{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "911a1be9",
   "metadata": {},
   "source": [
    "# 🤗 + 📚 dbmdz Turkish BERT model\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4f09b0f1",
   "metadata": {},
   "source": [
    "In this repository the MDZ Digital Library team (dbmdz) at the Bavarian State\n",
    "Library open sources a cased model for Turkish 🎉\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fa2a78a0",
   "metadata": {},
   "source": [
    "# 🇹🇷 BERTurk\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8b2f8c68",
   "metadata": {},
   "source": [
    "BERTurk is a community-driven cased BERT model for Turkish.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fe23e365",
   "metadata": {},
   "source": [
    "Some datasets used for pretraining and evaluation are contributed from the\n",
    "awesome Turkish NLP community, as well as the decision for the model name: BERTurk.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2da0ce24",
   "metadata": {},
   "source": [
    "## Stats\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d3f6af43",
   "metadata": {},
   "source": [
    "The current version of the model is trained on a filtered and sentence\n",
    "segmented version of the Turkish [OSCAR corpus](https://traces1.inria.fr/oscar/),\n",
    "a recent Wikipedia dump, various [OPUS corpora](http://opus.nlpl.eu/) and a\n",
    "special corpus provided by [Kemal Oflazer](http://www.andrew.cmu.edu/user/ko/).\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0d8d60c1",
   "metadata": {},
   "source": [
    "The final training corpus has a size of 35GB and 44,04,976,662 tokens.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ce42504f",
   "metadata": {},
   "source": [
    "For this model we use a vocab size of 128k.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4815bfdb",
   "metadata": {},
   "source": [
    "## Usage\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7a084604",
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install --upgrade paddlenlp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8d041c78",
   "metadata": {},
   "outputs": [],
   "source": [
    "import paddle\n",
    "from paddlenlp.transformers import AutoModel\n",
    "\n",
    "model = AutoModel.from_pretrained(\"dbmdz/bert-base-turkish-128k-cased\")\n",
    "input_ids = paddle.randint(100, 200, shape=[1, 20])\n",
    "print(model(input_ids))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "da82079c",
   "metadata": {},
   "source": [
    "# Reference"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b6632d46",
   "metadata": {},
   "source": [
    "> The model introduction and model weights originate from [https://huggingface.co/dbmdz/bert-base-turkish-128k-cased](https://huggingface.co/dbmdz/bert-base-turkish-128k-cased) and were converted to PaddlePaddle format for ease of use in PaddleNLP.\n"
   ]
  }
 ],
 "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
}