Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Dec 15, 2022

• Use larger zipf_oversample_ratio for small L in TBE device_with_spec (pytorch#1507)

  Show description for b0bb1cb
  

  sryap

  authored and

  facebook-github-bot

  committed
  b0bb1cb

  Copy full SHA for b0bb1cb

Commits on Dec 14, 2022

• Adding back docs generation (pytorch#1508)

  Show description for 521faf8
  

  mjanderson09

  authored and

  facebook-github-bot

  committed
  521faf8

  Copy full SHA for 521faf8

• Support k=0 in bottom_k_per_row (pytorch#1506)

  Show description for 8cae821
  

  sryap

  authored and

  facebook-github-bot

  committed
  8cae821

  Copy full SHA for 8cae821

• Fix the leftover test case for output_dtype (pytorch#1504)

  Show description for 12e85e2
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  12e85e2

  Copy full SHA for 12e85e2

Commits on Dec 13, 2022

• Add BF16 output support for inference TBE (pytorch#1498)

  Show description for 39a423e
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  39a423e

  Copy full SHA for 39a423e

• Add bag size generation from mean and std in TBE benchmark (pytorch#1450)

  Show description for fe0f7ab
  

  sryap

  authored and

  facebook-github-bot

  committed
  fe0f7ab

  Copy full SHA for fe0f7ab

Commits on Dec 12, 2022

• small fix to test CI failure on the trunk (pytorch#1499)

  Show description for 6ed7a68
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  6ed7a68

  Copy full SHA for 6ed7a68

Commits on Dec 8, 2022

• Low-overhead hit/miss counter for UVM_CACHING (pytorch#1420)

  Show description for 81ba6c5
  

  doehyun

  authored and

  facebook-github-bot

  committed
  81ba6c5

  Copy full SHA for 81ba6c5

Commits on Dec 7, 2022

• Minor fix to fbgemm all-to-one benchmark and related script. (pytorch#1496)

  Show description for 908a8f3
  

  doehyun

  authored and

  facebook-github-bot

  committed
  908a8f3

  Copy full SHA for 908a8f3

Commits on Dec 6, 2022

• Minor fix to benchmark_torch_function, when num_threads > 2 (pytorch#1493)

  Show description for bdccbcf
  

  doehyun

  authored and

  facebook-github-bot

  committed
  bdccbcf

  Copy full SHA for bdccbcf

Commits on Dec 5, 2022

• open source SSD implementation of TBE (pytorch#1485)

  Show description for 5e40a70
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  5e40a70

  Copy full SHA for 5e40a70

• Change from ubuntu latest (22.04) to ubuntu 20.04 (pytorch#1491)

  Show description for 8c6e82e
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  8c6e82e

  Copy full SHA for 8c6e82e

• Follow up on BC issue for open sourcing TBE inplace update op (pytorch#1492)

  Show description for 30bfa11
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  30bfa11

  Copy full SHA for 30bfa11

Commits on Dec 4, 2022

• Open source inplace update operators (pytorch#1482)

  Show description for 4cd267c
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  4cd267c

  Copy full SHA for 4cd267c

Commits on Dec 2, 2022

• Open source TBE inplace update operators (pytorch#1480)

  Show description for 6120f21
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  6120f21

  Copy full SHA for 6120f21

Commits on Dec 1, 2022

• Add FP16, INT8, INT4, INT2 support for SSD inference TBE (pytorch#1479)

  Show description for b7ee3d6
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  b7ee3d6

  Copy full SHA for b7ee3d6

• add back cudaHostAlloc for debugging the UVM OOM issue (pytorch#1486)

  Show description for 624381e
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  624381e

  Copy full SHA for 624381e

Commits on Nov 30, 2022

• Add FP32 + BF16 Hybrid Dtype Support for jagged_dense_dense_elementwise_add_jagged (pytorch#1487)

  Show description for 5d36625
  

  jiawenliu64

  authored and

  facebook-github-bot

  committed
  5d36625

  Copy full SHA for 5d36625

Commits on Nov 29, 2022

• Add ssd implementation for inference TBE (pytorch#1473)

  Show description for b6369c2
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  b6369c2

  Copy full SHA for b6369c2

• Add copyright header (pytorch#1483)

  Show description for f8a23de
  

  brad-mengchi

  authored and

  facebook-github-bot

  committed
  f8a23de

  Copy full SHA for f8a23de

• Add support for multi-thread/stream execution in feed_lower_benchmark (pytorch#1474)

  Show description for 13ea863
  

  Jiecao Yu

  authored and

  facebook-github-bot

  committed
  13ea863

  Copy full SHA for 13ea863

• batched_dense_vec_jagged_2d_mul add Meta and Autograd backend (pytorch#1468)

  Show description for e749c95
  

  brad-mengchi

  authored and

  facebook-github-bot

  committed
  e749c95

  Copy full SHA for e749c95

• jagged_dense_elementwise_mul add Meta and Autograd backend (pytorch#1469)

  Show description for e4f6ed8
  

  brad-mengchi

  authored and

  facebook-github-bot

  committed
  e4f6ed8

  Copy full SHA for e4f6ed8

• Jagged_dense_dense_elementwise_add_jagged add Meta and Autograd backend (pytorch#1467)

  Show description for 848c3c3
  

  brad-mengchi

  authored and

  facebook-github-bot

  committed
  848c3c3

  Copy full SHA for 848c3c3

• Vectorize to improve perf on {lru/lfu}_cache_insert_byte_kernel (pytorch#1475)

  Show description for 003d188
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  003d188

  Copy full SHA for 003d188

Commits on Nov 28, 2022

• Jagged_to_padded_dense add Meta and Autograd backend (pytorch#1466)

  Show description for b0375f5
  

  brad-mengchi

  authored and

  facebook-github-bot

  committed
  b0375f5

  Copy full SHA for b0375f5

• add bf16 support for jagged tensor op (pytorch#1472)

  Show description for 1debba9
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  1debba9

  Copy full SHA for 1debba9

• Remove inplace update in Python level to avoid BC issue (pytorch#1481)

  Show description for c281c62
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  c281c62

  Copy full SHA for c281c62

• add back torch.ops.fb.split_embedding_codegen_lookup_rowwise_adagrad_function API (pytorch#1476)

  Show description for 14ee909
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  14ee909

  Copy full SHA for 14ee909

• refactor test_utils and bench_utils (pytorch#68)

  Show description for 4c58137
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  4c58137

  Copy full SHA for 4c58137

Commits on Nov 17, 2022

• Minor fix to nbit_device_with_spec (pytorch#1465)

  Show description for afec5b4
  

  doehyun

  authored and

  facebook-github-bot

  committed
  afec5b4

  Copy full SHA for afec5b4

Commits on Nov 16, 2022

• Back out "add Meta and Autograd backends for jagged ops" (pytorch#1463)

  Show description for d0af623
  

  Xiao Sun

  authored and

  facebook-github-bot

  committed
  d0af623

  Copy full SHA for d0af623

• Fix UVMCacheStats error (pytorch#1462)

  Show description for 65d4508
  

  doehyun

  authored and

  facebook-github-bot

  committed
  65d4508

  Copy full SHA for 65d4508

Commits on Nov 15, 2022

• add Meta and Autograd backends for jagged ops (pytorch#1461)

  Show description for d47449d
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  d47449d

  Copy full SHA for d47449d

• Add nbit_device_with_spec to split_table_batche_embeddings_benchmark (pytorch#1455)

  Show description for a327a86
  

  doehyun

  authored and

  facebook-github-bot

  committed
  a327a86

  Copy full SHA for a327a86

Pagination

Previous

Next