update

app.py

@@ -173,12 +173,11 @@ if st.button("Run Benchmark"):
 
             # Extract billed input and output tokens from meta
             group["billed_input_tokens"] = group["meta"].apply(
-                lambda x: eval(x).get("billed_units", {}).get("input_tokens", "N/A")
+                lambda x: x.get("billed_units", {}).get("input_tokens", "N/A")
             )
             group["billed_output_tokens"] = group["meta"].apply(
-                lambda x: eval(x).get("billed_units", {}).get("output_tokens", "N/A")
+                lambda x: x.get("billed_units", {}).get("output_tokens", "N/A")
             )
-
             # Display the runs for each query in a table
             st.write(
                 group[