filter

app.py

@@ -71,8 +71,12 @@ def respond(
     # text_streamer = TextStreamer(tokenizer, skip_prompt = True)
     # model.generate(input_ids = inputs, streamer = text_streamer, max_new_tokens = max_tokens,
     #                 use_cache = True, temperature = temperature, min_p = top_p)
+    response = tokenizer.batch_decode(outputs, skip_special_tokens=True)[0]
+
+    if "assistant" in response:
+        response = response.split("assistant")[-1].strip()
     
-    yield tokenizer.batch_decode(outputs, skip_special_tokens = True)
+    print(response)
 
 
 """