diff --git a/docs/requirements.txt b/docs/requirements.txt
index ec57bd0ec495aa77660b74d54af30ec12f55530c..ee116a9b686494bf813b73fde0ed91ca914de4ba 100644
--- a/docs/requirements.txt
+++ b/docs/requirements.txt
@@ -37,7 +37,6 @@ soundfile~=0.10
 textgrid
 timer
 tqdm
-transformers==3.4.0
 typeguard
 visualdl
 webrtcvad
diff --git a/paddlespeech/t2s/frontend/g2pw/onnx_api.py b/paddlespeech/t2s/frontend/g2pw/onnx_api.py
index 8f4e89adc2d35a9f3ad61ee7ef43f4f8fdaf61b0..8861a80b28481206e40ed1b45c4670fc817dc74e 100644
--- a/paddlespeech/t2s/frontend/g2pw/onnx_api.py
+++ b/paddlespeech/t2s/frontend/g2pw/onnx_api.py
@@ -5,7 +5,7 @@ import numpy as np
 
 from opencc import OpenCC
 
-from transformers import BertTokenizer
+from paddlenlp.transformers import BertTokenizer
 
 from paddlespeech.t2s.frontend.g2pw.dataset import prepare_data, prepare_onnx_input, get_phoneme_labels, get_char_phoneme_labels
 from paddlespeech.t2s.frontend.g2pw.utils import load_config
diff --git a/setup.py b/setup.py
index ceb91dbc38ddd105142d15e6f3e64ef55e87c2f3..373484fbc382c646ab31d841370e529d9faeaa16 100644
--- a/setup.py
+++ b/setup.py
@@ -61,7 +61,6 @@ base = [
     "scipy",
     "sentencepiece~=0.1.96",
     "soundfile~=0.10",
-    "transformers==3.4.0",
     "textgrid",
     "timer",
     "tqdm",