course/videos/tensorflow_finetuning.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook regroups the code sample of the video below, which is a part of the [Hugging Face course](https://huggingface.co/course)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "cellView": "form"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<iframe width=\"560\" height=\"315\" src=\"https://www.youtube.com/embed/AUozVp78dhk?rel=0&amp;controls=0&amp;showinfo=0\" frameborder=\"0\" allowfullscreen></iframe>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#@title\n",
    "from IPython.display import HTML\n",
    "\n",
    "HTML('<iframe width=\"560\" height=\"315\" src=\"https://www.youtube.com/embed/AUozVp78dhk?rel=0&amp;controls=0&amp;showinfo=0\" frameborder=\"0\" allowfullscreen></iframe>')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Install the Transformers and Datasets libraries to run this notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! pip install datasets transformers[sentencepiece]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook regroups the code sample of the video below, which is a part of the [Hugging Face course](https://huggingface.co/course)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "cellView": "form"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<iframe width=\"560\" height=\"315\" src=\"https://www.youtube.com/embed/alq1l8Lv9GA?rel=0&amp;controls=0&amp;showinfo=0\" frameborder=\"0\" allowfullscreen></iframe>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#@title\n",
    "from IPython.display import HTML\n",
    "\n",
    "HTML('<iframe width=\"560\" height=\"315\" src=\"https://www.youtube.com/embed/alq1l8Lv9GA?rel=0&amp;controls=0&amp;showinfo=0\" frameborder=\"0\" allowfullscreen></iframe>')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Install the Transformers and Datasets libraries to run this notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! pip install datasets transformers[sentencepiece]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from datasets import load_dataset\n",
    "from transformers import AutoTokenizer\n",
    "import numpy as np\n",
    "\n",
    "raw_datasets = load_dataset(\"glue\", \"mrpc\")\n",
    "checkpoint = \"bert-base-uncased\"\n",
    "tokenizer = AutoTokenizer.from_pretrained(checkpoint)\n",
    "\n",
    "def tokenize_dataset(dataset):\n",
    "    encoded = tokenizer(\n",
    "        dataset[\"sentence1\"],\n",
    "        dataset[\"sentence2\"],\n",
    "        max_length=128,\n",
    "        truncation=True,\n",
    "    )\n",
    "    return encoded.data\n",
    "\n",
    "tokenized_datasets = raw_datasets.map(tokenize_dataset, batched=True)\n",
    "\n",
    "train_dataset = tokenized_datasets[\"train\"].to_tf_dataset(\n",
    "    columns=[\"input_ids\", \"attention_mask\", \"token_type_ids\"],\n",
    "    label_cols=[\"label\"],\n",
    "    shuffle=True,\n",
    "    batch_size=8)\n",
    "\n",
    "validation_dataset = tokenized_datasets[\"validation\"].to_tf_dataset(\n",
    "    columns=[\"input_ids\", \"attention_mask\", \"token_type_ids\"],\n",
    "    label_cols=[\"label\"],\n",
    "    shuffle=True,\n",
    "    batch_size=8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "next(iter(train_dataset))[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from transformers import TFAutoModelForSequenceClassification\n",
    "\n",
    "checkpoint = 'bert-base-cased'\n",
    "model = TFAutoModelForSequenceClassification.from_pretrained(checkpoint, num_labels=2)\n",
    "loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True)\n",
    "model.compile(optimizer='adam', loss=loss)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.fit(\n",
    "    train_dataset,\n",
    "    validation_data=validation_dataset,\n",
    "    epochs=3\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "name": "Fine-Tuning with TensorFlow",
   "provenance": []
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}