diff --git a/src/axolotl/cli/merge_lora.py b/src/axolotl/cli/merge_lora.py
index bc2dc84c7..0a4b3c612 100644
--- a/src/axolotl/cli/merge_lora.py
+++ b/src/axolotl/cli/merge_lora.py
@@ -26,7 +26,7 @@ def do_merge_lora(*, cfg: DictDefault) -> None:
     model, tokenizer, processor = load_model_and_tokenizer(cfg=cfg)
 
     LOG.info("Running merge of LoRA with base model...")
-    model = model.merge_and_unload(progressbar=True)
+    model = model.merge_and_unload(progressbar=True, safe_merge=True)
     try:
         model.to(dtype=cfg.torch_dtype)
     except ValueError as e:
diff --git a/src/axolotl/loaders/model.py b/src/axolotl/loaders/model.py
index 03c1f35bc..1b5131e19 100644
--- a/src/axolotl/loaders/model.py
+++ b/src/axolotl/loaders/model.py
@@ -226,7 +226,7 @@ class ModelLoader:
             isinstance(self.model, (peft.PeftModel, peft.PeftModelForCausalLM))
             and not self.is_qlora_and_fsdp_enabled
         ):
-            self.model = self.model.merge_and_unload()
+            self.model = self.model.merge_and_unload(safe_merge=True)
 
         self._configure_experts_implementation()
         self._apply_activation_checkpointing()
diff --git a/src/axolotl/train.py b/src/axolotl/train.py
index 16c3696c0..143509ce8 100644
--- a/src/axolotl/train.py
+++ b/src/axolotl/train.py
@@ -257,7 +257,7 @@ def save_trained_model(
     # Handle ReLoRA early return case
     if cfg.relora:
         if cfg.adapter == "lora" and not (cfg.load_in_4bit or cfg.load_in_8bit):
-            model = model.merge_and_unload()
+            model = model.merge_and_unload(safe_merge=True)
         else:
             # final model weights have already been saved by `ReLoRACallback.on_train_end`
             return
diff --git a/tests/utils/lora/test_merge_lora.py b/tests/utils/lora/test_merge_lora.py
index 8edccafb9..830581e2a 100644
--- a/tests/utils/lora/test_merge_lora.py
+++ b/tests/utils/lora/test_merge_lora.py
@@ -69,7 +69,7 @@ class TestAdapterMergeUnmerge:
 
         self.scaling = alpha / r
 
-        def mock_merge_and_unload(progressbar=False):
+        def mock_merge_and_unload(progressbar=False, safe_merge=False):
             """Simulate the actual merge operation"""
             # Apply LoRA delta to base weights: W_new = W_base + (B @ A) * scaling
             delta_q = (self.lora_B_q @ self.lora_A_q) * self.scaling