import gradio as gr
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

# Load the model and tokenizer
model_name = "indonlp/cendol-llama2-7b-chat"  
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)

# Define the chatbot function
def chatbot(input_text):
    inputs = tokenizer(input_text, return_tensors="pt")
    with torch.no_grad():
        outputs = model.generate(inputs.input_ids, max_length=1000, do_sample=True)
    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
    return response

# Create the Gradio interface
interface = gr.ChatInterface(
    fn=chatbot,
    title="Cendol Chatnot",
    description="Ask any coding-related questions and get helpful responses."
)

# Launch the web UI
interface.launch(share=True)