Update app.py

app.py

@@ -1,8 +1,35 @@
+from langchain.text_splitter import CharacterTextSplitter
+from langchain_community.document_loaders import TextLoader
+from langchain_community.document_loaders import DirectoryLoader
+from langchain_community.embeddings.sentence_transformer import SentenceTransformerEmbeddings
+from langchain_community.vectorstores import Chroma
 import streamlit as st
-from transformers import pipeline
-pipe = pipeline('sentiment-analysis')
+
+text_loader_kwargs={'autodetect_encoding': True}
+loader = DirectoryLoader("src_info", glob="./*.txt", loader_cls=TextLoader, loader_kwargs=text_loader_kwargs)
+docs = loader.load()
+
+# split it into chunks
+#text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
+#docs = text_splitter.split_documents(documents)
+
+# create the open-source embedding function
+#embedding_function = SentenceTransformerEmbeddings(model_name="all-MiniLM-L6-v2")
+embedding_function = SentenceTransformerEmbeddings(model_name="all-mpnet-base-v2")
+
+# load it into Chroma
+chdb = Chroma.from_documents(docs, embedding_function, collection_metadata={"hnsw:space": "cosine"}, persist_directory='chroma_db_info')
+
+
 text = st.text_area("enter text")
 if text:
-    out = pipe(text)
-    st.json(out)
+    docs = chdb.similarity_search_with_score(query, k=3)
+    docnum = len(docs)
+    index = 0
+    ret = ''
+    for ii in range(docnum):    
+        doc = docs[ii][0]
+        score = docs[ii][1]
+        ret += f"Return {index} ({score:.4f}) :\n{doc.page_content}\n"
+    st.ret