Update app.py

app.py

@@ -7,16 +7,19 @@ MODEL_ID = "newtechdevng/qwen-math-tutor"
 
 tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
 
-# Load config and strip quantization settings
+# Load and fix config — remove quantization completely
 config = AutoConfig.from_pretrained(MODEL_ID)
-config.quantization_config = {}  # ← remove baked-in quantization
+if hasattr(config, "quantization_config"):
+    delattr(config, "quantization_config")  # ← delete it entirely
 
 model = AutoModelForCausalLM.from_pretrained(
     MODEL_ID,
-    config  = config,
-    dtype   = torch.float32,  # ← fixed from torch_dtype
+    config        = config,
+    torch_dtype   = torch.float32,
+    low_cpu_mem_usage = True,
 )
 model = model.to("cpu")
+model.eval()
 print("Model loaded!")
 
 def solve(question, history):