Update app.py

app.py

@@ -7,16 +7,15 @@ import streamlit as st
 import torch
 
 # Load the BillSum dataset
-
-
 ds = load_dataset("FiscalNote/billsum")
+
 # Initialize models
 sbert_model = SentenceTransformer("all-mpnet-base-v2")
 t5_tokenizer = AutoTokenizer.from_pretrained("t5-small")
-t5_model = AutoModelForSeq2SeqLM.from_pretrained("t5-small")
+t5_model = AutoAutoModelForSeq2SeqLM.from_pretrained("t5-small")
 
 # Prepare data and build FAISS index
-texts = dataset["text"][:100]  # Limiting to 100 samples for speed
+texts = ds["train"]["text"][:100]  # Limiting to 100 samples for speed, and selecting the train split.
 case_embeddings = sbert_model.encode(texts, convert_to_tensor=True, show_progress_bar=True)
 
 # Convert embeddings to numpy array and handle deprecation warning