import gradio as gr
from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig
import torch
import spaces  # <-- правильный импорт (пакет spaces)

MODEL = "covaga/eplan-assistant-v3-merged"

print("Loading model...")
tokenizer = AutoTokenizer.from_pretrained(MODEL)
quant_config = BitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_compute_dtype=torch.float16,
)
model = AutoModelForCausalLM.from_pretrained(
    MODEL,
    device_map="auto",
    quantization_config=quant_config,
)
print("Model loaded.")

@spaces.GPU  # <-- используем декоратор из пакета spaces
def respond(message, history):
    messages = [{"role": "user", "content": message}]
    inputs = tokenizer.apply_chat_template(
        messages,
        add_generation_prompt=True,
        tokenize=True,
        return_dict=True,
        return_tensors="pt"
    ).to(model.device)
    outputs = model.generate(
        **inputs,
        max_new_tokens=512,
        do_sample=True,
        temperature=0.7
    )
    answer = tokenizer.decode(outputs[0][inputs.input_ids.shape[-1]:], skip_special_tokens=True)
    return answer

demo = gr.ChatInterface(
    fn=respond,
    title="EPLAN Assistant",
    description="Вопросы по EPLAN Electric 2026"
)
demo.launch()