| .. |
|
layers
|
Add --use-train-size flag to force use of train input_size (over test input size) for validation. Default test-time pooling to use train input size (fixes issues).
|
2 years ago |
|
pruned
|
…
|
|
|
__init__.py
|
Add initial EdgeNeXt import. Significant cleanup / reorg (like ConvNeXt). Fix #1320
|
3 years ago |
|
beit.py
|
Remove some redundant requires_grad=True from nn.Parameter in third party code
|
2 years ago |
|
byoanet.py
|
Transitioning default_cfg -> pretrained_cfg. Improving handling of pretrained_cfg source (HF-Hub, files, timm config, etc). Checkpoint handling tweaks.
|
3 years ago |
|
byobnet.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
cait.py
|
Remove some redundant requires_grad=True from nn.Parameter in third party code
|
2 years ago |
|
coat.py
|
Significant model refactor and additions:
|
3 years ago |
|
convit.py
|
Significant model refactor and additions:
|
3 years ago |
|
convmixer.py
|
Significant model refactor and additions:
|
3 years ago |
|
convnext.py
|
Add dilation support to convnext, allows output_stride=8 and 16 use. Fix #1341
|
2 years ago |
|
crossvit.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
cspnet.py
|
Add cs3sedarknet_x weights w/ 82.65 @ 288 top1. Add 2 cs3 edgenet models (w/ 3x3-1x1 block), remove aa from cspnet blocks (not needed)
|
2 years ago |
|
deit.py
|
Unbreak gamma remap impacting beit checkpoint load, version bump to 0.6.4
|
2 years ago |
|
densenet.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
dla.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
dpn.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
edgenext.py
|
x and xx small edgenext models do benefit from larger test input size
|
2 years ago |
|
efficientnet.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
efficientnet_blocks.py
|
Merge remote-tracking branch 'origin/master' into norm_norm_norm
|
3 years ago |
|
efficientnet_builder.py
|
Merge remote-tracking branch 'origin/master' into norm_norm_norm
|
3 years ago |
|
factory.py
|
Revert ml-decoder changes to model factory and train script
|
3 years ago |
|
features.py
|
Significant model refactor and additions:
|
3 years ago |
|
fx_features.py
|
Update type hint for `register_notrace_module`
|
2 years ago |
|
ghostnet.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
gluon_resnet.py
|
Transitioning default_cfg -> pretrained_cfg. Improving handling of pretrained_cfg source (HF-Hub, files, timm config, etc). Checkpoint handling tweaks.
|
3 years ago |
|
gluon_xception.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
hardcorenas.py
|
Significant model refactor and additions:
|
3 years ago |
|
helpers.py
|
no longer a point using kwargs for pretrain_cfg resolve, just pass explicit arg
|
3 years ago |
|
hrnet.py
|
Ensure all model entrypoint fn default to `pretrained=False` (a few didn't)
|
2 years ago |
|
hub.py
|
👽 use hf_hub_download instead of cached_download
|
2 years ago |
|
inception_resnet_v2.py
|
Significant model refactor and additions:
|
3 years ago |
|
inception_v3.py
|
Improve resolve_pretrained_cfg behaviour when no cfg exists, warn instead of crash. Improve usability ex #1311
|
3 years ago |
|
inception_v4.py
|
Significant model refactor and additions:
|
3 years ago |
|
levit.py
|
Fix pit implementation to be clsoer to deit/levit re distillation head handling
|
3 years ago |
|
mlp_mixer.py
|
Change models with mean 0,0,0 std 1,1,1 from int to float for consistency as mentioned in #1355
|
2 years ago |
|
mobilenetv3.py
|
Change models with mean 0,0,0 std 1,1,1 from int to float for consistency as mentioned in #1355
|
2 years ago |
|
mobilevit.py
|
Change models with mean 0,0,0 std 1,1,1 from int to float for consistency as mentioned in #1355
|
2 years ago |
|
nasnet.py
|
Significant model refactor and additions:
|
3 years ago |
|
nest.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
nfnet.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
pit.py
|
Fix pit implementation to be clsoer to deit/levit re distillation head handling
|
3 years ago |
|
pnasnet.py
|
Significant model refactor and additions:
|
3 years ago |
|
poolformer.py
|
Remove some redundant requires_grad=True from nn.Parameter in third party code
|
2 years ago |
|
registry.py
|
Transitioning default_cfg -> pretrained_cfg. Improving handling of pretrained_cfg source (HF-Hub, files, timm config, etc). Checkpoint handling tweaks.
|
3 years ago |
|
regnet.py
|
Fix group_matcher regex for regnet.py
|
3 years ago |
|
res2net.py
|
Significant model refactor and additions:
|
3 years ago |
|
resnest.py
|
Significant model refactor and additions:
|
3 years ago |
|
resnet.py
|
Ensure all model entrypoint fn default to `pretrained=False` (a few didn't)
|
2 years ago |
|
resnetv2.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
rexnet.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
selecsls.py
|
Significant model refactor and additions:
|
3 years ago |
|
senet.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
sequencer.py
|
Fix tests for rank-4 output where feature channels dim is -1 (3) and not 1
|
3 years ago |
|
sknet.py
|
Significant model refactor and additions:
|
3 years ago |
|
swin_transformer.py
|
Add initial swinv2 tiny / small weights
|
3 years ago |
|
swin_transformer_v2.py
|
Add updated vit_relpos weights, and impl w/ support for official swin-v2 differences for relpos. Add bias control support for MLP layers
|
3 years ago |
|
swin_transformer_v2_cr.py
|
Remove persistent buffers from Swin-V2. Change SwinV2Cr cos attn + tau/logit_scale to match official, add ckpt convert, init_value zeros resid LN weight by default
|
3 years ago |
|
tnt.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
tresnet.py
|
Change models with mean 0,0,0 std 1,1,1 from int to float for consistency as mentioned in #1355
|
2 years ago |
|
twins.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
vgg.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
visformer.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
vision_transformer.py
|
Take `no_emb_class` into account when calling `resize_pos_embed`
|
2 years ago |
|
vision_transformer_hybrid.py
|
Vision Transformer refactoring and Rel Pos impl
|
3 years ago |
|
vision_transformer_relpos.py
|
Update vit_relpos w/ some additional weights, some cleanup to match recent vit updates, more MLP log coord experiments.
|
2 years ago |
|
volo.py
|
Significant model refactor and additions:
|
3 years ago |
|
vovnet.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
xception.py
|
Significant model refactor and additions:
|
3 years ago |
|
xception_aligned.py
|
Fix correctness of some group matching regex (no impact on result), some formatting, missed forward_head for resnet
|
3 years ago |
|
xcit.py
|
Remove some redundant requires_grad=True from nn.Parameter in third party code
|
2 years ago |
Ross Wightman
c8f69e04a9