Buckets:

hf-doc-build
/

doc-dev

Files

xet

hf-doc-build/doc-dev / course /pr_1054 /id /chapter3 /2.html

rtrm

about 1 month ago

download

raw

87.2 kB

	<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{"title":"Memproses Data","local":"processing-the-data","sections":[{"title":"Memuat Dataset dari Hub","local":"loading-a-dataset-from-the-hub","sections":[],"depth":3},{"title":"Pra-pemrosesan Dataset","local":"preprocessing-a-dataset","sections":[{"title":"Padding Dinamis","local":"dynamic-padding","sections":[],"depth":4}],"depth":3},{"title":"Kuis Bagian","local":"section-quiz","sections":[{"title":"1. Apa keuntungan utama menggunakan Dataset.map() dengan batched=True ?","local":"1-apa-keuntungan-utama-menggunakan-datasetmap-dengan-batchedtrue-","sections":[],"depth":3},{"title":"2. Mengapa kita menggunakan padding dinamis daripada mem-pad semua urutan ke panjang maksimum dalam dataset?","local":"2-mengapa-kita-menggunakan-padding-dinamis-daripada-mem-pad-semua-urutan-ke-panjang-maksimum-dalam-dataset","sections":[],"depth":3},{"title":"3. Apa yang direpresentasikan oleh field token_type_ids dalam tokenisasi BERT?","local":"3-apa-yang-direpresentasikan-oleh-field-tokentypeids-dalam-tokenisasi-bert","sections":[],"depth":3},{"title":"4. Saat memuat dataset dengan load_dataset('glue', 'mrpc') , apa arti dari argumen kedua?","local":"4-saat-memuat-dataset-dengan-loaddatasetglue-mrpc--apa-arti-dari-argumen-kedua","sections":[],"depth":3},{"title":"5. Apa tujuan dari menghapus kolom seperti ‘sentence1’ dan ‘sentence2’ sebelum pelatihan?","local":"5-apa-tujuan-dari-menghapus-kolom-seperti-sentence1-dan-sentence2-sebelum-pelatihan","sections":[],"depth":3}],"depth":2}],"depth":1}">
	<link href="/docs/course/pr_1054/id/_app/immutable/assets/0.e3b0c442.css" rel="modulepreload">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/entry/start.4f92af03.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/chunks/scheduler.36a0863c.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/chunks/singletons.7dc7b9a4.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/chunks/index.733708bb.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/chunks/paths.cf097d06.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/entry/app.19cef1b6.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/chunks/index.156fee99.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/nodes/0.1203e4a0.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/chunks/each.e59479a4.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/nodes/24.913059f1.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/chunks/Tip.8a648467.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/chunks/Youtube.a5d6d567.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/chunks/CodeBlock.4cf998e6.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/chunks/CourseFloatingBanner.16bb8bff.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/chunks/Question.aeeaa180.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/chunks/stores.8f4efe8a.js">
	<link rel="modulepreload" href="/docs/course/pr_1054/id/_app/immutable/chunks/getInferenceSnippets.472bc46d.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{"title":"Memproses Data","local":"processing-the-data","sections":[{"title":"Memuat Dataset dari Hub","local":"loading-a-dataset-from-the-hub","sections":[],"depth":3},{"title":"Pra-pemrosesan Dataset","local":"preprocessing-a-dataset","sections":[{"title":"Padding Dinamis","local":"dynamic-padding","sections":[],"depth":4}],"depth":3},{"title":"Kuis Bagian","local":"section-quiz","sections":[{"title":"1. Apa keuntungan utama menggunakan Dataset.map() dengan batched=True ?","local":"1-apa-keuntungan-utama-menggunakan-datasetmap-dengan-batchedtrue-","sections":[],"depth":3},{"title":"2. Mengapa kita menggunakan padding dinamis daripada mem-pad semua urutan ke panjang maksimum dalam dataset?","local":"2-mengapa-kita-menggunakan-padding-dinamis-daripada-mem-pad-semua-urutan-ke-panjang-maksimum-dalam-dataset","sections":[],"depth":3},{"title":"3. Apa yang direpresentasikan oleh field token_type_ids dalam tokenisasi BERT?","local":"3-apa-yang-direpresentasikan-oleh-field-tokentypeids-dalam-tokenisasi-bert","sections":[],"depth":3},{"title":"4. Saat memuat dataset dengan load_dataset('glue', 'mrpc') , apa arti dari argumen kedua?","local":"4-saat-memuat-dataset-dengan-loaddatasetglue-mrpc--apa-arti-dari-argumen-kedua","sections":[],"depth":3},{"title":"5. Apa tujuan dari menghapus kolom seperti ‘sentence1’ dan ‘sentence2’ sebelum pelatihan?","local":"5-apa-tujuan-dari-menghapus-kolom-seperti-sentence1-dan-sentence2-sebelum-pelatihan","sections":[],"depth":3}],"depth":2}],"depth":1}"><!-- HEAD_svelte-u9bgzb_END --> <p></p> <h1 class="relative group"><a id="processing-the-data" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#processing-the-data"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Memproses Data</span></h1> <div class="flex space-x-1 absolute z-10 right-0 top-0"><a href="https://discuss.huggingface.co/t/chapter-3-questions" target="_blank"><img alt="Ask a Question" class="!m-0" src="https://img.shields.io/badge/Ask%20a%20question-ffcb4c.svg?logo=data:image/svg+xml;base64,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"></a> <a href="https://colab.research.google.com/github/huggingface/notebooks/blob/master/course/en/chapter3/section2.ipynb" target="_blank"><img alt="Open In Colab" class="!m-0" src="https://colab.research.google.com/assets/colab-badge.svg"></a> <a href="https://studiolab.sagemaker.aws/import/github/huggingface/notebooks/blob/master/course/en/chapter3/section2.ipynb" target="_blank"><img alt="Open In Studio Lab" class="!m-0" src="https://studiolab.sagemaker.aws/studiolab.svg"></a></div> <p data-svelte-h="svelte-m6xio4">Melanjutkan contoh dari <a href="/course/chapter2">bab sebelumnya</a>, berikut adalah cara kita melatih classifier urutan pada satu batch:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-keyword">import</span> torch
	<span class="hljs-keyword">from</span> torch.optim <span class="hljs-keyword">import</span> AdamW
	<span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoTokenizer, AutoModelForSequenceClassification

	<span class="hljs-comment"># Sama seperti sebelumnya</span>
	checkpoint = <span class="hljs-string">"bert-base-uncased"</span>
	tokenizer = AutoTokenizer.from_pretrained(checkpoint)
	model = AutoModelForSequenceClassification.from_pretrained(checkpoint)
	sequences = [
	<span class="hljs-string">"I've been waiting for a HuggingFace course my whole life."</span>,
	<span class="hljs-string">"This course is amazing!"</span>,
	]
	batch = tokenizer(sequences, padding=<span class="hljs-literal">True</span>, truncation=<span class="hljs-literal">True</span>, return_tensors=<span class="hljs-string">"pt"</span>)

	<span class="hljs-comment"># Ini yang baru</span>
	batch[<span class="hljs-string">"labels"</span>] = torch.tensor([<span class="hljs-number">1</span>, <span class="hljs-number">1</span>])

	optimizer = AdamW(model.parameters())
	loss = model(**batch).loss
	loss.backward()
	optimizer.step()<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-qtabp2">Tentu saja, hanya melatih model dengan dua kalimat tidak akan memberikan hasil yang baik. Untuk hasil yang lebih baik, Anda perlu menyiapkan dataset yang lebih besar.</p> <p data-svelte-h="svelte-iv3tbs">Di bagian ini kita akan menggunakan dataset MRPC (Microsoft Research Paraphrase Corpus) sebagai contoh, yang diperkenalkan dalam sebuah <a href="https://www.aclweb.org/anthology/I05-5002.pdf" rel="nofollow">makalah</a> oleh William B. Dolan dan Chris Brockett. Dataset ini terdiri dari 5.801 pasangan kalimat, dengan label yang menunjukkan apakah keduanya merupakan parafrasa (yaitu, apakah kedua kalimat memiliki makna yang sama). Dataset ini dipilih karena ukurannya kecil sehingga mudah untuk digunakan dalam eksperimen pelatihan.</p> <h3 class="relative group"><a id="loading-a-dataset-from-the-hub" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#loading-a-dataset-from-the-hub"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Memuat Dataset dari Hub</span></h3> <iframe class="w-full xl:w-4/6 h-80" src="https://www.youtube-nocookie.com/embed/_BZearw7f0w" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe> <p data-svelte-h="svelte-xgeiln">Hub tidak hanya berisi model; ia juga memiliki banyak dataset dalam berbagai bahasa. Anda dapat menjelajahi dataset <a href="https://huggingface.co/datasets" rel="nofollow">di sini</a>, dan kami sarankan Anda mencoba memuat dan memproses dataset baru setelah menyelesaikan bagian ini (lihat dokumentasi umum <a href="https://huggingface.co/docs/datasets/loading" rel="nofollow">di sini</a>). Untuk saat ini, kita fokus pada dataset MRPC! Dataset ini adalah salah satu dari 10 dataset dalam <a href="https://gluebenchmark.com/" rel="nofollow">benchmark GLUE</a>, yang digunakan untuk mengukur performa model ML dalam 10 tugas klasifikasi teks berbeda.</p> <p data-svelte-h="svelte-1fd6ii0">Modul 🤗 Datasets menyediakan perintah sederhana untuk mengunduh dan menyimpan dataset dari Hub. Kita bisa mengunduh dataset MRPC seperti ini:</p> <div class="course-tip bg-gradient-to-br dark:bg-gradient-to-r before:border-green-500 dark:before:border-green-800 from-green-50 dark:from-gray-900 to-white dark:to-gray-950 border border-green-50 text-green-700 dark:text-gray-400"><p data-svelte-h="svelte-9mb3s2">💡 <strong>Sumber Tambahan</strong>: Untuk teknik dan contoh pemuatan dataset lainnya, lihat <a href="https://huggingface.co/docs/datasets/" rel="nofollow">dokumentasi 🤗 Datasets</a>.</p></div> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-keyword">from</span> datasets <span class="hljs-keyword">import</span> load_dataset

	raw_datasets = load_dataset(<span class="hljs-string">"glue"</span>, <span class="hljs-string">"mrpc"</span>)
	raw_datasets<!-- HTML_TAG_END --></pre></div> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->DatasetDict({
	train: Dataset({
	features: [<span class="hljs-string">'sentence1'</span>, <span class="hljs-string">'sentence2'</span>, <span class="hljs-string">'label'</span>, <span class="hljs-string">'idx'</span>],
	num_rows: <span class="hljs-number">3668</span>
	})
	validation: Dataset({
	features: [<span class="hljs-string">'sentence1'</span>, <span class="hljs-string">'sentence2'</span>, <span class="hljs-string">'label'</span>, <span class="hljs-string">'idx'</span>],
	num_rows: <span class="hljs-number">408</span>
	})
	test: Dataset({
	features: [<span class="hljs-string">'sentence1'</span>, <span class="hljs-string">'sentence2'</span>, <span class="hljs-string">'label'</span>, <span class="hljs-string">'idx'</span>],
	num_rows: <span class="hljs-number">1725</span>
	})
	})<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-ex4pxw">Seperti yang Anda lihat, kita mendapatkan objek <code>DatasetDict</code> yang berisi data pelatihan, validasi, dan pengujian. Masing-masing memiliki kolom (<code>sentence1</code>, <code>sentence2</code>, <code>label</code>, dan <code>idx</code>) dan sejumlah baris (misalnya, 3.668 pasangan kalimat di pelatihan, 408 di validasi, dan 1.725 di pengujian).</p> <div class="course-tip bg-gradient-to-br dark:bg-gradient-to-r before:border-green-500 dark:before:border-green-800 from-green-50 dark:from-gray-900 to-white dark:to-gray-950 border border-green-50 text-green-700 dark:text-gray-400"><p data-svelte-h="svelte-6mrsv2">Perintah ini akan mengunduh dan menyimpan dataset secara lokal di <em>~/.cache/huggingface/datasets</em> secara default. Ingat dari Bab 2 bahwa Anda dapat menyesuaikan folder cache dengan mengatur variabel lingkungan <code>HF_HOME</code>.</p></div> <p data-svelte-h="svelte-yj32ip">Kita dapat mengakses setiap pasangan kalimat dengan melakukan indexing seperti dictionary:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->raw_train_dataset = raw_datasets[<span class="hljs-string">"train"</span>]
	raw_train_dataset[<span class="hljs-number">0</span>]<!-- HTML_TAG_END --></pre></div> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->{<span class="hljs-string">'idx'</span>: <span class="hljs-number">0</span>,
	<span class="hljs-string">'label'</span>: <span class="hljs-number">1</span>,
	<span class="hljs-string">'sentence1'</span>: <span class="hljs-string">'Amrozi accused his brother , whom he called " the witness " , of deliberately distorting his evidence .'</span>,
	<span class="hljs-string">'sentence2'</span>: <span class="hljs-string">'Referring to him as only " the witness " , Amrozi accused his brother of deliberately distorting his evidence .'</span>}<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-vpv77k">Kita dapat melihat bahwa label-labelnya sudah berupa bilangan bulat, jadi kita tidak perlu melakukan praproses pada bagian tersebut. Untuk mengetahui bilangan bulat mana yang sesuai dengan label tertentu, kita bisa memeriksa <code>features</code> dari <code>raw_train_dataset</code> kita. Ini akan memberi tahu kita tipe dari setiap kolom:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->raw_train_dataset.features<!-- HTML_TAG_END --></pre></div> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->{<span class="hljs-string">'sentence1'</span>: Value(dtype=<span class="hljs-string">'string'</span>, <span class="hljs-built_in">id</span>=<span class="hljs-literal">None</span>),
	<span class="hljs-string">'sentence2'</span>: Value(dtype=<span class="hljs-string">'string'</span>, <span class="hljs-built_in">id</span>=<span class="hljs-literal">None</span>),
	<span class="hljs-string">'label'</span>: ClassLabel(num_classes=<span class="hljs-number">2</span>, names=[<span class="hljs-string">'not_equivalent'</span>, <span class="hljs-string">'equivalent'</span>], names_file=<span class="hljs-literal">None</span>, <span class="hljs-built_in">id</span>=<span class="hljs-literal">None</span>),
	<span class="hljs-string">'idx'</span>: Value(dtype=<span class="hljs-string">'int32'</span>, <span class="hljs-built_in">id</span>=<span class="hljs-literal">None</span>)}<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1qq5oxx">Di balik layar, <code>label</code> bertipe <code>ClassLabel</code>, dan pemetaan dari bilangan bulat ke nama label disimpan di dalam folder <em>names</em>. <code>0</code> berarti <code>not_equivalent</code>, dan <code>1</code> berarti <code>equivalent</code>.</p> <div class="course-tip bg-gradient-to-br dark:bg-gradient-to-r before:border-green-500 dark:before:border-green-800 from-green-50 dark:from-gray-900 to-white dark:to-gray-950 border border-green-50 text-green-700 dark:text-gray-400"><p data-svelte-h="svelte-x5yd5y">✏️ <strong>Coba Sendiri!</strong> Lihat elemen ke-15 dari data pelatihan dan elemen ke-87 dari data validasi. Apa label mereka?</p></div> <h3 class="relative group"><a id="preprocessing-a-dataset" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#preprocessing-a-dataset"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Pra-pemrosesan Dataset</span></h3> <iframe class="w-full xl:w-4/6 h-80" src="https://www.youtube-nocookie.com/embed/0u3ioSwev3s" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe> <p data-svelte-h="svelte-1dt0sxw">Untuk melakukan praproses pada dataset, kita perlu mengubah teks menjadi angka yang dapat dipahami oleh model. Seperti yang telah Anda lihat di <a href="/course/chapter2">bab sebelumnya</a>, hal ini dilakukan dengan menggunakan tokenizer. Kita dapat memberikan satu kalimat atau daftar kalimat ke tokenizer, sehingga kita bisa langsung melakukan tokenisasi pada semua kalimat pertama dan semua kalimat kedua dari setiap pasangan seperti ini:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoTokenizer

	checkpoint = <span class="hljs-string">"bert-base-uncased"</span>
	tokenizer = AutoTokenizer.from_pretrained(checkpoint)
	tokenized_sentences_1 = tokenizer(raw_datasets[<span class="hljs-string">"train"</span>][<span class="hljs-string">"sentence1"</span>])
	tokenized_sentences_2 = tokenizer(raw_datasets[<span class="hljs-string">"train"</span>][<span class="hljs-string">"sentence2"</span>])<!-- HTML_TAG_END --></pre></div> <div class="course-tip bg-gradient-to-br dark:bg-gradient-to-r before:border-green-500 dark:before:border-green-800 from-green-50 dark:from-gray-900 to-white dark:to-gray-950 border border-green-50 text-green-700 dark:text-gray-400"><p data-svelte-h="svelte-p3l6d6">💡 <strong>Penjelasan Lanjutan</strong>: Untuk teknik tokenisasi lanjutan, kunjungi <a href="https://huggingface.co/docs/transformers/main/en/tokenizer_summary" rel="nofollow">dokumentasi 🤗 Tokenizers</a> dan <a href="https://huggingface.co/learn/cookbook/en/advanced_rag#tokenization-strategies" rel="nofollow">panduan tokenisasi di cookbook</a>.</p></div> <p data-svelte-h="svelte-btequf">Namun, kita tidak bisa langsung memberikan dua kalimat ke model dan mendapatkan prediksi apakah kedua kalimat tersebut merupakan parafrase atau tidak. Kita perlu menangani kedua kalimat tersebut sebagai pasangan, dan menerapkan praproses yang sesuai. Untungnya, tokenizer juga dapat menerima pasangan kalimat dan mempersiapkannya sesuai dengan yang diharapkan oleh model BERT kita:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->inputs = tokenizer(<span class="hljs-string">"This is the first sentence."</span>, <span class="hljs-string">"This is the second one."</span>)
	inputs<!-- HTML_TAG_END --></pre></div> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->{
	<span class="hljs-string">'input_ids'</span>: [<span class="hljs-number">101</span>, <span class="hljs-number">2023</span>, <span class="hljs-number">2003</span>, <span class="hljs-number">1996</span>, <span class="hljs-number">2034</span>, <span class="hljs-number">6251</span>, <span class="hljs-number">1012</span>, <span class="hljs-number">102</span>, <span class="hljs-number">2023</span>, <span class="hljs-number">2003</span>, <span class="hljs-number">1996</span>, <span class="hljs-number">2117</span>, <span class="hljs-number">2028</span>, <span class="hljs-number">1012</span>, <span class="hljs-number">102</span>],
	<span class="hljs-string">'token_type_ids'</span>: [<span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>],
	<span class="hljs-string">'attention_mask'</span>: [<span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>]
	}<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1asb3ll">Kita telah membahas kunci <code>input_ids</code> dan <code>attention_mask</code> di <a href="/course/chapter2">Bab 2</a>, tetapi kita menunda pembahasan tentang <code>token_type_ids</code>. Dalam contoh ini, <code>token_type_ids</code> memberi tahu model bagian mana dari input yang merupakan kalimat pertama dan mana yang merupakan kalimat kedua.</p> <div class="course-tip bg-gradient-to-br dark:bg-gradient-to-r before:border-green-500 dark:before:border-green-800 from-green-50 dark:from-gray-900 to-white dark:to-gray-950 border border-green-50 text-green-700 dark:text-gray-400"><p data-svelte-h="svelte-3s2ko5">✏️ <strong>Coba Sendiri!</strong> Ambil elemen ke-15 dari data pelatihan dan tokenisasi kalimatnya secara terpisah serta sebagai pasangan. Apa perbedaannya?</p></div> <p data-svelte-h="svelte-9r00it">Jika kita mendekode ID di dalam <code>input_ids</code> kembali menjadi kata-kata:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->tokenizer.convert_ids_to_tokens(inputs[<span class="hljs-string">"input_ids"</span>])<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-x06neq">Kita akan mendapatkan:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->[<span class="hljs-string">'[CLS]'</span>, <span class="hljs-string">'this'</span>, <span class="hljs-string">'is'</span>, <span class="hljs-string">'the'</span>, <span class="hljs-string">'first'</span>, <span class="hljs-string">'sentence'</span>, <span class="hljs-string">'.'</span>, <span class="hljs-string">'[SEP]'</span>, <span class="hljs-string">'this'</span>, <span class="hljs-string">'is'</span>, <span class="hljs-string">'the'</span>, <span class="hljs-string">'second'</span>, <span class="hljs-string">'one'</span>, <span class="hljs-string">'.'</span>, <span class="hljs-string">'[SEP]'</span>]<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-135m9jy">Jadi kita melihat bahwa model mengharapkan input dalam format <code>[CLS] kalimat1 [SEP] kalimat2 [SEP]</code> ketika terdapat dua kalimat. Menyesuaikan ini dengan <code>token_type_ids</code> memberi kita:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->[<span class="hljs-string">'[CLS]'</span>, <span class="hljs-string">'this'</span>, <span class="hljs-string">'is'</span>, <span class="hljs-string">'the'</span>, <span class="hljs-string">'first'</span>, <span class="hljs-string">'sentence'</span>, <span class="hljs-string">'.'</span>, <span class="hljs-string">'[SEP]'</span>, <span class="hljs-string">'this'</span>, <span class="hljs-string">'is'</span>, <span class="hljs-string">'the'</span>, <span class="hljs-string">'second'</span>, <span class="hljs-string">'one'</span>, <span class="hljs-string">'.'</span>, <span class="hljs-string">'[SEP]'</span>]
	[ <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>]<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-12rftm1">Seperti yang Anda lihat, bagian input yang sesuai dengan <code>[CLS] kalimat1 [SEP]</code> semuanya memiliki token type ID <code>0</code>, sementara bagian lainnya, yang sesuai dengan <code>kalimat2 [SEP]</code>, semuanya memiliki token type ID <code>1</code>.</p> <p data-svelte-h="svelte-1059l9o">Perlu dicatat bahwa jika Anda memilih checkpoint yang berbeda, Anda mungkin tidak akan mendapatkan <code>token_type_ids</code> dalam hasil tokenisasi (misalnya, <code>token_type_ids</code> tidak disediakan jika Anda menggunakan model DistilBERT). <code>token_type_ids</code> hanya dikembalikan jika model mengetahui cara menggunakannya, yaitu jika model tersebut telah melihatnya selama pelatihan pralatih.</p> <p data-svelte-h="svelte-3dufb0">Dalam hal ini, BERT telah dilatih sebelumnya dengan <code>token_type_ids</code>, dan selain tujuan masked language modeling yang telah kita bahas di <a href="/course/chapter1">Bab 1</a>, model ini juga memiliki tujuan tambahan yang disebut <em>next sentence prediction</em>. Tujuan dari tugas ini adalah untuk memodelkan hubungan antara pasangan kalimat.</p> <p data-svelte-h="svelte-k2ft94">Dalam next sentence prediction, model diberikan pasangan kalimat (dengan token yang secara acak dimasker) dan diminta untuk memprediksi apakah kalimat kedua mengikuti kalimat pertama. Untuk membuat tugas ini tidak sepele, setengah dari waktu pasangan kalimat berasal dari dokumen yang sama secara berurutan, dan setengah lainnya berasal dari dokumen yang berbeda.</p> <p data-svelte-h="svelte-13k8n3q">Secara umum, Anda tidak perlu khawatir apakah ada <code>token_type_ids</code> dalam input yang telah ditokenisasi: selama Anda menggunakan checkpoint yang sama untuk tokenizer dan model, semuanya akan berjalan baik karena tokenizer tahu apa yang perlu diberikan ke modelnya.</p> <p data-svelte-h="svelte-g27nwd">Sekarang setelah kita melihat bagaimana tokenizer dapat menangani satu pasangan kalimat, kita bisa menggunakannya untuk melakukan tokenisasi pada seluruh dataset: seperti di <a href="/course/chapter2">bab sebelumnya</a>, kita dapat memberikan daftar pasangan kalimat ke tokenizer dengan memberikan daftar kalimat pertama dan kemudian daftar kalimat kedua. Ini juga kompatibel dengan opsi padding dan truncation yang telah kita bahas di <a href="/course/chapter2">Bab 2</a>. Jadi, salah satu cara untuk melakukan praproses pada dataset pelatihan adalah:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->tokenized_dataset = tokenizer(
	raw_datasets[<span class="hljs-string">"train"</span>][<span class="hljs-string">"sentence1"</span>],
	raw_datasets[<span class="hljs-string">"train"</span>][<span class="hljs-string">"sentence2"</span>],
	padding=<span class="hljs-literal">True</span>,
	truncation=<span class="hljs-literal">True</span>,
	)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1bvpc0c">Pendekatan ini bekerja dengan baik, tetapi memiliki kelemahan yaitu mengembalikan sebuah dictionary (dengan kunci seperti <code>input_ids</code>, <code>attention_mask</code>, dan <code>token_type_ids</code>, serta nilai berupa daftar dari daftar). Selain itu, cara ini hanya akan berhasil jika Anda memiliki cukup RAM untuk menyimpan seluruh dataset selama proses tokenisasi (sementara dataset dari pustaka 🤗 Datasets disimpan sebagai berkas <a href="https://arrow.apache.org/" rel="nofollow">Apache Arrow</a> di disk, sehingga hanya sampel yang Anda minta saja yang dimuat ke dalam memori).</p> <p data-svelte-h="svelte-1gnryc5">Untuk menjaga data tetap dalam bentuk dataset, kita akan menggunakan metode <a href="https://huggingface.co/docs/datasets/package_reference/main_classes#datasets.Dataset.map" rel="nofollow"><code>Dataset.map()</code></a>. Metode ini juga memberi kita fleksibilitas tambahan jika kita ingin melakukan praproses lebih dari sekadar tokenisasi. Metode <code>map()</code> bekerja dengan menerapkan sebuah fungsi pada setiap elemen dalam dataset, jadi mari kita definisikan sebuah fungsi untuk melakukan tokenisasi pada input kita:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-keyword">def</span> <span class="hljs-title function_">tokenize_function</span>(<span class="hljs-params">example</span>):
	<span class="hljs-keyword">return</span> tokenizer(example[<span class="hljs-string">"sentence1"</span>], example[<span class="hljs-string">"sentence2"</span>], truncation=<span class="hljs-literal">True</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-tgyuwg">Fungsi ini menerima sebuah dictionary (seperti item dalam dataset kita) dan mengembalikan dictionary baru dengan kunci <code>input_ids</code>, <code>attention_mask</code>, dan <code>token_type_ids</code>. Perlu dicatat bahwa fungsi ini juga akan berfungsi jika dictionary <code>example</code> berisi beberapa sampel sekaligus (setiap kunci berisi daftar kalimat), karena <code>tokenizer</code> dapat menangani daftar pasangan kalimat, seperti yang telah kita lihat sebelumnya. Ini memungkinkan kita menggunakan opsi <code>batched=True</code> saat memanggil <code>map()</code>, yang secara signifikan akan mempercepat proses tokenisasi. <code>tokenizer</code> ini didukung oleh tokenizer yang ditulis dalam bahasa Rust dari pustaka <a href="https://github.com/huggingface/tokenizers" rel="nofollow">🤗 Tokenizers</a>. Tokenizer ini bisa sangat cepat, tetapi hanya jika kita memberinya banyak input sekaligus.</p> <p data-svelte-h="svelte-z6prwu">Perhatikan bahwa kita belum menyertakan argumen <code>padding</code> dalam fungsi tokenisasi kita saat ini. Hal ini karena melakukan padding pada semua sampel hingga panjang maksimum tidaklah efisien: akan lebih baik jika kita melakukan padding saat membangun sebuah batch, karena kita hanya perlu menyesuaikan padding hingga panjang maksimum dalam batch tersebut, bukan panjang maksimum dalam seluruh dataset. Ini bisa menghemat banyak waktu dan daya komputasi terutama saat panjang input sangat bervariasi!</p> <div class="course-tip bg-gradient-to-br dark:bg-gradient-to-r before:border-green-500 dark:before:border-green-800 from-green-50 dark:from-gray-900 to-white dark:to-gray-950 border border-green-50 text-green-700 dark:text-gray-400"><p data-svelte-h="svelte-1w8q7d2">📚 <strong>Tips Performa</strong>: Pelajari lebih lanjut tentang teknik pemrosesan data yang efisien di <a href="https://huggingface.co/docs/datasets/about_arrow" rel="nofollow">panduan performa 🤗 Datasets</a>.</p></div> <p data-svelte-h="svelte-19posmw">Berikut adalah cara kita menerapkan fungsi tokenisasi pada seluruh dataset sekaligus. Kita menggunakan <code>batched=True</code> dalam pemanggilan <code>map</code>, agar fungsi diterapkan pada beberapa elemen dataset sekaligus, bukan satu per satu. Ini memungkinkan praproses berjalan lebih cepat.</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->tokenized_datasets = raw_datasets.<span class="hljs-built_in">map</span>(tokenize_function, batched=<span class="hljs-literal">True</span>)
	tokenized_datasets<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-122jvot">Cara pustaka 🤗 Datasets menerapkan proses ini adalah dengan menambahkan field baru ke dalam dataset, satu untuk setiap kunci dalam dictionary yang dikembalikan oleh fungsi praproses:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->DatasetDict({
	train: Dataset({
	features: [<span class="hljs-string">'attention_mask'</span>, <span class="hljs-string">'idx'</span>, <span class="hljs-string">'input_ids'</span>, <span class="hljs-string">'label'</span>, <span class="hljs-string">'sentence1'</span>, <span class="hljs-string">'sentence2'</span>, <span class="hljs-string">'token_type_ids'</span>],
	num_rows: <span class="hljs-number">3668</span>
	})
	validation: Dataset({
	features: [<span class="hljs-string">'attention_mask'</span>, <span class="hljs-string">'idx'</span>, <span class="hljs-string">'input_ids'</span>, <span class="hljs-string">'label'</span>, <span class="hljs-string">'sentence1'</span>, <span class="hljs-string">'sentence2'</span>, <span class="hljs-string">'token_type_ids'</span>],
	num_rows: <span class="hljs-number">408</span>
	})
	test: Dataset({
	features: [<span class="hljs-string">'attention_mask'</span>, <span class="hljs-string">'idx'</span>, <span class="hljs-string">'input_ids'</span>, <span class="hljs-string">'label'</span>, <span class="hljs-string">'sentence1'</span>, <span class="hljs-string">'sentence2'</span>, <span class="hljs-string">'token_type_ids'</span>],
	num_rows: <span class="hljs-number">1725</span>
	})
	})<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-larw98">Anda bahkan bisa menggunakan multiprocessing saat menerapkan fungsi praproses dengan <code>map()</code> dengan menambahkan argumen <code>num_proc</code>. Kami tidak melakukannya di sini karena pustaka 🤗 Tokenizers sudah menggunakan beberapa thread untuk melakukan tokenisasi sampel dengan lebih cepat. Namun, jika Anda tidak menggunakan tokenizer cepat yang didukung oleh pustaka tersebut, penggunaan multiprocessing ini bisa mempercepat proses praproses Anda.</p> <p data-svelte-h="svelte-10y4ss7">Fungsi <code>tokenize_function</code> kita mengembalikan sebuah dictionary dengan kunci <code>input_ids</code>, <code>attention_mask</code>, dan <code>token_type_ids</code>, sehingga ketiga field tersebut akan ditambahkan ke semua bagian (split) dari dataset kita. Perlu dicatat bahwa kita juga bisa mengganti field yang sudah ada jika fungsi praproses kita mengembalikan nilai baru untuk kunci yang sudah ada dalam dataset yang kita terapkan <code>map()</code> padanya.</p> <p data-svelte-h="svelte-1bcbfq">Hal terakhir yang perlu kita lakukan adalah melakukan padding pada semua contoh hingga panjang elemen terpanjang saat kita menggabungkan elemen-elemen ke dalam batch — teknik ini disebut <em>padding dinamis</em> (<em>dynamic padding</em>).</p> <h4 class="relative group"><a id="dynamic-padding" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#dynamic-padding"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Padding Dinamis</span></h4> <iframe class="w-full xl:w-4/6 h-80" src="https://www.youtube-nocookie.com/embed/7q5NyFT8REg" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe> <p data-svelte-h="svelte-1e2wd7p">Fungsi yang bertanggung jawab untuk menggabungkan sampel-sampel ke dalam sebuah batch disebut <em>collate function</em>. Ini adalah argumen yang bisa Anda berikan saat membangun sebuah <code>DataLoader</code>, dengan fungsi default yang hanya akan mengonversi sampel Anda menjadi tensor PyTorch dan menggabungkannya (secara rekursif jika elemen Anda berupa list, tuple, atau dictionary). Namun, ini tidak akan bisa dilakukan dalam kasus kita karena input yang kita miliki tidak semuanya memiliki ukuran yang sama. Kita sengaja menunda proses padding agar hanya dilakukan seperlunya pada setiap batch, guna menghindari input yang terlalu panjang dengan banyak padding. Ini akan mempercepat proses pelatihan secara signifikan, tetapi perlu dicatat bahwa jika Anda melakukan pelatihan di TPU, ini bisa menimbulkan masalah — TPU lebih menyukai bentuk input yang tetap (fixed shapes), meskipun itu berarti harus menambahkan padding ekstra.</p> <div class="course-tip bg-gradient-to-br dark:bg-gradient-to-r before:border-green-500 dark:before:border-green-800 from-green-50 dark:from-gray-900 to-white dark:to-gray-950 border border-green-50 text-green-700 dark:text-gray-400"><p data-svelte-h="svelte-a9i3lh">🚀 <strong>Panduan Optimasi</strong>: Untuk detail lebih lanjut tentang mengoptimalkan performa pelatihan, termasuk strategi padding dan pertimbangan saat menggunakan TPU, lihat <a href="https://huggingface.co/docs/transformers/main/en/performance" rel="nofollow">dokumentasi performa 🤗 Transformers</a>.</p></div> <p data-svelte-h="svelte-11l2j2n">Untuk menerapkannya dalam praktik, kita harus mendefinisikan sebuah <em>collate function</em> yang akan menerapkan jumlah padding yang tepat pada item-item dataset yang ingin kita batch-kan bersama. Untungnya, pustaka 🤗 Transformers menyediakan fungsi seperti itu melalui <code>DataCollatorWithPadding</code>. Fungsi ini menerima sebuah tokenizer saat diinisialisasi (untuk mengetahui token padding yang digunakan, dan apakah model mengharapkan padding di sebelah kiri atau kanan input), dan akan melakukan semua yang Anda butuhkan:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> DataCollatorWithPadding

	data_collator = DataCollatorWithPadding(tokenizer=tokenizer)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1bbogdw">Contoh padding:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->samples = tokenized_datasets[<span class="hljs-string">"train"</span>][:<span class="hljs-number">8</span>]
	samples = {k: v <span class="hljs-keyword">for</span> k, v <span class="hljs-keyword">in</span> samples.items() <span class="hljs-keyword">if</span> k <span class="hljs-keyword">not</span> <span class="hljs-keyword">in</span> [<span class="hljs-string">"idx"</span>, <span class="hljs-string">"sentence1"</span>, <span class="hljs-string">"sentence2"</span>]}
	[<span class="hljs-built_in">len</span>(x) <span class="hljs-keyword">for</span> x <span class="hljs-keyword">in</span> samples[<span class="hljs-string">"input_ids"</span>]]<!-- HTML_TAG_END --></pre></div> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->[<span class="hljs-number">50</span>, <span class="hljs-number">59</span>, <span class="hljs-number">47</span>, <span class="hljs-number">67</span>, <span class="hljs-number">59</span>, <span class="hljs-number">50</span>, <span class="hljs-number">62</span>, <span class="hljs-number">32</span>]<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-s0lj6x">Tidak mengejutkan, kita mendapatkan sampel dengan panjang yang bervariasi, dari 32 hingga 67. Padding dinamis berarti semua sampel dalam batch ini akan dipadkan hingga panjang 67, yaitu panjang maksimum di dalam batch tersebut. Tanpa padding dinamis, semua sampel harus dipadkan hingga panjang maksimum dari seluruh dataset, atau hingga panjang maksimum yang dapat diterima oleh model. Mari kita periksa kembali apakah <code>data_collator</code> kita benar-benar melakukan padding dinamis dengan benar:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->batch = data_collator(samples)
	{k: v.shape <span class="hljs-keyword">for</span> k, v <span class="hljs-keyword">in</span> batch.items()}<!-- HTML_TAG_END --></pre></div> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->{<span class="hljs-string">'attention_mask'</span>: torch.Size([<span class="hljs-number">8</span>, <span class="hljs-number">67</span>]),
	<span class="hljs-string">'input_ids'</span>: torch.Size([<span class="hljs-number">8</span>, <span class="hljs-number">67</span>]),
	<span class="hljs-string">'token_type_ids'</span>: torch.Size([<span class="hljs-number">8</span>, <span class="hljs-number">67</span>]),
	<span class="hljs-string">'labels'</span>: torch.Size([<span class="hljs-number">8</span>])}<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-oma23l">Terlihat bagus! Sekarang, setelah kita berhasil mengubah teks mentah menjadi batch yang bisa diproses oleh model, kita siap untuk melakukan fine-tuning!</p> <div class="course-tip bg-gradient-to-br dark:bg-gradient-to-r before:border-green-500 dark:before:border-green-800 from-green-50 dark:from-gray-900 to-white dark:to-gray-950 border border-green-50 text-green-700 dark:text-gray-400"><p data-svelte-h="svelte-ws9bd6">✏️ <strong>Coba Sendiri!</strong> Ulangi proses pra-pemrosesan pada dataset GLUE SST-2. Dataset ini sedikit berbeda karena terdiri dari kalimat tunggal, bukan pasangan kalimat, tetapi sisanya akan terlihat serupa. Untuk tantangan yang lebih sulit, coba tulis fungsi pra-pemrosesan yang bisa digunakan untuk semua tugas dalam GLUE.</p> <p data-svelte-h="svelte-1w61t74">📖 <strong>Latihan Tambahan</strong>: Lihat contoh langsung di <a href="https://huggingface.co/docs/transformers/main/en/notebooks" rel="nofollow">🤗 Transformers examples</a>.</p></div> <p data-svelte-h="svelte-109gpje">Sempurna! Sekarang setelah kita memproses data kita menggunakan praktik terbaik terbaru dari pustaka 🤗 Datasets, kita siap untuk melatih model menggunakan API Trainer modern. Bagian selanjutnya akan menunjukkan cara melakukan fine-tuning model secara efektif menggunakan fitur dan optimasi terbaru dalam ekosistem Hugging Face.</p> <h2 class="relative group"><a id="section-quiz" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#section-quiz"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Kuis Bagian</span></h2> <p data-svelte-h="svelte-1ie60p2">Uji pemahamanmu tentang konsep pemrosesan data:</p> <h3 class="relative group"><a id="1-apa-keuntungan-utama-menggunakan-datasetmap-dengan-batchedtrue-" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#1-apa-keuntungan-utama-menggunakan-datasetmap-dengan-batchedtrue-"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>1. Apa keuntungan utama menggunakan Dataset.map() dengan batched=True ?</span></h3> <div><form><label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="0"> <!-- HTML_TAG_START -->Itu menggunakan lebih sedikit memori.<!-- HTML_TAG_END --></label> <label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="1"> <!-- HTML_TAG_START -->Itu memproses beberapa contoh sekaligus, membuat tokenisasi jauh lebih cepat.<!-- HTML_TAG_END --></label> <label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="2"> <!-- HTML_TAG_START -->Itu secara otomatis menangani padding untuk Anda.<!-- HTML_TAG_END --></label> <label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="3"> <!-- HTML_TAG_START -->Itu mengonversi data menjadi tensor PyTorch.<!-- HTML_TAG_END --></label> <div class="flex flex-row items-center mt-3"><button class="btn px-4 mr-4" type="submit" disabled>Submit</button> </div></form></div> <h3 class="relative group"><a id="2-mengapa-kita-menggunakan-padding-dinamis-daripada-mem-pad-semua-urutan-ke-panjang-maksimum-dalam-dataset" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#2-mengapa-kita-menggunakan-padding-dinamis-daripada-mem-pad-semua-urutan-ke-panjang-maksimum-dalam-dataset"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>2. Mengapa kita menggunakan padding dinamis daripada mem-pad semua urutan ke panjang maksimum dalam dataset?</span></h3> <div><form><label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="0"> <!-- HTML_TAG_START -->Padding dinamis diperlukan oleh arsitektur model.<!-- HTML_TAG_END --></label> <label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="1"> <!-- HTML_TAG_START -->Itu mengurangi beban komputasi dengan hanya padding ke panjang maksimum dalam setiap batch.<!-- HTML_TAG_END --></label> <label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="2"> <!-- HTML_TAG_START -->Itu meningkatkan akurasi model.<!-- HTML_TAG_END --></label> <label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="3"> <!-- HTML_TAG_START -->Itu wajib digunakan saat memakai DataCollatorWithPadding.<!-- HTML_TAG_END --></label> <div class="flex flex-row items-center mt-3"><button class="btn px-4 mr-4" type="submit" disabled>Submit</button> </div></form></div> <h3 class="relative group"><a id="3-apa-yang-direpresentasikan-oleh-field-tokentypeids-dalam-tokenisasi-bert" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#3-apa-yang-direpresentasikan-oleh-field-tokentypeids-dalam-tokenisasi-bert"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>3. Apa yang direpresentasikan oleh field token_type_ids dalam tokenisasi BERT?</span></h3> <div><form><label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="0"> <!-- HTML_TAG_START -->Posisi setiap token dalam urutan.<!-- HTML_TAG_END --></label> <label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="1"> <!-- HTML_TAG_START -->Kalimat mana yang dimiliki setiap token saat memproses pasangan kalimat.<!-- HTML_TAG_END --></label> <label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="2"> <!-- HTML_TAG_START -->Attention mask untuk setiap token.<!-- HTML_TAG_END --></label> <label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="3"> <!-- HTML_TAG_START -->ID kosakata dari setiap token.<!-- HTML_TAG_END --></label> <div class="flex flex-row items-center mt-3"><button class="btn px-4 mr-4" type="submit" disabled>Submit</button> </div></form></div> <h3 class="relative group"><a id="4-saat-memuat-dataset-dengan-loaddatasetglue-mrpc--apa-arti-dari-argumen-kedua" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#4-saat-memuat-dataset-dengan-loaddatasetglue-mrpc--apa-arti-dari-argumen-kedua"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>4. Saat memuat dataset dengan load_dataset('glue', 'mrpc') , apa arti dari argumen kedua?</span></h3> <div><form><label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="0"> <!-- HTML_TAG_START -->Versi dataset yang akan dimuat.<!-- HTML_TAG_END --></label> <label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="1"> <!-- HTML_TAG_START -->Tugas atau subset spesifik dalam benchmark GLUE.<!-- HTML_TAG_END --></label> <label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="2"> <!-- HTML_TAG_START -->Split dataset (train/validation/test).<!-- HTML_TAG_END --></label> <label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="3"> <!-- HTML_TAG_START -->Format data yang akan dikembalikan.<!-- HTML_TAG_END --></label> <div class="flex flex-row items-center mt-3"><button class="btn px-4 mr-4" type="submit" disabled>Submit</button> </div></form></div> <h3 class="relative group"><a id="5-apa-tujuan-dari-menghapus-kolom-seperti-sentence1-dan-sentence2-sebelum-pelatihan" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#5-apa-tujuan-dari-menghapus-kolom-seperti-sentence1-dan-sentence2-sebelum-pelatihan"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>5. Apa tujuan dari menghapus kolom seperti ‘sentence1’ dan ‘sentence2’ sebelum pelatihan?</span></h3> <div><form><label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="0"> <!-- HTML_TAG_START -->Untuk menghemat memori selama pelatihan.<!-- HTML_TAG_END --></label> <label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="1"> <!-- HTML_TAG_START -->Model tidak mengharapkan kolom teks mentah ini dan akan menghasilkan error.<!-- HTML_TAG_END --></label> <label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="2"> <!-- HTML_TAG_START -->Kolom-kolom ini tidak diperlukan untuk evaluasi.<!-- HTML_TAG_END --></label> <label class="block"><input autocomplete="off" class="form-input -mt-1.5 mr-2" name="choice" type="checkbox" value="3"> <!-- HTML_TAG_START -->Ini secara signifikan meningkatkan kecepatan pelatihan.<!-- HTML_TAG_END --></label> <div class="flex flex-row items-center mt-3"><button class="btn px-4 mr-4" type="submit" disabled>Submit</button> </div></form></div> <div class="course-tip bg-gradient-to-br dark:bg-gradient-to-r before:border-green-500 dark:before:border-green-800 from-green-50 dark:from-gray-900 to-white dark:to-gray-950 border border-green-50 text-green-700 dark:text-gray-400"><p data-svelte-h="svelte-1f1eedo">💡 <strong>Inti Penting</strong>:</p> <ul data-svelte-h="svelte-1yfamec"><li>Gunakan <code>batched=True</code> dengan <code>Dataset.map()</code> untuk mempercepat pra-pemrosesan secara signifikan</li> <li>Padding dinamis dengan <code>DataCollatorWithPadding</code> lebih efisien dibandingkan padding dengan panjang tetap</li> <li>Selalu pra-proses data Anda agar sesuai dengan ekspektasi model (tensor numerik, nama kolom yang benar)</li> <li>Pustaka 🤗 Datasets menyediakan alat-alat canggih untuk memproses data secara efisien dalam skala besar</li></ul></div> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/course/blob/main/chapters/id/chapter3/2.mdx" target="_blank"><span data-svelte-h="svelte-1kd6by1"><</span> <span data-svelte-h="svelte-x0xyl0">></span> <span data-svelte-h="svelte-1dajgef"><span class="underline ml-1.5">Update</span> on GitHub</span></a> <p></p>

	<script>
	{
	__sveltekit_ojy514 = {
	assets: "/docs/course/pr_1054/id",
	base: "/docs/course/pr_1054/id",
	env: {}
	};

	const element = document.currentScript.parentElement;

	const data = [null,null];

	Promise.all([
	import("/docs/course/pr_1054/id/_app/immutable/entry/start.4f92af03.js"),
	import("/docs/course/pr_1054/id/_app/immutable/entry/app.19cef1b6.js")
	]).then(([kit, app]) => {
	kit.start(app, element, {
	node_ids: [0, 24],
	data,
	form: null,
	error: null
	});
	});
	}
	</script>

Xet Storage Details

Size:: 87.2 kB
Xet hash:: 55ed8a51fdbac8bd766e5399d2ac0077a4bf77ed306e57cc9aea13e35de88f73

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.