app.py

import gradio as gr
from huggingface_hub import InferenceClient

"""
For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
"""
# Criando o cliente para interagir com o modelo no Hugging Face.
client = InferenceClient("qwen2.5:0.5b")


def respond(
    message,
    history: list[tuple[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
):
    # Prepara as mensagens para a API
    messages = [{"role": "system", "content": system_message}]

    for val in history:
        if val[0]:
            messages.append({"role": "user", "content": val[0]})
        if val[1]:
            messages.append({"role": "assistant", "content": val[1]})

    messages.append({"role": "user", "content": message})

    response = ""  # Resposta acumulada

    try:
        # Chama a API de completamento com streaming
        # A API do Hugging Face usa o método `client.chat_completion`.
        response_stream = client.chat_completion(
            messages=messages,
            max_tokens=max_tokens,
            temperature=temperature,
            top_p=top_p,
            stream=True
        )

        for message in response_stream:
            # Verifica se a resposta contém o conteúdo esperado
            if 'choices' not in message or len(message['choices']) == 0 or 'delta' not in message['choices'][0]:
                raise ValueError("Resposta inesperada do modelo.")

            token = message['choices'][0]['delta']['content']
            response += token  # Acumula o conteúdo

            # Retorna a resposta incrementalmente
            yield response

    except ValueError as e:
        print(f"Erro de valor: {e}")
    except ConnectionError as e:
        print(f"Erro de conexão: {e}")
    except TimeoutError as e:
        print(f"Erro de tempo: {e}")
    except Exception as e:
        print(f"Erro inesperado: {e}")

    return response  # Retorna a resposta final ao final do processamento


"""
For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
"""
# Criando a interface Gradio
demo = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
        gr.Slider(
            minimum=0.1,
            maximum=1.0,
            value=0.95,
            step=0.05,
            label="Top-p (nucleus sampling)",
        ),
    ],
)


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