Commits

Branch selector

 main

User selector

jianyuh

Datepicker

All time

Commit History

Commits on Sep 22, 2021

• Add FP32 support in infer version of table batched embedding op (pytorch#707)

  Show description for 87f4f4f
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  87f4f4f

Commits on Sep 20, 2021

• Fix OSS build errors with Tesla T4 using sm75 compute (pytorch#703)

  Show description for 0108d4f
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  0108d4f

Commits on Sep 14, 2021

• update README

  Show description for 4d9c006
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  4d9c006

Commits on Sep 13, 2021

• Support weights_uvm on CPU ops

  Show description for 7ceee02
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  7ceee02

• make the CPU and CUDA op interface of infer version of table batched embedding bag the same (pytorch#696)

  Show description for 5d100bd
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  5d100bd

• Rename ad_pooled_embeddings to pooled_embeddings (pytorch#695)

  Show description for 9c12ad2
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  9c12ad2

• Fix merge embeddings benchmark (pytorch#699)

  Show description for eb5ba80
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  eb5ba80

• Remove unused numpy deps

  Show description for c1d0db2
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  c1d0db2

Commits on Sep 12, 2021

• Add UVM support for infer version of table batched embedding bag (pytorch#694)

  Show description for 62d3f4e
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  62d3f4e

• Re-apply "Move merge_pooled_embeddings op into FBGEMM_GPU." (pytorch#689)

  Show description for 98a7c14
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  98a7c14

Commits on Sep 7, 2021

• Apply pyfmt formatting (pytorch#690)

  Show description for 3ce04fc
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  3ce04fc

Commits on Sep 3, 2021

• Back out "Move merge_pooled_embeddings op into FBGEMM_GPU."

  Show description for 7b49986
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  7b49986

Commits on Sep 2, 2021

• Quick fix OSS workflow (pytorch#687)

  Show description for f744bd6
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  f744bd6

Commits on Jul 17, 2021

• Minor changes for benchmarking (pytorch#650)

  Show description for 133c5c2
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  133c5c2

Commits on Jul 3, 2021

• Fix README instruction on OSS install (pytorch#644)

  Show description for 7495417
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  7495417

Commits on Jul 1, 2021

• Add quantize workflow for inference table batched embeddings (pytorch#641)

  Show description for 5dc753c
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  5dc753c

Commits on Jun 29, 2021

• Prepare for Using of FP16/INT8/IN4 inference version of table batched embeddings (pytorch#640)

  Show description for da7fbbc
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  da7fbbc

Commits on Jun 17, 2021

• Fix OSS build (pytorch#635)

  Show description for 201df1d
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  201df1d

Commits on Jun 11, 2021

• Fix the TorchScript issue for IntNBitTableBatchedEmbeddingBagsCodegen (pytorch#627)

  Show description for ae8ad8f
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  ae8ad8f

Commits on May 17, 2021

• Fix the int8 embedding bench/test build (pytorch#610)

  Show description for d4a6222
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  d4a6222

Commits on May 16, 2021

• Fix the int4 table batched embedding benchmark with mixed dim (pytorch#609)

  Show description for c48ba9b
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  c48ba9b

• Andrew's int4 table batched embedding implementation (pytorch#608)

  Show description for 0fe80ee
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  0fe80ee

Commits on Apr 26, 2021

• Support for sequence embeddings (Embedding list no bag) (pytorch#582)

  Show description for 95805fb
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  95805fb

Commits on Mar 31, 2021

• Re-apply "[fbgemm_gpu] have bf16 quantize kernels use 2d blocks with loop to cover large tensors" (pytorch#583)

  Show description for 5a8d11f
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  5a8d11f

Commits on Mar 30, 2021

• remove fp16 flag (pytorch#581)

  Show description for 2c2ae1f
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  2c2ae1f

Commits on Mar 24, 2021

• Back out "have bf16 quantize kernels use 2d blocks with loop to cover large tensors" (pytorch#572)

  Show description for fc2c734
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  fc2c734

Commits on Mar 23, 2021

• Fix OSS build failure due to approx sgd (pytorch#570)

  Show description for 28b130f
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  28b130f

• Move permute_pooled_embs_kernel into fbgemm_gpu (pytorch#569)

  Show description for 3657619
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  3657619

Commits on Mar 19, 2021

• Also include the dependency to FBGEMM CPU SPMDM routine (pytorch#559)

  Show description for e297346
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  e297346

Commits on Mar 18, 2021

• Optimize the performance of recat_embedding_grad_output (pytorch#488)

  Show description for d4f2dcc
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  d4f2dcc

• Add FBGEMM_CPU dependency in setup.py for table batched embedding cpu op (pytorch#556)

  Show description for e4ca7c4
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  e4ca7c4

• Fix the OSS CMakefile build (pytorch#555)

  Show description for 7716e15
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  7716e15

• Fix OSS build and run issues (pytorch#554)

  Show description for 4ef1ec0
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  4ef1ec0

• Add per-feature precision support in the model level (pytorch#552)

  Show description for 334134c
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  334134c

Commits on Mar 9, 2021

• Add the cmake file for searching sphinx (pytorch#536)

  Show description for 4b88f40
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  4b88f40

Pagination

PreviousNext