Commits

Branch selector

 main

User selector

ezyang

Datepicker

All time

Commit History

Commits on Nov 8, 2023

• SymIntify {}_embedding{}_codegen_forward_{}{}_cuda autograd function

  Show description for e5236c8
  

  ezyang

  authored and

  facebook-github-bot

  committed
  e5236c8

  Copy full SHA for e5236c8

• SymIntify {}_embedding{}_codegen_forward_{}{}_cuda meta function

  Show description for 9b13b5a
  

  ezyang

  authored and

  facebook-github-bot

  committed
  9b13b5a

  Copy full SHA for 9b13b5a

Commits on Nov 7, 2023

• Derandomize all tests in fbgemm_gpu

  Show description for 0457bb7
  

  ezyang

  authored and

  facebook-github-bot

  committed
  0457bb7

  Copy full SHA for 0457bb7

Commits on Nov 2, 2023

• Add meta functions for {}_embedding{}_codegen_forward_{}{}_cuda (pytorch#2094)

  Show description for 824ef10
  

  ezyang

  authored and

  facebook-github-bot

  committed
  824ef10

  Copy full SHA for 824ef10

Commits on Oct 27, 2023

• Add some more tests for impl_abstract in other parts of test code (pytorch#2098)

  Show description for d4eec11
  

  ezyang

  authored and

  facebook-github-bot

  committed
  d4eec11

  Copy full SHA for d4eec11

Commits on Oct 26, 2023

• No-op impl_abstract if upstream PyTorch doesn't have it (pytorch#2097)

  Show description for 79f38e4
  

  ezyang

  authored and

  facebook-github-bot

  committed
  79f38e4

  Copy full SHA for 79f38e4

• Turn on opcheck for split_table_batched_embeddings_test (pytorch#2092)

  Show description for 6c2be88
  

  ezyang

  authored and

  facebook-github-bot

  committed
  6c2be88

  Copy full SHA for 6c2be88

• SymInt'ify all operator schemas in sparse_ops (pytorch#2089)

  Show description for bbc5358
  

  ezyang

  authored and

  facebook-github-bot

  committed
  bbc5358

  Copy full SHA for bbc5358

• impl_abstract expand_into_jagged_permute (pytorch#2090)

  Show description for 3283a68
  

  ezyang

  authored and

  facebook-github-bot

  committed
  3283a68

  Copy full SHA for 3283a68

• impl_abstract for permute_1D_sparse_data (pytorch#2087)

  Show description for 34f62ad
  

  ezyang

  authored and

  facebook-github-bot

  committed
  34f62ad

  Copy full SHA for 34f62ad

Commits on Oct 24, 2023

• Remove impl_abstract_pystub for now, while old PyTorch is in support window

  Show description for 97e6263
  

  ezyang

  authored and

  facebook-github-bot

  committed
  97e6263

  Copy full SHA for 97e6263

• Add scaffolding for Python impl_abstract in fbgemm, implement fbgemm.permute_1D_sparse_data (pytorch#2084)

  Show description for f94254d
  

  ezyang

  authored and

  facebook-github-bot

  committed
  f94254d

  Copy full SHA for f94254d

Commits on Sep 27, 2023

• More removal of const ref on SymInt in dispatcher call sites (pytorch#2041)

  Show description for 0346155
  

  ezyang

  authored and

  facebook-github-bot

  committed
  0346155

  Copy full SHA for 0346155

Commits on Sep 25, 2023

• Fix incorrect SymInt signature on dense_to_jagged (pytorch#2039)

  Show description for 85de33b
  

  ezyang

  authored and

  facebook-github-bot

  committed
  85de33b

  Copy full SHA for 85de33b

Commits on Aug 11, 2023

• Don't call avx512 implementation if we didn't build it (pytorch#1930)

  Show description for 1b2746f
  

  ezyang

  authored and

  facebook-github-bot

  committed
  1b2746f

  Copy full SHA for 1b2746f

Commits on Aug 10, 2023

• Minimum gcc version is 8+ (pytorch#1924)

  Show description for bc34501
  

  ezyang

  authored and

  facebook-github-bot

  committed
  bc34501

  Copy full SHA for bc34501

Commits on Aug 3, 2023

• Add meta for asynchronous_complete_cumsum (pytorch#1905)

  Show description for eb64159
  

  ezyang

  authored and

  facebook-github-bot

  committed
  eb64159

  Copy full SHA for eb64159

Commits on May 12, 2023

• SymIntify jagged_dense_bmm_forward_meta (pytorch#1756)

  Show description for 520d1ea
  

  ezyang

  authored and

  facebook-github-bot

  committed
  520d1ea

  Copy full SHA for 520d1ea

Commits on Mar 23, 2021

• Use more compact syntax for is_uvm_tensor (pytorch#577)

  Show description for 1d7eaf3
  

  ezyang

  authored and

  facebook-github-bot

  committed
  1d7eaf3

  Copy full SHA for 1d7eaf3

Commits on Mar 21, 2020

• Revert D16234008: Quantize should work even if the value after quantization becomes inf

  Show description for ab431ad
  

  ezyang

  authored and

  jspark1105

  committed
  ab431ad

  Copy full SHA for ab431ad

• Revert D18566687: Fix the OSS build issue for FP16Benchmark when MKL/BLAS is not available

  Show description for fdb4a99
  

  ezyang

  authored and

  jspark1105

  committed
  fdb4a99

  Copy full SHA for fdb4a99

Commits on Sep 4, 2019

• Revert D16968373: Introduced CodeCache container to share the microkernels among different threads.

  Show description for f9078fd
  

  ezyang

  authored and

  facebook-github-bot

  committed
  f9078fd

  Copy full SHA for f9078fd