Italian — Wikilangs Models
Open-source tokenizers, n-gram & Markov language models, vocabulary stats, and word embeddings trained on Italian Wikipedia by Wikilangs.
🌐 Language Page · 🎮 Playground · 📊 Full Research Report
Language Samples
Example sentences drawn from the Italian Wikipedia corpus:
Eventi, invenzioni e scoperte Personaggi nasce Dante Alighieri Altri progetti 07
Eventi, invenzioni e scoperte Periodo della Grande carestia del Personaggi Giovanni Boccaccio nasce nel luglio Altri progetti 02
Eventi, invenzioni e scoperte Fine della cattività avignonese A Vicenza venne sparato il primo fuoco d'artificio Europeo. Personaggi Altri progetti 08
Eventi, invenzioni e scoperte Personaggi ... Altri progetti 09
Eventi, invenzioni e scoperte Viene inventato il Lapis Benjamin Franklin inventa il Parafulmine. Personaggi Wolfgang Amadeus Mozart Altri progetti 06
Quick Start
Load the Tokenizer
import sentencepiece as spm
sp = spm.SentencePieceProcessor()
sp.Load("it_tokenizer_32k.model")
text = "Eventi, invenzioni e scoperte Viene inventato il Lapis Benjamin Franklin inventa"
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: Eventi, invenzioni e scoperte Viene inventato il Lapis Benjamin Franklin inventa…
| Vocab | Tokens | Count |
|---|---|---|
| 8k | ▁eventi , ▁invenzioni ▁e ▁scoperte ▁viene ▁inv entato ▁il ▁la … (+29 more) |
39 |
| 16k | ▁eventi , ▁invenzioni ▁e ▁scoperte ▁viene ▁inventato ▁il ▁la pis … (+21 more) |
31 |
| 32k | ▁eventi , ▁invenzioni ▁e ▁scoperte ▁viene ▁inventato ▁il ▁la pis … (+17 more) |
27 |
| 64k | ▁eventi , ▁invenzioni ▁e ▁scoperte ▁viene ▁inventato ▁il ▁la pis … (+17 more) |
27 |
Sample 2: Eventi, invenzioni e scoperte Roma - Inaugurazione del Colosseo Personaggi 81 Ro…
| Vocab | Tokens | Count |
|---|---|---|
| 8k | ▁eventi , ▁invenzioni ▁e ▁scoperte ▁roma ▁- ▁inaugu razione ▁del … (+19 more) |
29 |
| 16k | ▁eventi , ▁invenzioni ▁e ▁scoperte ▁roma ▁- ▁inaugu razione ▁del … (+19 more) |
29 |
| 32k | ▁eventi , ▁invenzioni ▁e ▁scoperte ▁roma ▁- ▁inaugurazione ▁del ▁colo … (+18 more) |
28 |
| 64k | ▁eventi , ▁invenzioni ▁e ▁scoperte ▁roma ▁- ▁inaugurazione ▁del ▁colosseo … (+16 more) |
26 |
Sample 3: Eventi, invenzioni e scoperte Fine della cattività avignonese A Vicenza venne sp…
| Vocab | Tokens | Count |
|---|---|---|
| 8k | ▁eventi , ▁invenzioni ▁e ▁scoperte ▁fine ▁della ▁ca tti vità … (+23 more) |
33 |
| 16k | ▁eventi , ▁invenzioni ▁e ▁scoperte ▁fine ▁della ▁ca ttività ▁avi … (+22 more) |
32 |
| 32k | ▁eventi , ▁invenzioni ▁e ▁scoperte ▁fine ▁della ▁ca ttività ▁avi … (+22 more) |
32 |
| 64k | ▁eventi , ▁invenzioni ▁e ▁scoperte ▁fine ▁della ▁cattività ▁avignon ese … (+18 more) |
28 |
Load Word Embeddings
from gensim.models import KeyedVectors
# Aligned embeddings (cross-lingual, mapped to English vector space)
wv = KeyedVectors.load("it_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("it_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 | 3.86x |
| Tokenizer | 16k BPE | Compression | 4.25x |
| Tokenizer | 32k BPE | Compression | 4.58x |
| Tokenizer | 64k BPE | Compression | 4.82x 🏆 |
| N-gram | 2-gram (subword) | Perplexity | 214 🏆 |
| N-gram | 2-gram (word) | Perplexity | 204,245 |
| N-gram | 3-gram (subword) | Perplexity | 1,722 |
| N-gram | 3-gram (word) | Perplexity | 980,193 |
| N-gram | 4-gram (subword) | Perplexity | 10,064 |
| N-gram | 4-gram (word) | Perplexity | 1,937,953 |
| N-gram | 5-gram (subword) | Perplexity | 43,596 |
| N-gram | 5-gram (word) | Perplexity | 1,090,157 |
| Markov | ctx-1 (subword) | Predictability | 0.0% |
| Markov | ctx-1 (word) | Predictability | 0.0% |
| Markov | ctx-2 (subword) | Predictability | 32.2% |
| Markov | ctx-2 (word) | Predictability | 53.2% |
| Markov | ctx-3 (subword) | Predictability | 27.9% |
| Markov | ctx-3 (word) | Predictability | 79.8% |
| Markov | ctx-4 (subword) | Predictability | 32.0% |
| Markov | ctx-4 (word) | Predictability | 92.6% 🏆 |
| Vocabulary | full | Size | 511,837 |
| Vocabulary | full | Zipf R² | 0.9968 |
| Embeddings | mono_32d | Isotropy | 0.7834 |
| Embeddings | mono_64d | Isotropy | 0.7465 |
| Embeddings | mono_128d | Isotropy | 0.6690 |
| Embeddings | aligned_32d | Isotropy | 0.7834 🏆 |
| Embeddings | aligned_64d | Isotropy | 0.7465 |
| Embeddings | aligned_128d | Isotropy | 0.6690 |
| Alignment | aligned_32d | R@1 / R@5 / R@10 | 39.2% / 64.2% / 74.8% |
| Alignment | aligned_64d | R@1 / R@5 / R@10 | 60.6% / 81.4% / 85.8% |
| Alignment | aligned_128d | R@1 / R@5 / R@10 | 67.8% / 88.8% / 93.4% 🏆 |
📊 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}
}
Links
- 🌐 wikilangs.org
- 🌍 Language page
- 🎮 Playground
- 🤗 HuggingFace models
- 📊 wikipedia-monthly dataset
- 👤 Omar Kamali
- 🤝 Sponsor: Featherless AI
License: MIT — free for academic and commercial use.
Generated by Wikilangs Pipeline · 2026-03-03 11:41:08
