app.py

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "microsoft/phi-2"  # или другую маленькую
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    device_map="auto",
    torch_dtype="auto",
)