app_low.py

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

# ============================================================
# 1️⃣ Load model and tokenizer
# ============================================================
MODEL_ID = "gokaygokay/prompt-enhancer-gemma-3-270m-it"

# Use GPU if available
device = 0 if torch.cuda.is_available() else -1

tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
model = AutoModelForCausalLM.from_pretrained(MODEL_ID)

pipe = pipeline(
    "text-generation",
    model=model,
    tokenizer=tokenizer,
    device=device,  # 0 for GPU, -1 for CPU
)


# ============================================================
# 2️⃣ Define the generation function (chat-template style)
# ============================================================
def enhance_prompt(user_prompt, temperature, max_tokens, chat_history):
    chat_history = chat_history or []

    # Build messages using proper roles
    messages = [
        {"role": "system", "content": "Enhance and expand the following prompt with more details and context:"},
        {"role": "user", "content": user_prompt}
    ]

    # Use tokenizer chat template to build the input
    prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)

    # Generate output
    output = pipe(
        prompt,
        max_new_tokens=int(max_tokens),
        temperature=float(temperature),
        do_sample=True,
    )[0]["generated_text"].strip()
    print(output)
    print(output[0])

    # Append conversation to history
    chat_history.append({"role": "user", "content": user_prompt})
    chat_history.append({"role": "assistant", "content": output})

    return chat_history

import re

def extract_later_part(user_prompt, generated_text):
    """
    Cleans the model output and extracts only the enhanced (later) portion.
    Removes prompt echoes and system tags like <end_of_turn>, <start_of_turn>, etc.
    """
    # Step 1: Clean up internal tags
    cleaned = re.sub(r"<.*?>", "", generated_text)  # Remove <end_of_turn>, <start_of_turn>, etc.
    cleaned = cleaned.strip()

    # Step 2: Normalize spaces
    cleaned = re.sub(r"\s+", " ", cleaned)

    # Step 3: Try removing the original prompt if repeated
    user_prompt_clean = user_prompt.strip().lower()
    cleaned_lower = cleaned.lower()

    if cleaned_lower.startswith(user_prompt_clean):
        cleaned = cleaned[len(user_prompt):].strip(",. ").strip()

    return cleaned


# ===================== Prompt Enhancer Function =====================
def enhance_prompt1(user_prompt, temperature, max_tokens, chat_history):
    messages = [
        {"role": "system", "content": "Enhance and expand the following prompt with more details and context:"},
        {"role": "user", "content": user_prompt}
    ]

    prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
    output = pipe(prompt, max_new_tokens=256)
    raw_output = output[0]['generated_text']

    print("=== RAW MODEL OUTPUT ===")
    print(raw_output)

    # Extract the cleaned, later portion
    later_part = extract_later_part(user_prompt, raw_output)
    print("=== EXTRACTED CLEANED OUTPUT ===")
    print(later_part)

    # Append to chat history for Gradio
    chat_history = chat_history or []
    chat_history.append({"role": "user", "content": user_prompt})
    chat_history.append({"role": "assistant", "content": later_part})

    return chat_history

# ============================================================
# 3️⃣ Gradio UI
# ============================================================
with gr.Blocks(title="Prompt Enhancer – Gemma 3 270M", theme=gr.themes.Soft()) as demo:
    gr.Markdown(
        """
        # ✨ Prompt Enhancer (Gemma 3 270M)
        Enter a short prompt, and the model will **expand it with details and creative context**  
        using the Gemma chat-template interface.
        """
    )

    with gr.Row():
        chatbot = gr.Chatbot(height=400, label="Enhanced Prompts", type="messages")
        with gr.Column(scale=1):
            user_prompt = gr.Textbox(
                placeholder="Enter a short prompt...",
                label="Your Prompt",
                lines=3,
            )
            temperature = gr.Slider(0.0, 1.0, value=0.7, step=0.05, label="Temperature")
            max_tokens = gr.Slider(32, 256, value=128, step=16, label="Max Tokens")
            send_btn = gr.Button("🚀 dev dont click", variant="primary")
            clear_btn = gr.Button("🧹 Clear Chat")
            add_btn = gr.Button("🚀 Enchance  Prompt", variant="primary")

    # Bind UI actions
    #send_btn.click(enhance_prompt, [user_prompt, temperature, max_tokens, chatbot], chatbot)
    user_prompt.submit(enhance_prompt, [user_prompt, temperature, max_tokens, chatbot], chatbot)
    clear_btn.click(lambda: [], None, chatbot)
    add_btn.click(enhance_prompt1, [user_prompt, temperature, max_tokens, chatbot], chatbot)

    gr.Markdown(
        """
        ---
        💡 **Tips:**
        - Works best with short, descriptive prompts (e.g., "a cat sitting on a chair")
        - Increase *Temperature* for more creative output.
        """
    )

# ============================================================
# 4️⃣ Launch
# ============================================================
if __name__ == "__main__":
    demo.launch(show_error=True)