Seed 24 MultiBERTs (pretrained BERT) model on English language using a masked language modeling (MLM) objective. It was introduced in
this paper and first released in
this repository. This model is uncased: it does not make a difference
between english and English.
Disclaimer: The team releasing MultiBERTs did not write a model card for this model so this model card has been written by gchhablani.
Model description
MultiBERTs models are transformers model pretrained on a large corpus of English data in a self-supervised fashion. This means it
was pretrained on the raw texts only, with no humans labelling them in any way (which is why it can use lots of
publicly available data) with an automatic process to generate inputs and labels from those texts. More precisely, it
was pretrained with two objectives:
Masked language modeling (MLM): taking a sentence, the model randomly masks 15% of the words in the input then run
the entire masked sentence through the model and has to predict the masked words. This is different from traditional
recurrent neural networks (RNNs) that usually see the words one after the other, or from autoregressive models like
GPT which internally mask the future tokens. It allows the model to learn a bidirectional representation of the
sentence.
Next sentence prediction (NSP): the models concatenates two masked sentences as inputs during pretraining. Sometimes
they correspond to sentences that were next to each other in the original text, sometimes not. The model then has to
predict if the two sentences were following each other or not.
This way, the model learns an inner representation of the English language that can then be used to extract features
useful for downstream tasks: if you have a dataset of labeled sentences for instance, you can train a standard
classifier using the features produced by the MultiBERTs model as inputs.
Intended uses & limitations
You can use the raw model for either masked language modeling or next sentence prediction, but it's mostly intended to
be fine-tuned on a downstream task. See the model hub to look for
fine-tuned versions on a task that interests you.
Note that this model is primarily aimed at being fine-tuned on tasks that use the whole sentence (potentially masked)
to make decisions, such as sequence classification, token classification or question answering. For tasks such as text
generation you should look at model like GPT2.
How to use
Here is how to use this model to get the features of a given text in PyTorch:
python
1from transformers import BertTokenizer, BertModel
2tokenizer = BertTokenizer.from_pretrained('multiberts-seed-24')3model = BertModel.from_pretrained("multiberts-seed-24")4text ="Replace me by any text you'd like."5encoded_input = tokenizer(text, return_tensors='pt')6output = model(**encoded_input)
Limitations and bias
Even if the training data used for this model could be characterized as fairly neutral, this model can have biased
predictions. This bias will also affect all fine-tuned versions of this model. For an understanding of bias of this particular
checkpoint, please try out this checkpoint with the snippet present in the Limitation and bias section of the bert-base-uncased checkpoint.
Training data
The MultiBERTs models were pretrained on BookCorpus, a dataset consisting of 11,038
unpublished books and English Wikipedia (excluding lists, tables and
headers).
Training procedure
Preprocessing
The texts are lowercased and tokenized using WordPiece and a vocabulary size of 30,000. The inputs of the model are
then of the form:
[CLS] Sentence A [SEP] Sentence B [SEP]
With probability 0.5, sentence A and sentence B correspond to two consecutive sentences in the original corpus and in
the other cases, it's another random sentence in the corpus. Note that what is considered a sentence here is a
consecutive span of text usually longer than a single sentence. The only constrain is that the result with the two
"sentences" has a combined length of less than 512 tokens.
The details of the masking procedure for each sentence are the following:
15% of the tokens are masked.
In 80% of the cases, the masked tokens are replaced by [MASK].
In 10% of the cases, the masked tokens are replaced by a random token (different) from the one they replace.
In the 10% remaining cases, the masked tokens are left as is.
Pretraining
The model was trained on 16 Cloud TPU v2 chips for two million steps with a batch size
of 256. The sequence length was set to 512 throughout. The optimizer
used is Adam with a learning rate of 1e-4, \(\beta_{1} = 0.9\) and \(\beta_{2} = 0.999\), a weight decay of 0.01,
learning rate warmup for 10,000 steps and linear decay of the learning rate after.
BibTeX entry and citation info
bibtex
1@article{DBLP:journals/corr/abs-2106-16163,
2 author = {Thibault Sellam and
3 Steve Yadlowsky and
4 Jason Wei and
5 Naomi Saphra and
6 Alexander D'Amour and
7 Tal Linzen and
8 Jasmijn Bastings and
9 Iulia Turc and
10 Jacob Eisenstein and
11 Dipanjan Das and
12 Ian Tenney and
13 Ellie Pavlick},
14 title = {The MultiBERTs: {BERT} Reproductions for Robustness Analysis},
15 journal = {CoRR},
16 volume = {abs/2106.16163},
17 year = {2021},
18 url = {https://arxiv.org/abs/2106.16163},
19 eprinttype = {arXiv},
20 eprint = {2106.16163},
21 timestamp = {Mon, 05 Jul 2021 15:15:50 +0200},
22 biburl = {https://dblp.org/rec/journals/corr/abs-2106-16163.bib},
23 bibsource = {dblp computer science bibliography, https://dblp.org}
24}
