Update app.py

app.py

@@ -26,7 +26,7 @@ model_id = "google/gemma-2b"  # Use Google Gemma 2B
 tokenizer = AutoTokenizer.from_pretrained(model_id)
 
 # Load the model without quantization for CPU
-logger.info("Loading model for CPU...")
+logger.info("Loading model...")
 model = AutoModelForCausalLM.from_pretrained(
     model_id,
     torch_dtype=torch.float32,  # Use FP32 for CPU compatibility
@@ -37,8 +37,7 @@ model = AutoModelForCausalLM.from_pretrained(
 pipe = pipeline(
     "text-generation",
     model=model,
-    tokenizer=tokenizer,
-    device="cpu"  # Explicitly set device to CPU
+    tokenizer=tokenizer
 )
 
 # Define request body schema