diff --git a/modules/models.py b/modules/models.py
index f4bb11fd..a934514b 100644
--- a/modules/models.py
+++ b/modules/models.py
@@ -98,7 +98,7 @@ def load_model(model_name):
         command = "AutoModelForCausalLM.from_pretrained"
         params = ["low_cpu_mem_usage=True"]
         if not shared.args.cpu and not torch.cuda.is_available():
-            print("Warning: no GPU has been detected.\nFalling back to CPU mode.\n")
+            print("Warning: torch.cuda.is_available() returned False.\nThis means that no GPU has been detected.\nFalling back to CPU mode.\n")
             shared.args.cpu = True
 
         if shared.args.cpu: