From a86dad56bcc001fecc3f9d38c52f9117a1093818 Mon Sep 17 00:00:00 2001
From: Cyril Vallez <cyril.vallez@huggingface.co>
Date: Fri, 28 Mar 2025 17:57:16 +0100
Subject: [PATCH] Fix state_dict map location when quantized (#37086)

* Update modeling_utils.py

* Update modeling_utils.py
---
 src/transformers/modeling_utils.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/transformers/modeling_utils.py b/src/transformers/modeling_utils.py
index ac84297280..b14626cc65 100644
--- a/src/transformers/modeling_utils.py
+++ b/src/transformers/modeling_utils.py
@@ -4897,7 +4897,7 @@ class PreTrainedModel(nn.Module, ModuleUtilsMixin, GenerationMixin, PushToHubMix
 
             map_location = "cpu"
             if low_cpu_mem_usage:
-                if shard_file.endswith(".safetensors") and not is_quantized:
+                if shard_file.endswith(".safetensors"):
                     map_location = "meta"
                 elif (
                     device_map is not None