import streamlit as st
from transformers import AutoProcessor, AutoModelForCausalLM
from PIL import Image
import torch

# 设置模型和处理器
model_id = "OpenFace-CQUPT/Human_LLaVA"
processor = AutoProcessor.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id).to("cuda" if torch.cuda.is_available() else "cpu")

# Streamlit 界面设置
st.title("Visual Question Answering App")
st.write("Upload an image and ask a question about it!")

# 图片上传
uploaded_image = st.file_uploader("Choose an image...", type=["jpg", "jpeg", "png"])
question = st.text_input("Ask a question about the image:")

# 处理输入并获取答案
if uploaded_image is not None and question:
    image = Image.open(uploaded_image)

    # 显示图片和问题
    st.image(image, caption="Uploaded Image", use_column_width=True)
    st.write("Question:", question)

    # 使用模型生成答案
    with st.spinner("Generating answer..."):
        inputs = processor(images=image, text=question, return_tensors="pt").to("cuda" if torch.cuda.is_available() else "cpu")
        with torch.no_grad():
            output = model.generate(**inputs)
        answer = processor.decode(output[0], skip_special_tokens=True)

    # 显示答案
    st.write("Answer:", answer)