update

app.py

@@ -22,10 +22,10 @@ coquiTTS = CoquiTTS()
 
 
 # Driver function
-def fun_engine(audio) : 
+def driver_fun(audio) : 
   text1, lang = whisper_stt(audio)
   #text1 = model.transcribe(audio)["text"]
-  text2 = lang_model_response(text1)
+  text2 = lang_model_response(text1, lang)
   speech = tts(text2, lang) #'en')
   return text1, text2, speech
 
@@ -55,16 +55,17 @@ def whisper_stt(audio):
 
 
 # LLM - Bloom Response  
-def lang_model_response(prompt): 
+def lang_model_response(prompt, language): 
   print(f"*****Inside lang_model_response - Prompt is :{prompt}")
   p = """Question: How are you doing today?
   Answer: I am doing good, thanks.
   Question: """
-  prompt = p + prompt + "\n" + "Answer: "
-  #prompt = prompt + "\n"
-  
   if len(prompt) == 0:
-    prompt = """Can you help me please?"""
+    prompt = """Question: Can you help me please?
+    Answer: Sure, I am here for you.
+    Question: """
+  
+  prompt = p + prompt + "\n" + "Answer: "
   
   json_ = {"inputs": prompt,
             "parameters":
@@ -80,27 +81,28 @@ def lang_model_response(prompt):
           "wait_for_model": True, 
           },}
   response = requests.post(API_URL, headers=headers, json=json_)
-  print(f"Response  is : {response}")
+  #print(f"Response  is : {response}")
   output = response.json()
-  #print(f"output is : {output}") 
   output_tmp = output[0]['generated_text']
-  print(f"output_tmp is: {output_tmp}")
+  print(f"Bloom API Response is : {output_tmp}")
   solution = output_tmp.split("Answer: ")[2].split("\n")[0]
-    
-  print(f"Final response after splits is: {solution}")
+  print(f"Final Bloom Response after splits is: {solution}")
   return solution
 
 # Coqui - Text-to-Speech
 def tts(text, language):
-    print(f"Inside tts - language is : {language}")
-    with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as fp:
-        coquiTTS.get_tts(text, fp, speaker = {"language" : language})
-        return fp.name
+  print(f"Inside tts - language is : {language}")
+  coqui_langs = ['en' ,'es' ,'fr' ,'de' ,'pl' ,'uk' ,'ro' ,'hu' ,'bg' ,'nl' ,'fi' ,'sl' ,'lv' ,'ga']
+  if language not in coqui_langs:
+    language = 'en'
+  with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as fp:
+      coquiTTS.get_tts(text, fp, speaker = {"language" : language})
+      return fp.name
 
   
 gr.Interface(
     title = 'Testing Whisper', 
-    fn=fun_engine, 
+    fn=driver_fun, 
     inputs=[
         gr.Audio(source="microphone",  type="filepath"), #streaming = True,
        # "state"