Commits

Breadcrumbs

History for

onmain

User selector

All users

Datepicker

All time

Commit History

Commits on Dec 9, 2021

• Fix download link for raft_small (pytorch#5078)

  NicolasHug

  authored
  22a23e8

  Copy full SHA for 22a23e8

• Add our weights for raft_small on C+T (pytorch#5076)

  Show description for 2508a02
  

  NicolasHug

  and

  datumbox

  authored
  2508a02

  Copy full SHA for 2508a02

Commits on Dec 8, 2021

• support random seed for RA sampler (pytorch#5053)

  Show description for 4cacf5a
  

  xiaohu2015

  and

  datumbox

  authored
  4cacf5a

  Copy full SHA for 4cacf5a

• Add pretrained weights on Chairs and Things for raft_large (pytorch#5060)

  NicolasHug

  authored
  849d02b

  Copy full SHA for 849d02b

• Adding Repeated Augment Sampler (pytorch#5051)

  Show description for e250db3
  yiwen-song

  authored
  e250db3

  Copy full SHA for e250db3

Commits on Dec 7, 2021

• Add transforms and presets for optical flow models (pytorch#5026)

  NicolasHug

  authored
  47ae092

  Copy full SHA for 47ae092

• Add training reference for optical flow models (pytorch#5027)

  NicolasHug

  authored
  4dd8b5c

  Copy full SHA for 4dd8b5c

• Replace --apex with --amp in docs (pytorch#5037)

  datumbox

  authored
  3ceaff1

  Copy full SHA for 3ceaff1

Commits on Dec 2, 2021

• support amp training for video classification models (pytorch#5023)

  Show description for ba299e8
  

  kbozas

  and

  Konstantinos Bozas

  authored
  ba299e8

  Copy full SHA for ba299e8

Commits on Nov 30, 2021

• Refactor the get_weights API (pytorch#5006)

  Show description for 3d8723d
  datumbox

  authored
  3d8723d

  Copy full SHA for 3d8723d

Commits on Nov 29, 2021

• Use new torch.ao.quantization instead of torch.quantization (pytorch#4554)

  Show description for b3cdec1
  

  NicolasHug

  and

  datumbox

  authored
  b3cdec1

  Copy full SHA for b3cdec1

Commits on Nov 26, 2021

• support amp training for segmention models (pytorch#4994)

  Show description for cca1699
  

  xiaohu2015

  and

  datumbox

  authored
  cca1699

  Copy full SHA for cca1699

Commits on Nov 22, 2021

• Update README.md (pytorch#4972)

  Separius

  authored
  a96de03

  Copy full SHA for a96de03

• Add multi-type support on get_weight() (pytorch#4967)

  Show description for 220e0ff
  datumbox

  authored
  220e0ff

  Copy full SHA for 220e0ff

Commits on Nov 15, 2021

• support amp training for detection models (pytorch#4933)

  Show description for 59ec1df
  

  xiaohu2015

  and

  datumbox

  authored
  59ec1df

  Copy full SHA for 59ec1df

Commits on Nov 12, 2021

• Simplify the gradient clipping code. (pytorch#4896)

  datumbox

  authored
  1de53be

  Copy full SHA for 1de53be

• save grad_scaler if use amp for better resume (pytorch#4923)

  Show description for 22ff44f
  

  xiaohu2015

  and

  datumbox

  authored
  22ff44f

  Copy full SHA for 22ff44f

Commits on Nov 9, 2021

• fix bug in training model by amp (pytorch#4874)

  Show description for 031e129
  

  3 peopleauthored
  031e129

  Copy full SHA for 031e129

Commits on Nov 4, 2021

• Cleanup Quantized ShuffleNet (pytorch#4854)

  Show description for 9109d8d
  datumbox

  authored
  9109d8d

  Copy full SHA for 9109d8d

Commits on Nov 3, 2021

• Adding multiweight support to Quantized GoogLeNet (pytorch#4848)

  Show description for b43353e
  datumbox

  authored
  b43353e

  Copy full SHA for b43353e

• Moving the check for prototype support in all references. (pytorch#4849)

  datumbox

  authored
  3300692

  Copy full SHA for 3300692

• Adding multiweight support to FasterRCNN (pytorch#4847)

  Show description for dd1adb0
  datumbox

  authored
  dd1adb0

  Copy full SHA for dd1adb0

• Update Reference scripts to support the prototype models (pytorch#4837)

  Show description for b1fc290
  datumbox

  authored
  b1fc290

  Copy full SHA for b1fc290

Commits on Nov 2, 2021

• [references/classification] Adding gradient clipping (pytorch#4824)

  Show description for 0817f7f
  yiwen-song

  authored
  0817f7f

  Copy full SHA for 0817f7f

• Adding multiweight support to Quantized ResNet (pytorch#4827)

  Show description for 9a34c0c
  datumbox

  authored
  9a34c0c

  Copy full SHA for 9a34c0c

• Update training references from legacy models (pytorch#4830)

  Show description for 6a60b9b
  datumbox

  authored
  6a60b9b

  Copy full SHA for 6a60b9b

Commits on Nov 1, 2021

• Adding multiweight support for inception prototype model (pytorch#4821)

  Show description for 732fc0b
  datumbox

  authored
  732fc0b

  Copy full SHA for 732fc0b

Commits on Oct 28, 2021

• Use f-strings almost everywhere, and other cleanups by applying pyupgrade (pytorch#4585)

  Show description for d367a01
  

  Borda

  and

  NicolasHug

  authored
  d367a01

  Copy full SHA for d367a01

Commits on Oct 25, 2021

• Adding new ResNet50 weights (pytorch#4734)

  Show description for dc11399
  datumbox

  authored
  dc11399

  Copy full SHA for dc11399

• Fix reduce_across_processes inconsistent return type (pytorch#4733)

  NicolasHug

  authored
  3783f3f

  Copy full SHA for 3783f3f

Commits on Oct 24, 2021

• Add types and improve descriptions to ArgumentParser parameters (pytorch#4724)

  Show description for 9ae833a
  

  puhuk

  and

  datumbox

  authored
  9ae833a

  Copy full SHA for 9ae833a

Commits on Oct 22, 2021

• Additional SOTA ingredients on Classification Recipe (pytorch#4493)

  Show description for b280c31
  datumbox

  authored
  b280c31

  Copy full SHA for b280c31

Commits on Oct 17, 2021

• Replaced all 'no_grad()' instances with 'inference_mode()' (pytorch#4629)

  XoMute

  authored
  e08c9e3

  Copy full SHA for e08c9e3

Commits on Oct 13, 2021

• Reduce variance of classification references evaluation (pytorch#4609)

  NicolasHug

  authored
  5b81c05

  Copy full SHA for 5b81c05

Commits on Oct 8, 2021

• Updated classification README to refer to torch.cuda.amp (pytorch#4570)

  prabhat00155

  authored
  566bab5

  Copy full SHA for 566bab5

Pagination

Previous

Next