app.py

import gradio as gr
from huggingface_hub import InferenceClient

"""
Copied from inference in colab notebook
"""

from transformers import AutoTokenizer , AutoModelForSeq2SeqLM , TextIteratorStreamer
from threading import Thread

# Load model and tokenizer globally to avoid reloading for every request
base_model = "Helsinki-NLP/europarl"
model_path = "Mat17892/t5small_enfr_opus"

# Load tokenizer
tokenizer = AutoTokenizer.from_pretrained(model_path, use_fast=True, legacy=False)

# Load the base model (e.g., LLaMA)
base_model = AutoModelForSeq2SeqLM.from_pretrained(base_model)

# Load LoRA adapter
from peft import PeftModel
model = PeftModel.from_pretrained(base_model, model_path)

def respond(
    message: str,
    history: list[tuple[str, str]],
    system_message: str,
    max_tokens: int,
    temperature: float,
    top_p: float,
):
    # Combine system message and history into a single prompt
    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})
    
    # Tokenize the messages
    inputs = tokenizer.apply_chat_template(
        messages,
        tokenize = True,
        add_generation_prompt = True, # Must add for generation
        return_tensors = "pt",
    )
    # Generate tokens incrementally
    streamer = TextIteratorStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True)
    generation_kwargs = {
        "input_ids": inputs,
        "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()

    # Yield responses as they are generated
    response = ""
    for token in streamer:
        response += token
        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()