Commits

Branch selector

 nfnet

User selector

All users

Datepicker

All time

Commit History

Commits on Feb 13, 2021

• Update README.md and few more comments

  rwightman

  committed
  d86dbe4

  Copy full SHA for d86dbe4

• Fix issue with nfnet tests, bit more cleanup.

  rwightman

  committed
  0d253e2

  Copy full SHA for 0d253e2

• Add NFNet-F models and tweak existing NF models.

  rwightman

  committed
  cb06c7a

  Copy full SHA for cb06c7a

Commits on Feb 11, 2021

• Add first 'Normalizer Free' models. nf_regnet_b1 79.3 @ 288x288 test, and nf_resnet50 80.3 @ 256x256 test (80.68 @ 288x288).

  rwightman

  committed
  e4de077

  Copy full SHA for e4de077

Commits on Feb 10, 2021

• Merge pull request #419 from rwightman/byob_vgg_models

  Show description for d8e6920
  rwightman

  authored
  d8e6920

  Copy full SHA for d8e6920

• Update README.md and docs. Version bumped to 0.4.3

  rwightman

  committed
  ca9b078

  Copy full SHA for ca9b078

• Improve RegVGG block identity/vs non for clariy and fix attn usage. Add comments.

  rwightman

  committed
  6853b07

  Copy full SHA for 6853b07

• Default to native PyTorch AMP instead of APEX amp. Too many APEX issues cropping up lately.

  rwightman

  committed
  0356e77

  Copy full SHA for 0356e77

• Merge pull request #421 from aboutmedicine/master

  Show description for b9bd960
  rwightman

  authored
  b9bd960

  Copy full SHA for b9bd960

• update collections.abc import

  

  Reuben

  authored and

  Reuben

  committed
  94ca140

  Copy full SHA for 94ca140

• Merge pull request #418 from kecsap/master

  Show description for db6128d
  rwightman

  authored
  db6128d

  Copy full SHA for db6128d

• Fix a few small things.

  rwightman

  committed
  b4e216e

  Copy full SHA for b4e216e

• Add ByobNet w/ GPU-EfficientNets and RepVGG. Also add classic vgg models.

  rwightman

  committed
  dc85e5a

  Copy full SHA for dc85e5a

• Use in_channels for depthwise groups, allows using `out_channels=N * in_channels` (does not impact existing models). Fix #354.

  rwightman

  committed
  1bcc69e

  Copy full SHA for 1bcc69e

• Fix regression in models with 1001 class pretrained weights. Improve batchnorm arg and BatchNormAct layer handling in several models.

  rwightman

  committed
  9811e22

  Copy full SHA for 9811e22

Commits on Feb 9, 2021

• Change the Python interpreter to Python 3.x in the scripts

  kecsap

  committed
  5114c21

  Copy full SHA for 5114c21

Commits on Feb 8, 2021

• Update README.md

  rwightman

  authored
  aaa715b

  Copy full SHA for aaa715b

• Merge pull request #413 from rwightman/eca-weights

  Show description for a7f9581
  rwightman

  authored
  a7f9581

  Copy full SHA for a7f9581

• Merge branch 'master' into eca-weights

  rwightman

  authored
  a39c3ee

  Copy full SHA for a39c3ee

• Update README for new weights. Version 0.4.2

  rwightman

  committed
  e9d6fe2

  Copy full SHA for e9d6fe2

• Move stride in EdgeResidual block to 3x3 expansion conv. Fix #414

  rwightman

  committed
  666de85

  Copy full SHA for 666de85

• Fix some half removed resnet model defs, pooling for ecaresnet269d

  rwightman

  committed
  3b57490

  Copy full SHA for 3b57490

• Add validation script update for using test_input_size in model default_cfgs

  rwightman

  committed
  2a8c4dc

  Copy full SHA for 2a8c4dc

Commits on Feb 7, 2021

• Add new weights for ecaresnet26t/50t/269d models. Remove distinction between 't' and 'tn' (tiered models), tn is now t. Add test time img size spec to default cfg.

  rwightman

  committed
  68a4144

  Copy full SHA for 68a4144

Commits on Feb 5, 2021

• Update README.md

  rwightman

  authored
  3a7aa95

  Copy full SHA for 3a7aa95

Commits on Feb 4, 2021

• Merge pull request #282 from tigert1998/patch-1

  Show description for b9843f9
  rwightman

  authored
  b9843f9

  Copy full SHA for b9843f9

• Merge pull request #401 from hwangdeyu/deyu/add_HardSwishJitAutoFn_operator

  Show description for ea36a78
  rwightman

  authored
  ea36a78

  Copy full SHA for ea36a78

Commits on Feb 3, 2021

• add operator HardSwishJitAutoFn export to onnx

  

  hwangdeyu

  committed
  7a4be5c

  Copy full SHA for 7a4be5c

Commits on Feb 1, 2021

• Fix #387 so that checkpoint saver works with max history of 1. Add checkpoint-hist arg to train.py.

  rwightman

  committed
  4203efa

  Copy full SHA for 4203efa

Commits on Jan 31, 2021

• Merge pull request #389 from rwightman/norm_free_models

  Show description for 99b82ae
  rwightman

  authored
  99b82ae

  Copy full SHA for 99b82ae

• Fix NF-ResNet101 model defs

  rwightman

  committed
  f0e65e3

  Copy full SHA for f0e65e3

• Update README.md for NF-nets, bump version to 0.4.1 for merge

  rwightman

  committed
  2c988c3

  Copy full SHA for 2c988c3

• Fix pool size defs for NFNet models, add a comment.

  rwightman

  committed
  2de54d1

  Copy full SHA for 2de54d1

• Fix up a few details in NFResNet models, managed stable training. Add support for gamma gain to be applied in activation or ScaleStdConv. Some tweaks to ScaledStdConv.

  rwightman

  committed
  90980de

  Copy full SHA for 90980de

Commits on Jan 28, 2021

• Initial Normalizer-Free Reg/ResNet impl. A bit of related layer refactoring.

  rwightman

  committed
  5a8e1e6

  Copy full SHA for 5a8e1e6

Pagination

Previous

Next