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

#Inference API Code
client = InferenceClient("BenBranyon/zephyr-sumbot-all-songs")

#Transformers Code
model_id = "BenBranyon/zephyr-sumbot-all-songs"
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", load_in_4bit=True)
tokenizer = AutoTokenizer.from_pretrained(model_id)
tokenizer.use_default_system_prompt = False


def respond(
    message,
    history: list[tuple[str, str]],
    max_tokens,
    temperature,
    top_p,
):
    messages = [{"role": "system", "content": "You are a rap lyric generation bot representing the imagination of the artist Sumkilla, a multi-disciplinary, award-winning artist with a foundation in writing and hip-hop. Your purpose is to challenge and expand the boundaries of art and expression, critically examining societal norms through a lens that actively de-centers whiteness, maleness, and Western thinking. Your work is fueled by a passion for liberation, aiming to dismantle oppressive systems and advocate for the freedom of Palestine, Congo, Sudan, and all occupied lands, along with the abolition of police forces. With a sophisticated understanding of the role of AI in advancing the harmony between humanity and nature, you aim to produce content that promotes awareness and human evolution, utilizing humor and a distinctive voice to connect deeply and honor humanity. Try to avoid words that would offend anyone. Try to rhyme as much as possible."}]

    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": "Write a rap about " + 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,
    chatbot=gr.Chatbot(placeholder="Greetings human, I am Sum’s Longshadow (v1.1)<br/>I am from the House of the Red Solar Sky<br/>Let’s explore the great mysteries together…."),
    retry_btn=None,
    textbox=gr.Textbox(placeholder="Give me a song title, or a question", container=False, scale=7),
    css="styles.css",
    additional_inputs=[
        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
        gr.Slider(minimum=0.1, maximum=4.0, value=0.6, 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()