68a4144882
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.
2021-02-06 16:30:02 -08:00
f0e65e37b7
Fix NF-ResNet101 model defs
2021-01-30 23:26:19 -08:00
2de54d174a
Fix pool size defs for NFNet models, add a comment.
2021-01-30 18:02:33 -08:00
90980de4a9
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.
2021-01-30 16:32:07 -08:00
5a8e1e643e
Initial Normalizer-Free Reg/ResNet impl. A bit of related layer refactoring.
2021-01-27 22:06:57 -08:00
38d8f67570
Fix potential issue with change to num_classes arg in train/validate.py defaulting to None (rely on model def / default_cfg)
2021-01-25 11:53:34 -08:00
bb50ac4708
Add DeiT distilled weights and distilled model def. Remove some redudant ViT model args.
2021-01-25 11:05:23 -08:00
c16e965037
Add some ViT comments and fix a few minor issues.
2021-01-24 23:18:35 -08:00
22748f1a2d
Convert samples/targets in ParserImageInTar to numpy arrays, slightly less mem usage for massive datasets. Add a few more se/eca model defs to resnet.py
2021-01-22 16:54:33 -08:00
55f7dfa9ea
Refactor vision_transformer entrpy fns, add pos embedding resize support for fine tuning, add some deit models for testing
2021-01-18 16:11:02 -08:00
d55bcc0fee
Finishing adding stochastic depth support to BiT ResNetV2 models
2021-01-16 16:32:03 -08:00
855d6cc217
More dataset work including factories and a tensorflow datasets (TFDS) wrapper
...
* Add parser/dataset factory methods for more flexible dataset & parser creation
* Add dataset parser that wraps TFDS image classification datasets
* Tweak num_classes handling bug for 21k models
* Add initial deit models so they can be benchmarked in next csv results runs
2021-01-15 17:26:20 -08:00
20516abc18
Fix some broken tests for ResNetV2 BiT models
2021-01-04 23:21:39 -08:00
59ec7e6a53
Merge branch 'master' into imagenet21k_datasets_more
2021-01-04 12:11:05 -08:00
4e2533db77
Add 320x320 model default cfgs for 101D and 152D ResNets. Add SEResNet-152D weights and 320x320 cfg.
2021-01-03 12:10:25 -08:00
0167f749d3
Remove some old __future__ imports
2021-01-03 11:24:16 -08:00
ce69de70d3
Add 21k weight urls to vision_transformer. Cleanup feature_info for preact ResNetV2 (BiT) models
2020-12-28 16:59:15 -08:00
231d04e91a
ResNetV2 pre-act and non-preact model, w/ BiT pretrained weights and support for ViT R50 model. Tweaks for in21k num_classes passing. More to do... tests failing.
2020-12-28 16:59:15 -08:00
392595c7eb
Add pool_size to default cfgs for new models to prevent tests from failing. Add explicit 200D_320 model entrypoint for next benchmark run.
2020-12-18 21:28:47 -08:00
b1f1228a41
Add ResNet101D, 152D, and 200D weights, remove meh 66d model
2020-12-18 17:13:37 -08:00
cd72e66eff
Bug in last mod for features_only default_cfg
2020-12-03 12:33:01 -08:00
867a0e5a04
Add default_cfg back to models wrapped in feature extraction module as per discussion in #294 .
2020-12-03 10:24:35 -08:00
2ed8f24715
A few more changes for 0.3.2 maint release. Linear layer change for mobilenetv3 and inception_v3, support no bias for linear wrapper.
2020-11-30 16:19:52 -08:00
460eba7f24
Work around casting issue with combination of native torch AMP and torchscript for Linear layers
2020-11-30 13:30:51 -08:00
5f4b6076d8
Fix inplace arg compat for GELU and PreLU via activation factory
2020-11-30 13:27:40 -08:00
fd962c4b4a
Native SiLU (Swish) op doesn't export to ONNX
2020-11-29 21:56:55 -08:00
b401952caf
Add newly added vision transformer large/base 224x224 weights ported from JAX official repo
2020-10-29 17:31:01 -07:00
61200db0ab
in_chans=1 working w/ pretrained weights for vision_transformer
2020-10-29 15:49:36 -07:00
e90edce438
Support native silu activation (aka swish). An optimized ver is available in PyTorch 1.7.
2020-10-29 15:45:17 -07:00
da6cd2cc1f
Fix regression for pretrained classifier loading when using entrypt functions directly
2020-10-29 15:43:39 -07:00
f591e90b0d
Make sure num_features attr is present in vit models as with others
2020-10-29 15:33:47 -07:00
f944242cb0
Fix #262 , num_classes arg mixup. Make vision_transformers a bit closer to other models wrt get/reset classfier/forward_features. Fix torchscript for ViT.
2020-10-29 13:58:28 -07:00
736f209e7d
Update vision transformers to be compatible with official code. Port official ViT weights from jax impl.
2020-10-26 18:42:11 -07:00
27a93e9de7
Improve test crop for ViT models. Small now 77.85, added base weights at 79.35 top-1.
2020-10-21 23:35:25 -07:00
d4db9e7977
Add small vision transformer weights. 77.42 top-1.
2020-10-21 12:14:12 -07:00
f31933cb37
Initial Vision Transformer impl w/ patch and hybrid variants. Refactor tuple helpers.
2020-10-13 13:33:44 -07:00
fcb6258877
Add missing leaky_relu layer factory defn, update Apex/Native loss scaler interfaces to support unscaled grad clipping. Bump ver to 0.2.2 for pending release.
2020-10-02 16:19:39 -07:00
e8e2d9cabf
Add DropPath (stochastic depth) to ReXNet and VoVNet. RegNet DropPath impl tweak and dedupe se args.
2020-09-24 18:20:36 -07:00
e8ca45854c
More models in sotabench, more control over sotabench run, dataset filename extraction consistency
2020-09-24 15:56:57 -07:00
9c406532bd
Add EfficientNet-EdgeTPU-M (efficientnet_em) model trained natively in PyTorch. More sotabench fiddling.
2020-09-23 17:12:07 -07:00
c40384f5bd
Add ResNet weights. 80.5 (top-1) ResNet-50-D, 77.1 ResNet-34-D, 72.7 ResNet-18-D.
2020-09-18 12:05:37 -07:00
33f8a1bf36
Updated README, add wide_resnet50_2 and seresnext50_32x4d weights
2020-09-03 10:45:17 -07:00
751b0bba98
Add global_pool (--gp) arg changes to allow passing 'fast' easily for train/validate to avoid channels_last issue with AdaptiveAvgPool
2020-09-02 16:13:47 -07:00
9c297ec67d
Cleanup Apex vs native AMP scaler state save/load. Cleanup CheckpointSaver a bit.
2020-09-02 15:12:59 -07:00
80c9d9cc72
Add 'fast' global pool option, remove redundant SEModule from tresnet, normal one is now 'fast'
2020-09-02 09:11:48 -07:00
90a01f47d1
hrnet features_only pretrained weight loading issue. Fix #232 .
2020-09-01 17:37:55 -07:00
110a7c4982
AdaptiveAvgPool2d -> mean((2,3)) for all SE/attn layers to avoid NaN with AMP + channels_last layout. See https://github.com/pytorch/pytorch/issues/43992
2020-09-01 16:05:32 -07:00
470220b1f4
Fix MobileNetV3 crash with global_pool='', output consistent with other models but not equivalent due to efficient head.
2020-08-18 14:11:30 -07:00
fc8b8afb6f
Fix a silly bug in Sample version of EvoNorm missing x* part of swish, update EvoNormBatch to accumulated unbiased variance.
2020-08-13 18:25:01 -07:00
0f5d9d8166
Add CSPResNet50 weights, 79.6 top-1 at 256x256
2020-08-12 11:20:11 -07:00
Ross Wightman