Update app.py

app.py

@@ -7,9 +7,9 @@ model_name = "tanusrich/Mental_Health_Chatbot"
 model = AutoModelForCausalLM.from_pretrained(
     model_name,
     torch_dtype=torch.float16,  # Reduce memory usage
-    device_map="auto",  # Automatically assigns to GPU if available
+    device_map="cpu",  # Automatically assigns to GPU if available
     low_cpu_mem_usage=True,
-    max_memory={0: "4GiB", "cpu": "6GiB"}, # Optimize CPU memory
+    max_memory={0: "3.5GiB", "cpu": "12GiB"}, # Optimize CPU memory
     offload_folder=None
 )
 
@@ -23,7 +23,7 @@ model.save_pretrained(model_save_path)
 tokenizer.save_pretrained(model_save_path)'''
 
 def generate_response(user_input):
-    inputs = tokenizer(user_input, return_tensors="pt").to(device)
+    inputs = tokenizer(user_input, return_tensors="pt").to("cpu")
     with torch.no_grad():
         output = model.generate(
             **inputs,