Use AutoModel. CoNLLU annotations. Add mix_dropout.

yzhangcs / parser

:rocket: State-of-the-art parsers for natural language.

https://parser.yzhang.site/

MIT License

829 stars 141 forks source link

Closed attardi closed 4 years ago

attardi commented 4 years ago

Allow using any models from Huggingface.

Some model's tokenizers do not provide a vocab field, so I create it as a dictionary from tokeniser._convert_id_to_token().

yzhangcs commented 4 years ago

Great, thanks for your contribution.