Update app.py

app.py

@@ -57,7 +57,7 @@ def respond(
             prompt += f"{role}{hmessage['content']}\n<|end|>"        
         prompt += f"<|assistant|>"
 
-        # """
+        """
         # Generate the response
         response_output = generator(
             prompt,
@@ -73,14 +73,37 @@ def respond(
     
         # Extract the assistant's response
         result = generated_text[len(prompt):].strip()
-        # """
+        """
+        tokens_cnt = 0
+        tokens_inc = 3
+        last_token_len = 1
+        while ( (tokens_cnt < max_tokens) and (last_token_len > 0) ):
+            # Generate the response
+            response_output = generator(
+                prompt,
+                generation_config=generator_conf,
+                max_new_tokens=tokens_inc,
+                do_sample=True,
+                top_p=top_p,
+                repetition_penalty=1.2,
+                temperature=temperature
+            )        
+            generated_text = response_output[0]['generated_text']
+            # Extract the assistant's response
+            result = generated_text[len(prompt):]
+            prompt = prompt + result
+            tokens_cnt = tokens_cnt + tokens_inc
+            last_token_len = len(result)
+            yield result
+        
         # result = prompt +':'+result
     except Exception as error:
         exc_type, exc_obj, exc_tb = sys.exc_info()
         fname = os.path.split(exc_tb.tb_frame.f_code.co_filename)[1]
         result = str(error) +':'+ exc_type +':'+ fname +':'+ exc_tb.tb_lineno
+        yield result
 
-    yield result
+    
 
 """
 For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface