Commits

Branch selector

 master

User selector

All users

All time

Commit History

Commits on Feb 21, 2022

• Remove double underscore from identifiers

  

  9prady9

  authored and

  umar456

  committed
  24665c7

  Copy full SHA for 24665c7

• Use boost's epsilon difference when comparing floating point values

  Show description for 3330090
  umar456

  committed
  3330090

  Copy full SHA for 3330090

• Use c++11 when building tests

  umar456

  committed
  bb89234

  Copy full SHA for bb89234

Commits on Feb 18, 2022

• intl/uintl versions of select/replace to fix accuracy issues

  Show description for 13cbff8
  9prady9

  committed
  13cbff8

  Copy full SHA for 13cbff8

Commits on Feb 12, 2022

• Update windows GA job to use new VS toolchain for respective GH image

  Show description for f58b849
  9prady9

  committed
  f58b849

  Copy full SHA for f58b849

• Fix mkl_version inclusion guard in CPU and OpenCL backends

  Show description for d2b09a6
  9prady9

  committed
  d2b09a6

  Copy full SHA for d2b09a6

Commits on Feb 7, 2022

• Add CUDA 11.5 to max compute and compute capability arrays (#3203)

  Show description for 970f32a
  

  umar456

  and

  9prady9

  authored
  970f32a

  Copy full SHA for 970f32a

Commits on Oct 15, 2021

• Disk hash is now based on the full code + options, also for JIT code.

  

  willyborn

  authored and

  umar456

  committed
  1d03d07

  Copy full SHA for 1d03d07

Commits on Oct 14, 2021

• Update Intel MKL to oneMKL on github ci jobs

  

  pradeep

  authored and

  9prady9

  committed
  1ff07ca

  Copy full SHA for 1ff07ca

• Use appropriate MKL getrs_batch_strided API based on MKL Versions

  

  pradeep

  authored and

  9prady9

  committed
  72b73ff

  Copy full SHA for 72b73ff

Commits on Oct 13, 2021

• Allow moddims operations to be part of the JIT tree if possible (#3177)

  Show description for a800d9f
  umar456

  authored
  a800d9f

  Copy full SHA for a800d9f

Commits on Oct 5, 2021

• Update CLBlast to fix some errors on Turing cards

  

  umar456

  authored and

  9prady9

  committed
  1ad0400

  Copy full SHA for 1ad0400

• Fix default parameters for fftR2C and fftC2R from 0 to 1.0

  

  umar456

  authored and

  9prady9

  committed
  451331d

  Copy full SHA for 451331d

Commits on Oct 1, 2021

• fix transform operator for countByKey (#3175)

  Show description for 2444ef5
  syurkevi

  authored
  2444ef5

  Copy full SHA for 2444ef5

Commits on Sep 30, 2021

• Add assert that check if topk is called with a negative value for k

  umar456

  committed
  6fbf5fb

  Copy full SHA for 6fbf5fb

Commits on Sep 24, 2021

• Improve offline build experience for developers (#3162)

  Show description for 7995750
  9prady9

  authored
  7995750

  Copy full SHA for 7995750

Commits on Aug 18, 2021

• Fix reference count if array used in JIT operations.

  Show description for a57b291
  

  umar456

  authored and

  pradeep

  committed
  a57b291

  Copy full SHA for a57b291

• Create a hash function for Node objects for NodeMap

  Show description for fad0bce
  

  umar456

  authored and

  pradeep

  committed
  fad0bce

  Copy full SHA for fad0bce

• Use getArray instead of castArray if types are the same in arithOp

  

  umar456

  authored and

  pradeep

  committed
  fb23fc9

  Copy full SHA for fb23fc9

• Move cast and castArray to the common directory

  

  umar456

  authored and

  pradeep

  committed
  ea52651

  Copy full SHA for ea52651

• Move createBinaryNode to common

  

  umar456

  authored and

  pradeep

  committed
  92dd704

  Copy full SHA for 92dd704

• Add ASSERT_REF to check for reference counts of af::arrays

  

  umar456

  authored and

  pradeep

  committed
  1ce9429

  Copy full SHA for 1ce9429

Commits on Aug 17, 2021

• Improve canny's otsu helper by precomputing some arrays

  Show description for 4ea695f
  

  pradeep

  and

  umar456

  committed
  4ea695f

  Copy full SHA for 4ea695f

• Fix edgeTracking CPU kernel to handle batch support

  Show description for e7f000d
  

  pradeep

  authored and

  umar456

  committed
  e7f000d

  Copy full SHA for e7f000d

• Fix canny by resizing the sigma array to the correct size

  Show description for f6b06b7
  umar456

  committed
  f6b06b7

  Copy full SHA for f6b06b7

• Improve Readme (#3168)

  Show description for 7ddf462
  

  4 peopleauthored
  7ddf462

  Copy full SHA for 7ddf462

• Create ASSERT_IMAGE_NEAR which compares two images for equality

  Show description for 1b95366
  

  umar456

  authored and

  pradeep

  committed
  1b95366

  Copy full SHA for 1b95366

Commits on Jul 31, 2021

• Check symbols in MKL to enable solve batch functionality

  Show description for b6680d5
  

  umar456

  authored and

  pradeep

  committed
  b6680d5

  Copy full SHA for b6680d5

• Use correct assert macros in solve_dense tests

  

  pradeep

  committed
  dc2e023

  Copy full SHA for dc2e023

• Improve disabled linear algebra error message. minor header updates

  

  umar456

  authored and

  pradeep

  committed
  d14977f

  Copy full SHA for d14977f

• Allow MKL as a valid entry for AF_COMPUTE_LIBRARY

  

  umar456

  authored and

  pradeep

  committed
  2e54562

  Copy full SHA for 2e54562

• Use pinned memory to copy device pointers in CUDA solve

  

  umar456

  authored and

  pradeep

  committed
  854e5f3

  Copy full SHA for 854e5f3

• Add batch support for CPU and OpenCL solve. Add solver batch tests

  Show description for c826ddf
  

  umar456

  authored and

  pradeep

  committed
  c826ddf

  Copy full SHA for c826ddf

• Adding support for batched solve in CUDA backend

  

  pavanky

  authored and

  pradeep

  committed
  0d9ffc2

  Copy full SHA for 0d9ffc2

Commits on Jul 30, 2021

• Move array death test into a separate serially executed test

  Show description for 6633e10
  

  pradeep

  committed
  6633e10

  Copy full SHA for 6633e10