Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on May 24, 2022

• Add nvtx markers which can be useful in perf profiling (#64)

  tanmayv25

  authored
  df900ce

  Copy full SHA for df900ce

Commits on May 20, 2022

• Enforce ordering of I/O if naming convention is not followed (#63)

  Show description for ff103c4
  

  Hemant Jain

  authored
  ff103c4

  Copy full SHA for ff103c4

Commits on May 9, 2022

• Add support for String input/output (#60)

  Show description for ab59a37
  

  Hemant Jain

  authored
  ab59a37

  Copy full SHA for ab59a37

Commits on May 4, 2022

• Fix typos + cleanup ReadMe (#62)

  

  Hemant Jain

  authored
  510cc49

  Copy full SHA for 510cc49

Commits on Apr 26, 2022

• Fix libc10 missing issue on jetson (#56) (#58)

  

  Hemant Jain

  authored
  a4f3138

  Copy full SHA for a4f3138

Commits on Apr 4, 2022

• Disable NvFuser till fixed (Caused by move to 22.03 PyTorch) (#52)

  Show description for 9f1b35a
  

  Hemant Jain

  authored and

  dzier

  committed
  9f1b35a

  Copy full SHA for 9f1b35a

Commits on Apr 1, 2022

• Share Model Weights for Instances on the Same Device (#54)

  Show description for 73e8fc7
  

  the-david-oy

  and

  hemantj

  authored
  73e8fc7

  Copy full SHA for 73e8fc7

Commits on Mar 23, 2022

• Enable building of CPU only PyTorch (#53)

  Show description for 1c4db70
  

  Hemant Jain

  authored
  1c4db70

  Copy full SHA for 1c4db70

Commits on Mar 16, 2022

• Allow input names from model metadata (#51)

  Show description for b3cfc31
  

  Hemant Jain

  authored
  b3cfc31

  Copy full SHA for b3cfc31

Commits on Feb 4, 2022

• Add support for passing inputs as a dictionary to model (#50)

  Show description for e391cb4
  

  Hemant Jain

  authored
  e391cb4

  Copy full SHA for e391cb4

Commits on Feb 3, 2022

• Use tensor device instead of model device (#49)

  Show description for ee05ed1
  CoderHam

  authored
  ee05ed1

  Copy full SHA for ee05ed1

Commits on Jan 6, 2022

• Fix error handling to include failed counts in infer stat (#48)

  Show description for 1ec5155
  tanmayv25

  authored
  1ec5155

  Copy full SHA for 1ec5155

Commits on Dec 16, 2021

• Responder should explicitly use 'first dimension batching' (#47)

  

  deadeyegoodwin

  authored
  b2f5459

  Copy full SHA for b2f5459

Commits on Dec 9, 2021

• Adding a flag to control Torchvision support in non-docker builds (#46)

  Show description for d7a818f
  CoderHam

  authored
  d7a818f

  Copy full SHA for d7a818f

Commits on Nov 30, 2021

• TorchTRT support (#44)

  Show description for 6a2b070
  borisfom

  authored
  6a2b070

  Copy full SHA for 6a2b070

Commits on Nov 23, 2021

• Return error if no inputs / outputs are provided in model config (#45)

  Show description for 6de7d90
  CoderHam

  authored
  6de7d90

  Copy full SHA for 6de7d90

Commits on Nov 3, 2021

• Additional parameters to control PyTorch JIT and fuser options (#43)

  Show description for 052731c
  

  ryanleary

  and

  David Goodwin

  authored
  052731c

  Copy full SHA for 052731c

Commits on Oct 11, 2021

• Add option to enable Cuda fuser (NvFuser) (#40)

  Show description for b26c0d3
  CoderHam

  authored
  b26c0d3

  Copy full SHA for b26c0d3

Commits on Sep 28, 2021

• Add missing pytorch library for ARM64 (#39)

  

  deadeyegoodwin

  authored
  272f5c1

  Copy full SHA for 272f5c1

Commits on Sep 24, 2021

• Fix shared libraries to work with ARM64

  

  David Goodwin

  committed
  eea1841

  Copy full SHA for eea1841

• MKL libraries are not available on ARM64

  

  David Goodwin

  committed
  9eff559

  Copy full SHA for 9eff559

Commits on Sep 16, 2021

• Add verbose error for when output is a list / dict of tensors or scalar (#36)

  CoderHam

  authored
  6863749

  Copy full SHA for 6863749

Commits on Aug 12, 2021

• Add bool type (#35)

  Show description for d8afa67
  

  krishung5

  and

  Kris Hung

  authored
  d8afa67

  Copy full SHA for d8afa67

Commits on Aug 2, 2021

• Use InferenceMode guard instead of NoGradGuard (#33)

  Show description for a324393
  CoderHam

  authored
  a324393

  Copy full SHA for a324393

• Document issues in libtorch multi instance perf on backend (#34)

  Show description for b5cf9a0
  CoderHam

  authored
  b5cf9a0

  Copy full SHA for b5cf9a0

Commits on Jul 19, 2021

• Document effects of libtorch asynchronous execution on backend (#32)

  Show description for 78c96c6
  tanmayv25

  authored
  78c96c6

  Copy full SHA for 78c96c6

Commits on Jul 7, 2021

• Fix typo in NGC url (#30)

  CoderHam

  authored
  ac86471

  Copy full SHA for ac86471

Commits on Jul 2, 2021

• Modify backend to be host policy aware (#29)

  GuanLuo

  authored
  53a9228

  Copy full SHA for 53a9228

Commits on Jun 25, 2021

• Change omp library name, moving to PyTorch 21.06 (#27)

  Show description for a93bedb
  

  CoderHam

  authored and

  dzier

  committed
  a93bedb

  Copy full SHA for a93bedb

Commits on Jun 12, 2021

• Return appropriate error if the output is a scalar (#28)

  CoderHam

  authored
  e1699e0

  Copy full SHA for e1699e0

Commits on Jun 7, 2021

• Do not throw error when DISABLE_OPTIMIZED_EXECUTION is missing. (#26)

  Show description for 968e80f
  CoderHam

  authored
  968e80f

  Copy full SHA for 968e80f

Commits on Jun 4, 2021

• Fix typo that causes build failure (#25)

  CoderHam

  authored
  c29963c

  Copy full SHA for c29963c

• Add support for disabling optimized execution (#24)

  Show description for 74ccccb
  CoderHam

  authored
  74ccccb

  Copy full SHA for 74ccccb

Commits on May 28, 2021

• Fix timestamp collection to not impact perf (#23)

  Show description for a138552
  CoderHam

  authored
  a138552

  Copy full SHA for a138552

Commits on May 12, 2021

• Fix timing for GPU execution (#22)

  Show description for b96e04d
  tanmayv25

  authored
  b96e04d

  Copy full SHA for b96e04d

Pagination

Previous

Next