Update app.py

app.py

@@ -59,7 +59,7 @@ def index_docs(documents):
 def retrieve_docs(query):
     return vector_store.similarity_search(query)
 
-# Function to generate an answer based on retrieved documents
+# Function to generate an answer based on retrieved documents using text generation
 def answer_question(question, documents):
     context = "\n\n".join([doc.page_content for doc in documents])
     full_context = f"{context}"
@@ -68,9 +68,14 @@ def answer_question(question, documents):
     # Format the prompt with the user's question and context
     question_with_context = prompt.format(question=question, context=full_context)
     
-    # Use the Hugging Face InferenceClient to generate the response
-    # Replacing the `query` method with `predict` method
-    response = client.predict(question_with_context)
+    # Use the Hugging Face InferenceClient's text_generation method
+    generate_kwargs = {
+        "temperature": 0.7,  # Control the creativity of the generated response
+        "max_new_tokens": 150,  # Limit the length of the output
+        "top_p": 0.9  # Control diversity via nucleus sampling
+    }
+    
+    response = client.text_generation(question_with_context, **generate_kwargs)
     
     # Assuming the response contains a 'generated_text' field with the model's output
     return response["generated_text"]