Buckets:
| <meta charset="utf-8" /><meta name="hf:doc:metadata" content="{"title":"Search index","local":"search-index","sections":[{"title":"FAISS","local":"faiss","sections":[],"depth":2},{"title":"Elasticsearch","local":"elasticsearch","sections":[],"depth":2}],"depth":1}"> | |
| <link href="/docs/datasets/pr_7039/en/_app/immutable/assets/0.e3b0c442.css" rel="modulepreload"> | |
| <link rel="modulepreload" href="/docs/datasets/pr_7039/en/_app/immutable/entry/start.4ee36d94.js"> | |
| <link rel="modulepreload" href="/docs/datasets/pr_7039/en/_app/immutable/chunks/scheduler.bdbef820.js"> | |
| <link rel="modulepreload" href="/docs/datasets/pr_7039/en/_app/immutable/chunks/singletons.be7ae209.js"> | |
| <link rel="modulepreload" href="/docs/datasets/pr_7039/en/_app/immutable/chunks/index.8a885b74.js"> | |
| <link rel="modulepreload" href="/docs/datasets/pr_7039/en/_app/immutable/chunks/paths.047d52be.js"> | |
| <link rel="modulepreload" href="/docs/datasets/pr_7039/en/_app/immutable/entry/app.e1b00805.js"> | |
| <link rel="modulepreload" href="/docs/datasets/pr_7039/en/_app/immutable/chunks/index.c0aea24a.js"> | |
| <link rel="modulepreload" href="/docs/datasets/pr_7039/en/_app/immutable/nodes/0.3bc7195c.js"> | |
| <link rel="modulepreload" href="/docs/datasets/pr_7039/en/_app/immutable/chunks/each.e59479a4.js"> | |
| <link rel="modulepreload" href="/docs/datasets/pr_7039/en/_app/immutable/nodes/18.c3e499f7.js"> | |
| <link rel="modulepreload" href="/docs/datasets/pr_7039/en/_app/immutable/chunks/CodeBlock.6ccca92e.js"> | |
| <link rel="modulepreload" href="/docs/datasets/pr_7039/en/_app/immutable/chunks/EditOnGithub.725ee0c1.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{"title":"Search index","local":"search-index","sections":[{"title":"FAISS","local":"faiss","sections":[],"depth":2},{"title":"Elasticsearch","local":"elasticsearch","sections":[],"depth":2}],"depth":1}"><!-- HEAD_svelte-u9bgzb_END --> <p></p> <h1 class="relative group"><a id="search-index" 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="#search-index"><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>Search index</span></h1> <p data-svelte-h="svelte-5c7wnl"><a href="https://github.com/facebookresearch/faiss" rel="nofollow">FAISS</a> and <a href="https://www.elastic.co/elasticsearch/" rel="nofollow">Elasticsearch</a> enables searching for examples in a dataset. This can be useful when you want to retrieve specific examples from a dataset that are relevant to your NLP task. For example, if you are working on a Open Domain Question Answering task, you may want to only return examples that are relevant to answering your question.</p> <p data-svelte-h="svelte-hdwpul">This guide will show you how to build an index for your dataset that will allow you to search it.</p> <h2 class="relative group"><a id="faiss" 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="#faiss"><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>FAISS</span></h2> <p data-svelte-h="svelte-se9cy2">FAISS retrieves documents based on the similarity of their vector representations. In this example, you will generate the vector representations with the <a href="https://huggingface.co/transformers/model_doc/dpr.html" rel="nofollow">DPR</a> model.</p> <ol data-svelte-h="svelte-x1ghuo"><li>Download the DPR model from 🤗 Transformers:</li></ol> <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-meta">>>> </span><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> DPRContextEncoder, DPRContextEncoderTokenizer | |
| <span class="hljs-meta">>>> </span><span class="hljs-keyword">import</span> torch | |
| <span class="hljs-meta">>>> </span>torch.set_grad_enabled(<span class="hljs-literal">False</span>) | |
| <span class="hljs-meta">>>> </span>ctx_encoder = DPRContextEncoder.from_pretrained(<span class="hljs-string">"facebook/dpr-ctx_encoder-single-nq-base"</span>) | |
| <span class="hljs-meta">>>> </span>ctx_tokenizer = DPRContextEncoderTokenizer.from_pretrained(<span class="hljs-string">"facebook/dpr-ctx_encoder-single-nq-base"</span>)<!-- HTML_TAG_END --></pre></div> <ol start="2" data-svelte-h="svelte-1teerx9"><li>Load your dataset and compute the vector representations:</li></ol> <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-meta">>>> </span><span class="hljs-keyword">from</span> datasets <span class="hljs-keyword">import</span> load_dataset | |
| <span class="hljs-meta">>>> </span>ds = load_dataset(<span class="hljs-string">'crime_and_punish'</span>, split=<span class="hljs-string">'train[:100]'</span>) | |
| <span class="hljs-meta">>>> </span>ds_with_embeddings = ds.<span class="hljs-built_in">map</span>(<span class="hljs-keyword">lambda</span> example: {<span class="hljs-string">'embeddings'</span>: ctx_encoder(**ctx_tokenizer(example[<span class="hljs-string">"line"</span>], return_tensors=<span class="hljs-string">"pt"</span>))[<span class="hljs-number">0</span>][<span class="hljs-number">0</span>].numpy()})<!-- HTML_TAG_END --></pre></div> <ol start="3" data-svelte-h="svelte-poj19t"><li>Create the index with <a href="/docs/datasets/pr_7039/en/package_reference/main_classes#datasets.Dataset.add_faiss_index">Dataset.add_faiss_index()</a>:</li></ol> <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-meta">>>> </span>ds_with_embeddings.add_faiss_index(column=<span class="hljs-string">'embeddings'</span>)<!-- HTML_TAG_END --></pre></div> <ol start="4" data-svelte-h="svelte-wrk875"><li>Now you can query your dataset with the <code>embeddings</code> index. Load the DPR Question Encoder, and search for a question with <a href="/docs/datasets/pr_7039/en/package_reference/main_classes#datasets.Dataset.get_nearest_examples">Dataset.get_nearest_examples()</a>:</li></ol> <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-meta">>>> </span><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> DPRQuestionEncoder, DPRQuestionEncoderTokenizer | |
| <span class="hljs-meta">>>> </span>q_encoder = DPRQuestionEncoder.from_pretrained(<span class="hljs-string">"facebook/dpr-question_encoder-single-nq-base"</span>) | |
| <span class="hljs-meta">>>> </span>q_tokenizer = DPRQuestionEncoderTokenizer.from_pretrained(<span class="hljs-string">"facebook/dpr-question_encoder-single-nq-base"</span>) | |
| <span class="hljs-meta">>>> </span>question = <span class="hljs-string">"Is it serious ?"</span> | |
| <span class="hljs-meta">>>> </span>question_embedding = q_encoder(**q_tokenizer(question, return_tensors=<span class="hljs-string">"pt"</span>))[<span class="hljs-number">0</span>][<span class="hljs-number">0</span>].numpy() | |
| <span class="hljs-meta">>>> </span>scores, retrieved_examples = ds_with_embeddings.get_nearest_examples(<span class="hljs-string">'embeddings'</span>, question_embedding, k=<span class="hljs-number">10</span>) | |
| <span class="hljs-meta">>>> </span>retrieved_examples[<span class="hljs-string">"line"</span>][<span class="hljs-number">0</span>] | |
| <span class="hljs-string">'_that_ serious? It is not serious at all. It’s simply a fantasy to amuse\r\n'</span><!-- HTML_TAG_END --></pre></div> <ol start="5" data-svelte-h="svelte-1bcwjnj"><li>You can access the index with <a href="/docs/datasets/pr_7039/en/package_reference/main_classes#datasets.Dataset.get_index">Dataset.get_index()</a> and use it for special operations, e.g. query it using <code>range_search</code>:</li></ol> <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-meta">>>> </span>faiss_index = ds_with_embeddings.get_index(<span class="hljs-string">'embeddings'</span>).faiss_index | |
| <span class="hljs-meta">>>> </span>limits, distances, indices = faiss_index.range_search(x=question_embedding.reshape(<span class="hljs-number">1</span>, -<span class="hljs-number">1</span>), thresh=<span class="hljs-number">0.95</span>)<!-- HTML_TAG_END --></pre></div> <ol start="6" data-svelte-h="svelte-5z60e2"><li>When you are done querying, save the index on disk with <a href="/docs/datasets/pr_7039/en/package_reference/main_classes#datasets.Dataset.save_faiss_index">Dataset.save_faiss_index()</a>:</li></ol> <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-meta">>>> </span>ds_with_embeddings.save_faiss_index(<span class="hljs-string">'embeddings'</span>, <span class="hljs-string">'my_index.faiss'</span>)<!-- HTML_TAG_END --></pre></div> <ol start="7" data-svelte-h="svelte-1q3tlzf"><li>Reload it at a later time with <a href="/docs/datasets/pr_7039/en/package_reference/main_classes#datasets.Dataset.load_faiss_index">Dataset.load_faiss_index()</a>:</li></ol> <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-meta">>>> </span>ds = load_dataset(<span class="hljs-string">'crime_and_punish'</span>, split=<span class="hljs-string">'train[:100]'</span>) | |
| <span class="hljs-meta">>>> </span>ds.load_faiss_index(<span class="hljs-string">'embeddings'</span>, <span class="hljs-string">'my_index.faiss'</span>)<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="elasticsearch" 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="#elasticsearch"><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>Elasticsearch</span></h2> <p data-svelte-h="svelte-8ul8ma">Unlike FAISS, Elasticsearch retrieves documents based on exact matches.</p> <p data-svelte-h="svelte-195ztu1">Start Elasticsearch on your machine, or see the <a href="https://www.elastic.co/guide/en/elasticsearch/reference/current/setup.html" rel="nofollow">Elasticsearch installation guide</a> if you don’t already have it installed.</p> <ol data-svelte-h="svelte-3fd7ny"><li>Load the dataset you want to index:</li></ol> <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-meta">>>> </span><span class="hljs-keyword">from</span> datasets <span class="hljs-keyword">import</span> load_dataset | |
| <span class="hljs-meta">>>> </span>squad = load_dataset(<span class="hljs-string">'squad'</span>, split=<span class="hljs-string">'validation'</span>)<!-- HTML_TAG_END --></pre></div> <ol start="2" data-svelte-h="svelte-jpqhf4"><li>Build the index with <a href="/docs/datasets/pr_7039/en/package_reference/main_classes#datasets.Dataset.add_elasticsearch_index">Dataset.add_elasticsearch_index()</a>:</li></ol> <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-meta">>>> </span>squad.add_elasticsearch_index(<span class="hljs-string">"context"</span>, host=<span class="hljs-string">"localhost"</span>, port=<span class="hljs-string">"9200"</span>)<!-- HTML_TAG_END --></pre></div> <ol start="3" data-svelte-h="svelte-14evusb"><li>Then you can query the <code>context</code> index with <a href="/docs/datasets/pr_7039/en/package_reference/main_classes#datasets.Dataset.get_nearest_examples">Dataset.get_nearest_examples()</a>:</li></ol> <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-meta">>>> </span>query = <span class="hljs-string">"machine"</span> | |
| <span class="hljs-meta">>>> </span>scores, retrieved_examples = squad.get_nearest_examples(<span class="hljs-string">"context"</span>, query, k=<span class="hljs-number">10</span>) | |
| <span class="hljs-meta">>>> </span>retrieved_examples[<span class="hljs-string">"title"</span>][<span class="hljs-number">0</span>] | |
| <span class="hljs-string">'Computational_complexity_theory'</span><!-- HTML_TAG_END --></pre></div> <ol start="4" data-svelte-h="svelte-1tchn11"><li>If you want to reuse the index, define the <code>es_index_name</code> parameter when you build the index:</li></ol> <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-meta">>>> </span><span class="hljs-keyword">from</span> datasets <span class="hljs-keyword">import</span> load_dataset | |
| <span class="hljs-meta">>>> </span>squad = load_dataset(<span class="hljs-string">'squad'</span>, split=<span class="hljs-string">'validation'</span>) | |
| <span class="hljs-meta">>>> </span>squad.add_elasticsearch_index(<span class="hljs-string">"context"</span>, host=<span class="hljs-string">"localhost"</span>, port=<span class="hljs-string">"9200"</span>, es_index_name=<span class="hljs-string">"hf_squad_val_context"</span>) | |
| <span class="hljs-meta">>>> </span>squad.get_index(<span class="hljs-string">"context"</span>).es_index_name | |
| hf_squad_val_context<!-- HTML_TAG_END --></pre></div> <ol start="5" data-svelte-h="svelte-wgq9fb"><li>Reload it later with the index name when you call <a href="/docs/datasets/pr_7039/en/package_reference/main_classes#datasets.Dataset.load_elasticsearch_index">Dataset.load_elasticsearch_index()</a>:</li></ol> <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-meta">>>> </span><span class="hljs-keyword">from</span> datasets <span class="hljs-keyword">import</span> load_dataset | |
| <span class="hljs-meta">>>> </span>squad = load_dataset(<span class="hljs-string">'squad'</span>, split=<span class="hljs-string">'validation'</span>) | |
| <span class="hljs-meta">>>> </span>squad.load_elasticsearch_index(<span class="hljs-string">"context"</span>, host=<span class="hljs-string">"localhost"</span>, port=<span class="hljs-string">"9200"</span>, es_index_name=<span class="hljs-string">"hf_squad_val_context"</span>) | |
| <span class="hljs-meta">>>> </span>query = <span class="hljs-string">"machine"</span> | |
| <span class="hljs-meta">>>> </span>scores, retrieved_examples = squad.get_nearest_examples(<span class="hljs-string">"context"</span>, query, k=<span class="hljs-number">10</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1ajeg13">For more advanced Elasticsearch usage, you can specify your own configuration with custom settings:</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-meta">>>> </span><span class="hljs-keyword">import</span> elasticsearch <span class="hljs-keyword">as</span> es | |
| <span class="hljs-meta">>>> </span><span class="hljs-keyword">import</span> elasticsearch.helpers | |
| <span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> elasticsearch <span class="hljs-keyword">import</span> Elasticsearch | |
| <span class="hljs-meta">>>> </span>es_client = Elasticsearch([{<span class="hljs-string">"host"</span>: <span class="hljs-string">"localhost"</span>, <span class="hljs-string">"port"</span>: <span class="hljs-string">"9200"</span>}]) <span class="hljs-comment"># default client</span> | |
| <span class="hljs-meta">>>> </span>es_config = { | |
| <span class="hljs-meta">... </span> <span class="hljs-string">"settings"</span>: { | |
| <span class="hljs-meta">... </span> <span class="hljs-string">"number_of_shards"</span>: <span class="hljs-number">1</span>, | |
| <span class="hljs-meta">... </span> <span class="hljs-string">"analysis"</span>: {<span class="hljs-string">"analyzer"</span>: {<span class="hljs-string">"stop_standard"</span>: {<span class="hljs-string">"type"</span>: <span class="hljs-string">"standard"</span>, <span class="hljs-string">" stopwords"</span>: <span class="hljs-string">"_english_"</span>}}}, | |
| <span class="hljs-meta">... </span> }, | |
| <span class="hljs-meta">... </span> <span class="hljs-string">"mappings"</span>: {<span class="hljs-string">"properties"</span>: {<span class="hljs-string">"text"</span>: {<span class="hljs-string">"type"</span>: <span class="hljs-string">"text"</span>, <span class="hljs-string">"analyzer"</span>: <span class="hljs-string">"standard"</span>, <span class="hljs-string">"similarity"</span>: <span class="hljs-string">"BM25"</span>}}}, | |
| <span class="hljs-meta">... </span>} <span class="hljs-comment"># default config</span> | |
| <span class="hljs-meta">>>> </span>es_index_name = <span class="hljs-string">"hf_squad_context"</span> <span class="hljs-comment"># name of the index in Elasticsearch</span> | |
| <span class="hljs-meta">>>> </span>squad.add_elasticsearch_index(<span class="hljs-string">"context"</span>, es_client=es_client, es_config=es_config, es_index_name=es_index_name)<!-- HTML_TAG_END --></pre></div> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/datasets/blob/main/docs/source/faiss_es.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_1u8f530 = { | |
| assets: "/docs/datasets/pr_7039/en", | |
| base: "/docs/datasets/pr_7039/en", | |
| env: {} | |
| }; | |
| const element = document.currentScript.parentElement; | |
| const data = [null,null]; | |
| Promise.all([ | |
| import("/docs/datasets/pr_7039/en/_app/immutable/entry/start.4ee36d94.js"), | |
| import("/docs/datasets/pr_7039/en/_app/immutable/entry/app.e1b00805.js") | |
| ]).then(([kit, app]) => { | |
| kit.start(app, element, { | |
| node_ids: [0, 18], | |
| data, | |
| form: null, | |
| error: null | |
| }); | |
| }); | |
| } | |
| </script> | |
Xet Storage Details
- Size:
- 35.9 kB
- Xet hash:
- 180bc19b1902c74bacd6c05e69248b1121e376b112f01d0fc4fe2f421a12877a
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.