Ross Wightman
13e8da2b46
SelectKernel split_input works best when input channels split like grouped conv, but output is full width. Disable zero_init for SK nets, seems a bad combo.
5 years ago
Ross Wightman
7d07ebb660
Adding some configs to sknet, incl ResNet50 variants from 'Compounding ... Assembled Techniques' paper and original SKNet50
5 years ago
Ross Wightman
a9d2424fd1
Add separate zero_init_last_bn function to support more block variety without a mess
5 years ago
Ross Wightman
355aa152d5
Just leave it float for now, will look at fp16 later. Remove unused reference code.
5 years ago
Ross Wightman
ef457555d3
BlockDrop working on GPU
5 years ago
Ross Wightman
3ff19079f9
Missed nn_ops.py from last commit
5 years ago
Ross Wightman
9f11b4e8a2
Add ConvBnAct layer to parallel integrated SelectKernelConv, add support for DropPath and DropBlock to ResNet base and SK blocks
5 years ago
Ross Wightman
cefc9b7761
Move SelectKernelConv to conv2d_layers and more
...
* always apply attention in SelectKernelConv, leave MixedConv for no attention alternative
* make MixedConv torchscript compatible
* refactor first/previous dilation name to make more sense in ResNet* networks
5 years ago
Ross Wightman
9abe610931
Used wrong channel var for split
5 years ago
Ross Wightman
58e28dc7e7
Move Selective Kernel blocks/convs to their own sknet.py file
5 years ago
Ross Wightman
a93bae6dc5
A SelectiveKernelBasicBlock for more experiments
5 years ago
Ross Wightman
ad087b4b17
Missed bias=False in selection conv
5 years ago
Ross Wightman
c8b3d6b81a
Initial impl of Selective Kernel Networks. Very much a WIP.
5 years ago
Ross Wightman
1daa303744
Add support to Dataset for class id mapping file, clean up a bit of old logic. Add results file arg for validation and update script.
5 years ago
Ross Wightman
91534522f9
Add newly added TF ported EfficientNet-B8 weights (RandAugment)
5 years ago
Ross Wightman
12dbc74742
New ResNet50 JSD + RandAugment weights
5 years ago
Ross Wightman
2f41905ba5
Update ResNet50 weights to AuxMix trained 78.994 top-1. A few commentes re 'tiered_narrow' tn variant.
5 years ago
Ross Wightman
d9a6a9d0af
Merge pull request #74 from rwightman/augmix-jsd
...
AugMix, JSD loss, SplitBatchNorm (Auxiliary BN), and more
5 years ago
Ross Wightman
3eb4a96eda
Update AugMix, JSD, etc comments and references
5 years ago
Ross Wightman
a28117ea46
Add tiered narrow ResNet (tn) and weights for seresnext26tn_32x4d
5 years ago
Ross Wightman
833066b540
A few minor things in SplitBN
5 years ago
Ross Wightman
7547119891
Add SplitBatchNorm. AugMix, Rand/AutoAugment, Split (Aux) BatchNorm, Jensen-Shannon Divergence, RandomErasing all working together
5 years ago
Ross Wightman
2e955cfd0c
Update RandomErasing with some improved arg names, tweak to aspect range
5 years ago
Ross Wightman
3cc0f91e23
Fix augmix variable name scope overlap, default non-blended mode
5 years ago
Ross Wightman
ec0dd4053a
Add updated RandAugment trained EfficientNet-B0 trained weights from @michaelklachko
5 years ago
Ross Wightman
40fea63ebe
Add checkpoint averaging script. Add headers, shebangs, exec perms to all scripts
5 years ago
Ross Wightman
4666cc9aed
Add --pin-mem arg to enable dataloader pin_memory (showing more benefit in some scenarios now), also add --torchscript arg to validate.py for testing models with jit.script
5 years ago
Ross Wightman
53001dd292
ResNet / Res2Net additions:
...
* ResNet torchscript compat
* output_stride arg supported to limit network stride via dilations (support for dilation added to Res2Net)
* allow activation layer to be changed via act_layer arg
5 years ago
Ross Wightman
f96b3e5e92
InceptionResNetV2 torchscript compatible
5 years ago
Ross Wightman
19d93fe454
Add selecsls60 weights
5 years ago
Ross Wightman
0062c15fb0
Update checkpoint url with modelzoo compatible ones.
5 years ago
Ross Wightman
b5315e66b5
Streamline SelecSLS model without breaking checkpoint compat. Move cfg handling out of model class. Update feature/pooling behaviour to match current.
5 years ago
Ross Wightman
d59a756c16
Run PyCharm autoformat on selecsls and change mix cap variables and model names to all lower
5 years ago
Ross Wightman
fb3a0f4bb8
Merge pull request #65 from mehtadushy/selecsls
...
Incorporate SelecSLS Models
5 years ago
Ross Wightman
19fc205a4d
Update comments on the new SE-ResNeXt26 models
5 years ago
Ross Wightman
acc3ed2b8c
Add EfficientNet-B3 weights, trained from scratch with RA.
5 years ago
Dushyant Mehta
2404361f62
correct asset paths
5 years ago
Dushyant Mehta
31939311f6
Added SelecSLS models
5 years ago
rwightman
1f4498f217
Add ResNet deep tiered stem and model weights for seresnext26t_32x4d and seresnext26d_32x4d
5 years ago
Dushyant Mehta
32012a44fd
Added SelecSLS model
5 years ago
Ross Wightman
73b78459dc
Add update RandAugment MixNet-XL weights
5 years ago
Ross Wightman
3afc2a4dc0
Some cleanup/improvements to AugMix impl:
...
* make 'increasing' levels for Contrast, Color, Brightness, Saturation ops
* remove recursion from faster blending mix
* add config striing parsing for AugMix
5 years ago
Ross Wightman
232ab7fb12
Working on an implementation of AugMix with JensenShannonDivergence loss that's compatible with my AutoAugment and RandAugment impl
5 years ago
Ross Wightman
a435ea1327
Change reduce_bn to distribute_bn, add ability to choose between broadcast and reduce (mean). Add crop_pct arg to allow selecting validation crop while training.
5 years ago
Ross Wightman
3bff2b21dc
Add support for keeping running bn stats the same across distributed training nodes before eval/save
5 years ago
Ross Wightman
0161de0127
Switch RandoErasing back to on GPU normal sampling
5 years ago
Ross Wightman
ff421e5e09
New PyTorch trained EfficientNet-B2 weights with my RandAugment impl
5 years ago
Ross Wightman
3bef524f9c
Finish with HRNet, weights and models updated. Improve consistency in model classifier/global pool treatment.
5 years ago
Ross Wightman
6ca0828166
Update EfficientNet comments, MobileNetV3 non-TF create fns, fix factory arg checks, bump PyTorch version req to 1.2
5 years ago
Ross Wightman
eccbadca74
Update EfficientNet comments
5 years ago