How to byte-pair encode character sequences at prediction time?












0












$begingroup$


Suppose I have a vocabulary V = {'f', 'o', 'b', 'a', 'r', 'fo', 'ob', 'ar', 'foob', 'obar'} from BPE and at prediction time I come across an input 'foobar'. Either ['foob', 'ar'] or ['fo', 'obar'] would seem a reasonable encoding. What's the best way to handle this situation? If we greedily encode the character sequence by iteratively finding the longest next subsequence in V can it be suboptimal (e.g. there is an encoding with fewer subwords or one that our language model knows better)?









share







New contributor




Gareth Ari Aye is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.







$endgroup$

















    0












    $begingroup$


    Suppose I have a vocabulary V = {'f', 'o', 'b', 'a', 'r', 'fo', 'ob', 'ar', 'foob', 'obar'} from BPE and at prediction time I come across an input 'foobar'. Either ['foob', 'ar'] or ['fo', 'obar'] would seem a reasonable encoding. What's the best way to handle this situation? If we greedily encode the character sequence by iteratively finding the longest next subsequence in V can it be suboptimal (e.g. there is an encoding with fewer subwords or one that our language model knows better)?









    share







    New contributor




    Gareth Ari Aye is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
    Check out our Code of Conduct.







    $endgroup$















      0












      0








      0





      $begingroup$


      Suppose I have a vocabulary V = {'f', 'o', 'b', 'a', 'r', 'fo', 'ob', 'ar', 'foob', 'obar'} from BPE and at prediction time I come across an input 'foobar'. Either ['foob', 'ar'] or ['fo', 'obar'] would seem a reasonable encoding. What's the best way to handle this situation? If we greedily encode the character sequence by iteratively finding the longest next subsequence in V can it be suboptimal (e.g. there is an encoding with fewer subwords or one that our language model knows better)?









      share







      New contributor




      Gareth Ari Aye is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
      Check out our Code of Conduct.







      $endgroup$




      Suppose I have a vocabulary V = {'f', 'o', 'b', 'a', 'r', 'fo', 'ob', 'ar', 'foob', 'obar'} from BPE and at prediction time I come across an input 'foobar'. Either ['foob', 'ar'] or ['fo', 'obar'] would seem a reasonable encoding. What's the best way to handle this situation? If we greedily encode the character sequence by iteratively finding the longest next subsequence in V can it be suboptimal (e.g. there is an encoding with fewer subwords or one that our language model knows better)?







      natural-language-process language-model





      share







      New contributor




      Gareth Ari Aye is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
      Check out our Code of Conduct.










      share







      New contributor




      Gareth Ari Aye is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
      Check out our Code of Conduct.








      share



      share






      New contributor




      Gareth Ari Aye is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
      Check out our Code of Conduct.









      asked 4 mins ago









      Gareth Ari AyeGareth Ari Aye

      1




      1




      New contributor




      Gareth Ari Aye is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
      Check out our Code of Conduct.





      New contributor





      Gareth Ari Aye is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
      Check out our Code of Conduct.






      Gareth Ari Aye is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
      Check out our Code of Conduct.






















          0






          active

          oldest

          votes












          Your Answer








          StackExchange.ready(function() {
          var channelOptions = {
          tags: "".split(" "),
          id: "557"
          };
          initTagRenderer("".split(" "), "".split(" "), channelOptions);

          StackExchange.using("externalEditor", function() {
          // Have to fire editor after snippets, if snippets enabled
          if (StackExchange.settings.snippets.snippetsEnabled) {
          StackExchange.using("snippets", function() {
          createEditor();
          });
          }
          else {
          createEditor();
          }
          });

          function createEditor() {
          StackExchange.prepareEditor({
          heartbeatType: 'answer',
          autoActivateHeartbeat: false,
          convertImagesToLinks: false,
          noModals: true,
          showLowRepImageUploadWarning: true,
          reputationToPostImages: null,
          bindNavPrevention: true,
          postfix: "",
          imageUploader: {
          brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
          contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
          allowUrls: true
          },
          onDemand: true,
          discardSelector: ".discard-answer"
          ,immediatelyShowMarkdownHelp:true
          });


          }
          });






          Gareth Ari Aye is a new contributor. Be nice, and check out our Code of Conduct.










          draft saved

          draft discarded


















          StackExchange.ready(
          function () {
          StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fdatascience.stackexchange.com%2fquestions%2f49656%2fhow-to-byte-pair-encode-character-sequences-at-prediction-time%23new-answer', 'question_page');
          }
          );

          Post as a guest















          Required, but never shown

























          0






          active

          oldest

          votes








          0






          active

          oldest

          votes









          active

          oldest

          votes






          active

          oldest

          votes








          Gareth Ari Aye is a new contributor. Be nice, and check out our Code of Conduct.










          draft saved

          draft discarded


















          Gareth Ari Aye is a new contributor. Be nice, and check out our Code of Conduct.













          Gareth Ari Aye is a new contributor. Be nice, and check out our Code of Conduct.












          Gareth Ari Aye is a new contributor. Be nice, and check out our Code of Conduct.
















          Thanks for contributing an answer to Data Science Stack Exchange!


          • Please be sure to answer the question. Provide details and share your research!

          But avoid



          • Asking for help, clarification, or responding to other answers.

          • Making statements based on opinion; back them up with references or personal experience.


          Use MathJax to format equations. MathJax reference.


          To learn more, see our tips on writing great answers.




          draft saved


          draft discarded














          StackExchange.ready(
          function () {
          StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fdatascience.stackexchange.com%2fquestions%2f49656%2fhow-to-byte-pair-encode-character-sequences-at-prediction-time%23new-answer', 'question_page');
          }
          );

          Post as a guest















          Required, but never shown





















































          Required, but never shown














          Required, but never shown












          Required, but never shown







          Required, but never shown

































          Required, but never shown














          Required, but never shown












          Required, but never shown







          Required, but never shown







          Popular posts from this blog

          How to label and detect the document text images

          Vallis Paradisi

          Tabula Rosettana