Fix: Load CPU-safe model for HF Space

app.py

@@ -9,14 +9,16 @@ import io
 # Load model and tokenizer from local fine-tuned directory
 MODEL_PATH = "Tufan1/BioMedLM-Cardio-Fold4-CPU"
 tokenizer = AutoTokenizer.from_pretrained(MODEL_PATH)
-#model = AutoModelForCausalLM.from_pretrained(MODEL_PATH, device_map=None, low_cpu_mem_usage=True, torch_dtype=torch.float32)
-device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+
+# Force CPU-safe model loading
 model = AutoModelForCausalLM.from_pretrained(
     MODEL_PATH,
-    device_map="auto" if torch.cuda.is_available() else None,
-    torch_dtype=torch.float16 if torch.cuda.is_available() else torch.float32
-).to(device)
-#model.to(device)
+    device_map="cpu",                # Ensure model loads on CPU
+    torch_dtype=torch.float32        # Ensure safe dtype for CPU
+)
+
+device = torch.device("cpu")         # Explicitly define CPU
+
 
 # Dictionaries to decode user inputs
 gender_map = {1: "Female", 2: "Male"}