README.md

metadata

license: mit
language:
  - ita
tags:
  - tokenizer
  - bpe
  - flexitok
  - fineweb2

Byte-Level BPE Tokenizer: ['ita_Latn'] (8K)

A Byte-Level BPE tokenizer trained on ['ita_Latn'] data from Fineweb-2-HQ.

Training Details

Parameter	Value
Algorithm	Byte-Level BPE
Language	['ita_Latn']
Target Vocab Size	8,000
Final Vocab Size	9,011
Pre-tokenizer	custom:ita_Latn
Number handling	ltr_3digit
Contraction handling	False
Normalizer	NFC
Special Tokens	<s>, </s>, <pad>, <unk>
Training Shards	2, ['fineweb_2_hq.ita_Latn.chunk.00.jsonl', 'fineweb_2_hq.ita_Latn.chunk.01.jsonl']

Usage

from transformers import AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("flexitok/bpe_ltr_ita_Latn_8000_v3")
tokens = tokenizer.encode("Hello, world!")

Files

• tokenizer.json — Full HuggingFace tokenizer
• vocab.json — Vocabulary mapping
• merges.txt — BPE merge rules

Sample Encoding

Text	Tokens	Token IDs
Hello, world! 12345 This is a test. こんにちは	H, ello, ,, Ġw, orld, !, Ġ, 123, 45, ĠTh, is, Ġis, Ġa, Ġtest, ., Ġ, ãģ, ĵ, ã, Ĥ	42, 656, 14, 895, 4899, 3, 223, 8311, 3963, 2836, 341, 1641, 265, 2338, 16, 223, 6104, 244, 162, 227

Command used to create this tokenizer:

['/home/gsa/tokenizers2/flexitok/tokenizer_training/train_tokenizers.py', 'algorithm=bpe', 'vocab_size=8000', 'langs=[ita_Latn]', 'pretok_behavior=isolated', 'data_dir=/scratch/gsa/data/flexitok/', 'output_dir=/scratch/gsa/trained_tokenizers', 'pretokenizer=custom:ita_Latn', 'number_handling=ltr_3digit', 'add_numbers=true', 'handle_contractions=false', 'unicode_normalization=nfc', 'use_byte_level_regex=false', 'byte_fallback=false', 'strip_zero_width=false', 'cjk_char_split=false', 'cjk_char_coverage=0', 'add_cjk_chars=false', 'max_han_run=-1', 'max_lines=1_000_000', 'hf.publish_to_hf=true', 'hf_repo_prefix=flexitok/', 'hf.hf_repo_id=flexitok/bpe_ltr_ita_Latn_8000_v3', 'hf.collections=[flexitok/fineweb2-hq-tokenizers-v3,flexitok/8000-vocab-v3]']