DOC In-place modification through get_peft_model (#2313)

docs/source/tutorial/peft_model_config.md

@@ -135,6 +135,9 @@ lora_model.print_trainable_parameters()
 "trainable params: 1,572,864 || all params: 332,769,280 || trainable%: 0.472659014678278"
 ```
 
+> [!WARNING]
+> When calling [`get_peft_model`], the base model will be modified *in-place*. That means, when calling [`get_peft_model`] on a model that was already modified in the same way before, this model will be further mutated. Therefore, if you would like to modify your PEFT configuration after having called [`get_peft_model()`] before, you would first have to unload the model with [`~LoraModel.unload`] and then call [`get_peft_model()`] with your new configuration. Alternatively, you can re-initialize the model to ensure a fresh, unmodified state before applying a new PEFT configuration.
+
 Now you can train the [`PeftModel`] with your preferred training framework! After training, you can save your model locally with [`~PeftModel.save_pretrained`] or upload it to the Hub with the [`~transformers.PreTrainedModel.push_to_hub`] method.
 
 ```py

src/peft/mapping.py

@@ -154,7 +154,7 @@ def get_peft_model(
     low_cpu_mem_usage: bool = False,
 ) -> PeftModel | PeftMixedModel:
     """
-    Returns a Peft model object from a model and a config.
+    Returns a Peft model object from a model and a config, where the model will be modified in-place.
 
     Args:
         model ([`transformers.PreTrainedModel`]):