Update app.py

app.py

@@ -1,12 +1,19 @@
 import os
+import logging
 from dotenv import load_dotenv
 from huggingface_hub import InferenceClient
 import gradio as gr
 
+# Enable logging for debugging
+logging.basicConfig(level=logging.INFO)
+
 # Load .env file
 load_dotenv()
 HF_TOKEN = os.getenv("HF_TOKEN")
 
+if not HF_TOKEN:
+    raise ValueError("HF_TOKEN not found in .env file!")
+
 # Initialize Hugging Face client
 MODEL = "google/flan-t5-small"
 client = InferenceClient(model=MODEL, token=HF_TOKEN)
@@ -16,18 +23,27 @@ def ask_question(question: str) -> str:
     if not question.strip():
         return "Please ask something."
     try:
-        # For FLAN-T5, we need to format prompt as a Q&A
         prompt = f"Question: {question}\nAnswer:"
         resp = client.text_generation(
             prompt=prompt,
             max_new_tokens=100,
             temperature=0.7,
         )
-        return resp  # Already returns generated text
+        logging.info(f"HF response: {resp}")  # Debug log
+
+        # Safely extract generated text
+        if isinstance(resp, dict) and "generated_text" in resp:
+            return resp["generated_text"]
+        elif isinstance(resp, list) and len(resp) > 0 and "generated_text" in resp[0]:
+            return resp[0]["generated_text"]
+        else:
+            # Fallback: convert response to string
+            return str(resp)
     except Exception as e:
+        logging.error(f"Error in ask_question: {str(e)}")
         return f"Error: {str(e)}"
 
-# Build Gradio Web UI
+# Gradio Web UI
 iface = gr.Interface(
     fn=ask_question,
     inputs=gr.Textbox(lines=2, placeholder="Ask me anything..."),
@@ -37,4 +53,5 @@ iface = gr.Interface(
 )
 
 if __name__ == "__main__":
-    iface.launch()
+    # Serve on all interfaces for Docker/remote access, enable debug logs
+    iface.launch(server_name="0.0.0.0", server_port=7860, debug=True)