readme: add initial version

04297ad almost 3 years ago

12.8 kB

	---
	language: multilingual
	license: mit
	widget:
	- text: "and I cannot conceive the reafon why [MASK] hath"
	- text: "Täkäläinen sanomalehdistö [MASK] erit - täin"
	- text: "Det vore [MASK] häller nödvändigt att be"
	- text: "Comme, à cette époque [MASK] était celle de la"
	- text: "In [MASK] an atmosphärischen Nahrungsmitteln"
	---

	# Historic Language Models (HLMs)

	## Languages

	Our Historic Language Models Zoo contains support for the following languages - incl. their training data source:

	\| Language \| Training data \| Size
	\| -------- \| ------------- \| ----
	\| German \| [Europeana](http://www.europeana-newspapers.eu/) \| 13-28GB (filtered)
	\| French \| [Europeana](http://www.europeana-newspapers.eu/) \| 11-31GB (filtered)
	\| English \| [British Library](https://data.bl.uk/digbks/db14.html) \| 24GB (year filtered)
	\| Finnish \| [Europeana](http://www.europeana-newspapers.eu/) \| 1.2GB
	\| Swedish \| [Europeana](http://www.europeana-newspapers.eu/) \| 1.1GB

	## Models

	At the moment, the following models are available on the model hub:

	\| Model identifier \| Model Hub link
	\| --------------------------------------------- \| --------------------------------------------------------------------------
	\| `dbmdz/bert-base-historic-multilingual-cased` \| [here](https://huggingface.co/dbmdz/bert-base-historic-multilingual-cased)
	\| `dbmdz/bert-base-historic-english-cased` \| [here](https://huggingface.co/dbmdz/bert-base-historic-english-cased)
	\| `dbmdz/bert-base-finnish-europeana-cased` \| [here](https://huggingface.co/dbmdz/bert-base-finnish-europeana-cased)
	\| `dbmdz/bert-base-swedish-europeana-cased` \| [here](https://huggingface.co/dbmdz/bert-base-swedish-europeana-cased)

	We also released smaller models for the multilingual model:

	\| Model identifier \| Model Hub link
	\| ----------------------------------------------- \| ---------------------------------------------------------------------------
	\| `dbmdz/bert-tiny-historic-multilingual-cased` \| [here](https://huggingface.co/dbmdz/bert-tiny-historic-multilingual-cased)
	\| `dbmdz/bert-mini-historic-multilingual-cased` \| [here](https://huggingface.co/dbmdz/bert-mini-historic-multilingual-cased)
	\| `dbmdz/bert-small-historic-multilingual-cased` \| [here](https://huggingface.co/dbmdz/bert-small-historic-multilingual-cased)
	\| `dbmdz/bert-medium-historic-multilingual-cased` \| [here](https://huggingface.co/dbmdz/bert-base-historic-multilingual-cased)

	Notice: We have released language models for Historic German and French trained on more noisier data earlier - see
	[this repo](https://github.com/stefan-it/europeana-bert) for more information:

	\| Model identifier \| Model Hub link
	\| --------------------------------------------- \| --------------------------------------------------------------------------
	\| `dbmdz/bert-base-german-europeana-cased` \| [here](https://huggingface.co/dbmdz/bert-base-german-europeana-cased)
	\| `dbmdz/bert-base-french-europeana-cased` \| [here](https://huggingface.co/dbmdz/bert-base-french-europeana-cased)

	# Corpora Stats

	## German Europeana Corpus

	We provide some statistics using different thresholds of ocr confidences, in order to shrink down the corpus size
	and use less-noisier data:

	\| OCR confidence \| Size
	\| -------------- \| ----
	\| 0.60 \| 28GB
	\| 0.65 \| 18GB
	\| 0.70 \| 13GB

	For the final corpus we use a OCR confidence of 0.6 (28GB). The following plot shows a tokens per year distribution:

	![German Europeana Corpus Stats](stats/figures/german_europeana_corpus_stats.png)

	## French Europeana Corpus

	Like German, we use different ocr confidence thresholds:

	\| OCR confidence \| Size
	\| -------------- \| ----
	\| 0.60 \| 31GB
	\| 0.65 \| 27GB
	\| 0.70 \| 27GB
	\| 0.75 \| 23GB
	\| 0.80 \| 11GB

	For the final corpus we use a OCR confidence of 0.7 (27GB). The following plot shows a tokens per year distribution:

	![French Europeana Corpus Stats](stats/figures/french_europeana_corpus_stats.png)

	## British Library Corpus

	Metadata is taken from [here](https://data.bl.uk/digbks/DB21.html). Stats incl. year filtering:

	\| Years \| Size
	\| ----------------- \| ----
	\| ALL \| 24GB
	\| >= 1800 && < 1900 \| 24GB

	We use the year filtered variant. The following plot shows a tokens per year distribution:

	![British Library Corpus Stats](stats/figures/bl_corpus_stats.png)

	## Finnish Europeana Corpus

	\| OCR confidence \| Size
	\| -------------- \| ----
	\| 0.60 \| 1.2GB

	The following plot shows a tokens per year distribution:

	![Finnish Europeana Corpus Stats](stats/figures/finnish_europeana_corpus_stats.png)

	## Swedish Europeana Corpus

	\| OCR confidence \| Size
	\| -------------- \| ----
	\| 0.60 \| 1.1GB

	The following plot shows a tokens per year distribution:

	![Swedish Europeana Corpus Stats](stats/figures/swedish_europeana_corpus_stats.png)

	## All Corpora

	The following plot shows a tokens per year distribution of the complete training corpus:

	![All Corpora Stats](stats/figures/all_corpus_stats.png)

	# Multilingual Vocab generation

	For the first attempt, we use the first 10GB of each pretraining corpus. We upsample both Finnish and Swedish to ~10GB.
	The following tables shows the exact size that is used for generating a 32k and 64k subword vocabs:

	\| Language \| Size
	\| -------- \| ----
	\| German \| 10GB
	\| French \| 10GB
	\| English \| 10GB
	\| Finnish \| 9.5GB
	\| Swedish \| 9.7GB

	We then calculate the subword fertility rate and portion of `[UNK]`s over the following NER corpora:

	\| Language \| NER corpora
	\| -------- \| ------------------
	\| German \| CLEF-HIPE, NewsEye
	\| French \| CLEF-HIPE, NewsEye
	\| English \| CLEF-HIPE
	\| Finnish \| NewsEye
	\| Swedish \| NewsEye

	Breakdown of subword fertility rate and unknown portion per language for the 32k vocab:

	\| Language \| Subword fertility \| Unknown portion
	\| -------- \| ------------------ \| ---------------
	\| German \| 1.43 \| 0.0004
	\| French \| 1.25 \| 0.0001
	\| English \| 1.25 \| 0.0
	\| Finnish \| 1.69 \| 0.0007
	\| Swedish \| 1.43 \| 0.0

	Breakdown of subword fertility rate and unknown portion per language for the 64k vocab:

	\| Language \| Subword fertility \| Unknown portion
	\| -------- \| ------------------ \| ---------------
	\| German \| 1.31 \| 0.0004
	\| French \| 1.16 \| 0.0001
	\| English \| 1.17 \| 0.0
	\| Finnish \| 1.54 \| 0.0007
	\| Swedish \| 1.32 \| 0.0

	# Final pretraining corpora

	We upsample Swedish and Finnish to ~27GB. The final stats for all pretraining corpora can be seen here:

	\| Language \| Size
	\| -------- \| ----
	\| German \| 28GB
	\| French \| 27GB
	\| English \| 24GB
	\| Finnish \| 27GB
	\| Swedish \| 27GB

	Total size is 130GB.

	# Smaller multilingual models

	Inspired by the ["Well-Read Students Learn Better: On the Importance of Pre-training Compact Models"](https://arxiv.org/abs/1908.08962)
	paper, we train smaller models (different layers and hidden sizes), and report number of parameters and pre-training costs:

	\| Model (Layer / Hidden size) \| Parameters \| Pre-Training time
	\| --------------------------- \| ----------: \| ----------------------:
	\| hmBERT Tiny ( 2/128) \| 4.58M \| 4.3 sec / 1,000 steps
	\| hmBERT Mini ( 4/256) \| 11.55M \| 10.5 sec / 1,000 steps
	\| hmBERT Small ( 4/512) \| 29.52M \| 20.7 sec / 1,000 steps
	\| hmBERT Medium ( 8/512) \| 42.13M \| 35.0 sec / 1,000 steps
	\| hmBERT Base (12/768) \| 110.62M \| 80.0 sec / 1,000 steps

	We then perform downstream evaluations on the multilingual [NewsEye](https://zenodo.org/record/4573313#.Ya3oVr-ZNzU) dataset:

	![NewsEye hmBERT Evaluation](stats/figures/newseye-hmbert-evaluation.png)

	# Pretraining

	## Multilingual model - hmBERT Base

	We train a multilingual BERT model using the 32k vocab with the official BERT implementation
	on a v3-32 TPU using the following parameters:

	```bash
	python3 run_pretraining.py --input_file gs://histolectra/historic-multilingual-tfrecords/*.tfrecord \
	--output_dir gs://histolectra/bert-base-historic-multilingual-cased \
	--bert_config_file ./config.json \
	--max_seq_length=512 \
	--max_predictions_per_seq=75 \
	--do_train=True \
	--train_batch_size=128 \
	--num_train_steps=3000000 \
	--learning_rate=1e-4 \
	--save_checkpoints_steps=100000 \
	--keep_checkpoint_max=20 \
	--use_tpu=True \
	--tpu_name=electra-2 \
	--num_tpu_cores=32
	```

	The following plot shows the pretraining loss curve:

	![Training loss curve](stats/figures/pretraining_loss_historic-multilingual.png)

	## Smaller multilingual models

	We use the same parameters as used for training the base model.

	### hmBERT Tiny

	The following plot shows the pretraining loss curve for the tiny model:

	![Training loss curve](stats/figures/pretraining_loss_hmbert-tiny.png)

	### hmBERT Mini

	The following plot shows the pretraining loss curve for the mini model:

	![Training loss curve](stats/figures/pretraining_loss_hmbert-mini.png)

	### hmBERT Small

	The following plot shows the pretraining loss curve for the small model:

	![Training loss curve](stats/figures/pretraining_loss_hmbert-small.png)

	### hmBERT Medium

	The following plot shows the pretraining loss curve for the medium model:

	![Training loss curve](stats/figures/pretraining_loss_hmbert-medium.png)

	## English model

	The English BERT model - with texts from British Library corpus - was trained with the Hugging Face
	JAX/FLAX implementation for 10 epochs (approx. 1M steps) on a v3-8 TPU, using the following command:

	```bash
	python3 run_mlm_flax.py --model_type bert \
	--config_name /mnt/datasets/bert-base-historic-english-cased/ \
	--tokenizer_name /mnt/datasets/bert-base-historic-english-cased/ \
	--train_file /mnt/datasets/bl-corpus/bl_1800-1900_extracted.txt \
	--validation_file /mnt/datasets/bl-corpus/english_validation.txt \
	--max_seq_length 512 \
	--per_device_train_batch_size 16 \
	--learning_rate 1e-4 \
	--num_train_epochs 10 \
	--preprocessing_num_workers 96 \
	--output_dir /mnt/datasets/bert-base-historic-english-cased-512-noadafactor-10e \
	--save_steps 2500 \
	--eval_steps 2500 \
	--warmup_steps 10000 \
	--line_by_line \
	--pad_to_max_length
	```

	The following plot shows the pretraining loss curve:

	![Training loss curve](stats/figures/pretraining_loss_historic_english.png)

	## Finnish model

	The BERT model - with texts from Finnish part of Europeana - was trained with the Hugging Face
	JAX/FLAX implementation for 40 epochs (approx. 1M steps) on a v3-8 TPU, using the following command:

	```bash
	python3 run_mlm_flax.py --model_type bert \
	--config_name /mnt/datasets/bert-base-finnish-europeana-cased/ \
	--tokenizer_name /mnt/datasets/bert-base-finnish-europeana-cased/ \
	--train_file /mnt/datasets/hlms/extracted_content_Finnish_0.6.txt \
	--validation_file /mnt/datasets/hlms/finnish_validation.txt \
	--max_seq_length 512 \
	--per_device_train_batch_size 16 \
	--learning_rate 1e-4 \
	--num_train_epochs 40 \
	--preprocessing_num_workers 96 \
	--output_dir /mnt/datasets/bert-base-finnish-europeana-cased-512-dupe1-noadafactor-40e \
	--save_steps 2500 \
	--eval_steps 2500 \
	--warmup_steps 10000 \
	--line_by_line \
	--pad_to_max_length
	```

	The following plot shows the pretraining loss curve:

	![Training loss curve](stats/figures/pretraining_loss_finnish_europeana.png)

	## Swedish model

	The BERT model - with texts from Swedish part of Europeana - was trained with the Hugging Face
	JAX/FLAX implementation for 40 epochs (approx. 660K steps) on a v3-8 TPU, using the following command:

	```bash
	python3 run_mlm_flax.py --model_type bert \
	--config_name /mnt/datasets/bert-base-swedish-europeana-cased/ \
	--tokenizer_name /mnt/datasets/bert-base-swedish-europeana-cased/ \
	--train_file /mnt/datasets/hlms/extracted_content_Swedish_0.6.txt \
	--validation_file /mnt/datasets/hlms/swedish_validation.txt \
	--max_seq_length 512 \
	--per_device_train_batch_size 16 \
	--learning_rate 1e-4 \
	--num_train_epochs 40 \
	--preprocessing_num_workers 96 \
	--output_dir /mnt/datasets/bert-base-swedish-europeana-cased-512-dupe1-noadafactor-40e \
	--save_steps 2500 \
	--eval_steps 2500 \
	--warmup_steps 10000 \
	--line_by_line \
	--pad_to_max_length
	```

	The following plot shows the pretraining loss curve:

	![Training loss curve](stats/figures/pretraining_loss_swedish_europeana.png)

	# Acknowledgments

	Research supported with Cloud TPUs from Google's TPU Research Cloud (TRC) program, previously known as
	TensorFlow Research Cloud (TFRC). Many thanks for providing access to the TRC ❤️

	Thanks to the generous support from the [Hugging Face](https://huggingface.co/) team,
	it is possible to download both cased and uncased models from their S3 storage 🤗