Upload app.py with huggingface_hub

app.py

@@ -205,7 +205,7 @@ st.warning(
 with st.sidebar:
     st.header("About")
     st.markdown(
-        "Compares **Qwen 3 32B** and **Llama 4 Scout** for detecting errors "
+        "Compares **DeepSeek Reasoner** and **GPT-OSS-120B** for detecting errors "
         "in discharge letters."
     )
     st.markdown("---")
@@ -344,7 +344,7 @@ if st.session_state.result:
     with col_a:
         st.markdown(
             '<div class="model-header-a"><h4 style="color:#3182ce; margin:0">'
-            "Qwen 3 32B</h4></div>",
+            "DeepSeek Reasoner</h4></div>",
             unsafe_allow_html=True,
         )
         render_model_output(r.model_a_result, "model-header-a")
@@ -352,7 +352,7 @@ if st.session_state.result:
     with col_b:
         st.markdown(
             '<div class="model-header-b"><h4 style="color:#805ad5; margin:0">'
-            "Llama 4 Scout</h4></div>",
+            "GPT-OSS-120B</h4></div>",
             unsafe_allow_html=True,
         )
         render_model_output(r.model_b_result, "model-header-b")
@@ -374,8 +374,8 @@ if st.session_state.result:
     feedback_data = {}
 
     for model_key, model_label, res in [
-        ("model_a", "Qwen 3 32B", r.model_a_result),
-        ("model_b", "Llama 4 Scout", r.model_b_result),
+        ("model_a", "DeepSeek Reasoner", r.model_a_result),
+        ("model_b", "GPT-OSS-120B", r.model_b_result),
     ]:
         st.markdown(f"#### {model_label}")