Update app.py

app.py

@@ -47,18 +47,20 @@ def generate(query: str, temperature=0.9, max_new_tokens=256, top_p=0.95, repeti
         seed=42,
     )
 
-    stream = client.generate_stream(prompt, **generate_kwargs)
-    output = ""
-
-    previous_token = ""
-    for response in stream:
-        if response.token.text == "<|endoftext|>":
-            return output
-        else:
-            output += response.token.text
-        previous_token = response.token.text
-        yield output
-    return output
+    try:
+        stream = client.generate_stream(prompt, **generate_kwargs)
+        output = ""
+        previous_token = ""
+        for response in stream:
+            if response.token.text == "<|endoftext|>":
+                return output
+            else:
+                output += response.token.text
+            previous_token = response.token.text
+            yield output
+        return output
+    except errors.UnknownError:
+        return "Please wait for a while, The OctoCoder model is currently loading... 🐙"
 
 
 def process_example(**krwags):