Commits

Branch selector

 main

User selector

desertfire

Datepicker

All time

Commit History

Commits on Oct 3, 2024

• [Inductor] Use parametrize to break down some unit tests (pytorch#137156)

  Show description for 4513fb5
  

  desertfire

  authored and

  pytorchmergebot

  committed
  4513fb5

  Copy full SHA for 4513fb5

Commits on Oct 2, 2024

• [CI] Remove nanogpt from perf smoke test (pytorch#137176)

  Show description for bcaa0f5
  

  desertfire

  authored and

  pytorchmergebot

  committed
  bcaa0f5

  Copy full SHA for bcaa0f5

• [Inductor] Fix test_conv2d_unary_cpu_cpp_wrapper failure (pytorch#137158)

  Show description for 5c2c3ca
  

  desertfire

  authored and

  pytorchmergebot

  committed
  5c2c3ca

  Copy full SHA for 5c2c3ca

• [Inductor] Enable a SDPA pattern matching for CUDA (pytorch#137085)

  Show description for 8e39f2a
  

  desertfire

  authored and

  pytorchmergebot

  committed
  8e39f2a

  Copy full SHA for 8e39f2a

Commits on Sep 30, 2024

• [AOTI] Update sam_fast from timeout to fail_to_run (pytorch#136996)

  Show description for a15f3f5
  

  desertfire

  authored and

  pytorchmergebot

  committed
  a15f3f5

  Copy full SHA for a15f3f5

• [Inductor] Switch cpp_wrapper tests to ABI-compatible (pytorch#136904)

  Show description for d6d9183
  

  desertfire

  authored and

  pytorchmergebot

  committed
  d6d9183

  Copy full SHA for d6d9183

• [AOTI] Support test_open_device_registration in ABI-compatible (pytorch#136906)

  Show description for ad8fae2
  

  desertfire

  authored and

  pytorchmergebot

  committed
  ad8fae2

  Copy full SHA for ad8fae2

Commits on Sep 27, 2024

• [Inductor] Further solve missing aoti_torch_check symbole issue (pytorch#136775)

  Show description for 26869d3
  

  desertfire

  authored and

  pytorchmergebot

  committed
  26869d3

  Copy full SHA for 26869d3

Commits on Sep 26, 2024

• [AOTI] Set CUDA device for torch._export.aot_load (pytorch#136715)

  Show description for da5c7b6
  

  desertfire

  authored and

  pytorchmergebot

  committed
  da5c7b6

  Copy full SHA for da5c7b6

• [CI] Switch aarch64 dashboard run back to nightly (pytorch#136643)

  Show description for 5c78c6b
  

  desertfire

  authored and

  pytorchmergebot

  committed
  5c78c6b

  Copy full SHA for 5c78c6b

Commits on Sep 25, 2024

• [AOTI] Fix a missing aoti_torch_check symbol issue (pytorch#136669)

  Show description for ecc15c4
  

  desertfire

  authored and

  pytorchmergebot

  committed
  ecc15c4

  Copy full SHA for ecc15c4

• [AOTI] Move stack_allocation logic from PythonWrapperCodegen (pytorch#136463)

  Show description for 20a855b
  

  desertfire

  authored and

  pytorchmergebot

  committed
  20a855b

  Copy full SHA for 20a855b

• [AOTI] Move stack_allocation logic from CppWrapperCpu (pytorch#136462)

  Show description for 070952a
  

  desertfire

  authored and

  pytorchmergebot

  committed
  070952a

  Copy full SHA for 070952a

• [AOTI][reland] Create another wrapper class to handle ArrayRef (pytorch#136461)

  Show description for 5ad5f40
  

  desertfire

  authored and

  pytorchmergebot

  committed
  5ad5f40

  Copy full SHA for 5ad5f40

• [AOTI] Support MKL linear ops in cpp wrapper (pytorch#134974)

  Show description for 1c9a1a2
  

  2 peopleauthored and

  pytorchmergebot

  committed
  1c9a1a2

  Copy full SHA for 1c9a1a2

Commits on Sep 24, 2024

• [Inductor] Rename WrapperCodeGen to PythonWrapperCodegen (pytorch#136062)

  Show description for 95c0f74
  

  desertfire

  authored and

  pytorchmergebot

  committed
  95c0f74

  Copy full SHA for 95c0f74

Commits on Sep 23, 2024

• [AOTI] Create another wrapper class to handle ArrayRef (pytorch#136318)

  Show description for d21841d
  

  desertfire

  authored and

  pytorchmergebot

  committed
  d21841d

  Copy full SHA for d21841d

Commits on Sep 20, 2024

• [Inductor] Fix test_profiler_mark_wrapper_call_cuda_cuda_wrapper (pytorch#136356)

  Show description for f53a0f9
  

  desertfire

  authored and

  pytorchmergebot

  committed
  f53a0f9

  Copy full SHA for f53a0f9

Commits on Sep 16, 2024

• [reland][Inductor] Rename cpp_wrapper_cuda.py as cpp_wrapper_gpu.py (pytorch#136046)

  Show description for d833f49
  

  desertfire

  authored and

  pytorchmergebot

  committed
  d833f49

  Copy full SHA for d833f49

• [AOTI] Refactor how cpp_wrapper specific options are set (pytorch#136035)

  Show description for a803cb0
  

  desertfire

  authored and

  pytorchmergebot

  committed
  a803cb0

  Copy full SHA for a803cb0

Commits on Sep 14, 2024

• [AOTI] Fix a fallback op returning None issue (pytorch#135997)

  Show description for b4c84c3
  

  desertfire

  authored and

  pytorchmergebot

  committed
  b4c84c3

  Copy full SHA for b4c84c3

Commits on Sep 13, 2024

• [AOTI][reland] Fix assert_function call in cpu autotune template (pytorch#135920)

  Show description for ea2ecab
  

  desertfire

  authored and

  pytorchmergebot

  committed
  ea2ecab

  Copy full SHA for ea2ecab

Commits on Sep 11, 2024

• [Inductor] add _dynamo.reset to test_cat_slice_cat_cuda (pytorch#135694)

  Show description for 118d7e1
  

  desertfire

  authored and

  pytorchmergebot

  committed
  118d7e1

  Copy full SHA for 118d7e1

Commits on Sep 9, 2024

• [AOTI] Fix assert_function call in cpu autotune template (pytorch#135086)

  Show description for 16c3b8f
  

  desertfire

  authored and

  pytorchmergebot

  committed
  16c3b8f

  Copy full SHA for 16c3b8f

• [AOTI] Add C shim for aten.mkldnn_rnn_layer in cpp wrapper (pytorch#134857)

  Show description for 9c6dff4
  

  desertfire

  authored and

  pytorchmergebot

  committed
  9c6dff4

  Copy full SHA for 9c6dff4

Commits on Sep 6, 2024

• [AOTI] Support MKLDNN qconv ops in cpp wrapper (pytorch#134795)

  Show description for 1e57ef0
  

  desertfire

  authored and

  pytorchmergebot

  committed
  1e57ef0

  Copy full SHA for 1e57ef0

• [AOTI] Support MKLDNN qlinear ops in cpp wrapper (pytorch#134783)

  Show description for 614b86d
  

  desertfire

  authored and

  pytorchmergebot

  committed
  614b86d

  Copy full SHA for 614b86d

• [AOTI] Support MKLDNN conv ops in cpp wrapper (pytorch#134475)

  Show description for 0b96dfb
  

  desertfire

  authored and

  pytorchmergebot

  committed
  0b96dfb

  Copy full SHA for 0b96dfb

Commits on Sep 5, 2024

• [AOTI] Fix a unbacked symint retrieve bug (pytorch#134670)

  Show description for 50d1e37
  

  desertfire

  authored and

  pytorchmergebot

  committed
  50d1e37

  Copy full SHA for 50d1e37

Commits on Aug 29, 2024

• [AOTI] Switch benchmarking to use export non-strict mode (pytorch#130977)

  Show description for 387d3fc
  

  desertfire

  authored and

  pytorchmergebot

  committed
  387d3fc

  Copy full SHA for 387d3fc

Commits on Aug 28, 2024

• [AOTI] Fix test_aoti_inference CPU build issue (pytorch#134675)

  Show description for 310eb6d
  

  desertfire

  authored and

  pytorchmergebot

  committed
  310eb6d

  Copy full SHA for 310eb6d

• [Inductor][Refactor] Rename CPU benchmark test configs (pytorch#134639)

  Show description for e6bf171
  

  desertfire

  authored and

  pytorchmergebot

  committed
  e6bf171

  Copy full SHA for e6bf171

Commits on Aug 27, 2024

• [AOTI] Introduce DeferredCudaGridLine for cuda cpp wrapper (pytorch#129268)

  Show description for a4b44dd
  

  desertfire

  authored and

  pytorchmergebot

  committed
  a4b44dd

  Copy full SHA for a4b44dd

Commits on Aug 23, 2024

• [CI][dashboard] Change aarch64 perf run (pytorch#134265)

  Show description for 7c93c4f
  

  desertfire

  authored and

  pytorchmergebot

  committed
  7c93c4f

  Copy full SHA for 7c93c4f

Commits on Aug 22, 2024

• [CI][dashboard] Add jemalloc back for aarch64 (pytorch#134189)

  Show description for f0ba309
  

  desertfire

  authored and

  pytorchmergebot

  committed
  f0ba309

  Copy full SHA for f0ba309

Pagination

Previous

Next