linzehui / mRASP

165 stars 32 forks source link

数据集下载问题 #34

Open areaChun opened 2 years ago

areaChun commented 2 years ago

image Dataset:32-lang-pairs-TRAIN 和 32-lang-pairs-RAS-TRAIN 没法下载了,可以重新提供链接吗

linzehui commented 2 years ago

已更新链接

areaChun commented 2 years ago

image 好像下载还有问题哈 ~ http://lf3-nlp-opensource.bytetos.com/obj/nlp-opensource/emnlp2020/mrasp/pretrain/dataset/train/32-lang-pairs/download.sh

linzehui commented 2 years ago

把下载链接的前缀sf3-ttcdn-tos.pstatp.com全部替换成lf3-nlp-opensource.bytetos.com即可

areaChun commented 2 years ago

可以了 thanks.

surviveMiao commented 2 years ago

把下载链接的前缀sf3-ttcdn-tos.pstatp.com全部替换成lf3-nlp-opensource.bytetos.com即可

Hello, download according to your new address, and you will get 404 prompt. Is there any solution? If it's convenient, please give us a reply. Thank you very much

linzehui commented 2 years ago

把下载链接的前缀sf3-ttcdn-tos.pstatp.com全部替换成lf3-nlp-opensource.bytetos.com即可

Hello, download according to your new address, and you will get 404 prompt. Is there any solution? If it's convenient, please give us a reply. Thank you very much

你下载的是哪个数据?我这里试了一下没问题

surviveMiao commented 2 years ago

把下载链接的前缀sf3-ttcdn-tos.pstatp.com全部替换成lf3-nlp-opensource.bytetos.com即可

Hello, download according to your new address, and you will get 404 prompt. Is there any solution? If it's convenient, please give us a reply. Thank you very much

你下载的是哪个数据?我这里试了一下没问题

您好,近期我们项目需要中-阿拉伯文机翻模型,在论文中找到了您的mrasp模型,感觉非常适合我们。但由于我们本身不是机翻方向的,所以有些文件我们不太明白怎么去使用。 您说的那个连接我们已经修改好了,东西下载下来后不知道如何去使程序跑起来,所以恳请您在方便的时候可以帮我们简单的说一说复现的详细步骤吗? 不胜感激。祝您生活愉快。

linzehui commented 2 years ago

把下载链接的前缀sf3-ttcdn-tos.pstatp.com全部替换成lf3-nlp-opensource.bytetos.com即可

Hello, download according to your new address, and you will get 404 prompt. Is there any solution? If it's convenient, please give us a reply. Thank you very much

你下载的是哪个数据?我这里试了一下没问题

您好,近期我们项目需要中-阿拉伯文机翻模型,在论文中找到了您的mrasp模型,感觉非常适合我们。但由于我们本身不是机翻方向的,所以有些文件我们不太明白怎么去使用。 您说的那个连接我们已经修改好了,东西下载下来后不知道如何去使程序跑起来,所以恳请您在方便的时候可以帮我们简单的说一说复现的详细步骤吗? 不胜感激。祝您生活愉快。

直接调用mrasp做翻译请参考fairseq文档,用fairseq-generate来做;需要在mrasp基础上继续训练则参考fairseq的训练文档,通常包括fairseq-preprocess 对数据做处理 和 fairseq-train 用数据训模型 两步