Add CATSLU

app.py

@@ -35,7 +35,7 @@ speech2text_fsc = Speech2Text.from_pretrained(
     nbest=1
 )
 
-speech2text_fsc = Speech2Text.from_pretrained(
+speech2text_catslu = Speech2Text.from_pretrained(
     asr_train_config="catslu/config.yaml",
     asr_model_file="catslu/valid.acc.ave_5best.pth",
     # Decoding parameters are not included in the model file
@@ -75,7 +75,7 @@ def inference(wav,data):
             speech=speech[:,0]
             # soundfile.write("store_file.wav", speech, rate, subtype='FLOAT')
           print(speech.shape)
-          nbests = speech2text_fsc(speech)
+          nbests = speech2text_catslu(speech)
           text, *_ = nbests[0]
           text=text.split(" ")[0]
           # intent=text.split(" ")[0]