{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "ff0c69a5",
   "metadata": {},
   "source": [
    "# BERT multilingual base model (uncased)\n",
    "\n",
    "You can get more details from [Bert in PaddleNLP](https://github.com/PaddlePaddle/PaddleNLP/blob/develop/model_zoo/bert/README.md)。"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1ad499a9",
   "metadata": {},
   "source": [
    "Pretrained model on the top 102 languages with the largest Wikipedia using a masked language modeling (MLM) objective.\n",
    "It was introduced in [this paper](https://arxiv.org/abs/1810.04805) and first released in\n",
    "[this repository](https://github.com/google-research/bert). This model is uncased: it does not make a difference\n",
    "between english and English.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a8878d0c",
   "metadata": {},
   "source": [
    "Disclaimer: The team releasing BERT did not write a model card for this model so this model card has been written by\n",
    "the Hugging Face team.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4581e670",
   "metadata": {},
   "source": [
    "## Model description\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c8d5f59f",
   "metadata": {},
   "source": [
    "BERT is a transformers model pretrained on a large corpus of multilingual data in a self-supervised fashion. This means\n",
    "it was pretrained on the raw texts only, with no humans labelling them in any way (which is why it can use lots of\n",
    "publicly available data) with an automatic process to generate inputs and labels from those texts. More precisely, it\n",
    "was pretrained with two objectives:\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "836834df",
   "metadata": {},
   "source": [
    "- Masked language modeling (MLM): taking a sentence, the model randomly masks 15% of the words in the input then run\n",
    "the entire masked sentence through the model and has to predict the masked words. This is different from traditional\n",
    "recurrent neural networks (RNNs) that usually see the words one after the other, or from autoregressive models like\n",
    "GPT which internally mask the future tokens. It allows the model to learn a bidirectional representation of the\n",
    "sentence.\n",
    "- Next sentence prediction (NSP): the models concatenates two masked sentences as inputs during pretraining. Sometimes\n",
    "they correspond to sentences that were next to each other in the original text, sometimes not. The model then has to\n",
    "predict if the two sentences were following each other or not.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bafe70e4",
   "metadata": {},
   "source": [
    "This way, the model learns an inner representation of the languages in the training set that can then be used to\n",
    "extract features useful for downstream tasks: if you have a dataset of labeled sentences for instance, you can train a\n",
    "standard classifier using the features produced by the BERT model as inputs.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cf2a29e2",
   "metadata": {},
   "source": [
    "## How to use"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dc792d6e",
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install --upgrade paddlenlp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7a6faf50",
   "metadata": {},
   "outputs": [],
   "source": [
    "import paddle\n",
    "from paddlenlp.transformers import AutoModel\n",
    "\n",
    "model = AutoModel.from_pretrained(\"bert-base-multilingual-uncased\")\n",
    "input_ids = paddle.randint(100, 200, shape=[1, 20])\n",
    "print(model(input_ids))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5b616f23",
   "metadata": {},
   "source": [
    "```\n",
    "@article{DBLP:journals/corr/abs-1810-04805,\n",
    "author    = {Jacob Devlin and\n",
    "Ming{-}Wei Chang and\n",
    "Kenton Lee and\n",
    "Kristina Toutanova},\n",
    "title     = {{BERT:} Pre-training of Deep Bidirectional Transformers for Language\n",
    "Understanding},\n",
    "journal   = {CoRR},\n",
    "volume    = {abs/1810.04805},\n",
    "year      = {2018},\n",
    "url       = {http://arxiv.org/abs/1810.04805},\n",
    "archivePrefix = {arXiv},\n",
    "eprint    = {1810.04805},\n",
    "timestamp = {Tue, 30 Oct 2018 20:39:56 +0100},\n",
    "biburl    = {https://dblp.org/rec/journals/corr/abs-1810-04805.bib},\n",
    "bibsource = {dblp computer science bibliography, https://dblp.org}\n",
    "}\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "67e01093",
   "metadata": {},
   "source": [
    "> The model introduction and model weights originate from  https://huggingface.co/bert-base-multilingual-uncased  and were converted to PaddlePaddle format for ease of use in PaddleNLP."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}