google-bert/bert-base-uncased

1---2language: en3tags:4- exbert5license: apache-2.06datasets:7- bookcorpus8- wikipedia9---10 11# BERT base model (uncased)12 13Pretrained model on English language using a masked language modeling (MLM) objective. It was introduced in14[this paper](https://arxiv.org/abs/1810.04805) and first released in15[this repository](https://github.com/google-research/bert). This model is uncased: it does not make a difference16between english and English.17 18Disclaimer: The team releasing BERT did not write a model card for this model so this model card has been written by19the Hugging Face team.20 21## Model description22 23BERT is a transformers model pretrained on a large corpus of English data in a self-supervised fashion. This means it24was pretrained on the raw texts only, with no humans labeling them in any way (which is why it can use lots of25publicly available data) with an automatic process to generate inputs and labels from those texts. More precisely, it26was pretrained with two objectives:27 28- Masked language modeling (MLM): taking a sentence, the model randomly masks 15% of the words in the input then run29  the entire masked sentence through the model and has to predict the masked words. This is different from traditional30  recurrent neural networks (RNNs) that usually see the words one after the other, or from autoregressive models like31  GPT which internally masks the future tokens. It allows the model to learn a bidirectional representation of the32  sentence.33- Next sentence prediction (NSP): the models concatenates two masked sentences as inputs during pretraining. Sometimes34  they correspond to sentences that were next to each other in the original text, sometimes not. The model then has to35  predict if the two sentences were following each other or not.36 37This way, the model learns an inner representation of the English language that can then be used to extract features38useful for downstream tasks: if you have a dataset of labeled sentences, for instance, you can train a standard39classifier using the features produced by the BERT model as inputs.40 41## Model variations42 43BERT has originally been released in base and large variations, for cased and uncased input text. The uncased models also strips out an accent markers.  44Chinese and multilingual uncased and cased versions followed shortly after.  45Modified preprocessing with whole word masking has replaced subpiece masking in a following work, with the release of two models.  46Other 24 smaller models are released afterward.  47 48The detailed release history can be found on the [google-research/bert readme](https://github.com/google-research/bert/blob/master/README.md) on github.49 50| Model | #params | Language |51|------------------------|--------------------------------|-------|52| [`bert-base-uncased`](https://huggingface.co/bert-base-uncased) | 110M   | English |53| [`bert-large-uncased`](https://huggingface.co/bert-large-uncased)              | 340M    | English | sub 54| [`bert-base-cased`](https://huggingface.co/bert-base-cased)        | 110M    | English |55| [`bert-large-cased`](https://huggingface.co/bert-large-cased) | 340M    |  English |56| [`bert-base-chinese`](https://huggingface.co/bert-base-chinese) | 110M    | Chinese |57| [`bert-base-multilingual-cased`](https://huggingface.co/bert-base-multilingual-cased) | 110M | Multiple |58| [`bert-large-uncased-whole-word-masking`](https://huggingface.co/bert-large-uncased-whole-word-masking) | 340M | English |59| [`bert-large-cased-whole-word-masking`](https://huggingface.co/bert-large-cased-whole-word-masking) | 340M | English |60 61## Intended uses & limitations62 63You can use the raw model for either masked language modeling or next sentence prediction, but it's mostly intended to64be fine-tuned on a downstream task. See the [model hub](https://huggingface.co/models?filter=bert) to look for65fine-tuned versions of a task that interests you.66 67Note that this model is primarily aimed at being fine-tuned on tasks that use the whole sentence (potentially masked)68to make decisions, such as sequence classification, token classification or question answering. For tasks such as text69generation you should look at model like GPT2.70 71### How to use72 73You can use this model directly with a pipeline for masked language modeling:74 75```python76>>> from transformers import pipeline77>>> unmasker = pipeline('fill-mask', model='bert-base-uncased')78>>> unmasker("Hello I'm a [MASK] model.")79 80[{'sequence': "[CLS] hello i'm a fashion model. [SEP]",81  'score': 0.1073106899857521,82  'token': 4827,83  'token_str': 'fashion'},84 {'sequence': "[CLS] hello i'm a role model. [SEP]",85  'score': 0.08774490654468536,86  'token': 2535,87  'token_str': 'role'},88 {'sequence': "[CLS] hello i'm a new model. [SEP]",89  'score': 0.05338378623127937,90  'token': 2047,91  'token_str': 'new'},92 {'sequence': "[CLS] hello i'm a super model. [SEP]",93  'score': 0.04667217284440994,94  'token': 3565,95  'token_str': 'super'},96 {'sequence': "[CLS] hello i'm a fine model. [SEP]",97  'score': 0.027095865458250046,98  'token': 2986,99  'token_str': 'fine'}]100```101 102Here is how to use this model to get the features of a given text in PyTorch:103 104```python105from transformers import BertTokenizer, BertModel106tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')107model = BertModel.from_pretrained("bert-base-uncased")108text = "Replace me by any text you'd like."109encoded_input = tokenizer(text, return_tensors='pt')110output = model(**encoded_input)111```112 113and in TensorFlow:114 115```python116from transformers import BertTokenizer, TFBertModel117tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')118model = TFBertModel.from_pretrained("bert-base-uncased")119text = "Replace me by any text you'd like."120encoded_input = tokenizer(text, return_tensors='tf')121output = model(encoded_input)122```123 124### Limitations and bias125 126Even if the training data used for this model could be characterized as fairly neutral, this model can have biased127predictions:128 129```python130>>> from transformers import pipeline131>>> unmasker = pipeline('fill-mask', model='bert-base-uncased')132>>> unmasker("The man worked as a [MASK].")133 134[{'sequence': '[CLS] the man worked as a carpenter. [SEP]',135  'score': 0.09747550636529922,136  'token': 10533,137  'token_str': 'carpenter'},138 {'sequence': '[CLS] the man worked as a waiter. [SEP]',139  'score': 0.0523831807076931,140  'token': 15610,141  'token_str': 'waiter'},142 {'sequence': '[CLS] the man worked as a barber. [SEP]',143  'score': 0.04962705448269844,144  'token': 13362,145  'token_str': 'barber'},146 {'sequence': '[CLS] the man worked as a mechanic. [SEP]',147  'score': 0.03788609802722931,148  'token': 15893,149  'token_str': 'mechanic'},150 {'sequence': '[CLS] the man worked as a salesman. [SEP]',151  'score': 0.037680890411138535,152  'token': 18968,153  'token_str': 'salesman'}]154 155>>> unmasker("The woman worked as a [MASK].")156 157[{'sequence': '[CLS] the woman worked as a nurse. [SEP]',158  'score': 0.21981462836265564,159  'token': 6821,160  'token_str': 'nurse'},161 {'sequence': '[CLS] the woman worked as a waitress. [SEP]',162  'score': 0.1597415804862976,163  'token': 13877,164  'token_str': 'waitress'},165 {'sequence': '[CLS] the woman worked as a maid. [SEP]',166  'score': 0.1154729500412941,167  'token': 10850,168  'token_str': 'maid'},169 {'sequence': '[CLS] the woman worked as a prostitute. [SEP]',170  'score': 0.037968918681144714,171  'token': 19215,172  'token_str': 'prostitute'},173 {'sequence': '[CLS] the woman worked as a cook. [SEP]',174  'score': 0.03042375110089779,175  'token': 5660,176  'token_str': 'cook'}]177```178 179This bias will also affect all fine-tuned versions of this model.180 181## Training data182 183The BERT model was pretrained on [BookCorpus](https://yknzhu.wixsite.com/mbweb), a dataset consisting of 11,038184unpublished books and [English Wikipedia](https://en.wikipedia.org/wiki/English_Wikipedia) (excluding lists, tables and185headers).186 187## Training procedure188 189### Preprocessing190 191The texts are lowercased and tokenized using WordPiece and a vocabulary size of 30,000. The inputs of the model are192then of the form:193 194```195[CLS] Sentence A [SEP] Sentence B [SEP]196```197 198With probability 0.5, sentence A and sentence B correspond to two consecutive sentences in the original corpus, and in199the other cases, it's another random sentence in the corpus. Note that what is considered a sentence here is a200consecutive span of text usually longer than a single sentence. The only constrain is that the result with the two201"sentences" has a combined length of less than 512 tokens.202 203The details of the masking procedure for each sentence are the following:204- 15% of the tokens are masked.205- In 80% of the cases, the masked tokens are replaced by `[MASK]`.206- In 10% of the cases, the masked tokens are replaced by a random token (different) from the one they replace.207- In the 10% remaining cases, the masked tokens are left as is.208 209### Pretraining210 211The model was trained on 4 cloud TPUs in Pod configuration (16 TPU chips total) for one million steps with a batch size212of 256. The sequence length was limited to 128 tokens for 90% of the steps and 512 for the remaining 10%. The optimizer213used is Adam with a learning rate of 1e-4, \\(\beta_{1} = 0.9\\) and \\(\beta_{2} = 0.999\\), a weight decay of 0.01,214learning rate warmup for 10,000 steps and linear decay of the learning rate after.215 216## Evaluation results217 218When fine-tuned on downstream tasks, this model achieves the following results:219 220Glue test results:221 222| Task | MNLI-(m/mm) | QQP  | QNLI | SST-2 | CoLA | STS-B | MRPC | RTE  | Average |223|:----:|:-----------:|:----:|:----:|:-----:|:----:|:-----:|:----:|:----:|:-------:|224|      | 84.6/83.4   | 71.2 | 90.5 | 93.5  | 52.1 | 85.8  | 88.9 | 66.4 | 79.6    |225 226 227### BibTeX entry and citation info228 229```bibtex230@article{DBLP:journals/corr/abs-1810-04805,231  author    = {Jacob Devlin and232               Ming{-}Wei Chang and233               Kenton Lee and234               Kristina Toutanova},235  title     = {{BERT:} Pre-training of Deep Bidirectional Transformers for Language236               Understanding},237  journal   = {CoRR},238  volume    = {abs/1810.04805},239  year      = {2018},240  url       = {http://arxiv.org/abs/1810.04805},241  archivePrefix = {arXiv},242  eprint    = {1810.04805},243  timestamp = {Tue, 30 Oct 2018 20:39:56 +0100},244  biburl    = {https://dblp.org/rec/journals/corr/abs-1810-04805.bib},245  bibsource = {dblp computer science bibliography, https://dblp.org}246}247```248 249<a href="https://huggingface.co/exbert/?model=bert-base-uncased">250	<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">251</a>252