Commits

Branch selector

 compiler_gym

User selector

jspark1105

Datepicker

All time

Commit History

Commits on Feb 19, 2022

• use CUDA_KERNEL_LOOP (pytorch#943)

  Show description for 15e1985
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  15e1985

• move kernels unnecessarily defined in headers (pytorch#942)

  Show description for a296d9c
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  a296d9c

Commits on Feb 15, 2022

• use int64_t in fbgemmPartition1D (pytorch#933)

  Show description for f7050c9
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  f7050c9

Commits on Feb 11, 2022

• clean up unused variables in int_nbit_split_embedding_codegen_forward_cpu (pytorch#927)

  Show description for 19801c1
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  19801c1

Commits on Feb 5, 2022

• use thread local code cache in cpu quantized emb fwd (pytorch#903)

  Show description for e779169
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  e779169

• reduce test time by sampling bool configs (pytorch#905)

  Show description for 5a5964e
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  5a5964e

Commits on Jan 24, 2022

• use self.Assert in split_table_batched_embeddings_test (pytorch#880)

  Show description for ba72edd
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  ba72edd

Commits on Jan 20, 2022

• reduce memory used for test_stacked_jagged_2d_to_dense take 2 (pytorch#876)

  Show description for 17d6a15
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  17d6a15

• enable -Wall -Wextra -Werror (pytorch#868)

  Show description for b047231
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  b047231

Commits on Jan 18, 2022

• remove extra ; after marco (pytorch#866)

  Show description for 80e3f16
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  80e3f16

• actually clamp denorm in RoundToFloat16 when clamp_denorms is set (pytorch#871)

  Show description for c3ad33c
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  c3ad33c

• reduce memory used for test_stacked_jagged_2d_to_dense (pytorch#867)

  Show description for 9843c86
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  9843c86

Commits on Jan 13, 2022

• conv code refactoring (pytorch#856)

  Show description for 3fac636
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  3fac636

Commits on Jan 8, 2022

• use fbgemm cpu kernel in forward quantized cpu emb (pytorch#849)

  Show description for 0e5737d
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  0e5737d

• scale_bias_last option to support TBE layout (pytorch#848)

  Show description for 747fc4a
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  747fc4a

Commits on Jan 3, 2022

• simplify jagged tensor kernels (pytorch#847)

  Show description for c016cb4
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  c016cb4

Commits on Jan 2, 2022

• use DISPATCH_TO_CUDA macro (pytorch#845)

  Show description for 21cfdca
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  21cfdca

• increase parallelism in batched unary embeddings backward (pytorch#842)

  Show description for b09f287
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  b09f287

• extract embedding input transpose out of embedding_backward_split_template.cu (pytorch#841)

  Show description for a5866f1
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  a5866f1

Commits on Jan 1, 2022

• fix typo in CMakeLists.txt (pytorch#840)

  Show description for ac78056
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  ac78056

Commits on Dec 31, 2021

• fix warning count check in test_bounds_check (pytorch#839)

  Show description for ec0043a
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  ec0043a

Commits on Dec 30, 2021

• add missing new files to CMakeLists.txt (pytorch#838)

  Show description for a508e46
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  a508e46

• add missing C10_CUDA_KERNEL_LAUNCH_CHECK (pytorch#837)

  Show description for cc7f834
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  cc7f834

• minimize functions defined in headers (pytorch#836)

  Show description for 38a6c35
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  38a6c35

Commits on Dec 24, 2021

• fix copyright header of tensor_assert_test.cpp (pytorch#831)

  Show description for a221ff8
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  a221ff8

Commits on Dec 18, 2021

• fix copy right header of batch_benchmark_run.py (pytorch#820)

  Show description for f0d3105
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  f0d3105

Commits on Dec 11, 2021

• fix bounds_check_indices aten op signature definition to reflect offset is now mutable (pytorch#808)

  Show description for 7730397
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  7730397

Commits on Dec 10, 2021

• add bound check of offset values (pytorch#806)

  Show description for 8824ecc
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  8824ecc

Commits on Nov 30, 2021

• remove assumption number of rows is in 32 bit (#69066)

  Show description for ae24ea0
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  ae24ea0

Commits on Nov 12, 2021

• optimize csr2csc for non shared table (pytorch#757)

  Show description for 4399bed
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  4399bed

• use SOA in radix sort (pytorch#756)

  Show description for 8db978a
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  8db978a

Commits on Nov 10, 2021

• fix bug in CPU exact rowwise adagrad for shared table (pytorch#753)

  Show description for 7eff3ae
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  7eff3ae

Commits on Sep 22, 2021

• shuffle indices generated by zipf to avoid unintended spatial locality (pytorch#708)

  Show description for b119fc6
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  b119fc6

Commits on Sep 16, 2021

• remove pendatic to avoid compile error (pytorch#701)

  Show description for 91fc122
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  91fc122

Commits on Sep 8, 2021

• add option to benchmark dense backward (pytorch#691)

  Show description for d1ecc7d
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  d1ecc7d

Pagination

PreviousNext