Spaces:
Running
Running
File size: 1,518 Bytes
c2d8087 777ca73 c2d8087 777ca73 c2d8087 777ca73 c2d8087 777ca73 c2d8087 777ca73 c2d8087 777ca73 c2d8087 777ca73 c2d8087 777ca73 c2d8087 |
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 43 44 45 46 |
import streamlit as st
from transformers import AutoTokenizer, AutoModelForCausalLM
from huggingface_hub import login
import torch
import os
# Set model map
model_map = {
"FinGPT": "AI4Finance/FinGPT",
"FinanceConnect": "ceadar-ie/FinanceConnect-13B",
"Sujet-Finance": "sujet-ai/Sujet-Finance-8B-v0.1"
}
# Cache model loading for performance
@st.cache_resource
def load_model_and_tokenizer(model_id):
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
model_id,
torch_dtype=torch.float16 if torch.cuda.is_available() else torch.float32
)
return model, tokenizer
# Query model
def query_model(model_id, question):
model, tokenizer = load_model_and_tokenizer(model_id)
inputs = tokenizer(question, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=150)
return tokenizer.decode(outputs[0], skip_special_tokens=True)
# Streamlit app layout
st.title("💼 Financial LLM Evaluation Interface")
model_choice = st.selectbox("Select a Financial Model", list(model_map.keys()))
user_question = st.text_area("Enter your financial question:", "What is EBITDA?")
if st.button("Get Response"):
with st.spinner("Generating response..."):
try:
answer = query_model(model_map[model_choice], user_question)
st.subheader(f"Response from {model_choice}:")
st.write(answer)
except Exception as e:
st.error(f"Something went wrong: {e}")
|