Update app.py

app.py

@@ -64,6 +64,7 @@ def run_rnn(ctx, state):
 
 def generate_response(message, history, temperature=1.0, top_p=0.3):
     global model_tokens, model_state
+    model_state = None
     
     ctx = ""
     for human, assistant in history:
@@ -118,10 +119,6 @@ def chat_with_bot(message, history, temperature, top_p):
 with gr.Blocks(title="MiniRWKV_7 34.2M 🪿 2vGPU Space") as demo:
     gr.Markdown("# MiniRWKV_7 34.2M 🪿 ")
     gr.Markdown("### Only 34.2M Params!!! Use 2V CPU Backend to run this model. ")
-    @demo.load()
-    def on_load():
-        global model_state
-        model_state = None
     
     with gr.Row():
         with gr.Column(scale=3):