src/demo.py

import os
import random
from threading import Thread
from typing import Iterable

import torch
from huggingface_hub import HfApi
from datasets import load_dataset
from transformers import T5Tokenizer, T5ForConditionalGeneration
from transformers import AutoTokenizer, AutoModelForCausalLM


TOKEN = os.environ.get("HF_TOKEN", None)


model_id = "meta-llama/Llama-2-7b-chat-hf"
# tokenizer = AutoTokenizer.from_pretrained(model_id)
# model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.float16)


type2dataset = {
    "re2text-easy": load_dataset('3B-Group/ConvRe', "en-re2text", token=True, split="prompt1"),
    "re2text-hard": load_dataset('3B-Group/ConvRe', "en-re2text", token=True, split="prompt4"),
    "text2re-easy": load_dataset('3B-Group/ConvRe', "en-text2re", token=True, split="prompt1"),
    "text2re-hard": load_dataset('3B-Group/ConvRe', "en-text2re", token=True, split="prompt3")
}

# type2dataset = {}


def generate():
    return "1"


def random_examples(dataset_key) -> str:
    # target_dataset = type2dataset[f"{task.lower()}-{type.lower()}"]
    target_dataset = type2dataset[dataset_key]

    idx = random.randint(0, len(target_dataset) - 1)
    item = target_dataset[idx]
    return item['query']