app.py

import gradio as gr
from transformers import pipeline
import torch
import warnings
import numpy as np
from PIL import Image
warnings.filterwarnings('ignore')

device = "cuda" if torch.cuda.is_available() else "cpu"

# Dicionário expandido de modelos
models = {
    'transcription': pipeline("automatic-speech-recognition", model="openai/whisper-small", device=device),
    'translation': pipeline("translation", model="facebook/mbart-large-50-many-to-many-mmt", device=device),
    'summarization': pipeline("summarization", model="facebook/bart-large-cnn", device=device),
    'sentiment': pipeline("sentiment-analysis", model="nlptown/bert-base-multilingual-uncased-sentiment", device=device),
    'question_answering': pipeline("question-answering", model="deepset/roberta-base-squad2", device=device),
    'chat': pipeline("text-generation", model="facebook/opt-125m", device=device),
    'image_caption': pipeline("image-to-text", model="Salesforce/blip-image-captioning-base", device=device),
    'text_to_speech': pipeline("text-to-audio", model="facebook/mms-tts-eng", device=device),
    'zero_shot': pipeline("zero-shot-classification", model="facebook/bart-large-mnli", device=device),
    'ner': pipeline("token-classification", model="dslim/bert-base-NER", device=device)
}

def process_transcription(audio):
    if not audio:
        return "Por favor, forneça um arquivo de áudio."
    return models['transcription'](audio)["text"]

def process_translation(text, direction):
    if not text:
        return "Por favor, forneça um texto para tradução."
    return models['translation'](text, 
        src_lang="pt" if direction=="pt_en" else "en", 
        tgt_lang="en" if direction=="pt_en" else "pt")[0]['translation_text']

def process_summarization(text):
    if not text:
        return "Por favor, forneça um texto para resumir."
    return models['summarization'](text, max_length=130, min_length=30)[0]['summary_text']

def process_sentiment(text):
    if not text:
        return "Por favor, forneça um texto para análise."
    result = models['sentiment'](text)[0]
    return f"Sentimento: {result['label']} (Score: {result['score']:.2f})"

def process_qa(question, context):
    if not question or not context:
        return "Por favor, forneça tanto a pergunta quanto o contexto."
    return models['question_answering'](question=question, context=context)['answer']

def process_chat(message, history):
    if not message:
        return "", history
    response = models['chat'](message, max_length=100, do_sample=True)[0]['generated_text']
    history = history + [(message, response)]
    return "", history

def process_image_caption(image):
    if image is None:
        return "Por favor, forneça uma imagem."
    return models['image_caption'](image)[0]['generated_text']

def process_tts(text):
    if not text:
        return None
    return models['text_to_speech'](text)

def process_zero_shot(text, labels):
    if not text or not labels:
        return "Por favor, forneça texto e categorias."
    result = models['zero_shot'](text, labels.split(','))
    return f"Classificação: {result['labels'][0]} (Confiança: {result['scores'][0]:.2f})"

def process_ner(text):
    if not text:
        return "Por favor, forneça um texto para análise."
    results = models['ner'](text)
    entities = []
    for result in results:
        entities.append(f"{result['word']}: {result['entity']}")
    return "\n".join(entities)

with gr.Blocks(theme=gr.themes.Soft()) as demo:
    gr.HTML(open("index.html").read())
    
    with gr.Tabs():
        # Aba de Processamento de Áudio
        with gr.TabItem("🎤 Áudio"):
            with gr.Tabs():
                with gr.TabItem("Transcrição"):
                    audio_input = gr.Audio(type="filepath")
                    transcribe_button = gr.Button("Transcrever")
                    transcription_output = gr.Textbox(label="Resultado")
                    transcribe_button.click(process_transcription, inputs=audio_input, outputs=transcription_output)
                
                with gr.TabItem("Texto para Fala"):
                    tts_input = gr.Textbox(label="Texto para Converter")
                    tts_button = gr.Button("Converter para Áudio")
                    tts_output = gr.Audio(label="Áudio Gerado")
                    tts_button.click(process_tts, inputs=tts_input, outputs=tts_output)
        
        # Aba de Processamento de Texto
        with gr.TabItem("📝 Texto"):
            with gr.Tabs():
                with gr.TabItem("Tradução"):
                    with gr.Row():
                        text_to_translate = gr.Textbox(label="Texto")
                        translation_direction = gr.Radio(["en_pt", "pt_en"], value="en_pt", label="Direção")
                    translate_button = gr.Button("Traduzir")
                    translation_output = gr.Textbox(label="Resultado")
                    translate_button.click(process_translation, inputs=[text_to_translate, translation_direction], outputs=translation_output)
                
                with gr.TabItem("Resumo"):
                    text_sum = gr.Textbox(label="Texto", lines=5)
                    sum_button = gr.Button("Resumir")
                    sum_output = gr.Textbox(label="Resultado")
                    sum_button.click(process_summarization, inputs=text_sum, outputs=sum_output)
                
                with gr.TabItem("Perguntas e Respostas"):
                    qa_context = gr.Textbox(label="Contexto", lines=5)
                    qa_question = gr.Textbox(label="Pergunta")
                    qa_button = gr.Button("Obter Resposta")
                    qa_output = gr.Textbox(label="Resposta")
                    qa_button.click(process_qa, inputs=[qa_question, qa_context], outputs=qa_output)

        # Aba de Análise
        with gr.TabItem("📊 Análise"):
            with gr.Tabs():
                with gr.TabItem("Sentimento"):
                    text_sent = gr.Textbox(label="Texto")
                    sent_button = gr.Button("Analisar Sentimento")
                    sent_output = gr.Textbox(label="Resultado")
                    sent_button.click(process_sentiment, inputs=text_sent, outputs=sent_output)
                
                with gr.TabItem("Entidades Nomeadas"):
                    ner_input = gr.Textbox(label="Texto")
                    ner_button = gr.Button("Identificar Entidades")
                    ner_output = gr.Textbox(label="Entidades Identificadas")
                    ner_button.click(process_ner, inputs=ner_input, outputs=ner_output)
                
                with gr.TabItem("Classificação Zero-Shot"):
                    zero_text = gr.Textbox(label="Texto")
                    zero_labels = gr.Textbox(label="Categorias (separadas por vírgula)")
                    zero_button = gr.Button("Classificar")
                    zero_output = gr.Textbox(label="Resultado")
                    zero_button.click(process_zero_shot, inputs=[zero_text, zero_labels], outputs=zero_output)

        # Aba de IA Avançada
        with gr.TabItem("🤖 IA Avançada"):
            with gr.Tabs():
                with gr.TabItem("Chat"):
                    chatbot = gr.Chatbot()
                    msg = gr.Textbox(label="Mensagem")
                    clear = gr.Button("Limpar Conversa")
                    msg.submit(process_chat, inputs=[msg, chatbot], outputs=[msg, chatbot])
                    clear.click(lambda: None, None, chatbot, queue=False)
                
                with gr.TabItem("Descrição de Imagens"):
                    image_input = gr.Image()
                    caption_button = gr.Button("Gerar Descrição")
                    caption_output = gr.Textbox(label="Descrição")
                    caption_button.click(process_image_caption, inputs=image_input, outputs=caption_output)

demo.launch(share=True)