Update app.py

app.py

@@ -2,18 +2,21 @@ import time
 import gradio as gr
 from ctransformers import AutoModelForCausalLM
 
-# Set gpu_layers to the number of layers to offload to GPU. Set to 0 if no GPU acceleration is available on your system.
 llm = AutoModelForCausalLM.from_pretrained("TheBloke/WizardLM-7B-uncensored-GGUF", model_file="WizardLM-7B-uncensored.Q3_K_M.gguf", model_type="llama", stream=True)
 history = ["Chatbot:"]
 
 def generate_response(message):
     global history
+    token_count = 0
     for text in llm(message):
         response = ''.join(text)
         history.append(response)
-        yield ' '.join(history)
-    # Clear the history list after the last response
-    history = ["Chatbot:"]
+        token_count += len(response.split())
+        if token_count >= 2:
+            yield ' '.join(history)
+            token_count = 0
+            # Clear the history list after generating a response
+            history = ["Chatbot:"]
 
 def chatbot(message, history):
     response_generator = generate_response(message)