Add `library_name` and `pipeline_tag` metadata

This PR adds the missing `library_name` (`transformers`) and `pipeline_tag` (`image-to-text`) to the YAML metadata. This improves discoverability and enables the correct automatic usage snippet on the Hugging Face Hub. The existing model card content and links are unchanged.

README.md

@@ -1,8 +1,10 @@
 ---
-license: apache-2.0
 language:
 - en
 - zh
+license: apache-2.0
+library_name: transformers
+pipeline_tag: image-to-text
 ---
 
 # WeVisDoc
@@ -185,4 +187,4 @@ Use `Tencent/WeVisDoc-4B` for the 4B version. `--model` can be omitted when `WEV
   year    = {2026},
   journal = {arXiv preprint arXiv: 2609.20423}
 }
-```
+```