app.py

import gradio as gr
from transformers import GPT2LMHeadModel, GPT2Tokenizer

# Load model and tokenizer from Hugging Face Hub
model_name = "Electricarchmage/cookbookgpt"
model = GPT2LMHeadModel.from_pretrained(model_name)
tokenizer = GPT2Tokenizer.from_pretrained(model_name)

# Define the respond function
def respond(
    message,
    history: list[tuple[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
):
    # Preparing the messages for context (the history and the new message)
    input_text = system_message + "\n"
    
    for val in history:
        if val[0]:
            input_text += f"User: {val[0]}\n"
        if val[1]:
            input_text += f"Assistant: {val[1]}\n"

    input_text += f"User: {message}\nAssistant:"

    # Tokenize the input and generate a response
    inputs = tokenizer(input_text, return_tensors="pt", max_length=1024, truncation=True)

    # Generate output tokens
    output = model.generate(
        inputs["input_ids"],
        max_length=max_tokens + len(inputs["input_ids"][0]),
        temperature=temperature,
        top_p=top_p,
        num_return_sequences=1,
        no_repeat_ngram_size=2,
    )

    # Decode the output tokens into text
    response = tokenizer.decode(output[0], skip_special_tokens=True)

    # Extract only the assistant's reply
    assistant_reply = response.split("Assistant:")[-1].strip()

    return assistant_reply

# Define the Gradio interface
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)",
        ),
    ],
)

# Launch the app
if __name__ == "__main__":
    demo.launch()