Merge pull request #2480 from brianhou0208/tnt

Update TNT-(S/B) model weights and add feature extraction support

Author: rwightman

tests/test_models.py

@@ -53,7 +53,7 @@
     'vision_transformer', 'vision_transformer_sam', 'vision_transformer_hybrid', 'vision_transformer_relpos',
     'beit', 'mvitv2', 'eva', 'cait', 'xcit', 'volo', 'twins', 'deit', 'swin_transformer', 'swin_transformer_v2',
     'swin_transformer_v2_cr', 'maxxvit', 'efficientnet', 'mobilenetv3', 'levit', 'efficientformer', 'resnet',
-    'regnet', 'byobnet', 'byoanet', 'mlp_mixer', 'hiera', 'fastvit', 'hieradet_sam2', 'aimv2*',
+    'regnet', 'byobnet', 'byoanet', 'mlp_mixer', 'hiera', 'fastvit', 'hieradet_sam2', 'aimv2*', 'tnt',
     'tiny_vit', 'vovnet', 'tresnet', 'rexnet', 'resnetv2', 'repghost', 'repvit', 'pvt_v2', 'nextvit', 'nest',
     'mambaout', 'inception_next', 'inception_v4', 'hgnet', 'gcvit', 'focalnet', 'efficientformer_v2', 'edgenext',
     'davit', 'rdnet', 'convnext', 'pit'