wechsel

wechsel

Wechselのコード：NAACL2022で公開された単一言語モデルの横断的伝達のためのサブワード埋め込みの効果的な初期化。

論文：https：//aclanthology.org/2022.naacl-main.293/

論文のモデルは、ハグFaceハブで入手できます。

• roberta-base-wechsel-french
• roberta-base-wechsel-german
• roberta-base-wechsel-chinese
• roberta-base-wechsel-swahili
• gpt2-wechsel-french
• gpt2-wechsel-german
• gpt2-wechsel-chinese
• gpt2-wechsel-swahili

インストール

Pypiを介してPythonパッケージを配布します。

 pip install wechsel

または、リポジトリをクローンし、 requirements.txtをインストールし、 wechsel/でコードを実行します。

使用の例

イングリッシュroberta-baseをスワヒリ語に移す：

 import torch
from transformers import AutoModel , AutoTokenizer
from datasets import load_dataset
from wechsel import WECHSEL , load_embeddings

source_tokenizer = AutoTokenizer . from_pretrained ( "roberta-base" )
model = AutoModel . from_pretrained ( "roberta-base" )

target_tokenizer = source_tokenizer . train_new_from_iterator (
    load_dataset ( "oscar" , "unshuffled_deduplicated_sw" , split = "train" )[ "text" ],
    vocab_size = len ( source_tokenizer )
)

wechsel = WECHSEL (
    load_embeddings ( "en" ),
    load_embeddings ( "sw" ),
    bilingual_dictionary = "swahili"
)

target_embeddings , info = wechsel . apply (
    source_tokenizer ,
    target_tokenizer ,
    model . get_input_embeddings (). weight . detach (). numpy (),
)

model . get_input_embeddings (). weight . data = torch . from_numpy ( target_embeddings )
model . config . vocab_size = len ( target_embeddings )

# if the model has separate output embeddings, also copy those
if not model . config . tie_word_embeddings :
    target_out_embeddings , info = wechsel . apply (
        source_tokenizer ,
        target_tokenizer ,
        model . get_output_embeddings (). weight . detach (). numpy (),
    )
    model . get_output_embeddings (). weight . data = torch . from_numpy ( target_out_embeddings )

# use `model` and `target_tokenizer` to continue training in Swahili! 

バイリンガル辞書

3276のバイリンガル辞書を英語から他の言語に配布して、 dicts/で使用するために使用します。

引用

wechselを引用してください

 @inproceedings{minixhofer-etal-2022-wechsel,
    title = "{WECHSEL}: Effective initialization of subword embeddings for cross-lingual transfer of monolingual language models",
    author = "Minixhofer, Benjamin  and
      Paischer, Fabian  and
      Rekabsaz, Navid",
    booktitle = "Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies",
    month = jul,
    year = "2022",
    address = "Seattle, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2022.naacl-main.293",
    pages = "3992--4006",
    abstract = "Large pretrained language models (LMs) have become the central building block of many NLP applications. Training these models requires ever more computational resources and most of the existing models are trained on English text only. It is exceedingly expensive to train these models in other languages. To alleviate this problem, we introduce a novel method {--} called WECHSEL {--} to efficiently and effectively transfer pretrained LMs to new languages. WECHSEL can be applied to any model which uses subword-based tokenization and learns an embedding for each subword. The tokenizer of the source model (in English) is replaced with a tokenizer in the target language and token embeddings are initialized such that they are semantically similar to the English tokens by utilizing multilingual static word embeddings covering English and the target language. We use WECHSEL to transfer the English RoBERTa and GPT-2 models to four languages (French, German, Chinese and Swahili). We also study the benefits of our method on very low-resource languages. WECHSEL improves over proposed methods for cross-lingual parameter transfer and outperforms models of comparable size trained from scratch with up to 64x less training effort. Our method makes training large language models for new languages more accessible and less damaging to the environment. We make our code and models publicly available.",
}

謝辞

GoogleのTPU Research Cloud（TRC）のクラウドTPUでサポートされている研究。追加の計算リソースを提供してくれたAndy KohとArtus Krohn-Grimbergheに感謝します。機械学習研究所であるLIT AIラボであるEllis Unit Linzは、連邦国家上部オーストリアによってサポートされています。プロジェクトIncentrol-RL（FFG-881064）に感謝します。