{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Models (TensorFlow)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Install the Transformers, Datasets, and Evaluate libraries to run this notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install datasets evaluate transformers[sentencepiece]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from transformers import BertConfig, TFBertModel\n",
    "\n",
    "# Building the config\n",
    "config = BertConfig()\n",
    "\n",
    "# Building the model from the config\n",
    "model = TFBertModel(config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "BertConfig {\n",
       "  [...]\n",
       "  \"hidden_size\": 768,\n",
       "  \"intermediate_size\": 3072,\n",
       "  \"max_position_embeddings\": 512,\n",
       "  \"num_attention_heads\": 12,\n",
       "  \"num_hidden_layers\": 12,\n",
       "  [...]\n",
       "}"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from transformers import BertConfig, TFBertModel\n",
    "\n",
    "config = BertConfig()\n",
    "model = TFBertModel(config)\n",
    "\n",
    "# Model is randomly initialized!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from transformers import TFBertModel\n",
    "\n",
    "model = TFBertModel.from_pretrained(\"bert-base-cased\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save_pretrained(\"directory_on_my_computer\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sequences = [\"Hello!\", \"Cool.\", \"Nice!\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "encoded_sequences = [\n",
    "    [101, 7592, 999, 102],\n",
    "    [101, 4658, 1012, 102],\n",
    "    [101, 3835, 999, 102],\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "\n",
    "model_inputs = tf.constant(encoded_sequences)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "output = model(model_inputs)"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "name": "Models (TensorFlow)",
   "provenance": []
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}