c9

Commits on Jan 6, 2016

• Change: 111522571

  

  A. Unique TensorFlower

  authored and

  Vijay Vasudevan

  committed
  a1d4149

  Copy full SHA for a1d4149

• New GraphDef version 4: TensorFlow is now scalar strict

  Show description for 0ce489f
  

  A. Unique TensorFlower

  authored and

  Vijay Vasudevan

  committed
  0ce489f

  Copy full SHA for 0ce489f

Commits on Jan 5, 2016

• Added 'logging' import to control_flow_ops which is used in the file but not imported.

  Show description for 1c57936
  

  A. Unique TensorFlower

  authored and

  Vijay Vasudevan

  committed
  1c57936

  Copy full SHA for 1c57936

Commits on Dec 11, 2015

• TensorFlow: merge changes from internal

  Show description for d9cfc64
  

  Vijay Vasudevan

  committed
  d9cfc64

  Copy full SHA for d9cfc64

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 7, 2015

• TensorFlow: Initial commit of TensorFlow library.

  Show description for f41959c
  keveman

  committed
  f41959c

  Copy full SHA for f41959c

Loading