import gradio as gr
from transformers import AutoTokenizer, AutoModelForMaskedLM
import torch


model_name = "yangheng/PlantRNA-FM"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForMaskedLM.from_pretrained(model_name)


def predict_rna(sequence):
    
    inputs = tokenizer(sequence, return_tensors="pt")
    mask_token_index = torch.where(inputs.input_ids == tokenizer.mask_token_id)[1]  # 找到 <mask> 的位置

    
    with torch.no_grad():
        outputs = model(**inputs)

    
    mask_token_logits = outputs.logits[0, mask_token_index, :]
    predicted_token_ids = torch.argmax(mask_token_logits, dim=-1)
    predicted_tokens = tokenizer.convert_ids_to_tokens(predicted_token_ids)

    
    return " ".join(predicted_tokens)


input_text = gr.Textbox(lines=2, placeholder="Input RNA Sequence with <mask>, e.g., AAAGAGTCATATACGATATTGTCGACCGTGG<mask>AGAGAGAAGAATGTACGATTGGAGT")
output_text = gr.Textbox()

app = gr.Interface(
    fn=predict_rna,
    inputs=input_text,
    outputs=output_text,
    title="Zero-shot PlantFM-RNA MNM Inference",
    description="Zero-shot PlantFM-RNA MNM Inference: Predicts only the <mask> tokens."
)

if __name__ == "__main__":
    app.launch()