app.py

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

"""
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
"""
checkpoint = "LemiSt/SmolLM-135M-instruct-de-merged"
tokenizer = AutoTokenizer.from_pretrained(checkpoint)
model = AutoModelForCausalLM.from_pretrained(checkpoint, torch_dtype=torch.bfloat16)


class CustomIterable:
    def __init__(self):
        self._queue = queue.Queue()  # Thread-safe queue
        self.first = True

    def put(self, item):
        """Add an element to the internal queue."""
        if self.first:
            self.first = False
        else:
            self._queue.put(item)

    def end(self):
        """Signal that no more elements will be added."""
        self._queue.put(None)  # Sentinel value to indicate the end of the queue

    def __iter__(self):
        """Return the iterator (self in this case)."""
        return self

    def __next__(self):
        """Return the next element from the queue, blocking if necessary."""
        try:
            item = self._queue.get(block=True)  # Wait for an item
        except queue.Empty:
            raise StopIteration

        if item is None:  # Sentinel value to end the iteration
            raise StopIteration

        return item

def respond(
    message,
    history: list[tuple[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
    top_k,
    repetition_penalty
):
    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})

    streamer = CustomIterable()

    inputs = tokenizer.apply_chat_template(messages, tokenize=True, return_tensors="pt", add_generation_prompt=True)
    thread = threading.Thread(target=model.generate, args=([inputs]), kwargs={"max_new_tokens": max_tokens, "do_sample": True, "temperature": temperature, "top_p": top_p, "top_k": top_k, "repetition_penalty": repetition_penalty, "streamer": streamer})
    thread.start()
    response = ""

    for token in streamer:
        decoded = tokenizer.decode(token, skip_special_tokens=True)
        response += decoded
        yield response

    thread.join()

"""
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="Du bist ein hilfreicher Assistent.", label="System message"),
        gr.Slider(minimum=1, maximum=1024, value=512, step=1, label="Max new tokens"),
        gr.Slider(minimum=0.1, maximum=4.0, value=0.4, step=0.1, label="Temperature"),
        gr.Slider(
            minimum=0.1,
            maximum=1.0,
            value=0.9,
            step=0.05,
            label="Top-p (nucleus sampling)",
        ),
        gr.Slider(
            minimum=16,
            maximum=1024,
            value=512,
            step=1,
            label="Top-k",
        ),
        gr.Slider(
            minimum=0.1,
            maximum=2.0,
            value=1.1,
            step=0.05,
            label="Repetition penalty",
        ),
    ],
)


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