Update app.py

app.py

@@ -3,8 +3,11 @@ from transformers import RagTokenizer, RagRetriever, RagSequenceForGeneration
 from datasets import load_dataset
 from transformers import XLMRobertaTokenizer, XLMRobertaForSequenceClassification
 
-# Load the XNLI dataset (Multilingual NLI dataset) for demonstration
-dataset = load_dataset("xnli", split="validation")
+# Load a multilingual dataset (use "xnli" or "tydi_qa")
+try:
+    dataset = load_dataset("xnli", "en", split="validation")  # Using English subset as an example
+except Exception as e:
+    st.error(f"Error loading the dataset: {e}")
 
 # Initialize tokenizer and retriever for multilingual support (using XLM-Roberta)
 tokenizer = XLMRobertaTokenizer.from_pretrained("xlm-roberta-base")
@@ -37,6 +40,3 @@ if user_query:
     st.write(f"Answer: {answer}")
 
     # Display the most relevant documents
-    st.subheader("Relevant Documents:")
-    for doc in retrieved_docs:
-        st.write(doc['text'][:300] + '...')  # Display first 300 characters of each doc