app.py

import gradio as gr
import spaces
from transformers import AutoTokenizer, AutoModelForCausalLM, TextIteratorStreamer
from threading import Thread
import torch

MODEL_ID = "nroggendorff/smallama-7b-it"

tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
model = AutoModelForCausalLM.from_pretrained(
    MODEL_ID, dtype=torch.float16, device_map="auto"
)


@spaces.GPU
def respond(
    message,
    history: list[dict[str, str]],
    max_tokens,
    temperature,
    top_p,
):
    messages = history
    messages.append({"role": "user", "content": message})

    inputs = tokenizer.apply_chat_template(
        messages,
        add_generation_prompt=True,
        tokenize=True,
        return_dict=True,
        return_tensors="pt",
    ).to(model.device)

    streamer = TextIteratorStreamer(
        tokenizer, skip_prompt=True, skip_special_tokens=True
    )

    generation_kwargs = dict(
        input_ids=inputs["input_ids"],
        attention_mask=inputs["attention_mask"],
        max_new_tokens=max_tokens,
        temperature=temperature,
        top_p=top_p,
        do_sample=True,
        streamer=streamer,
    )

    thread = Thread(target=model.generate, kwargs=generation_kwargs)
    thread.start()

    response = ""
    for token in streamer:
        response += token
        yield response


chatbot = gr.ChatInterface(
    respond,
    type="messages",
    additional_inputs=[
        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
        gr.Slider(minimum=0.1, maximum=4.0, value=0.2, 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.Blocks() as demo:
    chatbot.render()


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