| Parameter | Value |
|---|---|
| Algorithm | Byte-Level BPE |
| Language | ell_Grek |
| Target Vocab Size | 8,000 |
| Final Vocab Size | 9,038 |
| Pre-tokenizer | custom:ell_Grek |
| Number handling | ltr_3digit |
| Contraction handling | True |
| Normalizer | NFC |
| Special Tokens | <s>, </s>, <pad>, <unk> |
| Training Shards | 2 |
1from transformers import AutoTokenizer
2
3tokenizer = AutoTokenizer.from_pretrained("flexitok/bpe_ltr_ell_Grek_8000_v2")
4tokens = tokenizer.encode("Hello, world!")tokenizer.json — Full HuggingFace tokenizervocab.json — Vocabulary mappingmerges.txt — BPE merge rules| Text | Tokens | Token IDs |
|---|---|---|
Hello, world! 12345 This is a test. こんにちは | H, ell, o, ,, Ġ, w, or, ld, !, Ġ, 123, 45, Ġ, Th, is, Ġ, is, Ġ, a, Ġ | 42, 2576, 81, 14, 223, 89, 693, 5583, 3, 223, 8522, 4808, 223, 3829, 933, 223, 933, 223, 67, 223 |