app.py

import threading

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

MODEL_ID = "HuggingFaceTB/nanowhale-100m"

print(f"Loading model {MODEL_ID} ...")
import torch
from safetensors.torch import load_file
from transformers import AutoConfig, AutoModelForCausalLM, AutoTokenizer
from huggingface_hub import hf_hub_download

# Load model (recommended: manual load for reliability)
config = AutoConfig.from_pretrained("HuggingFaceTB/nanowhale-100m", trust_remote_code=True)
model = AutoModelForCausalLM.from_config(config, trust_remote_code=True).float()

# Download and load weights
weights_path = hf_hub_download("HuggingFaceTB/nanowhale-100m", "model.safetensors")
state_dict = load_file(weights_path)
model.load_state_dict(state_dict, strict=True)
model = model.eval()

tokenizer = AutoTokenizer.from_pretrained("HuggingFaceTB/nanowhale-100m")


print("Model loaded.")

DEVICE = next(model.parameters()).device


def build_prompt(system_message: str, history: list[dict[str, str]], user_message: str) -> str:
    """
    Try to use the tokenizer's built-in chat template.
    Fall back to a simple newline-delimited format if none exists.
    """
    messages = [{"role": "system", "content": system_message}]
    messages.extend(history)
    messages.append({"role": "user", "content": user_message})

    if tokenizer.chat_template is not None:
        return tokenizer.apply_chat_template(
            messages,
            tokenize=False,
            add_generation_prompt=True,
        )

    # Fallback format
    parts = [f"System: {system_message}\n"]
    for msg in history:
        role = "User" if msg["role"] == "user" else "Assistant"
        parts.append(f"{role}: {msg['content']}\n")
    parts.append(f"User: {user_message}\nAssistant:")
    return "".join(parts)


def respond(
    message: str,
    history: list[dict[str, str]],
    system_message: str,
    max_new_tokens: int,
    temperature: float,
    top_p: float,
):
    prompt = build_prompt(system_message, history, message)

    inputs = tokenizer(prompt, return_tensors="pt").to(DEVICE)
    input_len = inputs["input_ids"].shape[-1]

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

    generation_kwargs = dict(
        **inputs,
        max_new_tokens=max_new_tokens,
        temperature=temperature,
        top_p=top_p,
        do_sample=temperature > 0.0,
        streamer=streamer,
    )

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

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

    thread.join()


chatbot = gr.ChatInterface(
    respond,
    title="Timmy, chat powered by nanowhale-100m",
    additional_inputs=[
        gr.Textbox(value="You are a friendly chatbot.", label="System message"),
        gr.Slider(minimum=1, maximum=512, value=128, 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__":
    chatbot.launch()