[vlm] fix loading of retrieval VLMs (#39242)

* fix vlm with retrieval

* we can't use AutoModel because new ColQwen was released after refactor

* no need for colqwen

* tied weight keys are necessary, if using IMageTextToText

* need to apply renaming in tied weights, only for ColPali

* overwrite tied keys in ColPali

* fix copies, modular can't handle if-statements

src/transformers/modeling_utils.py

@@ -231,6 +231,7 @@ TORCH_INIT_FUNCTIONS = {
 VLMS = [
     "aria",
     "ayavision",
+    "colpali",
     "emu3",
     "fuyu",
     "gotocr2",