Commits

Branch selector

 master

User selector

gchanan

Datepicker

All time

Commit History

Commits on Aug 15, 2017

• Support __neg__, .neg(), and neg_() for Long, Int, Short tensor types.

  

  gchanan

  authored and

  soumith

  committed
  0252bcd

  Copy full SHA for 0252bcd

Commits on Jun 22, 2017

• Remove THCTensor_(expand2) and THCTensor_(expand3).

  Show description for c394a56
  

  gchanan

  authored and

  soumith

  committed
  c394a56

  Copy full SHA for c394a56

Commits on Jun 11, 2017

• Remove raiseErrors from THTensor functions, have THStorage functions take an error_buffer to return a proper error message while being able to handle memory management correctly from calling function.

  

  gchanan

  authored and

  soumith

  committed
  fa790ef

  Copy full SHA for fa790ef

• Incorporate review comments:

  Show description for f9df4f0
  

  gchanan

  authored and

  soumith

  committed
  f9df4f0

  Copy full SHA for f9df4f0

• Add broadcasting support for copy_, simplify code generation by moving a lot of currently generated code to expand_utils.

  

  gchanan

  authored and

  soumith

  committed
  747be77

  Copy full SHA for 747be77

• Support "fused" ops: addcmul/addcdiv.

  

  gchanan

  authored and

  soumith

  committed
  99dbaec

  Copy full SHA for 99dbaec

• Expand improvements

  Show description for 4817e53
  

  gchanan

  authored and

  soumith

  committed
  4817e53

  Copy full SHA for 4817e53

Commits on May 9, 2017

• Add keepdim to lua cwrap. (#763)

  

  gchanan

  authored and

  soumith

  committed
  640ea15

  Copy full SHA for 640ea15

• Add a keepdim parameter for reduction functions over a single dimension.

  Show description for 9acf4d4
  

  gchanan

  authored and

  soumith

  committed
  9acf4d4

  Copy full SHA for 9acf4d4

Commits on Apr 19, 2017

• Include THCNumerics.cuh in THCAtomics.cuh. (#752)

  

  gchanan

  authored and

  soumith

  committed
  76dc265

  Copy full SHA for 76dc265

Commits on Apr 12, 2017

• Fix abs with char and short cuda types. (#747)

  

  gchanan

  authored and

  soumith

  committed
  8f9a4fa

  Copy full SHA for 8f9a4fa

Commits on Feb 27, 2017

• Add getter for cuda device allocator.

  gchanan

  committed
  c7bc7af

  Copy full SHA for c7bc7af

Commits on Feb 15, 2017

• Fix gcc 4.4.7 build.

  gchanan

  committed
  f5c353e

  Copy full SHA for f5c353e

Commits on Feb 14, 2017

• Free set of stored streams and handle NULL streams.

  gchanan

  committed
  d386954

  Copy full SHA for d386954

Commits on Feb 8, 2017

• Make THCCachingHostAllocator less aggressive.

  Show description for 0159f88
  gchanan

  committed
  0159f88

  Copy full SHA for 0159f88

Commits on Feb 1, 2017

• Small fixups:

  Show description for 8ea68c2
  gchanan

  committed
  8ea68c2

  Copy full SHA for 8ea68c2

Commits on Jan 13, 2017

• Avoid strict aliasing warning in float/half conversions.

  gchanan

  committed
  1a09693

  Copy full SHA for 1a09693

Commits on Jan 11, 2017

• Run checkHalf in test setUp so it will be called from testers outside

  Show description for 2fd04c1
  gchanan

  committed
  2fd04c1

  Copy full SHA for 2fd04c1

Commits on Jan 10, 2017

• Add THCThrustAllocator.cuh to install files to downstream projects can use it.

  gchanan

  committed
  b08bdec

  Copy full SHA for b08bdec

• Re-route thrust memory allocation to THCudaMalloc / THCudaFree

  Show description for a3916f5
  gchanan

  committed
  a3916f5

  Copy full SHA for a3916f5

Commits on Jan 9, 2017

• Get rid of a few unused imports.

  gchanan

  committed
  bd9f612

  Copy full SHA for bd9f612

Commits on Dec 29, 2016

• Add THHalfTensor support to cutorch (#655)

  Show description for 1ac0668
  

  gchanan

  authored and

  soumith

  committed
  1ac0668

  Copy full SHA for 1ac0668

Commits on Nov 17, 2016

• Add half support for addmv and addr.

  gchanan

  committed
  cd8e209

  Copy full SHA for cd8e209

Commits on Nov 9, 2016

• Add THCTensor_(sizeDesc) for better debug messages.

  gchanan

  committed
  00096dc

  Copy full SHA for 00096dc

Commits on Oct 27, 2016

• Add hdot support for CUDA 8.

  Show description for 24faa05
  gchanan

  committed
  24faa05

  Copy full SHA for 24faa05

• Enable '*' for CudaDoubleTensors.

  gchanan

  committed
  429e10d

  Copy full SHA for 429e10d

Commits on Oct 19, 2016

• Add generic type support for toDeviceTensor.

  gchanan

  committed
  0d6ae18

  Copy full SHA for 0d6ae18

Commits on Oct 12, 2016

• Make atomicAdd functions static inline.

  gchanan

  committed
  1396909

  Copy full SHA for 1396909

Commits on Oct 7, 2016

• Move atomicAdd functions to THCAtomics.cuh in order to share

  Show description for 0012963
  gchanan

  committed
  0012963

  Copy full SHA for 0012963

Commits on Sep 28, 2016

• Add support for GenerateFloatTypes, for use with cunn.

  gchanan

  committed
  d21bc0a

  Copy full SHA for d21bc0a

Commits on Sep 19, 2016

• Use bitwise operations for atomicAdd rather than byte_perm or pointer deferences.

  Show description for 4727104
  gchanan

  committed
  4727104

  Copy full SHA for 4727104

Commits on Sep 16, 2016

• Add support for CUDA indexAdd

  Show description for 388d1cd
  gchanan

  committed
  388d1cd

  Copy full SHA for 388d1cd