app.py

import os
import threading
import gradio as gr
from huggingface_hub import InferenceClient

"""
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
"""

os.environ["HF_HOME"] = "/tmp/huggingface"
os.environ["TRANSFORMERS_CACHE"] = "/tmp/huggingface"
MODEL_ID = os.getenv("MODEL_ID", "tianzhechu/BookQA-7B-Instruct")
TOKENIZER_ID = os.getenv("TOKENIZER_ID", "Qwen/Qwen2.5-0.5B-Instruct")  # Optional: tokenizer repo to use locally
USE_LOCAL_TRANSFORMERS = bool(TOKENIZER_ID) or os.getenv("USE_LOCAL_TRANSFORMERS") == "1"

# Remote inference (default)
client = None if USE_LOCAL_TRANSFORMERS else InferenceClient(MODEL_ID)

# Lazy-loaded local model/tokenizer when TOKENIZER_ID is provided
local_model = None
local_tokenizer = None


def _ensure_local_model_loaded():
    global local_model, local_tokenizer
    if local_model is not None and local_tokenizer is not None:
        return
    from transformers import AutoModelForCausalLM, AutoTokenizer

    if not TOKENIZER_ID:
        raise RuntimeError(
            "Local transformers backend requires TOKENIZER_ID to be set to a tokenizer repo."
        )
    local_tokenizer = AutoTokenizer.from_pretrained(TOKENIZER_ID, use_fast=True)
    local_model = AutoModelForCausalLM.from_pretrained(MODEL_ID)


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

    response = ""

    if not USE_LOCAL_TRANSFORMERS:
        for message in client.chat_completion(
            messages,
            max_tokens=max_tokens,
            stream=True,
            temperature=temperature,
            top_p=top_p,
        ):
            token = message.choices[0].delta.content
            if token:
                response += token
            yield response
        return

    # Local generation using transformers with an alternate tokenizer
    _ensure_local_model_loaded()

    try:
        from transformers import TextIteratorStreamer
    except Exception as e:
        raise RuntimeError(
            "transformers TextIteratorStreamer is required for local streaming; ensure transformers is installed."
        ) from e

    # Use chat template if available; otherwise fall back to a simple concatenation
    try:
        prompt_text = local_tokenizer.apply_chat_template(
            messages,
            tokenize=False,
            add_generation_prompt=True,
        )
    except Exception:
        convo_parts = []
        for m in messages:
            role = m.get("role", "user")
            content = m.get("content", "")
            if role == "system":
                convo_parts.append(f"<system>\n{content}\n</system>")
            elif role == "assistant":
                convo_parts.append(f"<assistant>\n{content}\n</assistant>")
            else:
                convo_parts.append(f"<user>\n{content}\n</user>")
        prompt_text = "\n".join(convo_parts) + "\n<assistant>\n"

    inputs = local_tokenizer(prompt_text, return_tensors="pt")

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

    generate_kwargs = dict(
        inputs=inputs.input_ids,
        attention_mask=inputs.get("attention_mask"),
        max_new_tokens=max_tokens,
        do_sample=temperature > 0,
        temperature=temperature,
        top_p=top_p,
        streamer=streamer,
    )

    thread = threading.Thread(target=local_model.generate, kwargs=generate_kwargs)
    thread.start()

    for new_text in streamer:
        if new_text:
            response += new_text
            yield response


"""
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="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()