# 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
# """
# client = InferenceClient("harsh4733/Llama-2-7b-chat-finetune-webglm")


# 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

# """
# For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
# """
# 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)",
#         ),
#     ],
# )

# import gradio as gr
# from transformers import pipeline

# def chat_with_model(question, prompt, system_message, max_tokens, temperature, top_p):
#     prompt_template = f"<s>[INST] <<SYS>>\n{system_message} <</SYS>> {prompt} [/INST]"

#     pipe = pipeline(
#         task="text-generation",
#         model="harsh4733/Llama-2-7b-chat-finetune-webglm",
#         tokenizer="harsh4733/Llama-2-7b-chat-finetune-webglm",
#         max_length=max_tokens,
#         temperature=temperature,
#         top_p=top_p,
#     )

#     result = pipe(prompt_template)
#     return result[0]['generated_text']

# def respond(
#     question,
#     prompt,
#     system_message,
#     max_tokens,
#     temperature,
#     top_p,
# ):
#     response = chat_with_model(question, prompt, system_message, max_tokens, temperature, top_p)
#     return response

# # Define Gradio interface
# demo = gr.Interface(
#     fn=respond,
#     inputs=[
#         gr.Textbox(value="What is a large language model?", label="Question"),
#         gr.Textbox(value="You are a helpful assistant that provides answers to the questions given based on the references provided to you regarding the question.", label="System message"),
#         gr.Textbox(value="You are a friendly Chatbot.", label="Prompt"),
#         gr.Slider(minimum=1, maximum=2048, value=512, label="Max new tokens"),
#         gr.Slider(minimum=0.1, maximum=4.0, value=0.7, label="Temperature"),
#         gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
#     ],
#     outputs=gr.Textbox(label="Response"),
#     title="Chat with Large Language Model",
#     description="Interact with a large language model to generate responses based on your input.",
# )

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


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

import gradio as gr
from transformers import TFAutoModelForCausalLM, AutoTokenizer
import tensorflow as tf

def chat_with_model(question, prompt, system_message, max_tokens, temperature, top_p):
    tokenizer = AutoTokenizer.from_pretrained("harsh4733/Llama-2-7b-chat-finetune-webglm")
    model = TFAutoModelForCausalLM.from_pretrained("harsh4733/Llama-2-7b-chat-finetune-webglm")

    prompt_template = f"<s>[INST] <<SYS>>\n{system_message} <</SYS>> {prompt} [/INST]"

    input_ids = tokenizer.encode(prompt_template, return_tensors="tf", max_length=512, truncation=True)
    output = model.generate(input_ids, max_length=max_tokens, temperature=temperature, top_p=top_p, num_return_sequences=1)

    response = tokenizer.decode(output[0], skip_special_tokens=True)
    return response

def respond(
    question,
    prompt,
    system_message,
    max_tokens,
    temperature,
    top_p,
):
    response = chat_with_model(question, prompt, system_message, max_tokens, temperature, top_p)
    return response

# Define Gradio interface
demo = gr.Interface(
    fn=respond,
    inputs=[
        gr.Textbox(value="What is a large language model?", label="Question"),
        gr.Textbox(value="You are a helpful assistant that provides answers to the questions given based on the references provided to you regarding the question.", label="System message"),
        gr.Textbox(value="You are a friendly Chatbot.", label="Prompt"),
        gr.Slider(minimum=1, maximum=2048, value=512, label="Max new tokens"),
        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, label="Temperature"),
        gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
    ],
    outputs=gr.Textbox(label="Response"),
    title="Chat with Large Language Model",
    description="Interact with a large language model to generate responses based on your input.",
)

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