Upload app.py

app.py

@@ -3,6 +3,7 @@ import streamlit_scrollable_textbox as stx
 from ast import literal_eval
 import pinecone
 import streamlit as st
+import json
 
 st.set_page_config(layout="wide")  # isort: split
 
@@ -89,17 +90,21 @@ sparse_scores = np.argsort(bm25.get_scores(tokenized_query), axis=0)[::-1]
 indices = get_bm25_search_hits(corpus, sparse_scores, 50)
 
 
-dense_embedding_output = instructor_model.predict(
+json_output_embedding = instructor_model.predict(
     query_embedding_instruction,
     query_text,
     api_name="/predict",
 )
 
-dense_embedding_arr = np.array(
-    literal_eval(dense_embedding_output)[0], dtype=np.float64
-)
-dense_embedding = dense_embedding_arr.tolist()
-print(type(dense_embedding[2]))
+json_file = open(json_output_embedding, "r")
+json_dict = json.load(json_file)
+dense_embedding = json_dict["data"]
+
+#dense_embedding_arr = np.array(
+#    literal_eval(dense_embedding_output)[0], dtype=np.float64
+#)
+#dense_embedding = dense_embedding_arr.tolist()
+#print(type(dense_embedding[2]))
 # dense_embedding = [float(x) for x in dense_embedding_output[0]]
 
 text_embedding_instructions_choice = [
@@ -151,13 +156,10 @@ with col2:
                     unsafe_allow_html=True,
                 )
 
-tab1 = st.tabs(["View transcript"])
-
 
-with tab1:
-    file_text = retrieve_transcript()
-    with st.expander("See Transcript"):
-        st.subheader("AMD Q1 2020 Earnings Call Transcript:")
-        stx.scrollableTextbox(
-            file_text, height=700, border=False, fontFamily="Helvetica"
-        )
+file_text = retrieve_transcript()
+with st.expander("See Transcript"):
+    st.subheader("AMD Q1 2020 Earnings Call Transcript:")
+    stx.scrollableTextbox(
+        file_text, height=700, border=False, fontFamily="Helvetica"
+    )