Buckets:
| <meta charset="utf-8" /><meta name="hf:doc:metadata" content="{"title":"Tokenizere","local":"tokenizere","sections":[{"title":"Încarcă un tokenizer","local":"încarcă-un-tokenizer","sections":[],"depth":2},{"title":"Encodare și decodare","local":"encodare-și-decodare","sections":[],"depth":2},{"title":"Token-uri speciale","local":"token-uri-speciale","sections":[],"depth":2},{"title":"Procesare în batch","local":"procesare-în-batch","sections":[{"title":"Padding","local":"padding","sections":[],"depth":3},{"title":"Trunchiere","local":"trunchiere","sections":[],"depth":3}],"depth":2},{"title":"Backend-uri","local":"backend-uri","sections":[],"depth":2},{"title":"Inspectează arhitectura tokenizer-ului","local":"inspectează-arhitectura-tokenizer-ului","sections":[],"depth":2},{"title":"Resurse","local":"resurse","sections":[],"depth":2}],"depth":1}"/> | |
| <link href="/docs/transformers/main/ro/_app/immutable/entry/start.DwnOcLC_.js" rel="modulepreload"> | |
| <link href="/docs/transformers/main/ro/_app/immutable/chunks/CL87Xr5_.js" rel="modulepreload"> | |
| <link href="/docs/transformers/main/ro/_app/immutable/chunks/Cl6kX2N0.js" rel="modulepreload"> | |
| <link href="/docs/transformers/main/ro/_app/immutable/entry/app.Dw7RcaNl.js" rel="modulepreload"> | |
| <link href="/docs/transformers/main/ro/_app/immutable/chunks/BzzeShVL.js" rel="modulepreload"> | |
| <link href="/docs/transformers/main/ro/_app/immutable/chunks/DWwSmaYb.js" rel="modulepreload"> | |
| <link href="/docs/transformers/main/ro/_app/immutable/chunks/DM0RASQS.js" rel="modulepreload"> | |
| <link href="/docs/transformers/main/ro/_app/immutable/chunks/DsnmJJEf.js" rel="modulepreload"> | |
| <link href="/docs/transformers/main/ro/_app/immutable/nodes/0.DMmsXSdS.js" rel="modulepreload"> | |
| <link href="/docs/transformers/main/ro/_app/immutable/chunks/C8aKV6he.js" rel="modulepreload"> | |
| <link href="/docs/transformers/main/ro/_app/immutable/nodes/11.CjpB3bO2.js" rel="modulepreload"> | |
| <link href="/docs/transformers/main/ro/_app/immutable/chunks/CjTxHwdC.js" rel="modulepreload"> | |
| <link href="/docs/transformers/main/ro/_app/immutable/chunks/hgCuwoId.js" rel="modulepreload"> | |
| <!--1ny7yeu--><meta name="hf:doc:metadata" content="{"title":"Tokenizere","local":"tokenizere","sections":[{"title":"Încarcă un tokenizer","local":"încarcă-un-tokenizer","sections":[],"depth":2},{"title":"Encodare și decodare","local":"encodare-și-decodare","sections":[],"depth":2},{"title":"Token-uri speciale","local":"token-uri-speciale","sections":[],"depth":2},{"title":"Procesare în batch","local":"procesare-în-batch","sections":[{"title":"Padding","local":"padding","sections":[],"depth":3},{"title":"Trunchiere","local":"trunchiere","sections":[],"depth":3}],"depth":2},{"title":"Backend-uri","local":"backend-uri","sections":[],"depth":2},{"title":"Inspectează arhitectura tokenizer-ului","local":"inspectează-arhitectura-tokenizer-ului","sections":[],"depth":2},{"title":"Resurse","local":"resurse","sections":[],"depth":2}],"depth":1}"/><!----> | |
| <link href="/docs/transformers/main/ro/_app/immutable/assets/0.tn0RQdqM.css" rel="modulepreload"> <!--[--><!--[0--><!--[--><!--[0--><!--[--><p></p> <div class="items-center shrink-0 min-w-[100px] max-sm:min-w-[50px] justify-end ml-auto flex" style="float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;"><div class="inline-flex rounded-md max-sm:rounded-sm"><button class="inline-flex items-center gap-1 h-7 max-sm:h-7 px-2 max-sm:px-1.5 text-sm font-medium text-gray-800 border border-r-0 rounded-l-md max-sm:rounded-l-sm border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-live="polite"><span class="inline-flex items-center justify-center rounded-md p-0.5 max-sm:p-0 hover:text-gray-800 dark:hover:text-gray-200"><svg class="sm:size-3.5 size-3" 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><!----></span> <span>Copy page</span></button> <button class="inline-flex items-center justify-center w-6 max-sm:w-5 h-7 max-sm:h-7 disabled:pointer-events-none text-sm text-gray-500 hover:text-gray-700 dark:hover:text-white rounded-r-md max-sm:rounded-r-sm border border-l transition border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-haspopup="menu" aria-expanded="false" aria-label="Open copy menu"><svg class="transition-transform text-gray-400 overflow-visible sm:size-3.5 size-3 rotate-0" width="1em" height="1em" viewBox="0 0 12 7" fill="none" xmlns="http://www.w3.org/2000/svg"><path d="M1 1L6 6L11 1" stroke="currentColor"></path></svg><!----></button></div> <!--[-1--><!--]--></div><!----> <!--[0--><h1 class="relative group"><a id="tokenizere" 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="#tokenizere"><span><svg 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>Tokenizere</span></h1><!--]--><!----> <p>Un tokenizer convertește text în tensori, care sunt input-urile unui model. Acesta normalizează și împarte textul, aplică algoritmul de tokenization, adaugă token-uri speciale și decodifică id-urile de output înapoi în text.</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 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="language-py "><!----><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoTokenizer | |
| tokenizer = AutoTokenizer.from_pretrained(<span class="hljs-string">"google/gemma-2-2b"</span>) | |
| tokenizer(<span class="hljs-string">"Sphinx of black quartz, judge my vow."</span>, return_tensors=<span class="hljs-string">"pt"</span>) | |
| { | |
| <span class="hljs-string">'input_ids'</span>: tensor([[ <span class="hljs-number">2</span>, <span class="hljs-number">235277</span>, <span class="hljs-number">82913</span>, <span class="hljs-number">576</span>, <span class="hljs-number">2656</span>, <span class="hljs-number">30407</span>, <span class="hljs-number">235269</span>, <span class="hljs-number">11490</span>, <span class="hljs-number">970</span>, <span class="hljs-number">29871</span>, <span class="hljs-number">235265</span>]]), | |
| <span class="hljs-string">'attention_mask'</span>: tensor([[<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>]]) | |
| }<!----></pre></div><!----> <p>Acest ghid acoperă încărcarea, encodarea, decodarea, procesarea în batch și backend-urile disponibile pentru tokenizere.</p> <!--[1--><h2 class="relative group"><a id="încarcă-un-tokenizer" 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="#încarcă-un-tokenizer"><span><svg 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>Încarcă un tokenizer</span></h2><!--]--><!----> <p>Încarcă un tokenizer cu clasa <code>AutoTokenizer</code> sau cu o clasă de tokenizer specifică modelului.</p> <div class="flex space-x-2 items-center my-1.5 mr-8 h-7 !pl-0 -mx-3 md:mx-0"><!--[--><div class="flex items-center border rounded-lg px-1.5 py-1 leading-none select-none text-smd border-gray-800 bg-black dark:bg-gray-700 text-white">AutoTokenizer</div><div class="flex items-center border rounded-lg px-1.5 py-1 leading-none select-none text-smd text-gray-500 cursor-pointer opacity-90 hover:text-gray-700 dark:hover:text-gray-200 hover:shadow-sm">model-specific tokenizer</div><!--]--></div> <div class="language-select"><!--[0--><p><code>AutoTokenizer.from_pretrained()</code> citește config-ul modelului, rezolvă clasa corectă de tokenizer și returnează o instanță a ei. Nu trebuie să știi clasa de tokenizer dinainte. Majoritatea tokenizere-lor se rezolvă la o subclasă a <code>TokenizersBackend</code>, un tokenizer rapid bazat pe Rust din biblioteca <a href="https://huggingface.co/docs/tokenizers/index" rel="nofollow">Tokenizers</a>.</p> <p>Încărcarea cu <code>AutoTokenizer</code> este abordarea recomandată.</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 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="language-py "><!----><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoTokenizer | |
| tokenizer = AutoTokenizer.from_pretrained(<span class="hljs-string">"google/gemma-2-2b"</span>)<!----></pre></div><!----><!----><!--]--><!----> <!--[-1--><!--]--><!----><!----></div><!----> <!--[1--><h2 class="relative group"><a id="encodare-și-decodare" 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="#encodare-și-decodare"><span><svg 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>Encodare și decodare</span></h2><!--]--><!----> <p>Metoda <code>TokenizersBackend.__call__()</code> encodează text sau un batch de text în <code>input_ids</code>, <code>attention_mask</code> și alte input-uri pentru model. Controlează și padding-ul, trunchierea și inserarea token-urilor speciale.</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 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="language-py "><!----><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoTokenizer | |
| tokenizer = AutoTokenizer.from_pretrained(<span class="hljs-string">"google/gemma-2-2b"</span>) | |
| tokenizer(<span class="hljs-string">"Sphinx of black quartz, judge my vow."</span>, return_tensors=<span class="hljs-string">"pt"</span>) | |
| { | |
| <span class="hljs-string">'input_ids'</span>: tensor([[ <span class="hljs-number">2</span>, <span class="hljs-number">235277</span>, <span class="hljs-number">82913</span>, <span class="hljs-number">576</span>, <span class="hljs-number">2656</span>, <span class="hljs-number">30407</span>, <span class="hljs-number">235269</span>, <span class="hljs-number">11490</span>, <span class="hljs-number">970</span>, <span class="hljs-number">29871</span>, <span class="hljs-number">235265</span>]]), | |
| <span class="hljs-string">'attention_mask'</span>: tensor([[<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>]]) | |
| }<!----></pre></div><!----> <p><code>TokenizersBackend.encode()</code> este similar, dar returnează doar <code>input_ids</code>.</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 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="language-py "><!----><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoTokenizer | |
| tokenizer = AutoTokenizer.from_pretrained(<span class="hljs-string">"google/gemma-2-2b"</span>) | |
| tokenizer.encode(<span class="hljs-string">"Sphinx of black quartz, judge my vow."</span>) | |
| [<span class="hljs-number">2</span>, <span class="hljs-number">235277</span>, <span class="hljs-number">82913</span>, <span class="hljs-number">576</span>, <span class="hljs-number">2656</span>, <span class="hljs-number">30407</span>, <span class="hljs-number">235269</span>, <span class="hljs-number">11490</span>, <span class="hljs-number">970</span>, <span class="hljs-number">29871</span>, <span class="hljs-number">235265</span>]<!----></pre></div><!----> <p><code>TokenizersBackend.decode()</code> convertește o secvență sau un batch de <code>input_ids</code> tokenizate înapoi în text.</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 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="language-py "><!---->tokenizer.decode(outputs[<span class="hljs-string">"input_ids"</span>]) | |
| [<span class="hljs-string">'<bos>Sphinx of black quartz, judge my vow.'</span>]<!----></pre></div><!----> <p><code>TokenizersBackend.decode()</code> păstrează spațierea exactă a operației de tokenization. Setează <code>clean_up_tokenization_spaces</code> ca să elimini spațiile de dinaintea punctuației și <code>skip_special_tokens</code> ca să scoți token-urile speciale din output.</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 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="language-py "><!---->tokenizer.decode(outputs[<span class="hljs-string">"input_ids"</span>], skip_special_tokens=<span class="hljs-literal">True</span>) | |
| [<span class="hljs-string">'Sphinx of black quartz, judge my vow.'</span>]<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="token-uri-speciale" 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="#token-uri-speciale"><span><svg 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>Token-uri speciale</span></h2><!--]--><!----> <p>Token-urile speciale marchează limitele structurale dintr-o secvență, cum ar fi începutul secvenței sau pozițiile de padding. Fiecare model definește propriul set de token-uri speciale. Tokenizer-ul le adaugă când îl apelezi.</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 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="language-py "><!---->tokenizer.encode(<span class="hljs-string">"Sphinx of black quartz, judge my vow."</span>) | |
| [<span class="hljs-number">2</span>, <span class="hljs-number">235277</span>, <span class="hljs-number">82913</span>, <span class="hljs-number">576</span>, <span class="hljs-number">2656</span>, <span class="hljs-number">30407</span>, <span class="hljs-number">235269</span>, <span class="hljs-number">11490</span>, <span class="hljs-number">970</span>, <span class="hljs-number">29871</span>, <span class="hljs-number">235265</span>] | |
| tokenizer.decode(outputs[<span class="hljs-string">"input_ids"</span>]) | |
| [<span class="hljs-string">'<bos>Sphinx of black quartz, judge my vow.'</span>]<!----></pre></div><!----> <p>Înregistrează token-uri speciale cu nume suplimentare folosind argumentul <code>extra_special_tokens</code>. Modelele multimodale le folosesc ca placeholders pentru imagini, video sau audio.</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 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="language-py "><!----><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoTokenizer | |
| tokenizer = AutoTokenizer.from_pretrained( | |
| <span class="hljs-string">"google/gemma-3-4b-pt"</span>, | |
| extra_special_tokens={<span class="hljs-string">"image_token"</span>: <span class="hljs-string">"<image>"</span>} | |
| )<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="procesare-în-batch" 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="#procesare-în-batch"><span><svg 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>Procesare în batch</span></h2><!--]--><!----> <p>Procesarea în batch tokenizează mai multe secvențe într-un singur apel. <code>TokenizersBackend</code> gestionează batch-uri mari mai rapid pentru că backend-ul său bazat pe Rust paralelizează munca pe thread-uri.</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 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="language-py "><!----><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoTokenizer | |
| tokenizer = AutoTokenizer.from_pretrained(<span class="hljs-string">"google/gemma-2-2b"</span>) | |
| tokenizer( | |
| [ | |
| <span class="hljs-string">"Sphinx of black quartz, judge my vow."</span>, | |
| <span class="hljs-string">"Pack my box with five dozen liquor jugs."</span>, | |
| <span class="hljs-string">"How vexingly quick daft zebras jump!"</span> | |
| ], | |
| return_tensors=<span class="hljs-string">"pt"</span> | |
| )<!----></pre></div><!----> <p>Procesarea în batch presupune că toate secvențele au aceeași lungime. Padding-ul și trunchierea sunt strategii ca să gestionezi secvențe de lungimi diferite.</p> <!--[2--><h3 class="relative group"><a id="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="#padding"><span><svg 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</span></h3><!--]--><!----> <p>Padding-ul adaugă token-uri speciale ca secvențele mai scurte să aibă aceeași lungime cu cea mai lungă din batch. Masca de atenție marchează pozițiile de padding cu <code>0</code> pentru ca modelul să le ignore. Setează <code>padding=True</code> ca să faci padding până la cea mai lungă secvență sau pasează <code>max_length</code> ca să faci padding la o dimensiune fixă.</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 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="language-py "><!----><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoTokenizer | |
| tokenizer = AutoTokenizer.from_pretrained(<span class="hljs-string">"google/gemma-2-2b"</span>) | |
| tokenizer( | |
| [ | |
| <span class="hljs-string">"Sphinx of black quartz, judge my vow."</span>, | |
| <span class="hljs-string">"Pack my box with five dozen liquor jugs."</span>, | |
| <span class="hljs-string">"How vexingly quick daft zebras jump!"</span> | |
| ], | |
| return_tensors=<span class="hljs-string">"pt"</span>, | |
| padding=<span class="hljs-literal">True</span>, | |
| ) | |
| { | |
| <span class="hljs-string">'input_ids'</span>: tensor([ | |
| [ <span class="hljs-number">2</span>, <span class="hljs-number">235277</span>, <span class="hljs-number">82913</span>, <span class="hljs-number">576</span>, <span class="hljs-number">2656</span>, <span class="hljs-number">30407</span>, <span class="hljs-number">235269</span>, <span class="hljs-number">11490</span>, <span class="hljs-number">970</span>, <span class="hljs-number">29871</span>, <span class="hljs-number">235265</span>], | |
| [ <span class="hljs-number">0</span>, <span class="hljs-number">2</span>, <span class="hljs-number">6519</span>, <span class="hljs-number">970</span>, <span class="hljs-number">3741</span>, <span class="hljs-number">675</span>, <span class="hljs-number">4105</span>, <span class="hljs-number">25955</span>, <span class="hljs-number">42184</span>, <span class="hljs-number">225789</span>, <span class="hljs-number">235265</span>], | |
| [ <span class="hljs-number">0</span>, <span class="hljs-number">2</span>, <span class="hljs-number">2299</span>, <span class="hljs-number">73378</span>, <span class="hljs-number">17844</span>, <span class="hljs-number">4320</span>, <span class="hljs-number">224463</span>, <span class="hljs-number">4949</span>, <span class="hljs-number">48977</span>, <span class="hljs-number">9902</span>, <span class="hljs-number">235341</span>] | |
| ]), | |
| <span class="hljs-string">'attention_mask'</span>: tensor([ | |
| [<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">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-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</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-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>] | |
| ]) | |
| }<!----></pre></div><!----> <blockquote class="note"><p>Modelele lingvistice mari fac padding pe partea <em>stângă</em> ca să nu perturbe generarea, care prezice token-ul următor din partea <em>dreaptă</em>.</p></blockquote> <!--[2--><h3 class="relative group"><a id="trunchiere" 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="#trunchiere"><span><svg 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>Trunchiere</span></h3><!--]--><!----> <p>Trunchierea taie token-uri ca o secvență să se încadreze într-o lungime maximă. Setează <code>truncation=True</code> și specifică <code>max_length</code> ca să o activezi.</p> <p>Padding-ul și trunchierea funcționează împreună. Secvențele scurte primesc token-uri de padding, iar cele lungi pierd token-uri de la coadă. Împreună, produc un tensor dreptunghiular compact.</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 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="language-py "><!----><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoTokenizer | |
| tokenizer = AutoTokenizer.from_pretrained(<span class="hljs-string">"google/gemma-2-2b"</span>) | |
| tokenizer( | |
| [ | |
| <span class="hljs-string">"Sphinx of black quartz, judge my vow."</span>, | |
| <span class="hljs-string">"Pack my box with five dozen liquor jugs."</span>, | |
| <span class="hljs-string">"How vexingly quick daft zebras jump!"</span> | |
| ], | |
| return_tensors=<span class="hljs-string">"pt"</span>, | |
| padding=<span class="hljs-literal">True</span>, | |
| truncation=<span class="hljs-literal">True</span>, | |
| max_length=<span class="hljs-number">5</span> | |
| ) | |
| { | |
| <span class="hljs-string">'input_ids'</span>: tensor([ | |
| [ <span class="hljs-number">2</span>, <span class="hljs-number">235277</span>, <span class="hljs-number">82913</span>, <span class="hljs-number">576</span>, <span class="hljs-number">2656</span>], | |
| [ <span class="hljs-number">2</span>, <span class="hljs-number">6519</span>, <span class="hljs-number">970</span>, <span class="hljs-number">3741</span>, <span class="hljs-number">675</span>], | |
| [ <span class="hljs-number">2</span>, <span class="hljs-number">2299</span>, <span class="hljs-number">73378</span>, <span class="hljs-number">17844</span>, <span class="hljs-number">4320</span>] | |
| ]), | |
| <span class="hljs-string">'attention_mask'</span>: tensor([ | |
| [<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>] | |
| ]) | |
| }<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="backend-uri" 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="#backend-uri"><span><svg 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>Backend-uri</span></h2><!--]--><!----> <p>Fiecare tokenizer de model este definit într-un singur fișier și suportă patru backend-uri de tokenization.</p> <table><thead><tr><th>backend</th><th>implementare</th><th>descriere</th></tr></thead><tbody><tr><td><code>TokenizersBackend</code></td><td><a href="https://huggingface.co/docs/tokenizers" rel="nofollow">Tokenizers</a></td><td>implicit pentru majoritatea modelelor</td></tr><tr><td><code>SentencePieceBackend</code></td><td><a href="https://github.com/google/sentencepiece" rel="nofollow">SentencePiece</a></td><td>modele care necesită SentencePiece</td></tr><tr><td><code>PythonBackend</code></td><td>Python</td><td>modele care necesită tokenizere personalizate specializate</td></tr><tr><td><code>MistralCommonBackend</code></td><td><a href="https://mistralai.github.io/mistral-common/" rel="nofollow">mistral-common</a></td><td>modele Mistral și Pixtral</td></tr></tbody></table> <p>Toate backend-urile moștenesc din <code>PreTrainedTokenizerBase</code> și împart aceleași API-uri pentru encodare, decodare, padding, trunchiere, salvare și încărcare. Diferența constă în pipeline-ul de tokenization care rulează pe dedesubt.</p> <p><code>AutoTokenizer</code> selectează cel mai bun backend disponibil când apelezi <code>from_pretrained()</code>.</p> <ol><li><p>Citește fișierul <code>tokenizer_config.json</code> pentru câmpul <code>tokenizer_class</code>.</p></li> <li><p>Registrul potrivește <code>tokenizer_class</code> cu un nume de clasă. Clasa rezolvată moștenește din unul din cele patru backend-uri. De exemplu, <code>GemmaTokenizer</code> moștenește din <code>TokenizersBackend</code>, iar <code>SiglipTokenizer</code> moștenește din <code>SentencePieceBackend</code>.</p> <p>Unele modele, ca GLM, se mapează direct la <code>TokenizersBackend</code> pentru că fișierul <code>tokenizer.json</code> descrie complet pipeline-ul. <code>GemmaTokenizer</code> există ca subclasă deoarece definește setări suplimentare specifice modelului în Python pe care <code>tokenizer.json</code> nu le captează.</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 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="language-py "><!---->TOKENIZER_MAPPING_NAMES = OrderedDict([ | |
| (<span class="hljs-string">"gemma2"</span>, <span class="hljs-string">"GemmaTokenizer"</span> <span class="hljs-keyword">if</span> is_tokenizers_available() <span class="hljs-keyword">else</span> <span class="hljs-literal">None</span>), | |
| (<span class="hljs-string">"glm"</span>, <span class="hljs-string">"TokenizersBackend"</span> <span class="hljs-keyword">if</span> is_tokenizers_available() <span class="hljs-keyword">else</span> <span class="hljs-literal">None</span>), | |
| ( | |
| <span class="hljs-string">"mistral"</span>, | |
| <span class="hljs-string">"MistralCommonBackend"</span> | |
| <span class="hljs-keyword">if</span> is_mistral_common_available() | |
| <span class="hljs-keyword">else</span> (<span class="hljs-string">"TokenizersBackend"</span> <span class="hljs-keyword">if</span> is_tokenizers_available() <span class="hljs-keyword">else</span> <span class="hljs-literal">None</span>), | |
| ), | |
| (<span class="hljs-string">"siglip"</span>, <span class="hljs-string">"SiglipTokenizer"</span> <span class="hljs-keyword">if</span> is_sentencepiece_available() <span class="hljs-keyword">else</span> <span class="hljs-literal">None</span>), | |
| ... | |
| ]<!----></pre></div><!----> <p>Când un backend ca mistral-common nu este instalat, <code>AutoTokenizer</code> revine la <code>TokenizersBackend</code>.</p></li></ol> <p>Verifică ce backend folosește un tokenizer cu proprietatea <code>backend</code>.</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 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="language-py "><!----><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoTokenizer | |
| tokenizer = AutoTokenizer.from_pretrained(<span class="hljs-string">"google/gemma-2-2b"</span>) | |
| tokenizer.backend | |
| <span class="hljs-string">'tokenizers'</span><!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="inspectează-arhitectura-tokenizer-ului" 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="#inspectează-arhitectura-tokenizer-ului"><span><svg 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>Inspectează arhitectura tokenizer-ului</span></h2><!--]--><!----> <p>Inspectează componentele interne ale unui tokenizer (normalizator, pre-tokenizator, model, decoder) cu atributul <code>_tokenizer</code>.</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 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="language-py "><!----><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoTokenizer | |
| tokenizer = AutoTokenizer.from_pretrained(<span class="hljs-string">"google/gemma-2-2b"</span>) | |
| <span class="hljs-built_in">print</span>(tokenizer._tokenizer.normalizer) | |
| <span class="hljs-built_in">print</span>(tokenizer._tokenizer.pre_tokenizer) | |
| <span class="hljs-built_in">print</span>(tokenizer._tokenizer.model) | |
| <span class="hljs-built_in">print</span>(tokenizer._tokenizer.decoder)<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="resurse" 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="#resurse"><span><svg 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>Resurse</span></h2><!--]--><!----> <ul><li>Postarea <a href="https://huggingface.co/blog/tokenizers" rel="nofollow">Tokenization in Transformers v5</a> discută motivația din spatele noilor backend-uri de tokenization.</li> <li>Consultă <a href="https://github.com/huggingface/transformers/blob/main/MIGRATION_GUIDE_V5.md#tokenization" rel="nofollow">ghidul de migrare</a> pentru o prezentare a schimbărilor de tokenization.</li></ul> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/transformers/blob/main/docs/source/ro/fast_tokenizers.md" target="_blank"><svg class="mr-1" 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="M31,16l-7,7l-1.41-1.41L28.17,16l-5.58-5.59L24,9l7,7z"></path><path d="M1,16l7-7l1.41,1.41L3.83,16l5.58,5.59L8,23l-7-7z"></path><path d="M12.419,25.484L17.639,6.552l1.932,0.518L14.351,26.002z"></path></svg><!----> <span><span class="underline">Update</span> on GitHub</span></a><!----> <p></p><!--]--><!----><!--]--><!--]--><!--]--> <!--[-1--><!--]--><!--]--> | |
| <script> | |
| { | |
| __sveltekit_10j3czc = { | |
| base: "/docs/transformers/main/ro", | |
| assets: "/docs/transformers/main/ro" | |
| }; | |
| const element = document.currentScript.parentElement; | |
| Promise.all([ | |
| import("/docs/transformers/main/ro/_app/immutable/entry/start.DwnOcLC_.js"), | |
| import("/docs/transformers/main/ro/_app/immutable/entry/app.Dw7RcaNl.js") | |
| ]).then(([kit, app]) => { | |
| kit.start(app, element, { | |
| node_ids: [0, 11], | |
| data: [null,null], | |
| form: null, | |
| error: null | |
| }); | |
| }); | |
| } | |
| </script> | |
Xet Storage Details
- Size:
- 56.3 kB
- Xet hash:
- 9aa05918de34cdc17fea039e297194e4807b188c64815b2ad01d29245697b568
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.