Decouples `XLMProphet` model from `Prophet` (#19406)

* decouples xlm_prophet from prophet and adds copy patterns that pass the copy check

* adds copy patterns to copied docstrings too

* restores autodoc for XLMProphetNetModel

* removes all-casing in a bunch of places to ensure that the model is compatible with all checkpoints on the hub

* adds missing model to main init

* adds autodocs to make document checker happy

* adds missing pretrained model import

* adds missing pretrained model import to main init

* adds XLMProphetNetPreTrainedModel to the dummy pt objects

* removes examples from the source-doc file since docstrings contain them already

* adds a missing new line to make check_repo happy

src/transformers/__init__.py

@@ -2113,6 +2113,7 @@ else:
             "XLMProphetNetForCausalLM",
             "XLMProphetNetForConditionalGeneration",
             "XLMProphetNetModel",
+            "XLMProphetNetPreTrainedModel",
         ]
     )
     _import_structure["models.xlm_roberta"].extend(
@@ -4796,6 +4797,7 @@ if TYPE_CHECKING:
             XLMProphetNetForCausalLM,
             XLMProphetNetForConditionalGeneration,
             XLMProphetNetModel,
+            XLMProphetNetPreTrainedModel,
         )
         from .models.xlm_roberta import (
             XLM_ROBERTA_PRETRAINED_MODEL_ARCHIVE_LIST,

src/transformers/models/xlm_prophetnet/__init__.py

@@ -45,6 +45,7 @@ else:
         "XLMProphetNetForCausalLM",
         "XLMProphetNetForConditionalGeneration",
         "XLMProphetNetModel",
+        "XLMProphetNetPreTrainedModel",
     ]
 
 
@@ -72,6 +73,7 @@ if TYPE_CHECKING:
             XLMProphetNetForCausalLM,
             XLMProphetNetForConditionalGeneration,
             XLMProphetNetModel,
+            XLMProphetNetPreTrainedModel,
         )
 
 else:

src/transformers/utils/dummy_pt_objects.py

@@ -5637,6 +5637,13 @@ class XLMProphetNetModel(metaclass=DummyObject):
         requires_backends(self, ["torch"])
 
 
+class XLMProphetNetPreTrainedModel(metaclass=DummyObject):
+    _backends = ["torch"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch"])
+
+
 XLM_ROBERTA_PRETRAINED_MODEL_ARCHIVE_LIST = None