Update app.py

app.py

@@ -2,6 +2,8 @@ import nemo.collections.asr as nemo_asr
 import gradio as gr
 
 asr_model = nemo_asr.models.EncDecCTCModelBPE.from_pretrained(model_name="stt_rw_conformer_ctc_large")
+df = pd.read_csv("amasaku_data.tsv",sep='\t')
+amasaku_mapping = {key.lower():val.lower() for key,val in zip(df.iloc[:,0],df.iloc[:,1])}
 
 def transcribe(file):
 
@@ -10,8 +12,12 @@ def transcribe(file):
     
     print("filename: ",file)
     transcription= asr_model.transcribe([file])
-    print(transcription)
-    return  transcription[0]
+    transcription = transcription.lower().split()
+    transcribed_with_amasuku = []
+    for word in transcription[0]:
+        transcribed_with_amasuku(amasaku_mapping.get(word,word))
+    transcribed_with_amasuku = " ".join(transcribed_with_amasuku)
+    return  transcription_with_amasaku.capitalize()
 
 
 with gr.Blocks() as demo: