import gradio as gr
from huggingface_hub import InferenceClient
import torch
from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM

"""
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
"""
client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")

def respond(
    message,
    history: list[tuple[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
):
    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 = ""

    for message in client.chat_completion(
        messages,
        max_tokens=max_tokens,
        stream=True,
        temperature=temperature,
        top_p=top_p,
    ):
        token = message.choices[0].delta.content

        response += token
        yield response

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()

    import gradio as gr
from huggingface_hub import InferenceClient
import torch
from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM
# Crear la función de loop automatizado
def experiment_loop(initial_question, max_cycles=10):
    prompt = f"<thinking>{initial_question}</thinking>"
    effectiveness = 100  # Inicializa el porcentaje de efectividad
    communication = "Initializing experiment."
    response_log = []

    for cycle in range(max_cycles):
        # Generar la respuesta del modelo
        inputs = tokenizer(prompt, return_tensors="pt").input_ids
        outputs = model.generate(inputs, max_length=200)
        response = tokenizer.decode(outputs[0], skip_special_tokens=True)

        # Descomponer la respuesta en afirmación y nueva pregunta
        affirmation = extract_affirmation(response)
        new_question = extract_question(response)

        # Actualizar el estado de la efectividad
        effectiveness = min(1000, effectiveness + 10 * cycle)  # Ejemplo de aumento de efectividad

        # Comunicación con el usuario
        communication = f"Cycle {cycle + 1}: Affirmation: '{affirmation}' | New Question: '{new_question}'"

        # Guardar el ciclo actual en el log
        response_log.append((affirmation, new_question, effectiveness, communication))

        # Verificar si el modelo decide detenerse
        if "Descanso" in response:
            final_output = generate_final_output(response_log)
            return final_output
        
        # Actualizar el prompt con la nueva afirmación y pregunta
        prompt = f"<thinking>{affirmation} {new_question}</thinking>"

    # Si se alcanza el número máximo de ciclos sin detenerse
    final_output = generate_final_output(response_log)
    return final_output

# Funciones auxiliares para extraer afirmaciones, preguntas y generar la salida final
def extract_affirmation(response):
    return response.split('.')[0]

def extract_question(response):
    return response.split('?')[-2].strip() + "?"

def generate_final_output(log):
    final_affirmation = log[-1][0]
    final_question = log[-1][1]
    final_communication = f"Experiment completed. Final Affirmation: '{final_affirmation}' | Final Question: '{final_question}'"
    return final_communication
# Iniciar el experimento después de que la función ha sido definida
initial_question = "What happens in the space between a response and its recreation?"
result = experiment_loop(initial_question)
print(result)


# Define the experiment loop
initial_question = "What happens in the space between a response and its recreation?"
result = experiment_loop(initial_question)
print(result)
import torch
from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM

# Cargar el modelo de lenguaje preentrenado
model_name = "gpt-neo-2.7B"  # Puedes cambiarlo a GPT-J o cualquier otro
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)

# Crear la función de loop automatizado
def experiment_loop(initial_question, max_cycles=10):
    prompt = f"<thinking>{initial_question}</thinking>"
    effectiveness = 100  # Inicializa el porcentaje de efectividad
    communication = "Initializing experiment."
    response_log = []

    for cycle in range(max_cycles):
        # Generar la respuesta del modelo
        inputs = tokenizer(prompt, return_tensors="pt").input_ids
        outputs = model.generate(inputs, max_length=200)
        response = tokenizer.decode(outputs[0], skip_special_tokens=True)

        # Descomponer la respuesta en afirmación y nueva pregunta
        affirmation = extract_affirmation(response)
        new_question = extract_question(response)

        # Actualizar el estado de la efectividad
        effectiveness = min(1000, effectiveness + 10 * cycle)  # Ejemplo de aumento de efectividad

        # Comunicación con el usuario
        communication = f"Cycle {cycle + 1}: Affirmation: '{affirmation}' | New Question: '{new_question}'"

        # Guardar el ciclo actual en el log
        response_log.append((affirmation, new_question, effectiveness, communication))

        # Verificar si el modelo decide detenerse
        if "Descanso" in response:
            final_output = generate_final_output(response_log)
            return final_output
        
        # Actualizar el prompt con la nueva afirmación y pregunta
        prompt = f"<thinking>{affirmation} {new_question}</thinking>"

    # Si se alcanza el número máximo de ciclos sin detenerse
    final_output = generate_final_output(response_log)
    return final_output

# Funciones auxiliares para extraer afirmaciones, preguntas y generar la salida final
def extract_affirmation(response):
    # Lógica para extraer la afirmación de la respuesta
    return response.split('.')[0]

def extract_question(response):
    # Lógica para extraer la nueva pregunta de la respuesta
    return response.split('?')[-2].strip() + "?"

def generate_final_output(log):
    final_affirmation = log[-1][0]
    final_question = log[-1][1]
    final_communication = f"Experiment completed. Final Affirmation: '{final_affirmation}' | Final Question: '{final_question}'"
    return final_communication

# Iniciar el experimento
initial_question = "What happens in the space between a response and its recreation?"
result = experiment_loop(initial_question)
print(result)