File size: 1,290 Bytes
2069ee0 925a98e 2069ee0 b83f839 2069ee0 b6ea566 2069ee0 b83f839 2069ee0 4931548 054992f 4931548 b83f839 9a9d27e 4931548 9a9d27e 2069ee0 b83f839 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 |
import gradio as gr
import duckdb
from langchain_community.embeddings import HuggingFaceEmbeddings
from langchain_community.vectorstores import DuckDB
conn = duckdb.connect('your_database.duckdb')
embedding_function = HuggingFaceEmbeddings()
vector_store = DuckDB(
connection=conn,
embedding=embedding_function)
# Define a data structure for user data
class User:
def __init__(self, phone: str, features: str):
self.phone = phone
self.features = features
def create(self):
vector_store.add_texts([f'#features\n{self.features}\n\n#phone\n{self.phone}'])
def search(self):
return vector_store.similarity_search(self.features, k=1)[0].page_content
def greet(a, b, c):
u = User(a, b)
if c:
return u.search()
u.create()
return 1
# demo = gr.Interface(fn=greet, inputs=["textbox", "button"], outputs="text")
demo = gr.Blocks()
with demo:
p = gr.Textbox()
f = gr.Textbox()
with gr.Row():
btn = gr.Button("insert")
btn2 = gr.Button("query")
n = gr.Textbox()
btn.click(greet, inputs=[p, f, gr.Number(value=0, visible=False)], outputs=[n])
btn2.click(greet, inputs=[p, f, gr.Number(value=1, visible=False)], outputs=[n])
# demo.launch()
demo.launch(auth=("username", "password")) |