deepfocus 1.0.1

Offcial Python implementation of "FOCUS: Effective Embedding Initialization for Monolingual Specialization of Multilingual Models" published at EMNLP 2023.

FOCUS

Code for "FOCUS: Effective Embedding Initialization for Monolingual Specialization of Multilingual Models" accepted at the EMNLP 2023 main conference.

Paper on arXiv: https://arxiv.org/abs/2305.14481.

Installation

We provide the package via pip install deepfocus.

Alternatively, you can simply copy the deepfocus folder and drop it into your project. The necessary dependencies are listed in requirements.txt (pip install -r requirements.txt).

Usage

The following example shows how to use FOCUS to specialize xlm-roberta-base on German with a custom, language-specific tokenizer. The code is also available in focus_example.py.

from transformers import AutoModelForMaskedLM, AutoTokenizer
from deepfocus import FOCUS

source_tokenizer = AutoTokenizer.from_pretrained("xlm-roberta-base")
source_model = AutoModelForMaskedLM.from_pretrained("xlm-roberta-base")

target_tokenizer = AutoTokenizer.from_pretrained(
    "./tokenizers/de/xlmr-unigram-50k"
)

# Example for training a new tokenizer:
# target_tokenizer = source_tokenizer.train_new_from_iterator(
#     load_dataset("cc100", lang="de", split="train")["text"],
#     vocab_size=50_048
# )
# target_tokenizer.save_pretrained("./target_tokenizer_test")

target_embeddings = FOCUS(
    source_embeddings=source_model.get_input_embeddings().weight,
    source_tokenizer=source_tokenizer,
    target_tokenizer=target_tokenizer,
    target_training_data_path="/path/to/data.txt"
    # fasttext_model_path="/path/to/fasttext.bin", # or directly provide path to token-level fasttext model 

    # In the paper, we use `target_training_data_path` but we also implement using
    # WECHSEL's word-to-subword mapping if the language has pretrained fasttext word embeddings available online
    # To use, supply a two-letter `language_identifier` (e.g. "de" for German) instead of `target_training_data_path` and set:
    # auxiliary_embedding_mode="fasttext-wordlevel",
    # language_identifier="de",

)
source_model.resize_token_embeddings(len(target_tokenizer))
source_model.get_input_embeddings().weight.data = target_embeddings

# Continue training the model on the target language with `target_tokenizer`.
# ...

Checkpoints

We publish the checkpoints trained with FOCUS on HuggingFace:

Language	Vocabulary Replacement (preferred)	Vocabulary Extension
German	konstantindobler/xlm-roberta-base-focus-german	konstantindobler/xlm-roberta-base-focus-extend-german
Arabic	konstantindobler/xlm-roberta-base-focus-arabic	konstantindobler/xlm-roberta-base-focus-extend-arabic
Kiswahili	konstantindobler/xlm-roberta-base-focus-kiswahili	konstantindobler/xlm-roberta-base-focus-extend-kiswahili
Hausa	konstantindobler/xlm-roberta-base-focus-hausa	konstantindobler/xlm-roberta-base-focus-extend-hausa
isiXhosa	konstantindobler/xlm-roberta-base-focus-isixhosa	konstantindobler/xlm-roberta-base-focus-extend-isixhosa

In our experiments, full vocabulary replacement coupled with FOCUS outperformed extending XLM-R's original vocabulary, while also resulting in a smaller model and being faster to train.

Citation

You can cite FOCUS like this:

@misc{dobler-demelo-2023-focus,
    title={FOCUS: Effective Embedding Initialization for Monolingual Specialization of Multilingual Models},
    author={Konstantin Dobler and Gerard de Melo},
    year={2023},
    eprint={2305.14481},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}

If you use the "WECHSEL-style" word-to-subword mapping, please consider also citing their original work.