diff --git a/paddlespeech/t2s/exps/syn_utils.py b/paddlespeech/t2s/exps/syn_utils.py
index 75b98f1e0..d29dd8110 100644
--- a/paddlespeech/t2s/exps/syn_utils.py
+++ b/paddlespeech/t2s/exps/syn_utils.py
@@ -591,7 +591,7 @@ def get_predictor(
 
     config = inference.Config(
         str(Path(model_dir) / model_file), str(Path(model_dir) / params_file))
-    if device != "npu":
+    if paddle.__version__ <= "2.5.2" and paddle.__version__ != "0.0.0":
         config.enable_memory_optim()
     config.switch_ir_optim(True)
     if device == "gpu":
diff --git a/paddlespeech/t2s/exps/synthesize.py b/paddlespeech/t2s/exps/synthesize.py
index 5e403e7e7..fa3ac1569 100644
--- a/paddlespeech/t2s/exps/synthesize.py
+++ b/paddlespeech/t2s/exps/synthesize.py
@@ -219,18 +219,18 @@ def parse_args():
     )
     # other
     parser.add_argument(
-        "--ngpu", type=int, default=1, help="if ngpu == 0, use cpu or xpu or npu.")
+        "--ngpu", type=int, default=1, help="if wish to use gpu, set ngpu > 0, otherwise use xpu, npu or cpu.")
     parser.add_argument(
         "--nxpu",
         type=int,
         default=0,
-        help="if wish to use xpu, set ngpu == 0 and nxpu > 0, and if ngpu == 0 and nxpu == 0 and nnpu == 0, use cpu."
+        help="if wish to use xpu, set ngpu == 0 and nxpu > 0, otherwise use gpu, npu or cpu."
     )
     parser.add_argument(
         "--nnpu",
         type=int,
         default=0,
-        help="if wish to use npu, set ngpu == 0 and nxpu == 0 and nnpu > 0, and if ngpu == 0 and nxpu == 0 and nnpu == 0, use cpu."
+        help="if wish to use npu, set ngpu == 0 and nnpu > 0, otherwise use gpu, xpu or cpu."
     )
     parser.add_argument("--test_metadata", type=str, help="test metadata.")
     parser.add_argument("--output_dir", type=str, help="output dir.")