Remove generalization example images from model card

README.md

@@ -17,17 +17,4 @@ The Table Transformer is equivalent to [DETR](https://huggingface.co/docs/transf
 
 ## Usage
 
-You can use the raw model for detecting tables in documents. See the [documentation](https://huggingface.co/docs/transformers/main/en/model_doc/table-transformer) for more info.
-
-## Generalization example: general model vs. a customer-specific fine-tune
-
-The examples below compare this general-purpose model (right, trained on PubTables-1M)
-against a YOLO detector fine-tuned only on one customer's financial reports (left) — run
-side by side on a page from an unrelated, non-financial technical report the fine-tuned
-model had never seen. The fine-tuned detector misses the table entirely (0 tables found);
-the general model detects it correctly (1 table found), illustrating that a model tuned
-tightly to one document distribution can generalize worse to out-of-distribution documents
-than a general-purpose detector.
-
-![Generalization example — page 2](examples/page2_detection_compare.png)
-![Generalization example — page 38](examples/page38_detection_compare.png)
+You can use the raw model for detecting tables in documents. See the [documentation](https://huggingface.co/docs/transformers/main/en/model_doc/table-transformer) for more info.