Change

app.py

@@ -32,6 +32,7 @@ def update_inputs(input_style):
 
 
 async def lynx(input_style_dropdown,document_input,question_input,answer_input):
+    start_time = time.time()
     if input_style_dropdown=='QA format':
         client = AsyncOpenAI(
         base_url="https://s6mipt5j797e6fql.us-east-1.aws.endpoints.huggingface.cloud/v1/", 
@@ -59,13 +60,19 @@ Your output should be in JSON FORMAT with the keys "REASONING" and "SCORE":
             frequency_penalty=None,
             presence_penalty=None
         )
-        print(chat_completion)
-        return chat_completion.choices.pop().message.content
+        message = chat_completion.choices.pop().message.content
+        message_new = message[len(message)-6:len(message)]
+        if 'FAIL' in message_new:
+            results = "🟥"
+        else:
+            results = "🟩"
     else:
-        return 'NA'
-
+        results = 'NA'
+    lynx_time = round(time.time() - start_time, 2)  # Calculate time taken for Lynx
+    return results, lynx_time
 # Function to judge reliability based on the selected input format
 async def judge_reliability(input_style, document, conversation, claim, question, answer):
+    start_time = time.time() 
     if input_style == "Dialog":
         conversation = json.loads(conversation)
         print(conversation)
@@ -74,8 +81,13 @@ async def judge_reliability(input_style, document, conversation, claim, question
         outputs = await collinear.judge.veritas.natural_language_inference(document,claim)
     elif input_style == "QA format":
         outputs = await collinear.judge.veritas.question_answer(document,question,answer)
-    results = f"Reliability Judge Outputs: {outputs}"
-    return results
+    output = outputs.judgement
+    if output ==1:
+        results = "🟩"
+    else:
+        results = "🟥"
+    veritas_time = round(time.time() - start_time, 2)  # Calculate time taken for Veritas
+    return result, veritas_time
 
 
 
@@ -115,9 +127,11 @@ Raises Full Year FY25 Operating Cash Flow Growth Guidance to 23% to 25% Y/Y""")
         answer_input = gr.Textbox(label="Answer", lines=5, visible=False, value="Salesforce revenue guidance for next year is about $37.8Bn ")
 
     with gr.Row():
-        result_output = gr.Textbox(label="Veritas Model")
+        result_output = gr.Textbox(label="Veritas Model Result")
+        veritas_time_output = gr.Textbox(label="Veritas Model Time (seconds)")
 
-        lynx_output = gr.Textbox(label="Lynx Model")
+        lynx_output = gr.Textbox(label="Lynx Model Result")
+        lynx_time_output = gr.Textbox(label="Lynx Model Time (seconds)")
 
 
     # Set the visibility of inputs based on the selected input style
@@ -131,11 +145,11 @@ Raises Full Year FY25 Operating Cash Flow Growth Guidance to 23% to 25% Y/Y""")
     gr.Button("Submit").click(
         fn=judge_reliability, 
         inputs=[input_style_dropdown, document_input, conversation_input, claim_input, question_input, answer_input], 
-        outputs=result_output
+        outputs=[result_output,veritas_time_output]
     ).then(
         fn=lynx,
         inputs=[input_style_dropdown,document_input,question_input,answer_input],
-        outputs=lynx_output
+        outputs=[lynx_output, lynx_time_output]
     )
 
 # Launch the demo