CODEFIX: Switched model and added more debug info to figure out why it doesn't work for embedded versions.

app.py

@@ -81,8 +81,9 @@ def handle_submit():
 
         # List of models to try in order
         models = [
+            #"mistralai/Mistral-7B-Instruct-v0.2", #not enough tokens. 32k while 35k is passed. But else works.
             "Qwen/Qwen2.5-72B-Instruct", #Will not use by default, as using a secondary model spends two requests. Will rather update the code if needed.
-            "Qwen/Qwen2.5-Coder-32B-Instruct",  # Primary model
+            #"Qwen/Qwen2.5-Coder-32B-Instruct",  # Primary model
         ]
 
         for model in models:
@@ -106,7 +107,8 @@ def handle_submit():
             except Exception as e:
                 if "422" in str(e):  # Service Unavailable error
                     if model == models[-1]:  # If this was the last model to try
-                        error_message = "All models are currently unavailable. Please try again later."
+                        error_message = f"422 An error occurred: {str(e)}"
+                        #error_message = "All models are currently unavailable. Please try again later."
                         placeholder.markdown(f'<div class="scrollable-div"><p>{error_message}</p></div>', unsafe_allow_html=True)
                     continue  # Try the next model
                 else: