Update app.py

app.py

@@ -79,6 +79,40 @@ def predict(message, history):
             yield partial_message
 
 
+def predict2(message, history):
+
+    history_transformer_format = history + [[message, ""]]
+    stop = StopOnTokens()
+
+    messages = "".join(["".join(["\nuser :"+item[0], "\nbot:"+item[1]])  # curr_system_message +
+                        for item in history_transformer_format])
+
+    print(messages)
+
+    model_inputs = tokenizer([messages], return_tensors="pt")
+    streamer = TextIteratorStreamer(
+        tokenizer, timeout=10., skip_prompt=True, skip_special_tokens=True)
+    generate_kwargs = dict(
+        model_inputs,
+        streamer=streamer,
+        max_new_tokens=1024,
+        do_sample=True,
+        top_p=0.95,
+        top_k=1000,
+        temperature=0.5,
+        num_beams=1,
+        stopping_criteria=StoppingCriteriaList([stop])
+    )
+    t = Thread(target=model.generate, kwargs=generate_kwargs)
+    t.start()
+
+    partial_message = ""
+    for new_token in streamer:
+        if new_token != '<':
+            partial_message += new_token
+            yield partial_message
+
+
 def generate_text(prompt, tokenizer, model):
     # Tokenize the input
     input_ids = tokenizer.encode(prompt, return_tensors="pt")
@@ -127,7 +161,7 @@ with gr.Blocks(css=css) as demo:
     # with gr.Tab("E-Commerce"):
     #     e_commerce_interface()
     with gr.Tab("OpenBB"):
-        gr.ChatInterface(predict, examples=[
+        gr.ChatInterface(predict2, examples=[
                          examples_openbb[0], examples_openbb[1], examples_openbb[2], examples_openbb[3]])