Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -1,87 +1,72 @@
|
|
1 |
-
import os
|
2 |
-
import warnings
|
3 |
-
import nest_asyncio
|
4 |
-
import streamlit as st
|
5 |
-
from dotenv import load_dotenv
|
6 |
-
from DataLoading.Data import get_data
|
7 |
-
from llama_index.core import Settings
|
8 |
-
from llama_index.llms.groq import Groq
|
9 |
-
from llama_index.vector_stores.faiss import FaissVectorStore
|
10 |
-
from llama_index.embeddings.huggingface import HuggingFaceEmbedding
|
11 |
-
from llama_index.core import StorageContext, load_index_from_storage
|
12 |
-
|
13 |
-
nest_asyncio.apply()
|
14 |
-
load_dotenv()
|
15 |
-
warnings.filterwarnings("ignore")
|
16 |
-
|
17 |
-
def init_llm(model_name):
|
18 |
-
return Groq(model=model_name, api_key=os.getenv("GROQ_API_KEY"))
|
19 |
-
|
20 |
-
@st.cache_resource
|
21 |
-
def load_index(selected_model):
|
22 |
-
curr_direc = os.getcwd()
|
23 |
-
file_path = os.path.join(curr_direc, 'processed_data.csv')
|
24 |
-
# print(file_path)
|
25 |
-
get_data(file_path)
|
26 |
-
model = init_llm(selected_model)
|
27 |
-
embedding_model = HuggingFaceEmbedding(model_name="BAAI/bge-small-en-v1.5")
|
28 |
-
|
29 |
-
Settings.embed_model = embedding_model
|
30 |
-
Settings.llm = model
|
31 |
-
|
32 |
-
vector_store = FaissVectorStore.from_persist_dir('storage')
|
33 |
-
storage_context = StorageContext.from_defaults(
|
34 |
-
vector_store=vector_store, persist_dir='storage'
|
35 |
-
)
|
36 |
-
index = load_index_from_storage(storage_context=storage_context)
|
37 |
-
return index.as_query_engine()
|
38 |
-
|
39 |
-
st.title("Chatbot from ClienterAI")
|
40 |
-
|
41 |
-
st.sidebar.header("Settings")
|
42 |
-
selected_model = st.sidebar.selectbox(
|
43 |
-
"Select Groq Model:",
|
44 |
-
options=["mixtral-8x7b-32768", "gemma2-9b-it", "llama-3.1-70b-versatile", "llama3-8b-8192", "llava-v1.5-7b-4096-preview"],
|
45 |
-
index=0
|
46 |
-
)
|
47 |
-
|
48 |
-
query_engine = load_index(selected_model)
|
49 |
-
|
50 |
-
if "messages" not in st.session_state:
|
51 |
-
st.session_state["messages"] = []
|
52 |
-
|
53 |
-
|
54 |
-
with st.form("chat_form", clear_on_submit=True):
|
55 |
-
user_input = st.text_input("Ask a question based on your data:", "")
|
56 |
-
submitted = st.form_submit_button("Send")
|
57 |
-
|
58 |
-
if submitted and user_input:
|
59 |
-
st.session_state["messages"].append({"role": "user", "content": user_input})
|
60 |
-
response = query_engine.query(user_input)
|
61 |
-
ai_response = response
|
62 |
-
st.session_state["messages"].append({"role": "assistant", "content": ai_response})
|
63 |
-
|
64 |
-
for message in st.session_state["messages"]:
|
65 |
-
if message["role"] == "user":
|
66 |
-
st.markdown(f"**You:** {message['content']}")
|
67 |
-
else:
|
68 |
-
st.markdown(f"**Assistant:** {message['content']}")
|
69 |
-
|
70 |
-
if st.sidebar.button("Clear Chat"):
|
71 |
-
st.session_state["messages"] = []
|
72 |
-
st.sidebar.success("Chat cleared!")
|
73 |
-
|
74 |
-
|
75 |
-
st.markdown("""
|
76 |
-
<style>
|
77 |
-
.stForm {
|
78 |
-
position: fixed;
|
79 |
-
align-self: center;
|
80 |
-
bottom: 0;
|
81 |
-
width: 50%;
|
82 |
-
left: 25%;
|
83 |
-
right: 50%;
|
84 |
-
padding: 10px;
|
85 |
-
}
|
86 |
-
<style>
|
87 |
-
""", unsafe_allow_html=True)
|
|
|
1 |
+
import os
|
2 |
+
import warnings
|
3 |
+
import nest_asyncio
|
4 |
+
import streamlit as st
|
5 |
+
from dotenv import load_dotenv
|
6 |
+
from DataLoading.Data import get_data
|
7 |
+
from llama_index.core import Settings
|
8 |
+
from llama_index.llms.groq import Groq
|
9 |
+
from llama_index.vector_stores.faiss import FaissVectorStore
|
10 |
+
from llama_index.embeddings.huggingface import HuggingFaceEmbedding
|
11 |
+
from llama_index.core import StorageContext, load_index_from_storage
|
12 |
+
|
13 |
+
nest_asyncio.apply()
|
14 |
+
load_dotenv()
|
15 |
+
warnings.filterwarnings("ignore")
|
16 |
+
|
17 |
+
def init_llm(model_name):
|
18 |
+
return Groq(model=model_name, api_key=os.getenv("GROQ_API_KEY"))
|
19 |
+
|
20 |
+
@st.cache_resource
|
21 |
+
def load_index(selected_model):
|
22 |
+
curr_direc = os.getcwd()
|
23 |
+
file_path = os.path.join(curr_direc, 'processed_data.csv')
|
24 |
+
# print(file_path)
|
25 |
+
get_data(file_path)
|
26 |
+
model = init_llm(selected_model)
|
27 |
+
embedding_model = HuggingFaceEmbedding(model_name="BAAI/bge-small-en-v1.5")
|
28 |
+
|
29 |
+
Settings.embed_model = embedding_model
|
30 |
+
Settings.llm = model
|
31 |
+
|
32 |
+
vector_store = FaissVectorStore.from_persist_dir('storage')
|
33 |
+
storage_context = StorageContext.from_defaults(
|
34 |
+
vector_store=vector_store, persist_dir='storage'
|
35 |
+
)
|
36 |
+
index = load_index_from_storage(storage_context=storage_context)
|
37 |
+
return index.as_query_engine()
|
38 |
+
|
39 |
+
st.title("Chatbot from ClienterAI")
|
40 |
+
|
41 |
+
st.sidebar.header("Settings")
|
42 |
+
selected_model = st.sidebar.selectbox(
|
43 |
+
"Select Groq Model:",
|
44 |
+
options=["mixtral-8x7b-32768", "gemma2-9b-it", "llama-3.1-70b-versatile", "llama3-8b-8192", "llava-v1.5-7b-4096-preview"],
|
45 |
+
index=0
|
46 |
+
)
|
47 |
+
|
48 |
+
query_engine = load_index(selected_model)
|
49 |
+
|
50 |
+
if "messages" not in st.session_state:
|
51 |
+
st.session_state["messages"] = []
|
52 |
+
|
53 |
+
|
54 |
+
with st.form("chat_form", clear_on_submit=True):
|
55 |
+
user_input = st.text_input("Ask a question based on your data:", "")
|
56 |
+
submitted = st.form_submit_button("Send")
|
57 |
+
|
58 |
+
if submitted and user_input:
|
59 |
+
st.session_state["messages"].append({"role": "user", "content": user_input})
|
60 |
+
response = query_engine.query(user_input)
|
61 |
+
ai_response = response
|
62 |
+
st.session_state["messages"].append({"role": "assistant", "content": ai_response})
|
63 |
+
|
64 |
+
for message in st.session_state["messages"]:
|
65 |
+
if message["role"] == "user":
|
66 |
+
st.markdown(f"**You:** {message['content']}")
|
67 |
+
else:
|
68 |
+
st.markdown(f"**Assistant:** {message['content']}")
|
69 |
+
|
70 |
+
if st.sidebar.button("Clear Chat"):
|
71 |
+
st.session_state["messages"] = []
|
72 |
+
st.sidebar.success("Chat cleared!")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|