from unsloth import FastLanguageModel
import gradio as gr

# Declare necessary variables
max_seq_length = 2048  # Choose any! We auto support RoPE Scaling internally!
dtype = None  # None for auto detection. Float16 for Tesla T4, V100, Bfloat16 for Ampere+
load_in_4bit = True  # Use 4bit quantization to reduce memory usage. Can be False.

# Load the model and tokenizer
model, tokenizer = FastLanguageModel.from_pretrained(
    model_name="abdfajar707/llama3_8B_lora_model_rkp_pn2025_v3",  # YOUR MODEL YOU USED FOR TRAINING
    max_seq_length=max_seq_length,
    dtype=dtype,
    load_in_4bit=load_in_4bit,
)
FastLanguageModel.for_inference(model)  # Enable native 2x faster inference

# Define the respond function
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 msg in model.chat_completion(
        messages,
        max_tokens=max_tokens,
        stream=True,
        temperature=temperature,
        top_p=top_p,
    ):
        token = msg.choices[0].delta.content
        response += token
        yield response

# Create the Gradio interface
demo = gr.ChatInterface(
    fn=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)",
        ),
    ],
)

# Launch the Gradio interface
if __name__ == "__main__":
    demo.launch()