English — Wikilangs Models

Open-source tokenizers, n-gram & Markov language models, vocabulary stats, and word embeddings trained on English Wikipedia by Wikilangs.

🌐 Language Page · 🎮 Playground · 📊 Full Research Report

Language Samples

Example sentences drawn from the English Wikipedia corpus:

Alexander V may refer to: Alexander V of Macedon (died 294 BCE) Antipope Alexander V Alexander V of Imereti

Alfonso IV may refer to: Alfonso IV of León (924–931) Afonso IV of Portugal Alfonso IV of Aragon Alfonso IV of Ribagorza Alfonso IV d'Este Duke of Modena and Regg

Anastasius I or Anastasios I may refer to: Anastasius I Dicorus (–518), Roman emperor Anastasius I of Antioch (died 599), Patriarch of Antioch Pope Anastasius I (died 401), pope

Angula may refer to: Aṅgula, a measure equal to a finger's breadth Eel, a biological order of fish Nahas Angula, former Prime Minister of Namibia Helmut Angula See also Angul (disambiguation)

Two antipopes used the regnal name Victor IV: Antipope Victor IV Antipope Victor IV

Quick Start

Load the Tokenizer

import sentencepiece as spm

sp = spm.SentencePieceProcessor()
sp.Load("en_tokenizer_32k.model")

text = "Albrecht Achilles may refer to: Albrecht III Achilles, Elector of Brandenburg Al"
tokens = sp.EncodeAsPieces(text)
ids    = sp.EncodeAsIds(text)

print(tokens)  # subword pieces
print(ids)     # integer ids

# Decode back
print(sp.DecodeIds(ids))

Tokenization examples (click to expand)

Sample 1: Albrecht Achilles may refer to: Albrecht III Achilles, Elector of Brandenburg Al…

Vocab	Tokens	Count
8k	`▁alb recht ▁ach illes ▁may ▁refer ▁to : ▁alb recht … (+27 more)`	37
16k	`▁alb recht ▁ach illes ▁may ▁refer ▁to : ▁alb recht … (+26 more)`	36
32k	`▁albrecht ▁achilles ▁may ▁refer ▁to : ▁albrecht ▁iii ▁achilles , … (+17 more)`	27
64k	`▁albrecht ▁achilles ▁may ▁refer ▁to : ▁albrecht ▁iii ▁achilles , … (+16 more)`	26

Sample 2: Alexander V may refer to: Alexander V of Macedon (died 294 BCE) Antipope Alexand…

Vocab	Tokens	Count
8k	`▁alexander ▁v ▁may ▁refer ▁to : ▁alexander ▁v ▁of ▁maced … (+20 more)`	30
16k	`▁alexander ▁v ▁may ▁refer ▁to : ▁alexander ▁v ▁of ▁macedon … (+18 more)`	28
32k	`▁alexander ▁v ▁may ▁refer ▁to : ▁alexander ▁v ▁of ▁macedon … (+15 more)`	25
64k	`▁alexander ▁v ▁may ▁refer ▁to : ▁alexander ▁v ▁of ▁macedon … (+15 more)`	25

Sample 3: Two antipopes used the regnal name Victor IV: Antipope Victor IV Antipope Victor…

Vocab	Tokens	Count
8k	`▁two ▁antip op es ▁used ▁the ▁reg nal ▁name ▁victor … (+8 more)`	18
16k	`▁two ▁antip opes ▁used ▁the ▁reg nal ▁name ▁victor ▁iv … (+7 more)`	17
32k	`▁two ▁antip opes ▁used ▁the ▁regnal ▁name ▁victor ▁iv : … (+6 more)`	16
64k	`▁two ▁antipopes ▁used ▁the ▁regnal ▁name ▁victor ▁iv : ▁antipope … (+5 more)`	15

Load Word Embeddings

from gensim.models import KeyedVectors

# Aligned embeddings (cross-lingual, mapped to English vector space)
wv = KeyedVectors.load("en_embeddings_128d_aligned.kv")

similar = wv.most_similar("word", topn=5)
for word, score in similar:
    print(f"  {word}: {score:.3f}")

Load N-gram Model

import pyarrow.parquet as pq

df = pq.read_table("en_3gram_word.parquet").to_pandas()
print(df.head())

Models Overview

Category	Assets
Tokenizers	BPE at 8k, 16k, 32k, 64k vocab sizes
N-gram models	2 / 3 / 4 / 5-gram (word & subword)
Markov chains	Context 1–5 (word & subword)
Embeddings	32d, 64d, 128d — mono & aligned
Vocabulary	Full frequency list + Zipf analysis
Statistics	Corpus & model statistics JSON

Metrics Summary

Component	Model	Key Metric	Value
Tokenizer	8k BPE	Compression	3.84x
Tokenizer	16k BPE	Compression	4.22x
Tokenizer	32k BPE	Compression	4.51x
Tokenizer	64k BPE	Compression	4.70x 🏆
N-gram	2-gram (subword)	Perplexity	257 🏆
N-gram	2-gram (word)	Perplexity	386,225
N-gram	3-gram (subword)	Perplexity	2,180
N-gram	3-gram (word)	Perplexity	4,093,782
N-gram	4-gram (subword)	Perplexity	12,758
N-gram	4-gram (word)	Perplexity	14,465,722
N-gram	5-gram (subword)	Perplexity	55,700
N-gram	5-gram (word)	Perplexity	12,820,936
Markov	ctx-1 (subword)	Predictability	0.0%
Markov	ctx-1 (word)	Predictability	6.2%
Markov	ctx-2 (subword)	Predictability	46.4%
Markov	ctx-2 (word)	Predictability	48.3%
Markov	ctx-3 (subword)	Predictability	45.8%
Markov	ctx-3 (word)	Predictability	75.9%
Markov	ctx-4 (subword)	Predictability	36.8%
Markov	ctx-4 (word)	Predictability	89.2% 🏆
Vocabulary	full	Size	1,867,537
Vocabulary	full	Zipf R²	0.9862
Embeddings	mono_32d	Isotropy	0.7693 🏆
Embeddings	mono_64d	Isotropy	0.7388
Embeddings	mono_128d	Isotropy	0.6687

📊 Full ablation study, per-model breakdowns, and interpretation guide →

About

Trained on wikipedia-monthly — monthly snapshots of 300+ Wikipedia languages.

A project by Wikilangs · Maintainer: Omar Kamali · Omneity Labs

Citation

@misc{wikilangs2025,
  author    = {Kamali, Omar},
  title     = {Wikilangs: Open NLP Models for Wikipedia Languages},
  year      = {2025},
  doi       = {10.5281/zenodo.18073153},
  publisher = {Zenodo},
  url       = {https://huggingface.co/wikilangs},
  institution = {Omneity Labs}
}

wikilangs
/

en