Commits

Breadcrumbs

History for

onthrow_js

User selector

All users

All time

Commit History

Commits on Oct 27, 2019

• webgpu: Enable rectangle workgroup size for matMul (#2256)

  Show description for b66bded
  

  qjia7

  authored and

  annxingyuan

  committed
  b66bded

  Copy full SHA for b66bded

Commits on Oct 22, 2019

• Convert all usages of ENV in all packages to env(). (#2251)

  Show description for 74b93f0
  

  Nikhil Thorat

  authored
  74b93f0

  Copy full SHA for 74b93f0

• webgpu: Add conv2dByMatMul (#2242)

  Show description for 86b2713
  

  qjia7

  authored and

  annxingyuan

  committed
  86b2713

  Copy full SHA for 86b2713

Commits on Oct 18, 2019

• [webgpu] Make condition for not checking out of bounds stricter. (#2223)

  Show description for e60d583
  annxingyuan

  authored
  e60d583

  Copy full SHA for e60d583

Commits on Oct 16, 2019

• [webgpu] Avoid checking coordIsValid where possible. (#2220)

  Show description for 4230a7a
  annxingyuan

  authored
  4230a7a

  Copy full SHA for 4230a7a

Commits on Oct 10, 2019

• [webgpu] Add CPU forwarding flag, upgrade tfjs-core dep. (#2176)

  Show description for 089eb23
  annxingyuan

  authored
  089eb23

  Copy full SHA for 089eb23

Commits on Oct 8, 2019

• webgpu: Optimize concat/resizeBilinear/transpose (#2115)

  Show description for 6171e2f
  

  qjia7

  authored and

  annxingyuan

  committed
  6171e2f

  Copy full SHA for 6171e2f

• webgpu: Implement clip,less ops (#2118)

  Show description for dea738a
  

  qjia7

  authored and

  annxingyuan

  committed
  dea738a

  Copy full SHA for dea738a

Commits on Sep 30, 2019

• Fallback to cpu when the inputs size is small (#2061)

  Show description for 45e30c7
  

  axinging

  authored and

  annxingyuan

  committed
  45e30c7

  Copy full SHA for 45e30c7

• webgpu: Optimize maxpool with groupsize(4, 4, 4) (#2112)

  Show description for 6b4e24f
  

  qjia7

  authored and

  annxingyuan

  committed
  6b4e24f

  Copy full SHA for 6b4e24f

Commits on Sep 26, 2019

• webgpu: Use DataStorage class for storing tensor map. (#2095)

  Show description for 45d6cbe
  annxingyuan

  authored
  45d6cbe

  Copy full SHA for 45d6cbe

Commits on Sep 25, 2019

• webgpu: Lazily upload tensor data to SSBOs. (#2085)

  Show description for 17b9aef
  annxingyuan

  authored
  17b9aef

  Copy full SHA for 17b9aef

• webgpu: Fix argMax benchmark doesn't work (#2087)

  Show description for dd34663
  

  qjia7

  authored and

  annxingyuan

  committed
  dd34663

  Copy full SHA for dd34663

Commits on Sep 18, 2019

• Revert "Replace mul+add with fma" (#2060)

  Show description for 035319a
  

  axinging

  authored and

  annxingyuan

  committed
  035319a

  Copy full SHA for 035319a

Commits on Sep 17, 2019

• Replace mul+add with fma (#2047)

  Show description for 9529b15
  

  axinging

  authored and

  annxingyuan

  committed
  9529b15

  Copy full SHA for 9529b15

• Implement operator greater and greaterEqual (#2028)

  Show description for 0ffc2b5
  

  axinging

  authored and

  annxingyuan

  committed
  0ffc2b5

  Copy full SHA for 0ffc2b5

Commits on Sep 13, 2019

• Bring back the dist importing lint rule (#2039)

  Show description for 6d3a9a9
  

  Nikhil Thorat

  authored
  6d3a9a9

  Copy full SHA for 6d3a9a9

• webgpu: Upgrade tfjs-core dependency. (#2038)

  Show description for d3d805a
  annxingyuan

  authored
  d3d805a

  Copy full SHA for d3d805a

Commits on Sep 12, 2019

• webgpu: Add realDivide (#2026)

  Show description for 06292c7
  

  haoyunfeix

  authored and

  annxingyuan

  committed
  06292c7

  Copy full SHA for 06292c7

• webgpu: Replace uint with int (#2025)

  Show description for 26de956
  

  haoyunfeix

  authored and

  annxingyuan

  committed
  26de956

  Copy full SHA for 26de956

Commits on Sep 11, 2019

• webgpu: Increase pad operation work group size to 16. (#2018)

  Show description for df6f3c8
  annxingyuan

  authored
  df6f3c8

  Copy full SHA for df6f3c8

• webgpu: Use (16, 1, 1) as the work group size for binary and un… (#2013)

  Show description for eef8a32
  

  qjia7

  authored and

  annxingyuan

  committed
  eef8a32

  Copy full SHA for eef8a32

Commits on Sep 9, 2019

• webgpu: Fix the compile error in conv2d_naive shader (#2002)

  Show description for bff4539
  

  qjia7

  authored and

  annxingyuan

  committed
  bff4539

  Copy full SHA for bff4539

Commits on Sep 5, 2019

• webgpu: Don't release the uniform buffer under delayed mode (#1968)

  Show description for ca35b08
  

  qjia7

  authored and

  annxingyuan

  committed
  ca35b08

  Copy full SHA for ca35b08

Commits on Sep 3, 2019

• webgpu: Bump dependency and version. (#1962)

  Show description for fdb923d
  annxingyuan

  authored
  fdb923d

  Copy full SHA for fdb923d

Commits on Aug 23, 2019

• webgpu: Filter tests by env.specFilter. (#1869)

  Show description for 06a2659
  annxingyuan

  authored
  06a2659

  Copy full SHA for 06a2659

Commits on Aug 22, 2019

• [global] Switch half the projects to Typescript 3.5.3 and add a… (#1886)

  Show description for b42d122
  dsmilkov

  authored
  b42d122

  Copy full SHA for b42d122

• webgpu: Increase WPT for pad. (#1888)

  Show description for 11abc43
  annxingyuan

  authored
  11abc43

  Copy full SHA for 11abc43

Commits on Aug 20, 2019

• Add depthwiseConv2D op for webgpu (#1843)

  Show description for 0f05dbc
  

  axinging

  authored and

  annxingyuan

  committed
  0f05dbc

  Copy full SHA for 0f05dbc

Commits on Aug 19, 2019

• webgpu: Increase work per thread for binary / unary ops. (#1845)

  Show description for 4858434
  annxingyuan

  authored
  4858434

  Copy full SHA for 4858434

Commits on Aug 15, 2019

• tfjs-webgpu: Rename tfjs-webgpu to tfjs-backend-webgpu. (#1839)

  Show description for b9e43bb
  annxingyuan

  authored
  b9e43bb

  Copy full SHA for b9e43bb

Loading