Update app_chat.py

app_chat.py

@@ -92,10 +92,14 @@ def generate(
     t.start()
 
     outputs = []
-    for text in streamer:
-        outputs.append(text)
-        yield "".join(outputs)
-
+    try:
+        for text in streamer:
+            outputs.append(text)
+            yield "".join(outputs)
+    except:
+        print("restarting the model, got some error")
+        model = AutoModelForCausalLM.from_pretrained(model_id, device_map="cuda", trust_remote_code=True)
+        model = model.cuda().to(torch.bfloat16)
 
 chat_interface = gr.ChatInterface(
     fn=generate,