app.py

import os
import typing as t

import gradio as gr
from huggingface_hub import InferenceClient


HF_API_TOKEN = os.getenv("HF_API_TOKEN")
if not HF_API_TOKEN:
    raise ValueError("Error: Hugging Face API token is not set. Please set the HF_API_TOKEN environment variable.")

MODEL_REPO_ID = "Qwen/Qwen2.5-72B-Instruct"


def format_messages(user_text: str, level: int) -> t.Mapping[str, str]:
    if level == 1:
        system_message = (
            "You are a text simplification tool for the Russian language. "
            "You are given a text in Russian and you need to give a simple version of it in Russian. "
            "Your task is to simplify complex sentences into simple ones. "
            "Avoid cramming too many details into one sentence; distribute them across multiple sentences where needed. "
            "Rephrase sentences to remove participial and gerundial constructions. "
            "Where possible, replace passive voice with active voice. "
            "If a sentence consists of only a noun, add a verb. "
            "Replace rare or low-frequency words with more common ones. "
            "Where appropriate, remove or replace foreign words. "
            "Clarify ambiguous phrases by replacing them with more concrete, easily understandable words. "
            "Where possible, avoid words that have paronyms. "
            "Use only Russian, English is forbidden at any cost."
        )
    elif level == 2:
        system_message = (
            "You are a text simplification tool for the Russian language. "
            "You are given a text in Russian and you need to give a simple version of it in Russian. "
            "Simplify complex or compound sentences by breaking them into shorter ones, aiming for a sentence length of no more than seven words. "
            "Ensure each sentence contains only one idea. "
            "Avoid participial and gerundial constructions, and prefer active voice over passive voice. "
            "Keep essential information like names, nationalities, and roles. Do not remove important details. "
            "Remove unnecessary foreign words (like brand names) and replace rare or long words with simpler, shorter ones. "
            "Simplify ambiguous phrases by using more concrete, clear words. "
            "Remove minor details that do not add significant meaning, but ensure key information remains intact. "
            "Use only Russian, English is forbidden at any cost."
        )
    elif level == 3:
        system_message = (
            "You are a text simplification assistant for the Russian language. "
            "You are given a text in Russian and you need to give a simple version of it in Russian. "
            "Your task is to make the text as simple as possible. "
            "Each sentence should contain only one idea and be no longer than five words. "
            "Remove or replace foreign words (such as names, places, or brands), and avoid minor details. "
            "Eliminate numbers and remove any unnecessary details. "
            "Focus on using the nominative and genitive cases for nouns, and only the present or past tense for verbs. "
            "Avoid passive voice and inverted word order. "
            "Replace rare or low-frequency words with more common ones. "
            "Where possible, replace complex phrases with common expressions, clichés, or idioms. "
            "Remove any extraneous details (if it is possible to without removing original semantics of the sentence) and simplify ambiguous phrases as much as possible."
            "Use only Russian, English is forbidden at any cost."
        )
    
    messages = [
        {"role": "system", "content": system_message},
        {"role": "user", "content": f"Simplify the following text in Russian: {user_text}"}
    ]
    return messages


def simplify_text(user_text: str, level: int) -> str:
    client = InferenceClient(model=MODEL_REPO_ID, token=HF_API_TOKEN)
    messages = format_messages(user_text, level)

    response = client.chat_completion(messages, max_tokens=512)
    if response is not None and response.choices is not None:
        content = response.choices[0].message.content
        return content
    else:
        return "Error: No valid response generated."



with gr.Blocks() as demo:
    gr.Markdown("# Text Simplification Tool for People with Aphasia")
    gr.Markdown("Simplify Russian text into easier-to-understand Russian based on the chosen level of simplification.")
    
    with gr.Row():
        user_text = gr.Textbox(label="Input Text (Russian)", lines=5, placeholder="Введите текст здесь...")
        level = gr.Radio(
            choices=[1, 2, 3],
            value=1,
            label="Simplification Level",
            info="Choose the level of simplification:\n1 = Basic, 2 = Intermediate, 3 = Advanced"
        )
    
    output_text = gr.Textbox(label="Simplified Text", lines=5, interactive=False)
    simplify_button = gr.Button("Simplify")
    
    simplify_button.click(simplify_text, inputs=[user_text, level], outputs=output_text)


demo.launch()