Upload app.py

app.py

@@ -7,6 +7,7 @@ from langchain.vectorstores import Chroma
 import openai
 import streamlit as st
 import gradio as gr
+from gradio.components import Textbox, Slider
 
 openai.api_key = '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'
 
@@ -73,12 +74,17 @@ def save_in_DB(splitted_text):
     return db
 
 
-def query(query_text):
+def query(query_text, num_docs):
     st.title('RAG system')
 
     # query_text = st.text_input("Enter your question", "Cynthia W. Harris is a citizen of which state?", key="question")
-    docs = db.similarity_search(query_text)
+    docs = db.similarity_search(query_text, k=num_docs)
     print("len(docs)", len(docs))
+    # print each docs .page_content with klar abgrenzen
+    for doc in docs:
+        print("doc", doc.page_content)
+        print()
+        print()
 
     # Store the first 10 results as context
     context = '\n\n'.join([doc.page_content for doc in docs[:5]])
@@ -102,7 +108,7 @@ def query(query_text):
     # Return the generated answer
     st.subheader("Answer:")
     st.write(predicted)
-    return predicted, context
+    return predicted
 
 
 
@@ -116,7 +122,14 @@ def run():
     db = save_in_DB(splitted_text)
     print("type db", type(db))
 
-    demo = gr.Interface(fn=query, inputs="text", outputs=["text", "text"])
+    demo = gr.Interface(
+        fn=query,
+        inputs=[
+            Textbox(lines=1, placeholder="Type your question here...", label="Question"),
+            Slider(minimum=1, maximum=20, default=4, step=1, label="Number of Documents in Context")
+            ],
+        outputs="text"
+        )
 
     demo.launch()
     # query(db)