Commits

Breadcrumbs

History for

on20e7027

User selector

All users

Datepicker

All time

Commit History

Commits on Nov 7, 2019

• Sink Transpose below Clip node. (pytorch#3748)

  Show description for 20e7027
  

  mciprian13

  authored and

  facebook-github-bot

  committed
  20e7027

  Copy full SHA for 20e7027

• Relax single-use constant restrictions (pytorch#3739)

  Show description for 181fd26
  

  bertmaher

  authored and

  facebook-github-bot

  committed
  181fd26

  Copy full SHA for 181fd26

• re-enable the Sigmoid NNPI test (pytorch#3735)

  Show description for ebf03b1
  

  Hector Yuen

  authored and

  facebook-github-bot

  committed
  ebf03b1

  Copy full SHA for ebf03b1

• Add check for partial tensor support to RepeatedSLSWithPartialTensors (pytorch#3724)

  Show description for 1d44837
  

  jfix71

  authored and

  facebook-github-bot

  committed
  1d44837

  Copy full SHA for 1d44837

Commits on Nov 6, 2019

• Test configuration is not supported by Habana backend.

  Show description for 65f8825
  

  Roman Dzhabarov

  authored and

  facebook-github-bot

  committed
  65f8825

  Copy full SHA for 65f8825

• Remove SinkCode from NNPI opt pipeline (pytorch#3743)

  Show description for 684be45
  

  jfix71

  authored and

  facebook-github-bot

  committed
  684be45

  Copy full SHA for 684be45

• Change pipeline so we only run the backend's preferred optimizations at the end of the pipeline (pytorch#3740)

  Show description for f742646
  

  jfix71

  authored and

  facebook-github-bot

  committed
  f742646

  Copy full SHA for f742646

• Handroll make_unique impl for future-compat with LLVM (pytorch#3566)

  Show description for f299e9c
  

  Bram Wasti

  authored and

  facebook-github-bot

  committed
  f299e9c

  Copy full SHA for f299e9c

Commits on Nov 5, 2019

• Fix parallel clone count for quantized tests (pytorch#3732)

  Show description for 5b89f3d
  

  jfix71

  authored and

  facebook-github-bot

  committed
  5b89f3d

  Copy full SHA for 5b89f3d

• Add AdaptiveAvgPool on CPU backend (pytorch#3734)

  Show description for fc07c38
  

  jackm321

  authored and

  facebook-github-bot

  committed
  fc07c38

  Copy full SHA for fc07c38

• Fix CI for habana. Those tests should not run on the backend.

  Show description for 3598ca9
  

  Roman Dzhabarov

  authored and

  facebook-github-bot

  committed
  3598ca9

  Copy full SHA for 3598ca9

Commits on Nov 4, 2019

• (trivial) Change bundle weights file names to ".weights.bin" / ".weights.txt" (pytorch#3721)

  Show description for 1c6782f
  

  mciprian13

  authored and

  facebook-github-bot

  committed
  1c6782f

  Copy full SHA for 1c6782f

Commits on Nov 3, 2019

• Refactored Provisioner to reduce complexity and memory consumption (pytorch#3615)

  Show description for 57a4427
  

  gcatron

  authored and

  facebook-github-bot

  committed
  57a4427

  Copy full SHA for 57a4427

• Add support for 4-bit fused-rowwise-quantized SLWS to Interpreter (pytorch#3719)

  Show description for cd38342
  

  jfix71

  authored and

  facebook-github-bot

  committed
  cd38342

  Copy full SHA for cd38342

Commits on Nov 1, 2019

• glow: add another unittest for back to back inference with slws ops (pytorch#3720)

  Show description for 2115d5a
  

  tracelogfb

  authored and

  facebook-github-bot

  committed
  2115d5a

  Copy full SHA for 2115d5a

• Add optimization for folding Clip(Splat(args)) -> Splat(args') (pytorch#3718)

  Show description for befb3ac
  

  jfix71

  authored and

  facebook-github-bot

  committed
  befb3ac

  Copy full SHA for befb3ac

• v0.3.2H0 Update for the Intel NNPI Backend (pytorch#3700)

  Show description for db91bb9
  

  jsubag

  authored and

  facebook-github-bot

  committed
  db91bb9

  Copy full SHA for db91bb9

• Add isDataParallel to node generation (pytorch#3715)

  Show description for 2b960af
  

  shajrawi

  authored and

  facebook-github-bot

  committed
  2b960af

  Copy full SHA for 2b960af

Commits on Oct 31, 2019

• Add command-line options to provide a code model to be used for JIT and bundles and a relocation model (pytorch#3706)

  Show description for 6078fab
  

  opti-mix

  authored and

  facebook-github-bot

  committed
  6078fab

  Copy full SHA for 6078fab

• evalBatch helper function in EE (pytorch#3607)

  Show description for 5488918
  

  mortzur

  authored and

  facebook-github-bot

  committed
  5488918

  Copy full SHA for 5488918

• Adding a new SLWS node using offsets (pytorch#3697)

  Show description for 935f664
  

  Qi Zhou

  authored and

  facebook-github-bot

  committed
  935f664

  Copy full SHA for 935f664

• Support Offline Tensors through ONNXIFI layer

  Show description for e09da24
  

  Yinghai Lu

  authored and

  facebook-github-bot

  committed
  e09da24

  Copy full SHA for e09da24

• Fix crash with --instrument-debug (pytorch#3704)

  Show description for c8de4b0
  

  jackm321

  authored and

  facebook-github-bot

  committed
  c8de4b0

  Copy full SHA for c8de4b0

Commits on Oct 29, 2019

• NNPI InferencePool tracing (pytorch#3696)

  Show description for 4a09831
  

  virtan

  authored and

  facebook-github-bot

  committed
  4a09831

  Copy full SHA for 4a09831

• Add network_execution_e2e event trace output and report execution time and (pytorch#3686)

  Show description for a75c883
  

  virtan

  authored and

  facebook-github-bot

  committed
  a75c883

  Copy full SHA for a75c883

• Add flag for disabling NNPI transforms (pytorch#3693)

  Show description for 66b8172
  

  jfix71

  authored and

  facebook-github-bot

  committed
  66b8172

  Copy full SHA for 66b8172

• Add in NNPI::transformPostLowering() stub and NNPI::transformPrivate() (pytorch#3692)

  Show description for 2d68a41
  

  jfix71

  authored and

  facebook-github-bot

  committed
  2d68a41

  Copy full SHA for 2d68a41

Commits on Oct 28, 2019

• Minor improvements (pytorch#3689)

  Show description for 63a1792
  

  opti-mix

  authored and

  facebook-github-bot

  committed
  63a1792

  Copy full SHA for 63a1792

• Add opt for eliminating concat->slices (pytorch#3688)

  Show description for a13d88d
  

  jfix71

  authored and

  facebook-github-bot

  committed
  a13d88d

  Copy full SHA for a13d88d

Commits on Oct 26, 2019

• Fix various bugs in ONNX Exporter (pytorch#3683)

  Show description for e00a8b4
  

  Yinghai Lu

  authored and

  facebook-github-bot

  committed
  e00a8b4

  Copy full SHA for e00a8b4

• Tell the different flavors of int8 types during ONNX export (pytorch#3684)

  Show description for 2b6f245
  

  Yinghai Lu

  authored and

  facebook-github-bot

  committed
  2b6f245

  Copy full SHA for 2b6f245

• Add option to dump input/output across ONNXIFI layer (pytorch#3682)

  Show description for 39ea3cc
  

  Yinghai Lu

  authored and

  facebook-github-bot

  committed
  39ea3cc

  Copy full SHA for 39ea3cc

Commits on Oct 25, 2019

• repro binary (pytorch#3681)

  Show description for 0f59089
  

  Yinghai Lu

  authored and

  facebook-github-bot

  committed
  0f59089

  Copy full SHA for 0f59089

• Dump glow function right after ONNXIFI (pytorch#3665)

  Show description for 6afa42b
  

  Yinghai Lu

  authored and

  facebook-github-bot

  committed
  6afa42b

  Copy full SHA for 6afa42b

• Add Zip support to ONNX Exporter (pytorch#3634)

  Show description for a8631f2
  

  Yinghai Lu

  authored and

  facebook-github-bot

  committed
  a8631f2

  Copy full SHA for a8631f2

Pagination

Previous

Next