first commit

app.py

@@ -0,0 +1,18 @@
+import streamlit as st
+
+st.markdown("# Buscador de Noticias Salvadoreñas")
+
+search_text = st.text_input(label="Búsqueda")
+
+# ds = load_dataset("justinian336/salvadoran-news-embedded")
+
+
+# def search(text, model, ds, n):
+#   encoded_text = model.encode(text)
+#   scores, retrieved_examples = ds.get_nearest_examples('embedding', encoded_text, k=n)
+#   matching_titles = retrieved_examples["title"]
+#   urls = retrieved_examples["link"]
+#   contents = retrieved_examples["content"]
+#   return list(zip(matching_titles, [c[:150] for c in contents], urls, scores))
+
+# search_results = search(search_text, model, embedded_ds, 10)

requirements.txt

@@ -0,0 +1,3 @@
+datasets
+sentence-transformers
+faiss-cpu