app.py

import gradio as gr
from transformers import AutoTokenizer, AutoModelForCausalLM
from spaces import GPU as gpu


class Delta:
    def __init__(self, content):
        self.content = content


class Choice:
    def __init__(self, delta):
        self.delta = delta


class InferenceClient:
    def __init__(self, model_id="nroggendorff/smallama-it"):
        self.tokenizer = AutoTokenizer.from_pretrained(model_id)
        self.model = AutoModelForCausalLM.from_pretrained(model_id)

    class ModelOutput:
        def __init__(self, client, inputs):
            self.client = client
            self.inputs = inputs
            self.choices = []

        def decode(self, output):
            decoded_output = self.client.tokenizer.decode(
                output[0][self.inputs["input_ids"].shape[-1] :],
                skip_special_tokens=True,
            )
            self.choices = [Choice(Delta(decoded_output))]
            return self

    @gpu
    def chat_completion(
        self, messages, max_tokens=256, stream=True, temperature=0.2, top_p=0.95
    ):
        inputs = self.tokenizer.apply_chat_template(
            messages,
            add_generation_prompt=True,
            tokenize=True,
            return_dict=True,
            return_tensors="pt",
        ).to(self.model.device)

        model_output = self.ModelOutput(self, inputs)

        for _ in range(max_tokens):
            output = self.model.generate(
                **inputs, max_new_tokens=1, temperature=temperature, top_p=top_p
            )
            yield model_output.decode(output)


def respond(
    message,
    history: list[dict[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
):
    client = InferenceClient()
    messages = [{"role": "system", "content": system_message}]
    messages.extend(history)
    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,
    ):
        choices = message.choices
        token = ""
        if len(choices) and choices[0].delta.content:
            token = choices[0].delta.content
        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()