Commits

Branch selector

 main

User selector

shintaro-iwasaki

Datepicker

All time

Commit History

Commits on Mar 1, 2024

• Support CPU for split_table_batched_embeddings_benchmark with device

  Show description for f5ffa86
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  f5ffa86

  Copy full SHA for f5ffa86

Commits on Mar 22, 2023

• jagged_dense_bmm: fix ROCm test failures (pytorch#1655)

  Show description for f3af571
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  f3af571

  Copy full SHA for f3af571

Commits on Jan 23, 2023

• fix a corner case of keyed_jagged_index_select_dim1 (pytorch#1554)

  Show description for 4cdb3f4
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  4cdb3f4

  Copy full SHA for 4cdb3f4

Commits on Jan 20, 2023

• Fix test_amd_gpu (pytorch#1552)

  Show description for 5e029f8
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  5e029f8

  Copy full SHA for 5e029f8

Commits on Jan 19, 2023

• Fix uvm_test.py when a machine has only 1 GPU (pytorch#1549)

  Show description for 0237a8a
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  0237a8a

  Copy full SHA for 0237a8a

Commits on Jan 14, 2023

• Add secret inheritance setting to pass PYPI_TOKEN (pytorch#1546)

  Show description for bf6c1ef
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  bf6c1ef

  Copy full SHA for bf6c1ef

• Run NVIDIA GPU tests on A10 per PR (pytorch#1544)

  Show description for 3848313
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  3848313

  Copy full SHA for 3848313

Commits on Jan 12, 2023

• Set PYPI_TOKEN in build_wheel.yml (pytorch#1542)

  Show description for dc328f2
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  dc328f2

  Copy full SHA for dc328f2

Commits on Jan 11, 2023

• Update the nightly-release mechanism (pytorch#1541)

  Show description for 92ad9bc
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  92ad9bc

  Copy full SHA for 92ad9bc

Commits on Jan 9, 2023

• Enhance FBGEMM nightly CI (pytorch#1407)

  Show description for a8eeac7
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  a8eeac7

  Copy full SHA for a8eeac7

Commits on Jan 6, 2023

• Remove CircleCI config (pytorch#1539)

  Show description for ea859ee
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  ea859ee

  Copy full SHA for ea859ee

Commits on Jan 4, 2023

• Use ROCm 5.2 (pytorch#1537)

  Show description for 9132dd3
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  9132dd3

  Copy full SHA for 9132dd3

Commits on Oct 21, 2022

• update cuDNN installation method (pytorch#1415)

  Show description for edbeb2b
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  edbeb2b

  Copy full SHA for edbeb2b

Commits on Sep 24, 2022

• Update the PyTorch installation method (pytorch#1349)

  Show description for aefa24c
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  aefa24c

  Copy full SHA for aefa24c

Commits on Sep 13, 2022

• fix FBGEMM compilation for AMD GPUs (pytorch#1299)

  Show description for 4057a76
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  4057a76

  Copy full SHA for 4057a76

Commits on Aug 16, 2022

• fix a 137 error in sparse_ops_test.py (pytorch#1247)

  Show description for 868946b
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  868946b

  Copy full SHA for 868946b

Commits on Aug 3, 2022

• support multiple CUDA versions (pytorch#1234)

  Show description for 5160e7d
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  5160e7d

  Copy full SHA for 5160e7d

Commits on Jul 26, 2022

• add --no-dkms to install ROCm on CI (pytorch#1220)

  Show description for 51ea00e
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  51ea00e

  Copy full SHA for 51ea00e

Commits on May 23, 2022

• reduce test size for test_expand_into_jagged_permute (pytorch#1132)

  Show description for 3c1fc79
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  3c1fc79

  Copy full SHA for 3c1fc79

• skip cudaMemcpy2DAsync() if no copy is needed (pytorch#1133)

  Show description for b0f903f
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  b0f903f

  Copy full SHA for b0f903f

Commits on May 20, 2022

• minor refactoring (pytorch#1130)

  Show description for 9aab56d
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  9aab56d

  Copy full SHA for 9aab56d

Commits on Apr 28, 2022

• use shfl_sync instead of __shfl_sync (pytorch#1080)

  Show description for 7fcc520
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  7fcc520

  Copy full SHA for 7fcc520

Commits on Apr 27, 2022

• allow FP16-type grad_t (pytorch#1072)

  Show description for 2d16f61
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  2d16f61

  Copy full SHA for 2d16f61

Commits on Jan 31, 2022

• reduce compilation size and time (pytorch#886)

  Show description for bf48eea
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  bf48eea

  Copy full SHA for bf48eea

Commits on Jan 25, 2022

• remove grad_t to reduce binary size + compilation time (pytorch#882)

  Show description for f5bff95
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  f5bff95

  Copy full SHA for f5bff95

Commits on Jan 19, 2022

• add bfloat16 gradient option to split_table_batched_embeddings_benchmark (2/2)

  Show description for 48515f9
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  48515f9

  Copy full SHA for 48515f9

Commits on Jan 13, 2022

• move ATen/CUDAGeneratorImpl.h to ATen/cuda (#71224)

  Show description for eb17d45
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  eb17d45

  Copy full SHA for eb17d45

Commits on Jan 4, 2022

• add bfloat16 gradient option to split_table_batched_embeddings_benchmark (pytorch#721)

  Show description for 127f813
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  127f813

  Copy full SHA for 127f813

Commits on Dec 24, 2021

• Add output_dtype to training TBE op for CPU (pytorch#829)

  Show description for d1907f1
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  d1907f1

  Copy full SHA for d1907f1

Commits on Dec 6, 2021

• open-source segment_sum_csr() (pytorch#795)

  Show description for 8d2314d
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  8d2314d

  Copy full SHA for 8d2314d

Commits on Nov 18, 2021

• Revert D31843250: open-source segment_sum_csr()

  Show description for 03d891d
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  03d891d

  Copy full SHA for 03d891d

Commits on Nov 17, 2021

• open-source segment_sum_csr() (pytorch#754)

  Show description for 858365f
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  858365f

  Copy full SHA for 858365f

Commits on Oct 19, 2021

• export quantize_ops_benchmark.py (pytorch#736)

  Show description for e165d5b
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  e165d5b

  Copy full SHA for e165d5b

Commits on Sep 13, 2021

• call proper sparse_adagrad_ref on non-AVX2/512 machines (pytorch#693)

  Show description for 742ec0e
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  742ec0e

  Copy full SHA for 742ec0e

Commits on Sep 8, 2021

• vectorize split_embedding_backward_exact_cpu_dense_kernel() (pytorch#692)

  Show description for fde782c
  

  shintaro-iwasaki

  authored and

  facebook-github-bot

  committed
  fde782c

  Copy full SHA for fde782c