Commits

Branch selector

 main

User selector

SS-JIA

Datepicker

All time

Commit History

Commits on Dec 20, 2023

• [pt-vulkan] Enable Python code blocks in shader templates and upgrade shader template generation (pytorch#115948)

  Show description for 545d212
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  545d212

  Copy full SHA for 545d212

Commits on Oct 20, 2023

• [PyTorch][Vulkan] Allow 0-size tensors to be represented in PyTorch Vulkan (pytorch#111512)

  Show description for 8df42f9
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  8df42f9

  Copy full SHA for 8df42f9

Commits on Oct 11, 2023

• [core ATen IR] Add decompositions for max, min, var_mean (pytorch#110906)

  Show description for 9606cda
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  9606cda

  Copy full SHA for 9606cda

Commits on Oct 10, 2023

• [ATen core IR] De-register full_like and empty_like as core (pytorch#110924)

  Show description for a425307
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  a425307

  Copy full SHA for a425307

Commits on Oct 9, 2023

• [ATen core IR] Register additional ATen operators as core (pytorch#110882)

  Show description for 6c8096e
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  6c8096e

  Copy full SHA for 6c8096e

Commits on Oct 7, 2023

• [core IR] Add decomps for aten.sum and aten.squeeze variants (pytorch#110645)

  Show description for c2e7a0d
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  c2e7a0d

  Copy full SHA for c2e7a0d

Commits on Oct 3, 2023

• [core IR][reland] Add split.Tensor and unbind decompositions to core ATen decomp table (pytorch#110323)

  Show description for ff96f6d
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  ff96f6d

  Copy full SHA for ff96f6d

Commits on Sep 28, 2023

• [core IR] Add lift_fresh, split.Tensor, and unbind decompositions to core ATen decomp table (pytorch#110102)

  Show description for 22e706f
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  22e706f

  Copy full SHA for 22e706f

Commits on Sep 27, 2023

• [core IR] Add a core decomposition for aten.all (pytorch#110093)

  Show description for dec140f
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  dec140f

  Copy full SHA for dec140f

• [core IR] Add glu as a core decomposition (pytorch#110043)

  Show description for 9928c10
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  9928c10

  Copy full SHA for 9928c10

Commits on Sep 26, 2023

• [core IR] Add a core decomposition for floor_divide (pytorch#110046)

  Show description for 5df8aca
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  5df8aca

  Copy full SHA for 5df8aca

Commits on Sep 25, 2023

• [core IR] Remove trunc decomp and add trunc to core (pytorch#109902)

  Show description for 7de669f
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  7de669f

  Copy full SHA for 7de669f

Commits on Sep 22, 2023

• [core IR] Add div.Tensor_mode, div.Scalar_mode, and copy as core operators (pytorch#109812)

  Show description for e42d450
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  e42d450

  Copy full SHA for e42d450

Commits on Aug 25, 2023

• Avoid decomposing _unsafe_index in Inductor (pytorch#107882)

  Show description for 86f9fec
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  86f9fec

  Copy full SHA for 86f9fec

Commits on Aug 10, 2023

• [Vulkan] Introduce GPU Memory Layout qualifier (pytorch#106978)

  Show description for 62b3018
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  62b3018

  Copy full SHA for 62b3018

• Add _unsafe_index decomp (pytorch#106814)

  Show description for 8c8477e
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  8c8477e

  Copy full SHA for 8c8477e

Commits on Jul 27, 2023

• Tag functions to core IR in native_functions.yaml (pytorch#105849)

  Show description for 70bc1b0
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  70bc1b0

  Copy full SHA for 70bc1b0

Commits on May 22, 2023

• [vulkan] Fix concat op in feature dimension (pytorch#101721)

  Show description for c618093
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  c618093

  Copy full SHA for c618093

Commits on Mar 17, 2023

• [PyTorch][XNNPACK] Update wrappers for internal only x86 SSE2 kernels (pytorch#96896)

  Show description for 4de1bc1
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  4de1bc1

  Copy full SHA for 4de1bc1

Commits on Feb 23, 2023

• [vulkan] Pad channels when using texture storage instead of "tight packing" (pytorch#95251)

  Show description for 0eeb046
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  0eeb046

  Copy full SHA for 0eeb046

Commits on Feb 7, 2023

• [vulkan] Add core graph components (pytorch#94222)

  Show description for 6b8eb0e
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  6b8eb0e

  Copy full SHA for 6b8eb0e

Commits on Jan 30, 2023

• [vulkan] Enable command buffer reuse and add keys to Tensor/StorageBuffer objects (pytorch#92993)

  Show description for 7dabb8b
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  7dabb8b

  Copy full SHA for 7dabb8b

• [vulkan] Create separate BUCK target for command buffer recording (pytorch#92157)

  Show description for e17bfde
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  e17bfde

  Copy full SHA for e17bfde

Commits on Jan 5, 2023

• [vulkan] Move Tensor.* from ops/ folder to api/ folder (pytorch#91033)

  Show description for 13b3d86
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  13b3d86

  Copy full SHA for 13b3d86

• [vulkan] Remove dependencies from op/ in vTensor and move it to higher level namespace (pytorch#91023)

  Show description for aa562f9
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  aa562f9

  Copy full SHA for aa562f9

• [vulkan] Remove ATen dependencies in vTensor class (pytorch#91022)

  Show description for ca62ed9
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  ca62ed9

  Copy full SHA for ca62ed9

• [vulkan] Remove external dependencies in core API and introduce torch_vulkan_api target (pytorch#91021)

  Show description for 64a3738
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  64a3738

  Copy full SHA for 64a3738

Commits on Nov 9, 2022

• [vulkan] Add option for buffer representations in vTensor (pytorch#87622)

  Show description for c4a3aa8
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  c4a3aa8

  Copy full SHA for c4a3aa8

Commits on Nov 4, 2022

• [vulkan] enable prepacking for Batchnorm op (pytorch#88433)

  Show description for b78b872
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  b78b872

  Copy full SHA for b78b872

Commits on Oct 18, 2022

• [vulkan] Use 2D texture types for convolution weights and biases (pytorch#86972)

  Show description for 9d1a8ed
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  9d1a8ed

  Copy full SHA for 9d1a8ed

• [vulkan] Enable 2D texture types (pytorch#86971)

  Show description for 5b58803
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  5b58803

  Copy full SHA for 5b58803

Commits on Oct 11, 2022

• [vulkan][ez] fix always printing out a warning when retrieving the global context (pytorch#86697)

  Show description for 5ffe24f
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  5ffe24f

  Copy full SHA for 5ffe24f

Commits on Oct 8, 2022

• [vulkan] Add buffer to buffer copies (pytorch#86424)

  Show description for b0ceb8e
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  b0ceb8e

  Copy full SHA for b0ceb8e

• [vulkan] Clean up convolution code (pytorch#86423)

  Show description for 511d81c
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  511d81c

  Copy full SHA for 511d81c

Commits on Sep 27, 2022

• [vulkan] Rewrite prepacking functions using aten functions + some code cleanup (pytorch#84973)

  Show description for cf5699f
  

  SS-JIA

  authored and

  pytorchmergebot

  committed
  cf5699f

  Copy full SHA for cf5699f

Pagination

Previous

Next