[modular] Simplify logic and docstring handling (#39185)

* simplify a lot

* Update modular_model_converter.py

* finalize

* remove outdated functions

* apply it

* and examples

examples/modular-transformers/modeling_new_task_model.py

@@ -437,32 +437,6 @@ class NewTaskModelForNewTask(NewTaskModelPreTrainedModel, GenerationMixin):
         num_logits_to_keep: int = 0,
     ) -> Union[tuple, NewTaskModelCausalLMOutputWithPast]:
         r"""
-        labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
-            Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
-            config.text_config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
-            (masked), the loss is only computed for the tokens with labels in `[0, ..., config.text_config.vocab_size]`.
-
-        Example:
-
-        ```python
-        >>> from PIL import Image
-        >>> import requests
-        >>> from transformers import AutoProcessor, NewTaskModelForNewTask
-
-        >>> model = NewTaskModelForNewTask.from_pretrained("google/new_task_model2-3b-mix-224")
-        >>> processor = AutoProcessor.from_pretrained("google/new_task_model2-3b-mix-224")
-
-        >>> prompt = "Where is the cat standing?"
-        >>> url = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/pipeline-cat-chonk.jpeg"
-        >>> image = Image.open(requests.get(url, stream=True).raw)
-
-        >>> inputs = processor(images=image, text=prompt,  return_tensors="pt")
-
-        >>> # Generate
-        >>> generate_ids = model.generate(**inputs,)
-        >>> processor.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
-        "Where is the cat standing?\nsnow"
-        ```
         Returns:
         """
         vlm_outputs = super().forward(