python 离线翻译中译英

最新推荐文章于 2025-04-14 10:16:45 发布

feeday

最新推荐文章于 2025-04-14 10:16:45 发布

阅读量698

点赞数 3

CC 4.0 BY-SA版权

文章标签： python 机器翻译

本文链接：https://blog.csdn.net/feeday/article/details/138236169

本文介绍了如何在Python环境中配置虚拟环境，并使用HuggingFace的transformers库加载预训练的Helsinki-NLP模型进行中英文翻译，提供了一个简单的交互式脚本实现翻译功能。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

https://huggingface.co/Helsinki-NLP/opus-mt-zh-en/tree/main

配置虚拟环境

# conda create -n fys python=3.11.5
# conda activate fys
# pip install --upgrade pip
# pip install transformers sentencepiece tensorflow torch
# pip install tensorflow -i https://mirrors.aliyun.com/pypi/simple
# pip install --force-reinstall sacremoses

翻译脚本

from transformers import AutoModelForSeq2SeqLM, AutoTokenizer
# Load model
model_name = "/Helsinki-NLP/opus-mt-zh-en"  # Example model
tokenizer = AutoTokenizer.from_pretrained(model_name, local_files_only=True)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name, local_files_only=True)

def translate(text):
    inputs = tokenizer.encode(text, return_tensors="pt")
    outputs = model.generate(inputs)
    translated_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
    return translated_text

if __name__ == '__main__':
    try:
        while True:
            source_text = input("请输入要中文文本: ")
            if source_text.lower() == 'q':
                break
            translation = translate(source_text)
            print("翻译结果: ", translation)
    except Exception as e:
        print("发生错误:", e)
        input("按 Enter 键退出...")

启动脚本

python.exe -s cntoen.py
pause

F:\python\fys
F:\python\fys\Helsinki-NLP\opus-mt-zh-en
F:\python\fys\fy.bat
F:\python\fys\cntoen.py