Improve model card: Add pipeline tag, library name, paper link, and fix sample usage

This PR enhances the model card for StainNet by:
- Adding `pipeline_tag: image-feature-extraction` to the metadata, improving discoverability on the Hugging Face Hub under the relevant task.
- Adding `library_name: timm` to the metadata, as evidenced by the sample usage, which will enable the automated "how to use" widget for `timm` models.
- Including a direct link to the paper, [StainNet: A Special Staining Self-Supervised Vision Transformer for Computational Pathology](https://huggingface.co/papers/2512.10326), in the main content for easier access.
- Correcting the sample usage code snippet by adding `import torchvision.transforms as transforms`, which is necessary for the provided `transforms.Compose` to function correctly.

README.md

@@ -1,8 +1,11 @@
 ---
-license: cc-by-nc-nd-4.0
 language:
 - en
+license: cc-by-nc-nd-4.0
+pipeline_tag: image-feature-extraction
+library_name: timm
 ---
+
 # Model Card for StainNet
 
 <!-- Provide a quick summary of what the model is/does. -->
@@ -11,11 +14,14 @@ language:
 
 The model is a Vision Transformer Small/16 with DINO [1] self-supervised pre-training on 1,418,938 patch images from 20,231 special staining whole slide images (WSIs) in HISTAI [2].
 
+This model was presented in the paper [StainNet: A Special Staining Self-Supervised Vision Transformer for Computational Pathology](https://huggingface.co/papers/2512.10326).
+
 ## Using StainNet to extract features from special staining pathology image
 
 ```python
 import timm
 import torch
+import torchvision.transforms as transforms
 
 model = timm.create_model('hf_hub:JWonderLand/StainNet', pretrained=True)