app.py

import streamlit as st
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer, pipeline
from diffusers import StableDiffusionPipeline
import torch

@st.cache_resource
def load_all_models():
    # Load IndicTrans2 Tamil-to-English model
    trans_model_id = "ai4bharat/indictrans2-indic-en-dist-200M"
    trans_tokenizer = AutoTokenizer.from_pretrained(trans_model_id, trust_remote_code=True)
    trans_model = AutoModelForSeq2SeqLM.from_pretrained(trans_model_id, trust_remote_code=True)

    # Load English text generation model (you can use GPT2 or Falcon, etc.)
    text_gen = pipeline("text-generation", model="gpt2")

    # Load Stable Diffusion for image generation
    img_pipe = StableDiffusionPipeline.from_pretrained(
        "runwayml/stable-diffusion-v1-5",
        torch_dtype=torch.float16,
        revision="fp16"
    ).to("cuda" if torch.cuda.is_available() else "cpu")

    return trans_tokenizer, trans_model, text_gen, img_pipe

def translate_text(text, tokenizer, model):
    input_text = f"translate Tamil to English: {text}"
    inputs = tokenizer(input_text, return_tensors="pt", padding=True)
    outputs = model.generate(**inputs, max_length=128)
    return tokenizer.decode(outputs[0], skip_special_tokens=True)

def main():
    st.title("Multimodal Tamil to Image Generator 🚀")
    st.markdown("Enter Tamil text, we translate it to English, continue the sentence, and generate an image!")

    user_input = st.text_area("Enter Tamil text:", "")

    if st.button("Generate"):
        with st.spinner("Loading models..."):
            tokenizer, model, text_gen, img_pipe = load_all_models()

        with st.spinner("Translating to English..."):
            english_text = translate_text(user_input, tokenizer, model)
            st.subheader("Translated English:")
            st.write(english_text)

        with st.spinner("Generating continuation..."):
            continuation = text_gen(english_text, max_length=50, do_sample=True)[0]['generated_text']
            st.subheader("Generated Text:")
            st.write(continuation)

        with st.spinner("Generating Image..."):
            image = img_pipe(continuation).images[0]
            st.subheader("Generated Image:")
            st.image(image)

if __name__ == "__main__":
    main()