facebookresearch / LASER

Language-Agnostic SEntence Representations
Other
3.58k stars 460 forks source link

failed to download tokenizer/detokenizer.perl & normalize-punctuation.perl #171

Closed dcmouth closed 2 years ago

dcmouth commented 3 years ago

when I run similarity.py.it raise error normalize-punctuation.perl: not found. I have already run install_external_tools to download it. I found that the URL “https://raw.githubusercontent.com/moses-smt/mosesdecoder/RELEASE-4.0/scripts” where we download normalize-punctuation.perl is unreachable.when we visit it,it raise 404 . how can I do to solve this problem. Thanks

heffernankevin commented 2 years ago

Hi @dcmouth, when you follow the full path the files are available e.g., https://raw.githubusercontent.com/moses-smt/mosesdecoder/RELEASE-4.0/scripts/tokenizer/normalize-punctuation.perl. The downloaded files should then be located in: LASER/tools-external/moses-tokenizer/tokenizer/ after running install_external_tools.sh.

heffernankevin commented 2 years ago

Closing as user mentioned this issue was resolved in https://github.com/facebookresearch/LASER/issues/212#issuecomment-1193476594.