app.py

import gradio as gr
from huggingface_hub import InferenceClient
from transformers import AutoModelForCausalLM, Trainer, TrainingArguments
from datasets import load_dataset
import os

"""
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("meta-llama/Meta-Llama-3-8B-Instruct")


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


def train_model(hf_token_value):
    os.environ["HUGGINGFACE_TOKEN"] = hf_token_value

    # Load dataset
    dataset = load_dataset('json', data_files={
                           'train': 'training_set.json'})

    # Load model
    model = AutoModelForCausalLM.from_pretrained(
        'meta-llama/Meta-Llama-3-8B-Instruct')

    # Define training arguments
    training_args = TrainingArguments(
        output_dir='./results',
        num_train_epochs=3,
        per_device_train_batch_size=16,
        save_steps=10_000,
        save_total_limit=2,
    )

    # Initialize Trainer
    trainer = Trainer(
        model=model,
        args=training_args,
        train_dataset=dataset['train'],
        eval_dataset=dataset['test']
    )

    # Start training
    trainer.train()
    return "Training complete"


"""
For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
"""
demo = gr.Blocks()

with demo:
    gr.Markdown("# Llama3training Chatbot and Model Trainer")
    with gr.Tab("Chat"):
        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)",
                ),
            ],
        )
    with gr.Tab("Train"):
        hf_token = gr.Textbox(label="Hugging Face Token", type="password")
        train_button = gr.Button("Start Training")
        train_output = gr.Textbox(label="Training Output")

        train_button.click(train_model, inputs=hf_token, outputs=train_output)

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