| 7149a2e2e2f549035f23e21224ee41afe8df3876 |
|
30-Jan-2018 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Cleanup: Ran clang-format on files in tensorflow/core/.../*.{cc,h}.
PiperOrigin-RevId: 183848459
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|
| 83c6e0c63acdcab2c58c4ed7220bfa58879b1d57 |
|
12-Jan-2017 |
Jonathan Hseu <jhseu@google.com> |
Switch open-source to use jemalloc for CPU Tensor memory allocation, gRPC, and other places where we call malloc/free.
- Only enabled on Linux for now.
- Added as a ./configure option defaulting to enabled.
Change: 144266237
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|
| edaf3b342db4afa1c872da541fb0ac176a4e8ef9 |
|
10-Oct-2016 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Merge changes from github.
Change: 135698415
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|
| 08a6ebc3ad2194a61ef310a1da97e4402a5a00bb |
|
21-Sep-2016 |
Derek Murray <mrry@google.com> |
Clean up warnings and IWYU in tensorflow/core/lib.
Change: 133826999
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|
| 46231cf242c19d74af75370eefd9e9b7c504c08a |
|
24-Aug-2016 |
Vijay Vasudevan <vrv@google.com> |
Move host_info.h internal, move public functions into cpu_info.h
and make public.
Hostname() is not public, but kLittleEndian arguably is (some code
might need to know what the endianness is, like decode_raw_op,
so it probably should be public). Similarly, NumSchedulableCPUs
is probably something a lot of code wants to know. It is also
already used widely and seems hard to remove in short notice,
which suggests something about its utility.
Change: 131203680
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|
| 4e37b20f629ab00d66adc283b67b2b31cdadfddd |
|
22-Aug-2016 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Introduce support for move-only types in InlinedVector, and use move during resize.
Note: the previous code in Move seemed to use std::move, but without effect, since the argument was a const&.
This cl fixes the problem, shown by my test with a MoveOnly type.
From:
Benchmark Time(ns) CPU(ns) Iterations
---------------------------------------------------------------
BM_InlinedVectorFillString/0 2 2 357702278
BM_InlinedVectorFillString/1 23 23 30020181 40.8M items/s
BM_InlinedVectorFillString/8 177 177 3938795 43.1M items/s
BM_InlinedVectorFillString/64 2515 2516 278714 24.3M items/s
BM_InlinedVectorFillString/512 21497 21478 31994 22.7M items/s
BM_InlinedVectorFillString/1k 43361 43312 16184 22.5M items/s
To:
Benchmark Time(ns) CPU(ns) Iterations
---------------------------------------------------------------
BM_InlinedVectorFillString/0 2 2 357070962
BM_InlinedVectorFillString/1 23 23 31134543 42.0M items/s
BM_InlinedVectorFillString/8 166 166 4246010 45.9M items/s
BM_InlinedVectorFillString/64 1645 1644 427302 37.1M items/s
BM_InlinedVectorFillString/512 13830 13806 49712 35.4M items/s
BM_InlinedVectorFillString/1k 28164 28114 24588 34.7M items/s
Change: 130922812
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|
| c8b59c046895fa5b6d79f73e0b5817330fcfbfc1 |
|
02-Jun-2016 |
A. Unique TensorFlower <nobody@tensorflow.org> |
Update copyright for 3p/tf/core.
Change: 123900938
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|
| fe056f0b5e52db86766761f5e6446a89c1aa3938 |
|
17-Feb-2016 |
Vijay Vasudevan <vrv@google.com> |
Merge changes from github.
Change: 114882676
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|
| 422f4492dd93af9b09fb0b27c57c65467c30f798 |
|
27-Jan-2016 |
A. Unique TensorFlower <nobody@tensorflow.org> |
Change InlineVector representation to better use inlined space.
We can now fit up to 15 1-byte elements in an InlinedVector that
occupies 16 bytes. Previously, this would require 32 bytes.
Removed Allocator support.
Removed old-style array() methods (use data() instead).
Benchmark Base (ns) New (ns) Improvement
------------------------------------------------------------------
BM_InlinedVectorFill/0 1 2 -100.0%
BM_InlinedVectorFill/1 2 2 +0.0%
BM_InlinedVectorFill/8 8 8 +0.0%
BM_InlinedVectorFill/64 184 261 -41.8%
BM_InlinedVectorFill/512 1211 1923 -58.8%
BM_InlinedVectorFill/1k 2346 3792 -61.6%
BM_InlinedVectorFillRange/0 1 1 +0.0%
BM_InlinedVectorFillRange/1 5 4 +20.0%
BM_InlinedVectorFillRange/8 5 4 +20.0%
BM_InlinedVectorFillRange/64 52 22 +57.7%
BM_InlinedVectorFillRange/512 242 55 +77.3%
BM_InlinedVectorFillRange/1k 458 88 +80.8%
BM_InlinedVectorTenAssignments/0 32 25 +21.9%
BM_InlinedVectorTenAssignments/1 51 47 +7.8%
BM_InlinedVectorTenAssignments/2 53 51 +3.8%
BM_InlinedVectorTenAssignments/3 59 73 -23.7%
BM_InlinedVectorTenAssignments/4 59 73 -23.7%
BM_InlinedVectorTenAssignments/20 129 151 -17.1%
BM_CreateFromInitializerList ?????? 1 ----
BM_LargeSwappableElements/0 3 4 -33.3%
BM_LargeSwappableElements/1 91 94 -3.3%
BM_LargeSwappableElements/8 1230 1243 -1.1%
BM_LargeSwappableElements/64 4 2 +50.0%
BM_LargeSwappableElements/512 4 2 +50.0%
BM_LargeSwappableElements/1k 4 2 +50.0%
Change: 113172547
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|
| fc0235e04a040f039f9caf5d8e28ee0db0b8abfd |
|
27-Jan-2016 |
Vijay Vasudevan <vrv@google.com> |
Rollback of"
Change InlineVector representation to better use inlined space.
We can now fit up to 15 1-byte elements in an InlinedVector that
occupies 16 bytes. Previously, this would require 32 bytes.
Removed Allocator support.
Removed old-style array() methods (use data() instead).
Benchmark Base (ns) New (ns) Improvement
------------------------------------------------------------------
BM_InlinedVectorFill/0 1 2 -100.0%
BM_InlinedVec...
"
Change: 113100787
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|
| 6d2a4182ff6d6fbb1b209037ab0df87ecd884ad5 |
|
26-Jan-2016 |
A. Unique TensorFlower <nobody@tensorflow.org> |
Change InlineVector representation to better use inlined space.
We can now fit up to 15 1-byte elements in an InlinedVector that
occupies 16 bytes. Previously, this would require 32 bytes.
Removed Allocator support.
Removed old-style array() methods (use data() instead).
Benchmark Base (ns) New (ns) Improvement
------------------------------------------------------------------
BM_InlinedVectorFill/0 1 2 -100.0%
BM_InlinedVectorFill/1 2 2 +0.0%
BM_InlinedVectorFill/8 8 8 +0.0%
BM_InlinedVectorFill/64 184 261 -41.8%
BM_InlinedVectorFill/512 1211 1923 -58.8%
BM_InlinedVectorFill/1k 2346 3792 -61.6%
BM_InlinedVectorFillRange/0 1 1 +0.0%
BM_InlinedVectorFillRange/1 5 4 +20.0%
BM_InlinedVectorFillRange/8 5 4 +20.0%
BM_InlinedVectorFillRange/64 52 22 +57.7%
BM_InlinedVectorFillRange/512 242 55 +77.3%
BM_InlinedVectorFillRange/1k 458 88 +80.8%
BM_InlinedVectorTenAssignments/0 32 25 +21.9%
BM_InlinedVectorTenAssignments/1 51 47 +7.8%
BM_InlinedVectorTenAssignments/2 53 51 +3.8%
BM_InlinedVectorTenAssignments/3 59 73 -23.7%
BM_InlinedVectorTenAssignments/4 59 73 -23.7%
BM_InlinedVectorTenAssignments/20 129 151 -17.1%
BM_CreateFromInitializerList ?????? 1 ----
BM_LargeSwappableElements/0 3 4 -33.3%
BM_LargeSwappableElements/1 91 94 -3.3%
BM_LargeSwappableElements/8 1230 1243 -1.1%
BM_LargeSwappableElements/64 4 2 +50.0%
BM_LargeSwappableElements/512 4 2 +50.0%
BM_LargeSwappableElements/1k 4 2 +50.0%
Change: 113094648
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|
| 5a24d3a2514698b0ae11563b2ea21e368de48a4f |
|
26-Jan-2016 |
Josh Levenberg <josh11b@tensorflow.org> |
Global search & replace to move to the new location for
tensorflow/core/ files and build targets.
Change: 113075110
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|
| b481783fe0e00a86f6feb20a8dcad5fc4fc936a4 |
|
21-Jan-2016 |
Josh Levenberg <josh11b@tensorflow.org> |
Move #include <vector> out of port.h to users of std::vector<>.
After this we can replace port.h with types.h.
Change: 112727463
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|
| 9c3043ff3bf31a6a81810b4ce9e87ef936f1f529 |
|
20-Nov-2015 |
Manjunath Kudlur <keveman@gmail.com> |
TensorFlow: Improve performance of Alexnet
Changes:
* error message that refers to removed `DefaultSession` method.
* -Wnull-conversion warnings
* the "_start_time" attr for recvs when the flag "--brain_enable_scheduling_for_recvs" is set.
* typo in tutorial data download progress message.
* a typo ("however their installing"=>"however installing").
* typo, rename "TensorFlow Mechanics" to "How To" to be consistent with the website.
* a typo ("subtact"=>"subtract").
* protobuf examples in comments in tensorflow::Example.proto.
* formula formatting in MNIST beginner tutorial
* negative fraction-of-queue-full stats
* protobuf inclusion path so that Android demo will build under Blaze.
* small typo (moderatly > moderately)
* Session.run() to check that tensor arguments come from the session's graph.
* another six import
* seq2seq typo in bazel command
Base CL: 108349164
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|
| 56313def004795f75ef8281a0294c958d28f1e06 |
|
16-Nov-2015 |
Vijay Vasudevan <vrv@google.com> |
TensorFlow: Doc and linter fixes, some additional tests and
error handling, updates to website.
Changes:
- Removes redundant reshape from image models by @mrry
- Default TensorBoard to localhost by @danmane
- Reformatting of tensorflow/core by @josh11b
- Make tutorials backwards compatible to 0.5.0 by @girving
- Improve print documentation (md files not updated).
- Add proper scrolling to sitemap by @martinwicke
Base CL: 107956254
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|
| f41959ccb2d9d4c722fe8fc3351401d53bcf4900 |
|
07-Nov-2015 |
Manjunath Kudlur <keveman@gmail.com> |
TensorFlow: Initial commit of TensorFlow library.
TensorFlow is an open source software library for numerical computation
using data flow graphs.
Base CL: 107276108
/external/tensorflow/tensorflow/core/lib/gtl/inlined_vector.h
|