diff --git a/examples/aishell3/tts3/local/lite_predict.sh b/examples/aishell3/tts3/local/lite_predict.sh
index 8654b2876..e77e8b6c2 100755
--- a/examples/aishell3/tts3/local/lite_predict.sh
+++ b/examples/aishell3/tts3/local/lite_predict.sh
@@ -5,6 +5,7 @@ train_output_path=$1
 stage=0
 stop_stage=0
 
+# pwgan
 if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
     python3 ${BIN_DIR}/../lite_predict.py \
         --inference_dir=${train_output_path}/pdlite \
@@ -17,6 +18,7 @@ if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
         --spk_id=0
 fi
 
+# hifigan
 if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
     python3 ${BIN_DIR}/../lite_predict.py \
         --inference_dir=${train_output_path}/pdlite \
diff --git a/examples/csmsc/tts3/README.md b/examples/csmsc/tts3/README.md
index a2bd62494..39926259d 100644
--- a/examples/csmsc/tts3/README.md
+++ b/examples/csmsc/tts3/README.md
@@ -239,6 +239,7 @@ The ONNX model can be downloaded here:
 - [fastspeech2_cnndecoder_csmsc_streaming_onnx_1.0.0.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/fastspeech2/fastspeech2_cnndecoder_csmsc_streaming_onnx_1.0.0.zip)
 
 The Paddle-Lite model can be downloaded here:
+> please compile develop version of Paddle-Lite to export and run TTS models, cause TTS models are supported by https://github.com/PaddlePaddle/Paddle-Lite/pull/9587 and https://github.com/PaddlePaddle/Paddle-Lite/pull/9706
 - [fastspeech2_csmsc_pdlite_1.3.0.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/fastspeech2/fastspeech2_csmsc_pdlite_1.3.0.zip)
 - [fastspeech2_cnndecoder_csmsc_pdlite_1.3.0.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/fastspeech2/fastspeech2_cnndecoder_csmsc_pdlite_1.3.0.zip)
 - [fastspeech2_cnndecoder_csmsc_streaming_pdlite_1.3.0.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/fastspeech2/fastspeech2_cnndecoder_csmsc_streaming_pdlite_1.3.0.zip)
diff --git a/examples/csmsc/tts3/run.sh b/examples/csmsc/tts3/run.sh
index ce083e660..14308af4e 100755
--- a/examples/csmsc/tts3/run.sh
+++ b/examples/csmsc/tts3/run.sh
@@ -64,6 +64,9 @@ fi
 
 # must run after stage 3 (which stage generated static models)
 if [ ${stage} -le 7 ] && [ ${stop_stage} -ge 7 ]; then
+    # NOTE by yuantian 2022.11.21: please compile develop version of Paddle-Lite to export and run TTS models,
+    #                   cause TTS models are supported by https://github.com/PaddlePaddle/Paddle-Lite/pull/9587 
+    #                   and https://github.com/PaddlePaddle/Paddle-Lite/pull/9706
     ./local/export2lite.sh ${train_output_path} inference pdlite fastspeech2_csmsc x86
     ./local/export2lite.sh ${train_output_path} inference pdlite pwgan_csmsc x86
     # ./local/export2lite.sh ${train_output_path} inference pdlite mb_melgan_csmsc x86
diff --git a/examples/vctk/tts3/local/lite_predict.sh b/examples/vctk/tts3/local/lite_predict.sh
index a4f70fe78..eb608535b 100755
--- a/examples/vctk/tts3/local/lite_predict.sh
+++ b/examples/vctk/tts3/local/lite_predict.sh
@@ -5,6 +5,7 @@ train_output_path=$1
 stage=0
 stop_stage=0
 
+# pwgan
 if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
     python3 ${BIN_DIR}/../lite_predict.py \
         --inference_dir=${train_output_path}/pdlite \
@@ -18,6 +19,7 @@ if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
         --lang=en
 fi
 
+# hifigan
 if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
     python3 ${BIN_DIR}/../lite_predict.py \
         --inference_dir=${train_output_path}/pdlite \
diff --git a/paddlespeech/t2s/exps/syn_utils.py b/paddlespeech/t2s/exps/syn_utils.py
index d256c6347..cea125291 100644
--- a/paddlespeech/t2s/exps/syn_utils.py
+++ b/paddlespeech/t2s/exps/syn_utils.py
@@ -520,7 +520,6 @@ def get_lite_predictor(model_dir: Optional[os.PathLike]=None,
                        cpu_threads: int=1):
     config = MobileConfig()
     config.set_model_from_file(str(Path(model_dir) / model_file))
-    # config.set_threads(cpu_threads)
     predictor = create_paddle_predictor(config)
     return predictor