GV

GeezerVison

The World Wide Web, restored to 800 pixels of glory.

Home|About|Gaming cabinet|Archives|How it works|Starter sites

On this site

  • Sign in
  • ActionsCI/CD and automation
  • CodespacesCloud development environments
  • All Features
  • Documentation
  • Training
  • Certifications
  • Build with Copilot
  • Customize Copilot
  • Integrations
  • Trending repos
  • GitHub Sponsors
  • Developer articles
  • Enterprises

This view

  • Original page
  • New address

GeezerVison is showing github.com — Pull requests · huggingface/tokenizers

Pull requests

Open
102 (102)
Closed
1,332 (1,332)
More items
  • Comfortable display density
  • Compact display density
  • perf(bitcannon): add RVV UTF-8 character classification

    #2677·xiazhuozhao opened Nov 11, 2026·
    ·
  • fix(bpe): fuse unk tokens on models with affixes

    #2676·Iamsohungrynow opened Nov 11, 2026·
    ·
  • fix(vocab): add spare slots to small perfect hash tables

    #2675·Iamsohungrynow opened Nov 11, 2026·
    ·
  • fix(convert): preserve input vocabulary order

    #2674·YeexiaoZheng opened Nov 11, 2026·
    ·
  • fix(bpe): allow incomplete byte fallback vocabularies

    #2673·rleungx opened Nov 10, 2026·
    ·
    1
  • Fix FalconMamba tokenizer conversion with nonterminal ByteLevel

    #2672·voropaevv opened Nov 10, 2026·
    ·
    1
  • fix(regex): refuse overlapping regex matches

    #2671·SBrandeis opened Nov 9, 2026Collaborator·
    ·
    1
  • chore(python): drop support for python 3.10

    #2668·SBrandeis opened Nov 9, 2026Collaborator·
    ·
  • feat(python): expose the ByteLevel normalizer alphabet

    #2662·Mist-wu opened Nov 9, 2026·
    ·
  • fix(parallelism): release the pool lock around fork()

    #2656·McPatate opened Nov 6, 2026Member··
    ·
    2
  • docs: fix typo in README

    #2654·dheiver2 opened Nov 6, 2026·
    ·
  • perf(bitcannon): add RVV literal matching

    #2650·xiazhuozhao opened Nov 4, 2026·
    ·
    7
  • perf(parallel): only cut sequences that could hold up the batch

    #2649·McPatate opened Nov 4, 2026Member·
    ·
    1
  • perf(parallelism): read the cached pool without a lock

    #2648·McPatate opened Nov 4, 2026Member··
    ·
    1
  • refactor(parallelism)!: move the MaybeParallel* iterators to tk-train

    #2647·McPatate opened Nov 4, 2026Member··
    ·
    1
  • feat(python): add encode_iter to stream a batch's encodings

    #2644·McPatate opened Nov 3, 2026Member·
    ·
    1
  • feat(python): add set_max_threads / get_max_threads and a per-call max_threads

    #2643·McPatate opened Nov 3, 2026Member··
    ·
    1
  • feat(parallelism)!: stop reading TOKENIZERS_PARALLELISM

    #2642·McPatate opened Nov 3, 2026Member··
    ·
    1
  • perf(parallel): cut around normalized added tokens

    #2641·McPatate opened Nov 3, 2026Member··
    ·
    1
  • feat(parallelism): default forked children to one thread

    #2640·McPatate opened Nov 3, 2026Member··
    ·
    1
  • feat(parallel): add per-call EncodeOptions::max_threads

    #2639·McPatate opened Nov 3, 2026Member··
    ·
    1
  • perf(parallel): reassemble chunked sequences in one copy

    #2638·McPatate opened Nov 3, 2026Member·
    ·
    1
  • perf(parallel): split large inputs at safe pre-token boundaries

    #2637·McPatate opened Nov 3, 2026Member··
    ·
    1
  • feat(bitcannon): add safe cut points after letter runs

    #2636·McPatate opened Nov 3, 2026Member··
    ·
    1
  • fix(parallel): error instead of hanging when a handle crosses a fork

    #2635·McPatate opened Nov 3, 2026Member··
    ·
    1
You are visitor 01402503 — GeezerVison does not store the pages you look up.
Best viewed with Netscape Navigator 4.7 or Internet Explorer 6 at 800×600. No cookies required.