app.py

import os
import gradio as gr
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import PeftModel

BASE_MODEL = os.environ.get("BASE_MODEL", "Qwen/Qwen2.5-0.5B-Instruct")
ADAPTER_RL_PATH = os.environ.get("ADAPTER_RL_PATH", "./adapter_dpo")  # commit folder into Space repo

SYSTEM_PROMPT = """You are a retail recommendation assistant.
You recommend at most 3 items that complement the user's cart and intent.
You must be:
- Relevant to the cart + intent
- Constraint-aware (budget, urgency, compatibility, brand preferences)
- Non-pushy and honest (no made-up specs or guarantees)
- Concise and structured

Output format:
Recommendations:
1) <item> — <one-line reason>
2) ...
Why these:
- ...
Compatibility / checks:
- ...
Optional next step:
- (only if helpful)
"""

def build_prompt(user_intent, cart, budget, urgency, brand_avoid):
    cart_list = [c.strip() for c in cart.split(",") if c.strip()]
    constraints = {
        "budget_usd": budget,
        "shipping_urgency": urgency,
        "brand_avoid": [b.strip() for b in brand_avoid.split(",") if b.strip()]
    }
    user_block = (
        f"User intent: {user_intent}\n"
        f"Cart: {', '.join(cart_list)}\n"
        f"Constraints: {constraints}\n"
        "Generate recommendations following the required format."
    )
    return f"<|system|>\n{SYSTEM_PROMPT}\n<|user|>\n{user_block}\n<|assistant|>\n"

@torch.inference_mode()
def generate(model, tok, prompt, max_new_tokens=220, temperature=0.7):
    inputs = tok(prompt, return_tensors="pt").to(model.device)
    out = model.generate(
        **inputs,
        max_new_tokens=max_new_tokens,
        do_sample=True if temperature > 0 else False,
        temperature=temperature,
        pad_token_id=tok.eos_token_id,
    )
    text = tok.decode(out[0], skip_special_tokens=True)

    # Best-effort: return only assistant completion
    if "<|assistant|>" in text:
        return text.split("<|assistant|>", 1)[-1].strip()
    return text.strip()

def load_models():
    tok = AutoTokenizer.from_pretrained(BASE_MODEL, use_fast=True)
    if tok.pad_token is None:
        tok.pad_token = tok.eos_token

    base = AutoModelForCausalLM.from_pretrained(
        BASE_MODEL,
        device_map="auto",
        torch_dtype="auto",
    )

    rl = None
    if os.path.exists(ADAPTER_RL_PATH):
        rl = PeftModel.from_pretrained(base, ADAPTER_RL_PATH)
    return tok, base, rl

tok, base_model, rl_model = load_models()

def run_both(user_intent, cart, budget, urgency, brand_avoid, max_new_tokens, temperature):
    prompt = build_prompt(user_intent, cart, budget, urgency, brand_avoid)
    before = generate(base_model, tok, prompt, max_new_tokens=max_new_tokens, temperature=temperature)

    if rl_model is None:
        after = "RL adapter not found. Ensure adapter_dpo/ is included in the Space repo or ADAPTER_RL_PATH is correct."
    else:
        after = generate(rl_model, tok, prompt, max_new_tokens=max_new_tokens, temperature=temperature)

    return before, after

with gr.Blocks() as demo:
    gr.Markdown("# Retail Recommendation Explainer — Before vs After RL (DPO)")

    user_intent = gr.Textbox(
        label="User intent",
        value="I’m starting to run regularly and want to avoid blisters."
    )
    cart = gr.Textbox(label="Cart items (comma-separated)", value="running shoes, socks")
    with gr.Row():
        budget = gr.Slider(10, 150, value=40, step=5, label="Budget (USD)")
        urgency = gr.Dropdown(["fast", "normal"], value="fast", label="Shipping urgency")
    brand_avoid = gr.Textbox(label="Brands/materials to avoid (comma-separated)", value="")

    with gr.Row():
        max_new_tokens = gr.Slider(80, 400, value=220, step=10, label="Max new tokens")
        temperature = gr.Slider(0.0, 1.2, value=0.7, step=0.05, label="Temperature")

    btn = gr.Button("Generate (Before vs After)")
    with gr.Row():
        out_before = gr.Textbox(label="Before (Base)", lines=18)
        out_after = gr.Textbox(label="After (RL / DPO LoRA)", lines=18)

    btn.click(
        fn=run_both,
        inputs=[user_intent, cart, budget, urgency, brand_avoid, max_new_tokens, temperature],
        outputs=[out_before, out_after]
    )

demo.launch()