Commits

Breadcrumbs

History for

onmaster

User selector

All users

All time

Commit History

Commits on Jan 24, 2017

• Add options argument for DNN activation

  Show description for eb65eff
  tensorflower-gardener

  committed
  eb65eff

  Copy full SHA for eb65eff

Commits on Jan 19, 2017

• Add convolve quantized ops to StreamExecutor API

  Show description for 598583c
  tensorflower-gardener

  committed
  598583c

  Copy full SHA for 598583c

Commits on Jan 17, 2017

• Add several operations to the StreamExecutor API

  Show description for b398f90
  tensorflower-gardener

  committed
  b398f90

  Copy full SHA for b398f90

Commits on Sep 16, 2016

• Add the interface in steam executor to call cuDNN batch normalization functions.

  Show description for 04df8d8
  

  Yao Zhang

  authored and

  tensorflower-gardener

  committed
  04df8d8

  Copy full SHA for 04df8d8

Commits on Aug 19, 2016

• Add stream-executor changes to enable Cudnn fused LSTM/RNN support.

  Show description for 859e47f
  

  zheng-xq

  authored and

  tensorflower-gardener

  committed
  859e47f

  Copy full SHA for 859e47f

Commits on Jul 13, 2016

• Roll-forward of "Local Response Normalization GPU support via Stream Executor."

  Show description for 35df3ed
  

  rryan

  authored and

  tensorflower-gardener

  committed
  35df3ed

  Copy full SHA for 35df3ed

Commits on Jul 11, 2016

• Automated rollback of change 127123966

  Show description for 3f1b887
  

  Vijay Vasudevan

  authored and

  tensorflower-gardener

  committed
  3f1b887

  Copy full SHA for 3f1b887

• Local Response Normalization GPU support via Stream Executor.

  Show description for 71c9fb3
  

  rryan

  authored and

  tensorflower-gardener

  committed
  71c9fb3

  Copy full SHA for 71c9fb3

Commits on Jun 21, 2016

• Improve convolution autotune process. The max batch size VGG model can handle

  Show description for 466eb29
  

  zheng-xq

  authored and

  tensorflower-gardener

  committed
  466eb29

  Copy full SHA for 466eb29

Commits on Jun 7, 2016

• Enable fp16 for most of the pooling ops (MaxPool, AvgPool, associated

  Show description for 5a84537
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  5a84537

  Copy full SHA for 5a84537

Commits on Jun 3, 2016

• Enable fp16 for most of the pooling ops (MaxPool, AvgPool, associated

  Show description for 8abe7db
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  8abe7db

  Copy full SHA for 8abe7db

• Enable fp16 for most of the pooling ops (MaxPool, AvgPool, associated

  Show description for 9bedadc
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  9bedadc

  Copy full SHA for 9bedadc

Commits on Jun 2, 2016

• Update copyright for 3p/tf.

  Show description for 122cdce
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  122cdce

  Copy full SHA for 122cdce

Commits on May 25, 2016

• StreamExecutor add CUDA support for cudnnConvolutionBackwardBias

  Show description for a9f3979
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  a9f3979

  Copy full SHA for a9f3979

Commits on May 17, 2016

• Extend the StreamExecutor interface with ...WithAlgorithm() calls for fp16;

  Show description for 00bc81a
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  00bc81a

  Copy full SHA for 00bc81a

Commits on May 11, 2016

• Add fp16 cuDNN convolution support to StreamExecutor. (TensorFlow ops will

  Show description for 61b12f5
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  61b12f5

  Copy full SHA for 61b12f5

• Add fp16 matrix multiplication (GEMM) support to StreamExecutor, gated on

  Show description for 5230554
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  5230554

  Copy full SHA for 5230554

• Add the autotune version for the backward passes. It is currently

  Show description for 939ede0
  

  zheng-xq

  authored and

  tensorflower-gardener

  committed
  939ede0

  Copy full SHA for 939ede0

Commits on May 9, 2016

• Adding autotune to the Cudnn conv algorithm selection.

  Show description for 6ff265e
  

  zheng-xq

  authored and

  tensorflower-gardener

  committed
  6ff265e

  Copy full SHA for 6ff265e

Commits on Mar 18, 2016

• Support ScratchAllocator in BLAS Batched GEMM

  Show description for 05ea40f
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  05ea40f

  Copy full SHA for 05ea40f

Commits on Dec 8, 2015

• TensorFlow: upstream changes to git.

  Show description for ddd4aaf
  

  Vijay Vasudevan

  committed
  ddd4aaf

  Copy full SHA for ddd4aaf

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