from openmind import AutoTokenizer, AutoModelForCausalLM, is_torch_npu_available
from openmind_hub import snapshot_download
from transformers import AutoModelForSeq2SeqLM
import torch
import argparse
import torch.nn.functional as F


# 均值池化 - 考虑注意力掩码以进行正确的平均
def mean_pooling(model_output, attention_mask):
    token_embeddings = model_output[0] # model_output的第一个元素包含所有token嵌入
    input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
    return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)

def parse_args():
    parser = argparse.ArgumentParser()
    parser.add_argument(
        "--model_name_or_path",
        type=str,
        help="Path to model",
        default="../",
    )
    args = parser.parse_args()
    return args

def main():
    args = parse_args()
    model_path = args.model_name_or_path

    if is_torch_npu_available():
        device = "npu:0"
    else:
        device = "cpu"
    # 从openmind_hub加载模型
    tokenizer = AutoTokenizer.from_pretrained(model_path)
    model = AutoModelForSeq2SeqLM.from_pretrained(model_path)

    # 对句子进行分词
    encoded_input = tokenizer(">>nl<< Your English text here", return_tensors="pt")
    # 计算token嵌入
    with torch.no_grad():
        model_output = model.generate(**encoded_input)
    print(tokenizer.batch_decode(model_output, skip_special_tokens=True))


if __name__ == "__main__":
    main()