From 683635523b581aa9b2d114e92b64966cd55cc5b6 Mon Sep 17 00:00:00 2001
From: HyunZ118 <156191095+HyunZ118@users.noreply.github.com>
Date: Fri, 30 May 2025 20:18:35 +0900
Subject: [PATCH 1/2] docs: add fine_tuning_vit_custom_dataset_ko

---
 notebooks/ko/_toctree.yml                     |    5 +
 .../fine_tuning_vit_custom_dataset_ko.ipynb   | 4367 +++++++++++++++++
 notebooks/ko/index.md                         |    1 +
 3 files changed, 4373 insertions(+)
 create mode 100644 notebooks/ko/fine_tuning_vit_custom_dataset_ko.ipynb

diff --git a/notebooks/ko/_toctree.yml b/notebooks/ko/_toctree.yml
index d420590c..c9a27475 100644
--- a/notebooks/ko/_toctree.yml
+++ b/notebooks/ko/_toctree.yml
@@ -15,6 +15,11 @@
           title: 지식 그래프를 활용한 RAG 추론 향상
         - local: rag_zephyr_langchain
           title: GitHub 이슈를 위한 EEVE와 LangChain을 사용한 간단한 RAG
+    - title: 컴퓨터 비전 레시피
+      isExpanded: false
+      sections:
+        - local: fine_tuning_vit_custom_dataset_ko
+          title: Custom Biomedical Dataset으로 Vision Transformer 모델 Fine-tuning 하기  
     - title: 에이전트 레시피
       isExpanded: false
       sections:
diff --git a/notebooks/ko/fine_tuning_vit_custom_dataset_ko.ipynb b/notebooks/ko/fine_tuning_vit_custom_dataset_ko.ipynb
new file mode 100644
index 00000000..fc5712ed
--- /dev/null
+++ b/notebooks/ko/fine_tuning_vit_custom_dataset_ko.ipynb
@@ -0,0 +1,4367 @@
+{
+  "cells": [
+    {
+      "cell_type": "markdown",
+      "id": "97bf8340-2c4f-4b32-9a64-5b8ed2d6247f",
+      "metadata": {
+        "id": "97bf8340-2c4f-4b32-9a64-5b8ed2d6247f"
+      },
+      "source": [
+        "# Custom Biomedical Dataset으로 Vision Transformer 모델 Fine-tuning 하기\n",
+        "_작성자 : [Emre Albayrak](https://github.com/emre570)_\n",
+        "\n",
+        "_번역자: [Hyun Ji Jeon](https://github.com/HyunZ118)_\n",
+        "\n",
+        "&nbsp;\n",
+        "\n",
+        "이 가이드는 Vision Transformer(ViT) 모델을 Custom Biomedical Dataset 에 Fine-tuning하는 과정을 설명합니다. 이 과정에서 데이터셋 로딩 및 전처리, 데이터 분할에 따른 이미지 변환 설정, ViT 모델 구성 및 초기화, 평가 및 시각화 도구를 포함한 학습 과정 정의 단계가 포함됩니다.  \n",
+        "\n",
+        "&nbsp;\n",
+        "## 데이터셋 정보\n",
+        "Custom Dataset은 수작업으로 제작되었으며, 총 780장의 이미지와 benign(양성), malignant(악성), normal(정상) 총 3개의 클래스를 포함합니다.\n",
+        "\n",
+        "![attachment:datasetinfo.png](https://huggingface.co/datasets/huggingface/cookbook-images/resolve/102d6c23e6cc24db857fbc60186461ded6cdfb75/datasetinfo.png)\n",
+        "\n",
+        "## 모델 정보\n",
+        "우리가 Fine-tuning 할 모델은 Google의 [`\"vit-large-patch16-224\"`](https://huggingface.co/google/vit-large-patch16-224)입니다. 이 모델은 ImageNet-21k(1,400만 장의 이미지, 21,843개의 클래스)으로 사전학습 되었으며, ImageNet 2012(100만 장의 이미지, 1,000개 클래스)로 해상도 224x224에서 Fine-tuning 되었습니다. Google은 이외에도 다양한 이미지 크기와 패치 크기를 가진 여러 ViT 모델을 보유하고 있습니다.\n",
+        "\n",
+        "\n",
+        "그럼 이제 시작해보겠습니다! 🤗"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "3cc02613-7bc6-4cd8-aa97-a21ba1970027",
+      "metadata": {
+        "id": "3cc02613-7bc6-4cd8-aa97-a21ba1970027"
+      },
+      "source": [
+        "## 시작\n",
+        "먼저 필요한 라이브러리들을 설치해야합니다."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 1,
+      "id": "7093dd4f-d0cb-44dc-935d-d54435187901",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "7093dd4f-d0cb-44dc-935d-d54435187901",
+        "outputId": "4dee826f-3841-4574-d4c2-7530a0c98dd2"
+      },
+      "outputs": [],
+      "source": [
+        "# !pip install datasets transformers accelerate torch torchvision scikit-learn matplotlib wandb"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "9b5019a8-d130-4c08-9503-cd8415f50ae9",
+      "metadata": {
+        "id": "9b5019a8-d130-4c08-9503-cd8415f50ae9"
+      },
+      "source": [
+        "(선택 사항) 모델을 Hugging Face Hub에 업로드할 예정이므로 먼저 로그인해야 합니다."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 2,
+      "id": "3d5acb41-a225-44a5-8c8f-f212c615008f",
+      "metadata": {
+        "colab": {
+          "referenced_widgets": [
+            "4a75a73de5234297a7e0e4e070eee6d9"
+          ]
+        },
+        "id": "3d5acb41-a225-44a5-8c8f-f212c615008f",
+        "outputId": "f2ceb9d4-4f4c-4ef4-b629-e929a8ae8108"
+      },
+      "outputs": [],
+      "source": [
+        "#from huggingface_hub import notebook_login\n",
+        "#notebook_login()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "80f2b730-ff8a-4d76-9baa-41c35b80fd38",
+      "metadata": {
+        "id": "80f2b730-ff8a-4d76-9baa-41c35b80fd38"
+      },
+      "source": [
+        "## 데이터셋 준비\n",
+        "'datasets' 라이브러리는 데이터셋에서 이미지와 클래스를 자동으로 불러옵니다. 자세한 내용은 [`이 링크`](https://huggingface.co/docs/datasets/image_load).를 참고하세요."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "id": "C2f-_NcrnYil",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "C2f-_NcrnYil",
+        "outputId": "27112e94-72ec-4dba-d848-a870d6699737"
+      },
+      "outputs": [],
+      "source": [
+        "# pip install --upgrade datasets # update by HJ"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 4,
+      "id": "f6e69664-1bb0-4ee4-932c-2f108550bfba",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 484,
+          "referenced_widgets": [
+            "27297aa211ef4564bcb9a23bd6e073c5",
+            "28e427d0fb3c42c0b9d55f9076f18918",
+            "9155b025badd40c284c07fbc87175169",
+            "40c47070f79b48d088622696128531cb",
+            "41a800c6e6c547a9ab528ee96e1ce137",
+            "43c0527631a64c89b171c2a0664dfa8b",
+            "95d5ad27410b4a689734c1593f71d02b",
+            "99893feb6a184879891978d111ebde49",
+            "fe8fca891aff479a9fb11aeeb1971e2d",
+            "8e67e87040944a878618f01a98294681",
+            "f3267441b9dc4f029a3bfdffaf491842",
+            "ff9d81edcfe9421bb1398ca284a197cc",
+            "ef44e725caed4ef38e9cc25621b76406",
+            "4358acace42c41519fc83cc67e814961",
+            "d0422dfef6a846c5bbd9045b7019ffb8",
+            "228c08de964d48fdbb3594c1f2ecde09",
+            "5aa5164fa3d040bc8180b122eb029d94",
+            "04fb4c819c3c44b59184c54ea44456bf",
+            "aa9365c6057a4137bb678ce4d05c448d",
+            "741d79cc18ae44259621f030e454ab99",
+            "039241d8c1c34103a655fbcb6de51ae6",
+            "fbcd9fb015d24543aedb6dbdb8e3d87a",
+            "09d1ea0c40184e648c75b10ee22f23e7",
+            "4b411454d68a436cbf04d8a97e2fe7da",
+            "a2f5fb56b54f4b7383fd85aed9920bcb",
+            "2a4a870ca4404c23bd92f9b1117d9a0e",
+            "c9077c3acb6342f2a248876f3ebf7464",
+            "c99b9fb60a4042b58b07a8524f8f555e",
+            "2f48026c77c9415292364602bff4ca18",
+            "c162aa27a92f496192a2335ab0496bc8",
+            "4901091780124b3dbba924779868f4c8",
+            "482d168e5e104fd48c5c762450c02930",
+            "466dfa0cc1344aa58bb1a4a8b8cf9aba",
+            "0fe075d0db0445aeb7e4d275860d39f2",
+            "64b29e54431246e98aeadf09c705b684",
+            "323db282be654fdeb25bf9a5b9f1733b",
+            "2d7ebc94bae04034aee3ee10f800c1df",
+            "33694de53716463b8c5b07720a40e067",
+            "566b4b97be32448fb502520a93976501",
+            "c83d52312d02407a8ea124fa25ebddb7",
+            "1eccd922da6e4b1682bdb061716ed27b",
+            "9b238f957be24a8e99b46baaa664cb5c",
+            "1173f9c2411640078fb20331845d2c87",
+            "db8dd16fc27a45d5a1ab81380b305389",
+            "f2566e5e755e4a72acbebab9c938aa34",
+            "d14c61501c7d49ce90cb3c1d073cbaa8",
+            "ee4079637e6643d48a9feba34c30d0d7",
+            "2ddd4fe77259428aaf7d6a56ed0b223e",
+            "a86b7af977bc40fea33a27748e57f432",
+            "d91cdc1965a74b7790e44126f204c708",
+            "2d028e6a22194fa28419e88e87573421",
+            "4113c7318d5d4484b3990dcbdb9bc1cd",
+            "1b8330ce7031406c95eac80280936dae",
+            "387d13f078f747fea3c747f55e0458f5",
+            "8454a7ef8c1140a78f544465577d7532"
+          ]
+        },
+        "id": "f6e69664-1bb0-4ee4-932c-2f108550bfba",
+        "outputId": "4f611cc7-ea8b-4d22-f6e0-1c2d2fb698bb"
+      },
+      "outputs": [
+        {
+          "data": {
+            "text/plain": [
+              "DatasetDict({\n",
+              "    train: Dataset({\n",
+              "        features: ['image', 'label'],\n",
+              "        num_rows: 624\n",
+              "    })\n",
+              "    test: Dataset({\n",
+              "        features: ['image', 'label'],\n",
+              "        num_rows: 156\n",
+              "    })\n",
+              "})"
+            ]
+          },
+          "execution_count": 4,
+          "metadata": {},
+          "output_type": "execute_result"
+        }
+      ],
+      "source": [
+        "from datasets import load_dataset\n",
+        "\n",
+        "dataset = load_dataset(\"emre570/breastcancer-ultrasound-images\")\n",
+        "dataset"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "3b0461c4",
+      "metadata": {},
+      "source": []
+    },
+    {
+      "cell_type": "markdown",
+      "id": "bf3176b2-65ff-44ea-bb1e-cd75b2d7609d",
+      "metadata": {
+        "id": "bf3176b2-65ff-44ea-bb1e-cd75b2d7609d"
+      },
+      "source": [
+        "데이터셋을 불러왔지만, 현재 검증 세트는 없습니다.\n",
+        "\n",
+        "검증 세트를 만들기 위해 테스트 세트의 크기를 기준으로 훈련 세트의 일부를 나누어 검증 세트를 생성할 것입니다.\n",
+        "\n",
+        "그 후 기존 훈련 세트를 새로운 훈련 세트와 검증 세트로 분할합니다."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 5,
+      "id": "4d385bab-dad0-462d-9339-11205fcd2e63",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "4d385bab-dad0-462d-9339-11205fcd2e63",
+        "outputId": "a634094c-6de2-4c55-b9f6-decd620236f8"
+      },
+      "outputs": [
+        {
+          "data": {
+            "text/plain": [
+              "DatasetDict({\n",
+              "    train: Dataset({\n",
+              "        features: ['image', 'label'],\n",
+              "        num_rows: 468\n",
+              "    })\n",
+              "    test: Dataset({\n",
+              "        features: ['image', 'label'],\n",
+              "        num_rows: 156\n",
+              "    })\n",
+              "})"
+            ]
+          },
+          "execution_count": 5,
+          "metadata": {},
+          "output_type": "execute_result"
+        }
+      ],
+      "source": [
+        "# Get the numbers of each set\n",
+        "test_num = len(dataset[\"test\"])\n",
+        "train_num = len(dataset[\"train\"])\n",
+        "\n",
+        "val_size = test_num / train_num\n",
+        "\n",
+        "train_val_split = dataset[\"train\"].train_test_split(test_size=val_size)\n",
+        "train_val_split"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "7dacfc60-6262-4e91-b9e9-e2af43210fde",
+      "metadata": {
+        "id": "7dacfc60-6262-4e91-b9e9-e2af43210fde"
+      },
+      "source": [
+        "분할된 훈련 세트를 준비했습니다. 이제 이를 테스트 세트와 함께 병합해보겠습니다."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 6,
+      "id": "b43bf130-49ad-4a4d-8230-436649045a16",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "b43bf130-49ad-4a4d-8230-436649045a16",
+        "outputId": "151731e9-6502-43b3-d198-1b64404d73ac"
+      },
+      "outputs": [
+        {
+          "data": {
+            "text/plain": [
+              "DatasetDict({\n",
+              "    train: Dataset({\n",
+              "        features: ['image', 'label'],\n",
+              "        num_rows: 468\n",
+              "    })\n",
+              "    validation: Dataset({\n",
+              "        features: ['image', 'label'],\n",
+              "        num_rows: 156\n",
+              "    })\n",
+              "    test: Dataset({\n",
+              "        features: ['image', 'label'],\n",
+              "        num_rows: 156\n",
+              "    })\n",
+              "})"
+            ]
+          },
+          "execution_count": 6,
+          "metadata": {},
+          "output_type": "execute_result"
+        }
+      ],
+      "source": [
+        "from datasets import DatasetDict\n",
+        "\n",
+        "dataset = DatasetDict({\n",
+        "    \"train\": train_val_split[\"train\"],\n",
+        "    \"validation\": train_val_split[\"test\"],\n",
+        "    \"test\": dataset[\"test\"]\n",
+        "})\n",
+        "dataset"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "4a0d7a89-099e-4925-97a5-51c91cdb7046",
+      "metadata": {
+        "id": "4a0d7a89-099e-4925-97a5-51c91cdb7046"
+      },
+      "source": [
+        "잘하셨습니다! 이제 데이터셋 준비가 완료되었습니다.\n",
+        "\n",
+        "각 하위 데이터셋을 서로 다른 변수에 할당해봅시다. (나중에 쉽게 참조할 수 있도록 하기 위함입니다.)"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 7,
+      "id": "9462e82b-0a12-4b0c-b73e-8dba76041ee6",
+      "metadata": {
+        "id": "9462e82b-0a12-4b0c-b73e-8dba76041ee6"
+      },
+      "outputs": [],
+      "source": [
+        "train_ds = dataset['train']\n",
+        "val_ds = dataset['validation']\n",
+        "test_ds = dataset['test']"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "27e5451f-e2fb-4d29-b63a-a8960686dcc2",
+      "metadata": {
+        "id": "27e5451f-e2fb-4d29-b63a-a8960686dcc2"
+      },
+      "source": [
+        "이미지를 확인해보면 'PIL.Image' 형식이며 이에 연결된 라벨이 함께 존재하는 것을 확인할 수 있습니다."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "id": "1d8ffe26-f128-4f63-962e-56349d45ff23",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "1d8ffe26-f128-4f63-962e-56349d45ff23",
+        "outputId": "411f3640-4150-44ae-e05e-bee054cbc606"
+      },
+      "outputs": [
+        {
+          "data": {
+            "text/plain": [
+              "{'image': <PIL.PngImagePlugin.PngImageFile image mode=RGB size=558x475>,\n",
+              " 'label': 1}"
+            ]
+          },
+          "execution_count": 8,
+          "metadata": {},
+          "output_type": "execute_result"
+        }
+      ],
+      "source": [
+        "train_ds[0]"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "384a09b0-1c47-411f-b91a-00acdd88b06b",
+      "metadata": {
+        "id": "384a09b0-1c47-411f-b91a-00acdd88b06b"
+      },
+      "source": [
+        "훈련 세트의 특성도 확인할 수 있습니다."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 9,
+      "id": "1e09647b-44e9-4f5f-800c-c333b0523b85",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "1e09647b-44e9-4f5f-800c-c333b0523b85",
+        "outputId": "287da3c7-0fe9-498f-d44c-ca7b9c9701ae"
+      },
+      "outputs": [
+        {
+          "data": {
+            "text/plain": [
+              "{'image': Image(mode=None, decode=True, id=None),\n",
+              " 'label': ClassLabel(names=['benign', 'malignant', 'normal'], id=None)}"
+            ]
+          },
+          "execution_count": 9,
+          "metadata": {},
+          "output_type": "execute_result"
+        }
+      ],
+      "source": [
+        "train_ds.features"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "a3fedab5-0e80-492c-9408-f629b230351d",
+      "metadata": {
+        "id": "a3fedab5-0e80-492c-9408-f629b230351d"
+      },
+      "source": [
+        "데이터셋에서 각 클래스별로 하나씩 이미지를 출력해봅시다."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 10,
+      "id": "5c901865-a876-4b4b-b1f2-8895b494cafb",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 248
+        },
+        "id": "5c901865-a876-4b4b-b1f2-8895b494cafb",
+        "outputId": "d9c1aef7-a4a4-427c-d902-63ef790d7c21"
+      },
+      "outputs": [
+        {
+          "data": {
+            "image/png": "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",
+            "text/plain": [
+              "<Figure size 1000x1000 with 3 Axes>"
+            ]
+          },
+          "metadata": {},
+          "output_type": "display_data"
+        }
+      ],
+      "source": [
+        "import matplotlib.pyplot as plt\n",
+        "\n",
+        "# Initialize a set to keep track of shown labels\n",
+        "shown_labels = set()\n",
+        "\n",
+        "# Initialize the figure for plotting\n",
+        "plt.figure(figsize=(10, 10))\n",
+        "\n",
+        "# Loop through the dataset and plot the first image of each label\n",
+        "for i, sample in enumerate(train_ds):\n",
+        "    label = train_ds.features['label'].names[sample['label']]\n",
+        "    if label not in shown_labels:\n",
+        "        plt.subplot(1, len(train_ds.features['label'].names), len(shown_labels) + 1)\n",
+        "        plt.imshow(sample['image'])\n",
+        "        plt.title(label)\n",
+        "        plt.axis('off')\n",
+        "        shown_labels.add(label)\n",
+        "        if len(shown_labels) == len(train_ds.features['label'].names):\n",
+        "            break\n",
+        "\n",
+        "plt.show()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "0300a4b7-4f8f-4155-bad8-42df6673eddc",
+      "metadata": {
+        "id": "0300a4b7-4f8f-4155-bad8-42df6673eddc"
+      },
+      "source": [
+        "\n",
+        "\n",
+        "## 데이터 전처리\n",
+        "데이터셋은 준비되었지만 아직 파인튜닝을 위한 준비는 완료되지 않았습니다. 아래와 같은 절차를 순차적으로 따라갈 것입니다:\n",
+        "\n",
+        "- **레이블 매핑 (Label Mapping)** : 레이블 ID와 해당 클래스 이름 간의 매핑을 수행합니다. 이는 모델 학습과 평가 시 유용하게 사용됩니다.\n",
+        "\n",
+        "- **이미지 전처리 (Image Processing)** : ViTImageProcessor를 사용해 입력 이미지의 크기를 표준화하고, 사전 학습된 모델에 맞는 정규화를 적용합니다. 또한, torchvision을 활용해 훈련, 검증, 테스트용으로 서로 다른 변환을 정의하여 모델의 일반화를 향상시킵니다.\n",
+        "\n",
+        "- **변환 함수 정의 (Transform Functions)** : ViT 모델에 맞는 형식과 크기로 이미지를 변환하기 위한 함수를 구현합니다.\n",
+        "\n",
+        "- **데이터 로딩 (Data Loading)** : 이미지와 라벨을 올바르게 배치할 수 있도록 사용자 정의 collate 함수를 설정하고, 학습 과정에서 효율적인 데이터 로딩과 배치를 위해 DataLoader를 구성합니다.\n",
+        "\n",
+        "- **배치 준비 (Batch Preparation)** : 샘플 배치의 데이터 형태(shape)를 확인하여 변환이 제대로 되었는지, 그리고 모델 입력으로 사용할 준비가 되었는지 검증합니다."
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "1463c15d-5b73-47dc-8113-a910e3cd38b9",
+      "metadata": {
+        "id": "1463c15d-5b73-47dc-8113-a910e3cd38b9"
+      },
+      "source": [
+        "### 레이블 매핑 (Label Mapping)"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 11,
+      "id": "43a6187f-0ba0-4cd9-a9ec-f06fca3a91bf",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "43a6187f-0ba0-4cd9-a9ec-f06fca3a91bf",
+        "outputId": "e9c81dd9-94a6-4cea-fe8d-47805a432bfd"
+      },
+      "outputs": [
+        {
+          "data": {
+            "text/plain": [
+              "({0: 'benign', 1: 'malignant', 2: 'normal'}, 'malignant')"
+            ]
+          },
+          "execution_count": 11,
+          "metadata": {},
+          "output_type": "execute_result"
+        }
+      ],
+      "source": [
+        "id2label = {id:label for id, label in enumerate(train_ds.features['label'].names)}\n",
+        "label2id = {label:id for id,label in id2label.items()}\n",
+        "id2label, id2label[train_ds[0]['label']]"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "0be0b208-5aad-48ff-97f5-ea484cfc8ad7",
+      "metadata": {
+        "id": "0be0b208-5aad-48ff-97f5-ea484cfc8ad7"
+      },
+      "source": [
+        "### 이미지 전처리 (Image Processing)"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 12,
+      "id": "a2d04160-93aa-425e-a06d-58c09ec6ffbd",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 49,
+          "referenced_widgets": [
+            "b7ea8f8585354f9bb4f11f43a6f1a8ac",
+            "15a9e8f3433d478ba8f73dd158515e7d",
+            "d4335bbecd3f4a9693a8c2742a765a4c",
+            "cec02057108742e0b0f5bf5704cf6c6d",
+            "5d2b2e10194043efadf1377e961b1e41",
+            "041c9438afa94e878b5678111b1cc87c",
+            "0bc23ea7af8046788212dd03fa6679f9",
+            "2658dc163a3942608f583622c44f28ea",
+            "9519479ac1234c50ac6b5459cb2db76b",
+            "8f0a3034925d409f85c64f5271187923",
+            "e0cb22c893ae4ed88b24224347751da7"
+          ]
+        },
+        "id": "a2d04160-93aa-425e-a06d-58c09ec6ffbd",
+        "outputId": "de82d69b-8010-4c23-d892-4fbe8353f151"
+      },
+      "outputs": [],
+      "source": [
+        "from transformers import ViTImageProcessor\n",
+        "\n",
+        "model_name = \"google/vit-large-patch16-224\"\n",
+        "processor = ViTImageProcessor.from_pretrained(model_name)"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 13,
+      "id": "136d114f-a054-467e-a034-cdddf9bf574b",
+      "metadata": {
+        "id": "136d114f-a054-467e-a034-cdddf9bf574b"
+      },
+      "outputs": [],
+      "source": [
+        "from torchvision.transforms import CenterCrop, Compose, Normalize, RandomHorizontalFlip, RandomResizedCrop, ToTensor, Resize\n",
+        "\n",
+        "image_mean, image_std = processor.image_mean, processor.image_std\n",
+        "size = processor.size[\"height\"]\n",
+        "\n",
+        "normalize = Normalize(mean=image_mean, std=image_std)\n",
+        "\n",
+        "train_transforms = Compose([\n",
+        "    RandomResizedCrop(size),\n",
+        "    RandomHorizontalFlip(),\n",
+        "    ToTensor(),\n",
+        "    normalize,\n",
+        "])\n",
+        "val_transforms = Compose([\n",
+        "    Resize(size),\n",
+        "    CenterCrop(size),\n",
+        "    ToTensor(),\n",
+        "    normalize,\n",
+        "])\n",
+        "test_transforms = Compose([\n",
+        "    Resize(size),\n",
+        "    CenterCrop(size),\n",
+        "    ToTensor(),\n",
+        "    normalize,\n",
+        "])"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "9e910499-84bf-4672-bbf0-ca78915b2821",
+      "metadata": {
+        "id": "9e910499-84bf-4672-bbf0-ca78915b2821"
+      },
+      "source": [
+        "### 변환 함수 생성 (Transform Functions)"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 14,
+      "id": "5ddc7ad4-bd09-4c76-ac00-ca7dafbd8417",
+      "metadata": {
+        "id": "5ddc7ad4-bd09-4c76-ac00-ca7dafbd8417"
+      },
+      "outputs": [],
+      "source": [
+        "def apply_train_transforms(examples):\n",
+        "    examples['pixel_values'] = [train_transforms(image.convert(\"RGB\")) for image in examples['image']]\n",
+        "    return examples\n",
+        "\n",
+        "def apply_val_transforms(examples):\n",
+        "    examples['pixel_values'] = [val_transforms(image.convert(\"RGB\")) for image in examples['image']]\n",
+        "    return examples\n",
+        "\n",
+        "def apply_test_transforms(examples):\n",
+        "    examples['pixel_values'] = [val_transforms(image.convert(\"RGB\")) for image in examples['image']]\n",
+        "    return examples"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "1ac74835-f883-46a3-877c-26265b27a325",
+      "metadata": {
+        "id": "1ac74835-f883-46a3-877c-26265b27a325"
+      },
+      "source": [
+        "### 각 데이터셋에 변환 함수 적용"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 15,
+      "id": "45f2b765-5258-4b44-b2ce-6bff952bdd1b",
+      "metadata": {
+        "id": "45f2b765-5258-4b44-b2ce-6bff952bdd1b"
+      },
+      "outputs": [],
+      "source": [
+        "train_ds.set_transform(apply_train_transforms)\n",
+        "val_ds.set_transform(apply_val_transforms)\n",
+        "test_ds.set_transform(apply_test_transforms)"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 16,
+      "id": "ce586ef6-5f48-4554-8dc4-48797a977674",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "ce586ef6-5f48-4554-8dc4-48797a977674",
+        "outputId": "0d6b46d7-9fd9-4825-f8ac-8b0eb0e5a273"
+      },
+      "outputs": [
+        {
+          "data": {
+            "text/plain": [
+              "{'image': Image(mode=None, decode=True, id=None),\n",
+              " 'label': ClassLabel(names=['benign', 'malignant', 'normal'], id=None)}"
+            ]
+          },
+          "execution_count": 16,
+          "metadata": {},
+          "output_type": "execute_result"
+        }
+      ],
+      "source": [
+        "train_ds.features"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 17,
+      "id": "731fccb7-7311-4175-8ce1-67141a662b1e",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "731fccb7-7311-4175-8ce1-67141a662b1e",
+        "outputId": "68be9976-f736-45dc-bf06-86255e9209a9"
+      },
+      "outputs": [
+        {
+          "data": {
+            "text/plain": [
+              "{'image': <PIL.PngImagePlugin.PngImageFile image mode=RGB size=558x475>,\n",
+              " 'label': 1,\n",
+              " 'pixel_values': tensor([[[ 0.0431,  0.0667,  0.0902,  ..., -0.2863, -0.2706, -0.2549],\n",
+              "          [ 0.0196,  0.0353,  0.0353,  ..., -0.3647, -0.3725, -0.3647],\n",
+              "          [ 0.0510,  0.0275,  0.0039,  ..., -0.3333, -0.3490, -0.3569],\n",
+              "          ...,\n",
+              "          [-0.9373, -0.9059, -0.8745,  ..., -0.7882, -0.7804, -0.7725],\n",
+              "          [-0.9059, -0.9059, -0.8824,  ..., -0.8196, -0.8118, -0.7961],\n",
+              "          [-0.8196, -0.8431, -0.8667,  ..., -0.8431, -0.8588, -0.8431]],\n",
+              " \n",
+              "         [[ 0.0431,  0.0667,  0.0902,  ..., -0.2863, -0.2706, -0.2549],\n",
+              "          [ 0.0196,  0.0353,  0.0353,  ..., -0.3647, -0.3725, -0.3647],\n",
+              "          [ 0.0510,  0.0275,  0.0039,  ..., -0.3333, -0.3490, -0.3569],\n",
+              "          ...,\n",
+              "          [-0.9373, -0.9059, -0.8745,  ..., -0.7882, -0.7804, -0.7725],\n",
+              "          [-0.9059, -0.9059, -0.8824,  ..., -0.8196, -0.8118, -0.7961],\n",
+              "          [-0.8196, -0.8431, -0.8667,  ..., -0.8431, -0.8588, -0.8431]],\n",
+              " \n",
+              "         [[ 0.0431,  0.0667,  0.0902,  ..., -0.2863, -0.2706, -0.2549],\n",
+              "          [ 0.0196,  0.0353,  0.0353,  ..., -0.3647, -0.3725, -0.3647],\n",
+              "          [ 0.0510,  0.0275,  0.0039,  ..., -0.3333, -0.3490, -0.3569],\n",
+              "          ...,\n",
+              "          [-0.9373, -0.9059, -0.8745,  ..., -0.7882, -0.7804, -0.7725],\n",
+              "          [-0.9059, -0.9059, -0.8824,  ..., -0.8196, -0.8118, -0.7961],\n",
+              "          [-0.8196, -0.8431, -0.8667,  ..., -0.8431, -0.8588, -0.8431]]])}"
+            ]
+          },
+          "execution_count": 17,
+          "metadata": {},
+          "output_type": "execute_result"
+        }
+      ],
+      "source": [
+        "train_ds[0]"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "49b06fa4-6f28-4e45-8af4-779c424583fe",
+      "metadata": {
+        "id": "49b06fa4-6f28-4e45-8af4-779c424583fe"
+      },
+      "source": [
+        "픽셀 값이 텐서 값으로 잘 변환하였습니다.\n",
+        "\n",
+        "### 데이터 로딩"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 18,
+      "id": "3f47f263-2046-4847-952d-728fa3fe5cf4",
+      "metadata": {
+        "id": "3f47f263-2046-4847-952d-728fa3fe5cf4"
+      },
+      "outputs": [],
+      "source": [
+        "import torch\n",
+        "from torch.utils.data import DataLoader\n",
+        "\n",
+        "def collate_fn(examples):\n",
+        "    pixel_values = torch.stack([example[\"pixel_values\"] for example in examples])\n",
+        "    labels = torch.tensor([example[\"label\"] for example in examples])\n",
+        "    return {\"pixel_values\": pixel_values, \"labels\": labels}\n",
+        "\n",
+        "train_dl = DataLoader(train_ds, collate_fn=collate_fn, batch_size=4)"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "aed986b0-e661-4c57-ad58-4fa73d795828",
+      "metadata": {
+        "id": "aed986b0-e661-4c57-ad58-4fa73d795828"
+      },
+      "source": [
+        "### 배치 준비"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 19,
+      "id": "14c6d3f8-48e6-423f-8193-7571f986f103",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "14c6d3f8-48e6-423f-8193-7571f986f103",
+        "outputId": "eb04bfbd-040b-4c78-d8ae-e8f302dfb9d3"
+      },
+      "outputs": [
+        {
+          "name": "stdout",
+          "output_type": "stream",
+          "text": [
+            "pixel_values torch.Size([4, 3, 224, 224])\n",
+            "labels torch.Size([4])\n"
+          ]
+        }
+      ],
+      "source": [
+        "batch = next(iter(train_dl))\n",
+        "for k,v in batch.items():\n",
+        "  if isinstance(v, torch.Tensor):\n",
+        "    print(k, v.shape)"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "35379d35-e567-4a30-8b91-6eac80b79044",
+      "metadata": {
+        "id": "35379d35-e567-4a30-8b91-6eac80b79044"
+      },
+      "source": [
+        "완벽합니다! 이제 파인튜닝 과정을 진행할 준비가 되었습니다."
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "45056d5e-8bca-4ece-b29b-bc772aeef49f",
+      "metadata": {
+        "id": "45056d5e-8bca-4ece-b29b-bc772aeef49f"
+      },
+      "source": [
+        "## 모델 파인튜닝\n",
+        "이어서 모델을 설정하고 파인튜닝하는 과정을 진행하겠습니다.\n",
+        "\n",
+        "우선, 특정 라벨 매핑과 사전 학습 설정(pre-trained settings)을 사용해 모델을 초기화하고, 크기 불일치를 조정해줍니다. 그다음, 학습 과정을 정의하기 위해 저장 전략, 배치 크기, 학습 epoch 등의 학습 파라미터를 설정하고, Weights & Biases를 통해 학습 결과를 로깅합니다.\n",
+        "이후 Hugging Face의 Trainer가 인스턴스화되어 학습과 평가를 관리하며, 커스텀 데이터 콜레이터와 모델에 내장된 프로세서를 함께 사용합니다. 마지막으로, 학습이 완료된 모델을 테스트 데이터셋으로 평가하고, 정확도를 확인하기 위한 지표(metrics)를 출력합니다.\n",
+        "\n",
+        "\n",
+        "\n",
+        "첫번째로, 모델을 먼저 호출해보겠습니다."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 20,
+      "id": "97d0b588-3e41-4852-9307-9e2ec7d5bb0b",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 205,
+          "referenced_widgets": [
+            "a507faf3af4a4b599008d6fde6558326",
+            "3aad39f8cc6f4c7aa04eae1e29c82cf2",
+            "6bc38493a08a4fdfb1a5c337bf9a56ea",
+            "a8f641150053483898246602e54592e1",
+            "2261070b86c7497fb2221e41ab323cce",
+            "5fb8aa77c7a14cf38f87eb4df6457a37",
+            "f511728e7deb40029950ff8019742162",
+            "00314085ba794a5d87eeca3dc7991a0c",
+            "40a209ad433f4c71803eab2d8a3a2f69",
+            "cb0206ba72d7496c91e2737a057f1fe1",
+            "9ede96f2fe244364980dd74861978874",
+            "ec03ea80c45b452fa9365c13eaafec35",
+            "8b2007e92a2f48e38c52e14a9e647528",
+            "e3c612d1b46f45cbbe35ecdc39df73bf",
+            "87af41d1980245e1a3edacfb4a0b2b12",
+            "b564a6710ab1477ba5aef3a87961913b",
+            "d88bcb77112743ed9c68f14af86d15f7",
+            "fb7a5bc6781c4794a8bef2ab3a7b9116",
+            "6e53b375bc41426788d7a094287fbc22",
+            "f3f6c7337c9c48eabc5ad3ba9a417dc1",
+            "b0eb68fadae24f569b813fb28d1f4611",
+            "cef9fd8581614026950f8495952ccd03",
+            "2b4abfd0f59e41309b99ae604e96738f",
+            "f1a5e1f980d9427095d24629d2a83ae4",
+            "a4efbed77ded4c4facbd3d79239d64d8",
+            "678e7fc2f9c24281a5e823a85847178e",
+            "2826ad430729434ea93ddf45280bdd29",
+            "c74b88aa20f846dda524c0a5e03dc66e",
+            "03871e390dfd485198461b5e3ba6faed",
+            "1d4ed206e715459989fed43f495ba8c2",
+            "3bc7a168ed034c6e8067af549b8979d8",
+            "8219eed6c80e4128a47d6f1489e92859",
+            "602cd980010d477ea5824d8cc6417bb0"
+          ]
+        },
+        "id": "97d0b588-3e41-4852-9307-9e2ec7d5bb0b",
+        "outputId": "28597115-735d-48b4-ca15-05aa557be02b"
+      },
+      "outputs": [
+        {
+          "name": "stderr",
+          "output_type": "stream",
+          "text": [
+            "Some weights of ViTForImageClassification were not initialized from the model checkpoint at google/vit-large-patch16-224 and are newly initialized because the shapes did not match:\n",
+            "- classifier.weight: found shape torch.Size([1000, 1024]) in the checkpoint and torch.Size([3, 1024]) in the model instantiated\n",
+            "- classifier.bias: found shape torch.Size([1000]) in the checkpoint and torch.Size([3]) in the model instantiated\n",
+            "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"
+          ]
+        }
+      ],
+      "source": [
+        "from transformers import ViTForImageClassification\n",
+        "\n",
+        "model = ViTForImageClassification.from_pretrained(model_name, id2label=id2label, label2id=label2id, ignore_mismatched_sizes=True)"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "2f773e18-63e8-41bd-885f-7ba95d074a3d",
+      "metadata": {
+        "id": "2f773e18-63e8-41bd-885f-7ba95d074a3d"
+      },
+      "source": [
+        "여기에는 미묘한 디테일이 하나 있습니다. 바로 `ignore_mismatched_sizes` 파라미터 입니다.\n",
+        "\n",
+        "사전 학습된(pre-trained) 모델을 새로운 데이터셋에 파인튜닝할 때, 종종 이미지의 입력 크기나 모델 구조(예: 분류 레이어의 클래스 수 등)가 원래 모델이 학습된 환경과 완전히 일치하지 않을 수 있습니다.\n",
+        "이러한 상황은 다양한 이유로 발생할 수 있는데, 예를 들어 ImageNet처럼 자연 이미지로 학습된 모델을 의료 영상이나 특수 카메라 이미지처럼 전혀 다른 유형의 이미지에 적용하려고 할 때 발생할 수 있습니다.\n",
+        "\n",
+        "\n",
+        "이때, `ignore_mismatched_sizes` 를 `True` 로 설정하면, 사이즈가 일치하지 않더라도 오류 없이 레이어를 자동으로 조정해 사용할 수 있게 해줍니다.\n",
+        "\n",
+        "\n",
+        "예를 들어, 해당 모델은 1000개의 클래스로 학습되어 있어서 출력 크기가 `torch.Size([1000])` 이고 입력 또한 `torch.Size([1000])` 클래스를 기대합니다. 그런데 우리가 사용하는 데이터셋은 클래스가 3개로 `torch.Size([3])`입니다. 이 경우 모델에 직접 입력하면 클래스 수가 맞지 않기 때문에 오류가 발생하게 됩니다."
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "e2c671bf-9978-46d8-82ef-0906d4e89d03",
+      "metadata": {
+        "id": "e2c671bf-9978-46d8-82ef-0906d4e89d03"
+      },
+      "source": [
+        "그다음 Google에서 제공하는 학습 인자(arguments)를 정의합니다."
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "d473af8a-4070-48d7-aecb-1a5a90f0b63f",
+      "metadata": {
+        "id": "d473af8a-4070-48d7-aecb-1a5a90f0b63f"
+      },
+      "source": [
+        "(선택 사항) `report_to` 파라미터를 `wandb`로 설정했기 때문에, 평가 지표는 Weights & Biases(W&B)에 저장됩니다. W&B는 API 키를 요구하므로 계정과 API 키를 미리 생성해두어야 합니다. 만약 W&B를 사용하고 싶지 않다면, `report_to` 파라미터를 제거하면 됩니다."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 21,
+      "id": "f16fc568-9fdc-4c60-acec-6ed3dbb85aef",
+      "metadata": {
+        "id": "f16fc568-9fdc-4c60-acec-6ed3dbb85aef"
+      },
+      "outputs": [
+        {
+          "data": {
+            "text/plain": [
+              "'\\n현재 환경에서 TrainingArguments가 해당 인자들을 인식하지 못하는 오류(TypeError)가 발생하여\\n최소한의 설정만으로 Trainer가 동작하도록 수정한 버전임.\\n(중간값 저장이나 자동 평가 없이 단순 학습만 수행됨.)\\n'"
+            ]
+          },
+          "execution_count": 21,
+          "metadata": {},
+          "output_type": "execute_result"
+        }
+      ],
+      "source": [
+        "from transformers import TrainingArguments, Trainer\n",
+        "import numpy as np\n",
+        "\n",
+        "train_args = TrainingArguments(\n",
+        "    output_dir = \"output-models\",\n",
+        "    #save_total_limit=2,\n",
+        "    #report_to=\"wandb\",\n",
+        "    #save_strategy=\"epoch\",\n",
+        "    #evaluation_strategy=\"epoch\",\n",
+        "    learning_rate=2e-5,\n",
+        "    per_device_train_batch_size=10,\n",
+        "    per_device_eval_batch_size=4,\n",
+        "    num_train_epochs=40,\n",
+        "    weight_decay=0.01,\n",
+        "    #load_best_model_at_end=True,\n",
+        "    logging_dir='logs',\n",
+        "    remove_unused_columns=False,\n",
+        "    report_to=\"none\"\n",
+        ")\n",
+        "\n",
+        "\"\"\"\n",
+        "현재 환경에서 TrainingArguments가 해당 인자들을 인식하지 못하는 오류(TypeError)가 발생하여\n",
+        "최소한의 설정만으로 Trainer가 동작하도록 수정한 버전임.\n",
+        "(중간값 저장이나 자동 평가 없이 단순 학습만 수행됨.)\n",
+        "\"\"\"\n"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "f74d4457-49fd-4e1b-9842-a3759ec524c9",
+      "metadata": {
+        "id": "f74d4457-49fd-4e1b-9842-a3759ec524c9"
+      },
+      "source": [
+        "이제 Trainer을 사용해 파인튜닝을 시작할 수 있습니다.\n",
+        "\n"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "id": "5a117d62-9054-4e14-b7b2-0703de17a741",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 181
+        },
+        "id": "5a117d62-9054-4e14-b7b2-0703de17a741",
+        "outputId": "55c9e8b4-eaf0-431a-a132-28af2227e059"
+      },
+      "outputs": [],
+      "source": [
+        "trainer = Trainer(\n",
+        "    model,\n",
+        "    train_args,\n",
+        "    train_dataset=train_ds,\n",
+        "    eval_dataset=val_ds,\n",
+        "    data_collator=collate_fn,\n",
+        "    tokenizer=processor,\n",
+        ")\n",
+        "trainer.train()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "e154fd79-1de7-4169-a6af-402b12881042",
+      "metadata": {
+        "id": "e154fd79-1de7-4169-a6af-402b12881042"
+      },
+      "source": [
+        "| Epoch | Training Loss | Validation Loss | Accuracy |\n",
+        "|-------|---------------|-----------------|----------|\n",
+        "| 40    | 0.174700      | 0.596288        | 0.903846 |\n",
+        "\n",
+        "파인튜닝이 완료되었습니다. 이제 테스트 세트를 사용해 모델을 평가해보겠습니다."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "id": "c19b6d99-0a89-45ac-a6d9-ec3e79edc041",
+      "metadata": {
+        "id": "c19b6d99-0a89-45ac-a6d9-ec3e79edc041",
+        "outputId": "2f5ab46c-d605-4cdb-e164-0bd8908462a4"
+      },
+      "outputs": [],
+      "source": [
+        "outputs = trainer.predict(test_ds)\n",
+        "print(outputs.metrics)"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "2c4ddecb-7ab0-493e-90b9-44bf4e2a530e",
+      "metadata": {
+        "id": "2c4ddecb-7ab0-493e-90b9-44bf4e2a530e"
+      },
+      "source": [
+        "`{'test_loss': 0.3219967782497406, 'test_accuracy': 0.9102564102564102, 'test_runtime': 4.0543, 'test_samples_per_second': 38.478, 'test_steps_per_second': 9.619}`"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "0be50a0b",
+      "metadata": {
+        "id": "0be50a0b"
+      },
+      "source": [
+        "### (선택 사항) 모델을 Hugging Face Hub에 업로드하기\n",
+        "\n",
+        "\n",
+        "\n",
+        "`push_to_hub`을 사용하면 Hugging Face Hub에 모델을 업로드할 수 있습니다."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "id": "f1d55e6a",
+      "metadata": {
+        "id": "f1d55e6a"
+      },
+      "outputs": [],
+      "source": [
+        "model.push_to_hub(\"your_model_name\")"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "5f74c058-e2a5-4c9c-8d70-1c2c574b933f",
+      "metadata": {
+        "id": "5f74c058-e2a5-4c9c-8d70-1c2c574b933f"
+      },
+      "source": [
+        "좋습니다! 이제 결과를 시각화해봅시다.\n",
+        "\n",
+        "## 결과\n",
+        "파인튜닝을 마쳤으니, 이제 모델이 각 클래스를 어떻게 예측했는지 확인해보겠습니다.\n",
+        "이를 위해 scikit-learn의 Confusion Matrix Display를 사용하고, Recall 점수도 함께 확인해보겠습니다."
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "ade5321d-ff63-4308-8317-d1e4da2219df",
+      "metadata": {
+        "id": "ade5321d-ff63-4308-8317-d1e4da2219df"
+      },
+      "source": [
+        "### Confusion Matrix 란?\n",
+        "Confusion Matrix(혼동 행렬)는 알고리즘, 특히 지도 학습 모델이 실제 정답을 알고 있는 테스트 데이터에 대해 얼마나 잘 작동하는지를 시각적으로 보여주는 표 형식입니다.\n",
+        "예측된 라벨과 실제 라벨의 빈도를 나타내므로, 분류 모델의 성능을 평가할 때 특히 유용합니다.\n",
+        "\n",
+        "우리 모델의 Confusion Matrix를 그려봅시다."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 25,
+      "id": "8efb0ece-92b3-498d-b47b-0f9c04d4ebb8",
+      "metadata": {
+        "id": "8efb0ece-92b3-498d-b47b-0f9c04d4ebb8",
+        "outputId": "eee6b79f-898b-466f-91c5-2c6dff61cf1f"
+      },
+      "outputs": [
+        {
+          "data": {
+            "text/plain": [
+              "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x7f57dad2f690>"
+            ]
+          },
+          "execution_count": 25,
+          "metadata": {},
+          "output_type": "execute_result"
+        },
+        {
+          "data": {
+            "image/png": "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",
+            "text/plain": [
+              "<Figure size 640x480 with 2 Axes>"
+            ]
+          },
+          "metadata": {},
+          "output_type": "display_data"
+        }
+      ],
+      "source": [
+        "from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay\n",
+        "\n",
+        "y_true = outputs.label_ids\n",
+        "y_pred = outputs.predictions.argmax(1)\n",
+        "\n",
+        "labels = train_ds.features['label'].names\n",
+        "cm = confusion_matrix(y_true, y_pred)\n",
+        "disp = ConfusionMatrixDisplay(confusion_matrix=cm, display_labels=labels)\n",
+        "disp.plot(xticks_rotation=45)"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "9178d2d1-b828-45a6-8873-039abc0419c2",
+      "metadata": {
+        "id": "9178d2d1-b828-45a6-8873-039abc0419c2"
+      },
+      "source": [
+        "### Recall Score 란?\n",
+        "Recall Score(재현율)는 분류 작업에서 모델이 실제로 긍정 클래스(Positive Class)인 샘플을 얼마나 잘 찾아냈는지를 측정하는 성능 지표입니다.\n",
+        "즉, 실제 정답 중에서 모델이 올바르게 예측한 비율을 나타냅니다.\n",
+        "\n",
+        "이제 scikit-learn을 사용해 재현율 점수를 출력해봅시다."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 26,
+      "id": "48d87ca7-8458-41d5-a773-38e2c9522f64",
+      "metadata": {
+        "id": "48d87ca7-8458-41d5-a773-38e2c9522f64",
+        "outputId": "da9056ed-1b64-4923-ba11-9752d5b596b3"
+      },
+      "outputs": [
+        {
+          "name": "stdout",
+          "output_type": "stream",
+          "text": [
+            "Recall for benign: 0.89\n",
+            "Recall for malignant: 0.90\n",
+            "Recall for normal: 0.81\n"
+          ]
+        }
+      ],
+      "source": [
+        "from sklearn.metrics import recall_score\n",
+        "\n",
+        "# Calculate the recall scores\n",
+        "# 'None' calculates recall for each class separately\n",
+        "recall = recall_score(y_true, y_pred, average=None)\n",
+        "\n",
+        "# Print the recall for each class\n",
+        "for label, score in zip(labels, recall):\n",
+        "    print(f'Recall for {label}: {score:.2f}')\n"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "c8b1a1b1-7de4-4eb6-98bf-de87e8cbbcec",
+      "metadata": {
+        "id": "c8b1a1b1-7de4-4eb6-98bf-de87e8cbbcec"
+      },
+      "source": [
+        "`Recall for benign: 0.90,\n",
+        "Recall for malignant: 0.86,\n",
+        "Recall for normal: 0.78`"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "67b76567-039d-467b-9cfc-0837fb3e1a1b",
+      "metadata": {
+        "id": "67b76567-039d-467b-9cfc-0837fb3e1a1b"
+      },
+      "source": [
+        "## 결론\n",
+        "이 cookbook에서는 의료 데이터셋을 활용해 ViT 모델을 학습하는 과정을 다루었습니다.\n",
+        "데이터셋 준비, 이미지 전처리, 모델 구성, 학습, 평가, 그리고 결과 시각화까지의 핵심 단계들을 모두 포함하고 있습니다.\n",
+        "\n",
+        "Hugging Face의 Transformers 라이브러리와 scikit-learn, PyTorch의 Torchvision을 활용하여 효율적인 모델 학습과 평가가 가능하며, 이를 통해 바이오메디컬 이미지를 얼마나 정확하게 분류할 수 있는지에 대한 유용한 통찰을 제공합니다."
+      ]
+    }
+  ],
+  "metadata": {
+    "accelerator": "GPU",
+    "colab": {
+      "gpuType": "T4",
+      "provenance": []
+    },
+    "kernelspec": {
+      "display_name": "hf",
+      "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": "3.11.11"
+    },
+    "widgets": {
+      "application/vnd.jupyter.widget-state+json": {
+        "00314085ba794a5d87eeca3dc7991a0c": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "03871e390dfd485198461b5e3ba6faed": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "039241d8c1c34103a655fbcb6de51ae6": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "041c9438afa94e878b5678111b1cc87c": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "04fb4c819c3c44b59184c54ea44456bf": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "09d1ea0c40184e648c75b10ee22f23e7": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HBoxModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HBoxModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HBoxView",
+            "box_style": "",
+            "children": [
+              "IPY_MODEL_4b411454d68a436cbf04d8a97e2fe7da",
+              "IPY_MODEL_a2f5fb56b54f4b7383fd85aed9920bcb",
+              "IPY_MODEL_2a4a870ca4404c23bd92f9b1117d9a0e"
+            ],
+            "layout": "IPY_MODEL_c9077c3acb6342f2a248876f3ebf7464"
+          }
+        },
+        "0bc23ea7af8046788212dd03fa6679f9": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "0fe075d0db0445aeb7e4d275860d39f2": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HBoxModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HBoxModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HBoxView",
+            "box_style": "",
+            "children": [
+              "IPY_MODEL_64b29e54431246e98aeadf09c705b684",
+              "IPY_MODEL_323db282be654fdeb25bf9a5b9f1733b",
+              "IPY_MODEL_2d7ebc94bae04034aee3ee10f800c1df"
+            ],
+            "layout": "IPY_MODEL_33694de53716463b8c5b07720a40e067"
+          }
+        },
+        "1173f9c2411640078fb20331845d2c87": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "15a9e8f3433d478ba8f73dd158515e7d": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_041c9438afa94e878b5678111b1cc87c",
+            "placeholder": "​",
+            "style": "IPY_MODEL_0bc23ea7af8046788212dd03fa6679f9",
+            "value": "preprocessor_config.json: 100%"
+          }
+        },
+        "1b8330ce7031406c95eac80280936dae": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "ProgressStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "ProgressStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "bar_color": null,
+            "description_width": ""
+          }
+        },
+        "1d4ed206e715459989fed43f495ba8c2": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "1eccd922da6e4b1682bdb061716ed27b": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "2261070b86c7497fb2221e41ab323cce": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "228c08de964d48fdbb3594c1f2ecde09": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "2658dc163a3942608f583622c44f28ea": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "27297aa211ef4564bcb9a23bd6e073c5": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HBoxModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HBoxModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HBoxView",
+            "box_style": "",
+            "children": [
+              "IPY_MODEL_28e427d0fb3c42c0b9d55f9076f18918",
+              "IPY_MODEL_9155b025badd40c284c07fbc87175169",
+              "IPY_MODEL_40c47070f79b48d088622696128531cb"
+            ],
+            "layout": "IPY_MODEL_41a800c6e6c547a9ab528ee96e1ce137"
+          }
+        },
+        "2826ad430729434ea93ddf45280bdd29": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "28e427d0fb3c42c0b9d55f9076f18918": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_43c0527631a64c89b171c2a0664dfa8b",
+            "placeholder": "​",
+            "style": "IPY_MODEL_95d5ad27410b4a689734c1593f71d02b",
+            "value": "README.md: 100%"
+          }
+        },
+        "2a4a870ca4404c23bd92f9b1117d9a0e": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_482d168e5e104fd48c5c762450c02930",
+            "placeholder": "​",
+            "style": "IPY_MODEL_466dfa0cc1344aa58bb1a4a8b8cf9aba",
+            "value": " 53.6M/53.6M [00:00&lt;00:00, 158MB/s]"
+          }
+        },
+        "2b4abfd0f59e41309b99ae604e96738f": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HBoxModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HBoxModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HBoxView",
+            "box_style": "",
+            "children": [
+              "IPY_MODEL_f1a5e1f980d9427095d24629d2a83ae4",
+              "IPY_MODEL_a4efbed77ded4c4facbd3d79239d64d8",
+              "IPY_MODEL_678e7fc2f9c24281a5e823a85847178e"
+            ],
+            "layout": "IPY_MODEL_2826ad430729434ea93ddf45280bdd29"
+          }
+        },
+        "2d028e6a22194fa28419e88e87573421": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "2d7ebc94bae04034aee3ee10f800c1df": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_1173f9c2411640078fb20331845d2c87",
+            "placeholder": "​",
+            "style": "IPY_MODEL_db8dd16fc27a45d5a1ab81380b305389",
+            "value": " 624/624 [00:01&lt;00:00, 538.15 examples/s]"
+          }
+        },
+        "2ddd4fe77259428aaf7d6a56ed0b223e": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_387d13f078f747fea3c747f55e0458f5",
+            "placeholder": "​",
+            "style": "IPY_MODEL_8454a7ef8c1140a78f544465577d7532",
+            "value": " 156/156 [00:00&lt;00:00, 362.72 examples/s]"
+          }
+        },
+        "2f48026c77c9415292364602bff4ca18": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "323db282be654fdeb25bf9a5b9f1733b": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "FloatProgressModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "FloatProgressModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "ProgressView",
+            "bar_style": "success",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_1eccd922da6e4b1682bdb061716ed27b",
+            "max": 624,
+            "min": 0,
+            "orientation": "horizontal",
+            "style": "IPY_MODEL_9b238f957be24a8e99b46baaa664cb5c",
+            "value": 624
+          }
+        },
+        "33694de53716463b8c5b07720a40e067": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "387d13f078f747fea3c747f55e0458f5": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "3aad39f8cc6f4c7aa04eae1e29c82cf2": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_5fb8aa77c7a14cf38f87eb4df6457a37",
+            "placeholder": "​",
+            "style": "IPY_MODEL_f511728e7deb40029950ff8019742162",
+            "value": "config.json: 100%"
+          }
+        },
+        "3bc7a168ed034c6e8067af549b8979d8": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "ProgressStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "ProgressStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "bar_color": null,
+            "description_width": ""
+          }
+        },
+        "40a209ad433f4c71803eab2d8a3a2f69": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "ProgressStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "ProgressStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "bar_color": null,
+            "description_width": ""
+          }
+        },
+        "40c47070f79b48d088622696128531cb": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_8e67e87040944a878618f01a98294681",
+            "placeholder": "​",
+            "style": "IPY_MODEL_f3267441b9dc4f029a3bfdffaf491842",
+            "value": " 1.18k/1.18k [00:00&lt;00:00, 64.1kB/s]"
+          }
+        },
+        "4113c7318d5d4484b3990dcbdb9bc1cd": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "41a800c6e6c547a9ab528ee96e1ce137": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "4358acace42c41519fc83cc67e814961": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "FloatProgressModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "FloatProgressModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "ProgressView",
+            "bar_style": "success",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_aa9365c6057a4137bb678ce4d05c448d",
+            "max": 211624597,
+            "min": 0,
+            "orientation": "horizontal",
+            "style": "IPY_MODEL_741d79cc18ae44259621f030e454ab99",
+            "value": 211624597
+          }
+        },
+        "43c0527631a64c89b171c2a0664dfa8b": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "466dfa0cc1344aa58bb1a4a8b8cf9aba": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "482d168e5e104fd48c5c762450c02930": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "4901091780124b3dbba924779868f4c8": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "ProgressStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "ProgressStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "bar_color": null,
+            "description_width": ""
+          }
+        },
+        "4b411454d68a436cbf04d8a97e2fe7da": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_c99b9fb60a4042b58b07a8524f8f555e",
+            "placeholder": "​",
+            "style": "IPY_MODEL_2f48026c77c9415292364602bff4ca18",
+            "value": "test-00000-of-00001.parquet: 100%"
+          }
+        },
+        "566b4b97be32448fb502520a93976501": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "5aa5164fa3d040bc8180b122eb029d94": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "5d2b2e10194043efadf1377e961b1e41": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "5fb8aa77c7a14cf38f87eb4df6457a37": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "602cd980010d477ea5824d8cc6417bb0": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "64b29e54431246e98aeadf09c705b684": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_566b4b97be32448fb502520a93976501",
+            "placeholder": "​",
+            "style": "IPY_MODEL_c83d52312d02407a8ea124fa25ebddb7",
+            "value": "Generating train split: 100%"
+          }
+        },
+        "678e7fc2f9c24281a5e823a85847178e": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_8219eed6c80e4128a47d6f1489e92859",
+            "placeholder": "​",
+            "style": "IPY_MODEL_602cd980010d477ea5824d8cc6417bb0",
+            "value": " 1.22G/1.22G [00:06&lt;00:00, 267MB/s]"
+          }
+        },
+        "6bc38493a08a4fdfb1a5c337bf9a56ea": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "FloatProgressModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "FloatProgressModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "ProgressView",
+            "bar_style": "success",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_00314085ba794a5d87eeca3dc7991a0c",
+            "max": 69667,
+            "min": 0,
+            "orientation": "horizontal",
+            "style": "IPY_MODEL_40a209ad433f4c71803eab2d8a3a2f69",
+            "value": 69667
+          }
+        },
+        "6e53b375bc41426788d7a094287fbc22": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "741d79cc18ae44259621f030e454ab99": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "ProgressStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "ProgressStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "bar_color": null,
+            "description_width": ""
+          }
+        },
+        "8219eed6c80e4128a47d6f1489e92859": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "8454a7ef8c1140a78f544465577d7532": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "87af41d1980245e1a3edacfb4a0b2b12": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_b0eb68fadae24f569b813fb28d1f4611",
+            "placeholder": "​",
+            "style": "IPY_MODEL_cef9fd8581614026950f8495952ccd03",
+            "value": " 1.22G/1.22G [00:06&lt;00:00, 187MB/s]"
+          }
+        },
+        "8b2007e92a2f48e38c52e14a9e647528": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_d88bcb77112743ed9c68f14af86d15f7",
+            "placeholder": "​",
+            "style": "IPY_MODEL_fb7a5bc6781c4794a8bef2ab3a7b9116",
+            "value": "pytorch_model.bin: 100%"
+          }
+        },
+        "8e67e87040944a878618f01a98294681": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "8f0a3034925d409f85c64f5271187923": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "9155b025badd40c284c07fbc87175169": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "FloatProgressModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "FloatProgressModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "ProgressView",
+            "bar_style": "success",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_99893feb6a184879891978d111ebde49",
+            "max": 1175,
+            "min": 0,
+            "orientation": "horizontal",
+            "style": "IPY_MODEL_fe8fca891aff479a9fb11aeeb1971e2d",
+            "value": 1175
+          }
+        },
+        "9519479ac1234c50ac6b5459cb2db76b": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "ProgressStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "ProgressStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "bar_color": null,
+            "description_width": ""
+          }
+        },
+        "95d5ad27410b4a689734c1593f71d02b": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "99893feb6a184879891978d111ebde49": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "9b238f957be24a8e99b46baaa664cb5c": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "ProgressStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "ProgressStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "bar_color": null,
+            "description_width": ""
+          }
+        },
+        "9ede96f2fe244364980dd74861978874": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "a2f5fb56b54f4b7383fd85aed9920bcb": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "FloatProgressModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "FloatProgressModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "ProgressView",
+            "bar_style": "success",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_c162aa27a92f496192a2335ab0496bc8",
+            "max": 53566328,
+            "min": 0,
+            "orientation": "horizontal",
+            "style": "IPY_MODEL_4901091780124b3dbba924779868f4c8",
+            "value": 53566328
+          }
+        },
+        "a4efbed77ded4c4facbd3d79239d64d8": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "FloatProgressModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "FloatProgressModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "ProgressView",
+            "bar_style": "success",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_1d4ed206e715459989fed43f495ba8c2",
+            "max": 1217353096,
+            "min": 0,
+            "orientation": "horizontal",
+            "style": "IPY_MODEL_3bc7a168ed034c6e8067af549b8979d8",
+            "value": 1217353096
+          }
+        },
+        "a507faf3af4a4b599008d6fde6558326": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HBoxModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HBoxModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HBoxView",
+            "box_style": "",
+            "children": [
+              "IPY_MODEL_3aad39f8cc6f4c7aa04eae1e29c82cf2",
+              "IPY_MODEL_6bc38493a08a4fdfb1a5c337bf9a56ea",
+              "IPY_MODEL_a8f641150053483898246602e54592e1"
+            ],
+            "layout": "IPY_MODEL_2261070b86c7497fb2221e41ab323cce"
+          }
+        },
+        "a86b7af977bc40fea33a27748e57f432": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "a8f641150053483898246602e54592e1": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_cb0206ba72d7496c91e2737a057f1fe1",
+            "placeholder": "​",
+            "style": "IPY_MODEL_9ede96f2fe244364980dd74861978874",
+            "value": " 69.7k/69.7k [00:00&lt;00:00, 5.43MB/s]"
+          }
+        },
+        "aa9365c6057a4137bb678ce4d05c448d": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "b0eb68fadae24f569b813fb28d1f4611": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "b564a6710ab1477ba5aef3a87961913b": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "b7ea8f8585354f9bb4f11f43a6f1a8ac": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HBoxModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HBoxModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HBoxView",
+            "box_style": "",
+            "children": [
+              "IPY_MODEL_15a9e8f3433d478ba8f73dd158515e7d",
+              "IPY_MODEL_d4335bbecd3f4a9693a8c2742a765a4c",
+              "IPY_MODEL_cec02057108742e0b0f5bf5704cf6c6d"
+            ],
+            "layout": "IPY_MODEL_5d2b2e10194043efadf1377e961b1e41"
+          }
+        },
+        "c162aa27a92f496192a2335ab0496bc8": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "c74b88aa20f846dda524c0a5e03dc66e": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "c83d52312d02407a8ea124fa25ebddb7": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "c9077c3acb6342f2a248876f3ebf7464": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "c99b9fb60a4042b58b07a8524f8f555e": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "cb0206ba72d7496c91e2737a057f1fe1": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "cec02057108742e0b0f5bf5704cf6c6d": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_8f0a3034925d409f85c64f5271187923",
+            "placeholder": "​",
+            "style": "IPY_MODEL_e0cb22c893ae4ed88b24224347751da7",
+            "value": " 160/160 [00:00&lt;00:00, 13.2kB/s]"
+          }
+        },
+        "cef9fd8581614026950f8495952ccd03": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "d0422dfef6a846c5bbd9045b7019ffb8": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_039241d8c1c34103a655fbcb6de51ae6",
+            "placeholder": "​",
+            "style": "IPY_MODEL_fbcd9fb015d24543aedb6dbdb8e3d87a",
+            "value": " 212M/212M [00:01&lt;00:00, 148MB/s]"
+          }
+        },
+        "d14c61501c7d49ce90cb3c1d073cbaa8": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_d91cdc1965a74b7790e44126f204c708",
+            "placeholder": "​",
+            "style": "IPY_MODEL_2d028e6a22194fa28419e88e87573421",
+            "value": "Generating test split: 100%"
+          }
+        },
+        "d4335bbecd3f4a9693a8c2742a765a4c": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "FloatProgressModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "FloatProgressModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "ProgressView",
+            "bar_style": "success",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_2658dc163a3942608f583622c44f28ea",
+            "max": 160,
+            "min": 0,
+            "orientation": "horizontal",
+            "style": "IPY_MODEL_9519479ac1234c50ac6b5459cb2db76b",
+            "value": 160
+          }
+        },
+        "d88bcb77112743ed9c68f14af86d15f7": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "d91cdc1965a74b7790e44126f204c708": {
+          "model_module": "@jupyter-widgets/base",
+          "model_module_version": "1.2.0",
+          "model_name": "LayoutModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "db8dd16fc27a45d5a1ab81380b305389": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "e0cb22c893ae4ed88b24224347751da7": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "e3c612d1b46f45cbbe35ecdc39df73bf": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "FloatProgressModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "FloatProgressModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "ProgressView",
+            "bar_style": "success",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_6e53b375bc41426788d7a094287fbc22",
+            "max": 1217466031,
+            "min": 0,
+            "orientation": "horizontal",
+            "style": "IPY_MODEL_f3f6c7337c9c48eabc5ad3ba9a417dc1",
+            "value": 1217466031
+          }
+        },
+        "ec03ea80c45b452fa9365c13eaafec35": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HBoxModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HBoxModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HBoxView",
+            "box_style": "",
+            "children": [
+              "IPY_MODEL_8b2007e92a2f48e38c52e14a9e647528",
+              "IPY_MODEL_e3c612d1b46f45cbbe35ecdc39df73bf",
+              "IPY_MODEL_87af41d1980245e1a3edacfb4a0b2b12"
+            ],
+            "layout": "IPY_MODEL_b564a6710ab1477ba5aef3a87961913b"
+          }
+        },
+        "ee4079637e6643d48a9feba34c30d0d7": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "FloatProgressModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "FloatProgressModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "ProgressView",
+            "bar_style": "success",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_4113c7318d5d4484b3990dcbdb9bc1cd",
+            "max": 156,
+            "min": 0,
+            "orientation": "horizontal",
+            "style": "IPY_MODEL_1b8330ce7031406c95eac80280936dae",
+            "value": 156
+          }
+        },
+        "ef44e725caed4ef38e9cc25621b76406": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_5aa5164fa3d040bc8180b122eb029d94",
+            "placeholder": "​",
+            "style": "IPY_MODEL_04fb4c819c3c44b59184c54ea44456bf",
+            "value": "train-00000-of-00001.parquet: 100%"
+          }
+        },
+        "f1a5e1f980d9427095d24629d2a83ae4": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HTMLModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_c74b88aa20f846dda524c0a5e03dc66e",
+            "placeholder": "​",
+            "style": "IPY_MODEL_03871e390dfd485198461b5e3ba6faed",
+            "value": "model.safetensors: 100%"
+          }
+        },
+        "f2566e5e755e4a72acbebab9c938aa34": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HBoxModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HBoxModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HBoxView",
+            "box_style": "",
+            "children": [
+              "IPY_MODEL_d14c61501c7d49ce90cb3c1d073cbaa8",
+              "IPY_MODEL_ee4079637e6643d48a9feba34c30d0d7",
+              "IPY_MODEL_2ddd4fe77259428aaf7d6a56ed0b223e"
+            ],
+            "layout": "IPY_MODEL_a86b7af977bc40fea33a27748e57f432"
+          }
+        },
+        "f3267441b9dc4f029a3bfdffaf491842": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "f3f6c7337c9c48eabc5ad3ba9a417dc1": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "ProgressStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "ProgressStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "bar_color": null,
+            "description_width": ""
+          }
+        },
+        "f511728e7deb40029950ff8019742162": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "fb7a5bc6781c4794a8bef2ab3a7b9116": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "fbcd9fb015d24543aedb6dbdb8e3d87a": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "DescriptionStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "fe8fca891aff479a9fb11aeeb1971e2d": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "ProgressStyleModel",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "ProgressStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "bar_color": null,
+            "description_width": ""
+          }
+        },
+        "ff9d81edcfe9421bb1398ca284a197cc": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_module_version": "1.5.0",
+          "model_name": "HBoxModel",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HBoxModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HBoxView",
+            "box_style": "",
+            "children": [
+              "IPY_MODEL_ef44e725caed4ef38e9cc25621b76406",
+              "IPY_MODEL_4358acace42c41519fc83cc67e814961",
+              "IPY_MODEL_d0422dfef6a846c5bbd9045b7019ffb8"
+            ],
+            "layout": "IPY_MODEL_228c08de964d48fdbb3594c1f2ecde09"
+          }
+        }
+      }
+    }
+  },
+  "nbformat": 4,
+  "nbformat_minor": 5
+}
diff --git a/notebooks/ko/index.md b/notebooks/ko/index.md
index c68982d8..034075df 100644
--- a/notebooks/ko/index.md
+++ b/notebooks/ko/index.md
@@ -6,6 +6,7 @@
 
 최근 추가된 노트북을 살펴보세요:
 
+- [Custom Biomedical Dataset으로 Vision Transformer 모델 Fine-tuning 하기](fine_tuning_vit_custom_dataset_ko)
 - [한국어 Advanced RAG 구현: Hugging Face와 LangChain 활용한 Cookbook](advanced_ko_rag)
 - [구조화된 생성으로 근거 강조 표시가 있는 RAG 시스템 구축하기](structured_generation)
 - [지식 그래프를 활용한 RAG 추론 향상](ko_rag_with_knowledge_graphs_neo4j)

From 2bb56bf372617735227e090903b82e2efd58e5b2 Mon Sep 17 00:00:00 2001
From: HyunZ118 <156191095+HyunZ118@users.noreply.github.com>
Date: Mon, 9 Jun 2025 22:59:55 +0900
Subject: [PATCH 2/2] Edited part of the translation

---
 .../fine_tuning_vit_custom_dataset_ko.ipynb   | 157 ++++++++++--------
 1 file changed, 91 insertions(+), 66 deletions(-)

diff --git a/notebooks/ko/fine_tuning_vit_custom_dataset_ko.ipynb b/notebooks/ko/fine_tuning_vit_custom_dataset_ko.ipynb
index fc5712ed..53a4cc71 100644
--- a/notebooks/ko/fine_tuning_vit_custom_dataset_ko.ipynb
+++ b/notebooks/ko/fine_tuning_vit_custom_dataset_ko.ipynb
@@ -42,7 +42,7 @@
     },
     {
       "cell_type": "code",
-      "execution_count": 1,
+      "execution_count": null,
       "id": "7093dd4f-d0cb-44dc-935d-d54435187901",
       "metadata": {
         "colab": {
@@ -56,6 +56,16 @@
         "# !pip install datasets transformers accelerate torch torchvision scikit-learn matplotlib wandb"
       ]
     },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "id": "eec0007b",
+      "metadata": {},
+      "outputs": [],
+      "source": [
+        "# !pip install --upgrade datasets # Updated by HJ (colab 환경에서 실험 중일 경우 최신 버전 사용을 위해 업그레이드 필요)"
+      ]
+    },
     {
       "cell_type": "markdown",
       "id": "9b5019a8-d130-4c08-9503-cd8415f50ae9",
@@ -93,23 +103,7 @@
       },
       "source": [
         "## 데이터셋 준비\n",
-        "'datasets' 라이브러리는 데이터셋에서 이미지와 클래스를 자동으로 불러옵니다. 자세한 내용은 [`이 링크`](https://huggingface.co/docs/datasets/image_load).를 참고하세요."
-      ]
-    },
-    {
-      "cell_type": "code",
-      "execution_count": null,
-      "id": "C2f-_NcrnYil",
-      "metadata": {
-        "colab": {
-          "base_uri": "https://localhost:8080/"
-        },
-        "id": "C2f-_NcrnYil",
-        "outputId": "27112e94-72ec-4dba-d848-a870d6699737"
-      },
-      "outputs": [],
-      "source": [
-        "# pip install --upgrade datasets # update by HJ"
+        "`datasets` 라이브러리는 데이터셋에서 이미지와 클래스를 자동으로 불러옵니다. 자세한 내용은 [이 링크](https://huggingface.co/docs/datasets/image_load)를 참고하세요."
       ]
     },
     {
@@ -209,12 +203,6 @@
         "dataset"
       ]
     },
-    {
-      "cell_type": "markdown",
-      "id": "3b0461c4",
-      "metadata": {},
-      "source": []
-    },
     {
       "cell_type": "markdown",
       "id": "bf3176b2-65ff-44ea-bb1e-cd75b2d7609d",
@@ -222,16 +210,16 @@
         "id": "bf3176b2-65ff-44ea-bb1e-cd75b2d7609d"
       },
       "source": [
-        "데이터셋을 불러왔지만, 현재 검증 세트는 없습니다.\n",
+        "데이터셋을 불러왔지만, 해당 데이터셋에는 검증 세트가 분할되어 있지 않은 상황입니다.\n",
         "\n",
-        "검증 세트를 만들기 위해 테스트 세트의 크기를 기준으로 훈련 세트의 일부를 나누어 검증 세트를 생성할 것입니다.\n",
+        "검증 세트를 만들기 위해 테스트 세트의 크기를 기준으로 훈련 세트의 일부를 나누어 검증 세트를 생성하려고 합니다.\n",
         "\n",
         "그 후 기존 훈련 세트를 새로운 훈련 세트와 검증 세트로 분할합니다."
       ]
     },
     {
       "cell_type": "code",
-      "execution_count": 5,
+      "execution_count": null,
       "id": "4d385bab-dad0-462d-9339-11205fcd2e63",
       "metadata": {
         "colab": {
@@ -262,7 +250,7 @@
         }
       ],
       "source": [
-        "# Get the numbers of each set\n",
+        "# 각 세트의 데이터 수를 구해봅니다\n",
         "test_num = len(dataset[\"test\"])\n",
         "train_num = len(dataset[\"train\"])\n",
         "\n",
@@ -336,9 +324,9 @@
         "id": "4a0d7a89-099e-4925-97a5-51c91cdb7046"
       },
       "source": [
-        "잘하셨습니다! 이제 데이터셋 준비가 완료되었습니다.\n",
+        "좋습니다! 이제 데이터셋 준비가 완료되었네요 👍🏻\n",
         "\n",
-        "각 하위 데이터셋을 서로 다른 변수에 할당해봅시다. (나중에 쉽게 참조할 수 있도록 하기 위함입니다.)"
+        "나중에 쉽게 참조할 수 있도록 각 하위 집합을 변수에 할당해 둡시다."
       ]
     },
     {
@@ -362,7 +350,7 @@
         "id": "27e5451f-e2fb-4d29-b63a-a8960686dcc2"
       },
       "source": [
-        "이미지를 확인해보면 'PIL.Image' 형식이며 이에 연결된 라벨이 함께 존재하는 것을 확인할 수 있습니다."
+        "이미지를 확인해보면 `PIL.Image` 형식이며 이에 연결된 레이블이 함께 존재하는 것을 확인할 수 있습니다."
       ]
     },
     {
@@ -443,7 +431,7 @@
     },
     {
       "cell_type": "code",
-      "execution_count": 10,
+      "execution_count": null,
       "id": "5c901865-a876-4b4b-b1f2-8895b494cafb",
       "metadata": {
         "colab": {
@@ -468,13 +456,13 @@
       "source": [
         "import matplotlib.pyplot as plt\n",
         "\n",
-        "# Initialize a set to keep track of shown labels\n",
+        "# 표시된 레이블을 추적할 수 있게 set를 초기화해줍니다\n",
         "shown_labels = set()\n",
         "\n",
-        "# Initialize the figure for plotting\n",
+        "# 그래프를 그리기 위해 figure를 초기화해줍니다\n",
         "plt.figure(figsize=(10, 10))\n",
         "\n",
-        "# Loop through the dataset and plot the first image of each label\n",
+        "# 데이터셋을 하나씩 살펴보면서 각 라벨마다 첫 번째 이미지를 그려줍니다\n",
         "for i, sample in enumerate(train_ds):\n",
         "    label = train_ds.features['label'].names[sample['label']]\n",
         "    if label not in shown_labels:\n",
@@ -499,11 +487,11 @@
         "\n",
         "\n",
         "## 데이터 전처리\n",
-        "데이터셋은 준비되었지만 아직 파인튜닝을 위한 준비는 완료되지 않았습니다. 아래와 같은 절차를 순차적으로 따라갈 것입니다:\n",
+        "데이터셋은 준비되었지만 파인튜닝을 위한 준비는 아직 완료되지 않았습니다. 아래와 같은 절차를 순차적으로 따라가보겠습니다!\n",
         "\n",
         "- **레이블 매핑 (Label Mapping)** : 레이블 ID와 해당 클래스 이름 간의 매핑을 수행합니다. 이는 모델 학습과 평가 시 유용하게 사용됩니다.\n",
         "\n",
-        "- **이미지 전처리 (Image Processing)** : ViTImageProcessor를 사용해 입력 이미지의 크기를 표준화하고, 사전 학습된 모델에 맞는 정규화를 적용합니다. 또한, torchvision을 활용해 훈련, 검증, 테스트용으로 서로 다른 변환을 정의하여 모델의 일반화를 향상시킵니다.\n",
+        "- **이미지 전처리 (Image Processing)** : `ViTImageProcessor`를 사용해 입력 이미지의 크기를 표준화하고, 사전 학습된 모델에 맞는 정규화를 적용합니다. 또한, `torchvision`을 활용해 훈련, 검증, 테스트용으로 서로 다른 변환을 정의하여 모델의 일반화를 향상시킵니다.\n",
         "\n",
         "- **변환 함수 정의 (Transform Functions)** : ViT 모델에 맞는 형식과 크기로 이미지를 변환하기 위한 함수를 구현합니다.\n",
         "\n",
@@ -772,8 +760,14 @@
         "id": "49b06fa4-6f28-4e45-8af4-779c424583fe"
       },
       "source": [
-        "픽셀 값이 텐서 값으로 잘 변환하였습니다.\n",
-        "\n",
+        "픽셀 값이 텐서로 잘 변환되었네요 👍🏻"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "74b6c684",
+      "metadata": {},
+      "source": [
         "### 데이터 로딩"
       ]
     },
@@ -842,7 +836,9 @@
         "id": "35379d35-e567-4a30-8b91-6eac80b79044"
       },
       "source": [
-        "완벽합니다! 이제 파인튜닝 과정을 진행할 준비가 되었습니다."
+        "완벽합니다👏🏻 이제 파인튜닝 진행을 위한 모든 준비가 되었습니다.\n",
+        "\n",
+        "이어서 모델을 설정하고 파인튜닝하는 과정을 진행하겠습니다."
       ]
     },
     {
@@ -853,13 +849,17 @@
       },
       "source": [
         "## 모델 파인튜닝\n",
-        "이어서 모델을 설정하고 파인튜닝하는 과정을 진행하겠습니다.\n",
-        "\n",
-        "우선, 특정 라벨 매핑과 사전 학습 설정(pre-trained settings)을 사용해 모델을 초기화하고, 크기 불일치를 조정해줍니다. 그다음, 학습 과정을 정의하기 위해 저장 전략, 배치 크기, 학습 epoch 등의 학습 파라미터를 설정하고, Weights & Biases를 통해 학습 결과를 로깅합니다.\n",
-        "이후 Hugging Face의 Trainer가 인스턴스화되어 학습과 평가를 관리하며, 커스텀 데이터 콜레이터와 모델에 내장된 프로세서를 함께 사용합니다. 마지막으로, 학습이 완료된 모델을 테스트 데이터셋으로 평가하고, 정확도를 확인하기 위한 지표(metrics)를 출력합니다.\n",
-        "\n",
         "\n",
+        "우선, 특정 레이블 매핑과 사전 학습 설정(pre-trained settings)을 사용해 모델을 초기화하고, 크기 불일치를 조정해줍니다. 그 다음, 학습 과정을 정의하기 위해 저장 전략, 배치 크기, 학습 epoch 등의 학습 파라미터를 설정하고, Weights & Biases를 통해 학습 결과를 로깅합니다.\n",
         "\n",
+        "이후 Hugging Face의 Trainer가 인스턴스화되어 학습과 평가를 관리하며, 커스텀 데이터 collator와 모델에 내장된 Processor를 함께 사용합니다. 마지막으로 학습이 완료된 모델을 테스트 데이터셋으로 평가하고, 정확도를 확인하기 위한 지표(metrics)를 출력합니다.\n"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "bfb2482b",
+      "metadata": {},
+      "source": [
         "첫번째로, 모델을 먼저 호출해보겠습니다."
       ]
     },
@@ -935,16 +935,14 @@
         "id": "2f773e18-63e8-41bd-885f-7ba95d074a3d"
       },
       "source": [
-        "여기에는 미묘한 디테일이 하나 있습니다. 바로 `ignore_mismatched_sizes` 파라미터 입니다.\n",
-        "\n",
-        "사전 학습된(pre-trained) 모델을 새로운 데이터셋에 파인튜닝할 때, 종종 이미지의 입력 크기나 모델 구조(예: 분류 레이어의 클래스 수 등)가 원래 모델이 학습된 환경과 완전히 일치하지 않을 수 있습니다.\n",
-        "이러한 상황은 다양한 이유로 발생할 수 있는데, 예를 들어 ImageNet처럼 자연 이미지로 학습된 모델을 의료 영상이나 특수 카메라 이미지처럼 전혀 다른 유형의 이미지에 적용하려고 할 때 발생할 수 있습니다.\n",
+        "이 부분에서 언급하고 싶은 디테일이 있습니다.`ignore_mismatched_sizes` 파라미터에 주목해주세요 👀\n",
         "\n",
+        "사전 학습된(pre-trained) 모델을 새로운 데이터셋에 파인튜닝할 때, 종종 이미지의 입력 크기나 모델 구조(예: 분류 레이어의 클래스 수 등)가 원래 모델이 학습된 환경과 완전히 일치하지 않을 수 있습니다. 이러한 상황은 다양한 이유로 발생할 수 있습니다. 예를 들어 ImageNet처럼 자연 이미지로 학습된 모델을 의료 영상이나 특수 카메라 이미지처럼 전혀 다른 유형의 이미지에 적용하려고 할 때 발생하는데요.\n",
         "\n",
         "이때, `ignore_mismatched_sizes` 를 `True` 로 설정하면, 사이즈가 일치하지 않더라도 오류 없이 레이어를 자동으로 조정해 사용할 수 있게 해줍니다.\n",
         "\n",
         "\n",
-        "예를 들어, 해당 모델은 1000개의 클래스로 학습되어 있어서 출력 크기가 `torch.Size([1000])` 이고 입력 또한 `torch.Size([1000])` 클래스를 기대합니다. 그런데 우리가 사용하는 데이터셋은 클래스가 3개로 `torch.Size([3])`입니다. 이 경우 모델에 직접 입력하면 클래스 수가 맞지 않기 때문에 오류가 발생하게 됩니다."
+        "예를 들어, 해당 모델은 1000개의 클래스로 학습되어 있어서 출력 크기가 `torch.Size([1000])` 이고 입력 또한 `torch.Size([1000])` 클래스를 기대합니다. 그런데 우리가 사용하는 데이터셋은 클래스가 3개 `torch.Size([3])`이고, 이 경우 모델에 직접 입력하면 클래스 수가 맞지 않기 때문에 오류가 발생하게 됩니다."
       ]
     },
     {
@@ -954,7 +952,7 @@
         "id": "e2c671bf-9978-46d8-82ef-0906d4e89d03"
       },
       "source": [
-        "그다음 Google에서 제공하는 학습 인자(arguments)를 정의합니다."
+        "그다음 Google에서 제공하는 학습 인자(arguments)를 정의해봅시다."
       ]
     },
     {
@@ -969,7 +967,7 @@
     },
     {
       "cell_type": "code",
-      "execution_count": 21,
+      "execution_count": null,
       "id": "f16fc568-9fdc-4c60-acec-6ed3dbb85aef",
       "metadata": {
         "id": "f16fc568-9fdc-4c60-acec-6ed3dbb85aef"
@@ -1008,9 +1006,10 @@
         ")\n",
         "\n",
         "\"\"\"\n",
-        "현재 환경에서 TrainingArguments가 해당 인자들을 인식하지 못하는 오류(TypeError)가 발생하여\n",
-        "최소한의 설정만으로 Trainer가 동작하도록 수정한 버전임.\n",
-        "(중간값 저장이나 자동 평가 없이 단순 학습만 수행됨.)\n",
+        "# Updated by HJ\n",
+        "(현재 환경에서 TrainingArguments가 해당 인자들을 인식하지 못하는 오류(TypeError)가 발생하여\n",
+        "최소한의 설정만으로 Trainer가 동작하도록 수정한 버전임. \n",
+        "중간값 저장이나 자동 평가 없이 단순 학습만 수행됨.)\n",
         "\"\"\"\n"
       ]
     },
@@ -1121,10 +1120,17 @@
         "id": "5f74c058-e2a5-4c9c-8d70-1c2c574b933f"
       },
       "source": [
-        "좋습니다! 이제 결과를 시각화해봅시다.\n",
-        "\n",
+        "좋습니다! 이제 결과를 시각화해볼까요?"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "49f0ce76",
+      "metadata": {},
+      "source": [
         "## 결과\n",
         "파인튜닝을 마쳤으니, 이제 모델이 각 클래스를 어떻게 예측했는지 확인해보겠습니다.\n",
+        "\n",
         "이를 위해 scikit-learn의 Confusion Matrix Display를 사용하고, Recall 점수도 함께 확인해보겠습니다."
       ]
     },
@@ -1137,9 +1143,15 @@
       "source": [
         "### Confusion Matrix 란?\n",
         "Confusion Matrix(혼동 행렬)는 알고리즘, 특히 지도 학습 모델이 실제 정답을 알고 있는 테스트 데이터에 대해 얼마나 잘 작동하는지를 시각적으로 보여주는 표 형식입니다.\n",
-        "예측된 라벨과 실제 라벨의 빈도를 나타내므로, 분류 모델의 성능을 평가할 때 특히 유용합니다.\n",
-        "\n",
-        "우리 모델의 Confusion Matrix를 그려봅시다."
+        "예측된 라벨과 실제 라벨의 빈도를 나타내므로 분류 모델의 성능을 평가할 때 특히 유용합니다."
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "014ff009",
+      "metadata": {},
+      "source": [
+        "우리 모델의 Confusion Matrix를 시각화해봅시다."
       ]
     },
     {
@@ -1200,7 +1212,7 @@
     },
     {
       "cell_type": "code",
-      "execution_count": 26,
+      "execution_count": null,
       "id": "48d87ca7-8458-41d5-a773-38e2c9522f64",
       "metadata": {
         "id": "48d87ca7-8458-41d5-a773-38e2c9522f64",
@@ -1220,11 +1232,11 @@
       "source": [
         "from sklearn.metrics import recall_score\n",
         "\n",
-        "# Calculate the recall scores\n",
-        "# 'None' calculates recall for each class separately\n",
+        "# Recall score 계산\n",
+        "# ‘None’으로 설정하면 각 클래스별로 개별적으로 recall을 계산합니다.\n",
         "recall = recall_score(y_true, y_pred, average=None)\n",
         "\n",
-        "# Print the recall for each class\n",
+        "# 각 클래스별로 recall score를 출력합니다.\n",
         "for label, score in zip(labels, recall):\n",
         "    print(f'Recall for {label}: {score:.2f}')\n"
       ]
@@ -1250,10 +1262,23 @@
       "source": [
         "## 결론\n",
         "이 cookbook에서는 의료 데이터셋을 활용해 ViT 모델을 학습하는 과정을 다루었습니다.\n",
-        "데이터셋 준비, 이미지 전처리, 모델 구성, 학습, 평가, 그리고 결과 시각화까지의 핵심 단계들을 모두 포함하고 있습니다.\n",
         "\n",
-        "Hugging Face의 Transformers 라이브러리와 scikit-learn, PyTorch의 Torchvision을 활용하여 효율적인 모델 학습과 평가가 가능하며, 이를 통해 바이오메디컬 이미지를 얼마나 정확하게 분류할 수 있는지에 대한 유용한 통찰을 제공합니다."
+        "데이터셋 준비, 이미지 전처리, 모델 구성, 학습, 평가, 그리고 결과 시각화까지의 핵심 단계들을 모두 포함하고 있습니다."
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "9f359f99",
+      "metadata": {},
+      "source": [
+        "Hugging Face의 `transformers` 라이브러리와 `scikit-learn`, PyTorch의 `torchvision`을 활용하여 효율적인 모델 학습과 평가가 가능하며, 이를 통해 바이오 메디컬 이미지를 얼마나 정확하게 분류할 수 있는지에 대한 유용한 인사이트를 제공합니다."
       ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "009f90ee",
+      "metadata": {},
+      "source": []
     }
   ],
   "metadata": {