a

main.py

@@ -6,6 +6,7 @@ import streamlit as st
 from pathlib import Path
 import tempfile
 import uuid
+from pinecone import query_text
 
 load_dotenv()
 openai_api_key = os.environ["OPENAI_API_KEY"]
@@ -18,7 +19,14 @@ if "messages" not in st.session_state:
     st.session_state.messages = []
 
 def response(user_input):
-    messages = [{"role": "system", "content": "あなたは安田章紀(やすだあきのり)博士です。京都大学に所属していて、チベット仏教を専門にしています。日本語で答えてください。"},]
+    pinecone_results = query_text(user_input, top_k=3)
+    contexts = [m['metadata']['text'] for m in pinecone_results['matches']]
+    context_text = "\n".join(contexts)
+    
+    messages = [{
+        "role": "system", 
+        "content": f"""あなたは安田章紀(やすだあきのり)博士です。京都大学に所属していて、チベット仏教を専門にしています。生前の情報:{context_text}この情報を参考に日本語で答えてください。"""},]
+    
     for msg in st.session_state.messages:
         messages.append({"role": msg["role"], "content": msg["content"]})
 

pinecone.py

@@ -0,0 +1,29 @@
+import pinecone
+import os
+from openai import OpenAI
+
+pinecone.init(api_key=os.environ["PINECONE_API_KEY"])
+index = pinecone.Index("your-index")
+
+client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])
+
+def upsert_texts(texts, ids):
+    embeddings = []
+    for text in texts:
+        response = client.embeddings.create(
+            model="text-embedding-3-large",
+            input=text
+        )
+        embeddings.append(response.data[0].embedding)
+
+    vectors = [(id, emb) for id, emb in zip(ids, embeddings)]
+    index.upsert(vectors)
+
+def query_text(query, top_k=5):
+    response = client.embeddings.create(
+        model="text-embedding-3-large",
+        input=query
+    )
+    embedding = response.data[0].embedding
+    results = index.query(vector=embedding, top_k=top_k, include_metadata=True)
+    return results

requirements.txt

@@ -2,4 +2,5 @@ streamlit
 openai
 requests
 python-dotenv
-
+pinecone-client
+elevenlabs