tokenizers/pr_2113/en/_app/immutable/nodes/9.c46bbd81.js

import{s as ct,o as pt,n as G}from"../chunks/scheduler.7c59faff.js";import{S as dt,i as mt,e as g,s as c,c as k,h as ft,a as $,d,b as p,f as de,g as v,j as me,k as i,l as T,m as w,t as x,n as C,o as _,p as M}from"../chunks/index.09bb5655.js";import{C as ut,H as ue,E as gt}from"../chunks/MermaidChart.svelte_svelte_type_style_lang.1fbbab81.js";import{D as ke}from"../chunks/Docstring.6794b142.js";import{C as ge}from"../chunks/CodeBlock.73acbedb.js";import{T as $t,M as Ae}from"../chunks/TokenizersLanguageContent.0fc17a7a.js";import{E as fe}from"../chunks/ExampleCodeBlock.cd3055e5.js";function ht(b){let e,l="Example:",t,n,a;return n=new ge({props:{code:"ZnJvbSUyMHRva2VuaXplcnMucHJvY2Vzc29ycyUyMGltcG9ydCUyMEJlcnRQcm9jZXNzaW5nJTBBcHJvY2Vzc29yJTIwJTNEJTIwQmVydFByb2Nlc3NpbmcoKCUyMiU1QlNFUCU1RCUyMiUyQyUyMDEwMiklMkMlMjAoJTIyJTVCQ0xTJTVEJTIyJTJDJTIwMTAxKSklMEFwcm9jZXNzb3IucHJvY2VzcyhlbmNvZGluZyk=",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> tokenizers.processors <span class="hljs-keyword">import</span> BertProcessing
<span class="hljs-meta">&gt;&gt;&gt; </span>processor = BertProcessing((<span class="hljs-string">&quot;[SEP]&quot;</span>, <span class="hljs-number">102</span>), (<span class="hljs-string">&quot;[CLS]&quot;</span>, <span class="hljs-number">101</span>))
<span class="hljs-meta">&gt;&gt;&gt; </span>processor.process(encoding)
<span class="hljs-comment"># Encoding with [CLS] at start and [SEP] at end</span>`,lang:"python",wrap:!1}}),{c(){e=g("p"),e.textContent=l,t=c(),k(n.$$.fragment)},l(s){e=$(s,"P",{"data-svelte-h":!0}),M(e)!=="svelte-11lpom8"&&(e.textContent=l),t=p(s),v(n.$$.fragment,s)},m(s,y){T(s,e,y),T(s,t,y),w(n,s,y),a=!0},p:G,i(s){a||(x(n.$$.fragment,s),a=!0)},o(s){C(n.$$.fragment,s),a=!1},d(s){s&&(d(e),d(t)),_(n,s)}}}function Tt(b){let e,l="Example:",t,n,a;return n=new ge({props:{code:"ZnJvbSUyMHRva2VuaXplcnMucHJvY2Vzc29ycyUyMGltcG9ydCUyMEJ5dGVMZXZlbCUwQXByb2Nlc3NvciUyMCUzRCUyMEJ5dGVMZXZlbCh0cmltX29mZnNldHMlM0RUcnVlKSUwQSUyMyUyME9mZnNldHMlMjB3aWxsJTIwYmUlMjB0cmltbWVkJTIwdG8lMjBleGNsdWRlJTIwbGVhZGluZyUyMHdoaXRlc3BhY2UlMjBieXRlcw==",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> tokenizers.processors <span class="hljs-keyword">import</span> ByteLevel
<span class="hljs-meta">&gt;&gt;&gt; </span>processor = ByteLevel(trim_offsets=<span class="hljs-literal">True</span>)
<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-comment"># Offsets will be trimmed to exclude leading whitespace bytes</span>`,lang:"python",wrap:!1}}),{c(){e=g("p"),e.textContent=l,t=c(),k(n.$$.fragment)},l(s){e=$(s,"P",{"data-svelte-h":!0}),M(e)!=="svelte-11lpom8"&&(e.textContent=l),t=p(s),v(n.$$.fragment,s)},m(s,y){T(s,e,y),T(s,t,y),w(n,s,y),a=!0},p:G,i(s){a||(x(n.$$.fragment,s),a=!0)},o(s){C(n.$$.fragment,s),a=!1},d(s){s&&(d(e),d(t)),_(n,s)}}}function yt(b){let e,l="Example:",t,n,a;return n=new ge({props:{code:"ZnJvbSUyMHRva2VuaXplcnMucHJvY2Vzc29ycyUyMGltcG9ydCUyMFJvYmVydGFQcm9jZXNzaW5nJTBBcHJvY2Vzc29yJTIwJTNEJTIwUm9iZXJ0YVByb2Nlc3NpbmcoKCUyMiUzQyUyRnMlM0UlMjIlMkMlMjAyKSUyQyUyMCglMjIlM0NzJTNFJTIyJTJDJTIwMCkpJTBBcHJvY2Vzc29yLnByb2Nlc3MoZW5jb2Rpbmcp",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> tokenizers.processors <span class="hljs-keyword">import</span> RobertaProcessing
<span class="hljs-meta">&gt;&gt;&gt; </span>processor = RobertaProcessing((<span class="hljs-string">&quot;&lt;/s&gt;&quot;</span>, <span class="hljs-number">2</span>), (<span class="hljs-string">&quot;&lt;s&gt;&quot;</span>, <span class="hljs-number">0</span>))
<span class="hljs-meta">&gt;&gt;&gt; </span>processor.process(encoding)
<span class="hljs-comment"># Encoding with &lt;s&gt; at start and &lt;/s&gt; at end</span>`,lang:"python",wrap:!1}}),{c(){e=g("p"),e.textContent=l,t=c(),k(n.$$.fragment)},l(s){e=$(s,"P",{"data-svelte-h":!0}),M(e)!=="svelte-11lpom8"&&(e.textContent=l),t=p(s),v(n.$$.fragment,s)},m(s,y){T(s,e,y),T(s,t,y),w(n,s,y),a=!0},p:G,i(s){a||(x(n.$$.fragment,s),a=!0)},o(s){C(n.$$.fragment,s),a=!1},d(s){s&&(d(e),d(t)),_(n,s)}}}function Mt(b){let e,l="With the type ids as following:",t,n,a;return n=new ge({props:{code:"JTVCQ0xTJTVEJTIwJTIwJTIwLi4uJTIwJTIwJTIwJTVCU0VQJTVEJTIwJTIwJTIwLi4uJTIwJTIwJTIwJTVCU0VQJTVEJTBBMCUyMCUyMCUyMCUyMCUyMCUyMDAlMjAlMjAlMjAlMjAlMjAlMjAwJTIwJTIwJTIwJTIwJTIwJTIwMSUyMCUyMCUyMCUyMCUyMCUyMDE=",highlighted:`[CLS]   ...   [SEP]   ...   [SEP]
<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>`,lang:"python",wrap:!1}}),{c(){e=g("p"),e.textContent=l,t=c(),k(n.$$.fragment)},l(s){e=$(s,"P",{"data-svelte-h":!0}),M(e)!=="svelte-tdrlp0"&&(e.textContent=l),t=p(s),v(n.$$.fragment,s)},m(s,y){T(s,e,y),T(s,t,y),w(n,s,y),a=!0},p:G,i(s){a||(x(n.$$.fragment,s),a=!0)},o(s){C(n.$$.fragment,s),a=!1},d(s){s&&(d(e),d(t)),_(n,s)}}}function kt(b){let e,l="You can achieve such behavior using a TemplateProcessing:",t,n,a;return n=new ge({props:{code:"VGVtcGxhdGVQcm9jZXNzaW5nKCUwQSUyMCUyMCUyMCUyMHNpbmdsZSUzRCUyMiU1QkNMUyU1RCUyMCUyNDAlMjAlNUJTRVAlNUQlMjIlMkMlMEElMjAlMjAlMjAlMjBwYWlyJTNEJTIyJTVCQ0xTJTVEJTIwJTI0QSUyMCU1QlNFUCU1RCUyMCUyNEIlM0ExJTIwJTVCU0VQJTVEJTNBMSUyMiUyQyUwQSUyMCUyMCUyMCUyMHNwZWNpYWxfdG9rZW5zJTNEJTVCKCUyMiU1QkNMUyU1RCUyMiUyQyUyMDEpJTJDJTIwKCUyMiU1QlNFUCU1RCUyMiUyQyUyMDApJTVEJTJDJTBBKQ==",highlighted:`TemplateProcessing(
    single=<span class="hljs-string">&quot;[CLS] $0 [SEP]&quot;</span>,
    pair=<span class="hljs-string">&quot;[CLS] $A [SEP] $B:1 [SEP]:1&quot;</span>,
    special_tokens=[(<span class="hljs-string">&quot;[CLS]&quot;</span>, <span class="hljs-number">1</span>), (<span class="hljs-string">&quot;[SEP]&quot;</span>, <span class="hljs-number">0</span>)],
)`,lang:"python",wrap:!1}}),{c(){e=g("p"),e.textContent=l,t=c(),k(n.$$.fragment)},l(s){e=$(s,"P",{"data-svelte-h":!0}),M(e)!=="svelte-1qw3em7"&&(e.textContent=l),t=p(s),v(n.$$.fragment,s)},m(s,y){T(s,e,y),T(s,t,y),w(n,s,y),a=!0},p:G,i(s){a||(x(n.$$.fragment,s),a=!0)},o(s){C(n.$$.fragment,s),a=!1},d(s){s&&(d(e),d(t)),_(n,s)}}}function vt(b){let e,l,t,n,a,s,y=`This post-processor takes care of adding the special tokens needed by
a Bert model:`,S,U,q="<li>a SEP token</li> <li>a CLS token</li>",z,J,I,j,r,h,B,ve,W,Qe="This post-processor takes care of trimming the offsets.",we,Y,De=`By default, the ByteLevel BPE might include whitespaces in the produced tokens. If you don’t
want the offsets to include these whitespaces, then this PostProcessor must be used.`,xe,H,$e,A,he,P,Q,Ce,X,Ze=`This post-processor takes care of adding the special tokens needed by
a Roberta model:`,_e,F,Ge="<li>a SEP token</li> <li>a CLS token</li>",be,K,We=`It also takes care of trimming the offsets.
By default, the ByteLevel BPE might include whitespaces in the produced tokens. If you don’t
want the offsets to include these whitespaces, then this PostProcessor should be initialized
with <code>trim_offsets=True</code>`,Ue,V,Te,D,ye,m,Z,Pe,O,Ye=`Provides a way to specify templates in order to add the special tokens to each
input sequence as relevant.`,Je,ee,Xe=`Let’s take <code>BERT</code> tokenizer as an example. It uses two special tokens, used to
delimitate each sequence. <code>[CLS]</code> is always used at the beginning of the first
sequence, and <code>[SEP]</code> is added at the end of both the first, and the pair
sequences. The final result looks like this:`,Le,te,Fe="<li>Single sequence: <code>[CLS] Hello there [SEP]</code></li> <li>Pair sequences: <code>[CLS] My name is Anthony [SEP] What is my name? [SEP]</code></li>",je,N,ze,R,Ee,se,Ke=`In this example, each input sequence is identified using a <code>$</code> construct. This identifier
lets us specify each input sequence, and the type_id to use. When nothing is specified,
it uses the default values. Here are the different ways to specify it:`,Ie,ne,Oe="<li>Specifying the sequence, with default <code>type_id == 0</code>: <code>$A</code> or <code>$B</code></li> <li>Specifying the <em>type_id</em> with default <code>sequence == A</code>: <code>$0</code>, <code>$1</code>, <code>$2</code>, …</li> <li>Specifying both: <code>$A:0</code>, <code>$B:1</code>, …</li>",Be,oe,et="The same construct is used for special tokens: <code>&lt;identifier&gt;(:&lt;type_id&gt;)?</code>.",Se,re,tt=`<strong>Warning</strong>: You must ensure that you are giving the correct tokens/ids as these
will be added to the Encoding without any further check. If the given ids correspond
to something totally different in a <em>Tokenizer</em> using this <em>PostProcessor</em>, it
might lead to unexpected results.`,He,le,st="Types:",Ve,ae,nt="Template (<code>str</code> or <code>List</code>):",Ne,ie,ot="<li>If a <code>str</code> is provided, the whitespace is used as delimiter between tokens</li> <li>If a <code>List[str]</code> is provided, a list of tokens</li>",Re,ce,rt="Tokens (<code>List[Union[Tuple[int, str], Tuple[str, int], dict]]</code>):",qe,pe,lt=`<li><p>A <code>Tuple</code> with both a token and its associated ID, in any order</p></li> <li><p>A <code>dict</code> with the following keys:</p> <ul><li>“id”: <code>str</code> =&gt; The special token id, as specified in the Template</li> <li>“ids”: <code>List[int]</code> =&gt; The associated IDs</li> <li>“tokens”: <code>List[str]</code> =&gt; The associated tokens</li></ul> <p>The given dict expects the provided <code>ids</code> and <code>tokens</code> lists to have
the same length.</p></li>`,Me;return e=new ue({props:{title:"BertProcessing",local:"tokenizers.processors.BertProcessing",headingTag:"h2"}}),n=new ke({props:{name:"class tokenizers.processors.BertProcessing",anchor:"tokenizers.processors.BertProcessing",parameters:"",parametersDescription:[{anchor:"tokenizers.processors.BertProcessing.sep",description:`<strong>sep</strong> (<code>Tuple[str, int]</code>) &#x2014;
A tuple with the string representation of the SEP token, and its id`,name:"sep"},{anchor:"tokenizers.processors.BertProcessing.cls",description:`<strong>cls</strong> (<code>Tuple[str, int]</code>) &#x2014;
A tuple with the string representation of the CLS token, and its id`,name:"cls"}]}}),J=new fe({props:{anchor:"tokenizers.processors.BertProcessing.example",$$slots:{default:[ht]},$$scope:{ctx:b}}}),j=new ue({props:{title:"ByteLevel",local:"tokenizers.processors.ByteLevel",headingTag:"h2"}}),B=new ke({props:{name:"class tokenizers.processors.ByteLevel",anchor:"tokenizers.processors.ByteLevel",parameters:[{name:"add_prefix_space",val:" = None"},{name:"trim_offsets",val:" = None"},{name:"use_regex",val:" = None"}],parametersDescription:[{anchor:"tokenizers.processors.ByteLevel.trim_offsets",description:`<strong>trim_offsets</strong> (<code>bool</code>) &#x2014;
Whether to trim the whitespaces from the produced offsets.`,name:"trim_offsets"},{anchor:"tokenizers.processors.ByteLevel.add_prefix_space",description:`<strong>add_prefix_space</strong> (<code>bool</code>, <em>optional</em>, defaults to <code>True</code>) &#x2014;
If <code>True</code>, keeps the first token&#x2019;s offset as is. If <code>False</code>, increments
the start of the first token&#x2019;s offset by 1. Only has an effect if <code>trim_offsets</code>
is set to <code>True</code>.`,name:"add_prefix_space"}]}}),H=new fe({props:{anchor:"tokenizers.processors.ByteLevel.example",$$slots:{default:[Tt]},$$scope:{ctx:b}}}),A=new ue({props:{title:"RobertaProcessing",local:"tokenizers.processors.RobertaProcessing",headingTag:"h2"}}),Q=new ke({props:{name:"class tokenizers.processors.RobertaProcessing",anchor:"tokenizers.processors.RobertaProcessing",parameters:[{name:"sep",val:""},{name:"cls_token",val:""},{name:"trim_offsets",val:" = True"},{name:"add_prefix_space",val:" = True"}],parametersDescription:[{anchor:"tokenizers.processors.RobertaProcessing.sep",description:`<strong>sep</strong> (<code>Tuple[str, int]</code>) &#x2014;
A tuple with the string representation of the SEP token, and its id`,name:"sep"},{anchor:"tokenizers.processors.RobertaProcessing.cls",description:`<strong>cls</strong> (<code>Tuple[str, int]</code>) &#x2014;
A tuple with the string representation of the CLS token, and its id`,name:"cls"},{anchor:"tokenizers.processors.RobertaProcessing.trim_offsets",description:`<strong>trim_offsets</strong> (<code>bool</code>, <em>optional</em>, defaults to <code>True</code>) &#x2014;
Whether to trim the whitespaces from the produced offsets.`,name:"trim_offsets"},{anchor:"tokenizers.processors.RobertaProcessing.add_prefix_space",description:`<strong>add_prefix_space</strong> (<code>bool</code>, <em>optional</em>, defaults to <code>True</code>) &#x2014;
Whether the add_prefix_space option was enabled during pre-tokenization. This
is relevant because it defines the way the offsets are trimmed out.`,name:"add_prefix_space"}]}}),V=new fe({props:{anchor:"tokenizers.processors.RobertaProcessing.example",$$slots:{default:[yt]},$$scope:{ctx:b}}}),D=new ue({props:{title:"TemplateProcessing",local:"tokenizers.processors.TemplateProcessing",headingTag:"h2"}}),Z=new ke({props:{name:"class tokenizers.processors.TemplateProcessing",anchor:"tokenizers.processors.TemplateProcessing",parameters:[{name:"single",val:" = None"},{name:"pair",val:" = None"},{name:"special_tokens",val:" = None"}],parametersDescription:[{anchor:"tokenizers.processors.TemplateProcessing.single",description:`<strong>single</strong> (<code>Template</code>) &#x2014;
The template used for single sequences`,name:"single"},{anchor:"tokenizers.processors.TemplateProcessing.pair",description:`<strong>pair</strong> (<code>Template</code>) &#x2014;
The template used when both sequences are specified`,name:"pair"},{anchor:"tokenizers.processors.TemplateProcessing.special_tokens",description:`<strong>special_tokens</strong> (<code>Tokens</code>) &#x2014;
The list of special tokens used in each sequences`,name:"special_tokens"}]}}),N=new fe({props:{anchor:"tokenizers.processors.TemplateProcessing.example",$$slots:{default:[Mt]},$$scope:{ctx:b}}}),R=new fe({props:{anchor:"tokenizers.processors.TemplateProcessing.example-2",$$slots:{default:[kt]},$$scope:{ctx:b}}}),{c(){k(e.$$.fragment),l=c(),t=g("div"),k(n.$$.fragment),a=c(),s=g("p"),s.textContent=y,S=c(),U=g("ul"),U.innerHTML=q,z=c(),k(J.$$.fragment),I=c(),k(j.$$.fragment),r=c(),h=g("div"),k(B.$$.fragment),ve=c(),W=g("p"),W.textContent=Qe,we=c(),Y=g("p"),Y.textContent=De,xe=c(),k(H.$$.fragment),$e=c(),k(A.$$.fragment),he=c(),P=g("div"),k(Q.$$.fragment),Ce=c(),X=g("p"),X.textContent=Ze,_e=c(),F=g("ul"),F.innerHTML=Ge,be=c(),K=g("p"),K.innerHTML=We,Ue=c(),k(V.$$.fragment),Te=c(),k(D.$$.fragment),ye=c(),m=g("div"),k(Z.$$.fragment),Pe=c(),O=g("p"),O.textContent=Ye,Je=c(),ee=g("p"),ee.innerHTML=Xe,Le=c(),te=g("ul"),te.innerHTML=Fe,je=c(),k(N.$$.fragment),ze=c(),k(R.$$.fragment),Ee=c(),se=g("p"),se.innerHTML=Ke,Ie=c(),ne=g("ul"),ne.innerHTML=Oe,Be=c(),oe=g("p"),oe.innerHTML=et,Se=c(),re=g("p"),re.innerHTML=tt,He=c(),le=g("p"),le.textContent=st,Ve=c(),ae=g("p"),ae.innerHTML=nt,Ne=c(),ie=g("ul"),ie.innerHTML=ot,Re=c(),ce=g("p"),ce.innerHTML=rt,qe=c(),pe=g("ul"),pe.innerHTML=lt,this.h()},l(o){v(e.$$.fragment,o),l=p(o),t=$(o,"DIV",{class:!0});var f=de(t);v(n.$$.fragment,f),a=p(f),s=$(f,"P",{"data-svelte-h":!0}),M(s)!=="svelte-j5l3j3"&&(s.textContent=y),S=p(f),U=$(f,"UL",{"data-svelte-h":!0}),M(U)!=="svelte-19nfwue"&&(U.innerHTML=q),z=p(f),v(J.$$.fragment,f),f.forEach(d),I=p(o),v(j.$$.fragment,o),r=p(o),h=$(o,"DIV",{class:!0});var E=de(h);v(B.$$.fragment,E),ve=p(E),W=$(E,"P",{"data-svelte-h":!0}),M(W)!=="svelte-1pew3nx"&&(W.textContent=Qe),we=p(E),Y=$(E,"P",{"data-svelte-h":!0}),M(Y)!=="svelte-18el03w"&&(Y.textContent=De),xe=p(E),v(H.$$.fragment,E),E.forEach(d),$e=p(o),v(A.$$.fragment,o),he=p(o),P=$(o,"DIV",{class:!0});var L=de(P);v(Q.$$.fragment,L),Ce=p(L),X=$(L,"P",{"data-svelte-h":!0}),M(X)!=="svelte-13xluxv"&&(X.textContent=Ze),_e=p(L),F=$(L,"UL",{"data-svelte-h":!0}),M(F)!=="svelte-19nfwue"&&(F.innerHTML=Ge),be=p(L),K=$(L,"P",{"data-svelte-h":!0}),M(K)!=="svelte-5hdvuy"&&(K.innerHTML=We),Ue=p(L),v(V.$$.fragment,L),L.forEach(d),Te=p(o),v(D.$$.fragment,o),ye=p(o),m=$(o,"DIV",{class:!0});var u=de(m);v(Z.$$.fragment,u),Pe=p(u),O=$(u,"P",{"data-svelte-h":!0}),M(O)!=="svelte-17j0bbk"&&(O.textContent=Ye),Je=p(u),ee=$(u,"P",{"data-svelte-h":!0}),M(ee)!=="svelte-1jzrjo"&&(ee.innerHTML=Xe),Le=p(u),te=$(u,"UL",{"data-svelte-h":!0}),M(te)!=="svelte-bbb8xt"&&(te.innerHTML=Fe),je=p(u),v(N.$$.fragment,u),ze=p(u),v(R.$$.fragment,u),Ee=p(u),se=$(u,"P",{"data-svelte-h":!0}),M(se)!=="svelte-1px755u"&&(se.innerHTML=Ke),Ie=p(u),ne=$(u,"UL",{"data-svelte-h":!0}),M(ne)!=="svelte-1jv5met"&&(ne.innerHTML=Oe),Be=p(u),oe=$(u,"P",{"data-svelte-h":!0}),M(oe)!=="svelte-1h9guxl"&&(oe.innerHTML=et),Se=p(u),re=$(u,"P",{"data-svelte-h":!0}),M(re)!=="svelte-svd9p0"&&(re.innerHTML=tt),He=p(u),le=$(u,"P",{"data-svelte-h":!0}),M(le)!=="svelte-19y9q03"&&(le.textContent=st),Ve=p(u),ae=$(u,"P",{"data-svelte-h":!0}),M(ae)!=="svelte-fxju4z"&&(ae.innerHTML=nt),Ne=p(u),ie=$(u,"UL",{"data-svelte-h":!0}),M(ie)!=="svelte-hg8fj8"&&(ie.innerHTML=ot),Re=p(u),ce=$(u,"P",{"data-svelte-h":!0}),M(ce)!=="svelte-150u5tp"&&(ce.innerHTML=rt),qe=p(u),pe=$(u,"UL",{"data-svelte-h":!0}),M(pe)!=="svelte-1xe1sl4"&&(pe.innerHTML=lt),u.forEach(d),this.h()},h(){me(t,"class","docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8"),me(h,"class","docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8"),me(P,"class","docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8"),me(m,"class","docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8")},m(o,f){w(e,o,f),T(o,l,f),T(o,t,f),w(n,t,null),i(t,a),i(t,s),i(t,S),i(t,U),i(t,z),w(J,t,null),T(o,I,f),w(j,o,f),T(o,r,f),T(o,h,f),w(B,h,null),i(h,ve),i(h,W),i(h,we),i(h,Y),i(h,xe),w(H,h,null),T(o,$e,f),w(A,o,f),T(o,he,f),T(o,P,f),w(Q,P,null),i(P,Ce),i(P,X),i(P,_e),i(P,F),i(P,be),i(P,K),i(P,Ue),w(V,P,null),T(o,Te,f),w(D,o,f),T(o,ye,f),T(o,m,f),w(Z,m,null),i(m,Pe),i(m,O),i(m,Je),i(m,ee),i(m,Le),i(m,te),i(m,je),w(N,m,null),i(m,ze),w(R,m,null),i(m,Ee),i(m,se),i(m,Ie),i(m,ne),i(m,Be),i(m,oe),i(m,Se),i(m,re),i(m,He),i(m,le),i(m,Ve),i(m,ae),i(m,Ne),i(m,ie),i(m,Re),i(m,ce),i(m,qe),i(m,pe),Me=!0},p(o,f){const E={};f&2&&(E.$$scope={dirty:f,ctx:o}),J.$set(E);const L={};f&2&&(L.$$scope={dirty:f,ctx:o}),H.$set(L);const u={};f&2&&(u.$$scope={dirty:f,ctx:o}),V.$set(u);const at={};f&2&&(at.$$scope={dirty:f,ctx:o}),N.$set(at);const it={};f&2&&(it.$$scope={dirty:f,ctx:o}),R.$set(it)},i(o){Me||(x(e.$$.fragment,o),x(n.$$.fragment,o),x(J.$$.fragment,o),x(j.$$.fragment,o),x(B.$$.fragment,o),x(H.$$.fragment,o),x(A.$$.fragment,o),x(Q.$$.fragment,o),x(V.$$.fragment,o),x(D.$$.fragment,o),x(Z.$$.fragment,o),x(N.$$.fragment,o),x(R.$$.fragment,o),Me=!0)},o(o){C(e.$$.fragment,o),C(n.$$.fragment,o),C(J.$$.fragment,o),C(j.$$.fragment,o),C(B.$$.fragment,o),C(H.$$.fragment,o),C(A.$$.fragment,o),C(Q.$$.fragment,o),C(V.$$.fragment,o),C(D.$$.fragment,o),C(Z.$$.fragment,o),C(N.$$.fragment,o),C(R.$$.fragment,o),Me=!1},d(o){o&&(d(l),d(t),d(I),d(r),d(h),d($e),d(he),d(P),d(Te),d(ye),d(m)),_(e,o),_(n),_(J),_(j,o),_(B),_(H),_(A,o),_(Q),_(V),_(D,o),_(Z),_(N),_(R)}}}function wt(b){let e,l;return e=new Ae({props:{$$slots:{default:[vt]},$$scope:{ctx:b}}}),{c(){k(e.$$.fragment)},l(t){v(e.$$.fragment,t)},m(t,n){w(e,t,n),l=!0},p(t,n){const a={};n&2&&(a.$$scope={dirty:n,ctx:t}),e.$set(a)},i(t){l||(x(e.$$.fragment,t),l=!0)},o(t){C(e.$$.fragment,t),l=!1},d(t){_(e,t)}}}function xt(b){let e,l='The Rust API Reference is available directly on the <a href="https://docs.rs/tokenizers/latest/tokenizers/" rel="nofollow">Docs.rs</a> website.';return{c(){e=g("p"),e.innerHTML=l},l(t){e=$(t,"P",{"data-svelte-h":!0}),M(e)!=="svelte-4ytcyb"&&(e.innerHTML=l)},m(t,n){T(t,e,n)},p:G,d(t){t&&d(e)}}}function Ct(b){let e,l;return e=new Ae({props:{$$slots:{default:[xt]},$$scope:{ctx:b}}}),{c(){k(e.$$.fragment)},l(t){v(e.$$.fragment,t)},m(t,n){w(e,t,n),l=!0},p(t,n){const a={};n&2&&(a.$$scope={dirty:n,ctx:t}),e.$set(a)},i(t){l||(x(e.$$.fragment,t),l=!0)},o(t){C(e.$$.fragment,t),l=!1},d(t){_(e,t)}}}function _t(b){let e,l="The node API has not been documented yet.";return{c(){e=g("p"),e.textContent=l},l(t){e=$(t,"P",{"data-svelte-h":!0}),M(e)!=="svelte-1mrchm6"&&(e.textContent=l)},m(t,n){T(t,e,n)},p:G,d(t){t&&d(e)}}}function bt(b){let e,l;return e=new Ae({props:{$$slots:{default:[_t]},$$scope:{ctx:b}}}),{c(){k(e.$$.fragment)},l(t){v(e.$$.fragment,t)},m(t,n){w(e,t,n),l=!0},p(t,n){const a={};n&2&&(a.$$scope={dirty:n,ctx:t}),e.$set(a)},i(t){l||(x(e.$$.fragment,t),l=!0)},o(t){C(e.$$.fragment,t),l=!1},d(t){_(e,t)}}}function Ut(b){let e,l,t,n,a,s,y,S,U,q,z,J,I,j;return a=new ut({props:{containerStyle:"float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;"}}),y=new ue({props:{title:"Post-processors",local:"post-processors",headingTag:"h1"}}),U=new $t({props:{python:!0,rust:!0,node:!0,$$slots:{node:[bt],rust:[Ct],python:[wt]},$$scope:{ctx:b}}}),z=new gt({props:{source:"https://github.com/huggingface/tokenizers/blob/main/docs/source-doc-builder/api/post-processors.mdx"}}),{c(){e=g("meta"),l=c(),t=g("p"),n=c(),k(a.$$.fragment),s=c(),k(y.$$.fragment),S=c(),k(U.$$.fragment),q=c(),k(z.$$.fragment),J=c(),I=g("p"),this.h()},l(r){const h=ft("svelte-u9bgzb",document.head);e=$(h,"META",{name:!0,content:!0}),h.forEach(d),l=p(r),t=$(r,"P",{}),de(t).forEach(d),n=p(r),v(a.$$.fragment,r),s=p(r),v(y.$$.fragment,r),S=p(r),v(U.$$.fragment,r),q=p(r),v(z.$$.fragment,r),J=p(r),I=$(r,"P",{}),de(I).forEach(d),this.h()},h(){me(e,"name","hf:doc:metadata"),me(e,"content",Pt)},m(r,h){i(document.head,e),T(r,l,h),T(r,t,h),T(r,n,h),w(a,r,h),T(r,s,h),w(y,r,h),T(r,S,h),w(U,r,h),T(r,q,h),w(z,r,h),T(r,J,h),T(r,I,h),j=!0},p(r,[h]){const B={};h&2&&(B.$$scope={dirty:h,ctx:r}),U.$set(B)},i(r){j||(x(a.$$.fragment,r),x(y.$$.fragment,r),x(U.$$.fragment,r),x(z.$$.fragment,r),j=!0)},o(r){C(a.$$.fragment,r),C(y.$$.fragment,r),C(U.$$.fragment,r),C(z.$$.fragment,r),j=!1},d(r){r&&(d(l),d(t),d(n),d(s),d(S),d(q),d(J),d(I)),d(e),_(a,r),_(y,r),_(U,r),_(z,r)}}}const Pt='{"title":"Post-processors","local":"post-processors","sections":[{"title":"BertProcessing","local":"tokenizers.processors.BertProcessing","sections":[],"depth":2},{"title":"ByteLevel","local":"tokenizers.processors.ByteLevel","sections":[],"depth":2},{"title":"RobertaProcessing","local":"tokenizers.processors.RobertaProcessing","sections":[],"depth":2},{"title":"TemplateProcessing","local":"tokenizers.processors.TemplateProcessing","sections":[],"depth":2}],"depth":1}';function Jt(b){return pt(()=>{new URLSearchParams(window.location.search).get("fw")}),[]}class Ht extends dt{constructor(e){super(),mt(this,e,Jt,Ut,ct,{})}}export{Ht as component};