added Disclaimer to Description about app's limits in practical use

app.py

@@ -45,7 +45,7 @@ def tweet_detect(text):
 gradio_ui = gr.Interface(
     fn=tweet_detect,
     title="Detect Fake Tweets",
-    description="Enter a tweet and see if the transformer model can identify if it was written by state-backed trolls.",
+    description="Enter a tweet and see if the transformer model can identify if it was written by state-backed trolls. DISCLAIMER: While the model was fine tuned on 100k real and troll tweets, and achieved high accuracy in my tests, its performance drops significantly against the day-to-day barrage of content on Twitter. As such, this app is intended as an example for understanding the limits of AI/ML in highly complex problems like fake media detection, and not as a final arbiter of whether someone's tweet is real or not.",
     inputs=gr.inputs.Textbox(lines=10, label="Paste tweet text here [English Only]"),
     outputs=gr.outputs.Label(type="auto", label="Prediction"),
     interpretation="shap",