{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "85c2e1a7",
   "metadata": {},
   "source": [
    "# GerPT2\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "595fe7cb",
   "metadata": {},
   "source": [
    "See the GPT2 model card for considerations on limitations and bias. See the GPT2 documentation for details on GPT2.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5b4f950b",
   "metadata": {},
   "source": [
    "## Comparison to dbmdz/german-gpt2\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "95be6eb8",
   "metadata": {},
   "source": [
    "I evaluated both GerPT2-large and the other German GPT2, dbmdz/german-gpt2 on the [CC-100](http://data.statmt.org/cc-100/) dataset and on the German Wikipedia:\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8acd14be",
   "metadata": {},
   "source": [
    "|                   | CC-100 (PPL) | Wikipedia (PPL) |\n",
    "|-------------------|--------------|-----------------|\n",
    "| dbmdz/german-gpt2 | 49.47        | 62.92           |\n",
    "| GerPT2            | 24.78        | 35.33           |\n",
    "| GerPT2-large      | __16.08__    | __23.26__       |\n",
    "|                   |              |                 |\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6fa10d79",
   "metadata": {},
   "source": [
    "See the script `evaluate.py` in the [GerPT2 Github repository](https://github.com/bminixhofer/gerpt2) for the code.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a8514e1e",
   "metadata": {},
   "source": [
    "## Usage\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4bc62c63",
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install --upgrade paddlenlp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "63f78302",
   "metadata": {},
   "outputs": [],
   "source": [
    "import paddle\n",
    "from paddlenlp.transformers import AutoModel\n",
    "\n",
    "model = AutoModel.from_pretrained(\"benjamin/gerpt2-large\")\n",
    "input_ids = paddle.randint(100, 200, shape=[1, 20])\n",
    "print(model(input_ids))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "563152f3",
   "metadata": {},
   "source": [
    "```\n",
    "@misc{Minixhofer_GerPT2_German_large_2020,\n",
    "author = {Minixhofer, Benjamin},\n",
    "doi = {10.5281/zenodo.5509984},\n",
    "month = {12},\n",
    "title = {{GerPT2: German large and small versions of GPT2}},\n",
    "url = {https://github.com/bminixhofer/gerpt2},\n",
    "year = {2020}\n",
    "}\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0d67d21",
   "metadata": {},
   "source": [
    "## Acknowledgements\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "474c1c61",
   "metadata": {},
   "source": [
    "Thanks to [Hugging Face](https://huggingface.co) for awesome tools and infrastructure.\n",
    "Huge thanks to [Artus Krohn-Grimberghe](https://twitter.com/artuskg) at [LYTiQ](https://www.lytiq.de/) for making this possible by sponsoring the resources used for training.\n",
    "\n",
    "> The model introduction and model weights originate from [https://huggingface.co/benjamin/gerpt2-large](https://huggingface.co/benjamin/gerpt2-large) 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
}