Fix name parsing for int8 smoothquant model detection

Signed-off-by: Andrea Fasoli <andrea.fasoli@ibm.com>

Author: andrea-fasoli

scripts/inference.py

@@ -394,9 +394,9 @@ def select_int8_module(
 
     if args.int8_smoothquant:
         # TODO: consider saving this info into config during quantization
-        if any("granite" in p for p in [args.model_path, args.architecture]):
+        if any("granite" in p.lower() for p in [args.model_path, args.architecture]):
             smoothquant_layers = ["key", "value", "w1", "wg"]
-        elif any("roberta" in p for p in [args.model_path, args.architecture]):
+        elif any("roberta" in p.lower() for p in [args.model_path, args.architecture]):
             smoothquant_layers = ["query", "key", "value", "w1"]
         else:
             raise NotImplementedError(