app.py

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

# Load model and tokenizer directly
model_name = "jdowling/lora_model"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)

# Move the model to the appropriate device (GPU if available, else CPU)
device = "cuda" if torch.cuda.is_available() else "cpu"
model.to(device)


def respond(
    message,
    history: list[tuple[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
):
    # Prepare prompt with history
    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})

    # Convert conversation into a single input string
    prompt = f"{system_message}\n"
    for turn in messages[1:]:
        if turn["role"] == "user":
            prompt += f"User: {turn['content']}\n"
        elif turn["role"] == "assistant":
            prompt += f"Assistant: {turn['content']}\n"
    prompt += "Assistant:"

    # Tokenize input
    inputs = tokenizer(prompt, return_tensors="pt").to(device)

    # Generate response
    output = model.generate(
        inputs["input_ids"],
        max_length=inputs["input_ids"].shape[1] + max_tokens,
        temperature=temperature,
        top_p=top_p,
        pad_token_id=tokenizer.eos_token_id,
    )

    # Decode response and extract the new assistant message
    response = tokenizer.decode(output[0], skip_special_tokens=True)
    response = response[len(prompt):].strip()  # Strip the input part from the response

    yield response


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)",
        ),
    ],
)

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