diff --git a/tools/audio_sr.py b/tools/audio_sr.py
index b714ab1..d51f055 100644
--- a/tools/audio_sr.py
+++ b/tools/audio_sr.py
@@ -12,7 +12,8 @@ import numpy as np
 import torchaudio
 import time
 import torchaudio.functional as aF
-from attrdict import AttrDict
+# from attrdict import AttrDict####will be bug in py3.10
+
 from datasets1.dataset import amp_pha_stft, amp_pha_istft
 from models.model import APNet_BWE_Model
 import soundfile as sf
@@ -20,7 +21,7 @@ import matplotlib.pyplot as plt
 from rich.progress import track
 
 class AP_BWE():
-    def __init__(self,device,checkpoint_file=None):
+    def __init__(self,device,DictToAttrRecursive,checkpoint_file=None):
         if checkpoint_file==None:
             checkpoint_file="%s/24kto48k/g_24kto48k.zip"%(AP_BWE_main_dir_path)
             if os.path.exists(checkpoint_file)==False:
@@ -28,7 +29,8 @@ class AP_BWE():
         config_file = os.path.join(os.path.split(checkpoint_file)[0], 'config.json')
         with open(config_file) as f:data = f.read()
         json_config = json.loads(data)
-        h = AttrDict(json_config)
+        # h = AttrDict(json_config)
+        h = DictToAttrRecursive(json_config)
         model = APNet_BWE_Model(h).to(device)
         state_dict = torch.load(checkpoint_file,map_location="cpu",weights_only=False)
         model.load_state_dict(state_dict['generator'])
@@ -46,4 +48,4 @@ class AP_BWE():
             amp_wb_g, pha_wb_g, com_wb_g = self.model(amp_nb, pha_nb)
             audio_hr_g = amp_pha_istft(amp_wb_g, pha_wb_g, self.h.n_fft, self.h.hop_size, self.h.win_size)
             # sf.write(opt_path, audio_hr_g.squeeze().cpu().numpy(), self.h.hr_sampling_rate, 'PCM_16')
-            return audio_hr_g.squeeze().cpu().numpy(),self.h.hr_sampling_rate
\ No newline at end of file
+            return audio_hr_g.squeeze().cpu().numpy(),self.h.hr_sampling_rate